summaryrefslogtreecommitdiff
path: root/support/bib2gls/src/bib2gls.tex
blob: aadbf0c5c73248c0bf68f2b58e44637cbb721870 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
3485
3486
3487
3488
3489
3490
3491
3492
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
3514
3515
3516
3517
3518
3519
3520
3521
3522
3523
3524
3525
3526
3527
3528
3529
3530
3531
3532
3533
3534
3535
3536
3537
3538
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
3552
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
3566
3567
3568
3569
3570
3571
3572
3573
3574
3575
3576
3577
3578
3579
3580
3581
3582
3583
3584
3585
3586
3587
3588
3589
3590
3591
3592
3593
3594
3595
3596
3597
3598
3599
3600
3601
3602
3603
3604
3605
3606
3607
3608
3609
3610
3611
3612
3613
3614
3615
3616
3617
3618
3619
3620
3621
3622
3623
3624
3625
3626
3627
3628
3629
3630
3631
3632
3633
3634
3635
3636
3637
3638
3639
3640
3641
3642
3643
3644
3645
3646
3647
3648
3649
3650
3651
3652
3653
3654
3655
3656
3657
3658
3659
3660
3661
3662
3663
3664
3665
3666
3667
3668
3669
3670
3671
3672
3673
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
3686
3687
3688
3689
3690
3691
3692
3693
3694
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
3706
3707
3708
3709
3710
3711
3712
3713
3714
3715
3716
3717
3718
3719
3720
3721
3722
3723
3724
3725
3726
3727
3728
3729
3730
3731
3732
3733
3734
3735
3736
3737
3738
3739
3740
3741
3742
3743
3744
3745
3746
3747
3748
3749
3750
3751
3752
3753
3754
3755
3756
3757
3758
3759
3760
3761
3762
3763
3764
3765
3766
3767
3768
3769
3770
3771
3772
3773
3774
3775
3776
3777
3778
3779
3780
3781
3782
3783
3784
3785
3786
3787
3788
3789
3790
3791
3792
3793
3794
3795
3796
3797
3798
3799
3800
3801
3802
3803
3804
3805
3806
3807
3808
3809
3810
3811
3812
3813
3814
3815
3816
3817
3818
3819
3820
3821
3822
3823
3824
3825
3826
3827
3828
3829
3830
3831
3832
3833
3834
3835
3836
3837
3838
3839
3840
3841
3842
3843
3844
3845
3846
3847
3848
3849
3850
3851
3852
3853
3854
3855
3856
3857
3858
3859
3860
3861
3862
3863
3864
3865
3866
3867
3868
3869
3870
3871
3872
3873
3874
3875
3876
3877
3878
3879
3880
3881
3882
3883
3884
3885
3886
3887
3888
3889
3890
3891
3892
3893
3894
3895
3896
3897
3898
3899
3900
3901
3902
3903
3904
3905
3906
3907
3908
3909
3910
3911
3912
3913
3914
3915
3916
3917
3918
3919
3920
3921
3922
3923
3924
3925
3926
3927
3928
3929
3930
3931
3932
3933
3934
3935
3936
3937
3938
3939
3940
3941
3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
3953
3954
3955
3956
3957
3958
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
3969
3970
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
3981
3982
3983
3984
3985
3986
3987
3988
3989
3990
3991
3992
3993
3994
3995
3996
3997
3998
3999
4000
4001
4002
4003
4004
4005
4006
4007
4008
4009
4010
4011
4012
4013
4014
4015
4016
4017
4018
4019
4020
4021
4022
4023
4024
4025
4026
4027
4028
4029
4030
4031
4032
4033
4034
4035
4036
4037
4038
4039
4040
4041
4042
4043
4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
4068
4069
4070
4071
4072
4073
4074
4075
4076
4077
4078
4079
4080
4081
4082
4083
4084
4085
4086
4087
4088
4089
4090
4091
4092
4093
4094
4095
4096
4097
4098
4099
4100
4101
4102
4103
4104
4105
4106
4107
4108
4109
4110
4111
4112
4113
4114
4115
4116
4117
4118
4119
4120
4121
4122
4123
4124
4125
4126
4127
4128
4129
4130
4131
4132
4133
4134
4135
4136
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
4162
4163
4164
4165
4166
4167
4168
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
4179
4180
4181
4182
4183
4184
4185
4186
4187
4188
4189
4190
4191
4192
4193
4194
4195
4196
4197
4198
4199
4200
4201
4202
4203
4204
4205
4206
4207
4208
4209
4210
4211
4212
4213
4214
4215
4216
4217
4218
4219
4220
4221
4222
4223
4224
4225
4226
4227
4228
4229
4230
4231
4232
4233
4234
4235
4236
4237
4238
4239
4240
4241
4242
4243
4244
4245
4246
4247
4248
4249
4250
4251
4252
4253
4254
4255
4256
4257
4258
4259
4260
4261
4262
4263
4264
4265
4266
4267
4268
4269
4270
4271
4272
4273
4274
4275
4276
4277
4278
4279
4280
4281
4282
4283
4284
4285
4286
4287
4288
4289
4290
4291
4292
4293
4294
4295
4296
4297
4298
4299
4300
4301
4302
4303
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
4315
4316
4317
4318
4319
4320
4321
4322
4323
4324
4325
4326
4327
4328
4329
4330
4331
4332
4333
4334
4335
4336
4337
4338
4339
4340
4341
4342
4343
4344
4345
4346
4347
4348
4349
4350
4351
4352
4353
4354
4355
4356
4357
4358
4359
4360
4361
4362
4363
4364
4365
4366
4367
4368
4369
4370
4371
4372
4373
4374
4375
4376
4377
4378
4379
4380
4381
4382
4383
4384
4385
4386
4387
4388
4389
4390
4391
4392
4393
4394
4395
4396
4397
4398
4399
4400
4401
4402
4403
4404
4405
4406
4407
4408
4409
4410
4411
4412
4413
4414
4415
4416
4417
4418
4419
4420
4421
4422
4423
4424
4425
4426
4427
4428
4429
4430
4431
4432
4433
4434
4435
4436
4437
4438
4439
4440
4441
4442
4443
4444
4445
4446
4447
4448
4449
4450
4451
4452
4453
4454
4455
4456
4457
4458
4459
4460
4461
4462
4463
4464
4465
4466
4467
4468
4469
4470
4471
4472
4473
4474
4475
4476
4477
4478
4479
4480
4481
4482
4483
4484
4485
4486
4487
4488
4489
4490
4491
4492
4493
4494
4495
4496
4497
4498
4499
4500
4501
4502
4503
4504
4505
4506
4507
4508
4509
4510
4511
4512
4513
4514
4515
4516
4517
4518
4519
4520
4521
4522
4523
4524
4525
4526
4527
4528
4529
4530
4531
4532
4533
4534
4535
4536
4537
4538
4539
4540
4541
4542
4543
4544
4545
4546
4547
4548
4549
4550
4551
4552
4553
4554
4555
4556
4557
4558
4559
4560
4561
4562
4563
4564
4565
4566
4567
4568
4569
4570
4571
4572
4573
4574
4575
4576
4577
4578
4579
4580
4581
4582
4583
4584
4585
4586
4587
4588
4589
4590
4591
4592
4593
4594
4595
4596
4597
4598
4599
4600
4601
4602
4603
4604
4605
4606
4607
4608
4609
4610
4611
4612
4613
4614
4615
4616
4617
4618
4619
4620
4621
4622
4623
4624
4625
4626
4627
4628
4629
4630
4631
4632
4633
4634
4635
4636
4637
4638
4639
4640
4641
4642
4643
4644
4645
4646
4647
4648
4649
4650
4651
4652
4653
4654
4655
4656
4657
4658
4659
4660
4661
4662
4663
4664
4665
4666
4667
4668
4669
4670
4671
4672
4673
4674
4675
4676
4677
4678
4679
4680
4681
4682
4683
4684
4685
4686
4687
4688
4689
4690
4691
4692
4693
4694
4695
4696
4697
4698
4699
4700
4701
4702
4703
4704
4705
4706
4707
4708
4709
4710
4711
4712
4713
4714
4715
4716
4717
4718
4719
4720
4721
4722
4723
4724
4725
4726
4727
4728
4729
4730
4731
4732
4733
4734
4735
4736
4737
4738
4739
4740
4741
4742
4743
4744
4745
4746
4747
4748
4749
4750
4751
4752
4753
4754
4755
4756
4757
4758
4759
4760
4761
4762
4763
4764
4765
4766
4767
4768
4769
4770
4771
4772
4773
4774
4775
4776
4777
4778
4779
4780
4781
4782
4783
4784
4785
4786
4787
4788
4789
4790
4791
4792
4793
4794
4795
4796
4797
4798
4799
4800
4801
4802
4803
4804
4805
4806
4807
4808
4809
4810
4811
4812
4813
4814
4815
4816
4817
4818
4819
4820
4821
4822
4823
4824
4825
4826
4827
4828
4829
4830
4831
4832
4833
4834
4835
4836
4837
4838
4839
4840
4841
4842
4843
4844
4845
4846
4847
4848
4849
4850
4851
4852
4853
4854
4855
4856
4857
4858
4859
4860
4861
4862
4863
4864
4865
4866
4867
4868
4869
4870
4871
4872
4873
4874
4875
4876
4877
4878
4879
4880
4881
4882
4883
4884
4885
4886
4887
4888
4889
4890
4891
4892
4893
4894
4895
4896
4897
4898
4899
4900
4901
4902
4903
4904
4905
4906
4907
4908
4909
4910
4911
4912
4913
4914
4915
4916
4917
4918
4919
4920
4921
4922
4923
4924
4925
4926
4927
4928
4929
4930
4931
4932
4933
4934
4935
4936
4937
4938
4939
4940
4941
4942
4943
4944
4945
4946
4947
4948
4949
4950
4951
4952
4953
4954
4955
4956
4957
4958
4959
4960
4961
4962
4963
4964
4965
4966
4967
4968
4969
4970
4971
4972
4973
4974
4975
4976
4977
4978
4979
4980
4981
4982
4983
4984
4985
4986
4987
4988
4989
4990
4991
4992
4993
4994
4995
4996
4997
4998
4999
5000
5001
5002
5003
5004
5005
5006
5007
5008
5009
5010
5011
5012
5013
5014
5015
5016
5017
5018
5019
5020
5021
5022
5023
5024
5025
5026
5027
5028
5029
5030
5031
5032
5033
5034
5035
5036
5037
5038
5039
5040
5041
5042
5043
5044
5045
5046
5047
5048
5049
5050
5051
5052
5053
5054
5055
5056
5057
5058
5059
5060
5061
5062
5063
5064
5065
5066
5067
5068
5069
5070
5071
5072
5073
5074
5075
5076
5077
5078
5079
5080
5081
5082
5083
5084
5085
5086
5087
5088
5089
5090
5091
5092
5093
5094
5095
5096
5097
5098
5099
5100
5101
5102
5103
5104
5105
5106
5107
5108
5109
5110
5111
5112
5113
5114
5115
5116
5117
5118
5119
5120
5121
5122
5123
5124
5125
5126
5127
5128
5129
5130
5131
5132
5133
5134
5135
5136
5137
5138
5139
5140
5141
5142
5143
5144
5145
5146
5147
5148
5149
5150
5151
5152
5153
5154
5155
5156
5157
5158
5159
5160
5161
5162
5163
5164
5165
5166
5167
5168
5169
5170
5171
5172
5173
5174
5175
5176
5177
5178
5179
5180
5181
5182
5183
5184
5185
5186
5187
5188
5189
5190
5191
5192
5193
5194
5195
5196
5197
5198
5199
5200
5201
5202
5203
5204
5205
5206
5207
5208
5209
5210
5211
5212
5213
5214
5215
5216
5217
5218
5219
5220
5221
5222
5223
5224
5225
5226
5227
5228
5229
5230
5231
5232
5233
5234
5235
5236
5237
5238
5239
5240
5241
5242
5243
5244
5245
5246
5247
5248
5249
5250
5251
5252
5253
5254
5255
5256
5257
5258
5259
5260
5261
5262
5263
5264
5265
5266
5267
5268
5269
5270
5271
5272
5273
5274
5275
5276
5277
5278
5279
5280
5281
5282
5283
5284
5285
5286
5287
5288
5289
5290
5291
5292
5293
5294
5295
5296
5297
5298
5299
5300
5301
5302
5303
5304
5305
5306
5307
5308
5309
5310
5311
5312
5313
5314
5315
5316
5317
5318
5319
5320
5321
5322
5323
5324
5325
5326
5327
5328
5329
5330
5331
5332
5333
5334
5335
5336
5337
5338
5339
5340
5341
5342
5343
5344
5345
5346
5347
5348
5349
5350
5351
5352
5353
5354
5355
5356
5357
5358
5359
5360
5361
5362
5363
5364
5365
5366
5367
5368
5369
5370
5371
5372
5373
5374
5375
5376
5377
5378
5379
5380
5381
5382
5383
5384
5385
5386
5387
5388
5389
5390
5391
5392
5393
5394
5395
5396
5397
5398
5399
5400
5401
5402
5403
5404
5405
5406
5407
5408
5409
5410
5411
5412
5413
5414
5415
5416
5417
5418
5419
5420
5421
5422
5423
5424
5425
5426
5427
5428
5429
5430
5431
5432
5433
5434
5435
5436
5437
5438
5439
5440
5441
5442
5443
5444
5445
5446
5447
5448
5449
5450
5451
5452
5453
5454
5455
5456
5457
5458
5459
5460
5461
5462
5463
5464
5465
5466
5467
5468
5469
5470
5471
5472
5473
5474
5475
5476
5477
5478
5479
5480
5481
5482
5483
5484
5485
5486
5487
5488
5489
5490
5491
5492
5493
5494
5495
5496
5497
5498
5499
5500
5501
5502
5503
5504
5505
5506
5507
5508
5509
5510
5511
5512
5513
5514
5515
5516
5517
5518
5519
5520
5521
5522
5523
5524
5525
5526
5527
5528
5529
5530
5531
5532
5533
5534
5535
5536
5537
5538
5539
5540
5541
5542
5543
5544
5545
5546
5547
5548
5549
5550
5551
5552
5553
5554
5555
5556
5557
5558
5559
5560
5561
5562
5563
5564
5565
5566
5567
5568
5569
5570
5571
5572
5573
5574
5575
5576
5577
5578
5579
5580
5581
5582
5583
5584
5585
5586
5587
5588
5589
5590
5591
5592
5593
5594
5595
5596
5597
5598
5599
5600
5601
5602
5603
5604
5605
5606
5607
5608
5609
5610
5611
5612
5613
5614
5615
5616
5617
5618
5619
5620
5621
5622
5623
5624
5625
5626
5627
5628
5629
5630
5631
5632
5633
5634
5635
5636
5637
5638
5639
5640
5641
5642
5643
5644
5645
5646
5647
5648
5649
5650
5651
5652
5653
5654
5655
5656
5657
5658
5659
5660
5661
5662
5663
5664
5665
5666
5667
5668
5669
5670
5671
5672
5673
5674
5675
5676
5677
5678
5679
5680
5681
5682
5683
5684
5685
5686
5687
5688
5689
5690
5691
5692
5693
5694
5695
5696
5697
5698
5699
5700
5701
5702
5703
5704
5705
5706
5707
5708
5709
5710
5711
5712
5713
5714
5715
5716
5717
5718
5719
5720
5721
5722
5723
5724
5725
5726
5727
5728
5729
5730
5731
5732
5733
5734
5735
5736
5737
5738
5739
5740
5741
5742
5743
5744
5745
5746
5747
5748
5749
5750
5751
5752
5753
5754
5755
5756
5757
5758
5759
5760
5761
5762
5763
5764
5765
5766
5767
5768
5769
5770
5771
5772
5773
5774
5775
5776
5777
5778
5779
5780
5781
5782
5783
5784
5785
5786
5787
5788
5789
5790
5791
5792
5793
5794
5795
5796
5797
5798
5799
5800
5801
5802
5803
5804
5805
5806
5807
5808
5809
5810
5811
5812
5813
5814
5815
5816
5817
5818
5819
5820
5821
5822
5823
5824
5825
5826
5827
5828
5829
5830
5831
5832
5833
5834
5835
5836
5837
5838
5839
5840
5841
5842
5843
5844
5845
5846
5847
5848
5849
5850
5851
5852
5853
5854
5855
5856
5857
5858
5859
5860
5861
5862
5863
5864
5865
5866
5867
5868
5869
5870
5871
5872
5873
5874
5875
5876
5877
5878
5879
5880
5881
5882
5883
5884
5885
5886
5887
5888
5889
5890
5891
5892
5893
5894
5895
5896
5897
5898
5899
5900
5901
5902
5903
5904
5905
5906
5907
5908
5909
5910
5911
5912
5913
5914
5915
5916
5917
5918
5919
5920
5921
5922
5923
5924
5925
5926
5927
5928
5929
5930
5931
5932
5933
5934
5935
5936
5937
5938
5939
5940
5941
5942
5943
5944
5945
5946
5947
5948
5949
5950
5951
5952
5953
5954
5955
5956
5957
5958
5959
5960
5961
5962
5963
5964
5965
5966
5967
5968
5969
5970
5971
5972
5973
5974
5975
5976
5977
5978
5979
5980
5981
5982
5983
5984
5985
5986
5987
5988
5989
5990
5991
5992
5993
5994
5995
5996
5997
5998
5999
6000
6001
6002
6003
6004
6005
6006
6007
6008
6009
6010
6011
6012
6013
6014
6015
6016
6017
6018
6019
6020
6021
6022
6023
6024
6025
6026
6027
6028
6029
6030
6031
6032
6033
6034
6035
6036
6037
6038
6039
6040
6041
6042
6043
6044
6045
6046
6047
6048
6049
6050
6051
6052
6053
6054
6055
6056
6057
6058
6059
6060
6061
6062
6063
6064
6065
6066
6067
6068
6069
6070
6071
6072
6073
6074
6075
6076
6077
6078
6079
6080
6081
6082
6083
6084
6085
6086
6087
6088
6089
6090
6091
6092
6093
6094
6095
6096
6097
6098
6099
6100
6101
6102
6103
6104
6105
6106
6107
6108
6109
6110
6111
6112
6113
6114
6115
6116
6117
6118
6119
6120
6121
6122
6123
6124
6125
6126
6127
6128
6129
6130
6131
6132
6133
6134
6135
6136
6137
6138
6139
6140
6141
6142
6143
6144
6145
6146
6147
6148
6149
6150
6151
6152
6153
6154
6155
6156
6157
6158
6159
6160
6161
6162
6163
6164
6165
6166
6167
6168
6169
6170
6171
6172
6173
6174
6175
6176
6177
6178
6179
6180
6181
6182
6183
6184
6185
6186
6187
6188
6189
6190
6191
6192
6193
6194
6195
6196
6197
6198
6199
6200
6201
6202
6203
6204
6205
6206
6207
6208
6209
6210
6211
6212
6213
6214
6215
6216
6217
6218
6219
6220
6221
6222
6223
6224
6225
6226
6227
6228
6229
6230
6231
6232
6233
6234
6235
6236
6237
6238
6239
6240
6241
6242
6243
6244
6245
6246
6247
6248
6249
6250
6251
6252
6253
6254
6255
6256
6257
6258
6259
6260
6261
6262
6263
6264
6265
6266
6267
6268
6269
6270
6271
6272
6273
6274
6275
6276
6277
6278
6279
6280
6281
6282
6283
6284
6285
6286
6287
6288
6289
6290
6291
6292
6293
6294
6295
6296
6297
6298
6299
6300
6301
6302
6303
6304
6305
6306
6307
6308
6309
6310
6311
6312
6313
6314
6315
6316
6317
6318
6319
6320
6321
6322
6323
6324
6325
6326
6327
6328
6329
6330
6331
6332
6333
6334
6335
6336
6337
6338
6339
6340
6341
6342
6343
6344
6345
6346
6347
6348
6349
6350
6351
6352
6353
6354
6355
6356
6357
6358
6359
6360
6361
6362
6363
6364
6365
6366
6367
6368
6369
6370
6371
6372
6373
6374
6375
6376
6377
6378
6379
6380
6381
6382
6383
6384
6385
6386
6387
6388
6389
6390
6391
6392
6393
6394
6395
6396
6397
6398
6399
6400
6401
6402
6403
6404
6405
6406
6407
6408
6409
6410
6411
6412
6413
6414
6415
6416
6417
6418
6419
6420
6421
6422
6423
6424
6425
6426
6427
6428
6429
6430
6431
6432
6433
6434
6435
6436
6437
6438
6439
6440
6441
6442
6443
6444
6445
6446
6447
6448
6449
6450
6451
6452
6453
6454
6455
6456
6457
6458
6459
6460
6461
6462
6463
6464
6465
6466
6467
6468
6469
6470
6471
6472
6473
6474
6475
6476
6477
6478
6479
6480
6481
6482
6483
6484
6485
6486
6487
6488
6489
6490
6491
6492
6493
6494
6495
6496
6497
6498
6499
6500
6501
6502
6503
6504
6505
6506
6507
6508
6509
6510
6511
6512
6513
6514
6515
6516
6517
6518
6519
6520
6521
6522
6523
6524
6525
6526
6527
6528
6529
6530
6531
6532
6533
6534
6535
6536
6537
6538
6539
6540
6541
6542
6543
6544
6545
6546
6547
6548
6549
6550
6551
6552
6553
6554
6555
6556
6557
6558
6559
6560
6561
6562
6563
6564
6565
6566
6567
6568
6569
6570
6571
6572
6573
6574
6575
6576
6577
6578
6579
6580
6581
6582
6583
6584
6585
6586
6587
6588
6589
6590
6591
6592
6593
6594
6595
6596
6597
6598
6599
6600
6601
6602
6603
6604
6605
6606
6607
6608
6609
6610
6611
6612
6613
6614
6615
6616
6617
6618
6619
6620
6621
6622
6623
6624
6625
6626
6627
6628
6629
6630
6631
6632
6633
6634
6635
6636
6637
6638
6639
6640
6641
6642
6643
6644
6645
6646
6647
6648
6649
6650
6651
6652
6653
6654
6655
6656
6657
6658
6659
6660
6661
6662
6663
6664
6665
6666
6667
6668
6669
6670
6671
6672
6673
6674
6675
6676
6677
6678
6679
6680
6681
6682
6683
6684
6685
6686
6687
6688
6689
6690
6691
6692
6693
6694
6695
6696
6697
6698
6699
6700
6701
6702
6703
6704
6705
6706
6707
6708
6709
6710
6711
6712
6713
6714
6715
6716
6717
6718
6719
6720
6721
6722
6723
6724
6725
6726
6727
6728
6729
6730
6731
6732
6733
6734
6735
6736
6737
6738
6739
6740
6741
6742
6743
6744
6745
6746
6747
6748
6749
6750
6751
6752
6753
6754
6755
6756
6757
6758
6759
6760
6761
6762
6763
6764
6765
6766
6767
6768
6769
6770
6771
6772
6773
6774
6775
6776
6777
6778
6779
6780
6781
6782
6783
6784
6785
6786
6787
6788
6789
6790
6791
6792
6793
6794
6795
6796
6797
6798
6799
6800
6801
6802
6803
6804
6805
6806
6807
6808
6809
6810
6811
6812
6813
6814
6815
6816
6817
6818
6819
6820
6821
6822
6823
6824
6825
6826
6827
6828
6829
6830
6831
6832
6833
6834
6835
6836
6837
6838
6839
6840
6841
6842
6843
6844
6845
6846
6847
6848
6849
6850
6851
6852
6853
6854
6855
6856
6857
6858
6859
6860
6861
6862
6863
6864
6865
6866
6867
6868
6869
6870
6871
6872
6873
6874
6875
6876
6877
6878
6879
6880
6881
6882
6883
6884
6885
6886
6887
6888
6889
6890
6891
6892
6893
6894
6895
6896
6897
6898
6899
6900
6901
6902
6903
6904
6905
6906
6907
6908
6909
6910
6911
6912
6913
6914
6915
6916
6917
6918
6919
6920
6921
6922
6923
6924
6925
6926
6927
6928
6929
6930
6931
6932
6933
6934
6935
6936
6937
6938
6939
6940
6941
6942
6943
6944
6945
6946
6947
6948
6949
6950
6951
6952
6953
6954
6955
6956
6957
6958
6959
6960
6961
6962
6963
6964
6965
6966
6967
6968
6969
6970
6971
6972
6973
6974
6975
6976
6977
6978
6979
6980
6981
6982
6983
6984
6985
6986
6987
6988
6989
6990
6991
6992
6993
6994
6995
6996
6997
6998
6999
7000
7001
7002
7003
7004
7005
7006
7007
7008
7009
7010
7011
7012
7013
7014
7015
7016
7017
7018
7019
7020
7021
7022
7023
7024
7025
7026
7027
7028
7029
7030
7031
7032
7033
7034
7035
7036
7037
7038
7039
7040
7041
7042
7043
7044
7045
7046
7047
7048
7049
7050
7051
7052
7053
7054
7055
7056
7057
7058
7059
7060
7061
7062
7063
7064
7065
7066
7067
7068
7069
7070
7071
7072
7073
7074
7075
7076
7077
7078
7079
7080
7081
7082
7083
7084
7085
7086
7087
7088
7089
7090
7091
7092
7093
7094
7095
7096
7097
7098
7099
7100
7101
7102
7103
7104
7105
7106
7107
7108
7109
7110
7111
7112
7113
7114
7115
7116
7117
7118
7119
7120
7121
7122
7123
7124
7125
7126
7127
7128
7129
7130
7131
7132
7133
7134
7135
7136
7137
7138
7139
7140
7141
7142
7143
7144
7145
7146
7147
7148
7149
7150
7151
7152
7153
7154
7155
7156
7157
7158
7159
7160
7161
7162
7163
7164
7165
7166
7167
7168
7169
7170
7171
7172
7173
7174
7175
7176
7177
7178
7179
7180
7181
7182
7183
7184
7185
7186
7187
7188
7189
7190
7191
7192
7193
7194
7195
7196
7197
7198
7199
7200
7201
7202
7203
7204
7205
7206
7207
7208
7209
7210
7211
7212
7213
7214
7215
7216
7217
7218
7219
7220
7221
7222
7223
7224
7225
7226
7227
7228
7229
7230
7231
7232
7233
7234
7235
7236
7237
7238
7239
7240
7241
7242
7243
7244
7245
7246
7247
7248
7249
7250
7251
7252
7253
7254
7255
7256
7257
7258
7259
7260
7261
7262
7263
7264
7265
7266
7267
7268
7269
7270
7271
7272
7273
7274
7275
7276
7277
7278
7279
7280
7281
7282
7283
7284
7285
7286
7287
7288
7289
7290
7291
7292
7293
7294
7295
7296
7297
7298
7299
7300
7301
7302
7303
7304
7305
7306
7307
7308
7309
7310
7311
7312
7313
7314
7315
7316
7317
7318
7319
7320
7321
7322
7323
7324
7325
7326
7327
7328
7329
7330
7331
7332
7333
7334
7335
7336
7337
7338
7339
7340
7341
7342
7343
7344
7345
7346
7347
7348
7349
7350
7351
7352
7353
7354
7355
7356
7357
7358
7359
7360
7361
7362
7363
7364
7365
7366
7367
7368
7369
7370
7371
7372
7373
7374
7375
7376
7377
7378
7379
7380
7381
7382
7383
7384
7385
7386
7387
7388
7389
7390
7391
7392
7393
7394
7395
7396
7397
7398
7399
7400
7401
7402
7403
7404
7405
7406
7407
7408
7409
7410
7411
7412
7413
7414
7415
7416
7417
7418
7419
7420
7421
7422
7423
7424
7425
7426
7427
7428
7429
7430
7431
7432
7433
7434
7435
7436
7437
7438
7439
7440
7441
7442
7443
7444
7445
7446
7447
7448
7449
7450
7451
7452
7453
7454
7455
7456
7457
7458
7459
7460
7461
7462
7463
7464
7465
7466
7467
7468
7469
7470
7471
7472
7473
7474
7475
7476
7477
7478
7479
7480
7481
7482
7483
7484
7485
7486
7487
7488
7489
7490
7491
7492
7493
7494
7495
7496
7497
7498
7499
7500
7501
7502
7503
7504
7505
7506
7507
7508
7509
7510
7511
7512
7513
7514
7515
7516
7517
7518
7519
7520
7521
7522
7523
7524
7525
7526
7527
7528
7529
7530
7531
7532
7533
7534
7535
7536
7537
7538
7539
7540
7541
7542
7543
7544
7545
7546
7547
7548
7549
7550
7551
7552
7553
7554
7555
7556
7557
7558
7559
7560
7561
7562
7563
7564
7565
7566
7567
7568
7569
7570
7571
7572
7573
7574
7575
7576
7577
7578
7579
7580
7581
7582
7583
7584
7585
7586
7587
7588
7589
7590
7591
7592
7593
7594
7595
7596
7597
7598
7599
7600
7601
7602
7603
7604
7605
7606
7607
7608
7609
7610
7611
7612
7613
7614
7615
7616
7617
7618
7619
7620
7621
7622
7623
7624
7625
7626
7627
7628
7629
7630
7631
7632
7633
7634
7635
7636
7637
7638
7639
7640
7641
7642
7643
7644
7645
7646
7647
7648
7649
7650
7651
7652
7653
7654
7655
7656
7657
7658
7659
7660
7661
7662
7663
7664
7665
7666
7667
7668
7669
7670
7671
7672
7673
7674
7675
7676
7677
7678
7679
7680
7681
7682
7683
7684
7685
7686
7687
7688
7689
7690
7691
7692
7693
7694
7695
7696
7697
7698
7699
7700
7701
7702
7703
7704
7705
7706
7707
7708
7709
7710
7711
7712
7713
7714
7715
7716
7717
7718
7719
7720
7721
7722
7723
7724
7725
7726
7727
7728
7729
7730
7731
7732
7733
7734
7735
7736
7737
7738
7739
7740
7741
7742
7743
7744
7745
7746
7747
7748
7749
7750
7751
7752
7753
7754
7755
7756
7757
7758
7759
7760
7761
7762
7763
7764
7765
7766
7767
7768
7769
7770
7771
7772
7773
7774
7775
7776
7777
7778
7779
7780
7781
7782
7783
7784
7785
7786
7787
7788
7789
7790
7791
7792
7793
7794
7795
7796
7797
7798
7799
7800
7801
7802
7803
7804
7805
7806
7807
7808
7809
7810
7811
7812
7813
7814
7815
7816
7817
7818
7819
7820
7821
7822
7823
7824
7825
7826
7827
7828
7829
7830
7831
7832
7833
7834
7835
7836
7837
7838
7839
7840
7841
7842
7843
7844
7845
7846
7847
7848
7849
7850
7851
7852
7853
7854
7855
7856
7857
7858
7859
7860
7861
7862
7863
7864
7865
7866
7867
7868
7869
7870
7871
7872
7873
7874
7875
7876
7877
7878
7879
7880
7881
7882
7883
7884
7885
7886
7887
7888
7889
7890
7891
7892
7893
7894
7895
7896
7897
7898
7899
7900
7901
7902
7903
7904
7905
7906
7907
7908
7909
7910
7911
7912
7913
7914
7915
7916
7917
7918
7919
7920
7921
7922
7923
7924
7925
7926
7927
7928
7929
7930
7931
7932
7933
7934
7935
7936
7937
7938
7939
7940
7941
7942
7943
7944
7945
7946
7947
7948
7949
7950
7951
7952
7953
7954
7955
7956
7957
7958
7959
7960
7961
7962
7963
7964
7965
7966
7967
7968
7969
7970
7971
7972
7973
7974
7975
7976
7977
7978
7979
7980
7981
7982
7983
7984
7985
7986
7987
7988
7989
7990
7991
7992
7993
7994
7995
7996
7997
7998
7999
8000
8001
8002
8003
8004
8005
8006
8007
8008
8009
8010
8011
8012
8013
8014
8015
8016
8017
8018
8019
8020
8021
8022
8023
8024
8025
8026
8027
8028
8029
8030
8031
8032
8033
8034
8035
8036
8037
8038
8039
8040
8041
8042
8043
8044
8045
8046
8047
8048
8049
8050
8051
8052
8053
8054
8055
8056
8057
8058
8059
8060
8061
8062
8063
8064
8065
8066
8067
8068
8069
8070
8071
8072
8073
8074
8075
8076
8077
8078
8079
8080
8081
8082
8083
8084
8085
8086
8087
8088
8089
8090
8091
8092
8093
8094
8095
8096
8097
8098
8099
8100
8101
8102
8103
8104
8105
8106
8107
8108
8109
8110
8111
8112
8113
8114
8115
8116
8117
8118
8119
8120
8121
8122
8123
8124
8125
8126
8127
8128
8129
8130
8131
8132
8133
8134
8135
8136
8137
8138
8139
8140
8141
8142
8143
8144
8145
8146
8147
8148
8149
8150
8151
8152
8153
8154
8155
8156
8157
8158
8159
8160
8161
8162
8163
8164
8165
8166
8167
8168
8169
8170
8171
8172
8173
8174
8175
8176
8177
8178
8179
8180
8181
8182
8183
8184
8185
8186
8187
8188
8189
8190
8191
8192
8193
8194
8195
8196
8197
8198
8199
8200
8201
8202
8203
8204
8205
8206
8207
8208
8209
8210
8211
8212
8213
8214
8215
8216
8217
8218
8219
8220
8221
8222
8223
8224
8225
8226
8227
8228
8229
8230
8231
8232
8233
8234
8235
8236
8237
8238
8239
8240
8241
8242
8243
8244
8245
8246
8247
8248
8249
8250
8251
8252
8253
8254
8255
8256
8257
8258
8259
8260
8261
8262
8263
8264
8265
8266
8267
8268
8269
8270
8271
8272
8273
8274
8275
8276
8277
8278
8279
8280
8281
8282
8283
8284
8285
8286
8287
8288
8289
8290
8291
8292
8293
8294
8295
8296
8297
8298
8299
8300
8301
8302
8303
8304
8305
8306
8307
8308
8309
8310
8311
8312
8313
8314
8315
8316
8317
8318
8319
8320
8321
8322
8323
8324
8325
8326
8327
8328
8329
8330
8331
8332
8333
8334
8335
8336
8337
8338
8339
8340
8341
8342
8343
8344
8345
8346
8347
8348
8349
8350
8351
8352
8353
8354
8355
8356
8357
8358
8359
8360
8361
8362
8363
8364
8365
8366
8367
8368
8369
8370
8371
8372
8373
8374
8375
8376
8377
8378
8379
8380
8381
8382
8383
8384
8385
8386
8387
8388
8389
8390
8391
8392
8393
8394
8395
8396
8397
8398
8399
8400
8401
8402
8403
8404
8405
8406
8407
8408
8409
8410
8411
8412
8413
8414
8415
8416
8417
8418
8419
8420
8421
8422
8423
8424
8425
8426
8427
8428
8429
8430
8431
8432
8433
8434
8435
8436
8437
8438
8439
8440
8441
8442
8443
8444
8445
8446
8447
8448
8449
8450
8451
8452
8453
8454
8455
8456
8457
8458
8459
8460
8461
8462
8463
8464
8465
8466
8467
8468
8469
8470
8471
8472
8473
8474
8475
8476
8477
8478
8479
8480
8481
8482
8483
8484
8485
8486
8487
8488
8489
8490
8491
8492
8493
8494
8495
8496
8497
8498
8499
8500
8501
8502
8503
8504
8505
8506
8507
8508
8509
8510
8511
8512
8513
8514
8515
8516
8517
8518
8519
8520
8521
8522
8523
8524
8525
8526
8527
8528
8529
8530
8531
8532
8533
8534
8535
8536
8537
8538
8539
8540
8541
8542
8543
8544
8545
8546
8547
8548
8549
8550
8551
8552
8553
8554
8555
8556
8557
8558
8559
8560
8561
8562
8563
8564
8565
8566
8567
8568
8569
8570
8571
8572
8573
8574
8575
8576
8577
8578
8579
8580
8581
8582
8583
8584
8585
8586
8587
8588
8589
8590
8591
8592
8593
8594
8595
8596
8597
8598
8599
8600
8601
8602
8603
8604
8605
8606
8607
8608
8609
8610
8611
8612
8613
8614
8615
8616
8617
8618
8619
8620
8621
8622
8623
8624
8625
8626
8627
8628
8629
8630
8631
8632
8633
8634
8635
8636
8637
8638
8639
8640
8641
8642
8643
8644
8645
8646
8647
8648
8649
8650
8651
8652
8653
8654
8655
8656
8657
8658
8659
8660
8661
8662
8663
8664
8665
8666
8667
8668
8669
8670
8671
8672
8673
8674
8675
8676
8677
8678
8679
8680
8681
8682
8683
8684
8685
8686
8687
8688
8689
8690
8691
8692
8693
8694
8695
8696
8697
8698
8699
8700
8701
8702
8703
8704
8705
8706
8707
8708
8709
8710
8711
8712
8713
8714
8715
8716
8717
8718
8719
8720
8721
8722
8723
8724
8725
8726
8727
8728
8729
8730
8731
8732
8733
8734
8735
8736
8737
8738
8739
8740
8741
8742
8743
8744
8745
8746
8747
8748
8749
8750
8751
8752
8753
8754
8755
8756
8757
8758
8759
8760
8761
8762
8763
8764
8765
8766
8767
8768
8769
8770
8771
8772
8773
8774
8775
8776
8777
8778
8779
8780
8781
8782
8783
8784
8785
8786
8787
8788
8789
8790
8791
8792
8793
8794
8795
8796
8797
8798
8799
8800
8801
8802
8803
8804
8805
8806
8807
8808
8809
8810
8811
8812
8813
8814
8815
8816
8817
8818
8819
8820
8821
8822
8823
8824
8825
8826
8827
8828
8829
8830
8831
8832
8833
8834
8835
8836
8837
8838
8839
8840
8841
8842
8843
8844
8845
8846
8847
8848
8849
8850
8851
8852
8853
8854
8855
8856
8857
8858
8859
8860
8861
8862
8863
8864
8865
8866
8867
8868
8869
8870
8871
8872
8873
8874
8875
8876
8877
8878
8879
8880
8881
8882
8883
8884
8885
8886
8887
8888
8889
8890
8891
8892
8893
8894
8895
8896
8897
8898
8899
8900
8901
8902
8903
8904
8905
8906
8907
8908
8909
8910
8911
8912
8913
8914
8915
8916
8917
8918
8919
8920
8921
8922
8923
8924
8925
8926
8927
8928
8929
8930
8931
8932
8933
8934
8935
8936
8937
8938
8939
8940
8941
8942
8943
8944
8945
8946
8947
8948
8949
8950
8951
8952
8953
8954
8955
8956
8957
8958
8959
8960
8961
8962
8963
8964
8965
8966
8967
8968
8969
8970
8971
8972
8973
8974
8975
8976
8977
8978
8979
8980
8981
8982
8983
8984
8985
8986
8987
8988
8989
8990
8991
8992
8993
8994
8995
8996
8997
8998
8999
9000
9001
9002
9003
9004
9005
9006
9007
9008
9009
9010
9011
9012
9013
9014
9015
9016
9017
9018
9019
9020
9021
9022
9023
9024
9025
9026
9027
9028
9029
9030
9031
9032
9033
9034
9035
9036
9037
9038
9039
9040
9041
9042
9043
9044
9045
9046
9047
9048
9049
9050
9051
9052
9053
9054
9055
9056
9057
9058
9059
9060
9061
9062
9063
9064
9065
9066
9067
9068
9069
9070
9071
9072
9073
9074
9075
9076
9077
9078
9079
9080
9081
9082
9083
9084
9085
9086
9087
9088
9089
9090
9091
9092
9093
9094
9095
9096
9097
9098
9099
9100
9101
9102
9103
9104
9105
9106
9107
9108
9109
9110
9111
9112
9113
9114
9115
9116
9117
9118
9119
9120
9121
9122
9123
9124
9125
9126
9127
9128
9129
9130
9131
9132
9133
9134
9135
9136
9137
9138
9139
9140
9141
9142
9143
9144
9145
9146
9147
9148
9149
9150
9151
9152
9153
9154
9155
9156
9157
9158
9159
9160
9161
9162
9163
9164
9165
9166
9167
9168
9169
9170
9171
9172
9173
9174
9175
9176
9177
9178
9179
9180
9181
9182
9183
9184
9185
9186
9187
9188
9189
9190
9191
9192
9193
9194
9195
9196
9197
9198
9199
9200
9201
9202
9203
9204
9205
9206
9207
9208
9209
9210
9211
9212
9213
9214
9215
9216
9217
9218
9219
9220
9221
9222
9223
9224
9225
9226
9227
9228
9229
9230
9231
9232
9233
9234
9235
9236
9237
9238
9239
9240
9241
9242
9243
9244
9245
9246
9247
9248
9249
9250
9251
9252
9253
9254
9255
9256
9257
9258
9259
9260
9261
9262
9263
9264
9265
9266
9267
9268
9269
9270
9271
9272
9273
9274
9275
9276
9277
9278
9279
9280
9281
9282
9283
9284
9285
9286
9287
9288
9289
9290
9291
9292
9293
9294
9295
9296
9297
9298
9299
9300
9301
9302
9303
9304
9305
9306
9307
9308
9309
9310
9311
9312
9313
9314
9315
9316
9317
9318
9319
9320
9321
9322
9323
9324
9325
9326
9327
9328
9329
9330
9331
9332
9333
9334
9335
9336
9337
9338
9339
9340
9341
9342
9343
9344
9345
9346
9347
9348
9349
9350
9351
9352
9353
9354
9355
9356
9357
9358
9359
9360
9361
9362
9363
9364
9365
9366
9367
9368
9369
9370
9371
9372
9373
9374
9375
9376
9377
9378
9379
9380
9381
9382
9383
9384
9385
9386
9387
9388
9389
9390
9391
9392
9393
9394
9395
9396
9397
9398
9399
9400
9401
9402
9403
9404
9405
9406
9407
9408
9409
9410
9411
9412
9413
9414
9415
9416
9417
9418
9419
9420
9421
9422
9423
9424
9425
9426
9427
9428
9429
9430
9431
9432
9433
9434
9435
9436
9437
9438
9439
9440
9441
9442
9443
9444
9445
9446
9447
9448
9449
9450
9451
9452
9453
9454
9455
9456
9457
9458
9459
9460
9461
9462
9463
9464
9465
9466
9467
9468
9469
9470
9471
9472
9473
9474
9475
9476
9477
9478
9479
9480
9481
9482
9483
9484
9485
9486
9487
9488
9489
9490
9491
9492
9493
9494
9495
9496
9497
9498
9499
9500
9501
9502
9503
9504
9505
9506
9507
9508
9509
9510
9511
9512
9513
9514
9515
9516
9517
9518
9519
9520
9521
9522
9523
9524
9525
9526
9527
9528
9529
9530
9531
9532
9533
9534
9535
9536
9537
9538
9539
9540
9541
9542
9543
9544
9545
9546
9547
9548
9549
9550
9551
9552
9553
9554
9555
9556
9557
9558
9559
9560
9561
9562
9563
9564
9565
9566
9567
9568
9569
9570
9571
9572
9573
9574
9575
9576
9577
9578
9579
9580
9581
9582
9583
9584
9585
9586
9587
9588
9589
9590
9591
9592
9593
9594
9595
9596
9597
9598
9599
9600
9601
9602
9603
9604
9605
9606
9607
9608
9609
9610
9611
9612
9613
9614
9615
9616
9617
9618
9619
9620
9621
9622
9623
9624
9625
9626
9627
9628
9629
9630
9631
9632
9633
9634
9635
9636
9637
9638
9639
9640
9641
9642
9643
9644
9645
9646
9647
9648
9649
9650
9651
9652
9653
9654
9655
9656
9657
9658
9659
9660
9661
9662
9663
9664
9665
9666
9667
9668
9669
9670
9671
9672
9673
9674
9675
9676
9677
9678
9679
9680
9681
9682
9683
9684
9685
9686
9687
9688
9689
9690
9691
9692
9693
9694
9695
9696
9697
9698
9699
9700
9701
9702
9703
9704
9705
9706
9707
9708
9709
9710
9711
9712
9713
9714
9715
9716
9717
9718
9719
9720
9721
9722
9723
9724
9725
9726
9727
9728
9729
9730
9731
9732
9733
9734
9735
9736
9737
9738
9739
9740
9741
9742
9743
9744
9745
9746
9747
9748
9749
9750
9751
9752
9753
9754
9755
9756
9757
9758
9759
9760
9761
9762
9763
9764
9765
9766
9767
9768
9769
9770
9771
9772
9773
9774
9775
9776
9777
9778
9779
9780
9781
9782
9783
9784
9785
9786
9787
9788
9789
9790
9791
9792
9793
9794
9795
9796
9797
9798
9799
9800
9801
9802
9803
9804
9805
9806
9807
9808
9809
9810
9811
9812
9813
9814
9815
9816
9817
9818
9819
9820
9821
9822
9823
9824
9825
9826
9827
9828
9829
9830
9831
9832
9833
9834
9835
9836
9837
9838
9839
9840
9841
9842
9843
9844
9845
9846
9847
9848
9849
9850
9851
9852
9853
9854
9855
9856
9857
9858
9859
9860
9861
9862
9863
9864
9865
9866
9867
9868
9869
9870
9871
9872
9873
9874
9875
9876
9877
9878
9879
9880
9881
9882
9883
9884
9885
9886
9887
9888
9889
9890
9891
9892
9893
9894
9895
9896
9897
9898
9899
9900
9901
9902
9903
9904
9905
9906
9907
9908
9909
9910
9911
9912
9913
9914
9915
9916
9917
9918
9919
9920
9921
9922
9923
9924
9925
9926
9927
9928
9929
9930
9931
9932
9933
9934
9935
9936
9937
9938
9939
9940
9941
9942
9943
9944
9945
9946
9947
9948
9949
9950
9951
9952
9953
9954
9955
9956
9957
9958
9959
9960
9961
9962
9963
9964
9965
9966
9967
9968
9969
9970
9971
9972
9973
9974
9975
9976
9977
9978
9979
9980
9981
9982
9983
9984
9985
9986
9987
9988
9989
9990
9991
9992
9993
9994
9995
9996
9997
9998
9999
10000
10001
10002
10003
10004
10005
10006
10007
10008
10009
10010
10011
10012
10013
10014
10015
10016
10017
10018
10019
10020
10021
10022
10023
10024
10025
10026
10027
10028
10029
10030
10031
10032
10033
10034
10035
10036
10037
10038
10039
10040
10041
10042
10043
10044
10045
10046
10047
10048
10049
10050
10051
10052
10053
10054
10055
10056
10057
10058
10059
10060
10061
10062
10063
10064
10065
10066
10067
10068
10069
10070
10071
10072
10073
10074
10075
10076
10077
10078
10079
10080
10081
10082
10083
10084
10085
10086
10087
10088
10089
10090
10091
10092
10093
10094
10095
10096
10097
10098
10099
10100
10101
10102
10103
10104
10105
10106
10107
10108
10109
10110
10111
10112
10113
10114
10115
10116
10117
10118
10119
10120
10121
10122
10123
10124
10125
10126
10127
10128
10129
10130
10131
10132
10133
10134
10135
10136
10137
10138
10139
10140
10141
10142
10143
10144
10145
10146
10147
10148
10149
10150
10151
10152
10153
10154
10155
10156
10157
10158
10159
10160
10161
10162
10163
10164
10165
10166
10167
10168
10169
10170
10171
10172
10173
10174
10175
10176
10177
10178
10179
10180
10181
10182
10183
10184
10185
10186
10187
10188
10189
10190
10191
10192
10193
10194
10195
10196
10197
10198
10199
10200
10201
10202
10203
10204
10205
10206
10207
10208
10209
10210
10211
10212
10213
10214
10215
10216
10217
10218
10219
10220
10221
10222
10223
10224
10225
10226
10227
10228
10229
10230
10231
10232
10233
10234
10235
10236
10237
10238
10239
10240
10241
10242
10243
10244
10245
10246
10247
10248
10249
10250
10251
10252
10253
10254
10255
10256
10257
10258
10259
10260
10261
10262
10263
10264
10265
10266
10267
10268
10269
10270
10271
10272
10273
10274
10275
10276
10277
10278
10279
10280
10281
10282
10283
10284
10285
10286
10287
10288
10289
10290
10291
10292
10293
10294
10295
10296
10297
10298
10299
10300
10301
10302
10303
10304
10305
10306
10307
10308
10309
10310
10311
10312
10313
10314
10315
10316
10317
10318
10319
10320
10321
10322
10323
10324
10325
10326
10327
10328
10329
10330
10331
10332
10333
10334
10335
10336
10337
10338
10339
10340
10341
10342
10343
10344
10345
10346
10347
10348
10349
10350
10351
10352
10353
10354
10355
10356
10357
10358
10359
10360
10361
10362
10363
10364
10365
10366
10367
10368
10369
10370
10371
10372
10373
10374
10375
10376
10377
10378
10379
10380
10381
10382
10383
10384
10385
10386
10387
10388
10389
10390
10391
10392
10393
10394
10395
10396
10397
10398
10399
10400
10401
10402
10403
10404
10405
10406
10407
10408
10409
10410
10411
10412
10413
10414
10415
10416
10417
10418
10419
10420
10421
10422
10423
10424
10425
10426
10427
10428
10429
10430
10431
10432
10433
10434
10435
10436
10437
10438
10439
10440
10441
10442
10443
10444
10445
10446
10447
10448
10449
10450
10451
10452
10453
10454
10455
10456
10457
10458
10459
10460
10461
10462
10463
10464
10465
10466
10467
10468
10469
10470
10471
10472
10473
10474
10475
10476
10477
10478
10479
10480
10481
10482
10483
10484
10485
10486
10487
10488
10489
10490
10491
10492
10493
10494
10495
10496
10497
10498
10499
10500
10501
10502
10503
10504
10505
10506
10507
10508
10509
10510
10511
10512
10513
10514
10515
10516
10517
10518
10519
10520
10521
10522
10523
10524
10525
10526
10527
10528
10529
10530
10531
10532
10533
10534
10535
10536
10537
10538
10539
10540
10541
10542
10543
10544
10545
10546
10547
10548
10549
10550
10551
10552
10553
10554
10555
10556
10557
10558
10559
10560
10561
10562
10563
10564
10565
10566
10567
10568
10569
10570
10571
10572
10573
10574
10575
10576
10577
10578
10579
10580
10581
10582
10583
10584
10585
10586
10587
10588
10589
10590
10591
10592
10593
10594
10595
10596
10597
10598
10599
10600
10601
10602
10603
10604
10605
10606
10607
10608
10609
10610
10611
10612
10613
10614
10615
10616
10617
10618
10619
10620
10621
10622
10623
10624
10625
10626
10627
10628
10629
10630
10631
10632
10633
10634
10635
10636
10637
10638
10639
10640
10641
10642
10643
10644
10645
10646
10647
10648
10649
10650
10651
10652
10653
10654
10655
10656
10657
10658
10659
10660
10661
10662
10663
10664
10665
10666
10667
10668
10669
10670
10671
10672
10673
10674
10675
10676
10677
10678
10679
10680
10681
10682
10683
10684
10685
10686
10687
10688
10689
10690
10691
10692
10693
10694
10695
10696
10697
10698
10699
10700
10701
10702
10703
10704
10705
10706
10707
10708
10709
10710
10711
10712
10713
10714
10715
10716
10717
10718
10719
10720
10721
10722
10723
10724
10725
10726
10727
10728
10729
10730
10731
10732
10733
10734
10735
10736
10737
10738
10739
10740
10741
10742
10743
10744
10745
10746
10747
10748
10749
10750
10751
10752
10753
10754
10755
10756
10757
10758
10759
10760
10761
10762
10763
10764
10765
10766
10767
10768
10769
10770
10771
10772
10773
10774
10775
10776
10777
10778
10779
10780
10781
10782
10783
10784
10785
10786
10787
10788
10789
10790
10791
10792
10793
10794
10795
10796
10797
10798
10799
10800
10801
10802
10803
10804
10805
10806
10807
10808
10809
10810
10811
10812
10813
10814
10815
10816
10817
10818
10819
10820
10821
10822
10823
10824
10825
10826
10827
10828
10829
10830
10831
10832
10833
10834
10835
10836
10837
10838
10839
10840
10841
10842
10843
10844
10845
10846
10847
10848
10849
10850
10851
10852
10853
10854
10855
10856
10857
10858
10859
10860
10861
10862
10863
10864
10865
10866
10867
10868
10869
10870
10871
10872
10873
10874
10875
10876
10877
10878
10879
10880
10881
10882
10883
10884
10885
10886
10887
10888
10889
10890
10891
10892
10893
10894
10895
10896
10897
10898
10899
10900
10901
10902
10903
10904
10905
10906
10907
10908
10909
10910
10911
10912
10913
10914
10915
10916
10917
10918
10919
10920
10921
10922
10923
10924
10925
10926
10927
10928
10929
10930
10931
10932
10933
10934
10935
10936
10937
10938
10939
10940
10941
10942
10943
10944
10945
10946
10947
10948
10949
10950
10951
10952
10953
10954
10955
10956
10957
10958
10959
10960
10961
10962
10963
10964
10965
10966
10967
10968
10969
10970
10971
10972
10973
10974
10975
10976
10977
10978
10979
10980
10981
10982
10983
10984
10985
10986
10987
10988
10989
10990
10991
10992
10993
10994
10995
10996
10997
10998
10999
11000
11001
11002
11003
11004
11005
11006
11007
11008
11009
11010
11011
11012
11013
11014
11015
11016
11017
11018
11019
11020
11021
11022
11023
11024
11025
11026
11027
11028
11029
11030
11031
11032
11033
11034
11035
11036
11037
11038
11039
11040
11041
11042
11043
11044
11045
11046
11047
11048
11049
11050
11051
11052
11053
11054
11055
11056
11057
11058
11059
11060
11061
11062
11063
11064
11065
11066
11067
11068
11069
11070
11071
11072
11073
11074
11075
11076
11077
11078
11079
11080
11081
11082
11083
11084
11085
11086
11087
11088
11089
11090
11091
11092
11093
11094
11095
11096
11097
11098
11099
11100
11101
11102
11103
11104
11105
11106
11107
11108
11109
11110
11111
11112
11113
11114
11115
11116
11117
11118
11119
11120
11121
11122
11123
11124
11125
11126
11127
11128
11129
11130
11131
11132
11133
11134
11135
11136
11137
11138
11139
11140
11141
11142
11143
11144
11145
11146
11147
11148
11149
11150
11151
11152
11153
11154
11155
11156
11157
11158
11159
11160
11161
11162
11163
11164
11165
11166
11167
11168
11169
11170
11171
11172
11173
11174
11175
11176
11177
11178
11179
11180
11181
11182
11183
11184
11185
11186
11187
11188
11189
11190
11191
11192
11193
11194
11195
11196
11197
11198
11199
11200
11201
11202
11203
11204
11205
11206
11207
11208
11209
11210
11211
11212
11213
11214
11215
11216
11217
11218
11219
11220
11221
11222
11223
11224
11225
11226
11227
11228
11229
11230
11231
11232
11233
11234
11235
11236
11237
11238
11239
11240
11241
11242
11243
11244
11245
11246
11247
11248
11249
11250
11251
11252
11253
11254
11255
11256
11257
11258
11259
11260
11261
11262
11263
11264
11265
11266
11267
11268
11269
11270
11271
11272
11273
11274
11275
11276
11277
11278
11279
11280
11281
11282
11283
11284
11285
11286
11287
11288
11289
11290
11291
11292
11293
11294
11295
11296
11297
11298
11299
11300
11301
11302
11303
11304
11305
11306
11307
11308
11309
11310
11311
11312
11313
11314
11315
11316
11317
11318
11319
11320
11321
11322
11323
11324
11325
11326
11327
11328
11329
11330
11331
11332
11333
11334
11335
11336
11337
11338
11339
11340
11341
11342
11343
11344
11345
11346
11347
11348
11349
11350
11351
11352
11353
11354
11355
11356
11357
11358
11359
11360
11361
11362
11363
11364
11365
11366
11367
11368
11369
11370
11371
11372
11373
11374
11375
11376
11377
11378
11379
11380
11381
11382
11383
11384
11385
11386
11387
11388
11389
11390
11391
11392
11393
11394
11395
11396
11397
11398
11399
11400
11401
11402
11403
11404
11405
11406
11407
11408
11409
11410
11411
11412
11413
11414
11415
11416
11417
11418
11419
11420
11421
11422
11423
11424
11425
11426
11427
11428
11429
11430
11431
11432
11433
11434
11435
11436
11437
11438
11439
11440
11441
11442
11443
11444
11445
11446
11447
11448
11449
11450
11451
11452
11453
11454
11455
11456
11457
11458
11459
11460
11461
11462
11463
11464
11465
11466
11467
11468
11469
11470
11471
11472
11473
11474
11475
11476
11477
11478
11479
11480
11481
11482
11483
11484
11485
11486
11487
11488
11489
11490
11491
11492
11493
11494
11495
11496
11497
11498
11499
11500
11501
11502
11503
11504
11505
11506
11507
11508
11509
11510
11511
11512
11513
11514
11515
11516
11517
11518
11519
11520
11521
11522
11523
11524
11525
11526
11527
11528
11529
11530
11531
11532
11533
11534
11535
11536
11537
11538
11539
11540
11541
11542
11543
11544
11545
11546
11547
11548
11549
11550
11551
11552
11553
11554
11555
11556
11557
11558
11559
11560
11561
11562
11563
11564
11565
11566
11567
11568
11569
11570
11571
11572
11573
11574
11575
11576
11577
11578
11579
11580
11581
11582
11583
11584
11585
11586
11587
11588
11589
11590
11591
11592
11593
11594
11595
11596
11597
11598
11599
11600
11601
11602
11603
11604
11605
11606
11607
11608
11609
11610
11611
11612
11613
11614
11615
11616
11617
11618
11619
11620
11621
11622
11623
11624
11625
11626
11627
11628
11629
11630
11631
11632
11633
11634
11635
11636
11637
11638
11639
11640
11641
11642
11643
11644
11645
11646
11647
11648
11649
11650
11651
11652
11653
11654
11655
11656
11657
11658
11659
11660
11661
11662
11663
11664
11665
11666
11667
11668
11669
11670
11671
11672
11673
11674
11675
11676
11677
11678
11679
11680
11681
11682
11683
11684
11685
11686
11687
11688
11689
11690
11691
11692
11693
11694
11695
11696
11697
11698
11699
11700
11701
11702
11703
11704
11705
11706
11707
11708
11709
11710
11711
11712
11713
11714
11715
11716
11717
11718
11719
11720
11721
11722
11723
11724
11725
11726
11727
11728
11729
11730
11731
11732
11733
11734
11735
11736
11737
11738
11739
11740
11741
11742
11743
11744
11745
11746
11747
11748
11749
11750
11751
11752
11753
11754
11755
11756
11757
11758
11759
11760
11761
11762
11763
11764
11765
11766
11767
11768
11769
11770
11771
11772
11773
11774
11775
11776
11777
11778
11779
11780
11781
11782
11783
11784
11785
11786
11787
11788
11789
11790
11791
11792
11793
11794
11795
11796
11797
11798
11799
11800
11801
11802
11803
11804
11805
11806
11807
11808
11809
11810
11811
11812
11813
11814
11815
11816
11817
11818
11819
11820
11821
11822
11823
11824
11825
11826
11827
11828
11829
11830
11831
11832
11833
11834
11835
11836
11837
11838
11839
11840
11841
11842
11843
11844
11845
11846
11847
11848
11849
11850
11851
11852
11853
11854
11855
11856
11857
11858
11859
11860
11861
11862
11863
11864
11865
11866
11867
11868
11869
11870
11871
11872
11873
11874
11875
11876
11877
11878
11879
11880
11881
11882
11883
11884
11885
11886
11887
11888
11889
11890
11891
11892
11893
11894
11895
11896
11897
11898
11899
11900
11901
11902
11903
11904
11905
11906
11907
11908
11909
11910
11911
11912
11913
11914
11915
11916
11917
11918
11919
11920
11921
11922
11923
11924
11925
11926
11927
11928
11929
11930
11931
11932
11933
11934
11935
11936
11937
11938
11939
11940
11941
11942
11943
11944
11945
11946
11947
11948
11949
11950
11951
11952
11953
11954
11955
11956
11957
11958
11959
11960
11961
11962
11963
11964
11965
11966
11967
11968
11969
11970
11971
11972
11973
11974
11975
11976
11977
11978
11979
11980
11981
11982
11983
11984
11985
11986
11987
11988
11989
11990
11991
11992
11993
11994
11995
11996
11997
11998
11999
12000
12001
12002
12003
12004
12005
12006
12007
12008
12009
12010
12011
12012
12013
12014
12015
12016
12017
12018
12019
12020
12021
12022
12023
12024
12025
12026
12027
12028
12029
12030
12031
12032
12033
12034
12035
12036
12037
12038
12039
12040
12041
12042
12043
12044
12045
12046
12047
12048
12049
12050
12051
12052
12053
12054
12055
12056
12057
12058
12059
12060
12061
12062
12063
12064
12065
12066
12067
12068
12069
12070
12071
12072
12073
12074
12075
12076
12077
12078
12079
12080
12081
12082
12083
12084
12085
12086
12087
12088
12089
12090
12091
12092
12093
12094
12095
12096
12097
12098
12099
12100
12101
12102
12103
12104
12105
12106
12107
12108
12109
12110
12111
12112
12113
12114
12115
12116
12117
12118
12119
12120
12121
12122
12123
12124
12125
12126
12127
12128
12129
12130
12131
12132
12133
12134
12135
12136
12137
12138
12139
12140
12141
12142
12143
12144
12145
12146
12147
12148
12149
12150
12151
12152
12153
12154
12155
12156
12157
12158
12159
12160
12161
12162
12163
12164
12165
12166
12167
12168
12169
12170
12171
12172
12173
12174
12175
12176
12177
12178
12179
12180
12181
12182
12183
12184
12185
12186
12187
12188
12189
12190
12191
12192
12193
12194
12195
12196
12197
12198
12199
12200
12201
12202
12203
12204
12205
12206
12207
12208
12209
12210
12211
12212
12213
12214
12215
12216
12217
12218
12219
12220
12221
12222
12223
12224
12225
12226
12227
12228
12229
12230
12231
12232
12233
12234
12235
12236
12237
12238
12239
12240
12241
12242
12243
12244
12245
12246
12247
12248
12249
12250
12251
12252
12253
12254
12255
12256
12257
12258
12259
12260
12261
12262
12263
12264
12265
12266
12267
12268
12269
12270
12271
12272
12273
12274
12275
12276
12277
12278
12279
12280
12281
12282
12283
12284
12285
12286
12287
12288
12289
12290
12291
12292
12293
12294
12295
12296
12297
12298
12299
12300
12301
12302
12303
12304
12305
12306
12307
12308
12309
12310
12311
12312
12313
12314
12315
12316
12317
12318
12319
12320
12321
12322
12323
12324
12325
12326
12327
12328
12329
12330
12331
12332
12333
12334
12335
12336
12337
12338
12339
12340
12341
12342
12343
12344
12345
12346
12347
12348
12349
12350
12351
12352
12353
12354
12355
12356
12357
12358
12359
12360
12361
12362
12363
12364
12365
12366
12367
12368
12369
12370
12371
12372
12373
12374
12375
12376
12377
12378
12379
12380
12381
12382
12383
12384
12385
12386
12387
12388
12389
12390
12391
12392
12393
12394
12395
12396
12397
12398
12399
12400
12401
12402
12403
12404
12405
12406
12407
12408
12409
12410
12411
12412
12413
12414
12415
12416
12417
12418
12419
12420
12421
12422
12423
12424
12425
12426
12427
12428
12429
12430
12431
12432
12433
12434
12435
12436
12437
12438
12439
12440
12441
12442
12443
12444
12445
12446
12447
12448
12449
12450
12451
12452
12453
12454
12455
12456
12457
12458
12459
12460
12461
12462
12463
12464
12465
12466
12467
12468
12469
12470
12471
12472
12473
12474
12475
12476
12477
12478
12479
12480
12481
12482
12483
12484
12485
12486
12487
12488
12489
12490
12491
12492
12493
12494
12495
12496
12497
12498
12499
12500
12501
12502
12503
12504
12505
12506
12507
12508
12509
12510
12511
12512
12513
12514
12515
12516
12517
12518
12519
12520
12521
12522
12523
12524
12525
12526
12527
12528
12529
12530
12531
12532
12533
12534
12535
12536
12537
12538
12539
12540
12541
12542
12543
12544
12545
12546
12547
12548
12549
12550
12551
12552
12553
12554
12555
12556
12557
12558
12559
12560
12561
12562
12563
12564
12565
12566
12567
12568
12569
12570
12571
12572
12573
12574
12575
12576
12577
12578
12579
12580
12581
12582
12583
12584
12585
12586
12587
12588
12589
12590
12591
12592
12593
12594
12595
12596
12597
12598
12599
12600
12601
12602
12603
12604
12605
12606
12607
12608
12609
12610
12611
12612
12613
12614
12615
12616
12617
12618
12619
12620
12621
12622
12623
12624
12625
12626
12627
12628
12629
12630
12631
12632
12633
12634
12635
12636
12637
12638
12639
12640
12641
12642
12643
12644
12645
12646
12647
12648
12649
12650
12651
12652
12653
12654
12655
12656
12657
12658
12659
12660
12661
12662
12663
12664
12665
12666
12667
12668
12669
12670
12671
12672
12673
12674
12675
12676
12677
12678
12679
12680
12681
12682
12683
12684
12685
12686
12687
12688
12689
12690
12691
12692
12693
12694
12695
12696
12697
12698
12699
12700
12701
12702
12703
12704
12705
12706
12707
12708
12709
12710
12711
12712
12713
12714
12715
12716
12717
12718
12719
12720
12721
12722
12723
12724
12725
12726
12727
12728
12729
12730
12731
12732
12733
12734
12735
12736
12737
12738
12739
12740
12741
12742
12743
12744
12745
12746
12747
12748
12749
12750
12751
12752
12753
12754
12755
12756
12757
12758
12759
12760
12761
12762
12763
12764
12765
12766
12767
12768
12769
12770
12771
12772
12773
12774
12775
12776
12777
12778
12779
12780
12781
12782
12783
12784
12785
12786
12787
12788
12789
12790
12791
12792
12793
12794
12795
12796
12797
12798
12799
12800
12801
12802
12803
12804
12805
12806
12807
12808
12809
12810
12811
12812
12813
12814
12815
12816
12817
12818
12819
12820
12821
12822
12823
12824
12825
12826
12827
12828
12829
12830
12831
12832
12833
12834
12835
12836
12837
12838
12839
12840
12841
12842
12843
12844
12845
12846
12847
12848
12849
12850
12851
12852
12853
12854
12855
12856
12857
12858
12859
12860
12861
12862
12863
12864
12865
12866
12867
12868
12869
12870
12871
12872
12873
12874
12875
12876
12877
12878
12879
12880
12881
12882
12883
12884
12885
12886
12887
12888
12889
12890
12891
12892
12893
12894
12895
12896
12897
12898
12899
12900
12901
12902
12903
12904
12905
12906
12907
12908
12909
12910
12911
12912
12913
12914
12915
12916
12917
12918
12919
12920
12921
12922
12923
12924
12925
12926
12927
12928
12929
12930
12931
12932
12933
12934
12935
12936
12937
12938
12939
12940
12941
12942
12943
12944
12945
12946
12947
12948
12949
12950
12951
12952
12953
12954
12955
12956
12957
12958
12959
12960
12961
12962
12963
12964
12965
12966
12967
12968
12969
12970
12971
12972
12973
12974
12975
12976
12977
12978
12979
12980
12981
12982
12983
12984
12985
12986
12987
12988
12989
12990
12991
12992
12993
12994
12995
12996
12997
12998
12999
13000
13001
13002
13003
13004
13005
13006
13007
13008
13009
13010
13011
13012
13013
13014
13015
13016
13017
13018
13019
13020
13021
13022
13023
13024
13025
13026
13027
13028
13029
13030
13031
13032
13033
13034
13035
13036
13037
13038
13039
13040
13041
13042
13043
13044
13045
13046
13047
13048
13049
13050
13051
13052
13053
13054
13055
13056
13057
13058
13059
13060
13061
13062
13063
13064
13065
13066
13067
13068
13069
13070
13071
13072
13073
13074
13075
13076
13077
13078
13079
13080
13081
13082
13083
13084
13085
13086
13087
13088
13089
13090
13091
13092
13093
13094
13095
13096
13097
13098
13099
13100
13101
13102
13103
13104
13105
13106
13107
13108
13109
13110
13111
13112
13113
13114
13115
13116
13117
13118
13119
13120
13121
13122
13123
13124
13125
13126
13127
13128
13129
13130
13131
13132
13133
13134
13135
13136
13137
13138
13139
13140
13141
13142
13143
13144
13145
13146
13147
13148
13149
13150
13151
13152
13153
13154
13155
13156
13157
13158
13159
13160
13161
13162
13163
13164
13165
13166
13167
13168
13169
13170
13171
13172
13173
13174
13175
13176
13177
13178
13179
13180
13181
13182
13183
13184
13185
13186
13187
13188
13189
13190
13191
13192
13193
13194
13195
13196
13197
13198
13199
13200
13201
13202
13203
13204
13205
13206
13207
13208
13209
13210
13211
13212
13213
13214
13215
13216
13217
13218
13219
13220
13221
13222
13223
13224
13225
13226
13227
13228
13229
13230
13231
13232
13233
13234
13235
13236
13237
13238
13239
13240
13241
13242
13243
13244
13245
13246
13247
13248
13249
13250
13251
13252
13253
13254
13255
13256
13257
13258
13259
13260
13261
13262
13263
13264
13265
13266
13267
13268
13269
13270
13271
13272
13273
13274
13275
13276
13277
13278
13279
13280
13281
13282
13283
13284
13285
13286
13287
13288
13289
13290
13291
13292
13293
13294
13295
13296
13297
13298
13299
13300
13301
13302
13303
13304
13305
13306
13307
13308
13309
13310
13311
13312
13313
13314
13315
13316
13317
13318
13319
13320
13321
13322
13323
13324
13325
13326
13327
13328
13329
13330
13331
13332
13333
13334
13335
13336
13337
13338
13339
13340
13341
13342
13343
13344
13345
13346
13347
13348
13349
13350
13351
13352
13353
13354
13355
13356
13357
13358
13359
13360
13361
13362
13363
13364
13365
13366
13367
13368
13369
13370
13371
13372
13373
13374
13375
13376
13377
13378
13379
13380
13381
13382
13383
13384
13385
13386
13387
13388
13389
13390
13391
13392
13393
13394
13395
13396
13397
13398
13399
13400
13401
13402
13403
13404
13405
13406
13407
13408
13409
13410
13411
13412
13413
13414
13415
13416
13417
13418
13419
13420
13421
13422
13423
13424
13425
13426
13427
13428
13429
13430
13431
13432
13433
13434
13435
13436
13437
13438
13439
13440
13441
13442
13443
13444
13445
13446
13447
13448
13449
13450
13451
13452
13453
13454
13455
13456
13457
13458
13459
13460
13461
13462
13463
13464
13465
13466
13467
13468
13469
13470
13471
13472
13473
13474
13475
13476
13477
13478
13479
13480
13481
13482
13483
13484
13485
13486
13487
13488
13489
13490
13491
13492
13493
13494
13495
13496
13497
13498
13499
13500
13501
13502
13503
13504
13505
13506
13507
13508
13509
13510
13511
13512
13513
13514
13515
13516
13517
13518
13519
13520
13521
13522
13523
13524
13525
13526
13527
13528
13529
13530
13531
13532
13533
13534
13535
13536
13537
13538
13539
13540
13541
13542
13543
13544
13545
13546
13547
13548
13549
13550
13551
13552
13553
13554
13555
13556
13557
13558
13559
13560
13561
13562
13563
13564
13565
13566
13567
13568
13569
13570
13571
13572
13573
13574
13575
13576
13577
13578
13579
13580
13581
13582
13583
13584
13585
13586
13587
13588
13589
13590
13591
13592
13593
13594
13595
13596
13597
13598
13599
13600
13601
13602
13603
13604
13605
13606
13607
13608
13609
13610
13611
13612
13613
13614
13615
13616
13617
13618
13619
13620
13621
13622
13623
13624
13625
13626
13627
13628
13629
13630
13631
13632
13633
13634
13635
13636
13637
13638
13639
13640
13641
13642
13643
13644
13645
13646
13647
13648
13649
13650
13651
13652
13653
13654
13655
13656
13657
13658
13659
13660
13661
13662
13663
13664
13665
13666
13667
13668
13669
13670
13671
13672
13673
13674
13675
13676
13677
13678
13679
13680
13681
13682
13683
13684
13685
13686
13687
13688
13689
13690
13691
13692
13693
13694
13695
13696
13697
13698
13699
13700
13701
13702
13703
13704
13705
13706
13707
13708
13709
13710
13711
13712
13713
13714
13715
13716
13717
13718
13719
13720
13721
13722
13723
13724
13725
13726
13727
13728
13729
13730
13731
13732
13733
13734
13735
13736
13737
13738
13739
13740
13741
13742
13743
13744
13745
13746
13747
13748
13749
13750
13751
13752
13753
13754
13755
13756
13757
13758
13759
13760
13761
13762
13763
13764
13765
13766
13767
13768
13769
13770
13771
13772
13773
13774
13775
13776
13777
13778
13779
13780
13781
13782
13783
13784
13785
13786
13787
13788
13789
13790
13791
13792
13793
13794
13795
13796
13797
13798
13799
13800
13801
13802
13803
13804
13805
13806
13807
13808
13809
13810
13811
13812
13813
13814
13815
13816
13817
13818
13819
13820
13821
13822
13823
13824
13825
13826
13827
13828
13829
13830
13831
13832
13833
13834
13835
13836
13837
13838
13839
13840
13841
13842
13843
13844
13845
13846
13847
13848
13849
13850
13851
13852
13853
13854
13855
13856
13857
13858
13859
13860
13861
13862
13863
13864
13865
13866
13867
13868
13869
13870
13871
13872
13873
13874
13875
13876
13877
13878
13879
13880
13881
13882
13883
13884
13885
13886
13887
13888
13889
13890
13891
13892
13893
13894
13895
13896
13897
13898
13899
13900
13901
13902
13903
13904
13905
13906
13907
13908
13909
13910
13911
13912
13913
13914
13915
13916
13917
13918
13919
13920
13921
13922
13923
13924
13925
13926
13927
13928
13929
13930
13931
13932
13933
13934
13935
13936
13937
13938
13939
13940
13941
13942
13943
13944
13945
13946
13947
13948
13949
13950
13951
13952
13953
13954
13955
13956
13957
13958
13959
13960
13961
13962
13963
13964
13965
13966
13967
13968
13969
13970
13971
13972
13973
13974
13975
13976
13977
13978
13979
13980
13981
13982
13983
13984
13985
13986
13987
13988
13989
13990
13991
13992
13993
13994
13995
13996
13997
13998
13999
14000
14001
14002
14003
14004
14005
14006
14007
14008
14009
14010
14011
14012
14013
14014
14015
14016
14017
14018
14019
14020
14021
14022
14023
14024
14025
14026
14027
14028
14029
14030
14031
14032
14033
14034
14035
14036
14037
14038
14039
14040
14041
14042
14043
14044
14045
14046
14047
14048
14049
14050
14051
14052
14053
14054
14055
14056
14057
14058
14059
14060
14061
14062
14063
14064
14065
14066
14067
14068
14069
14070
14071
14072
14073
14074
14075
14076
14077
14078
14079
14080
14081
14082
14083
14084
14085
14086
14087
14088
14089
14090
14091
14092
14093
14094
14095
14096
14097
14098
14099
14100
14101
14102
14103
14104
14105
14106
14107
14108
14109
14110
14111
14112
14113
14114
14115
14116
14117
14118
14119
14120
14121
14122
14123
14124
14125
14126
14127
14128
14129
14130
14131
14132
14133
14134
14135
14136
14137
14138
14139
14140
14141
14142
14143
14144
14145
14146
14147
14148
14149
14150
14151
14152
14153
14154
14155
14156
14157
14158
14159
14160
14161
14162
14163
14164
14165
14166
14167
14168
14169
14170
14171
14172
14173
14174
14175
14176
14177
14178
14179
14180
14181
14182
14183
14184
14185
14186
14187
14188
14189
14190
14191
14192
14193
14194
14195
14196
14197
14198
14199
14200
14201
14202
14203
14204
14205
14206
14207
14208
14209
14210
14211
14212
14213
14214
14215
14216
14217
14218
14219
14220
14221
14222
14223
14224
14225
14226
14227
14228
14229
14230
14231
14232
14233
14234
14235
14236
14237
14238
14239
14240
14241
14242
14243
14244
14245
14246
14247
14248
14249
14250
14251
14252
14253
14254
14255
14256
14257
14258
14259
14260
14261
14262
14263
14264
14265
14266
14267
14268
14269
14270
14271
14272
14273
14274
14275
14276
14277
14278
14279
14280
14281
14282
14283
14284
14285
14286
14287
14288
14289
14290
14291
14292
14293
14294
14295
14296
14297
14298
14299
14300
14301
14302
14303
14304
14305
14306
14307
14308
14309
14310
14311
14312
14313
14314
14315
14316
14317
14318
14319
14320
14321
14322
14323
14324
14325
14326
14327
14328
14329
14330
14331
14332
14333
14334
14335
14336
14337
14338
14339
14340
14341
14342
14343
14344
14345
14346
14347
14348
14349
14350
14351
14352
14353
14354
14355
14356
14357
14358
14359
14360
14361
14362
14363
14364
14365
14366
14367
14368
14369
14370
14371
14372
14373
14374
14375
14376
14377
14378
14379
14380
14381
14382
14383
14384
14385
14386
14387
14388
14389
14390
14391
14392
14393
14394
14395
14396
14397
14398
14399
14400
14401
14402
14403
14404
14405
14406
14407
14408
14409
14410
14411
14412
14413
14414
14415
14416
14417
14418
14419
14420
14421
14422
14423
14424
14425
14426
14427
14428
14429
14430
14431
14432
14433
14434
14435
14436
14437
14438
14439
14440
14441
14442
14443
14444
14445
14446
14447
14448
14449
14450
14451
14452
14453
14454
14455
14456
14457
14458
14459
14460
14461
14462
14463
14464
14465
14466
14467
14468
14469
14470
14471
14472
14473
14474
14475
14476
14477
14478
14479
14480
14481
14482
14483
14484
14485
14486
14487
14488
14489
14490
14491
14492
14493
14494
14495
14496
14497
14498
14499
14500
14501
14502
14503
14504
14505
14506
14507
14508
14509
14510
14511
14512
14513
14514
14515
14516
14517
14518
14519
14520
14521
14522
14523
14524
14525
14526
14527
14528
14529
14530
14531
14532
14533
14534
14535
14536
14537
14538
14539
14540
14541
14542
14543
14544
14545
14546
14547
14548
14549
14550
14551
14552
14553
14554
14555
14556
14557
14558
14559
14560
14561
14562
14563
14564
14565
14566
14567
14568
14569
14570
14571
14572
14573
14574
14575
14576
14577
14578
14579
14580
14581
14582
14583
14584
14585
14586
14587
14588
14589
14590
14591
14592
14593
14594
14595
14596
14597
14598
14599
14600
14601
14602
14603
14604
14605
14606
14607
14608
14609
14610
14611
14612
14613
14614
14615
14616
14617
14618
14619
14620
14621
14622
14623
14624
14625
14626
14627
14628
14629
14630
14631
14632
14633
14634
14635
14636
14637
14638
14639
14640
14641
14642
14643
14644
14645
14646
14647
14648
14649
14650
14651
14652
14653
14654
14655
14656
14657
14658
14659
14660
14661
14662
14663
14664
14665
14666
14667
14668
14669
14670
14671
14672
14673
14674
14675
14676
14677
14678
14679
14680
14681
14682
14683
14684
14685
14686
14687
14688
14689
14690
14691
14692
14693
14694
14695
14696
14697
14698
14699
14700
14701
14702
14703
14704
14705
14706
14707
14708
14709
14710
14711
14712
14713
14714
14715
14716
14717
14718
14719
14720
14721
14722
14723
14724
14725
14726
14727
14728
14729
14730
14731
14732
14733
14734
14735
14736
14737
14738
14739
14740
14741
14742
14743
14744
14745
14746
14747
14748
14749
14750
14751
14752
14753
14754
14755
14756
14757
14758
14759
14760
14761
14762
14763
14764
14765
14766
14767
14768
14769
14770
14771
14772
14773
14774
14775
14776
14777
14778
14779
14780
14781
14782
14783
14784
14785
14786
14787
14788
14789
14790
14791
14792
14793
14794
14795
14796
14797
14798
14799
14800
14801
14802
14803
14804
14805
14806
14807
14808
14809
14810
14811
14812
14813
14814
14815
14816
14817
14818
14819
14820
14821
14822
14823
14824
14825
14826
14827
14828
14829
14830
14831
14832
14833
14834
14835
14836
14837
14838
14839
14840
14841
14842
14843
14844
14845
14846
14847
14848
14849
14850
14851
14852
14853
14854
14855
14856
14857
14858
14859
14860
14861
14862
14863
14864
14865
14866
14867
14868
14869
14870
14871
14872
14873
14874
14875
14876
14877
14878
14879
14880
14881
14882
14883
14884
14885
14886
14887
14888
14889
14890
14891
14892
14893
14894
14895
14896
14897
14898
14899
14900
14901
14902
14903
14904
14905
14906
14907
14908
14909
14910
14911
14912
14913
14914
14915
14916
14917
14918
14919
14920
14921
14922
14923
14924
14925
14926
14927
14928
14929
14930
14931
14932
14933
14934
14935
14936
14937
14938
14939
14940
14941
14942
14943
14944
14945
14946
14947
14948
14949
14950
14951
14952
14953
14954
14955
14956
14957
14958
14959
14960
14961
14962
14963
14964
14965
14966
14967
14968
14969
14970
14971
14972
14973
14974
14975
14976
14977
14978
14979
14980
14981
14982
14983
14984
14985
14986
14987
14988
14989
14990
14991
14992
14993
14994
14995
14996
14997
14998
14999
15000
15001
15002
15003
15004
15005
15006
15007
15008
15009
15010
15011
15012
15013
15014
15015
15016
15017
15018
15019
15020
15021
15022
15023
15024
15025
15026
15027
15028
15029
15030
15031
15032
15033
15034
15035
15036
15037
15038
15039
15040
15041
15042
15043
15044
15045
15046
15047
15048
15049
15050
15051
15052
15053
15054
15055
15056
15057
15058
15059
15060
15061
15062
15063
15064
15065
15066
15067
15068
15069
15070
15071
15072
15073
15074
15075
15076
15077
15078
15079
15080
15081
15082
15083
15084
15085
15086
15087
15088
15089
15090
15091
15092
15093
15094
15095
15096
15097
15098
15099
15100
15101
15102
15103
15104
15105
15106
15107
15108
15109
15110
15111
15112
15113
15114
15115
15116
15117
15118
15119
15120
15121
15122
15123
15124
15125
15126
15127
15128
15129
15130
15131
15132
15133
15134
15135
15136
15137
15138
15139
15140
15141
15142
15143
15144
15145
15146
15147
15148
15149
15150
15151
15152
15153
15154
15155
15156
15157
15158
15159
15160
15161
15162
15163
15164
15165
15166
15167
15168
15169
15170
15171
15172
15173
15174
15175
15176
15177
15178
15179
15180
15181
15182
15183
15184
15185
15186
15187
15188
15189
15190
15191
15192
15193
15194
15195
15196
15197
15198
15199
15200
15201
15202
15203
15204
15205
15206
15207
15208
15209
15210
15211
15212
15213
15214
15215
15216
15217
15218
15219
15220
15221
15222
15223
15224
15225
15226
15227
15228
15229
15230
15231
15232
15233
15234
15235
15236
15237
15238
15239
15240
15241
15242
15243
15244
15245
15246
15247
15248
15249
15250
15251
15252
15253
15254
15255
15256
15257
15258
15259
15260
15261
15262
15263
15264
15265
15266
15267
15268
15269
15270
15271
15272
15273
15274
15275
15276
15277
15278
15279
15280
15281
15282
15283
15284
15285
15286
15287
15288
15289
15290
15291
15292
15293
15294
15295
15296
15297
15298
15299
15300
15301
15302
15303
15304
15305
15306
15307
15308
15309
15310
15311
15312
15313
15314
15315
15316
15317
15318
15319
15320
15321
15322
15323
15324
15325
15326
15327
15328
15329
15330
15331
15332
15333
15334
15335
15336
15337
15338
15339
15340
15341
15342
15343
15344
15345
15346
15347
15348
15349
15350
15351
15352
15353
15354
15355
15356
15357
15358
15359
15360
15361
15362
15363
15364
15365
15366
15367
15368
15369
15370
15371
15372
15373
15374
15375
15376
15377
15378
15379
15380
15381
15382
15383
15384
15385
15386
15387
15388
15389
15390
15391
15392
15393
15394
15395
15396
15397
15398
15399
15400
15401
15402
15403
15404
15405
15406
15407
15408
15409
15410
15411
15412
15413
15414
15415
15416
15417
15418
15419
15420
15421
15422
15423
15424
15425
15426
15427
15428
15429
15430
15431
15432
15433
15434
15435
15436
15437
15438
15439
15440
15441
15442
15443
15444
15445
15446
15447
15448
15449
15450
15451
15452
15453
15454
15455
15456
15457
15458
15459
15460
15461
15462
15463
15464
15465
15466
15467
15468
15469
15470
15471
15472
15473
15474
15475
15476
15477
15478
15479
15480
15481
15482
15483
15484
15485
15486
15487
15488
15489
15490
15491
15492
15493
15494
15495
15496
15497
15498
15499
15500
15501
15502
15503
15504
15505
15506
15507
15508
15509
15510
15511
15512
15513
15514
15515
15516
15517
15518
15519
15520
15521
15522
15523
15524
15525
15526
15527
15528
15529
15530
15531
15532
15533
15534
15535
15536
15537
15538
15539
15540
15541
15542
15543
15544
15545
15546
15547
15548
15549
15550
15551
15552
15553
15554
15555
15556
15557
15558
15559
15560
15561
15562
15563
15564
15565
15566
15567
15568
15569
15570
15571
15572
15573
15574
15575
15576
15577
15578
15579
15580
15581
15582
15583
15584
15585
15586
15587
15588
15589
15590
15591
15592
15593
15594
15595
15596
15597
15598
15599
15600
15601
15602
15603
15604
15605
15606
15607
15608
15609
15610
15611
15612
15613
15614
15615
15616
15617
15618
15619
15620
15621
15622
15623
15624
15625
15626
15627
15628
15629
15630
15631
15632
15633
15634
15635
15636
15637
15638
15639
15640
15641
15642
15643
15644
15645
15646
15647
15648
15649
15650
15651
15652
15653
15654
15655
15656
15657
15658
15659
15660
15661
15662
15663
15664
15665
15666
15667
15668
15669
15670
15671
15672
15673
15674
15675
15676
15677
15678
15679
15680
15681
15682
15683
15684
15685
15686
15687
15688
15689
15690
15691
15692
15693
15694
15695
15696
15697
15698
15699
15700
15701
15702
15703
15704
15705
15706
15707
15708
15709
15710
15711
15712
15713
15714
15715
15716
15717
15718
15719
15720
15721
15722
15723
15724
15725
15726
15727
15728
15729
15730
15731
15732
15733
15734
15735
15736
15737
15738
15739
15740
15741
15742
15743
15744
15745
15746
15747
15748
15749
15750
15751
15752
15753
15754
15755
15756
15757
15758
15759
15760
15761
15762
15763
15764
15765
15766
15767
15768
15769
15770
15771
15772
15773
15774
15775
15776
15777
15778
15779
15780
15781
15782
15783
15784
15785
15786
15787
15788
15789
15790
15791
15792
15793
15794
15795
15796
15797
15798
15799
15800
15801
15802
15803
15804
15805
15806
15807
15808
15809
15810
15811
15812
15813
15814
15815
15816
15817
15818
15819
15820
15821
15822
15823
15824
15825
15826
15827
15828
15829
15830
15831
15832
15833
15834
15835
15836
15837
15838
15839
15840
15841
15842
15843
15844
15845
15846
15847
15848
15849
15850
15851
15852
15853
15854
15855
15856
15857
15858
15859
15860
15861
15862
15863
15864
15865
15866
15867
15868
15869
15870
15871
15872
15873
15874
15875
15876
15877
15878
15879
15880
15881
15882
15883
15884
15885
15886
15887
15888
15889
15890
15891
15892
15893
15894
15895
15896
15897
15898
15899
15900
15901
15902
15903
15904
15905
15906
15907
15908
15909
15910
15911
15912
15913
15914
15915
15916
15917
15918
15919
15920
15921
15922
15923
15924
15925
15926
15927
15928
15929
15930
15931
15932
15933
15934
15935
15936
15937
15938
15939
15940
15941
15942
15943
15944
15945
15946
15947
15948
15949
15950
15951
15952
15953
15954
15955
15956
15957
15958
15959
15960
15961
15962
15963
15964
15965
15966
15967
15968
15969
15970
15971
15972
15973
15974
15975
15976
15977
15978
15979
15980
15981
15982
15983
15984
15985
15986
15987
15988
15989
15990
15991
15992
15993
15994
15995
15996
15997
15998
15999
16000
16001
16002
16003
16004
16005
16006
16007
16008
16009
16010
16011
16012
16013
16014
16015
16016
16017
16018
16019
16020
16021
16022
16023
16024
16025
16026
16027
16028
16029
16030
16031
16032
16033
16034
16035
16036
16037
16038
16039
16040
16041
16042
16043
16044
16045
16046
16047
16048
16049
16050
16051
16052
16053
16054
16055
16056
16057
16058
16059
16060
16061
16062
16063
16064
16065
16066
16067
16068
16069
16070
16071
16072
16073
16074
16075
16076
16077
16078
16079
16080
16081
16082
16083
16084
16085
16086
16087
16088
16089
16090
16091
16092
16093
16094
16095
16096
16097
16098
16099
16100
16101
16102
16103
16104
16105
16106
16107
16108
16109
16110
16111
16112
16113
16114
16115
16116
16117
16118
16119
16120
16121
16122
16123
16124
16125
16126
16127
16128
16129
16130
16131
16132
16133
16134
16135
16136
16137
16138
16139
16140
16141
16142
16143
16144
16145
16146
16147
16148
16149
16150
16151
16152
16153
16154
16155
16156
16157
16158
16159
16160
16161
16162
16163
16164
16165
16166
16167
16168
16169
16170
16171
16172
16173
16174
16175
16176
16177
16178
16179
16180
16181
16182
16183
16184
16185
16186
16187
16188
16189
16190
16191
16192
16193
16194
16195
16196
16197
16198
16199
16200
16201
16202
16203
16204
16205
16206
16207
16208
16209
16210
16211
16212
16213
16214
16215
16216
16217
16218
16219
16220
16221
16222
16223
16224
16225
16226
16227
16228
16229
16230
16231
16232
16233
16234
16235
16236
16237
16238
16239
16240
16241
16242
16243
16244
16245
16246
16247
16248
16249
16250
16251
16252
16253
16254
16255
16256
16257
16258
16259
16260
16261
16262
16263
16264
16265
16266
16267
16268
16269
16270
16271
16272
16273
16274
16275
16276
16277
16278
16279
16280
16281
16282
16283
16284
16285
16286
16287
16288
16289
16290
16291
16292
16293
16294
16295
16296
16297
16298
16299
16300
16301
16302
16303
16304
16305
16306
16307
16308
16309
16310
16311
16312
16313
16314
16315
16316
16317
16318
16319
16320
16321
16322
16323
16324
16325
16326
16327
16328
16329
16330
16331
16332
16333
16334
16335
16336
16337
16338
16339
16340
16341
16342
16343
16344
16345
16346
16347
16348
16349
16350
16351
16352
16353
16354
16355
16356
16357
16358
16359
16360
16361
16362
16363
16364
16365
16366
16367
16368
16369
16370
16371
16372
16373
16374
16375
16376
16377
16378
16379
16380
16381
16382
16383
16384
16385
16386
16387
16388
16389
16390
16391
16392
16393
16394
16395
16396
16397
16398
16399
16400
16401
16402
16403
16404
16405
16406
16407
16408
16409
16410
16411
16412
16413
16414
16415
16416
16417
16418
16419
16420
16421
16422
16423
16424
16425
16426
16427
16428
16429
16430
16431
16432
16433
16434
16435
16436
16437
16438
16439
16440
16441
16442
16443
16444
16445
16446
16447
16448
16449
16450
16451
16452
16453
16454
16455
16456
16457
16458
16459
16460
16461
16462
16463
16464
16465
16466
16467
16468
16469
16470
16471
16472
16473
16474
16475
16476
16477
16478
16479
16480
16481
16482
16483
16484
16485
16486
16487
16488
16489
16490
16491
16492
16493
16494
16495
16496
16497
16498
16499
16500
16501
16502
16503
16504
16505
16506
16507
16508
16509
16510
16511
16512
16513
16514
16515
16516
16517
16518
16519
16520
16521
16522
16523
16524
16525
16526
16527
16528
16529
16530
16531
16532
16533
16534
16535
16536
16537
16538
16539
16540
16541
16542
16543
16544
16545
16546
16547
16548
16549
16550
16551
16552
16553
16554
16555
16556
16557
16558
16559
16560
16561
16562
16563
16564
16565
16566
16567
16568
16569
16570
16571
16572
16573
16574
16575
16576
16577
16578
16579
16580
16581
16582
16583
16584
16585
16586
16587
16588
16589
16590
16591
16592
16593
16594
16595
16596
16597
16598
16599
16600
16601
16602
16603
16604
16605
16606
16607
16608
16609
16610
16611
16612
16613
16614
16615
16616
16617
16618
16619
16620
16621
16622
16623
16624
16625
16626
16627
16628
16629
16630
16631
16632
16633
16634
16635
16636
16637
16638
16639
16640
16641
16642
16643
16644
16645
16646
16647
16648
16649
16650
16651
16652
16653
16654
16655
16656
16657
16658
16659
16660
16661
16662
16663
16664
16665
16666
16667
16668
16669
16670
16671
16672
16673
16674
16675
16676
16677
16678
16679
16680
16681
16682
16683
16684
16685
16686
16687
16688
16689
16690
16691
16692
16693
16694
16695
16696
16697
16698
16699
16700
16701
16702
16703
16704
16705
16706
16707
16708
16709
16710
16711
16712
16713
16714
16715
16716
16717
16718
16719
16720
16721
16722
16723
16724
16725
16726
16727
16728
16729
16730
16731
16732
16733
16734
16735
16736
16737
16738
16739
16740
16741
16742
16743
16744
16745
16746
16747
16748
16749
16750
16751
16752
16753
16754
16755
16756
16757
16758
16759
16760
16761
16762
16763
16764
16765
16766
16767
16768
16769
16770
16771
16772
16773
16774
16775
16776
16777
16778
16779
16780
16781
16782
16783
16784
16785
16786
16787
16788
16789
16790
16791
16792
16793
16794
16795
16796
16797
16798
16799
16800
16801
16802
16803
16804
16805
16806
16807
16808
16809
16810
16811
16812
16813
16814
16815
16816
16817
16818
16819
16820
16821
16822
16823
16824
16825
16826
16827
16828
16829
16830
16831
16832
16833
16834
16835
16836
16837
16838
16839
16840
16841
16842
16843
16844
16845
16846
16847
16848
16849
16850
16851
16852
16853
16854
16855
16856
16857
16858
16859
16860
16861
16862
16863
16864
16865
16866
16867
16868
16869
16870
16871
16872
16873
16874
16875
16876
16877
16878
16879
16880
16881
16882
16883
16884
16885
16886
16887
16888
16889
16890
16891
16892
16893
16894
16895
16896
16897
16898
16899
16900
16901
16902
16903
16904
16905
16906
16907
16908
16909
16910
16911
16912
16913
16914
16915
16916
16917
16918
16919
16920
16921
16922
16923
16924
16925
16926
16927
16928
16929
16930
16931
16932
16933
16934
16935
16936
16937
16938
16939
16940
16941
16942
16943
16944
16945
16946
16947
16948
16949
16950
16951
16952
16953
16954
16955
16956
16957
16958
16959
16960
16961
16962
16963
16964
16965
16966
16967
16968
16969
16970
16971
16972
16973
16974
16975
16976
16977
16978
16979
16980
16981
16982
16983
16984
16985
16986
16987
16988
16989
16990
16991
16992
16993
16994
16995
16996
16997
16998
16999
17000
17001
17002
17003
17004
17005
17006
17007
17008
17009
17010
17011
17012
17013
17014
17015
17016
17017
17018
17019
17020
17021
17022
17023
17024
17025
17026
17027
17028
17029
17030
17031
17032
17033
17034
17035
17036
17037
17038
17039
17040
17041
17042
17043
17044
17045
17046
17047
17048
17049
17050
17051
17052
17053
17054
17055
17056
17057
17058
17059
17060
17061
17062
17063
17064
17065
17066
17067
17068
17069
17070
17071
17072
17073
17074
17075
17076
17077
17078
17079
17080
17081
17082
17083
17084
17085
17086
17087
17088
17089
17090
17091
17092
17093
17094
17095
17096
17097
17098
17099
17100
17101
17102
17103
17104
17105
17106
17107
17108
17109
17110
17111
17112
17113
17114
17115
17116
17117
17118
17119
17120
17121
17122
17123
17124
17125
17126
17127
17128
17129
17130
17131
17132
17133
17134
17135
17136
17137
17138
17139
17140
17141
17142
17143
17144
17145
17146
17147
17148
17149
17150
17151
17152
17153
17154
17155
17156
17157
17158
17159
17160
17161
17162
17163
17164
17165
17166
17167
17168
17169
17170
17171
17172
17173
17174
17175
17176
17177
17178
17179
17180
17181
17182
17183
17184
17185
17186
17187
17188
17189
17190
17191
17192
17193
17194
17195
17196
17197
17198
17199
17200
17201
17202
17203
17204
17205
17206
17207
17208
17209
17210
17211
17212
17213
17214
17215
17216
17217
17218
17219
17220
17221
17222
17223
17224
17225
17226
17227
17228
17229
17230
17231
17232
17233
17234
17235
17236
17237
17238
17239
17240
17241
17242
17243
17244
17245
17246
17247
17248
17249
17250
17251
17252
17253
17254
17255
17256
17257
17258
17259
17260
17261
17262
17263
17264
17265
17266
17267
17268
17269
17270
17271
17272
17273
17274
17275
17276
17277
17278
17279
17280
17281
17282
17283
17284
17285
17286
17287
17288
17289
17290
17291
17292
17293
17294
17295
17296
17297
17298
17299
17300
17301
17302
17303
17304
17305
17306
17307
17308
17309
17310
17311
17312
17313
17314
17315
17316
17317
17318
17319
17320
17321
17322
17323
17324
17325
17326
17327
17328
17329
17330
17331
17332
17333
17334
17335
17336
17337
17338
17339
17340
17341
17342
17343
17344
17345
17346
17347
17348
17349
17350
17351
17352
17353
17354
17355
17356
17357
17358
17359
17360
17361
17362
17363
17364
17365
17366
17367
17368
17369
17370
17371
17372
17373
17374
17375
17376
17377
17378
17379
17380
17381
17382
17383
17384
17385
17386
17387
17388
17389
17390
17391
17392
17393
17394
17395
17396
17397
17398
17399
17400
17401
17402
17403
17404
17405
17406
17407
17408
17409
17410
17411
17412
17413
17414
17415
17416
17417
17418
17419
17420
17421
17422
17423
17424
17425
17426
17427
17428
17429
17430
17431
17432
17433
17434
17435
17436
17437
17438
17439
17440
17441
17442
17443
17444
17445
17446
17447
17448
17449
17450
17451
17452
17453
17454
17455
17456
17457
17458
17459
17460
17461
17462
17463
17464
17465
17466
17467
17468
17469
17470
17471
17472
17473
17474
17475
17476
17477
17478
17479
17480
17481
17482
17483
17484
17485
17486
17487
17488
17489
17490
17491
17492
17493
17494
17495
17496
17497
17498
17499
17500
17501
17502
17503
17504
17505
17506
17507
17508
17509
17510
17511
17512
17513
17514
17515
17516
17517
17518
17519
17520
17521
17522
17523
17524
17525
17526
17527
17528
17529
17530
17531
17532
17533
17534
17535
17536
17537
17538
17539
17540
17541
17542
17543
17544
17545
17546
17547
17548
17549
17550
17551
17552
17553
17554
17555
17556
17557
17558
17559
17560
17561
17562
17563
17564
17565
17566
17567
17568
17569
17570
17571
17572
17573
17574
17575
17576
17577
17578
17579
17580
17581
17582
17583
17584
17585
17586
17587
17588
17589
17590
17591
17592
17593
17594
17595
17596
17597
17598
17599
17600
17601
17602
17603
17604
17605
17606
17607
17608
17609
17610
17611
17612
17613
17614
17615
17616
17617
17618
17619
17620
17621
17622
17623
17624
17625
17626
17627
17628
17629
17630
17631
17632
17633
17634
17635
17636
17637
17638
17639
17640
17641
17642
17643
17644
17645
17646
17647
17648
17649
17650
17651
17652
17653
17654
17655
17656
17657
17658
17659
17660
17661
17662
17663
17664
17665
17666
17667
17668
17669
17670
17671
17672
17673
17674
17675
17676
17677
17678
17679
17680
17681
17682
17683
17684
17685
17686
17687
17688
17689
17690
17691
17692
17693
17694
17695
17696
17697
17698
17699
17700
17701
17702
17703
17704
17705
17706
17707
17708
17709
17710
17711
17712
17713
17714
17715
17716
17717
17718
17719
17720
17721
17722
17723
17724
17725
17726
17727
17728
17729
17730
17731
17732
17733
17734
17735
17736
17737
17738
17739
17740
17741
17742
17743
17744
17745
17746
17747
17748
17749
17750
17751
17752
17753
17754
17755
17756
17757
17758
17759
17760
17761
17762
17763
17764
17765
17766
17767
17768
17769
17770
17771
17772
17773
17774
17775
17776
17777
17778
17779
17780
17781
17782
17783
17784
17785
17786
17787
17788
17789
17790
17791
17792
17793
17794
17795
17796
17797
17798
17799
17800
17801
17802
17803
17804
17805
17806
17807
17808
17809
17810
17811
17812
17813
17814
17815
17816
17817
17818
17819
17820
17821
17822
17823
17824
17825
17826
17827
17828
17829
17830
17831
17832
17833
17834
17835
17836
17837
17838
17839
17840
17841
17842
17843
17844
17845
17846
17847
17848
17849
17850
17851
17852
17853
17854
17855
17856
17857
17858
17859
17860
17861
17862
17863
17864
17865
17866
17867
17868
17869
17870
17871
17872
17873
17874
17875
17876
17877
17878
17879
17880
17881
17882
17883
17884
17885
17886
17887
17888
17889
17890
17891
17892
17893
17894
17895
17896
17897
17898
17899
17900
17901
17902
17903
17904
17905
17906
17907
17908
17909
17910
17911
17912
17913
17914
17915
17916
17917
17918
17919
17920
17921
17922
17923
17924
17925
17926
17927
17928
17929
17930
17931
17932
17933
17934
17935
17936
17937
17938
17939
17940
17941
17942
17943
17944
17945
17946
17947
17948
17949
17950
17951
17952
17953
17954
17955
17956
17957
17958
17959
17960
17961
17962
17963
17964
17965
17966
17967
17968
17969
17970
17971
17972
17973
17974
17975
17976
17977
17978
17979
17980
17981
17982
17983
17984
17985
17986
17987
17988
17989
17990
17991
17992
17993
17994
17995
17996
17997
17998
17999
18000
18001
18002
18003
18004
18005
18006
18007
18008
18009
18010
18011
18012
18013
18014
18015
18016
18017
18018
18019
18020
18021
18022
18023
18024
18025
18026
18027
18028
18029
18030
18031
18032
18033
18034
18035
18036
18037
18038
18039
18040
18041
18042
18043
18044
18045
18046
18047
18048
18049
18050
18051
18052
18053
18054
18055
18056
18057
18058
18059
18060
18061
18062
18063
18064
18065
18066
18067
18068
18069
18070
18071
18072
18073
18074
18075
18076
18077
18078
18079
18080
18081
18082
18083
18084
18085
18086
18087
18088
18089
18090
18091
18092
18093
18094
18095
18096
18097
18098
18099
18100
18101
18102
18103
18104
18105
18106
18107
18108
18109
18110
18111
18112
18113
18114
18115
18116
18117
18118
18119
18120
18121
18122
18123
18124
18125
18126
18127
18128
18129
18130
18131
18132
18133
18134
18135
18136
18137
18138
18139
18140
18141
18142
18143
18144
18145
18146
18147
18148
18149
18150
18151
18152
18153
18154
18155
18156
18157
18158
18159
18160
18161
18162
18163
18164
18165
18166
18167
18168
18169
18170
18171
18172
18173
18174
18175
18176
18177
18178
18179
18180
18181
18182
18183
18184
18185
18186
18187
18188
18189
18190
18191
18192
18193
18194
18195
18196
18197
18198
18199
18200
18201
18202
18203
18204
18205
18206
18207
18208
18209
18210
18211
18212
18213
18214
18215
18216
18217
18218
18219
18220
18221
18222
18223
18224
18225
18226
18227
18228
18229
18230
18231
18232
18233
18234
18235
18236
18237
18238
18239
18240
18241
18242
18243
18244
18245
18246
18247
18248
18249
18250
18251
18252
18253
18254
18255
18256
18257
18258
18259
18260
18261
18262
18263
18264
18265
18266
18267
18268
18269
18270
18271
18272
18273
18274
18275
18276
18277
18278
18279
18280
18281
18282
18283
18284
18285
18286
18287
18288
18289
18290
18291
18292
18293
18294
18295
18296
18297
18298
18299
18300
18301
18302
18303
18304
18305
18306
18307
18308
18309
18310
18311
18312
18313
18314
18315
18316
18317
18318
18319
18320
18321
18322
18323
18324
18325
18326
18327
18328
18329
18330
18331
18332
18333
18334
18335
18336
18337
18338
18339
18340
18341
18342
18343
18344
18345
18346
18347
18348
18349
18350
18351
18352
18353
18354
18355
18356
18357
18358
18359
18360
18361
18362
18363
18364
18365
18366
18367
18368
18369
18370
18371
18372
18373
18374
18375
18376
18377
18378
18379
18380
18381
18382
18383
18384
18385
18386
18387
18388
18389
18390
18391
18392
18393
18394
18395
18396
18397
18398
18399
18400
18401
18402
18403
18404
18405
18406
18407
18408
18409
18410
18411
18412
18413
18414
18415
18416
18417
18418
18419
18420
18421
18422
18423
18424
18425
18426
18427
18428
18429
18430
18431
18432
18433
18434
18435
18436
18437
18438
18439
18440
18441
18442
18443
18444
18445
18446
18447
18448
18449
18450
18451
18452
18453
18454
18455
18456
18457
18458
18459
18460
18461
18462
18463
18464
18465
18466
18467
18468
18469
18470
18471
18472
18473
18474
18475
18476
18477
18478
18479
18480
18481
18482
18483
18484
18485
18486
18487
18488
18489
18490
18491
18492
18493
18494
18495
18496
18497
18498
18499
18500
18501
18502
18503
18504
18505
18506
18507
18508
18509
18510
18511
18512
18513
18514
18515
18516
18517
18518
18519
18520
18521
18522
18523
18524
18525
18526
18527
18528
18529
18530
18531
18532
18533
18534
18535
18536
18537
18538
18539
18540
18541
18542
18543
18544
18545
18546
18547
18548
18549
18550
18551
18552
18553
18554
18555
18556
18557
18558
18559
18560
18561
18562
18563
18564
18565
18566
18567
18568
18569
18570
18571
18572
18573
18574
18575
18576
18577
18578
18579
18580
18581
18582
18583
18584
18585
18586
18587
18588
18589
18590
18591
18592
18593
18594
18595
18596
18597
18598
18599
18600
18601
18602
18603
18604
18605
18606
18607
18608
18609
18610
18611
18612
18613
18614
18615
18616
18617
18618
18619
18620
18621
18622
18623
18624
18625
18626
18627
18628
18629
18630
18631
18632
18633
18634
18635
18636
18637
18638
18639
18640
18641
18642
18643
18644
18645
18646
18647
18648
18649
18650
18651
18652
18653
18654
18655
18656
18657
18658
18659
18660
18661
18662
18663
18664
18665
18666
18667
18668
18669
18670
18671
18672
18673
18674
18675
18676
18677
18678
18679
18680
18681
18682
18683
18684
18685
18686
18687
18688
18689
18690
18691
18692
18693
18694
18695
18696
18697
18698
18699
18700
18701
18702
18703
18704
18705
18706
18707
18708
18709
18710
18711
18712
18713
18714
18715
18716
18717
18718
18719
18720
18721
18722
18723
18724
18725
18726
18727
18728
18729
18730
18731
18732
18733
18734
18735
18736
18737
18738
18739
18740
18741
18742
18743
18744
18745
18746
18747
18748
18749
18750
18751
18752
18753
18754
18755
18756
18757
18758
18759
18760
18761
18762
18763
18764
18765
18766
18767
18768
18769
18770
18771
18772
18773
18774
18775
18776
18777
18778
18779
18780
18781
18782
18783
18784
18785
18786
18787
18788
18789
18790
18791
18792
18793
18794
18795
18796
18797
18798
18799
18800
18801
18802
18803
18804
18805
18806
18807
18808
18809
18810
18811
18812
18813
18814
18815
18816
18817
18818
18819
18820
18821
18822
18823
18824
18825
18826
18827
18828
18829
18830
18831
18832
18833
18834
18835
18836
18837
18838
18839
18840
18841
18842
18843
18844
18845
18846
18847
18848
18849
18850
18851
18852
18853
18854
18855
18856
18857
18858
18859
18860
18861
18862
18863
18864
18865
18866
18867
18868
18869
18870
18871
18872
18873
18874
18875
18876
18877
18878
18879
18880
18881
18882
18883
18884
18885
18886
18887
18888
18889
18890
18891
18892
18893
18894
18895
18896
18897
18898
18899
18900
18901
18902
18903
18904
18905
18906
18907
18908
18909
18910
18911
18912
18913
18914
18915
18916
18917
18918
18919
18920
18921
18922
18923
18924
18925
18926
18927
18928
18929
18930
18931
18932
18933
18934
18935
18936
18937
18938
18939
18940
18941
18942
18943
18944
18945
18946
18947
18948
18949
18950
18951
18952
18953
18954
18955
18956
18957
18958
18959
18960
18961
18962
18963
18964
18965
18966
18967
18968
18969
18970
18971
18972
18973
18974
18975
18976
18977
18978
18979
18980
18981
18982
18983
18984
18985
18986
18987
18988
18989
18990
18991
18992
18993
18994
18995
18996
18997
18998
18999
19000
19001
19002
19003
19004
19005
19006
19007
19008
19009
19010
19011
19012
19013
19014
19015
19016
19017
19018
19019
19020
19021
19022
19023
19024
19025
19026
19027
19028
19029
19030
19031
19032
19033
19034
19035
19036
19037
19038
19039
19040
19041
19042
19043
19044
19045
19046
19047
19048
19049
19050
19051
19052
19053
19054
19055
19056
19057
19058
19059
19060
19061
19062
19063
19064
19065
19066
19067
19068
19069
19070
19071
19072
19073
19074
19075
19076
19077
19078
19079
19080
19081
19082
19083
19084
19085
19086
19087
19088
19089
19090
19091
19092
19093
19094
19095
19096
19097
19098
19099
19100
19101
19102
19103
19104
19105
19106
19107
19108
19109
19110
19111
19112
19113
19114
19115
19116
19117
19118
19119
19120
19121
19122
19123
19124
19125
19126
19127
19128
19129
19130
19131
19132
19133
19134
19135
19136
19137
19138
19139
19140
19141
19142
19143
19144
19145
19146
19147
19148
19149
19150
19151
19152
19153
19154
19155
19156
19157
19158
19159
19160
19161
19162
19163
19164
19165
19166
19167
19168
19169
19170
19171
19172
19173
19174
19175
19176
19177
19178
19179
19180
19181
19182
19183
19184
19185
19186
19187
19188
19189
19190
19191
19192
19193
19194
19195
19196
19197
19198
19199
19200
19201
19202
19203
19204
19205
19206
19207
19208
19209
19210
19211
19212
19213
19214
19215
19216
19217
19218
19219
19220
19221
19222
19223
19224
19225
19226
19227
19228
19229
19230
19231
19232
19233
19234
19235
19236
19237
19238
19239
19240
19241
19242
19243
19244
19245
19246
19247
19248
19249
19250
19251
19252
19253
19254
19255
19256
19257
19258
19259
19260
19261
19262
19263
19264
19265
19266
19267
19268
19269
19270
19271
19272
19273
19274
19275
19276
19277
19278
19279
19280
19281
19282
19283
19284
19285
19286
19287
19288
19289
19290
19291
19292
19293
19294
19295
19296
19297
19298
19299
19300
19301
19302
19303
19304
19305
19306
19307
19308
19309
19310
19311
19312
19313
19314
19315
19316
19317
19318
19319
19320
19321
19322
19323
19324
19325
19326
19327
19328
19329
19330
19331
19332
19333
19334
19335
19336
19337
19338
19339
19340
19341
19342
19343
19344
19345
19346
19347
19348
19349
19350
19351
19352
19353
19354
19355
19356
19357
19358
19359
19360
19361
19362
19363
19364
19365
19366
19367
19368
19369
19370
19371
19372
19373
19374
19375
19376
19377
19378
19379
19380
19381
19382
19383
19384
19385
19386
19387
19388
19389
19390
19391
19392
19393
19394
19395
19396
19397
19398
19399
19400
19401
19402
19403
19404
19405
19406
19407
19408
19409
19410
19411
19412
19413
19414
19415
19416
19417
19418
19419
19420
19421
19422
19423
19424
19425
19426
19427
19428
19429
19430
19431
19432
19433
19434
19435
19436
19437
19438
19439
19440
19441
19442
19443
19444
19445
19446
19447
19448
19449
19450
19451
19452
19453
19454
19455
19456
19457
19458
19459
19460
19461
19462
19463
19464
19465
19466
19467
19468
19469
19470
19471
19472
19473
19474
19475
19476
19477
19478
19479
19480
19481
19482
19483
19484
19485
19486
19487
19488
19489
19490
19491
19492
19493
19494
19495
19496
19497
19498
19499
19500
19501
19502
19503
19504
19505
19506
19507
19508
19509
19510
19511
19512
19513
19514
19515
19516
19517
19518
19519
19520
19521
19522
19523
19524
19525
19526
19527
19528
19529
19530
19531
19532
19533
19534
19535
19536
19537
19538
19539
19540
19541
19542
19543
19544
19545
19546
19547
19548
19549
19550
19551
19552
19553
19554
19555
19556
19557
19558
19559
19560
19561
19562
19563
19564
19565
19566
19567
19568
19569
19570
19571
19572
19573
19574
19575
19576
19577
19578
19579
19580
19581
19582
19583
19584
19585
19586
19587
19588
19589
19590
19591
19592
19593
19594
19595
19596
19597
19598
19599
19600
19601
19602
19603
19604
19605
19606
19607
19608
19609
19610
19611
19612
19613
19614
19615
19616
19617
19618
19619
19620
19621
19622
19623
19624
19625
19626
19627
19628
19629
19630
19631
19632
19633
19634
19635
19636
19637
19638
19639
19640
19641
19642
19643
19644
19645
19646
19647
19648
19649
19650
19651
19652
19653
19654
19655
19656
19657
19658
19659
19660
19661
19662
19663
19664
19665
19666
19667
19668
19669
19670
19671
19672
19673
19674
19675
19676
19677
19678
19679
19680
19681
19682
19683
19684
19685
19686
19687
19688
19689
19690
19691
19692
19693
19694
19695
19696
19697
19698
19699
19700
19701
19702
19703
19704
19705
19706
19707
19708
19709
19710
19711
19712
19713
19714
19715
19716
19717
19718
19719
19720
19721
19722
19723
19724
19725
19726
19727
19728
19729
19730
19731
19732
19733
19734
19735
19736
19737
19738
19739
19740
19741
19742
19743
19744
19745
19746
19747
19748
19749
19750
19751
19752
19753
19754
19755
19756
19757
19758
19759
19760
19761
19762
19763
19764
19765
19766
19767
19768
19769
19770
19771
19772
19773
19774
19775
19776
19777
19778
19779
19780
19781
19782
19783
19784
19785
19786
19787
19788
19789
19790
19791
19792
19793
19794
19795
19796
19797
19798
19799
19800
19801
19802
19803
19804
19805
19806
19807
19808
19809
19810
19811
19812
19813
19814
19815
19816
19817
19818
19819
19820
19821
19822
19823
19824
19825
19826
19827
19828
19829
19830
19831
19832
19833
19834
19835
19836
19837
19838
19839
19840
19841
19842
19843
19844
19845
19846
19847
19848
19849
19850
19851
19852
19853
19854
19855
19856
19857
19858
19859
19860
19861
19862
19863
19864
19865
19866
19867
19868
19869
19870
19871
19872
19873
19874
19875
19876
19877
19878
19879
19880
19881
19882
19883
19884
19885
19886
19887
19888
19889
19890
19891
19892
19893
19894
19895
19896
19897
19898
19899
19900
19901
19902
19903
19904
19905
19906
19907
19908
19909
19910
19911
19912
19913
19914
19915
19916
19917
19918
19919
19920
19921
19922
19923
19924
19925
19926
19927
19928
19929
19930
19931
19932
19933
19934
19935
19936
19937
19938
19939
19940
19941
19942
19943
19944
19945
19946
19947
19948
19949
19950
19951
19952
19953
19954
19955
19956
19957
19958
19959
19960
19961
19962
19963
19964
19965
19966
19967
19968
19969
19970
19971
19972
19973
19974
19975
19976
19977
19978
19979
19980
19981
19982
19983
19984
19985
19986
19987
19988
19989
19990
19991
19992
19993
19994
19995
19996
19997
19998
19999
20000
20001
20002
20003
20004
20005
20006
20007
20008
20009
20010
20011
20012
20013
20014
20015
20016
20017
20018
20019
20020
20021
20022
20023
20024
20025
20026
20027
20028
20029
20030
20031
20032
20033
20034
20035
20036
20037
20038
20039
20040
20041
20042
20043
20044
20045
20046
20047
20048
20049
20050
20051
20052
20053
20054
20055
20056
20057
20058
20059
20060
20061
20062
20063
20064
20065
20066
20067
20068
20069
20070
20071
20072
20073
20074
20075
20076
20077
20078
20079
20080
20081
20082
20083
20084
20085
20086
20087
20088
20089
20090
20091
20092
20093
20094
20095
20096
20097
20098
20099
20100
20101
20102
20103
20104
20105
20106
20107
20108
20109
20110
20111
20112
20113
20114
20115
20116
20117
20118
20119
20120
20121
20122
20123
20124
20125
20126
20127
20128
20129
20130
20131
20132
20133
20134
20135
20136
20137
20138
20139
20140
20141
20142
20143
20144
20145
20146
20147
20148
20149
20150
20151
20152
20153
20154
20155
20156
20157
20158
20159
20160
20161
20162
20163
20164
20165
20166
20167
20168
20169
20170
20171
20172
20173
20174
20175
20176
20177
20178
20179
20180
20181
20182
20183
20184
20185
20186
20187
20188
20189
20190
20191
20192
20193
20194
20195
20196
20197
20198
20199
20200
20201
20202
20203
20204
20205
20206
20207
20208
20209
20210
20211
20212
20213
20214
20215
20216
20217
20218
20219
20220
20221
20222
20223
20224
20225
20226
20227
20228
20229
20230
20231
20232
20233
20234
20235
20236
20237
20238
20239
20240
20241
20242
20243
20244
20245
20246
20247
20248
20249
20250
20251
20252
20253
20254
20255
20256
20257
20258
20259
20260
20261
20262
20263
20264
20265
20266
20267
20268
20269
20270
20271
20272
20273
20274
20275
20276
20277
20278
20279
20280
20281
20282
20283
20284
20285
20286
20287
20288
20289
20290
20291
20292
20293
20294
20295
20296
20297
20298
20299
20300
20301
20302
20303
20304
20305
20306
20307
20308
20309
20310
20311
20312
20313
20314
20315
20316
20317
20318
20319
20320
20321
20322
20323
20324
20325
20326
20327
20328
20329
20330
20331
20332
20333
20334
20335
20336
20337
20338
20339
20340
20341
20342
20343
20344
20345
20346
20347
20348
20349
20350
20351
20352
20353
20354
20355
20356
20357
20358
20359
20360
20361
20362
20363
20364
20365
20366
20367
20368
20369
20370
20371
20372
20373
20374
20375
20376
20377
20378
20379
20380
20381
20382
20383
20384
20385
20386
20387
20388
20389
20390
20391
20392
20393
20394
20395
20396
20397
20398
20399
20400
20401
20402
20403
20404
20405
20406
20407
20408
20409
20410
20411
20412
20413
20414
20415
20416
20417
20418
20419
20420
20421
20422
20423
20424
20425
20426
20427
20428
20429
20430
20431
20432
20433
20434
20435
20436
20437
20438
20439
20440
20441
20442
20443
20444
20445
20446
20447
20448
20449
20450
20451
20452
20453
20454
20455
20456
20457
20458
20459
20460
20461
20462
20463
20464
20465
20466
20467
20468
20469
20470
20471
20472
20473
20474
20475
20476
20477
20478
20479
20480
20481
20482
20483
20484
20485
20486
20487
20488
20489
20490
20491
20492
20493
20494
20495
20496
20497
20498
20499
20500
20501
20502
20503
20504
20505
20506
20507
20508
20509
20510
20511
20512
20513
20514
20515
20516
20517
20518
20519
20520
20521
20522
20523
20524
20525
20526
20527
20528
20529
20530
20531
20532
20533
20534
20535
20536
20537
20538
20539
20540
20541
20542
20543
20544
20545
20546
20547
20548
20549
20550
20551
20552
20553
20554
20555
20556
20557
20558
20559
20560
20561
20562
20563
20564
20565
20566
20567
20568
20569
20570
20571
20572
20573
20574
20575
20576
20577
20578
20579
20580
20581
20582
20583
20584
20585
20586
20587
20588
20589
20590
20591
20592
20593
20594
20595
20596
20597
20598
20599
20600
20601
20602
20603
20604
20605
20606
20607
20608
20609
20610
20611
20612
20613
20614
20615
20616
20617
20618
20619
20620
20621
20622
20623
20624
20625
20626
20627
20628
20629
20630
20631
20632
20633
20634
20635
20636
20637
20638
20639
20640
20641
20642
20643
20644
20645
20646
20647
20648
20649
20650
20651
20652
20653
20654
20655
20656
20657
20658
20659
20660
20661
20662
20663
20664
20665
20666
20667
20668
20669
20670
20671
20672
20673
20674
20675
20676
20677
20678
20679
20680
20681
20682
20683
20684
20685
20686
20687
20688
20689
20690
20691
20692
20693
20694
20695
20696
20697
20698
20699
20700
20701
20702
20703
20704
20705
20706
20707
20708
20709
20710
20711
20712
20713
20714
20715
20716
20717
20718
20719
20720
20721
20722
20723
20724
20725
20726
20727
20728
20729
20730
20731
20732
20733
20734
20735
20736
20737
20738
20739
20740
20741
20742
20743
20744
20745
20746
20747
20748
20749
20750
20751
20752
20753
20754
20755
20756
20757
20758
20759
20760
20761
20762
20763
20764
20765
20766
20767
20768
20769
20770
20771
20772
20773
20774
20775
20776
20777
20778
20779
20780
20781
20782
20783
20784
20785
20786
20787
20788
20789
20790
20791
20792
20793
20794
20795
20796
20797
20798
20799
20800
20801
20802
20803
20804
20805
20806
20807
20808
20809
20810
20811
20812
20813
20814
20815
20816
20817
20818
20819
20820
20821
20822
20823
20824
20825
20826
20827
20828
20829
20830
20831
20832
20833
20834
20835
20836
20837
20838
20839
20840
20841
20842
20843
20844
20845
20846
20847
20848
20849
20850
20851
20852
20853
20854
20855
20856
20857
20858
20859
20860
20861
20862
20863
20864
20865
20866
20867
20868
20869
20870
20871
20872
20873
20874
20875
20876
20877
20878
20879
20880
20881
20882
20883
20884
20885
20886
20887
20888
20889
20890
20891
20892
20893
20894
20895
20896
20897
20898
20899
20900
20901
20902
20903
20904
20905
20906
20907
20908
20909
20910
20911
20912
20913
20914
20915
20916
20917
20918
20919
20920
20921
20922
20923
20924
20925
20926
20927
20928
20929
20930
20931
20932
20933
20934
20935
20936
20937
20938
20939
20940
20941
20942
20943
20944
20945
20946
20947
20948
20949
20950
20951
20952
20953
20954
20955
20956
20957
20958
20959
20960
20961
20962
20963
20964
20965
20966
20967
20968
20969
20970
20971
20972
20973
20974
20975
20976
20977
20978
20979
20980
20981
20982
20983
20984
20985
20986
20987
20988
20989
20990
20991
20992
20993
20994
20995
20996
20997
20998
20999
21000
21001
21002
21003
21004
21005
21006
21007
21008
21009
21010
21011
21012
21013
21014
21015
21016
21017
21018
21019
21020
21021
21022
21023
21024
21025
21026
21027
21028
21029
21030
21031
21032
21033
21034
21035
21036
21037
21038
21039
21040
21041
21042
21043
21044
21045
21046
21047
21048
21049
21050
21051
21052
21053
21054
21055
21056
21057
21058
21059
21060
21061
21062
21063
21064
21065
21066
21067
21068
21069
21070
21071
21072
21073
21074
21075
21076
21077
21078
21079
21080
21081
21082
21083
21084
21085
21086
21087
21088
21089
21090
21091
21092
21093
21094
21095
21096
21097
21098
21099
21100
21101
21102
21103
21104
21105
21106
21107
21108
21109
21110
21111
21112
21113
21114
21115
21116
21117
21118
21119
21120
21121
21122
21123
21124
21125
21126
21127
21128
21129
21130
21131
21132
21133
21134
21135
21136
21137
21138
21139
21140
21141
21142
21143
21144
21145
21146
21147
21148
21149
21150
21151
21152
21153
21154
21155
21156
21157
21158
21159
21160
21161
21162
21163
21164
21165
21166
21167
21168
21169
21170
21171
21172
21173
21174
21175
21176
21177
21178
21179
21180
21181
21182
21183
21184
21185
21186
21187
21188
21189
21190
21191
21192
21193
21194
21195
21196
21197
21198
21199
21200
21201
21202
21203
21204
21205
21206
21207
21208
21209
21210
21211
21212
21213
21214
21215
21216
21217
21218
21219
21220
21221
21222
21223
21224
21225
21226
21227
21228
21229
21230
21231
21232
21233
21234
21235
21236
21237
21238
21239
21240
21241
21242
21243
21244
21245
21246
21247
21248
21249
21250
21251
21252
21253
21254
21255
21256
21257
21258
21259
21260
21261
21262
21263
21264
21265
21266
21267
21268
21269
21270
21271
21272
21273
21274
21275
21276
21277
21278
21279
21280
21281
21282
21283
21284
21285
21286
21287
21288
21289
21290
21291
21292
21293
21294
21295
21296
21297
21298
21299
21300
21301
21302
21303
21304
21305
21306
21307
21308
21309
21310
21311
21312
21313
21314
21315
21316
21317
21318
21319
21320
21321
21322
21323
21324
21325
21326
21327
21328
21329
21330
21331
21332
21333
21334
21335
21336
21337
21338
21339
21340
21341
21342
21343
21344
21345
21346
21347
21348
21349
21350
21351
21352
21353
21354
21355
21356
21357
21358
21359
21360
21361
21362
21363
21364
21365
21366
21367
21368
21369
21370
21371
21372
21373
21374
21375
21376
21377
21378
21379
21380
21381
21382
21383
21384
21385
21386
21387
21388
21389
21390
21391
21392
21393
21394
21395
21396
21397
21398
21399
21400
21401
21402
21403
21404
21405
21406
21407
21408
21409
21410
21411
21412
21413
21414
21415
21416
21417
21418
21419
21420
21421
21422
21423
21424
21425
21426
21427
21428
21429
21430
21431
21432
21433
21434
21435
21436
21437
21438
21439
21440
21441
21442
21443
21444
21445
21446
21447
21448
21449
21450
21451
21452
21453
21454
21455
21456
21457
21458
21459
21460
21461
21462
21463
21464
21465
21466
21467
21468
21469
21470
21471
21472
21473
21474
21475
21476
21477
21478
21479
21480
21481
21482
21483
21484
21485
21486
21487
21488
21489
21490
21491
21492
21493
21494
21495
21496
21497
21498
21499
21500
21501
21502
21503
21504
21505
21506
21507
21508
21509
21510
21511
21512
21513
21514
21515
21516
21517
21518
21519
21520
21521
21522
21523
21524
21525
21526
21527
21528
21529
21530
21531
21532
21533
21534
21535
21536
21537
21538
21539
21540
21541
21542
21543
21544
21545
21546
21547
21548
21549
21550
21551
21552
21553
21554
21555
21556
21557
21558
21559
21560
21561
21562
21563
21564
21565
21566
21567
21568
21569
21570
21571
21572
21573
21574
21575
21576
21577
21578
21579
21580
21581
21582
21583
21584
21585
21586
21587
21588
21589
21590
21591
21592
21593
21594
21595
21596
21597
21598
21599
21600
21601
21602
21603
21604
21605
21606
21607
21608
21609
21610
21611
21612
21613
21614
21615
21616
21617
21618
21619
21620
21621
21622
21623
21624
21625
21626
21627
21628
21629
21630
21631
21632
21633
21634
21635
21636
21637
21638
21639
21640
21641
21642
21643
21644
21645
21646
21647
21648
21649
21650
21651
21652
21653
21654
21655
21656
21657
21658
21659
21660
21661
21662
21663
21664
21665
21666
21667
21668
21669
21670
21671
21672
21673
21674
21675
21676
21677
21678
21679
21680
21681
21682
21683
21684
21685
21686
21687
21688
21689
21690
21691
21692
21693
21694
21695
21696
21697
21698
21699
21700
21701
21702
21703
21704
21705
21706
21707
21708
21709
21710
21711
21712
21713
21714
21715
21716
21717
21718
21719
21720
21721
21722
21723
21724
21725
21726
21727
21728
21729
21730
21731
21732
21733
21734
21735
21736
21737
21738
21739
21740
21741
21742
21743
21744
21745
21746
21747
21748
21749
21750
21751
21752
21753
21754
21755
21756
21757
21758
21759
21760
21761
21762
21763
21764
21765
21766
21767
21768
21769
21770
21771
21772
21773
21774
21775
21776
21777
21778
21779
21780
21781
21782
21783
21784
21785
21786
21787
21788
21789
21790
21791
21792
21793
21794
21795
21796
21797
21798
21799
21800
21801
21802
21803
21804
21805
21806
21807
21808
21809
21810
21811
21812
21813
21814
21815
21816
21817
21818
21819
21820
21821
21822
21823
21824
21825
21826
21827
21828
21829
21830
21831
21832
21833
21834
21835
21836
21837
21838
21839
21840
21841
21842
21843
21844
21845
21846
21847
21848
21849
21850
21851
21852
21853
21854
21855
21856
21857
21858
21859
21860
21861
21862
21863
21864
21865
21866
21867
21868
21869
21870
21871
21872
21873
21874
21875
21876
21877
21878
21879
21880
21881
21882
21883
21884
21885
21886
21887
21888
21889
21890
21891
21892
21893
21894
21895
21896
21897
21898
21899
21900
21901
21902
21903
21904
21905
21906
21907
21908
21909
21910
21911
21912
21913
21914
21915
21916
21917
21918
21919
21920
21921
21922
21923
21924
21925
21926
21927
21928
21929
21930
21931
21932
21933
21934
21935
21936
21937
21938
21939
21940
21941
21942
21943
21944
21945
21946
21947
21948
21949
21950
21951
21952
21953
21954
21955
21956
21957
21958
21959
21960
21961
21962
21963
21964
21965
21966
21967
21968
21969
21970
21971
21972
21973
21974
21975
21976
21977
21978
21979
21980
21981
21982
21983
21984
21985
21986
21987
21988
21989
21990
21991
21992
21993
21994
21995
21996
21997
21998
21999
22000
22001
22002
22003
22004
22005
22006
22007
22008
22009
22010
22011
22012
22013
22014
22015
22016
22017
22018
22019
22020
22021
22022
22023
22024
22025
22026
22027
22028
22029
22030
22031
22032
22033
22034
22035
22036
22037
22038
22039
22040
22041
22042
22043
22044
22045
22046
22047
22048
22049
22050
22051
22052
22053
22054
22055
22056
22057
22058
22059
22060
22061
22062
22063
22064
22065
22066
22067
22068
22069
22070
22071
22072
22073
22074
22075
22076
22077
22078
22079
22080
22081
22082
22083
22084
22085
22086
22087
22088
22089
22090
22091
22092
22093
22094
22095
22096
22097
22098
22099
22100
22101
22102
22103
22104
22105
22106
22107
22108
22109
22110
22111
22112
22113
22114
22115
22116
22117
22118
22119
22120
22121
22122
22123
22124
22125
22126
22127
22128
22129
22130
22131
22132
22133
22134
22135
22136
22137
22138
22139
22140
22141
22142
22143
22144
22145
22146
22147
22148
22149
22150
22151
22152
22153
22154
22155
22156
22157
22158
22159
22160
22161
22162
22163
22164
22165
22166
22167
22168
22169
22170
22171
22172
22173
22174
22175
22176
22177
22178
22179
22180
22181
22182
22183
22184
22185
22186
22187
22188
22189
22190
22191
22192
22193
22194
22195
22196
22197
22198
22199
22200
22201
22202
22203
22204
22205
22206
22207
22208
22209
22210
22211
22212
22213
22214
22215
22216
22217
22218
22219
22220
22221
22222
22223
22224
22225
22226
22227
22228
22229
22230
22231
22232
22233
22234
22235
22236
22237
22238
22239
22240
22241
22242
22243
22244
22245
22246
22247
22248
22249
22250
22251
22252
22253
22254
22255
22256
22257
22258
22259
22260
22261
22262
22263
22264
22265
22266
22267
22268
22269
22270
22271
22272
22273
22274
22275
22276
22277
22278
22279
22280
22281
22282
22283
22284
22285
22286
22287
22288
22289
22290
22291
22292
22293
22294
22295
22296
22297
22298
22299
22300
22301
22302
22303
22304
22305
22306
22307
22308
22309
22310
22311
22312
22313
22314
22315
22316
22317
22318
22319
22320
22321
22322
22323
22324
22325
22326
22327
22328
22329
22330
22331
22332
22333
22334
22335
22336
22337
22338
22339
22340
22341
22342
22343
22344
22345
22346
22347
22348
22349
22350
22351
22352
22353
22354
22355
22356
22357
22358
22359
22360
22361
22362
22363
22364
22365
22366
22367
22368
22369
22370
22371
22372
22373
22374
22375
22376
22377
22378
22379
22380
22381
22382
22383
22384
22385
22386
22387
22388
22389
22390
22391
22392
22393
22394
22395
22396
22397
22398
22399
22400
22401
22402
22403
22404
22405
22406
22407
22408
22409
22410
22411
22412
22413
22414
22415
22416
22417
22418
22419
22420
22421
22422
22423
22424
22425
22426
22427
22428
22429
22430
22431
22432
22433
22434
22435
22436
22437
22438
22439
22440
22441
22442
22443
22444
22445
22446
22447
22448
22449
22450
22451
22452
22453
22454
22455
22456
22457
22458
22459
22460
22461
22462
22463
22464
22465
22466
22467
22468
22469
22470
22471
22472
22473
22474
22475
22476
22477
22478
22479
22480
22481
22482
22483
22484
22485
22486
22487
22488
22489
22490
22491
22492
22493
22494
22495
22496
22497
22498
22499
22500
22501
22502
22503
22504
22505
22506
22507
22508
22509
22510
22511
22512
22513
22514
22515
22516
22517
22518
22519
22520
22521
22522
22523
22524
22525
22526
22527
22528
22529
22530
22531
22532
22533
22534
22535
22536
22537
22538
22539
22540
22541
22542
22543
22544
22545
22546
22547
22548
22549
22550
22551
22552
22553
22554
22555
22556
22557
22558
22559
22560
22561
22562
22563
22564
22565
22566
22567
22568
22569
22570
22571
22572
22573
22574
22575
22576
22577
22578
22579
22580
22581
22582
22583
22584
22585
22586
22587
22588
22589
22590
22591
22592
22593
22594
22595
22596
22597
22598
22599
22600
22601
22602
22603
22604
22605
22606
22607
22608
22609
22610
22611
22612
22613
22614
22615
22616
22617
22618
22619
22620
22621
22622
22623
22624
22625
22626
22627
22628
22629
22630
22631
22632
22633
22634
22635
22636
22637
22638
22639
22640
22641
22642
22643
22644
22645
22646
22647
22648
22649
22650
22651
22652
22653
22654
22655
22656
22657
22658
22659
22660
22661
22662
22663
22664
22665
22666
22667
22668
22669
22670
22671
22672
22673
22674
22675
22676
22677
22678
22679
22680
22681
22682
22683
22684
22685
22686
22687
22688
22689
22690
22691
22692
22693
22694
22695
22696
22697
22698
22699
22700
22701
22702
22703
22704
22705
22706
22707
22708
22709
22710
22711
22712
22713
22714
22715
22716
22717
22718
22719
22720
22721
22722
22723
22724
22725
22726
22727
22728
22729
22730
22731
22732
22733
22734
22735
22736
22737
22738
22739
22740
22741
22742
22743
22744
22745
22746
22747
22748
22749
22750
22751
22752
22753
22754
22755
22756
22757
22758
22759
22760
22761
22762
22763
22764
22765
22766
22767
22768
22769
22770
22771
22772
22773
22774
22775
22776
22777
22778
22779
22780
22781
22782
22783
22784
22785
22786
22787
22788
22789
22790
22791
22792
22793
22794
22795
22796
22797
22798
22799
22800
22801
22802
22803
22804
22805
22806
22807
22808
22809
22810
22811
22812
22813
22814
22815
22816
22817
22818
22819
22820
22821
22822
22823
22824
22825
22826
22827
22828
22829
22830
22831
22832
22833
22834
22835
22836
22837
22838
22839
22840
22841
22842
22843
22844
22845
22846
22847
22848
22849
22850
22851
22852
22853
22854
22855
22856
22857
22858
22859
22860
22861
22862
22863
22864
22865
22866
22867
22868
22869
22870
22871
22872
22873
22874
22875
22876
22877
22878
22879
22880
22881
22882
22883
22884
22885
22886
22887
22888
22889
22890
22891
22892
22893
22894
22895
22896
22897
22898
22899
22900
22901
22902
22903
22904
22905
22906
22907
22908
22909
22910
22911
22912
22913
22914
22915
22916
22917
22918
22919
22920
22921
22922
22923
22924
22925
22926
22927
22928
22929
22930
22931
22932
22933
22934
22935
22936
22937
22938
22939
22940
22941
22942
22943
22944
22945
22946
22947
22948
22949
22950
22951
22952
22953
22954
22955
22956
22957
22958
22959
22960
22961
22962
22963
22964
22965
22966
22967
22968
22969
22970
22971
22972
22973
22974
22975
22976
22977
22978
22979
22980
22981
22982
22983
22984
22985
22986
22987
22988
22989
22990
22991
22992
22993
22994
22995
22996
22997
22998
22999
23000
23001
23002
23003
23004
23005
23006
23007
23008
23009
23010
23011
23012
23013
23014
23015
23016
23017
23018
23019
23020
23021
23022
23023
23024
23025
23026
23027
23028
23029
23030
23031
23032
23033
23034
23035
23036
23037
23038
23039
23040
23041
23042
23043
23044
23045
23046
23047
23048
23049
23050
23051
23052
23053
23054
23055
23056
23057
23058
23059
23060
23061
23062
23063
23064
23065
23066
23067
23068
23069
23070
23071
23072
23073
23074
23075
23076
23077
23078
23079
23080
23081
23082
23083
23084
23085
23086
23087
23088
23089
23090
23091
23092
23093
23094
23095
23096
23097
23098
23099
23100
23101
23102
23103
23104
23105
23106
23107
23108
23109
23110
23111
23112
23113
23114
23115
23116
23117
23118
23119
23120
23121
23122
23123
23124
23125
23126
23127
23128
23129
23130
23131
23132
23133
23134
23135
23136
23137
23138
23139
23140
23141
23142
23143
23144
23145
23146
23147
23148
23149
23150
23151
23152
23153
23154
23155
23156
23157
23158
23159
23160
23161
23162
23163
23164
23165
23166
23167
23168
23169
23170
23171
23172
23173
23174
23175
23176
23177
23178
23179
23180
23181
23182
23183
23184
23185
23186
23187
23188
23189
23190
23191
23192
23193
23194
23195
23196
23197
23198
23199
23200
23201
23202
23203
23204
23205
23206
23207
23208
23209
23210
23211
23212
23213
23214
23215
23216
23217
23218
23219
23220
23221
23222
23223
23224
23225
23226
23227
23228
23229
23230
23231
23232
23233
23234
23235
23236
23237
23238
23239
23240
23241
23242
23243
23244
23245
23246
23247
23248
23249
23250
23251
23252
23253
23254
23255
23256
23257
23258
23259
23260
23261
23262
23263
23264
23265
23266
23267
23268
23269
23270
23271
23272
23273
23274
23275
23276
23277
23278
23279
23280
23281
23282
23283
23284
23285
23286
23287
23288
23289
23290
23291
23292
23293
23294
23295
23296
23297
23298
23299
23300
23301
23302
23303
23304
23305
23306
23307
23308
23309
23310
23311
23312
23313
23314
23315
23316
23317
23318
23319
23320
23321
23322
23323
23324
23325
23326
23327
23328
23329
23330
23331
23332
23333
23334
23335
23336
23337
23338
23339
23340
23341
23342
23343
23344
23345
23346
23347
23348
23349
23350
23351
23352
23353
23354
23355
23356
23357
23358
23359
23360
23361
23362
23363
23364
23365
23366
23367
23368
23369
23370
23371
23372
23373
23374
23375
23376
23377
23378
23379
23380
23381
23382
23383
23384
23385
23386
23387
23388
23389
23390
23391
23392
23393
23394
23395
23396
23397
23398
23399
23400
23401
23402
23403
23404
23405
23406
23407
23408
23409
23410
23411
23412
23413
23414
23415
23416
23417
23418
23419
23420
23421
23422
23423
23424
23425
23426
23427
23428
23429
23430
23431
23432
23433
23434
23435
23436
23437
23438
23439
23440
23441
23442
23443
23444
23445
23446
23447
23448
23449
23450
23451
23452
23453
23454
23455
23456
23457
23458
23459
23460
23461
23462
23463
23464
23465
23466
23467
23468
23469
23470
23471
23472
23473
23474
23475
23476
23477
23478
23479
23480
23481
23482
23483
23484
23485
23486
23487
23488
23489
23490
23491
23492
23493
23494
23495
23496
23497
23498
23499
23500
23501
23502
23503
23504
23505
23506
23507
23508
23509
23510
23511
23512
23513
23514
23515
23516
23517
23518
23519
23520
23521
23522
23523
23524
23525
23526
23527
23528
23529
23530
23531
23532
23533
23534
23535
23536
23537
23538
23539
23540
23541
23542
23543
23544
23545
23546
23547
23548
23549
23550
23551
23552
23553
23554
23555
23556
23557
23558
23559
23560
23561
23562
23563
23564
23565
23566
23567
23568
23569
23570
23571
23572
23573
23574
23575
23576
23577
23578
23579
23580
23581
23582
23583
23584
23585
23586
23587
23588
23589
23590
23591
23592
23593
23594
23595
23596
23597
23598
23599
23600
23601
23602
23603
23604
23605
23606
23607
23608
23609
23610
23611
23612
23613
23614
23615
23616
23617
23618
23619
23620
23621
23622
23623
23624
23625
23626
23627
23628
23629
23630
23631
23632
23633
23634
23635
23636
23637
23638
23639
23640
23641
23642
23643
23644
23645
23646
23647
23648
23649
23650
23651
23652
23653
23654
23655
23656
23657
23658
23659
23660
23661
23662
23663
23664
23665
23666
23667
23668
23669
23670
23671
23672
23673
23674
23675
23676
23677
23678
23679
23680
23681
23682
23683
23684
23685
23686
23687
23688
23689
23690
23691
23692
23693
23694
23695
23696
23697
23698
23699
23700
23701
23702
23703
23704
23705
23706
23707
23708
23709
23710
23711
23712
23713
23714
23715
23716
23717
23718
23719
23720
23721
23722
23723
23724
23725
23726
23727
23728
23729
23730
23731
23732
23733
23734
23735
23736
23737
23738
23739
23740
23741
23742
23743
23744
23745
23746
23747
23748
23749
23750
23751
23752
23753
23754
23755
23756
23757
23758
23759
23760
23761
23762
23763
23764
23765
23766
23767
23768
23769
23770
23771
23772
23773
23774
23775
23776
23777
23778
23779
23780
23781
23782
23783
23784
23785
23786
23787
23788
23789
23790
23791
23792
23793
23794
23795
23796
23797
23798
23799
23800
23801
23802
23803
23804
23805
23806
23807
23808
23809
% arara: lualatex
% arara: bib2gls: {group: on} if missing("glstex") || changed("bib") || found("log", "Warning: Glossary entry")
% arara: bibtex
% arara: lualatex if changed("glstex") || missing("toc")
% arara: bib2gls: {group: on, packages: [mfirstuc-english]}
% arara: lualatex
% arara: lualatex if found ("log", "Rerun")
\documentclass[titlepage=false,index=totoc,bibliography=totoc,
 fontsize=12pt,captions=tableheading]{scrreprt}

\usepackage[autooneside=false]{scrlayer-scrpage}
\usepackage{pifont}

% Need support for extended characters
% and need a mono font that supports hyphenation
\usepackage[no-math]{fontspec}
\setmainfont{Linux Libertine O}

\newfontface\cyrillicmono{FreeMono}[Scale=MatchLowercase]
\newcommand{\textcyrillicmono}[1]{{\cyrillicmono #1}}

\newfontface\greekmono{FreeMono}[Scale=MatchLowercase]
\newfontface\freeserif{FreeSerif}
\newcommand{\devone}{{\freeserif\char"0967}}
\newcommand{\devtwo}{{\freeserif\char"0968}}
\newcommand{\devsix}{{\freeserif\char"096C}}
\newcommand{\insularG}{{\freeserif\char"A77D}}
\newcommand{\insularg}{{\freeserif\char"1D79}}
\newcommand{\longs}{\char"017F}
\newcommand{\Wynn}{\char"01F7}
\newcommand{\wynn}{\char"01BF}

\usepackage[x11names]{xcolor}
\usepackage{alltt}
\usepackage{array}
\usepackage{upquote}
\usepackage{etoolbox}

\usepackage{hologo}
\usepackage{amsmath}
\usepackage{accents}
\usepackage{tikz}
\usepackage{tcolorbox}
\usepackage{datetime2}
\usepackage{listings}
\usepackage[hidelinks]{hyperref}
\usepackage{cleveref}
\usepackage[% v1.21+
   record,% use bib2gls
   index,% create index glossary
   entrycounter,% enable entry counting
   subentrycounter,% enable entry counting
   %debug=showtargets,% debugging information
   %debug=showwrgloss,% debugging information
   stylemods={mcols,bookindex},% adjust predefined styles and load glossary-mcols.sty and glossary-bookindex.sty
   style=bookindex
 ]
 {glossaries-extra}

\lstset{language={[LaTeX]TeX},upquote,basicstyle={\ttfamily\small},
 commentstyle={\color{gray}}}

\renewcommand*{\glsshowtarget}[1]{\texttt{\small [#1]}}

\glsxtrprovidestoragekey{note}{}{}
\glsxtrprovidestoragekey{package}{}{}
\glsxtrprovidestoragekey{nametitle}{}{}

\providecommand{\omicron}{o}

% put break before dash to avoid confusion with a hyphen
\newcommand{\dhyphen}{%
 \texorpdfstring
 {\discretionary{}{}{}\texttt{-}}%
 {-}%
}

% Redefine \- (used in entry names) so that hyphenation in command names, 
% switches options etc uses a serif hyphen to reduce confusion with an 
% actual hyphen in the code.
\renewrobustcmd{\-}{%
 \discretionary
 {{\rmfamily\char\ifnum\hyphenchar\font<0
  \defaulthyphenchar\else\hyphenchar\font\fi
 }}%
 {}{}%
}

\renewcommand{\glsentrycounterlabel}{}
\renewcommand{\glssubentrycounterlabel}{}
\renewcommand\glstreeitem{\par\hangindent10pt}
\renewcommand\glstreesubitem{\par\hangindent30pt\hspace*{15pt}}

\newcommand{\bibglsseealsosep}{\glstreesubitem}

\renewcommand*{\glsseesep}{, \glstreesubitem\quad}
\renewcommand*{\glsseelastsep}{\glstreesubitem\quad\strut\llap{\andname\space}}

\newcommand{\glsxtrpostnameabbreviation}{%
 \space(\glsentrylong{\glscurrententrylabel})%
}

\newcommand{\glsxtrpostnameenvironment}{\space environment}
\newcommand{\glsxtrpostnamecounter}{\space counter}

\newcommand*{\csfmtfont}[1]{\texttt{#1}}
\newcommand*{\csfmtcolourfont}[1]{\texttt{\textcolor{cs}{#1}}}

\newcommand*{\csfmt}[1]{%
 \texorpdfstring
 {\csfmtfont{\char`\\ #1}}%
 {\string\\#1}%
}

\newcommand{\texparserdefprefix}{}
\newcommand{\texparserdefnotetarget}{%
  \raisebox{2ex}{\hypertarget{\texparserdefprefix texparserdef}{}}%
  \textsuperscript{*}Indicates command is
  recognised by \bibgls's interpreter although it may have
  a slightly different implementation.
}
\newcommand{\texparserdefnote}{\hyperlink{\texparserdefprefix texparserdef}{*}}

\setabbreviationstyle{long-short-sc}
\setabbreviationstyle[common]{short-sc-nolong}

\glsxtrprovidestoragekey{unsortedprogeny}{}{}

\newcommand*{\longargfmt}[1]{%
 \texorpdfstring{\texttt{\longswitch #1}}%
 {\string-\string-#1}%
}

\GlsXtrLoadResources[
  entry-type-aliases={
   bibglscommand=dualindexentry,
   mainglscommand=spawndualindexentry,
   glscommand=spawnentry,
   abbrvstylecommand=spawnentry,
   glostylecommand=spawnentry,
   resourceoption=dualindexentry,
   packageoption=dualindexentry,
   field=dualindexentry,
   entrytype=dualindexentry,
   switch=dualindexentry,
   samplefile=dualindexentry,
   examplecommand=dualindexentry,
   examplesymbol=symbol,
   exampleabbreviation=abbreviation,
   exampleentry=entry
  },
  max-loc-diff={2},
  see=omit,seealso=omit,alias=omit,
  field-aliases={topics=adoptparents,annote=user2},
  replicate-fields={name=nametitle,progeny=unsortedprogeny},
  field-case-change={nametitle=title},
  interpret-fields={nametitle},
  no-case-change-cs={fieldfmt,abbrstylefmt,glostylefmt},
  short-case-change={lc},
  dual-short-case-change={lc},
  unknown-entry-alias={index},
  save-original-entrytype,
  label-prefix={idx.},
  save-loclist=false,
  sort={letternumber-nocase},
  sort-replace={{\glshex2423}{ },{\string\\([a-zA-Z])}{\glscapturedgroup1}},
  type={index},
  dual-field={dualid},
  match-action={add},
  match-op={or},
  match={
   {category}={.*field},
   {original entrytype}={.*(gls|style)command},
   {original entrytype}={.*option},
   {original entrytype}={abbreviationstyle},
   {original entrytype}={glossarystyle},
  },
  dual-prefix={},
  dual-type={main},
  combine-dual-locations={primary},
  selection={recorded and deps and see not also},
  symbol-sort-fallback={name},
  sort-label-list={progeny:letter-nocase:glsxtrentryparentname},
  strip-missing-parents,
  save-child-count
]

\glsxtraddlabelprefix{}
\glsxtraddlabelprefix{idx.}

\DTMsavetimestamp{creation}{2017-01-20T15:39:00Z}

\IfFileExists{../java/Bib2Gls.java}
{
  \DTMsavefilemoddate{moddate}{../java/Bib2Gls.java}
}
{
  \DTMsavenow{moddate}
}

\newif\ifmainmatter

% ignore locations outside of main matter
\GlsXtrSetDefaultNumberFormat{glsignore}

\GlsXtrSetAltModifier{!}{format=glsignore}

\newcommand{\frontmatter}{%
 \clearpage\pagenumbering{roman}%
 \pagestyle{scrheadings}%
 \automark[section]{chapter}%
 \mainmatterfalse
}
\newcommand{\mainmatter}{%
 \clearpage\pagenumbering{arabic}%
 \GlsXtrSetDefaultNumberFormat{glsnumberformat}%
 \mainmattertrue
}

\newcommand{\rightleftmark}[2]{\ifstrequal{#1}{#2}{#1}{#1: #2}}
\makeatletter
\newcommand{\setsecdepth}{\@ifstar\s@setsecdepth\@setsecdepth}
\newcommand{\s@setsecdepth}[1]{%
 \setcounter{secnumdepth}{#1}%
 \automark[section]{chapter}%
 \let\@rightmark\rightleftmark
}
\newcommand{\@setsecdepth}[1]{%
 \setcounter{secnumdepth}{#1}%
 \ifnum#1 > 0
  \automark[section]{chapter}%
 \else
  \automark{chapter}%
 \fi
 \let\@rightmark\@secondoftwo
}
\makeatother

\newtcolorbox{important}{colback=red!5!white,colframe=red}

\newcommand{\bibgls}{\appfmt{bib2gls}}

\newcommand*{\BibTeX}{\hologo{BibTeX}}
\newcommand*{\eTeX}{\hologo{eTeX}}
\newcommand*{\XeLaTeX}{\hologo{XeLaTeX}}
\newcommand*{\LuaLaTeX}{\hologo{LuaLaTeX}}
\newcommand*{\pdfLaTeX}{\hologo{pdfLaTeX}}

\newcommand{\langxml}{\hyperref[sec:lang.xml]{language resource file}}

\newcommand{\hex}[1]{0x#1}
\newcommand{\hexsb}[1]{\textsubscript{\scriptsize\rmfamily#1}}
\newcommand{\subfigfmt}[1]{(\emph{#1})}

\newcommand{\qt}[1]{``#1''}

\newcommand{\qtt}[1]{\qt{\,\texttt{#1}\,}}

\newcommand{\qtdelim}[1]{\idx{doublequotechardelim}#1\idx{doublequotechardelim}}

\newcommand{\ifnumhex}[1]{\cs{ifnum}\idx{doublequotecharhex}#1}

\newcommand{\charhex}[1]{\cs{char}\idx{doublequotecharhex}#1}
\newcommand{\icharhex}[1]{\ics{char}\idx[format=glsnumberformat,noindex=false]{doublequotecharhex}#1}

\newcommand{\era}[1]{\texorpdfstring{\textsc{\MakeLowercase{#1}}}{#1}}

\newcommand{\dequals}{%
 \texorpdfstring
 {\discretionary{}{}{}\texttt{=}\discretionary{}{}{}}%
 {=}%
}

\newcommand{\dcomma}{%
 \texorpdfstring
 {\texttt{,}\discretionary{}{}{}}%
 {,}%
}

\newcommand{\dcolon}{%
 \texorpdfstring
 {\texttt{:}\discretionary{}{}{}}%
 {:}%
}

\pdfstringdefDisableCommands{%
  \def\dhyphen{-}%
  \def\dcolon{:}%
  \def\dcomma{,}%
  \def\dequals{,}%
  \let\-\empty
}

% Provide commands that work like \gls[#1]{idx.#2}[#3]

\glsxtrnewglslike{idx.}{\idx}{\idxpl}{\Idx}{\Idxpl}
\newcommand{\idxlink}[3][]{\glslink[#1]{idx.#2}{#3}}

\newcommand{\stringu}{\idx{cs.string}\idx{uhex}}

\newcommand*{\filefmt}[1]{\texorpdfstring{\texttt{#1}}{#1}}
\newcommand*{\metafilefmt}[3]{%
  \filefmt{#1}\discretionary{}{}{}\meta{#2}\discretionary{}{}{}\filefmt{#3}%
}
\newcommand*{\metametafilefmt}[5]{%
  \filefmt{#1}\discretionary{}{}{}\meta{#2}\discretionary{}{}{}\filefmt{#3}%
  \discretionary{}{}{}\meta{#4}\discretionary{}{}{}\filefmt{#5}%
}

\newcommand*{\primaryresourcefmt}{%
 \texorpdfstring
  {\texttt{\char`\\ jobname.glstex}}%
  {\string\\jobname.glstex}%
}

\newcommand*{\suppresourcefmt}[1]{%
 \texorpdfstring
  {\texttt{\char`\\ jobname-#1.glstex}}%
  {\string\\jobname-#1.glstex}%
}

\definecolor{field}{named}{DarkSlateGray4}
\definecolor{cs}{named}{DarkSeaGreen4}
\definecolor{styopt}{named}{DarkOrchid4}
\definecolor{entry}{named}{SteelBlue4}
\definecolor{comment}{named}{gray}
\definecolor{attribute}{named}{Purple4}
\definecolor{style}{named}{Blue4}

\newcommand{\extstyopt}{\textsuperscript{\textdagger}}

\newcommand*{\appfmt}[1]{\texorpdfstring{\texttt{#1}}{#1}}
\newcommand*{\styfmt}[1]{\texorpdfstring{\textsf{#1}}{#1}}
\newcommand*{\envfmt}[1]{\texorpdfstring{\textsf{#1}}{#1}}
\newcommand*{\optfmt}[1]{\texorpdfstring{\texttt{#1}}{#1}}
\newcommand*{\csoptfmt}[1]{\texorpdfstring{\textcolor{cs}{\optfmt{#1}}}{#1}}
\newcommand*{\styoptfmt}[1]{\texorpdfstring{\textcolor{styopt}{\optfmt{#1}}}{#1}}
\newcommand*{\fieldfmt}[1]{\texorpdfstring{\texttt{\color{field}#1}}{#1}}
\newcommand*{\entryfmt}[1]{\texorpdfstring{\texttt{\color{entry}#1}}{#1}}
\newcommand*{\atentryfmt}[1]{\entryfmt{@#1}}
\newcommand*{\abbrstylefmt}[1]{\texorpdfstring{\textsf{#1}}{#1}}
\newcommand*{\glostylefmt}[1]{\texorpdfstring{\textsf{#1}}{#1}}
\newcommand*{\catattrfmt}[1]{\texorpdfstring{\textsf{#1}}{#1}}
\newcommand*{\counterfmt}[1]{\texorpdfstring{\textsf{#1}}{#1}}
\newcommand*{\idprefixfmt}[1]{%
 \texorpdfstring{\texttt{#1\spacefactor999 .}}{#1.}}
\newcommand*{\xmltagfmt}[1]{\texorpdfstring{\texttt{#1}}{#1}}

\newcommand*{\argor}{\texorpdfstring{\protect\textbar}{|}}

\newrobustcmd*{\texmeta}[1]{{\normalfont\normalcolor$\langle$\emph{#1}$\rangle$}}

\newcommand*{\meta}[1]{%
 \texorpdfstring{\ifmmode\text{\texmeta{#1}}\else\texmeta{#1}\fi}{#1}%
}

\newcommand*{\oarg}[1]{\discretionary{}{}{}[#1]}
\newcommand*{\oargm}[1]{\oarg{\meta{#1}}}

\newcommand*{\marg}[1]{\texorpdfstring
 {\discretionary{}{}{}\char`\{#1\char`\} }%
 {\{#1\}}%
}

\newcommand*{\margm}[1]{\marg{\meta{#1}}}

\newcommand*{\file}[1]{%
 \texorpdfstring
 {\idx{file.#1}}%
 {#1}%
}

\newcommand{\texmfcnffmt}[1]{\texttt{#1}}

\glsxtrnewgls{idx.texmfcnf.}{\texmfcnf}

\newcommand*{\extfmt}[1]{\filefmt{.#1}}%

\glsxtrnewgls[format=glsignore]{idx.ext.}{\ext}

\newcommand*{\iext}[1]{%
 \glsxtrtitleorpdforheading{\idx{ext.#1}}{.#1}{\extfmt{#1}}%
}

\glsxtrnewgls[format=glsignore]{idx.}{\sty}

\newcommand*{\isty}[1]{%
  \texorpdfstring{\idx[noindex=false]{#1}}{#1}%
}

\newcommand*{\env}[1]{%
  \texorpdfstring{\idx{env.#1}}{#1}%
}

\newcommand{\glsxtrpostdescenvironment}{environment}

\newcommand*{\abbrstyle}[1]{%
  \texorpdfstring{\idx{abbrstyle.#1}}{#1}%
}

\newcommand*{\topicabbrstyle}[1]{%
  %\texorpdfstring{\idx[prefix=topic.]{abbrstyle.#1}}{#1}%
  \texorpdfstring{\idx{abbrstyle.#1}}{#1}%
}

\newcommand*{\glostyle}[1]{%
  \texorpdfstring{\idx{glostyle.#1}}{#1}%
}

\newcommand*{\topicglostyle}[1]{%
 %\texorpdfstring{\idx[prefix=topic.]{glostyle.#1}}{#1}%
 \texorpdfstring{\idx{glostyle.#1}}{#1}%
}

\newcommand*{\iglostyle}[1]{%
  \texorpdfstring{\idx[noindex=false]{glostyle.#1}}{#1}%
}

\newcommand*{\catattr}[1]{%
  \texorpdfstring{\idx{catattr.#1}}{#1}%
}

\newcommand*{\counter}[1]{%
  \texorpdfstring{\idx{ctr.#1}}{#1}%
}

\newcommand{\glsxtrpostdesccounter}{counter}

\newcommand*{\idprefix}[1]{%
 \texorpdfstring{\idx{idprefix.#1}}{#1}%
}


\newcommand*{\styopt}[2][]{%
  \texorpdfstring%
  {%
    \gls{styopt.#2}\styoptfmt{\ifblank{#1}{}{\dequals\marg{#1}}}%
  }%
  {#2\ifblank{#1}{}{=#1}}%
}

\newcommand*{\istyopt}[2][]{%
  \texorpdfstring%
  {%
    \gls[noindex=false]{styopt.#2}\styoptfmt{\ifblank{#1}{}{\dequals\marg{#1}}}%
  }%
  {#2\ifblank{#1}{}{=#1}}%
}

\newcommand*{\styoptref}[1]{\gls[format=glsignore]{styopt.#1}}

\newcommand*{\keyvallist}{%
 \texorpdfstring
 {key\dequals value list}%
 {key=value list}%
}

\newcommand{\cssection}[2][\subsection]{%
 #1{\texorpdfstring{\glsxtrglossentry{#2}}{\string\\#2}}%
}

\newcommand{\format}[1]{%
  \gls{#1}\glsentryuseri{#1}%
}

\newcommand{\formatdef}[1]{%
  \begin{definition}%
  \gls[hyper=false]{#1}\glsentryuseri{#1}%
  \end{definition}\ignorespaces
}

\newcommand{\nosecformatdef}[1]{%
  \begin{definition}
   \renewcommand{\GlsXtrStandaloneEntryName}[1]{\glstarget{##1}{\strut}\cs{##1}}%
   \glsadd{#1}\glsxtrglossentry{#1}%
   \glsentryuseri{#1}%
  \end{definition}\ignorespaces
}

\newcommand{\nosecdef}[1]{%
  \begin{flushleft}\ttfamily
   \renewcommand{\GlsXtrStandaloneEntryName}[1]{\glstarget{##1}{\strut}\cs{##1}}%
   \glsadd{#1}\glsxtrglossentry{#1}%
   \glsentryuseri{#1}%
  \end{flushleft}\ignorespaces
}

\newcommand{\inlinedef}[1]{%
  \code
  {%
   \glsadd{#1}\glsxtrglossentry{#1}%
   \glsentryuseri{#1}%
  }%
}

\newcommand*{\csref}{\gls}

\newcommand*{\cs}[1]{%
  \glsdoifexists{idx.#1}%
  {\glsdohyperlink{\glolinkprefix cs.idx.#1}{\glsentryname{idx.#1}}}%
}

\glsxtridentifyglslike{idx.}{\cs}

\newcommand*{\ics}[2][]{\glsadd[#1]{idx.#2}\cs{#2}}

\newcommand*{\csdisp}[2]{%
 \texorpdfstring
 {\glsdohyperlink{\glolinkprefix cs.idx.#1}{\csfmt{#2}}}%
 {\string\\#2}%
}

\newcommand*{\icsdisp}[3][]{%
 \glsadd[#1]{idx.#2}\glsdohyperlink{\glolinkprefix cs.idx.#2}{\csfmt{#3}}%
}

\newcommand*{\encapdisp}[2]{%
 \texorpdfstring
 {\glsdohyperlink{\glolinkprefix cs.idx.#1}{\code{#2}}}%
 {#2}%
}

\glsxtridentifyglslike{idx.}{\ics}

\newcommand*{\icswithargs}[2][]{\glsadd[#1]{idx.#2}\cs{#2}\glsentryuseri{#2}}

\newcommand*{\postdeschook}[1]{%
 \glsdohyperlink{\glolinkprefix cs.idx.glsxtrpostdesccategory}{\csfmt{glsxtrpostdesc#1}}}
\newcommand*{\postnamehook}[1]{%
 \glsdohyperlink{\glolinkprefix cs.idx.glsxtrpostnamecategory}{\csfmt{glsxtrpostname#1}}}
\newcommand*{\postlinkhook}[1]{%
 \glsdohyperlink{\glolinkprefix cs.idx.glsxtrpostlinkcategory}{\csfmt{glsxtrpostlink#1}}}

\newcommand*{\encap}[2][]{\glsadd[#1]{idx.#2}%
 {%
   \let\csfmt\code
   \glsdohyperlink{\glolinkprefix cs.idx.#2}{\glsentryname{idx.#2}}%
 }%
}

\newcommand{\optsection}[2][\subsection]{%
 \glsxtrifhasfield*{useri}{opt.#2}%
 {%
   #1[\texorpdfstring
       {\glsxtrheadname{opt.#2}\csoptfmt{=\glscurrentfieldvalue}}%
       {\glsentryname{opt.#2}=\glscurrentfieldvalue}]
     {\glsadd{opt.#2}\glsxtrglossentry{opt.#2}%
       \csoptfmt{=\glscurrentfieldvalue}% 
     }%
 }%
 {%
   #1[\texorpdfstring{\glsxtrheadname{opt.#2}}{\glsentryname{opt.#2}}]
     {\glsadd{opt.#2}\glsxtrglossentry{opt.#2}}%
 }%
}

\glsxtrnewgls{file.}{\exfile}

\newcommand{\filesection}[2][\section]{%
 #1[\texorpdfstring{\glsxtrheadname{file.#2}}{\glsentryname{file.#2}}]
   {\glsadd{file.#2}\glsxtrglossentry{file.#2}}%
}

\newcommand*{\csopt}[2][]{\gencsopt{#1}{opt}{#2}}%
\newcommand*{\glsopt}[2][]{\gencsopt{#1}{idx.gls}{#2}}%
\newcommand*{\glsaddopt}[2][]{\gencsopt{#1}{idx.glsadd}{#2}}%
\newcommand*{\printglossopt}[2][]{\gencsopt{#1}{idx.printgloss}{#2}}%
\newcommand*{\iprintglossopt}[2][]{\glsadd{idx.printgloss.#2}\gencsopt{#1}{idx.printgloss}{#2}}%

\newcommand*{\gencsopt}[3]{%
  \texorpdfstring%
  {%
    \gls{#2.#3}%
    \csoptfmt{\ifblank{#1}{}{\dequals\marg{#1}}}%
  }%
  {#3\ifblank{#1}{}{=#1}}%
}

\newcommand*{\field}[1]{%
 \texorpdfstring
 {\gls{field.#1}}%
 {#1}%
}

\newcommand*{\fielddisp}[2]{%
 \texorpdfstring
 {\glsdisp{field.#1}{\fieldfmt{#2}}}%
 {#2}%
}

\newcommand*{\fielddef}[1]{%
 \glsadd{field.#1}%
 \glsxtrglossentry{field.#1}{}%
}

\newcommand*{\fieldref}[1]{\gls[format=glsignore]{field.#1}}

\newglossarystyle{fieldstyle}{%
  \renewenvironment{theglossary}%
   {%
     \setlength{\glsdescwidth}{\dimexpr\linewidth-\maxnamewidth-4\tabcolsep}%
     \begin{longtable}{l>{\raggedright}p{\glsdescwidth}}%
   }%
   {\end{longtable}}%
  \renewcommand*{\glossaryheader}{%
    \noalign{\xdef\dofieldtabcaption{%
      \noexpand\caption\expandonce\fieldtablelotcaption
        {\expandonce\fieldtablecaption}}}%
    \dofieldtabcaption
    \fieldtablecaptionlabel
    \fieldtablepostcaption\tabularnewline
    \bfseries Field & \bfseries Description\tabularnewline
    \endfirsthead
    \caption*{\fieldtablecaption\ (Continued)}\tabularnewline
    \bfseries Field & \bfseries Description\tabularnewline
    \endhead
  }%
  \renewcommand*{\glsgroupheading}[1]{}%
  \renewcommand{\glossentry}[2]{%
    \glsadd{##1}%
    \glsentryitem{##1}\glstarget{##1}{\glossentryname{##1}} &
    \glossentrydesc{##1}\tabularnewline
  }%
  \renewcommand{\subglossentry}[3]{\glossentry{##2}{##3}}%
  \renewcommand*{\glsgroupskip}{}%
}

\newcommand*{\fieldhook}[1]{%
  \edef\currentcategory{\glscategory{#1}}%
  \glsxtriflabelinlist{\currentcategory}{\fieldcategories}%
  {%
     \settowidth{\dimen0}{\glsentryname{#1}}%
     \ifdim\dimen0>\maxnamewidth \setlength{\maxnamewidth}{\dimen0}\fi
  }%
  {\printunsrtglossaryskipentry}%
}

\newlength\maxnamewidth

%\printfields[postcaption]{category list}[lot caption]{caption}{label}
\newcommand*{\printfields}[2][]{%
  \ifstrempty{#1}%
  {%
    \def\fieldtablepostcaption{}%
  }%
  {%
    \def\fieldtablepostcaption{%
      \tabularnewline
      \multicolumn{2}{p{\linewidth}}{#1}%
      \tabularnewline
    }%
  }%
  \def\fieldcategories{#2}%
  \innerprintfields
}

\newcommand*{\innerprintfields}[3][]{%
 \printunsrtglossary*[style=fieldstyle]{%
   \renewcommand*\glossarysection[2][]{}%
   \ifstrempty{#1}%
   {%
     \def\fieldtablelotcaption{}%
   }%
   {%
     \def\fieldtablelotcaption{[#1]}%
   }%
   \def\fieldtablecaption{#2}%
   \def\fieldtablecaptionlabel{\label{#3}}%
   \setlength{\maxnamewidth}{0pt}%
   \let\printunsrtglossaryentryprocesshook\fieldhook
 }%
}

\renewcommand*{\glsxtrbookindexname}[1]{% 
 \glsxtrifhasfield{dualid}{#1}%
 {%
   \glshyperlink[\glossentryname{#1}]{\glscurrentfieldvalue}%
 }
 {%
   \GlsXtrIfFieldEqStr{category}{#1}{command}%
   {\glsdohyperlink{\glolinkprefix cs.#1}{\glossentryname{#1}}}%
   {\glossentryname{#1}}%
 }%
}

\newcommand{\summarysubheaderfont}[1]{\textbf{\sffamily #1}}

\newcommand{\summaryoptionhandler}[1]{%
 \par\smallskip\nopagebreak
 \par\noindent\hspace{1em}%
 \code{\gls{#1}\glsxtrifhasfield{useri}{#1}{=\glscurrentfieldvalue}{}}%
 \glsxtrifhasfield{note}{#1}%
 {\nolinebreak\hfill\mbox{\footnotesize\glscurrentfieldvalue}}%
 {}%
 \nopagebreak\par\hangindent2em\noindent\hspace{2em}\Glossentrydesc{#1}%
}

\newrobustcmd{\donote}[1]{%
 \glsxtrifhasfield{note}{#1}%
 {\nolinebreak\hfill\mbox{\normalfont\footnotesize\glscurrentfieldvalue}}%
 {}%
}

\newglossarystyle{commandsummary}%
{%
  \setglossarystyle{index}%
  \renewcommand*{\glossaryheader}{\raggedright}%
  \renewcommand*{\glsgroupheading}[1]{%
    \glsxtrgetgrouptitle{##1}{\thisgrptitle}%
    \glsxtrbookindexbookmark{\thisgrptitle}{summary.##1}%
    \glsxtrbookindexformatheader{\thisgrptitle}%
    \nopagebreak\indexspace\nopagebreak\csuse{@afterheading}%
  }%
  \renewcommand*{\glossentry}[2]{%
   \glstreeitem \glstarget{cs.##1}{\strut}%
   \glsxtrifhasfield{dualid}{##1}%
   {%
     {%
       \let\csfmtfont\csfmtcolourfont
       \glshyperlink{\glscurrentfieldvalue}%
     }%
     \glsxtrifhasfield{useri}{##1}{\glscurrentfieldvalue}{}%
     \donote{##1}%
     \nopagebreak\par\hspace{10pt}%
     \Glossentrydesc{\glscurrentfieldvalue}%
     \ifglshasdesc{\glscurrentfieldvalue}{\@. }{}%
   }%
   {%
     {\let\csfmtfont\csfmtcolourfont\gls[hyper=false]{##1}}%
     \glsxtrifhasfield{useri}{##1}{\glscurrentfieldvalue}{}%
     \donote{##1}%
     \nopagebreak\par\hspace{10pt}%
     \Glossentrydesc{##1}%
     \ifglshasdesc{##1}{\@. }{}%
   }%
   \GlsXtrIfFieldEqStr{originalentrytype}{##1}{glscommand}%
   {%
     % Does this entry have options?
     \GlsXtrIfHasNonZeroChildCount{##1}%
     {%
       \glspar\medskip\glspar 
       \summarysubheaderfont{Options:} 
       \glsxtrfieldforlistloop{##1}{childlist}{\summaryoptionhandler}%
     }%
     {}%
   }%
   {%
     \GlsXtrIfHasNonZeroChildCount{##1}%
     {This command has \glscurrentfieldvalue\ forms:}{}%
   }%
   %\listtopics{##1}%
   \par\medskip
  }%
  \renewcommand*{\subglossentry}[3]{%
   \glstreesubitem \glstarget{cs.##2}{\strut}%
   \glsxtrifhasfield{dualid}{##2}%
   {%
     {%
       \let\csfmtfont\csfmtcolourfont
       \glshyperlink{\glscurrentfieldvalue}%
     }%
     \glsxtrifhasfield{useri}{##2}{\glscurrentfieldvalue}{}%
     \donote{##2}%
     \nopagebreak\par\hspace{20pt}%
     \ifglshasdesc{\glscurrentfieldvalue}%
     {\Glossentrydesc{\glscurrentfieldvalue}.}{}%
   }%
   {%
     {\let\csfmtfont\csfmtcolourfont\gls[hyper=false]{##2}}%
     \glsxtrifhasfield{useri}{##2}{\glscurrentfieldvalue}{}%
     \donote{##2}%
     \nopagebreak\par\hspace{20pt}%
     \ifglshasdesc{##2}{\Glossentrydesc{##2}.}{}%
   }%
   \par\medskip
  }%
}

\newcommand*{\commandsummaryhook}[1]{%
  \edef\currentcategory{\glscategory{#1}}%
  \ifdefstring{\currentcategory}{command}%
  {%
    \glsxtrifhasfield*{progenitor}{#1}%
    {\printunsrtglossaryskipentry}{}%
  }%
  {\printunsrtglossaryskipentry}%
}

\newcommand*{\printcommandsummary}{%
  \printunsrtglossary*[type=index,
    style=commandsummary,
    title={General Command Summary},
  ]%
  {%
   \def\glossarypreamble{This is an alphabetical summary of 
    commands referenced in this document. See the
    relevent user guides for further details.\endgraf\medskip
    \texparserdefnotetarget}%
   \let\printunsrtglossaryentryprocesshook\commandsummaryhook
   \let\csref\cs
   \glsxtrlocalsetgrouptitle{glssymbols}{}%
  }%
}

\newcommand{\topicxrhandler}[1]{%
 \glsxtrifhasfield{parent}{#1}%
 {%
   \let\field\fieldfmt
   \hyperlink
   {topic.\glscurrentfieldvalue}%
   {\glsentryname{\glscurrentfieldvalue}}%
 }%
 {%
   \hyperlink{topic.#1}{\glsentryname{#1}}%
 }% 
}

% \iffirstprogeny{entry label}{topic label}{true}{false}
\makeatletter
\newcommand{\iffirstprogeny}[2]{%
  \let\doiffirstprogeny\@secondoftwo
  \def\firstprogeny{}%
  \glsxtrifhasfield*{unsortedprogeny}{#1}%
  {%
    \edef\suppliedprogeny{#2}%
    \@for\thisprogeny:=\glscurrentfieldvalue\do{%
      \let\firstprogeny\thisprogeny
      \ifdefequal{\thisprogeny}{\suppliedprogeny}%
      {\let\doiffirstprogeny\@firstoftwo}%
      {\let\doiffirstprogeny\@secondoftwo}%
      \@endfortrue
    }%
  }%
  {}%
  \doiffirstprogeny
}
\makeatother

\newcommand{\ifstylecommand}[1]{%
  \letcs\doifstylecommand{@firstoftwo}%
  \GlsXtrIfFieldEqStr{originalentrytype}{#1}{abbrvstylecommand}%
  {}%
  {%
    \GlsXtrIfFieldEqStr{originalentrytype}{#1}{glostylecommand}%
    {}%
    {\global\letcs\doifstylecommand{@secondoftwo}}%
  }%
  \doifstylecommand
}

\newcommand{\listtopics}[1]{%
 \glsxtrifhasfield{progeny}{#1}% 
 {%
   \ifdefequal\glscurrentfieldvalue\glscurrententrylabel
   {}%
   {%
     \par\nopagebreak\medskip\par\footnotesize
     \sloppy
     \ifstylecommand{#1}%
     {\summarysubheaderfont{Styles:}}%
     {\summarysubheaderfont{Topics:}}%
     \space
     \let\DTLlistformatitem\topicxrhandler
     \DTLformatlist{\glscurrentfieldvalue}.\par
   }%
 }%
 {}%
}

\newglossarystyle{commandtopic}%
{%
  \setglossarystyle{index}%
  %\renewcommand*{\glossaryheader}{\raggedright}%
  \renewcommand*{\glsgroupheading}[1]{}%
  \renewcommand*{\glossentry}[2]{%
   \glsfieldfetch{##1}{nametitle}{\thisvalue}%
   \section*{%
     \raisebox{2ex}%
     {\pdfbookmark[1]{\thisvalue}{topic.##1}\hypertarget{topic.##1}{}}%
     \glsdisp{##1}{\thisvalue}\donote{##1}%
   }%
   \expandafter\sectionmark\expandafter{\thisvalue}%
   \ifglshasdesc{##1}{\Glossentrydesc{##1}.}{}%
   \glsxtrifhasfield{seealso}{##1}%
   {%
    \par\smallskip
     \emph{See also:} 
     \let\DTLlistformatitem\topicxrhandler
     \DTLformatlist{\glscurrentfieldvalue}.%
     \par\medskip\par
   }%
   {}%
  }%
  \renewcommand*{\subglossentry}[3]{%
   \glsifcategory{##2}{command}%
   {%
     \glsdohypertarget{topic.##2}{\strut}%
     \glsxtrifhasfield*{progenitor}{##2}%
     {\let\originallabel\glscurrentfieldvalue}%
     {\def\originallabel{##2}}%
%
     \ifstylecommand{##2}%
     {%
        \iffirstprogeny{\originallabel}{##2}%
        {\letcs\dothisentry{@firstofone}}%
        {%
          \glspar
          \ifdefempty\firstprogeny
          {%
            {\let\csfmtfont\csfmtcolourfont\gls[noindex]{\originallabel}}%
            \glsxtrifhasfield{useri}{##2}{\glscurrentfieldvalue}{}%
          }%
          {%
            {\let\csfmtfont\csfmtcolourfont
              \hyperlink{topic.\firstprogeny}{\glsentryname{\firstprogeny}}}%
            \glsxtrifhasfield{useri}{##2}{\glscurrentfieldvalue}{}%
            \glsxtrifhasfield{parent}{\firstprogeny}%
            {\hfill
             {\footnotesize\emph{see} \hyperlink{topic.\glscurrentfieldvalue}%
             {\glsentryname{\glscurrentfieldvalue}}.}}
            {}%
          }%
          \letcs\dothisentry{@gobble}%
        }%
     }%
     {\letcs\dothisentry{@firstofone}}%
     \dothisentry
     {%
       \par\smallskip\par 
       % hyperlink to index:
       {\let\csfmtfont\csfmtcolourfont\gls[format=glsignore]{\originallabel}}%
       \glsxtrifhasfield{useri}{##2}{\glscurrentfieldvalue}{}%
       \glsxtrifhasfield{note}{##2}%
       {\nolinebreak\hfill\mbox{\footnotesize\glscurrentfieldvalue}}%
       {}%
       \nopagebreak\par\smallskip\par\nopagebreak
       \ifglshasdesc{##2}{\Glossentrydesc{##2}.}{}%
       % Topics
       \listtopics\originallabel
       \par\smallskip
     }%
   }%
   {%
     \glsfieldfetch{##2}{nametitle}{\thisvalue}%
     \subsection*{%
       \raisebox{2ex}%
       {\hypertarget{topic.##2}{}\pdfbookmark[2]{\thisvalue}{topic.##2}}%
       \glsdisp{##2}{\thisvalue}\donote{##2}%
     }%
     \expandafter\subsectionmark\expandafter{\thisvalue}%
     \ifglshasdesc{##2}{\Glossentrydesc{##2}.\glspar\nopagebreak\smallskip}{}%
   }%
  }%
}

\newcommand*{\commandtopichook}[1]{%
  \glsxtrifhasfield*{progenitor}{#1}%
  {}%
  {%
    \GlsXtrIfFieldEqStr*{originalentrytype}{#1}{topic}%
    {}%
    {%
      \glsxtrifhasfield*{parent}{#1}%
      {%
        \GlsXtrIfFieldEqStr*{originalentrytype}{\glscurrentfieldvalue}{topic}%
        {}%
        {\printunsrtglossaryskipentry}%
      }%
      {\printunsrtglossaryskipentry}%
    }%
  }%
}

\newcommand*{\printcommandtopic}{%
  \setsecdepth*{0}%
  \automark[subsection]{section}%
  \printunsrtglossary*[type=index,
    style=commandtopic,
    title={Summary by Topic of Glossary Commands},
    label={sec:topics},
    nogroupskip,
    nonumberlist
  ]%
  {%
   \def\texparserdefprefix{topic.}%
   \def\glossarypreamble{This is a summary of commands provided
    by \styfmt{glossaries}, \styfmt{glossaries-extra} and
    their supplementary packages that may be useful to 
    \bibgls\ users. Commands specific to other indexing methods
    aren't listed unless they've been referenced in this manual.
    See the appropriate user guides for further details.
    \endgraf\medskip
    \texparserdefnotetarget}%
   \let\abbrstyle\topicabbrstyle
   \let\glostyle\topicglostyle
   \let\printunsrtglossaryentryprocesshook\commandtopichook
   \let\csref\cs
  }%
  \setsecdepth{0}%
}

\newglossarystyle{styoptsummary}%
{%
  \setglossarystyle{mcolindex}%
  \renewcommand{\glossarypreamble}{%
    Most options are in the form \meta{option}=\meta{value} and
    may have a default if \meta{value} is omitted, 
    but some options don't have values and should not have
    one assigned.  For boolean options, 
    if the value is omitted \styoptfmt{true} is assumed. 
    \extstyopt Indicates a value that's only provided by
    \styfmt{glossaries-extra} and not by the base
    \styfmt{glossaries} package. \par
  }%
  \renewcommand*{\glossaryheader}{}%
  \renewcommand*{\glsgroupheading}[1]{%
    \glsxtrgetgrouptitle{##1}{\thisgrptitle}%
    \glsxtrbookindexbookmark{\thisgrptitle}{styopt.##1}%
    \glsxtrbookindexformatheader{\thisgrptitle}%
    \nopagebreak\indexspace\nopagebreak\csuse{@afterheading}%
  }%
  \renewcommand*{\glossentry}[2]{%
   \item \glstarget{##1}{\strut}%
   \glsxtrifhasfield{dualid}{##1}%
   {%
     \gls{\glscurrentfieldvalue}%
   }%
   {%
     \gls[hyper=false]{##1}%
   }%
   \glsxtrifhasfield{useri}{##1}{=\glscurrentfieldvalue}{}%
   \nopagebreak\par\hspace{10pt}%
   \Glossentrydesc{##1}.%
   \glsxtrifhasfield{package}{##1}%
   {\nopagebreak\par \ding{43}Provided by 
     \ifdefstring{\glscurrentfieldvalue}{glossaries,glossaries-extra}
     {\styfmt{glossaries} and modified by \styfmt{glossaries-extra}}%
     {\styfmt{\glscurrentfieldvalue}}.}%
   {}%
   \glsxtrifhasfield{note}{##1}%
   {\space \xmakefirstuc{\glscurrentfieldvalue}.}%
   {}%
   \par\medskip
  }%
}

\newcommand*{\styoptsummaryhook}[1]{%
  \edef\currentcategory{\glscategory{#1}}%
  \ifdefstring{\currentcategory}{packageoption}%
  {}%
  {\printunsrtglossaryskipentry}%
}

\newcommand*{\printstyoptsummary}{%
  \printunsrtglossary*[type=main,
    style=styoptsummary,
    title={Package Option Summary}
  ]%
  {%
   \let\printunsrtglossaryentryprocesshook\styoptsummaryhook
   \glsxtrlocalsetgrouptitle{glssymbols}{@}%
  }%
}

\newcommand{\entrysection}[2][\subsection]{%
 #1[\texorpdfstring{\glsxtrheadname{entry.#2}}{\glsentryname{entry.#2}}]
   {\glsadd{entry.#2}\glsxtrglossentry{entry.#2}}%
}

\newcommand{\entrydef}[1]{\glsadd{entry.#1}\glsxtrglossentry{entry.#1}}

\newcommand*{\atentry}[2][]{%
 \texorpdfstring
 {\gls[#1]{entry.#2}}%
 {#2}%
}

\newcommand*{\atentrypageref}[1]{%
 \atentry{#1} (page~\glsxtrpageref{entry.#1})}%

\newcommand*{\atentryref}[1]{\gls[format=glsignore]{entry.#1}}

\newcommand*{\entryref}[1]{\glslink[format=glsignore]{entry.#1}{\entryfmt{#1}}}

\newcommand{\convertglsbibarg}[2][\subsection]{\argsection[#1]{gls2bib.#2}}

\newcommand{\argsection}[2][\section]{%
 \def\switcharg{}%
 \def\switchalt{}%
 \def\switchaltpdf{}%
 \glsxtrifhasfield*{useri}{switch.#2}%
 {%
   \edef\switcharg{ \expandonce\glscurrentfieldvalue}%
 }%
 {}%
 \ifglshassymbol{switch.#2}%
 {%
   \def\switchalt{ (or \protect\glsentrysymbol{switch.#2}\switcharg)}%
 }%
 {}%
 #1[\texorpdfstring
     {\glsxtrheadname{switch.#2}}%
     {\glsentryname{switch.#2}}%
     \switcharg\switchalt
   ]
   {\glsadd{switch.#2}%
    \glsxtrglossentry{switch.#2}\switcharg\switchalt}%
}

\newrobustcmd{\longswitch}{\string-{}\string-}

\newcommand*{\shortargfmt}[1]{%
 \texorpdfstring{\texttt{\string-#1}}%
 {\string-#1}%
}

\newcommand*{\longarg}[1]{%
  \texorpdfstring
  {\gls{switch.#1}}%
  {\string-\string-#1}%
}

\newcommand*{\convertglsbiblongarg}[1]{%
  \texorpdfstring
  {\gls{switch.gls2bib.#1}}%
  {\string-\string-#1}%
}

\newcommand*{\longargpageref}[1]{%
 \longarg{#1} (page~\glsxtrpageref{switch.#1})}%

\newcommand*{\longargorshort}[1]{%
  \texorpdfstring
  {\gls{switch.#1}\ifglshassymbol{switch.#1}{ or \glssymbol{switch.#1}}{}}%
  {\string-\string-#1}%
}

\newcommand*{\unicodecategoryfmt}[1]{\qt{\textsf{#1}}}

\definecolor{defbackground}{rgb}{1,1,0.75}
\newsavebox\defsbox
\newlength\defwidth

\newenvironment{definition}%
{%
  \setlength{\fboxsep}{4pt}\setlength{\fboxrule}{1.25pt}%
  \begin{lrbox}{\defsbox}%
   \setlength\defwidth\linewidth
   \addtolength\defwidth{-2\fboxrule}%
   \addtolength\defwidth{-2\fboxsep}%
   \begin{minipage}{\defwidth}
   \flushleft\ttfamily\ignorespaces
}%
{%
   \end{minipage}%
  \end{lrbox}\par\medskip\noindent
  \fcolorbox{black}{defbackground}{\usebox\defsbox}%
  \medskip\par\noindent
  \ignorespacesafterend
}

\newcommand{\introguide}{\href{bib2gls-begin.pdf}{\qt{\styfmt{glossaries-extra}
and \bibgls: An Introductory Guide} (\filefmt{bib2gls-begin.pdf})}}

\setcounter{tocdepth}{4}

\newcommand*{\sectionref}[1]{\cref{#1}}
\newcommand*{\Sectionref}[1]{\Cref{#1}}
\newcommand*{\figureref}[1]{\hyperlink{#1-top}{figure~\ref*{#1}}}
\newcommand*{\Figureref}[1]{\hyperlink{#1-top}{Figure~\ref*{#1}}}
\newcommand*{\tableref}[1]{\cref{#1}}
\newcommand*{\Tableref}[1]{\Cref{#1}}

\newcommand{\figcontents}[3]{%
  \hypertarget{#3-top}{}%
  \centering
  #1\par
  #2\label{#3}%
}

\newlength\imagewidth
\newcount\tmpctr

\newcommand*{\pdftwocol}[3][1]{%
  \renewcommand{\tabcolsep}{2pt}%
  \setlength{\imagewidth}{\dimexpr.5\linewidth-2\tabcolsep}%
  \def\tabcontents{\begin{tabular}{@{}cc@{}}}%
  \tmpctr\numexpr#1-1\relax
  \loop
  \advance\tmpctr by 1\relax
  \eappto\tabcontents{\noexpand\frame{%
   \noexpand\includegraphics[page=\the\tmpctr,width=\noexpand\imagewidth]{#2}}}%
  \ifodd\tmpctr
    \appto\tabcontents{&}%
  \else
    \appto\tabcontents{\\}%
  \fi
  \ifnum\tmpctr<#3
  \repeat
  \appto\tabcontents{\end{tabular}}%
  \tabcontents
}

\makeatletter
\newcommand*{\tablelistcs}[2]{%
 \tmpctr=0\relax
 \def\tabcontents{}%
 \loop
   \advance\tmpctr by 1\relax
   \appto\tabcontents{l}%
 \ifnum\tmpctr<#1
 \repeat
%
 \edef\tabcontents{\noexpand\begin{tabular}{@{}\tabcontents @{}}}%
 \def\thislist{#2}%
 \dtlsortlist{\thislist}{\dtlicompare}%
 \tmpctr=1\relax
 \@for\thislabel:=\thislist\do{%
  \eappto\tabcontents{\noexpand\ics{\thislabel}}%
  \ifnum\tmpctr<#1\relax
   \appto\tabcontents{&}%
   \advance\tmpctr by 1\relax
  \else
   \appto\tabcontents{\\}%
   \tmpctr=1\relax
  \fi
 }%
 \appto\tabcontents{\end{tabular}}%
 \tabcontents
}
\makeatother

% This is a bit fiddly. Need to represent \vec{v} in typewriter
% font for the interpreter examples. ($\mathtt{\vec{v}}$ doesn't
% work)
\newsavebox\varrow
\sbox\varrow{$\mathtt{\accentset{\to}{v}}$}

\newcommand*{\mtx}[1]{\boldsymbol{#1}}
\newcommand*{\set}[1]{\mathcal{#1}}
\newcommand*{\card}[1]{|\set{#1}|}
\newcommand*{\imaginary}{i}

\newcommand{\addr}[1]{\\\href{https://www.#1/}{\nolinkurl{#1}}}
\title{\bibgls: a command line Java application to convert
\filefmt{.bib} files to \filefmt{glossaries-extra.sty} resource
files}
\author{Nicola Talbot\addr{dickimaw-books.com}}
\date{\DTMusedate{moddate}}

\makeatletter
\renewcommand{\fps@figure}{htbp}

\newcommand{\code}[1]{\texorpdfstring{{\ttfamily\obeyspaces #1}}{#1}}

\newcommand{\setupcodeenvfmts}{%
   \def\cmd{\char`\\}%
   \def\comment##1{\mbox{\textcolor{comment}{\idx{commentchar}\ ##1}}}%
   \renewcommand*{\styfmt}[1]{##1}%
   \renewcommand*{\counterfmt}[1]{##1}%
   \renewcommand*{\catattrfmt}[1]{\textcolor{attribute}{##1}}%
   \renewcommand*{\abbrstylefmt}[1]{\textcolor{style}{##1}}%
   \renewcommand*{\csfmtfont}[1]{\textcolor{cs}{##1}}%
}

\newenvironment{codeenv}
 {%
   \GlsXtrStartUnsetBuffering
   \renewcommand{\glslinkpresetkeys}{\setkeys{glslink}{noindex}}%
   \setupcodeenvfmts
   \begin{flushleft}\ttfamily\obeylines\frenchspacing\@vobeyspaces
   \parindent\z@\parfillskip\@flushglue\parskip\z@skip
 }
 {\end{flushleft}\GlsXtrDiscardUnsetBuffering\ignorespacesafterend}

\newenvironment{codeenv*}
 {%
   \GlsXtrStartUnsetBuffering
   \setupcodeenvfmts
   \begin{flushleft}\ttfamily\obeylines\frenchspacing\@vobeyspaces
   \parindent\z@\parfillskip\@flushglue\parskip\z@skip
 }
 {\end{flushleft}\GlsXtrDiscardUnsetBuffering\ignorespacesafterend}

\begingroup
 \renewcommand{\addr}[1]{}
 \let\texorpdfstring\@secondoftwo
 \DTMsetstyle{pdf}
 \protected@edef\x{\endgroup
   \noexpand\hypersetup{%
     pdfinfo={
       Title={\@title},
       Author={\@author},
       CreationDate={\DTMuse{creation}},
       ModDate={\DTMuse{moddate}},
     }%
   }%
 }\x

\makeatother

\renewcommand{\titlepagestyle}{empty}

\begin{document}
\maketitle
\pagenumbering{alph}
\pagestyle{empty}

\begin{abstract}
The \bibgls\ command line application can be used to extract
glossary information stored in a \ext{bib} file and convert it
into glossary entry definitions that can be read using
\styfmt{glossaries-extra}'s \gls{GlsXtrLoadResources} command. When used
in combination with the \styoptfmt{record} package option, \bibgls\
can select only those entries that have been used in the document,
as well as any dependent entries, which reduces the \TeX\ resources
required by not defining unnecessary commands.

Since \bibgls\ can also sort and collate the recorded locations
present in the \ext{aux} file, it can simultaneously by-pass the
need to use \idx!{makeindex} or \idx!{xindy}, although \bibgls\ 
can be used together with an external indexing application if required. (For
example, if a custom \idx{xindy} rule is needed.)

An additional build may be required to ensure the locations are
up-to-date as the page-breaking may be slightly different on the
first \LaTeX\ run due to the unknown references being replaced with
?? which can be significantly shorter than the actual text produced
when the reference is known. 

Note that \bibgls\ is a Java application, and requires at least
Java~8.\footnote{The \code{List.sort} method used to sort the
entries was only introduced to Java~8.}  Additionally,
\sty{glossaries-extra} must be at least version 1.12.  (Although
the latest version is recommended.) This application was
developed in response to the question
\qt{\href{http://tex.stackexchange.com/q/342544}{Is there a program
for managing glossary tags?}} on \TeX\ on
StackExchange~\cite{tex.sx}. The \ext{bib} file can be managed in an
application such as JabRef.

If you already have a \ext{tex} file containing 
entry definitions using commands like \gls{newglossaryentry}
then you can use the supplementary tool \idx{convertgls2bib}
to convert the entries to the \ext{bib} format required by
\bibgls. See \sectionref{sec:gls2bib} for further details.

The supplementary file \introguide\ is an
introductory guide to the \isty{glossaries-extra} package, which you
may prefer to start with if you are unfamiliar with the
\sty{glossaries} and \sty{glossaries-extra} packages.
\end{abstract}

\frontmatter

\tableofcontents
\listoftables
\listoffigures

\mainmatter
\chapter{Introduction}

If you have extensively used the \styfmt{glossaries}~\cite{glossaries} or
\styfmt{glossaries-extra}~\cite{glossaries-extra} package, 
you may have found yourself creating a large \iext{tex} file
containing many definitions that you frequently use in documents.
This file can then simply be loaded using \ics{input} or
\ics{loadglsentries}, but a large file like this can be difficult to
maintain and if the document only actually uses a small proportion
of those entries, the document build is unnecessarily slow due to
the time and resources taken on defining the unwanted entries.

The aim of \bibgls\ is to allow the entries to be stored in a
\iext{bib} file, which can be maintained using a reference system
such as JabRef. The document build process
can now be analogous to that used with \idx{bibtex} (or
\appfmt{biber}), where only those entries that have been recorded in the
document (and possibly their dependent entries) will be extracted
from the \ext{bib} file. Since \bibgls\ can also perform
hierarchical sorting and can collate location lists, it doubles as
an indexing application, which means that the \idx{makeglossaries}
step can be skipped.
Note that \bibgls\ doesn't warn you if an entry that's referenced
in the document doesn't exist in any of the supplied \ext{bib}
files, but instead relies on the \styfmt{glossaries-extra} package
to generate the warning. So at the end of the document build check
the \ext{log} file for warnings.

You can't use \ics{glsaddall} with \bibgls\ as that command works
by iterating over all defined entries and calling
\code{\ics{glsadd}\margm{label}}. On the first \LaTeX\ run there are no
entries defined, so \ics{glsaddall} does nothing. If you want to
select all entries, just use \csopt[all]{selection} instead (which
has the advantage over \ics{glsaddall} in that it doesn't create a
redundant location for each entry).

Note that \bibgls\ requires the extension package
\isty{glossaries-extra} and can't be used with just the base
\isty{glossaries} package, since it requires some of the extension
commands. See the \styfmt{glossaries-extra} user
manual~\cite{glossaries-extra} for information
on the differences between the basic package and the extended
package, as some of the default settings are different.

Since the information used by \bibgls\ is
written to the \iext{aux} file, it's not possible to run \bibgls\
through \TeX's shell escape while the \ext{aux} file is open for
write access. (The \ext{aux} file is closed \emph{after} the end
document hook, so it can't be deferred with \ics{AtEndDocument}.) 
This means that if you really want to run \bibgls\
through \ics{write18} it must be done in the preamble with
\ics{immediate}. For example:
\begin{verbatim}
\immediate\write18{bib2gls \jobname}
\end{verbatim}
As from version 1.14 of \styfmt{glossaries-extra}, this can be done
automatically with the \styopt{automake} option if the \ext{aux}
file exists. (Remember that this will require the shell escape to be
enabled.)

\section{Example Use}
The glossary entries are stored in a \ext{bib} file. For
example, the file \filefmt{entries.bib} might contain:
\begin{codeenv}
\atentry{entry}\marg{bird,
  \field{name}=\marg{bird},
  \field{description} = \marg{feathered animal}
}
\strut
\atentry{abbreviation}\marg{html,
  \field{short}=\qtdelim{html},
  \field{long}=\marg{hypertext markup language}
}
\strut
\atentry{symbol}\marg{v,
  \field{name}=\marg{\idx!{mshiftchar}\cs{vec}\marg{v}\idx!{mshiftchar}},
  \field{text}=\marg{\cs{vec}\marg{v}},
  \field{description}=\marg{a vector}
}
\strut
\atentry{index}\marg{goose,\field{plural}=\qtdelim{geese}}
\end{codeenv}
Here's an example document that uses this data:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cs{usepackage}\oarg{\styoptref{record}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[en-GB]{sort}\comment{sort according to 'en-GB' locale}
}
\strut
\cmd{begin}\marg{document}
\cs{Gls}\marg{bird} and \cs{gls}\marg{goose}.
Symbol: \idx{mshiftchar}\cs{gls}\marg{v}\idx{mshiftchar}.
Abbreviation: \cs{gls}\marg{html}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
If this document is called \filefmt{myDoc.tex}, the build 
process is:
\begin{verbatim}
pdflatex myDoc
bib2gls myDoc
pdflatex myDoc
\end{verbatim}
(This manual assumes \appfmt{pdflatex} for simplicity. Replace
with \appfmt{latex}, \appfmt{xelatex} or \appfmt{lualatex} as
appropriate.) If you want \idxpl{lettergroup} (either headed, with
styles like \glostyle{indexgroup}, or just a blank line separator
with \styopt[false]{nogroupskip}) then you need to use the
\longarg{group} switch:
\begin{verbatim}
pdflatex myDoc
bib2gls --group myDoc
pdflatex myDoc
\end{verbatim}

You can have multiple instances of \gls{GlsXtrLoadResources}. For
example:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{\styopt{record},\styopt{index},\styopt{abbreviations},\styopt{symbols}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[en-GB]{sort},\comment{sort according to 'en-GB' locale}
  \csopt[entrytype=\marg{entry}]{match},\comment{only select \atentry{entry}}
  \csopt[main]{type}\comment{put these entries in the 'main' glossary}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[en-GB]{sort},\comment{sort according to 'en-GB' locale}
  \csopt[entrytype=\marg{abbreviation}]{match},\comment{only select \atentry{abbreviation}}
  \csopt[abbreviations]{type}\comment{put these in the 'abbreviations' glossary}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[letter-case]{sort},\comment{case-sensitive letter sort}
  \csopt[entrytype=\marg{symbol}]{match},\comment{only select \atentry{symbol}}
  \csopt[symbols]{type}\comment{put these entries in the 'symbols' glossary}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[en-GB]{sort},\comment{sort according to 'en-GB' locale}
  \csopt[entrytype=\marg{index}]{match},\comment{only select \atentry{index}}
  \csopt[index]{type}\comment{put these entries in the 'index' glossary}
}
\strut
\cmd{begin}\marg{document}
\cs{Gls}\marg{bird} and \cs{gls}\marg{goose}.
Symbol: \idx!{mshiftchar}\cs{gls}\marg{v}\idx!{mshiftchar}. Abbreviation: \cs{gls}\marg{html}.
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
There are more examples provided in \sectionref{sec:examples}.

Note that there's no need to called \idx!{xindy} or \idx!{makeindex}
since \bibgls\ automatically sorts the entries and collates the locations
after selecting the required entries from the \ext{bib} file and
before writing the temporary file that's input with \gls{glsxtrresourcefile}
(or the more convenient shortcut
\gls{GlsXtrLoadResources}).\footnote{This document will mostly use
the more convenient \gls{GlsXtrLoadResources}.}  This
means the entries are already defined in the correct order, and only
those entries that are required in the document are defined, so
\ics{printunsrtglossary} (or \ics{printunsrtglossaries}) may be used.
(The \qtt{unsrt} part of the command name indicates that all
defined entries should be listed in the order of definition from
\styfmt{glossaries-extra}'s point of view, see the supplementary
document \introguide\ for further details.)

If you don't provide a value with the \styopt{record} option, then
\styopt[only]{record} is assumed. This saves the same indexing
information that's used with the \ics{cs.makeglossaries} and
\ics{makenoidxglossaries} methods (described in the main
\styfmt{glossaries} user manual~\cite{glossaries}). As from
\styfmt{glossaries-extra} version 1.37, you can instead use
\styopt[nameref]{record}, which saves some extra information for
each location that's not available for the other indexing methods. 
See \longarg{merge-nameref-on} for further details.

If you additionally want to use an indexing application, such
as \idx{xindy}, you need the package option
\styopt[alsoindex]{record} and use \ics{cs.makeglossaries}
and \ics{printglossary} (or the iterative \ics{printglossaries}) as usual.
This requires a more complicated build process:
\begin{verbatim}
pdflatex myDoc
bib2gls myDoc
pdflatex myDoc
makeglossaries myDoc
pdflatex myDoc
\end{verbatim}
(The entries aren't defined until the second \LaTeX\ run, so the
indexing files required by \idx!{xindy} or \idx!{makeindex} can't be
created until then.) In this case, \bibgls\ is simply being used to
fetch the entry definitions from one or more \ext{bib} files, with
the sorting and collating performed by the other indexing
application (so the resource option list would need
\csopt[none]{sort} and \csopt[false]{save-locations}).  In general, 
it's best to avoid this hybrid method unless you have a particular
set of \idx{xindy} rules that can't be replicated with \bibgls.

\section{Logical Divisions: \field{type} vs \field{group} vs \field{parent}}
\label{sec:logicaldivisions}

If you have a document with many terms that need listing, it's
likely that you may want to divide the terms into separate blocks or
units for easier reading. There are three fields that are used
for this.
\begin{description}
\item[\field{type}] The highest division is the glossary to which
the entry belongs. The glossary must first be defined (see
\sectionref{sec:newglossary}) with an associated label used to
identify it. The title is assigned to the glossary when it is
defined or it can be overridden with the \printglossopt{title} key.
The glossary is displayed using \cs{printunsrtglossary} and the
title is placed in a sectioning command by default.

\begin{important}
\bibgls\ does not provide any means of sorting glossary types. If
you use \ics{printunsrtglossaries} the order will be according to
the order in which the glossaries were defined. You may use
\cs{printunsrtglossary} to list individual glossaries in your own
preferred order.
\end{important}

\item[\field{group}] The entries within a glossary can form groups
as a by-product of the sorting method. This must be enabled with the
\longarg{group} switch and isn't available for the sort methods
listed in \tableref{tab:sortoptionsnosort}. The group label is
stored in the \field{group} field. This is an internal field that
typically shouldn't be set in the \ext{bib} file.

You can specify your own custom groups but if you do so you must
ensure that the terms are ordered in such a way that they are
gathered according to group. This is typically done by splitting the
glossary into blocks using a separate \gls{GlsXtrLoadResources} with
the \csopt{group} option set. You control the order of the groups by
your ordering of \gls{GlsXtrLoadResources}. The group title can be
assigned using \gls{glsxtrsetgrouptitle} within the document.

\begin{important}
\bibgls\ does not sort by group title. At most it can sort by the
group label (by changing the \csopt{sort-field}) but this is usually
an indication that you actually have a hierarchical glossary and you
ought to be using the \field{parent} field instead. (Compare
\exfile{sample-textsymbols.tex} and
\exfile{sample-textsymbols2.tex}.)
\end{important}

\item[\field{parent}] An entry may have one or more sub-entries.
Most of the sort methods will produce a hierarchical ordering that
ensures that the sub-entries are listed immediately after their
parent entry. The parent entry is identified by the \field{parent}
field which should contain the parent's label.

\begin{important}
\bibgls\ sorts the parent and child entries using the same
comparator. The sort methods listed
in \tableref{tab:sortoptionsnosort} disregard the hierarchical
level, which can result in child entries becoming detached from
their parent entry. The other methods sort hierarchically using
the same comparator but take the hierarchical level into account.
\end{important}

\end{description}

Suppose you have a mixture of terms, abbreviations and symbols, then
you might want to have three glossaries that are listed in the table
of contents. In this case, you use the \field{type} field or the
\field{type} resource option. The ordering of the glossaries is
determined by the ordering of the \cs{printunsrtglossary}
commands within the document. For example:
\begin{codeenv}
\cs{printunsrtglossary}
\cs{printunsrtglossary}\oarg{\printglossopt[abbreviations]{type}}
\cs{printunsrtglossary}\oarg{\printglossopt[symbols]{type}}
\end{codeenv}

Suppose that your list of terms spans many pages and you feel it
would be helpful to the reader to split it up into letter groups
then you would need to run \bibgls\ with the \longarg{group} switch
and use a glossary style that supports letter groups for that
glossary. For example:
\begin{codeenv}
\cs{printunsrtglossary}\oarg{\printglossopt[indexgroup]{style}}
\end{codeenv}

Suppose that your list of symbols consists of pictographs, Latin characters
and Greek characters and you want them grouped together
in that order. Then you would use a separate
\gls{GlsXtrLoadResources} for each block and assign your own custom 
group. This means ensuring that each \idx{resourceset} only selects
the terms for that group. The simplest way of doing this is to
have a separate \ext{bib} file for each set. For example:
\begin{codeenv}
\gls{glsxtrsetgrouptitle}\marg{pictographs}\marg{Pictographs}
\gls{glsxtrsetgrouptitle}\marg{latinsymbols}\marg{Latin Characters}
\gls{glsxtrsetgrouptitle}\marg{greeksymbols}\marg{Greek Characters}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[generalsymbols]{src},\comment{data in generalsymbols.bib}
  \csopt[pictographs]{group},
  \csopt[symbols]{type}
}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[latinsymbols]{src},\comment{data in latinsymbols.bib}
  \csopt[latin]{group},
  \csopt[symbols]{type}
}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[greeksymbols]{src},\comment{data in greeksymbols.bib}
  \csopt[greek]{group},
  \csopt[symbols]{type}
}
\end{codeenv}

Suppose instead that you have many of these logical blocks and you
want them ordered according to the block title. In this case you
have a hierarchical glossary and you need to use the \field{parent}
field. You then need to select an appropriate \idx{glossarystyle}.

If you only want to have a single \ext{bib} file that contains all
your entries and you want to share it across multiple documents then
the most flexible approach is to use custom fields and entry types
that can be aliased according to the needs of the
\idxpl{resourceset}.

For example, the file \filefmt{entries.bib}:
\begin{codeenv}
\comment{Encoding: UTF-8}
\strut
\atentryfmt{indexplural}\marg{latin,\field{text}=\marg{Latin character}}
\atentryfmt{indexplural}\marg{greek,\field{text}=\marg{Greek character}}
\atentryfmt{indexplural}\marg{pictograph}
\strut
\atentry{symbol}\marg{fx,
  \field{name}=\marg{\cs{ensuremath}\marg{f(x)}},
  \field{description}=\marg{function of \idx{mshiftchar}x\idx{mshiftchar}},
  \fieldfmt{identifier}=\marg{latin}
}
\strut
\atentry{symbol}\marg{f\idx{aposchar}x,
  \field{name}={\cs{ensuremath}\marg{f\idx{aposchar}(x)}},
  \field{description}=\marg{derivative of \cs{gls}\marg{fx}},
  \fieldfmt{identifier}=\marg{latin}
}
\strut
\atentry{symbol}\marg{pi,
  \field{name}=\marg{\cs{ensuremath}\marg{\cs{pi}}},
  \field{description}=\marg{ratio of circumference to diameter},
  \fieldfmt{identifier}=\marg{greek}
}
\strut
\atentry{symbol}\marg{heart,
  \field{name}=\marg{\cs{ensuremath}\marg{\ics{heartsuit}}},
  \field{description}=\marg{heart},
  \fieldfmt{identifier}=\marg{pictograph}
}
\strut
\atentry{symbol}\marg{diamond,
  \field{name}=\marg{\cs{ensuremath}\marg{\ics{diamondsuit}}},
  \field{description}=\marg{diamond},
  \fieldfmt{identifier}=\marg{pictograph}
}
\strut
\atentry{abbreviation}\marg{html,
  \field{short}=\marg{html},
  \field{long}=\marg{hypertext markup language},
  \fieldfmt{identifier}=\marg{markuplanguage}
}
\strut
\atentry{abbreviation}\marg{xml,
  \field{short}=\marg{xml},
  \field{long}=\marg{extensible markup language},
  \fieldfmt{identifier}=\marg{markuplanguage}
}
\strut
\atentry{entry}\marg{duck,
  \field{name}=\marg{duck},
  \field{description}=\marg{a waterbird with webbed feet},
  \fieldfmt{identifier}=\marg{animal}
}
\strut
\atentry{entry}\marg{parrot,
  \field{name}=\marg{parrot},
  \field{description}=\marg{mainly tropical bird with bright plumage},
  \fieldfmt{identifier}=\marg{animal}
}
\end{codeenv}
This has a custom field \fieldfmt{identifier}. This will be ignored
by \bibgls\ unless defined or aliased in the document.

Here's an example document that creates three glossary types (the
default \code{main} glossary and the glossaries created with the
\styopt{abbreviations} and \styopt{symbols} options). They are
listed in the order of \cs{printunsrtglossary} and their titles are
added to the table of contents.

The custom \fieldfmt{identifier} fields are ignored for the
main and abbreviation glossaries, but they are aliased for the
symbols to the \field{group} field. Since I've split the symbols
glossary into blocks with each block only containing entries that
have the same \field{group} value, this isn't a problem. It also
won't trigger a warning with \longarg{warn-non-bib-fields} as it's
being aliased rather than set in the \ext{bib} file. The blocks
appear in the same order as the corresponding \gls{GlsXtrLoadResources} 
commands. The title for each block is provided in the document 
using \gls{glsxtrsetgrouptitle}.
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{\styopt{record},\styopt{abbreviations},\styopt{symbols}}\marg{glossaries-extra}
\strut
\cs{renewcommand}\marg{\gls{GlsXtrDefaultResourceOptions}}\marg{
 \csopt[all]{selection},\csopt[entries]{src},\csopt[false]{save-locations}}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[main]{type},\csopt[entrytype=entry]{match}}
\gls{GlsXtrLoadResources}\oarg{\csopt[abbreviations]{type},
 \csopt[entrytype=abbreviation]{match}}
\strut
\gls{glsxtrsetgrouptitle}\marg{pictograph}\marg{Pictographs}
\gls{GlsXtrLoadResources}\oarg{\csopt[symbols]{type},
 \csopt[identifier=\field{group}]{field-aliases},
 \csopt[\field{group}=pictograph]{match}}
\strut
\gls{glsxtrsetgrouptitle}\marg{latin}\marg{Latin Characters}
\gls{GlsXtrLoadResources}\oarg{\csopt[symbols]{type},
 \csopt[identifier=\field{group}]{field-aliases},
 \csopt[\field{group}=latin]{match}}
\strut
\gls{glsxtrsetgrouptitle}\marg{greek}\marg{Greek Characters}
\gls{GlsXtrLoadResources}\oarg{\csopt[symbols]{type},
 \csopt[identifier=\field{group}]{field-aliases},
 \csopt[\field{group}=greek]{match}}
\strut
\cmd{begin}\marg{document}
\cs{tableofcontents}
\cs{printunsrtglossary}\oarg{\printglossopt[abbreviations]{type}}
\cs{printunsrtglossary}
\cs{printunsrtglossary}\oarg{\printglossopt[symbols]{type},\printglossopt[\glostyle{treegroup}]{style}}
\cmd{end}\marg{document}
\end{codeenv}

In the above example document, the symbols list is divided into
three groups, listed in the order: Pictographs, Latin characters and
Greek characters. If you want these titles ordered alphabetically
then you need a hierarchical structure instead. This can be obtained
by aliasing the custom \fieldfmt{identifier} field to
\field{parent}:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{\styopt{record},\styopt[topic]{stylemods},\styopt{abbreviations},\styopt{symbols}}\marg{glossaries-extra}
\strut
\cs{renewcommand}\marg{\gls{GlsXtrDefaultResourceOptions}}\marg{\comment{}
 \csopt[all]{selection},\csopt[entries]{src},\csopt[false]{save-locations}}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[main]{type},\csopt[entrytype=entry]{match}}
\gls{GlsXtrLoadResources}\oarg{\csopt[abbreviations]{type},
 \csopt[entrytype=abbreviation]{match}}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[symbols]{type},
 \csopt[identifier=\field{parent}]{field-aliases},
 \csopt[entrytype=symbol,entrytype=indexplural]{match}}
\strut
\cmd{begin}\marg{document}
\cs{tableofcontents}
\cs{printunsrtglossary}\oarg{\printglossopt[abbreviations]{type}}
\cs{printunsrtglossary}
\cs{printunsrtglossary}\oarg{\printglossopt[symbols]{type},\printglossopt[\glostyle{topic}]{style}}
\cmd{end}\marg{document}
\end{codeenv}
The style used for the symbols list is now \glostyle{topic} rather
than \glostyle{treegroup}. This results in a slightly different
appearance. You can select the most appropriate style according to
your needs (see the gallery of predefined
styles~\cite{glossarystylesgallery}). The topic ordering is now:
Greek characters, Latin characters and Pictographs.

\section{Defining a New Glossary}
\label{sec:newglossary}

Some of the examples in this manual use \ics{newglossary*} to define
a new glossary type and some use \ics{newignoredglossary} or
\ics{newignoredglossary*}. Why the starred forms and why define an
\idx{ignoredglossary}?

The base \styfmt{glossaries} package was originally designed to work
with \idx!{makeindex}. Support for \idx!{xindy} was later added, but
both require three files per glossary type: the transcript file
(created by the indexing application), the file written by \LaTeX\
(and input by the indexing application) and the file input by
\LaTeX\ (and written by the indexing application). So when a new glossary
is defined with \ics{newglossary}, this not only defines internal
control sequences that store the list of entry labels associated
with that glossary, the title and the entry format but also has to
define internal control sequences that store the three file
extensions. The starred form \cs{newglossary*} is just a shortcut
that forms the extensions from the glossary label. For the purposes
of \bibgls, this is simpler than the unstarred version since the
extensions are now irrelevant as they are only applicable to
\idx!{makeindex} and \idx!{xindy}. (Unless, of course, you are using
a hybrid method with \styopt[alsoindex]{record}.)

Since some users wanted the ability to define entries that were
common enough to not be worth including in any glossary lists, the
concept of an \idx{ignoredglossary} was introduced, defined with
\cs{newignoredglossary}. This only requires an internal control
sequence to store the list of entry labels associated with that
glossary\footnote{All entries must be assigned to a glossary. If you
don't use the \field{type} field the default is used.}\ and the
associated internal command that governs the way that
commands like \cs{gls} are displayed for that glossary type. 
Since this type of glossary has no associated files, it can't be
used with \cs{printglossary} and therefore isn't included in the
list of glossary labels that's iterated over by commands like
\cs{printglossaries}. Since there's no glossary list (and therefore
no targets), \cs{newignoredglossary} additionally disables hyperlinks for that
glossary type, but it doesn't disable indexing.  The indexing macro
is still called, but because there's no associated file to write to,
it has no effect. With \bibgls, the indexing is written to the
\ext{aux} file and so does have an effect.

Although \idxpl{ignoredglossary} can't be used with
\cs{printglossary}, they can be used with \cs{printunsrtglossary},
which is designed to work without any indexing, but you need to
explicitly set the title in the optional argument to override the
default. \Idxpl{ignoredglossary} still can't be used in
\cs{printunsrtglossaries}, since they're not included in the list
that this command iterates over.

So \cs{newignoredglossary} (or \cs{provideignoredglossary}) is
useful with \bibgls\ if you're happy to use \cs{printunsrtglossary}
with the \printglossopt{type} and \printglossopt{title} options as it reduces the overall
number of internal control sequences.  \Idxpl{ignoredglossary} are
also useful for stand-alone definitions (\ics{glsxtrglossentry}) or
with \ics{printunsrtinnerglossary} as no title is required in those
cases (see \exfile{sample-nested.tex} for an example). 

Since there is now the possibility of targets (created within \cs{printunsrtglossary} or
\cs{printunsrtinnerglossary} or \cs{glsxtrglossentry}), it's convenient to
have an \idx{ignoredglossary} that doesn't suppress the hyperlinks,
which can be obtained with the starred form \ics{newignoredglossary*}
provided by \styfmt{glossaries-extra} (or \cs{provideignoredglossary*}).

Some resource options, such as \csopt{master}, \csopt{secondary} and
\csopt{trigger-type}, need to ensure that a
required glossary is defined. In this case, \bibgls\ uses
\ics{provideignoredglossary*} in the \iext{glstex} file 
even if \longarg{no-provide-glossaries} is set. If you
haven't already defined that glossary in the document with \cs{newglossary*},
you'll need to set the title in the optional argument of
\cs{printunsrtglossary} if you don't want the default. The glossary
won't be defined on the first run (if the definition is only
provided in the \ext{glstex} file) but \cs{printunsrtglossary} will
just give a warning if the type is undefined so it won't interrupt
the document build.

If you want \bibgls\ to automatically provide unknown glossaries for
all entries that have the \field{type} field set (unrelated to the
\csopt{master}, \csopt{secondary} and \csopt{trigger-type} options)
then use the \longarg{provide-glossaries} switch.

The base \sty{glossaries} package provides a command that can be
used to test the existence of a glossary:
\begin{codeenv}
\ics{ifglossaryexists}\margm{label}\margm{true}\margm{false}
\end{codeenv}
The unstarred version considers ignored glossaries as non-existent
(and so will do \meta{false} for an ignored glossary). As from
v4.46, this command now has a starred version
\ics{ifglossaryexists*} that considers ignored glossaries as
existing (and so will do \meta{true} for an ignored glossary). In
the event that you have an older version of \sty{glossaries}, the
\sty{glossaries-extra} package (v1.44+) will provide the starred
form if it hasn't been defined. (In general, it's best to have
up-to-date versions of both \sty{glossaries} and
\sty{glossaries-extra}.)

\section{Resource Sets}
\label{sec:resourcesets}

Each instance of \gls{glsxtrresourcefile} or
\gls{GlsXtrLoadResources} in the document represents a \idx{resourceset}. Each
\idx{resourceset} has one or more associated \iext{bib} files
that provides the data for that set. Command line switches
(\sectionref{sec:switches}) are applied to all \idxpl{resourceset}.
Resource options (\sectionref{sec:resourceopts}) are only applied to
that specific \idx{resourceset}. Each \idx{resourceset} is processed
in stages:
\begin{description}

\item[Stage 1 (Initialisation)] Occurs after the \ext{aux} file has
been read, this stage parses the resource option list and ensures
options are valid and don't cause a conflict. The transcript will
show the message
\begin{alltt}
Initialising resource \meta{resource-name}
\end{alltt}
at this point.

\item[Stage 2 (Parsing)] All the \iext{bib} files associated with
the \idx{resourceset} are parsed. Entry aliases (identified by
\csopt{entry-type-aliases}) are performed. The
\hyperref[sec:multientry]{multi-entry types}, such as
\atentry{bibtexentry} and \atentry{progenitor}, spawn their
associated primary entries. Preamble information (provided by
\atentry{preamble}) is saved but is not interpreted at this stage.
The transcript will show the message
\begin{alltt}
Parsing bib files for resource \meta{resource-name}
\end{alltt}
at this point.

\item[Stage 3 (Processing Entries)] 
The transcript will show the message
\begin{alltt}
Processing resource \meta{resource-name}
\end{alltt}
at this point.
For each entry that was found in
the corresponding set of \ext{bib} files: 
\begin{itemize}
\item Records are transferred to aliases if required (\csopt{alias-loc}).
\item Field checks and modifications are performed:
  \begin{itemize}
  \item field aliases are performed (\csopt{field-aliases}); 

  \item ignored fields (identified by \csopt{ignore-fields}) are
  removed;

  \item case-changes (for example, \csopt{short-case-change}) are
  performed, except for the \field{name} field and fields identified
  with \csopt{field-case-change};

  \item suffixes are appended if required (for example, with
  \csopt{short-plural-suffix});

  \item field replications are made (\csopt{replicate-fields}),
    and any of the above case-change or suffixes required
    on the replicated fields are performed;

  \item the \field{group} field is assigned if \csopt[\meta{label}]{group} 
    is set;

  \item any variables (identified by \atentry{string}) are expanded
  (if not already done in any of the previous steps);

  \item any fields that have been identified by
  \csopt{bibtex-contributor-fields} are converted;

  \item any fields that have been identified with
  \csopt{encapsulate-fields} are converted;

  \item any fields that have been identified with
  \csopt{encapsulate-fields*} are converted;

  \item any fields that must be converted into a label form
  (\csopt{labelify} or \csopt{labelify-list}) are processed;

  \item any fields identified by \csopt{dependency-fields} are parsed for
  dependent entries;

  \item any fields whose value must be a label are interpreted
  if \csopt{interpret-label-fields} is set;

  \item the \field{parent} field is adjusted according to the label prefix 
  settings (\csopt{label-prefix} etc);

  \item \ics{makefirstuc} protection is applied according to 
  \longarg{mfirstuc-protection} and \longarg{mfirstuc-math-protection};

  \item fields are parsed for commands like \cs{gls} or
  \cs{glshyperlink} and also checked for nested links if
  \longarg{nested-link-check} is set;

  \item the \field{description} field is adjusted according to
  \csopt{strip-trailing-nopost};

  \item end punctuation is checked according to
  \csopt{check-end-punctuation};

  \item \field{name} case-change is performed if
  \csopt{name-case-change} is set;

  \item if \csopt[true]{copy-alias-to-see} the \field{alias} is copied 
  to the \field{see} field;

  \item general field case changes identified by
  \csopt{field-case-change} are performed;

  \item any fields that have been identified with
  \csopt{interpret-fields} are replaced with their interpreted
  values;

  \item check for \field{nonumberlist}.
  \end{itemize}
\item The dual version (if appropriate) is created.
\item Records are added to the entry's location list (or transferred
to the dual\slash primary according to \csopt{combine-dual-locations}).
\item The \field{type}, \field{category} and \field{counter} fields
are set according to \csopt{type}, \csopt{dual-type},
\csopt{category}, \csopt{dual-category}, \csopt{counter} and
\csopt{dual-counter}.
\item Filtering is applied (according to options like \csopt{match}
but not \csopt{selection} or \csopt{limit}).
\item Required fields are checked for existence.
\item Dependencies are registered (if
\csopt[recorded and deps]{selection} or \csopt[recorded and deps and
see]{selection}). 
\item Any fields that have been identified by
\csopt{date-time-fields}, \csopt{date-fields} or \csopt{time-fields}
are converted.
\end{itemize}
If \csopt[recorded and deps and see]{selection} then any recorded
entries that have been cross-referenced by an unrecorded entry, will
register a dependency with the unrecorded entry. Finally,
supplemental records are added to entries.

\item[Stage 4 (Selection, Sorting, Writing)] Entries are selected
from the list according to the \csopt{selection} setting, sorting is
performed (if required), truncation is applied (if \csopt{limit} is
set) and the \iext{glstex} file is written.
The transcript will show the message
\begin{alltt}
Selecting entries for resource \meta{resource-name}
\end{alltt}
or (if \csopt{master})
\begin{alltt}
Processing master \meta{resource-name}
\end{alltt}
at this point.

\end{description}

Parent entries must always be in the same \idx{resourceset} as their
child entries. (They may be defined in different \ext{bib} files as
long as all those \ext{bib} files are listed in the same \csopt{src}.) 
Other forms of dependencies may be in a
different \idx{resourceset} under certain circumstances. These types
of dependencies are instances of commands such as \cs{gls} being
found (for example, in the \field{description} field), or the
cross-reference fields (\field{see}, \field{seealso} or
\field{alias} or fields identified with \csopt{dependency-fields}) 
in recorded entries that reference unrecorded entries.

The \qt{cross-referenced by} dependencies enabled with
\csopt[recorded and deps and see]{selection} (where an unrecorded
entry references a recorded entry through the cross-reference
fields) \emph{aren't supported} across \idxpl{resourceset} (even with
\longarg{force-cross-resource-refs}).

A \idx{crossresourceref} is a reference from a recorded entry provided in one
\idx{resourceset} to an unrecorded entry in another \idx{resourceset}.
Since
the contents of each \idx{resourceset}['s] preamble must be
processed before fields can be interpreted and one
\idx{resourceset}['s] preamble may contain definitions that override
another, \idxpl{crossresourceref} can't be supported if fields
containing cross-referencing information need to be interpreted.

The \idx{crossresourceref} mode determines whether or not \bibgls\
can support \idxpl{crossresourceref}. If enabled, the message
\begin{verbatim}
Cross-resource references allowed.
\end{verbatim}
will be written to the transcript otherwise the message is
\begin{verbatim}
Cross-resource references disabled.
\end{verbatim}
The mode can only be enabled if the following condition is satisfied:
\begin{itemize}
\item the interpreter is off (\longarg{no-interpret}), or
\item every \idx{resourceset} either doesn't have a preamble
(\atentry{preamble}) or has \csopt[false]{interpret-preamble} set.
\end{itemize}
If you know the preamble contents won't cause a problem, you
can force the \idxpl{crossresourceref} mode on with 
\longarg{force-cross-resource-refs}.

If you don't use either \csopt[recorded and
deps]{selection} or \csopt[recorded and deps and see]{selection}
then the dependencies aren't picked up for that \idx{resourceset}
(and so can't be cross-referenced from another \idx{resourceset}).

Trails don't work with \idxpl{crossresourceref}. For example, if
entry $A$ has been recorded and depends on entry $B$ that hasn't been
recorded, then $B$ can be picked up from a different
\idx{resourceset}, but if $A$ and $B$ are in the same
\idx{resourceset} and $B$ is dependent on $C$ which is in a
different \idx{resourceset} then $C$ won't be picked up if it hasn't
been recorded because $B$ hasn't been recorded and is in a different
\idx{resourceset}.

If the \idx{crossresourceref} mode is enabled then stage~3 and
stage~4 are processed in separate \idxpl{loop}, otherwise they are
processed in the same \idx{loop}.

\section{Indexing}

The dual index entries such as \atentry{dualindexentry} (described in
\sectionref{sec:dualentry}) are designed to provide a way of
including an entry in a glossary (with a description) and also
include the term (without the description) in an index.  Additional
terms that should only appear in the index can be defined with
\atentry{index}.  (See, for example, the \exfile{sample-multi1.tex}
and \exfile{sample-multi2.tex} sample files.)

Although \bibgls\ is designed to create indexes as well as glossary
lists using the same interface (\cs{gls} etc), it is
possible to have a mixture of \bibgls\ and \ics{index}. For example:
\begin{codeenv}
\cmd{documentclass}\marg{report}
\strut
\cmd{usepackage}\marg{makeidx}
\cmd{usepackage}\oarg{\styopt{record}}\marg{glossaries-extra}
\strut
\cmd{makeindex}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\strut
\ics{glssetcategoryattribute}\marg{general}\marg{dualindex}\marg{true}
\cs{glssetcategoryattribute}\marg{symbol}\marg{dualindex}\marg{true}
\cs{glssetcategoryattribute}\marg{abbreviation}\marg{dualindex}\marg{true}
\strut
\cs{glssetcategoryattribute}\marg{general}\marg{indexname}\marg{hyperbf}
\cs{glssetcategoryattribute}\marg{symbol}\marg{indexname}\marg{hyperbf}
\cs{glssetcategoryattribute}\marg{abbreviation}\marg{indexname}\marg{hyperbf}
\strut
\cmd{begin}\marg{document}
\cmd{chapter}\marg{Example}
\cs{gls}\marg{bird}, \cs{gls}\marg{html}, \idx{mshiftchar}\cs{gls}\marg{v}\idx{mshiftchar} and \cs{glspl}\marg{goose}.
\strut
\cs{printunsrtglossaries}
\cmd{printindex}
\cmd{end}\marg{document}
\end{codeenv}
If the document is called \filefmt{myDoc.tex} then the document
build is:
\begin{verbatim}
pdflatex myDoc
bib2gls myDoc
pdflatex myDoc
makeindex myDoc.idx
pdflatex myDoc
\end{verbatim}
This requires an additional \LaTeX\ call between \bibgls\ and
\idx{makeindex} since the entries must be defined before they can be
indexed (and they can't be defined until \bibgls\ creates the
associated \ext{glstex} files).

Note that this method will use the \field{sort} value obtained by \bibgls\
as the \meta{sort} part within
\code{\cs{index}\marg{\meta{sort}@\meta{actual}}}.
Be careful if you use \idx{makeindex} as this can result in Unicode
characters appearing in the sort value, which \idx{makeindex}
doesn't support.
The \meta{actual} part is given by
\code{\ics{glsentryname}\margm{label}}.
(You can change the \meta{sort} and \meta{actual} parts by
redefining \ics{glsxtrautoindexassignsort} and
\ics{glsxtrautoindexentry}. See the \sty{glossaries-extra} manual
for further details.)

\section{Security}

\TeX\ Live come with security settings
\texmfcnf{openinany} and \texmfcnf{openoutany} that, respectively,
govern read and write file access (in addition to the operating
system's file permissions). \bibgls\ uses \idx{kpsewhich} to
determine these values and honours them. MikTeX doesn't use these
settings, so if these values are unset, \bibgls\ will default to
\code{a} (any) for \texmfcnf{openinany} and \code{p} (paranoid) for
\texmfcnf{openoutany}.

The only external processes that are run by \bibgls\ are calls to
\idx{kpsewhich} to check the security settings and locate files on
\TeX's path. These are started with Java's \code{ProcessBuilder}
class so there should be no issues with spaces or shell special characters
in the argument.  The \longarg{debug} switch will write the process
call in the transcript file and will delimit the argument in the log
with single quote characters for convenience, but the process isn't
actually called in that way.

\bibgls\ creates files with the extension \ext{glstex}, which are
input by \gls{glsxtrresourcefile} (and therefore by the shortcut
\gls{GlsXtrLoadResources}). This extension is fixed and is imposed
by both \bibgls\ and \gls{glsxtrresourcefile}. \bibgls\ also creates
a transcript file with the default extension \ext{glg}. This may be
overridden by the \longarg{log-file} switch, but \bibgls\ always
forbids write access to any file with the following extensions:
\extfmt{tex}, \extfmt{ltx}, \extfmt{sty}, \extfmt{cls},
\extfmt{bib}, \extfmt{dtx}, \extfmt{ins}, \extfmt{def} and
\extfmt{ldf}.

\section{Localisation}
\label{sec:lang.xml}

The messages produced by \bibgls\ are fetched from a resource file
called \metafilefmt{bib2gls-}{lang}{.xml}, where \meta{lang} is a
valid \idx{IETF} language tag.

The appropriate file is searched for in the following order, where
\meta{locale} is the operating system's locale or the value supplied
by the \longarg{locale} switch:
\begin{enumerate}
\item \meta{lang} exactly matches \meta{locale}.
For example, my locale is \code{en-GB}, so \bibgls\ will first search
for \filefmt{bib2gls-en-GB.xml}. This file doesn't exist, so it will
try again.

\item If \meta{locale} has an associated script, the
next try is with \meta{lang} set to \meta{lang
code}\code{-}\meta{script} where \meta{lang code} is the two
letter ISO language code and \meta{script} is the script code.
For example, if \meta{locale} is \code{sr-RS-Latn}
then \bibgls\ will search for \filefmt{bib2gls-sr-Latn.xml} if
\filefmt{bib2gls-sr-RS-Latn.xml} doesn't exist.

\item The final attempt is with \meta{lang} set to just the two
letter ISO language code. For example, \filefmt{bib2gls-sr.xml}.
\end{enumerate}

If there is no match, \bibgls\ will fallback on the English resource file
\file{bib2gls-en.xml}.
(Currently only \file{bib2gls-en.xml} exists as my language skills aren't up
to translating it. Any volunteers who want to provide other language
resource files would be much appreciated.)

Note that if you use the \csopt[true]{loc-prefix} option, the
textual labels (\qt{Page} and \qt{Pages} in English) will be taken
from the resource file. In the event that the loaded resource file
doesn't match the document language, you will have to manually set
the correct translation (in English, this would be
\csopt[Page\dcomma Pages]{loc-prefix}). The default definition of
\gls!{bibglspassim} is also obtained from the resource file.

\section{Conditional Document Build}

If you are using a document build method that tries to determine
whether or not \bibgls\ should be run, you can find the information
by searching the \iext{aux} file for instances of
\nosecdef{glsxtr@resource}
Each instance corresponds to an instance of \gls{glsxtrresourcefile}
where \meta{filename} is the base name of the \iext{glstex} file
that \bibgls\ needs to create for this resource set. If the
\meta{options} part is missing the \csopt{src} option, then
\meta{filename} also indicates the base name for the \iext{bib} file.

So the simplest check to determine if \bibgls\ needs to be run is to test
if the \iext{aux} file contains \gls{glsxtr@resource}. For
example, with \idx{arara} version 4.0:
\begin{codeenv}
\% arara: bib2gls if found("aux", "glsxtr@resource")
\end{codeenv}
A sophisticated method could check if
\meta{filename}\ext{glstex} is missing or is older than the document
\ext{tex} file for each instance of \gls{glsxtr@resource} found in
the \ext{aux} file.

It might also be possible, although far more complex, to
parse the \meta{options} part in each instance of \gls{glsxtr@resource}
for \csopt{src} and determine if the corresponding \ext{bib} file or
files are newer than the \ext{tex} file.

It's not possible to determine if the location lists require
updating, just as it's not possible to do this for the
\idx{toc}, list of figures, list of tables etc. (Or, if it could
be implemented, the required code would make the document build far
more complicated.)

In general, the basic algorithm is:
\begin{enumerate}
\item Run \LaTeX\ (or PDF\LaTeX\ etc).

\item If \gls{glsxtr@resource} is found in the \ext{aux} file then:

   \begin{enumerate}
   \item run \bibgls;
   \item run \LaTeX\ (or PDF\LaTeX\ etc).
   \end{enumerate}

\item If \ics{@istfilename} is found in the \ext{aux} file then:

   \begin{enumerate}
   \item run \idx{makeglossaries} (or \idx{makeglossaries-lite});
   \item run \LaTeX\ (or PDF\LaTeX\ etc).
   \end{enumerate}

\end{enumerate}
This allows for the \styopt[alsoindex]{record} package option.  See
also
\href{https://www.dickimaw-books.com/latex/buildglossaries}{\qt{Incorporating
\idx[hyper=false,noindex]{makeglossaries} or
\idx[hyper=false,noindex]{makeglossaries-lite} or \bibgls\ into
the document build}}~\cite{buildglossaries}.

\section{Manual Installation}

If you are unable to install \bibgls\ through your \TeX\ package
manager, you can install manually using the instructions below.
Replace \meta{TEXMF} with the path to your local or home TEXMF tree 
(for example, \filefmt{\glstildechar/texmf}).

Copy the files provided to the following locations:
\begin{itemize}
\item \meta{TEXMF}\filefmt{/scripts/bib2gls/bib2gls.jar}
(Java application.)
\item \meta{TEXMF}\filefmt{/scripts/bib2gls/convertgls2bib.jar}
(Java application.)
\item \meta{TEXMF}\filefmt{/scripts/bib2gls/texparserlib.jar}
(Java library.)
\item \meta{TEXMF}\filefmt{/scripts/bib2gls/resources/bib2gls-en.xml}
(English resource file.)
\item \meta{TEXMF}\filefmt{/doc/support/bib2gls/bib2gls.pdf}
(This document.)
\item \meta{TEXMF}\filefmt{/doc/support/bib2gls/bib2gls-begin.pdf}
(Introductory guide.)
\end{itemize}
If you use the Unix \appfmt{man} command, copy the
\filefmt{bib2gls.1} and \filefmt{convertgls2bib.1} files to the
appropriate location.

If you are using a Unix-like system, there are also bash scripts
provided called \file{bib2gls.sh} and \file{convertgls2bib.sh}. 
Either copy then directly to somewhere on your path without the \iext{sh} 
extension, for example:
\begin{verbatim}
cp bib2gls.sh ~/bin/bib2gls
cp convertgls2bib.sh ~/bin/convertgls2bib
\end{verbatim}
or copy the files to
\meta{TEXMF}\filefmt{/scripts/bib2gls/} and create a
symbolic link to them called just \filefmt{bib2gls} and
\filefmt{convertgls2bib} from somewhere on
your path, for example:
\begin{verbatim}
cp bib2gls.sh ~/texmf/scripts/bib2gls/
cp convertgls2bib.sh ~/texmf/scripts/bib2gls/
cd ~/bin
ln -s ~/texmf/scripts/bib2gls/bib2gls.sh bib2gls
ln -s ~/texmf/scripts/bib2gls/convertgls2bib.sh convertgls2bib
\end{verbatim}
The \file{texparserlib.jar} file isn't an application but is a
library used by both \file{bib2gls.jar} and
\file{convertgls2bib.jar}, and so needs to be in the same class path. 
(The library is in a separate
\href{https://github.com/nlct/texparser}{GitHub
repository}~\cite{texparser} as it's
also used by some of my other applications.)

Windows users can create a \iext{bat} file that works in a
similar way to the bash scripts. To do this, create a file called
\file{bib2gls.bat} that contains the following:
\begin{verbatim}
@ECHO OFF
FOR /F "tokens=*" %%I IN ('kpsewhich --progname=bib2gls --format=texmfscripts
bib2gls.jar') DO SET JARPATH=%%I
java -Djava.locale.providers=CLDR,JRE -jar "%JARPATH%" %*
\end{verbatim}
Save this file to somewhere on your system's path.
(Similarly for \filefmt{convertgls2bib}.) Note that \TeX\
distributions for Windows usually convert \iext{jar} files to
executables.

You may need to refresh \TeX's database to ensure that
\idx{kpsewhich} can find the \iext{jar} files.
To test that the application has been successfully installed, open a
command prompt or terminal and run the following command:
\begin{verbatim}
bib2gls --version
convertgls2bib --version
\end{verbatim}
This should display the version information for both applications.

\chapter{\texorpdfstring{\TeX}{TeX}\ Parser Library}
\label{sec:texparserlib}

The \bibgls\ application requires the \TeX\ Parser Library
\file{texparserlib.jar}\footnote{\url{https://github.com/nlct/texparser}}
which is used to parse the \iext{aux} and \iext{bib} files.

With the \longarg{interpret} switch on (default), this library is
also used to interpret the sort value when it contains a backslash
\idx{escchar} or a tilde \idx{nbspchar} or a dollar symbol 
\idx{mshiftchar} or braces \idx{bgroupchar} \idx{egroupchar}
(and when the \csopt{sort} option is not
\optfmt{unsrt} or \optfmt{none} or \optfmt{use}).\footnote{The other
special characters are omitted from the check: the comment symbol
\idx{commentchar} is best avoided in field values, the subscript and
superscript characters \idx{sbchar} and \idx{spchar} should
either be encapsulated by \idx{mshiftchar} or by \cs{ensuremath}, which
will be picked up by the check for \idx{mshiftchar} or
\idx{escchar}, and the other special characters would
indicate something too complex for the interpreter to handle.}

The other cases that the interpreter is used for are:
\begin{itemize}
\item when \csopt{set-widest} is used to determine the width of
the \field{name} field;
\item if \csopt{labelify} or \csopt{labelify-list} are set the
identified field values are first interpreted (if they contain
\idx{escchar} \idx{bgroupchar} \idx{egroupchar} 
\idx{nbspchar} or \idx{mshiftchar}) before being converted to labels;
\item if \csopt[true]{interpret-label-fields} is set and the
\field{parent}, \field{category}, \field{type}, \field{group},
\field{seealso} or \field{alias} fields contain \idx{escchar}
or \idx{bgroupchar} or \idx{egroupchar} the interpreter is used 
since these fields must be just a label
(other special characters aren't checked as they won't expand to
characters allowed in a label).
\end{itemize}

Information in the \ext{aux} file is parsed for specific commands
but the arguments of those commands are not interpreted so, for
example, UTF-8 characters that occur in any resource options will
need to be detokenized when using \sty{inputenc} to prevent
expansion when they are written to the \ext{aux} file. (In some
options, such as \csopt{sort-rule}, you can use \cs{glshex}\meta{hex}
syntax to specify a UTF-8 character.)

The \longarg{no-interpret} switch will turn off the interpreter, but
the library will still be used to parse the \ext{aux} and \ext{bib}
files. Note that the \field{see} field doesn't use the interpreter
with \csopt[true]{interpret-label-fields} as
it may legitimately contain \LaTeX\ code in the optional tag part
(such as \cs{seealsoname} or \cs{alsoname}).

The parser has a different concept of expansion to \TeX\
and will expand some things that aren't expanded by \LaTeX\
(such as \ics{MakeUppercase} and \ics{char}) and won't expand other commands
that would be expanded by \LaTeX\ (such as commands defined
in terms of complicated internals).

If you get a \idx{StackOverflowError} while a field is being
interpreted (with a long stack trace that contains repeated file names and
line numbers) then it's likely you have an infinite \idx{loop}. For
example, this can be triggered if a field contains \csfmt{foo} that
has been defined as:
\begin{codeenv}
\cs{def}\cmd{foo}\marg{\cmd{foo}}
\end{codeenv}
This will obviously also cause an error in the \LaTeX\ document as
well (unless the document has a different definition that doesn't
have this unbounded recursion).

The \file{texparserlib.jar} library is not a \TeX\ engine and there
are plenty of situations where it doesn't work. In particular, in
this case it's being used in a fragmented context without knowing
most of the packages used by the document or any custom commands or
environments provided within the document.

\bibgls\ can detect from the log file a small
number of packages that the parser recognises. Note that in
some cases there's only very limited support. For example, 
\isty{siunitx}'s \ics{si} command is recognised but other
commands from that package aren't. See
\longargpageref{list-known-packages} for further details.

Since the parser doesn't have a full set of commands available
within the \LaTeX\ document, when it encounters \ics{renewcommand} it
won't check if the command is undefined. If the command isn't
defined, it will simply behave like \cs{newcommand}. Whereas
with \ics{providecommand} the parser will only define the command
if it's unrecognised.

The interpreter has its own internal implementation of the
glossary-related commands listed in \tableref{tab:bibglsdefs}. These
may be overridden by custom packages provided with the
\longarg{custom-packages} switch. Note that commands that reference
an entry, such as \cs{glsentryname}, aren't guaranteed to work
across \idxpl{resourceset} and will only be able to look up field
values that are known to \bibgls. (For example, the \field{name}
field for abbreviations is typically set by the associated
abbreviation style, which isn't available to \bibgls.)

\begin{table}[htbp]
\caption{Glossary-Related Commands Implemented by the \bibgls\ Interpreter}
\label{tab:bibglsdefs}
\setlength{\tabcolsep}{4pt}%
\centering
\tablelistcs{3}{glsxtrprovidecommand,bibglsuppercase,bibglslowercase,bibglsfirstuc,%
bibglstitlecase,GlsXtrEnableInitialTagging,bibglscontributorlist,%
bibglshyperlink,glshyperlink,bibglscontributor,bibglsdatetime,%
bibglsdate,bibglstime,glsentryname,glsentrytext,glsentryshort,%
glsentrylong,glsentryfirst,glsentrysymbol,glsentryuseri,%
glsentryuserii,glsentryuseriii,glsentryuseriv,glsentryuserv,%
glsentryuservi,glsentryplural,glsentryfirstplural,glsentryshortpl,%
glsentrylongpl,glsentrysymbolplural,Glsentryname,Glsentrytext,%
Glsentryshort,Glsentrylong,Glsentryfirst,Glsentrysymbol,%
Glsentryuseri,Glsentryuserii,Glsentryuseriii,Glsentryuseriv,%
Glsentryuserv,Glsentryuservi,Glsentryplural,Glsentryfirstplural,Glsentryshortpl,%
Glsentrylongpl,Glsentrysymbolplural,bibglshashchar,bibglsunderscorechar,%
bibglsdollarchar,bibglsampersandchar,bibglscircumchar,glsbackslash,glstildechar,glsopenbrace,glsclosebrace,glspercentchar,%
glsxtrusefield,Glsxtrusefield,GLSxtrusefield,glsentrytitlecase,%
glsxtrhiernamesep,glsxtrhiername,Glsxtrhiername,GlsXtrhiername,%
GLSxtrhiername,GLSXTRhiername}
\end{table}

If a command isn't recognised, you can provide it in the
\atentry{preamble} and use \ics{char} to map a symbol to the most
appropriate Unicode character. For example, suppose your document
loads a package that provides symbols for use on maps, such as
\csfmt{Harbour}, \csfmt{Battlefield} and \csfmt{Stadium}, then you
can provide versions of these commands just for \bibgls's
use:\footnote{These commands won't work with PDF\LaTeX, as the
\cs{char} values are too large, but they're fine for \bibgls.}
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cmd{Harbour}}\marg{\charhex{2693}}
\cs{providecommand}\marg{\cmd{Battlefield}}\marg{\charhex{2694}}
\cs{providecommand}\marg{\cmd{Stadium}}\marg{\charhex{26BD}}}}
\end{codeenv}
Since these use \cs{providecommand}, they won't overwrite the
document's version (provided these commands have been defined before
\gls!{GlsXtrLoadResources}). Alternatively, you can instruct \bibgls\
to not write the \atentry{preamble} contents to the resource file
using \csopt[false]{write-preamble}. Now you can either sort these
symbols by
their Unicode values (\csopt[letter-case]{sort}) or provide a custom
rule that recognises these Unicode characters (for example,
\csopt[custom]{sort},
\csopt[\cs{glshex}2694 \string< \cs{glshex}2693 \string< \cs{glshex}26BD]{sort-rule}).

\TeX\ syntax can be quite complicated and, in some cases, far too
complicated for simple \idxpl!{regex}. The \TeX\ parser library performs
better than a simple pattern match, and that's the purpose of
\file{texparserlib.jar} and why it's used by \bibgls\ (and by
\idx!{convertgls2bib}). When the \longarg{debug} mode is on, any
warnings or errors triggered by the interpreter will be written to
the transcript prefixed with \code{texparserlib:} (the results of
the conversions will be included in the transcript as informational
messages prefixed with \code{texparserlib:} even with
\longarg{no-debug}).

For example, suppose the \ext{bib} file includes:
\begin{codeenv}
\atentry{preamble}\marg{
\qtdelim{\cs{providecommand}\marg{\cmd{mtx}}[1]\marg{\cs{boldsymbol}\marg{\idx{param}1}}
\cs{providecommand}\marg{\cmd{set}}[1]\marg{\ics{mathcal}\marg{\idx{param}1}}
\cs{providecommand}\marg{\cmd{card}}[1]\marg{|\cmd{set}\marg{\idx{param}1}|}
\cs{providecommand}\marg{\cmd{imaginary}}\marg{i}}}
\strut
\atentry{entry}\marg{M,
  \field{name}=\marg{\marg{}\idx{mshiftchar}\cmd{mtx}\marg{M}\idx{mshiftchar}},
  \field{text}=\marg{\cmd{mtx}\marg{M}},
  \field{description}=\marg{a matrix}
}
\strut
\atentry{entry}\marg{v,
  \field{name}=\marg{\marg{}\idx{mshiftchar}\cs{vec}\marg{v}\idx{mshiftchar}},
  \field{text}=\marg{\cs{vec}\marg{v}},
  \field{description}=\marg{a vector}
}
\strut
\atentry{entry}\marg{S,
  \field{name}=\marg{\marg{}\idx{mshiftchar}\cmd{set}\marg{S}\idx{mshiftchar}},
  \field{text}=\marg{\cmd{set}\marg{S}},
  \field{description}=\marg{a set}
}
\strut
\atentry{entry}\marg{card,
  \field{name}=\marg{\marg{}\idx{mshiftchar}\cmd{card}\marg{S}\idx{mshiftchar}},
  \field{text}=\marg{\cmd{card}\marg{S}},
  \field{description}=\marg{the cardinality of the set \idx{mshiftchar}\cmd{set}\marg{S}\idx{mshiftchar}}
}
\strut
\atentry{entry}\marg{i,
  \field{name}=\marg{\marg{}\idx{mshiftchar}\cmd{imaginary}\idx{mshiftchar}},
  \field{text}=\marg{\cmd{imaginary}},
  \field{description}=\marg{square root of minus one (\idx{mshiftchar}\cmd{sqrt}\marg{-1}\idx{mshiftchar})}
}
\end{codeenv}
(The empty group at the start of the \field{name} fields
protects against the possibility that the \catattr{glossname}
category attribute might be set to \optfmt{firstuc}, which
automatically converts the first letter of the name to 
\idx{uppercase} when displaying the glossary. See also
\longarg{mfirstuc-protection} and
\longarg{mfirstuc-math-protection}.)

None of these entries have a \field{sort} field so the \field{name}
is used. If the entry type had been \atentry{symbol} instead,
the fallback would be the entry's label.
This means that with \atentry{symbol} instead of \atentry{entry}, and the default
\csopt[sort]{sort-field}, and with
\csopt[letter-case]{sort}, these entries will be defined in
the order: \code{M}, \code{S}, \code{card}, \code{i}, \code{v} (since
this is the case-sensitive letter order of the labels) whereas
with \csopt[letter-nocase]{sort-field}, the order will be:
\code{card}, \code{i}, \code{M}, \code{S}, \code{v} (since this
is the case-insensitive letter order of the labels).

However, with \atentry{entry}, the fallback field will be taken from
the \field{name} which in the above example contains \TeX\ code, so
\bibgls\ will use \file{texparserlib.jar} to interpret this code.
The library has several different ways of writing the processed
code. For simplicity, \bibgls\ uses the library's HTML output and
then strips the HTML markup and trims any leading or trailing
spaces. The library method that writes \idx{non-ASCII} characters
using \qtt{\&x\meta{hex};} markup is overridden by \bibgls\ to just
write the actual Unicode character, which means that the letter-based
sorting options will sort according to the integer value \meta{hex}
rather than the string \qtt{\&x\meta{hex};}.

The interpreter is first passed the code provided with
\atentry{preamble}:
\begin{codeenv}
\cs{providecommand}\marg{\cmd{mtx}}[1]\marg{\cs{boldsymbol}\marg{\idx{param}1}}
\cs{providecommand}\marg{\cmd{set}}[1]\marg{\cs{mathcal}\marg{\idx{param}1}}
\cs{providecommand}\marg{\cmd{card}}[1]\marg{|\cmd{set}\marg{\idx{param}1}|}
\cs{providecommand}\marg{\cmd{imaginary}}\marg{i}
\end{codeenv}
(unless \csopt[false]{interpret-preamble}).
This means that the provided commands are now recognised by the
interpreter when it has to parse the fields later.

In the case of the \code{M} entry in the example above, the 
code that's passed to the interpreter is:
\begin{verbatim}
{}$\mtx{M}$
\end{verbatim}
The transcript (\iext{glg}) file will show the results of the
conversion:
\begin{verbatim}
texparserlib: {}$\mtx{M}$ -> M
\end{verbatim}
So the \field{sort} value for this entry is set to \qtt{M}. The font
change (caused by math-mode and \ics{boldsymbol}) has been 
ignored. The sort value therefore consists of a single Unicode
character \hex{4D} (Latin \idx!{uppercase} letter \qtt{M}, decimal value 77).

For the \code{v} entry, the code is:
\begin{verbatim}
{}$\vec{v}$
\end{verbatim}
The transcript shows:
\begin{alltt}
texparserlib: \marg{}\$\ics{vec}\marg{v}\$ -> \usebox\varrow
\end{alltt}
So the \field{sort} value for this entry is set to \qtt{\usebox\varrow},
which consists of two Unicode characters \hex{76}
(Latin \idx!{lowercase} letter \qtt{v}, decimal value 118) and 
\hex{20D7} (combining right arrow above, decimal value 8407).

For the \code{set} entry, the code is:
\begin{verbatim}
{}$\set{S}$
\end{verbatim}
The transcript shows:
\begin{verbatim}
texparserlib: {}$\set{S}$ -> S
\end{verbatim}
So the \field{sort} value for this entry is set to \qtt{S} (again
ignoring the font change). This consists of a single Unicode
character \hex{53} (Latin \idx!{uppercase} letter \qtt{S}, decimal value~83).

For the \code{card} entry, the code is:
\begin{verbatim}
{}$\card{S}$
\end{verbatim}
The transcript shows:
\begin{verbatim}
texparserlib: {}$\card{S}$ -> |S|
\end{verbatim}
So the \field{sort} value for this entry is set to \qtt{|S|}
(the \textbar\ characters from the definition of \csfmt{card} provided
in \atentry{preamble} have been included, but the
font change has been discarded). In this case the sort value
consists of three Unicode characters \hex{7C} (vertical line, decimal
value 124),
\hex{53} (Latin \idx!{uppercase} letter \qtt{S}, decimal value 83) and
\hex{7C} again.
If \csopt[false]{interpret-preamble} had been used, \csfmt{card}
wouldn't be recognised and would be discarded leaving just \qtt{S}
as the sort value.

(Note that if \ics{vert} is used instead of \textbar\ then it would
be converted into the mathematical operator \hex{2223} and result in
a different order.)

For the \code{i} entry, the code is:
\begin{verbatim}
{}$\imaginary$
\end{verbatim}
The transcript shows:
\begin{verbatim}
texparserlib: {}$\imaginary$ -> i
\end{verbatim}
So the \field{sort} value for this entry is set to \qtt{i}.
If \csopt[false]{interpret-preamble} had been used, \csfmt{imaginary}
wouldn't be recognised and would be discarded, leaving an empty sort
value.

This means that in the case of the default \csopt[sort]{sort-field} with
\csopt[letter-case]{sort}, these entries will be defined in
the order: \code{M} ($\mtx{M}$), \code{S} ($\set{S}$),
\code{i} ($\imaginary$), \code{v} ($\vec{v}$) and
\code{card} ($\card{S}$).
In this case, the entries have been sorted according to
the character codes. If you run \bibgls\ with \longarg{verbose}
the decimal character codes will be included in the transcript.
For this example:
\begin{alltt}
i -> 'i' [105]
card -> '|S|' [124 83 124]
M -> 'M' [77]
S -> 'S' [83]
v -> '\usebox\varrow' [118 8407]
\end{alltt}

The \longarg{group} option (in addition to \longarg{verbose}) 
will place the letter group in 
parentheses before the character code list:
\begin{alltt}
i -> 'i' (i) [105]
card -> '|S|' [124 83 124]
M -> 'M' (M) [77]
S -> 'S' (S) [83]
v -> '\usebox\varrow' (v) [118 8407]
\end{alltt}
(Note that the \code{card} entry doesn't have a letter
group since the vertical bar character isn't considered a 
letter.)

If \csopt[letter-nocase]{sort} is used instead then, after conversion
by the interpreter, the sort values will all be changed to
\idx{lowercase}. The order is now: \code{i} ($\imaginary$), 
\code{M} ($\mtx{M}$), \code{S} ($\set{S}$),
\code{v} ($\vec{v}$) and \code{card} ($\card{S}$).
The transcript (with \longarg{verbose}) now shows
\begin{alltt}
i -> 'i' [105]
card -> '|s|' [124 115 124]
M -> 'm' [109]
S -> 's' [115]
v -> '\usebox\varrow' [118 8407]
\end{alltt}
With \longarg{group} (in addition to \longarg{verbose}) 
the letter groups are again included:
\begin{alltt}
i -> 'i' (I) [105]
card -> '|s|' [124 115 124]
M -> 'm' (M) [109]
S -> 's' (S) [115]
v -> '\usebox\varrow' (V) [118 8407]
\end{alltt}
Note that the letter groups are \idx{uppercase} not \idx{lowercase}.
Again the \code{card} entry doesn't have an associated
letter group.

If a locale-based sort is used, the ordering will follow the
locale's alphabet rules. For example, with \csopt[en]{sort}
(English, no region or variant), the order becomes:
\code{card} ($\card{S}$), \code{i} ($\imaginary$), 
\code{M} ($\mtx{M}$), \code{S} ($\set{S}$) and
\code{v} ($\vec{v}$). The transcript (with \longarg{verbose}) 
shows the collation keys instead:
\begin{alltt}
i -> 'i' [0 92 0 0 0 0]
card -> '|S|' [0 66 0 102 0 66 0 0 0 0]
M -> 'M' [0 96 0 0 0 0]
S -> 'S' [0 102 0 0 0 0]
v -> '\usebox\varrow' [0 105 0 0 0 0]
\end{alltt}
Again the addition of the \longarg{group} switch will show
the letter groups.\footnote{For more information on collation keys see the
\href{http://docs.oracle.com/javase/8/docs/api/java/text/CollationKey.html}{CollationKey}
class in Java's API~\cite{collationkey}.}

Suppose I add a new symbol to my \ext{bib} file:
\begin{codeenv}
\atentry{symbol}\marg{angstrom,
  \field{name}=\marg{\ics{AA}},
  \field{description}=\marg{\cs{AA} ngstr\ics{umlaut}om}
}
\end{codeenv}
and I also use this entry in the document.\footnote{A better method
is to use \sty{siunitx} instead.} Then with
\csopt[en]{sort}, the order is: \code{card} ($\card{S}$),
\code{angstrom} (\AA), \code{i} ($\imaginary$), \code{M}
($\mtx{M}$), \code{S} ($\set{S}$), and \code{v} ($\vec{v}$).
The \longarg{group} switch shows that the \code{angstrom} entry
(\AA) has been placed in the \qt{A} letter group.

However, if I change the locale to \csopt[sv]{sort}, the
\code{angstrom} entry is moved to the end of the list and the
\longarg{group} switch shows that it's been placed in the \qt{\AA}
letter group.

\label{locale.provider}If you are using Java~8, you can set the
\code{java.locale.providers} property~\cite{javacldr} to use
the \idx{CLDR} \idx{localeprovider}, which has more extensive
support for locales than the native \idx{JRE}. For example:
\begin{verbatim}
java.locale.providers=CLDR,JRE
\end{verbatim}
This should be enabled by default for Java~9. The property
can either be set in a script that runs \bibgls, for example,
\begin{verbatim}
java -Djava.locale.providers=CLDR,JRE,SPI -jar "$jarpath" "$@"
\end{verbatim}
(where \verb|$jarpath| is the path to the \file{bib2gls.jar}
file and \verb|"$@"| is the argument list) or you can set the
property as the default for all Java applications by adding
the definition to the \code{JAVA\_TOOL\_OPTIONS} environment
variable~\cite{javaoptions}.
For example, in a bash shell:
\begin{verbatim}
export JAVA_TOOL_OPTIONS='-Djava.locale.providers=CLDR,JRE,SPI'
\end{verbatim}
or in Windows:
\begin{verbatim}
set JAVA_TOOL_OPTIONS=-Djava.locale.providers=CLDR,JRE,SPI
\end{verbatim}

\chapter{Command Line Options}
\label{sec:switches}
\setsecdepth{0}

The syntax of \bibgls\ is:
\begin{alltt}
bib2gls \oargm{options} \meta{filename}
\end{alltt}
where \meta{filename} is the name of the \ext{aux} file. (The
extension may be omitted.) Only one \meta{filename} is permitted.
Available options are listed below.

\argsection{help}

Display the help message and quit.

\argsection{version}

Display the version information and quit. As from v2.5, this now
includes the version number of the \file{texparserlib.jar} library.

\argsection{debug}

Switch on debugging mode. If \meta{n} is present, it must be a
non-negative integer indicating the debugging level. If omitted 1 is
assumed. This option also switches on the verbose mode. A value of 0
is equivalent to \longargfmt{no-debug}.

\argsection{no-debug}

Switches off the debugging mode.

\argsection{verbose}

Switches on the verbose mode. This writes extra information to the
terminal and transcript file.

\argsection{no-verbose}

Switches off the verbose mode. This is the default behaviour.
Some messages are written to the terminal. To completely suppress
all messages (except errors), switch on the silent mode.
For additional information messages, switch on the verbose mode.

\argsection{silent}

Suppresses all messages except for errors that would normally be
written to the terminal. Warnings and informational messages are
written to the transcript file, which can be inspected afterwards.

\argsection{locale}

Specify the preferred \langxml, where \meta{lang} is a valid \idx{IETF} language tag.
This option requires an appropriate \metafilefmt{bib2gls-}{lang}{.xml}
resource file otherwise \bibgls\ will fallback on English.
This also sets the default document locale when \csopt[doc]{sort} is
used and the document doesn't have any language support.
Note that \csopt[locale]{sort} uses the \idx{JVM}['s] default
locale and is not governed by this switch.

If a document doesn't have any locale support or has support
for more than one language then it's best to explicitly set
the required locale in the appropriate \idx{resourceset}.

\argsection{log-file}

Sets the name of the transcript file. By default, the name is the
same as the \iext{aux} file but with a \iext{glg} extension. Note that
if you use \bibgls\ in combination with \idx{xindy} or
\idx{makeindex}, you will need to change the transcript file name to
prevent conflict.

\argsection{dir}

By default \bibgls\ assumes that the output files should be written
in the current working directory. The input \iext{bib} files are assumed to be
either in the current working directory or on \TeX's path (in which
case \idx{kpsewhich} will be used to find them).

If your \iext{aux} file isn't in the current working directory (for
example, you have run \TeX\ with \shortargfmt{output-directory})
then you need to take care how you invoke \bibgls.

Suppose I have a file called \filefmt{test-entries.bib} that
contains my entry definitions and a document called
\filefmt{mydoc.tex} that selects the \ext{bib} file using:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[test-entries]{src}}
\end{codeenv}
(\filefmt{test-entries.bib} is in the same directory as
\filefmt{mydoc.tex}).
If I compile this document using
\begin{verbatim}
pdflatex -output-directory tmp mydoc
\end{verbatim}
then the auxiliary file \filefmt{mydoc.aux} will be written to the
\filefmt{tmp} sub-directory. The resource information is listed in
the \ext{aux} file as
\begin{codeenv}
\gls{glsxtr@resource}\marg{\csopt[test-entries]{src}}\marg{mydoc}
\end{codeenv}
If I run \bibgls\ from the \filefmt{tmp} directory, then it won't
be able to find the \filefmt{test-entries.bib} file (since it's in
the parent directory).

If I run \bibgls\ from the same directory as \filefmt{mydoc.tex}
using
\begin{verbatim}
bib2gls tmp/mydoc
\end{verbatim}
then the \ext{aux} file is found and the transcript file is
\filefmt{tmp/mydoc.glg} (since the default path name is the same as the
\ext{aux} file but with the extension changed to \iext{glg}) but the
output file \filefmt{mydoc.glstex} will be written to the current
directory.

This works fine from \TeX's point of view as it can find the
\iext{glstex} file, but it may be that you'd rather the \ext{glstex}
file was tidied away into the \filefmt{tmp} directory along with all
the other files.  In this case you need to invoke \bibgls\ with the
\longargorshort{dir} option:
\begin{verbatim}
bib2gls -d tmp mydoc
\end{verbatim}

\argsection{interpret}

Switch on the interpreter mode (default). See \sectionref{sec:texparserlib}
for more details.

\argsection{no-interpret}

Switch off the interpreter mode. See \sectionref{sec:texparserlib}
for more details about the interpreter. 

\argsection{no-break-space}

The interpreter treats a tilde character \idx{nbspchar} as a non-breakable
space (default). Similarly \ics{nobreakspace} produces a
non-breakable space character (\hex{00A0}).

\argsection{break-space}

The interpreter treats a tilde character \idx{nbspchar} as a normal
space. Similarly \ics{nobreakspace} just produces a space.

\argsection{cite-as-record}

Treat instances of \code{\ics{citation}\margm{label}} found in the 
\ext{aux} file as though it was actually an \idx{ignoredrecord}:
\begin{codeenv}
\gls{glsxtr@record}\margm{label}\marg{}\marg{\counter{page}}\marg{\encap{glsignore}}\marg{}
\end{codeenv}
Note that \code{\cs{citation}\marg{*}} will always be skipped. Use
\csopt[all]{selection} to select all entries.
This switch is most useful in conjunction with
\atentrypageref{bibtexentry}.

\argsection{no-cite-as-record}

Don't check for instances of \ics{citation} in the \ext{aux} file
(default).

\argsection{warn-non-bib-fields}

If any internal fields are found in the \ext{bib} file, this setting
will issue a warning as their use can cause unexpected results.
The fields checked for are those listed in Tables~\ref{tab:internalfields}
and \ref{tab:baseinternalfields} with a few exceptions, notably
\field{type} and \field{sort}. Ideally you shouldn't need to use
\field{sort} as there should be an appropriate fallback set up to
use if \field{sort} isn't set, such as the label for symbols or the
name for terms or the short form for abbreviations.

This is the default setting and was added as some users were
confused over which fields could be used in the \ext{bib} file.
The use of these fields can break \bibgls's normal behaviour and
cause unexpected results.

The check is performed before field aliasing, so it's possible to
alias a field to an internal field, such as \field{group}, without
triggering this warning. If you do this you need to make sure you
have taken appropriate precautions to avoid unexpected results.

\argsection{no-warn-non-bib-fields}

Switches off the check for non-bib fields. If you use this option
you need to make sure you have taken appropriate precautions to
avoid unexpected results.

\argsection{warn-unknown-entry-types}

If any unknown entry types are found in the \ext{bib} file, \bibgls\
will issue a warning with this option set (default).

\argsection{no-warn-unknown-entry-types}

This option will suppress the warning if an unknown entry types are
found in the \ext{bib} file.

\argsection{merge-wrglossary-records}

For use with the \styopt{indexcounter} package option
(\styfmt{glossaries-extra} v1.29+), this switch merges an entry's
\counter{wrglossary} records for the same page location. This is the
default setting. (See also \csopt{save-index-counter}.)

\argsection{no-merge-wrglossary-records}

Don't merge an entry's \counter{wrglossary} records. This means that you
may end up with duplicate page numbers in the entry's location list,
but they will link to different parts of the page.

\argsection{merge-nameref-on}

The \styopt[nameref]{record} package option (introduced to 
\sty{glossaries-extra} version 1.37) provides extra information
in the record when indexing, obtained from \ics{@currentlabelname},
\ics{@currentHref} and \ics{theHentrycounter}. Instead of writing the record as:
\begin{codeenv}
\format{glsxtr@record}
\end{codeenv}
the record is written as:
\nosecdef{glsxtr@record@nameref}
If \isty{hyperref} hasn't been loaded \meta{title} and \meta{href}
will always be empty. The most reliable target is given by
\code{\meta{counter}.\meta{hcounter}}, where \meta{counter} is the
associated counter name and \meta{hcounter} is
obtained from \cs{theHentrycounter}, which is set to the hyper
target command \csfmt{theH}\meta{counter} during indexing. Since
this information can't be included in the location when indexing
with \idx!{makeindex} or \idx!{xindy}, the base \sty{glossaries}
package tries to obtain a prefix from which the target name can be
formed. This doesn't work if \csfmt{theH}\meta{counter} can't be
formed from \meta{prefix}\csfmt{the}\meta{counter}, which results in
broken links. Since \bibgls\ doesn't have the same restrictions, the
actual target can be included in the record. You can then customize
the document to choose whether to use \meta{href} (to link to the
nearest anchor) or \meta{hcounter} to link to the place where the
indexing counter was incremented.

The \code{nameref} record will be written to the \idx{locationlist} using:
\nosecdef{glsxtrdisplaylocnameref}
The \meta{file} part will be empty for normal internal locations,
and will be set to the corresponding file name for supplemental
locations.

With \sty{hyperref}, \meta{title} is initially empty. The \meta{href} will be
\code{Doc-Start} at the start of the document and is updated
globally on every instance of \ics{refstepcounter}. The 
\meta{title} is updated locally by certain commands, such as 
\ics{section} or \ics{caption}. This means that the \meta{href}
may not always correspond to the \meta{title}, so using
the \styopt[nameref]{record} package option can have unpredictable
results if the \meta{title} is used as link text with \meta{href} as
the target. 

For compactness, \bibgls\ tries to merge duplicate or near
duplicate records. There are four possible rules that it will
use for \code{nameref} records, identified by \meta{rule} in the
\longarg{merge-nameref-on} switch:
\begin{itemize}
\item \optfmt{location}: merge records that match on the
\meta{prefix}, \meta{counter} and \meta{location} parts (as regular
records);
\item \optfmt{title}: merge records that match on the \meta{counter}
and \meta{title} parts;
\item \optfmt{href}: merge records that match on the \meta{counter}
and \meta{href} parts;
\item \optfmt{hcounter}: merge records that match on the \meta{counter}
and \meta{hcounter} parts.
\end{itemize}
The default \meta{rule} is \optfmt{hcounter}. Note that for all
rules the \meta{counter} must match. See the \qt{Nameref Record}
section of the \sty{glossaries-extra} user manual for further
details.

\argsection{force-cross-resource-refs}

Force \idx{crossresourceref} mode on (see
\sectionref{sec:resourcesets}).

\argsection{no-force-cross-resource-refs}

Don't force \idx{crossresourceref} mode on (default). 
The mode will be enabled if applicable (see
\sectionref{sec:resourcesets}).

\argsection{support-unicode-script}

Text superscript (\ics{textsuperscript}) and subscript
(\ics{textsubscript}) will use Unicode super/subscript characters
if available (default). For example,
\begin{codeenv}
\cs{textsuperscript}\marg{(2)}
\end{codeenv}
will be converted to \code{\textsuperscript{(2)}}, which consists
of: \hex{207D} (superscript left parenthesis)
\hex{00B2} (superscript two) \hex{207E} (superscript right
parenthesis). If the entire contents of the argument can't be
represented by Unicode characters, the interpreter uses \verb|<sup>|
and \verb|<sub>| markup, which is then stripped by \bibgls. For
example,
\begin{codeenv}
\cs{textsuperscript}\marg{(2,3)}
\end{codeenv}
will be converted to
\begin{verbatim}
<sup>(2,3)</sup>
\end{verbatim}
(since there's no superscript comma). The markup is stripped leaving
just \code{(2,3)}.

Superscripts and subscripts in maths mode always use markup
regardless of this setting. Some supported packages that use
\idx{spchar} or \idx{sbchar} as shortcuts within an encapsulating
command may internally use the same code as \cs{textsuperscript} and
\cs{textsubscript}, in which case they will be sensitive to this
setting.

\argsection{no-support-unicode-script}

Text superscript (\cs{textsuperscript}) and subscript
(\cs{textsubscript}) won't use Unicode super/subscript characters.
Note that if other commands are provided that expand to Unicode
superscript or subscript characters, then they won't be affected by
this setting. For example, if \csfmt{superiortwo} is defined as
\begin{codeenv}
\cs{providecommand}\marg{\cmd{superiortwo}}\marg{\charhex{B2}}
\end{codeenv}
then it will be interpreted as \hex{00B2} (superscript two) even if
this setting is on.

\argsection{list-known-packages}

This option will list all the packages supported by the \TeX\ parser
library and will then exit \bibgls. The results are divided into two
sections: those packages that are searched for in the \iext{log}
file and those packages that aren't searched for in the \iext{log}
file but have some support available. Some of the support is very
limited. Package options aren't detected.
The transcript file is always searched for \isty{glossaries-extra}
to ensure that the version is new enough to support \bibgls. 

Packages that fall into the first category are:
\isty{amsmath}, \isty{amssymb}, \isty{bpchem}, \isty{fontenc},
\isty{fontspec}, \isty{fourier}, 
\isty{hyperref}, \isty{lipsum}, \isty{MnSymbol}, \isty{mhchem}, \isty{natbib},
\isty{pifont}, \isty{siunitx} (limited), \isty{stix},
\isty{textcase}, \isty{textcomp}, \isty{tipa}, \isty{upgreek} and
\isty{wasysym}.  (You can omit checking for specific packages with
\longarg{ignore-packages}.) These are packages that provide commands
that might be needed within entry fields. The check for
\isty{fontspec} is to simply determine whether or not UTF-8
characters are allowed in labels (for \csopt{labelify} and 
\csopt{labelify-list}).

Packages that fall into the second category are:
\isty{booktabs}, \isty{color}, \isty{datatool-base} (very limited),
\isty{datatool} (very limited), \isty{etoolbox} (very limited), \isty{graphics},
\isty{graphicx}, \isty{ifthen}, \isty{jmlrutils},
\isty{mfirstuc-english}, \isty{probsoln}, \isty{shortvrb},
and \isty{xspace}. These are less likely to be needed within fields
and so aren't checked for by default. If they are needed then you
can instruct \bibgls\ to support them with \longarg{packages}.

Note that \sty{mfirstuc} is always automatically loaded, but
\sty{mfirstuc-english} is not implemented unless explicitly
requested with \code{\longarg{packages} mfirstuc-english}.

If you're wondering about the selection, the \file{texparserlib.jar}
library was originally written for another application that required
support for some of them.

\argsection{packages}

Instruct the interpreter to assume the packages listed 
in \meta{list} have been used by the document.
This option has a cumulative action so \code{\longarg{packages}
"wasysym,pifont"} is the same as \code{\longarg{packages} wasysym 
\longarg{packages} pifont}.

Note that there's only a limited number of packages supported by the
\TeX\ parser library. This option is provided for cases where you're
using a command from a package that the interpreter doesn't support
but it happens to have the same name and meaning as a command from
a package that the interpreter does support. You can also use it to
provide support for known packages that aren't checked for when the
\ext{log} file is parsed. If you want \bibgls\ to parse an
unsupported package use \longarg{custom-packages}.

\argsection{custom-packages}

Instruct the interpreter to parse the package files identified in
\meta{list}. The package files need to be quite simple. When this
switch is used, the interpreter can recognise \ics{ProvidesPackage},
\ics{DeclareOption} (and \ics{DeclareOption*}),
\ics{ProcessOptions}, \ics{PackageError} and \ics{RequirePackage},
but it can't deal with complicated code.  In the case of
\ics{RequirePackage}, support will also be governed by
\longargfmt{custom-packages}.  This option has a cumulative action.

\argsection{ignore-packages}

This option is cumulative. When the document \iext{log} file is
parsed for known packages, \bibgls\ will skip the check for any
listed in \meta{list}. Note that this option simply instructs
\bibgls\ to ignore the package information in the log file. Any packages
that are identified with \longarg{packages} will be passed to the
interpreter if support is available, even if the package is also
listed in \longargfmt{ignore-packages}.  Note that
unknown packages can't be included in the ignored \meta{list}.

\argsection{mfirstuc-protection}

Commands like \ics{Gls} use \ics{makefirstuc} provided by the
\isty{mfirstuc} package. This command has limitations and one of the
things that can break it is the use of a referencing command 
at the start of its argument. The \sty{glossaries-extra} package has
more detail about the problem in the \qt{Nested Links} section of
the user manual~\cite{glossaries-extra}. If a glossary field starts
with one of these problematic commands, the recommended method (if
the command can't be replaced) is to insert an empty group in front
of it.

For example, the following definition
\begin{codeenv}
\gls{newabbreviation}\marg{shtml}\marg{shtml}\marg{\ics{glsps}\marg{ssi} enabled \cs{glsps}\marg{short}\marg{html}}
\end{codeenv}
will cause a problem for \code{\cs{Gls}\marg{shtml}} on first use.
The above example would be written in a \ext{bib} file as:
\begin{codeenv}
\atentry{abbreviation}\marg{shtml,
  \field{short}=\marg{shtml},
  \field{long}=\marg{\cs{glsps}\marg{ssi} enabled \cs{glsps}\marg{html}}
}
\end{codeenv}
The default \sty{mfirstuc} protection will automatically insert an empty
group before \code{\cs{glsps}\marg{ssi}} when writing the definition
in the \ext{glstex} file.

The argument for this switch should either be a comma-separated list
of fields or the keyword \code{all} (which indicates all fields).
\bibgls\ will automatically insert an empty group at the start of
the listed fields that start with a problematic command, and a
warning will be written to the transcript. Unknown fields are
skipped even if they're included in the list. An empty argument is
equivalent to \longarg{no-mfirstuc-protection}. The default value is
\code{all}.

\argsection{no-mfirstuc-protection}

Switches off the \isty{mfirstuc} protection mechanism described
above.

\argsection{mfirstuc-math-protection}

This works in the same way as \longarg{mfirstuc-protection} but
guards against fields starting with inline maths
(\idx{mshiftchar}\ldots\idx{mshiftchar}). For example, if the
\field{name} field starts with
\code{\idx{mshiftchar}x\idx{mshiftchar}} and the glossary style
automatically tries to convert the first letter of the name to
\idx{uppercase}, then this will cause a problem.

With \longarg{mfirstuc-math-protection} set, \bibgls\ will
automatically insert an empty group at the start of the field and
write a warning in the transcript. This setting is on by default.

\argsection{no-mfirstuc-math-protection}

Switches off the above.

\argsection{nested-link-check}

By default, \bibgls\ will parse certain fields for potential nested links.
(See the section \qt{Nested Links} in the \sty{glossaries-extra}
user manual~\cite{glossaries-extra}.)

The default set of fields to check are: \field{name}, \field{text},
\field{plural}, \field{first}, \field{firstplural}, \field{long},
\field{longplural}, \field{short}, \field{shortplural} and
\field{symbol}.

You can change this set of fields using
\longarg{nested-link-check} \meta{value} where \meta{value} may be
\optfmt{none} (don't parse any of the fields) or a comma-separated
list of fields to be checked.

\argsection{no-nested-link-check}

Equivalent to \longarg{nested-link-check} \optfmt{none}.

\argsection{shortcuts}

Some entries may reference another entry within a field, using
commands like \ics{gls}, so \bibgls\ parses the fields for these
commands to determine dependent entries to allow them to be selected
even if they haven't been used within the document.
The \styopt{shortcuts} package option provided by
\styfmt{glossaries-extra} defines various synonyms, such as \ics{ac}
which is equivalent to \ics{gls}. By default the value of the
\styopt{shortcuts} option will be picked up by \bibgls\ when parsing the
\iext{aux} file. This then allows \bibgls\ to additionally search for
those shortcut commands while parsing the fields.

You can override the \styopt{shortcuts} setting using
\longarg{shortcuts} \meta{value} (where \meta{value} may take
any of the allowed values for the \styopt{shortcuts} package option), 
but in general there is little need to use this switch.

\argsection{map-format}

This sets up the rule of precedence for partial location
matches (see \sectionref{sec:locationopts}). The argument may be
a comma-separated list of \meta{map}\code{:}\meta{value} pairs.
Alternatively, you can have multiple instances of 
\longarg{map-format} \meta{map}\dcolon\meta{value} which have a
cumulative effect.

For example, 
\begin{verbatim}
bib2gls --map-format "emph:hyperbf" mydoc
\end{verbatim}
This essentially means that if there's a record conflict
involving \encap{emph}, try replacing \encap{emph} with
\encap{hyperbf} and see if that resolves the conflict.

Note that if the conflict includes a range formation, the
range takes precedence.
The mapping tests are applied as the records are
read. For example, suppose the records are listed in the \ext{aux}
file as:
\begin{codeenv}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{emph}}\marg{3}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{hypersf}}\marg{3}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{hyperbf}}\marg{3}
\end{codeenv}
and \bibgls\ is invoked with
\begin{verbatim}
bib2gls --map-format "emph:hyperbf,hypersf:hyperit" mydoc
\end{verbatim}
or
\begin{verbatim}
bib2gls --map-format emph:hyperbf --map-format hypersf:hyperit mydoc
\end{verbatim}
then \bibgls\ will process these records as follows:
\begin{enumerate}
\item Accept the first record (\encap{emph}) since there's
currently no conflict. (This is the first record for page~3 for the
entry given by \code{gls.sample}.)
\item The second record (\encap{hypersf}) conflicts 
with the existing record (\encap{emph}). Neither has
the format \encap{glsnumberformat} or \encap{glsignore} so \bibgls\ consults
the mappings provided by \longargfmt{map-format}.
\begin{itemize}
\item The \encap{hypersf} format (from the new record) is mapped to
\encap{hyperit},
so \bibgls\ checks if the existing record
has this format. In this case it doesn't (the format is
\code{emph}). So \bibgls\ moves on to the next test:

\item The \encap{emph} format (from the existing record) is mapped
to \encap{hyperbf},
so \bibgls\ checks if the new record has this format.
In this case it doesn't (the format is \encap{hypersf}).

Since the provided mappings haven't resolved this conflict,
the new record is discarded with a warning. Note that there's 
no look ahead to the next record. (There may be other 
records for other entries also used on page~3 interspersed between these records.)
\end{itemize}
\item The third record (\encap{hyperbf}) conflicts 
with the existing record (\encap{emph}). Neither has
the format \encap{glsnumberformat} or \encap{glsignore} so \bibgls\ again consults
the mappings provided by \longargfmt{map-format}.
\begin{itemize}
\item  The new record's \encap{hyperbf} format has no mapping provided,
so \bibgls\ moves on to the next test:

\item The existing record's \encap{emph} format has a mapping
provided (\encap{hyperbf}). This matches the new record's format,
so the new record takes precedence.

This means that the location list ends up with the \encap{hyperbf}
location for page~3.
\end{itemize}
\end{enumerate}

If, on the other hand, the mappings are given as
\begin{verbatim}
--map-format "emph:hyperit,hypersf:hyperit,hyperbf:hyperit"
\end{verbatim}
then all the three conflicting records (\encap{emph},
\encap{hypersf} and \encap{hyperbf}) will end up being replaced
by a single record with \encap{hyperit} as the format.

Multiple conflicts will typically be rare as there's usually little
reason for more than two or three different location formats within
the same list. (For example, \encap{glsnumberformat} as the default
and \encap{hyperbf} or \encap{hyperit} for a primary reference.)

\argsection{group}

The \styfmt{glossaries-extra} \styopt{record} package option
automatically creates a new internal field called \field{group}. If the
\longarg{group} switch is used with the default \csopt[auto]{group}
option then, when sorting, \bibgls\ will try
to determine the letter group for each entry and assign it to the
\field{group} field.  (Some \csopt{sort} options ignore this
setting.) This value will be picked up by \ics{printunsrtglossary}
if group headings are required (for example with the
\glostyle{indexgroup} style) or if group separators are required
(for example, the \glostyle{index} style with the default 
\styopt[false]{nogroupskip}). If you don't require grouping within
the glossary, there's no need to use this switch. Note that this
switch doesn't automatically select an appropriate glossary style.

\begin{important}
The \field{group} field should typically not be set in the \ext{bib}
file and will trigger a warning if found. The explicit use of the
\field{group} key will override \bibgls's normal group formation behaviour,
which can cause unexpected results. The custom use of the
\field{group} field requires some care. As a general rule, if you
find yourself wanting to use the \field{group} field in the \ext{bib} file,
then the chances are that what you actually have is a hierarchical
glossary (list of topics) and what you really need is the \field{parent} field.
Compare the example files \exfile{sample-textsymbols.tex} and
\exfile{sample-textsymbols2.tex}. See also \sectionref{sec:logicaldivisions}.
\end{important}
There are eight types of groups:
\begin{description}
\item[\idx{lettergroup}] The first non-ignored character of the
sort value is alphabetic. This type of group occurs when using the
alphabetic sort methods listed in \tableref{tab:sortoptionsrule} or
with the letter sort methods listed in
\tableref{tab:sortoptionsletter} or with the letter-number sort
methods listed in \tableref{tab:sortoptionsletternumber}. The group
label is obtained from \gls!{bibglslettergroup}.

\item[\idx{nonlettergroup} (or \idx{symbolgroup})]
The first non-ignored character of all the sort values within this group
are non-alphabetical. This type of group occurs when using the
alphabetic sort methods listed in \tableref{tab:sortoptionsrule} or
with the letter sort methods listed in
\tableref{tab:sortoptionsletter} or with the letter-number sort
methods listed in \tableref{tab:sortoptionsletternumber}.
The alphabetic sort methods ignore many punctuation
characters, so an entry that has a non-alphabetic initial
character in the sort value may actually be placed in a
\idx{lettergroup}.
The group label is obtained from \gls!{bibglsothergroup}.

\item[\idx{emptygroup}] The sort value is empty when sorting with an
alphabetical, letter or letter-number method, typically a
result of the original value consisting solely of commands that \bibgls\
can't interpret. The group label is obtained from \gls!{bibglsemptygroup}.

\item[\idx{numbergroup}] The entries were sorted by one of the
numeric comparisons listed in \tableref{tab:sortoptionsnumerical}.
The group label is obtained from \gls!{bibglsnumbergroup}.

\item[\idx{datetimegroup}] The entries were sorted by one of the date-time
comparisons listed in \tableref{tab:sortoptionsdatetime} (where both
date and time are present). 
The group label is obtained from \gls!{bibglsdatetimegroup}.

\item[\idx{dategroup}] The entries were sorted by one of the date
comparisons (where the time is omitted). 
The group label is obtained from \gls!{bibglsdategroup}.

\item[\idx{timegroup}] The entries were sorted by one of the time
comparisons (where the date is omitted). 
The group label is obtained from \gls!{bibglstimegroup}.

\item[\idx{customgroup}] The group label is explicitly set either by
aliasing a field (with \csopt{field-aliases}) or by using the
\csopt[\meta{label}]{group} resource option. You will need to use
\gls{glsxtrsetgrouptitle} in the document to provide an associated title if the
\meta{label} isn't the same as the title.  Remember that the label
can't contain any active characters, so you can't use non-ASCII
characters in \meta{label} with \sty{inputenc} (but you can use
non-ASCII alphanumerics with \sty{fontspec}).

\end{description}

The \idx{lettergroup} titles will typically have the first character
converted to \idx{uppercase} for the alphabet sort methods
(\tableref{tab:sortoptionsrule}). A \qt{letter} may not necessarily
be a single character (depending on the sort rule), but may be
composed of multiple characters, such as a \idx{digraph} (two
characters) or \idx{trigraph} (three characters).

For example, if the sort rule recognises the digraph \qt{dz} as a
letter, then it will be converted to \qt{Dz} for the group title. 
There are some exceptions to this. For example, the Dutch digraph 
\qt{ij} should be \qt{IJ} rather than \qt{Ij}. This is indicated 
by the following line in the \langxml:
\begin{verbatim}
<entry key="grouptitle.case.ij">IJ</entry>
\end{verbatim}
If there isn't a \idx{grouptitle.case.lc} key (where 
\meta{lc} is the \idx!{lowercase} version), then only the first character
will be converted to \idx{uppercase} otherwise the value supplied by the
resource file is used. This resource key is only checked for
the alphabetical comparisons listed in
\tableref{tab:sortoptionsrule}. If the initial part of the sort
value isn't recognised as a letter according to the sort rule, then
the entry will be in a \idx{nonlettergroup} (even if the character is
alphabetical).

The letter (\tableref{tab:sortoptionsletter}) and letter-number
(\tableref{tab:sortoptionsletternumber}) methods only select the
first character of the sort value for the group. If the character is
alphabetical\footnote{according to Java's
\code{Character.isAlphabetic(int)} method} then it will be a 
\idx{lettergroup} otherwise it's a \idx{nonlettergroup}.  The case-insensitive
ordering (such as \csopt[letter-nocase]{sort}) will convert the
letter group character to \idx{uppercase}. The case-sensitive ordering
(such as \csopt[letter-case]{sort}) won't change the case.

Glossary styles with navigational links to groups (such as
\glostyle{indexhypergroup}) require an extra run for the ordinary
\cs{cs.makeglossaries} and \cs{makenoidxglossaries} methods. For
example, for the document \filefmt{myDoc.tex}:
\begin{verbatim}
pdflatex myDoc
makeglossaries myDoc
pdflatex myDoc
pdflatex myDoc
\end{verbatim}
On the first \appfmt{pdflatex} call, there's no glossary. On the second
\appfmt{pdflatex}, there's a glossary but the glossary must be
processed to find the group information, which is written to the
\iext{aux} file as
\nosecdef{@gls@hypergroup}
The third \appfmt{pdflatex} reads this information and is then able
to create the navigation links.

With \bibgls, if the \optfmt{type} is provided (through the
\field{type} field or via options such as \csopt{type} and
\csopt{dual-type}) then this information can be determined when
\bibgls\ is ready to write the \iext{glstex} file, which means that
the extra \LaTeX\ run isn't necessary. If \bibgls\ doesn't know
the glossary type then it will fallback on the original method
which requires an extra \LaTeX\ run.

For example:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\cmd{usepackage}\oarg{colorlinks}\marg{hyperref}
\cmd{usepackage}\oarg{\styopt{record},\styopt{abbreviations},\styopt[indexhypergroup]{style}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\comment{data in entries.bib}
  \csopt[main]{type}\comment{put these entries in the 'main' glossary}
}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[abbrvs]{src},\comment{data in abbrvs.bib}
  \csopt[abbreviations]{type}\comment{put entries in the 'abbreviations' glossary}
}
\end{codeenv}
Here the \csopt{type} is set and \bibgls\ can detect that
\isty{hyperref} has been loaded, so if the \longargfmt{group} switch
is used, then the group hyperlinks can be set (using
\gls!{bibglshypergroup}). This means that the build process is
just:
\begin{verbatim}
pdflatex myDoc
bibtex --group myDoc
pdflatex myDoc
\end{verbatim}
Note that this requires \isty{glossaries} v4.32+. If your version of
\sty{glossaries} is too old then \bibgls\ can't override
the default behaviour of \isty{glossary-hypernav}'s
\ics{glsnavhypertarget}.

If \isty{hyperref} isn't loaded or the \longargfmt{group} switch
isn't used or the \field{type} isn't set or your version of
\sty{glossaries} is too old, then the information can't be saved in
the \ext{glstex} file.

For example:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\cmd{usepackage}\oarg{colorlinks}\marg{hyperref}
\cmd{usepackage}\oarg{\styopt{record},\styopt{abbreviations},\styopt[indexhypergroup]{style}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}\comment{data in entries.bib}
\gls{GlsXtrLoadResources}\oarg{\csopt[abbrvs]{src}}\comment{data in abbrvs.bib}
\end{codeenv}
This requires the build process:
\begin{verbatim}
pdflatex myDoc
bibtex --group myDoc
pdflatex myDoc
pdflatex myDoc
\end{verbatim}
because the group hyperlink information can't be determined by
\bibgls, so it's best to always set the \optfmt{type} if you want
hyper-group styles, and make sure you have an up-to-date version of 
\styfmt{glossaries} (and \styfmt{glossaries-extra}).

\argsection{no-group}

Don't automatically set the \field{group} field with
\csopt[auto]{group} (default). The glossary won't have groups even if a
group style, such as \glostyle{indexgroup}, is used (unless the
\field{group} field is set to a custom value).

\argsection{tex-encoding}

\bibgls\ tries to determine the character encoding to use for the
output files. If the document has loaded the \isty{inputenc} package then
\bibgls\ can obtain the value of the encoding from the
\iext{aux} file. This is then converted to a name
recognised by Java. For example, \code{utf8} will be mapped to
\code{UTF-8}. If the \isty{fontspec} package has been loaded,
\styfmt{glossaries-extra} will assume the encoding is \code{utf8} and
write that value to the \ext{aux} file.

If neither package has been loaded, \bibgls\ will assume the \idx{JVM}['s]
default encoding (identified by the \code{file.encoding} property). If this is
incorrect or if \bibgls\ can't work out the appropriate mapping then you can
specify the correct encoding using \longargfmt{tex-encoding} \meta{name} where
\meta{name} is the encoding name (such as \code{UTF-8}).

If you have a problem with non-ASCII characters not displaying
correctly in your document:
\begin{itemize}
\item Check that the file encoding of your document \ext{tex} file (or files)
has been correctly set by your text editor.
\item Check that your document supports that encoding (for example,
through the \isty{inputenc} package).
\item Check \bibgls's transcript file (\ext{glg}) for the line that
starts
\begin{verbatim}
TeX character encoding:
\end{verbatim}
This should be followed by the encoding used by \bibgls\ when
creating the \ext{glstex} files. If this is incorrect use
\longargfmt{tex-encoding}.
\item Check that the encoding of the \ext{bib} files (set by your
text editor or bibliographic management system) matches the encoding
line in the \ext{bib} file or the \csopt{charset} resource option.
\end{itemize}

\argsection{no-expand-fields}

By default, \gls{newglossaryentry} and similar commands expand field values
(except for \field{name}, \field{symbol} and \field{description}).
This is useful if constructing field values programmatically
(for example in a \idx!{loop}) but can cause a problem if certain fragile
commands are included in the field. 

The switch \longarg{no-expand-fields} makes \bibgls\ write
\gls{glsnoexpandfields} to the \iext{glstex} file, which
switches off the expansion.  Since \bibgls\ is simply
fetching the data from \iext{bib} files, it's unlikely 
that this automatic expansion is required and since it can
also be problematic this option is on by default. You can
switch it off with \longarg{expand-fields}.

\argsection{expand-fields}

Don't write \gls{glsnoexpandfields} to the \iext{glstex} file,
allowing fields to expand when the entries are defined.
Remember that this doesn't include the \field{name}, \field{symbol}
or \field{description} fields, which need to have their
expansion switched on with \gls{glssetexpandfield}
before the entries are defined (that is, before using
\gls!{GlsXtrLoadResources}).

\argsection{trim-fields}

Trim leading and trailing spaces from all field values. For example,
if the \ext{bib} file contains:
\begin{codeenv}
\atentry{entry}\marg{sample,
  \field{name} = \marg{sample},
  \field{description} = \marg{
    an example
  }
}
\end{codeenv}
This will cause spurious spaces in the \field{description} field. 
Using \longarg{trim-fields} will automatically trim the values
before writing the \iext{glstex} file.

Note that even without this trimming option on, fields that are
set as keys within \cs{longnewglossaryentry} or the optional
argument of \cs{newabbreviation} will automatically have the leading and
trailing spaces internally trimmed by the \isty{xkeyval} package, so
this trimming action only affects fields that aren't set in this
way, such as the \field{description}, \field{long} and \field{short}
fields. If you specifically require a space at the start or end of a
field then use a spacing command, such as \cs{cs.space} or \cs{space} 
or \idx{nbspchar}.

\argsection{trim-only-fields}

Only trim leading and trailing spaces from the fields identified in
the comma-separated \meta{list}. This option has a cumulative effect
but is cancelled by \longarg{no-trim-fields} (which switches off all
trimming) and by \longarg{trim-fields} (which switches on trimming
for all fields). This option may not be used with
\longarg{trim-except-fields}.

For example, to only trim the \field{description} field:
\begin{verbatim}
bib2gls --trim-only-fields description myDoc
\end{verbatim}

\argsection{trim-except-fields}

Trim all leading and trailing spaces from fields except those identified in
the comma-separated \meta{list}. This option has a cumulative effect
but is cancelled by \longarg{no-trim-fields} (which switches off all
trimming) and by \longarg{trim-fields} (which switches on trimming
for all fields). This option may not be used with
\longarg{trim-only-fields}. See the above note about \sty{xkeyval}.

For example, to trim all fields except \field{short} and
\field{long}:
\begin{verbatim}
bib2gls --trim-except-fields short,long myDoc
\end{verbatim}
Or
\begin{verbatim}
bib2gls --trim-except-fields short --trim-except-fields long myDoc
\end{verbatim}

\argsection{no-trim-fields}

Don't trim any leading or trailing spaces from field values (but see
the above note about \sty{xkeyval}). This is the default setting.

\argsection{provide-glossaries}

This setting will make \bibgls\ add the line
\begin{codeenv}
\ics{provideignoredglossary*}\margm{type}
\end{codeenv}
to the \ext{glstex} file before an entry is defined where that entry
has the \field{type} field set to an unknown glossary type (\bibgls\
can detect from the \ext{aux} file all glossaries that have been
defined with \cs{newglossary} but not those defined with \cs{newignoredglossary}).

This ensures that the glossary exists, but the use of
\ics{provideignoredglossary} (rather than \cs{newignoredglossary})
will prevent an error if the glossary has already been defined.

\argsection{no-provide-glossaries}

This setting prevents \bibgls\ from providing unknown glossaries,
except in a few documented situations (the \csopt{master},
\csopt{trigger-type} and \csopt{secondary} options). This is the
default since it's a useful way of detecting misspelt glossary
labels. It's harder to detect the problem if a misspelt label has
caused an entry to be added to a hidden glossary.

\argsection{record-count}

Switch on record counting. This will ensure that when each entry
is written to the \iext{glstex} file, \bibgls\ will additionally
set the following fields
\begin{itemize}
\item \field{recordcount}: set to the total
number of records found for the entry;
\item \field{recordcount.counter}: set to the total
number of records found for the entry for the given counter.
\end{itemize}
These fields can then be used with the \gls{rgls}-like commands.
The default behaviour of
\nosecdef{rgls}
is to check the \field{recordcount} field against the \catattr{recordcount}
attribute value. This attribute can be set with
\nosecdef{GlsXtrSetRecordCountAttribute}
where \meta{category list} is a comma-separated list
of category labels and \meta{value} is a positive integer.
If the value of the \field{recordcount} field is greater than
\meta{value} then \gls{rgls} behaves like \ics{gls}, otherwise
it does
\nosecdef{rglsformat}
instead.
If the use of \gls{rglsformat} is triggered in this way,
then \gls{rgls} writes a record to the \iext{aux} file
with the \glsopt{format} set to \encap{glstriggerrecordformat}.
This ensures that the record count is correct on the next run,
but the record isn't added to the location list as
\bibgls\ recognises it as a special \idx{ignoredrecord}.
Note that the entry will still appear in the usual glossary unless
you assign it to a different one with \csopt{trigger-type}.

If the \catattr{recordcount} attribute hasn't been set
\gls{rgls} behaves like \ics{gls}. (That is, \gls{rgls}
uses the same internal command used by \ics{gls}.) You can use
\ics{glsxtrenablerecordcount} to redefine \ics{gls}
to \gls{rgls}, so that you can continue to use \ics{gls}
without having to switch command name.

For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[abbrevs]{src},\comment{entries defined in abbrevs.bib}
 \csopt[ignored]{trigger-type},
 \csopt[abbreviation]{category}
}
\cs{glsxtrenablerecordcount}
\gls{GlsXtrSetRecordCountAttribute}\marg{abbreviation}\marg{1}
\end{codeenv}
See the \sty{glossaries-extra} user manual~\cite{glossaries-extra} 
for further details.

\argsection{no-record-count}

Switch off record counting. (Default.)

\argsection{record-count-unit}

Automatically implements \longarg{record-count} and additionally
sets the \field{recordcount.counter.location} fields.
These fields can then be used with the \gls{rgls}-like
commands.

\argsection{no-record-count-unit}

Switches off unit record counting. (Default.)
Note that you need \longarg{no-record-count} to completely
switch off record counting.

\chapter{\iext{bib} Format}
\label{sec:bib}
\setsecdepth{1}

\bibgls\ recognises certain entry types. Any unrecognised types will
be ignored and a warning will be written to the transcript file.
Entries are defined in the usual \ext{bib} format:
\begin{codeenv*}
\idx{atchar}\meta{entry-type}\marg{\meta{id},
  \meta{field-name-1} = \margm{text},
  \ldots
  \meta{field-name-n} = \margm{text}
}
\end{codeenv*}
where \meta{entry-type} is the entry type (listed below),
\meta{field-name-1}, \ldots, \meta{field-name-n} are the field names and
\meta{id} is a unique label. The label can't contain any spaces or
commas, and most special characters are forbidden. The hyphen character and
some other punctuation characters are allowed by \bibgls, but you need to make 
sure that your document hasn't made them active. In general it's best to
stick with alpha-numeric labels.  The field values may be delimited by braces
\margm{text} or double-quotes \code{\qtdelim{\meta{text}}}.

The \csopt{label-prefix} option can be used to instruct \bibgls\ to
insert prefixes to the labels (\meta{id}) when the data is read.
Remember to use these prefixes when you reference the entries in the
document, but don't include them when you reference them in the
\ext{bib} file.  There are some special prefixes that have a
particular meaning to \bibgls: \qt{\idprefix{dual}} and
\qt{\idprefix{extn}} where \meta{n} is a positive integer.
In the first case, \idprefix{dual} references the dual element of a
dual entry (see \atentry{dualentry}). This prefix will be
replaced by the value of the \csopt{dual-prefix} option. The
\idprefix{extn} prefix is used to reference an entry from a
different set of resources (loaded by another
\gls{GlsXtrLoadResources} command). This prefix is replaced by the
corresponding element of the list supplied by \csopt{ext-prefixes},
but this is only supported if the \idx{crossresourceref} mode is enabled
(see \sectionref{sec:resourcesets}).

In the event that the \field{sort} value falls back on the label,
the original label supplied in the \ext{bib} file is used, not the
prefixed label.

\section{Encoding}
\label{sec:bibencoding}

Avoid \idx{non-ASCII} characters in the \meta{id} if your document uses the 
\isty{inputenc} package. (This isn't a problem for \XeLaTeX\ or
\LuaLaTeX, but you still need to avoid special characters.)
You can set the character \idx{encoding} in the \ext{bib} file using:
\begin{codeenv}
\idx{commentchar} Encoding: \meta{encoding-name}
\end{codeenv}
where \meta{encoding-name} is the name of the character encoding.
For example:
\begin{verbatim}
% Encoding: UTF-8
\end{verbatim}
You can also set the encoding using the \csopt{charset} option,
but it's simpler to include the above comment on the first line of
the \ext{bib} file. (This comment is also searched for by JabRef
to determine the encoding, so it works for both applications.) 
If you don't use either method \bibgls\ will
have to search the entire \ext{bib} file, which is inefficient and
you may end up with a mismatched encoding.

\section{Comments}
\label{sec:bibcomments}

The original \ext{bib} file format as defined by \BibTeX\ doesn't
have a designated comment character, but instead treats anything
outside of \code{@\meta{entry}\margm{data}} as unwanted material
that's ignored. This can catch out users who try to do something
like:
\begin{verbatim}
%@misc{sample, title={Sample} }
\end{verbatim}
In this case, the percent character is simply discarded and the
line is treated as:
\begin{verbatim}
@misc{sample, title={Sample} }
\end{verbatim}
Some applications that parse \ext{bib} files are less tolerant of
unwanted material. In the case of \bibgls, the percent character is
treated as a comment character and other unwanted material
should be omitted.  Avoid using comments within field values.
Comments are best placed outside of entry definitions.

The most common type of comment is the encoding comment, described
above.  \BibTeX's \entrydef{comment} is also supported by \bibgls\
for general comments, but not for the encoding. For example, JabRef
uses \atentryfmt{Comment} for metadata.
\begin{codeenv}
\atentryfmt{Comment}\marg{jabref-meta: databaseType:bib2gls;}
\end{codeenv}

\section{Fields}
\label{sec:fields}

Each entry type may have required fields, optional fields and
ignored fields. These are set using a \keyvallist\ within
\code{@\meta{entry-type}\marg{\meta{id},\meta{fields}}} in the
\ext{bib} file. Most keys recognised by
\gls{newglossaryentry} may be used as a field unless \bibgls\
considers them an internal field (see below). In general, you
shouldn't need to use the \field{sort} field.

If an optional field is missing and \bibgls\ needs to access it for
some reason, \bibgls\ will try to fallback on another value. The
actual fallback value depends on the entry type. The most common
fallback is that used if the \field{sort} field is missing, which is
typically the case. This approach allows different entry types to
have different fields used for sorting.

Predefined fields for use in \ext{bib} files are listed in
Tables~\ref{tab:fields}, \ref{tab:bib2glsfields},
\ref{tab:prefixfields} and~\ref{tab:accsuppfields}.  If you add any
custom keys in your document using \ics{glsaddkey} or
\ics{glsaddstoragekey}, those commands must be placed before the
first use of \gls{GlsXtrLoadResources} to ensure that \bibgls\
recognises them as a valid field name. 

Internal fields that may be assigned within the document (the
\LaTeX\ assignment code having been written by \bibgls\ in the
\iext{glstex} file) are listed in Table~\ref{tab:internalfields}.
These typically shouldn't be used in the \ext{bib} file. Some of
these fields can be set for a particular document using a resource
option, such as \csopt{type} or \csopt{group}. With
\longarg{warn-non-bib-fields} set, \bibgls\ will check for internal
fields that can cause interference with its normal operations and
will warn if any are found in the \ext{bib} file.

There are also some fields that are set and used by 
\styfmt{glossaries} or \styfmt{glossaries-extra} listed in
Table~\ref{tab:baseinternalfields} that aren't recognised by \bibgls.
In most cases these fields don't have a designated key and are only
intended for internal use by \bibgls\ or by the \styfmt{glossaries} or
\styfmt{glossaries-extra} package. Note that the value of the \field{sort} field
written to the \ext{bib} file doesn't always exactly match the sort
value used by \bibgls\ (which is stored in \field{bib2gls@sort}).
Any special characters found in the sort value are always substituted
before writing the \ext{bib} file to avoid syntax errors.

Any unrecognised fields will be ignored by \bibgls. This is more
convenient than using \ics{input} or \ics{loadglsentries}, which
requires all the keys used in the file to be defined, regardless of
whether or not you actually need them in the document.

Other entries can be cross-referenced using the \field{see},
\field{seealso} or \field{alias} fields or
by using commands like \ics{gls} or \ics{glsxtrp} in any of the
recognised fields. These will automatically be selected if the
\csopt{selection} setting includes dependencies, but you may need to
rebuild the document to ensure the location lists are correct.  Use
of the \ics{glssee} command will create an \idx{ignoredrecord} and the
\field{see} field will be set to the relevant information.  If an
entry has the \field{see} field already set, any instance of
\ics{glssee} in the document for that entry will be appended to the
\field{see} field (provided you have at least v1.14 of
\sty{glossaries-extra}). In general, it's best just to use the
\field{see} field and not use \ics{glssee}.

The \field{seealso} key was only added to \sty{glossaries-extra}
v1.16, but this field may be used with \bibgls\ even if you only
have version 1.14 or 1.15. If the key isn't available,
\code{\field{seealso}=\margm{xr-list}} will be treated as
\code{\field{see}=\marg{[\ics{seealsoname}]\meta{xr-list}}} (the resource
option \csopt{seealso} won't have an effect).  You can't use both
\field{see} and \field{seealso} for the same entry with \bibgls.
Note that the \field{seealso} field doesn't allow for the optional
\oargm{tag} part. If you need a different tag, either use \field{see}
or change the definition of \ics{seealsoname} or
\ics{glsxtruseseealsoformat}.  Note that, unless you are using
\idx!{xindy}, \ics{glsxtrindexseealso} just does
\ics{glssee}\oarg{\ics{seealsoname}}, and so will be treated as \field{see}
rather than \field{seealso} by \bibgls. Again, it's better to just
use the \field{seealso} field directly.

You can identify an arbitrary field as containing a list of dependent entry 
labels with \csopt{dependency-fields}. This instructs \bibgls\ to parse the 
listed fields for dependencies in a similar manner to the \field{see} field, 
but it doesn't add any information to the cross-referencing part of the 
location list. The option may be used in combination with the \field{see} or
\field{seealso} fields.

\clearpage

\printfields
 {basefield,extrafield}
 {Fields Provided by \styfmt{glossaries-extra}}
 {tab:fields}

\clearpage

\printfields
 {bib2glsfield}
 {Fields Provided by \bibgls}
 {tab:bib2glsfields}

\printfields
 {prefixfield}%
 [Fields Provided by \styfmt{glossaries-prefix}]%
 {Fields Provided by \isty{glossaries-prefix}}
 {tab:prefixfields}

\printfields
 [%
   Don't load \isty{glossaries-accsupp} directly (with \ics{usepackage}) 
   when using \styfmt{glossaries-extra}. Load using the 
   \glsadd{idx.accsupp}\styopt{accsupp} 
   package option instead.
%\begin{codeenv}
%\cs{usepackage}\oarg{\styopt{record},\styopt{accsupp}}\marg{glossaries-extra}
%\end{codeenv}
 ]%
 {accessfield}%
 [Fields Provided by \styfmt{glossaries-accsupp}]%
 {Fields Provided by \isty{glossaries-accsupp}}
 {tab:accsuppfields}

\printfields
 [%
  You may define and assign \field{bibtextype} as a key (although it's more
  likely to be aliased). Don't define any of the others listed in this
  table, and don't use any of them in the \ext{bib} file. A possible
  exception is the \field{type} field, but it's more flexible to set
  that through a resource option. The explicit use of \field{group} 
  within a \ext{bib} file can cause unpredictable
  results and is best set through a resource option or by \bibgls.
  In general, you shouldn't need to set the \field{sort} field as
  appropriate fallbacks should produce useful sort values.%
 ]%
 {internalfield}%
 [Fields Set by \bibgls]%
 {Fields Sometimes Set by \bibgls\ in the \iext{glstex} File}%
 {tab:internalfields}

\printfields
 [%
   Don't define any of these as keys and don't use any of them in the \ext{bib}
   file.%
 ]%
 {baseinternalfield}%
 [Internal Fields Set by \styfmt{glossaries} or \styfmt{glossaries-extra} 
  or \bibgls]%
 {Internal Fields Set by \isty{glossaries} or \isty{glossaries-extra} 
  or \bibgls}%
 {tab:baseinternalfields}

\clearpage

\section{Standard Entry Types}
\label{sec:standardentry}

\entrysection{string}

The standard \atentry{string} is available and can be used to define
variables that may be used in field values. Don't include braces or
double-quote delimiters when referencing a variable. You can use
\idx{stringconcat} to concatenate strings.
For example:
\begin{codeenv}
\atentry{string}\marg{ssi=\marg{server-side includes}}
\atentry{string}\marg{html=\marg{hypertext markup language}}
\strut
\atentry{abbreviation}\marg{shtml,
  \field{short}=\qtdelim{shtml},
  \field{long}=ssi \idx{stringconcat} \qtdelim{ enabled } \idx{stringconcat} html,
  \field{see}=\marg{ssi,html}
}
\strut
\atentry{abbreviation}\marg{html,
  \field{short}=\qtdelim{html},
  \field{long}=html
}
\strut
\atentry{abbreviation}\marg{ssi,
  \field{short}=\qtdelim{ssi},
  \field{long}=ssi
}
\end{codeenv}
Note the difference between \code{=\qtdelim{ssi}} (a field value delimited by
double-quotes), the undelimited \code{=ssi} (a reference to the
variable), the grouped \code{=\marg{ssi,html}} (a field value
delimited by braces) and \code{ssi} the entry label.

\entrysection{preamble}

The standard \atentry{preamble} is available and can be used to
provide command definitions used within field values.
For example:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cmd{mtx}}[1]\marg{\cs{boldsymbol}\marg{\idx{param}1}}}}
\strut
\atentry{entry}\marg{matrix,
  \field{name}=\marg{matrix},
  \field{plural}=\marg{matrices},
  \field{description}=\marg{rectangular array of values, denoted \idx{mshiftchar}\cmd{mtx}\marg{M}\idx{mshiftchar}}
}
\end{codeenv}
Alternatively you can use \ics{glsxtrprovidecommand} which behaves
the same as \ics{providecommand} within the document but behaves
like \ics{renewcommand} within \bibgls, which allows you to change
\bibgls's internal definition of a command without affecting the
definition within the document (if it's already been defined before
the resource file is input). In general, it's best to just use
\cs{providecommand}.

The \TeX\ parser library used by \bibgls\ will parse the contents of
\atentry{preamble} before trying to interpret the field value used
as a fallback when \field{sort} is omitted (unless
\csopt[false]{interpret-preamble} is set in the resource options).
For example:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cmd{set}}[1]\marg{\cs{mathcal}\marg{\idx{param}1}}
\cs{providecommand}\marg{\cmd{card}}[1]\marg{|\cmd{set}\marg{\idx{param}1}|}}}
\strut
\atentry{entry}\marg{S,
  \field{name}=\marg{\marg{}\idx{mshiftchar}\cmd{set}\marg{S}\idx{mshiftchar}},
  \field{text}=\marg{\cmd{set}\marg{S}},
  \field{description}=\marg{a set}
}
\atentry{entry}\marg{card,
  \field{name}=\marg{\marg{}\idx{mshiftchar}\cmd{card}\marg{S}\idx{mshiftchar}},
  \field{text}=\marg{\cmd{card}\marg{S}},
  \field{description}=\marg{the cardinality of \cs{gls}\marg{S}}
}
\end{codeenv}
Neither entry has the \field{sort} field, so \bibgls\ has to fall
back on the \field{name} field and, since this contains the special
characters \idx{escchar} (backslash), \idx{mshiftchar} (maths
shift), \idx{bgroupchar} (begin group) and
\idx{egroupchar} (end group), the \TeX\ parser library is used to interpret it.
The definitions provided by \atentry{preamble} allow \bibgls\ to
deduce that the \field{sort} value of the \code{S} entry is just
\code{S} and the \field{sort} value of the \code{card} entry is
\verb"|S|" (see \sectionref{sec:texparserlib}).

What happens if you also need to use these commands in the document?
The definitions provided in \atentry{preamble} won't be available
until the \iext{glstex} file has been created, which means the
commands won't be defined on the first \LaTeX\ run.

There are several approaches:
\begin{enumerate}
\item Just define the commands in the document. This means the
commands are available, but \bibgls\ won't be able to correctly
interpret the \field{name} fields.

\item Define the commands in both the document and in
\atentry{preamble}. For example:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{set}}[1]\marg{\cs{mathcal}\marg{\idx{param}1}}
\cs{newcommand}\marg{\cmd{card}}[1]\marg{|\cmd{set}\marg{\idx{param}1}|}
\gls{GlsXtrLoadResources}\oarg{\csopt[my-data]{src}}
\end{codeenv}
Alternatively:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[my-data]{src}}
\cs{providecommand}\marg{\cmd{set}}[1]\marg{\cs{mathcal}\marg{\idx{param}1}}
\cs{providecommand}\marg{\cmd{card}}[1]\marg{|\cmd{set}\marg{\idx{param}1}|}
\end{codeenv}
If the provided definitions match those given in the \ext{bib} file,
there's no difference. If they don't match then in the first example 
the document definitions will take precedence (but the interpreter
will use the \atentry{preamble} definitions) and in the second
example the \atentry{preamble} definitions will take precedence.
For example, the document may define \csfmt{card} as:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{card}}[1]\marg{\ics{vert}\cmd{set}\marg{\idx{param}1}\cs{vert}}
\end{codeenv}

\item Make use of \gls{glsxtrfmt} provided by \styfmt{glossaries-extra} which
allows you to store the name of the formatting command in a field.
The default is the \field{user1} field, but this can be changed to 
another field by redefining \ics{GlsXtrFmtField}.

The \ext{bib} file can now look like this:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cmd{set}}[1]\marg{\cs{mathcal}\marg{\idx{param}1}}
\cs{providecommand}\marg{\cmd{card}}[1]\marg{|\cmd{set}\marg{\idx{param}1}|}}}
\strut
\atentry{symbol}\marg{S,
  \field{name}=\marg{\marg{}\idx{mshiftchar}\cmd{set}\marg{S}\idx{mshiftchar}},
  \field{text}=\marg{\cmd{set}\marg{S}},
  \field{user1}=\marg{set},
  \field{description}=\marg{a set}
}
\atentry{symbol}\marg{cardS,
  \field{name}=\marg{\marg{}\idx{mshiftchar}\cmd{card}\marg{S}\idx{mshiftchar}},
  \field{text}=\marg{\cmd{card}\marg{S}},
  \field{user1}=\marg{card},
  \field{description}=\marg{the cardinality of \cs{gls}\marg{S}}
}
\end{codeenv}
Within the document, you can format \meta{text} using the formatting
command provided in the \field{user1} field with:
\nosecformatdef{glsxtrfmt}
(which internally uses \ics{glslink}) or
\nosecformatdef{glsxtrentryfmt}
which just applies the appropriate formatting command to
\meta{text}. Version 1.23+ of \sty{glossaries-extra} also provides a
starred form of the linking command:
\nosecformatdef{glsxtrfmt*}
which inserts additional material inside the link text but outside
the formatting command.

If the entry given by \meta{label} hasn't been defined,
then \gls!{glsxtrfmt} just does \meta{text} (followed by \meta{insert} for
the starred version) and a warning is issued. (There's no warning 
if the entry is defined but the field hasn't been set.)
The \meta{options} are as for \ics{glslink} but \ics{glslink} will
actually be using:
\begin{codeenv}
\ics{glslink}\oarg{\meta{def-options},\meta{options}}\margm{label}\marg{\cmd{}\meta{csname}\margm{text}\meta{insert}}
\end{codeenv}
where the default options \meta{def-options} are given by
\ics{GlsXtrFmtDefaultOptions}. The default definition of this is
just \code{noindex} which suppresses the automatic indexing or
recording action. (See the \sty{glossaries-extra}
manual~\cite{glossaries-extra} for further details.) The
\meta{insert} part is omitted for the unstarred form.

This means that the document doesn't need to actually provide
\verb|\set| or \verb|\card| but can instead use, for example,
\begin{codeenv}
\gls{glsxtrfmt}\marg{S}\marg{A}
\gls{glsxtrentryfmt}\marg{cardS}\marg{B}
\end{codeenv}
instead of:
\begin{verbatim}
\set{A}
\card{B}
\end{verbatim}
The first \LaTeX\ run will simply ignore the formatting and produce
a warning.

Since this is a bit cumbersome to write, you can provide shortcut
commands. For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[my-data]{src}}
\cs{newcommand}\marg{\cmd{gset}}[2][]\marg{\gls{glsxtrfmt}\oarg{\idx{param}1}\marg{S}\marg{\idx{param}2}}
\cs{newcommand}\marg{\cmd{gcard}}[2][]\marg{\gls{glsxtrfmt}\oarg{\idx{param}1}\marg{cardS}\marg{\idx{param}2}}
\end{codeenv}
Whilst this doesn't seem a great deal different from simply
providing the definitions of \csfmt{set} and \csfmt{card} in the
document, this means you don't have to worry about remembering 
the names of the actual commands provided in the \ext{bib} file
(just the entry labels) and the use of \gls{glsxtrfmt} will
automatically produce a hyperlink to the glossary entry if the
\isty{hyperref} package has been loaded.
\end{enumerate}

Here's an alternative \ext{bib} that defines entries with a term, a
description and a symbol:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cmd{setfmt}}[1]\marg{\cs{mathcal}\marg{\idx{param}1}}
\cs{providecommand}\marg{\cmd{cardfmt}}[1]\marg{|\cmd{setfmt}\marg{\idx{param}1}|}}}
\strut
\atentry{entry}\marg{set,
  \field{name}=\marg{set},
  \field{symbol}=\marg{\cmd{setfmt}\marg{S}},
  \field{user1}=\marg{setfmt},
  \field{description}=\marg{collection of values}
}
\atentry{entry}\marg{cardinality,
  \field{name}=\marg{cardinality},
  \field{symbol}=\marg{\cmd{cardfmt}\marg{S}},
  \field{user1}=\marg{cardfmt},
  \field{description}=\marg{the number of elements in the \cs{gls}\marg{set} \idx{mshiftchar}\cs{glssymbol}\marg{set}\idx{mshiftchar}}
}
\end{codeenv}
I've changed the entry labels and the names of the formatting commands.
The definitions in the document need to reflect the change in label
but not the change in the formatting commands:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{gset}}[2][]\marg{\gls{glsxtrfmt}\oarg{\idx{param}1}\marg{set}\marg{\idx{param}2}}
\cs{newcommand}\marg{\cmd{gcard}}[2][]\marg{\gls{glsxtrfmt}\oarg{\idx{param}1}\marg{cardinality}\marg{\idx{param}2}}
\end{codeenv}

Here's another approach that allows for a more complicated argument
for the cardinality. (For example, if the argument is an expression
involving set unions or intersections.)
The \ext{bib} file is now:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cmd{setfmt}}[1]\marg{\cs{mathcal}\marg{\idx{param}1}}
\cs{providecommand}\marg{\cmd{cardfmt}}[1]\marg{|\idx{param}1|}}}
\strut
\atentry{entry}\marg{set,
  \field{name}=\marg{set},
  \field{symbol}=\marg{\cmd{setfmt}\marg{S}},
  \field{user1}=\marg{setfmt},
  \field{description}=\marg{collection of values}
}
\atentry{entry}\marg{cardinality,
  \field{name}=\marg{cardinality},
  \field{symbol}=\marg{\cmd{cardfmt}\marg{\cmd{setfmt}\marg{S}}},
  \field{user1}=\marg{cardfmt},
  \field{description}=\marg{the number of elements in the \cs{gls}\marg{set} \idx{mshiftchar}\cs{glssymbol}\marg{set}\idx{mshiftchar}}
}
\end{codeenv}
This has removed the \csfmt{setfmt} command from the definition of
\csfmt{cardfmt}. Now the definitions in the document are:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{gset}}[1]\marg{\gls{glsxtrentryfmt}\marg{set}\marg{\idx{param}1}}
\cs{newcommand}\marg{\cmd{gcard}}[2][]\marg{\gls{glsxtrfmt}\oarg{\idx{param}1}\marg{cardinality}\marg{\idx{param}2}}
\end{codeenv}
This allows for code such as:
\begin{verbatim}
\[ \gcard{\gset{A} \cap \gset{B}} \]
\end{verbatim}
which will link back to the \code{cardinality} entry in the
glossary and avoids any hyperlinking with \csfmt{gset}.
Alternatively to avoid links with \csfmt{gcard} as well:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{gset}}[1]\marg{\gls{glsxtrentryfmt}\marg{set}\marg{\idx{param}1}}
\cs{newcommand}\marg{\cmd{gcard}}[1]\marg{\gls{glsxtrentryfmt}\marg{cardinality}\marg{\idx{param}1}}
\end{codeenv}
Now \csfmt{gset} and \csfmt{gcard} are simply formatting commands,
but their actual definitions are determined in the \ext{bib} file.

\section{Single Entry Types}
\label{sec:singleentry}

The entry types described in this section create a single glossary
definition per entry (from \styfmt{glossaries-extra}'s point of view). 
For example:
\begin{codeenv}
\atentry{entry}\marg{matrix,
  \field{name}=\marg{matrix},
  \field{plural}=\marg{matrices},
  \field{description}=\marg{rectangular array of values}
}
\end{codeenv}
is analogous to:
\begin{codeenv}
\gls{newglossaryentry}\marg{matrix}\comment{label}
\marg{\comment{fields}
  \field{name}=\marg{matrix},
  \field{plural}=\marg{matrices},
  \field{description}=\marg{rectangular array of values}
}
\end{codeenv}
The \csopt{secondary} option allows the creation of a fake glossary
with the entry labels in its internal list in a different order.
This means that the same data can be displayed in two separate lists
without duplicating the resources required by each glossary entry.

\Sectionref{sec:dualentry} describes \bibgls\ entry types
that create two separate (but related) \styfmt{glossaries-extra}
definitions per \ext{bib} entry.

\entrysection{entry}

Regular terms are defined by the \atentry{entry} field. This requires the
\field{description} field and either \field{name} or \field{parent}.
For example:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cmd{mtx}}[1]\marg{\cs{boldsymbol}\marg{\idx{param}1}}}}
\strut
\atentry{entry}\marg{matrix,
  \field{name}=\marg{matrix},
  \field{plural}=\marg{matrices},
  \field{description}=\marg{rectangular array of values, denoted \cs{gls}\marg{M}},
  \field{seealso}=\marg{vector}
}
\strut
\atentry{entry}\marg{M,
  \field{name}=\marg{\cs{ensuremath}\marg{M}},
  \field{description}=\marg{a \cs{gls}\marg{matrix}}
}
\strut
\atentry{entry}\marg{vector,
  \field{name} = \qtdelim{vector},
  \field{description} = \marg{column or row of values, denoted \cs{gls}\marg{v}},
  \field{seealso}=\marg{matrix}
}
\strut
\atentry{entry}\marg{v,
  \field{name}=\marg{\cs{ensuremath}\marg{\cs{vec}\marg{v}}},
  \field{description}=\marg{a \cs{gls}\marg{vector}}
}
\end{codeenv}

If the \field{sort} field is missing the default is obtained from
the \field{name} field (unless overridden by options like
\csopt{entry-sort-fallback}). For hierarchical entries, if the
\field{name} field is omitted it will be obtained from the parent's
\field{name}. 

Terms defined using \atentry{entry} will be written to the output
(\ext{glstex}) file using the command \gls!{bibglsnewentry}.

\entrysection{symbol}

The \atentry{symbol} entry type is much like \atentry{entry}, but it's
designed specifically for symbols, so in the previous example, the
\code{M} and \code{v} terms would be better defined using the
\atentry{symbol} entry type instead. For example:
\begin{codeenv}
\atentry{symbol}\marg{M,
  \field{name}=\marg{\cs{ensuremath}\marg{M}},
  \field{description}=\marg{a \cs{gls}\marg{matrix}}
}
\end{codeenv}

The required fields are \field{name} or \field{parent}. The
\field{description} field is required if the \field{name} field is
missing. If the \field{sort} field is omitted, the default sort is
given by the entry label (unless overridden by options like
\csopt{symbol-sort-fallback}). Note that this is different from
\atentry{entry} where the sort defaults to \field{name} if omitted.

Terms that are defined using \atentry{symbol} will be written to
the output file using the command \gls!{bibglsnewsymbol}.

\entrysection{number}

The \atentry{number} entry type is like \atentry{symbol}, but it's for
numbers. The numbers don't have to be explicit digits and may have a
symbolic representation. There's no real difference between the
behaviour of \atentry{number} and \atentry{symbol} except that terms
defined using \atentry{number} will be written to the output file
using the command \gls!{bibglsnewnumber}.

For example, the file \filefmt{constants.bib} might define
mathematical constants like this:
\begin{codeenv}
\atentry{number}\marg{pi,
   \field{name}=\marg{\cs{ensuremath}\marg{\ics{pi}}},
   \field{description}=\marg{the ratio of the length of the circumference
    of a circle to its diameter},
   \field{user1}=\marg{3.14159}
}
\strut
\atentry{number}\marg{e,
  \field{name}=\marg{\cs{ensuremath}\marg{e}},
  \field{description}=\marg{base of natural logarithms},
  \field{user1}=\marg{2.71828}
}
\end{codeenv}
This stores the approximate value in the \field{user1} field. This
can be used to sort the entries in numerical order according to the
values rather than the symbols:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[constants]{src},\comment{constants.bib}
  \csopt[number]{category},\comment{set the category for all selected entries}
  \csopt[double]{sort},\comment{numerical double-precision sort}
  \csopt[user1]{sort-field}\comment{sort according to 'user1' field}
}
\end{codeenv}
The \csopt[number]{category} option makes it easy to adjust the
glossary format to include the \field{user1} field:
\begin{codeenv}
\ics{glsdefpostdesc}\marg{number}\marg{\comment{}
  \cs{ifglshasfield}\marg{useri}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{ (approximate value: \cs{glscurrentfieldvalue})}\comment{}
  \marg{}\comment{}
}
\end{codeenv}

\entrysection{index}

The \atentry{index} entry type is designed for entries that don't
have a description. Only the label is required. If \field{name} is
omitted, it's assumed to be the same as the label, even if
\field{parent} is present. (Note this is different to the fallback
behaviour of \atentry{entry}, which fetches the name from the parent
entry.) If the name contains any characters that can't be used in
the label, you must use the \field{name} field. If the \field{sort}
field is missing the default is obtained from the \field{name}.
Note that the \atentry{index} entry type is \emph{not} governed by
\csopt{entry-sort-fallback} (but it is governed by
\csopt{custom-sort-fallbacks}). This allows \atentry{index} and
\atentry{entry} to have different fallbacks if the \field{sort}
field is missing.

Example:
\begin{codeenv}
\atentry{index}\marg{duck}
\atentry{index}\marg{goose,\field{plural}=\marg{geese}}
\atentry{index}\marg{sealion,\field{name}=\marg{sea lion}}
\atentry{index}\marg{facade,\field{name}=\marg{fa\ics{c}\marg{c}ade}}
\end{codeenv}

Terms that are defined using \atentry{index} will be written to the output
file using the command \gls!{bibglsnewindex}.

\entrysection{indexplural}

The \atentry{indexplural} entry type is similar to the
\atentry{index} entry type except that the \field{name} field, if
missing, is obtained from the \field{plural} field. If the
\field{plural} field is missing it's obtained from the \field{text}
field with the plural suffix appended. If the \field{text} field is
missing, it's obtained from the original entry label. If the
\field{sort} field is missing the default is obtained from the
\field{name} field. (As with \atentry{index}, \atentry{indexplural}
is \emph{not} governed by \csopt{entry-sort-fallback}, but it is
governed by \csopt{custom-sort-fallbacks}.) 
All fields are optional. For example:
\begin{codeenv}
\atentry{indexplural}\marg{goose,
  \field{plural} = \marg{geese}
}
\strut
\atentry{indexplural}\marg{duck}
\strut
\atentry{indexplural}\marg{chateau,
  \field{text} = \marg{ch\ics{cs.circum}ateau},
  \field{plural} = \marg{ch\cs{cs.circum}ateaux}
}
\end{codeenv}
This is equivalent to:
\begin{codeenv}
\atentry{indexplural}\marg{goose,
  \field{name} = \marg{geese},
  \field{text} = \marg{goose},
  \field{plural} = \marg{geese}
}
\strut
\atentry{indexplural}\marg{duck,
  \field{name} = \marg{ducks},
  \field{text} = \marg{duck},
  \field{plural} = \marg{ducks}
}
\strut
\atentry{indexplural}\marg{chateau,
  \field{name} = \marg{ch\cs{cs.circum}ateaux},
  \field{text} = {ch\cs{cs.circum}ateau},
  \field{plural} = {ch\cs{cs.circum}ateaux}
}
\end{codeenv}

Terms that are defined using \atentry{indexplural} will be written to the output
file using the command \gls!{bibglsnewindexplural}.

\entrysection{abbreviation}

The \atentry{abbreviation} entry type is designed for abbreviations.
The required fields are \field{short} and \field{long}. If the
\field{sort} key is missing, \bibgls\ will use the field given by
\csopt{abbreviation-sort-fallback}, which defaults to the \field{short} field.
(If you want an equivalent of \gls!{newdualentry}, use
\atentry{dualabbreviationentry} instead.)

If you use \csopt[name]{sort-field} (rather than the default
\csopt[sort]{sort-field}), then the fallback for the
\field{name} field is always the \field{short} field, regardless of
the \csopt{abbreviation-sort-fallback} setting, unless you use
\csopt{abbreviation-name-fallback} to change the fallback for the
\field{name} field.

Note that you must set the abbreviation style before loading the
resource file to ensure that the abbreviations are defined
correctly, however \bibgls\ has no knowledge of the abbreviation
style so it doesn't know if the \field{description} field must be
included or if the default \field{sort} value isn't simply the value
of the \field{short} field.

You can instruct \bibgls\ to sort by the \field{long} field instead
using \csopt[long]{abbreviation-sort-fallback}. You can also tell \bibgls\ to
ignore certain fields using \csopt{ignore-fields}, so you can
include a \field{description} field in the \ext{bib} file if
you sometimes need it, and then instruct \bibgls\ to ignore it when
you don't want it.

For example:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
  \field{short} = \marg{html},
  \field{long}  = \marg{hypertext markup language},
  \field{description} = \marg{a markup language for creating web pages}
}
\end{codeenv}
If you want the \abbrstyle{long-noshort-desc} style, then you can put
the following in your document (where the \ext{bib} file is called
\filefmt{entries-abbrv.bib}):
\begin{codeenv}
\ics{setabbreviationstyle}\marg{long-noshort-desc}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-abbrv]{src},
 \csopt[long]{abbreviation-sort-fallback}}
\end{codeenv}
Whereas, if you want the \abbrstyle{long-short-sc} style, then you can
instead do:
\begin{codeenv}
\cs{setabbreviationstyle}\marg{long-short-sc}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-abbrv]{src},\csopt[description]{ignore-fields}}
\end{codeenv}
or to convert the short value to \idx{uppercase} and use the
\abbrstyle{long-short-sm} style instead:
\begin{codeenv}
\cs{setabbreviationstyle}\marg{long-short-sm}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-abbrv]{src},
 \csopt[uc]{short-case-change},\comment{convert short value to upper case}
 \csopt[description]{ignore-fields}}
\end{codeenv}

Case-changing can be applied with \csopt{short-case-change} to
convert the case of the \field{short} field, as illustrated above. 
If you use a style that obtains the \field{description} from the
\field{long} form, but you want to apply a case-change to the
\field{description} field with \csopt{description-case-change}, then
you can copy the \field{long} field to the \field{description} with
\csopt[\field{long}=\field{description}]{replicate-fields}.

For example, if \filefmt{entries-abbrv.bib} contains:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
  \field{short} = \marg{html},
  \field{long}  = \marg{hypertext markup language}
}
\end{codeenv}
then the document may include:
\begin{codeenv}
\cs{setabbreviationstyle}\marg{long-short-sc}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-abbrv]{src},
 \csopt[firstuc]{description-case-change},
 \csopt[\field{long}=\field{description}]{replicate-fields}}
\end{codeenv}
Note that this can cause a problem for styles that set the
\field{description} field to the \field{long} form encapsulated by a
style command (such as with the \abbrstyle{long-em-short-em} style)
as this will override the style setting.

Similarly, if you want to change the case of the \field{name} field:
\begin{codeenv}
\cs{setabbreviationstyle}\marg{long-short-sc}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-abbrv]{src},
 \csopt[firstuc]{description-case-change},
 \csopt[uc]{name-case-change},
 \csopt[\field{long}=\field{description},\field{short}=\field{name}]{replicate-fields}}
\end{codeenv}
Again, this will lose any custom formatting command that would
usually be applied by the abbreviation style to the \field{name}
field (and \field{description}, if applicable).

Terms defined using \atentry{abbreviation} will be written to the output
file using the command \gls!{bibglsnewabbreviation}.

\entrysection{acronym}

The \atentry{acronym} entry type is like \atentry{abbreviation} except that
the term is written to the output file using the command
\gls!{bibglsnewacronym}.

\entrysection{contributor}

The \atentry{contributor} entry type is primarily provided for use 
by the \atentry{bibtexentry} type. You may use it explicitly
if you want, but you need to take care that it doesn't clash with
\atentry{bibtexentry}. It behaves much like \atentry{index} except that the
term is written to the \ext{glstex} file using the command
\gls!{bibglsnewcontributor}. There are no required fields. As with
\atentry{index}, if the \field{name} field is missing, the fallback
value is the entry's label. When this entry type is automatically
created by \atentry{bibtexentry}, the \field{name} is set to 
\begin{codeenv}
\gls{bibglscontributor}\margm{forenames}\margm{von}\margm{surname}\margm{suffix}
\end{codeenv}

If you do explicitly use \atentry{contributor} you need to make sure it's
defined \emph{before} the first instance of \atentry{bibtexentry} that
tries to access it, but within the same resource set. If you ensure that 
the label of \atentry{contributor} matches the contributor label generated by
\atentry{bibtexentry} then they can have their dependency lists
updated, and
the \field{bibtexentry} and \field{bibtexentry@entrytype} internal
fields can be set for the
\atentry{contributor} entry.  For example:
\begin{codeenv}
\atentry{contributor}\marg{KnuthDonaldE,
  \field{name}=\marg{\gls{bibglscontributor}\marg{Donald E.}\marg{}\marg{Knuth}\marg{}},
  \field{description}=\marg{Famous mathematician and computer scientist who
   created \cmd{TeX}}
}
\strut
@book\marg{texbook,
  title = \marg{The {\cmd{TeX} book}},
  author = \marg{Donald E. Knuth},
  publisher = \marg{Addison-Wesley},
  year = 1986
}
\end{codeenv}
The resource options then need to include:
\begin{codeenv}
 \csopt[\ics{GlsXtrBibTeXEntryAliases}]{entry-type-aliases},
 \csopt[
  \marg{[ \cs{cs.string}\csfmt{-}\cs{cs.string}\csfmt{.}]}\marg{}
 ]{labelify-replace}
\end{codeenv}

If the \atentry{contributor} entry is deferred until after the
corresponding \atentry{bibtexentry} then you will end up with a label clash.

\section{Dual Entry Types}
\label{sec:dualentry}

The entry types described in this section create two separate (but
related) \styfmt{glossaries-extra} entry definitions per \ext{bib}
entry. The first of these entries is considered the
\idx{primary} entry, and the second is the \idx{dual} entry
(also referred to as the \idx{secondary} entry, but is not
related to the \csopt{secondary} option). The naming scheme is
\code{@dual}\meta{entry-type} where both the primary and dual are
considered to have the same type of entry (such as
\atentry{dualsymbol} where both the primary and dual are
functionally like \atentry{symbol}) or
\code{@dual}\meta{primary}\meta{dual} where the primary
is functionally like \code{@}\meta{primary} and the dual is
functionally like \code{@}\meta{dual}.

If the fields provided by the \isty{glossaries-prefix} are defined,
there will be additional mappings for the special internal fields
\field{dualprefix}, \field{dualprefixfirst},
\field{dualprefixplural}, and \field{dualprefixfirstplural}.

For example:
\begin{codeenv}
\atentry{dualabbreviationentry}\marg{svm,
  \field{short} = \marg{SVM},
  \field{long} = \marg{support vector machine},
  \field{description} = \marg{statistical pattern recognition technique}
}
\end{codeenv}
is like:
\begin{codeenv}
\atentry{abbreviation}\marg{svm,
  \field{short} = \marg{SVM},
  \field{long} = \marg{support vector machine},
}
\atentry{entry}\marg{dual.svm,
  \field{text} = \marg{SVM},
  \field{name} = \marg{support vector machine},
  \field{description} = \marg{statistical pattern recognition technique}
}
\end{codeenv}
and is analogous to:
\begin{codeenv}
\gls{newabbreviation}\marg{svm}\marg{SVM}\marg{support vector machine}
\gls{newglossaryentry}\marg{dual.svm}\marg{\field{name}=\marg{support vector machine},\field{text}=\marg{SVM},
 \field{description}=\marg{statistical pattern recognition technique}}
\end{codeenv}
but both entries are considered dependent on each other. This means
that if you only reference the primary entry (using \ics{gls} etc)
then the dual entry will still be selected if the \csopt{selection}
setting includes dependencies.

The creation of the dual entry involves mapping or copying fields
from the primary entry. Each dual entry type has a set of mappings.
If a field in the set of mappings is missing, its fallback value is
used.  Any fields that aren't listed in the mappings are simply copied,
except for the \field{alias} field, which will never be copied to
the dual entry, nor can it be mapped. The alias will only apply to
the primary entry.  The dual entry is given the label
\meta{prefix}\meta{id} where \meta{prefix} is set by the
\csopt{dual-prefix} option and \meta{id} is the label supplied in
the \ext{bib} file.

If \csopt[combine]{dual-sort} then the dual entries will be sorted
along with the primary entries, otherwise the \csopt{dual-sort}
indicates how to sort the dual entries and the dual entries will be
appended to the end of the \ext{glstex} file. The
\csopt{dual-sort-field} determines what field to use for the sort
value if the dual entries should be sorted separately.

Take care if you have a mixture of entry types (such as 
\atentry{dualindexentry}, \atentry{dualindexsymbol} 
and \atentry{index}) and you're not
using the default \csopt[combine]{dual-sort}. Remember that the
primary entries are all sorted together along with the single
entries types described in \sectionref{sec:dualentry} (but they may
be assigned to different glossary types), and then the dual entries
are sorted together (but may be assigned to different glossary
types). This may result in an odd ordering if some of the primaries
and some of the duals are assigned to the same glossary. For
example, don't mix \atentry{dualindexabbreviation} (duals are
abbreviations) with \atentry{dualabbreviationentry} (primaries
are abbreviations) when you aren't using \csopt[combine]{dual-sort}
(unless you have two different glossaries for the primary vs dual
abbreviations).

Remember that \bibgls\ is designed to take advantage of
\ics{printunsrtglossary}, which simply iterates over all defined
entries in the order in which they were defined (or, more precisely,
the order of the internal list of entry labels associated with that
glossary). The aim of \bibgls\ is to write the entry definitions to
the \ext{glstex} file so that the internal list of labels is in the
appropriate order.

For example, suppose the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{index}\marg{aardvark}
\atentry{index}\marg{mouse}
\atentry{index}\marg{zebra}
\atentry{dualindexabbreviation}\marg{xml,
 \field{short}=\marg{XML},
 \field{long}=\marg{extensible markup language}
}
\atentry{dualabbreviationentry}\marg{ssi,
  \field{short}=\marg{SSI},
  \field{long}=\marg{server-side includes},
  \field{description}=\marg{directives placed in \cs{gls}\marg{html} pages
   evaluated by the server}
}
\atentry{dualindexabbreviation}\marg{html,
 \field{short}=\marg{HTML},
 \field{long}=\marg{hypertext markup language}
}
\atentry{dualabbreviationentry}\marg{css,
 \field{short}=\marg{CSS},
 \field{long}=\marg{cascading stylesheets},
 \field{description}=\marg{a language that describes the style of an
  \cs{gls}\marg{html} document}
}
\end{codeenv}
This contains a mixture of entry types, including
\atentry{dualindexabbreviation} (where the dual is the
abbreviation) and \atentry{dualabbreviationentry} (where the
primary is the abbreviation).

Now consider the following document:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[\styopt{record},\styopt{abbreviations}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[all]{selection},\csopt[entries]{src}}
\strut
\cmd{begin}\marg{document}
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
This uses the default \csopt[combine]{sort}, so all the entries are
sorted together, resulting in the order: 
\code{aardvark},
\code{dual.css},
\code{css},
\code{html},
\code{dual.html},
\code{mouse},
\code{dual.ssi},
\code{ssi},
\code{xml},
\code{dual.xml},
\code{zebra}.

The \LaTeX\ code written to the \iext{glstex} file is essentially
(but not exactly):
\begin{codeenv}
\comment{from \atentry{index}\marg{aardvark}:}
\gls{newglossaryentry}\marg{aardvark}\marg{\field{name}=\marg{aardvark},\field{description}=\marg{}}
\strut
\comment{dual of \atentry{dualabbreviationentry}\marg{css,\ldots}:}
\gls{newglossaryentry}\marg{dual.css}\marg{\field{name}=\marg{cascading stylesheets},\marg{text}=\marg{CSS},
 \field{description}=\marg{a language that describes the style of an
  \cs{glsxtrshort}\marg{html} document}}
\strut
\comment{primary of \atentry{dualabbreviationentry}\marg{css,\ldots}:}
\gls{newabbreviation}\marg{css}\marg{CSS}\marg{cascading stylesheets}
\strut
\comment{primary of \atentry{dualindexabbreviation}\marg{html,\ldots}:}
\gls{newglossaryentry}\marg{html}\marg{\field{name}=\marg{HTML},\field{description}=\marg{}}
\strut
\comment{dual of \atentry{dualindexabbreviation}\marg{html,\ldots}:}
\gls{newabbreviation}\marg{dual.html}\marg{HTML}\marg{hypertext markup language}
\strut
\comment{from \atentry{index}\marg{mouse}:}
\gls{newglossaryentry}\marg{mouse}{\marg{name}=\marg{mouse},\field{description}=\marg{}}
\strut
\comment{dual of \atentry{dualabbreviationentry}\marg{ssi,\ldots}:}
\gls{newglossaryentry}\marg{dual.ssi}\marg{\field{name}=\marg{server-side includes},\field{text}=\marg{SSI},
 \field{description}=\marg{directives placed in \cs{glsxtrshort}\marg{html} pages
 evaluated by the server}}
\strut
\comment{primary of \atentry{dualabbreviationentry}\marg{ssi,\ldots}:}
\gls{newabbreviation}\marg{ssi}\marg{SSI}\marg{server-side includes}
\strut
\comment{primary of \atentry{dualindexabbreviation}\marg{xml,\ldots}:}
\gls{newglossaryentry}\marg{xml}\marg{\field{name}=\marg{XML},\field{description}=\marg{}}
\strut
\comment{dual of \atentry{dualindexabbreviation}\marg{xml,\ldots}:}
\gls{newabbreviation}\marg{dual.xml}\marg{XML}\marg{extensible markup language}
\strut
\comment{from \atentry{index}\marg{zebra}:}
\gls{newglossaryentry}\marg{zebra}\marg{\field{name}=\marg{zebra},\field{description}=\marg{}}
\end{codeenv}
Since the document uses the \styopt{abbreviations} package option,
\gls{newabbreviation} automatically assigns the abbreviation 
to the \code{abbreviations} glossary (created through that package
option). This means that the \code{main} (default) glossary
contains the entries (in order): 
\begin{itemize}
\item \code{aardvark} (name: aardvark),
\item \code{dual.css} (name: cascading stylesheets), 
\item \code{html} (name: HTML),
\item \code{mouse} (name: mouse), 
\item \code{dual.ssi} (name: server-side includes), 
\item \code{xml} (name: XML),
\item \code{zebra} (name: zebra).
\end{itemize}
The \code{abbreviations} glossary contains:
\begin{itemize}
\item \code{css} (short: CSS), 
\item \code{dual.html} (short: HTML),
\item \code{ssi} (short: SSI), 
\item \code{dual.xml} (short: XML).
\end{itemize}

Since all the entries were combined and sorted together, the
resulting glossaries are both ordered alphabetically (using
\field{short} for the abbreviations and \field{name} for the rest),
but note that you need to take care when referencing the
abbreviations if you want to make use of the abbreviation style. You
need \verb|\gls{css}| and \verb|\gls{ssi}| for the primary
abbreviations created with \atentry{dualabbreviationentry} and
\verb|\gls{dual.html}| and \verb|\gls{dual.xml}| for the dual
abbreviations created with \atentry{dualindexabbreviation}. Also
the \field{name} of the primary\slash dual alternative of the
abbreviations is also inconsistent (short form for \code{html} and
\code{xml} and long form for \code{dual.css} and \code{dual.ssi}), 
as different field mappings are used.

If the document is changed so that the dual entries are now
sorted and written after all the primary entries have been dealt
with:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries]{src},
 \csopt[letter-nocase]{dual-sort},
 \csopt[all]{selection}
}
\end{codeenv}
then \bibgls\ first orders the primaries:
\begin{itemize}
\item \code{aardvark} (name: aardvark),
\item \code{css} (short: CSS),
\item \code{html} (name: HTML),
\item \code{mouse} (name: mouse),
\item \code{ssi} (short: SSI),
\item \code{xml} (name: XML),
\item \code{zebra} (name: zebra)
\end{itemize}
and writes them to the \ext{glstex} file
(functionally like):
\begin{codeenv}
\comment{from \atentry{index}\marg{aardvark}:}
\gls{newglossaryentry}\marg{aardvark}\marg{\field{name}=\marg{aardvark},\field{description}=\marg{}}
\strut
\comment{primary of \atentry{dualabbreviationentry}\marg{css,\ldots}:}
\gls{newabbreviation}\marg{css}\marg{CSS}\marg{cascading stylesheets}
\strut
\comment{primary of \atentry{dualindexabbreviation}\marg{html,\ldots}:}
\gls{newglossaryentry}\marg{html}\marg{\field{name}=\marg{HTML},\field{description}=\marg{}}
\strut
\comment{from \atentry{index}\marg{mouse}:}
\gls{newglossaryentry}\marg{mouse}\marg{\field{name}=\marg{mouse},\field{description}=\marg{}}
\strut
\comment{primary of \atentry{dualabbreviationentry}\marg{ssi,\ldots}:}
\gls{newabbreviation}\marg{ssi}\marg{SSI}\marg{server-side includes}
\strut
\comment{primary of \atentry{dualindexabbreviation}\marg{xml,\ldots}:}
\gls{newglossaryentry}\marg{xml}\marg{\field{name}=\marg{XML},\field{description}=\marg{}}
\strut
\comment{from \atentry{index}\marg{zebra}:}
\gls{newglossaryentry}\marg{zebra}\marg{\field{name}=\marg{zebra},\field{description}=\marg{}}
\end{codeenv}
Then \bibgls\ orders the duals:
\begin{itemize}
\item \code{dual.css} (name: cascading stylesheets),
\item \code{dual.html} (short: HTML),
\item \code{dual.ssi} (name: server-side includes),
\item \code{dual.xml} (short: XML)
\end{itemize}
and writes them to the \ext{glstex} file
(functionally like):
\begin{codeenv}
\comment{dual of \atentry{dualabbreviationentry}\marg{css,\ldots}:}
\gls{newglossaryentry}\marg{dual.css}\marg{\field{name}=\marg{cascading stylesheets},\field{text}=\marg{CSS},
 \field{description}=\marg{a language that describes the style of an
  \cs{glsxtrshort}\marg{html} document}}
\strut
\comment{dual of \atentry{dualindexabbreviation}\marg{html,\ldots}:}
\gls{newabbreviation}\marg{dual.html}\marg{HTML}\marg{hypertext markup language}
\strut
\comment{dual of \atentry{dualabbreviationentry}\marg{ssi,\ldots}:}
\gls{newglossaryentry}\marg{dual.ssi}\marg{\field{name}=\marg{server-side includes},\field{text}=\marg{SSI},
 \field{description}=\marg{directives placed in \cs{glsxtrshort}\marg{html} pages
 evaluated by the server}}
\strut
\comment{dual of \atentry{dualindexabbreviation}\marg{xml,\ldots}:}
\gls{newabbreviation}\marg{dual.xml}\marg{XML}\marg{extensible markup language}
\end{codeenv}
When the \ext{glstex} file is input (during the next \LaTeX\ run) 
the entries are defined in the order:
\begin{enumerate}
\item \code{aardvark} (type: \code{main}),
\item \code{css} (type: \code{abbreviations}),
\item \code{html} (type: \code{main}),
\item \code{mouse} (type: \code{main}),
\item \code{ssi} (type: \code{abbreviations}),
\item \code{xml} (type: \code{main}),
\item \code{zebra} (type: \code{main}),
\item \code{dual.css} (type: \code{main}),
\item \code{dual.html} (type: \code{abbreviations}),
\item \code{dual.ssi} (type: \code{main}),
\item \code{dual.xml} (type: \code{abbreviations}).
\end{enumerate}
This means that the \code{main} glossary's internal list is in the
order:
\begin{itemize}
\item \code{aardvark} (aardvark),
\item \code{html} (HTML),
\item \code{mouse} (mouse),
\item \code{xml} (XML),
\item \code{zebra} (zebra), 
\item \code{dual.css} (cascading stylesheets), 
\item \code{dual.ssi} (server-side includes)
\end{itemize}
and the \code{abbreviations} glossary's internal list is in the order:
\begin{itemize}
\item \code{css} (CSS),
\item \code{ssi} (SSI),
\item \code{dual.html} (HTML), 
\item \code{dual.xml} (XML).
\end{itemize}
The lists are no longer in alphabetical order as they have a mixture
of primary and dual entries that were separated before sorting.

The above is a fairly contrived example as it wouldn't make sense
in a real document to have glossary terms (that include a
description) mixed with index terms (that don't include a
description). 
A better solution would be to use 
\atentry{tertiaryindexabbreviationentry} instead of
\atentry{dualabbreviationentry}.

\entrysection{dualentry}

The \atentry{dualentry} entry type is similar to \atentry{entry} but
actually defines two entries.
The dual entry contains the same information as the primary entry
but some of the fields are swapped around.
The default mappings are:
\begin{itemize}
\item \field{name} $\mapsto$ \field{description}
\item \field{plural} $\mapsto$ \field{descriptionplural}
\item \field{description} $\mapsto$ \field{name}
\item \field{descriptionplural} $\mapsto$ \field{plural}
\end{itemize}
If the prefix fields are defined, then the default mappings
additionally include:
\begin{itemize}
\item \field{prefix} $\mapsto$ \field{dualprefix}
\item \field{prefixplural} $\mapsto$ \field{dualprefixplural}
\item \field{prefixfirst} $\mapsto$ \field{dualprefixfirst}
\item \field{prefixfirstplural} $\mapsto$ \field{dualprefixfirstplural}
\item \field{dualprefix} $\mapsto$ \field{prefix}
\item \field{dualprefixplural} $\mapsto$ \field{prefixplural}
\item \field{dualprefixfirst} $\mapsto$ \field{prefixfirst}
\item \field{dualprefixfirstplural} $\mapsto$ \field{prefixfirstplural}
\end{itemize}
The required fields are as for \atentry{entry}.

For example:
\begin{codeenv}
\atentry{dualentry}\marg{child,
  \field{name}=\marg{child},
  \field{plural}=\marg{children},
  \field{description}=\marg{enfant}
}
\end{codeenv}
is like:
\begin{codeenv}
\atentry{entry}\marg{child,
  \field{name}=\marg{child},
  \field{plural}=\marg{children},
  \field{description}=\marg{enfant}
  \field{descriptionplural}=\marg{enfants}
}
\strut
\atentry{entry}\marg{dual.child,
  \field{description}=\marg{child},
  \field{descriptionplural}=\marg{children},
  \field{name}=\marg{enfant}
  \field{plural}=\marg{enfants}
}
\end{codeenv}
where \idprefix{dual} is replaced by the value of the
\csopt{dual-prefix} option. However, instead of defining the entries
with \csfmt{bibglsnewentry} both the primary and dual entries are
defined using \gls!{bibglsnewdualentry}. The \field{category} 
and \field{type} fields can be set for the dual entry using the
\csopt{dual-category} and \csopt{dual-type} options.

For example:
\begin{codeenv}
\cs{newglossary*}\marg{english}\marg{English}
\cs{newglossary*}\marg{french}\marg{French}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries-dual]{src},\comment{data in entries-dual.bib}
 \csopt[english]{type},\comment{put primary entries in glossary 'english'}
 \csopt[french]{dual-type},\comment{put dual entries in glossary 'french'}
 \csopt[dictionary]{category},\comment{set the primary category to 'dictionary'}
 \csopt[dictionary]{dual-category},\comment{set the dual category to 'dictionary'}
 \csopt[en]{sort},\comment{sort primary entries according to language 'en'}
 \csopt[fr]{dual-sort}\comment{sort dual entries according to language 'fr'}
}
\end{codeenv}

\entrysection{dualindexentry}

There are no required fields.  The primary
entry behaves like \atentry{index} and the dual entry behaves
like \atentry{entry}.  The default field mapping is:
\begin{itemize}
\item \field{name} $\mapsto$ \field{name}
\end{itemize}
If the prefix fields are defined, then the default mappings
additionally include:
\begin{itemize}
\item \field{prefix} $\mapsto$ \field{dualprefix}
\item \field{prefixplural} $\mapsto$ \field{dualprefixplural}
\item \field{prefixfirst} $\mapsto$ \field{dualprefixfirst}
\item \field{prefixfirstplural} $\mapsto$ \field{dualprefixfirstplural}
\item \field{dualprefix} $\mapsto$ \field{prefix}
\item \field{dualprefixplural} $\mapsto$ \field{prefixplural}
\item \field{dualprefixfirst} $\mapsto$ \field{prefixfirst}
\item \field{dualprefixfirstplural} $\mapsto$ \field{prefixfirstplural}
\end{itemize}
This doesn't actually perform any swapping of fields, but it
provides the field used for backlinks (if
\csopt{dual-indexentry-backlink} is set). The reason that the primary
(rather than the dual) is like \atentry{index} is to allow the
primaries to merge with any \atentry{index} entries found in the
resource set, since glossary entries with descriptions are likely to 
be a subset of all indexed entries.

If no \field{name} is given, the dual entry is assigned the
(unprefixed) entry label. For example:
\begin{codeenv}
\atentry{dualindexentry}\marg{array,
  \field{description}=\marg{ordered list of values}
}
\end{codeenv}
This is effectively like:
\begin{codeenv}
\atentry{index}\marg{array}
\strut
\atentry{entry}\marg{dual.array,
  \field{name}=\marg{array},
  \field{description}=\marg{ordered list of values}
}
\end{codeenv}
The primary entries are defined using
\gls!{bibglsnewdualindexentry},
which by default sets the \field{category} to \optfmt{index}
(although this may be overridden, for example, by the \csopt{category} option).
The dual entries are defined with
\gls!{bibglsnewdualindexentrysecondary}.

This is the most convenient way of having an entry that's also
automatically indexed. For example, suppose the file
\filefmt{terms.bib} contains:
\begin{codeenv}
\atentry{index}\marg{duck}
\atentry{index}\marg{zebra}
\atentry{index}\marg{aardvark}
\end{codeenv}
and suppose the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{dualindexentry}\marg{array,
  \field{description}=\marg{ordered list of values}
}
\strut
\atentry{dualindexentry}\marg{vector,
  \field{name}=\marg{vector},
  \field{description}=\marg{column or row of values}
}
\strut
\atentry{dualindexentry}\marg{set,
  \field{description}=\marg{collection of values}
}
\strut
\atentry{dualindexentry}\marg{matrix,
  \field{plural}=\marg{matrices},
  \field{description}=\marg{rectangular array of values}
}
\end{codeenv}
These entries can be used in an example document that has an 
index and a glossary:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}[\styopt{record},\styopt{index},\styopt[mcols]{stylemods}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[terms,entries]{src},
 \csopt[index]{type},
 \csopt[idx.]{label-prefix},
 \csopt[gls.]{dual-prefix},
 \csopt[primary]{combine-dual-locations},
 \csopt[main]{dual-type}
}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{gls.array}, \cs{gls}\marg{gls.vector}, \cs{gls}\marg{gls.set}, \cs{gls}\marg{gls.matrix}.
\strut
\cs{gls}\marg{idx.duck}, \cs{gls}\marg{idx.aardvark}, \cs{gls}\marg{idx.zebra}.
\strut
\cs{renewcommand}\marg{\cs{glstreenamefmt}}[1]\marg{\ics{textsc}\marg{\idx{param}1}}
\cs{printunsrtglossary}\oarg{\printglossopt[main]{type},\printglossopt[index]{style},\iprintglossopt{nogroupskip}}
\strut
\cs{renewcommand}\marg{\cs{glstreenamefmt}}[1]\marg{\idx{param}1}
\cs{renewcommand}\marg{\cs{glstreegroupheaderfmt}}[1]\marg{\cs{textbf}\marg{\idx{param}1}}
\cs{printunsrtglossary}\oarg{\printglossopt[index]{type},\printglossopt[mcolindexgroup]{style}}
\cmd{end}\marg{document}
\end{codeenv}
This uses \csopt{combine-dual-locations} to combine the locations
for the primary and dual entries so that they only appear in the
index.

To avoid the inconvenience of remembering which prefix to use, you can
set up the prefixes with \ics{glsxtraddlabelprefix} and reference entries
with \ics{dgls}, \ics{dGls} etc instead of \cs{gls}, \cs{Gls} etc.

\entrysection{dualindexabbreviation}

The \atentry{dualindexabbreviation} entry type is similar to 
\atentry{dualindexentry} and again, by default, the
field mapping is:
\begin{itemize}
\item \field{name} $\mapsto$ \field{name}
\end{itemize}
If the prefix fields are defined, then the default mappings
additionally include:
\begin{itemize}
\item \field{prefix} $\mapsto$ \field{dualprefix}
\item \field{prefixplural} $\mapsto$ \field{dualprefixplural}
\item \field{prefixfirst} $\mapsto$ \field{dualprefixfirst}
\item \field{prefixfirstplural} $\mapsto$ \field{dualprefixfirstplural}
\item \field{dualprefix} $\mapsto$ \field{prefix}
\item \field{dualprefixplural} $\mapsto$ \field{prefixplural}
\item \field{dualprefixfirst} $\mapsto$ \field{prefixfirst}
\item \field{dualprefixfirstplural} $\mapsto$ \field{prefixfirstplural}
\end{itemize}
However in this case the required fields are \field{short} and
\field{long}. The \field{name} for the primary entry defaults to
\field{short} if omitted. (This may be changed with the
\csopt{abbreviation-name-fallback} option.) The fallback for the
\field{sort} field is given by \csopt{abbreviation-sort-fallback},
which defaults to the \field{short} field.

For example:
\begin{codeenv}
\atentry{dualindexabbreviation}\marg{html,
  \field{short} = \marg{HTML},
  \field{long}  = \marg{hypertext markup language}
}
\end{codeenv}
is like:
\begin{codeenv}
\atentry{index}\marg{html,\field{name}=\marg{HTML}}
\strut
\atentry{abbreviation}\marg{dual.html,
  \field{short} = \marg{HTML},
  \field{long}  = \marg{hypertext markup language}
}
\end{codeenv}
The primary term is defined using
\gls!{bibglsnewdualindexabbreviation}, which encapsulates the
\field{name} to match the font used by
the dual abbreviation. The encapsulation command depends
on the \csopt{abbreviation-name-fallback} value. If it's
the \field{short} field then \gls{bibglsuseabbrvfont} is
used, otherwise \gls{bibglsuselongfont} is used.

The primary definition also
by default sets the \field{category} to \optfmt{index} (although
this again may be overridden).
The dual term is defined using
\gls!{bibglsnewdualindexabbreviationsecondary}.

\entrysection{dualindexsymbol}

The \atentry{dualindexsymbol} entry type is similar to 
\atentry{dualindexentry}, but by default the
field mappings are:
\begin{itemize}
\item \field{symbol} $\mapsto$ \field{name}
\item \field{name} $\mapsto$ \field{symbol}
\item \field{symbolplural} $\mapsto$ \field{plural}
\item \field{plural} $\mapsto$ \field{symbolplural}
\end{itemize}
If the prefix fields are defined, then the default mappings
additionally include:
\begin{itemize}
\item \field{prefix} $\mapsto$ \field{dualprefix}
\item \field{prefixplural} $\mapsto$ \field{dualprefixplural}
\item \field{prefixfirst} $\mapsto$ \field{dualprefixfirst}
\item \field{prefixfirstplural} $\mapsto$ \field{dualprefixfirstplural}
\item \field{dualprefix} $\mapsto$ \field{prefix}
\item \field{dualprefixplural} $\mapsto$ \field{prefixplural}
\item \field{dualprefixfirst} $\mapsto$ \field{prefixfirst}
\item \field{dualprefixfirstplural} $\mapsto$ \field{prefixfirstplural}
\end{itemize}
The required field is: \field{symbol}.
If the \field{name} field is omitted, the dual entry is assigned a
symbol from the original (unprefixed) label.
The primary entries are defined using
\gls!{bibglsnewdualindexsymbol},
which by default sets the \field{category} to \optfmt{index},
and the dual entries are defined using
\gls!{bibglsnewdualindexsymbolsecondary},
which by default sets the \field{category} to \optfmt{symbol}.
For example:
\begin{codeenv}
\atentry{dualindexsymbol}\marg{pi,
  \field{symbol}=\marg{\cs{ensuremath}\marg{\ics{pi}}},
  \field{description}=\marg{ratio of a circle's circumference to its diameter}
}
\end{codeenv}
is like:
\begin{codeenv}
\atentry{index}\marg{pi,\field{symbol}=\marg{\cs{ensuremath}\marg{\cs{pi}}}}
\strut
\atentry{symbol}\marg{dual.pi,
  \field{name}=\marg{\cs{ensuremath}\marg{\cs{pi}}},
  \field{symbol}=\marg{pi},
  \field{description}=\marg{ratio of a circle's circumference to its diameter}
}
\end{codeenv}

For example, suppose I have a file called \filefmt{symbols.bib} that
contains:
\begin{codeenv}
\atentry{dualindexsymbol}\marg{pi,
  \field{symbol}=\marg{\cs{ensuremath}\marg{\cs{pi}}},
  \field{description}=\marg{ratio of a circle's circumference to its diameter}
}
\strut
\atentry{dualindexsymbol}\marg{e,
  \field{name}=\marg{Euler's number},
  \field{symbol}=\marg{\cs{ensuremath}\marg{e}},
  \field{description}=\marg{base of the natural logarithm}
}
\end{codeenv}
Then the previous example document can be modified to have an index,
a glossary and a list of symbols:
\begin{codeenv}
\cmd{documentclass}\marg{report}
\strut
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}[\styopt{record},\styopt{symbols},\styopt{index},\styopt[mcols]{stylemods}]\marg{glossaries-extra}
\strut
\cs{newcommand}\marg{\gls{bibglsnewdualindexsymbolsecondary}}[5]\marg{\comment{}
  \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{category}=symbol,\comment{}
    \field{symbol}=\marg{\idx{param}4},\idx{param}2,\field{type}=\marg{symbols}}\marg{\idx{param}5}\comment{}
}
\strut
\cs{newcommand}\marg{\cmd{indexprimary}}[1]\marg{\ics{glsadd}\oarg{\glsaddopt[\encap{hyperbf}]{format}}\marg{idx.\idx{param}1}}
\strut
\cs{glsdefpostdesc}\marg{symbol}\marg{\cmd{indexprimary}\marg{\cs{glscurrententrylabel}}}
\cs{glsdefpostdesc}\marg{general}\marg{\cmd{indexprimary}\marg{\cs{glscurrententrylabel}}}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries,terms,symbols]{src},
 \csopt[index]{type},
 \csopt{set-widest},
 \csopt[idx.]{label-prefix},
 \csopt[\empty]{dual-prefix},
 \csopt[primary]{combine-dual-locations},
 \csopt[letter-case]{dual-sort},
 \csopt[main]{dual-type}
}
\strut
\gls{glsxtrnewglslike}\oarg{\glsopt[false]{hyper}}\marg{idx.}\marg{\cmd{idx}}\marg{\cmd{idxpl}}\marg{\cmd{Idx}}\marg{\cmd{Idxpl}}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{array}, \cs{gls}\marg{vector}, \cs{gls}\marg{set}, \cs{glspl}\marg{matrix}.
\cmd{idx}\marg{duck}, \cmd{idx}\marg{aardvark}, \cmd{idx}\marg{zebra}.
\cs{gls}\marg{e} and \cs{gls}\marg{pi}.
\strut
\cmd{newpage}
\cs{gls}\marg{array}, \cmd{idx}\marg{vector}, \cmd{idx}\marg{set}, \cs{gls}\marg{matrix}.
\strut
\cmd{newpage}
\cs{gls}\marg{array}, \cs{gls}\marg{vector}, \cs{gls}\marg{set}, \cs{gls}\marg{matrix}.
\strut
\cs{renewcommand}\marg{\cs{glstreenamefmt}}[1]\marg{\cs{textsc}\marg{\idx{param}1}}
\cs{printunsrtglossary}\oarg{\printglossopt[main]{type},\printglossopt{nogroupskip},\printglossopt[\glostyle{alttree}]{style}}
\strut
\cs{renewcommand}\marg{\cs{glstreenamefmt}}[1]\marg{\idx{param}1}
\cs{printunsrtglossary}\oarg{\printglossopt[symbols]{type},\printglossopt{nogroupskip},\printglossopt[index]{style}}
\strut
\cs{renewcommand}\marg{\cs{glstreenamefmt}}[1]\marg{\idx{param}1}
\cs{renewcommand}\marg{\cs{glstreegroupheaderfmt}}[1]\marg{\cs{textbf}\marg{\idx{param}1}}
\cs{printunsrtglossary}\oarg{\printglossopt[index]{type},\printglossopt[\glostyle{mcolindexgroup}]{style}}
\strut
\cmd{end}\marg{document}
\end{codeenv}
Here I've provided some convenient commands for referencing the
primary (index) terms (\csfmt{idx}, \csfmt{idxpl}, \csfmt{Idx}
and \csfmt{Idxpl}). This means I don't need to worry about the
label prefix and it also switches off the hyperlinks (with 
\code{\glsopt[false]{hyper}}). These custom
commands are defined using:
\nosecformatdef{glsxtrnewglslike}
which, in this case, essentially does:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{idx}}[2][]\marg{\cs{gls}\oarg{\glsopt[false]{hyper},\idx{param}1}\marg{idx.\idx{param}2}}
\cs{newcommand}\marg{\cmd{Idx}}[2][]\marg{\cs{Gls}\oarg{\glsopt[false]{hyper},\idx{param}1}\marg{idx.\idx{param}2}}
\cs{newcommand}\marg{\cmd{idxpl}}[2][]\marg{\cs{glspl}\oarg{\glsopt[false]{hyper},\idx{param}1}\marg{idx.\idx{param}2}}
\cs{newcommand}\marg{\cmd{Idxpl}}[2][]\marg{\cs{Glspl}\oarg{\glsopt[false]{hyper},\idx{param}1}\marg{idx.\idx{param}2}}
\end{codeenv}
but the new commands will also recognise the \ics{gls} modifiers, so
\verb|\idx+| will behave like \verb|\gls+| which wouldn't be
possible if \csfmt{idx} was defined using \ics{newcommand} in the
above manner. There's a similar command:
\nosecformatdef{glsxtrnewgls}
if no case-changing versions are required.

I've also redefined \gls!{bibglsnewdualindexsymbolsecondary} to put
the dual entries created with \atentry{dualindexsymbol} into
the \code{symbols} glossary (which is created with the
\styopt{symbols} package option), so it overrides the
\csopt[main]{dual-type} setting.

This command also sets the \field{category}
to \code{symbol}, so I can redefine the post-description hook
for symbols (\ics{glsxtrpostdescsymbol}) to automatically index 
the symbol definition. Similarly for the \code{general}
post-description hook \ics{glsxtrpostdescgeneral}. 

Since the post-description hook isn't done until the glossary has
been created, this requires a slightly longer build process. If the
document file is called \filefmt{myDoc.tex}, then the complete
document build is:
\begin{verbatim}
pdflatex myDoc
bib2gls -g myDoc
pdflatex myDoc
bib2gls -g myDoc
pdflatex myDoc
\end{verbatim}

As from \sty{glossaries-extra-bib2gls} version 1.37, an alternative
method is to identify possible label prefixes with
\ics{glsxtraddlabelprefix} or \ics{glsxtrprependlabelprefix} and use
\ics{dgls}, \ics{dglspl}, \ics{dGls} or \ics{dGlspl}. See the
\styfmt{glossaries-extra} user manual~\cite{glossaries-extra} for
further details.

\entrysection{dualindexnumber}

The \atentry{dualindexnumber} entry type is almost identical to
\atentry{dualindexsymbol}, but the primary entries are defined using
\gls!{bibglsnewdualindexnumber},
which by default sets the \field{category} to \optfmt{index},
and the dual entries are defined using
\gls!{bibglsnewdualindexnumbersecondary},
which by default sets the \field{category} to \optfmt{number}.

\entrysection{dualabbreviationentry}

The \atentry{dualabbreviationentry} entry type is similar to 
\atentry{dualentry}, but by default the
field mappings are:
\begin{itemize}
\item \field{long} $\mapsto$ \field{name}
\item \field{longplural} $\mapsto$ \field{plural}
\item \field{short} $\mapsto$ \field{text}
\end{itemize}
If the prefix fields are defined, then the default mappings
additionally include:
\begin{itemize}
\item \field{prefix} $\mapsto$ \field{dualprefix}
\item \field{prefixplural} $\mapsto$ \field{dualprefixplural}
\item \field{prefixfirst} $\mapsto$ \field{dualprefixfirst}
\item \field{prefixfirstplural} $\mapsto$ \field{dualprefixfirstplural}
\item \field{dualprefix} $\mapsto$ \field{prefix}
\item \field{dualprefixplural} $\mapsto$ \field{prefixplural}
\item \field{dualprefixfirst} $\mapsto$ \field{prefixfirst}
\item \field{dualprefixfirstplural} $\mapsto$ \field{prefixfirstplural}
\end{itemize}
You may need to add a mapping from \field{shortplural} to
\field{plural} if the default is inappropriate.
(In \bibgls\ version 1.0 this entry type was originally called
\atentry{dualentryabbreviation}. In version 1.1, it was renamed
\atentry{dualabbreviationentry} which makes for a more consistent naming scheme
\code{@dual}\meta{primary}\meta{dual}.)

The required fields are: \field{short}, \field{long} and
\field{description}. This entry type is designed to emulate the
example \gls{newdualentry} command given in the \sty{glossaries}
user manual~\cite{glossaries}. The primary entry is an abbreviation with the given
\field{short} and \field{long} fields (but not the 
\field{description}) and the secondary entry is a regular entry with
the \field{name} copied from the \field{long} field.
The fallback for the \field{sort} is given by
\csopt{abbreviation-sort-fallback}, which defaults to the
\field{short} field.

For example:
\begin{codeenv}
\atentry{dualabbreviationentry}\marg{svm,
   \field{long} = \marg{support vector machine},
   \field{short} = \marg{SVM},
   \field{description} = \marg{statistical pattern recognition technique}
}
\end{codeenv}
is rather like doing:
\begin{codeenv}
\atentry{abbreviation}\marg{svm,
   \field{long} = \marg{support vector machine},
   \field{short} = \marg{SVM}
}
\strut
\atentry{entry}\marg{dual.svm,
   \field{name} = \marg{support vector machine},
   \field{description} = \marg{statistical pattern recognition technique}
}
\end{codeenv}
but \code{dual.svm} will automatically be selected if \code{svm}
is indexed in the document. If \code{dual.svm} isn't explicitly
indexed, it won't have a location list.

If the \field{sort} field is missing \bibgls\ by default falls back
on the \field{name} field. If this is missing, this sort value will
fallback on the \field{short} field. This means that if \field{name}
isn't explicitly given in \atentry{dualabbreviationentry}, then the 
primary entry will be sorted according to \field{short} but the dual
will be sorted according its \field{name} (which has been copied
from the primary \field{long}).

Entries provided using \atentry{dualabbreviationentry} will be defined
with:
\begin{codeenv}
\gls{bibglsnewdualabbreviationentry}
\end{codeenv}
(which uses \gls{newabbreviation}) for the primary entries and with :
\begin{codeenv}
\gls{bibglsnewdualabbreviationentrysecondary}
\end{codeenv}
(which uses \gls{longnewglossaryentry}) for the secondary entries.
This means that if the \styopt{abbreviations}
package option is used, the primary entry will be put in the
\optfmt{abbreviations} glossary and the secondary entry in the
\optfmt{main} glossary.  Use the \csopt{type} and
\csopt{dual-type} options to override this.

\entrysection{dualentryabbreviation}

This entry type is deprecated as from \bibgls\ version 1.1. It's functionally
equivalent to \atentry{dualabbreviationentry} but its name
doesn't fit the general dual entry naming scheme.

\entrysection{dualsymbol}

This is like \atentry{dualentry} but the default mappings are:
\begin{itemize}
\item \field{name} $\mapsto$ \field{symbol}
\item \field{plural} $\mapsto$ \field{symbolplural}
\item \field{symbol} $\mapsto$ \field{name}
\item \field{symbolplural} $\mapsto$ \field{plural}
\end{itemize}
If the prefix fields are defined, then the default mappings
additionally include:
\begin{itemize}
\item \field{prefix} $\mapsto$ \field{dualprefix}
\item \field{prefixplural} $\mapsto$ \field{dualprefixplural}
\item \field{prefixfirst} $\mapsto$ \field{dualprefixfirst}
\item \field{prefixfirstplural} $\mapsto$ \field{dualprefixfirstplural}
\item \field{dualprefix} $\mapsto$ \field{prefix}
\item \field{dualprefixplural} $\mapsto$ \field{prefixplural}
\item \field{dualprefixfirst} $\mapsto$ \field{prefixfirst}
\item \field{dualprefixfirstplural} $\mapsto$ \field{prefixfirstplural}
\end{itemize}
The \field{name} and \field{symbol} fields are required.
For example:
\begin{codeenv}
\atentry{dualsymbol}\marg{pi,
   \field{name}=\marg{pi},
   \field{symbol}=\marg{\cs{ensuremath}\marg{\cs{pi}}},
   \field{description}=\marg{the ratio of the length of the circumference
    of a circle to its diameter}
}
\end{codeenv}

Entries are defined using \gls!{bibglsnewdualsymbol}, which by 
default sets the \field{category} to \optfmt{symbol}.

\entrysection{dualnumber}

This is almost identical to \atentry{dualsymbol} but entries are
defined using \gls!{bibglsnewdualnumber}, which by default sets
the \field{category} to \optfmt{number}.

The above example could be defined as a number since $\pi$ is a
constant:
\begin{codeenv}
\atentry{dualnumber}\marg{pi,
   \field{name}=\marg{pi},
   \field{symbol}=\marg{\cs{ensuremath}\marg{\cs{pi}}},
   \field{description}=\marg{the ratio of the length of the circumference
    of a circle to its diameter},
   \field{user1}=\marg{3.14159}
}
\end{codeenv}

This has stored the approximate value in the \field{user1} field.
The post-description hook could then be adapted to show this.
\begin{codeenv}
\cs{glsdefpostdesc}\marg{number}\marg{\comment{}
  \cs{ifglshasfield}\marg{useri}\marg{\cs{glscurrententrylabel}}
  \marg{ (approximate value: \cs{glscurrentfieldvalue})}\comment{}
  \marg{}\comment{}
}
\end{codeenv}

This use of the \field{user1} field means that the dual entries
could be sorted numerically according to the approximate value:
\begin{codeenv}
\cmd{usepackage}[\styopt{record},\styopt{postdot},\styopt{numbers},\styopt[index]{style}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{entries.bib}
  \csopt[numbers]{dual-type},
  \csopt[double]{dual-sort},\comment{decimal sort}
  \csopt[user1]{dual-sort-field}
}
\end{codeenv}

\entrysection{dualabbreviation}

The \atentry{dualabbreviation} entry type is similar to 
\atentry{dualentry}, but by default the
field mappings are:
\begin{itemize}
\item \field{short} $\mapsto$ \field{dualshort}
\item \field{shortplural} $\mapsto$ \field{dualshortplural}
\item \field{long} $\mapsto$ \field{duallong}
\item \field{longplural} $\mapsto$ \field{duallongplural}
\item \field{dualshort} $\mapsto$ \field{short}
\item \field{dualshortplural} $\mapsto$ \field{shortplural}
\item \field{duallong} $\mapsto$ \field{long}
\item \field{duallongplural} $\mapsto$ \field{longplural}
\end{itemize}
If the prefix fields are defined, then the default mappings
additionally include:
\begin{itemize}
\item \field{prefix} $\mapsto$ \field{dualprefix}
\item \field{prefixplural} $\mapsto$ \field{dualprefixplural}
\item \field{prefixfirst} $\mapsto$ \field{dualprefixfirst}
\item \field{prefixfirstplural} $\mapsto$ \field{dualprefixfirstplural}
\item \field{dualprefix} $\mapsto$ \field{prefix}
\item \field{dualprefixplural} $\mapsto$ \field{prefixplural}
\item \field{dualprefixfirst} $\mapsto$ \field{prefixfirst}
\item \field{dualprefixfirstplural} $\mapsto$ \field{prefixfirstplural}
\end{itemize}
The required fields are: \field{short}, \field{long}, 
\field{dualshort} and \field{duallong}.
This includes some new fields: \field{dualshort},
\field{dualshortplural}, \field{duallong} and
\field{duallongplural}. If these aren't already defined, they
will be provided in the \iext{glstex} file with
\begin{codeenv}
\ics{glsxtrprovidestoragekey}\margm{key}\marg{}\marg{}
\end{codeenv}
Note that this use with an empty third argument prevents
the creation of a field access command (analogous to
\ics{glsentrytext}). The value can be accessed with
\ics{glsxtrusefield} instead. Remember that the field won't be
available until the \ext{glstex} file has been created.

Note that \bibgls\ doesn't know what abbreviation styles
are in used, so if the \field{sort} field is missing
it will fallback on the \field{short} field. If the abbreviations
need to be sorted according to the \field{long} field instead,
use \csopt[long]{abbreviation-sort-fallback}.

Terms that are defined using \atentry{dualabbreviation} will be
written to the output file using \gls!{bibglsnewdualabbreviation}.

If the \csopt{dual-abbrv-backlink} option is on, the default field
used for the backlinks is the \field{dualshort} field, so you'll need
to make sure you adapt the glossary style to show that field. The
simplest way to do this is through the category \idx{postdescriptionhook}.

For example, if the entries all have the \field{category} set
to \optfmt{abbreviation}, then this requires redefining
\ics{glsxtrpostdescabbreviation} (either with \cs{renewcommand}
or via \cs{glsdefpostdesc}).

Here's an example dual abbreviation for a document where English is
the primary language and German is the secondary language:
\begin{codeenv}
\atentry{dualabbreviation}\marg{rna,
  \field{short}=\marg{RNA},
  \field{dualshort}=\marg{RNS},
  \field{long}=\marg{ribonucleic acid},
  \field{duallong}=\marg{Ribonukleins\"aure}
}
\end{codeenv}
If the abbreviation is in the file called
\filefmt{entries-dual-abbrv.bib}, then here's an example document:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[T1]\marg{fontenc}
\cmd{usepackage}[utf8]\marg{inputenc}
\strut
\cmd{usepackage}[ngerman,main=english]\marg{babel}
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}[\styopt{record},\styopt{nomain}]\marg{glossaries-extra}
\strut
\cs{newglossary*}\marg{english}\marg{English}
\cs{newglossary*}\marg{german}\marg{German}
\strut
\cs{setabbreviationstyle}\marg{long-short}
\strut
\cs{glsdefpostdesc}\marg{abbreviation}\marg{\comment{}
  \cs{ifglshasfield}\marg{dualshort}\marg{\cs{glscurrententrylabel}}
  \marg{\comment{}
    \cmd{space}(\cs{glscurrentfieldvalue})\comment{}
  }\comment{}
  \marg{}\comment{}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries-dual-abbrv]{src},\comment{entries-dual-abbrv.bib}
 \csopt[english]{type},\comment{put primary entries in glossary 'english'}
 \csopt[german]{dual-type},\comment{put primary entries in glossary 'german'}
 \csopt[en.]{label-prefix},\comment{primary label prefix}
 \csopt[de.]{dual-prefix},\comment{dual label prefix}
 \csopt[en]{sort},\comment{sort primary entries according to language 'en'}
 \csopt[de-1996]{dual-sort},\comment{sort dual entries according to 'de-1996'}
                     \comment{(German new orthography)}
 \csopt{dual-abbrv-backlink}\comment{add links in the glossary to the opposite entry}
}
\strut
\cmd{begin}\marg{document}
\strut
English: \cs{gls}\marg{en.rna}; \cs{gls}\marg{en.rna}.
\strut
German: \cs{gls}\marg{de.rna}; \cs{gls}\marg{de.rna}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
If the \csopt{label-prefix} is omitted, then only the dual entries
will have a prefix:
\begin{codeenv}
English: \cs{gls}\marg{rna}; \cs{gls}\marg{rna}.
\strut
German: \cs{gls}\marg{de.rna}; \cs{gls}\marg{de.rna}.
\end{codeenv}
Another variation is to use the \abbrstyle{long-short-user}
abbreviation style and modify the associated \ics{glsxtruserfield} so that
the \field{duallong} field is selected for the parenthetical
material:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtruserfield}}\marg{duallong}
\end{codeenv}
This means that the first use of the primary entry is displayed as
\begin{quote}
ribonucleic acid (RNA, Ribonukleins\"aure)
\end{quote}
and the first use of the dual entry is displayed as:
\begin{quote}
Ribonukleins\"aure (RNS, ribonucleic acid)
\end{quote}

Here's an example to be used with the \abbrstyle{long-short-desc}
style:
\begin{codeenv}
\atentry{dualabbreviation}\marg{rna,
  \field{short}=\marg{RNA},
  \field{dualshort}=\marg{RNS},
  \field{long}=\marg{ribonucleic acid},
  \field{duallong}=\marg{Ribonukleins\"aure}
  \field{description}=\marg{a polymeric molecule},
  \field{user1}=\marg{Ein polymeres Molek\"ul}
}
\end{codeenv}
This stores the dual description in the \field{user1} field,
so this needs a mapping.
The new example document is much the same as the previous one, except
that the \csopt{dual-abbrv-map} option is needed to include the 
mapping between the \field{description} and \field{user1} fields:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[T1]\marg{fontenc}
\cmd{usepackage}[utf8]\marg{inputenc}
\strut
\cmd{usepackage}[ngerman,main=english]\marg{babel}
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}[\styopt{record},\styopt{nomain}]\marg{glossaries-extra}
\strut
\cs{newglossary*}\marg{english}\marg{English}
\cs{newglossary*}\marg{german}\marg{German}
\strut
\cs{setabbreviationstyle}\marg{long-short-desc}
\strut
\cs{glsdefpostdesc}\marg{abbreviation}\marg{\comment{}
  \cs{ifglshasfield}\marg{dualshort}\marg{\cs{glscurrententrylabel}}
  \marg{\comment{}
    \cmd{space}(\cs{glscurrentfieldvalue})\comment{}
  }\comment{}
  \marg{}\comment{}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries-dual-abbrv-desc]{src},\comment{entries-dual-abbrv-desc.bib}
 \csopt[english]{type},\comment{put primary entries in glossary 'english'}
 \csopt[german]{dual-type},\comment{put primary entries in glossary 'german'}
 \csopt[en.]{label-prefix},\comment{primary label prefix}
 \csopt[de.]{dual-prefix},\comment{dual label prefix}
 \csopt[en]{sort},\comment{sort primary entries according to language 'en'}
 \csopt[long]{abbreviation-sort-fallback},\comment{fallback on 'long' field}
 \csopt[de-1996]{dual-sort},\comment{sort dual entries according to 'de-1996'}
                   \comment{(German new orthography)}
 \csopt{dual-abbrv-backlink},\comment{add links in the glossary to the opposite entry}
\comment{ dual key mappings:}
 \csopt[\comment{}
   \marg{short,shortplural,long,longplural,dualshort,dualshortplural,
     duallong,duallongplural,description,user1},
   \marg{dualshort,dualshortplural,duallong,duallongplural,short,shortplural,
     long,longplural,user1,description}
 ]{dual-abbrv-map}
}
\strut
\cmd{begin}\marg{document}
\strut
English: \cs{gls}\marg{en.rna}; \cs{gls}\marg{en.rna}.
\strut
German: \cs{gls}\marg{de.rna}; \cs{gls}\marg{de.rna}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
Note that since this document uses the \abbrstyle{long-short-desc}
abbreviation style, the \csopt{abbreviation-sort-fallback} needs to be changed
to \optfmt{long}.

If I change the order of the mapping to:
\begin{codeenv}
 \csopt[\comment{}
   \marg{long,longplural,short,shortplural,dualshort,dualshortplural,
     duallong,duallongplural,description,user1},
   \marg{duallong,duallongplural,dualshort,dualshortplural,short,shortplural,
     long,longplural,user1,description}
 ]{dual-abbrv-map}
\end{codeenv}
Then the back-link field will switch to \field{duallong}. The
\idx{postdescriptionhook} can be modified to allow for this:
\begin{codeenv}
\cs{glsdefpostdesc}\marg{abbreviation}\marg{\comment{}
  \cs{ifglshasfield}\marg{duallong}\marg{\cs{glscurrententrylabel}}
  \marg{\comment{}
    \cmd{space}(\cs{glscurrentfieldvalue})\comment{}
  }\comment{}
  \marg{}\comment{}
}
\end{codeenv}

An alternative is to use the \abbrstyle{long-short-user-desc} style
without the \idx!{postdescriptionhook}:
\begin{codeenv}
\cs{setabbreviationstyle}\marg{long-short-user-desc}
\cs{renewcommand}*\marg{\cs{glsxtruserfield}}\marg{duallong}
\end{codeenv}
However be careful with this approach as it can cause nested
hyperlinks. In this case it's better to use the
\abbrstyle{long-postshort-user-desc} style which defers the
parenthetical material until after the link-text:
\begin{codeenv}
\cs{setabbreviationstyle}\marg{long-postshort-user-desc}
\cs{renewcommand}*\marg{\cs{glsxtruserfield}}\marg{duallong}
\end{codeenv}
If the back-link field has been switched to \field{duallong} then
the \idx!{postdescriptionhook} is no longer required.

\entrysection{dualacronym}

As \atentry{dualabbreviation} but defines the entries with
\gls!{bibglsnewdualacronym}.

\section{Tertiary Entry Types}
\label{sec:tertiaryentry}

A tertiary entry type is essentially a dual entry that creates three
separate (but related) \styfmt{glossaries-extra} entry definitions per 
\ext{bib} entry. As with dual entries, the first and second of these
are the primary and secondary. The third of these is the
\idx{tertiary} which is effectively an appendage of the
secondary, and is defined by the same associated
\csfmt{bibglsnew\ldots secondary} command that defines the secondary
entry. Therefore the secondary and tertiary are both considered the
dual and are treated as a single entry for the purposes of sorting
and collating.

The tertiary entry will never have any locations. Any records found
will be assigned to the secondary (and may then be moved to the
primary with \csopt[primary]{combine-dual-locations}). The tertiary
will always have the same order as the secondary and will have the same
\field{group} value. You can set the \field{type} for the tertiary
with \csopt{tertiary-type} and the \field{category} with
\csopt{tertiary-category}. The label prefix defaults to
\idprefix{tertiary} and can be changed with \csopt{tertiary-prefix}.

\entrysection{tertiaryindexabbreviationentry}

This entry type is very similar to
\atentry{dualindexabbreviation} but creates a tertiary entry as
well.  The required fields are: \field{short} and \field{long} (as for
\atentry{dualindexabbreviation}) and also \field{description}.  The
mappings are shared by both entry types.  For example:
\begin{codeenv}
\atentry{tertiaryindexabbreviationentry}\marg{html,
  \field{short} = \marg{HTML},
  \field{long} = \marg{hypertext markup language},
  \field{description} = \marg{a markup language for creating web pages}
}
\end{codeenv}
is analogous to:
\begin{codeenv}
\gls{newglossaryentry}\marg{html,\field{name}=\marg{HTML},\field{description}=\marg{}}
\strut
\gls{newabbreviation}\marg{dual.html}\marg{HTML}\marg{hypertext markup language}
\strut
\gls{newglossaryentry}\marg{tertiary.html, 
  \field{name}=\marg{hypertext markup language},
  \field{description}=\marg{a markup language for creating web pages}
}
\end{codeenv}
The last two are actually defined using one command:
\begin{codeenv}
\gls{bibglsnewtertiaryindexabbreviationentrysecondary}
  \marg{dual.html}\comment{secondary label}
  \marg{tertiary.html}\comment{tertiary label}
  \marg{\ldots}\comment{secondary fields}
  \marg{\ldots}\comment{tertiary fields}
  \marg{HTML}\comment{primary name}
  \marg{HTML}\comment{short}
  \marg{hypertext markup language}\comment{long}
  \marg{a markup language for creating web pages}\comment{description}
\end{codeenv}
The \gls!{bibglsnewtertiaryindexabbreviationentrysecondary}
command is provided in the \iext{glstex} file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewtertiaryindexabbreviationentrysecondary}}[8]\marg{\comment{}
  \gls{newabbreviation}\oarg{\idx{param}3}\marg{\idx{param}1}\marg{\idx{param}6}\marg{\idx{param}7}\comment{}
  \cs{longnewglossaryentry*}\marg{\idx{param}2}\comment{}
  \marg{\field{name}=\marg{\cmd{protect}\gls{bibglsuselongfont}\marg{\idx{param}7}\marg{\ics{glscategory}\marg{\idx{param}1}}},\idx{param}4}\comment{}
  \marg{\idx{param}8}\comment{}
}
\end{codeenv}
which defines the secondary as an abbreviation using
\csfmt{newabbreviation} and the tertiary as a regular entry using
\csfmt{longnewglossaryentry}. This means that the tertiary entry is
always defined immediately after the corresponding secondary entry.
The primary may be defined earlier or later in the file depending on the
way the entries are sorted and on the \csopt{dual-sort} setting.

\section{Multi-Entry Types}
\label{sec:multientry}

A multi-entry type is an entry that may spawn multiple 
primary entries. This means that both the main entry and the
spawned entries are sorted together along with all the other primary
entries. In the case of \atentry{spawndualindexentry}, the main and
spawned entries are primary. The main entry's dual is created as per
\atentry{dualindexentry}.

\entrysection{bibtexentry}

The \atentry{bibtexentry} type will typically need to be aliased
as it's designed for converting \BibTeX\ entries into \bibgls\
entries. For example, to make \bibgls\ treat \atentryfmt{article}
and \atentryfmt{book} as though they were both
\atentry{bibtexentry}:
\begin{codeenv}
\csopt[
  article=bibtexentry,
  book=bibtexentry
]{entry-type-aliases}
\end{codeenv}
For convenience, \isty{glossaries-extra-bib2gls} v1.29+
provides \ics{GlsXtrBibTeXEntryAliases} which covers all the
standard \BibTeX\ entry types. Alternatively, you can use
\csopt[bibtexentry]{unknown-entry-alias} to alias all entries
that aren't recognised by \bibgls. If you use 
\csopt[same as original entry]{category}, the \field{category}
field will be set to the original entry type (for example,
\code{article} or \code{book}). Similarly you can use
\csopt[same as original entry]{type} to set the \field{type}
field (but remember that the glossary types will need to be defined
in the document).

There are no required fields. The fallback for the \field{sort}
field is given by \csopt{bibtexentry-sort-fallback}. If you want to
access any of the \BibTeX\ fields, you will need to alias or define
them.  For example:
\begin{codeenv}
\csopt[
  \fieldfmt{title}=\field{name}
]{field-aliases}
\end{codeenv}
Since \BibTeX's \fieldfmt{type} field conflicts with \bibgls's
\field{type} field, when \bibgls\ parses \atentry{bibtexentry}
if will convert \fieldfmt{type} to \field{bibtextype}, so you
must use \field{bibtextype} as the identifier when aliasing.

Alternatively, you can use \ics{GlsXtrProvideBibTeXFields} which
uses \ics{glsaddstoragekey} to provide all the standard \BibTeX\
fields. (Remember that new fields must be defined before the first 
\idx{resourceset}.)

The \atentry{bibtexentry} essentially creates an \atentry{index}
form of entry, but it additionally defines a \atentry{contributor}
entry for each listed author or editor
and updates the dependency lists: each \atentry{contributor} is
added to the main
\atentry{bibtexentry}'s list of dependencies (so if the
\atentry{bibtexentry} has a record then all its satellite
\atentry{contributor}s are selected with the default
\csopt[recorded and deps]{selection}), and 
each \atentry{contributor} is treated as having a cross-reference to
the main \atentry{bibtexentry} (so if a \atentry{contributor}
has a record then all the linked \atentry{bibtexentry} terms will
be selected if \csopt[recorded and deps and see]{selection}).
You can instruct \bibgls\ to treat \ics{citation} as an 
\idx{ignoredrecord} using \longarg{cite-as-record}.

Each contributor is effectively defined as:
\begin{codeenv}
\atentry{contributor}\marg{\meta{label},
  \field{name}=\marg{\gls{bibglscontributor}\margm{forenames}\margm{von}\margm{surname}\margm{suffix}}
}
\end{codeenv}
The label is obtained by converting the \field{name}
to a label, using the same function as \csopt{labelify} (which
means it's governed by \csopt{labelify-replace}).

The \fieldfmt{author} and \fieldfmt{editor} fields are always
checked, even if those fields aren't recognised by \bibgls, (which
they aren't by default). These checks are performed before field
aliases are applied. If neither field is present, no additional
entries are spawned. If the dependent \atentry{contributor} entry 
has already been defined, it won't be redefined, but will have the
new \atentry{bibtexentry} added to its internal \field{bibtexentry}
field.

The main \atentry{bibtexentry} is defined using
\gls{bibglsnewbibtexentry} and is followed by:
\begin{codeenv}
\gls{glsxtrfieldlistadd}\margm{id}\marg{bibtexcontributor}\margm{contributor-id}
\end{codeenv}
where \meta{id} is the label identifying the main
\atentry{bibtexentry} and \meta{contributor-id} is the
label identifying the contributor, for each contributor that has
been selected.

Each contributor is defined using \gls{bibglsnewcontributor}. The definition 
is followed by:
\begin{codeenv}
\gls{glsxtrfieldlistadd}\margm{contributor-id}\marg{bibtexentry}\margm{id}
\gls{glsxtrfieldlistadd}\margm{contributor-id}\marg{bibtexentry@\meta{entry-type}}\margm{id}
\end{codeenv}
for each selected \atentry{bibtexentry} associated with that contributor. The
second line provides the internal list field
\field{bibtexentry@entrytype}, where \meta{entry-type} is the
original entry type (before it was aliased to
\atentry{bibtexentry} and converted to \idx!{lowercase}).
For example \code{article} or \code{book}.

You can iterate over these internal list fields using
\glsadd{idx.loop}\ics{glsxtrfielddolistloop} or \ics{glsxtrfieldforlistloop}.
For example:
\begin{codeenv}
\cs{newcommand}\marg{\csfmt{contributorhandler}}[1]\marg{\csfmt{par}\cs{glsentryname}\marg{\#1}}

\cs{newcommand}\marg{\postdeschook{contributor}}\marg{\comment{}
  \cs{glsxtrifhasfield}\marg{bibtexentry}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{\comment{}
    \cs{glsxtrfieldforlistloop}
    \marg{\cs{glscurrententrylabel}}\marg{bibtexentry}\comment{}
    \marg{\csfmt{contributorhandler}}\comment{}
  }\comment{}
  \marg{\csfmt{par} No titles.}\comment{}
}
\end{codeenv}
(where the resource option \csopt[\fieldfmt{title}=\field{name}]{field-aliases}
has been used).

Here's an example that uses the test \file{xampl.bib} file that's
provided with \TeX\ distributions:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[\styopt{record},\styopt{nomain}]\marg{glossaries-extra}
\strut
\cs{newglossary*}\marg{contributors}\marg{Authors/Editors}
\cs{newglossary*}\marg{titles}\marg{Titles}
\strut
\cs{newcommand}\marg{\gls{bibglsnewbibtexentry}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}={\idx{param}3},\idx{param}2,\field{type}=\marg{titles}}\marg{\idx{param}4}\comment{}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[xampl]{src},
  \csopt[false]{write-preamble},
  \csopt[
   \ics{GlsXtrBibTeXEntryAliases}
  ]{entry-type-aliases},
  \csopt[
    \fieldfmt{title}=\field{name}
  ]{field-aliases},
  \csopt[
    \fieldfmt{note}=\field{name}
  ]{replicate-fields},
  \csopt[
    \marg{[ \cs{cs.string}\csfmt{-}\cs{cs.string}\csfmt{.}]}\marg{}
  ]{labelify-replace},
  \csopt[contributors]{type},
  \csopt[same as original entry]{category},
  \csopt[\field{category}]{sort-field},
  \csopt[\field{name}]{sort-suffix}
}
\strut
\gls{glsxtrsetgrouptitle}\marg{article}\marg{Articles}
\gls{glsxtrsetgrouptitle}\marg{booklet}\marg{Booklets}
\gls{glsxtrsetgrouptitle}\marg{book}\marg{Books}
\gls{glsxtrsetgrouptitle}\marg{inbook}\marg{Book Chapters}
\gls{glsxtrsetgrouptitle}\marg{misc}\marg{Miscellaneous}
\strut
\cs{newcommand}\marg{\csfmt{contributorhandler}}[1]\marg{\csfmt{par}\ics{glsentryname}\marg{\idx{param}1} (\idx{param}1)}
\strut
\cs{newcommand}\marg{\postdeschook{contributor}}{\comment{}
  \cs{glsxtrifhasfield}\marg{bibtexentry}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{\comment{}
    \ics{glsxtrfieldforlistloop}
    \marg{\cs{glscurrententrylabel}}\marg{bibtexentry}\comment{}
    \marg{\csfmt{contributorhandler}}\comment{}
  }\comment{}
  \marg{\csfmt{par} No titles.}\comment{}
}
\strut
\csfmt{begin}\marg{document}
Sample\idx{nbspchar}\ics{cite}\marg{book-minimal,article-full,inbook-full,misc-minimal}.
Another sample\idx{nbspchar}\cs{cite}\marg{booklet-minimal,misc-full,article-minimal}.
\strut
\csfmt{bibliographystyle}\marg{plain}
\ics{bibliography}\marg{xampl}
\strut
\cs{printunsrtglossary}[\printglossopt[contributors]{type},\printglossopt[altlist]{style}]
\cs{printunsrtglossary*}[\printglossopt[titles]{type},\printglossopt[indexgroup]{style}]
\marg{\comment{}
 \cs{renewcommand}\marg{\ics{glsxtrgroupfield}}\marg{category}\comment{}
 \cs{renewcommand}\marg{\cs{glstreenamefmt}}[1]\marg{\cs{emph}\marg{\idx{param}1}}\comment{}
 \cs{renewcommand}\marg{\cs{glstreegroupheaderfmt}}[1]\marg{\cs{textbf}\marg{\idx{param}1}}\comment{}
}
\strut
\csfmt{end}\marg{document}
\end{codeenv}
If the file is called \filefmt{myDoc.tex} then the document build
is:
\begin{codeenv}
pdflatex myDoc
bib2gls \longarg{cite-as-record} myDoc
bibtex myDoc
pdflatex myDoc
pdflatex myDoc
\end{codeenv}

\entrysection{progenitor}

The \atentryfmt{progenitor} type of entries are the only place where
the \field{adoptparents} field is permitted. The value should be a
comma-separated list of labels. The \field{adoptparents} field
must be set and must contain a least one label. If the value
contains any of the characters \idx{backslashchar} (backslash), 
\idx{bgroupchar} (open brace) or \idx{egroupchar} (close brace) then
the field will be interpreted (if the default \longarg{interpret} settings 
is on).

Since entries are spawned before fields are processed, the
\field{adoptparents} field is parsed before any field aliases
(\csopt{field-aliases}) or replication (\csopt{replicate-fields})
takes place. However, if the \field{adoptparents} field isn't found,
\bibgls\ will check for a simple mapping in both the
\csopt{field-aliases} and \csopt{replicate-fields} settings.

This entry type creates a main \idx[textformat=emph]{progenitor}
term (with all the given fields except \field{adoptparents})
and $n$ spawned \idx[textformat=emph]{progeny} terms, where
$n$ is the number of elements in the \field{adoptparents} field,
that are dependent on the main term.

Each of the spawned \idx{progeny} entries have the field identified by
\csopt{adopted-parent-field} (\field{parent} by default) set to the
corresponding element in the \field{adoptparents} field.

All fields from the original definition are copied except for the
\field{adoptparents}, \field{alias} and \field{parent} fields. The
\field{parent} field is never copied, regardless of the value of
\csopt{adopted-parent-field}.  If the adopted parent field is
changed to one that's contained in the original entry, it's value
will be from \field{adoptparents} not the value from the original
entry.

The copied fields follow the same conditions as normal
entries. (For example, unknown fields are ignored, case-changes are
applied, if appropriate, and the \field{type} field must reference a
valid glossary, if set.) If \csopt{progenitor-type} is set, then
this assignment is made after the \idx{progeny} are created
and only applies to the main \idx{progenitor} entry. The
type for the \idx{progeny} can be set with \csopt{progeny-type}.
For example, \csopt[same as parent]{progeny-type} will ensure
that the \idx{progeny} are in the same glossary type as
their parent entry.

For example, an entry defined as:
\begin{codeenv}
\atentry{progenitor}\marg{\meta{id},
  \field{adoptparents} = \marg{\meta{parent-1 id},\ldots,\meta{parent-N id}},
  \meta{field-name-1} = \margm{text},
  \ldots
  \meta{field-name-n} = \margm{text}
}
\end{codeenv}
is essentially like:
\begin{codeenv}
\atentry{index}\marg{\meta{id},
  \field{progeny} = \marg{\meta{parent-1 id}.\meta{id},\ldots,\meta{parent-N id}.\meta{id}},
  \meta{field-name-1} = \margm{text},
  \ldots
  \meta{field-name-n} = \margm{text}
}
\strut
\atentry{index}\marg{\meta{parent-1 id}.\meta{id},
  \field{progenitor} = \margm{id},
  \field{parent} = \margm{parent-1 id},
  \meta{field-name-1} = \margm{text},
  \ldots
  \meta{field-name-n} = \margm{text}
}
\strut
\ldots
\strut
\atentry{index}\marg{\meta{parent-N id}.\meta{id},
  \field{progenitor} = \margm{id},
  \field{parent} = \margm{parent-N id},
  \meta{field-name-1} = \margm{text},
  \ldots
  \meta{field-name-n} = \margm{text}
}
\end{codeenv}
This creates the main (\idx{progenitor}) \meta{id} entry, which
contains all the fields (except for \field{adoptparents}) that were
in the original \atentry{progenitor} definition and has the new
field \field{progeny} set to the comma-separated list of spawned
entry labels. The main entries are defined in the \ext{glstex} file
with \gls{bibglsnewprogenitor}.

In addition to the main \meta{id} entry, the above also creates the
spawned \idx{progeny} entries \code{\meta{parent-1 id}.\meta{id}}, \ldots, 
\code{\meta{parent-N id}.\meta{id}} that are dependent on the main
\meta{id} entry.

The spawned entries have the \field{parent} field set to the
corresponding label obtained from the \field{adoptparents} list.
This parent entry must also be defined, as usual for the
\field{parent} field.  (This restriction obviously doesn't apply if
\csopt{adopted-parent-field} is changed from the default
\field{parent}.) The spawned entries are defined in the \ext{glstex}
file with \gls{bibglsnewspawnedindex}

If the main \idx{progenitor} entry is referenced in the document
then (assuming the default selection criteria) the spawned entries
will also be automatically selected. You can check for the existence
of the \field{progenitor} field using \cs{glsxtrifhasfield} and
fetch the \field{location} field from the main entry, if required.

Although the spawned entries are considered dependents of the main
entry, the reverse doesn't apply. If a spawned entry is referenced
in the document (with \code{\meta{parent-id}.\meta{id}}) then the
main entry and its other spawned entries aren't automatically
selected.

For example, suppose the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{indexplural}\marg{stylesheet, \field{text}=\marg{stylesheet language}}
\strut
\atentry{index}\marg{webdesign, \field{name}=\marg{web design}}
\strut
\atentry{indexplural}\marg{markup, \field{text}=\marg{markup language}}
\strut
\atentry{progenitor}\marg{xml,
  \field{name}=\marg{XML},
  \field{adoptparents}=\marg{markup}
}
\strut
\atentry{progenitor}\marg{css,
  \field{name}=\marg{CSS},
  \field{adoptparents}=\marg{stylesheet,webdesign}
}
\strut
\atentry{progenitor}\marg{html,
  \field{name}=\marg{HTML},
  \field{adoptparents}=\marg{markup,webdesign}
}
\strut
\atentry{progenitor}\marg{xsl,
  \field{name}=\marg{XSL},
  \field{adoptparents}=\marg{stylesheet}
}
\end{codeenv}
and if the document contains:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{\styopt{record},\styopt[tree]{stylemods},\styopt[index]{style}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[all]{selection}}
\strut
\cs{newcommand}*\marg{\cs{glstreenamefmt}}[1]\marg{\idx{param}1}
\cmd{begin}\marg{document}
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
Then the resulting list will be:
\begin{quote}
\setlength{\parindent}{0pt}%
\setlength{\parskip}{0pt plus 0.3pt}%
\glstreeitem CSS\par
\glstreeitem HTML\par
\glstreeitem markup language\par
\glstreesubitem HTML\par
\glstreesubitem XML\par
\glstreeitem stylesheet language\par
\glstreesubitem CSS\par
\glstreesubitem XSL\par
\glstreeitem web design\par
\glstreesubitem CSS\par
\glstreesubitem HTML\par
\glstreeitem XML\par
\glstreeitem XSL\par
\end{quote}
This allows the HTML and CSS entries to be listed under multiple
parents.

The following \atentryfmt{spawn\meta{single-type}} commands are all forms
of \atentry{progenitor} that create the given
\atentryfmt{\meta{single-type}} of entry. The spawned entries are actually
created with the private entry type \atentryfmt{spawned\meta{type}}. In the
case of \atentry{progenitor}, the spawned entries are defined as 
a \atentryfmt{spawnedindex} entry. These special
\atentryfmt{spawned\meta{type}} entry types aren't intended for use
in the \ext{bib} file, but if you reference the entry type (for
example, with \csopt[same as entry]{category}) you will get 
\atentryfmt{spawned\meta{type}} as the entry type. The
original entry type for the spawned entries is the same as
the original entry for the main \atentry{progenitor} entry.

There is currently only one form of dual \atentry{progenitor} entry and that's
\atentry{spawndualindexentry}. Only the main \idx{progenitor} entry
is a dual entry. The spawned \idx{progeny} are all \atentry{index} 
primary entries.

\entrysection{spawnindex}

As \atentry{progenitor}, but the main entries are defined in the
\ext{glstex} file with \gls{bibglsnewspawnindex} and the
spawned entries are defined with \gls{bibglsnewspawnedindex}.

\entrysection{spawnindexplural}

As \atentry{progenitor}, except that it creates
\atentry{indexplural} terms instead of \atentry{index}.
As with \atentry{indexplural}, if the \field{name} field isn't set,
it's assigned to the same value as the \field{plural} field (or the
fallback for the \field{plural}, if not defined).

The main entries are defined in the
\ext{glstex} file with \gls{bibglsnewspawnindexplural} and the
spawned entries are defined with \gls{bibglsnewspawnedindexplural}.

\entrysection{spawnentry}

As \atentry{progenitor}, except that it creates
\atentry{entry} terms instead of \atentry{index}.
As with \atentry{entry}, the \field{description} field is required
and either \field{name} or \field{parent}.

The main entries are defined in the
\ext{glstex} file with \gls{bibglsnewspawnentry} and the
spawned entries are defined with \gls{bibglsnewspawnedentry}.

\entrysection{spawnabbreviation}

As \atentry{progenitor}, except that it creates
\atentry{abbreviation} terms instead of \atentry{index}.
As with \atentry{abbreviation}, the \field{short} and \field{long} 
fields are required.

The main entries are defined in the
\ext{glstex} file with \gls{bibglsnewspawnabbreviation} and the
spawned entries are defined with \gls{bibglsnewspawnedabbreviation}.

\entrysection{spawnacronym}

As \atentry{progenitor}, except that it creates
\atentry{acronym} terms instead of \atentry{index}.
As with \atentry{acronym}, the \field{short} and \field{long} 
fields are required.

The main entries are defined in the
\ext{glstex} file with \gls{bibglsnewspawnacronym} and the
spawned entries are defined with \gls{bibglsnewspawnedacronym}.

\entrysection{spawnsymbol}

As \atentry{progenitor}, except that it creates
\atentry{symbol} terms instead of \atentry{index}.
As with \atentry{symbol}, the required fields are \field{name} or
\field{parent}, and the \field{description} field is required if the
\field{name} field is missing.

The main entries are defined in the
\ext{glstex} file with \gls{bibglsnewspawnsymbol} and the
spawned entries are defined with \gls{bibglsnewspawnedsymbol}.

\entrysection{spawnnumber}

As \atentry{progenitor}, except that it creates
\atentry{number} terms instead of \atentry{index}.
As with \atentry{number}, the required fields are \field{name} or
\field{parent}, and the \field{description} field is required if the
\field{name} field is missing.

The main entries are defined in the
\ext{glstex} file with \gls{bibglsnewspawnnumber} and the
spawned entries are defined with \gls{bibglsnewspawnednumber}.

\entrysection{spawndualindexentry}

As \atentry{progenitor}, except that the main (\idx{progenitor})
entry behaves like \atentry{dualindexentry}. The spawned
\idx{progeny} behave like \atentry{index} are so are all considered primary
entries. The \field{adoptparents} field should therefore reference
primary entries with the default \csopt[parent]{adopted-parent-field}.

The main primary and secondary (dual) entries are defined in the
\ext{glstex} file with
\gls{bibglsnewspawndualindexentry} and
\gls{bibglsnewspawndualindexentrysecondary}. The spawned
\idx{progeny} are defined with \gls{bibglsnewspawnedindex}.

\chapter{Resource File Options}
\label{sec:resourceopts}

Make sure that you use \isty{glossaries-extra} with the
\styopt{record} package option. This ensures that \bibgls\ can pick
up the required information from the \iext{aux} file, and both 
\styopt[only]{record} and \styopt[nameref]{record} additionally
load the supplementary \isty{glossaries-extra-bib2gls} package. 
These two \styopt{record} option values also
switch on the \styopt[none]{sort} package option (if you have a new
enough version of the base \isty{glossaries} package), which means
that there's no attempt to assign or process the \field{sort} key if
it's omitted from \gls{newglossaryentry} (or similar commands). The
\field{sort} key will be provided by \bibgls\ for informational
purposes, but there's no need for \LaTeX\ to write it to any
external files (unless you use the hybrid \styopt[alsoindex]{record}, 
in which case you need to prevent \bibgls\ from sorting using the
\csopt[none]{sort} resource option).

The \iext{glstex} resource files created by \bibgls\ are loaded in
the document using
\nosecformatdef{glsxtrresourcefile}
where \meta{filename} is the name of the resource file without the
\ext{glstex} extension.
You can have multiple \gls{glsxtrresourcefile} commands within your
document, but each \meta{filename} must be unique, otherwise \LaTeX\
would attempt to input the same \ext{glstex} file multiple times
(\bibgls\ checks for non-unique file names). The associated data for each
resource file is called the \idx{resourceset} (see
\sectionref{sec:resourcesets}).

There's a shortcut command that uses
\ics{jobname} in the \meta{filename}:
\nosecformatdef{GlsXtrLoadResources}
The first instance of this command is equivalent to:
\begin{codeenv}
\gls{glsxtrresourcefile}\oargm{options}\marg{\ics{jobname}}
\end{codeenv}
Any additional use of \gls{GlsXtrLoadResources} is equivalent to:
\begin{codeenv}
\gls{glsxtrresourcefile}\oargm{options}\marg{\ics{jobname}-\meta{n}}
\end{codeenv}
where \meta{n} is number. For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-en]{src},\csopt[en]{sort}}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-fr]{src},\csopt[fr]{sort}}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-de]{src},\csopt[de-1996]{sort}}
\end{codeenv}
This is equivalent to:
\begin{codeenv}
\gls{glsxtrresourcefile}\oarg{\csopt[entries-en]{src},\csopt[en]{sort}}\marg{\cs{jobname}}
\gls{glsxtrresourcefile}\oarg{\csopt[entries-fr]{src},\csopt[fr]{sort}}\marg{\cs{jobname}-1}
\gls{glsxtrresourcefile}\oarg{\csopt[entries-de]{src},\csopt[de-1996]{sort}}\marg{\cs{jobname}-2}
\end{codeenv}
In general, it's simplest just to use \gls{GlsXtrLoadResources}.

The optional argument \meta{options} is a comma-separated
\keyvallist.  Allowed options are listed below.  The option list
applies only to that specific \meta{filename}\ext{glstex} and are
not carried over to the next instance of \gls{glsxtrresourcefile}.
Only the definitions provided in \atentry{preamble} (if the
interpreter is on and \csopt[true]{interpret-preamble}) are carried
over to the next resource set and, possibly,
\idxpl{crossresourceref} if permitted (see
\sectionref{sec:resourcesets}).  The \sty{glossaries-extra} package
doesn't parse the options, but just writes the information to the
\ext{aux} file. This means that any invalid options will be reported
by \bibgls\ not by \sty{glossaries-extra}.

As from \sty{glossaries-extra} v1.40 you can provide a default set
of options by redefining:
\nosecdef{GlsXtrDefaultResourceOptions}
This command will be inserted at the start of the options list for
all resource commands (and will expand as it's written to
the \ext{aux} file).  For example:
\begin{codeenv}
\cs{renewcommand}\marg{\gls{GlsXtrDefaultResourceOptions}}\marg{\comment{}
  \csopt[all]{selection},\csopt[entries]{src}}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[symbols]{type},
  \csopt[entrytype=symbol]{match}}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[abbreviations]{type},
  \csopt[entrytype=abbreviation]{match}}
\end{codeenv}
This acts like:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[all]{selection},\csopt[entries]{src},
  \csopt[symbols]{type},
  \csopt[entrytype=symbol]{match}}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[all]{selection},\csopt[entries]{src},
  \csopt[abbreviations]{type},
  \csopt[entrytype=abbreviation]{match}}
\end{codeenv}

If you have multiple \ext{bib} files you can either select them all
using \csopt[\meta{bib list}]{src} in a single
\gls{glsxtrresourcefile} call, if they all require the same settings,
or you can load them separately with different settings applied.

For example, if the files \filefmt{entries-terms.bib} and
\filefmt{entries-symbols.bib} have the same settings:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-terms,entries-symbols]{src}}
\end{codeenv}
Alternatively, if they have different settings:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-terms]{src},\csopt[main]{type}}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-symbols]{src},\csopt[use]{sort},\csopt[symbols]{type}}
\end{codeenv}

Note that the sorting is applied to each \idx{resourceset} independently
of other \idxpl{resourceset}. This means that if you have multiple instances
of \gls{glsxtrresourcefile} but only one glossary type, the glossary
will effectively contain blocks of sorted entries. For example, if
\filefmt{file1.bib} contains:
\begin{codeenv}
\atentry{index}\marg{duck}
\atentry{index}\marg{zebra}
\atentry{index}\marg{aardvark}
\end{codeenv}
and \filefmt{file2.bib} contains:
\begin{codeenv}
\atentry{index}\marg{caterpillar}
\atentry{index}\marg{bee}
\atentry{index}\marg{wombat}
\end{codeenv}
then
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[file1,file2]{src}}
\end{codeenv}
will result in the list: aardvark, bee, caterpillar, duck, wombat,
zebra. These six entries are all defined when
\csfmt{jobname}\filefmt{.glstex} is read.
Whereas
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[file1]{src}}
\gls{GlsXtrLoadResources}\oarg{\csopt[file2]{src}}
\end{codeenv}
will result in the list: aardvark, duck, zebra, bee, caterpillar,
wombat. The first three (aardvark, duck, zebra) are defined when 
\csfmt{jobname}\filefmt{.glstex} is read. The second three (bee,
caterpillar, wombat) are defined when \csfmt{jobname}\filefmt{-1.glstex}
is read. Since \ics{printunsrtglossary} simply iterates over all
defined entries, this is the ordering used.

Abbreviation styles must be set (using \ics{setabbreviationstyle})
before the resource command that selects the abbreviations from the
appropriate \ext{bib} file, since the entries are defined (through
\gls{newabbreviation} or \gls{newacronym}) when
\gls{glsxtrresourcefile} inputs the \ext{glstex} file. (Similarly for any
associated abbreviation style commands that must be set before
abbreviations are defined, such as \cs{glsxtrlongshortdescname}.)

Note \bibgls\ allows \iext{bib} files that don't provide any entries.
This can be used to provide commands in \atentry{preamble}.
For example, suppose I have \filefmt{defs.bib} that just contains:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cmd{strong}}[1]\marg{\cs{textbf}\marg{\ics{cs.color}\marg{red}\idx{param}1}}
\cs{providecommand}\marg{\cmd{parenswap}}[2]\marg{\idx{param}2 (\idx{param}1)}}}
\end{codeenv}
This provides two commands:
\nosecdef{strong}
(which sets the font weight and colour) and 
\nosecdef{parenswap}
(which just displays its second argument followed by the first in parentheses).

Suppose I also have \filefmt{entries.bib} that contains:
\begin{codeenv}
\atentry{index}\marg{example,
  \field{name}=\marg{\gls{strong}\marg{\gls{parenswap}\marg{stuff}\marg{example}}}
}
\atentry{index}\marg{sample}
\atentry{index}\marg{test}
\atentry{index}\marg{foo}
\atentry{index}\marg{bar}
\end{codeenv}
This contains an entry that requires the commands provided in
\filefmt{defs.bib}, so to ensure those commands are defined, I can
do:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[defs,entries]{src}}
\end{codeenv}
Unfortunately this results in the sort value for \code{example} being
set to \code{redexample (stuff)} because the interpreter has
detected the provided commands and expanded:
\begin{codeenv}
\gls{strong}\marg{\gls{parenswap}\marg{stuff}\marg{example}}
\end{codeenv}
to:
\begin{codeenv}
\cs{textbf}\marg{\cs{cs.color}\marg{red}example (stuff)}
\end{codeenv}
It discards font changes, so \cs{textbf} is ignored, but it doesn't 
recognise \cs{cs.color} and so doesn't know that the first argument is
just the colour specifier and therefore doesn't discard it.
This means that \qt{\textbf{\color{red}example (stuff)}} is placed
between \qt{foo} and \qt{sample} instead of between \qt{bar} and
\qt{foo}.

I can prevent the interpreter from parsing \atentry{preamble}:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[defs,entries]{src},\csopt[false]{interpret-preamble}}
\end{codeenv}
Now when the sort value for \code{example} is obtained from:
\begin{codeenv}
\gls{strong}\marg{\gls{parenswap}\marg{stuff}\marg{example}}
\end{codeenv}
no expansion occurs (since \gls{strong} and \gls{parenswap} are now
unrecognised) so the sort value ends up as:
\code{stuffexample}
which places \qt{\textbf{\color{red}example (stuff)}} between 
\qt{sample} and \qt{test}, which is again incorrect.

The best thing to do in this situation is to split the provided
commands into two \iext{bib} files: one that shouldn't be interpreted
and one that should.

For example, \filefmt{defs-nointerpret.bib}:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\gls{strong}}[1]\marg{\cs{textbf}\marg{\cs{cs.color}\marg{red}\idx{param}1}}}}
\end{codeenv}
and \filefmt{defs-interpret.bib}:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\gls{parenswap}}[2]\marg{\idx{param}2 (\idx{param}1)}}}
\end{codeenv}
Now the first one can be loaded with \csopt[false]{interpret-preamble}:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[defs-nointerpret]{src},\csopt[false]{interpret-preamble}}
\end{codeenv}
This creates a \iext{glstex} file that provides \gls{strong} but
doesn't define any entries. The other file \filefmt{defs-interpret.bib} can
then be loaded with the default \csopt[true]{interpret-preamble}:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[defs-interpret,entries]{src}}
\end{codeenv}
The provided commands are remembered by the interpreter, so you can
also do:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[defs-interpret]{src}}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\end{codeenv}
The \emph{contents} of \atentry{preamble} are only written to the
associated \iext{glstex} file, but the definitions contained within
the \atentry{preamble} are retained by the interpreter for subsequent
\idxpl{resourceset}.

\section{General Options}
\label{sec:generalopts}

\optsection{charset}

If the character \idx{encoding} hasn't been supplied in the \iext{bib} file
with the encoding comment
\begin{alltt}
\idx{commentchar} Encoding: \meta{encoding-name}
\end{alltt}
then you can supply the correct encoding using
\csopt[encoding-name]{charset}. In general, it's better to include
the encoding in the \ext{bib} file where it can also be read by
a \ext{bib} managing systems, such as JabRef.

See \longarg{tex-encoding} for the encoding used to write the \ext{glstex}
file.

\optsection{interpret-preamble}

This is a boolean option that determines whether or not the
interpreter should parse the contents of \atentry{preamble}.
The default is \optfmt{true}. If \optfmt{false}, the preamble
contents will still be written to the \iext{glstex} file, but any
commands provided in the preamble won't be recognised by the
interpreter (see \sectionref{sec:texparserlib}).

Related options are: \csopt{set-widest} (which uses the interpreter to determine
the widest name for the \glostyle{alttree} style or the
\isty{glossary-longextra} styles), \csopt{interpret-label-fields} (which governs
whether or not fields that must only contain a label should be interpreted),
\csopt{labelify} (which converts a field into a string suitable for use as a
label), and \csopt{labelify-list} (which converts a field into a string suitable
for use as a comma-separated list of labels).

\optsection{write-preamble}

This is a boolean option that determines whether or not the
preamble should be written to the \ext{glstex} file. The default
is \optfmt{true}. Note that the preamble will
still be parsed if \csopt[true]{interpret-preamble} even if
\csopt[false]{write-preamble}. This means it's possible to 
provide \bibgls\ command definitions in \atentry{preamble}
that don't get seen by \LaTeX.

\optsection{set-widest}

The \glostyle{alttree} glossary style needs to know the widest
\field{name} (for each level, if hierarchical). This can be set
using \ics{glssetwidest} provided by the \isty{glossary-tree}
package (or similar commands like \ics{glsupdatewidest} provided by
\isty{glossaries-extra-stylemods}), but this requires knowing
which name is the widest. Alternatively, one of the iterative commands
such as \ics{glsFindWidestTopLevelName} can be used, which slows the
document build as it has to iterate over all defined entries.

The \isty{glossary-longextra} package, provided with
\sty{glossaries-extra} v1.37+, also needs to know the widest name,
but in this case only the top-level is needed. If this has already
been found through the commands provided with the \glostyle{alttree}
style then that value will be used as the default, but you can set another
value that's only used for the \sty{glossary-longextra} styles with
\ics{glslongextraSetWidest}.

The \sty{glossaries-extra-bib2gls} package provides \ics{glsxtrSetWidest},
which sets the widest name for those styles that need it. As from version
1.8, \bibgls\ now checks for the existence of this command and will use it with
\csopt{set-widest} to allow for the new styles provided by the
\sty{glossary-longextra} package.

The boolean option \csopt[true]{set-widest} will try to calculate
the widest names for each hierarchical level to help remove the need
to determine the correct value within the document. 
Since \bibgls\ doesn't know the fonts that will be used
in the document or if there are any non-standard commands that
aren't provided in the \ext{bib} files preamble, \emph{this option may not
work}. For example, if one entry has the \field{name} defined as:
\begin{codeenv}
\field{name}=\marg{some \marg{\cmd{Huge} huge} text}
\end{codeenv}
and another entry has the \field{name} defined as:
\begin{codeenv}
\field{name}=\marg{some \marg{\cmd{small} small} text}
\end{codeenv}
then \bibgls\ will determine that the second name is the widest
although the first will actually be wider when it's rendered in the
document. 

When using this option, the transcript file will include the message:
\begin{alltt}
Calculated width of '\meta{text}': \meta{number}
\end{alltt}
where \meta{text} is \bibgls's interpretation of the contents of the
\field{name} field and \meta{number} is a rough guide to the width
of \meta{text} assuming the operating system's default serif font.
The entry that has the largest \meta{number} is the one that will be
selected. This will then be implemented as follows:
\begin{itemize}
\item If the \field{type} is unknown then:
  \begin{itemize}
   \item if the interpreter resolves all \field{name} fields to the
    empty string (that is the \field{name} fields all consist of
    unknown commands) then
    \begin{itemize}
     \item if there are child entries \gls{bibglssetwidestfallback}
     is used,
     \item otherwise \gls{bibglssetwidesttoplevelfallback} is used;
    \end{itemize}
   \item otherwise \gls{bibglssetwidest} is used.
  \end{itemize}
\item If the \field{type} is known then:
  \begin{itemize}
   \item if the interpreter resolves all \field{name} fields for that
    type to the empty string (that is the \field{name} fields all consist of
     unknown commands) then
    \begin{itemize}
     \item if there are child entries \gls{bibglssetwidestfortypefallback}
     is used,
     \item otherwise 
      \gls{bibglssetwidesttoplevelfortypefallback} is used;
    \end{itemize}
   \item otherwise \gls{bibglssetwidestfortype} is used.
  \end{itemize}
\end{itemize}
This leaves \TeX\ to compute the width according to the document
fonts. If \bibgls\ can't correctly determine the widest entry then
you will need to use one of the commands provided by
\sty{glossary-tree}, \sty{glossary-longextra} or 
\sty{glossaries-extra-stylemods} to set it.

In general, if you have more than one glossary it's best to
set the \field{type} using options like \csopt{type} and
\csopt{dual-type} if you use \csopt{set-widest}.

\optsection{entry-type-aliases}

In the \ext{bib} file, the data is identified by 
\code{@\meta{entry-type}}, such as \atentry[noindex]{abbreviation}.
It may be that you want to replace all instances of
\code{@\meta{entry-type}} with a different 
type of entry. For example, suppose my \ext{bib} file
contains abbreviations defined in the form:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
  \field{short} = \marg{html},
  \field{long}  = \marg{hypertext markup language},
  \field{description} = \marg{a markup language for creating web pages}
}
\end{codeenv}
but suppose in one of my documents I actually want all these
abbreviations defined with \atentry[noindex]{dualabbreviationentry}
instead of \atentry[noindex]{abbreviation}. Instead of editing
the \ext{bib} file I can just supply a mapping:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[abbreviation=dualabbreviationentry]{entry-type-aliases}
}
\end{codeenv}
This makes all instances of \atentry[noindex]{abbreviation}
behave as \atentry[noindex]{dualabbreviationentry}. You can have more than
one mapping. For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[
    \comment{\atentry{abbreviation} -> \atentry{dualabbreviationentry}:}
    abbreviation=dualabbreviationentry,
    \comment{\atentry{entry} -> \atentry{index}:}
    entry=index
  ]{entry-type-aliases}
}
\end{codeenv}
This option isn't cumulative.  Multiple instances of
\csopt{entry-type-aliases} override previous instances.
If \meta{\keyvallist} is empty there will be no mappings.
You can save the original entry type in the
\field{originalentrytype} field with
\csopt{save-original-entrytype}.

Here's another example entry in a \ext{bib} file:
\begin{codeenv}
@foo\marg{html,
   \field{name} = \marg{HTML},
   \field{short} = \marg{HTML},
   \field{long} = \marg{hypertext markup language},
   \field{description} = \marg{hypertext markup language}
}
\end{codeenv}
Ordinarily this entry would be ignored since \code{@foo}
isn't recognised, but it can be mapped like this:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[short,long]{ignore-fields},
  \csopt[foo=entry]{entry-type-aliases}
}
\end{codeenv}
This treats the entry as though it had been defined as:
\begin{codeenv}
\atentry{entry}\marg{html,
   \field{name} = \marg{HTML},
   \field{description} = \marg{hypertext markup language}
}
\end{codeenv}
whereas:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[name,description]{ignore-fields},
  \csopt[foo=abbreviation]{entry-type-aliases}
}
\end{codeenv}
treats the entry as though it had been defined as:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
   \field{short} = \marg{HTML},
   \field{long} = \marg{hypertext markup language}
}
\end{codeenv}

\optsection{unknown-entry-alias}

If this option is set, the \meta{value} is used as the alias for any
unknown entry types (after any aliases provided with
\csopt{entry-type-aliases} have been applied). If the value is
missing or empty, unknown entry types will be ignored with a
warning.

\optsection{action}

This governs how the entries are written in the \iext{glstex} file.
The \meta{value} may be one of: 
\begin{itemize}
\item\code{define}: define the entries;
\item\code{copy}: copy the entries;
\item\code{define or copy}: copy existing entries and define
non-existing entries.
\end{itemize}
The default setting is \csopt[define]{action}, which writes the entry
definition to the \ext{glstex} file using one of the commands
described in \sectionref{sec:newentrydefs}. Since the \styopt{record}
package option automatically switches on the \styopt[warn]{undefaction} 
option, any attempt at defining an entry that's already been defined
will generate a warning rather than an error. The duplicate
definition will be ignored. (The warnings can be found in the
\ext{log} file since they are warnings produce by
\sty{glossaries-extra} not by \bibgls.)

For example, if you try:
\begin{codeenv}
\cs{newglossary*}\marg{copies}\marg{Copies}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\gls{GlsXtrLoadResources}\oarg{\csopt[use]{sort},\csopt[copies]{type},\csopt[entries]{src}}
\end{codeenv}
you'll find that the \code{copies} glossary is empty and there will
be warnings in the \ext{log} file when the second resource file is
loaded.

There are various ways of having the same entries in multiple
glossaries. The simplest method is to use \csopt{secondary}, but
another method is to use \csopt[copy]{action} which simply writes
\nosecformatdef{glsxtrcopytoglossary}
instead of using one of the commands listed in
\sectionref{sec:newentrydefs}. This copies the entries rather than
defining them, which means the entries must already have been
defined. The \meta{type} is determined as follows:
\begin{itemize}
\item if the entry has the \field{type} field set, that's used;
\item if the entry is a tertiary and \csopt{tertiary-type} is set, that's
used;
\item if the entry is a dual and \csopt{dual-type} is set, that's
used;
\item otherwise the value of the \csopt{type} option is used.
\end{itemize}
If you're not sure whether the entries may already be defined, you
can use \csopt[define or copy]{action} which will use
\ics{ifglsentryexists} in the resource file to determine whether to
define or copy the entry.

Options that set or modify fields, such as \csopt{category},
\csopt{group}, \csopt{save-locations}, \csopt{flatten} or
\csopt{name-case-change}, will be ignored if entries are copied.
However the \csopt{copy-action-group-field} may be used to copy
the \field{group} field (which may have been locally set by the 
\csopt{sort} method) to another field. This ensures that the
original \field{group} value from the entry definition in an earlier
resource set won't be overwritten (unless you set
\csopt[group]{copy-action-group-field}).

Remember that \gls{glsxtrcopytoglossary} simply copies the entry's
label to the glossary's internal list. The only checks that \bibgls\
performs if \csopt{action} is not \code{define} is to ensure that
the \csopt{master} or \csopt{secondary} options have not been used,
since they're incompatible, and that the \csopt{type} option is set,
since it's required as a fallback for any entries that don't have
the \field{type} field set. (There are too many options that alter
field values to check them all and some may be used to alter the
sorting.) The purpose of the copy action is simply to provide a
duplicate list in a different order.

Remember that if you are using \sty{hyperref}, you need to use
\code{\printglossopt[false]{target}} in the optional argument of
\ics{printunsrtglossary} for the glossary containing the copies to
prevent duplicate hypertargets. Commands like \cs{gls} will link to
the original entries. For example, in the preamble:
\begin{codeenv}
\cs{newignoredglossary}\marg{copies}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[use]{sort},
 \csopt[copy]{action},
 \csopt[copies]{type},
 \csopt[entries]{src}
}
\end{codeenv}
and later in the document:
\begin{codeenv}
\cs{printunsrtglossary}\oarg{\printglossopt[Glossary (Alphabetical)]{title},\printglossopt[\glostyle{indexgroup}]{style}}
\cs{printunsrtglossary}\oarg{\printglossopt[copies]{type},\printglossopt[Glossary (Order of Use)]{title},
 \printglossopt[index]{style},\printglossopt{nogroupskip},\comment{no grouping}
 \printglossopt[false]{target}}
\end{codeenv}
Note also the need to use \code{nogroupskip} and a non-group style
for the duplicates since the \field{group} field will have been
assigned in the first resource set if \bibgls\ was invoked with 
\longarg{group}. The grouping is appropriate for alphabetical
ordering but not for order of use.

If you want different grouping for the duplicates, you can specify
the field name to use in which to store the group information using
\csopt{copy-action-group-field}. Unlike \csopt{secondary}, you
will need to redefine \ics{glsxtrgroupfield} to the relevant field
before you display the glossary. The simplest way to do this is with
the starred form of \cs{printunsrtglossary}.
For example, if \csopt[dupgroup]{copy-action-group-field} is added to
the options for the second resource set:
\begin{codeenv}
\ics{printunsrtglossary*}\oarg{\printglossopt[copies]{type},\printglossopt[Duplicates]{title},\printglossopt[\glostyle{indexgroup}]{style}}
 \marg{\cs{renewcommand}\marg{\cs{glsxtrgroupfield}}\marg{dupgroup}}
\end{codeenv}
This just does:
\begin{codeenv}
\cmd{begingroup}
 \cs{renewcommand}\marg{\ics{glsxtrgroupfield}}\marg{dupgroup}\comment{}
 \cs{printunsrtglossary}\oarg{\printglossopt[copies]{type},\printglossopt[Duplicates]{title},\printglossopt[indexgroup]{style}}
\cmd{endgroup}
\end{codeenv}


\section{Selection Options}
\label{sec:selectionopts}

\optsection{src}

This identifies the \iext{bib} files containing the entry
definitions.  The value should be a comma-separated list of the
required \ext{bib} files. These may either be in the current working
directory or in the directory given by the \longarg{dir} switch or on
\TeX's path (in which case \idx{kpsewhich} will be used to find
them).  The \ext{bib} extension may be omitted.  Remember that if
\meta{list} contains multiple files it must be grouped to protect
the comma from the \meta{options} list.

For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-terms,entries-symbols]{src}}
\end{codeenv}
indicates that \bibgls\ must read the files
\filefmt{entries-terms.bib} and \filefmt{entries-symbols.bib} and
create the file given by \ics{jobname}\iext{glstex} on the first
instance or \ics{jobname}\code{-}\meta{n}\ext{glstex} on subsequent
use.

With \code{\gls!{glsxtrresourcefile}\oargm{options}\margm{filename}}, if 
the \csopt{src} option is omitted, 
the \ext{bib} file is assumed to be \meta{filename}\ext{bib}. For example:
\begin{codeenv}
\gls{glsxtrresourcefile}\marg{entries-symbols}
\end{codeenv}
indicates that \bibgls\ needs to read the file
\filefmt{entries-symbols.bib}, which contains the entry data, and create the file
\filefmt{entries-symbols.glstex}. If the \ext{bib} file is
different or if you have multiple \ext{bib} files, you need to use
the \csopt{src} option.

\gls{GlsXtrLoadResources} uses \ics{jobname} as the argument of 
\gls{glsxtrresourcefile} on the first instance, so:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{}
\end{codeenv}
will assume \csopt[\ics{jobname}]{src}. Remember that subsequent uses
of \gls{GlsXtrLoadResources} append a suffix, so in general it's
best to always supply \csopt{src}.

If you have non-ASCII characters in the \ext{bib} filename but
aren't using \XeLaTeX\ or \LuaLaTeX, then you will need to use
\ics{detokenize} to prevent expansion when the information is
written to the \ext{aux} file. Similarly for any special characters
that need protecting (although it's better not to use special
characters in filenames). For example:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{T2A}\marg{fontenc}
\cmd{usepackage}\oarg{utf8}\marg{inputenc}
\cmd{usepackage}\oarg{russian}\marg{babel}
\cmd{usepackage}\oarg{record}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[\cs{detokenize}\marg{\textcyrillicmono{кириллица}}]{src},\comment{data in \textcyrillicmono{кириллица.bib}}
  \csopt[all]{selection}
}
\strut
\cmd{begin}\marg{document}
\cs{printunsrtglossary}
\cmd{end}\marg{document}
\end{codeenv}

\optsection{selection}

By default all entries that have records in the \iext{aux} file will
be selected as well as all their dependent entries. The dependent
entries that don't have corresponding records on the first \LaTeX\
run, may need an additional build to ensure their location lists
are updated.

Remember that on the first \LaTeX\ run the \iext{glstex} files don't
exist. This means that the entries aren't defined at that point. The
\styopt{record} package option additionally switches on the
\styopt[warn]{undefaction} option, which means that you'll only get
warnings rather than errors when you reference entries in the
document. You can't use \ics{glsaddall} with
\bibgls\ because the glossary lists are empty on the first run, so
there's nothing for \ics{glsaddall} to iterate over.
Instead, if you want to add all defined entries, you need to
instruct \bibgls\ to do this with the \csopt{selection} option. The
following values are allowed:
\begin{itemize}
\item \optfmt{recorded and deps}: add all recorded entries and
their dependencies (default).

\item \optfmt{recorded and deps and see}: as above but will also
add unrecorded entries whose \field{see}, \field{seealso} or
\field{alias} field refers to a recorded entry.

\item \optfmt{recorded and deps and see not also}: as above but will 
add unrecorded entries whose \field{see} or \field{alias} (but not
\field{seealso}) field refers to a recorded entry.

\item \optfmt{recorded no deps}: add all recorded entries but not
their dependencies. The dependencies include those referenced in the
\field{see} or \field{seealso} field or fields identified by 
\csopt{dependency-fields}, \field{parent} entries and those found referenced
with commands like \ics{gls} in the field values that are parsed by
\bibgls. With this setting, parents will be omitted unless they've
been referenced in the document through commands like \ics{gls}.

\item \optfmt{recorded and ancestors}: this is like the previous
setting but parents are added even if they haven't been referenced
in the document. The other dependent entries are omitted if they
haven't been referenced in the document.

\item \optfmt{deps but not recorded}: this first selects entries as
though \optfmt{recorded and deps} had been used, but after all
ancestors and dependencies have been added it then removes all
entries that have records. This means that you end up with only the
unrecorded dependencies. (Recorded entries will need to be selected
in a different resource set.)

\item \optfmt{ancestors but not recorded}: this first selects
entries as though \optfmt{recorded and ancestors} had been used, but
after all ancestors have been added it then removes all entries that
have records. This means that you end up with only the unrecorded
ancestors. (Recorded entries will need to be selected
in a different resource set.) See the \exfile{sample-nested.tex}
example document.

\item \optfmt{all}: add all entries found in the \ext{bib} files
supplied in the \csopt{src} option.
\end{itemize}
The \meta{value} must be supplied.

For example, suppose the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{index}\marg{run}
\atentry{index}\marg{sprint,\field{see}=\marg{run}}
\atentry{index}\marg{dash,\field{see}=\marg{sprint}}
\end{codeenv}
If the document only references the \qt{run} entry (for example,
using \code{\cs{gls}\marg{run}}) then:
\begin{itemize}
\item If \csopt[recorded and deps]{selection}, only the \qt{run}
entry is selected. The \qt{run} entry has a record, so it's
selected, but it has no dependencies. Neither \qt{sprint} nor 
\qt{dash} have records, so they're not selected.
\item If \csopt[recorded and deps and see]{selection}, the \qt{run}
and \qt{sprint} entries are selected, but not the \qt{dash} entry.
The \qt{run} entry is selected because it has a record. The
\qt{sprint} entry doesn't have a record but its \field{see} field
includes \qt{run}, which does have a record, so \qt{sprint} is also
selected. The \qt{dash} entry doesn't have a record. Its
\field{see} field references \qt{sprint}. Although \qt{sprint} has
been selected, it doesn't have any records, so \qt{dash} isn't
selected.
\end{itemize}
The above is just an example. The circuitous redirection of
\qt{dash} to \qt{sprint} to \qt{run} is unhelpful to the reader and
is best avoided (especially for an index where there are no accompanying 
descriptions and no \idx{locationlist} for the intermediate \qt{sprint}). 
A better method would be:
\begin{codeenv}
\atentry{index}\marg{run}
\atentry{index}\marg{sprint,\field{see}=\marg{run}}
\atentry{index}\marg{dash,\field{see}=\marg{run}}
\end{codeenv}
The \csopt[recorded and deps and see]{selection} in this case will
select all three entries, and the document won't send the reader on a
long-winded detour.

Now suppose that the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{entry}\marg{run,
 \field{name} = \marg{run},
 \field{description}=\marg{move fast using legs}
}
\strut
\atentry{entry}\marg{sprint,
 \field{name} = \marg{sprint},
 \field{description}=\marg{run at full speed over short distance},
 \field{seealso}=\marg{run}
}
\strut
\atentry{entry}\marg{dash,
 \field{name} = \marg{dash},
 \field{description}=\marg{run in a great hurry},
 \field{seealso}=\marg{sprint}
}
\end{codeenv}
and suppose the document only references \qt{dash} (for example, with
\code{\cs{gls}\marg{dash}}), then with the default \csopt[recorded and deps]{selection}
\qt{dash} will be selected because it has a record, and \qt{sprint}
will be selected because \qt{dash} requires it (for the cross-reference),
and \qt{run} will be selected because \qt{sprint} requires it
(for the cross-reference). In this case, neither \qt{sprint} nor \qt{run}
have a \idx{locationlist} but they do both provide additional information
for the reader in their descriptions.

A better method here would be for each entry to have a cross-reference
list that includes all related terms:
\begin{codeenv}
\atentry{entry}\marg{run,
 \field{name} = \marg{run},
 \field{description}=\marg{move fast using legs},
 \field{seealso}=\marg{sprint,dash}
}
\strut
\atentry{entry}\marg{sprint,
 \field{name} = \marg{sprint},
 \field{description}=\marg{run at full speed over short distance},
 \field{seealso}=\marg{run,dash}
}
\strut
\atentry{entry}\marg{dash,
 \field{name} = \marg{dash},
 \field{description}=\marg{run in a great hurry},
 \field{seealso}=\marg{sprint,run}
}
\end{codeenv}
Now, whichever one is indexed in the document, the other two will automatically
be selected.

\optsection{match}

It's possible to filter the selection by matching field values.
The value is required for this key but may be empty, which indicates 
that the setting is switched off, otherwise
\meta{\keyvallist} should be a \meta{key}=\meta{regexp} list, where
\meta{key} is the name of a field or \optfmt{id} for the entry's
label or \optfmt{entrytype} for the \bibgls\ entry type (as in
the part after \verb|@| identifying the entry not the \field{type}
field identifying the glossary label). If you've used
\csopt{entry-type-aliases}, this refers to the target entry type not
the original entry type specified in the \ext{bib} file.

The \meta{regex} part should be a \idx{regex} conforming
to
\href{http://docs.oracle.com/javase/8/docs/api/java/util/regex/Pattern.html}{Java's
Pattern class}~\cite{pattern}. The pattern is anchored 
(\code{oo}\idx{matchanydot}\idx{zeroormore} matches
\code{oops} but not \code{loops}) and \meta{regexp} can't be
empty. Remember that \TeX\ will expand the option list as
it writes the information to the \ext{aux} file so take
care with special characters. For example, to match a literal
period use \cs{cs.string}\idx{cs.period} not \idx{cs.period} (backslash dot).

If the field is missing its value it is assumed to be empty for
the purposes of the pattern match even if it will be assigned a
non-empty default value when the entry is defined. If the field is
unrecognised by \bibgls\ any reference to it in \meta{\keyvallist} will
be ignored.

If a field is listed multiple times, the pattern for that
field is concatenated using:
\begin{alltt}
(?:\meta{pattern-1})|(?:\meta{pattern-2})
\end{alltt}
where \meta{pattern-1} is the current pattern for that field
and \meta{pattern-2} is the new pattern. This means it performs a
logical OR\@. For the non-duplicate
fields the logical operator is given by \csopt{match-op}.
For example:
\begin{codeenv}
\csopt[and]{match-op},
\csopt[
  \field{category}=animals,
  \fieldfmt{topic}=biology,
  \field{category}=vegetables
]{match}
\end{codeenv}
This will keep all the selected entries that satisfy:
\begin{itemize}
\item \field{category} matches \verb"(?:animals)|(?:vegetables)"

(the \field{category} is either \optfmt{animals} or
\optfmt{vegetables})

\item[AND]

\item \fieldfmt{topic} (custom key provided by user) is \code{biology}.
\end{itemize}
and will discard any entries that don't satisfy this condition.
A message will be written to the log file for each entry
that's discarded.

Patterns for unknown fields will be ignored. If the entire
list consists of patterns for unknown fields it will be
treated as \csopt[\empty]{match}. That is, no filtering will be
applied. In the above example, the custom \fieldfmt{topic} key must
be provided before the first \gls{GlsXtrLoadResources} with
\ics{glsaddkey} or \ics{glsaddstoragekey}.

\optsection{match-op}

If the value of \csopt{match} contains more than one 
\meta{key}=\meta{pattern} element, the \csopt{match-op}
determines whether to apply a logical AND or a logical OR.
The \meta{value} may be either \optfmt{and} or \optfmt{or}.
The default is \csopt[and]{match-op}.

\optsection{not-match}

If \csopt[\meta{\keyvallist}]{match} would cause an entry to be
selected then \csopt[\meta{\keyvallist}]{not-match} would cause that
entry to be ignored.  The value is required for this key but may be
empty, which indicates that the setting is switched off.  If you
have both \csopt{match} and \csopt{not-match} in the same resource
set, the last one listed takes precedence.

\optsection{match-action}

The default behaviour with \csopt{match} or \csopt{not-match}
is to filter the selection. This may be changed to append to the
selection instead. The \meta{value} may be one of:
\begin{itemize}
\item\optfmt{filter}: (default) filter selection;
\item\optfmt{add}: append any matches (with \csopt{match}) or
non-matches (with \csopt{not-match}) to the selection.
This setting can't be used with \csopt[use]{sort}.
\end{itemize}
For example, if I want to select all record entries and their
dependencies, but I also want to make sure that any entries with
the category set to \code{important} are always selected regardless
of whether or not they have any records:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[add]{match-action},
  \csopt[\field{category}=important]{match}
}
\end{codeenv}

\optsection{limit}

If \meta{number} is greater than 0 then this will truncate the
list of selected entries after sorting to \meta{number} (if the list
size is greater than that value). The transcript will show the
message:
\begin{alltt}
Truncating according to limit=\meta{number}
\end{alltt}
When used with \csopt{shuffle}, this provides a means of randomly
selecting at most \meta{number} entries. The default setting is
\csopt[0]{limit} (no truncation). A negative value of \meta{number} is not
permitted.

If you have any dual entries, then the truncation will be applied to
the combined list of primary and duals if \csopt[combine]{dual-sort}
otherwise each list will be truncated separately by \meta{number},
which results in a maximum of $2 \times \meta{number}$. Remember
that tertiary entries are created when dual entries are defined in
the \ext{glstex} file, so this will increase the total number of
entries.

\section{Hierarchical Options}
\label{sec:hierarchicalopts}

\optsection{save-child-count}

This is a boolean option. The default setting is
\csopt[false]{save-child-count}. 
If \csopt[true]{save-child-count},
each entry will be assigned a field called \field{childcount} with
the value equal to the number of child entries that have been
selected. As from version 1.5, this option also creates
the \field{childlist} field for entries that have children
selected. This field is in \sty{etoolbox}'s internal list
format and can be iterated over using \glsadd{idx.loop}\ics{glsxtrfieldforlistloop}.

The assignment is done using \ics{GlsXtrSetField} so there's
no associated key. You can test if the field is set and non-zero
using:
\nosecformatdef{GlsXtrIfHasNonZeroChildCount}
which is provided with \sty{glossaries-extra-bib2gls} v1.31+.
Within \meta{true}, you can access the actual value with
\ics{glscurrentfieldvalue}. If \csopt[false]{save-child-count},
this command will do \meta{false} as the \field{childcount} field
won't be set.

For example, suppose \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{index}\marg{birds}
\atentry{index}\marg{duck,\field{parent}=\marg{birds}}
\atentry{index}\marg{goose,\field{plural}=\marg{geese},\field{parent}=\marg{birds}}
\atentry{index}\marg{swan,\field{parent}=\marg{birds}}
\strut
\atentry{index}\marg{minerals}
\atentry{index}\marg{quartz,\field{parent}=\marg{minerals}}
\atentry{index}\marg{corundum,\field{parent}=\marg{minerals}}
\atentry{index}\marg{amethyst,\field{parent}=\marg{minerals}}
\atentry{index}\marg{gypsum,\field{parent}=\marg{minerals}}
\atentry{index}\marg{gold,\field{parent}=\marg{minerals}}
\end{codeenv}
and the document contains:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[\styopt{record},\styopt[indexgroup]{style}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt{save-child-count}}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{duck} and \cs{gls}\marg{goose}.
\cs{gls}\marg{quartz}, \cs{gls}\marg{corundum}, \cs{gls}\marg{amethyst}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
Then the \ext{glstex} file will contain:
\begin{codeenv}
\cs{GlsXtrSetField}\marg{birds}\marg{childcount}\marg{2}
\cs{GlsXtrSetField}\marg{duck}\marg{childcount}\marg{0}
\gls{glsxtrfieldlistadd}\marg{birds}\marg{childlist}\marg{duck}
\cs{GlsXtrSetField}\marg{goose}\marg{childcount}\marg{0}
\gls{glsxtrfieldlistadd}\marg{birds}\marg{childlist}\marg{goose}
\cs{GlsXtrSetField}\marg{minerals}\marg{childcount}\marg{3}
\cs{GlsXtrSetField}\marg{amethyst}\marg{childcount}\marg{0}
\gls{glsxtrfieldlistadd}\marg{minerals}\marg{childlist}\marg{amethyst}
\cs{GlsXtrSetField}\marg{corundum}\marg{childcount}\marg{0}
\gls{glsxtrfieldlistadd}\marg{minerals}\marg{childlist}\marg{corundum}
\cs{GlsXtrSetField}\marg{quartz}\marg{childcount}\marg{0}
\gls{glsxtrfieldlistadd}\marg{minerals}\marg{childlist}\marg{quartz}
\end{codeenv}
Note that although \code{birds} has three children defined in the
\ext{bib} file, only two have been selected, so the child count is
set to~2. Similarly the \code{minerals} entry has five children
defined in the \ext{bib} file, but only three have been selected, so
the child count is~3.

The following uses the post-description hook to show the child count
in parentheses:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[general]{category},\csopt{save-child-count}}
\strut
\cs{glsdefpostdesc}\marg{general}\marg{\comment{}
  \cs{glsxtrifhasfield}\marg{childcount}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{ (child count: \cs{glscurrentfieldvalue}.)}\comment{}
  \marg{}\comment{}
}
\end{codeenv}
\ics{glsxtrifhasfield} requires at least \sty{glossaries-extra} v1.19.
It's slightly more efficient than \ics{ifglshasfield} provided by
the base \styfmt{glossaries} package, and it doesn't complain if the
entry or field don't exist, but note that \ics{glsxtrifhasfield}
implicitly scopes its content. Use the starred version to omit the
grouping.  With \sty{glossaries-extra} v1.31+ you can perform a
numerical test with \ics{GlsXtrIfFieldNonZero} or
\ics{GlsXtrIfFieldEqNum}.

\optsection{save-sibling-count}

This is a boolean option. The default setting is
\csopt[false]{save-sibling-count}. This is like
\csopt{save-child-count} but saves the sibling count in 
\field{siblingcount} and the sibling list in \field{siblinglist}.
As with the child list, the sibling list is in \sty{etoolbox}['s]
internal list format. The sibling information is only saved for
entries that have a parent.

The advantage with \field{siblinglist} over accessing the parent's
\field{childlist} is that the entry itself is excluded from the list.

\optsection{flatten}

This is a boolean option. The default value is \csopt[false]{flatten}. 
If \csopt[true]{flatten}, the sorting will ignore hierarchy and
the \field{parent} field will be omitted when writing
the definitions to the \ext{glstex} file, but the parent entries
will still be considered a dependent ancestor from the 
\csopt{selection} point of view.

Note the difference between this option and using
\csopt[parent]{ignore-fields} which will remove the dependency
(unless a dependency is established through another field).

\optsection{flatten-lonely}

This may take one of three values: \optfmt{false} (default),
\optfmt{presort} and \optfmt{postsort}. The value must be supplied. 

Unlike the \csopt{flatten} option, which completely
removes the hierarchy, the \csopt{flatten-lonely} option can be used
to selectively alter the hierarchy.  In this case only those entries
that have a parent but have no siblings are considered. This option
is affected by the \csopt{flatten-lonely-rule} setting. The conditions for
moving a child up one hierarchical level are as follows:
\begin{itemize}
\item The child must have a parent, and
\item the child can't have any selected siblings, and
\item if \csopt[only unrecorded parents]{flatten-lonely-rule}
      then the parent can't have a location list, where the location list
      includes records and \field{see} or \field{seealso}
      cross-references (for the other rules the parent may have a location
      list as long as it only has the one child selected).
\end{itemize}
If the child is selected for hierarchical adjustment, the parent
will be removed if:
\begin{itemize}
\item The parent has no location list, and
\item \csopt{flatten-lonely-rule} isn't set to \optfmt{no discard}.
\end{itemize}

The value of \csopt{flatten-lonely} determines whether the
adjustment should be made before sorting (\optfmt{presort})
or after sorting (\optfmt{postsort}). To disable this function
use \csopt[false]{flatten-lonely}.

For example, suppose the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{index}\marg{birds}
\atentry{index}\marg{duck,\field{parent}=\marg{birds}}
\atentry{index}\marg{goose,\field{plural}=\marg{geese},\field{parent}=\marg{birds}}
\atentry{index}\marg{swan,\field{parent}=\marg{birds}}
\atentry{index}\marg{chicken,\field{parent}=\marg{birds}}
\strut
\atentry{index}\marg{vegetable}
\atentry{index}\marg{cabbage,\field{parent}=\marg{vegetable}}
\strut
\atentry{index}\marg{minerals}
\atentry{index}\marg{quartz,\field{parent}=\marg{minerals}}
\atentry{index}\marg{corundum,\field{parent}=\marg{minerals}}
\atentry{index}\marg{amethyst,\field{parent}=\marg{minerals}}
\atentry{index}\marg{gypsum,\field{parent}=\marg{minerals}}
\strut
\atentry{index}\marg{aardvark}
\atentry{index}\marg{bard}
\atentry{index}\marg{buzz}
\strut
\atentry{index}\marg{item}
\atentry{index}\marg{subitem,\field{parent}=\marg{item}}
\atentry{index}\marg{subsubitem,\field{parent}=\marg{subitem}}
\end{codeenv}
and suppose the document contains:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[\styopt{record},\styopt[indexgroup]{style}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries.bib]{src}}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{duck}.
\cs{gls}\marg{quartz}, \cs{gls}\marg{corundum}, \cs{gls}\marg{amethyst}.
\cs{gls}\marg{aardvark}, \cs{gls}\marg{bard}, \cs{gls}\marg{buzz}.
\cs{gls}\marg{vegetable}, \cs{gls}\marg{cabbage}.
\cs{gls}\marg{subsubitem}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
Although the \code{duck} entry has siblings in the
\filefmt{entries.bib} file, none of them have been recorded
(indexed) in the document, nor has the parent \code{birds} entry.

This document hasn't used \csopt{flatten-lonely}, so the default
\csopt[false]{flatten-lonely} is assumed. This results in the
hierarchical structure:
\begin{description}\setlength\itemsep{0pt}%
\item[A] \null
\item[aardvark] 1
\item[B] \null
\item[bard] 1
\item[birds] \mbox{}
  \begin{description}\setlength\itemsep{0pt}%
  \item[duck] 1
  \end{description}
\item[buzz] 1
\item[I] \null
\item[item] \mbox{}
  \begin{description}\setlength\itemsep{0pt}%
  \item[subitem] \mbox{}
    \begin{description}\setlength\itemsep{0pt}%
    \item[subsubitem] 1
    \end{description}
  \end{description}
\item[M] \null
\item[minerals] \mbox{}
  \begin{description}\setlength\itemsep{0pt}%
  \item[amethyst] 1
  \item[corundum] 1
  \item[quartz] 1
  \end{description}
\item[V] \null
\item[vegetable] 1
  \begin{description}\setlength\itemsep{0pt}%
  \item[cabbage] 1
  \end{description}
\end{description}
(The \qt{1} in the above indicates the page number.)
There are some entries here that look a little odd: \code{duck},
\code{cabbage} and \code{subsubitem}. In each case they are a
lone child entry. It would look better if they could be compressed,
but I don't want to use the \csopt{flatten} option, as I still want
to keep the mineral hierarchy.

If I now add \csopt[postsort]{flatten-lonely}:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries.bib]{src},\csopt[postsort]{flatten-lonely}}
\end{codeenv}
the hierarchy becomes:
\begin{description}\setlength\itemsep{0pt}%
\item[A] \null
\item[aardvark] 1
\item[B] \null
\item[bard] 1
\item[birds, duck] 1
\item[buzz] 1
\item[I] \null
\item[item, subitem, subsubitem] 1
\item[M] \null
\item[minerals] \mbox{}
  \begin{description}\setlength\itemsep{0pt}%
  \item[amethyst] 1
  \item[corundum] 1
  \item[quartz] 1
  \end{description}
\item[V] \null
\item[vegetable] 1
  \begin{description}\setlength\itemsep{0pt}%
  \item[cabbage] 1
  \end{description}
\end{description}
The \field{name} field of the \code{duck} entry has been
set to:
\begin{codeenv}
\field{name}=\marg{\gls{bibglsflattenedchildpostsort}\marg{birds}\marg{duck}}
\end{codeenv}
the \field{text} field has been set to:
\begin{codeenv}
\field{text}=\marg{duck}
\end{codeenv}
the \field{group} field is copied over from the parent entry (\qt{B}),
and the \field{parent} field has been adjusted, moving \code{duck}
up one hierarchical level.
Finally, the former parent \code{birds} entry has been removed (the default
\csopt[only unrecorded parents]{flatten-lonely-rule} is in effect).

The default definition of \gls!{bibglsflattenedchildpostsort}
formats its arguments so that they are separated by a comma and
space (\qt{birds, duck}). If the \field{text} field had been set
in the original \atentry{index} definition of \text{duck}, it
wouldn't have been altered. This adjustment ensures that in the
document \code{\cs{gls}\marg{duck}} still produces \qt{duck} rather than
\qt{birds, duck}.
(If the child and parent \field{name} fields are identical,
the terms are considered homographs. See below for further details.)

The \code{subsubitem} entry has also been adjusted. This was done
in a multi-stage process, starting with sub-items and then moving down
the hierarchical levels:
\begin{itemize}
\item The \code{subitem} entry was adjusted, moving it from a
sub-entry to a top-level entry. The \field{name} field was then
modified to:
\begin{codeenv}
\field{name}=\marg{\gls{bibglsflattenedchildpostsort}\marg{item}\marg{subitem}}
\end{codeenv}
This now means that the \code{subsubitem} entry is now a sub-entry
(rather than a sub-sub-entry). The \code{subitem} entry now has no
parent, but at this stage the \code{subsubitem} entry still has
\code{subitem} as its parent.
\item The \code{subsubitem} entry is then adjusted moving from a
sub-entry to a top-level entry. The \field{name} field was then
modified to:
\begin{codeenv}
\field{name}=
\marg{\comment{}
 \gls{bibglsflattenedchildpostsort}
 \marg{\comment{}
   \comment{name from former parent}
   \gls{bibglsflattenedchildpostsort}\marg{item}\marg{subitem}\comment{}
 }\comment{}
 \marg{subsubitem}\comment{original name}
}
\end{codeenv}
The first argument of \gls!{bibglsflattenedchildpostsort} is
obtained from the \field{name} field of the entry's former parent
(which is removed from the child's set of ancestors). This field
value was changed in the previous step, and the change is
reflected here.

This means that the name for \code{subitem} will be displayed as
\qt{item, subitem} and the name for \code{subsubitem} will be
displayed as \qt{item, subitem, subsubitem}.

\item The parent entries \code{item} and \code{subitem}
are removed from the selection as they have no location lists.
\end{itemize}

Note that the \code{cabbage} sub-entry hasn't been adjusted. It
doesn't have any siblings but its parent entry (\code{vegetable})
has a location list so it can't be discarded.
If I change the rule:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries.bib]{src},
 \csopt[discard unrecorded]{flatten-lonely-rule},
 \csopt[postsort]{flatten-lonely}
}
\end{codeenv}
then this will move the \code{cabbage} entry up a level but the
original parent entry \code{vegetable} will remain:
\begin{description}\setlength\itemsep{0pt}%
\item[A] \null
\item[aardvark] 1
\item[B] \null
\item[bard] 1
\item[birds, duck] 1
\item[buzz] 1
\item[I] \null
\item[item, subitem, subsubitem] 1
\item[M] \null
\item[minerals] \mbox{}
  \begin{description}\setlength\itemsep{0pt}%
  \item[amethyst] 1
  \item[corundum] 1
  \item[quartz] 1
  \end{description}
\item[V] \null
\item[vegetable] 1
\item[vegetable, cabbage] 1
\end{description}

Remember that \csopt[postsort]{flatten-lonely} performs the
adjustment after sorting. This means that the entries are still in
the same relative location that they were in with the original
\csopt[false]{flatten-lonely} setting. For example, \code{duck}
remains in the B letter group before \qt{buzz}.

With \csopt[presort]{flatten-lonely} the adjustments are made before
the sorting is performed. For example, using:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries.bib]{src},
 \csopt[discard unrecorded]{flatten-lonely-rule},
 \csopt[presort]{flatten-lonely}
}
\end{codeenv}
the hierarchical order is now:
\begin{description}\setlength\itemsep{0pt}%
\item[A] \null
\item[aardvark] 1
\item[B] \null
\item[bard] 1
\item[buzz] 1
\item[C] \null
\item[cabbage] 1
\item[D] \null
\item[duck] 1
\item[M] \null
\item[minerals] \mbox{}
  \begin{description}\setlength\itemsep{0pt}%
  \item[amethyst] 1
  \item[corundum] 1
  \item[quartz] 1
  \end{description}
\item[S] \null
\item[subsubitem] 1
\item[V] \null
\item[vegetable] 1
\end{description}
This method uses a different format for the modified \field{name}
field. For example, the \code{duck} entry now has:
\begin{codeenv}
\field{name}=\marg{\gls{bibglsflattenedchildpresort}\marg{duck}\marg{birds}}
\end{codeenv}
The default definition of \gls{bibglsflattenedchildpresort} simply
does the first argument and ignores the second. The sorting is then
performed, but the interpreter recognises this command and can
deduce that the sort value for this entry should be \code{duck},
so \qt{duck} now ends up in the D letter group.

If you provide a definition of \gls{bibglsflattenedchildpresort} in
the \atentry{preamble}, it will be picked up by the interpreter.
For example:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\gls{bibglsflattenedchildpresort}}[2]\marg{\idx{param}1 (\idx{param}2)}}}
\end{codeenv}

Note that the \field{text} field is only changed if not already set.
This option may have unpredictable results for abbreviations as the
\field{name} field (and sometimes the \field{text} field) is
typically set by the abbreviation style.  Remember that
if the parent entry doesn't have a \idx{locationlist} and the rule isn't
set to \optfmt{no discard} then the parent entry will be discarded
after all relevant entries and their dependencies have been
selected, so any cross-references within the parent entry (such as
\ics{gls} occurring in the description) may end up being selected
even if they wouldn't be selected if the parent entry didn't exist.

With both \optfmt{presort} and \optfmt{postsort},
if the parent \field{name} is the same as the child's \field{name}
then the child is considered a homograph and
the child's name is set to:
\begin{codeenv*}
\format{bibglsflattenedhomograph}
\end{codeenv*}
instead of the corresponding \csfmt{bibglsflattenedchild\ldots sort}. 
This defaults to just \meta{name}.


\optsection{flatten-lonely-rule}

This option governs the rule used by \csopt{flatten-lonely} to
determine which sub-entries (that have no siblings) to adjust and 
which parents to remove. The value may be one of the following:
\begin{description}
\item[\optfmt{only unrecorded parents}] Only the sub-entries
that have a parent without a \idx{locationlist} will be altered.
The parent entry will be removed from the selection.
This value is the default setting.
\item[\optfmt{discard unrecorded}] This setting will adjust all
sub-entries that have no siblings regardless of whether or not the
parent has a \idx{locationlist}.
Only the parent entries that don't have a location list will be
removed from the selection.
\item[\optfmt{no discard}] This setting will adjust all
sub-entries that don't have siblings regardless of whether or not the
parent has a \idx{locationlist}. No entries will be discarded, so parent
entries that don't have a \idx{locationlist} will still appear in the
glossary.
\end{description}
In the above, the \idx{locationlist} includes records and
cross-references obtained from the \field{see} or \field{seealso}
fields. See \csopt{flatten-lonely} for further details.

\optsection{strip-missing-parents}

The \sty{glossaries} package requires that all child entries must be
defined after the parent entry. An error occurs otherwise, so
\bibgls\ will omit the \field{parent} field if it can't be found in
the given \idx{resourceset}. However, when the default
\csopt[false]{strip-missing-parents} is on, this omission only occurs
while writing the definitions in the \ext{glstex} file (after
selection and sorting).

Sorting is performed hierarchically and the \field{group} field is set
accordingly for the top-level entries (but not for child entries),
which means that an entry with a \field{parent} field will be
treated by the sort method as a child entry. This can lead to a
strange result, which \bibgls\ warns about:
\begin{alltt}
Parent '\meta{parent id}' not found for entry \meta{child-id}
\end{alltt}

This is the default behaviour as it may simply be a result of a
typing mistake in the \field{parent} field. If you actually want
missing parents to be stripped before sorting (but after the
selection process) then use \csopt[true]{strip-missing-parents}. If
you want all parents stripped then use \csopt{flatten} or
\csopt[parent]{ignore-fields} instead. As from version 1.4, if you
want \bibgls\ to create the missing parents, then you can use 
\csopt[create]{missing-parents}.

\optsection{missing-parents}

As an alternative to \csopt{strip-missing-parents}, as from version
1.4 you can now use \csopt[\meta{value}]{missing-parents} where
\meta{value} may be one of:
\begin{itemize}
\item \optfmt{strip}: this is equivalent to
\csopt[true]{strip-missing-parents};

\item \optfmt{warn}: this is equivalent to the default
\csopt[false]{strip-missing-parents};

\item \optfmt{create}: this will create a new \atentry{index}
entry with the missing parent's label (after it's been processed by
options such as \csopt{labelify}) with the \field{name} obtained
from the \emph{original} value of the \field{parent} field (before being
processed by options like \csopt{labelify}).
If the child entry has the \field{type} field set, then the new
parent entry will be given the same value. The \field{category} for
the new parent entry can be assigned with
\csopt{missing-parent-category}.
\end{itemize}

For example, consider the \exfile!{books.bib} file which contains
entries like:
\begin{codeenv}
\atentry{entry}\marg{ubik,
  \field{name}=\marg{Ubik},
  \field{description}=\marg{novel by Philip K. Dick},
  \fieldfmt{identifier}=\marg{book},
  \fieldfmt{author}=\marg{\gls{sortmediacreator}\marg{Philip K.}\marg{Dick}},
  \fieldfmt{year}=\marg{1969}
}
\end{codeenv}
then the field alias:
\begin{codeenv}
\csopt[\fieldfmt{author}=\field{parent}]{field-aliases}
\end{codeenv}
will treat:
\begin{codeenv}
  \fieldfmt{author}=\marg{\gls{sortmediacreator}\marg{Philip K.}\marg{Dick}},
\end{codeenv}
as though it had been defined as:
\begin{codeenv}
  \field{parent}=\marg{\gls{sortmediacreator}\marg{Philip K.}\marg{Dick}},
\end{codeenv}
This can be converted into a label with the options:
\begin{codeenv}
 \csopt[parent]{labelify},
 \csopt[
   \marg{[ \cs{cs.string}\csfmt{.}]}\marg{}
 ]{labelify-replace}
\end{codeenv}
If the interpreter has been provided with the definition:
\begin{codeenv}
\cs{providecommand}*\marg{\gls{sortmediacreator}}[2]\marg{\idx{param}2 \idx{param}1}
\end{codeenv}
then the \field{parent} field for the \code{ubik} entry
will become \code{DickPhilipK} but the original value is stored
internally when \csopt[create]{missing-parents} is set so that it
can be used as the \field{name} if the parent needs to be created.
Once all the entries have been processed, if \code{ubik} has been
selected but no entry can be found with the label \code{DickPhilipK}
then a new entry will be added as though it had been defined with:
\begin{codeenv}
\atentry{index}\marg{DickPhilipK,
  \field{name}=\marg{\gls{sortmediacreator}\marg{Philip K.}\marg{Dick}}
}
\end{codeenv}
This is an alternative approach to the \exfile{sample-authors.tex}
document from the \hyperref[sec:examples]{examples chapter}.

\optsection{missing-parent-category}

If a missing parent entry is created through the use of
\csopt[create]{missing-parents} then the \field{category} field can
be assigned to the new parent entry with this option. The
\meta{value} may be one of:
\begin{itemize}
 \item \code{same as child}: the parent entry's \field{category}
  field is set to the same value as the child's (if set);
 \item \code{same as base}: the parent entry's \field{category} is
  set to the base name of the \ext{bib} file that provided the
  child entry's definition;
 \item \code{no value}: don't set the \field{category} field;
 \item \meta{label}: the parent entry's \field{category} field is set to
  \meta{label} (which shouldn't contain any special characters). 
\end{itemize}
The default setting is \csopt[no value]{missing-parent-category}.

\section{Master Documents}
\label{sec:master}

Suppose you have two documents \filefmt{mybook.tex} and
\filefmt{myarticle.tex} that share a common glossary that's shown
in \filefmt{mybook.pdf} but not in \filefmt{myarticle.pdf}.
Furthermore, you'd like to use \isty{hyperref} and be able to click
on a term in \filefmt{myarticle.pdf} and be taken to the relevant 
page in \filefmt{mybook.pdf} where the term is listed in the
glossary.

This can be achieved with the \catattr{targeturl} and
\catattr{targetname} category attributes. For example, without
\bibgls\ the file \filefmt{mybook.tex} might look like:
\begin{codeenv}
\cmd{documentclass}\marg{book}
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}\marg{glossaries-extra}
\strut
\cs{cs.makeglossaries}
\strut
\gls{newglossaryentry}\marg{sample}\marg{\field{name}=\marg{sample},\field{description}=\marg{an example}}
\strut
\cmd{begin}\marg{document}
\cs{chapter}\marg{Example}
\cs{gls}\marg{sample}.
\strut
\cs{printglossaries}
\cmd{end}\marg{document}
\end{codeenv}
The other document \filefmt{myarticle.tex} might look like:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}\marg{glossaries-extra}
\strut
\cs{newignoredglossary*}\marg{external}
\ics{glssetcategoryattribute}\marg{external}\marg{targeturl}\marg{mybook.pdf}
\cs{glssetcategoryattribute}\marg{external}\marg{targetname}\marg{\cs{glolinkprefix}\cs{glslabel}}
\strut
\gls{newglossaryentry}\marg{sample}\marg{\field{type}=external,\field{category}=external,
 \field{name}=\marg{sample},\field{description}=\marg{an example}}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{sample}.
\cmd{end}\marg{document}
\end{codeenv}
In this case the \optfmt{main} glossary isn't used, but the category 
attributes allow a mixture of internal and external references, so
the \optfmt{main} glossary could be used for the internal
references. (In which case, \csfmt{makeglossaries} and
\csfmt{printglossaries} would need to be added back to
\filefmt{myarticle.tex}.)

Note that both documents had to define the common terms. The above
documents can be rewritten to work with \bibgls. First a \ext{bib}
file needs to be created:
\begin{codeenv}
\atentry{entry}\marg{sample,
  \field{name}=\marg{sample},
  \field{description}=\marg{an example}
}
\end{codeenv}
Assuming this file is called \filefmt{myentries.bib}, then
\filefmt{mybook.tex} can be changed to:
\begin{codeenv}
\cmd{documentclass}\marg{book}
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}[record]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[myentries]{src}}
\strut
\cmd{begin}\marg{document}
\cs{chapter}\marg{Example}
\cs{gls}\marg{sample}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
and \filefmt{myarticle.tex} can be changed to:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}[record]\marg{glossaries-extra}
\strut
\cs{newignoredglossary*}\marg{external}
\cs{glssetcategoryattribute}\marg{external}\marg{targeturl}\marg{mybook.pdf}
\cs{glssetcategoryattribute}\marg{external}\marg{targetname}\marg{\cs{glolinkprefix}\cs{glslabel}}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[myentries]{src},
 \csopt[none]{sort},
 \csopt[external]{type},
 \csopt[external]{category}
}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{sample}.
\cmd{end}\marg{document}
\end{codeenv}
Most of the options related to sorting and the glossary format are
unneeded here since the glossary isn't being displayed. This may be
sufficient for your needs, but it may be that the book has changed
various settings that have been written to \filefmt{mybook.glstex}
but aren't present in the \iext{bib} file (such as
\csopt[uc]{short-case-change}). In this case, you could just remember to copy
over the settings from \filefmt{mybook.tex} to
\filefmt{myarticle.tex}, but
another possibility is to simply make \filefmt{myarticle.tex} input
\filefmt{mybook.glstex} instead of using \gls{GlsXtrLoadResources}. This
can work but it's not so convenient to set the label prefix, the
type and the category.  The \optfmt{master} option allows this, but
it has limitations (see below), so in complex cases (in particular
different label prefixes combined with hierarchical entries or cross-references) you'll have to use
the method shown in the example code above.

\optsection{master}

This option will disable most of the options that relate to
parsing and processing data contained in \iext{bib} files
(since this option doesn't actually read any \ext{bib} files).
It also can't be used with \csopt[copy]{action} or \csopt[define
or copy]{action}.

The use of \optfmt{master} isn't always suitable. In particular
if any of the terms cross-reference each other, such as through
the \field{see} or \field{seealso} field or the \field{parent} field or
using commands like \ics{gls} in any of the other fields when
the labels have been assigned prefixes. In this
case you will need to use the method described in the example above.

The \meta{name} is the name of the \iext{aux} file for the master
document without the extension (in this case, \filefmt{mybook}). It
needs to be relative to the document referencing it or an absolute
path using forward slashes as the directory divider. Remember that
if it's a relative path, the PDF files (\filefmt{mybook.pdf} and
\filefmt{myarticle.pdf}) will also need to be located in the same
relative position.

When \bibgls\ detects the \csopt{master} option, it won't search for
entries in any \ext{bib} files (for that particular resource set)
but will create a \iext{glstex} file that inputs the master
document's
\ext{glstex} files, but it will additionally temporarily
adjust the internal commands used to define entries so that
the prefix given by \csopt{label-prefix}, the glossary type and the
category type are all automatically inserted. If the \csopt{type}
or \csopt{category} options haven't been used, the corresponding
value will default to \optfmt{master}. The \catattr{targeturl}
and \catattr{targetname} category attributes will automatically be
set, and the glossary type will be provided using
\code{\ics{provideignoredglossary*}\margm{type}} (even if
\longarg{no-provide-glossaries} is set).

The above \filefmt{myarticle.tex} can be changed to:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}[record]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[book.]{label-prefix},\csopt[mybook]{master}}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{book.sample}.
\cmd{end}\marg{document}
\end{codeenv}

There are some settings from the master document that you 
still need to repeat in the other document. These include
the label prefixes set when the master document loaded
the resource files, and any settings in the master document
that relate to the master document's entries.

For example, if the master document loaded a resource file
with \csopt[term.]{label-prefix} then you also need this
prefix when you reference the entries in the dependent document
in addition to the \optfmt{label-prefix} for the dependent document.
Suppose \filefmt{mybook.tex} loads the resources using:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[myentries]{src},\csopt[term.]{label-prefix}}
\end{codeenv}
and \filefmt{myarticle.tex} loads the resources using:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[book.]{label-prefix},\csopt[mybook]{master}}
\end{codeenv}
Then the entries referenced in \filefmt{myarticle.tex} need
to use the prefix \optfmt{book.term.}\ as in:
\begin{codeenv}
This is a \cs{gls}\marg{book.term.sample} term.
\end{codeenv}

Remember that the category labels will need adjusting to reflect the
change in category label in the dependent document.

For example, if \filefmt{mybook.tex} included:
\begin{codeenv}
\cs{setabbreviationstyle}\marg{long-short-sc}
\end{codeenv}
then \filefmt{myarticle.tex} will need:
\begin{codeenv}
\cs{setabbreviationstyle}\oarg{master}\marg{long-short-sc}
\end{codeenv}
(change \optfmt{master} to \meta{value} if you have used
\csopt[\meta{value}]{category}). You can, of course, choose
a different abbreviation style for the dependent document, 
but the category in the optional argument needs to be correct.

\optsection{master-resources}

If the master document has multiple resource files
then by default all the master document's
\iext{glstex} files will be input. If you don't want them all
you can use \optfmt{master-resources} to specify
only those files that should be included. The value \meta{list} is 
a comma-separated list of names, where each name corresponds
to the final argument of \gls!{glsxtrresourcefile}.
Remember that \gls{GlsXtrLoadResources} is just a shortcut
for \gls{glsxtrresourcefile} that bases the name on \ics{jobname}.
(Note that, as with the argument of \gls{glsxtrresourcefile},
the \ext{glstex} extension should not be included in \meta{list}.) The file
\primaryresourcefmt\ is considered the primary
resource file and the files \suppresourcefmt{\meta{n}}
(starting with \meta{n} equal to 1) are considered the supplementary resource files.

For example, to just select the first and third of the
supplementary resource files (omitting the primary
\filefmt{mybook.glstex}):
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[mybook]{master},
  \csopt[mybook-1,mybook-3]{master-resources}
}
\end{codeenv}

\section{Field and Label Options}
\label{sec:fieldlabelopts}

The options in this section may be used to set or adjust field
values or labels.  Some field values are expected to be labels (such
as \field{group}).  These labels must not contain special
characters or commands, but it's possible to convert a field value
into a valid label using options such as \csopt{labelify}.

\subsection{Label Options}
\label{sec:labelopts}

\optsection[\subsubsection]{interpret-label-fields}

This is a boolean option that determines whether or not the fields
that may only contain labels should have their values interpreted
(\field{parent}, \field{category}, \field{type}, \field{group},
\field{seealso} and \field{alias}). Although this option interprets
commands within those fields, it doesn't strip any characters that
can't be used within a label. The \field{see} field isn't included
as it may optionally start with \code{\oargm{tag}} where \meta{tag}
may legitimately contain \LaTeX\ code that shouldn't be interpreted.

The default setting is \csopt[false]{interpret-label-fields}.
Note that if this setting is on, \idxpl{crossresourceref} aren't
permitted. This setting has no effect if the interpreter has been
disabled.

Related settings are \csopt{labelify} and
\csopt{labelify-list} which can be used to strip content that
can't be used in labels and may be used more generally for other
fields. The \csopt{labelify} and \csopt{labelify-list} options
are performed before \csopt{interpret-label-fields}.

\optsection[\subsubsection]{labelify}

This option should take a comma-separated list of recognised field names as the
value. (If a field is present in both \csopt{labelify} and
\csopt{labelify-list}, then \csopt{labelify-list} takes precedence.) 
Note that if this setting is on, \idxpl{crossresourceref} aren't
permitted. The value is required for this key but may be empty,
which indicates an empty set of fields (that is, the setting is 
switched off).

Each listed field will be converted into a string suitable
for use as a label. (Not necessarily a glossary entry label, but
any label that may be used in the construction of a control sequence
name.) 

The conversion is performed in the following order:
\begin{enumerate}
\item If the interpreter is on and the field value contains
any of the characters \idx{escchar}~(backslash), 
\idx{bgroupchar}~(begin group), \idx{egroupchar}~(end group), 
\idx{nbspchar}~(non-breakable space) or \idx{mshiftchar}~(maths shift), 
then the value is interpreted.

\item Any substitutions that have been specified with
\csopt{labelify-replace} are performed.

\item All characters that aren't alphanumeric or the space character
 or any of the following punctuation
characters \code{.}~(full stop), \code{-}~(hyphen), \code{+}~(plus),
\code{:}~(colon), \code{;}~(semi-colon), \code{\string|}~(pipe),
\code{/}~(forward slash), \code{!}~(exclamation mark), 
\code{?}~(question mark), \code{*}~(asterisk), \code{<}~(less than), 
\code{>}~(greater than), \code{\textasciigrave}~(backtick), 
\code{'}~(apostrophe) or \code{@}~(at-sign) are stripped. If you want to
retain commas, use \csopt{labelify-list} instead. If you want to
strip any of the allowed punctuation, use \csopt{labelify-replace} to
remove the unwanted characters. (Remember that \isty{babel} can make
some of these punctuation characters active, in which case they need
to be stripped.)

\item If \bibgls\ hasn't detected \sty{fontspec} in the document's
transcript file, the value is then decomposed and all non-ASCII
characters are removed.
\end{enumerate}

For example, suppose the \ext{bib} file contains:
\begin{codeenv}
\atentry{index}\marg{sample,
  \field{name}=\marg{\ics{AA} ngstr\ics{umlaut}om, \ics{O} stergaard, d'Arcy, and Fotheringay-Smythe}
}
\end{codeenv}
Then:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries]{src},\comment{data in entries.bib}
 \csopt[name]{labelify}
}
\end{codeenv}
will convert the \field{name} field into:
\begin{verbatim}
Angstrom stergaard d'Arcy and Fotheringay-Smythe
\end{verbatim}
if the document hasn't used \sty{fontspec} otherwise it will be:
\begin{verbatim}
Ångström Østergaard d'Arcy and Fotheringay-Smythe
\end{verbatim}
Note that \O\ is considered an unmodified letter and so can't
be decomposed into a basic Latin letter with a combining diacritic.
It's therefore removed completely from the non-\sty{fontspec}
version. Whereas \AA\ can be decomposed into \qt{A} followed by the
\qt{combining ring above} character and \"o can be decomposed into \qt{o}
followed by the \qt{combining diaresis} character. You can use
\csopt{labelify-replace} to replace non-ASCII characters into
the closest match. Alternatively, switch to using \XeLaTeX\ or
\LuaLaTeX.

You can use this option with \csopt{replicate-fields} if you need
to retain the original:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries]{src},\comment{data in entries.bib}
 \csopt[\field{name}=\marg{\field{user1}}]{replicate-fields},
 \csopt[\field{user1}]{labelify}
}
\end{codeenv}

\optsection[\subsubsection]{labelify-list}

This option is like \csopt{labelify} but it retains commas, as it's
designed for fields that should be converted into a comma-separated
list of labels. Any empty elements are removed. For example, with
the \ext{bib} entry from above:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries]{src},\comment{data in entries.bib}
 \csopt[\field{name}=\marg{\field{user1}}]{replicate-fields},
 \csopt[\field{user1}]{labelify-list}
}
\end{codeenv}
will convert the \field{user1} field into:
\begin{verbatim}
Angstrom, stergaard, d'Arcy, and Fotheringay-Smythe
\end{verbatim}
or:
\begin{verbatim}
Ångström, Østergaard, d'Arcy, and Fotheringay-Smythe
\end{verbatim}
depending on whether or not \sty{fontspec} was detected.

\optsection[\subsubsection]{labelify-replace}

This option takes a comma-separated list as a value with each
element in the list in the form \code{\margm{regex}\margm{replacement}}
where \meta{regex} is a \idx{regex} (that conforms to \href{http://docs.oracle.com/javase/8/docs/api/java/util/regex/Pattern.html}{Java's
Pattern class}~\cite{pattern}) and \meta{replacement}
is the replacement text. The value is required for this key but
may be empty, which indicates that the
setting is switched off.

Remember that the argument of
\gls{GlsXtrLoadResources} is expanded when written to the \ext{aux}
file so take care to protect any special characters. For example, to
match a literal \idx{full-stop} use \cs{cs.string}\idx{cs.period}
rather than just \idx{cs.period} (backslash dot).

In the \meta{replacement} part, you can use
\ics{glscapturedgroup}\meta{n} to reference a captured sub-sequence.
For example:
\begin{codeenv}
\csopt[\marg{([A-Z])\ics{cs.string}\idx{cs.period}}\marg{\cs{glscapturedgroup}1}]{labelify-replace}
\end{codeenv} 
This removes any \idx{full-stop} that follows any of the characters
A,\ldots,Z. Alternatively, you can just use \verb|\string\$|
instead of \cs{glscapturedgroup}. If you want a literal dollar
character, you need to use \cs{glshex}\code{24} (or
\verb|\string\u24|). This isn't recommended for labels (since special
characters are automatically stripped), but
\csopt{sort-replace} follows the same rules as
\csopt{labelify-replace}, and it may be needed
for that.

Both \csopt{labelify} and \csopt{labelify-list} use the
\csopt{labelify-replace} setting to 
perform substitutions. For example, to replace the sub-string \qt{ and }
(including spaces) with a comma:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries]{src},\comment{data in entries.bib}
 \csopt[\field{name}=\marg{\field{user1}}]{replicate-fields},
 \csopt[\marg{ and }\marg{,}]{labelify-replace},
 \csopt[\field{user1}]{labelify-list}
}
\end{codeenv}
The earlier example will now end up as:
\begin{verbatim}
Angstrom, stergaard, d'Arcy,Fotheringay-Smythe
\end{verbatim}
or:
\begin{verbatim}
Ångström, Østergaard, d'Arcy,Fotheringay-Smythe
\end{verbatim}
depending on whether or not \sty{fontspec} was detected.

Note that this produces the same result regardless of whether or not
the Oxford comma is present as \verb*|, and | would first be
converted to \verb|,,| and then the empty element is removed
resulting in a single comma.

You can have more than one replacement:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries]{src},\comment{data in entries.bib}
 \csopt[\field{name}=\marg{\field{user1}}]{replicate-fields},
 \csopt[
   \marg{ and }\marg{,},\comment{first substitution}
   \marg{[ '\cs{cs.string}\cmd{-}]}\marg{},\comment{second substitution}
   \marg{\ics{glshex}00D8}\marg{O}\comment{third substitution}
 ]{labelify-replace},
 \csopt[\field{user1}]{labelify-list}
}
\end{codeenv}
This additionally removes the space, apostrophe and hyphen
characters (second substitution) and replaces \qt{\O} (\hex{00D8})
with \qt{O} (third substitution) so the string now ends up as:
\begin{verbatim}
Angstrom,Ostergaard,dArcy,FotheringaySmythe
\end{verbatim}
or:
\begin{verbatim}
Ångström,Ostergaard,dArcy,FotheringaySmythe
\end{verbatim}
depending on whether or not \sty{fontspec} was detected.

\optsection[\subsubsection]{label-prefix}

The \csopt{label-prefix} option prepends \meta{tag} to each entry's label. This
\meta{tag} will also be inserted in front of any cross-references, unless they
start with \idprefix{dual} or \idprefix{tertiary} or \idprefix{extn} (where
\meta{n} is an integer). Use \csopt{dual-prefix} to change the dual label
prefixes and \csopt{ext-prefixes} to change the external label prefixes.

As from version 1.8, the primary label prefix is identified
in the \ext{glstex} file with:
\begin{codeenv}
\format{bibglsprimaryprefixlabel}
\end{codeenv}

For example, if the \ext{bib} file contains:
\begin{codeenv}
\atentry{entry}\marg{bird,
  \field{name}=\marg{bird},
  \field{description} = \marg{feathered animal, such as a \cs{gls}\marg{duck} or \cs{gls}\marg{goose}}
}
\strut
\atentry{entry}\marg{waterfowl,
  \field{name}=\marg{waterfowl},
  \field{description}=\marg{Any \cs{gls}\marg{bird} that lives in or about water},
  \field{see}=\marg{[see also]\marg{duck,goose}}
}
\strut
\atentry{index}\marg{duck}
\strut
\atentry{index}\marg{goose,\field{plural}=\qtdelim{geese}}
\end{codeenv}
Then if this \ext{bib} file is loaded with \csopt[gls.]{label-prefix}
it's as though the entries had been defined as:
\begin{codeenv}
\atentry{entry}\marg{gls.bird,
  \field{name}=\marg{bird},
  \field{description} = \marg{feathered animal, such as a \cs{gls}\marg{gls.duck} or \cs{gls}\marg{gls.goose}}
}
\strut
\atentry{entry}\marg{gls.waterfowl,
  \field{name}=\marg{waterfowl},
  \field{description}=\marg{Any \cs{gls}\marg{gls.bird} that lives in or about water},
  \field{see}=\marg{[see also]\marg{gls.duck,gls.goose}}
}
\strut
\atentry{index}\marg{gls.duck,\field{name}=\marg{duck}}
\strut
\atentry{index}\marg{gls.goose,\field{name}=\marg{goose},\field{plural}=\qtdelim{geese}}
\end{codeenv}

Remember to use this prefix when you reference the terms in the
document with commands like \ics{gls}.

\optsection[\subsubsection]{duplicate-label-suffix}

The \sty{glossaries} package doesn't permit entries with duplicate
labels (even if they're in different glossaries). If you
have multiple \idxpl{resourceset} and an entry that's selected
in one \idx{resourceset} is also selected in another, by
default, \bibgls\ will issue a warning, but it will still write the
entry definition to the \ext{glstex} file, which means you'll also
get a warning from \sty{glossaries-extra} and the duplicate
definition will be ignored, but associated internal fields
set with commands like \cs{GlsXtrSetField} may still be set.

If you actually want the duplicate, you need to specify a
suffix with \csopt{duplicate-label-suffix}. This suffix is only
set just before writing the entry definition to the \ext{glstex}
file, so it doesn't affect selection criteria nor can label
substitutions be performed in any cross-references. Options such as
\csopt{set-widest} that reference entry labels are incompatible
as they will use the unsuffixed label.

The actual suffix is formed from \meta{value}\meta{n} where \meta{n}
is an integer that's incremented in the event of multiple
duplicates.  For example, \csopt[.copy]{duplicate-label-suffix} will
change the label to \meta{id}\code{.copy1} for the first duplicate
of the entry whose label is \meta{id}, and \meta{id}\code{.copy2} for
the second duplicate, etc.

\optsection[\subsubsection]{record-label-prefix}

If set, this option will cause \bibgls\ to pretend that each record
label starts with \meta{tag}, if it doesn't already. For example, suppose
the records in the \ext{aux} file are:
\begin{codeenv}
\gls{glsxtr@record}\marg{bird}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\gls{glsxtr@record}\marg{waterfowl}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\gls{glsxtr@record}\marg{idx.duck}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\gls{glsxtr@record}\marg{idx.goose}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\end{codeenv}
The use of \csopt[idx.]{record-label-prefix} makes \bibgls\ act as 
though the records were given as:
\begin{codeenv}
\gls{glsxtr@record}\marg{idx.bird}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\gls{glsxtr@record}\marg{idx.waterfowl}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\gls{glsxtr@record}\marg{idx.duck}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\gls{glsxtr@record}\marg{idx.goose}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\end{codeenv}


\optsection[\subsubsection]{cs-label-prefix}

If you have commands such as \code{\ics{gls}\margm{label}} or
\code{\ics{glstext}\margm{label}} in field
values (in situations where nested link text won't cause a problem)
the \meta{label} will be converted as follows:
\begin{itemize}
\item if \meta{label} starts with \idprefix{dual} then
\idprefix{dual} will be replaced by the \csopt{dual-prefix} value;
\item if \meta{label} starts with \idprefix{tertiary} then
\idprefix{tertiary} will be replaced by the \csopt{tertiary-prefix} value;
\item if \meta{label} starts with \idprefix{extn} then
\idprefix{extn} will be replaced by the corresponding
\csopt{ext-prefixes} setting (if \idx{crossresourceref} mode is
enabled, see \sectionref{sec:resourcesets});
\item if \meta{label} doesn't start with one of the above recognised
prefixes then, if \csopt{cs-label-prefix} has been used the supplied
value will be inserted otherwise the \csopt{label-prefix} setting
will be inserted.
\end{itemize}
For example, given:
\begin{codeenv}
\atentry{entry}\marg{bird,
  \field{name}=\marg{bird},
  \field{description} = \marg{feathered animal, such as a \cs{gls}\marg{duck} or \cs{gls}\marg{goose}}
}
\end{codeenv}
then if \csopt[idx.]{label-prefix} is set but \csopt{cs-label-prefix}
isn't included in the resource option list this will convert the
\field{description} field to:
\begin{codeenv}
\field{description} = \marg{feathered animal, such as a \cs{gls}\marg{idx.duck} or
\cs{gls}\marg{idx.goose}}
\end{codeenv}
However with \csopt[gls.]{cs-label-prefix} the \field{description}
field will be converted to:
\begin{codeenv}
\field{description} = \marg{feathered animal, such as a \cs{gls}\marg{gls.duck} or
\cs{gls}\marg{gls.goose}}
\end{codeenv}
regardless of the \csopt{label-prefix} setting. Whereas if the
original entry definition is:
\begin{codeenv}
\atentry{entry}\marg{bird,
  \field{name}=\marg{bird},
  \field{description} = \marg{feathered animal, such as a \cs{gls}\marg{dual.duck} or
\cs{gls}\marg{dual.goose}}
}
\end{codeenv}
then \idprefix{dual} will be replaced by the value of the
\csopt{dual-prefix} option regardless of the \csopt{cs-label-prefix}
setting.

The \csopt{cs-label-prefix} setting doesn't affect labels in the
fields that have an entry label or label list as the value
(\field{parent}, \field{alias}, \field{see} and \field{seealso}).

\optsection[\subsubsection]{ext-prefixes}

Any cross-references in the \iext{bib} file that start with
\idprefix{extn} (where \meta{n} is a positive integer) will be 
substituted with the \meta{n}th tag listed in the comma-separated
\meta{list}. If there aren't that many items in the list, the
\idprefix{extn} will simply be removed. The default setting is
an empty list, which will strip all \idprefix{extn} prefixes.
Remember that \idx{crossresourceref} mode needs to be enabled for
this option to work (see \sectionref{sec:resourcesets}).

As from version 1.8, the external label prefixes are identified
in the \ext{glstex} file with:
\begin{codeenv}
\format{bibglsexternalprefixlabel}
\end{codeenv}

For example, suppose the file \filefmt{entries-terms.bib} contains:
\begin{codeenv}
\atentry{entry}\marg{set,
  \field{name}=\marg{set},
  \field{description}=\marg{collection of values, denoted \cs{gls}\marg{ext1.set}}
}
\end{codeenv}
and the file \filefmt{entries-symbols.bib} contains:
\begin{codeenv}
\atentry{symbol}\marg{set,
  \field{name}=\marg{\cs{ensuremath}\marg{\cs{mathcal}\marg{S}}},
  \field{description}=\marg{a \cs{gls}\marg{ext1.set}}
}
\end{codeenv}

These files both contain an entry with the label \code{set}
but the \field{description} field includes \code{\cs{gls}\marg{ext1.set}} 
which is referencing the entry from the other file. These
two files can be loaded without conflict using:
\begin{codeenv}
\cmd{usepackage}[\styopt{record},\styopt{symbols}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-terms]{src},
 \csopt[gls.]{label-prefix},
 \csopt[sym.]{ext-prefixes}
}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-symbols]{src},
 \csopt[symbols]{type},
 \csopt[sym.]{label-prefix},
 \csopt[gls.]{ext-prefixes}
}
\end{codeenv}

Now the \code{set} entry from \filefmt{entries-terms.bib}
will be defined with the label \code{gls.set} and the
description will be:
\begin{codeenv}
collection of values, denoted \cs{gls}\marg{sym.set}
\end{codeenv}
The \code{set} entry
from \filefmt{entries-symbols.bib} will be defined with the label
\code{sym.set} and the description will be: 
\begin{codeenv}
a \cs{gls}\marg{gls.set}
\end{codeenv}

Note that in this case the \ext{bib} files have to be loaded 
as two separate resources. They can't be combined into a 
single \csopt{src} list as the labels aren't unique. 

If you want to allow the flexibility to choose between 
loading them together or separately, you'll have to give them
unique labels. For example, \filefmt{entries-terms.bib} could
contain:
\begin{codeenv}
\atentry{entry}\marg{set,
  \field{name}=\marg{set},
  \field{description}=\marg{collection of values, denoted \cs{gls}\marg{ext1.S}}
}
\end{codeenv}
and \filefmt{entries-symbols.bib} could contain:
\begin{codeenv}
\atentry{symbol}\marg{S,
  \field{name}=\marg{\cs{ensuremath}\marg{\cs{mathcal}\marg{S}}},
  \field{description}=\marg{a \cs{gls}\marg{ext1.set}}
}
\end{codeenv}
Now they can be combined with:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-terms,entries-symbols]{src}}
\end{codeenv}
which will simply strip the \glsdisp{idx.idprefix.extn}{\idprefixfmt{ext1}} 
prefix from the cross-references. Alternatively:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-terms,entries-symbols]{src},
 \csopt[gls.]{label-prefix},
 \csopt[gls.]{ext-prefixes}
}
\end{codeenv}
which will insert the supplied \csopt{label-prefix} at the
start of the labels in the entry definitions and will replace
the \idprefixfmt{ext1} prefix with \idprefixfmt{gls} in the
cross-references.

\optsection[\subsubsection]{save-original-id}

The \meta{value} may be either the keyword \code{false} or
the name of an internal field in which to store the entry's original
label (as given in the \ext{bib} file). The default setting is
\csopt[false]{save-original-id}. If \meta{value} is omitted,
\csopt[originalid]{save-original-id} is assumed.

If \meta{value} is a known field, it will be set after the field
aliases, otherwise it will simply be added to the \ext{glstex} file
using \ics{GlsXtrSetField} after the entry definition. 

\optsection[\subsubsection]{dependency-fields}

The \meta{list} should be a comma-separated list of fields that
have values in the form \code{\oargm{tag}\meta{id-list}} where
\meta{id-list} is a comma-separated list of entry labels. The value
is required for this key but may be empty, which indicates an empty
set of fields (that is, the setting is switched off).

This setting makes those fields act like the \field{see} field by
identifying the listed entries as dependencies, but the information
isn't added to the cross-reference part of the location list.  This
action is performed after \csopt{labelify-list}, if that's also set.

For example, suppose the file \filefmt{entries-en.bib} contains:
\begin{codeenv}
\atentry{index}\marg{cat,
  \fieldfmt{translations-pt}=\marg{gato,gatinho},
  \field{seealso}=\marg{kitten}
}
\strut
\atentry{index}\marg{kitten,
  \fieldfmt{translations-pt}=\marg{gato,gatinho}
}
\strut
\atentry{index}\marg{staple}
\atentry{index}\marg{rivet}
\atentry{index}\marg{mat}
\atentry{index}\marg{carpet}
\atentry{index}\marg{rug}
\atentry{index}\marg{tapestry}
\atentry{index}\marg{doormat}
\atentry{index}\marg{matting}
\atentry{index}\marg{coconut-matting,
  \field{name}=\marg{coconut matting}
}
\atentry{index}\marg{track}
\atentry{index}\marg{furrow}
\end{codeenv}
and suppose the file \filefmt{entries-pt.bib} contains:
\begin{codeenv}
\atentry{index}\marg{gato,
 \field{prefix}=\marg{o},
 \fieldfmt{translations-en}=\marg{cat,staple,rivet},
 \field{seealso}=\marg{gatinho}
}
\strut
\atentry{index}\marg{gatinho,
 \fieldfmt{translations-en}=\marg{kitten}
}
\strut
\atentry{index}\marg{tapete,
  \fieldfmt{translations-en}=\marg{carpet,rug,mat,tapestry}
}
\strut
\atentry{index}\marg{esteira,
  \field{prefix}=\marg{a},
  \fieldfmt{translations-en}=\marg{mat,track,matting,furrow}
}
\strut
\atentry{index}\marg{capacho,
  \field{prefix}=\marg{o},
  \fieldfmt{translations-en}=\marg{doormat,matting,mat,coconut-matting}
}
\end{codeenv}
The aim here is to have a document containing an 
English-to-Portuguese and a Portuguese-to-English dictionary. The
custom \fieldfmt{translations-pt} and \fieldfmt{translations-pt}
fields contain comma-separated lists of possible translations. In
this case I don't want to use the \field{see} field (and, in fact,
can't for the entries that have the \field{seealso} field set), but
I can identify the values of those fields as dependent entries to
ensure that they are selected even if they're not referenced in the
document.

For convenience I've aliased the custom fields to \field{user1}:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[T1]\marg{fontenc}
\cmd{usepackage}[utf8]\marg{inputenc}
\cmd{usepackage}[british,brazilian]\marg{babel}
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}\oarg{\styopt{record},
 \styopt{nomain},
 \styopt{nostyles},
 \styopt[bookindex]{stylemods},
 \styopt[\glostyle{bookindex}]{style}
}\marg{glossaries-extra}
\cmd{usepackage}\marg{\isty{glossaries-prefix}}
\strut
\cs{newglossary*}\marg{en}\marg{English Terms}
\cs{newglossary*}\marg{pt}\marg{Portuguese Terms}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[en]{type},
  \csopt[entries-en]{src},
  \csopt[en-GB]{sort},
  \csopt[en]{category},
  \csopt[\fieldfmt{translations-pt}=\field{user1}]{field-aliases},
  \csopt[\field{user1}]{dependency-fields},
  \csopt[\field{user1}:pt-BR:\encap{glsentryname}]{sort-label-list}
}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[pt]{type},
  \csopt[entries-pt]{src},
  \csopt[pt-BR]{sort},
  \csopt[pt]{category},
  \csopt[\fieldfmt{translations-en}=\field{user1}]{field-aliases},
  \csopt[\field{user1}]{dependency-fields},
  \csopt[\field{user1}:en-GB:\encap{glsentryname}]{sort-label-list}
}
\strut
\cs{apptoglossarypreamble}\oarg{en}\marg{\cs{selectlanguage}\marg{british}}
\cs{apptoglossarypreamble}\oarg{pt}\marg{\cs{selectlanguage}\marg{brazilian}}
\strut
\cmd{begin}\marg{document}
\cs{selectlanguage}\marg{british}
The \cs{gls}\marg{cat} sat on the \cs{gls}\marg{mat}.
\strut
\cs{selectlanguage}\marg{brazilian}
O \cs{gls}\marg{gato} sentou-se no \cs{gls}\marg{tapete}.
\strut
\cs{renewcommand}*\marg{\cs{glsxtrbookindexname}}[1]\marg{\comment{}
 \cs{glsxtrifhasfield}\marg{\field{prefix}}\marg{\idx{param}1}\marg{\ics{xmakefirstuc}\cs{glscurrentfieldvalue}\cs{cs.space}}\marg{}\comment{}
 \cs{glossentryname}\marg{\idx{param}1}\comment{}
 \cs{glsxtrifhasfield}\marg{\field{useri}}\marg{\idx{param}1}
 \marg{; translations: \cs{glsxtrseelist}\cs{glscurrentfieldvalue}}\marg{}\comment{}
}
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}

\subsection{Assignments}
\label{sec:fieldassignments}

\optsection[\subsubsection]{group}

The \csopt{group} option will set the \field{group} field to \meta{label}
unless \meta{label} is \optfmt{auto}. If \csopt[auto]{group} then if
the \longarg{group} switch is used the value of the \field{group}
field is set automatically during the sorting (see also \csopt{group-formation}
and \sectionref{sec:logicaldivisions}). If the \longarg{no-group}
setting is on then \csopt[auto]{group} does nothing.

The corresponding group title can be set with
\gls{glsxtrsetgrouptitle} in the document if the title is different
from the label. The default behaviour is \csopt[auto]{group}.

For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[integer]{sort},\csopt[Constants]{group},
 \csopt[entries-constants]{src}\comment{data in entries-constants.bib}
}
\gls{GlsXtrLoadResources}\oarg{\csopt[letter-case]{sort},\csopt[Variables]{group},
 \csopt[entries-variables]{src}\comment{data in entries-variables.bib}
}
\end{codeenv}
In this case, if the \field{type} field hasn't been set in the \ext{bib} files,
these entries will be added to the same glossary, but will
be grouped according to each instance of \gls{GlsXtrLoadResources},
with the provided group label.


\optsection[\subsubsection]{category}

The selected entries may all have their \field{category} field
changed before writing their definitions to the \ext{glstex} file.
The \meta{value} may be:
\begin{itemize}
\item \optfmt{same as entry}: set the
\field{category} to the \ext{bib} entry type used to define it
(\idx!{lowercase} and without the initial \code{@}) after any aliasing,
if applicable;

\item \optfmt{same as original entry}: (new to v1.4) set the \field{category} 
to the original entry type (\idx!{lowercase} and without 
the initial \code{@}) before it was aliased (behaves like 
\optfmt{same as entry} if the entry type wasn't aliased);

\item \optfmt{same as base}: (new to v1.1) set the \field{category}
to the base name of the \ext{bib} file (without the extension) 
that provided the entry definition;

\item \optfmt{same as type}: set the \field{category} to the same
value as the \field{type} field (if that field has been provided
either in the \ext{bib} file or through the \csopt{type} option);

\item \meta{label}: the \field{category} is set to
\meta{label} (which mustn't contain any special characters).
\end{itemize}
This will override any
\field{category} fields supplied in the \ext{bib} file.

When used with \csopt{entry-type-aliases}, the option \csopt[same as
entry]{category} refers to the \emph{target} entry type whereas
\csopt[same as original entry]{category} refers to the
\emph{original} entry type given in the \ext{bib} file.  In both
cases, the value is converted to \idx!{lowercase} to ensure consistency.

For example, if the \ext{bib} file contains:
\begin{codeenv}
\atentry{entry}\marg{bird,
  \field{name}=\marg{bird},
  \field{description} = \marg{feathered animal}
}
\strut
\atentry{index}\marg{duck}
\strut
\atentry{index}\marg{goose,\field{plural}=\qtdelim{geese}}
\strut
\atentry{dualentry}\marg{dog,
  \field{name}=\marg{dog},
  \field{description}=\marg{chien}
}
\end{codeenv}
then if the document contains:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[same as entry]{category},\csopt[entries]{src}}
\end{codeenv}
this will set the \field{category} of the \code{bird} term to \optfmt{entry}
(since it was defined with \atentry{entry}), the \field{category} of the
\code{duck} and \code{goose} terms to \optfmt{index} (since they were defined
with \atentry{index}), and the \field{category} of the \code{dog} term to
\optfmt{dualentry} (since it was defined with \atentry{dualentry}). Note that
the dual entry \code{dual.dog} doesn't have the category set, since that's
governed by \csopt{dual-category} instead.

If, instead, the document contains:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[animals]{category},\csopt[entries]{src}}
\end{codeenv}
then the \field{category} of all the primary selected entries will 
be set to \optfmt{animals}. Again the dual entry \code{dual.dog}
doesn't have the \field{category} set.

Note that the categories may be overridden by the commands that are used to
actually define the entries (such as \gls!{bibglsnewindex}).

For example, if the document contains:
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglsnewdualentry}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2,\field{category}=\marg{dual}}\marg{\idx{param}4}\comment{}
}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[animals]{category},\csopt[entries]{src}}
\end{codeenv}
then both the \code{dog} and \code{dual.dog} entries will
have their \field{category} field set to \optfmt{dual} since the
new definition of \gls{bibglsnewdualentry} has overridden
the \csopt[animals]{category} option.

\optsection[\subsubsection]{type}

The \meta{value} may be one of:
\begin{itemize}
 \item \optfmt{same as entry} set the \field{type} field
  to the entry type (\idx!{lowercase} and without the initial \code{@}); 

 \item \optfmt{same as original entry} set the \field{type} 
  to the original entry type (\idx!{lowercase} and without 
  the initial \code{@}) before it was aliased (behaves like 
  \optfmt{same as entry} if the entry type wasn't aliased);

 \optfmt{same as base} set the \field{type} field
  to the base name of the corresponding \ext{bib} file
  (without the extension);

 \item \optfmt{same as category} set the \field{type} field
   to the same value as the \field{category} field
   (\field{type} unchanged if \field{category} not set);

\item \optfmt{same as parent}: sets the \field{type} to the same as
the entry's parent (new to v1.9). If the entry doesn't have a parent
or if the parent doesn't have the \field{type} field set, then
no change is made. Entries should always have the same type
as their parent, but it's possible for spawned entries to 
pick up the \field{type} field from their \idx{progenitor} entry
(if it was explicitly set in the \ext{bib} file),
which may be inappropriate.

 \item\meta{label} sets the \field{type} field to the glossary 
   identified by \meta{label}.
\end{itemize}
When used with \csopt{entry-type-aliases}, the option \csopt[same as
entry]{type} refers to the \emph{target} entry type and \csopt[same
as original entry]{type} refers to the \emph{original} entry type
given in the \ext{bib} file.  It's not possible to have both
\csopt[same as type]{category} and \csopt[same as category]{type}.

Note that this setting only changes the \field{type} field for
primary entries. Use \csopt{dual-type} for dual entries.

For example:
\begin{codeenv}
\cmd{usepackage}[\styopt{record},\styopt{symbols}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-symbols]{src},\csopt[symbols]{type}}
\end{codeenv}

Make sure that the glossary type has already been defined
(see \sectionref{sec:newglossary}). In the above, the
\styopt{symbols} option defines the \code{symbols} glossary.
If you want to use a custom glossary, you need to provide it. For
example:
\begin{codeenv}
\cmd{usepackage}[\styopt{record},\styopt{nomain}]\marg{glossaries-extra}
\strut
\cs{newglossary*}\marg{dictionary}\marg{Dictionary}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries-symbols]{src},\csopt[dictionary]{type}}
\end{codeenv}
(The \styopt{nomain} option was added to suppress the
creation of the default \code{main} glossary.)

\optsection[\subsubsection]{trigger-type}

The record counting commands, such as \gls{rgls}, use the special
format \ics{glstriggerrecordformat}, which \bibgls\ also treats
as an \idx{ignoredrecord}. This means the entry will still be
identified as having a record for selection purposes, which is
necessary for the entry to be defined for use in the document, but
in order to prevent it from appearing in the glossary you need to
transfer the entry with \csopt[\meta{type}]{trigger-type}.
This will override the \csopt{type}, \csopt{dual-type},
\csopt{tertiary-type} and the type specification in
\csopt{secondary}.

The provided value \meta{type} must be a glossary label (not one of
the keywords allowed by \csopt{type}).
You can define the glossary before loading the resource, but
it's not required as \bibgls\ will write
\code{\ics{provideignoredglossary*}\margm{type}} to the \ext{glstex} file
even if \longarg{no-provide-glossaries} is set
(see \sectionref{sec:newglossary}).

\optsection[\subsubsection]{progenitor-type}

This sets the default \field{type} field for the main term defined by 
\atentry{progenitor}-like entries. The \meta{value} is as for
\csopt{type}. This doesn't change the \field{type} for the spawned
\idx{progeny}.

\optsection[\subsubsection]{progeny-type}

This sets the default \field{type} field for the \idx{progeny} term
spawned by 
\atentry{progenitor}-like entries. The \meta{value} is as for
\csopt{type}. This doesn't change the \field{type} for the main
\idx{progenitor}. Remember that with the default
\csopt[parent]{adopted-parent-field} setting, the given type 
should match the type of the parent entry.

\optsection[\subsubsection]{adopted-parent-field}

This identifies the target field to be set to the corresponding
value of the \field{adoptparents} list by the \idx{progeny}
entries spawned by the \atentry{progenitor} type of entry.
The default is \field{parent}.

\optsection[\subsubsection]{abbreviation-name-fallback}

The entry types that define abbreviations (such as
\atentry{abbreviation} and \atentry{acronym}) will, by default,
fallback on the \field{short} field if the \field{name} field is
missing and it's required for some reason (for example, with
\csopt[name]{sort-field}).  If you prefer to
fallback on a different field, then you can use this option to
specify the field. For example,
\csopt[long]{abbreviation-name-fallback}.
The \meta{field} value must be a known field (not an internal field)
but can't be the \field{sort} field.

\optsection[\subsubsection]{ignore-fields}

The \csopt{ignore-fields} key indicates that you want \bibgls\ to
skip the fields listed in the supplied comma-separated \meta{list} of field
labels. Remember that unrecognised fields will always be skipped.

For example, suppose my \ext{bib} file contains:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
  \field{short} =\qtdelim{html},
  \field{long}  = \marg{hypertext markup language},
  \field{description}=\marg{a markup language for creating web pages},
  \field{seealso}=\marg{xml}
}
\end{codeenv}
but I want to use the \abbrstyle{short-long} style and I don't want
the cross-referenced term, then I can use
\csopt[seealso,description]{ignore-fields}.

Note that \csopt[parent]{ignore-fields} removes the \field{parent}
before determining the dependency lists. This means that
\csopt[recorded and deps]{selection} and 
\csopt[recorded and ancestors]{selection} won't pick up the
label in the \field{parent} field.

If you want to maintain the dependency and ancestor relationship but
omit the \field{parent} field when writing the entries to the
\ext{glstex} file, you need to use \csopt{flatten} instead.

\optsection[\subsubsection]{field-aliases}

You can instruct \bibgls\ to treat one field as though it
was another using this option. The value should be a comma-separated
list of \meta{field1}\dequals\meta{field2} pairs, where
\meta{field1} and \meta{field2} are field names. Identical
mappings and trails aren't permitted. (That is, \meta{field1}
and \meta{field2} can't be the same nor can you have both
\meta{field1}\dequals\meta{field2} and
\meta{field2}\dequals\meta{field3}.) If you want to swap
fields you need to use one of the dual entry types instead.
Field aliases are performed before \csopt{ignore-fields},
so if \meta{field1} is listed in \csopt{ignore-fields} it won't
be ignored (unless \meta{field2} is in \csopt{ignore-fields}).

For example, suppose \filefmt{people.bib} contains:
\begin{codeenv}
\atentry{entry}\marg{alexander,
  \field{name}=\marg{Alexander III of Macedon},
  \field{description}=\marg{Ancient Greek king of Macedon},
  \fieldfmt{born}=\marg{20 July 356 BC},
  \fieldfmt{died}=\marg{10 June 323 BC},
  \fieldfmt{othername}=\marg{Alexander the Great}
}
\end{codeenv}
This contains three non-standard fields: \fieldfmt{born},
\fieldfmt{died} and \fieldfmt{othername}. I could define
these fields using \ics{glsaddkey}, but another possibility
is to map these onto the user keys \field{user1}, \field{user2}
and \field{user3}, which saves the overhead of providing new 
keys:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[people]{src},\comment{data in people.bib}
  \csopt[\fieldfmt{born}=\field{user1},\fieldfmt{died}=\field{user2},\fieldfmt{othername}=\field{user3}]{field-aliases}
}
\end{codeenv}

\optsection[\subsubsection]{replicate-fields}

The value of one field can be copied to other fields using
this option where each \meta{key}\dequals\meta{value} pair
is in the form
\code{\meta{field1}\dequals\marg{\meta{field2},\meta{field3},\ldots}}
where all values are field names. The value is required for this key
but may be empty, which indicates that the setting is switched off.

This option copies the contents
of \meta{field1} to \meta{field2}, \meta{field3}, \ldots\
(only if the target field isn't already set with
\csopt[false]{replicate-override}). This action is
performed after \csopt{ignore-fields} (see
\sectionref{sec:resourcesets}). If the source field is missing, the
\csopt{replicate-missing-field-action} setting determines the
action.

For example, suppose \filefmt{people.bib} contains:
\begin{codeenv}
\atentry{entry}\marg{alexander,
  \field{name}=\marg{Alexander III of Macedon (Alexander the Great)},
  \field{text}=\marg{Alexander},
  \field{description}=\marg{Ancient Greek king of Macedon}
}
\end{codeenv}
Since the \field{first} field hasn't been supplied, it
will default to the value of the \field{text} field, but
perhaps for one of my documents I'd like the \field{first}
field to be the same as the \field{name} field. Rather than
editing the \ext{bib} file, I can just do:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[people]{src},\comment{data in people.bib}
  \csopt[\field{name}=\field{first}]{replicate-fields}
}
\end{codeenv}
This copies the contents of the \field{name} field into the
\field{first} field. If you have more than one field
in the list take care to brace the lists to avoid confusion.
For example, if for some reason I want to copy the value
of the \field{name} field to both \field{first} and
\field{firstplural} and copy the value of the \field{text}
field to the \field{plural} field, then this requires braces
for the inner list:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[people]{src},\comment{data in people.bib}
  \csopt[\field{name}=\marg{\field{first},\field{firstplural}},\field{text}=\field{plural}]{replicate-fields}
}
\end{codeenv}

If my \filefmt{people.bib} file instead contained:
\begin{codeenv}
\atentry{entry}\marg{alexander,
  \field{name}=\marg{Alexander III of Macedon (Alexander the Great)},
  \field{first}=\marg{Alexander the Great},
  \field{text}=\marg{Alexander},
  \field{description}=\marg{Ancient Greek king of Macedon}
}
\end{codeenv}
then:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[people]{src},\comment{data in people.bib}
  \csopt[\field{name}=\field{first}]{replicate-fields}
}
\end{codeenv}
won't alter the \field{first} field since \csopt{replicate-fields}
doesn't override existing values by default. You can use 
\csopt{replicate-override} to change this. Alternatively, since 
\csopt{replicate-fields} is always performed after \csopt{ignore-fields} 
it's possible to ignore the \field{first} field which means that the 
\field{name} value can then be copied into it:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[people]{src},\comment{data in people.bib}
  \csopt[\field{first}]{ignore-fields},
  \csopt[\field{name}=\field{first}]{replicate-fields}
}
\end{codeenv}
Note that the ordering within the resource options doesn't
make a difference. The same result occurs with:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[people]{src},\comment{data in people.bib}
  \csopt[\field{name}=\field{first}]{replicate-fields},
  \csopt[\field{first}]{ignore-fields}
}
\end{codeenv}

\optsection[\subsubsection]{replicate-override}

This is a boolean option. The default setting is
\csopt[false]{replicate-override}. If \optfmt{true},
\csopt{replicate-fields} will override the existing value if the
target field is already set.

\optsection[\subsubsection]{replicate-missing-field-action}

This option indicates what to do if a source field identified in 
\csopt{replicate-fields} is missing. The value may be one of:
\begin{itemize}
\item \optfmt{skip}: skip the replication of the missing field
(default);
\item \optfmt{fallback}: use the fallback for the missing field, if
one is available (otherwise skip);
\item \optfmt{empty}: make the target field empty.
\end{itemize}

\optsection[\subsubsection]{counter}

The \csopt{counter} option assigns the default counter to use
for the selected entries. (This can be overridden with the
\glsopt{counter} key when using commands like \csfmt{gls}.)
The value must be the name of a counter. Since \bibgls\ doesn't know
which counters are defined within the document, there's no check to
determine if the value is valid (except for ensuring that
\meta{value} is non-empty).

Note that this will require an extra \LaTeX\ and \bibgls\ call since
the counter can't be used for the indexing until the entry has been
defined.

\optsection[\subsubsection]{copy-action-group-field}

This option may only be used when invoking \bibgls\ with the
\longarg{group} (or \shortargfmt{g}) switch. If an action
other than the default \csopt[define]{action} is set,
this option can be used to identify a field in which to save
the letter group information 
where \meta{value} is the name of the field. This just uses
\cs{GlsXtrSetField}. You will need to redefine
\ics{glsxtrgroupfield} to \meta{value} before displaying the glossary.
For example, if \csopt[dupgroup]{copy-action-group-field},
\csopt[copy]{action} and \csopt[copies]{type} are set in
the resource options and \code{copies} identifies a custom
glossary:
\begin{codeenv}
\ics{printunsrtglossary*}\oarg{\printglossopt[copies]{type},\printglossopt[\glostyle{indexgroup}]{style}}
 \marg{\cs{renewcommand}\marg{\cs{glsxtrgroupfield}}\marg{dupgroup}}
\end{codeenv}

This option is ignored when used with \csopt[define]{action}.
This option is not used by \csopt{secondary} which will
always save the group information in the \field{secondarygroup}
field. When used with \csopt[define or copy]{action}, entries
that are defined will have both \field{group} and
the field given by \csopt{copy-action-group-field} set.

Note that you may do \csopt[group]{copy-action-group-field} which
will override the \field{group} field from the original definition.
This may be useful if you don't use grouping in the primary
glossary. That is, you use \code{nogroupskip} and a non-group
style. For example:
\begin{codeenv}
\cs{printunsrtglossary}\oarg{\printglossopt{nogroupskip},\printglossopt[\glostyle{index}]{style}}
\cs{printunsrtglossary}\oarg{\printglossopt[copies]{type},\printglossopt[indexgroup]{style}}
\end{codeenv}

\optsection[\subsubsection]{copy-alias-to-see}

If set, the value of the \field{alias} field is copied to
the \field{see} field. The default setting is \csopt[false]{copy-alias-to-see}.

\optsection[\subsubsection]{save-original-entrytype}

If this is set to \optfmt{true}, the original entry type
(without the leading \idx{atchar}) is stored in the 
\field{originalentrytype} field. If no entry aliasing has been
applied (with \csopt{entry-type-aliases}) the field will be set to
the actual entry type.

\subsection{Field Adjustments}
\label{sec:fieldmods}

\optsection[\subsubsection]{post-description-dot}

The \styopt{postdot} package option (or \styopt[false]{nopostdot})
can be used to append a \idx{full-stop} (\idx{periodchar})\ to the
end of all the descriptions. This can be awkward if some of the
descriptions end with punctuation characters. This resource option
can be used instead. The \meta{value} may be one of:
\begin{itemize}
\item\optfmt{none}: don't append a \idx{full-stop} (default);
\item\optfmt{all}: append a \idx{full-stop} to all \field{description}
  fields in this resource set;
\item\optfmt{check}: selectively append a \idx{full-stop} (see below).
\end{itemize}
Note that if you have dual entries and you use this option to
append a \idx{full-stop}, then it will be copied over to the mapped field.
This is different to the \styopt{postdot} option which doesn't
add the dot to the field but incorporates it in the 
\idx{postdescriptionhook}. This means that a dot inserted with
\csopt{post-description-dot} will come before the
\idx{postdescriptionhook} whereas with \styopt{postdot} the
punctuation comes after any category-specific hook.

The \csopt[check]{post-description-dot} setting determines whether
to append the dot as follows:
\begin{itemize}
\item If the \field{description} field ends with
\ics{nopostdesc} or \ics{glsxtrnopostpunc}, then a dot isn't appended.
\item If the \field{description} field doesn't end with a regular
(ungrouped letter or other) character, then a dot is appended.
(For example, if the description ends with a control sequence
or an end group token.)
\item If the \field{description} field ends with a character 
that belongs to the Unicode category \idx{punctuationclose}
or \idx{punctuationfinalquote} then the token preceding
that character is checked.
\item If the \field{description} field doesn't end with
a character that belongs to the Unicode category
\idx{punctuationother} then the dot is added.
\end{itemize}
Note that the interpreter isn't used during the check.
If the \field{description} ends with a command then a dot will be
appended (unless it's \cs{glsxtrnopostpunc} or \cs{nopostdesc}) even if
that command expands in such a way that it ends with a terminating
punctuation character. This option only applies to the
\field{description} field.

\optsection[\subsubsection]{strip-trailing-nopost}

This option is always performed before \csopt{post-description-dot}. 
The default setting is \csopt[false]{strip-trailing-nopost}. If \optfmt{true}
any trailing ungrouped \ics{nopostdesc} or \ics{glsxtrnopostpunc} found in the
\field{description} field will be removed.  Note that the command (possibly
followed by ignored space) must be at the very end of the description for it to
be removed. A description should not contain both commands. This option only
applies to the \field{description} field.

For example, \cs{nopostdesc} will be stripped from:
\begin{codeenv}
\field{description}=\marg{sample\cs{nopostdesc}}
\end{codeenv}
since it's at the end. It will also be stripped from:
\begin{codeenv}
\field{description}=\marg{sample\cs{nopostdesc} }
\end{codeenv}
since the trailing space is ignored as it follows a control
word. It won't be stripped from:
\begin{codeenv}
\field{description}=\marg{sample\cs{nopostdesc}\marg{} }
\end{codeenv}
because the final space is now significant, but even without the
space it still won't be stripped as the field ends with
an empty group not with \cs{nopostdesc}. Similarly it won't be
stripped from:
\begin{codeenv}
\field{description}=\marg{sample\cs{nopostdesc}\cmd{relax}}
\end{codeenv}
because again it's not at the end.

\optsection[\subsubsection]{check-end-punctuation}

This options checks the end of all the fields given in \meta{list} for
end of sentence punctuation. This is determined as follows, for each
\meta{field} in the comma-separated \meta{list}:
\begin{itemize}
\item if the last character is of type \idx{punctuationclose}
or \idx{punctuationfinalquote}, check the character that comes
before it;
\item if the character is of type \idx{punctuationother}, then check
if it's listed in the entry given by \idx{sentence.terminators} 
in \bibgls's \langxml.
\end{itemize}
If a sentence terminator is found, an internal field is created
called \field{fieldendpunc} that contains the punctuation
character. Fields whose values must be labels (such as
\field{parent}, \field{category} and \field{type}) aren't checked,
even if they're included in \meta{list}.

The default \idx{sentence.terminators} is defined in \file{bib2gls-en.xml}
as:
\begin{verbatim}
<entry key="sentence.terminators">.?!</entry>
\end{verbatim}
Any character that isn't of type \idx{punctuationother} won't
match.

For example, the sample \exfile!{books.bib} file contains:
\begin{codeenv}
\atentry{entry}\marg{whydidnttheyaskevans,
  \field{name}=\marg{Why Didn't They Ask Evans?},
  \field{description}=\marg{novel by Agatha Christie},
  \fieldfmt{identifier}=\marg{book},
  \fieldfmt{author}=\marg{\gls{sortmediacreator}\marg{Agatha}\marg{Christie}},
  \fieldfmt{year}=\marg{1934}
}
\end{codeenv}
With \csopt[name]{check-end-punctuation}, this entry will be
assigned an internal field called 
\fielddisp{fieldendpunc}{nameendpunc} set to \code{?}\
as that's included in \idx{sentence.terminators} and is found
at the end of the \field{name} field:
\begin{codeenv}
\cs{GlsXtrSetField}\marg{whydidnttheyaskevans}\marg{nameendpunc}\marg{?}
\end{codeenv}
(Note that \csopt[first,text]{check-end-punctuation} won't match
as there's no \field{first} or \field{text} field supplied.)

If you have a field that ends with an abbreviation followed by a
\idx{full-stop}, this will be considered an end of sentence terminator, but
the main purpose of this option is to provide a way to deal with
cases like:
\begin{codeenv}
Agatha Christie wrote \cs{gls}\marg{whydidnttheyaskevans}.
\end{codeenv}
where the end of sentence punctuation following \cs{gls} needs to be
discarded. This is needed regardless of whether or not
the link text ends with an abbreviation or is a complete sentence.

It's then possible to hook into the \idx{postlinkhook} \qt{discard
period} check. By default this just checks the category attributes
that govern whether or not to discard a following period, but
(with \styfmt{glossaries-extra} v1.23+) it's possible to provide
an additional check by redefining:
\nosecdef{glsxtrifcustomdiscardperiod}
This should expand to \meta{true} if the check should be performed
otherwise it should expand to \meta{false}. You can reference the
label using \cs{glslabel}. For example:
\begin{codeenv}
\cs{renewcommand}*\marg{\gls{glsxtrifcustomdiscardperiod}}[2]\marg{\comment{}
 \cs{GlsXtrIfFieldUndef}\marg{nameendpunc}\marg{\cs{glslabel}}\marg{\idx{param}2}\marg{\idx{param}1}\comment{}
}
\end{codeenv}
This uses \ics{GlsXtrIfFieldUndef} rather than
\ics{glsxtrifhasfield*} since there's no need to access the field's
value. (The unstarred form \ics{glsxtrifhasfield} can't be used
as it introduces implicit scoping, which would interfere with the
punctuation lookahead.) The other difference between
\ics{GlsXtrIfFieldUndef} and the other \csfmt{\ldots hasfield} tests
is the case where the field is set to an empty value. In this case
the field is defined (so \ics{GlsXtrIfFieldUndef} does the
\meta{false} argument) but it's considered unset (so commands like
\ics{ifglshasfield} do the \meta{false} argument).

\optsection[\subsubsection]{sort-label-list}

This option takes a list as the value with each element in the list
in the form:
\begin{codeenv}
\code{\meta{field-list}:\meta{sort}:\meta{csname}}
\end{codeenv}
or:
\begin{codeenv}
\code{\meta{field-list}:\meta{sort}}
\end{codeenv}
where:
\begin{itemize}
\item \meta{field-list} is a comma-separated list of valid fields;
\item \meta{sort} is a valid sort method as per the \csopt{sort}
option, but not including \optfmt{none} or \optfmt{unsrt};
\item \meta{csname} is the name (without a leading backslash) of a
command that takes a label as its sole mandatory argument that's
recognised by \bibgls' interpreter (such as those listed in
\tableref{tab:bibglsdefs}).
\end{itemize}
The final \code{:\meta{csname}} part may be omitted if no command
need be applied. (That is, sort by label.) The value is required for this key but
may be empty, which indicates the setting is switched off.

The sorting options are as those for the main list.  For example,
for entries in the primary list the break point is obtained from the
\csopt{break-at} setting and for entries in the dual list the break
point is obtained from \csopt{dual-break-at}.  (Remember that if
\csopt[combine]{dual-sort} then there is only one list that contains both
the primary and dual entries, which is governed by the primary
options only.)

If the \meta{field-list} has more than one element
take care to use braces \code{\marg{}} to avoid confusion for the
list-parser. For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[\marg{see,seealso}:en:glsentryname]{sort-label-list}
}
\end{codeenv}

Note that strange results may occur if this setting is used on any
fields that don't simply contain a list of entry labels or if any of
the referenced entries are processed in different
\idxpl{resourceset} (see \sectionref{sec:resourcesets}).

After the main sorting of each set of selected entries is performed
(as per \csopt{sort} or \csopt{dual-sort}), if this
option is set, then for each 
\code{\margm{field-list}:\meta{sort}:\meta{csname}} the following
steps are performed:
\begin{enumerate}
\item For each entry \meta{id}:
 \begin{enumerate}
  \item For each \meta{field} in \meta{field-list}, if the field is
    set for entry \meta{id} then:

  \begin{enumerate}
   \item The field value must be in the form
     \code{\oargm{tag}\meta{label-list}} where
     \code{\oargm{tag}} is optional and
     \meta{label-list} is a comma-separated list of entry labels
     \meta{label$_1$}, \ldots, \meta{label$_n$};
   \item A new list is constructed where the $i$th element is:
    \code{\marg{\csfmt{}\meta{csname}\margm{label$_i$}}}
    unless \meta{csname} hasn't been set, in which case the
    $i$th element is just \margm{label$_i$} (the optional
    \oargm{tag} part is omitted);
   \item This new list is sorted according to the interpreter's
    definition of the command given by \meta{csname} (if provided) 
    and the designated \meta{sort} method;
   \item The field value is reconstructed with the labels in the
    corresponding order (prefixed with \code{\oargm{tag}} if it was
    present in the original).
  \end{enumerate}
 \end{enumerate}
\end{enumerate}
Note that there is no hierarchical structure in the sorting of the field list
even if any of the referenced entries has a parent.

For example, suppose the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{index}\marg{bird}
\strut
\atentry{index}\marg{waterfowl, \field{parent}=\marg{bird} }
\strut
\atentry{index}\marg{duck,
  \field{parent}=\marg{waterfowl},
  \field{seealso}=\marg{swan,duckling,parrot,goose}
}
\strut
\atentry{index}\marg{swan,
  \field{parent}=\marg{waterfowl},
  \field{seealso}=\marg{goose,duck}
}
\strut
\atentry{index}\marg{goose,
 \field{parent}=\marg{waterfowl},
 \field{seealso}=\marg{duck}
}
\strut
\atentry{index}\marg{parrot, \field{parent}=\marg{bird} }
\strut
\atentry{index}\marg{duckling,
 \field{see}=\marg{[related terms]fluffy,velociraptor,duck,tardigrade}
}
\strut
\atentry{index}\marg{fluffy}
\strut
\atentry{index}\marg{tardigrade, \field{name}=\marg{water bear} }
\strut
\atentry{index}\marg{velociraptor}
\end{codeenv}
And suppose the document contains:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[\styopt{record},\styopt[tree]{style}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{ 
 \csopt[entries]{src},
 \csopt[en]{sort},
 \csopt[\marg{seealso,see}:en:glsentryname]{sort-label-list}
}
\strut
\cmd{begin}\marg{document}
\cs{Gls}\marg{parrot}, \cs{gls}\marg{tardigrade}, \cs{gls}\marg{swan}, \cs{gls}\marg{duck}, 
\cs{gls}\marg{goose}, \cs{gls}\marg{fluffy} \cs{gls}\marg{duckling}, \cs{gls}\marg{velociraptor}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
Then this reorders the \field{see} and \field{seealso} fields according to
the referenced entry's name (obtained with \ics{glsentryname}).

For example, the \field{see} field for the \code{duckling} entry was
originally:
\begin{codeenv}
\field{see}=\marg{[related terms]fluffy,velociraptor,duck,tardigrade}
\end{codeenv}
but in the \ext{glstex} file it's written as:
\begin{codeenv}
\field{see}=\marg{[related terms]duck,fluffy,velociraptor,tardigrade}
\end{codeenv}
The reason for \code{tardigrade} being placed after
\code{velociraptor} is because
\code{\ics{glsentryname}\marg{tardigrade}} is expanded to \qt{water
bear} (and \qt{W} comes after \qt{V}). If no encapsulating command 
was specified:
\begin{codeenv}
\csopt[\marg{seealso,see}:en]{sort-label-list}
\end{codeenv}
then the list would have been sorted according to the labels
instead (and so \code{tardigrade} would come before
\code{velociraptor}). Note that the optional tag is kept at the start of the 
list.

The \field{seealso} fields have also been changed. For example, 
the \code{duck} entry originally had:
\begin{codeenv}
\field{seealso}=\marg{swan,duckling,parrot,goose}
\end{codeenv}
but in the \ext{glstex} file it's written as:
\begin{codeenv}
\field{seealso}=\marg{duckling,goose,parrot,swan}
\end{codeenv}
Note that the hierarchical structure hasn't been maintained. The glossary
lists \qt{duckling} (a top-level entry) after \qt{swan} (a level~2 entry)
but the \field{seealso} field has \code{duckling} first.

If you want to maintain the hierarchy you can use \ics{glsxtrhiername} instead
of \ics{glsentryname}:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{ 
 \csopt[entries]{src},
 \csopt[en]{sort},
 \csopt[\marg{seealso,see}:en:glsxtrhiername]{sort-label-list}
}
\end{codeenv}
The separator between the levels is
given by \ics{glsxtrhiernamesep} which is defined by
\sty{glossaries-extra} to produce \qt{\glsxtrhiernamesep}.
The \bibgls\ interpreter's definition of this command is different
to assist sorting and simply expands to a \idx{full-stop} to prevent
it from being replaced by the default word break marker.

In this case \code{\ics{glsxtrhiername}\marg{swan}} would be
displayed as \qt{bird\glsxtrhiernamesep waterfowl\glsxtrhiernamesep
swan} if used in the document, but the interpreter converts it to
\qt{bird.waterfowl.swan}, so with the default \csopt{break-at}
setting the actual sort value becomes \code{bird.waterfowl.swan|}
(instead of \code{bird|waterfowl|swan|} which would be the result if
the interpreter used the same definition as \sty{glossaries-extra}).

Therefore the \field{seealso} field for the \code{duck} entry ends up as:
\begin{codeenv}
\field{seealso}=\marg{parrot,goose,swan,duckling}
\end{codeenv}
Now \code{swan} comes before \code{duckling} because the actual sort
value started with a \qt{B} not \qt{S}.

This hierarchical information isn't shown in the cross-reference by
default, so the \code{duck} cross-reference list appears in the
document as: parrot, goose, swan \& duckling.

If you want the hierarchical information to appear to help assist
the reader, you can redefine \ics{glsseeitemformat} in the document
to use \ics{glsxtrhiername}:
\begin{codeenv}
\cmd{renewcommand}*\marg{\cs{glsseeitemformat}}[1]\marg{\cs{glsxtrhiername}\marg{\idx{param}1}}
\end{codeenv}
This means that the \code{duck} cross-reference now appears in the
document as: bird\glsxtrhiernamesep parrot, bird\glsxtrhiernamesep
waterfowl\glsxtrhiernamesep goose, bird\glsxtrhiernamesep waterfowl
\glsxtrhiernamesep swan \& duckling.

This next example document has two languages, English and
Portuguese. The file \filefmt{entries-en.bib} contains the English
terms, such as:
\begin{codeenv}
\atentry{index}\marg{cat, \fieldfmt{translations}=\marg{gato,gatinho} }
\atentry{index}\marg{kitten, \fieldfmt{translations}=\marg{gatinho} }
\atentry{index}\marg{staple, \fieldfmt{translations}=\marg{grampo}}
\atentry{index}\marg{rivet, \fieldfmt{translations}=\marg{rebite}}
\end{codeenv}
The file \filefmt{entries-pt.bib} contains the Portuguese
terms, such as:
\begin{codeenv}
\atentry{index}\marg{gato, \fieldfmt{translations}=\marg{cat,staple,rivet} }
\atentry{index}\marg{gatinho, \fieldfmt{translations}=\marg{kitten} }
\end{codeenv}

Both files have a custom field called \fieldfmt{translations} that
will need to be either defined or aliased.  This field contains a
comma-separated list of labels for the corresponding entries in the
other language file that provide a possible translation.  Where a
word has multiple possible translations, I'd like the list sorted
alphabetically. (In practice, it would make more sense to sort them
according to how likely the translation is, but this is for
illustrative purposes.) For convenience, the
custom field is simply aliased to the \field{user1} field.

The document has two glossaries for each set of terms. The English
terms are sorted according to \csopt[en-GB]{sort} in one
\idx{resourceset} and the Portuguese terms are sorted according to
\csopt[pt-BR]{sort} in another \idx{resourceset}. This means that there
are cross-resource references, but since there are no instances of
\atentry{preamble} it should be possible to resolve the references.

The document code is:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[T1]\marg{fontenc}
\cmd{usepackage}[utf8]\marg{inputenc}
\cmd{usepackage}[british,brazilian]\marg{babel}
\cmd{usepackage}\oarg{\styopt{record},
 \styopt{nomain},
 \styopt{nostyles},
 \styopt[bookindex]{stylemods},
 \styopt[bookindex]{style}
}\marg{glossaries-extra}
\cmd{usepackage}\marg{glossaries-prefix}
\strut
\cs{newglossary*}\marg{en}\marg{English Terms}
\cs{newglossary*}\marg{pt}\marg{Portuguese Terms}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[all]{selection},
  \csopt[en]{type},
  \csopt[entries-en]{src},
  \csopt[en-GB]{sort},
  \csopt[\fieldfmt{translations}=\field{user1}]{field-aliases},
  \csopt[\field{user1}:pt-BR:\encap{glsentryname}]{sort-label-list}
}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[all]{selection},
  \csopt[pt]{type},
  \csopt[entries-pt]{src},
  \csopt[pt-BR]{sort},
  \csopt[\fieldfmt{translations}=\field{user1}]{field-aliases},
  \csopt[\field{user1}:en-GB:\encap{glsentryname}]{sort-label-list}
}
\strut
\cs{apptoglossarypreamble}\oarg{en}\marg{\ics{selectlanguage}\marg{british}}
\cs{apptoglossarypreamble}\oarg{pt}\marg{\cs{selectlanguage}\marg{brazilian}}
\strut
\cmd{begin}\marg{document}
\cs{renewcommand}*\marg{\cs{glsxtrbookindexname}}[1]\marg{\comment{}
 \cs{glossentryname}\marg{\idx{param}1}\comment{}
 \cs{glsxtrifhasfield}\marg{useri}\marg{\idx{param}1}\marg{: \ics{glsxtrseelist}\cs{glscurrentfieldvalue}}\marg{}\comment{}
}
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
In verbose mode, the transcript file indicates that it's performing
the label list sorting. For example, when sorting according to
\csopt[user1:pt-BR:glsentryname]{sort-label-list}, the transcript
file contains:
\begin{verbatim}
Label list sort method 'pt-BR' on field: user1
\end{verbatim}
The \code{cat} entry has a list of two elements in this field:
\code{gato,gatinho}. This is converted into a new list where the
first element is:
\begin{codeenv}
\marg{\cs{glsentryname}\marg{gato}}
\end{codeenv}
and the second element is:
\begin{codeenv}
\marg{\cs{glsentryname}\marg{gatinho}}
\end{codeenv}
Regardless of the level of verbosity, the transcript file will
contain the conversions obtained by the interpreter:
\begin{verbatim}
texparserlib: {\glsentryname{gato}} -> gato
texparserlib: {\glsentryname{gatinho}} -> gatinho
\end{verbatim}
The \code{kitten} entry has the same list, and the same process is
repeated for that entry. The \longarg{verbose} mode will provide additional
information. The \longarg{debug} mode will indicate whether the
referenced label was found in the current \idx{resourceset} or if it had
to be fetched from another \idx{resourceset}. So if the resulting
order isn't what you expect, check the transcript file for messages.

\optsection[\subsubsection]{bibtex-contributor-fields}

This option indicates that the listed fields all use \BibTeX's name
syntax (as used in \BibTeX's \code{author} and \code{editor} fields). The value is required for this key but
may be empty, which indicates an empty set of fields (that is, the
setting is switched off).

The values of these fields will be converted into the form:
\begin{codeenv}
\gls{bibglscontributorlist}\margm{contributor list}\margm{n}
\end{codeenv}
where \meta{n} is the number of names in the list and
\meta{contributor-list} is a comma-separated list of names in the
form:
\begin{codeenv}
\gls{bibglscontributor}\margm{forenames}\margm{von-part}\margm{surname}\margm{suffix}
\end{codeenv}

The \gls{bibglscontributorlist} command is initially defined in 
\bibgls's interpreter to just do the first argument and ignore the
second. This means that if you're sorting on this field, the \qt{and}
part between the final names doesn't appear in the sort value.
The actual definition of \gls{bibglscontributorlist} provided in the 
\ext{glstex} file depends on whether or not \ics{DTLformatlist} is defined.
(Note that \styfmt{glossaries} automatically loads \sty{datatool-base}
so this command will be defined if you have at least v2.28 of
\sty{datatool-base}.)

For example, if the \field{name} field is specified as:
\begin{codeenv}
\field{name}=\marg{John Smith and Jane Doe and Dickie von Duck}
\end{codeenv}
then \csopt[name]{bibtex-contributor-fields} will convert the
\field{name} field value to:
\begin{codeenv}
\gls{bibglscontributorlist}\marg{\comment{}
 \gls{bibglscontributor}\marg{John}\marg{}\marg{Smith}\marg{},\comment{}
 \gls{bibglscontributor}\marg{Jane}\marg{}\marg{Doe}\marg{},\comment{}
 \gls{bibglscontributor}\marg{Dickie}\marg{von}\marg{Duck}\marg{}}\marg{3}
\end{codeenv}
With \csopt[von]{contributor-order} the sort value obtained from
this field will be:
\begin{verbatim}
Smith, John,Doe, Jane,von Duck, Dickie
\end{verbatim}
With one of the locale sort methods and with the default
\csopt[word]{break-at}, this will end up as:
\begin{verbatim}
Smith|John|Doe|Jane|von|Duck|Dickie
\end{verbatim}

\optsection[\subsubsection]{contributor-order}

The \gls{bibglscontributor} command is defined in
\bibgls's interpreter and its definition is dependent on this
setting. The \meta{value} may be one of (where the parts in square 
brackets are omitted if that argument is empty):
\begin{itemize}
\item \optfmt{surname}: \gls{bibglscontributor} expands to
\meta{surname}\oarg{, \meta{suffix}}\oarg{, \meta{forenames}}\oarg{,
\meta{von-part}};
\item \optfmt{von}: \gls{bibglscontributor} expands to
\oarg{\meta{von-part} }\meta{surname}\oarg{, \meta{suffix}}\oarg{, \meta{forenames}};
\item \optfmt{forenames}: \gls{bibglscontributor} expands to
\oarg{\meta{forenames} }\oarg{\meta{von-part} }\meta{surname}\oarg{,
\meta{suffix}}.
\end{itemize}
The default value is \optfmt{von}. Note that if you have multiple
resource sets, this option governs the way \bibgls's version of
\gls{bibglscontributor} behaves. The actual definition is written to
the \ext{glstex} using \cs{providecommand}, which means that \LaTeX\
will only pick up the first definition.

For example:
\begin{codeenv}
\cs{newcommand}*\marg{\gls{bibglscontributor}}[4]\marg{\comment{}
  \idx{param}1\ics{ifstrempty}\marg{\idx{param}2}\marg{}\marg{ \idx{param}2} \idx{param}3\cs{ifstrempty}\marg{\idx{param}4}\marg{}\marg{, \idx{param}4}\comment{}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[name]{bibtex-contributor-fields}
}
\end{codeenv}
This will display the names in the glossary with the forenames
first, but \bibgls\ will sort according to surname.

An alternative approach, if you need an initial resource set
such as with the \exfile{no-interpret-preamble.bib} file:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[no-interpret-preamble]{src},
  \csopt[false]{interpret-preamble},
  \csopt[name]{bibtex-contributor-fields},
  \csopt[forenames]{contributor-order}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[name]{bibtex-contributor-fields}
}
\end{codeenv}
Note the need to use \csopt[name]{bibtex-contributor-fields}
in the first resource set even though there are no entries in 
the \ext{bib} file. This is because the definition of
\gls{bibglscontributor} is only written to the \ext{glstex} file if
\csopt{bibtex-contributor-fields} has been set to a non-empty list.
The second resource set will use the default 
\csopt[von]{bibtex-contributor-fields} setting when obtaining the
sort value.

\optsection[\subsubsection]{encapsulate-fields}

This option should take a comma-separated list of 
\optfmt{\meta{field}\dequals\meta{cs-name-1arg}} values, where
\meta{cs-name-1arg} is the name of a control sequence that takes
one argument. The value is required for this key but
may be empty, which indicates an empty set (that is, the
setting is switched off). 

During the processing stage, each field identified in
the list (if defined) will have its value replaced with:
\begin{codeenv}
\cmd{\meta{cs-name-1arg}}\margm{value}
\end{codeenv}
where \meta{value} was its previous value. An
empty list switches off encapsulation (the default).

This action overrides any previous use of \csopt{encapsulate-fields}
within the same \idx{resourceset} and is always performed before
\csopt{encapsulate-fields*}, regardless of the order in the 
\idx{resourceset}['s] list of options.

\optsection[\subsubsection]{encapsulate-fields*}

This option should take a comma-separated list of 
\optfmt{\meta{field}\dequals\meta{cs-name-2arg}} values, where 
\meta{cs-name-2arg} is the name of a control sequence that takes two
arguments. The value is required for this key but
may be empty, which indicates an empty set (that is, the
setting is switched off). 

During the processing stage, each field identified in the
list (if defined) will have its value replaced with:
\begin{codeenv}
\cmd{\meta{cs-name-2arg}}\margm{value}\margm{label}
\end{codeenv}
where \meta{value} was its previous value and \meta{label} is the
entry's label (including prefix, if appropriate). An
empty list switches off encapsulation (the default). 

This action overrides any previous use of
\csopt{encapsulate-fields*} within the same \idx{resourceset}, and
is always performed after \csopt{encapsulate-fields}, regardless of
the order in the \idx{resourceset}['s] list of options, so if the
same field is listed in both settings, its value will end up as:
\begin{codeenv}
\cmd{\meta{cs-name-2arg}}\marg{\cmd{\meta{cs-name-1arg}}\margm{value}}\margm{label}
\end{codeenv}

\optsection[\subsubsection]{interpret-fields}

This option indicates that the listed fields should be replaced by
their interpreted values. The value is required for this key but
may be empty, which indicates an empty set of fields (that is, the
setting is switched off). Other fields not listed may still be
interpreted depending on other settings. As with the \field{sort}
field, any special characters are replaced with commands like
\ics{glsbackslash} and \ics{bibglsdollarchar}. This option is
applied after \csopt{field-case-change} (if set).

For example, suppose I have a file \filefmt{entries.bib} that
contains definitions like:
\begin{codeenv}
\atentry{symbol}\marg{pi,
  \field{name}=\marg{\cs{ensuremath}\marg{\ics{pi}}},
  \field{description}=\marg{the ratio of a circle's circumference to its diameter},
}
\strut
\atentry{symbol}\marg{sigma,
  \field{name} = \marg{\cs{ensuremath}\marg{\ics{sigma}}},
  \field{description} = \marg{standard deviation}
}
\end{codeenv}
Instead of having a list of terms (glossary), suppose I want to have
stand-alone definitions, where the term appears in a section heading.
I could define a command like this:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{definition}}[1]\marg{\comment{}
  \cs{ifglsentryexists}\marg{\idx{param}1}\comment{}
  \marg{\comment{}
    \cs{section}\oarg{\cs{glsentryname}\marg{\idx{param}}}\marg{\cs{glsadd}\marg{\idx{param}1}\ics{glsxtrglossentry}\marg{\idx{param}1}}\comment{}
    \ics{Glossentrydesc}\marg{\idx{param}1}\cs{glspostdescription}
  }\comment{}
  \marg{\cs{section}\oarg{Missing `\idx{param}1'}\marg{\cs{glsadd}\marg{\idx{param}1}}}\comment{}
}
\end{codeenv}
which can be used in the document:
\begin{codeenv}
\ics{tableofcontents}
\cmd{definition}\marg{pi}
\cmd{definition}\marg{sigma}
\end{codeenv}
A problem with this definition of my custom command occurs if I add
\sty{hyperref} to the document, because this tries to write \cs{pi}
and \cs{sigma} to the PDF bookmarks, which doesn't work because
those commands can't be automatically converted to characters
permitted in a PDF string. This leads to a warning from
\sty{hyperref}:
\begin{alltt}
Token not allowed in a PDF string (Unicode)
\end{alltt}
Ideally I'd like to be able to convert these symbols to Unicode so
that they can appear in the bookmarks. Since \bibgls' interpreter
recognises these commands, I can get it to make the conversion
instead of trying to implement a method within \TeX:
\begin{codeenv}
\cs{glsaddstoragekey}\marg{pdfname}\marg{}\marg{\cmd{pdfname}}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},
  \csopt[\field{name}=\fieldfmt{pdfname}]{replicate-fields},
  \csopt[fallback]{replicate-missing-field-action},
  \csopt[\fieldfmt{pdfname}]{interpret-fields}
}
\end{codeenv}
This first copies the \field{name} field to the custom
\fieldfmt{pdfname} and then interprets the copy. This leaves
the \field{name} field with the \LaTeX\ code to produce the symbol
in the document, but the \fieldfmt{pdfname} field ends up with all markup
stripped by the interpreter and the \cs{pi} and \cs{sigma} are
converted to the Unicode characters \hex{1D70B} (mathematical italic
small pi) and \hex{1D70E} (mathematical italic small sigma). With
\XeLaTeX\ or \LuaLaTeX\ these characters can be written to the PDF
bookmarks by adjusting the definition of the custom command:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{definition}}[1]\marg{\comment{}
  \cs{ifglsentryexists}\marg{\idx{param}1}\comment{}
  \marg{\comment{}
    \cs{section}
    \oarg{\ics{texorpdfstring}\marg{\cs{glsentryname}\marg{\idx{param}1}}\marg{\cmd{pdfname}\marg{\idx{param}1}}}
    \marg{\cs{glsadd}\marg{\idx{param}1}\cs{glsxtrglossentry}\marg{\idx{param}1}}\comment{}
    \cs{Glossentrydesc}\marg{\idx{param}1}\cs{glspostdescription}
  }\comment{}
  \marg{\cs{section}\oarg{Missing `\idx{param}1'}\marg{\cs{glsadd}\marg{\idx{param}1}}}%
}
\end{codeenv}

\optsection[\subsubsection]{interpret-fields-action}

This option governs the behaviour of \csopt{interpret-fields}.
Available values are: 
\begin{itemize}
\item\code{replace}: replace the field content with its interpreted
value (default);
\item\code{replace non empty}: only replace the field content with
its interpreted value if the interpreted value isn't an empty
string.
\end{itemize}
If a field value consists solely of commands that are unknown to the
interpreter, then the resulting value will end up empty. In this
case, it may be more appropriate to leave the field unchanged.

\optsection[\subsubsection]{date-time-fields}

This option indicates that the listed fields all contain 
date and time information. Primary entries will have these fields
parsed according to \csopt{date-time-field-format} and
\csopt{date-time-field-locale} and dual entries will have these
fields parsed according to \csopt{dual-date-time-field-format} and
\csopt{dual-date-time-field-locale}. If the field value is missing
or doesn't match the given pattern it remains unchanged, otherwise
it's converted into the form:
\formatdef{bibglsdatetime}
where \meta{original} is the value of the field before conversion.
If the interpreter is on, the value will be interpreted before
being parsed if it contains \idx{escchar}, \idx{mshiftchar},
\idx{bgroupchar}, \idx{egroupchar} or \idx{nbspchar}. (Remember that
\idx{nbspchar} is converted to the non-breaking space character
\hex{A0} unless \longarg{break-space} is used.)

\optsection[\subsubsection]{date-fields}

As \csopt{date-time-fields} but for fields that only contain date
(not time) information. If parsed correctly, the field is converted
to:
\formatdef{bibglsdate}
The fields are parsed according to
\csopt{date-field-format} and
\csopt{date-field-locale} for primary entries and according to
\csopt{dual-date-field-format} and
\csopt{dual-date-field-locale} for dual entries.

\optsection[\subsubsection]{time-fields}

As \csopt{date-time-fields} but for fields that only contain time
(not date) information. If parsed correctly, the field is converted
to:
\formatdef{bibglstime}
The fields are parsed according to
\csopt{time-field-format} and
\csopt{time-field-locale} for primary entries and according to
\csopt{dual-time-field-format} and
\csopt{dual-time-field-locale} for dual entries.

\optsection[\subsubsection]{date-time-field-format}

This option also sets
\csopt[\meta{value}]{dual-date-time-field-format}.
The value is the format pattern used when parsing fields identified
by \csopt{date-time-fields}. The \meta{value} is as for
\csopt{date-sort-format}.

\optsection[\subsubsection]{date-field-format}

This option also sets
\csopt[\meta{value}]{dual-date-field-format}.
The value is the format pattern used when parsing fields identified
by \csopt{date-fields}. The \meta{value} is as for
\csopt{date-sort-format}.

\optsection[\subsubsection]{time-field-format}

This option also sets
\csopt[\meta{value}]{dual-time-field-format}.
The value is the format pattern used when parsing fields identified
by \csopt{time-fields}. The \meta{value} is as for
\csopt{date-sort-format}.

\optsection[\subsubsection]{date-time-field-locale}

This option also sets
\csopt[\meta{value}]{dual-date-time-field-locale}.
The value is the locale used when parsing fields identified
by \csopt{date-time-fields}. The \meta{value} is as for
\csopt{date-sort-locale}.

\optsection[\subsubsection]{date-field-locale}

This option also sets
\csopt[\meta{value}]{dual-date-field-locale}.
The value is the locale used when parsing fields identified
by \csopt{date-fields}. The \meta{value} is as for
\csopt{date-sort-locale}.

\optsection[\subsubsection]{time-field-locale}

This option also sets
\csopt[\meta{value}]{dual-time-field-locale}.
The value is the locale used when parsing fields identified
by \csopt{time-fields}. The \meta{value} is as for
\csopt{date-sort-locale}.

\subsection{Prefix Fields}
\label{sec:prefixfields}

If you use the \isty{glossaries-prefix} package, the prefix
set of fields become available (\field{prefix},
\field{prefixplural}, \field{prefixfirst} and
\field{prefixfirstplural}). The default behaviour of \ics{pgls} is for no
separator between the prefix and the text produced with \cs{gls}.
This is because there are situations where there shouldn't be a
space, although a space is more commonly required.

This means that a space needs to be appended to the required prefix
fields, but an actual space character can't be used because
\sty{xkeyval} trims leading and trailing spaces. The \ics{space}
command needs to be used instead, but there are also situations
where an non-breakable space should be used (for example, where the
prefix is a single character). It's a bit tiresome having to
remember to put \ics{space} or \idx{nbspchar} at the end of the
field value. 

The \csopt{append-prefix-field} option allows the automatic insertion of a
space, but it may be used without the \sty{glossaries-prefix}
package. The fields that contain prefixes are identified by
\csopt{prefix-fields}.

If you have any dual entries, then \bibgls\ will also recognise the
special internal fields \field{dualprefix},
\field{dualprefixplural}, \field{dualprefixfirst} and
\field{dualprefixfirstplural}.

\optsection[\subsubsection]{prefix-fields}

Identifies the fields that are used to store prefixes. The default
set is: \field{prefix}, \field{prefixfirst}, \field{prefixplural},
\field{prefixfirstplural}, and their dual counterparts
\field{dualprefix}, \field{dualprefixfirst}, \field{dualprefixplural}
and \field{dualprefixfirstplural}.

\optsection[\subsubsection]{append-prefix-field}

Allowed values are:
\begin{itemize}
\item \optfmt{none}: don't append a space to the prefix fields (default);

\item \optfmt{space}: append the command identified by
\csopt{append-prefix-field-cs} (\ics{space} by default) to the
prefix field unless the field value ends with a character identified
by \csopt{append-prefix-field-exceptions} or a command identified by
\csopt{append-prefix-field-cs-exceptions}. Note that if the field
value ends with anything else (such as an empty group) then these
exceptions won't apply.

\item \optfmt{space or nbsp}: as above but uses \idx{nbspchar}
instead of \ics{space} if the field value matches the pattern given
by \csopt{append-prefix-field-nbsp-match}.

\end{itemize}

\optsection[\subsubsection]{append-prefix-field-cs}

Identifies the command \meta{cs} that should be used to append to
the prefix fields. The default value is \ics{space}. Remember to use
\cs{cs.string} or \cs{protect} to prevent the command from being
expanded as it's written to the \ext{aux} file.

\optsection[\subsubsection]{append-prefix-field-exceptions}

This setting identifies the set of characters that, if found at the
end of a prefix field, prevent \csopt{append-prefix-field} from
appending a space (either \ics{space} or \idx{nbspchar}).

The value should be a sequence of characters. You may use
\stringu\meta{hex} to identify a character by its hexadecimal code.
Spaces are ignored, so \csopt[ ' - ]{append-prefix-field-exceptions}
is equivalent to \csopt['-]{append-prefix-field-exceptions}.

The default set is the straight apostrophe character (\hex{0027}),
the hyphen-minus character (\hex{002D}), the tilde character
(\idx{nbspchar}), the hyphen character (\hex{2010}), the
non-breaking hyphen (\hex{2011}), and the right single quotation
mark (\hex{2019}).

\optsection[\subsubsection]{append-prefix-field-cs-exceptions}

This setting identifies the set of commands that, if found at the
end of a prefix field, prevent \csopt{append-prefix-field} from
appending a space (either \ics{space} or \idx{nbspchar}). Any spaces
found in \meta{sequence} are ignored. The default setting is the
set: \cs{space}, \ics{nobreakspace} and \cs{cs.space}.

Remember that you will need to use \cs{cs.string} or \cs{protect} to
prevent the command from being expanded while the resource options
are written to the \ext{aux} file.

\optsection[\subsubsection]{append-prefix-field-nbsp-match}

The value is the \idx{regex} that identifies prefixes that should be
followed by \idx{nbspchar} instead of \ics{space}. The default
is \csopt[\idx{matchanydot}]{append-prefix-field-nbsp-match} which indicates a
single character.

\subsection{Case-Changing}
\label{sec:fieldcase}

The \sty{glossaries-extra} package comes with the category
attributes \catattr{glossdesc} and \catattr{glossname}, which may
take the values \optfmt{firstuc} or \optfmt{title}. These don't
change the actual \field{name} or \field{description} fields, but
instead \ics{glossentryname} and \ics{glossentrydesc} (which are
used by the default glossary styles) check for the corresponding
attribute and apply the appropriate \idx{case-change} to the field value.

So \ics{glossentryname} will use \ics{Glsentryname} if the
\catattr{glossname} attribute for the given entry is set to \optfmt{firstuc} 
and \ics{glossentrydesc} will use \ics{Glsentrydesc} if the
\catattr{glossdesc} attribute is set to \optfmt{firstuc}.
The \optfmt{title} setting will instead use \ics{capitalisewords}
applied to the field value.

The resource options described in this section provide an
alternative to those attributes that actually modify the relevant
field (rather than just adjusting the style code used to display it). 
There are two forms of modification: the field is adjusted so that
the original value is encapsulated by a command or \bibgls\ will
perform the actual \idx{case-change} according to its own algorithm.
The results can vary according to the field content.

Only a subset of known fields have a resource option that can be
used to apply a \idx{case-change}. For example, \csopt{name-case-change}
can be used to change the case of the \field{name} field, but
there's no equivalent option for the \field{text} field.

Each of the case-changing resource options may take one of the
following values:
\begin{itemize}
\item \optfmt{none}: don't apply any case-changing (default);

\item \optfmt{lc-cs}: make \bibgls\ behave as though the field
assignment:
\begin{codeenv}
\meta{field} = \margm{text}
\end{codeenv}
had actually been specified as:
\begin{codeenv*}
\meta{field} = \marg{\gls{bibglslowercase}\margm{text}}
\end{codeenv*}
which uses \TeX\ to convert the field to \idx{lowercase};

\item \optfmt{uc-cs}: make \bibgls\ behave as though the field
assignment:
\begin{codeenv}
\meta{field} = \margm{text}
\end{codeenv}
had actually been specified as:
\begin{codeenv*}
\meta{field} = \marg{\gls{bibglsuppercase}\margm{text}}
\end{codeenv*}
which uses \TeX\ to convert the field to \idx{uppercase};

\item \optfmt{firstuc-cs}: make \bibgls\ behave as though the field
assignment:
\begin{codeenv}
\meta{field} = \margm{text}
\end{codeenv}
had actually been specified as:
\begin{codeenv*}
\meta{field} = \marg{\gls{bibglsfirstuc}\margm{text}}
\end{codeenv*}
which uses \TeX\ to convert the field to first-letter \idx{uppercase};

\item \optfmt{title-cs}: make \bibgls\ behave as though the field
assignment:
\begin{codeenv}
\meta{field} = \margm{text}
\end{codeenv}
had actually been specified as:
\begin{codeenv*}
\meta{field} = \marg{\gls{bibglstitlecase}\margm{text}}
\end{codeenv*}
which uses \TeX\ to convert the field to \idx{titlecase};

\item \optfmt{lc}: convert to \idx{lowercase} by making the
appropriate modifications to tokens in the field value that have
a known \idx{lowercase} alternative (see below);

\item \optfmt{uc}: convert to \idx{uppercase} by making the
appropriate modifications to tokens in the field value that have
a known \idx{uppercase} alternative (see below);

\item \optfmt{firstuc}: convert to first letter \idx{uppercase} 
by making the appropriate modification, if it has 
a known \idx{uppercase} alternative (see below);

\item \optfmt{title}: convert to \idx{titlecase} by making the
appropriate modifications to the first letter of each identified
word in the field value that has a known \idx{uppercase}
alternative (see below).

A word-boundary is identified according to the
\csopt{word-boundaries} setting. Words to be excluded from the
case-changing (unless they occur at the start) can be identified
with \ics{MFUnocap} in the \atentry{preamble} or you can use
\code{\longarg{packages} mfirstuc-english} for the exclusion list
provided by the \isty{mfirstuc-english} package. Alternatively, you
can use \longarg{custom-packages} to load a simple package that
contains the required \cs{MFUnocap} commands (in a similar style to
\sty{mfirstuc-english}).

The \bibgls\ word-boundary implementation is slightly different with
this setting than with the \cs{capitalisewords} command (implemented
in \TeX\ or by the \TeX\ parser library when interpreting field
values).  Only words in the exclusion list that start with an
alphabetical character can be matched.  Punctuation following a
word-boundary is not considered part of the next word.

\end{itemize}
The \optfmt{firstuc-cs} and \optfmt{firstuc} options are essentially
a \idx{sentencecase} change, but there's no check for
sentence-breaks within the value, so even if the value contains
multiple sentences, only the first is changed.

The \optfmt{\meta{option}-cs} settings defer the actual case-changing
to \TeX, which means that the case-changing has to be applied every
time the field is typeset (and it introduces non-expandable content
to the field value). Be aware of the limitations of using any of the
case-changing commands. See the \isty{textcase} and \isty{mfirstuc}
package documentation for further details~\cite{textcase,mfirstuc}.
You may use \code{\ics{NoCaseChange}\margm{content}} (provided by
\sty{textcase}) to prevent any \idx{case-change} to \meta{content}.

For the settings where \bibgls\ itself performs the \idx{case-change}, then
\bibgls\ will iterate over each token of the field value and apply
the following rules:
\begin{enumerate}

\item If the token is a normal Unicode alphabetic character, it will
be replaced with the corresponding upper or lower case character, as
appropriate.
For \optfmt{title} and \optfmt{firstuc}, the \idx{titlecase} character is
used as the replacement, for \optfmt{uc} the \idx{uppercase} character is
used as the replacement, and for \optfmt{lc} the \idx{lowercase} character
is used as the replacement. Many characters have the same upper and
title case alternative (for example, \qt{a} will be converted to \qt{A}
for the \optfmt{title}, \optfmt{firstuc} and \optfmt{uc} settings), but
some characters have different title and upper versions (for example, 
the digraph \qt{\char"01F3} has the title version \qt{\char"01F2}
and \idx{uppercase} version \qt{\char"01F1}).

If the option is \optfmt{firstuc} then all the remaining tokens are
skipped. If the option is \optfmt{title} then the subsequent tokens
are skipped until a word-boundary is found.

\item If the token is a normal Unicode character that isn't alphabetical,
then this token will be skipped for all options.

\item If \code{\idx{mshiftchar}\meta{maths}\idx{mshiftchar}} is
encountered, it will be skipped. If the option is \optfmt{firstuc}
then all remaining tokens are skipped, so no \idx{case-change} will be
performed.

\item\label{case-change-group} If a group \code{\margm{content}} is
found, then the \idx{case-change} is applied to the entire \meta{content}
(which may be empty). This corresponds to the way \ics{makefirstuc}
and \ics{capitalisewords} work if a~word starts with a group. Note
that with \optfmt{firstuc} and \optfmt{title} the group content will
be converted according to \optfmt{uc}, so the normal \idx{uppercase}
character is used rather than the title case character (if they are
different).

If the option is \optfmt{firstuc} then all the remaining tokens are
skipped. If the option is \optfmt{title} then the subsequent tokens
are skipped until a word-boundary is found.

\item If a control sequence \csfmt{}\meta{csname} is found, then:

  \begin{enumerate}
  \item If the control sequence is \cs{protect}, this token is skipped
  for all options.

  \item If the control sequence is one of: \ics{o}, \ics{O}, \ics{l},
  \ics{L}, \ics{ae}, \ics{AE}, \ics{oe}, \ics{OE}, \ics{aa}, \ics{AA},
  \ics{ss}, \ics{SS}, \ics{ng}, \ics{NG}, \ics{th}, \ics{TH},
  \ics{dh}, \ics{DH}, \ics{dj} or \ics{DJ}, then
  it's replaced with its \idx{case-change} counterpart (if not already the
  correct case).

  If the option is \optfmt{firstuc} then all the remaining tokens are
  skipped. If the option is \optfmt{title} then the subsequent tokens
  are skipped until a word-boundary is found.

  \item If the control sequence is \ics{NoCaseChange} or is in the
  \csopt{no-case-change-cs} list, then the 
  control sequence and its argument is ignored. With \optfmt{firstuc}
  and \optfmt{title}, if \code{\ics{NoCaseChange}\margm{text}}
  (but not any \csopt{no-case-change-cs} command)
  occurs at the start of a word, then \bibgls\ will act as though
  the word hasn't started yet (so the next token will be considered
  for a \idx{case-change}). This is different to the way \ics{makefirstuc}
  and \ics{capitalisewords} work.

  \item If the control sequence is \ics{ensuremath}, \ics{si} or if 
  \meta{csname} ends with \qtt{ref} (for example, \ics{ref} or 
  \ics{pageref}) then the control sequence and its argument is ignored.
  In the case where \meta{csname} ends with \qtt{ref}, a following
  star (\code{*}) or optional argument before the mandatory argument 
  will also be skipped. This allows for some common cross-referencing 
  commands, such as \sty{hyperref}['s] \ics{autoref}, which may have 
  a starred form, but does not allow for more complicated commands with 
  multiple arguments.

  If the option is \optfmt{firstuc} then all the remaining tokens are
  skipped (so no \idx{case-change} will be performed). If the option is 
  \optfmt{title} then the subsequent tokens
  are skipped until a word-boundary is found (so no \idx{case-change} is
  performed for this word).

  \item If the control sequence is \ics{glsentrytitlecase} then:
   \begin{description}
   \item[\optfmt{lc}] the control sequence is converted to
    \ics{glsxtrusefield};
   \item[\optfmt{uc}] the control sequence is converted to
    \ics{GLSxtrusefield};
   \item[\optfmt{firstuc}] the control sequence is converted to
    \ics{Glsxtrusefield} and the remaining tokens are skipped;
   \item[\optfmt{title}] the control sequence is left unchanged
   and subsequent tokens are skipped until a word-boundary is found.
   \end{description}
  The field and entry label arguments are skipped.

  \item If the control sequence is \ics{glshyperlink} then the
  \idx{case-change} is applied to its optional argument. (If there was no
  optional argument in the original field value, one will be inserted.)
  The label argument is skipped.

  If the option is \optfmt{firstuc} then all the remaining tokens are
  skipped. If the option is \optfmt{title} then the subsequent tokens
  are skipped until a word-boundary is found.

  \item If the control sequence is \ics{glsdisp}, \ics{glslink},
  \ics{dglsdisp} or \ics{dglslink} then the \idx{case-change} will be
  applied to the appropriate argument. The optional argument (if
  present) and the label are skipped.

  If the option is \optfmt{firstuc} then all the remaining tokens are
  skipped. If the option is \optfmt{title} then the subsequent tokens
  are skipped until a word-boundary is found.

  \item If the control sequence has a known case variant, it will be
  substituted. For example, \ics{gls} will be changed to \ics{Gls}
  or \ics{GLS}. In some cases there isn't an appropriate variant.
  For example, \ics{glsentrytext} has a first-letter upper case 
  version \ics{Glsentrytext}, but not an all-caps version.

  If the option is \optfmt{firstuc} then all the remaining tokens are
  skipped. If the option is \optfmt{title} then the subsequent tokens
  are skipped until a word-boundary is found.

  \item If the control sequence is followed by a group, then the
   appropriate \idx{case-change} is applied to the group contents.
   Unlike step~\ref{case-change-group}, the \idx{case-change} isn't
   applied to the entire group content with \optfmt{firstuc} and
   \optfmt{title}. (Again, this follows the way that
   \ics{makefirstuc} and \ics{capitalisewords} work.)

   If there are subsequent groups, they won't be considered arguments, 
   but will be treated as groups, as per step~\ref{case-change-group}. 
   (This will only affect the \optfmt{title} setting as they will be 
   skipped by the \optfmt{firstuc} setting.) For complex cases,
   consider using a semantic command that hides non-textual context
   such as the \csfmt{strong} example described on page~\glsxtrpageref{strong}.

  \item Otherwise the control sequence is skipped.
  \end{enumerate}

\item Anything else is skipped.

\end{enumerate}

For example, if an entry is defined as:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
 \field{short} = \marg{HTML},
 \field{long}  = \marg{hypertext markup language},
 \field{description}=\marg{a markup language for creating web pages}
}
\end{codeenv}
then:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[lc]{short-case-change},
  \csopt[title]{long-case-change},
  \csopt[firstuc]{description-case-change}
}
\end{codeenv}
will make the entry behave as if it had been defined as:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
 \field{short} = \marg{html},
 \field{long}  = \marg{Hypertext Markup Language},
 \field{description}=\marg{A markup language for creating web pages}
}
\end{codeenv}
whereas: 
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[lc-cs]{short-case-change},
  \csopt[title-cs]{long-case-change},
  \csopt[firstuc-cs]{description-case-change}
}
\end{codeenv}
will make the entry behave as if it had been defined as:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
 \field{short} = \marg{\gls{bibglslowercase}\marg{HTML}},
 \field{long}  = \marg{\gls{bibglstitlecase}\marg{hypertext markup language}},
 \field{description}=\marg{\gls{bibglsfirstuc}\marg{a markup language for creating web pages}}
}
\end{codeenv}
If the given field is missing, no change is made, except under
certain circumstances (see the relevant resource option for details). 
For example, if an abbreviation is simply defined as:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
 \field{short} = \marg{html},
 \field{long}  = \marg{hypertext markup language}
}
\end{codeenv}
then:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[uc]{name-case-change},
  \csopt[title]{description-case-change}
}
\end{codeenv}
won't have an effect. Although the default \abbrstyle{long-short} abbreviation
style sets the \field{name} and \field{description} fields, \bibgls\ doesn't
have access to this information.

Remember that you can create missing fields by copying the value
from another field. So if the resource options are changed to:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[uc]{name-case-change},
  \csopt[title]{description-case-change},
  \csopt[\field{short}=\field{name},\field{long}=\field{description}]{replicate-fields}
}
\end{codeenv}
then \bibgls\ will act as though the entry had been defined as:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
 \field{short} = \marg{html},
 \field{long}  = \marg{hypertext markup language},
 \field{name} = \marg{HTML},
 \field{description}  = \marg{Hypertext Markup Language}
}
\end{codeenv}
If the \abbrstyle{long-short-sc} abbreviation style is set (before
\gls{GlsXtrLoadResources}) then this will override the default style
for the \field{name} and \field{description}, so
\code{\cs{gls}\marg{html}} will display the short form using
\code{\ics{textsc}\marg{html}} but the \field{name} in the glossary
will be displayed using just \code{HTML}.

Note that with \atentry{index} the \field{name} and \field{text}
fields will automatically be created if they are missing and
\csopt{name-case-change} is used. For example, if an entry is
defined as:
\begin{codeenv}
\atentry{index}\marg{duck}
\end{codeenv}
then \csopt[firstuc]{name-case-change} will make this entry behave
as though it was defined as:
\begin{codeenv}
\atentry{index}\marg{duck,
  \field{name} = \marg{Duck},
  \field{text} = \marg{duck}
}
\end{codeenv}

Suppose I have a slightly eccentric abbreviation definition:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
  \field{short} = \qtdelim{ht\cs{emph}\marg{ml}},
  \field{long}  = \qtdelim{hypertext markup language}
}
\end{codeenv}
then \csopt[uc]{short-case-change} would convert the value of the
\field{short} field into:
\begin{codeenv}
HT\cs{emph}{ML}
\end{codeenv}
Note that \csfmt{emph} isn't modified as it's recognised as a
command. There's a difference between a group that follows a control
sequence and one that doesn't. For example:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
  \field{short} = \qtdelim{\marg{ht}ml},
  \field{long}  = \qtdelim{hypertext markup language}
}
\end{codeenv}
In this case \csopt[firstuc]{short-case-change} will convert the
\field{short} field value to:
\begin{codeenv}
\marg{HT}ml
\end{codeenv}
(The entire contents of the group \code{\marg{ht}} has been
converted.) Whereas with:
\begin{codeenv}
\atentry{abbreviation}\marg{html,
  \field{short} = \qtdelim{\cs{emph}\marg{ht}ml},
  \field{long}  = \qtdelim{hypertext markup language}
}
\end{codeenv}
then \csopt[firstuc]{short-case-change} will convert the
\field{short} field value to:
\begin{codeenv}
\cs{emph}\marg{Ht}ml
\end{codeenv}
(Only the first letter of the argument \code{\marg{ht}} has been
converted.)

There's no attempt at interpreting the field contents at this point (but
the value may later be interpreted during sorting).
For example, suppose a \field{name} field is defined using:
\begin{codeenv}
\field{name} = \qtdelim{z\ics{ae}\ics{oe}},
\end{codeenv}
then with \csopt[uc]{name-case-change}, the value would be converted to
\begin{codeenv}
Z\ics{AE}\ics{OE}
\end{codeenv}
because \cs{ae} and \cs{oe} have known \idx{uppercase} versions.

With \csopt[uc-cs]{name-case-change}, the \field{name}
value would be converted to:
\begin{codeenv}
\gls{bibglsuppercase}\marg{z\cs{ae}\cs{oe}}
\end{codeenv}
If the interpreter is used during sorting, the sort value will be set to 
\code{Z\AE\OE} because the interpreter recognises all three commands.

You can use \code{\ics{NoCaseChange}\margm{text}} to prevent the given 
\meta{text} from having the case changed. For example, if the \field{short}
field is defined as:
\begin{codeenv}
\field{short} = \marg{a\cs{NoCaseChange}\marg{bc}d}
\end{codeenv}
then with \csopt[uc]{short-case-change}, this would be converted to
\begin{codeenv}
A\cs{NoCaseChange}\marg{bc}D
\end{codeenv}

Note that with \optfmt{firstuc} and \optfmt{title}, if
\code{\ics{NoCaseChange}\margm{text}} occurs at the start of a word
then it's skipped, and the case change is
applied to the material following its argument. For example,
suppose the \field{short} field is defined as:
\begin{codeenv}
\field{short}=\marg{\cs{NoCaseChange}\marg{h}tml}
\end{codeenv}
then the result is:
\begin{codeenv}
\cs{NoCaseChange}\marg{h}Tml
\end{codeenv}
whereas with:
\begin{codeenv}
\field{short}=\marg{\marg{}html}
\end{codeenv}
then the result is just \code{\marg{}html} (since the case change is
applied to the empty group, which has no effect).

If you have a command that takes a label or identifier as an argument then it's
best to hide the label in a custom command. For example, if the
\field{short} field in the \ext{bib} definition is defined as:
\begin{codeenv}
  \field{short} = \qtdelim{ht\ics{textcolor}\marg{red}\marg{ml}},
\end{codeenv}
then with \csopt[uc]{short-case-change} this would end up as:
\begin{codeenv}
HT\cs{textcolor}\marg{RED}\marg{ML}
\end{codeenv}
which is incorrect. Instead, provide a command that hides the label
(such as the \csfmt{strong} example described on page~\glsxtrpageref{strong}).

\optsection[\subsubsection]{no-case-change-cs}

Instructs the non-\TeX\ case-changing options (where \bibgls, not
\TeX, performs the modification) to treat the commands whose control
sequence names are given in the comma-separated \meta{list} in the
same way as it treats \cs{ensuremath} etc. That is, the case-change
is omitted for the argument that follows any of those commands.

For example, this manual defines some semantic commands such as
\csfmt{fieldfmt} (to format field names), \csfmt{abbrstylefmt}
(to format abbreviation style names) and \csfmt{glostylefmt} (to format
glossary style names). If any these occur in section
and subsection headings (which are converted to \idx{titlecase})
then the case-change would produce an inappropriate result. 
These formatting commands shouldn't have their argument changed so they 
are identified with:
\begin{codeenv}
\csopt[fieldfmt,abbrstylefmt,glostylefmt]{no-case-change-cs}
\end{codeenv}

\optsection[\subsubsection]{word-boundaries}

Governs how the \optfmt{title} \idx{case-change} option
determines word boundaries. The \meta{list} must contain one or
more of the following keywords:
\begin{description}
\item[\optfmt{white space}] any white space Unicode character
that is not a non-breakable space indicates a word-boundary;
\item[\optfmt{cs space}] the control sequences
\ics{space} or \ics{cs.space} indicate a word-boundary;
\item[\optfmt{dash}] a Unicode character that belongs to the
\idx{punctuationdash} block indicates a word-boundary;
\item[\optfmt{nbsp}] the \idx{nbspchar} active character or the Unicode
non-breakable characters \hex{00A0}, \hex{2007} and \hex{202F}
indicate a word-boundary.
\end{description}
Any keyword that is not listed indicates that particular setting is off.  This
option is not cumulative. Any subsequent use of
\csopt{word-boundaries} within the same set of resource options will
override previous settings.

The default setting is \csopt[white space,cs
space]{word-boundaries}, which excludes non-breakable
spaces and dashes.

\optsection[\subsubsection]{short-case-change}

Applies a case-change to the \field{short} field (if present).
This option may take one of the values described above.

See \csopt{dual-short-case-change} to adjust the \field{dualshort}
field.

\optsection[\subsubsection]{long-case-change}

Applies a case-change to the \field{long} field (if present).
This option may take one of the values described above.

See \csopt{dual-long-case-change} to adjust the \field{duallong}
field.

\optsection[\subsubsection]{name-case-change}

Applies a case-change to the \field{name} field.
This option may take one of the values described above.

If the \field{text} field hasn't been set, the \field{name}
value is first copied to the \field{text} field. If the \field{name}
field hasn't been set (for example, with the \atentry{index} entry
type), it's copied from the fallback value (which
depends on the entry type) unless the entry type is
\atentry{abbreviation} or \atentry{acronym}, in which case if
the \field{name} field is missing no action is performed.

\optsection[\subsubsection]{description-case-change}

Applies a case-change to the \field{description} field (if present).
This option may take one of the values described above.

\optsection[\subsubsection]{field-case-change}

A general case-change instruction. The value should be a
comma-separated list of \code{\meta{field}\dequals\meta{setting}} for each
field that needs a case-change applied.  The value is required for this key but
may be empty, which indicates this option is switched off.

The \meta{setting} should be
the same as the permitted values for the above options. This
option is applied after all fields have been parsed but before
\csopt{interpret-fields}. If the
specified field is missing, the fallback for that field (if known) is copied
into the field. For example:
\begin{codeenv}
\csopt[\field{user1}=uc,\field{user2}={firstuc}]{field-case-change}
\end{codeenv}
This manual provides a custom storage key called
\fieldfmt{nametitle}:
\begin{codeenv}
\cs{glsxtrprovidestoragekey}\marg{\fieldfmt{nametitle}}\marg{}\marg{}
\end{codeenv}
The resource options copy the \field{name} value to this custom
field and convert \fieldfmt{nametitle} to \idx{titlecase}:
\begin{codeenv}
\csopt[\field{name}=\fieldfmt{nametitle}]{replicate-fields},
\csopt[\fieldfmt{nametitle}=title]{field-case-change},
\end{codeenv}
This means that it's possible to fetch the value of
\fieldfmt{nametitle} instead of \field{name}, which provides an
expandable \idx{titlecase} form that's suitable for the PDF
bookmarks.

This option isn't cumulative. If used multiple times in the same
\idx{resourceset}, the last instance will be the one used. If
the \keyvallist\ is missing, no general case-changing is applied
(the default).

\section{Plurals}
\label{sec:plurals}

Some languages, such as English, have a general rule that plurals
are formed from the singular with a suffix appended. This isn't 
an absolute rule. There are plenty of exceptions (for example,
geese, children, churches, elves, fairies, sheep, mice), so a
simplistic approach of just doing \code{\ics{gls}\margm{label}[s]}
will sometimes produce inappropriate results, so the \styfmt{glossaries}
package provides a \field{plural} key with the corresponding command
\ics{glspl}.

In some cases a plural may not make any sense (for example, if the
term is a verb or symbol), so the \field{plural} key is optional, but to
make life easier for languages where the majority of plurals can
simply be formed by appending a suffix to the singular, the
\styfmt{glossaries} package lets the \field{plural} field default
to the value of the \field{text} field with \ics{glspluralsuffix}
appended. This command is defined to be just the letter \qt{s}.
This means that the majority of terms in such languages don't need to have the
\field{plural} supplied as well, and you only need to use it for the
exceptions.

For languages that don't have this general rule, the \field{plural}
field will always need to be supplied for nouns.

There are other plural fields, such as \field{firstplural},
\field{longplural} and \field{shortplural}. Again, if you are using
a language that doesn't have a simple suffix rule, you'll have to
supply the plural forms if you need them (and if a plural makes
sense in the context).

If these fields are omitted, the \styfmt{glossaries} package follows
these rules:
\begin{itemize}
\item If \field{firstplural} is missing, then \ics{glspluralsuffix}
is appended to the \field{first} field, if that field has been
supplied. If the \field{first} field hasn't been supplied but the
\field{plural} field has been supplied, then the \field{firstplural}
field defaults to the \field{plural} field. If the \field{plural}
field hasn't been supplied, then both the \field{plural} and
\field{firstplural} fields default to the \field{text} field (or
\field{name}, if no \field{text} field) with \ics{glspluralsuffix}
appended.

\item If the \field{longplural} field is missing, then 
\ics{glspluralsuffix} is appended to the \field{long} field, if the
\field{long} field has been supplied.

\item If the \field{shortplural} field is missing then, \emph{with
the base \styfmt{glossaries} acronym mechanism}, \ics{acrpluralsuffix}
is appended to the \field{short} field.

\end{itemize}

The last case is different with the \isty{glossaries-extra} extension 
package. The \field{shortplural} field defaults to the \field{short}
field with \ics{abbrvpluralsuffix} appended \emph{unless overridden
by category attributes}. This suffix command is set by the
abbreviation styles. This means that every time an abbreviation
style is implemented, \ics{abbrvpluralsuffix} is redefined. Most
styles simply define this command as:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{abbrvpluralsuffix}}\marg{\cs{glsxtrabbrvpluralsuffix}}
\end{codeenv}
where \ics{glsxtrabbrvpluralsuffix} expands to \ics{glspluralsuffix}.
The \qt{sc} styles (such as \abbrstyle{long-short-sc}) use a different
definition:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{abbrvpluralsuffix}}\marg{\cs{protect}\ics{glsxtrscsuffix}}
\end{codeenv}
This allows the suffix to be reverted back to the upright font,
counteracting the affect of the small-caps font.

This means that if you want to change or strip the suffix used for
the plural short form, it's usually not sufficient to redefine
\ics{abbrvpluralsuffix}, as the change will be undone the next time
the style is applied. Instead, for a document-wide solution, you
need to redefine \ics{glsxtrabbrvpluralsuffix}. Alternatively you can
use the category attributes.

There are two attributes that affect the short plural suffix
formation. The first is \catattr{aposplural} which uses the suffix
\begin{codeenv*}
\idx{aposchar}\cs{abbrvpluralsuffix}
\end{codeenv*}
That is, an \idx{apostrophe} followed by \ics{abbrvpluralsuffix} is
appended. The second attribute is \catattr{noshortplural} which
suppresses the suffix and simply sets \field{shortplural} to the
same as \field{short}.

With \bibgls, if you have some abbreviations where the plural should
have a suffix and some where the plural shouldn't have a suffix
(for example, the document has both English and French abbreviations)
then there are two approaches.

The first approach is to use the category attributes. For example:
\begin{codeenv}
\cs{glssetcategoryattribute}\marg{french}\marg{\catattr{noshortplural}}
\end{codeenv}
Now just make sure all the French abbreviations are have their
\field{category} field set to \optfmt{french}:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[fr-abbrvs]{src},\csopt[french]{category}}
\end{codeenv} 

The other approach is to use the options listed below for the given
\idx{resourceset}. For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[fr-abbrvs]{src},\csopt[\empty]{short-plural-suffix}}
\end{codeenv} 

\optsection{short-plural-suffix}

Sets the plural suffix for the default \field{shortplural} to
\meta{value}. The \meta{value} may be one of:
\begin{itemize}
\item \meta{suffix}: add the
\field{shortplural} field, if missing, with the given \meta{suffix}. 
\item \meta{empty}: add the
\field{shortplural} field, if missing, with no suffix.
\item \optfmt{use-default}: leave
it to \styfmt{glossaries-extra} to determine the appropriate default.
\end{itemize}
The default setting is \csopt[use-default]{short-plural-suffix}.
If the \code{=\meta{value}} part is omitted, then
\csopt[\empty]{short-plural-suffix} is assumed.

\optsection{dual-short-plural-suffix}

Sets the plural suffix for the default \field{dualshortplural} field to
\meta{value}. As with \csopt{short-plural-suffix}, the default
setting is \csopt[use-default]{dual-short-plural-suffix}. If the
\meta{value} is omitted or empty, the suffix is set to empty.

\section{Location List Options}
\label{sec:locationopts}

The \styopt{record} package option automatically adds two new keys: 
\field{loclist} and \field{location}. These two fields are set by
\bibgls\ from the information supplied in the \iext{aux} file (unless
the option \csopt[false]{save-locations} is used). The
\field{location} field contains the code to typeset the formatted
\idx{locationlist}.

The \field{loclist} field has the syntax of an \isty{etoolbox} internal list
and includes every location (except for the discarded duplicates and
\idxpl{ignoredrecord}) with no range formations. Any explicit range markup 
is stripped from the \glsopt{format} information to leave just the \idx{encap}
name, so you just get the start and end locations added as individual 
elements but they are still encapsulated with the associated formatting
command. Each item in the list is provided in one of the following forms:
\begin{codeenv}
\ics{glsseeformat}\oargm{tag}\margm{label list}\marg{}
\end{codeenv}
for the cross-reference supplied by the \field{see} field,
\begin{codeenv}
\ics{glsxtruseseealsoformat}\margm{xr list}
\end{codeenv}
for the cross-reference supplied by the \field{seealso} field,
\nosecdef{glsnoidxdisplayloc}
for standard the internal locations, 
\begin{codeenv}
\format{glsxtrdisplaysupploc}
\end{codeenv}
for supplemental (external) locations and
\begin{codeenv}
\format{glsxtrdisplaylocnameref}
\end{codeenv}
for \code{nameref} records. (See \sectionref{sec:supplementalopts}
for more information about supplemental locations and
\longarg{merge-nameref-on} for more information about \code{nameref}
records.)

You can iterate through the \field{loclist} value
using one of \sty{etoolbox}'s internal list \idxpl{loop} (either
by first fetching the list using \ics{glsfieldfetch}
or through \styfmt{glossaries-extra}'s \ics{glsxtrfielddolistloop}
or \ics{glsxtrfieldforlistloop} shortcuts).

The \meta{format} is that supplied by the \glsopt{format} key
when using commands like \ics{gls} or \ics{glsadd} (the encapsulator or 
\idx{encap} in \idx!{makeindex} parlance). If omitted, the default
\glsopt[glsnumberformat]{format} is assumed (unless this
default value is changed with \ics{GlsXtrSetDefaultNumberFormat}.
The value of the \glsopt{format} key must be the name of 
a text-block command without the leading backslash that takes 
a single argument (the location). The location
is encapsulated by that command. For example, 
\begin{codeenv}
\cs{gls}\oarg{\glsopt[\encap{textbf}]{format}}\marg{sample}
\end{codeenv}
will display the corresponding location in bold, but note that this
will no longer have a hyperlink if you've used \sty{hyperref}.
If you want to retain the hyperlink you need the location
encapsulated with \cs{hyperbf} instead of \cs{textbf}:
\begin{codeenv}
\cs{gls}\oarg{\glsopt[\encap{hyperbf}]{format}}\marg{sample}
\end{codeenv}
The \csfmt{hyper}\meta{xx} set of commands all internally use
\cs{glshypernumber} which adds the appropriate hyperlink to the
location. See Table~6.1 in the \styfmt{glossaries}~\cite{glossaries}
user manual for a list of all the \csfmt{hyper}\meta{xx} commands.

Ranges can be explicitly formed using the parenthetical 
syntax \glsopt[\idx{openrange}]{format} and
\glsopt[\idx{openrange}]{format} or \glsopt[\idx{openrange}\meta{csname}]{format} and
\glsopt[\idx{closerange}\meta{csname}]{format} (where \meta{csname} is again the name of 
a text-block command without the initial backslash) in the optional
argument of commands like \ics{gls} or \ics{glsadd}. These will always
form a range, regardless of \csopt{min-loc-range}, and will be
encapsulated by \gls!{bibglsrange}. (This command is not used with
ranges that are formed by collating consecutive locations.) The
initial marker is stripped from the \meta{format} argument of the
location formatting commands, such as \gls{glsnoidxdisplayloc}, to
allow for easy conversion to the corresponding text-block command.

Explicit ranges don't merge with neighbouring locations, but will
absorb any single locations within the range that don't conflict.
(Conflicts will be moved to the start of the explicit range.)
For example, if \code{\cs{gls}\marg{sample}} is used on page~1,
\code{\cs{gls}\oarg{\glsopt[\idx{openrange}]{format}}\marg{sample}} is used on page~2, 
\code{\cs{gls}\marg{sample}} is used on page~3, and
\code{\cs{gls}\oarg{\glsopt[\idx{closerange}]{format}}\marg{sample}} is used on page~4, then the location
list will be 1, 2--4. The entry on page~3 is absorbed into the
explicit range, but the range can't be expanded to include page~1.
If the entry on page~3 had a different format to the explicit range,
for example \code{\cs{gls}\oarg{\glsopt[\encap{textbf}]{format}}\marg{sample}} 
then it would cause a warning and be moved before the start of the
range so that the location list would then be 1, \textbf{3}, 2--4. 

An \idx{ignoredrecord} identifies a term that needs to be treated as
though it has a record for selection purposes, but the record
should not be included in the location list.
The special format \glsaddopt[\encap{glsignore}]{format} is provided
by the \styfmt{glossaries} package for cases where the location
should be ignored. (The command \ics{glsignore} simply ignores its
argument.) This works reasonably well if an entry only has the one
location, but if the entry happens to be indexed again, it can lead
to an odd empty gap in the location list with a spurious comma. If
\bibgls\ encounters a record with this special format, the entry
will be selected but the record will be discarded.

This means that the location list will be empty if the entry was
only indexed with the special ignored format, but if the entry was also
indexed with another format then the location list won't include the
\idxpl{ignoredrecord}. (This format is used by \ics{glsaddallunused} but
remember that iterative commands like this don't work with \bibgls.
Instead, just use \csopt[all]{selection} to select all entries.
Those that don't have records won't have a location list.)

For example, suppose you only want main matter locations in the
number list, but you want entries that only appear in the back matter 
to still appear in the glossary (without a location list), then you could do:
\begin{codeenv}
\ics{backmatter}
\ics{GlsXtrSetDefaultNumberFormat}\marg{\encap{glsignore}}
\end{codeenv}
If you also want to drop front matter locations as well:
\begin{codeenv}
\ics{frontmatter}
\cs{GlsXtrSetDefaultNumberFormat}\marg{\encap{glsignore}}
\ldots
\ics{mainmatter}
\cs{GlsXtrSetDefaultNumberFormat}\marg{\encap{glsnumberformat}}
\ldots
\cs{backmatter}
\cs{GlsXtrSetDefaultNumberFormat}\marg{\encap{glsignore}}
\end{codeenv}

Note that explicit range formations aren't discarded, so if
\encap{glsignore} is used in a range, such as:
\begin{codeenv}
\ics{glsadd}\oarg{\glsopt[\idx{openrange}\encap{glsignore}]{format}}\marg{sample}
\ldots 
\cs{glsadd}\oarg{\glsopt[\idx{closerange}\encap{glsignore}]{format}}\marg{sample} 
\end{codeenv}
then the range will be included in the location list (encapsulated
with \ics{glsignore}), but this case would be a rather odd use of
this special format and is not recommended.

The record counting commands, such as \gls{rgls}, use the special
format \encap{glstriggerrecordformat}, which \bibgls\ also treats
as an \idx{ignoredrecord} and the same rules as for \encap{glsignore} apply.

The locations are always listed in the order in which they were indexed,
(except for the cross-reference which may be placed at the start or
end of the list or omitted).
This is different to \idx!{xindy} and \idx!{makeindex} where you can
specify the ordering (such as \idx!{lowercase} Roman first, then digits,
etc), but unlike those applications, \bibgls\ allows any location,
although it may not be able to work out an integer representation.
(With \idx!{xindy}, you can define new location formats, but you need
to remember to add the appropriate code to the custom module.)

It's possible to define a custom glossary style where 
\ics{glossentry} (and the child form \ics{subglossentry}) ignore the
final argument (which will be the \field{location} field)
and instead parse the \field{loclist} field and re-order the
locations or process them in some other way. Remember that you can
also use \ics{glsnoidxloclist}
provided by \styfmt{glossaries}. For example:
\begin{codeenv}
\cs{glsfieldfetch}\marg{gls.sample}\marg{loclist}\marg{\cmd{loclist}}\comment{fetch location list}
\cs{glsnoidxloclist}\marg{\cmd{loclist}}\comment{iterate over locations}
\end{codeenv}
This uses \ics{glsnoidxloclisthandler} as the list's \idx{handler}
macro, which simply displays each location separated by \ics{delimN}.
(See also
\href{http://www.dickimaw-books.com/latex/admin/html/foreachtips.shtml}{Iteration
Tips and Tricks}~\cite{iterationtips}.)

Each regular location is listed in the \iext{aux} file in the form:
\nosecdef{glsxtr@record}
(See \longarg{merge-nameref-on} for \code{nameref} records.)
Exact duplicates are discarded. For example, if \code{cat}
is indexed twice on page 1:
\begin{codeenv}
\gls{glsxtr@record}\marg{cat}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\gls{glsxtr@record}\marg{cat}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\end{codeenv}
then the second record is discarded. Only the first record is added
to the location list.

Partial duplicates, where all arguments match except for
\meta{format}, may be discarded depending on the value
of \meta{format}. For example, if page~1 of the document
uses \code{\cs{gls}\marg{cat}} and
\code{\cs{gls}\oarg{\glsopt[\encap{hyperbf}]{format}}\marg{cat}}
then the \ext{aux} file will contain:
\begin{codeenv}
\gls{glsxtr@record}\marg{cat}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\gls{glsxtr@record}\marg{cat}\marg{}\marg{\counter{page}}\marg{\encap{hyperbf}}\marg{1}
\end{codeenv}
This is a partial record match. In this case, \bibgls\ 
makes the following tests:
\begin{itemize}
\item If one of the formats includes a range formation, the
range takes precedence.

\item If one of the formats is \encap{glsnumberformat} (as in the
above example) or an \idx{ignoredrecord} format such as
\encap{glsignore}, that format will be skipped.  So in the above
example, the second record will be added to the location list, but
not the first. (A message will only be written to the transcript if
the \longarg{debug} switch is used.) The default
\encap{glsnumberformat} will take precedence over the
\idx{ignoredrecord} formats (\encap{glsignore} and
\encap{glstriggerrecordformat}).

\item If a mapping has been set with the \longarg{map-format}
switch that mapping will be checked.
\item Otherwise the duplicate record will be discarded with a
warning.
\end{itemize}

The \field{location} field is used to store the formatted location
list. The code for this list is generated by \bibgls\ based on the
information provided in the \iext{aux} file, the presence of the
\field{see} or \field{seealso} field and the various settings described in this
chapter. When you display the glossary using \ics{printunsrtglossary},
if the \field{location} field is present it will be displayed
according to the glossary style (and other factors, such as whether the
\styopt{nonumberlist} option has been used, either as a package
option or supplied in the optional argument of \ics{printunsrtglossary}).
For more information on adjusting the formatting see the
\styfmt{glossaries}~\cite{glossaries} and 
\styfmt{glossaries-extra}~\cite{glossaries-extra} user manuals.

\optsection{save-locations}

By default, the locations will be processed and stored in the
\field{location} and \field{loclist} fields. However, if you don't
want the location lists (for example, you are using the
\styopt{nonumberlist} option or you are using \idx!{xindy} with a
custom location rule), then there's no need for \bibgls\ to process
the locations. To switch this function off, just use
\csopt[false]{save-locations}. Note that with this setting, if
you're not additionally using \idx!{makeindex} or \idx!{xindy}, then
the locations won't be available even if you don't have the
\styopt{nonumberlist} option set.

The boolean \field{nonumberlist} key that may be used in
\gls{newglossaryentry} can also be used in a \ext{bib} file, but in
this case it can't have an empty value. The value must be either
\code{true} or \code{false}. If \code{true} then \bibgls\ won't
save the \field{location} or \field{loclist} fields.

The \field{nonumberlist} key provided by the base \sty{glossaries}
package doesn't represent a real field. The value isn't saved but,
if used, it will alter the indexing information that's written to
the \idx!{makeindex} or \idx!{xindy} file. It's a little hack to
ensure that the location is hidden for a specific entry when used
with \idx!{makeindex} and \idx!{xindy}.

\bibgls\ will look for this key to determine if the location should
be omitted for the given entry, but it won't write the key to the 
\ext{glstex} file.

\optsection{save-loclist}

If you want the \field{location} field but don't need
\field{loclist}, you can use \csopt[false]{save-loclist}.
This can help to save resources and build time.

\optsection{save-primary-locations}

It's sometimes useful to identify primary locations with a different
format, such as bold or italic. This helps the reader select which
location to try first in the event of a long location list. However,
you may prefer to store the primary location in a different field to
give it a more prominent position. In order to do this you 
need to specify the format (or formats) used to identify primary
locations with \csopt{primary-location-formats} and 
use \csopt{save-primary-locations} to determine how to deal with
these locations.

This option may take one of the following values:
\begin{itemize}
\item \optfmt{false}: don't save primary locations (default);
\item \optfmt{retain}: save primary locations in the
\field{primarylocations} field but don't remove from the usual
\idx{locationlist};
\item \optfmt{default format}: similar to \optfmt{retain} but the
format for the primary records in the \field{location} field is
converted to the default \encap{glsnumberformat} \idx{encap} (the records in the
\field{primarylocations} field retain their given format);
\item \optfmt{start}: save primary locations in the
\field{primarylocations} field and also move to the start of the
usual \idx{locationlist};
\item \optfmt{remove}: save primary locations in the
\field{primarylocations} field and remove from the usual
\idx{locationlist}.
\end{itemize}

The primary locations are copied to the \field{primarylocations}
field and encapsulated with \gls!{bibglsprimary}.
If you use \csopt[remove]{save-primary-locations}, the
\field{location} field will end up empty if the locations for the
associated entry were all identified as primary. If you use
\csopt[start]{save-primary-locations}, all primary locations will be
moved to the start of the \idx{locationlist} stored in the
\field{location} field, but there will be no additional markup
(other than the given format) to identify them. If you need
additional markup, then use \csopt[remove]{save-primary-locations}
and adjust the \idx{locationlist} format to insert the primary
locations at the start. This can be done by modifying the glossary
style.

For example, the \glostyle{bookindex} style inserts
\ics{glsxtrbookindexprelocation} before the location, so you could
redefine this:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrbookindexprelocation}}[1]\marg{\comment{}
  \cs{glsxtrifhasfield}\marg{primarylocations}\marg{\idx{param}1}\comment{}
  \marg{\comment{}
    \ics{glsxtrprelocation}
    \cs{glscurrentfieldvalue}
    \cs{glsxtrifhasfield}\marg{location}\marg{\idx{param}1}\marg{;}\marg{}\comment{}
  }\comment{}
  \marg{}\comment{}
  \cs{glsxtrprelocation}
}
\end{codeenv}
(Note that if \csopt{loc-prefix} is used, the prefix will be
in the \field{location} field and so will come after the primary
locations in the above example. Similarly for cross-references
unless they've been omitted.)

You can switch from using the \field{location} field to the
\field{primarylocations} field by locally changing
\ics{GlsXtrLocationField}:
\begin{codeenv}
\cs{printunsrtglossary*}\marg{\comment{}
  \cs{renewcommand}\marg{\cs{GlsXtrLocationField}}\marg{primarylocations}\comment{}
}
\end{codeenv}
Remember that the \idx{handler} used by \cs{printunsrtglossary} will
fallback on the \field{loclist} field if the field identified
by \cs{GlsXtrLocationField} is missing or empty. You may want to
consider using \csopt[false]{save-loclist} to prevent this.

\optsection{primary-location-formats}

This option will automatically set
\csopt[retain]{save-primary-locations} unless it has already been
changed from the default \csopt[false]{save-primary-locations}
setting. The argument should be a comma-separated list of 
formats. If a record's format is contained in this list then it will
be considered a primary location and it will be included in the
associated entry's \field{primarylocations} field.

For example, suppose the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{entry}\marg{bird,
  \field{name}=\marg{bird},
  \field{description}=\marg{feathered animal}
}
\atentry{entry}\marg{waterfowl,
  \field{name}=\marg{waterfowl},
  \field{description}=\marg{any bird that lives in or about water}
}
\atentry{entry}\marg{zebra,
  \field{name}=\marg{zebra},
  \field{description}=\marg{striped African horse}
}
\atentry{entry}\marg{parrot,
  \field{name}=\marg{parrot},
  \field{description}=\marg{mainly tropical bird with bright plumage}
}
\end{codeenv}
and the document \filefmt{test.tex} contains:
\begin{codeenv}
\cmd{documentclass}\marg{report}
\strut
\cmd{usepackage}\oarg{colorlinks}\marg{hyperref}
\cmd{usepackage}\oarg{\styopt{record},
 \styopt[dot]{postpunc},
 \styopt{nostyles},
 \styopt[tree,bookindex]{stylemods},
 \styopt[bookindex]{style}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},
  \csopt[hyperbf,hyperemph]{primary-location-formats},
  \csopt[remove]{save-primary-locations}
}
\strut
\cmd{renewcommand}*\marg{\cs{glsxtrbookindexprelocation}}[1]\marg{\comment{}
  \cs{glsxtrifhasfield}\marg{\field{primarylocations}}\marg{\idx{param}1}\comment{}
  \marg{\comment{}
    \ics{glsxtrprelocation}
    \cs{glscurrentfieldvalue}
    \cs{glsxtrifhasfield}\marg{\field{location}}\marg{\idx{param}1}\marg{;}\marg{}\comment{}
  }\comment{}
  \marg{}\comment{}
  \cs{glsxtrprelocation}
}
\strut
\gls{glsxtrnewglslike}\oarg{\glsopt[\encap{hyperbf}]{format}}\marg{}\marg{\cmd{primary}}\marg{\cmd{primarypl}}\marg{\cmd{Primary}}\marg{\cmd{Primarypl}}
\strut
\cmd{begin}\marg{document}
\cmd{chapter}\marg{Sample}
\cmd{Primary}\marg{waterfowl}, \cs{gls}\marg{bird} and \cs{gls}\marg{zebra}.
\strut
\cmd{chapter}\marg{Another Sample}
\cs{Gls}\marg{waterfowl}, \cmd{primary}\marg{bird} and \cs{gls}\marg{zebra}.
\strut
\cmd{chapter}\marg{Yet Another Sample}
\cs{Gls}\marg{waterfowl}, \cs{gls}\marg{bird} and \cmd{primary}\marg{zebra}.
\strut
\cmd{chapter}\marg{Yet Another Sample Again}
\cs{Gls}\marg{waterfowl}, \cs{gls}\marg{bird}, \cmd{primarypl}\marg{parrot} and \cs{gls}\marg{zebra}.
\strut
\ics{printunsrtglossary*}\oarg{\printglossopt[\iglostyle{tree}]{style},\printglossopt{nonumberlist}}\marg{\comment{}
 \cmd{renewcommand}*\marg{\ics{glsextrapostnamehook}}[1]\marg{\ics{glsadd}\oarg{\glsopt[\encap{hyperemph}]{format}}\marg{\idx{param}1}}\comment{}
}
\strut
\cs{printunsrtglossary}\oarg{\printglossopt[Index]{title},\printglossopt[false]{target}}
\cmd{end}\marg{document}
\end{codeenv}

The \csopt[\encap{hyperbf},\encap{hyperemph}]{primary-location-formats} setting
in the above indicates that locations encapsulated with \ics{hyperbf} and
\ics{hyperemph} are primary records. In this case, the bold format is used to
indicate the primary location in the main document text and the emphasized
format is used to indicate the location in the main glossary.

The primary records are removed from the \field{location} field
due to the \csopt[remove]{save-primary-locations} setting. This can
lead to a ragged location list. The option 
\csopt[default format]{save-primary-locations} can allow the primary
location to be absorbed into a range.

The main glossary records are added through the category-independent post-name
hook with \cs{glsadd}. This won't be implemented until the entries are actually
defined as the page number can't be determined until the glossary can be
displayed. This means that the document build requires an extra \bibgls\ and
\LaTeX\ run:
\begin{verbatim}
pdflatex test
bib2gls --group test
pdflatex test
bib2gls --group test
pdflatex test
\end{verbatim}

For consistency, I've used \gls{glsxtrnewglslike} to provide commands
used to indicate a primary reference in the text. This means that if
I decide to change the optional arguments used for primary
references I only need to edit one line. For example, I might want to 
change the default counter:
\begin{codeenv}
\gls{glsxtrnewglslike}\oarg{\glsopt[\encap{hyperbf}]{format},\glsopt[\counter{chapter}]{counter}}\marg{}\marg{\cmd{primary}}\marg{\cmd{primarypl}}\marg{\cmd{Primary}}\marg{\cmd{Primarypl}}
\end{codeenv}

Here's another example that only has one primary format (\encap{hyperrm}) 
that's indexed through the use of \ics{GlsXtrAutoAddOnFormat}, which sets
up a hook that automatically inserts:
\begin{codeenv*}
\ics{glsadd}\oarg{\glsaddopt[\counter{chapter}]{counter},\glsaddopt[\encap{hyperrm}]{format}}\margm{label}
\end{codeenv*}
on each instance of \code{\cs{gls}\oarg{\glsopt[primaryfmt]{format}}\margm{label}} (or
similar).  This means that the entry is indexed twice when this particular
format is used: first with the \encap{hyperrm} format and \counter{chapter} 
counter (from the \cs{glsadd} command in the hook), and then with the
\code{primaryfmt} format and the default counter (as per normal behaviour):
\begin{codeenv}
\cmd{documentclass}\marg{report}
\strut
\cmd{usepackage}\oarg{colorlinks}\marg{hyperref}
\cmd{usepackage}\oarg{
 \styopt[nameref]{record},
 \styopt[dot]{postpunc},
 \styopt{nostyles},
 \styopt[tree,bookindex]{stylemods},
 \styopt[bookindex]{style}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[topics]{src},
  \csopt[hyperrm]{primary-location-formats},
  \csopt[remove]{save-primary-locations},
  \csopt[false]{save-loclist}
}
\strut
\cmd{newcommand}\marg{\cmd{primaryfmt}}\oarg{1}\marg{\cs{hyperbf}\marg{\idx{param}1}}
\strut
\ics{GlsXtrAutoAddOnFormat}\marg{primaryfmt}\marg{\glsaddopt[\counter{chapter}]{counter},\glsaddopt[\encap{hyperrm}]{format}}
\strut
\gls{glsxtrnewglslike}\oarg{\glsaddopt[primaryfmt]{format}}\marg{}\marg{\cmd{primary}}\marg{\cmd{primarypl}}\marg{\cmd{Primary}}\marg{\cmd{Primarypl}}
\strut
\cmd{begin}\marg{document}
\cmd{chapter}\marg{Sample}
\cmd{Primary}\marg{waterfowl}, \cs{gls}\marg{bird} and \cs{gls}\marg{zebra}.
\strut
\cmd{chapter}\marg{Another Sample}
\cs{Gls}\marg{waterfowl}, \cmd{primary}\marg{bird} and \cs{gls}\marg{zebra}.
\strut
\cmd{chapter}\marg{Yet Another Sample}
\cs{Gls}\marg{waterfowl}, \cs{gls}\marg{bird} and \cmd{primary}\marg{zebra}.
\strut
\cmd{chapter}\marg{Yet Another Sample Again}
\cs{Gls}\marg{waterfowl}, \cs{gls}\marg{bird}, \cmd{primarypl}\marg{parrot} and \cs{gls}\marg{zebra}.
\strut
\ics{printunsrtglossary*}\oarg{\printglossopt[\glostyle{tree}]{style},\printglossopt[Summary]{title}}\marg{\comment{}
 \cmd{renewcommand}*\marg{\ics{glsextrapostnamehook}}[1]\marg{\ics{glsadd}\oarg{\glsaddopt[\encap{hyperemph}]{format}}\marg{\idx{param}1}}\comment{}
 \cmd{renewcommand}\marg{\ics{GlsXtrLocationField}}\marg{\field{primarylocations}}\comment{}
}
\strut
\cs{printunsrtglossary}\oarg{\printglossopt[Index]{title},\printglossopt[false]{target}}
\cmd{end}\marg{document}
\end{codeenv}

Note that in this case, from \bibgls' point of view, the primary format is
\encap{hyperrm} not \code{primaryfmt}. This picks out the records created with
the automated \ics{glsadd}, which have the counter set to \counter{chapter}.
The first glossary (with the title \qt{Summary}) switches the location field to
\field{primarylocations} so that only the primary records are listed. Since
\styopt[nameref]{record} has been used this means that the chapter title is
shown rather than the chapter number.

The second glossary (\qt{Index}) shows the location lists that only
have the \counter{page} counter (because the automated \cs{glsadd}
records with the \counter{chapter} counter have been removed because
they were identified as primary records). These just show the page
number as that's the default display with \styopt[nameref]{record} 
for records with the \counter{page} counter.

An alternative to \ics{GlsXtrAutoAddOnFormat} would be to simply
define the custom commands as follows:
\begin{codeenv}
\cmd{newcommand}\marg{\cmd{primary}}[2][]\marg{\comment{}
 \cs{glsadd}\oarg{\glsaddopt[\counter{chapter}]{counter},\glsaddopt[\encap{hyperrm}]{format}}\marg{\idx{param}2}\comment{}
 \cs{gls}\oarg{\glsopt[primaryfmt]{format},\idx{param}1}\marg{\idx{param}2}\comment{}
}
\cmd{newcommand}\marg{\cmd{primarypl}}[2][]\marg{\comment{}
 \cs{glsadd}\oarg{\glsaddopt[\counter{chapter}]{counter},\glsaddopt[\encap{hyperrm}]{format}}\marg{\idx{param}2}\comment{}
 \cs{glspl}\oarg{\glsopt[primaryfmt]{format},\idx{param}1}\marg{\idx{param}2}\comment{}
}
\cmd{newcommand}\marg{\cmd{Primary}}[2][]\marg{\comment{}
 \cs{glsadd}\oarg{\glsaddopt[\counter{chapter}]{counter},\glsaddopt[\encap{hyperrm}]{format}}\marg{\idx{param}2}\comment{}
 \cs{Gls}\oarg{\glsopt[primaryfmt]{format},\idx{param}1}\marg{\idx{param}2}\comment{}
}
\cmd{newcommand}\marg{\cmd{Primarypl}}[2][]\marg{\comment{}
 \cs{glsadd}\oarg{\glsaddopt[\counter{chapter}]{counter},\glsaddopt[\encap{hyperrm}]{format}}\marg{\idx{param}2}\comment{}
 \cs{Glspl}\oarg{\glsopt[primaryfmt]{format},\idx{param}1}\marg{\idx{param}2}\comment{}
}
\end{codeenv}
This is more useful if you want to simply omit the
\code{\glsopt[primaryfmt]{format}} option (just remove it from the above four 
definitions), which makes it easier to merge the locations into
ranges in the index.

\optsection{min-loc-range}

By default, three or more consecutive locations \meta{loc-1},
\meta{loc-2}, \ldots, \meta{loc-n} are compressed into
the range \code{\meta{loc-1}\ics{delimR} \meta{loc-n}}. Otherwise
the locations are separated by \gls!{bibglsdelimN} or \gls!{bibglslastDelimN}.
As mentioned above, these aren't merged with explicit range formations.

You can change this with the \csopt{min-loc-range} setting where
\meta{value} is either \optfmt{none} (don't form ranges) or an
integer greater than one indicating how many consecutive locations should be
converted into a range.

\bibgls\ determines if one location
\code{\margm{prefix-2}\margm{counter-2}\margm{format-2}\margm{location-2}}
is one unit more than another location
\code{\margm{prefix-1}\margm{counter-1}\margm{format-1}\margm{location-1} }
according to the following:
\begin{enumerate}
\item\label{itm:pre} If \meta{prefix-1} is not equal to \meta{prefix-2} or
\meta{counter-1} is not equal to \meta{counter-2} or \meta{format-1}
is not equal to \meta{format-2}, then the locations aren't
considered consecutive.
\item\label{itm:emptyloc} If either \meta{location-1} or \meta{location-2} are empty,
then the locations aren't considered consecutive.
\item\label{itm.csmatch} If both \meta{location-1} and \meta{location-2} match the
pattern (line break for clarity only)\footnote{The Java class \code{\csfmt{p}\marg{javaDigit}}
used in the \idx{regex} will match any digits in the 
Unicode \idx{numberdecimaldigit} category not just the digits in the Basic Latin set. Similarly \code{\csfmt{p}\marg{javaAlphabetic}} will also match alphabetic
characters outside the Basic Latin set.}
\begin{verbatim}
(.*?)(?:\\protect\s*)?(\\[\p{javaAlphabetic}@]+)\s*\{([\p{javaDigit}
\p{javaAlphabetic}]+)\}
\end{verbatim}
then:
  \begin{itemize}
  \item if the control sequence matched by group 2 isn't the same for
  both locations, the locations aren't considered consecutive;
  \item if the argument of the control sequence (group 3) is the same for
  both locations, then the test is retried with \meta{location-1}
set to group 1 of the first pattern match and \meta{location-2}
set to group 1 of the second pattern match;
  \item otherwise the test is retried with \meta{location-1} set to
group 3 of the first pattern match and \meta{location-2} set to
group 3 of the second pattern match.
  \end{itemize}
\item\label{itm:decmatch} If both \meta{location-1} and \meta{location-2} match the
pattern
\begin{verbatim}
(.*?)([^\p{javaDigit}]?)(\p{javaDigit}+)
\end{verbatim}
then:
  \begin{enumerate}
  \item\label{itm:decgrp3eq} if group 3 of both pattern matches are
equal then:
  \begin{enumerate}
   \item\label{itm:decgrp3nz} if group 3 isn't zero, the locations
aren't considered consecutive;
   \item if the separators (group 2) are different the test is
retried with \meta{location-1} set to the concatenation of the first
two groups \meta{group-1}\meta{group-2} of the first pattern match 
and \meta{location-2} set to the concatenation of the first two groups
\meta{group-1}\meta{group-2} of the second pattern
match;
   \item\label{itm:decgrp3eqsepeq} if the separators (group 2) are the same the test is
retried with \meta{location-1} set to the first
group \meta{group-1} of the first pattern match 
and \meta{location-2} set to the first group
\meta{group-1} of the second pattern match.
  \end{enumerate}
  \item If \meta{group-1} of the first pattern match (of
\meta{location-1}) doesn't equal
\meta{group-1} of the second pattern match (of \meta{location-2})
or \meta{group-2} of the first pattern match (of
\meta{location-1}) doesn't equal
\meta{group-2} of the second pattern match (of \meta{location-2})
then the locations aren't considered consecutive;
   \item\label{itm:decgrp3} If $0 < l_2 - l_1 \leq d $
where $l_2$ is \meta{group 3} of the second pattern match,
$l_1$ is \meta{group 3} of the first pattern match and
$d$ is the value of \optfmt{max-loc-diff} then the locations
are consecutive otherwise they're not consecutive.
  \end{enumerate}
\item\label{itm:rommatch} The next pattern matches for \meta{prefix}\meta{sep}\meta{n}
where \meta{n} is a \idx!{lowercase} Roman numeral, which is converted to
a decimal value and the test is performed in the same way as the 
above \hyperref[itm:decmatch]{decimal test}.
\item\label{itm:Rommatch} The next pattern matches for \meta{prefix}\meta{sep}\meta{n}
where \meta{n} is an \idx!{uppercase} Roman numeral, which is converted to
a decimal value and the test is performed
in the same way as the above \hyperref[itm:decmatch]{decimal test}.
\item\label{itm:alphmatch} The next pattern matches for \meta{prefix}\meta{sep}\meta{c}
where \meta{c} is either a \idx!{lowercase} letter from \code{a} to
\code{z} or an \idx!{uppercase} letter from \code{A} to \code{Z}.
The character is converted to its code point and the test is
performed in the same way as the \hyperref[itm:decmatch]{decimal pattern} above.
\item\label{itm:nomatch} If none of the above, the locations aren't considered
consecutive.
\end{enumerate} 

Examples:
\begin{enumerate}
\item
\begin{codeenv}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{2}
\end{codeenv}
These records are consecutive. The prefix, counter and format are
identical (so the test passes step~\ref{itm:pre}), the locations match
the \hyperref[itm:decmatch]{decimal pattern} and the test in
step~\ref{itm:decgrp3} passes.

\item
\begin{codeenv}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{textbf}}\marg{2}
\end{codeenv}
These records aren't consecutive since the formats are different.

\item
\begin{codeenv}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{A.i}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{A.ii}
\end{codeenv}
These records are consecutive. The prefix, counter and format are
identical (so it passes step~\ref{itm:pre}). The locations match
the \hyperref[itm:rommatch]{lower case Roman numeral pattern}, where
\code{A} is considered a prefix and the dot is consider a
separator. The Roman numerals i and ii are converted to decimal and
the test is retried with the locations set to 1 and 2, respectively.
This now passes the decimal pattern test (step~\ref{itm:decgrp3}).

\item
\begin{codeenv}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{i.A}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{ii.A}
\end{codeenv}
These records aren't consecutive. They match the
\hyperref[itm:alphmatch]{alpha pattern}. The first location is
considered to consist of the prefix \code{i}, the separator
\code{.}\ (dot) and the number given by the character code of A. 
The second location is considered to consist of the prefix
\code{ii}, the separator \code{.}\ (dot) and the number 
given by the character code of A.

The test fails because the numbers are equal and the prefixes are
different.

\item
\begin{codeenv}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1.0}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{2.0}
\end{codeenv}
These records are consecutive. They match the \hyperref[itm:decmatch]{decimal
pattern}, and then step~\ref{itm:decgrp3eq} followed by
step~\ref{itm:decgrp3eqsepeq}. The \code{.0} part is discarded and
the test is retried with the first location set to 1 and the second
location set to 2.

\item
\begin{codeenv}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{1.1}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{2.1}
\end{codeenv}
These records aren't consecutive as the test branches off into
step~\ref{itm:decgrp3nz}.

\item
\begin{codeenv}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{\cmd{@alph}\marg{1}}
\gls{glsxtr@record}\marg{gls.sample}\marg{}\marg{\counter{page}}\marg{\encap{glsnumberformat}}\marg{\cmd{@alph}\marg{2}}
\end{codeenv}
These records are consecutive. The locations match the 
\hyperref[itm.csmatch]{control sequence pattern}. The control
sequences are the same, so the test is retried with the first
location set to 1 and the second location set to 2.

In this example, the location has been written to the file as
\code{\csfmt{@alph}\margm{number}} instead of fully expanding according to the
normal behaviour of \code{\ics{alph}\margm{counter}}.  (Note that
\gls!{glsxtrresourcefile} changes the category code of \code{@} to allow for
internal commands in locations.) This unusual case is for illustrative purposes.
\end{enumerate}

\optsection{max-loc-diff}

This setting is used to determine whether two locations are
considered consecutive.
The value must be an integer greater than or equal to 1.
(The default is \optfmt{1}.)

For two locations, \meta{location-1} and \meta{location-2},
that have numeric values $n_1$ and $n_2$ (and identical prefix,
counter and format), then the sequence \meta{location-1},
\meta{location-2} is considered consecutive if
\[
0 < n_2 - n_1 \leq \text{\meta{max-loc-diff}}
\]
The default value of 1 means that \meta{location-2} immediately
follows \meta{location-1} if $n_2 = n_1+1$.

For example, if \meta{location-1} is \qt{B} and \meta{location-2} 
is \qt{C}, then $n_1 = 66$ and $n_2 = 67$. Since $n_2 = 67 = 66+1=
n_1+1$ then \meta{location-2} immediately follows \meta{location-1}.

This is used in the range formations within the location lists (as described
in the above section).
So, for example, the list \qt{1, 2, 3, 5, 7, 8, 10, 11, 12, 58, 59,
61} becomes
\qt{1--3, 5, 7, 8, 10--12, 58, 59, 61}.

The automatically indexing of commands like \ics{gls} means that
the location lists can become long and ragged. You could
deal with this by switching off the automatic indexing and
only explicitly index pertinent use or you can adjust
the value of \optfmt{max-loc-diff} so that a range can be formed even
if there are one or two gaps in it.
By default, any location ranges that have skipped gaps in this
manner will be followed by \gls!{bibglspassim}. The default
definition of this command is obtained from the resource file.
For English, this is \verb*| passim| (space followed by \qt{passim}).

So with the above set of locations, if \csopt[2]{max-loc-diff} then 
the list becomes \qt{1--12 passim, 58--61 passim} which now highlights that
there are two blocks within the document related to that
term.

\optsection{suffixF}

If set, a range consisting of two consecutive locations 
\meta{loc-1} and \meta{loc-2} will be
displayed in the location list as \meta{loc-1}\meta{value}.

Note that \csopt[\empty]{suffixF} sets the suffix to the
empty string. To remove the suffix formation use
\csopt[none]{suffixF}.

The default is \csopt[none]{suffixF}.

\optsection{suffixFF}

If set, a range consisting of three or more consecutive locations 
\meta{loc-1} and \meta{loc-2} will be
displayed in the location list as \meta{loc-1}\meta{value}.

Note that \csopt[\empty]{suffixFF} sets the suffix to the
empty string. To remove the suffix formation use
\csopt[none]{suffixFF}.

The default is \csopt[none]{suffixFF}.

\optsection{compact-ranges}

The \meta{value} may be an integer \meta{n} or \optfmt{false} (equivalent to
\csopt[0]{compact-ranges}) or \optfmt{true} (equivalent to
\csopt[3]{compact-ranges}). If no \meta{value} is specified,
\optfmt{true} is assumed.

This setting allows location ranges such as 184--189 to appear
more compactly as 184--9. The end location is encapsulated
in the command \gls{bibglscompact}, so the range would actually
become:
\begin{codeenv}
184\ics{delimR}\gls{bibglscompact}\marg{digit}\marg{18}\marg{9}
\end{codeenv}
If the location is in the form \code{\meta{cs}\margm{loc}}
(where \meta{cs} is a command)
then \gls{bibglscompact} will be inside the argument.
For example, if the range would normally be:
\begin{codeenv}
\cmd{custom}\marg{184}\ics{delimR}\cmd{custom}\marg{189}
\end{codeenv}
then it would become:
\begin{codeenv}
\cmd{custom}\marg{184}\ics{delimR}\cmd{custom}\marg{\gls{bibglscompact}\marg{digit}\marg{18}\marg{9}}
\end{codeenv}
The numerical value given in \csopt[\meta{n}]{compact-ranges}
indicates that compaction should only occur if the actual location
consists of at least \meta{n} characters, for $\meta{n} \geq 2$. 
Any value of \meta{n} less than 2 will switch off compaction.

For example, \code{189} consists of 3 characters, so it will be
compacted with \csopt[3]{compact-ranges} but not with
\csopt[4]{compact-ranges}. Whereas \code{\csfmt{custom}\marg{89}}
would only be compacted with \csopt[2]{compact-ranges}
because \code{89} only consists of 2 characters.

The compaction isn't limited to decimal digits but it will only
occur if both the start and end location have the same number of
characters. For example, xvi--xviii can't be compacted because
the start consists of three characters and the end consists
of five characters, whereas xxv--xxx can be compacted to xxv--x,
which may look a little strange. In this case, you may want to
consider changing the definition of \cs{bibglscompact} so that
it only performs the compaction for digits.

\optsection{see}

If an entry has a \field{see} field, this can be placed before or
after the location list, or completely omitted (but the value will
still be available in the \field{see} field for use with
\ics{glsxtrusesee}). The required \meta{value} must be one of:
\begin{itemize}
\item \optfmt{omit}: omit the see reference from the location
list.
\item \optfmt{before}: place the see reference before the location
list.
\item \optfmt{after}: place the see reference after the location
list (default).
\end{itemize}

The separator between the location list and the cross-reference is
provided by \gls!{bibglsseesep}. This separator is omitted if the
location list is empty. The cross-reference is written to the
\field{location} field using \code{\gls{bibglsusesee}\margm{label}}.

\optsection{seealso}

This is like \csopt{see} but governs the location of the cross-references
provided by the \field{seealso} field. You need at least v1.16
of \sty{glossaries-extra} for this option. The values are the
same as for \csopt{see} but the separator is given by
\gls!{bibglsseealsosep}. The cross-reference is written to the
\field{location} field using \code{\gls{bibglsuseseealso}\margm{label}}.

\optsection{alias}

This is like \csopt{alias} but governs the location of the cross-references
provided by the \field{alias} field. The separator is given by
\gls!{bibglsaliassep}. The cross-reference is written to the
\field{location} field using \code{\gls{bibglsusealias}\margm{label}}.

\optsection{alias-loc}

If an entry has an \field{alias} field, the location list
may be retained or omitted or transferred to the target entry.
The required \meta{value} must be one of:
\begin{itemize}
\item\optfmt{keep}: keep the location list;
\item\optfmt{transfer}: transfer the location list;
\item\optfmt{omit}: omit the location list.
\end{itemize}
The default setting is \csopt[transfer]{alias-loc}.
In all cases, the target entry will be added to the \field{see}
field of the entry with the \field{alias} field, unless it
already has a \field{see} field (in which case the \field{see} value
is left unchanged).

Note that with \csopt[transfer]{alias-loc}, both the aliased
entry and the target entry must be in the same resource set.
(That is, both entries have been selected by the same instance of
\gls!{glsxtrresourcefile}.) If you have \sty{glossaries-extra} version~1.12, 
you may need to redefine \ics{glsxtrsetaliasnoindex} to do
nothing if the location lists aren't showing correctly
with aliased entries. (This was corrected in version~1.13.)

\optsection{loc-prefix}

The \csopt{loc-prefix} setting indicates that the location lists
should begin with \code{\gls!{bibglslocprefix}\margm{n}}. The \meta{value} may
be one of the following:
\begin{itemize}
\item \optfmt{false}: don't insert \code{\gls{bibglslocprefix}\margm{n}} at the
start of the location lists (default).

\item \optfmt{\margm{prefix-1},\margm{prefix-2},\ldots,\margm{prefix-n}}:
insert \code{\gls{bibglslocprefix}\margm{n}} (where \meta{n} is the number of
locations in the list) at the start of each location list and the
definition of \gls{bibglslocprefix} will be appended to the glossary
preamble providing an \ics{ifcase} condition:
\begin{codeenv}
\cmd{providecommand}\marg{\gls{bibglslocprefix}}[1]\marg{\comment{}
  \ics{ifcase}\idx{param}1
  \cmd{or} \meta{prefix-1}\gls{bibglspostlocprefix}
  \cmd{or} \meta{prefix-2}\gls{bibglspostlocprefix}
  \ldots
  \cmd{else} \meta{prefix-n}\gls{bibglspostlocprefix}
  \cmd{fi}
}
\end{codeenv}

\item \optfmt{comma}: equivalent to \csopt[\marg{,~}]{loc-prefix} 
but avoids confusion with the list syntax. That is, the prefix is a comma
followed by a space for non-empty locations.

\item \optfmt{list}: equivalent to \csopt[\ics{pagelistname} ]{loc-prefix}.

\item \optfmt{true}: equivalent to 
\csopt[\gls{bibglspagename},\gls{bibglspagesname}]{loc-prefix},
where the definitions of \gls{bibglspagename} and
\gls{bibglspagesname} are obtained from the \idx{tag.page} and
\idx{tag.pages} entries in \bibgls's \langxml.
This setting works best if the document's language matches the 
language file. However, you can redefine these commands within
the document's language hooks or in the glossary preamble.

\end{itemize}

If \meta{value} is omitted, \optfmt{true} is assumed. Take care not
to mix different values of \csopt{loc-prefix} for entries for the
same \csopt{type} setting. It's okay to mix \csopt[false]{loc-prefix} with 
another value, but don't mix non-\optfmt{false} values. 
See the description of \gls!{bibglslocprefix} for
further details.

For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[main]{type},\csopt[entries1]{src},\csopt[false]{loc-prefix}}
\gls{GlsXtrLoadResources}\oarg{\csopt[main]{type},\csopt[entries2]{src},\csopt{loc-prefix}}
\gls{GlsXtrLoadResources}\oarg{\csopt[symbols]{type},\csopt[entries3]{src},\csopt[p.,pp.]{loc-prefix}}
\end{codeenv}
This works since the conflicting \csopt[p.,pp.]{loc-prefix} and
\csopt[true]{loc-prefix} are in different glossaries (assigned through the
\csopt{type} key). The entries fetched from \filefmt{entries1.bib}
won't have a location prefix. The entries fetched from
\filefmt{entries2.bib} will have the location prefix obtained from
the language resource file. The entries fetched from
\filefmt{entries3.bib} will have the location prefix \qt{p.}\ or
\qt{pp.} (Note that using the \csopt{type} option isn't the same as
setting the \field{type} field for each entry in the \ext{bib}
file.)

If the \csopt{type} option isn't used:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries1]{src},\csopt[false]{loc-prefix}}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries2]{src},\csopt{loc-prefix}}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries3]{src},\csopt[p.,pp.]{loc-prefix}}
\end{codeenv}
then \csopt[true]{loc-prefix} takes precedence over
\csopt[p.,pp.]{loc-prefix} (since it was used first). The entries fetched from
\filefmt{entries1.bib} still won't have a location prefix, but the entries
fetched from both \filefmt{entries2.bib} and \filefmt{entries3.bib}
have the location prefixes obtained from the language resource file.

\optsection{loc-suffix}

This is similar to \csopt{loc-prefix} but there are some subtle
differences. In this case \meta{value} may either be the keyword 
\optfmt{false} (in which case the location suffix is omitted)
or a comma-separated list
\optfmt{\meta{suffix-0}\dcomma\meta{suffix-1}\dcomma\ldots\dcomma\meta{suffix-n}}
where \meta{suffix-0} is the suffix to use when the location list
only has a cross-reference with no locations, \meta{suffix-1} is the suffix to use
when the location list has one location (optionally with a
cross-reference), and so on. The final \meta{suffix-n} in the list
is the suffix when the location list has \meta{n} or more locations
(optionally with a cross-reference).

This option will append \code{\gls!{bibglslocsuffix}\margm{n}} to location
lists that either have a cross-reference or have at least one location.
Unlike \gls{bibglslocprefix}, this command isn't used when the
location list is completely empty. Also, unlike
\gls{bibglslocprefix}, this suffix command doesn't have an equivalent
to \gls{bibglspostlocprefix}.

If \meta{value} omitted, \csopt[\ics{at}\idx{periodchar}]{loc-suffix} 
is assumed.  The default is \csopt[false]{loc-suffix}.

As with \csopt{loc-prefix}, take care not to mix different values
of \csopt{loc-suffix} for entries in the same glossary type.

\optsection{loc-counters}

Commands like \ics{gls} allow you to select a different 
counter to use for the location for that specific instance
(overriding the default counter for the entry's glossary type).
This is done with the \glsopt{counter} option. For example,
consider the following document:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{colorlinks}\marg{hyperref}
\cmd{usepackage}\oarg{\styopt{record},\styopt[tree]{style}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src}\comment{data in entries.bib}
}
\strut
\cmd{begin}\marg{document}
\strut
\cs{gls}\marg{pi}.
\cmd{begin}\marg{equation}
\cs{gls}\oarg{\glsopt[\counter{equation}]{counter}}\marg{pi}
\cmd{end}\marg{equation}
\cmd{begin}\marg{equation}
\cs{gls}\oarg{\glsopt[\counter{equation}]{counter}}\marg{pi}
\cmd{end}\marg{equation}
\strut
\cmd{newpage}
\cmd{begin}\marg{equation}
\cs{gls}\oarg{\glsopt[\counter{equation}]{counter}}\marg{pi}
\cmd{end}\marg{equation}
\strut
\cmd{newpage}
\cs{gls}\marg{pi}.
\strut
\cmd{newpage}
\cs{gls}\marg{pi}.
\strut
\cmd{newpage}
\cs{gls}\marg{pi}.
\strut
\cmd{newpage}
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
This results in the location list \qt{1, 1--3, 3--5}. This
looks a little odd and it may seem as though the range formation
hasn't worked, but the locations are actually: page~1, equation~1,
equation~2, equation~3, page~3, page~4 and page~5. Ranges can't
be formed across different counters.

The \csopt[\meta{list}]{loc-counters} option instructs \bibgls\ 
to group the locations according to the counters given in
the comma-separated \meta{list}. If a location has a counter
that's not listed in \meta{list}, then the location is discarded.

For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[\counter{equation},\counter{page}]{loc-counters},\comment{group locations by counter}
  \csopt[entries]{src}\comment{data in entries.bib}
}
\end{codeenv}
This will first list the locations for the \counter{equation}
counter and then the locations for the \counter{page} counter.
Each group of locations is encapsulated within the command
\gls!{bibglslocationgroup}\margm{n}\margm{counter}\margm{locations}.
The groups are separated by \gls!{bibglslocationgroupsep}.

The \meta{list} value must be non-empty.  Use
\csopt[as-use]{loc-counters} to restore the default behaviour, where
the locations are listed in the document order of use, or
\csopt[false]{save-locations} to omit the location lists.  Note that
you can't form counter groups from
\hyperref[sec:supplementalopts]{supplemental location lists}.

\optsection{save-index-counter}

\emph{This option requires at least version 1.29 of
\isty{glossaries-extra}.} The \meta{value} may be one of:
\begin{itemize}
\item \optfmt{false}: don't create the \field{indexcounter} field
(default);
\item \optfmt{true}: create the \field{indexcounter} field with the
value set to the first \counter{wrglossary} location;
\item \meta{encap}: create the \field{indexcounter} field with the
value set to the first \counter{wrglossary} location
where the \glsopt{format} is \meta{encap}.
\end{itemize}
This setting will have no effect if the \styopt{indexcounter} package
option hasn't been used. In the case where the \meta{value} is
\meta{encap}, make sure that this format takes priority in the
location precedence rules (\longarg{map-format}). If the location
with that \meta{encap} format value is discarded then it can't be
saved.

The \styopt{indexcounter} package option
(\sty{glossaries-extra} v1.29+) creates a new counter called
\counter{wrglossary} that's incremented every time a term is
indexed (recorded), except for cross-references such as \cs{glssee}.
The increment is performed using \ics{refstepcounter} and is
followed by \code{\ics{label}\marg{wrglossary.\meta{n}}} where
\meta{n} is the value of the \counter{wrglossary} counter.
This option is intended for use with the \sty{hyperref} package to
allow locations to link back to the particular part of the page where
the term was referenced rather than to the top of the page.

The \styopt{indexcounter} package option also automatically
implements the option \styopt[wrglossary]{counter}, which means that
each instance of \code{\ics{gls}\margm{id}} writes the label
information to the \iext{aux} file:
\begin{codeenv}
\cmd{newlabel}\marg{wrglossary.\meta{n}}\marg{\margm{n}\margm{page}\marg{}\marg{wrglossary.\meta{n}}\marg{}}
\end{codeenv}
(where \meta{page} is the page number) followed by the record:
\begin{codeenv}
\gls{glsxtr@record}\margm{id}\marg{}\marg{wrglossary}\marg{\encap{glsnumberformat}}\margm{n}
\end{codeenv}
The location here is actually the value of the \counter{wrglossary}
counter not the page number, but \bibgls\ can pick up the
corresponding \meta{page} from the \csfmt{newlabel} command. It then
replaces the record's location \meta{n} with:
\nosecdef{glsxtr@wrglossarylocation}
(but it only does this for records that have the \counter{wrglossary}
counter).

The \sty{glossaries-extra} package (v1.29+) adjusts the definition
of \ics{glshypernumber} (which is internally used by
\cs{glsnumberformat}, \cs{hyperbf} etc when \sty{hyperref} has been
loaded) so that if the counter is \counter{wrglossary} then
\ics{pageref} is used instead of \ics{hyperlink}. This means that the
page number is displayed in the location list but it links back to
the place where the corresponding \cs{label} occurred.

This method works partially with \idx!{makeindex} and \idx!{xindy}
but from their point of view the location is the value of the
\counter{wrglossary} counter, which interferes with their ability to
merge duplicate page numbers and form ranges. Since \bibgls\ is
designed specifically to work with \sty{glossaries-extra}, it's
aware of this special counter and will merge and collate the
locations according to the corresponding page number instead.

With the default \longarg{merge-wrglossary-records} switch, if a
term has multiple \counter{wrglossary} records for a given page they
will be merged. The reference link will be the dominant record for
that page.

The \csopt{save-index-counter} option allows you to save the first
of the \counter{wrglossary} locations for a given entry or the first
instance of a specific format of the \counter{wrglossary} locations
for a given entry. This location is stored in the
\field{indexcounter} internal field using:
\begin{codeenv}
\ics{GlsXtrSetField}\margm{id}\marg{\field{indexcounter}}\marg{\gls{glsxtr@wrglossarylocation}\margm{n}\margm{page}}
\end{codeenv}
Since \gls{glsxtr@wrglossarylocation} simply expands to its first
argument, the corresponding label can be obtained with:
\begin{codeenv}
wrglossary.\gls{glsxtr@wrglossarylocation}\margm{n}\margm{page}
\end{codeenv}
For convenience, \sty{glossaries-extra-bib2gls} provides:
\nosecdef{GlsXtrIndexCounterLink}
which will do:
\begin{codeenv*}
\ics{cs.hyperref.int}\oarg{wrglossary.\meta{value}}\margm{text}
\end{codeenv*}
where \meta{value} is the value of the \field{indexcounter} field if
it has been set. If the \field{indexcounter} field hasn't been set
(or if \sty{hyperref} hasn't been loaded) then just \meta{text} is
done.

This provides a convenient way of encapsulating the \field{name} in
the glossary so that it links back to the first \counter{wrglossary}
entry or the first \glsopt[\meta{encap}]{format} \counter{wrglossary}
entry. This encapsulation can be done by providing a new glossary
style or more simply by redefining \ics{glsnamefont}:
\begin{codeenv}
\cmd{renewcommand}\marg{\cs{glsnamefont}}[1]\marg{\comment{}
 \gls{GlsXtrIndexCounterLink}\marg{\idx{param}1}\marg{\cs{glscurrententrylabel}}}
\end{codeenv}
Here's a complete example:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\marg{lipsum}\comment{dummy filler text}
\cmd{usepackage}\oarg{colorlinks}\marg{hyperref}
\cmd{usepackage}\oarg{\styopt{record},\styopt{indexcounter}}\marg{glossaries-extra}
\strut
\cmd{newcommand}\marg{\cmd{primary}}[1]\marg{\cs{hyperbf}\marg{\idx{param}1}}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries]{src},\comment{terms defined in entries.bib}
 \csopt[primary]{save-index-counter}
}
\strut
\cmd{renewcommand}\marg{\cs{glsnamefont}}[1]\marg{\comment{}
 \gls{GlsXtrIndexCounterLink}\marg{\idx{param}1}\marg{\cs{glscurrententrylabel}}}
\strut
\cmd{begin}\marg{document}
\strut
A \cs{gls}\marg{sample}. \cmd{lipsum}*[1] A \cs{gls}\marg{duck}.
\strut
An equation:
\cmd{begin}\marg{equation}
\cs{gls}\oarg{\glsopt[\counter{equation}]{counter}}\marg{pi}
\cmd{end}\marg{equation}
\strut
\cmd{lipsum}[2]
\strut
Another \cs{gls}\oarg{\glsopt[primary]{format}}\marg{sample}. \cmd{lipsum}*[3] Another
\cs{gls}\marg{duck}.
\strut
\cs{gls}\marg{pi}. \cmd{lipsum}[4]
\strut
A \cs{gls}\marg{sample}. \cmd{lipsum}*[5] A \cs{gls}\marg{duck} and
\cs{gls}\oarg{\glsopt[primary]{format}}\marg{pi}.
\strut
\cmd{lipsum}*[6] A \cs{gls}\oarg{\glsopt[primary]{format}}\marg{duck}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
Note that the \glsopt[\counter{equation}]{counter} entry will have its own
independent location. In this example, it's difficult to tell the
difference between 1 (the equation reference) and 1 (the page
reference) in the location list for the \code{pi} entry.

The \glsopt[primary]{format} instances indicate primary references.
They're displayed in bold (since \csfmt{primary} is defined to use
\csfmt{hyperbf}) and these are the locations saved in the
\field{indexcounter} field because that's the \meta{encap}
identified by the \csopt[primary]{save-index-counter} setting.

\section{Supplemental Locations}
\label{sec:supplementalopts}

\emph{These options require at least version 1.14 of
\isty{glossaries-extra}.} If you require locations from multiple
external sources, then you need at least version 1.36 of
\isty{glossaries-extra} (or, more specifically,
\isty{glossaries-extra-bib2gls}, which is automatically loaded
by the \styopt[only]{record} package option).

The \sty{glossaries-extra} package (from v1.14) provides a way of
manually adding locations in supplemental documents through the use
of the \glsaddopt{thevalue} option in the optional argument of
\ics{glsadd}.  Setting values manually is inconvenient and can result
in errors, so \bibgls\ provides a way of doing this automatically.
Both the main document and the supplementary document need to use
the \styopt{record} option. The entries provided in the \csopt{src}
set must have the same labels as those used in the supplementary
document. (The simplest way to achieve this is to ensure that both
documents use the same \ext{bib} files and the same prefixes.)

For example, suppose the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{entry}\marg{sample,
  \field{name}=\marg{sample},
  \field{description}=\qtdelim{an example entry}
}
\strut
\atentry{abbreviation}\marg{html,
  \field{short}=\qtdelim{html},
  \field{long}=\marg{hypertext markup language}
}
\strut
\atentry{abbreviation}\marg{ssi,
  \field{short}=\qtdelim{ssi},
  \field{long}=\qtdelim{server-side includes}
}
\strut
\atentry{index}\marg{goose,\field{plural}=\qtdelim{geese}}
\end{codeenv}
Now suppose the supplementary document is contained in the file
\filefmt{suppl.tex}:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{colorlinks}\marg{hyperref}
\cmd{usepackage}\oarg{\styopt{record},\styopt[section]{counter}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\strut
\cmd{renewcommand}\marg{\cmd{thesection}}\marg{S\cmd{arabic}\marg{section}}
\cmd{renewcommand}\marg{\cmd{theHsection}}\marg{\cmd{thepart}.\cmd{thesection}}
\strut
\cmd{begin}\marg{document}
\cmd{part}\marg{Sample Part}
\cmd{section}\marg{Sample Section}
\cs{gls}\marg{goose}. \cs{gls}\marg{sample}.
\strut
\cmd{part}\marg{Another Part}
\cmd{section}\marg{Another Section}
\cs{gls}\marg{html}.
\cs{gls}\marg{ssi}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
This uses the \counter{section} counter for the locations and has a
prefix (\code{\csfmt{thepart}.})\ for the section hyperlinks.

Now let's suppose I have another document called \filefmt{main.tex}
that uses the \code{sample} entry, but also needs to include the
location (S1) from the supplementary document. The manual approached
offered by \styfmt{glossaries-extra} is quite cumbersome and requires
setting the \catattr{externallocation} attribute and using
\ics{glsadd} with \glsaddopt[S1]{thevalue}, \glsaddopt[I.S1]{theHvalue}
and \glsaddopt[\encap{glsxtrsupphypernumber}]{format}.

This can be simplified with \bibgls\ by using the
\optfmt{supplemental-locations} option, described below. 

Version 1.36 of \isty{glossaries-extra-bib2gls} introduces some
special location formatting commands that don't use the
\catattr{externallocation} attribute, but instead have an extra
argument that indicates the external reference. The additional
argument means that it can't be used by the \glsaddopt{format}
key, but with \bibgls\ you don't use \cs{glsadd} to record
the external locations. Instead it obtains the records from
the corresponding supplementary \ext{aux} file, and adjusts the
location encapsulator as appropriate.

If \bibgls\ detects an older version of \sty{glossaries-extra},
it will only allow one external supplemental source, and
will set the \catattr{externallocation} attribute and use
the \encap{glsxtrsupphypernumber} format. Otherwise \bibgls\
will allow multiple sources and use the newer method.

\optsection{supplemental-locations}

The value should be the base name (without the extension) of the
supplementary document (\optfmt{suppl} in the above example).
If you have at least version 1.36 of \sty{glossaries-extra},
the value may be a comma-separated list of base names (without the
extensions) of the supplementary documents. If an older version is
detected, \bibgls\ will issue a warning and only accept the first
element of the list.

For example:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{colorlinks}\marg{hyperref}
\cmd{usepackage}\oarg{\styopt{record}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[suppl]{supplemental-locations},\comment{fetch records from suppl.aux}
 \csopt[entries]{src}}
\strut
\cmd{begin}\marg{document}
\cs{Gls}\marg{sample} document.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
The location list for \code{sample} will now be \qt{1, S1} (page~1
from the main document and S1 from the supplementary document).

With \sty{glossaries-extra} v1.36+, a regular location from the
supplementary document will be encapsulated with:
\nosecdef{glsxtrdisplaysupploc}
By default, this simply creates an external hyperlink to the
supplementary document with the location as the hyperlink text.
The hyperlink is created using \meta{src} as the target path
with the fragment part (anchor) formed from the prefix and 
location. The \catattr{externallocation} attribute is not set in
this case. The actual formatting is done via:
\nosecdef{glsxtrmultisupplocation}
which ignores the \meta{format} argument by default. Its
definition is simply:
\begin{codeenv}
\cmd{newcommand}*\marg{\gls{glsxtrmultisupplocation}}[3]\marg{\comment{}
 \marg{\comment{scope required to localise changes}
   \cmd{def}\cs{glsxtrsupplocationurl}\marg{\idx{param}2}\comment{}
   \cs{glshypernumber}\marg{\idx{param}1}\comment{}
 }\comment{}
}
\end{codeenv}
This locally sets the command \ics{glsxtrsupplocationurl}, which is checked 
by \ics{glshypernumber} to establish an external rather than internal link. 
You can redefine the supplemental location command to retain the original 
\idx{encap} used in the target document:
\begin{codeenv}
\cmd{renewcommand}*\marg{\gls{glsxtrmultisupplocation}}[3]\marg{\comment{}
 \marg{\comment{scope required to localise changes}
   \cmd{def}\cs{glsxtrsupplocationurl}\marg{\idx{param}2}\comment{}
   \ics{csuse}\marg{\idx{param}3}\marg{\idx{param}1}\comment{}
 }\comment{}
}
\end{codeenv}
but remember that if a hyperlink is required, the identified
control sequence name must correspond to a command that
uses \cs{glshyperlink} (such as \cs{hyperbf}), otherwise you
will lose the hyperlink.

With older versions of \sty{glossaries-extra},
the original location format from the supplementary document
will be replaced by \encap{glsxtrsupphypernumber}, which 
again produces an external hyperlink. The \catattr{externallocation}
attribute also needs to be set (this can be done automatically with
\csopt{supplemental-category}) to identify the external document.
The original format can't be accessed.

In both cases, if the document hasn't loaded the \isty{hyperref} package, the
location will simply be displayed without a hyperlink.  Even if both the main
and the supplementary documents have loaded \sty{hyperref}, note that not all
PDF viewers can handle external hyperlinks, and some that can open the external
PDF file may not recognise the destination within that file.

The special \code{nameref} locations (see
\longarg{merge-nameref-on}) are still identified with
\gls!{glsxtrdisplaylocnameref} but the \meta{file} argument will now
be set.

As from \bibgls\ v1.7, any awkward characters in the file path are
replaced with \gls{bibglshrefchar} or (for \idx{non-ASCII} characters when
\sty{fontspec} is loaded) \gls{bibglshrefunicode}. Both commands
take two arguments: the hexadecimal character code and the actual
character. In the case of \gls{bibglshrefchar}, the second
argument is ignored, and the first is preceded by a literal percent
character, so \filefmt{file name.pdf} will be converted to:
\begin{codeenv}
file\gls{bibglshrefchar}\marg{20}\marg{ }name.pdf
\end{codeenv}
which will expand to \verb|file%20name.pdf|.
In the case of \gls{bibglshrefunicode}, the first argument is
ignored, so \filefmt{skr\'aarnafn.pdf} will be converted to:
\begin{codeenv}
skr\gls{bibglshrefunicode}\marg{E1}\marg{\'a}arnafn.pdf
\end{codeenv}
which will expand to \code{skr\'aarnafn.pdf}.


The supplementary locations lists are encapsulated within
\gls{bibglssupplemental}. With \sty{glossaries-extra} v1.36+,
this command will encapsulate the sub-lists with
\gls{bibglssupplementalsublist}.

So the above example with an old version of \sty{glossaries-extra}
(pre 1.36) will set the supplemental location list (which only consists
of one location) to:
\begin{codeenv}
\gls{bibglssupplemental}
\marg{1}\marg{\ics{setentrycounter}\oarg{I}\marg{\counter{section}}\cs{glsxtrsupphypernumber}\marg{S1}}
\end{codeenv}
and the external target must be supplied through the
\catattr{externallocation} attribute, which can be set 
with the \csopt{supplemental-category} option.

Whereas with at least version 1.36, the list will be:
\begin{codeenv}
\gls{bibglssupplemental}\marg{1}\marg{\gls{bibglssupplementalsublist}\marg{1}\marg{suppl.pdf}
\marg{\gls{glsxtrdisplaysupploc}\marg{I}\marg{\counter{section}}\marg{\encap{glsnumberformat}}\marg{suppl.pdf}\marg{S1}}}
\end{codeenv}

If an entry has both a main location list and a supplementary
location list (such as the \code{sample} entry above), the lists
will be separated by \gls!{bibglssupplementalsep}. The sub-lists
(when supported) are separated by \gls{bibglssupplementalsubsep}.

\optsection{supplemental-selection}

In the above example, only the \code{sample} entry is listed in
the main document, even though the supplementary document also
references the \code{goose}, \code{html} and \code{ssi}
entries. By default, only those entries that are referenced in the
main document will have supplementary locations added (if found in
the supplementary document's \iext{aux} file). You can additionally
include other entries that are referenced in the supplementary
document but not in the main document using
\optfmt{supplemental-selection}. The \meta{value} may be one of the
following:
\begin{itemize}
\item \optfmt{all}: add all the entries in the supplementary
document that have been defined in the \ext{bib} files listed in
\csopt{src} for this resource set in the main document.
\item \optfmt{selected}: only add supplemental locations for entries
that have already been selected by this resource set.
\item \meta{label-1},\ldots,\meta{label-2}: in addition to all those
entries that have already been selected by this resource set, also
add the entries identified in the comma-separated list. If a label
in this list doesn't have a record in the supplementary document's
\ext{aux} file, it will be ignored.
\end{itemize}
Any records in the supplementary \ext{aux} file that aren't defined
by the current resource set (through the \ext{bib} files listed in
\csopt{src}) will be ignored. Entry aliases aren't taken into
account when including supplementary locations.

For example:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{colorlinks}\marg{hyperref}
\cmd{usepackage}\oarg{\styopt{record}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[suppl]{supplemental-locations},
 \csopt[html,ssi]{supplemental-selection},
 \csopt[entries]{src}}
\strut
\cmd{begin}\marg{document}
\cs{Gls}\marg{sample} document.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
This will additionally add the \code{html} and \code{ssi} entries
even though they haven't been used in this document. The
\code{goose} entry used in the supplementary document won't be
included.

\optsection{supplemental-category}

The \field{category} field for entries containing supplemental location
lists may be set using this option. If unset,
\meta{value} defaults to the same as that given by the
\csopt{category} option. The \meta{value} may either be a known
identifier (as per \csopt{category}) or the category label. For example:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{colorlinks}\marg{hyperref}
\cmd{usepackage}\oarg{\styopt{record}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[suppl]{supplemental-locations},
 \csopt[html,ssi]{supplemental-selection},
 \csopt[supplemental]{supplemental-category},
 \csopt[entries]{src}}
\strut
\cmd{begin}\marg{document}
\cs{Gls}\marg{sample} document.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}

\section{Sorting}
\label{sec:sortingopts}

Entries are typically displayed in an ordered list, but the
\styfmt{glossaries-extra} package is versatile enough to be used in
wider contexts than simple terms, symbols or abbreviations. For
example, entries could contain theorems or problems where the
\field{name} supplies the title and the \field{description} provides
a description of the theorem or problem. Another field might then
contain the proof or solution.  Therefore, somewhat unusually for an
indexing application, \bibgls\ also provides the option to shuffle
the entries instead of sorting them.

This section covers the resource options for sorting primary
entries. See \sectionref{sec:dualoptssort} for sorting dual entries
and also \csopt{sort-label-list} for sorting field values that
contain a comma-separated list of entry labels (such as the
\field{see} or \field{seealso} fields).

The sort methods that use a comparison function (that is, all the
sort methods except those listed in \tableref{tab:sortoptionsnosort})
require a sort value for each entry. The function compares these
values to determine the order. By default, this sort value is
obtained from the \field{sort} field but for greater flexibility
it's best to not actually set this field. \bibgls\ has a set of fallbacks
that it uses if a field it needs to access is missing. These
fallbacks depend on the entry type and resource settings.

For example, if a term defined with \atentry{index} doesn't have the
\field{sort} field set then \bibgls\ will use the value given by the
\field{name} field because \field{name} is the fallback field for
\field{sort} for \atentry{index} entries. If the \field{name} field
isn't set either then \bibgls\ will use the fallback for that field.
In the case of \atentry{index} that's the entry's label. If the
\field{sort} field is explicitly set then there's no need to use the
fallback.

If, on the other hand, a term defined with \atentry{symbol} doesn't
have the \field{sort} field set then \bibgls\ will use the value
from the field identified by \csopt{symbol-sort-fallback}, which is
the entry's label by default (not the \field{name} field).

This means that if I don't explicitly set the \field{sort} field for
any entries then I can, for example, sort terms defined with
\atentry{index} by \field{name} and those defined with
\atentry{symbol} by \field{description} with the setting:
\begin{codeenv}
\csopt[description]{symbol-sort-fallback}
\end{codeenv}

If the field used to obtain the sort value is changed
(with \csopt{sort-field}) then the \field{sort} field won't be
queried. This reduces the flexibility of selecting the most
appropriate field for given entry types. For example,
\csopt[name]{sort-field} will force all entries to be sorted by the
\field{name} field, which may not be appropriate for symbols.

\begin{important}
If you choose a field whose value must be a label (such as
\field{parent} or \field{group}) then the sort value will be that label.
\end{important}

You can have \atentry{preamble} definitions that
can be hidden from \bibgls's interpreter. For example,
\exfile{no-interpret-preamble.bib} might contain:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cmd{providecommand}\marg{\gls{sortop}}[2]\marg{\idx{param}1 \idx{param}2}}}
\end{codeenv}
which is loaded using:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[no-interpret-preamble]{src},
 \csopt[false]{interpret-preamble}}
\end{codeenv}
This provides a custom command:
\nosecdef{sortop}
for internal use in the document. (Remember it won't be defined on
the first \LaTeX\ run before the \iext{glstex} file has been
created and so is only used within entry fields.)

Another file, say, \filefmt{interpret-preamble.bib} may provide
a definition for \bibgls:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cmd{providecommand}\marg{\gls{sortop}}[2]\marg{\idx{param}2, \idx{param}1}}}
\end{codeenv}
which can be processed with:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[interpret-preamble]{src}}
\end{codeenv}
to provide \bibgls\ with this definition.
The \filefmt{entries.bib} file could contain:
\begin{codeenv}
\atentry{entry}\marg{caesar,
  \field{name}=\marg{\gls{sortop}\marg{Gaius Julius}\marg{Caesar}},
  \field{first}=\marg{Julius Caesar},
  \field{text}=\marg{Caesar},
  \field{description}=\marg{Roman politician and general}
}
\end{codeenv}
and then be processed with:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\end{codeenv}

The definition provided in \filefmt{interpret-preamble.bib}, which
swaps the two arguments around, is now picked up by \bibgls, so the
sort value becomes \code{Caesar, Gaius Julius}, but this new
definition doesn't affect the document since \LaTeX\ has already
defined \gls{sortop} from the first resource set, so the name will
appear as \qt{Gaius Julius Caesar} in the glossary. (If you have
\cs{renewcommand} rather than \cs{providecommand}, you can prevent
the redefinition occurring in the document with
\csopt[false]{write-preamble}.)

Alternatively both of these \ext{bib} files can be loaded in one
\idx{resourceset}:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[interpret-preamble,entries]{src}}
\end{codeenv}
Another possibility is to provide a custom package that contains the
command definitions for the \bibgls\ interpreter and load it with
\longarg{custom-packages} instead of having the
\filefmt{interpret-preamble.bib} file.

\optsection{sort}

The \csopt{sort} key indicates how primary entries should be sorted.
If the \meta{value} is omitted, \csopt[doc]{sort} is assumed. If the
\csopt{sort} option isn't used then \csopt[doc]{sort} is assumed if the 
document has a language that's been detected by \sty{tracklang}, 
otherwise \csopt[locale]{sort} is assumed.

The \optfmt{\meta{method}\dhyphen reverse} options reverse the
result returned by the corresponding \meta{method} comparator.
However \optfmt{\meta{method}\dhyphen reverse} may not produce a
list that's the exact reverse of the underlying non-reversed
\meta{method} as the hierarchical structure or associated settings
can affect the order.

\begin{table}[p]
\caption{Summary of Available Sort Options: No Sort Field}
\label{tab:sortoptionsnosort}
\centering
\begin{tabular}{ll}
\optfmt{none} or \optfmt{unsrt} & don't sort\tabularnewline
\optfmt{random} & shuffle entries\tabularnewline
\optfmt{use} & order of use\tabularnewline
\optfmt{use-reverse} & reverse order of use\tabularnewline
\optfmt{recordcount}\textsuperscript\textdagger & order of record count\tabularnewline
\optfmt{recordcount-reverse}\textsuperscript\textdagger & reverse order of record count
\end{tabular}
\par
\medskip
\textsuperscript\textdagger Requires \longarg{record-count} switch.
\end{table}

\begin{table}[p]
\caption{Summary of Available Sort Options: Alphabet}
\label{tab:sortoptionsrule}
\centering
\begin{tabular}{ll}
\meta{lang tag} & sort according to this language tag\tabularnewline
\meta{lang tag}\optfmt{-reverse} & reverse sort according to this language tag\tabularnewline
\optfmt{doc} & sort according to the document
language\tabularnewline
\optfmt{doc-reverse} & reverse sort according to the document
language\tabularnewline
\optfmt{locale} & sort according to the default
locale\tabularnewline
\optfmt{locale-reverse} & reverse sort according to the default
locale\tabularnewline
\optfmt{custom} & sort according to \csopt[\meta{custom
rule}]{sort-rule}\tabularnewline
\optfmt{custom-reverse} & reverse sort according to \csopt[\meta{custom rule}]{sort-rule}
\end{tabular}
\end{table}

\begin{table}[p]
\caption{Summary of Available Sort Options: Letter (Non-Locale)}
\label{tab:sortoptionsletter}
\centering
\begin{tabular}{ll}
\optfmt{letter-case} & case-sensitive letter sort\tabularnewline
\optfmt{letter-case-reverse} & reverse case-sensitive letter
sort\tabularnewline
\optfmt{letter-nocase} & case-insensitive letter sort\tabularnewline
\optfmt{letter-nocase-reverse} & reverse case-insensitive letter
sort\tabularnewline
\optfmt{letter-upperlower} & upper-lower letter sort\tabularnewline
\optfmt{letter-upperlower-reverse} & reverse upper-lower letter
sort\tabularnewline
\optfmt{letter-lowerupper} & lower-upper letter sort\tabularnewline
\optfmt{letter-lowerupper-reverse} & reverse lower-upper letter sort
\end{tabular}
\end{table}

\begin{table}[p]
\caption{Summary of Available Sort Options: Letter-Number}
\label{tab:sortoptionsletternumber}
\centering
\begin{tabular}{ll}
\optfmt{letternumber-case} & case-sensitive letter-number
sort\tabularnewline
\optfmt{letternumber-case-reverse} & reverse case-sensitive
letter-number sort\tabularnewline
\optfmt{letternumber-nocase} & case-insensitive letter-number
sort\tabularnewline
\optfmt{letternumber-nocase-reverse} & reverse case-insensitive
letter-number sort\tabularnewline
\optfmt{letternumber-upperlower} & upper-lower letter-number
sort\tabularnewline
\optfmt{letternumber-upperlower-reverse} & reverse upper-lower
letter-number sort\tabularnewline
\optfmt{letternumber-lowerupper} & lower-upper letter-number
sort\tabularnewline
\optfmt{letternumber-lowerupper-reverse} & reverse lower-upper letter-number sort
\end{tabular}
\end{table}

\begin{table}[p]
\caption{Summary of Available Sort Options: Numerical}
\label{tab:sortoptionsnumerical}
\centering
\begin{tabular}{ll}
\optfmt{integer} & integer sort\tabularnewline
\optfmt{integer-reverse} & reverse integer sort\tabularnewline
\optfmt{hex} & hexadecimal sort\tabularnewline
\optfmt{hex-reverse} & reverse hexadecimal sort\tabularnewline
\optfmt{octal} & octal sort\tabularnewline
\optfmt{octal-reverse} & reverse octal sort\tabularnewline
\optfmt{binary} & binary sort\tabularnewline
\optfmt{binary-reverse} & reverse binary sort\tabularnewline
\optfmt{float} & float sort\tabularnewline
\optfmt{float-reverse} & reverse float sort\tabularnewline
\optfmt{double} & double sort\tabularnewline
\optfmt{double-reverse} & reverse double sort\tabularnewline
\optfmt{numeric} & locale-sensitive numeric sort\tabularnewline
\optfmt{numeric-reverse} & reverse locale-sensitive numeric
sort\tabularnewline
\optfmt{currency} & locale-sensitive currency sort\tabularnewline
\optfmt{currency-reverse} & reverse locale-sensitive currency
sort\tabularnewline
\optfmt{percent} & locale-sensitive percent sort\tabularnewline
\optfmt{percent-reverse} & reverse locale-sensitive percent
sort\tabularnewline
\optfmt{numberformat} & locale-sensitive custom numeric
sort\tabularnewline
\optfmt{numberformat-reverse} & reverse locale-sensitive custom numeric sort
\end{tabular}
\end{table}

\begin{table}[p]
\caption{Summary of Available Sort Options: Date-Time}
\label{tab:sortoptionsdatetime}
\centering
\begin{tabular}{ll}
\optfmt{date} & locale-sensitive date sort\tabularnewline
\optfmt{date-reverse} & reverse locale-sensitive date
sort\tabularnewline
\optfmt{datetime} & locale-sensitive date-time sort\tabularnewline
\optfmt{datetime-reverse} & reverse locale-sensitive date-time
sort\tabularnewline
\optfmt{time} & locale-sensitive time sort\tabularnewline
\optfmt{time-reverse} & reverse locale-sensitive time sort
\end{tabular}
\end{table}

\subsubsection{No Sort Field}

Most of the sort methods listed in \tableref{tab:sortoptionsnosort} 
don't actually perform any sorting. This may cause a problem for
hierarchical entries. In some cases this can lead to detached child
entries or an attempt to define a child entry before its parent.
The methods listed in this section all ignore the \csopt{sort-field}
setting and all the various sort fallback settings, except where
noted below.
\begin{itemize}

\item \optfmt{none} (or \optfmt{unsrt}): don't sort the entries.
(The entries will be in the order they were processed when parsing
the data.)

\item \optfmt{random}: shuffles rather than sorts the entries.
This won't work if there are hierarchical entries, so it's best
to use this option with \csopt{flatten}. The seed for the
random generator can be set using \csopt{shuffle} (which
also automatically sets \csopt[random]{sort} and \csopt{flatten}).

\item \optfmt{use}: order of use. This order is determined
by the records written to the \iext{aux} file by the \styopt{record}
package option. Dependencies and cross-references (including those
identified with \ics{glssee}) come after entries with records.

Note that this is different from using the analogous option with
\idx{makeindex} or \idx{xindy}, which does actually sort
numerically, where each entry has an associated number set on the
first use of that term that's used as the sort value.

\item \optfmt{use-reverse}: reverses the order that would be
obtained with \csopt[use]{sort} without reference to hierarchy.

\item \optfmt{recordcount}: order of record count (starting from 0). This order is
determined by the total number of records written to the \iext{aux}
file for each entry. Unlike the above methods, this performs a hierarchical sort.
If letter groups are enabled with \longarg{group}, this method will assign the entries to
the \idx{numbergroup}.

This option requires the \longarg{record-count} switch. Although
that switch makes \bibgls\ write the total record count to the
\ext{glstex} file in the \field{recordcount} internal field (so that it
can be accessed in the document), \bibgls\ doesn't actually have a
field itself that contains the information. So although this option
behaves much like \csopt[integer]{sort} it's not possible to select
a field containing the required value. In the event of two or more
entries having the same record count, the
\csopt{identical-sort-action} option is used to determine the
relative ordering between them.

\item \optfmt{recordcount-reverse}: reverse order of record count
(ending with 0). All the above notes applying to
\optfmt{recordcount} also apply here.

\end{itemize}

Suppose the file \filefmt{entries.bib} contains
definitions of a set of symbols that don't have any intuitive
ordering (for example, they are all pictographs) then there may be
no point in trying to order them, in which case you can do:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[none]{sort}}
\end{codeenv}
Alternatively, you could list them in order of use:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[use]{sort}}
\end{codeenv}
or by frequency of use. For example, starting with entries that
don't have any records followed by the least used entries (a rarely-used symbol 
may be harder to remember and most likely to be looked up in the glossary):
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[recordcount]{sort}}
\end{codeenv}
Or starting with the most used entries:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[recordcount-reverse]{sort}}
\end{codeenv}
It all depends on what's likely to be most useful to the reader.

Consider the following:
\begin{codeenv}
\ics{newglossary*}\marg{frequent}\marg{Most Frequently Used Terms}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[use]{sort},
 \csopt[recordcount-reverse:frequent]{secondary}
}
\cs{newcommand}\marg{\csfmt{filterhook}}[1]\marg{\comment{}
  \ics{GlsXtrIfFieldCmpNum}*\marg{\field{recordcount}}\marg{\idx{hashchar}1}\marg{>}\marg{10}\comment{}
  \marg{}\comment{}
  \marg{\ics{printunsrtglossaryskipentry}}\comment{}
}
\csfmt{begin}\marg{document}
\cs{printunsrtglossary*}\oarg{\printglossopt[false]{target},\printglossopt[frequent]{type}}\marg{\comment{}
  \cs{let}\ics{printunsrtglossaryentryprocesshook}\csfmt{filterhook}
}
\comment{Main body of the document \ldots}
\cs{printunsrtglossary}
\csfmt{end}\marg{document}
\end{codeenv}
This has a summary at the start of the document that only contains
entries that have at least 10 records and is ordered according to
the total number of records (starting with the most frequently used
entry). The main glossary at the end of the document is ordered
according to use and contains all selected entries.

Compare this with the following:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[use]{sort}}
\cs{newcommand}\marg{\csfmt{filterhook}}[1]\marg{\comment{}
  \ics{GlsXtrIfFieldCmpNum}*\marg{\field{recordcount}}\marg{\idx{hashchar}1}\marg{>}\marg{10}\comment{}
  \marg{}\comment{}
  \marg{\ics{printunsrtglossaryskipentry}}\comment{}
}
\csfmt{begin}\marg{document}
\cs{printunsrtglossary*}\oarg{\printglossopt[false]{target},
 \printglossopt[Most Frequently Used Terms]{title}}\marg{\comment{}
  \cs{let}\ics{printunsrtglossaryentryprocesshook}\csfmt{filterhook}
}
\comment{Main body of the document \ldots}
\cs{printunsrtglossary}
\csfmt{end}\marg{document}
\end{codeenv}
This again has a summary at the start of the document that only
contains entries that have at least 10 records but is now ordered
according to use. 

Both examples assume there are no child entries as the filtering can
cause parent entries to be omitted. Both examples require
\longarg{record-count} but only the first example sorts according to
the record count.

\subsubsection{Alphabet}

The sort methods listed in \tableref{tab:sortoptionsrule} are for
alphabets that are defined by a rule. These usually ignore most
punctuation and may ignore modifiers (such as accents).  Use with
\csopt{break-at} to determine whether or not to split at word
boundaries. The collation rules (except for the custom options) are obtained
from the \idx{localeprovider} (see page~\pageref{locale.provider}).

\begin{itemize}
\item \meta{lang tag}: sort according to the rules of the locale
given by the \idx{IETF} language tag \meta{lang tag}.

\item \meta{lang tag}\optfmt{-reverse}: reverse sort according to the rules of the locale
given by the \idx{IETF} language tag \meta{lang tag}.

\item \optfmt{locale}: equivalent to 
\csopt[\meta{lang tag}]{sort} where \meta{lang tag} is obtained
from the \idx{JRE}['s] default locale (which usually matches the
operating system's locale).

\item \optfmt{locale-reverse}: equivalent to 
\csopt[\meta{lang tag}-reverse]{sort} where \meta{lang tag} is obtained
from the \idx{JRE}.

\item \optfmt{doc}: sort the entries according to the document
language. This is equivalent to \csopt[\meta{lang tag}]{sort}
where \meta{lang tag} is the locale associated with the document
language. In the case of a multi-lingual document, \meta{lang tag}
is the locale of the last language resource file to be loaded through
\isty{tracklang}'s interface. It's best to explicitly set the locale
for multi-lingual documents to avoid confusion. If no languages have been tracked, this
option is equivalent to \csopt[locale]{sort}.

\item \optfmt{doc-reverse}: as \optfmt{doc} but in reverse order.

\item \optfmt{custom}: sort the entries according to the 
rule provided by \csopt{sort-rule}.

\item \optfmt{custom-reverse}: reverse sort the entries according to the 
rule provided by \csopt{sort-rule}.
\end{itemize}

Note that \csopt[\meta{locale}]{sort} can provide more detail about the
given locale than \csopt[doc]{sort}, depending on how the document
language has been specified.  For example, with:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\cmd{usepackage}\oarg{ngerman}\marg{babel}
\cmd{usepackage}\oarg{\styopt{record}}\marg{glossaries}
\gls{GlsXtrLoadResources}\oarg{\csopt[german-terms]{src}}
\end{codeenv}
the language tag will be \code{de-1996}, which doesn't have an
associated region, so this is equivalent to using \csopt[de-1996]{sort}. 
Whereas with:
\begin{codeenv}
\cmd{documentclass}\oarg{de-DE-1996}\marg{article}
\cmd{usepackage}\oarg{ngerman}\marg{babel}
\cmd{usepackage}\oarg{\styopt{record}}\marg{glossaries}
\gls{GlsXtrLoadResources}\oarg{\csopt[german-terms]{src}}
\end{codeenv}
the language tag will be \code{de-DE-1996} because \isty{tracklang}
has picked up the locale from the document class options, so this is 
equivalent to using \csopt[de-DE-1996]{sort}. This is
only likely to cause a difference if a language has different
sorting rules according to the region or if the language may be
written in multiple scripts.

If no language package is loaded then the default is \csopt[locale]{sort}
rather than \csopt[doc]{sort}. For example, with:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\cmd{usepackage}\oarg{\styopt{record}}\marg{glossaries}
\gls{GlsXtrLoadResources}\oarg{\csopt[german-terms]{src}}
\end{codeenv}
the language tag will be whatever is the default locale for the \idx{JVM}.
For a user in Germany, this could be \code{de-DE-1996} and for a user
in Austria this could be \code{de-AT-1996}.

A multilingual document will need to have the \csopt{sort} specified
when loading the \idx{resourceset} to ensure the correct language is chosen. 
For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[english-terms]{src},\csopt[en-GB]{sort}}
\gls{GlsXtrLoadResources}\oarg{\csopt[german-terms]{src},\csopt[de-DE-1996]{sort}}
\end{codeenv}

\subsubsection{Letter (Non Locale)}

The sort methods listed in \tableref{tab:sortoptionsletter} 
use letter comparators. These simply compare the
character codes. The \optfmt{-nocase} options first convert the
\field{sort} field to \idx{lowercase} before performing the sort to
provide a case-insensitive comparison. 

Punctuation isn't ignored.
Use \csopt[\meta{lang tag}]{sort} with \csopt[none]{break-at} to
emulate \idx!{xindy}['s] locale letter ordering. The examples below
show the ordering of the list \code{antelope}, \code{bee}, \code{Africa}, 
\code{aardvark} and \code{Brazil}.
\begin{itemize}
\item \optfmt{letter-case}: case-sensitive letter sort.
\Idx{uppercase} and \idx{lowercase} are in separate letter groups.
Example: 

\code{Africa} (letter group \idx{uppercase} \qt{A}), \code{Brazil}
(letter group \idx{uppercase} \qt{B}), \code{aardvark} (letter group 
\idx{lowercase} \qt{a}), \code{antelope} (letter group
\idx{lowercase} \qt{a}), \code{bee} (letter group \idx{lowercase} \qt{b}).

\item \optfmt{letter-case-reverse}: reverse case-sensitive letter sort.
Example:

\code{bee} (letter group \idx{lowercase} \qt{b}), \code{antelope}
(letter group \idx{lowercase} \qt{a}), \code{aardvark} (letter group
\idx{lowercase} \qt{a}) \code{Brazil} (letter group \idx{uppercase}
\qt{B}), \code{Africa} (letter group \idx{uppercase} \qt{A}).

\item \optfmt{letter-nocase}: case-insensitive letter sort. (All
\idx{uppercase} characters will have first been converted to
\idx{lowercase} in the sort value.)
Example:

\code{aardvark} (letter group \qt{A}), \code{Africa}
(letter group \qt{A}), \code{antelope} (letter group \qt{A}), 
\code{bee} (letter group \qt{B}),
\code{Brazil} (letter group \qt{B}).

\item \optfmt{letter-nocase-reverse}: reverse case-insensitive letter sort.
Example:

\code{Brazil} (letter group \qt{B}), \code{bee} (letter group
\qt{B}), \code{antelope} (letter group \qt{A}), \code{Africa}
(letter group \qt{A}), \code{aardvark} (letter group \qt{A}).

\item \optfmt{letter-upperlower}: each character pair is first
compared according to their \idx{lowercase} values. If these are equal,
then they are compared according to case. This puts upper and lower
case in the same letter group but the \idx{uppercase} comes first.
Example:

\code{Africa} (letter group \qt{A}), \code{aardvark} (letter group
\qt{A}), \code{antelope} (letter group \qt{A}), \code{Brazil}
(letter group \qt{B}), \code{bee} (letter group \qt{B}).

\item \optfmt{letter-upperlower-reverse}: reverse upper-lower letter sort.
This now puts the \idx{lowercase} letters first within the letter group.
Example:

\code{bee} (letter group \qt{B}), \code{Brazil} (letter group
\qt{B}), \code{antelope} (letter group \qt{A}), \code{aardvark}
(letter group \qt{A}), \code{Africa} (letter group \qt{A}).

\item \optfmt{letter-lowerupper}: each character pair is first
compared according to their \idx{lowercase} values. If these are equal,
then they are compared according to case. This puts
\idxlink{uppercase}{upper} and \idx{lowercase} in the same letter group but the \idx{lowercase} comes first.
Example:

\code{aardvark} (letter group \qt{A}), \code{antelope} (letter group \qt{A}),
\code{Africa} (letter group \qt{A}), \code{bee} (letter group
\qt{B}), \code{Brazil} (letter group \qt{B}).

\item \optfmt{letter-lowerupper-reverse}: reverse lower-upper letter sort.
This now puts the \idx{uppercase} letters first within the letter group.
Example:

\code{Brazil} (letter group \qt{B}), \code{bee} (letter group
\qt{B}), \code{Africa} (letter group \qt{A}), \code{antelope}
(letter group \qt{A}), \code{aardvark} (letter group \qt{A}).

\end{itemize}

\subsubsection{Letter-Number}
\label{sec:letternumber}

The sort methods listed in \tableref{tab:sortoptionsletternumber} 
use a letter-integer hybrid. They behave in a similar way to the
above letter sort methods, but if an integer number pattern is detected in
the string then the sub-string containing the number will be compared.
This only detects base 10 integers (unlike the numeric methods
such as \csopt[hexadecimal]{sort} or \csopt[float]{sort}) but in
addition to recognising all the digits in the Unicode \idx{numberdecimaldigit} 
category it also recognises the subscript and
superscript digits, such as \textsuperscript{1} (\hex{00B9})
and \textsuperscript{2} (\hex{00B2}).

As with the letter sort methods, letters are compared using a
character code comparison not by a locale alphabet. The closest
locale-sensitive equivalent is to use \csopt{sort-number-pad} with a
locale sort method.

\begin{figure}
\figcontents
{%
\begin{tikzpicture}[every node/.style={minimum size=9mm},
 ampersand replacement=\&]
\matrix[font={\Large\ttfamily}]
{
  \node[draw]{a\hexsb{61}}; \& \node{$=$}; \& \node[draw]{a\hexsb{61}}; \&[20mm]
  \node[draw]{a\hexsb{61}}; \& \node{$=$}; \& \node[draw]{a\hexsb{61}};
  \\
  \node[draw]{b\hexsb{62}}; \& \node{$=$}; \& \node[draw]{b\hexsb{62}}; \&
  \node[draw]{b\hexsb{62}}; \& \node{$=$}; \& \node[draw]{b\hexsb{62}};
  \\
  \node[draw]{c\hexsb{63}}; \& \node{$=$}; \& \node[draw]{c\hexsb{63}}; \&
  \node[draw]{c\hexsb{63}}; \& \node{$=$}; \& \node[draw]{c\hexsb{63}};
  \\
  \node[draw,ultra thick]{1\hexsb{31}}; \& \node{$<$}; \&
  \node[draw,ultra thick]{6\hexsb{36}}; \&
  \node (n1) {1}; \& \node{$>$}; \& \node[draw,ultra thick]{6};
  \\
  \node[draw]{2\hexsb{32}}; \&  \& \node[draw]{b\hexsb{62}}; \&
  \node (n2) {2}; \&  \& \node[draw]{b\hexsb{62}};
  \\
  \node[draw]{f\hexsb{66}}; \&  \& \node[draw]{a\hexsb{61}}; \&
  \node[draw]{f\hexsb{66}}; \&  \& \node[draw]{a\hexsb{61}};
  \\
  \node[draw]{o\hexsb{6f}}; \&  \& \node[draw]{r\hexsb{72}}; \&
  \node[draw]{o\hexsb{6f}}; \&  \& \node[draw]{r\hexsb{72}};
  \\
  \node[draw]{o\hexsb{6f}}; \&  \& \&
  \node[draw]{o\hexsb{6f}}; \&  \& 
  \\[10pt]
   \& \node[font=\rmfamily]{\subfigfmt{a}};  \& \&
   \& \node[font=\rmfamily]{\subfigfmt{b}};  \& 
  \\
};
\draw[ultra thick] (n1.north east) rectangle (n2.south west);
\end{tikzpicture}%
}
{%
 \caption[Regular letter comparison vs letter-number
 comparison]{Regular letter comparison vs letter-number comparison.
 Comparing the strings \code{abc12foo} and \code{abc6bar}:
 \subfigfmt{a} \optfmt{letter-case}; \subfigfmt{b}
 \optfmt{letternumber-case}.}%
}
{fig:letternumber}
\end{figure}

For example, suppose the first string is \code{abc12foo}
and the second string is \code{abc6bar}.
\Figureref{fig:letternumber}\subfigfmt{a} shows the regular letter
comparison using \csopt[letter-case]{sort}, where the subscript
indicates the hexadecimal character code. The first three characters
from each string are identical (\code{abc}). At this point there's
no difference detected, so the comparator moves on to the next
character, \code{1\hexsb{31}} for the first string and
\code{6\hexsb{36}} for the second string. Since \hex{31} is less than 
\hex{36}, the first string (\code{abc12foo}) is considered less than the second
(\code{abc6bar}).

With the letter-number comparison using
\csopt[letternumber-case]{sort}, the comparator starts in much the
same way. The first three characters from each string are still
identical, so the comparator moves on to the next character,
\code{1} for the first string and \code{6} for the second. 
These are now both recognised as digits, so the comparator
looks ahead and reads in any following digits (if present).  For the
first case, this is the sub-string \code{12} and, for the second
case, \code{6} (\figureref{fig:letternumber}\subfigfmt{b}).  These
are both compared according to their integer representation $12 >
6$, so \code{abc12bar} is considered greater than \code{abc6foo}
(that is, \code{abc12bar} comes after \code{abc6foo}). 

The same result occurs for other numbering systems, for example if
the Basic Latin digits 1, 2 and 6 are replaced with the
corresponding Devanagari digits \devone, \devtwo\ and
\devsix. (But note that the letter comparisons will still be
based on their Unicode values not according to a particular locale.
This type of sort method is intended primarily for symbolic values,
such as chemical formulae, rather than for words or phrases.)

Signed integers are also recognised, so \code{abc-12foo} is less
than \code{abc+6bar}, which is again different from the result
obtained with a straight letter comparator where the
character~\code{+} (\hex{2B})
comes before the character~\code{-} (\hex{2D}). The sign must be followed by at
least one digit for it to be recognised as a number otherwise it's
treated as a punctuation character.

If only one sub-string is numeric
then the \csopt{letter-number-rule} is used to determine the
result. Where both sub-strings are non-numeric, then the 
\csopt{letter-number-punc-rule} setting is used to determine the result
according to the category of the characters, which may be one of the
following:
\begin{itemize}
\item white space: belongs to the Unicode \idx{separatorspace}
category. If both characters are white space, then they are compared
according to their Unicode values otherwise they are ordered according to 
the \csopt{letter-number-punc-rule} setting.

\item letter: belongs to one of the Unicode categories \idx{letteruppercase}, 
\idx{letterlowercase}, \idx{lettertitlecase},
\idx{lettermodifier} or \idx{letterother}. If both characters
are letters then, for sort method \optfmt{letternumber-\meta{modifier}}, 
the characters are compared in the same way as the
corresponding \optfmt{letter-\meta{modifier}} sort method otherwise
they are ordered according to the 
\csopt{letter-number-punc-rule} setting.

\item punctuation: everything else. If both characters are
punctuation, then they are compared according to their Unicode value
otherwise they are ordered according to 
the \csopt{letter-number-punc-rule} setting.
\end{itemize}
For simplicity, the actual sort value used during sorting isn't a
simple string but is converted into a list of objects that represent
one of: letter, integer, space or other (punctuation). This reduces
the amount of parsing of substrings that needs to be performed.

The examples below show the ordering of the list:
\code{CH\textsubscript{2}O}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{C\textsubscript{5}H\textsubscript{4}NCOOH}, 
\code{CO}, \code{Cl}, 
\code{Co}, \code{Co\textsubscript{2}O\textsubscript{3}}, 
\code{Co\textsubscript{2}}, \code{CO\textsubscript{2}},
\code{CoMoO\textsubscript{4}} and \code{CoCl\textsubscript{2}},
for the setting
\csopt[between]{letter-number-rule}, where the subscripts are
the Unicode subscript characters.

\begin{itemize}
\item \optfmt{letternumber-case}: case-sensitive
letter-number sort. Example:

\code{CH\textsubscript{2}O}, 
\code{CO},
\code{CO\textsubscript{2}},
\code{C\textsubscript{5}H\textsubscript{4}NCOOH}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{Cl}, 
\code{Co}, 
\code{CoCl\textsubscript{2}},
\code{CoMoO\textsubscript{4}}, 
\code{Co\textsubscript{2}},
\code{Co\textsubscript{2}O\textsubscript{3}}.

(Order determined by: $\mathtt{H} < \mathtt{O} < 5 < 10 < \mathtt{l} < \mathtt{o}$.)

\item \optfmt{letternumber-case-reverse}: reverse case-sensitive
letter-number sort. Example:

\code{Co\textsubscript{2}O\textsubscript{3}},
\code{Co\textsubscript{2}},
\code{CoMoO\textsubscript{4}}, 
\code{CoCl\textsubscript{2}},
\code{Co}, 
\code{Cl}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{C\textsubscript{5}H\textsubscript{4}NCOOH}, 
\code{CO\textsubscript{2}},
\code{CO},
\code{CH\textsubscript{2}O}.

\item \optfmt{letternumber-nocase}: case-insensitive
letter-number sort. The sort value is first converted to lower
case. Note that \csopt[between]{letter-number-rule} doesn't make
sense in this context as there won't be any \idx!{uppercase} characters in
the sort value, so numbers will always come before letters. Example:

\code{C\textsubscript{5}H\textsubscript{4}NCOOH}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{CH\textsubscript{2}O}, 
\code{Cl}, 
\code{CO}, 
\code{Co}, 
\code{CO\textsubscript{2}},
\code{Co\textsubscript{2}},
\code{Co\textsubscript{2}O\textsubscript{3}},
\code{CoCl\textsubscript{2}},
\code{CoMoO\textsubscript{4}}. 

(Order determined by: $5 < 10 < \mathtt{h} < \mathtt{l} < \mathtt{o}$.)

\item \optfmt{letternumber-nocase-reverse}: reverse case-insensitive
letter-number sort, so numbers will now always come after letters. Example:

\code{CoMoO\textsubscript{4}}, 
\code{CoCl\textsubscript{2}},
\code{Co\textsubscript{2}O\textsubscript{3}},
\code{Co\textsubscript{2}},
\code{CO\textsubscript{2}},
\code{Co}, 
\code{CO}, 
\code{Cl}, 
\code{CH\textsubscript{2}O}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{C\textsubscript{5}H\textsubscript{4}NCOOH}. 

\item \optfmt{letternumber-upperlower}: upper-lower
letter-number sort. This behaves slightly differently to
\optfmt{letter-upperlower} when used with \csopt[between]{letter-number-rule}
and has a more complicated rule that's determined by the character
following the number and implied numbers inserted between letters. 
(There was a bug in earlier versions that has been corrected in v1.8
so you may find a slightly different ordering when upgrading.)
Example:

\code{CH\textsubscript{2}O}, 
\code{C\textsubscript{5}H\textsubscript{4}NCOOH}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{Cl}, 
\code{CO}, 
\code{CO\textsubscript{2}},
\code{Co}, 
\code{Co\textsubscript{2}},
\code{CoCl\textsubscript{2}},
\code{CoMoO\textsubscript{4}}, 
\code{Co\textsubscript{2}O\textsubscript{3}}.

(Order determined by: $\mathtt{H} < 5\mathtt{H} < 10\mathtt{H} <
\mathtt{l} < \mathtt{O} < \mathtt{o}$, and for the terms starting
with \code{CO} or \code{Co}: 2 comes after null and $\mathtt{C} <
\mathtt{M} < 2\mathtt{O}$.)

Compare this with \csopt[before letter]{letter-number-rule} which
results in the order:

\code{C\textsubscript{5}H\textsubscript{4}NCOOH}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{CH\textsubscript{2}O}, 
\code{Cl}, 
\code{CO}, 
\code{CO\textsubscript{2}},
\code{Co}, 
\code{Co\textsubscript{2}},
\code{Co\textsubscript{2}O\textsubscript{3}},
\code{CoCl\textsubscript{2}},
\code{CoMoO\textsubscript{4}}.

\item \optfmt{letternumber-upperlower-reverse}: reverse upper-lower
letter-number sort. Example (with  \csopt[between]{letter-number-rule}):

\code{Co\textsubscript{2}O\textsubscript{3}},
\code{CoMoO\textsubscript{4}}, 
\code{CoCl\textsubscript{2}},
\code{Co\textsubscript{2}},
\code{Co}, 
\code{CO\textsubscript{2}},
\code{CO}, 
\code{Cl}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{C\textsubscript{5}H\textsubscript{4}NCOOH}, 
\code{CH\textsubscript{2}O}. 

Compare this with \csopt[before letter]{letter-number-rule} which
results in the order:

\code{CoMoO\textsubscript{4}}, 
\code{CoCl\textsubscript{2}},
\code{Co\textsubscript{2}O\textsubscript{3}},
\code{Co\textsubscript{2}},
\code{Co}, 
\code{CO\textsubscript{2}},
\code{CO}, 
\code{Cl}, 
\code{CH\textsubscript{2}O}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{C\textsubscript{5}H\textsubscript{4}NCOOH}. 

Remember that the associated settings are reversed as
well. So \csopt[before letter]{letter-number-rule} 
results in numbers \emph{after} letters.

\item \optfmt{letternumber-lowerupper}: lower-upper letter-number
sort. As with the upper-lower option, this behaves slightly differently to
\optfmt{letter-lowerupper} when used with \csopt[between]{letter-number-rule}
and has a more complicated rule.
Example:

\code{CH\textsubscript{2}O}, 
\code{C\textsubscript{5}H\textsubscript{4}NCOOH}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{Cl}, 
\code{Co}, 
\code{Co\textsubscript{2}},
\code{CoCl\textsubscript{2}},
\code{CoMoO\textsubscript{4}}, 
\code{Co\textsubscript{2}O\textsubscript{3}},
\code{CO}, 
\code{CO\textsubscript{2}}.

Compare this with \csopt[before letter]{letter-number-rule} which
results in the order:

\code{C\textsubscript{5}H\textsubscript{4}NCOOH}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{CH\textsubscript{2}O}, 
\code{Cl}, 
\code{Co}, 
\code{Co\textsubscript{2}},
\code{Co\textsubscript{2}O\textsubscript{3}},
\code{CoCl\textsubscript{2}},
\code{CoMoO\textsubscript{4}}, 
\code{CO}, 
\code{CO\textsubscript{2}}.

\item \optfmt{letternumber-lowerupper-reverse}: reverse lower-upper
letter-number sort. Example (with \csopt[between]{letter-number-rule}):

\code{CO\textsubscript{2}},
\code{CO}, 
\code{Co\textsubscript{2}O\textsubscript{3}},
\code{CoMoO\textsubscript{4}}, 
\code{CoCl\textsubscript{2}},
\code{Co\textsubscript{2}},
\code{Co}, 
\code{Cl}, 
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}}, 
\code{C\textsubscript{5}H\textsubscript{4}NCOOH}, 
\code{CH\textsubscript{2}O}. 

\end{itemize}

\subsubsection{Numerical}
\label{sec:numerical}

The sort methods listed in \tableref{tab:sortoptionsnumerical} 
use numeric comparisons. The sort value is expected to
be a numeric value. If it can't be parsed then it's treated as 0
(and a warning will be written to the transcript).
These all recognise the digits in the Unicode \qt{Number,
Decimal Digit} category but, unlike the hybrid letter-number
comparators above, they don't recognise the superscript or subscript
digits. The \qt{non-locale} in some of the descriptions below
indicates that the method doesn't recognise locale-sensitive
formatting, such as group separators.
\begin{itemize}
\item \optfmt{integer}: integer sort. This is for non-locale integer sort
values.

\item \optfmt{integer-reverse}: as above but reverses the order.

\item \optfmt{hex}: hexadecimal integer sort. This is for non-locale
hexadecimal sort values.

\item \optfmt{hex-reverse}: as above but reverses the order.

\item \optfmt{octal}: octal integer sort. This is for non-locale
octal sort values.

\item \optfmt{octal-reverse}: as above but reverses the order.

\item \optfmt{binary}: binary integer sort. This is for non-locale binary sort
values.

\item \optfmt{binary-reverse}: as above but reverses the order.

\item \optfmt{float}: single-precision sort. This is for non-locale
decimal sort values.

\item \optfmt{float-reverse}: as above but reverses the order.

\item \optfmt{double}: double-precision sort. This is for non-locale
decimal sort values.

\item \optfmt{double-reverse}: as above but reverses the order.

\item \optfmt{numeric}: locale-sensitive numeric sort. Use
\csopt{numeric-locale} to set the locale.

\item \optfmt{numeric-reverse}: as above but reverses the order.

\item \optfmt{currency}: locale-sensitive currency sort. Use
\csopt{numeric-locale} to set the locale.

\item \optfmt{currency-reverse}: as above but reverses the order.

\item \optfmt{percent}: locale-sensitive percent sort. Use
\csopt{numeric-locale} to set the locale.

\item \optfmt{percent-reverse}: as above but reverses the order.

\item \optfmt{numberformat}: locale-sensitive custom numeric sort. Use
\csopt{numeric-locale} to set the locale and
\csopt{numeric-sort-pattern} to set the number pattern.

\item \optfmt{numberformat-reverse}: as above but reverses the order.

\end{itemize}

In general, it doesn't make much sense to have hierarchical entries
that need to be sorted by a number, but it is possible as long as each
level uses the same type of numbering.

\subsubsection{Date-Time}

The sort methods listed in \tableref{tab:sortoptionsdatetime} 
are for dates and times.  Use \csopt{date-sort-format}
and \csopt{date-sort-locale} to specify the date format and locale.
\begin{itemize}
\item \optfmt{date}: sort dates.

\item \optfmt{date-reverse}: as above but reverses the order.

\item \optfmt{datetime}: sort date and time information.

\item \optfmt{datetime-reverse}: as above but reverses the order.

\item \optfmt{time}: sort times.

\item \optfmt{time-reverse}: as above but reverses the order.

\end{itemize}

If the field you want to sort by contains a date then the
simplest way to sort is to ensure the date is in ISO format
and then just use a letter sort. However it may be that
the date is in the format particular to your locale or you have a
mix of \era{AD} and \era{BC}. In which
case you can use one of the date/time sort options (such as
\csopt[date]{sort} or \csopt[date-reverse]{sort}). The locale
is assumed to be your default locale (as given by the \idx{JVM}) 
but if you are using a different locale this can
be set with \csopt{date-sort-locale}. The pattern is assumed
to be the default for that locale but you can change this with
\csopt{date-sort-format}. If you provide your
own custom pattern you must make sure that it matches the selected
\optfmt{sort} option.

Take care if you switch from using the \idx{JRE} to the \idx{CLDR}
 \idx{localeprovider} as you may find the default pattern changes.

The locale and pattern information is used by \bibgls\ to parse the
field.  If the field value can't be parsed then \bibgls\ will issue
a warning and assume the current date (or time).

The actual sort value that's used by the comparator is numeric.  In
the case of the time-based \csopt[datetime]{sort} and
\csopt[time]{sort} (or their \optfmt{-reverse} versions), this value
is the number of milliseconds since 1st~January, 1970. In the case
of \csopt[date]{sort} (or \csopt[date-reverse]{sort}), this value is
obtained from $a(y\times10000 + m\times100 + d)$ where $y$ is the
year, $m$ is the month number, $d$ is the day of month number, and
$a$ is an integer representation of the era ($-1$ for \era{BC} and
$+1$ for \era{AD}).

Unlike the numeric sort methods (such as \csopt[integer]{sort})
the date-time sort methods set the \field{sort} field to a value
that can be more easily parsed within the document and that should
mostly achieve the same ordering if a letter comparator were to be used
with it (except for \era{BC} dates, where the order needs to be
reversed). This has the by-product of providing a field that
you can access within the document that can be more easily parsed by
\LaTeX.

In general, it doesn't make much sense to have hierarchical entries
that need to be sorted by date, but it is possible as long as each
level uses the same date format.

For example, suppose my \ext{bib} file contains:
\begin{codeenv}
\atentry{entry}\marg{journalentry,
  \field{name}=\marg{10 Jan 2017},
  \field{description}=\marg{an interesting journal entry}
}
\end{codeenv}
The \field{name} field uses an abbreviated UK date format.
If all my other entries also use this format in the \field{name}
then I can sort them chronologically:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[date]{sort},
  \csopt[en-GB]{date-sort-locale},
  \csopt[medium]{date-sort-format}
}
\end{codeenv}
(The medium format is actually the default for this locale,
and the locale matches my system locale, so I could omit
both \csopt{date-sort-locale} and \csopt{date-sort-format}.)

If \longarg{verbose} mode is on, the transcript will show
the label, sort value and numeric value for each entry.
In this case, the information is:
\begin{verbatim}
journalentry -> '+1 2017-01-10' [20170110]
\end{verbatim}
The first value is the label (\code{journalentry}), the second
value is assigned to the \field{sort} field
(\code{+1 2017-01-10}) and the number in square brackets 
is the actual numeric value used by the comparator. The signed
number at the start of the sort field \code{+1} is the numeric
representation of the era as used for the $a$ variable in the
computation of the numeric value (as described earlier).

If I change the format to \csopt[short]{date-sort-format},
then the date can't be parsed correctly and \bibgls\ will 
issue the following warning:
\begin{verbatim}
Warning: Can't parse sort value '10 Jan 2017' for 'journalentry'
(pattern: 'dd/MM/yyyy')
\end{verbatim}
This shows the value that \bibgls\ is trying to
parse (\code{10 Jan 2017}) for the entry identified by
the given label (\code{journalentry}). The pattern \bibgls\ 
expects is also given (\code{dd/MM/yyyy}).

\optsection{shuffle}

Automatically sets \csopt[random]{sort} and \csopt{flatten}.
The value \meta{seed} may be omitted. If present, it should
be an integer used as a seed for the random number generator.

\optsection{sort-field}

The \csopt{sort-field} key indicates which field provides the sort
value. The \meta{field} must be a recognised field name or you
may use \csopt[id]{sort-field} to sort according to the label.
The default value is the \field{sort} field (which is typically
inferred rather than explicitly set).

Example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries-terms]{src},\comment{data in entries-terms.bib}
 \csopt[symbol]{sort-field},\comment{sort by 'symbol' field}
 \csopt[letter-case]{sort}\comment{case-sensitive letter sort}
}
\end{codeenv}
This sorts the entries according to the \field{symbol} field using
a case-sensitive letter comparison.

\begin{important}
In general it's better to use the default \csopt[sort]{sort-field} 
and adjust the fallbacks instead. The \csopt{sort-field} option is
provided if you want to use a specific field regardless of the entry
type.
\end{important}

If an entry is missing a value for \meta{field}, then the value of
the fallback field will be used instead. If \csopt{missing-sort-fallback} is
set, then that's used as the fallback, otherwise it depends on the
entry type. 
If no fallback field can be found, the entry's label will be used.

For the specific case with the default \csopt[sort]{sort-field}
setting, the fallback for the \field{sort} field is governed not
only by the entry type but also by some associated settings:

\begin{itemize}
\item If the entry's original type (before being aliased with
\csopt{entry-type-aliases}) is identified in
\csopt{custom-sort-fallbacks}, then if the \field{sort} field is missing
the value is obtained from the supplied custom mapping.

\item If the entry is defined using \atentry{entry} (or a dual form
that acts like \atentry{entry}), then if the \field{sort} field is
missing the value is obtained from the field identified by
\csopt{entry-sort-fallback}. If that field is also missing then
that field's fallback is used.

\item For the index entry types like \atentry{index} or
\atentry{indexplural}, then if the \field{sort} field is missing the
value is obtained from the \field{name} field. If that field is also
missing, then the value is obtained from the particular entry type's
fallback for the \field{name} field. (For example, the entry's label
for \atentry{index} or the \field{plural} field for
\atentry{indexplural}.)

\item If the entry is defined with an abbreviation type (for
example, \atentry{abbreviation} or \atentry{acronym}) then if the
\field{sort} field is missing, \bibgls\ will fallback on the field
given by \csopt{abbreviation-sort-fallback}.

\item The symbol-like entry types fallback on the field given by
\csopt{symbol-sort-fallback} if the \field{sort} field is missing.

\item Entries defined using \atentry{bibtexentry} fallback on the field
given by \csopt{bibtexentry-sort-fallback}, which defaults to the
\field{name} field. Note that this only applies to the main entry.
The spawned \atentry{contributor} entries behave like
\atentry{index}.

\end{itemize}
Use \csopt{dual-sort-field} when sorting dual entries.

\optsection{missing-sort-fallback}

With \csopt[\meta{sort-field}]{sort-field}, if the value of the field
identified by \meta{sort-field} is missing, then \bibgls\ behaves as
follows:
\begin{enumerate}
\item If \csopt[\meta{fallback-field}]{missing-sort-fallback} is set, then
\bibgls\ will fallback on the value provided by the field
\meta{fallback-field}.  If \meta{fallback-field} is
missing, then \bibgls\ will query the entry type's fallback for
\meta{fallback-field} (not for \meta{sort-field}).

The \meta{fallback-field} must be a known field but not an internal
field. It can't be the \field{sort} field. (Take care not to cause
an infinite loop if \csopt{sort-field} has been changed.) Unlike the
other sort fallback options such as \csopt{entry-sort-fallback}, the
\meta{fallback-field} can't be a keyword (to identify the label) and can't be a
composite.

\item If the entry type has a fallback rule for \meta{sort-field},
then that rule is used. When \csopt[sort]{sort-field} this means:
\begin{itemize}
\item If the entry's original entry type has been identified in
\csopt{custom-sort-fallbacks}, then \bibgls\ will fallback on the
designated custom setting.

\item If the entry was defined using one of the index types (such as
\atentry{index}), then \bibgls\ will fallback on the \field{name}
field.

\item If the entry was defined using the \atentry{entry} type (or a
dual form that acts like \atentry{entry}), then \bibgls\ will
fallback on the field given by \csopt{entry-sort-fallback}.

\item If the entry was defined using
one of the symbol types (such as \atentry{symbol}), then \bibgls\
will fallback on the field given by \csopt{symbol-sort-fallback}.

\item If the entry was defined using
one of the abbreviation types (such as \atentry{abbreviation}), then \bibgls\
will fallback on the field given by
\csopt{abbreviation-sort-fallback}.

\item If the entry was defined using \atentry{bibtexentry} (but not
the spawned \atentry{contributor} entries), then \bibgls\ will
fallback on the field given by
\csopt{bibtexentry-sort-fallback}.
\end{itemize}
If \meta{sort-field} is not \field{sort}, then there may not be a
fallback, in which case the next condition applies:
\item Otherwise the sort value will be set to the entry label and \bibgls\
will issue a warning.
\end{enumerate}
The default setting is \csopt[\empty]{missing-sort-fallback}, which
means that step~1 above is omitted.

Use \csopt{dual-missing-sort-fallback} when sorting dual entries
separately from primaries, and use
\csopt{secondary-missing-sort-fallback} for \csopt{secondary}
sorting.

\optsection{custom-sort-fallbacks}

The value should be a \keyvallist\ in the form
\meta{entrytype}\dequals\meta{field} where \meta{entrytype} is the
\emph{original} entry type (before being aliased with
\csopt{entry-type-aliases}). This will override any of the sort fallback
options listed below for entries whose original entry type matches
\meta{entrytype}.

The \meta{field} may be a known field but not an internal field. For
obvious reasons, it can't be the \field{sort} field (since
\meta{field} is the fallback a missing \field{sort} field). It may also be one of the
keywords: \optfmt{id} (for the entry's label) or \optfmt{original id}
(for the entry's original label). The \meta{field} may also be a
composite in the form \optfmt{\meta{field1}+\meta{field2}+\ldots}
which indicates that the sort value should be obtained by
concatenating the values of the given fields, where the separator is
given by \csopt{field-concat-sep}.

For example, if the \ext{bib} file contains:
\begin{codeenv}
\atentryfmt{unit}\marg{ohm,
  \field{name}=\marg{\cs{si}\marg{\csfmt{ohm}}},
  \field{description}=\marg{electrical resistance}
}
\strut
\atentryfmt{constant}\marg{pi,
  \field{name}=\marg{\cs{ensuremath}\marg{\cs{pi}}},
  \field{description}=\marg{the ratio of the length of the circumference
    of a circle to its diameter},
  \field{user1}=\marg{3.14159}
}
\strut
\atentry{symbol}\marg{fx,
  \field{name}=\marg{\cs{ensuremath}\marg{f(x)}},
  \field{description}=\marg{a function of \idx{mshiftchar}x\idx{mshiftchar}}
}
\strut
\atentry{number}\marg{zero,
  \field{name}=\marg{0},
  \field{description}=\marg{nothing or no quantity}
}
\end{codeenv}
Then the resource options:
\begin{codeenv}
\csopt[\entryfmt{unit}=\entryref{symbol},\entryfmt{constant}=\entryref{number}]{entry-type-aliases},
\csopt[\entryfmt{unit}=\field{name},\entryfmt{constant}=\field{user1}]{custom-sort-fallbacks}
\end{codeenv}
will treat the custom \atentryfmt{unit} and \atentryfmt{constant}
entries as though they had been defined with \atentry{symbol}
and \atentry{number}, respectively, but the fallback for the
\field{sort} field is different: the \code{ohm} entry will use the
\field{name} field for the sort fallback (because its original entry
type was \entryfmt{unit}), the \code{pi} entry will use the
\field{user1} field for the sort fallback (because its original
entry type was \entryfmt{constant}) and the \code{fx} and
\code{zero} entries will use the label for the sort fallback (since
neither \entryfmt{symbol} nor \entryfmt{number} were identified in
\csopt{custom-sort-fallbacks} so the \csopt{symbol-sort-fallback} is
used).

If an entry hasn't had its entry type aliased then \meta{entrytype}
is its actual entry type. For example, consider the following
definitions:
\begin{codeenv}
\atentry{abbreviation}\marg{svm,
   \field{short}=\marg{SVM},
   \field{long}=\marg{support vector machine}
}
\atentry{acronym}\marg{laser,
  \field{short}=\marg{laser},
  \field{long}=\marg{light amplification by stimulated emission of radiation}
}
\end{codeenv}
Then \csopt[short]{abbreviation-sort-fallback} will make both
entries fallback on the \field{short} field (since
\csopt{abbreviation-sort-fallback} applies to both \atentry{acronym}
and \atentry{abbreviation}), but the option:
\begin{codeenv}
\csopt[\entryref{abbreviation}=\field{long},\entryref{acronym}=\field{short}]{custom-sort-fallbacks} 
\end{codeenv}
will make the entry defined with \atentry{abbreviation} fallback on
the \field{long} field and the entry defined with \atentry{acronym}
will fallback on the \field{short} field.

Since the default setting is \csopt[short]{abbreviation-sort-fallback}
this only needs to be:
\begin{codeenv}
\csopt[\entryref{abbreviation}=\field{long}]{custom-sort-fallbacks} 
\end{codeenv}
In this case, the entry defined with \atentry{abbreviation} (\qt{SVM}) will use
the setting given in \csopt{custom-sort-fallbacks}, but the entry
defined with \atentry{acronym} (\qt{laser}) will use the setting given by
\csopt{abbreviation-sort-fallback} since \atentry{acronym} hasn't
been identified in \csopt{custom-sort-fallbacks}.

This option also covers dual entries. For example:
\begin{codeenv}
\csopt[
   \entryref{dualindexnumber}=\field{description},
   \entryfmt{dualindexnumbersecondary}=\field{user1}
]{custom-sort-fallbacks}
\end{codeenv}
Note that the entry type for the dual is in the form 
\code{\meta{primary entry type}secondary}.

\begin{important}
The \csopt{custom-sort-fallbacks} setting is only used when \bibgls\
tries to access the \field{sort} field for an entry (whose original
entry type has been identified in this setting) and finds that the field
hasn't been set. This means that this setting has no effect if you
explicitly set the \field{sort} field or if you change the field
used for sorting (\csopt{sort-field}).
\end{important}

\optsection{entry-sort-fallback}

The regular entry types (such as \atentry{entry} and
\atentry{dualentry}) will, by default, fallback on the \field{name}
field if the \field{sort} field is missing (assuming
\csopt[sort]{sort-field}). If you prefer to
fallback on a different field, then you can use this option to
specify the field.
Note that \csopt{missing-sort-fallback} and
\csopt{custom-sort-fallbacks} override this setting.

The \meta{field} may be a known field but not an internal
field. It can't be the \field{sort} field. It may also be
one of the keywords: \optfmt{id} (for the entry's label) or
\optfmt{original id} (for the entry's original label). The
\meta{field} may also be a composite in the form
\optfmt{\meta{field1}+\meta{field2}+\ldots} which indicates that the
sort value should be obtained by concatenating the values of given fields,
where the separator is given by \csopt{field-concat-sep}.

This setting doesn't affect the index type of entries, such as
\atentry{index} or \atentry{indexplural}. This is useful if your
glossary contains homographs (terms with the same spelling) which
can't be distinguished by the sort comparators. For example, suppose
my file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{index}\marg{pagelist,
  \field{name}=\marg{page list},
  \field{description}=\marg{a list of individual pages or page ranges}
}
\strut
\atentry{index}\marg{glossary}
\strut
\atentry{entry}\marg{glossarylist,
  \field{parent}=\marg{glossary},
  \field{description}=\marg{list of technical words}
}
\strut
\atentry{entry}\marg{glosscol,
  \field{parent}=\marg{glossary},
  \field{description}=\marg{collection of glosses}
}
\end{codeenv}
Now first consider a document that uses the default settings:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}\oarg{\styopt{record},\styopt{subentrycounter},\styopt[\glostyle{treenoname}]{style}}\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\strut
\cmd{begin}\marg{document}
A test document describing \cs{glspl}\marg{pagelist} and
\cs{gls}\marg{glosscol} (collection) vs \cs{gls}\marg{glossarylist} (list).
\strut
\cs{printunsrtglossary}
\cmd{end}\marg{document}
\end{codeenv}
The default behaviour for \atentry{entry} if the \field{sort} field
is missing is to fallback on the \field{name} field. If the
\field{name} field is missing (as with \code{glossarylist} and
\code{glosscol}), then the value is obtained from the
\field{name} field from the parent entry. The parent entry for these
homographs is the \code{glossary} entry, which was defined with
\atentry{index} and doesn't have the \field{name} field. For the
\atentry{index} entries, if \field{name} is missing the value is
obtained from the label.

Therefore both \code{glossarylist} and \code{glosscol} end
up with the same sort value: \code{glossary}. This triggers a
message in verbose mode (\longarg{verbose}) which can be found 
in the transcript file:
\begin{verbatim}
Identical sort values for 'glossarylist' and 'glosscol'
Falling back on ID
\end{verbatim}
So the actual sort values used are \qt{glossarylist} and
\qt{glosscol}. This puts the \code{glossarylist} entry
before the \code{glosscol} entry.

Now suppose a minor modification is made to the document:
\begin{codeenv}
\gls{GlsXtrLoadResources}
 \oarg{
   \csopt[entries]{src},
   \csopt[description]{entry-sort-fallback}
 }
\end{codeenv}
This means that when the sort function fails to find the
\field{sort} field for the terms defined with \atentry{entry}, it
will fallback on the \field{description} field. This doesn't affect
the terms defined with \atentry{index}, which still fallback on the
\field{name} field. This time there's no message in the transcript
file and the \code{glosscol} entry now comes before the
\code{glossarylist} entry.

\begin{important}
The \csopt{entry-sort-fallback} setting is only used when \bibgls\
tries to access the \field{sort} field for a term defined with
\atentry{entry} and finds that the field hasn't been set. This means
that this setting has no effect if you explicitly set the \field{sort}
field or if you change the field used for sorting
(\csopt{sort-field}).
\end{important}

\optsection{abbreviation-sort-fallback}

The entry types that define abbreviations (such as
\atentry{abbreviation} and \atentry{acronym}) will, by default,
fallback on the \field{short} field if the \field{sort} field is
missing (assuming \csopt[sort]{sort-field}). If you prefer to
fallback on a different field, then you can use this option to
specify the field. For example,
\csopt[long]{abbreviation-sort-fallback}. Note that if you use
\csopt[name]{sort-field}, then the fallback field will be given by
\csopt{abbreviation-name-fallback} if the \field{name} field is
omitted.

The \meta{field} may be a known field but not an internal
field. It can't be the \field{sort} field. It may also be
one of the keywords: \optfmt{id} (for the entry's label) or
\optfmt{original id} (for the entry's original label). The
\meta{field} may also be a composite in the form
\optfmt{\meta{field1}+\meta{field2}+\ldots} which indicates that the
sort value should be obtained by concatenating the values of given fields,
where the separator is given by \csopt{field-concat-sep}.

Note that \csopt{missing-sort-fallback} and
\csopt{custom-sort-fallbacks} override this setting.

\begin{important}
The \csopt{abbreviation-sort-fallback} setting is only used when \bibgls\
tries to access the \field{sort} field for an abbreviation and finds
that the field hasn't been set. This means that this setting has no effect 
if you explicitly set the \field{sort} field or if you change the field used 
for sorting (\csopt{sort-field}).
\end{important}

\optsection{symbol-sort-fallback}

The entry types that define symbols (such as \atentry{symbol} and
\atentry{number}) will, by default, fallback on the entry's original label if the
\field{sort} field is missing (assuming the default
\csopt[sort]{sort-field}). If you prefer to fallback on a different
field, then you can use this option to specify the field. For
example, \csopt[name]{symbol-sort-fallback}.

The \meta{field} may be a known field but not an internal
field. It can't be the \field{sort} field. It may also be
one of the keywords: \optfmt{id} (for the entry's label) or
\optfmt{original id} (for the entry's original label). The
\meta{field} may also be a composite in the form
\optfmt{\meta{field1}+\meta{field2}+\ldots} which indicates that the
sort value should be obtained by concatenating the values of given fields,
where the separator is given by \csopt{field-concat-sep}.

Note that \csopt{missing-sort-fallback} and
\csopt{custom-sort-fallbacks} override this setting.

\begin{important}
The \csopt{symbol-sort-fallback} setting is only used when \bibgls\
tries to access the \field{sort} field for a symbol and finds
that the field hasn't been set. This means that this setting has no effect 
if you explicitly set the \field{sort} field or if you change the field used 
for sorting (\csopt{sort-field}).
\end{important}

\optsection{bibtexentry-sort-fallback}

The main \atentry{bibtexentry} entry types will, by default,
fallback on the \field{name} if the
\field{sort} field is missing (assuming the default
\csopt[sort]{sort-field}). If you prefer to fallback on a different
field, then you can use this option to specify the field.

The \meta{field} may be a known field but not an internal
field. It can't be the \field{sort} field. It may also be
one of the keywords: \optfmt{id} (for the entry's label) or
\optfmt{original id} (for the entry's original label). The
\meta{field} may also be a composite in the form
\optfmt{\meta{field1}+\meta{field2}+\ldots} which indicates that the
sort value should be obtained by concatenating the values of given fields,
where the separator is given by \csopt{field-concat-sep}.

Note that \csopt{missing-sort-fallback} and
\csopt{custom-sort-fallbacks} override this setting.

\begin{important}
The \csopt{bibtexentry-sort-fallback} setting is only used when
\bibgls\ tries to access the \field{sort} field for a main entry
defined with \atentry{bibtexentry} and finds that the field hasn't
been set. This means that this setting has no effect if you
explicitly set the \field{sort} field or if you change the field used for
sorting (\csopt{sort-field}).
\end{important}

\optsection{field-concat-sep}

This option sets the field concatenation separator to \meta{value}.
The default is a space. An empty value indicates no separator. You
may use \ics{uhex}\meta{hex} to indicate a character by its hexadecimal
code. For example, suppose the \ext{bib} file contains:
\begin{codeenv}
\atentry{abbreviation}\marg{ac,
  \field{short}=\marg{AC},
  \field{long}=\marg{alternating current}
}
\atentry{index}\marg{acacia}
\end{codeenv}
Then the resource option:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[letter-nocase]{sort},
  \csopt[\field{short}+\field{long}]{abbreviation-sort-fallback}
}
\end{codeenv}
will set the \code{ac} sort value to \qt{AC alternating current}.
That is, the \field{short} value concatenated with the \field{long}
value using the default space separator. With the
\optfmt{letter-nocase} sort method, this will put the \code{ac}
entry before the \code{acacia} entry (because the space character
comes before \qt{a}).

If the resource options are changed to:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[letter-nocase]{sort},
  \csopt[\empty]{field-concat-sep},
  \csopt[\field{short}+\field{long}]{abbreviation-sort-fallback}
}
\end{codeenv}
This will obtain the sort value for abbreviations from a
concatenation of the \meta{short} and \meta{long} values without a
separator. This means that the \code{ac} sort value will be
\qt{ACalternating current} and so the \code{ac} entry will
come after the \code{acacia} entry (since \qt{l} comes after \qt{c}).

This setting is only used for the sort fallback options that allow
field concatenation (such as \csopt{entry-sort-fallback} but not
\csopt{missing-sort-fallback}). 

Note that due to the way that the \keyvallist\ parser trims leading and
trailing spaces, you can't simply do \csopt[\space]{field-concat-sep}
to indicate a space character as the value will end up as an empty
string. You can instead do \csopt[\stringu20]{field-concat-sep} but
since this is the default value there shouldn't be much need for it.

Remember that the separator may be replaced with a break point
marker depending on the sort method and \csopt{break-at} setting.

\optsection{trim-sort}

If the interpreter is used to determine the sort value, this setting
governs whether or not the interpreter should trim leading and
trailing spaces. The default setting is \csopt[true]{trim-sort}.

This option automatically sets \csopt[\meta{boolean}]{dual-trim-sort}
and \csopt[\meta{boolean}]{secondary-trim-sort}.

\optsection{sort-replace}

This option may be used to perform \idx{regex} substitutions
on the sort value and has the same syntax as \csopt{labelify-replace}. 
The value is required for this key but may be empty, which indicates
that the setting is switched off.

This action is done after the interpreter parses the
sort value (if applicable) and before \csopt{sort-number-pad} (if
applicable). For example, suppose the sort value is:
\begin{codeenv}
\cs{ensuremath}\marg{\ics{approx} 3.14}
\end{codeenv}
then the interpreter will convert this to $\mathtt{\approx}$\texttt{3.14} but:
\begin{codeenv}
\csopt[\marg{\cs{glshex}2248}\marg{}]{sort-replace}
\end{codeenv}
can be used to strip the $\approx$ symbol (\hex{2248}) so that the value can
now be parsed as a number if \csopt[double]{sort} has been used.

Use \csopt{dual-sort-replace} for dual
and \csopt{secondary-sort-replace} for secondary sort methods.

\optsection{sort-rule}

If the \csopt[custom]{sort} option is used, the sort rule must be
provided with \optfmt{sort-rule}. If \csopt{sort} is not set to
\optfmt{custom}, the \optfmt{sort-rule} setting will be ignored.
This setting uses Java's
\href{http://docs.oracle.com/javase/8/docs/api/java/text/RuleBasedCollator.html}{RuleBasedCollator}
class~\cite{rulebasedcollator}, and the rule syntax needs to conform
to that format.

Remember that the options will be expanded as they are written to
the \iext{aux} file, so be careful of any special characters that
occur in the rule. For the special characters \idx{param}
\idx{commentchar} \idx{sbchar} \idx{colsep} \idx{bgroupchar} and
\idx{egroupchar} you can use \ics{cs.hash}, \ics{cs.percent},
\ics{cs.underscore}, \ics{cs.amp}, \ics{cs.openbrace} and
\ics{cs.closebrace}. These will be written to the \ext{aux} file
with the leading backslash, but \bibgls\ will remove it for this
resource option. Remember that the \styfmt{glossaries} package
provides \ics{glsbackslash} and \ics{glstildechar} which can be used
to produce a literal backslash (\idx{backslashchar}) and tilde
(\idx{tildechar}).

You can also
use \stringu\meta{hex} (where \meta{hex} is a hexadecimal
code) to represent a Unicode character. For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[custom]{sort},
  \csopt[< a,A < b,B < c,C < ch,Ch,CH < d,D
   < dd,Dd,DD < e,E < f,F < ff,Ff,FF
   < g,G < ng,Ng,NG < h,H < ij,Ij,IJ
   < i,I < j,J < k,K < l,L < ll,Ll,LL < m,M
   < n,N < o,O < p,P < ph,Ph,PH < q,Q < r,R < rh,Rh,RH
   < s,S < t,T < th,Th,TH < u,U < v,V < w,W < x,X < y,Y < z,Z
   < \stringu00E6,\stringu00C6]{sort-rule}
}
\end{codeenv}
It's best to use \ics{cs.string} rather than \ics{protect} to avoid
unwanted spaces interfering with \meta{hex}. Note that
\sty{glossaries-extra} v1.21+ provides\footnote{The command
definition was moved to \isty{glossaries-extra-bib2gls} from version
1.27 since it's only needed with \bibgls.}\ \ics{glshex}
which just does \stringu\ so you can do
\code{\ics{glshex} 00E6} instead of \code{\stringu00E6}.
This is only one character different, but you can redefine
\ics{glsxtrresourceinit} to locally set \ics{u} to \ics{glshex} while
the protected write is performed. For example:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrresourceinit}}\marg{\cmd{let}\cmd{u}\cs{glshex}}
\end{codeenv}
Then you can just do \code{\csfmt{u}00E6} instead of \code{\stringu00E6}.

The \isty{glossaries-extra-bib2gls} package (which is automatically loaded by
the \styopt{record} option) provides some commands for common rule blocks 
that may be used in the construction of custom rules. For example:
\begin{codeenv}
\csopt[\ics{glsxtrcontrolrules}
 ;\ics{glsxtrspacerules}
 ;\ics{glsxtrnonprintablerules}
 ;\ics{glsxtrcombiningdiacriticrules}
 ,\ics{glsxtrhyphenrules}
 <\ics{glsxtrgeneralpuncrules}
 <\ics{glsxtrdigitrules}
 <\ics{glsxtrfractionrules}
 <\ics{glsxtrMathItalicGreekIrules}
 <\ics{glsxtrGeneralLatinIVrules}
 <\ics{glsxtrLatinAA}
 <\ics{glsxtrLatinOslash}
]{sort-rule}
\end{codeenv}
This places the Greek maths symbols (such as \ics{alpha}) before the
Latin block. See the \sty{glossaries-extra} documentation for
further details of these commands.

You might find it convenient to provide similar commands in a
package for rules you may often need. For example, suppose I have a
package called, say, \styfmt{mapsymbols} for providing map symbols:
\begin{codeenv}
\cmd{NeedsTeXFormat}\marg{LaTeX2e}
\cmd{ProvidesPackage}\marg{mapsymbols}
\comment{some package or font loading stuff here to provide}
\comment{the appropriate symbols}
\cmd{newcommand}\marg{\cmd{Stadium}}\marg{\ldots}
\cmd{newcommand}\marg{\cmd{Battlefield}}\marg{\ldots}
\cmd{newcommand}\marg{\cmd{Harbour}}\marg{\ldots}
\comment{etc}
\strut
\comment{Provide a rule block:}
\cmd{newcommand}\marg{\cmd{MapSymbolOrder}}\marg{\comment{}
 \cs{glshex} 2694 \comment{crossed-swords 0x2694}
 < \cs{glshex} 2693 \comment{anchor 0x2693}
 < \cs{glshex} 26BD \comment{football 0x26BD}
}
\end{codeenv}
In addition to \filefmt{mapsymbols.sty}, I also need to create
\filefmt{mapsymbols.bib} to provide the appropriate definitions for
\bibgls:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{glsxtrprovidecommand}\marg{\cmd{Harbour}}\marg{\icharhex{2693}}
\cs{glsxtrprovidecommand}\marg{\cmd{Battlefield}}\marg{\charhex{2694}}
\cs{glsxtrprovidecommand}\marg{\cmd{Stadium}}\marg{\charhex{26BD}}}}
\end{codeenv}
The use of \ics{glsxtrprovidecommand} will override any previous
definitions of these commands in \bibgls's interpreter but will act
like \ics{providecommand} within the document, and so won't
interfere with the commands defined in \filefmt{mapsymbols.sty}.
Now I can just do:
\begin{codeenv}
\cmd{usepackage}\marg{mapsymbols}\comment{my custom package}
\cmd{usepackage}[\styopt{record}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[mapsymbols,\comment{<--- my custom mapsymbols.bib}
   entries\comment{data in entries.bib}
  ]{src},
  \csopt[custom]{sort},
  \csopt[\cs{glsxtrcontrolrules}
 ;\cs{glsxtrspacerules}
 ;\cs{glsxtrnonprintablerules}
 ;\cs{glsxtrcombiningdiacriticrules}
 ,\cs{glsxtrhyphenrules}
 <\cs{glsxtrgeneralpuncrules}
 <\cs{glsxtrdigitrules}
 <\cs{glsxtrfractionrules}
 <\cmd{MapSymbolOrder} \comment{<--- custom map symbols}
 <\cs{glsxtrMathItalicGreekIrules}
 <\ics{glsxtrGeneralLatinIrules}
 ]{sort-rule}
}
\end{codeenv}

An alternative to providing \filefmt{mapsymbols.bib} is to provide a
custom package just for \bibgls' use. For example,
\filefmt{mapsymbols-bib2gls.sty}:
\begin{codeenv}
\comment{Provided for \bibgls\ only.}
\comment{Use \cmd{usepackage}\marg{mapsymbols} in the document.}
\cmd{NeedsTeXFormat}\marg{LaTeX2e}
\cmd{ProvidesPackage}\marg{mapsymbols-bib2gls}
\cs{glsxtrprovidecommand}\marg{\cmd{Harbour}}\marg{\charhex{2693}}
\cs{glsxtrprovidecommand}\marg{\cmd{Battlefield}}\marg{\charhex{2694}}
\cs{glsxtrprovidecommand}\marg{\cmd{Stadium}}\marg{\charhex{26BD}}
\cmd{endinput}
\end{codeenv}
and instruct \bibgls\ to parse it with 
\code{\longarg{custom-packages} mapsymbols-bib2gls} (and use
\filefmt{mapsymbols.sty} in the document). Remember that \bibgls\
isn't a \TeX\ engine so make sure to only use simple commands in
this file.

\optsection{break-at}

This option automatically implements
\csopt[\meta{option}]{dual-break-at} and
\csopt[\meta{option}]{secondary-break-at}.

The alphabet sort options (\tableref{tab:sortoptionsrule}) typically
list non-letter characters before alphabetical characters and spaces
are quite often in the ignored set. This means that the alphabet
sort options are naturally in a letter order, similar to
\idx!{xindy}['s] \code{ord/letorder} module.  (This isn't the same as
\csopt[letter-nocase]{sort}, which just sorts according to the
Unicode value not according to a particular alphabet.)

In order to replicate \idx!{makeindex} and \idx!{xindy}['s] default word
order, \bibgls\ splits up the sort value at word boundaries and
inserts a marker (identified by \csopt{break-marker}).
For example, if the sort value is \qt{sea lion} then it's actually
converted to \verb"sea|lion|" whereas \qt{sea} becomes \verb"sea|"
and \qt{seal} becomes \verb"seal|". The default marker is \verb"|"
which is commonly placed in collation rules before digits but 
after the ignored characters, such as spaces and hyphens.

Note that this action removes non-letters, so for example,
if the sort value is \verb"# (parameter)" then it will be converted
to \verb"parameter|" (hash, space and parentheses removed).
If you only want to break at spaces (optionally following a comma)
use the following instead:
\begin{codeenv}
 \csopt[none]{break-at},
 \csopt[\marg{,? +}\marg{|}]{sort-replace}
\end{codeenv}

You can change the construction of the break points with
\csopt[\meta{option}]{break-at} where \meta{option} may be one of:
\begin{itemize}
\item \optfmt{word}: break at word boundaries (default).
Note that what constitutes a word varies
according to the locale but usually anything that's not alphanumeric 
will designate a word-boundary. The characters between words are
discarded.
For example, the sort value \qt{Tom, Dick, and Harry} becomes 
\verb"Tom|Dick|and|Harry", which has discarded the comma and space
characters.
\item \optfmt{character}: break after each character.
\item \optfmt{sentence}: break after each sentence.
\item \optfmt{upper-notlower}: break after any \idx{uppercase} character
that's not followed by a \idx{lowercase} character. For example, 
\qt{MathML} becomes \verb"MathM|L|" and \qt{W3C} becomes
\verb"W|3C|".
\item \optfmt{upper-upper}: break after any \idx{uppercase} character
that's followed by an \idx{uppercase} character.
\item \optfmt{upper-notlower-word}: first applies break-points
according to \optfmt{upper-notlower} and then according to
\optfmt{word}.
\item \optfmt{upper-upper-word}: first applies break-points
according to \optfmt{upper-upper} and then according to
\optfmt{word}.
\item \optfmt{none}: don't create break points. Use this option to
emulate \idx!{makeindex} or \idx!{xindy}'s letter ordering, or combine
with \csopt{sort-replace} to insert custom break points.
\end{itemize}

This option is ignored when used with the non-alphabetic
\csopt{sort} options. You can find the break points in the
\field{sort} field for the entry's definition in the \ext{glstex}
file (which is provided for information rather than for use in the
document).  Alternatively, use the \longarg{debug} switch to show
the break points in the transcript. (This will also show the
collation rule.) 

\optsection{break-marker}

This option automatically implements the dual and secondary settings
\csopt[\meta{marker}]{dual-break-marker} and
\csopt[\meta{marker}]{secondary-break-marker}.

The break marker can be changed using
\csopt[\meta{marker}]{break-marker}, where \meta{marker} is 
the character to use. For example, \csopt[-]{break-marker} will use a
hyphen. The marker may be empty, which effectively strips the
inter-word punctuation. For example, with
\csopt[\empty]{break-marker}, \qt{Tom, Dick, and Harry} becomes
\code{TomDickandHarry} and \qt{sea lion} simply becomes
\code{sealion}. If \meta{marker} is omitted,
\csopt[\empty]{break-marker} is assumed.

\optsection{sort-number-pad}

This option automatically implements the dual and secondary settings
\csopt[\meta{number}]{dual-sort-number-pad},
\csopt[\meta{number}]{secondary-sort-number-pad}.

If \meta{number} is greater than 1, any integer sub-strings found
in the sort value will be zero-padded up to this value. Since the
\code{-} character is often ignored by rule-based sort methods, 
any signs found will be replaced with the markers given by
\csopt{sort-pad-plus} and \csopt{sort-pad-minus}, which should be
chosen to ensure that negative numbers are ordered before positive
numbers (if this is desired). An unsigned number will have the
\csopt{sort-pad-plus} marker inserted before it. The default value
is \csopt[0]{sort-number-pad}, which doesn't implement any padding.

If you use this with a locale sort method, it's best to also set
\csopt[none]{break-at}, as the default word boundary break points
will likely be confused by a mix of alphanumerics.

\optsection{sort-pad-plus}

This option automatically implements the dual and secondary settings
\csopt[\meta{marker}]{dual-sort-pad-plus},
\csopt[\meta{marker}]{secondary-sort-pad-plus}.

This option only has an effect when used with 
\csopt[\meta{number}]{sort-number-pad} where \meta{number} is
greater than 1. Positive numbers will have their sign replaced with 
\meta{marker}. The default setting is \csopt[>]{sort-pad-plus}.

\optsection{sort-pad-minus}

This option automatically implements the dual and secondary settings
\csopt[\meta{marker}]{dual-sort-pad-minus},
\csopt[\meta{marker}]{secondary-sort-pad-minus}.

This option only has an effect when used with 
\csopt[\meta{number}]{sort-number-pad} where \meta{number} is
greater than 1. Negative numbers will have their sign replaced with 
\meta{marker}. The default setting is \csopt[<]{sort-pad-plus}.

\optsection{identical-sort-action}

This option automatically implements the dual and secondary settings
\csopt[\meta{value}]{dual-identical-sort-action} and
\csopt[\meta{value}]{secondary-identical-sort-action}.

This option determines what the comparator should do if
two entries at the same hierarchical level are considered
equal.  The \meta{value} may be one of:
\begin{itemize}
\item\optfmt{none}: don't take any further action if sort values are
identical;
\item\optfmt{def} if sort values are identical, order them according to definition;
\item\optfmt{use}: if sort values are identical, order them
according to use in the document (order determine by a normal record);
\item\optfmt{id}: if sort values are identical, compare
the entry labels;
\item\optfmt{original id}: if sort values are identical, compare the
original unprefixed entry labels (as given in the \ext{bib} file);
\item\meta{field}: if sort values are identical, compare
the values from the given \meta{field}.
\end{itemize}
For the last three cases, a simple case-sensitive string comparison is used. 
If \meta{value} isn't a recognised keyword or valid field an error will occur.  The
default setting is \csopt[id]{identical-sort-action}.  If you're
using one of the sort rules listed in \tableref{tab:sortoptionsrule}
and you also want a locale-sensitive sort used on the fallback, then
you need to use \csopt{sort-suffix} instead.

\bibgls\ allows duplicate sort values, but this can cause a problem
for hierarchical entries where parent entries with duplicate sort
fields are clumped together and their children follow. To prevent
this from happening, the \csopt[id]{identical-sort-action}
setting will fallback on comparing the labels.  Since all labels
must be unique, this means comparisons between two different entries
are all either strictly higher or strictly lower.

This action occurs after any suffixes have been appended through
\csopt{sort-suffix}.

\optsection{sort-suffix}

This option automatically implements the dual and secondary settings
\csopt[\meta{value}]{dual-sort-suffix} and
\csopt[\meta{value}]{secondary-sort-suffix}.
The value may be one of:
\begin{itemize}
\item\optfmt{none}: don't append a suffix to any \field{sort} value;
\item\optfmt{non-unique}: append a numeric suffix to non-unique
  \field{sort} values;
\item\meta{field}: append the value of the given field (if set) to
  the \field{sort} field. The given field must be defined (has an
  associated key for use in \gls{newglossaryentry}) but may be unset.
  If the interpreter is on, the field contents will be interpreted.
  If the field is just a label (such as the \field{category} field)
  you may find it simpler to use
  \csopt[\meta{field}]{identical-sort-action} instead.
\end{itemize}
The default setting is \csopt[none]{sort-suffix}.

This option only affects the alphabetic
(\tableref{tab:sortoptionsrule}), letter
(\tableref{tab:sortoptionsletter}) and letter-number
(\tableref{tab:sortoptionsletternumber}) sort rules.  For the other
types of sort methods (not including the no-sort options listed in
\tableref{tab:sortoptionsnosort}) you'll need to use
\csopt{identical-sort-action} to prevent problems occurring with
duplicate sort values.

In the case of \csopt[non-unique]{sort-suffix}, this will only append a
suffix to the duplicate sort values (within the same hierarchical
level). The first sort value to be encountered isn't given a suffix.

The \csopt[\meta{field}]{sort-suffix} setting will only append a suffix
if that field is set, but (if set) it will apply the suffix to all
\field{sort} values, even those that are unique.

If you use \longarg{verbose}, then \bibgls\ will write information
in the transcript when it appends a suffix to the sort value.  The
message:
\begin{alltt}
Sort value '\meta{sort}' (entry '\meta{id}') not unique for the entry's
hierarchical level.
\end{alltt}
indicates that an entry with the given \meta{sort} value has already
been found within the same hierarchical level as the currently
processed entry (whose label is given by \meta{id}). The same
hierarchical level in this context means that either both entries
don't have a parent or both entries have the same parent. (That is,
the entries are considered siblings.)

This message will then be followed by:
\begin{alltt}
Appending suffix '\meta{suffix}' to the sort value '\meta{sort}' 
for entry '\meta{id}'.
\end{alltt}
which indicates that the entry (identified by the label \meta{id})
has been assigned the sort value given by \meta{sort}\meta{suffix}.
If any break markers are applied, this is done after the suffix has
been appended.

For example, suppose in my document I want to write about \appfmt{makeglossaries}
(the application) and \csfmt{makeglossaries} (the command). I might
decide to define semantic commands:
\begin{codeenv}
\cs{newcommand}*\marg{\cmd{application}}[1]\marg{\ics{texttt}\marg{\idx{param}1}}
\cs{newcommand}*\marg{\cmd{command}}[1]\marg{\cs{texttt}\marg{\cs{glsbackslash} \idx{param}1}}
\end{codeenv}
In my \ext{bib} file I might have:
\begin{codeenv}
\atentry{entry}\marg{cs.makeglossaries,
  \field{name}=\marg{\cmd{command}\marg{makeglossaries}},
  \field{category}=\marg{command},
  \field{description}=\marg{opens glossary files}
}
\strut
\atentry{entry}\marg{ap.makeglossaries,
  \field{name}=\marg{\cmd{application}\marg{makeglossaries}},
  \field{category}=\marg{application},
  \field{description}=\marg{Perl script}
}
\end{codeenv}
If \bibgls\ is provided with the definitions of \csfmt{application}
and \csfmt{command} (by interpreting the \atentry{preamble} or a
package provided with \longarg{custom-packages}) then it will
determine that the sort value for \code{cs.makeglossaries} is
\csfmt{makeglossaries} and the sort value for
\code{ap.makeglossaries} is just \code{makeglossaries}. These are 
two distinct sort values from \bibgls's point of view although the
sort rule may consider them identical if the rule ignores the
\idx{escchar} character (such as the locale sort methods), in which case,
\bibgls\ will then act according to \csopt{identical-sort-action}.

If \bibgls\ isn't provided with these custom definitions, then it
will ignore those semantic commands and both entries will end up with
the sort value \code{makeglossaries}. The second instance will be
recognised as a duplicate and the sort value will be converted to
\code{makeglossaries1} (where the automated suffix is \code{1} and
the suffix marker, see below, is the empty string). Whereas with,
say, \csopt[.]{sort-suffix-marker}\ then the sort value would become
\code{makeglossaries.1}.

For comparison, consider the following document:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[\styopt[indexgroup]{style}]\marg{glossaries}
\strut
\ics{cs.makeglossaries}
\strut
\cs{newcommand}*\marg{\cmd{application}}[1]\marg{\cs{texttt}\marg{\idx{param}1}}
\cs{newcommand}*\marg{\cmd{command}}[1]\marg{\cs{texttt}\marg{\cs{glsbackslash} \idx{param}1}}
\strut
\gls{newglossaryentry}\marg{cs.makeglossaries}\marg{\comment{}
  \field{name}=\marg{\cmd{command}\marg{makeglossaries}},
  \field{description}=\marg{opens glossary files}}
\strut
\gls{newglossaryentry}\marg{ap.makeglossaries}\marg{\comment{}
  \field{name}=\marg{\cmd{application}\marg{makeglossaries}},
  \field{description}=\marg{Perl script}}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{cs.makeglossaries} and \cs{gls}\marg{ap.makeglossaries}.
\ics{printglossaries}
\cmd{end}\marg{document}
\end{codeenv}

This uses \idx{makeindex}, which puts both entries in the
\qt{Symbols} group (since they both start with \idx{escchar} from the
start of \csfmt{command} and \csfmt{application}, respectively).
The ordering is \appfmt{makeglossaries}, \csfmt{makeglossaries}
because \qt{a} (second character of \csfmt{application}) comes
before \qt{c} (second character of \csfmt{command}).

The switch to \idx{xindy} just involves adding the \styopt{xindy}
package option:
\begin{codeenv}
\cs{usepackage}[\styopt{xindy},\styopt[indexgroup]{style}]\marg{glossaries}
\end{codeenv}
This results in a glossary that only contains one entry,
\csfmt{makeglossaries}, because \idx{xindy} merges entries with
duplicate sort values and the sort values end up as duplicates
because \idx{xindy} discards the \csfmt{application} and
\csfmt{command} control sequences. Although \bibgls\ also ignores
unknown control sequences, it doesn't perform this merger. 

If I add: 
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}*\marg{\cmd{application}}[1]\marg{\cs{texttt}\marg{\idx{param}1}}
\cs{providecommand}\marg{\cmd{command}}[1]\marg{\cs{texttt}\marg{\cs{glsbackslash} \idx{param}1}}}}
\end{codeenv}
to the earlier \ext{bib} file (called, say, \filefmt{entries.bib})
then the document can be altered to use \bibgls:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[\styopt{record},\styopt[indexgroup]{style}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries.bib]{src},
  \csopt[non-unique]{sort-suffix},
  \csopt[none]{identical-sort-action}
}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{cs.makeglossaries} and \cs{gls}\marg{ap.makeglossaries}.
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
This uses the default \csopt[locale]{sort} which considers \idx{escchar}
an ignored (punctuation) character, so both \csfmt{makeglossaries} and
\appfmt{makeglossaries} are listed in the \qt{M} letter group, even
though the interpreter has determined that the sort value for
\code{cs.makeglossaries} is the literal string \csfmt{makeglossaries}. 
Note that in this case \bibgls\ doesn't detect duplicate sort values
since it only uses a simple string comparison to detect duplicates
rather than using the collator.

If I switch to using a letter-based sort rule instead, for example
\csopt[letter-nocase]{sort}, then \csfmt{makeglossaries} will be
listed in the \qt{Symbols} letter group since the leading \idx{escchar}
from the sort value \csfmt{makeglossaries} isn't ignored with this
rule.

Now let's suppose I use \csopt[false]{interpret-preamble} to prevent 
\bibgls\ from interpreting the preamble:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries.bib]{src},\csopt[false]{interpret-preamble}}
\end{codeenv}
This means that the custom commands won't be recognised and will
therefore be ignored, so both entries will have their sort values reduced to 
\code{makeglosssaries}.

The first entry to be processed is \code{cs.makeglossaries} because
it's the first to be selected. This is assigned the sort value
\code{makeglossaries}. (Note that, unless you use
\csopt[unsrt]{sort}, the initial selection order is based on the record
order. In this example, \code{cs.makeglossaries} has the first
record in the \ext{aux} file.)

The next entry to be processed is \code{ap.makeglossaries}. This
also ends up with the sort value \code{makeglossaries} so \bibgls\
converts this to \code{makeglossaries1} and (with verbose mode on)
the following messages are written to the transcript:
\begin{verbatim}
Sort value 'makeglossaries' (entry 'ap.makeglossaries') not unique
for the entry's hierarchical level.
Appending suffix '1' to the sort value 'makeglossaries' for entry
'ap.makeglossaries'.
\end{verbatim}
Both entries are listed in the \qt{M} letter group in the order
\csfmt{makeglossaries}, \appfmt{makeglossaries}.

If the records are reversed:
\begin{codeenv}
\cs{gls}\marg{ap.makeglossaries} and \cs{gls}\marg{cs.makeglossaries}.
\end{codeenv}
then the sort value for \code{cs.makeglossaries} is now considered
the duplicate and the order is reversed: \appfmt{makeglossaries},
\csfmt{makeglossaries}.

Suppose now I modify the \ext{bib} file so that
\code{ap.makeglossaries} is defined as:
\begin{codeenv}
\atentry{entry}\marg{ap.makeglossaries,
  \field{name}=\marg{\cmd{application}\marg{makeglossaries}},
  \field{category}=\marg{application},
  \field{description}=\marg{Perl script (must be used with \cs{gls}\marg{cs.makeglossaries})}
}
\end{codeenv}
and suppose the document only contains an explicit reference to
\code{ap.makeglossaries}:
\begin{codeenv}
\cmd{begin}\marg{document}
\cs{gls}\marg{ap.makeglossaries}
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
Now \code{ap.makeglossaries} is the first entry to be selected
because entries with records are always selected before any
(unrecorded) dependencies. In this case \code{cs.makeglossaries} is
only selected because it's required by \code{ap.makeglossaries}. Now
\code{ap.makeglossaries} is the first to have its sort value
assigned, and it's \code{cs.makeglossaries} that has the duplicate.
This means that the ordering in the glossary is now:
\appfmt{makeglossaries}, \csfmt{makeglossaries}.

An oddity occurs if the glossary is moved to the start of the
document:
\begin{codeenv}
\cmd{begin}\marg{document}
\cs{printunsrtglossaries}
\cs{gls}\marg{ap.makeglossaries}
\cmd{end}\marg{document}
\end{codeenv}
In this case, the first document build:
\begin{verbatim}
pdflatex myDoc
bibgls --group --verbose myDoc
pdflatex myDoc
\end{verbatim}
leads to the ordering described above: 
\appfmt{makeglossaries}, \csfmt{makeglossaries}.
However, the next document build has a new record for
\code{cs.makeglossaries} occurring in the glossary (within the
description of \code{ap.makeglossaries}) which means it's now the
first entry to be selected so the ordering switches to:
\csfmt{makeglossaries}, \appfmt{makeglossaries}.
In this type of situation you might be better off with the
\csopt[id]{identical-sort-action} option instead.

Remember that you can temporarily switch off the indexing by locally
setting:
\begin{codeenv*}
\ics{GlsXtrSetDefaultGlsOpts}\marg{\glsopt{noindex}}
\end{codeenv*}
Since the glossary preamble is scoped, you can simply do
\begin{codeenv}
\ics{appto}\ics{glossarypreamble}\marg{\cs{GlsXtrSetDefaultGlsOpts}\marg{\glsopt{noindex}}}
\end{codeenv}
to switch off the indexing within the glossary (or use
\ics{apptoglossarypreamble}). Note that this is
different to using:
\begin{codeenv}
\ics{GlsXtrSetDefaultNumberFormat}\marg{\encap{glsignore}}
\end{codeenv}
which creates an \idx{ignoredrecord}. Even though the record is ignored
(and so won't show in the location list) the record still influences
the selection order and the record count.

\optsection{sort-suffix-marker}

This automatically implements the dual and secondary settings
\csopt[\meta{value}]{dual-sort-suffix-marker} and
\csopt[\meta{value}]{secondary-sort-suffix-marker}.

If a suffix is appended to the sort value (see above) then it will
be separated by the suffix marker, which can be set with
\csopt[\meta{value}]{sort-suffix-marker} where \meta{value} is the
marker. By default the marker is empty. You can use \stringu\meta{hex}
or \cs{glshex}\meta{hex} to indicate Unicode characters outside the
\idx{ASCII} range. If, for some reason, you want to use a special
character, such as \idx{param}, you will need to precede it with
\ics{cs.string} (for example \code{\cs{cs.string}\idx{param}}) or
use the above hexadecimal markup. If you use \ics{cs.hash} it will
be treated as a literal string containing a backslash followed by a
hash character.

\optsection{strength}

This option automatically implements
\csopt[\meta{value}]{dual-strength} and
\csopt[\meta{value}]{secondary-strength}.

The collation strength used by the alphabet sort methods 
(\tableref{tab:sortoptionsrule}) can be set
to the following values: \optfmt{primary} (default),
\optfmt{secondary}, \optfmt{tertiary} or \optfmt{identical}. These
indicate the difference between two characters, but the exact
assignment is locale dependent. See the documentation for Java's
\href{http://docs.oracle.com/javase/8/docs/api/java/text/Collator.html}{\code{Collator}
class}~\cite{collator} for further details.

For example, suppose the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{index}\marg{resume}
\atentry{index}\marg{RESUME}
\atentry{index}\marg{resumee, \field{name}=\marg{r\ics{acute}esum\cs{acute}e}}
\atentry{index}\marg{rat}
\atentry{index}\marg{rot}
\atentry{index}\marg{aardvark}
\atentry{index}\marg{zoo}
\end{codeenv}
and the document contains:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[\styopt{record}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[en]{sort},\csopt[entries]{src}}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{resumee}, \cs{gls}\marg{resume}, \cs{gls}\marg{RESUME},
\cs{gls}\marg{aardvark}, \cs{gls}\marg{rat}, \cs{gls}\marg{rot}, \cs{gls}\marg{zoo}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
then this uses the default \csopt[primary]{strength}, so the entries
are listed as aardvark, rat, r\'esum\'e, resume, RESUME, rot, zoo.

If the strength is changed to \optfmt{secondary}:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[en]{sort},\csopt[entries]{src},\csopt[secondary]{strength}}
\end{codeenv}
then the entries are listed as aardvark, rat, resume, RESUME,
r\'esum\'e, rot, zoo.

If the strength is changed to \optfmt{tertiary} or
\optfmt{identical}, there's no difference from
\csopt[secondary]{strength} for this particular example.

This option is ignored by non-alphabet sorts (such as letter or numeric).

\optsection{decomposition}

This option automatically implements the dual and secondary settings
\csopt[\meta{value}]{dual-decomposition} and
\csopt[\meta{value}]{secondary-decomposition}.

The collation decomposition used by alphabet sort methods
(\tableref{tab:sortoptionsrule}) can be set
to the following values: \optfmt{canonical} (default),
\optfmt{full} or \optfmt{none}. This determines how Unicode composed
characters are handled. The fastest mode is \optfmt{none} but is
only appropriate for languages without accents. The slowest mode is
\optfmt{full} but is the most complete for languages with \idx{non-ASCII}
characters. See the documentation for Java's
\href{http://docs.oracle.com/javase/8/docs/api/java/text/Collator.html}{\code{Collator}
class}~\cite{collator} for further details.
This option is ignored by non-alphabet sorts (such as letter or numeric).

\optsection{letter-number-rule}

This automatically implements the dual and secondary settings
\csopt[\meta{value}]{dual-letter-number-rule}
and
\csopt[\meta{value}]{secondary-letter-number-rule}.

If you use one of the letter-number sort methods
(\tableref{tab:sortoptionsletternumber}), then you can determine the
comparison between a number and letter. The \meta{value} may be
one of:
\begin{itemize}
\item\optfmt{before letter}: numbers are considered less than any
letter.

\item\optfmt{after letter}: numbers are considered greater than any
letter.

\item\optfmt{between}: (default) numbers come between letter cases.
With the \optfmt{letternumber\dhyphen case} sort option, this will put
numbers after \idx{uppercase} and before \idx{lowercase}.  This
setting doesn't make much sense with the
\optfmt{letternumber\dhyphen nocase} option but, if used, this will put
numbers before letters. The \optfmt{letternumber\dhyphen upperlower} and
\optfmt{letternumber\dhyphen lowerupper} options are more complicated. See 
\sectionref{sec:letternumber} for more detail.

\item\optfmt{first}: numbers are considered less than all
characters (including punctuation and spaces).

\item\optfmt{last}: numbers are considered greater than all
characters (including punctuation and spaces).
\end{itemize}
Note that the reverse sort methods will invert this setting.
Remember also that the case-insensitive letter-number sort methods always
first convert the \field{sort} field to \idx{lowercase}, which means that
if you use one of them then there won't be any \idx{uppercase}
characters. 

Use \csopt{letter-number-punc-rule} to determine the relative
position of white space and punctuation.

\optsection{letter-number-punc-rule}

This automatically implements the dual and secondary 
\csopt[\meta{value}]{dual-letter-number-punc-rule}
and
\csopt[\meta{value}]{secondary-letter-number-punc-rule}.

If you use one of the letter-number sort methods
(\tableref{tab:sortoptionsletternumber}), then you can determine the
order of white space and punctuation. In this context, punctuation
means any character that's not considered a letter, a number
or white space. This means that characters such as combining marks
are considered punctuation.

The \meta{value} may be one of the following:
\begin{itemize}
\item\optfmt{punc-space-first}: punctuation comes first, followed by
white space (then letters and optionally numbers according to the letter-number
rule);
\item\optfmt{punc-space-last}: punctuation followed by white space
come last (after letters and optionally numbers according to the letter-number
rule);
\item\optfmt{space-punc-first}: white space comes first, followed by
punctuation (then letters and optionally numbers according to the letter-number
rule);
\item\optfmt{space-punc-last}: white space followed by punctuation
come last (after letters and optionally numbers according to the letter-number
rule); 
\item\optfmt{space-first-punc-last}: white space comes first
(followed by letters and optionally numbers according to the letter-number
rule) and punctuation comes last; 
\item\optfmt{punc-first-space-last}: punctuation comes first
(followed by letters and optionally numbers according to the letter-number
rule) and white space comes last;
\item\optfmt{punc-first-space-zero}: punctuation comes first
(although numbers may come before)
and white space is replaced by the digit~\code{0} (\hex{30});
\item\optfmt{punc-last-space-zero}: punctuation comes last
(although numbers may come after)
and white space is replaced by the digit~\code{0} (\hex{30}).
\item\optfmt{punc-first-space-zero-match-next}: punctuation comes first
(although numbers may come before)
and white space is replaced by the appropriate zero character (see below);
\item\optfmt{punc-last-space-zero-match-next}: punctuation comes last
(although numbers may come after)
and white space is replaced by the appropriate zero character (see below).
\end{itemize}
Remember that the reverse sort methods will invert order governed by this setting.

For the \optfmt{space-zero-match-next} settings, the sort value will
have all spaces replaced with a digit that represents zero. If the
space isn't followed by a digit, the basic Latin \code{0} (\hex{30})
will be used, otherwise \bibgls\ will try to match the zero with the
following digit group. For example, if the space is followed by
\textsuperscript{1} (\hex{B9}) the space will be replaced by
\textsuperscript{0} (\hex{2070}),  resulting in the sub-string
\code{\textsuperscript{01}} (\hex{B9} \hex{2070}). 

If just the \optfmt{space-zero}
(without the \optfmt{-match-next}) is used then the space will just
be replaced with \code{0} resulting in the sub-string
\code{0\textsuperscript{1}} (\hex{30} \hex{2070}). In this case, the
\code{0} will be distinct from \textsuperscript{1} (rather than
being considered a leading zero). However, for 
other numbering systems the \code{0} will be treated as a leading
zero. For example, if the space is followed by the Devanagari digit
one (\hex{0967}) then the sub-string will be \hex{30} \hex{0967} but here the
mixture is allowed to form a number (with a leading zero) as both
characters belong to the Unicode category \idx{numberdecimaldigit}.

This means that the \optfmt{-match-next} settings are only really
needed if the sort string contains the superscript or subscript
digits that don't belong to the \qt{Number, Decimal Digit} category.
The plain \optfmt{space-zero} alternatives are more efficient as
they just perform a simple substitution.

The \TeX\ parser library used by \bibgls\ recognises the standard
\LaTeX\ text-mode commands \ics{textsuperscript}\margm{text} and 
\ics{textsubscript}\margm{text}
and will use the Unicode superscript or subscript characters if they cover every
character in \meta{text}, otherwise HTML markup is used, but that's
then stripped by \bibgls. This means that:
\begin{codeenv}
C\cs{textsubscript}\marg{10}H\cs{textsubscript}\marg{10}O\cs{textsubscript}\marg{4}
\end{codeenv}
will be converted to:
\code{C\textsubscript{10}H\textsubscript{10}O\textsubscript{4}} but:
\begin{codeenv}
X\cs{textsubscript}\marg{1, 2}
\end{codeenv}
will be converted to:
\begin{verbatim}
X<sub>1, 2</sub>
\end{verbatim}
which ends up as \code{X1, 2}.


Note that \csopt[first]{letter-number-rule} and
\csopt[last]{letter-number-rule} overrides this option when
comparing a number with white space or punctuation.

\optsection{numeric-sort-pattern}

If you use the custom \csopt[numberformat]{sort} or
\csopt[numberformat-reverse]{sort}, you need to 
specify the format pattern with this option where
\meta{value} is a pattern recognised by Java's
\href{http://docs.oracle.com/javase/8/docs/api/java/text/DecimalFormat.html}{\code{java.text.DecimalFormat}}
class~\cite{decimalformat}. You can use \stringu\meta{hex} or
\cs{glshex}\meta{hex} to indicate Unicode characters by their
hexadecimal code. You can also use \ics{cs.hash}, \ics{cs.percent},
\ics{cs.underscore}, \ics{cs.amp}, \ics{cs.openbrace} and
\ics{cs.closebrace} to indicate \idx{hashchar}, \idx{percentchar},
\idx{underscorechar}, \idx{ampchar}, \idx{openbracechar} and
\idx{closebracechar}. 

Where the dual or secondary sort uses \optfmt{numberformat}
or \optfmt{numberformat-reverse}, use \csopt{dual-numeric-sort-pattern} for 
\csopt{dual-sort} and \csopt{secondary-numeric-sort-pattern} 
for \csopt{secondary}.

\optsection{numeric-locale}

If you use any of the locale-sensitive numeric sort methods
described in \sectionref{sec:numerical},
such as \csopt[numeric]{sort}, use this option to set
the locale. The value may be:
\begin{itemize}
\item\optfmt{locale}: use Java's default locale (which is usually
the operating system's locale);
\item\optfmt{doc}: use the document's locale or, if not set, assume
\csopt[locale]{numeric-locale};
\item\meta{lang-tag}: set to the locale identified by the given
a valid language tag \meta{lang-tag}.
\end{itemize}
Use \csopt{dual-numeric-locale} for 
\csopt{dual-sort} and \csopt{secondary-numeric-locale} 
for \csopt{secondary}.

\optsection{date-sort-locale}

If you use a date/time sort method (\tableref{tab:sortoptionsdatetime}), 
then you can set the locale used by Java's date-time parser.
The default setting is \csopt[locale]{date-sort-locale}.

The value may be \optfmt{locale} (use Java's default locale),
\optfmt{doc} (use the document's locale) or a valid language tag
\meta{lang-tag} identifying the locale.

Use \csopt{dual-date-sort-locale} and \csopt{secondary-date-sort-locale}
for the dual and secondary.

\optsection{date-sort-format}

If you use a date/time sort method (\tableref{tab:sortoptionsdatetime}), 
then you can set the format used by Java's date-time parser. 
If omitted, \csopt[default]{date-sort-format} is assumed.
The \meta{value} may be one of:
\begin{itemize}
\item\optfmt{default}: use the locale's default format.
\item\optfmt{short}: use the locale's short format.
\item\optfmt{medium}: use the locale's medium format.
\item\optfmt{long}: use the locale's long format.
\item\optfmt{full}: use the locale's full format.
\item\meta{pattern}: provide a custom pattern.
This should match the specifications for Java's
\href{http://docs.oracle.com/javase/8/docs/api/java/text/SimpleDateFormat.html}{\code{SimpleDateFormat}}
class~\cite{simpledateformat}. You may use \stringu\meta{hex} or
\code{\cs{glshex} \meta{hex}} to indicate Unicode characters
or \ics{cs.hash}, \ics{cs.percent}, \ics{cs.underscore}, \ics{cs.amp},
\ics{cs.openbrace} and \ics{cs.closebrace} to indicate
\idx{hashchar}, \idx{percentchar}, \idx{underscorechar},
\idx{ampchar}, \idx{openbracechar} and~\idx{closebracechar}.
\end{itemize}

With the custom setting, if the pattern only contains date (but not
time) information, then it must be used with \csopt[date]{sort}
or \csopt[date-reverse]{sort}. If the pattern only contains time
(but not date) information, then it must be used with
\csopt[time]{sort} or \csopt[time\dhyphen reverse]{sort}. If the pattern
contains date and time information, then it must be used with
\csopt[datetime]{sort} or \csopt[datetime-reverse]{sort}.

For example, suppose each entry provides information about a person
and the \field{user1} field is used to store their date of birth:
\begin{codeenv}
\atentry{entry}\marg{caesar,
  \field{name}=\marg{Gaius Julius Caesar},
  \field{first}=\marg{Julius Caesar},
  \field{text}=\marg{Caesar},
  \field{description}=\marg{Roman politician and general},
  \field{user1}=\marg{13 July 100 BC}
}
\strut
\atentry{entry}\marg{wellington,
  \field{name}=\marg{Arthur Wellesley, 1st Duke of Wellington},
  \field{first}=\marg{Arthur Wellesley (Duke of Wellington)},
  \field{text}=\marg{Wellington},
  \field{description}=\marg{Anglo-Irish soldier and statesman},
  \field{user1}=\marg{1 May 1769 AD}
}
\end{codeenv}
Then the entries can be sorted by date of birth using:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{data in entries.bib}
  \csopt[user1]{sort-field},
  \csopt[date]{sort},
  \csopt[d MMM y G]{date-sort-format}
}
\end{codeenv}

The \code{G} (era) date pattern specifier expects a string, such as
\qt{AD}. It will match \idx{lowercase} forms, such as \qt{ad}, so if you
have \code{\cs{textsc}\marg{ad}} the interpreter will convert this to
\code{ad} (stripping the text-block command). However, in general
it's best to supply a semantic command that ensures that the
interpreted result matches the required format.

For example, if \csfmt{era} is provided with:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cmd{era}}[1]\marg{\cs{textsc}\marg{\cs{MakeLowercase}\marg{\idx{param}1}}}}}
\end{codeenv}
If the definition is hidden from the interpreter
(\csopt[false]{interpret-preamble}) and the field
value contains \verb|\era{AD}| then the custom command will simply be stripped
leaving \code{AD} which can be matched by \code{G}. 

If the definition is picked up by the interpreter then the field
value will contain \code{ad} (from \ics{MakeLowercase}) but this can
be matched by \code{G}, so it isn't a problem.
However, if the definition of \csfmt{era} is changed so that the era label
supplied in the argument is converted to something that doesn't
match \code{G} then the definition should be hidden from the
interpreter.

Here's a complete document that changes the \field{group} fields to use
the year and era:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[\styopt{record},\styopt[indexgroup]{style}]\marg{glossaries-extra}
\strut
\cs{newcommand}\marg{\gls{bibglsdategroup}}[7]\marg{\idx{param}1\idx{param}4\idx{param}7}
\cs{newcommand}\marg{\gls{bibglsdategrouptitle}}[7]\marg{\ics{number}\idx{param}1\cs{cs.space}\idx{param}4}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},
  \csopt[user1]{sort-field},
  \csopt[date]{sort},
  \csopt[d MMM y G]{date-sort-format},
  \csopt[all]{selection}
}
\strut
\cmd{begin}\marg{document}
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
(The use of \ics{number} strips the leading zero from the year.)

\optsection{group-formation}

If the \field{group} field hasn't been set in
the \ext{bib} file or through options like \csopt{group}, then it is 
assigned according to this option's setting during sorting if
\longarg{group} has been used. Permitted values:
\begin{itemize}
\item\optfmt{default}: the group is assigned according to 
the sort method's default group formation. This is the default
setting.

\item\optfmt{codepoint}: the group is set to
\format{bibglsunicodegroup}, where the first argument is the first
significant character (converted to \idx{lowercase} and decomposed,
if applicable) of the sort value.

\item\optfmt{unicode category}: the group is set to
\format{bibglsunicodegroup}, where the first argument is the label
identifying the Unicode category of the first significant character
of the sort value. For example, the label \code{Ll} signifies a
\idx!{lowercase} letter and \code{Lu} signifies an \idx!{uppercase}
letter.

\item\optfmt{unicode script}: the group is set to
\format{bibglsunicodegroup}, where the first argument is the
label identifying the Unicode script of the first significant
character of the sort value. For example, the label \code{LATIN}
indicates Latin, \code{GREEK} indicates Greek and \code{COMMON}
indicates common characters (such as mathematical Greek characters
that are often used with non-Greek scripts).

\item\optfmt{unicode category and script}: the group is set to
\format{bibglsunicodegroup}, where the first argument is the
label corresponding to the Unicode category and script of the first
significant character of the sort value. For example, the label
\code{Ll.LATIN} indicates a \idx!{lowercase} Latin letter.
\end{itemize}
This option has no effect with \longarg{no-group} or if no sorting
is applied. Use \csopt{secondary-group-formation} for secondary
sorting and \csopt{dual-group-formation} for dual entries.
Settings other than the default can cause the groups to become
fragmented, so care is needed if you use this option.
See also \sectionref{sec:logicaldivisions}.

\section{Secondary Glossary}
\label{sec:secondaryopts}

The secondary glossary may only be used with \csopt[define]{action}
(within the same resource set) since it's incompatible with the copy actions.
You may use \csopt{secondary} in the first resource set and a copy
action in a subsequent resource set.

\optsection{secondary}

It may be that you want to display a glossary twice but with a
different order. For example, the first time alphabetically and the
second time by category.  One way to do this is to have two
\gls{GlsXtrLoadResources} that both load the same \ext{bib} file with
different \csopt{label-prefix} and \csopt{sort} settings, but this
is only possible with \csopt[all]{selection} or by ensuring you
reference each entry with both label prefixes. Another method is
to use \csopt[copy]{action} but this requires a second resource
command with the same selection criteria.

A simpler method is to use a single \gls{GlsXtrLoadResources}
with the \csopt{secondary} option. The value
(which must be supplied) should be in the format:
\begin{definition}
\meta{sort}:\meta{field}:\meta{type}
\end{definition}
or
\begin{definition}
\meta{sort}:\meta{type}
\end{definition}
If the \meta{field} is omitted, the value of \csopt{sort-field} is
used. Remember that when the primary entries are sorted, the
\field{sort} field will be set, which means that the fallback field
(such as \field{name}) won't be used in the secondary sort. In
general it's best to supply the field unless one type is sorted and the
other isn't. (The actual sort value obtained by the secondary sort
will be saved in the \field{secondarysort} field in case you require it.)

The value of \meta{sort} is as for \csopt{sort}, but note
that in this case the sort value \optfmt{unsrt} or \optfmt{none}
means to use the same ordering as the primary entries. For
example, with \csopt[de-CH-1996]{sort},
\csopt[none:copies]{secondary} the \code{copies} list will be
ordered according to \code{de-CH-1996} and not according to the
order in which they were read when the \ext{bib} file or files were
parsed.  If \meta{sort} is \optfmt{custom}, then the rule should be
provided with \csopt{secondary-sort-rule}.

This option will copy all the selected entries into the glossary labelled
\meta{type} sorted according to \meta{sort} (using \meta{field} as
the sort value). Note that this \emph{just copies the entry's label}
to the second glossary list rather than creating a duplicate entry,
which saves resources but it means that all the fields will be
identical. If you want groups in your glossary, the group
information for the secondary glossary will be stored in the
internal \field{secondarygroup} field. The \field{group} field will
contain the group for the primary glossary.

In order to switch fields in \ics{printunsrtglossary}, you need at
least v1.21 of \sty{glossaries-extra} which provides
\ics{glsxtrgroupfield} to keep track of the appropriate field label.
If this command is defined, the preamble for the secondary glossary 
will be adjusted to locally change the field to
\field{secondarygroup}. With older versions, the group information
in the secondary glossary will be the same as for the primary
glossary.

(If the glossary \meta{type} doesn't exist, it will be
defined with \ics{provideignoredglossary*}\margm{type} 
even if \longarg{no-provide-glossaries} is set.)
Note that if the glossary already exists and contains entries,
the existing entries aren't re-ordered. The new entries are
simply appended to the list.

For example, suppose the \ext{bib} file contains entries like:
\begin{codeenv}
\atentry{entry}\marg{quartz,
  \field{name}=\marg{quartz},
  \field{description}=\marg{hard mineral consisting of silica},
  \field{category}=\marg{mineral}
}
\strut
\atentry{entry}\marg{cabbage,
  \field{name}=\marg{cabbage},
  \field{description}=\marg{vegetable with thick green or purple leaves},
  \field{category}=\marg{vegetable}
}
\strut
\atentry{entry}\marg{waterfowl,
  \field{name}=\marg{waterfowl},
  \field{description}=\marg{any bird that lives in or about water},
  \field{category}=\marg{animal}
}
\end{codeenv}
and the document preamble contains:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[en-GB]{sort},
  \csopt[en-GB:category:topic]{secondary}
}
\end{codeenv}
This sorts the primary entries according to the default
\csopt{sort-field} and then sorts the entries according
to the \field{category} field and copies this list to
the \code{topic} glossary (which will be provided if not defined.)

The secondary list can be displayed with the hypertargets switched
off to prevent duplicates. The cross-references will link to the
original glossary.

For example:
\begin{codeenv*}
\ics{printunsrtglossary}\oarg{\printglossopt[Summary (alphabetical)]{title}}
\ics{printunsrtglossary}\oarg{\printglossopt[Summary (by topic)]{title},\printglossopt[false]{target}}
\end{codeenv*}

The alternative (or if more than two lists are required) is to
reload the same \ext{bib} file with different label prefixes.
For example, if the entries are stored in \filefmt{entries.bib}:
\begin{codeenv}
\cs{newglossary*}\marg{nosort}\marg{Symbols (Unsorted)}
\cs{newglossary*}\marg{byname}\marg{Symbols (Letter Order)}
\cs{newglossary*}\marg{bydesc}\marg{Symbols (Ordered by Description)}
\cs{newglossary*}\marg{byid}\marg{Symbols (Ordered by Label)}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{entries.bib}
  \csopt[unsrt]{sort},
  \csopt[nosort]{type}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{entries.bib}
  \csopt[letter-case]{sort},
  \csopt[byname]{type},
  \csopt[byname.]{label-prefix}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{entries.bib}
  \csopt[locale]{sort},
  \csopt[description]{sort-field},
  \csopt[bydesc]{type},
  \csopt[bydesc.]{label-prefix}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[entries]{src},\comment{entries.bib}
  \csopt[letter]{sort},
  \csopt[id]{sort-field},
  \csopt[byid]{type},
  \csopt[byid.]{label-prefix}
}
\end{codeenv}

\optsection{secondary-missing-sort-fallback}

As \csopt{missing-sort-fallback} but for secondary sorting.

\optsection{secondary-trim-sort}

As \csopt{trim-sort} but for secondary sorting.

\optsection{secondary-sort-replace}

As \csopt{sort-replace} but for secondary sorting.

\optsection{secondary-sort-rule}

As \csopt{sort-rule} but for secondary custom sorting.

\optsection{secondary-break-at}

As \csopt{break-at} but for secondary entries.

\optsection{secondary-break-marker}

As \csopt{break-marker} but for secondary entries.

\optsection{secondary-sort-number-pad}

As \csopt{sort-number-pad} but for secondary entries.

\optsection{secondary-sort-pad-plus}

As \csopt{sort-pad-plus} but for secondary entries.

\optsection{secondary-sort-pad-minus}

As \csopt{sort-pad-minus} but for secondary entries.

\optsection{secondary-identical-sort-action}

As \csopt{identical-sort-action} but for secondary entries.

\optsection{secondary-sort-suffix}

As \csopt{sort-suffix} but for secondary entries.

\optsection{secondary-sort-suffix-marker}

As \csopt{sort-suffix-marker} but for secondary entries.

\optsection{secondary-strength}

As \csopt{strength} but for secondary entries.

\optsection{secondary-decomposition}

As \csopt{decomposition} but for secondary entries.

\optsection{secondary-letter-number-rule}

As \csopt{letter-number-rule} but for secondary letter-number sorting.

\optsection{secondary-letter-number-punc-rule}

As \csopt{letter-number-punc-rule} but for secondary letter-number sorting.

\optsection{secondary-numeric-sort-pattern}

As \csopt{numeric-sort-pattern} but for secondary
locale-sensitive numeric sorting.

\optsection{secondary-numeric-locale}

As \csopt{numeric-locale} but for secondary
locale-sensitive numeric sorting.

\optsection{secondary-date-sort-locale}

As \csopt{date-sort-locale} but for secondary date-time sorting.

\optsection{secondary-date-sort-format}

As \csopt{date-sort-format} but for secondary date-time sorting.

\optsection{secondary-group-formation}

As \csopt{group-formation} but for secondary sorting.

\section{Dual Entries}
\label{sec:dualopts}

\subsection{General Dual Settings}
\label{sec:dualoptsgeneral}

\optsection[\subsubsection]{dual-prefix}

This option indicates the prefix to use for the dual entries. The
default value is \idprefix{dual} (including the terminating period). 
Any references to dual entries within the \ext{bib} file should use
the prefix \idprefix{dual} which will be replaced by \meta{value}
when the \ext{bib} file is parsed.

As from version 1.8, the dual label prefix is identified
in the \ext{glstex} file with:
\begin{codeenv}
\format{bibglsdualprefixlabel}
\end{codeenv}

\optsection[\subsubsection]{primary-dual-dependency}

This is a boolean setting that determines whether or not primary and
dual entries should be considered mutual dependencies. The default value is
\csopt[true]{primary-dual-dependency}, which means that if a primary
has records then the dual is added as a dependency and vice versa.
The setting \csopt[false]{primary-dual-dependency} can't be used
with \csopt[none]{dual-sort} or \csopt[use]{dual-sort} (but may be
used with \csopt[combine]{dual-sort} and \csopt[none]{sort} or
\csopt[use]{sort}).

\optsection[\subsubsection]{combine-dual-locations}

This setting allows the location lists for each primary entry
to be merged with that of the corresponding dual entry.
The \meta{value} may be one of:
\begin{itemize}
\item\optfmt{false} This is the default setting. The location lists
aren't combined.
\item\optfmt{both} Both the primary and dual are given the combined
location list.
\item\optfmt{dual} Only the dual is given the combined location
list. The primary's location list is emptied.
\item\optfmt{primary} Only the primary is given the combined
location list. The dual's location list is emptied.
\end{itemize}

For example, suppose the file \filefmt{entries.bib} contains:
\begin{codeenv}
\atentry{dualindexentry}\marg{array,
  \field{description}=\marg{ordered list of values}
}
\strut
\atentry{dualindexentry}\marg{vector,
  \field{name}=\marg{vector},
  \field{description}=\marg{column or row of values}
}
\strut
\atentry{dualindexentry}\marg{set,
  \field{description}=\marg{collection of values}
}
\strut
\atentry{dualindexentry}\marg{matrix,
  \field{plural}=\marg{matrices},
  \field{description}=\marg{rectangular array of values}
}
\end{codeenv}
and the document contains:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[colorlinks]\marg{hyperref}
\cmd{usepackage}[\styopt{record},\styopt{index},\styopt[indexgroup]{style}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries]{src},
 \csopt[index]{type},
 \csopt[idx.]{label-prefix},
 \csopt[gls.]{dual-prefix},
 \csopt[main]{dual-type}
}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{gls.array}, \cs{gls}\marg{gls.vector}, \cs{gls}\marg{gls.set}, \cs{gls}\marg{gls.matrix}.
\strut
\cmd{newpage}
\cs{gls}\marg{gls.array}, \cs{gls}\marg{idx.vector}, \cs{gls}\marg{idx.set}, \cs{gls}\marg{gls.matrix}.
\strut
\cmd{newpage}
\cs{gls}\marg{gls.array}, \cs{gls}\marg{gls.vector}, \cs{gls}\marg{gls.set}, \cs{gls}\marg{gls.matrix}.
\strut
\cs{printunsrtglossaries}
\cmd{end}\marg{document}
\end{codeenv}
In this case, the primary entries are placed in the \code{index}
glossary type and are assigned the prefix \idprefixfmt{idx} but only
two of the primary entries have been used in the document (both on
page~2).

The dual entries are assigned the prefix \idprefixfmt{gls} and are
placed in the \code{main} glossary. The \code{gls.array} 
and \code{gls.matrix} entries have been indexed on pages~1, 2 and~3.
The \code{gls.vector} and \code{gls.set} entries have been indexed
on pages~1 and~3.

With the default setting, some of the locations are in the
\code{main} glossary (corresponding to \code{\cs{gls}\marg{gls.array}},
\code{\cs{gls}\marg{gls.vector}}, \code{\cs{gls}\marg{gls.set}} and
\code{\cs{gls}\marg{gls.matrix}}) and some of the locations are in the
\code{index} glossary (corresponding to \code{\cs{gls}\marg{idx.vector}}
and \code{\cs{gls}\marg{idx.set}}).

If the option \csopt[primary]{combine-dual-locations} is added to
the resource set, then all the locations are moved to the
\code{index} glossary. The entries in the \code{main} glossary no
longer have locations. This is actually preferable for this type
of document and it's best not to reference the primary (index)
entries as the hyperlink created by \ics{gls} will point to the
index, but these entries don't have descriptions, so it's less
useful than referencing the dual (main) entries as then the
hyperlink can point to the definition in the \code{main} glossary.


\subsection{Dual Fields}
\label{sec:dualoptsfields}

\optsection[\subsubsection]{dual-type}

This option sets the \field{type} field for all dual
entries. (The primary entries obey the \csopt{type} option.) This
will override any value of \field{type} provided in the \ext{bib}
file (or created through a mapping). The \meta{value} is required and
should be one of:
\begin{itemize}
\item \optfmt{same as entry}: sets the \field{type} to the entry
type (\idx!{lowercase} and without 
the initial \code{@}). For example, if the entry was defined with
\atentry{dualentry}, the \field{type} will be set to
\optfmt{dualentry}.
If you've used \csopt{entry-type-aliases}, this refers to the target
entry type not the original entry type provided in the \ext{bib}
file.

\item \optfmt{same as original entry}: set the \field{type} field
to the original entry type (\idx!{lowercase} and without 
the initial \code{@}) before it was aliased (behaves like 
\optfmt{same as entry} if the entry type wasn't aliased).

\item \optfmt{same as base}: sets the \field{type} to the base name
of the \ext{bib} file (without the extension) that provided the
entry definition (new to v1.1);

\item \optfmt{same as primary}: sets the \field{type} to the same as
the corresponding primary entry's \field{type} (which may have been
set with \csopt{type}). If the primary entry doesn't have the
\field{type} field set, the dual's \field{type} will remain
unchanged.

\item \optfmt{same as parent}: sets the \field{type} to the same as
the entry's parent (new to v1.9). If the entry doesn't have a parent
or if the parent doesn't have the \field{type} field set, then
no change is made.

\item \meta{label}: sets the \field{type} field to \meta{label}.
\end{itemize}

Remember that the glossary with that label must have already 
been defined (see \sectionref{sec:newglossary}).
For example:
\begin{codeenv}
\cs{newglossary*}\marg{english}\marg{English}
\cs{newglossary*}\marg{french}\marg{French}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[en]{sort},\csopt[fr]{dual-sort},
 \csopt[english]{type},
 \csopt[french]{dual-type}}
\end{codeenv}
Alternatively:
\begin{codeenv}
\cs{newglossary*}\marg{dictionary}\marg{Dictionary}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt[en]{sort},\csopt[fr]{dual-sort},
 \csopt[dictionary]{type},
 \csopt[same as primary]{dual-type}}
\end{codeenv}

\optsection[\subsubsection]{dual-category}

This option sets the \field{category} field for all dual
entries. (The primary entries obey the \csopt{category} option.) This
will override any value of \field{category} provided in the \ext{bib}
file (or created through a mapping). The \meta{value} may be empty or
one of:
\begin{itemize}
\item \optfmt{same as entry}: sets the \field{category} to the entry
type (\idx{lowercase} and without 
the initial \code{@}). For example, if the entry was defined with
\atentry{dualentry}, the \field{category} will be set to
\optfmt{dualentry}.
If you've used \csopt{entry-type-aliases}, this refers to the target
entry type not the original entry type provided in the \ext{bib}
file.

\item \optfmt{same as original entry}: set the \field{category} field
to the original entry type (\idx{lowercase} and without 
the initial \code{@}) before it was aliased (behaves like 
\optfmt{same as entry} if the entry type wasn't aliased).

\item \optfmt{same as base}: sets the \field{category} to the base
name of the \ext{bib} file (without the extension) that provided the
entry definition (new to v1.1);

\item \optfmt{same as primary}: sets the \field{category} to the
same as the corresponding primary entry's \field{category} (which
may have been set with \csopt{category}). If the primary entry
doesn't have the \field{category} field set, the dual's
\field{category} will remain unchanged.

\item \optfmt{same as type}: sets the \field{category} to the same
as the value of the entry's \field{type} field (which may have been
set with \csopt{dual-type}). If the entry doesn't have the
\field{type} field set, the \field{category} will remain unchanged.

\item \meta{label}: sets the \field{category} field to \meta{label}.
\end{itemize}

\optsection[\subsubsection]{dual-counter}

As \csopt{counter} but for the dual entries. In this case
\meta{value} may be the name of the counter or 
\optfmt{same as primary} which uses the counter for the primary
entry.

\optsection[\subsubsection]{dual-short-case-change}

As \csopt{short-case-change} but applies to the \field{dualshort}
field instead.

\optsection[\subsubsection]{dual-long-case-change}

As \csopt{long-case-change} but applies to the \field{duallong}
field instead.

\optsection[\subsubsection]{dual-field}

If this option is used, this will add \ics{glsxtrprovidestoragekey}
to the start of the \iext{glstex} file providing the key given by
\meta{value}.  Any entries defined using a dual entry type, such as
\atentry{dualentry}, will be written to the \ext{glstex} file with
an extra field called \meta{value} that is set to the mirror entry.
If \meta{value} is omitted \csopt[dual]{dual-field} is assumed. If
you use a different value, you will need to redefine
\gls{GlsXtrDualField} (either locally or globally).

For example, if the \ext{bib} file contains:
\begin{codeenv}
\atentry{dualentry}\marg{child,
  \field{name}=\marg{child},
  \field{plural}=\marg{children},
  \field{description}=\marg{enfant}
}
\end{codeenv}
Then with \csopt[dual]{dual-field} (or simply \csopt{dual-field}
without a value) this will first add the line:
\begin{codeenv}
\cs{glsxtrprovidestoragekey}\marg{\field{dual}}\marg{}\marg{}
\end{codeenv}
at the start of the file and will include the line:
\begin{codeenv*}
\field{dual}=\marg{dual.child},
\end{codeenv*}
for the primary entry (\code{child}) and the line:
\begin{codeenv}
\field{dual}=\marg{child},
\end{codeenv}
for the dual entry (\code{dual.child}). It's then possible to
reference one entry from the other. For example, the post-description 
hook could contain:
\begin{codeenv}
 \cs{ifglshasfield}\marg{\field{dual}}\marg{\cs{glscurrententrylabel}}
 \marg{\comment{}
   \cs{space}
   (\cs{glshyperlink}\marg{\cs{glscurrentfieldvalue}})\comment{}
 }\comment{}
 \marg{}\comment{}
\end{codeenv}
Note that this new field won't be available for use within the
\iext{bib} file (unless it was previously defined in the document
before \gls!{glsxtrresourcefile}).

\optsection[\subsubsection]{dual-date-time-field-format}

As \csopt{date-time-field-format} but is used for dual entries.

\optsection[\subsubsection]{dual-date-field-format}

As \csopt{date-field-format} but is used for dual entries.

\optsection[\subsubsection]{dual-time-field-format}

As \csopt{time-field-format} but is used for dual entries.

\optsection[\subsubsection]{dual-date-time-field-locale}

As \csopt{date-time-field-locale} but is used for dual entries.

\optsection[\subsubsection]{dual-date-field-locale}

As \csopt{date-field-locale} but is used for dual entries.

\optsection[\subsubsection]{dual-time-field-locale}

As \csopt{time-field-locale} but is used for dual entries.

\subsection{Dual Sorting}
\label{sec:dualoptssort}

\optsection[\subsubsection]{dual-sort}

This option indicates how to sort the dual entries. The primary
entries are sorted with the normal entries according to
\csopt{sort}, and the dual entries are sorted according to
\csopt{dual-sort} unless \csopt[combine]{dual-sort} in which case the dual
entries will be combined with the primary entries and all the
entries will sorted together according to the \csopt{sort} option.

If \meta{value} isn't set to \optfmt{combine} then the dual
entries are sorted separately according to \meta{value} (as per
\csopt{sort}) and the dual entries will be appended at the end of
the \iext{glstex} file. The field used by the comparator is given by
\csopt{dual-sort-field}.
If \csopt[custom]{dual-sort}, then the dual entries are sorted according to the 
rule provided by \csopt{dual-sort-rule}.

For example:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries-dual]{src},
 \csopt[en]{sort},
 \csopt[de-CH-1996]{dual-sort}
}
\end{codeenv}
This will sort the primary entries according to \optfmt{en}
(English) and the secondary entries according to \optfmt{de-CH-1996} 
(Swiss German new orthography) whereas:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[entries-dual]{src},
 \csopt[en-GB]{sort},
 \csopt[combine]{dual-sort}
}
\end{codeenv}
will combine the dual entries with the primary entries and sort them
all according to the \optfmt{en-GB} locale (British English).

If not set, \csopt{dual-sort} defaults to \optfmt{combine}. If
\meta{value} is omitted, \optfmt{locale} is assumed.

\optsection[\subsubsection]{dual-sort-field}

This option indicates the field to use when sorting dual entries
(when they haven't been combined with the primary entries). The
default value is the same as the \csopt{sort-field} value.

\optsection[\subsubsection]{dual-missing-sort-fallback}

As \csopt{missing-sort-fallback} but for dual sorting.

\optsection[\subsubsection]{dual-trim-sort}

As \csopt{trim-sort} but for dual sorting.

\optsection[\subsubsection]{dual-sort-replace}

As \csopt{sort-replace} but for dual sorting.

\optsection[\subsubsection]{dual-sort-rule}

As \csopt{sort-rule} but for \csopt[custom]{dual-sort}.

\optsection[\subsubsection]{dual-break-at}

As \csopt{break-at} but for dual entries.

\optsection[\subsubsection]{dual-break-marker}

As \csopt{break-marker} but for dual entries.

\optsection[\subsubsection]{dual-sort-number-pad}

As \csopt{sort-number-pad} but for dual entries.

\optsection[\subsubsection]{dual-sort-pad-plus}

As \csopt{sort-pad-plus} but for dual entries.

\optsection[\subsubsection]{dual-sort-pad-minus}

As \csopt{sort-pad-minus} but for dual entries.

\optsection[\subsubsection]{dual-identical-sort-action}

As \csopt{identical-sort-action} but for dual entries.

\optsection[\subsubsection]{dual-sort-suffix}

As \csopt{sort-suffix} but for dual entries.

\optsection[\subsubsection]{dual-sort-suffix-marker}

As \csopt{sort-suffix-marker} but for dual entries.

\optsection[\subsubsection]{dual-strength}

As \csopt{strength} but for dual entries.

\optsection[\subsubsection]{dual-decomposition}

As \csopt{decomposition} but for dual entries.

\optsection[\subsubsection]{dual-letter-number-rule}

As \csopt{letter-number-rule} but for dual entries that use a
letter-number sort.

\optsection[\subsubsection]{dual-letter-number-punc-rule}

As \csopt{letter-number-punc-rule} but for dual entries that use a
letter-number sort.

\optsection[\subsubsection]{dual-numeric-sort-pattern}

As \csopt{numeric-sort-pattern} but for dual entries that use a
locale-sensitive numeric sort.

\optsection[\subsubsection]{dual-numeric-locale}

As \csopt{numeric-locale} but for dual entries that use a
locale-sensitive numeric sort.

\optsection[\subsubsection]{dual-date-sort-locale}

As \csopt{date-sort-locale} but for dual entries that
use a date/time sort.

\optsection[\subsubsection]{dual-date-sort-format}

As \csopt{date-sort-format} but for dual entries that
use a date/time sort.

\optsection[\subsubsection]{dual-group-formation}

As \csopt{group-formation} but for dual sorting.


\subsection{Dual Mappings}
\label{sec:dualoptsmap}

\optsection[\subsubsection]{dual-entry-map}

This setting governs the behaviour of \atentry{dualentry}
definitions. The value consists of two comma-separated lists of
equal length identifying the field mapping used to create the dual
entry from the primary one. Note that the \field{alias} field
can't be mapped.

The default setting is:
\begin{codeenv}
\csopt[
  \marg{\field{name},\field{plural},\field{description},\field{descriptionplural}},
  \marg{\field{description},\field{descriptionplural},\field{name},\field{plural}}
]{dual-entry-map}
\end{codeenv}
The dual entry is created by copying the value of the field in the
first list \meta{list1} to the field in the corresponding place in the second
list \meta{list2}. Any additional fields are copied over to the same
field.
For example:
\begin{codeenv}
\atentry{dualentry}\marg{cat,
  \field{name}=\marg{cat},
  \field{description}=\marg{chat},
  \field{see}=\marg{dog}
}
\end{codeenv}
defines two entries. The primary entry is essentially like:
\begin{codeenv}
\atentry{entry}\marg{cat,
  \field{name}=\marg{cat},
  \field{plural}=\marg{cat\cs{glspluralsuffix} },
  \field{description}=\marg{chat},
  \field{descriptionplural}=\marg{chat\cs{glspluralsuffix} },
  \field{see}=\marg{dog}
}
\end{codeenv}
and the dual entry is essentially like:
\begin{codeenv}
\atentry{entry}\marg{dual.cat,
  \field{description}=\marg{cat},
  \field{descriptionplural}=\marg{cat\cs{glspluralsuffix} },
  \field{name}=\marg{chat},
  \field{plural}=\marg{chat\cs{glspluralsuffix} },
  \field{see}=\marg{dog}
}
\end{codeenv}
(except they're defined using \gls{bibglsnewdualentry} instead of
\gls{bibglsnewentry}, and each is considered dependent on the other.)

The \field{see} field isn't listed in \csopt{dual-entry-map} so its
value is simply copied directly over to the \field{see} field in the
dual entry. Note that the missing \field{plural} and
\field{descriptionplural} fields have been filled in using their
fallback values.

In general \bibgls\ doesn't try to supply missing fields, but in the
dual entry cases it needs to do this for the mapped fields. This is
because the shuffled fields might have different default values from
the \styfmt{glossaries-extra} package's point of view. For example, 
\gls{longnewglossaryentry} doesn't provide a default for
\field{descriptionplural} if it hasn't been set.

\optsection[\subsubsection]{dual-abbrv-map}

This is like \csopt{dual-entry-map} but applies to
\atentry{dualabbreviation} rather than \atentry{dualentry}. 
Note that the \field{alias} field can't be mapped. The
default setting is:
\begin{codeenv}
\csopt[
  \marg{\field{short},\field{shortplural},\field{long},\field{longplural},\field{dualshort},\field{dualshortplural},
   \field{duallong},\field{duallongplural}},
  \marg{\field{dualshort},\field{dualshortplural},\field{duallong},\field{duallongplural},\field{short},\field{shortplural},
   \field{long},\field{longplural}}
]{dual-abbrv-map}
\end{codeenv}

This essentially flips the \field{short} field with the
\field{dualshort} field and the \field{long} field with the
\field{duallong} field. See \atentry{dualabbreviation}
for further details.

\optsection[\subsubsection]{dual-abbrventry-map}

This is like \csopt{dual-entry-map} but applies to
\atentry{dualabbreviationentry} rather than \atentry{dualentry}. 
Note that the \field{alias} field can't be mapped. The
default setting is:
\begin{codeenv}
\csopt[
  \marg{\field{long},\field{short},\field{shortplural}},
  \marg{\field{name},\field{text},\field{plural}}
]{dual-abbrventry-map}
\end{codeenv}
See \atentry{dualabbreviationentry} for further details.


\optsection[\subsubsection]{dual-symbol-map}

This is like \csopt{dual-entry-map} but applies to
\atentry{dualsymbol} rather than \atentry{dualentry}. 
Note that the \field{alias} field can't be mapped. The
default setting is:
\begin{codeenv}
\csopt[
  \marg{\field{name},\field{plural},\field{symbol},\field{symbolplural}},
  \marg{\field{symbol},\field{symbolplural},\field{name},\field{plural}}
]{dual-symbol-map}
\end{codeenv}
This essentially flips the \field{name} field with the
\field{symbol} field.

\optsection[\subsubsection]{dual-indexentry-map}

This is like \csopt{dual-entry-map} but applies to
\atentry{dualindexentry} rather than \atentry{dualentry}. 
The default setting is:
\begin{codeenv}
\csopt[
  \marg{\field{name}},
  \marg{\field{name}}
]{dual-indexentry-map}
\end{codeenv}
Note that there must always be at least one pair, even if it's the
same field, since this identifies the field to use for the backlink,
if set.

\optsection[\subsubsection]{dual-indexsymbol-map}

This is like \csopt{dual-entry-map} but applies to both
\atentry{dualindexsymbol} and \atentry{dualindexnumber}. 
The default setting is:
\begin{codeenv}
\csopt[
  \marg{\field{symbol},\field{name},\field{symbolplural},\field{plural}},
  \marg{\field{name},\field{symbol},\field{plural},\field{symbolplural}}
]{dual-indexsymbol-map}
\end{codeenv}

\optsection[\subsubsection]{dual-indexabbrv-map}

This is like \csopt{dual-entry-map} but applies to both the dual
\atentry{dualindexabbreviation} and tertiary
\atentry{tertiaryindexabbreviationentry} entry types. 
The default setting is:
\begin{codeenv}
\csopt[
  \marg{\field{name}},
  \marg{\field{name}}
]{dual-indexabbrv-map}
\end{codeenv}

\subsection{Dual Back-Links}
\label{sec:dualoptsbacklinks}

\optsection[\subsubsection]{dual-entry-backlink}

This is a boolean setting. If \meta{boolean} is missing \optfmt{true}
is assumed.

When used with \atentry{dualentry}, if \meta{boolean} is
\optfmt{true}, this will wrap the contents of the first mapped field
with \gls{bibglshyperlink}.  The field is obtained from the first
mapping listed in \csopt{dual-entry-map}.

For example, if the document contains:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt{dual-entry-backlink},
  \csopt[
    \marg{\field{name},\field{plural},\field{description},\field{descriptionplural}},
    \marg{\field{description},\field{descriptionplural},\field{name},\field{plural}}
  ]{dual-entry-map},
  \csopt[entries-dual]{src}}
\end{codeenv}
and if the \ext{bib} file contains:
\begin{codeenv}
\atentry{dualentry}\marg{child,
  \field{name}=\marg{child},
  \field{plural}=\marg{children},
  \field{description}=\marg{enfant}
}
\end{codeenv}
Then the definition of the primary entry (\code{child}) in the
\ext{glstex} file will set the \field{description} field to:
\begin{codeenv}
\gls{bibglshyperlink}\marg{enfant}\marg{dual.child}
\end{codeenv}
and the dual entry (\code{dual.child}) will have the
\field{description} field set to:
\begin{codeenv}
\gls{bibglshyperlink}\marg{child}\marg{child}
\end{codeenv}
This use of the wrapper \gls{bibglshyperlink} (rather than explicitly
using \ics{glshyperlink}) and inserting the actual field value
(rather than using commands like \ics{glsentryname}) allows it 
to work with \ics{makefirstuc} if the field requires a case-change.

The reason the \field{description} field is chosen for the modification is
because the first field listed in \meta{list1} of \csopt{dual-entry-map} 
is the \field{name} field which maps to \field{description} (the
first field in the second list \meta{list2}). This means that the hyperlink for
the dual entry should be put in the \field{description} field.

For the primary entry, the \field{name} field is looked up in the
second list from the \csopt{dual-entry-map} setting. This is the
third item in this second list, so the third item in the first list
is selected, which also happens to be the \field{description} field,
so the hyperlink for the primary entry is put in the
\field{description} field.

\optsection[\subsubsection]{dual-abbrv-backlink}

This is analogous to \csopt{dual-entry-backlink} but for entries
defined with \atentry{dualabbreviation} instead of
\atentry{dualentry}.

\optsection[\subsubsection]{dual-symbol-backlink}

This is analogous to \csopt{dual-entry-backlink} but for entries
defined with \atentry{dualsymbol} instead of \atentry{dualentry}.

\optsection[\subsubsection]{dual-abbrventry-backlink}

Analogous to \csopt{dual-entry-backlink} but for entries
defined with \atentry{dualabbreviationentry} instead of
\atentry{dualentry}. This setting can be problematic as the
backlinks rely on the relevant field being known to \bibgls.
Since the abbreviation style typically sets the \field{name}
field (and sometimes the \field{description} field as well), you may
find that no backlink appears. A simple workaround is to 
use \csopt{dual-field} (or \csopt[dual]{dual-field}) 
to store the dual label in the \field{dual} field, and then
use a style that checks for this field and adds the backlink.

With \sty{glossaries-extra} v1.30+ you can use:
\nosecdef{GlsXtrDualBackLink}
which encapsulates \meta{text} with a hyperlink to the dual.
The \meta{label} identifies the entry that requires a backlink.
The dual's label is obtained from the field given by:
\nosecdef{GlsXtrDualField}
which defaults to \field{dual}. Note that if you assign a 
different field label with \csopt{dual-field}, then you will
need to redefine \gls{GlsXtrDualField} as appropriate.

For example:
\begin{codeenv}
\cs{renewcommand}*\marg{\ics{glsuserdescription}}[2]\marg{\comment{}
  \gls{GlsXtrDualBackLink}\marg{\ics{glslonguserfont}\marg{\idx{param}1}}\marg{\idx{param}2}\comment{}
}
\cs{setabbreviationstyle}\marg{\abbrstyle{long-short-user}}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt{dual-field}}
\end{codeenv}

\optsection[\subsubsection]{dual-entryabbrv-backlink}

As \csopt{dual-abbrventry-backlink} but for entries
defined with \atentry{dualentryabbreviation} instead of
\atentry{dualabbreviationentry}.

\optsection[\subsubsection]{dual-indexentry-backlink}

This is analogous to \csopt{dual-entry-backlink} but for entries
defined with \atentry{dualindexentry} instead of
\atentry{dualentry}.

\optsection[\subsubsection]{dual-indexsymbol-backlink}

This is analogous to \csopt{dual-entry-backlink} but for entries
defined with \atentry{dualindexsymbol} and
\atentry{dualindexnumber}.

\optsection[\subsubsection]{dual-indexabbrv-backlink}

This is analogous to \csopt{dual-entry-backlink} but for entries
defined with \atentry{dualindexabbreviation} and
\atentry{tertiaryindexabbreviationentry}.

\optsection[\subsubsection]{dual-backlink}

Shortcut for:
\begin{flushleft}\obeylines
\csopt[\meta{boolean}]{dual-entry-backlink}, 
\csopt[\meta{boolean}]{dual-abbrventry-backlink}, 
\csopt[\meta{boolean}]{dual-abbrv-backlink},
\csopt[\meta{boolean}]{dual-symbol-backlink},
\csopt[\meta{boolean}]{dual-indexentry-backlink},
\csopt[\meta{boolean}]{dual-indexsymbol-backlink},
\csopt[\meta{boolean}]{dual-indexabbrv-backlink}
\end{flushleft}

\section{Tertiary Entries}

\optsection{tertiary-prefix}

This option indicates the prefix to use for the tertiary entries. The
default value is \idprefix{tertiary} (including the terminating period). 

As from version 1.8, the tertiary label prefix is identified
in the \ext{glstex} file with:
\begin{codeenv}
\format{bibglstertiaryprefixlabel}
\end{codeenv}

\optsection{tertiary-type}

This option indicates that the tertiary entries should have
their \field{type} field set to \meta{value}. If \meta{value} is
empty the \field{type} is left unchanged. Unlike 
the \csopt{type} and \csopt{dual-type} options, there are no
recognised keywords.

\optsection{tertiary-category}

This option indicates that the tertiary entries should have
their \field{category} field set to \meta{value}. If \meta{value} is
empty the \field{category} is left unchanged. Unlike 
the \csopt{category} and \csopt{dual-category} options, there are no
recognised keywords.

\chapter{Provided Commands}
\label{sec:bibglscs}

When \bibgls\ creates the \iext{glstex} file, it writes some
definitions for custom commands in the form \csfmt{bibgls\ldots} which
may be changed as required. The command definitions all use
\ics{providecommand} which means that you can define the command with
\ics{newcommand} before the resource file is loaded.

Note that if you try to redefine any of these commands after the
resource file has been loaded with \ics{renewcommand}, you will get
an error on the first \LaTeX\ run when the \ext{glstex} file doesn't
exist. You may prefer to use \ics{glsrenewcommand} instead, which
will generate a warning instead of an error.

Since many of the commands are actually used within the \ext{glstex}
file, it's best to use \cs{newcommand} before the first
\idx{resourceset} and \cs{renewcommand} between \idxpl{resourceset}
if adjustments are necessary.


\section{Entry Definitions}
\label{sec:newentrydefs}

This section lists the commands (\csfmt{bibglsnew\ldots}) used to
define entries.  Note that the entry definition commands are
actually used when \TeX\ inputs the resource file, so redefining
them after the resource file is loaded won't have an effect on the
entries defined in that resource file (but will affect entries defined
in subsequent resource files). Each provided command is defined in
the \ext{glstex} file immediately before the first entry that
requires it, so only the commands that are actually needed are
provided.

The \field{sort} key may be set within the \ext{glstex} entry
definition, but its value is usually not required in the document
unless you are using a hybrid method with \styopt[alsoindex]{record}.

After each entry is defined, if it has any associated locations and
the default \csopt[true]{save-loclist} is set, then the
locations are added using:
\nosecformatdef{glsxtrfieldlistadd}
Any additional fields that don't have associated keys are then set (if
required) with \ics{GlsXtrSetField}.

\cssection{bibglsnewentry}

\formatdef{bibglsnewentry}
This command is used to define terms identified with the
\atentry{entry} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewentry}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}
This uses the starred form \ics{longnewglossaryentry*} that
doesn't automatically append \ics{nopostdesc} (which interferes with
the post-description hooks provided by category attributes).

\cssection{bibglsnewsymbol}

\formatdef{bibglsnewsymbol}
This command is used to define terms identified with the
\atentry{symbol} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewsymbol}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{sort}=\marg{\idx{param}1},\field{category}=\marg{symbol},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}
Note that this sets the \field{sort} field to the label, but this
may be overridden by the \meta{options} if the \field{sort} field
was supplied or if \bibgls\ has determined the value whilst sorting
the entries.

This also sets the \field{category} to \optfmt{symbol}, but again this may
be overridden by \meta{options} if the entry had the \field{category}
field set in the \ext{bib} file or if the \field{category} was
overridden with \csopt[\meta{value}]{category}.

\cssection{bibglsnewnumber}

\formatdef{bibglsnewnumber}
This command is used to define terms identified with the
\atentry{number} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewnumber}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{sort}=\marg{\idx{param}1},\field{category}=\marg{number},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}
This is much the same as \gls{bibglsnewsymbol} above but sets the 
\field{category} to \optfmt{number}. Again the \field{sort} and
\field{category} keys may be overridden by \meta{options}.

\cssection{bibglsnewindex}

\formatdef{bibglsnewindex}
This command is used to define terms identified with the
\atentry{index} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}*\marg{\gls{bibglsnewindex}}[2]\marg{\comment{}
 \gls{newglossaryentry}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}1},\field{category}=\marg{index},\field{description}=\marg{},\idx{param}2}\comment{}
}
\end{codeenv}
This makes the \field{name} default to the \meta{label}, assigns 
the \field{category} to \code{index} and sets an empty
\field{description}. These settings may be overridden by
\meta{options}.

Note that the \field{description} doesn't include
\ics{nopostdesc} to allow for the \idx{postdescriptionhook} used by
category attributes.

\cssection{bibglsnewindexplural}

\formatdef{bibglsnewindexplural}
This command is used to define terms identified with the
\atentry{indexplural} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewindexplural}}[3]\marg{\comment{}
 \gls{newglossaryentry}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{category}=\marg{indexplural},\field{description}=\marg{},\idx{param}2}\comment{}
}
\end{codeenv}
This assigns the \field{category} to \code{indexplural} and sets an empty
\field{description}. These settings may be overridden by \meta{options}.

\cssection{bibglsnewabbreviation}

\formatdef{bibglsnewabbreviation}
This command is used to define terms identified with the
\atentry{abbreviation} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewabbreviation}}[4]\marg{\comment{}
  \gls{newabbreviation}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}3}\marg{\idx{param}4}\comment{}
}
\end{codeenv}
Since this uses \gls{newabbreviation}, it obeys the 
abbreviation style for its given \field{category} (which may have
been set in \meta{options}, either from the \field{category} field
in the \ext{bib} file or through the \csopt{category} option).
Similarly the \field{type} will obey \ics{glsxtrabbrvtype} unless 
the value is supplied in the \ext{bib} file or through the
\csopt{type} option.

\cssection{bibglsnewacronym}

\formatdef{bibglsnewacronym}
This command is used to define terms identified with the
\atentry{acronym} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewacronym}}[4]\marg{\comment{}
  \gls{newacronym}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}3}\marg{\idx{param}4}\comment{}
}
\end{codeenv}
This works in much the same way as \gls{bibglsnewabbreviation}.
Remember that with the \styfmt{glossaries-extra} package \gls{newacronym}
is redefined to just use \gls{newabbreviation} with the default \field{type}
set to \ics{acronymtype} and the default \field{category} set to
\code{acronym}.

\cssection{bibglsnewdualentry}

\formatdef{bibglsnewdualentry}
This command is used to define terms identified with the
\atentry{dualentry} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualentry}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewdualindexentry}

\formatdef{bibglsnewdualindexentry}
This command is used to define primary terms identified with the
\atentry{dualindexentry} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualindexentry}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{category}=\marg{index},\idx{param}2}\marg{}\comment{}
}
\end{codeenv}
Note that this definition ignores the \meta{description} argument.

\cssection{bibglsnewdualindexentrysecondary}

\formatdef{bibglsnewdualindexentrysecondary}
This command is used to define secondary terms identified with the
\atentry{dualindexentry} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualindexentrysecondary}}[4]\marg{\comment{}
  \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewdualindexsymbol}

\formatdef{bibglsnewdualindexsymbol}
This command is used to define primary terms identified with the
\atentry{dualindexsymbol} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualindexsymbol}}[5]\marg{\comment{}
  \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{category}=\marg{index},\field{symbol}=\marg{\idx{param}4},\idx{param}2}\marg{}\comment{}
}
\end{codeenv}
Note that this definition ignores the \meta{description} argument.

\cssection{bibglsnewdualindexsymbolsecondary}

\formatdef{bibglsnewdualindexsymbolsecondary}
This command is used to define secondary terms identified with the
\atentry{dualindexsymbol} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualindexsymbolsecondary}}[5]\marg{\comment{}
  \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{category}=\marg{symbol},\field{symbol}=\marg{\idx{param}4},\idx{param}2}\marg{\idx{param}5}\comment{}
}
\end{codeenv}

\cssection{bibglsnewdualindexnumber}

\formatdef{bibglsnewdualindexnumber}
This command is used to define primary terms identified with the
\atentry{dualindexnumber} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualindexnumber}}[5]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{category}=\marg{index},\field{symbol}=\marg{\idx{param}4},\idx{param}2}\marg{}\comment{}
}
\end{codeenv}
Note that this definition ignores the \meta{description} argument.

\cssection{bibglsnewdualindexnumbersecondary}

\formatdef{bibglsnewdualindexnumbersecondary}
This command is used to define secondary terms identified with the
\atentry{dualindexnumber} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\cs{bibglsnewdualindexnumbersecondary}}[5]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{category}=\marg{number},\field{symbol}=\marg{\idx{param}4},\idx{param}2}\marg{\idx{param}5}\comment{}
}
\end{codeenv}

\cssection{bibglsnewdualindexabbreviation}

\formatdef{bibglsnewdualindexabbreviation}
This command is used to define primary terms identified with the
\atentry{dualindexabbreviation} type. The default definition provided in 
the \ext{glstex} file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualindexabbreviation}}[7]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\comment{}
   \field{name}=\marg{\cs{protect}\gls{bibglsuseabbrvfont}\marg{\idx{param}4}\marg{\cs{glscategory}\marg{\idx{param}2}}},\comment{}
   \field{category}=\marg{index},\idx{param}3}\marg{}\comment{}
}
\end{codeenv}
In this case \meta{dual-label} is the dual entry's label, which is used 
to fetch the category label in \gls{bibglsuseabbrvfont}. (The
\field{category} field for the dual isn't used since a custom
definition of
\gls!{bibglsnewdualindexabbreviationsecondary} may override the
value known to \bibgls.)

Note that (as shown above) with the default 
\csopt[short]{abbreviation-name-fallback} the \field{name} uses:
\nosecformatdef{bibglsuseabbrvfont}
to format the name, which ensures that it uses the same font as the
short form for the dual abbreviation. This will use
\ics{glsuseabbrvfont} if it's defined otherwise it will be defined
to replicate that command. If \csopt{abbreviation-name-fallback} is
set to some other field then the \field{name} uses:
\nosecformatdef{bibglsuselongfont}
instead, which ensures that it uses the same font as the
long form for the dual abbreviation.

\cssection{bibglsnewdualindexabbreviationsecondary}

\formatdef{bibglsnewdualindexabbreviationsecondary}
This command is used to define secondary terms identified with the
\atentry{dualindexabbreviation} entry type.
The definition provided in the \ext{glstex} file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualindexabbreviationsecondary}}[6]\marg{\comment{}
  \cs{ifstrempty}\marg{\idx{param}6}\comment{}
  \marg{\gls{newabbreviation}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}4}\marg{\idx{param}5}}\comment{}
  \marg{\gls{newabbreviation}\oarg{\idx{param}2,\field{description}=\marg{\idx{param}6}}\marg{\idx{param}1}\marg{\idx{param}4}\marg{\idx{param}5}}\comment{}
}
\end{codeenv}
This ensures that a missing or empty \field{description} doesn't
interfere with the abbreviation style.

\cssection{bibglsnewdualabbreviationentry}

\formatdef{bibglsnewdualabbreviationentry}
This command is used to define primary terms identified with the
\atentry{dualabbreviationentry} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualabbreviationentry}}[5]\marg{\comment{}
  \gls{newabbreviation}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}3}\marg{\idx{param}4}\comment{}
}
\end{codeenv}
Note that this definition ignores the \meta{description} argument.

\cssection{bibglsnewdualabbreviationentrysecondary}

\formatdef{bibglsnewdualabbreviationentrysecondary}
This command is used to define secondary terms identified with the
\atentry{dualabbreviationentry} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualabbreviationentrysecondary}}[5]\marg{\comment{}
  \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\idx{param}2}\marg{\idx{param}5}\comment{}
}
\end{codeenv}
Note that this definition ignores the \meta{short} and \meta{long}
arguments (which will typically be empty unless the default mappings
are changed).

\cssection{bibglsnewdualentryabbreviation}

\formatdef{bibglsnewdualentryabbreviation}
This command is used to define primary terms identified with the
(now deprecated) entry type
\atentry{dualentryabbreviation}. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualentryabbreviation}}[5]\marg{\comment{}
  \gls{newabbreviation}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}3}\marg{\idx{param}4}\comment{}
}
\end{codeenv}
Note that this definition ignores the \meta{description} argument.

\cssection{bibglsnewdualentryabbreviationsecondary}

\formatdef{bibglsnewdualentryabbreviationsecondary}
This command is used to define secondary terms identified with the
(now deprecated) entry type
\atentry{dualentryabbreviation}. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualentryabbreviationsecondary}}[5]\marg{\comment{}
  \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\idx{param}2}\marg{\idx{param}5}\comment{}
}
\end{codeenv}
Note that this definition ignores the \meta{short} and \meta{long}
arguments (which will typically be empty unless the default mappings
are changed).

\cssection{bibglsnewdualsymbol}

\formatdef{bibglsnewdualsymbol}
This command is used to define terms identified with the
\atentry{dualsymbol} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualsymbol}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{sort}=\marg{\idx{param}1},\field{category}=\marg{symbol},\idx{param}2}\marg{\idx{param}4}}
\end{codeenv}

\cssection{bibglsnewdualnumber}

\formatdef{bibglsnewdualnumber}
This command is used to define terms identified with the
\atentry{dualnumber} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualnumber}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{sort}=\marg{\idx{param}1},\field{category}=\marg{symbol},\idx{param}2}\marg{\idx{param}4}}
\end{codeenv}

\cssection{bibglsnewdualabbreviation}

\formatdef{bibglsnewdualabbreviation}
This command is used to define terms identified with the
\atentry{dualabbreviation} type where the \field{duallong} field
is swapped with the \field{long} field and the \field{dualshort}
field is swapped with the \field{short} field. The definition provided in the
\ext{glstex} file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualabbreviation}}[4]\marg{\comment{}
  \gls{newabbreviation}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}3}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewdualacronym}

\formatdef{bibglsnewdualacronym}
This command is used to define terms identified with the
\atentry{dualacronym} type. The definition provided in the \ext{glstex}
file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewdualacronym}}[4]\marg{\comment{}
  \gls{newacronym}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}3}\marg{\idx{param}4}\comment{}
}
\end{codeenv}
This works in much the same way as \gls{bibglsnewdualabbreviation}.
Remember that with the \styfmt{glossaries-extra} package \gls{newacronym}
is redefined to just use \gls{newabbreviation} with the default \field{type}
set to \ics{acronymtype} and the default \field{category} set to
\code{acronym}.

\cssection{bibglsnewtertiaryindexabbreviationentry}

\formatdef{bibglsnewtertiaryindexabbreviationentry}
This is used to define primary terms identified with the
\atentry{tertiaryindexabbreviationentry} type. It's essentially
the same as \gls!{bibglsnewdualindexabbreviation}.
The definition provided in the \ext{glstex} file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewtertiaryindexabbreviationentry}}[7]\marg{\comment{}
  \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\comment{}
      \field{name}=\marg{\cs{protect}\gls{bibglsuseabbrvfont}\marg{\idx{param}4}\marg{\cs{glscategory}\marg{\idx{param}2}}},\comment{}
      \field{category}=\marg{index},\idx{param}3}{}\comment{}
}
\end{codeenv}

\cssection{bibglsnewtertiaryindexabbreviationentrysecondary}

\formatdef{bibglsnewtertiaryindexabbreviationentrysecondary}
This command is used to define both the secondary and tertiary terms identified with the
\atentry{tertiaryindexabbreviationentry} type. The secondary
term is an abbreviation and the tertiary term is a regular entry.
The definition written to the \ext{glstex} file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewtertiaryindexabbreviationentrysecondary}}[8]\marg{\comment{}
  \gls{newabbreviation}\oarg{\idx{param}3}\marg{\idx{param}1}\marg{\idx{param}6}\marg{\idx{param}7}\comment{}
  \cs{longnewglossaryentry*}\marg{\idx{param}2}\comment{}
  \marg{\field{name}=\marg{\cs{protect}\gls{bibglsuselongfont}\marg{\idx{param}7}\marg{\cs{glscategory}\marg{\idx{param}1}}},\idx{param}4}\comment{}
  \marg{\idx{param}8}\comment{}
}
\end{codeenv}
The \meta{label} is the label for the secondary (abbreviation) entry
and \meta{tertiary-label} is the label for the tertiary (regular)
entry. The fifth argument (\meta{primary name}) isn't used but is
provided if required for a custom redefinition.
The \field{name} field for the tertiary is obtained from 
the \meta{long} argument encapsulated by \gls{bibglsuselongfont}
to format the name, which ensures that it uses the same font as the
long form for the dual abbreviation. This will use
\ics{glsuselongfont} if it's defined otherwise it will be defined to
replicate that command.

\cssection{bibglsnewbibtexentry}

\formatdef{bibglsnewbibtexentry}
This command is used to define the main term identified with
\atentry{bibtexentry}.
The definition written to the \ext{glstex} file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewbibtexentry}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewcontributor}

\formatdef{bibglsnewcontributor}
This command is used to define terms identified with
\atentry{contributor} (typically implicitly created through
\atentry{bibtexentry}).
The definition written to the \ext{glstex} file is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewcontributor}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewprogenitor}

\formatdef{bibglsnewprogenitor}
This command is used to define the main terms created by
\atentry{progenitor}. The definition is written to the \ext{glstex}
file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewprogenitor}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnindex}

\formatdef{bibglsnewspawnindex}
This command is used to define the main terms created by
\atentry{spawnindex}. The definition is written to the \ext{glstex}
file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnindex}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnedindex}

\formatdef{bibglsnewspawnedindex}
This command is used to define the terms spawned by
\atentry{progenitor} or \atentry{spawnindex}. The definition is written 
to the \ext{glstex} file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnedindex}}[2]\marg{\comment{}
 \gls{newglossaryentry}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}1},\field{category}={index},\field{description}=\marg{},\idx{param}2}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnindexplural}

\formatdef{bibglsnewspawnindexplural}
This command is used to define the main terms created by
\atentry{spawnindexplural}. The definition is written to the \ext{glstex}
file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnindexplural}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnedindexplural}

\formatdef{bibglsnewspawnedindexplural}
This command is used to define the terms spawned by
\atentry{spawnindexplural}. 
The definition is written to the \ext{glstex} file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnedindexplural}}[3]\marg{\comment{}
 \gls{newglossaryentry}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{category}=\marg{indexplural},\field{description}=\marg{},\idx{param}2}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnentry}

\formatdef{bibglsnewspawnentry}
This command is used to define the main terms created by
\atentry{spawnentry}. The definition is written to the \ext{glstex}
file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnentry}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnedentry}

\formatdef{bibglsnewspawnedentry}
This command is used to define the terms spawned by \atentry{spawnentry}. 
The definition is written to the \ext{glstex} file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnedentry}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnabbreviation}

\formatdef{bibglsnewspawnabbreviation}
This command is used to define the main terms created by
\atentry{spawnabbreviation}. The definition is written to the \ext{glstex}
file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnabbreviation}}[4]\marg{\comment{}
  \gls{newabbreviation}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}3}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnedabbreviation}

\formatdef{bibglsnewspawnedabbreviation}
This command is used to define the terms spawned by \atentry{spawnabbreviation}. 
The definition is written to the \ext{glstex} file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnedabbreviation}}[4]\marg{\comment{}
  \gls{newabbreviation}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}3}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnacronym}

\formatdef{bibglsnewspawnacronym}
This command is used to define the main terms created by
\atentry{spawnacronym}. The definition is written to the \ext{glstex}
file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnacronym}}[4]\marg{\comment{}
  \gls{newacronym}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}3}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnedacronym}

\formatdef{bibglsnewspawnedacronym}
This command is used to define the terms spawned by \atentry{spawnacronym}. 
The definition is written to the \ext{glstex} file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnedacronym}}[4]\marg{\comment{}
  \gls{newacronym}\oarg{\idx{param}2}\marg{\idx{param}1}\marg{\idx{param}3}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnsymbol}

\formatdef{bibglsnewspawnsymbol}
This command is used to define the main terms created by
\atentry{spawnsymbol}. The definition is written to the \ext{glstex}
file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnsymbol}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnedsymbol}

\formatdef{bibglsnewspawnedsymbol}
This command is used to define the terms spawned by \atentry{spawnsymbol}. 
The definition is written to the \ext{glstex} file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnedsymbol}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{sort}=\marg{\idx{param}1},\field{category}=\marg{spawnedsymbol},\idx{param}2}\marg{\idx{param}4}}
\end{codeenv}

\cssection{bibglsnewspawnnumber}

\formatdef{bibglsnewspawnnumber}
This command is used to define the main terms created by
\atentry{spawnnumber}. The definition is written to the \ext{glstex}
file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnnumber}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\cssection{bibglsnewspawnednumber}

\formatdef{bibglsnewspawnednumber}
This command is used to define the terms spawned by \atentry{spawnnumber}. 
The definition is written to the \ext{glstex} file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawnednumber}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{sort}=\marg{\idx{param}1},\field{category}=\marg{spawnednumber},\idx{param}2}\marg{\idx{param}4}}
\end{codeenv}

\cssection{bibglsnewspawndualindexentry}

\formatdef{bibglsnewspawndualindexentry}
This command is used to define the \idx{progenitor}['s] primary term created by 
\atentry{spawndualindexentry}. 
The definition is written to the \ext{glstex} file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawndualindexentry}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\field{category}=\marg{index},\idx{param}2}{}\comment{}
}
\end{codeenv}
The \meta{description} argument is ignored.

\cssection{bibglsnewspawndualindexentrysecondary}

\formatdef{bibglsnewspawndualindexentrysecondary}
This command is used to define the \idx{progenitor}['s] secondary
(dual) term created by 
\atentry{spawndualindexentry}. 
The definition is written to the \ext{glstex} file as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsnewspawndualindexentrysecondary}}[4]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2}\marg{\idx{param}4}\comment{}
}
\end{codeenv}

\section{Location Lists and Cross-References}
\label{sec:loclistdefs}

These commands deal with the way the location lists and cross
references are formatted. The commands typically aren't used until
the entry information is displayed in the glossary, so you may
redefine these commands after the resource file has been loaded.

\cssection{bibglsseesep}

\formatdef{bibglsseesep}
Any entries that provide a \field{see} field (and that field hasn't
be omitted from the location list with \csopt[omit]{see}) will
have \gls{bibglsseesep} inserted between the \field{see} part and the
location list (unless there are no locations, in which case just
the \field{see} part is displayed without \gls{bibglsseesep}).

This command is provided with:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsseesep}}\marg{, }
\end{codeenv}
You can define this before you load the \ext{bib} file:
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglsseesep}}\marg{; }
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\end{codeenv}
Or you can redefine it afterwards:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\cs{glsrenewcommand}\marg{\gls{bibglsseesep}}\marg{; }
\end{codeenv}

\cssection{bibglsseealsosep}

\formatdef{bibglsseealsosep}
This is like \gls{bibglsseesep} but is used with cross-reference
lists provided with the \field{seealso} field, if supported.

\cssection{bibglsaliassep}

\formatdef{bibglsaliassep}
This is like \gls{bibglsseesep} but is used with cross-reference
lists provided with the \field{alias} field.

\cssection{bibglsusesee}

\formatdef{bibglsusesee}
Displays the formatted cross-reference list stored in the
\field{see} field for the given entry. This just defaults to 
\ics{glsxtrusesee}\margm{label}.

\cssection{bibglsuseseealso}

\formatdef{bibglsuseseealso}
Displays the formatted cross-reference list stored in the
\field{seealso} field for the given entry. This just defaults to 
\ics{glsxtruseseealso}\margm{label}.

\cssection{bibglsusealias}

\formatdef{bibglsusealias}
Displays the formatted cross-reference stored in the
\field{alias} field for the given entry. This is defined to use
\ics{glsseeformat}.

\cssection{bibglsdelimN}

\formatdef{bibglsdelimN}
Separator between individual locations, except for the last.
This defaults to \ics{delimN}.

\cssection{bibglslastDelimN}

\formatdef{bibglslastDelimN}
Separator between penultimate and final individual locations.
This defaults to \code{,\idx{nbspchar}} to discourage lonely locations.

\cssection{bibglscompact}

\formatdef{bibglscompact}
The first argument \meta{pattern} indicates the location pattern:
\code{digit} for digits, \code{roman} for \idx!{lowercase}
Roman numerals, \code{ROMAN} for \idx!{uppercase} Roman
numerals and \code{alpha} for alphabetical locations. The actual
location is split into two parts, \meta{part1} and \meta{part2}.
The string concatenation \meta{part1}\meta{part2} forms the
actual location.

This just does \meta{part2} by default.

\cssection{bibglspassim}

\formatdef{bibglspassim}
If \csopt{max-loc-diff} is greater than~1, then any ranges that have
skipped over gaps will be followed by \gls{bibglspassim}, which is
defined as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglspassim}}\marg{ \gls{bibglspassimname}}
\end{codeenv}
You can define this before you load the \ext{bib} file:
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglspassim}}\marg{}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\end{codeenv}
Or you can redefine it afterwards:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\cs{glsrenewcommand}\marg{\gls{bibglspassim}}\marg{}
\end{codeenv}

\cssection{bibglspassimname}

\formatdef{bibglspassimname}
The default definition is obtained from the \langxml. For example, with 
\file{bib2gls-en.xml} the provided definition is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglspassimname}}\marg{passim}
\end{codeenv}

\cssection{bibglsrange}

\formatdef{bibglsrange}
Explicit ranges formed using \glsopt[\idx{openrange}]{format} and
\glsopt[\idx{closerange}]{format} or \glsopt[\idx{openrange}\meta{csname}]{format} and
\glsopt[\idx{closerange}\meta{csname}]{format} (where \meta{csname} matches and is a
text-block command without the initial backslash) in the optional
argument of commands like \ics{gls} or \ics{glsadd} are encapsulated within
the argument of \gls{bibglsrange}. By default this simply does its
argument. This command is not used with ranges that are formed by collating
consecutive locations.

\cssection{bibglsinterloper}

\formatdef{bibglsinterloper}
If an explicit range conflicts with a record, a warning will be
issued and the conflicting record will be shifted to the front
of the range inside the argument of \gls{bibglsinterloper}.
The default definition just does \meta{location}\gls!{bibglsdelimN}
so that it fits neatly into the list.

For example, suppose on page~4 of my document I start a range with:
\begin{codeenv*}
\cs{glsadd}\oarg{\glsaddopt[\idx{openrange}]{format}}\marg{sample}
\end{codeenv*}
and end it on page~9 with:
\begin{codeenv*}
\cs{glsadd}\oarg{\glsaddopt[\idx{closerange}]{format}}\marg{sample}
\end{codeenv*}
This forms an explicit range, but let's suppose on page~6 I
have:
\begin{codeenv}
\cs{gls}\oarg{\glsopt[\encap{hyperbf}]{format}}\marg{sample}
\end{codeenv}
This record conflicts with the explicit range (which doesn't include
\encap{hyperbf} in the format). This causes a warning and
the conflicting entry will be moved before the start of the
explicit range resulting in \textbf{6}, 4--9.

Note that implicit ranges can't be formed from interlopers (nor can
implicit ranges be merged with explicit ones),
so if \code{\cs{gls}\oarg{\glsopt[\encap{hyperbf}]{format}}\marg{sample}} 
also occurs on pages~7 and~8 then the result will be \textbf{6}, \textbf{7},
\textbf{8}, 4--9. Either remove the explicit range or
remove the conflicting entries. (Alternatively, redefine
\gls{bibglsinterloper} to ignore its argument, which will
discard the conflicting entries.)

\cssection{bibglspostlocprefix}

\formatdef{bibglspostlocprefix}
If the \csopt{loc-prefix} option is on, \gls!{bibglslocprefix} will
be inserted at the start of location lists, and its default
definition includes \gls{bibglspostlocprefix}
placed after the prefix text. This command is provided with:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglspostlocprefix}}\marg{\cs{cs.space}}
\end{codeenv}
which puts a space between the prefix text and the location list.
You can define this before you load the \ext{bib} file:
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglspostlocprefix}}\marg{: }
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt{loc-prefix}}
\end{codeenv}
Or you can redefine it afterwards:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src},\csopt{loc-prefix}}
\cs{glsrenewcommand}\marg{\gls{bibglspostlocprefix}}\marg{: }
\end{codeenv}

\cssection{bibglslocprefix}

\formatdef{bibglslocprefix}
If the \csopt{loc-prefix} option is on, this command will be
provided. If the glossary type has been provided by \csopt{type}
(and \csopt{dual-type} if there are any dual entries) then the
definition of \gls{bibglslocprefix} will be appended to the glossary
preamble for the given type (or types if there are dual entries).
For example, if the document has:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[main]{type},\csopt[p.,pp.]{loc-prefix},\csopt[entries]{src}}
\end{codeenv}
and there are no dual entries, then the following will be added to
the \ext{glstex} file:
\begin{codeenv}
\cs{apptoglossarypreamble}\oarg{main}\marg{\comment{}
 \cs{providecommand}\marg{\gls{bibglslocprefix}}[1]\marg{\comment{}
  \cs{ifcase}\idx{param}\idx{param}1
  \cmd{or} p.\gls{bibglspostlocprefix}
  \cmd{else} pp.\gls{bibglspostlocprefix}
  \cmd{fi}
 }\comment{}
}
\end{codeenv}
However, if the \csopt{type} key is missing, then the following will
be added instead:
\begin{codeenv}
\cs{appto}\cs{glossarypreamble}\marg{\comment{}
 \cs{providecommand}\marg{\gls{bibglslocprefix}}[1]\marg{\comment{}
  \cs{ifcase}\idx{param}1
  \cmd{or} p.\gls{bibglspostlocprefix}
  \cmd{else} pp.\gls{bibglspostlocprefix}
  \cmd{fi}
 }\comment{}
}
\end{codeenv}

\cssection{bibglspagename}

\formatdef{bibglspagename}
If \csopt[true]{loc-prefix} is used, then this command is provided
using the value of \idx{tag.page} from the \langxml. For example with \file{bib2gls-en.xml} the definition is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglspagename}}\marg{Page}
\end{codeenv}

\cssection{bibglspagesname}

\formatdef{bibglspagesname}
If \csopt[true]{loc-prefix} is used, then this command is provided
using the value of \idx{tag.pages} from the \langxml. For example with \file{bib2gls-en.xml} the definition is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglspagesname}}\marg{Pages}
\end{codeenv}

\cssection{bibglslocsuffix}

\formatdef{bibglslocsuffix}
If the \csopt{loc-suffix} option is on, this command will be
provided. If the glossary type has been provided by \csopt{type}
(and \csopt{dual-type} if there are any dual entries) then the
definition of \gls{bibglslocsuffix} will be appended to the glossary
preamble for the given type (or types if there are dual entries).

This commands definition depends on the value provided by
\csopt{loc-suffix}. For example, with 
\csopt[\ics{at}{\idx{periodchar}}]{loc-suffix}
the command is defined as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglslocsuffix}}[1]\marg{\ics{at}.}
\end{codeenv}
(which ignores the argument).

Whereas with \csopt[\meta{A}\dcomma\meta{B}\dcomma\meta{C}]{loc-suffix}
the command is defined as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglslocsuffix}}[1]\marg{\cs{ifcase}\idx{param}1 \meta{A}\cmd{or} \meta{B}\cmd{else} \meta{C}\cmd{fi}}
\end{codeenv}

Note that this is slightly different from \gls{bibglslocprefix} as
it includes the 0 case, which in this instance means that there were
no locations but there was a cross-reference. This command isn't
added when the location list is empty.

\cssection{bibglsprimary}

\formatdef{bibglsprimary}
When the \csopt{save-primary-locations} option is used, the primary
locations are stored in the \field{primarylocations} field
encapsulated with this command. The first argument is the number of
locations in the list. The second argument is the list of locations
formatted in the usual way. The default definition is to ignore the
first argument and simply do the second.

\cssection{bibglslocationgroup}

\formatdef{bibglslocationgroup}
When the \csopt{loc-counters} option is used, the locations
for each entry are grouped together according to the counter
(in the order specified in the value of \csopt{loc-counters}).
Each group of locations is encapsulated within
\gls{bibglslocationgroup}, where \meta{n} is the number
of locations within the group, \meta{counter} is the
counter name and \meta{list} is the formatted location sub-list.
By default, this simply does \meta{list}, but may be
defined (before the resources are loaded) or redefined 
(after the resources are loaded) as required.

For example:
\begin{codeenv}
\cs{newcommand}*\marg{\gls{bibglslocationgroup}}[3]\marg{\comment{}
  \cs{ifnum}\idx{param}1=1
   \idx{param}2:
  \cmd{else}
   \idx{param}2s:
  \cmd{fi}
  \idx{param}3\comment{}
}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[equation,page]{loc-counters},\comment{group locations by counter}
  \csopt[entries]{src}\comment{data in entries.bib}
}
\end{codeenv}
This will prefix each group with the counter name, if there's
only one location, or the counter name followed by \qt{s},
if there are multiple locations within the group.

There are various ways to adapt this to translate the counter
name to a different textual label, such as:
\begin{codeenv}
\cs{providecommand}\marg{\cmd{pagename}}\marg{Page}
\cs{providecommand}\marg{\cmd{pagesname}}\marg{Pages}
\cs{providecommand}\marg{\cmd{equationname}}\marg{Equation}
\cs{providecommand}\marg{\cmd{equationsname}}\marg{Equations}
\strut
\cs{newcommand}*\marg{\gls{bibglslocationgroup}}[3]\marg{\comment{}
  \ics{ifnum}\idx{param}1=1
   \ics{ifcsdef}\marg{\idx{param}2name}\marg{\cs{csuse}\marg{\idx{param}2name}}\marg{\idx{param}2}:
  \cmd{else}
   \cs{ifcsdef}\marg{\idx{param}2sname}\marg{\cs{csuse}\marg{\idx{param}2sname}}\marg{\idx{param}2s}:
  \cmd{fi}
  \idx{param}3\comment{}
}
\end{codeenv}

\cssection{bibglslocationgroupsep}

\formatdef{bibglslocationgroupsep}
When the \csopt{loc-counters} option is set, this command
is used to separate each location sub-group. It may be defined
before the resources are loaded:
\begin{codeenv}
\cs{newcommand}*\marg{\gls{bibglslocationgroupsep}}\marg{; }
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[equation,page]{loc-counters},\comment{group locations by counter}
  \csopt[entries]{src}\comment{data in entries.bib}
}
\end{codeenv}
or redefined after the resources are loaded:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[equation,page]{loc-counters},\comment{group locations by counter}
  \csopt[entries]{src}\comment{data in entries.bib}
}
\strut
\cs{glsrenewcommand}*\marg{\gls{bibglslocationgroupsep}}\marg{; }
\end{codeenv}

\cssection{bibglssupplemental}

\formatdef{bibglssupplemental}
When the \csopt{supplemental-locations} option is used, the locations
from a supplementary document are encapsulated within the \meta{list}
part of \gls{bibglssupplemental}. The first argument \meta{n}
(ignored by default) is the number of supplementary locations.

If multiple supplemental sources are permitted (that is, 
\bibgls\ has detected that the document is using at least
version 1.36 of \sty{glossaries-extra}), then the \meta{list}
part will consist of sub-lists for each external source. In this
case, \meta{n} will be the total number of elements across
all the sub-lists.

\cssection{bibglssupplementalsublist}

\formatdef{bibglssupplementalsublist}
If multiple supplemental sources are permitted, this will be used
to format each sub-list, where \meta{n} (ignored by default)
is the number of elements in the sub-list, \meta{external document}
(ignored by default) is the external source and \meta{list}
is the list of supplementary locations in \meta{external document}.

\cssection{bibglssupplementalsep}

\formatdef{bibglssupplementalsep}
The separator between the main location list and the supplementary
location list. By default this is just \gls!{bibglsdelimN}. This may be
defined before the resources are loaded:
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglssupplementalsep}}\marg{; }
\strut
\gls{GlsXtrLoadResources}\oarg{
 \csopt[supplDoc]{supplemental-locations},
 \csopt[entries]{src}}
\end{codeenv}
or redefined after the resources are loaded:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[supplDoc]{supplemental-locations},
 \csopt[entries]{src}}
\strut
\cs{glsrenewcommand}\marg{\gls{bibglssupplementalsep}}\marg{; }
\end{codeenv}

\cssection{bibglssupplementalsubsep}

\formatdef{bibglssupplementalsubsep}
The separator between the supplementary
location sub-lists. By default this is just \gls!{bibglsdelimN}.

\cssection{bibglshrefchar}

\formatdef{bibglshrefchar}
Expands to a literal percent character followed by \meta{hex}. The
second argument is ignored.

\cssection{bibglshrefunicode}

\formatdef{bibglshrefunicode}
Expands to the second argument. The first argument is ignored.

\section{Letter Groups}
\label{sec:lettergroupdefs}

The commands listed in this section are provided for use with the
\longarg{group} switch and glossary styles that display the letter
group title. If these need their definitions altered, they should
be defined before the resource file is loaded if field expansion
is on (\longarg{expand-fields}) otherwise they may be redefined
afterwards.

The base \isty{glossaries} package determines group titles through a fairly
simplistic rule. Both \idx!{makeindex} and \idx!{xindy} write the
line:
\nosecformatdef{glsgroupheading}
to the associated glossary file at the start of each new letter group.
For example, the \qt{A} letter group will be written as:
\begin{codeenv}
\gls{glsgroupheading}\marg{A}
\end{codeenv}
This is quite straightforward and the heading title can just be
\qt{A}. The \qt{Symbols} group is written as:
\begin{codeenv}
\gls{glsgroupheading}\marg{glssymbols}
\end{codeenv}
To allow for easy translation, the base \sty{glossaries} package has
the simple rule: 
\begin{itemize}
\item if \csfmt{\meta{heading}groupname} exists use that;
\item otherwise just use \meta{heading}.
\end{itemize}
There's no \csfmt{Agroupname} provided, but \ics{glssymbolsgroupname}
is provided and is supported by the associated language modules,
such as \styfmt{glossaries-french}. (Similarly for the \qt{Numbers}
group.)

The glossary styles that provide hyperlinks to the groups (such as 
\glostyle{indexhypergroup}) use \meta{heading} to form the target
name. A problem arises when active characters occur in
\meta{heading}, which happens with extended characters and
\isty{inputenc}.

The \sty{glossaries-extra} package (as from version 1.14) provides:
\nosecformatdef{glsxtrsetgrouptitle}
to set the title for a group with the given label. The internal
workings of \gls{glsgroupheading} are modified to use a slightly
altered rule:
\begin{itemize}
\item if a title has been set using
\gls{glsxtrsetgrouptitle}\margm{heading}\margm{title} for the given
\meta{heading}, use that;
\item if \csfmt{\meta{heading}groupname} exists, use that;
\item just use \meta{heading} for the title.
\end{itemize}
So if \gls{glsxtrsetgrouptitle} hasn't been used, it falls back on
the original rule.

The problem is now how to make the indexing application use the
desired label in the argument of \gls{glsgroupheading} instead of selecting
the heading based on the first character of each sort value for each
top-level entry in that group. This can't be done with
\idx!{makeindex}, and with \idx!{xindy} it requires a custom language
module, which isn't a trivial task.

With \bibgls, a different approach is used. The \iext{glstex} file
created isn't comparable to the \iext{gls} file created by
\idx!{makeindex} or \idx!{xindy}. There's nowhere for \bibgls\ to
write the \gls{glsgroupheading} line as it isn't creating the code
that typesets the glossary list. Instead it's creating the code that
defines the entries. The actual group heading is inserted by 
\ics{printunsrtglossary} and it's only able to do this by checking if
the entry has a \field{group} field and comparing it to the previous
entry's \field{group} field. 

The behaviour of the group formation implemented by the sort
methods may be changed with \csopt{group-formation}. With any
setting other than \csopt[default]{group-formation}, the group label
is set to \format{bibglsunicodegroup} and the title is set to
\format{bibglsunicodegrouptitle} (see below) otherwise the label and
title are determined by the sort method.

The collators used by the locale and letter-based rules save the
following information for each entry based on the first significant letter of
the \field{sort} field (if the letter is recognised as alphabetical,
according to the rule):
\begin{itemize}
\item \meta{title} The group's title. This is typically title-cased.
For example, if the rule recognises the digraph \qt{dz}, then the
title is \qt{Dz}. Exceptions to this are included in the \langxml.
If the key \idx{grouptitle.case.lc} exists, where
\meta{lc} is the \idx!{lowercase} version of \meta{title}, then the value
of that key is used instead. For example, the Dutch digraph \qt{ij}
should be converted to \qt{IJ}, so \file{bib2gls-en.xml} includes:
\begin{verbatim}
<entry key="grouptitle.case.ij">IJ</entry>
\end{verbatim}
(See the \longarg{group} switch for more details.)
\item \meta{letter} This is the actual letter at the start of the
given entry's \field{sort} field, which may be \idx!{lowercase} or may
contain diacritics that don't appear in \meta{title}.
\item \meta{id} A numeric identifier. This may be the collation key
or the code point for the given letter, depending on the sort
method.
\item \meta{type} The entry's glossary type. If not known, this will
be empty. (\bibgls\ won't know if you've modified the associated 
\csfmt{bibglsnew\ldots} command to set the \field{type}. It can only
know the type if it's in the original \ext{bib} definition or is set
using resource options such as \csopt{type}.)
\end{itemize}

The \field{group} field is then set using:
\begin{codeenv}
\field{group}=\marg{\gls{bibglslettergroup}\margm{title}\margm{letter}\margm{id}\margm{type}}
\end{codeenv}
This field needs to expand to a simple label, which \gls!{bibglslettergroup}
is designed to do.  Note that non-letter groups are dealt with
separately (see below).

\cssection{bibglssetlastgrouptitle}

In the last resource (\ext{glstex}) file, after all the relevant
group titles have been set with the commands listed below, there's a
final title setting:
\formatdef{bibglssetlastgrouptitle}
This does nothing by default, but the arguments are set to
correspond to the group with the maximum id for that resource file.
It's provided as a convenient way of overriding the final group
title without the inconvenience of looking up the group label in the
\ext{glstex} file. If you have multiple glossaries or if you want to
override a different group, then you need to inspect the
\ext{glstex} file to work out the corresponding label (by finding
the \field{group} assignment for one of the entries in that group).

The \meta{cs} argument is the control sequence used in the
\field{group} field to obtain the label from \meta{specs}. For
example, if the highest \meta{id} is 2147418112 from:
\begin{codeenv}
\field{group}=\marg{\gls{bibglslettergroup}\marg{Ø}\marg{Ø}\marg{2147418112}\marg{}}
\end{codeenv}
then the last group is identified with:
\begin{codeenv}
\gls{bibglssetlastgrouptitle}\marg{\gls{bibglslettergroup}}\marg{\marg{Ø}\marg{Ø}\marg{2147418112}\marg{}}
\end{codeenv}
In this case \meta{cs} is \gls{bibglslettergroup} and \meta{specs}
are the arguments for that command. If you want \gls{bibglssetlastgrouptitle}
to change the group title then you need to define it before the
\idx{resourceset}. For example:
\begin{codeenv}
\cs{newcommand}*\marg{\gls{bibglssetlastgrouptitle}}[2]\marg{\comment{}
 \cs{glsxtrsetgrouptitle}\marg{\idx{param}1\idx{param}2}\marg{Foreign Words}}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\end{codeenv}
If you need to change a particular group title, then it has to be
done after the \idx{resourceset}:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\cs{glsxtrsetgrouptitle}
 \marg{\gls{bibglslettergroup}\marg{\marg{Ø}\marg{Ø}\marg{2147418112}\marg{}}}\comment{label}
 \marg{Foreign Words}\comment{title}
\end{codeenv}

\cssection{bibglssetlettergrouptitle}

For each \idx{lettergroup} that's detected, \bibgls\ will write the line:
\formatdef{bibglssetlettergrouptitle}
in the \ext{glstex} file, which sets the group's title using:
\begin{codeenv}
\format{glsxtrsetgrouptitle}
\end{codeenv}
where the \meta{group label} part matches the corresponding \field{group}
value.

Note that \gls{bibglssetlettergrouptitle} only has a single
argument, but that argument contains the four arguments needed by
\gls!{bibglslettergroup} and \gls!{bibglslettergrouptitle}.
These arguments are as described above.

If \gls{glsxtrsetgrouptitle} has been defined (\sty{glossaries-extra}
version 1.14 onwards), then \gls{bibglssetlettergrouptitle} will be
defined as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglssetlettergrouptitle}}[1]\marg{\comment{}
  \cs{glsxtrsetgrouptitle}\marg{\gls{bibglslettergroup}\idx{param}1}\marg{\gls{bibglslettergrouptitle}\idx{param}1}}
\end{codeenv}
If an earlier version of \sty{glossaries-extra} is used, then this
function can't be supported and the command will be defined to
simply ignore its argument. This will fall back on the original
method of just using \meta{title} as the label.

Since \gls{bibglssetlettergrouptitle} is used in the \ext{glstex} file to
set the group titles, the associated commands need to be defined
before the resource file is loaded if their definitions require
modification. After the resource file has been loaded, you can
adjust the title of a specific group, but you'll need to check the
\ext{glstex} file for the appropriate arguments. For example, if the
\ext{glstex} file contains:
\begin{codeenv}
\gls{bibglssetlettergrouptitle}\marg{\marg{\AE}\marg{\ae}\marg{7274496}\marg{}}
\end{codeenv}
but you actually want the group title to appear as \qt{\AE\ (AE)}
instead of just \qt{\AE}, then after the resource file has been
loaded you can do:
\begin{codeenv}
\gls{glsxtrsetgrouptitle}
 \marg{\gls{bibglslettergroup}\marg{\AE}\marg{\ae}\marg{7274496}\marg{}}\comment{label}
 \marg{\AE\ (AE)}\comment{title}
\end{codeenv}

\cssection{bibglslettergroup}

\formatdef{bibglslettergroup}
This command is used to determine the \idx{lettergroup} label. The
default definition is \meta{type}\meta{id}, which ensures that no
problematic characters occur in the label since \meta{type} can't
contains special characters and \meta{id} is numeric. The
\meta{type} is included in case there are multiple glossaries, since
the hyperlink name must be unique.

\cssection{bibglslettergrouptitle}

\formatdef{bibglslettergrouptitle}
This command is used to determine the \idx{lettergroup} title.
The default definition is \ics{unexpanded}\margm{title}, which guards
against any expansion issues that may arise with characters outside
the basic Latin set.

For example:
\begin{codeenv}
\atentry{entry}\marg{angstrom,
  \field{name}=\marg{\cs{AA} ngstr\cs{umlaut}om}
  \field{description}=\marg{a unit of length equal to one hundred-millionth 
of a centimetre}
}
\end{codeenv}
The \field{sort} value is \qtt{\AA ngstr\"om}. With \csopt[en]{sort}
the \meta{title} part will be \code{A} but with \csopt[sv]{sort}
the \meta{title} part will be \code{\AA}. In both cases the
\meta{letter} argument will be \code{\AA}.

Take care if you are using a script that needs encapsulating. For
example, with the \isty{CJKutf8} package the \idx{CJK} characters need to
be placed within the \env{CJK} environment, so any letter group
titles that contain \idx{CJK} characters will need special attention.

For example, suppose the \ext{bib} file contains entries in the
form:
\begin{codeenv}
\atentry{dualentry}\marg{\meta{label},
  \field{name} = \marg{\gls{cjkname}\margm{\idx{CJK} characters}},
  \field{description} = \margm{English translation}
}
\end{codeenv}
and the document contains:
\begin{codeenv}
\cmd{usepackage}\marg{CJKutf8}
\cmd{usepackage}[\styopt{record},\styopt[indexgroup]{style},\styopt{nomain}]\marg{glossaries-extra}
\strut
\cs{newglossary*}\marg{japanese}\marg{Japanese to English}
\cs{newglossary*}\marg{english}\marg{English to Japanese}
\strut
\ics{newrobustcmd}\marg{\gls{cjkname}}[1]\marg{\cmd{begin}\marg{CJK}\marg{UTF8}\marg{min}\idx{param}1\cmd{end}\marg{CJK}}
\strut
\gls{GlsXtrLoadResources}\oarg{
  \csopt[testcjk]{src},\comment{bib file}
  \csopt[ja-JP]{sort},\comment{locale used to sort primary entries}
  \csopt[en-GB]{dual-sort},\comment{locale used to sort secondary entries}
  \csopt[japanese]{type},\comment{put the primary entries in the 'japanese' glossary}
  \csopt[english]{dual-type},\comment{put the dual entries in the 'english' glossary}
  \csopt[en.]{dual-prefix}
}
\end{codeenv}
then \idx{CJK} characters will appear in the \meta{title} argument of 
\gls{bibglslettergrouptitle} which causes a problem because they need
to be encapsulated within the \env{CJK} environment. This can be more
conveniently done with the user supplied \inlinedef{cjkname}, but the \idx{CJK}
characters need to be protected from expansion so \ics{unexpanded} is
also needed. The new definition of \gls{bibglslettergrouptitle} needs
to be defined before \gls{GlsXtrLoadResources}. For example:
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglslettergrouptitle}}[4]\marg{\cs{unexpanded}\marg{\gls{cjkname}\marg{\idx{param}1}}}
\end{codeenv}
There's a slight problem here in that the English letter group titles 
also end up encapsulated. An alternative approach is to use the
\meta{type} part to provide different forms. For example:
\begin{codeenv}
\cs{newcommand}*\marg{\cmd{englishlettergroup}}[1]\marg{\idx{param}1}
\cs{newcommand}*\marg{\cmd{japaneselettergroup}}[1]\marg{\gls{cjkname}\marg{\idx{param}1}}
\cs{newcommand}\marg{\gls{bibglslettergrouptitle}}[4]\marg{\comment{}
 \cs{unexpanded}\marg{\cs{csuse}\marg{\idx{param}4lettergroup}\marg{\idx{param}1}}}
\end{codeenv}

\cssection{bibglssetothergrouptitle}

The label and title for \idxpl{symbolgroup}
are dealt with in a similar way to the \idxpl!{lettergroup}, but in this
case the title is set using:
\formatdef{bibglssetothergrouptitle}
This is defined in an analogous manner:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglssetothergrouptitle}}[1]\marg{\comment{}
 \cs{glsxtrsetgrouptitle}\marg{\gls{bibglsothergroup}\idx{param}1}\marg{\gls{bibglsothergrouptitle}\idx{param}1}}
\end{codeenv}
where the group label is obtained using \gls!{bibglsothergroup} and
the group title is obtained from \gls!{bibglsothergrouptitle}.
Note that since non-alphabetic characters don't have upper or lower
case versions, there are only three arguments. The other difference
between this and the letter group version is that the \meta{id} is
given in hexadecimal format (corresponding to the character code).

For example, suppose my \ext{bib} file contains:
\begin{codeenv}
\atentry{entry}\marg{sauthor,
  \field{name}=\marg{/Author},
  \field{description} = \marg{author string}
}
\end{codeenv}
If a locale sort is used, the leading slash \code{/} will be
ignored and this entry will belong to the \qt{A} letter group using
the letter commands described above. If, instead, one of the
character code sort methods are used, such as
\csopt[letter-case]{sort}, then this entry will be identified as
belonging to a symbol (or \qt{other}) group and the title will be
set using:
\begin{codeenv}
\gls{bibglssetothergrouptitle}\marg{\marg{/}\marg{2F}\marg{}}
\end{codeenv}

\cssection{bibglsothergroup}

\formatdef{bibglsothergroup}
This expands to the label for \idxpl{symbolgroup}. This just
defaults to \code{glssymbols} (ignoring all arguments), which
replicates the label used when \idx!{makeindex} or \idx!{xindy}
generate the glossary files.

\cssection{bibglsothergrouptitle}

\formatdef{bibglsothergrouptitle}
This expands to the title for \idxpl{symbolgroup}. This just
expands to \ics{glssymbolsgroupname} by default.

\cssection{bibglssetemptygrouptitle}

Used when the sort value degenerates to an empty string. This command
sets the label and title.
\formatdef{bibglssetemptygrouptitle}
(Note the inner group, as with the other similar 
\csfmt{bibglsset\ldots grouptitle} commands.)

\cssection{bibglsemptygroup}

\formatdef{bibglsemptygroup}
This expands to the label for \idxpl{emptygroup}. This
defaults to \code{glssymbols} to make it consistent with 
\idxpl{nonlettergroup} (since the sort value likely contained
unknown symbol commands).

\cssection{bibglsemptygrouptitle}

\formatdef{bibglsemptygrouptitle}
This expands to the group title for \idx{emptygroup}. This just
expands to \ics{glssymbolsgroupname} by default.

\cssection{bibglssetnumbergrouptitle}

The numeric sort methods (\tableref{tab:sortoptionsnumerical}) 
all create \idxpl{numbergroup} instead of letter
or symbol groups. These behave in an analogous way to the above.
\formatdef{bibglssetnumbergrouptitle}
In this case \meta{value} is the actual numeric sort value, and
\meta{id} is a decimal number obtained from converting \meta{value}
to an integer. This command is defined as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglssetnumbergrouptitle}}[1]\marg{\comment{}
  \cs{glsxtrsetgrouptitle}\marg{\gls{bibglsnumbergroup}\idx{param}1}\marg{\gls{bibglsnumbergrouptitle}\idx{param}1}}
\end{codeenv}

\cssection{bibglsnumbergroup}

The \idx{numbergroup} label is obtained from:
\formatdef{bibglsnumbergroup}
This just defaults to \code{glsnumbers}.

\cssection{bibglsnumbergrouptitle}

The \idx{numbergroup} title is obtained from:
\formatdef{bibglsnumbergrouptitle}
This just defaults to \ics{glsnumbersgroupname}.

\cssection{bibglssetdatetimegrouptitle}

The date-time sort methods (\tableref{tab:sortoptionsdatetime}) 
create \idxpl{datetimegroup}. These behave in an analogous way to the above.
\formatdef{bibglssetdatetimegrouptitle}
This command is defined as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglssetdatetimegrouptitle}}[1]\marg{\comment{}
 \cs{glsxtrsetgrouptitle}
  \marg{\gls{bibglsdatetimegroup}\idx{param}1}\comment{}
  \marg{\gls{bibglsdatetimegrouptitle}\idx{param}1}\comment{}
}
\end{codeenv}

\cssection{bibglsdatetimegroup}

\formatdef{bibglsdatetimegroup}

This command is used for \idx{datetimegroup} labels with \optfmt{datetime} sorting
(\tableref{tab:sortoptionsdatetime}). This has ten arguments, which means a
little trickery is needed to deal with the tenth argument.
The default definition is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsdatetimegroup}}[9]\marg{\idx{param}1\idx{param}2\idx{param}3\ics{@firstofone}}
\end{codeenv}
This forms the group label from the year \meta{YYYY}, month \meta{MM}, 
day \meta{DD} and \meta{type}.

\cssection{bibglsdatetimegrouptitle}

\formatdef{bibglsdatetimegrouptitle}

This command is used for \idx{datetimegroup} titles with \optfmt{datetime} sorting
(\tableref{tab:sortoptionsdatetime}).
The default definition is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsdatetimegrouptitle}}[9]\marg{\idx{param}1-\idx{param}2-\idx{param}3\ics{@gobble}}
\end{codeenv}
This sets the title to the numeric
\code{\meta{YYYY}-\meta{MM}-\meta{DD}}
but may be redefined as appropriate.

\cssection{bibglssetdategrouptitle}

The date sort methods (\tableref{tab:sortoptionsdatetime}) 
create \idxpl{dategroup} (the time isn't included). These behave in an 
analogous way to the above.
\formatdef{bibglssetdategrouptitle}
This command is defined as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglssetdategrouptitle}}[1]\marg{\comment{}
 \cs{glsxtrsetgrouptitle}\marg{\gls{bibglsdategroup}\idx{param}1}\marg{\gls{bibglsdategrouptitle}\idx{param}1}}
\end{codeenv}

\cssection{bibglsdategroup}

\formatdef{bibglsdategroup}

This command is used for \idx{dategroup} labels with \optfmt{date} (no time) sorting 
(\tableref{tab:sortoptionsdatetime}). The default definition is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsdategroup}}[7]\marg{\idx{param}1\idx{param}2\idx{param}4\idx{param}7}
\end{codeenv}
This forms the group label from the year, month, era and type.
In this case, the era is a textual representation not the numeric
value used in calculating the sort value.

\cssection{bibglsdategrouptitle}

\formatdef{bibglsdategrouptitle}
This command is used for \idx{dategroup} titles with \optfmt{date} (no time) sorting
(\tableref{tab:sortoptionsdatetime}).  The default definition is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsdategrouptitle}}[7]\marg{\idx{param}1-\idx{param}2}
\end{codeenv}
This just sets the title to the numeric year-month form 
\code{\meta{YYYY}-\meta{MM}}.

\cssection{bibglssettimegrouptitle}

The time sort methods (\tableref{tab:sortoptionsdatetime}) 
create \idxpl{timegroup} (the date isn't included). These behave in an 
analogous way to the above.
\formatdef{bibglssettimegrouptitle}
This command is defined as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglssettimegrouptitle}}[1]\marg{\comment{}
 \cs{glsxtrsetgrouptitle}\marg{\gls{bibglstimegroup}\idx{param}1}\marg{\gls{bibglstimegrouptitle}\idx{param}1}}
\end{codeenv}

\cssection{bibglstimegroup}

\formatdef{bibglstimegroup}
This command is used for \idx{timegroup} labels with \optfmt{time} (no date) sorting
(\tableref{tab:sortoptionsdatetime}).
This command is defined as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglstimegroup}}[7]\marg{\idx{param}1\idx{param}2\idx{param}7}
\end{codeenv}

\cssection{bibglstimegrouptitle}

\formatdef{bibglstimegrouptitle}

This command is used for \idx{timegroup} titles with \optfmt{time} (no date) sorting
(\tableref{tab:sortoptionsdatetime}).
This command is defined as:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglstimegrouptitle}}[7]\marg{\idx{param}1}
\end{codeenv}

\cssection{bibglssetunicodegrouptitle}

\formatdef{bibglssetunicodegrouptitle}
This command is used to assign the group titles when the group
formation is set to any value other than the default. For example,
this command will be used with \csopt[codepoint]{group-formation}.
The label is obtained from \gls{bibglsunicodegroup} and the title is
obtained from \gls{bibglsunicodegrouptitle}.

\cssection{bibglsunicodegroup}

\formatdef{bibglsunicodegroup}
The \meta{label} depends on the \csopt{group-formation} setting:
\begin{itemize}
\item\csopt[codepoint]{group-formation}: the \meta{label} is the
Unicode value of \meta{character} (converted to \idx!{lowercase} and
decomposed, if applicable);
\item\csopt[unicode category]{group-formation}: the \meta{label} is the
Unicode category of \meta{character} (for example, \code{Lu} means
an \idx!{uppercase} letter);
\item\csopt[unicode script]{group-formation}: the \meta{label} is the
Unicode script associated with \meta{character} (for example,
\code{LATIN});
\item\csopt[unicode category and script]{group-formation}: the \meta{label}
identifies both the Unicode category and script associated with 
\meta{character} (for example, \code{Lu.LATIN}).
\end{itemize}
(Similarly for \csopt{secondary-group-formation} and
\csopt{dual-group-formation}.) By default this command expands to
\meta{type}\meta{label}.

The \meta{character} is the first significant character of the sort
value. The \meta{id} is the hexadecimal code of (possibly decomposed) 
\meta{character}. The case of codepoint \meta{id} may or may not correspond 
to the case of \meta{character}.

For example, with \csopt[codepoint]{group-formation}, an unset
\field{type} and a sort value of \qt{\AA ngstr\"om} with \qt{\AA} as a
significant character distinct from \qt{A} then the \field{group} 
field will be assigned using:
\begin{codeenv}
\field{group}=\marg{\gls{bibglsunicodegroup}\marg{å}\marg{Å}\marg{C5}\marg{}}
\end{codeenv}
whereas with \csopt[unicode category and script]{group-formation} it will be:
\begin{codeenv}
\field{group}=\marg{\gls{bibglsunicodegroup}\marg{Lu.LATIN}\marg{Å}\marg{C5}\marg{}}
\end{codeenv}
(\idx!{uppercase} Latin letter). 

If instead \qt{\AA} is considered equivalent to \qt{A} according to the 
collator, then with \csopt[codepoint]{group-formation}, the value will be:
\begin{codeenv}
\field{group}=\marg{\gls{bibglsunicodegroup}\marg{a}\marg{Å}\marg{61}\marg{}}
\end{codeenv}
Note that the \meta{id} is now \hex{61} (the decomposed \qt{A}
converted to \idx{lowercase}) not \hex{C5}.

\cssection{bibglsunicodegrouptitle}

\formatdef{bibglsunicodegrouptitle}
The title for Unicode group formations is simply defined as 
\code{\cs{unexpanded}\margm{label}} so you will need to change it to
something more appropriate. For example (before the \idx{resourceset}):
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglsunicodegrouptitle}}[4]\marg{\comment{}
  \ifnumhex{\idx{param}3}>64 
   \ifnumhex{\idx{param}3} < 91 
     A-{}-Z\comment{}
   \cmd{else} 
     \ifnumhex{\idx{param}3} > 96
       \ifnumhex{\idx{param}3} < 123
         A-{}-Z\comment{}
       \cmd{fi}
     \cmd{fi}
   \cmd{fi}
  \cmd{fi}
}
\end{codeenv}
This will make the title \qt{A--Z} if \meta{id} is greater
than 64 and less than 91 or greater than 96 and less than 123 
(and will be empty otherwise).

Note that this setting can create an odd effect if the sorting
causes the groups to be split up. For example, if some of the sort values
start with extended or non-Latin characters this can break up the
groups. First check how the group labels are assigned using:
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglsunicodegrouptitle}}\marg{\gls{bibglsunicodegroup}}
\end{codeenv}
then adjust the definition of \gls{bibglsunicodegroup} until the
grouping is correct, and then change the definition of 
\gls{bibglsunicodegrouptitle} so that the title is correct.

\cssection{bibglshypergroup}

\formatdef{bibglshypergroup}

If the \iext{log} file indicates that \isty{hyperref} has been loaded
and the \longarg{group} switch is used, then this command will be
used to create the navigation information for glossary styles such
as \glostyle{indexhypergroup}.

\section{Flattened Entries}
\label{sec:flattendefs}

These commands relate to the way the \field{name} field is altered
when flattening lonely child entries with the \csopt{flatten-lonely}
option.

\cssection{bibglsflattenedhomograph}

\formatdef{bibglsflattenedhomograph}
The default definition simply does \meta{name}.

This command is used if the child and parent name's are identical.
For example, suppose the \ext{bib} file contains:
\begin{codeenv}
\atentry{index}\marg{super.glossary, \field{name}=\marg{glossary}}
\strut
\atentry{entry}\marg{glossarycol,
  \field{parent}=\marg{super.glossary},
  \field{description}=\marg{collection of glosses}
}
\strut
\atentry{entry}\marg{glossarylist,
  \field{parent}=\marg{super.glossary},
  \field{description}=\marg{list of technical words}
}
\end{codeenv}
The child entries don't have a \field{name} field, so the value is assumed
to be the same as the parent's \field{name} field. Here's an example 
document where both child entries are used:
\begin{codeenv}
\cmd{documentclass}\marg{article}
\strut
\cmd{usepackage}[\styopt{record},\istyopt{subentrycounter},\styopt[treenoname]{style}]\marg{glossaries-extra}
\strut
\gls{GlsXtrLoadResources}\oarg{\csopt[entries]{src}}
\strut
\cmd{begin}\marg{document}
\cs{gls}\marg{glossarycol} (collection) vs \cs{gls}\marg{glossarylist} (list).
\strut
\cs{printunsrtglossary}
\cmd{end}\marg{document}
\end{codeenv}
This uses one of the glossary styles designed for homographs and the
glossary has the structure:
\begin{flushleft}
\textbf{glossary}\par
\quad 1) collection of glosses 1\par
\quad 2) list of technical words 1
\end{flushleft}
If only one child entry is selected, then the result looks a little
odd. For example:
\begin{flushleft}
\textbf{glossary}\par
\quad 1) collection of glosses 1
\end{flushleft}
With the \csopt{flatten-lonely} option, the parent is removed and
the child is moved up a hierarchical level. With 
\csopt[postsort]{flatten-lonely} this would normally adjust the
name so that it appears as \meta{parent name}, \meta{child name}
but in this case it would look a little odd for the name to 
appear as \qt{glossary, glossary} so instead the name is
set to:
\begin{codeenv}
\gls{bibglsflattenedhomograph}\marg{glossary}\marg{super.glossary}
\end{codeenv}
(where the first argument is the original name and the second argument is the
label of the parent entry).

This means that the name simply appears as \qt{glossary}, even if
the \csopt[postsort]{flatten-lonely} option is used. Note that if
the parent entry is removed, the parent label won't be of much use.
You can test for existence using \ics{ifglsentryexists} in case you
want to vary the way the name is displayed according to whether or
not the parent is still present.

\cssection{bibglsflattenedchildpresort}

\formatdef{bibglsflattenedchildpresort}

Used by the \csopt[presort]{flatten-lonely} option. This defaults
to just \meta{child name}. If you want to change this, remember that
you can let the interpreter know by adding the definition to
\atentry{preamble}. For example:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\gls{bibglsflattenedchildpresort}}[2]\marg{\idx{param}1 (\idx{param}2)}}}
\end{codeenv}

\cssection{bibglsflattenedchildpostsort}

\formatdef{bibglsflattenedchildpostsort}

Used by the \csopt[postsort]{flatten-lonely} option. This defaults
to \code{\meta{parent name}, \meta{child name}}.

Note that the arguments are in the reverse order to those of the
previous command. This is done to assist the automated first letter
upper-casing. If either command is redefined to alter the ordering,
then this can confuse the case-changing mechanism, in which case you
may want to consider switching on the expansion of the \field{name}
field using:
\begin{codeenv}
\gls{glssetexpandfield}\marg{name}
\end{codeenv}
(before \gls{GlsXtrLoadResources}).

\section{Other}

\cssection{bibglshyperlink}

\formatdef{bibglshyperlink}
Used by the \hyperref[sec:dualoptsbacklinks]{back link options},
this just defaults to: 
\begin{codeenv}
\ics{glshyperlink}\oargm{text}\margm{label}
\end{codeenv}

\cssection{bibglssetwidest}

\formatdef{bibglssetwidest}
This is used by \csopt{set-widest} to set the widest name for the
given hierarchical level where the glossary type can't be determined.
This is defined as:
\begin{codeenv}
\cs{providecommand}*\marg{\gls{bibglssetwidest}}[2]\marg{\cs{glsxtrSetWidest}\marg{}\marg{\idx{param}1}\marg{\idx{param}2}}
\end{codeenv}
if \ics{glsxtrSetWidest} has been defined, or:
\begin{codeenv}
\cs{providecommand}*\marg{\gls{bibglssetwidest}}[2]\marg{\cs{glsupdatewidest}\oarg{\idx{param}1}\marg{\idx{param}2}}
\end{codeenv}
if \ics{glsupdatewidest} is defined, otherwise it will be
defined to use \ics{glssetwidest}:
\begin{codeenv}
\cs{providecommand}*\marg{\gls{bibglssetwidest}}[2]\marg{\cs{glssetwidest}\oarg{\idx{param}1}\marg{\idx{param}2}}
\end{codeenv}
Since this isn't scoped, this will affect other glossaries.
In general, if you have more than one glossary it's best to
set the \field{type} using options like \csopt{type}.

\cssection{bibglssetwidestfortype}

\formatdef{bibglssetwidestfortype}
This is used by \csopt{set-widest} to set the widest name for the
given hierarchical level where the glossary type is known. This
is defined as:
\begin{codeenv}
\cs{providecommand}*\marg{\gls{bibglssetwidestfortype}}[3]\marg{\comment{}
  \cs{glsxtrSetWidest}\marg{\idx{param}1}\marg{\idx{param}2}\marg{\idx{param}3}\comment{}
}
\end{codeenv}
if \ics{glsxtrSetWidest} has been defined, or:
\begin{codeenv}
\cs{providecommand}*\marg{\gls{bibglssetwidestfortype}}[3]\marg{\comment{}
  \cs{apptoglossarypreamble}\oarg{\idx{param}1}\marg{\cs{glsupdatewidest}\oarg{\idx{param}2}\marg{\idx{param}3}}\comment{}
}
\end{codeenv}
if \ics{glsupdatewidest} is defined, otherwise it will be
defined to use \ics{glssetwidest}:
\begin{codeenv}
\cs{providecommand}*\marg{\gls{bibglssetwidestfortype}}[3]\marg{\comment{}
  \cs{apptoglossarypreamble}\oarg{\idx{param}1}\marg{\cs{glssetwidest}\oarg{\idx{param}2}\marg{\idx{param}3}}\comment{}
}
\end{codeenv}
Since the glossary preamble is scoped, this won't affect other
glossaries.

\cssection{bibglssetwidestfallback}

\formatdef{bibglssetwidestfallback}
This is used by \csopt{set-widest} instead of \gls{bibglssetwidest}
when all \field{name} fields end up as an empty string when
interpreted by \bibgls. This typically means that all the
\field{name} fields contain unknown commands. This fallback command
will use:
\begin{codeenv}
\ics{glsxtrSetWidestFallback}\marg{2}\margm{glossary list}
\end{codeenv}
if defined otherwise it will use \ics{glsFindWidestLevelTwo}, which
sets the widest name for the top-level and first two sub-levels
across all the listed glossaries.

\cssection{bibglssetwidestfortypefallback}

\formatdef{bibglssetwidestfortypefallback}
This is used by \csopt{set-widest} instead of
\gls{bibglssetwidestfortype}
when all \field{name} fields end up as an empty string when
interpreted by \bibgls. This typically means that all the
\field{name} fields contain unknown commands. This fallback command
will append \gls{bibglssetwidestfallback} to the glossary preamble
for the given type.

\cssection{bibglssetwidesttoplevelfallback}

\formatdef{bibglssetwidesttoplevelfallback}
This is used by \csopt{set-widest} instead of \gls{bibglssetwidest}
when all \field{name} fields end up as an empty string when
interpreted by \bibgls. This typically means that all the
\field{name} fields contain unknown commands. This fallback command
will use:
\begin{codeenv}
\ics{glsxtrSetWidestFallback}\marg{0}\margm{glossary list}
\end{codeenv}
if defined otherwise it will use \ics{glsFindWidestTopLevelName},
which sets the widest name for the top-level.

\cssection{bibglssetwidesttoplevelfortypefallback}

\formatdef{bibglssetwidesttoplevelfortypefallback}
This is used by \csopt{set-widest} instead of
\gls{bibglssetwidestfortype}
when all \field{name} fields end up as an empty string when
interpreted by \bibgls. This typically means that all the
\field{name} fields contain unknown commands. This fallback command
will append \gls{bibglssetwidesttoplevelfallback} to the glossary
preamble of the given type.

\cssection{bibglscontributorlist}

\formatdef{bibglscontributorlist}
This is used when \csopt{bibtex-contributor-fields} is set. The
definition depends on whether or not \ics{DTLformatlist} has been
defined:
\begin{codeenv}
\ics{ifdef}\cs{DTLformatlist}
\marg{\comment{datatool v2.28+}
 \cs{providecommand}*\marg{\gls{bibglscontributorlist}}[2]\marg{\cs{DTLformatlist}\marg{\idx{param}1}}
}
\marg{\comment{datatool v2.27 or earlier}
 \cs{providecommand}*\marg{\gls{bibglscontributorlist}}[2]\marg{\comment{}
  \ics{def}\cmd{bibgls@sep}\marg{}\comment{}
  \ics{@for}\cmd{bibgls@item}:=\idx{param}1\cmd{do}\marg{\cmd{bibgls@sep}\cmd{bibgls@item}\cs{def}\cmd{bibgls@sep}\marg{, }}\comment{}
 }
}
\end{codeenv}
The second argument allows you to provide definitions like:
\begin{codeenv}
\cs{newcommand}*\marg{\gls{bibglscontributorlist}}[2]\marg{\comment{}
 \cs{ifcase}\idx{param}2
 \cmd{or}
   name:
 \cmd{else}
   names:
 \cmd{fi}
 \cs{DTLformatlist}\marg{\idx{param}1}\comment{}
}
\end{codeenv}

\cssection{bibglscontributor}

\formatdef{bibglscontributor}
This is used when \csopt{bibtex-contributor-fields} is set. The
definition depends on the value of \csopt{contributor-order}. Note that if you
have multiple resource sets, that option governs the way \bibgls's version of
\gls{bibglscontributor} behaves. The definition is written to the \ext{glstex}
using \cs{providecommand}, so \LaTeX\ will only pick up the first definition.

\cssection{bibglsdatetime}

\formatdef{bibglsdatetime}
Used to encapsulate date-time fields identified with
\csopt{date-time-fields}.
Since \gls{bibglsdatetime} requires more than nine arguments, the
remaining four arguments are picked up with:
\nosecdef{bibglsdatetimeremainder}
The default definitions are:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsdatetime}}[9]\marg{\gls{bibglsdatetimeremainder}}
\cs{providecommand}\marg{\gls{bibglsdatetimeremainder}}[4]\marg{\idx{param}4}
\end{codeenv}

\cssection{bibglsdate}

\formatdef{bibglsdate}
Used to encapsulate date fields identified with
\csopt{date-fields}.
The default definition is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglsdate}}[7]\marg{\idx{param}7}
\end{codeenv}

\cssection{bibglstime}

\formatdef{bibglstime}
Used to encapsulate date fields identified with
\csopt{time-fields}.
The default definition is:
\begin{codeenv}
\cs{providecommand}\marg{\gls{bibglstime}}[7]\marg{\idx{param}7}
\end{codeenv}

\cssection{bibglsprimaryprefixlabel}

\formatdef{bibglsprimaryprefixlabel}
A hook to pick up the primary prefix label (identified with
\csopt{label-prefix}) if required. This does nothing by default. If
required, this command should be defined before the
\idx{resourceset} is loaded.

\cssection{bibglsdualprefixlabel}

\formatdef{bibglsdualprefixlabel}
A hook to pick up the dual prefix label (identified with
\csopt{dual-prefix}) if required. This does nothing by default. If
required, this command should be defined before the
\idx{resourceset} is loaded.

\cssection{bibglstertiaryprefixlabel}

\formatdef{bibglstertiaryprefixlabel}
A hook to pick up the tertiary prefix label (identified with
\csopt{tertiary-prefix}) if required. This does nothing by default. If
required, this command should be defined before the
\idx{resourceset} is loaded.

\cssection{bibglsexternalprefixlabel}

\formatdef{bibglsexternalprefixlabel}
A hook to pick up the \meta{n}th external prefix label (identified with
\csopt{ext-prefixes}) if required. This does nothing by default and
won't be used if the list of external prefixes is empty. If required,
this command should be defined before the \idx{resourceset} is
loaded.

\cssection{bibglshashchar}

\formatdef{bibglshashchar}
Expands to a literal hash character (\,\idx{hashchar}\,).

\cssection{bibglsunderscorechar}

\formatdef{bibglsunderscorechar}
Expands to a literal underscore character (\,\idx{underscorechar}\,).

\cssection{bibglsdollarchar}

\formatdef{bibglsdollarchar}
Expands to a literal dollar character (\,\idx{dollarchar}\,).

\cssection{bibglsampersandchar}

\formatdef{bibglsampersandchar}
Expands to a literal ampersand character (\,\idx{ampchar}\,).

\cssection{bibglscircumchar}

\formatdef{bibglscircumchar}
Expands to a literal circumflex character (\,\idx{circumchar}\,).

\cssection{bibglsuppercase}

\formatdef{bibglsuppercase}
Converts \meta{text} to \idx{uppercase}. This just uses
\ics{MakeTextUppercase} by default.

\cssection{bibglslowercase}

\formatdef{bibglslowercase}
Converts \meta{text} to \idx{lowercase}. This just uses
\ics{MakeTextLowercase} by default.

\cssection{bibglstitlecase}

\formatdef{bibglstitlecase}
Converts \meta{text} to \idx{titlecase}. This just uses
\ics{capitalisewords} by default.

\cssection{bibglsfirstuc}

\formatdef{bibglsfirstuc}
Converts the first letter of \meta{text} to \idx{uppercase}. This just uses
\ics{makefirstuc} by default.

\chapter{Converting Existing \iext{tex} to \iext{bib}}
\label{sec:gls2bib}
\setsecdepth{1}

If you have already been using the \styfmt{glossaries} or
\styfmt{glossaries-extra} package with a large file containing all your
definitions using commands like \gls{newglossaryentry}, then you can
use the supplementary tool \idx{convertgls2bib} to convert the definitions
to the \ext{bib} format required by \bibgls. The syntax is:
\begin{alltt}
convertgls2bib \oargm{options} \meta{tex file} \meta{bib file}
\end{alltt}
where \meta{tex file} is the \ext{tex} file and \meta{bib file} is
the \ext{bib} file. This application is less secure than \bibgls\ as
it doesn't use \idx{kpsewhich} to check \texmfcnf{openinany} and 
\texmfcnf{openoutany}. Take care not to accidentally overwrite
existing \ext{bib} files as there's no check to determine if
\meta{bib file} already exists with the default \convertglsbiblongarg{overwrite}. 

If the \ext{bib} extension is missing from \meta{bib file}, it will
be added. The extension is required for \meta{tex file}.

\section{Command Line Arguments}

The \meta{options} recognised by \idx{convertgls2bib} are described below.

\convertglsbibarg{texenc}

The character encoding of the \ext{tex} file. If omitted, the
operating system's default encoding is assumed (or the
\idx{JVM}['s]).

\convertglsbibarg{bibenc}

The character encoding of the \ext{bib} file. If omitted, the same
encoding as the \ext{tex} file is assumed.

\convertglsbibarg{space-sub}

The \ext{bib} format doesn't allow spaces in labels. If your original
definitions in your \ext{tex} file have spaces, use this option to
replace spaces in labels. Each space will be substituted with
\meta{replacement}. The cross-referencing fields, \field{see},
\field{seealso} and \field{alias}, will also be adjusted, but any
references using \ics{gls} etc will have to be substituted manually 
(or use a global search and replace in your text editor).
If you want to strip the spaces, use an empty string for
\meta{replacement}. You'll need to delimit this according to your
operating system. For example:
\begin{verbatim}
convertgls2bib --space-sub '' entries.tex entries.bib
\end{verbatim}

\convertglsbibarg{ignore-sort}

Omit the \field{sort} field.  This is the default since \bibgls\ can
work out a more intuitive sort value than either \idx!{makeindex} or
\idx!{xindy}.

\convertglsbibarg{no-ignore-sort}

Don't ignore the \field{sort} field.

\convertglsbibarg{ignore-type}

Omit the \field{type} field in the \ext{bib} file.  You may find it
more flexible not to be locked into a specific glossary type if you
have a large database of entries.

\convertglsbibarg{no-ignore-type}

 Don't omit the \field{type} field (default unless
\convertglsbiblongarg{split-on-type}).

\convertglsbibarg{split-on-type}

Split the entries into separate files according to the \field{type}
field. Any entries that have the \field{type} field set to \ics{glsdefaulttype}
or that don't have the \field{type} field set and there's no default provided
by the command used to define the entry (see below) then the
\atentryfmt{\meta{entry}} data will be written to the main \meta{bib
file}. Otherwise entries will be written to the split file
(in the same directory as \meta{bib file}). 

The split file name depends on whether or not the
\convertglsbiblongarg{split-on-category} switch has also been used.  With both
and if the category and field values are different then
the file name is \metametafilefmt{}{type}{-}{category}{.bib} otherwise
it's \metafilefmt{}{type}{.bib}.

Commands that have a default type are as follows:
\begin{itemize}
\item \gls{newabbreviation}, \gls{newacronym}, \ics{oldacronym},
\gls{newdualentry}:
the default type is assumed to be \code{abbreviations} (regardless
of the definition of \ics{acronymtype} or \ics{glsxtrabbrvtype});
\item \gls{glsxtrnewsymbol}: the default type is assumed to be \code{symbols}; 
\item \gls{glsxtrnewnumber}: the default type is assumed to be \code{numbers}; 
\item \gls{newterm}: the default type is assumed to be \code{index}. 
\end{itemize}

This option automatically implements \convertglsbiblongarg{ignore-type} and
\convertglsbiblongarg{no-overwrite}.

\convertglsbibarg{no-split-on-type}

Don't split the entries into separate files according to their type
(default).

\convertglsbibarg{ignore-category}

Omit the \field{category} field in the \ext{bib} file.

\convertglsbibarg{no-ignore-category}

Don't omit the \field{category} field (default unless
\convertglsbiblongarg{split-on-category}).

\convertglsbibarg{split-on-category}

Split the entries into separate files according to the \field{category}.
If the \field{category} field isn't present and there's no default provided
by the command used to define the entry (see below) then the
\atentryfmt{\meta{entry}} data will be written to the main \meta{bib
file}. Otherwise entries will be written to the split file
(in the same directory as \meta{bib file}). 

The split file name depends on whether or not the
\convertglsbiblongarg{split-on-type} switch has also been used.  With both and
if the category and field values are different then
the file name is \metametafilefmt{}{type}{}{category}{.bib} otherwise
it's \metafilefmt{}{category}{.bib}.

Commands that have a default category are as follows:
\begin{itemize}
\item \gls{newabbreviation}, \gls{newacronym}, \ics{oldacronym},
\gls{newdualentry}:
the default category is assumed to be \code{abbreviation};
\item \gls{glsxtrnewsymbol}: the default category is assumed to be \code{symbol}; 
\item \gls{glsxtrnewnumber}: the default category is assumed to be \code{number}; 
\item \gls{newterm}: the default category is assumed to be \code{index}. 
\end{itemize}

For example, if you have both \convertglsbiblongarg{split-on-type} and
\convertglsbiblongarg{split-on-category}, then the default file name for
\gls{newabbreviation} will be
\filefmt{abbreviations-abbreviation.bib} but the default file name
for \gls{newterm} will be \filefmt{index.bib}. Whereas if you only
have \convertglsbiblongarg{split-on-category} and not
\convertglsbiblongarg{split-on-type}, then then default file name for
\gls{newabbreviation} will be \filefmt{abbreviation.bib}.

This option automatically implements \convertglsbiblongarg{ignore-category} and
\convertglsbiblongarg{no-overwrite}.

\convertglsbibarg{no-split-on-category}

Don't split the entries into separate files according to their
category (default).

\convertglsbibarg{ignore-fields}

Omit all the fields listed in \meta{list} from the \ext{bib} file.
If \field{sort}, \field{type} or \field{category} are included
in the list, this will automatically implement the corresponding
\convertglsbiblongarg{ignore-sort},
\convertglsbiblongarg{ignore-type} or
\convertglsbiblongarg{ignore-category} option.

This option is not cumulative. If the list is empty it will unset
any previous list but won't unset any 
\convertglsbiblongarg{ignore-sort},
\convertglsbiblongarg{ignore-type} or
\convertglsbiblongarg{ignore-category} option.

For example:
\begin{alltt}
convertgls2bib \longargfmt{ignore-fields} 'user1,sort' entries.bib
\end{alltt}
is equivalent to:
\begin{alltt}
convertgls2bib \longargfmt{ignore-fields} user1 \longargfmt{ignore-sort entries.bib}
\end{alltt}
and
\begin{alltt}
convertgls2bib \longargfmt{ignore-fields} 'user1,sort' \longargfmt{ignore-fields} '' entries.bib
\end{alltt}
is equivalent to:
\begin{alltt}
convertgls2bib \longargfmt{ignore-sort} entries.bib
\end{alltt}

\convertglsbibarg{preamble-only}

Stop parsing if the start of the \env{document} environment is found.

\convertglsbibarg{no-preamble-only}

Parse the entire file (default). Be prepared for a lot of
unknown command warnings if you make \idx{convertgls2bib} parse an
entire document.

\convertglsbibarg{absorb-see}

Absorb any cross-referencing information identified with 
\ics{glssee} or \ics{glsxtrindexseealso} commands into
the corresponding entry (default).

\convertglsbibarg{no-absorb-see}

Don't absorb any cross-referencing information identified with \ics{glssee} 
or \ics{glsxtrindexseealso} commands.

\convertglsbibarg{index-conversion}

Use \atentry{index} instead of \atentry{entry} if the
\field{description} is empty or simply \ics{nopostdesc} or
\ics{glsxtrnopostpunc}. (Only applies to terms that would otherwise
be converted to \atentry{entry}, such as those defined with
\gls{newglossaryentry}.)

\convertglsbibarg{no-index-conversion}

Don't convert \atentry{entry} to \atentry{index} (default).

\convertglsbibarg{locale}

Identifies the \langxml\ to use for \idx{convertgls2bib}['s]
messages.

\convertglsbibarg{overwrite}

Allow existing \ext{bib} files to be overwritten. (Default unless
\convertglsbiblongarg{split-on-type}.)

\convertglsbibarg{no-overwrite}

Don't allow existing \ext{bib} files to be overwritten. (Default if
\convertglsbiblongarg{split-on-type}.)

\convertglsbibarg{silent}

Suppress all messages except for errors.

\convertglsbibarg{verbose}

Display messages and warnings (default).

\convertglsbibarg{debug}

Display debugging messages (stack traces and other information in
addition to \convertglsbiblongarg{verbose}).

\convertglsbibarg{help}

Display help message and quit.

\convertglsbibarg{version}

Display version information and quit.

\section{Recognised Commands}

This application recognises the commands listed below as well as
some standard commands such as \ics{newcommand}. Avoid any overly
complicated code within the \ext{tex} file. The
\hyperref[sec:texparserlib]{\TeX\ parser library} isn't a \TeX\ engine! The
\ext{tex} file doesn't need to be a complete document, but if you want certain
commands recognised from packages that the \TeX\ parser library supports,
you'll need to include \ics{usepackage} in the \ext{tex} file. If
you want to quit parsing the \ext{tex} file at the start of the
document, use the \convertglsbiblongarg{preamble-only} switch.

In all cases below, if \meta{\keyvallist} contains:
\begin{codeenv*}
\field{see}=\marg{[\ics{seealsoname}]\meta{label(s)}}
\end{codeenv*}
or
\begin{codeenv*}
\field{see}=\marg{[\ics{alsoname}]\meta{label(s)}}
\end{codeenv*}
this will be substituted with:
\begin{codeenv*}
\field{seealso}=\margm{label(s)}
\end{codeenv*}
For example:
\begin{codeenv}
\gls{newterm}\oarg{\field{see}=\marg{[\cs{seealsoname}]goose}}\marg{duck}
\end{codeenv}
will be written as:
\begin{codeenv}
\atentry{index}\marg{duck,
  \field{seealso} = \marg{goose}
}
\end{codeenv}
Note that it won't convert \code{\field{see}=\marg{[see also]\meta{labels}}}.
If you have used explicit text instead of \ics{seealsoname} or
\ics{alsoname} then consider performing a global search and replace
on your file using your text editor.

Additionally, if \meta{\keyvallist} contains:
\begin{codeenv}
\field{type}=\marg{\ics{glsdefaulttype}}
\end{codeenv}
then this field will be ignored. (This \field{type} value is
recommended in \meta{\keyvallist} when loading files with
\ics{loadglsentries}\oargm{type}\margm{file} to allow the optional
argument to set the \field{type}. With \bibgls\ you can use the
\csopt{type} option instead.)

\cssection{glsexpandfields}
The base \styfmt{glossaries} package provides:
\formatdef{glsexpandfields}
If present, this instructs \idx{convertgls2bib} to expand all fields
except for those explicitly identified by \gls{glssetnoexpandfield}.
Remember that there are many commands that aren't recognised by
\idx{convertgls2bib} so it may not be possible to correctly expand
field values. Conversely, there are some commands that will be
expanded by \idx{convertgls2bib} that aren't expandable in \TeX\
(such as \ics{MakeUppercase} and \ics{char}).

\cssection{glsnoexpandfields}
The base \styfmt{glossaries} package provides:
\formatdef{glsnoexpandfields}
If present, this instructs \idx{convertgls2bib} to not expand fields
unless explicitly identified by \gls{glssetexpandfield}.

\cssection{glssetexpandfield}
The base \styfmt{glossaries} package provides:
\formatdef{glssetexpandfield}
If present, this instructs \idx{convertgls2bib} to expand the
given field, even if \gls{glsnoexpandfields} has been used.

\cssection{glssetnoexpandfield}
The base \styfmt{glossaries} package provides:
\formatdef{glssetnoexpandfield}
If present, this instructs \idx{convertgls2bib} to not expand the
given field, even if \gls{glsexpandfields} has been used.
Unlike the default behaviour with the \sty{glossaries} package, there are no
fields switched explicitly switched off by default with
\idx{convertgls2bib}.

\cssection{newglossaryentry}
The base \styfmt{glossaries} package provides:
\formatdef{newglossaryentry}
This is converted to:
\begin{codeenv}
\atentry{entry}\marg{\meta{label},
  \meta{\keyvallist}
}
\end{codeenv}
\ics{newentry} is recognised as a synonym of \gls{newglossaryentry}.

\cssection{provideglossaryentry}
The base \styfmt{glossaries} package provides:
\formatdef{provideglossaryentry}
This is converted to:
\begin{codeenv}
\atentry{entry}\marg{\meta{label},
  \meta{\keyvallist}
}
\end{codeenv}
but only if \meta{label} hasn't already been defined.

\cssection{longnewglossaryentry}
The base \styfmt{glossaries} package provides:
\formatdef{longnewglossaryentry}
This is converted to:
\begin{codeenv}
\atentry{entry}\marg{\meta{label},
 \meta{\keyvallist},
 \field{description} = \margm{description}
}
\end{codeenv}
The starred version provided by the \styfmt{glossaries-extra} package
is also recognised. The unstarred version strips trailing spaces
from \meta{description}. (This doesn't add \ics{nopostdesc}, but
\styfmt{glossaries-extra} defaults to \styopt{nopostdot}.)

\cssection{longprovideglossaryentry}

The base \styfmt{glossaries} package provides:
\formatdef{longprovideglossaryentry}
As above, but only if \meta{label} hasn't already been defined.

\cssection{newterm}

The base \styfmt{glossaries} package provides:
\formatdef{newterm}
(when the \styopt{index} option is used).
This is converted to:
\begin{codeenv}
\atentry{index}\marg{\meta{label},
 \meta{\keyvallist}
}
\end{codeenv}
if the optional argument is present, otherwise it's just converted
to:
\begin{codeenv}
\atentry{index}\marg{\meta{label}}
\end{codeenv}

If \longargfmt{space-sub} is used and \meta{label} contains one or
more spaces, then \field{name} will be set if not included in
\meta{\keyvallist}. For example, if \filefmt{entries.bib}
contains:
\begin{codeenv}
\gls{newterm}\marg{sea lion}
\gls{newterm}\oarg{\field{seealso}=\marg{sea lion}}\marg{seal}
\end{codeenv}
then:
\begin{verbatim}
convertgls2bib --space-sub '-' entries.bib entries.tex
\end{verbatim}
will write the terms to \filefmt{entries.tex} as:
\begin{codeenv}
\atentry{index}\marg{sea-lion,
  \field{name} = \marg{sea lion}
}
\strut
\atentry{index}\marg{seal,
  \field{seealso} = \marg{sea-lion}
}
\end{codeenv}
whereas just:
\begin{verbatim}
convertgls2bib entries.bib entries.tex
\end{verbatim}
will write the terms to \filefmt{entries.tex} as:
\begin{codeenv}
\atentry{index}\marg{sea lion}
\strut
\atentry{index}\marg{seal,
  \field{seealso} = \marg{sea lion}
}
\end{codeenv}
which will cause a problem when the \ext{bib} file is parsed
by \bibgls\ (and will probably also cause a problem for 
bibliographic management systems).

\cssection{newabbreviation}
The \styfmt{glossaries-extra} package provides:
\formatdef{newabbreviation}
This is converted to:
\begin{codeenv}
\atentry{abbreviation}\marg{\meta{label},
 \field{short} = \margm{short},
 \field{long} = \margm{long},
 \meta{\keyvallist}
}
\end{codeenv}
if the optional argument is present, otherwise it's converted to:
\begin{codeenv}
\atentry{abbreviation}\marg{\meta{label},
 \field{short} = \margm{short},
 \field{long} = \margm{long}
}
\end{codeenv}

\cssection{newacronym}
The base \styfmt{glossaries} package provides:
\formatdef{newacronym}
(which is redefined by \styfmt{glossaries-extra} to use
\gls{newabbreviation}).

As above but uses \atentry{acronym} instead. The base package also
provides \ics{oldacronym}, which emulates the way abbreviations were
defined with the precursor \styfmt{glossary} package. This has
different syntax to \gls{newacronym} but is also recognised by \idx{convertgls2bib}
and is converted to \atentry{acronym}.

\cssection{glsxtrnewsymbol}
The \styfmt{glossaries-extra} package provides:
\formatdef{glsxtrnewsymbol}
(when the \styopt{symbols} option is used).
This is converted to:
\begin{codeenv}
\atentry{symbol}\marg{\meta{label},
 \field{name} = \margm{symbol}
}
\end{codeenv}
if the optional argument is missing, otherwise it's converted to:
\begin{codeenv}
\atentry{symbol}\marg{\meta{label},
 \field{name} = \margm{symbol},
 \meta{\keyvallist}
}
\end{codeenv}
unless \meta{\keyvallist} contains the \field{name} field,
in which case it's converted to:
\begin{codeenv}
\atentry{symbol}\marg{\meta{label},
 \meta{\keyvallist}
}
\end{codeenv}
\ics{newsym} is recognised as a synonym for \gls{glsxtrnewsymbol}.

\cssection{glsxtrnewnumber}

The \styfmt{glossaries-extra} package provides:
\formatdef{glsxtrnewnumber}
(when the \styopt{numbers} option is used).
This is converted to:
\begin{codeenv}
\atentry{number}\marg{\meta{label},
 \field{name} = \margm{label}
}
\end{codeenv}
if the optional argument is missing, otherwise it's converted to:
\begin{codeenv}
\atentry{number}\marg{\meta{label},
 \field{name} = \margm{label},
 \meta{\keyvallist}
}
\end{codeenv}
if \field{name} isn't listed in \meta{\keyvallist},
otherwise it's converted to:
\begin{codeenv}
\atentry{number}\marg{\meta{label},
 \meta{\keyvallist}
}
\end{codeenv}
\ics{newnum} is recognised as a synonym for \gls{glsxtrnewnumber}.

\cssection{newdualentry}

\formatdef{newdualentry}
This command isn't provided by either \styfmt{glossaries} or
\styfmt{glossaries-extra} but is used as an example in the
\sty{glossaries} user manual~\cite{glossaries} and in the sample file
\file{sample-dual.tex} that accompanies the \sty{glossaries}
package. Since this command seems to be used quite a bit (given the
number of times it crops up on sites like
\href{https://tex.stackexchange.com/}{\TeX\ on StackExchange}), 
\idx{convertgls2bib} also supports it unless this command is
defined using \ics{newcommand} or \ics{renewcommand} in the 
input file. In which case the default definition will be overridden.

If the command definition isn't overridden, then it's converted to:
\begin{codeenv}
\atentry{dualabbreviationentry}\marg{\meta{label},
  \field{short} = \margm{short},
  \field{long} = \margm{long},
  \field{description} = \margm{description},
  \meta{\keyvallist}
}
\end{codeenv}
if \meta{\keyvallist} is supplied, otherwise it's converted to:
\begin{codeenv}
\atentry{dualabbreviationentry}\marg{\meta{label},
  \field{short} = \margm{short},
  \field{long} = \margm{long},
  \field{description} = \margm{description}
}
\end{codeenv}

For example, if the original \ext{tex} file contains:
\begin{codeenv}
\cs{newcommand}*\marg{\gls{newdualentry}}[5][]\marg{\comment{}
  \gls{newglossaryentry}\marg{main-\idx{param}2}\marg{\field{name}=\marg{\idx{param}4},\comment{}
    \field{text}=\marg{\idx{param}3\cs{glsadd}\marg{\idx{param}2}},\comment{}
    \field{description}=\marg{\idx{param}5},\comment{}
    \idx{param}1
  }\comment{}
  \gls{newacronym}\marg{\idx{param}2}\marg{\idx{param}3\cs{glsadd}\marg{main-\idx{param}2}}\marg{\idx{param}4}\comment{}
}
\strut
\gls{newdualentry}\marg{svm}\comment{label}
  \marg{SVM}\comment{abbreviation}
  \marg{support vector machine}\comment{long form}
  \marg{Statistical pattern recognition technique}\comment{description}
\end{codeenv}
then the \ext{bib} file will contain:
\begin{codeenv}
\atentry{entry}\marg{main-svm,
  \field{name} = \marg{support vector machine},
  \field{description} = \marg{Statistical pattern recognition technique},
  \field{text} = \marg{SVM\cs{glsadd}\marg{svm}}
}
\strut
\atentry{acronym}\marg{svm,
  \field{short} = \marg{SVM\cs{glsadd}\marg{main-svm}},
  \field{long} = \marg{support vector machine}
}
\end{codeenv}
since \gls{newdualentry} was defined with \ics{newcommand}. However,
if the original file uses \ics{providecommand} or omits the
definition of \gls{newdualentry}, then the \ext{bib} file will
contain:
\begin{codeenv}
\atentry{dualabbreviationentry}\marg{svm,
  \field{short} = \marg{SVM},
  \field{description} = \marg{Statistical pattern recognition technique},
  \field{long} = \marg{support vector machine}
}
\end{codeenv}

\chapter{Examples}
\label{sec:examples}
\setsecdepth*{0}

The example files described here can be found in the
\filefmt{examples} sub-directory. The \ext{bib} files are listed
first and then sample files that use the \ext{bib} data.
Make sure you have the latest versions of \styfmt{glossaries},
\styfmt{mfirstuc}, \styfmt{glossaries-extra} and \bibgls\ if you
want to try these out.  (The \exfile{sample-media.tex} file requires
at least \styfmt{datatool} v2.28.) If you get any undefined control
sequence or undefined style errors then you need to update your
\TeX\ distribution.  Use the \longarg{group} switch when invoking
\bibgls\ for all these examples if you want the glossaries divided
into groups.  The set of system calls for the document build in the
examples below may require an extra \LaTeX\ run to ensure the PDF
bookmarks are up-to-date when \sty{hyperref} is used.

These files are just examples of how to use \bibgls. There are other
ways of defining similar entries and sometimes alternatives are
suggested. Use the code here as a starting point if you need data
like this and adapt it to a format appropriate for your
requirements.

\filesection{no-interpret-preamble.bib}

The \exfile{no-interpret-preamble.bib} file contains command
definitions used in some of the \field{name} fields. Although these
commands aren't used explicitly in the document, they need to be
defined when the names are displayed in the document (typically in
the glossary). These commands are much like the \gls{sortop} command 
described on \glsxtrpageref{sortop} and need to be hidden 
from \bibgls's interpreter. This file doesn't contain any entry
definitions and must be loaded first with
\csopt[false]{interpret-preamble}. The
\exfile{interpret-preamble.bib} or \exfile{interpret-preamble2.bib} 
file can then be loaded to provide alternative definitions for
\bibgls's interpreter.

The first command is:
\nosecdef{sortname}
This is used in the \field{name} fields for entries containing
information about a person. The aim here is for \bibgls\ to sort
according to \meta{surname}, \meta{first name(s)} but for the
glossary to display \meta{first name(s)} \meta{surname}. For names
with a \qt{von} part, there's another command:
\nosecdef{sortvonname}
which has a similar purpose.
The third command is:
\nosecdef{sortart}
This is the same as \gls{sortname} but is designed for titles,
phrases or sentences that start with an article (such as \qt{a} or
\qt{the}). Although it has the same definition as \gls{sortname}
in this file, in the interpreted files the article part is omitted
to completely ignore them in the sorting.
The fourth command is:
\nosecdef{sortmediacreator}
which again is functionally the same as \gls{sortname}.

The names could be specified using \BibTeX's
syntax instead with \csopt{bibtex-contributor-fields} to convert it,
but the aim here is to show a variety of ways to use \bibgls. For an
example of \csopt{bibtex-contributor-fields}, see the way the 
\fieldfmt{cast} field in \exfile{films.bib} is dealt with.

Although the file only contains ASCII characters, it starts with 
an encoding line to prevent \bibgls\ from searching the entire file
for it. (That's not so much of an issue with a short file, but may
cause an unnecessary delay for much longer files.)

The contents of \filefmt{no-interpret-preamble.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/no-interpret-preamble.bib}

\filesection{interpret-preamble.bib}

This provides definitions of \gls{sortname}, \gls{sortvonname},
\gls{sortart} and \gls{sortmediacreator} in \atentry{preamble} 
that can be picked up by the interpreter and used during sorting.
Note that in this case \gls{sortart} is defined to ignore the
article to completely ignore it from sorting. If you happen to have 
\qt{a \meta{something}} and \qt{the \meta{something}} where the 
\meta{something}s are identical, you may want to append the article
to disambiguate them.

The contents of \filefmt{interpret-preamble.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/interpret-preamble.bib}

\filesection{interpret-preamble2.bib}

An alternative to \exfile{interpret-preamble.bib} with a different
definition of \gls{sortmediacreator}. This uses \cs{renewcommand}
instead of \cs{providecommand} so \csopt[false]{write-preamble}
is required to prevent \LaTeX\ from picking up the definitions.

The contents of \filefmt{interpret-preamble2.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/interpret-preamble2.bib}

\filesection{constants.bib}

The \exfile{constants.bib} file contains mathematical constants.
These all use a custom entry type \atentryfmt{constant}, which must
be aliased otherwise the entries will all be ignored. The entries
all have custom fields, which also need to be aliased.
For example:
\begin{codeenv}
\csopt[constant=entry]{entry-type-aliases},
\csopt[
  \fieldfmt{constantname}=\field{name},
  \fieldfmt{constantsymbol}=\field{symbol},
  \fieldfmt{definition}=\field{description},
  \fieldfmt{identifier}=\field{category},
  \fieldfmt{value}=\field{user1}
]{field-aliases}
\end{codeenv}
This setting means that, for example,
\begin{codeenv}
\atentryfmt{constant}\marg{root2,
  \fieldfmt{constantname}=\marg{Pythagoras' constant},
  \fieldfmt{constantsymbol}=\marg{\cs{ensuremath}\marg{\ics{surd}2}},
  \fieldfmt{definition}=\marg{the square root of 2},
  \fieldfmt{value}=\marg{1.41421},
  \fieldfmt{identifier}=\marg{constant}
}
\end{codeenv}
is treated as though it was defined as:
\begin{codeenv}
\atentry{entry}\marg{root2,
  \field{name}=\marg{Pythagoras' constant},
  \field{symbol}=\marg{\cs{ensuremath}\marg{\cs{surd}2}},
  \field{description}=\marg{the square root of 2},
  \field{user1}=\marg{1.41421},
  \field{category}=\marg{constant}
}
\end{codeenv}
This use of custom fields and entry types allows more flexibility.
For example, I may have another document that uses the same
\ext{bib} file but requires a different definition:
\begin{codeenv}
\atentry{number}\marg{root2,
  \field{description}=\marg{Pythagoras' constant},
  \field{name}=\marg{\cs{ensuremath}\marg{\cs{surd}2}}
}
\end{codeenv}
which can be obtained with:
\begin{codeenv}
\csopt[constant=number]{entry-type-aliases},
\csopt[
  \fieldfmt{constantname}=\field{description},
  \fieldfmt{constantsymbol}=\field{name}
]{field-aliases}
\end{codeenv}
Since the other custom fields haven't be aliased, they're ignored.

The custom fields are: \fieldfmt{identifier} (set to \code{constant}
for all the entries), \fieldfmt{constantname} (the constant's name),
\fieldfmt{definition} (a definition of the constant),
\fieldfmt{value} (the approximate numeric value of the constant),
\fieldfmt{constantsymbol} (the symbolic representation of the
constant) and \fieldfmt{alternative} (alternative symbol). 
There are three entries that don't have the custom
\fieldfmt{value} field: \code{zero} and \code{one} (the exact value
is in the \fieldfmt{constantsymbol} field in both cases) and \code{imaginary}
(where there's no real number value).

I've provided some commands in the \atentry{preamble} for
constants that are represented by Latin and Greek letters.
These can be defined in the document before the \idx{resourceset}
if different notation is required. The upright Greek commands require
the \isty{upgreek} package.

If it's likely that there may be a need to sort according to
\fieldfmt{definition}, then it would be better to use \gls!{sortart}
describe above:
\begin{codeenv}
\atentryfmt{constant}\marg{root2,
  \fieldfmt{constantname}=\marg{Pythagoras' constant},
  \fieldfmt{constantsymbol}=\marg{\cs{ensuremath}\marg{\cs{surd}2}},
  \fieldfmt{definition}=\marg{\gls{sortart}\marg{the}\marg{square root of 2}},
  \fieldfmt{value}=\marg{1.41421},
  \fieldfmt{identifier}=\marg{constant}
}
\end{codeenv}
Remember that this would need \exfile{no-interpret-preamble.bib} to
ensure the command is recognised in the document.

The contents of \filefmt{constants.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/constants.bib}

\filesection{chemicalformula.bib}

The \exfile{chemicalformula.bib} file contains chemical formulae.
Each entry has a field that uses \ics{ce} provided by \sty{mhchem}
so the document will need to load that package. Since all resource
files must be loaded in the preamble, it's possible to ensure that
the package is loaded using:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{usepackage}\marg{mhchem}}}
\end{codeenv}
However, it's best just to load it in the document otherwise it
won't be available before the \ext{glstex} file has been loaded.
Also, \sty{glossaries} (and therefore \sty{glossaries-extra}) must be
loaded after \sty{hyperref}, which usually needs to be loaded last
so most packages should be loaded before \sty{glossaries-extra}.
Instead, I've just put a comment in the \ext{bib} file as a
reminder.

All entries are defined using a custom entry type
\atentryfmt{chemical}. This must be aliased using
\csopt{entry-type-aliases} or the entries will be ignored. For
example, to make \atentryfmt{chemical} behave like \atentry{symbol}:
\begin{codeenv}
\csopt[\fieldfmt{chemical}=\field{symbol}]{entry-type-aliases}
\end{codeenv}
Remember that with the \atentry{symbol} type, if the \field{sort}
field is omitted \bibgls\ will fallback on the label by default. It
can be changed to fallback on the \field{name} field instead using
\csopt[name]{symbol-sort-fallback}.  This will require the use of
the interpreter if the name contains a command but \bibgls\
recognises the \sty{mhchem} package and has a limited ability to
interpret \cs{ce}.  If \atentryfmt{chemical} is changed to
\atentry{entry} instead then the fallback for the \field{sort} will be the
entry's \field{name}.

All entries only contain custom fields, which will all be ignored by
\bibgls\ unless defined or aliased: \fieldfmt{identifier}, which is set to
\code{chemical} for all entries, \fieldfmt{formula}, which is set to the
chemical formula, and \fieldfmt{chemicalname}, which is set to the chemical
name. This allows the flexibility of determining whether the
\field{name} or \field{symbol} field should contain the chemical
formula on a per-resource basis. For example:
\begin{codeenv}
\csopt[\fieldfmt{formula}=\field{name},\fieldfmt{chemicalname}=\field{description}]{field-aliases}
\end{codeenv}
or
\begin{codeenv}
\csopt[\fieldfmt{chemicalname}=\field{name},\fieldfmt{formula}=\field{symbol}]{field-aliases}
\end{codeenv}

The contents of \filefmt{chemicalformula.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/chemicalformula.bib}

\filesection{bacteria.bib}

The \exfile{bacteria.bib} file contains bacteria abbreviations.
These all use the \atentry{abbreviation} entry type with a
\field{short} and \field{long} field.

The entries all have a custom field \fieldfmt{identifier} set to 
\code{bacteria}. This will be ignored by \bibgls\ unless it's
defined using \cs{glsaddkey} or \cs{glsaddstoragekey} or if it's
aliased with \csopt{field-aliases}.

The contents of \filefmt{bacteria.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/bacteria.bib}

\filesection{baseunits.bib}

The \exfile{baseunits.bib} file contains base \idxpl{SIunit}. The entries
are all defined using the custom \atentryfmt{unit} entry
type. This must be aliased with \csopt{entry-type-aliases} otherwise
\bibgls\ will ignore all the entries. For example:
\begin{codeenv}
\csopt[\fieldfmt{unit}=\field{symbol}]{entry-type-aliases}
\end{codeenv}
will make \bibgls\ treat the entries as though they were defined
using \atentry{symbol}. (Remember that \atentry{symbol} entry types use the label
as the fallback field for \field{sort}.)

The entries all have custom fields \fieldfmt{unitname},
\fieldfmt{unitsymbol} and \fieldfmt{measurement}, one of which must
be aliased or copied to \field{name} if \atentryfmt{unit} is aliased
to an entry type that requires it. The other custom fields may be aliased or
copied to \field{symbol} and \field{description} as required. The
\fieldfmt{unitsymbol} fields all use \ics{si} provided by the
\isty{siunitx} package, so that package must be loaded in the
document. This is one of the small number of packages recognised by
\bibgls, so it's possible to sort according to the symbol if
required.

The entries also all have a custom field \fieldfmt{identifier} set to 
\code{baseunit}. This will be ignored by \bibgls\ unless it's
defined using \cs{glsaddkey} or \cs{glsaddstoragekey} or if it's
aliased with \csopt{field-aliases}.

The contents of \filefmt{baseunits.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/baseunits.bib}

\filesection{derivedunits.bib}

The \exfile{derivedunits.bib} file is much like \exfile{baseunits.bib}
but contains derived units and in this case the custom entry type
is \atentryfmt{measurement}, which must be aliased
otherwise the entries will all be ignored.
The entries all have a custom field \fieldfmt{identifier} set to 
\code{derivedunit}. This will be ignored by \bibgls\ unless it's
defined using \cs{glsaddkey} or \cs{glsaddstoragekey} or if it's
aliased with \csopt{field-aliases}.

The contents of \filefmt{derivedunits.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/derivedunits.bib}

\filesection{people.bib}

The \exfile{people.bib} file contains details about people. 
The \field{name} fields contain custom commands provided in
\exfile{no-interpret-preamble.bib} and
\exfile{interpret-preamble.bib}.
Remember that if \exfile{no-interpret-preamble.bib} is
loaded first, the definitions provided in that file will be the one
in use in the document. The \exfile{interpret-preamble.bib} file
then needs to be loaded to provide the definitions for \bibgls's
interpreter.

The information for each person is supplied in an \atentry{entry}
type. There are some non-standard fields: \fieldfmt{born},
\fieldfmt{died} and \fieldfmt{othername}. These fields will be ignored
unless keys are provided (using \cs{glsaddkey} or
\cs{glsaddstoragekey}) or the fields are aliased (using
\csopt{field-aliases}). The \fieldfmt{born} and \fieldfmt{died}
fields have dates that are \emph{almost} in the default \code{en-GB}
locale format with the \idx{JRE} \idx{localeprovider}, but they include a
tilde \idx{nbspchar} to prevent awkward line breaks. By default \bibgls's
interpreter converts \idx{nbspchar} to the non-breaking space character
\hex{A0} which isn't recognised by the date format.  This can easily
be fixed with the \longarg{break-space} switch which will interpret
\idx{nbspchar} as a normal breakable space (\hex{20}), so with that switch
\csopt[date]{sort} or \csopt[date-reverse]{sort} can be used on
either of those fields. However, the \idx{CLDR} has a slightly
different default format than the \idx{JRE} for dates with
\code{en-GB}, so it's probably simplest to actually specify the
required format.

An alternative approach would be to provide a command that can be
modified in the document to adjust the date style. For example, the
\fieldfmt{born} field could be specified as:
\begin{codeenv}
\fieldfmt{born}=\marg{\cmd{formatdate}\marg{13}\marg{7}\marg{100}\marg{BC}}
\end{codeenv}
The definition provided for the document could then be, for example:
\begin{codeenv}
\cs{providecommand}\marg{\cmd{formatdate}}[4]\marg{\ics{DTMdisplaydate}\marg{\idx{param}3}\marg{\idx{param}2}\marg{\idx{param}1}\marg{-1} \idx{param}4}
\end{codeenv}
(where \ics{DTMdisplaydate} is provided by the \isty{datetime2}
package) and a definition could be provided for \bibgls's
interpreter, for example:
\begin{codeenv}
\cs{providecommand}\marg{\cmd{formatdate}}[4]\marg{\idx{param}1/\idx{param}2/\idx{param}3 \idx{param}4}
\end{codeenv}
This would need the date format set.  For example,
\csopt[d/M/y G]{date-sort-format}.

Some of the entries, such as \code{caesar}, have a \field{first}
field. In those cases the \field{first} field is slightly different
from the \field{name} field (for example, \qt{Gaius} is omitted in
\code{caesar}'s \field{first} field). The other entries don't have a
\field{first} field. They can simply have the \field{name} copied to
\field{first} with the \csopt{replicate-fields} option (so that the
full name is shown on first use) or the \field{first} field can be
ignored with \csopt{ignore-fields} (so all entries will use the
\field{text} field on first use). The \csopt{replicate-override}
option can be used to force the \field{name} field to be copied to
the \field{first} field, even if the \field{first} field is already
set. Alternatively, with \csopt[true]{replicate-override} and
\csopt[\field{first}=\field{name}]{replicate-fields}, the \field{first} field be
copied to the \field{name} field. For consistency, the \field{first}
fields use the same custom commands as used in the \field{name}
field.

There's one name with a \qt{von} part. In this case the \field{name}
field is set to:
\begin{codeenv}
\gls{sortvonname}\marg{Manfred}\marg{von}\marg{Richthofen}
\end{codeenv}
which will come under the \qt{V} letter group since \gls{sortvonname}
is defined as \meta{von} \meta{surname}, \meta{first name(s)}

If you prefer that this name should come under \qt{R} instead, then
you need to adjust the definition of \gls{sortvonname}:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cmd{sortname}}[2]\marg{\idx{param}2, \idx{param}1}
\cs{providecommand}\marg{\cmd{sortvonname}}[3]\marg{\idx{param}3, \idx{param}1 \idx{param}2}}}
\end{codeenv}
An alternative approach would be to format the names using
\BibTeX's contributor syntax and use
\csopt[name]{bibtex-contributor-fields} to convert them.

There are also some synonyms provided with \atentry{index} entry
types that have the \field{alias} field to redirect to the main
entry. These don't include a \field{description} or any of the other
fields as that would be redundant. All the information can be found
in the main entry.

Except for the aliases, the entries have a custom field
\fieldfmt{identifier} set to \code{person}. This will be ignored by
\bibgls\ unless it's defined using \cs{glsaddkey} or
\cs{glsaddstoragekey} or if it's aliased with \csopt{field-aliases}.

The contents of \filefmt{people.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/people.bib}

\filesection{books.bib}

The \exfile{books.bib} file contains details about books.
As above, the entries use custom commands provided in
\exfile{no-interpret-preamble.bib} and
\exfile{interpret-preamble.bib} or
\exfile{interpret-preamble2.bib}.
The entries all have a custom field \fieldfmt{identifier} set to 
\code{book} and other custom fields \fieldfmt{author} and
\fieldfmt{year}.
These will be ignored by \bibgls\ unless they're
defined using \cs{glsaddkey} or \cs{glsaddstoragekey} or if they're
aliased with \csopt{field-aliases}.

There are other ways in which this data could be specified. For
example, the \field{description} field could contain a brief summary
(or \qt{log line}). The \fieldfmt{author} field could use \BibTeX's
syntax instead with \csopt{bibtex-contributor-fields} to convert it.
Alternatively, the entries could be defined using standard \BibTeX\
entry types that are all aliased to \atentry{bibtexentry}.

The contents of \filefmt{books.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/books.bib}

\filesection{films.bib}

The \exfile{films.bib} file contains details about films.
As above, the entries use custom commands provided in
\exfile{no-interpret-preamble.bib} and
\exfile{interpret-preamble.bib}.
The entries all have a custom field \fieldfmt{identifier} set to 
\code{film} and other custom fields \fieldfmt{cast}, \fieldfmt{director} and
\fieldfmt{year}. These will be ignored by \bibgls\ unless they're
defined using \cs{glsaddkey} or \cs{glsaddstoragekey} or if they're
aliased with \csopt{field-aliases}.

This example file references entries defined in \exfile{books.bib}
through the use of the special 
\glsdisp{idx.idprefix.extn}{\idprefixfmt{ext1}} prefix.
To avoid a label conflict \exfile{films.bib} prefixes all labels
with \idprefixfmt{film} rather than relying on \csopt{label-prefix}.
This ensures that both \exfile{books.bib} and \exfile{films.bib}
can be loaded in the same resource set (otherwise they'd have to be
loaded in separate resource sets with different prefixes). Remember
that you can use \gls{glsxtrnewgls}. For example:
\begin{codeenv}
\gls{glsxtrnewgls}\marg{film.}\marg{\cmd{film}}
\end{codeenv}
This means you can do, for example, just \verb|\film{bladerunner}|
if you want to reference a film without worrying about the prefix.

As with all the example files, there are other ways in which
to specify the data, depending on your requirements. For example,
the \fieldfmt{director} field could use \BibTeX's contributor syntax
(as the \fieldfmt{cast} field does). Some of
the films actually had more than one director but only one is listed
per film in this sample file for simplicity. Similarly, the
\fieldfmt{cast} field only contains the principle actors rather than the
complete list. The book on which the film is based
could be contained in a cross-reference field or a custom
\fieldfmt{basedon} field.

The book \qt{Do Androids Dream of Electric Sheep?}\ referenced at the end 
of the \qt{Blade Runner} film's \field{description} ends with 
a question mark. (Similarly for \qt{Why Didn't They Ask Evans?}) 
If the \field{description} field is simply set as:
\begin{codeenv}
\field{description}=\marg{a film starring Harrison Ford, Rutger Hauer
 and Sean Young loosely based on the novel
 \cs{gls}\marg{ext1.doandroidsdreamofelectricsheep}},
\end{codeenv}
then the \styopt{postdot} package option will produce an odd
result as the inserted \idx{full-stop} immediately follows the question
mark. This is an awkward situation. One possibility is to explicitly 
put the \idx{full-stop} at the end of the \field{description} field for all the 
other entries and omit it for the problematic entries, but this
interferes with the possibility of a category-dependent
\idx{postdescriptionhook}. 

Another option is to put \ics{nopostdesc} in the 
problematic entries. For example:
\begin{codeenv}
\field{description}=\marg{a film starring Harrison Ford, Rutger Hauer
 and Sean Young loosely based on the novel
 \cs{gls}\marg{ext1.doandroidsdreamofelectricsheep}\cs{nopostdesc}},
\end{codeenv}
Be careful with this as it will completely suppress the
\idx{postdescriptionhook}.
A third possibility is to use \ics{glsxtrnopostpunc} instead:
\begin{codeenv}
\field{description}=\marg{a film starring Harrison Ford, Rutger Hauer
 and Sean Young loosely based on the novel
 \cs{gls}\marg{ext1.doandroidsdreamofelectricsheep}\cs{glsxtrnopostpunc}},
\end{codeenv}
This doesn't interfere with the \idx{postdescriptionhook} but if a hook
is provided the post-punctuation may then be required. In both of
the above two cases, \csopt{strip-trailing-nopost} could be used
to remove the suppression commands from the \field{description}
fields if a hook is defined. However this doesn't deal with hooks
that only conditionally append text.

The best solution is with \styfmt{glossaries-extra} v1.23+ which
provides \ics{glsxtrrestorepostpunc} for use in the category
\idxpl{postdescriptionhook} that counteracts \cs{glsxtrnopostpunc}.
This can be placed inside a \idx{conditional}, as used in
\exfile{sample-media.tex}, and does nothing if \cs{glsxtrnopostpunc}
doesn't occur in the \field{description} field. (Note that 
\cs{glsxtrrestorepostpunc} can't be used to counteract
\cs{nopostdesc}, since that completely suppresses the hook.)

The contents of \filefmt{films.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/films.bib}

\filesection{citations.bib}

The \exfile{citations.bib} file is actually a \BibTeX\ file, but it
can be parsed by \bibgls\ if the \BibTeX\ entry types are converted
to \atentry{bibtexentry}, which can easily be done with:
\begin{codeenv}
\csopt[\ics{GlsXtrBibTeXEntryAliases}]{entry-type-aliases}
\end{codeenv}
The field names will also need to be defined or aliased. For
example:
\begin{codeenv}
\csopt[\fieldfmt{title}=\field{name}]{field-aliases}
\end{codeenv}
If \bibgls\ is then run with \longarg{cite-as-record} any
\ics{citation} commands found in the \ext{aux} file will be treated
as ignored records. The \atentry{preamble} provides a formatting
command that's used by both \BibTeX\ and \bibgls, so \cs{providecommand}
is required rather than \cs{newcommand} as it will appear in both
the \iext{bbl} and the \iext{glstex} files. (In general it's best to
use \cs{providecommand} rather than \cs{newcommand} in the
\atentry{preamble} but in this case it's essential.)
The contents of \filefmt{citations.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/citations.bib}

\filesection{mathgreek.bib}

The \exfile{mathgreek.bib} file contains Greek letters for use in
maths mode. These are all defined with \atentry{symbol}, which means
that by default the \field{sort} field will be obtained from the
label not from the \field{name} field. However, if you want to sort
by the \field{name} field (for example, with \csopt[name]{sort-field})
the \TeX\ parser library recognises all the mathematical Greek letter commands 
provided in the \LaTeX\ kernel. Additionally it recognises
\ics{omicron} which isn't provided by \LaTeX\ (the symbol can be
reproduced with a \idx!{lowercase} Latin \qt{o}). Note that
\isty{glossaries-extra-bib2gls} (\sty{glossaries-extra} v1.27+)
provides all the missing Greek letters (such as \cs{omicron}).

The \ext{bib} file could just use \code{o}:
\begin{codeenv}
\atentry{symbol}\marg{omicron,
  \field{name}=\marg{\cs{ensuremath}\marg{o}},
  \field{description}=\marg{omicron},
  \fieldfmt{identifier}=\marg{mathgreek}
}
\end{codeenv}
but this means that if \bibgls\ sorts according to the \field{name}
field using a letter sort, this entry will come before all the other
Greek letters since the character \qt{o} has Unicode value \hex{6F}
whereas, for example, mathematical italic small alpha ($\alpha$)
has Unicode value \hex{1D6FC}. This means that for sorting purposes
it's better to use \ics{omicron}:
\begin{codeenv}
\atentry{symbol}\marg{omicron,
  \field{name}=\marg{\cs{ensuremath}\marg{\cs{omicron}}},
  \field{description}=\marg{omicron},
  \fieldfmt{identifier}=\marg{mathgreek}
}
\end{codeenv}
but \LaTeX\ needs a definition for this, so it's provided in the
\atentry{preamble}:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cs{providecommand}\marg{\cs{omicron}}\marg{o}}}
\end{codeenv}
(With \sty{glossaries-extra} v1.27+, this is no longer needed.)
The \TeX\ parser library and \sty{glossaries-extra-bib2gls} 
similarly provide the missing \idx!{uppercase}
Greek letters, and these can be dealt with in the same way.

The contents of \filefmt{mathgreek.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/mathgreek.bib}

\filesection{bigmathsymbols.bib}

The \exfile{bigmathsymbols.bib} file contains mathematical symbols
that have a large version in display mode. As with
\exfile{mathgreek.bib} the entries are defined using
\atentry{symbol}. This example file requires the \isty{stix} package
as not all of the commands are provided by the \LaTeX\ kernel. 
This file also has a preamble:
\begin{codeenv}
\atentry{preamble}\marg{\qtdelim{\cmd{providecommand}\marg{\cmd{bigoperatornamefmt}}[1]\marg{\comment{}
 \idx{mshiftchar}\ics{displaystyle}\idx{param}1\ics{textstyle}\idx{param}1\idx{mshiftchar}}
\cmd{providecommand}\marg{\cmd{nary}}[1]\marg{\idx{mshiftchar}\idx{param}1\idx{mshiftchar}-ary}}}
\end{codeenv}
The first command \inlinedef{bigoperatornamefmt}
is used in the \field{name} field to display both the in-line and display
versions of the symbol. The \TeX\ parser library only has a limited ability to
interpret this as not all the symbols have Unicode in-line and large versions.
In some cases, such as the integral symbol $\int$, there is only a small 
version. (A large version would require construction from \hex{2320}, 
\hex{23AE} and \hex{2321}, which is too complicated in this context.) 
However, the interpreter works well enough to guess at the widest
name if \csopt{set-widest} is used.  There's no advantage in sorting
according to the \field{name} field here, unless a custom rule is
provided, as the Unicode symbols are scattered about different
blocks. Better approaches are to sort according to document use
(\csopt[use]{sort}) or to sort according to the \field{description}
field.

The other custom command is \inlinedef{nary} to provide semantic markup for
\qt{$n$-ary}. This could be defined without an argument:
\begin{codeenv}
\cmd{providecommand}\marg{\cmd{nary}}\marg{\idx{mshiftchar}n\idx{mshiftchar}-ary}
\end{codeenv}
but providing an argument will allow \code{\csfmt{nary}\marg{n}} to work
with first letter upper-casing in the event that the
\field{description} field has a case-change applied (otherwise
it would end up as \qt{$N$-ARY}). Of course, it may be that no case-change 
should be applied, but this example is just for illustrative purposes.

As with the other sample \ext{bib} files, each entry is given a
custom \fieldfmt{identifier} field, which by default will be
ignored. In this case, \fieldfmt{identifier} is either set to 
\code{naryoperator} (for $n$-ary operators) or \code{integral}
for integrals.

The contents of \filefmt{bigmathsymbols.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/bigmathsymbols.bib}

\filesection{mathsrelations.bib}

The \exfile{mathsrelations.bib} file contains mathematical
relational symbols. These use the maths shift character
\idx{mshiftchar} in the \field{name} field and just the symbol in
the \field{text} field. This just illustrates an alternative way of
defining symbols.  Since \ics{ensuremath} isn't used, commands like
\cs{gls} must be explicitly placed in maths mode. For example,
\code{\idx{mshiftchar}\cs{gls}\marg{leq}\idx{mshiftchar}} rather than simply 
\code{\cs{gls}\marg{leq}}. The custom \fieldfmt{identifier} field is
set to \code{relation}.

The contents of \filefmt{mathsrelations.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/mathsrelations.bib}

\filesection{binaryoperators.bib}

The \exfile{binaryoperators.bib} file contains mathematical binary
operators. The format is much like the above
\exfile{mathsrelations.bib} file. The custom \fieldfmt{identifier}
field is set to \code{binaryoperator}.

The contents of \filefmt{binaryoperators.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/binaryoperators.bib}

\filesection{unaryoperators.bib}

The \exfile{unaryoperators.bib} file contains mathematical unary
operators. As above, this again uses \atentry{symbol} to define the symbols,
but in this case \ics{ensuremath} is used in the \field{name}
field and there's no \field{text} field. I've also used \ics{mathord}
to ensure the symbol is treated as a unary (rather than binary)
operator, except for the \ics{forall} entry which is already defined as an
ordinary maths symbol.

The contents of \filefmt{unaryoperators.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/unaryoperators.bib}

\filesection{mathsobjects.bib}

The \exfile{mathsobjects.bib} file contains entries related to 
mathematical objects (sets, spaces, vectors and matrices). This
provides some custom formatting commands in the preamble:
\nosecdef{setfmt}
which is used to format \meta{symbol} as a set,
\nosecdef{setcontentsfmt}
which is used to format the set contents,
\nosecdef{setmembershipfmt}
which is used to format the set membership criteria,
\nosecdef{setcardfmt}
which is used to format the cardinality of a set,
(Note this uses \ics{vert} not \textbar\ as in some of the earlier
examples.)
\nosecdef{numspacefmt}
which is used to format \meta{symbol} as a number space,
\nosecdef{transposefmt}
which is used to format matrix and vector transposes,
\nosecdef{invfmt}
which is used to format inverses,
\nosecdef{vecfmt}
which is used to format \meta{symbol} as a vector, and
\nosecdef{mtxfmt}
which is used to format \meta{symbol} as a matrix.
These commands are intended for use with \gls{glsxtrfmt},
but \gls!{setmembershipfmt} causes a problem as it has two
arguments and \gls{glsxtrfmt} requires the control sequence to 
have exactly one argument. This means employing a little trick.
A command with just one argument is provided:
\nosecdef{setmembershiponeargfmt}
that requires the actual two arguments to be supplied inside
\code{\idx!{param}1}. The outer grouping is removed and the two-argument
\gls!{setmembershipfmt} command is applied:
\begin{codeenv}
\cs{providecommand}\marg{\gls{setmembershiponeargfmt}}[1]\marg{\gls{setmembershipfmt}\idx{param}1}
\end{codeenv}
This means that the entry needs to be referenced in the document
using:
\begin{codeenv}
\gls{glsxtrfmt}\marg{setmembership}\marg{\margm{variable(s)}\margm{condition}}
\end{codeenv}
The simplest thing to do here is to provide a wrapper command in the
document, for example:
\begin{codeenv}
\cs{newcommand}*\marg{\cmd{setmembership}}[2]\marg{\gls{glsxtrfmt}\marg{setmembership}\marg{\marg{\idx{param}1}\marg{\idx{param}2}}}
\end{codeenv}
Now this can be used as:
\begin{codeenv}
\cmd{setmembership}\margm{variable(s)}\margm{condition}
\end{codeenv}

There are essentially two types of entry defined in this file: 
entries that demonstrate the formatting for the objects
and entries that represent specific objects. In the first case
there's a custom \fieldfmt{format} field
that's set to the control sequence name of the relevant semantic
command. If this field is defined or aliased then it can be used
with \gls!{glsxtrfmt} (as in the example above).

In both cases there's a custom \fieldfmt{identifier} field that
reflects the type of object: \code{set} for sets, 
\code{numberspace} for number spaces,
\code{matrix} for matrices or vectors.

Be careful with the set cardinality example. Remember that nested
links cause problems and the \sty{glossaries-extra} manual advises
against using commands like \cs{gls} or \gls!{glsxtrfmt} within
link text and that includes within the \meta{text} argument of 
\gls!{glsxtrfmt}. See \exfile{sample-maths.tex} for suggested
usage.

Some of the \field{description} fields use \gls!{sortart}, so
\exfile{no-interpret-preamble.bib} and
\exfile{interpret-preamble.bib} are also needed.

The contents of \filefmt{mathsobjects.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/mathsobjects.bib}

\filesection{miscsymbols.bib}

The \exfile{miscsymbols.bib} file contains text symbols provided
by the \isty{marvosym} and \isty{ifsym} packages. The \sty{ifsym}
package needs to be loaded with the \styoptfmt{weather} option
to provide the weather commands. Unfortunately both packages define
\csfmt{Sun} and \csfmt{Lightning}, which causes a conflict. See
\exfile{sample-textsymbols.tex} for a workaround. Alternatively,
you can load \sty{ifsym} without the \styoptfmt{weather} option
and use the internal definition of \sty{ifsym}'s \csfmt{Sun}
and \csfmt{Lightning} commands:
\begin{codeenv}
\atentryfmt{icon}\marg{sun,
  \fieldfmt{icon}=\marg{\ics{textweathersymbol}\marg{16}},
  \fieldfmt{description}=\marg{sunny},
  \fieldfmt{identifier}=\marg{weather}
}
\strut
\atentryfmt{icon}\marg{lightning,
  \fieldfmt{icon}=\marg{\cs{textweathersymbol}\marg{26}},
  \field{description}=\marg{thunderstorm},
  \fieldfmt{identifier}=\marg{weather}
}
\end{codeenv}
This removes the conflict, and \csfmt{Sun} and \csfmt{Lightning}
are as defined by \sty{marvosym}.

This file uses a custom entry type \atentryfmt{icon}, which must be
aliased to a recognised entry identifier otherwise the entries will
all be ignored. For example:
\begin{codeenv}
\csopt[\fieldfmt{icon}=\field{symbol}]{entry-type-aliases}
\end{codeenv}

There are three types of symbols defined: media controls, information
and weather. They have the custom \fieldfmt{identifier}
field set to \code{mediacontrol}, \code{information} and
\code{weather}, respectively. There are two other custom fields: 
\fieldfmt{icon} and \fieldfmt{icondescription}. These will need to
be aliased to \field{name} and \field{description}.

Neither of these packages are recognised by \bibgls, which means
that \csopt{set-widest} won't be able to determine the widest name
nor is this data suitable for sorting according to the
\fieldfmt{icon} field (or its alias). Instead, either sort by label
(which is the default for \atentry{symbol}) or by the
\field{description}. If you want to use one of the
\glostyle{alttree} styles you can still use \csopt{set-widest}, but
it will have to use the fallback command.  Alternatively, you can
omit \csopt{set-widest} and explicitly use
\cs{glsFindWidestTopLevelName}.

The contents of \filefmt{miscsymbols.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/miscsymbols.bib}

\filesection{markuplanguages.bib}

The \exfile{markuplanguages.bib} file includes a mixture of 
\atentry{entry} and \atentry{abbreviation} definitions. A~custom
command is provided in \atentry{preamble} to tag the letters
in the \field{long} field that are used to form the abbreviation.
This simply does its argument and is provided in case it's not set
up in the document. If you do want to enable tagging using
\ics{GlsXtrEnableInitialTagging}, remember that this command must be
used before the abbreviations are defined, which means before
the resource file is input with \gls{GlsXtrLoadResources}.
Similarly, the abbreviation style must be set before the
abbreviations are defined.

For convenience \atentry{string} is also used to define a \ext{bib}
variable, which may be appended to fields using the \ext{bib}
concatenation character \idx{stringconcat}. As with the other sample
\ext{bib} files, there's a custom field \fieldfmt{identifier}
which will be ignored unless defined or aliased.

The empty braces at the start some of the fields are there to
protect against first letter uppercasing within \TeX, where it might cause a
problem. (For example, with the \catattr{glossname} attribute.)

The contents of \filefmt{markuplanguages.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/markuplanguages.bib}

\filesection{usergroups.bib}

The \exfile{usergroups.bib} file requires either \XeLaTeX\ or
\LuaLaTeX\ as some of the entry labels use \idx{non-ASCII} characters.
This file has a mixture of \atentry{abbreviation} and
\atentry{index} entries. It also uses \atentry{string} for
convenience and provides a custom command \csfmt{dash} in 
\atentry{preamble}. Each entry is the name of a \TeX\ user group:
the international \idx{TUG} and all the local groups.
Most of them have an abbreviated name, so they're defined with
\atentry{abbreviation}. There are a few without an abbreviation, so
they're defined with \atentry{index} instead. There's one alias.
(The information was obtained from \idx{TUG}['s] 
\href{http://tug.org/usergroups.html}{user groups
page}~\cite{tugusergroups}.)

As with the other examples, there are some custom fields which will
be ignored if they aren't defined or aliased: \fieldfmt{identifier}
(set to \code{texusergroup}), \fieldfmt{language} (a comma-separated
list of language tags) and \fieldfmt{translation} (provides a
translation if the user group name isn't in English). 

Not all entries have a \fieldfmt{translation} field. It it's
omitted, then the user group name is in English, otherwise it's
in the first language listed in the \fieldfmt{language} field.
Most of the language tags are just the ISO 639-1 language code, but
a few of them include the ISO 3166-1 region code as well.

The contents of \filefmt{usergroups.bib} are as follows:
\begin{lstlisting}[escapechar=|]
% Encoding: UTF-8

% Requires XeLaTeX/LuaLaTeX for non-ASCII labels

@string{tug={\TeX\ Users Group}}

@preamble{"\providecommand{\dash}{\,---\,}"}

@abbreviation{TUG,
  short={TUG},
  long=tug,
  language={en},
  identifier={texusergroup}
}

@abbreviation{bgTeX,
  short={bgTeX},
  long={Bulgarian \LaTeX\ Users Group},
  language={bg},
  identifier={texusergroup}
}

@abbreviation{latex-br,
  short={latex-br},
  long={Grupo de Usuários},
  language={pt-BR},
  identifier={texusergroup},
  translation={Brazilian }#tug
}

@abbreviation{CTeX,
  short={CTeX},
  long={Chinese \TeX\ Society},
  identifier={texusergroup},
  language={zh}
}

@abbreviation{CSTUG,
  short={CSTUG},
  long={|Československé sdružení uživatelů TeXu|, z.~s.},
  language={cs},
  identifier={texusergroup},
  translation={Czech Republic }#tug
}

@abbreviation{DANTE,
  short={DANTE e.V.},
  long={Deutschsprachige Anwendervereinigung \TeX\ e.V.},
  language={de},
  identifier={texusergroup},
  translation={German Speaking }#tug
}

@abbreviation{DKTUG,
  short={DK-TUG},
  long={Danish }#tug,
  language={da},
  identifier={texusergroup}
}

@index{EUG,
  name={Estonian User Group},
  language={et},
  identifier={texusergroup}
}

@abbreviation{CervanTeX,
  short={CervanTeX},
  long={Grupo de Usuarios de \TeX\ Hispanohablantes},
  language={es},
  identifier={texusergroup},
  translation={Spanish Speaking }#tug
}

@abbreviation{TirantloTeX,
  short={Tirant lo \TeX},
  long={Catalan }#tug,
  language={ca},
  identifier={texusergroup}
}

@abbreviation{GUTenberg,
  short={GUTenberg},
  long={Groupe francophone des utilisateurs de \TeX},
  language={fr},
  identifier={texusergroup},
  translation={French Speaking }#tug
}

@abbreviation{UKTUG,
  short={UK-TUG},
  long={UK }#tug,
  language={en-GB},
  identifier={texusergroup}
}

@abbreviation{|\greekmono ɛϕτ|,
  short={|\greekmono ɛϕτ|},
  long={|\greekmono Σύλλογος Ελλήνων Φίλων του| \TeX},
  language={el},
  identifier={texusergroup},
  translation={Greek \TeX\ Friends}
}

@abbreviation{MaTeX,
  short={MaTeX},
  long={Magyar \TeX\ Egyesület},
  language={hu},
  identifier={texusergroup},
  translation={Hungarian }#tug
}

@abbreviation{ITALIC,
  short={ITALIC},
  long={Irish \TeX\ and \LaTeX\ In-print Community},
  language={en-IE,en-GB},
  identifier={texusergroup}
}

@abbreviation{ÍsTeX,
  short={ÍsTeX},
  long={Vefur íslenskra \TeX\ notenda},
  language={is},
  identifier={texusergroup},
  translation={Icelandic }#tug
}

@abbreviation{GuIT,
  short={GuIT},
  long={Gruppo Utilizzatori Italiani di \TeX},
  language={it},
  identifier={texusergroup},
  translation={Italian }#tug
}

@abbreviation{KTS,
  short={KTS},
  identifier={texusergroup},
  long={Korean \TeX\ Society},
  language={ko}
}

@index{KTUG,
  alias={KTS},
  identifier={texusergroup}
}

@index{LTVG,
  name={Lietuvos \TeX'o |Vartotojų Grupė|},
  language={lt},
  identifier={texusergroup},
  translation={Lithuanian }#tug
}

@index{mxTeX,
  name={\TeX\ México},
  language={es-MX},
  identifier={texusergroup},
  translation={Mexican }#tug
}

@abbreviation{NTG,
  short={NTG},
  long={Nederlandstalige \TeX\ Gebruikersgroep},
  language={nl},
  identifier={texusergroup},
  translation={Netherlands }#tug
}

@index{NTUG,
  name={Nordic \TeX\ Users Group},
  language={da,et,fi,fo,is,nb,nn,sv},
  identifier={texusergroup}
}

@abbreviation{GUST,
  short={GUST},
  long={Polska Grupa |Użytkowników| Systemu \TeX},
  language={pl},
  identifier={texusergroup},
  translation={Polish }#tug
}

@abbreviation{GUTpt,
  short={GUTpt},
  long={Grupo de Utilizadores de \TeX},
  language={pt},
  identifier={texusergroup},
  translation={Portuguese }#tug
}

@abbreviation{VietTUG,
  short={VietTUG},
  long={Vietnamese }#tug,
  language={vi},
  identifier={texusergroup}
}

@abbreviation{LUGSA,
  short={LUGSA},
  long={\LaTeX\ User Group\dash South Africa},
  language={en-ZA},
  identifier={texusergroup}
}


\end{lstlisting}

\filesection{animals.bib}

The \exfile{animals.bib} file contains entries defined using
\atentry{entry}. As with the above example \ext{bib} files, there's
a custom \fieldfmt{identifier} field that will be ignored unless
defined or aliased.

The contents of \filefmt{animals.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/animals.bib}

\filesection{minerals.bib}

The \exfile{minerals.bib} file contains entries defined using
\atentry{entry}. As with the above example \ext{bib} files, there's
a custom \fieldfmt{identifier} field that will be ignored unless
defined or aliased.

The contents of \filefmt{minerals.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/minerals.bib}

\filesection{vegetables.bib}

The \exfile{vegetables.bib} file contains entries defined using
\atentry{entry} and an entry defined with \atentry{index} with just
the \field{alias} field. As with the above example \ext{bib} files, there's
a custom \fieldfmt{identifier} field that will be ignored unless
defined or aliased.

The contents of \filefmt{vegetables.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/vegetables.bib}

\filesection{terms.bib}

The \exfile{terms.bib} file contains entries defined using
\atentry{index}. Unlike the above sample \ext{bib} files, there
are no custom fields here.

The contents of \filefmt{terms.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/terms.bib}

\filesection{topics.bib}

The \exfile{topics.bib} file contains entries defined using
\atentry{index}. Again there are no custom fields here.

The contents of \filefmt{topics.bib} are as follows:
\lstinputlisting[firstline=5]{../examples/topics.bib}

\filesection{sample-hierarchical.tex}

This example uses the \exfile{terms.bib}, \exfile{animals.bib},
\exfile{minerals.bib} and \exfile{vegetables.bib} files to create a
hierarchical glossary. These are specified with the resource option:
\begin{codeenv}
\csopt[{terms,animals,minerals,vegetables}]{src}
\end{codeenv}
The custom \fieldfmt{identifier} field is
aliased to the \field{parent} field since it conveniently matches
the labels of the animal, mineral and vegetable entries in the
\exfile{terms.bib} file:
\begin{codeenv}
\csopt[{identifier=parent}]{field-aliases}
\end{codeenv}
The default \csopt{selection} setting means that only those terms
referenced in the document and their dependencies are selected. The
referenced entries simply have \qt{1} in the \idx{locationlist} as
it's only a trivial single-paged example.

The dependencies that haven't actually been referenced in the
document don't have a \idx{locationlist}. (The \qt{seal} entry is a
dependency, but it's also been referenced in the document, so it has a
\idx{locationlist}.) The \qt{quartz}, \qt{beryl} and \qt{marrow}
entries are dependencies because they occur in the description of
some of the referenced entries. Normally this would mean that they
have no \idx{locationlist} after the first \LaTeX+\bibgls+\LaTeX\
build but once the glossary has been created the references to those
dependent entries in the descriptions will create records and so on
the next \bibgls+\LaTeX\ they will also have \idxpl{locationlist}.
This would make the complete document build:
\begin{verbatim}
pdflatex sample-hierarchical
bib2gls --group sample-hierarchical
pdflatex sample-hierarchical
bib2gls --group sample-hierarchical
pdflatex sample-hierarchical
\end{verbatim}
However, in this example I've decided to ignore any records created
in the glossary:
\begin{codeenv}
\ics{GlsXtrSetDefaultNumberFormat}\marg{\encap{glsignore}}
\end{codeenv}
This means that the document build is the usual \LaTeX+\bibgls+\LaTeX.

I've used the \glostyle{treegroup} style so I need to invoke
\bibgls\ with the \longarg{group} switch. This creates letter groups
for the top-level entries. Note that sub-entries never have letter groups.

The complete code is listed below.  The document build is:
\begin{verbatim}
pdflatex sample-hierarchical
bib2gls --group sample-hierarchical
pdflatex sample-hierarchical
\end{verbatim}
The complete document is shown in \figureref{fig:sample-hierarchical.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-hierarchical.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-hierarchical.pdf}}%
}
{\caption{\filefmt{sample-hierarchical.pdf}}}
{fig:sample-hierarchical.pdf}
\end{figure}

\filesection{sample-nested.tex}

As discussed in \sectionref{sec:logicaldivisions} there are three
ways of creating logical divisions when displaying the entries
through the use of the \field{type}, \field{group} and \field{parent} fields.
In general, hierarchical glossaries are created with
the \field{parent} field and an appropriate glossary style (as in
the previous \exfile{sample-hierarchical.tex} example).

This example creates a hierarchical effect but the entries don't
actually have a hierarchical structure as none of them have the
\field{parent} field set. Instead, what were the child entries in
\exfile{sample-hierarchical.tex} now have the \field{type} field
set. The hierarchical effect is achieved with
\ics{printunsrtinnerglossary} (which requires at least
\sty{glossaries-extra} v1.44).

\begin{important}
The \ics{printunsrtinnerglossary} command is unsuitable for use
with tabular-like styles, such as \glostyle{long}, and can be
problematic with \glostyle{list} styles. However, those styles
aren't suitable for hierarchical glossaries anyway.
\end{important}

Normally, hierarchy is achieve through definitions like:
\begin{codeenv}
\atentry{index}\marg{animal}
\atentry{entry}\marg{duck,\field{name}=\marg{duck},
  \field{description}=\marg{a waterbird with webbed feet},
  \field{parent}=\marg{animal}
}
\end{codeenv}
The previous example did this by loading both the \exfile{terms.bib}
and \exfile{animals.bib} files and aliasing the custom
\fieldfmt{identifier} field to \field{parent}. In this example, the
custom field is aliased to \field{type}, which effectively makes the
definitions behave like:
\begin{codeenv}
\atentry{index}\marg{animal}
\atentry{entry}\marg{duck,\field{name}=\marg{duck},
  \field{description}=\marg{a waterbird with webbed feet},
  \field{type}=\marg{animal}
}
\end{codeenv}
The aim here is for the \code{animal} entry to be placed in the main
glossary so that it's listed with \cs{printunsrtglossary}. The
\code{duck} entry is placed in a glossary that has a label
(\code{animal}) that matches the label of the \qt{parent} entry
(even though it's technically not a parent). This new glossary
(\code{animal}) can be automatically defined by invoking \bibgls\
with the \longarg{provide-glossaries} switch.

This example document defines a custom handler function that
will do the current entry as normal (with \cs{glsxtrunsrtdo}) but
will then check for the existence of a glossary that has the same
label as the current entry. This requires the starred version of
\ics{ifglossaryexists} to included ignored glossaries in the
existence check. If the glossary exists, it's then displayed using
\ics{printunsrtinnerglossary}:
\begin{codeenv}
\cs{newcommand}\marg{\csfmt{nestedhandler}}[1]\marg{\comment{}
  \cs{glsxtrunsrtdo}\marg{\idx{hashchar}1}\comment{}
  \ics{ifglossaryexists*}\marg{\idx{hashchar}1}\comment{}
  \marg{\comment{}
    \cs{printunsrtinnerglossary}\oarg{\printglossopt[\idx{hashchar}1]{type},\printglossopt[++1]{leveloffset},\printglossopt[false]{groups}}
    \marg{}\marg{}\comment{}
  }\comment{}
  \marg{}\comment{}
}
\end{codeenv}
The \printglossopt{leveloffset} option is required to achieve a
hierarchical effect (provided the glossary style supports it) and
the \printglossopt[false]{groups} option is needed to prevent letter
groups showing for the nested glossary, which would otherwise create
a strange effect. (This example uses the \glostyle{treegroup} style,
which provides a hierarchical glossary with letter groups.)

The \cs{printunsrtglossary} handler macro then needs to be set to
this custom macro when the main glossary is displayed:
\begin{codeenv}
\cs{printunsrtglossary*}\marg{\cs{let}\cs{printunsrtglossaryhandler}\csfmt{nestedhandler}}
\end{codeenv}
The main difficulty comes with ensuring that all the necessary
entries are selected. Now that the custom \fieldfmt{identifier}
field has been aliased to \field{type} rather than \field{parent},
the \code{animal} entry is no longer considered a dependent. The
\code{duck} entry has been referenced in the document with \cs{gls}
but the \code{animal} entry hasn't. The previous example ensured
that the \code{animal} entry was selected because it was a parent of
a selected entry. If the same resource options are used in this
example, the main glossary will be empty, which means that the
nested glossaries won't be displayed either.

One way to ensure that the \code{animal}, \code{mineral} and
\code{vegetable} entries are selected is to identify the \field{type}
field as a dependency field:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[{terms,animals,minerals,vegetables}]{src},
  \csopt[{identifier=type}]{field-aliases},
  \csopt[{type}]{dependency-fields}
}
\end{codeenv}
This will achieve the same effect as the
\exfile{sample-hierarchical.tex} document, but it's a far more
convoluted method. The reason this example document is listed here
is to demonstrate a sightly modified hierarchical effect that can't
be achieved through the normal method.

Suppose that, for some strange reason, I want the \qt{animal},
\qt{mineral} and \qt{vegetable} entries to be listed in a different
order (say, reverse alphabetical). The other entries (\qt{duck} etc)
need to be sorted in normal alphabetical order.

The \csopt{sort} option applies the same sort method to all
hierarchical levels. The sort \emph{value} chosen for particular entries
can be altered through the use of fallbacks (such as the
\csopt{entry-sort-fallback} or \csopt{symbol-sort-fallback} options)
and a letter comparator may be used to resolve identical sort
values (\csopt{identical-sort-action}), but the same sort algorithm
is applied to all entries in the same set (primary, secondary or dual
within the same resource set). The only way to apply different sort
methods is to separate the entries into different resource sets (or
use dual or secondary sorting).

This can be achieved by having one resource set for the main entries
with one sort method and another resource set for all the other
entries with a different sort method:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[terms]{src},\csopt[en-reverse]{sort}}

\gls{GlsXtrLoadResources}\oarg{\csopt[animals,minerals,vegetables]{src},\csopt[en]{sort},
  \csopt[{identifier=type}]{field-aliases},\csopt[{type}]{dependency-fields}
}
\end{codeenv}
This works when cross-resource dependencies are permitted (see
\sectionref{sec:resourcesets}). In the event that cross-resource
dependencies aren't permitted, the selection criteria is more
complicated:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{\csopt[terms,animals,minerals,vegetables]{src},
 \csopt[en-reverse]{sort},
 \csopt[{identifier=parent}]{field-aliases},
 \csopt[ancestors but not recorded]{selection}
}

\gls{GlsXtrLoadResources}\oarg{
  \csopt[animals,minerals,vegetables]{src},
  \csopt[{identifier=type}]{field-aliases},
  \csopt[{type}]{dependency-fields}
  \csopt[en]{sort}
}
\end{codeenv}
Fortunately in this example, cross-resource dependencies are
permitted so the simpler alternative works. (If they're not
permitted, the \bibgls\ transcript file will contain
\qt{Cross-resource references can't be supported for resource set
\meta{filename}}.)

The complete code is listed below.  The document build is:
\begin{verbatim}
pdflatex sample-nested
bib2gls --group --provide-glossaries sample-nested
pdflatex sample-nested
\end{verbatim}
The complete document is shown in \figureref{fig:sample-nested.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-nested.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-nested.pdf}}%
}
{\caption{\filefmt{sample-nested.pdf}}}
{fig:sample-nested.pdf}
\end{figure}

\filesection{sample-constants.tex}

This example uses the \exfile{constants.bib} file. The aim here
is to just have a list of all the constants defined in the \ext{bib}
file. (There are no references in the document.) This means I need
to use:
\begin{codeenv}
\csopt[all]{selection}
\end{codeenv}
in order to select all entries. I also need to alias the custom
\atentryfmt{constant} entry type otherwise all the entries will be
ignored. I decided to make \atentryfmt{constant} behave like
\atentry{number} for semantic reasons:
\begin{codeenv}
\csopt[constant=number]{entry-type-aliases}
\end{codeenv}
The custom fields also need aliasing:
\begin{codeenv}
\csopt[
  \fieldfmt{identifier}=\field{category},
  \fieldfmt{constantsymbol}=\field{name},
  \fieldfmt{constantname}=\field{description},
  \fieldfmt{value}=\field{user1},
  \fieldfmt{definition}=\field{user2},
  \fieldfmt{alternative}=\field{user3},
]{field-aliases}
\end{codeenv}
I decided to use the \glostyle{altlist} style, so I've instructed
\bibgls\ to determine the widest name:
\begin{codeenv}
\csopt{set-widest}
\end{codeenv}
It's always a good idea to specify the glossary type when using
\csopt{set-widest}, although in this example there's only one
glossary so it doesn't make much difference.
\begin{codeenv}
\csopt[main]{type}
\end{codeenv}
I decided to order the constants according to their (approximate)
numerical value. I've aliased the custom \fieldfmt{value} field to
\field{user1}, so I can sort by that field using a numerical
comparison:
\begin{codeenv}
\csopt[\field{user1}]{sort-field},
\csopt[double]{sort}
\end{codeenv}
There are three entries without the \fieldfmt{user1} field (as the
custom \fieldfmt{value} field is missing in the \ext{bib} file):
\code{zero}, \code{one} and \code{imaginary}. In the case of
\code{zero} and \code{one} the exact value can be obtained from the
\field{name} field. Since I've change the default \csopt{sort-field}, I
can't use \csopt{symbol-sort-fallback}. Instead I need to use:
\begin{codeenv}
\csopt[\field{name}]{missing-sort-fallback}
\end{codeenv}

What happens with the \code{imaginary} entry? It has no real
representation.  The transcript (\ext{glg}) file shows the message:
\begin{verbatim}
Warning: Can't parse sort value 'i' for: imaginary
\end{verbatim}
With the numerical sort methods, if the field can't be parsed the
value defaults to 0. This means that both \code{zero} and
\code{imaginary} have 0 as the sort value, so the
\csopt{identical-sort-action} is implemented. The default setting
means that \bibgls\ will fallback on comparing the entry labels, so
\code{imaginary} comes before \code{zero}.

Since I'm just using the \glostyle{alttree} style, I only need
\sty{glossary-tree}. I can improve efficiency in the document build
by preventing the other glossary style packages from being loaded
using the \styopt{nostyles} package option. This also prevents
\sty{glossary-tree} from being loaded, but I can both load it and
patch the styles with \sty{glossaries-extra-stylemods} through the
option \styopt[tree]{stylemods}. Since the default \glostyle{list}
style is no longer available, I need to set a new default with
\styopt[alttree]{style}. I also want to automatically insert a
\idx{full-stop} after the description, which can be done with
\styopt{postdot}. Don't forget that the \styopt{record} option is
always needed when using \bibgls. This means that the
\sty{glossaries-extra} package needs to be loaded as follows:
\begin{codeenv}
\cs{usepackage}\oarg{\styopt{record},\styopt{nostyles},\styopt{postdot},\styopt[tree]{stylemods},\styopt[alttree]{style}}
 \marg{glossaries-extra}
\end{codeenv}
I've assigned the custom \fieldfmt{constantname} field to the
\field{description} field and the custom \fieldfmt{constantsymbol} field to the
\field{name} field. This means that by default the glossary list
will just show the symbolic representation and the constant's name.
I'd like to append the value and definition after the description.
With the base \sty{glossaries} package this would require defining a
new glossary style but with \sty{glossaries-extra} it can easily be
achieved through the \idx{postdescriptionhook}.

I've aliased the custom \fieldfmt{identifier} field to
\field{category}, which means that all the entries will have the
\field{category} set to \code{constant}. The
\idx{postdescriptionhook} is obtained from
\idx{glsxtrpostdesccategory}, so I need to define
the command \csfmt{glsxtrpostdescconstant}. A simple definition is:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{glsxtrpostdescconstant}}\marg{\comment{}
 \cs{space} (approximately \cs{glsentryuseri}\marg{\cs{glscurrententrylabel}})\comment{}
 : \cs{glsentryuserii}\marg{\cs{glscurrententrylabel}}\comment{}
}
\end{codeenv}
This is fine if all entries have the \field{user1} and \field{user2}
fields set. A more generic approach tests for the existence of these
fields. This can either be done with \ics{ifglshasfield}:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{glsxtrpostdescconstant}}\marg{\comment{}
  \cs{ifglshasfield}\marg{\field{user1}}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{ (approximately \cs{glscurrentfieldvalue})}\comment{}
  \marg{}\comment{}
  \cs{ifglshasfield}\marg{\field{user2}}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{: \cs{glscurrentfieldvalue}}\comment{}
  \marg{}\comment{}
}
\end{codeenv}
or with \ics{glsxtrifhasfield}:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{glsxtrpostdescconstant}}{\comment{}
  \cs{glsxtrifhasfield}\marg{\field{useri}}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{ (approximately \cs{glscurrentfieldvalue})}\comment{}
  \marg{}\comment{}
  \cs{glsxtrifhasfield}\marg{\field{userii}}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{: \cs{glscurrentfieldvalue}}\comment{}
  \marg{}\comment{}
}
\end{codeenv}
(Note the need to use the internal field label \field{useri} and
\field{userii} with \ics{glsxtrifhasfield}.)

A modification can be made to also show the alternative
representation (obtained from the custom \fieldfmt{alternative}
field which has been aliased to \field{user3}):
\begin{codeenv}
\cs{newcommand}\marg{\cmd{glsxtrpostdescconstant}}{\comment{}
  \cs{glsxtrifhasfield}\marg{\field{useriii}}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{ (also denoted \cs{glscurrentfieldvalue}
      \cs{glsxtrifhasfield}\marg{\field{useri}}\marg{\cs{glscurrententrylabel}}\comment{}
      \marg{, approximately \cs{glscurrentfieldvalue}}\comment{}
      \marg{}\comment{}
   )\comment{}
  }\comment{}
  \marg{\comment{}
    \cs{glsxtrifhasfield}\marg{\field{useri}}\marg{\cs{glscurrententrylabel}}\comment{}
    \marg{ (approximately \cs{glscurrentfieldvalue})}\comment{}
    \marg{}\comment{}
  }\comment{}
  \cs{glsxtrifhasfield}\marg{\field{userii}}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{: \cs{glscurrentfieldvalue}}\comment{}
  \marg{}\comment{}
}
\end{codeenv}
If you have at least \styfmt{glossaries-extra} v1.31, it's better to use:
\begin{codeenv}
\ics{glsdefpostdesc}\marg{constant}
\end{codeenv}
instead of:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{glsxtrpostdescconstant}}
\end{codeenv}
as it can guard against accidental misspelling of the \code{glsxtrpostdesc}
part of the command name.

The complete code is listed below.  The document build is:
\begin{verbatim}
pdflatex sample-constants
bib2gls sample-constants
pdflatex sample-constants
\end{verbatim}
The complete document is shown in \figureref{fig:sample-constants.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-constants.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-constants.pdf}}%
}
{\caption{\filefmt{sample-constants.pdf}}}
{fig:sample-constants.pdf}
\end{figure}

\filesection{sample-chemical.tex}

This example just uses the \exfile{chemicalformula.bib} file. The aim here
is to have a list of chemical formulae referenced in the document
but not have a number list. I could use the \styopt{nonumberlist}
package option to suppress the number list display, but it's more
efficient to instruct \bibgls\ to not save the number list with:
\begin{codeenv}
\csopt[false]{save-locations}
\end{codeenv}
All entries are defined in \exfile{chemicalformula.bib} using
a custom entry type \atentryfmt{chemical} which needs to be aliased
in order for the entries to be recognised:
\begin{codeenv}
\csopt[chemical=symbol]{entry-type-aliases}
\end{codeenv}
Additionally, the entries only have custom fields, so these also
need to be aliased. In this case I want the formula in the \field{name} field
and the chemical name in the \field{description} field:
\begin{codeenv}
\csopt[\fieldfmt{formula}=\field{name},\fieldfmt{chemicalname}=\field{description}]{field-aliases}
\end{codeenv}
The \atentry{symbol} entry type falls back on the label for the
\field{sort} value by default, but I've decided
to fallback on the \field{name} field for sorting:
\begin{codeenv}
\csopt[\field{name}]{symbol-sort-fallback}
\end{codeenv}
An alternative approach would simply be to alias
\atentryfmt{chemical} to \atentry{entry} instead.

Since the \field{name} field contains chemical formulae rather than
words, it makes more sense to use one of the letter sort methods
rather than a locale collator. In this case the names contain
mixtures of letters and numbers, so one of the letter-number sort
methods (listed in \tableref{tab:sortoptionsletternumber}) would be
appropriate.

I want to use the \glostyle{alttreegroup} style (provided by
\sty{glossary-tree}). Since I don't require the other style
packages, I've used \styopt{nostyles} to suppress the automatic
loading and \styopt[tree]{stylemods} to both load
\sty{glossary-tree} and patch it. The \glostyle{alttreegroup} style
needs to know the widest name, so I've use \csopt{set-widest} for
convenience. The default behaviour of the tree styles is to 
format the name in bold. This is done through the command
\ics{glstreenamefmt} which is defined as:
\begin{codeenv}
\cs{newcommand}*\marg{\cs{glstreenamefmt}}[1]\marg{\cs{textbf}\marg{\idx{param}1}}
\end{codeenv}
The group headings use \ics{glstreegroupheaderfmt} which defaults to
\cs{glstreenamefmt}. Since I want to keep bold headings, I need to
redefine this as well:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glstreenamefmt}}[1]\marg{\idx{param}1}
\cs{renewcommand}*\marg{\cs{glstreegroupheaderfmt}}[1]\marg{\cs{textbf}\marg{\idx{param}1}}
\end{codeenv}
(For a more compact layout, you could use \glostyle{mcolalttreegroup}
instead.)
I also need the \longarg{group} switch to make the sort method
automatically assign letter groups.

The complete code is listed below.  The document build is:
\begin{verbatim}
pdflatex sample-chemical
bib2gls --group sample-chemical
pdflatex sample-chemical
\end{verbatim}
The complete document is shown in \figureref{fig:sample-chemical.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-chemical.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-chemical.pdf}}%
}
{\caption{\filefmt{sample-chemical.pdf}}}
{fig:sample-chemical.pdf}
\end{figure}

\filesection{sample-bacteria.tex}

This example just uses the \exfile{bacteria.bib} file. The aim here
is to have a simple list of the bacteria referenced in the document.
Bacteria names are often shown in the long form on first use
(without the short form) and then the short form on subsequent use.
This can easily be done with the \abbrstyle{long-only-short-only}
style. Bacteria are usually typeset in italic. It's best to create a
semantic command for this:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{bacteriafont}}[1]\marg{\cs{emph}\marg{\idx{param}1}}
\end{codeenv}
There are two methods to apply this to the bacteria entries. The
first is to redefine the formatting commands used by the
\abbrstyle{long-only-short-only} style:
\begin{codeenv}
\cs{renewcommand}*\marg{\ics{glsabbrvonlyfont}}[1]\marg{\cmd{bacteriafont}\marg{\idx{param}1}}
\cs{renewcommand}*\marg{\ics{glslongonlyfont}}[1]\marg{\cmd{bacteriafont}\marg{\idx{param}1}}
\end{codeenv}
This is fine if I don't intend to use this style for other types of
abbreviations. However, I may decide to extend the document at a
later date to include other abbreviations that need
\abbrstyle{long-only-short-only} but shouldn't be emphasized.
This can be done through the use of \idxpl{categoryattribute}.
The font used for the \field{name} in the glossary is governed by
the \catattr{glossnamefont} attribute, the font used for the
\field{description} in the glossary is governed by the
\catattr{glossdescfont} attribute and the font used by commands like
\cs{gls} in the document is governed by the \catattr{textformat}
attribute (\styfmt{glossaries-extra} v1.21+). So if I set
the \field{category} to \code{bacteria} then I can do:
\begin{codeenv}
\cs{setabbreviationstyle}\oarg{bacteria}\marg{\abbrstyle{long-only-short-only}}
\cs{glssetcategoryattribute}\marg{bacteria}\marg{\catattr{textformat}}\marg{bacteriafont}
\cs{glssetcategoryattribute}\marg{bacteria}\marg{\catattr{glossnamefont}}\marg{bacteriafont}
\end{codeenv}
and (if the \field{description} field is displayed in the glossary):
\begin{codeenv}
\cs{glssetcategoryattribute}\marg{bacteria}\marg{\catattr{glossdescfont}}\marg{bacteriafont}
\end{codeenv}
(Note that the attribute value is the control sequence name without
the initial backslash.)

I'd like to use the \glostyle{bookindex} style, which is
provided by the \isty{glossary-bookindex} 
package.\footnote{\styfmt{glossary-bookindex}
is distributed with \styfmt{glossaries-extra} v1.21+.} This isn't loaded
automatically, but it can be loaded through the
\styopt{stylemods} package option:
\begin{codeenv}
\cmd{usepackage}[\styopt{record},\comment{use bib2gls}
 \styopt{nostyles},\comment{don't load default style packages}
 \styopt[bookindex]{stylemods},\comment{load glossary-bookindex.sty and patch styles}
 \styopt[bookindex]{style}]\marg{glossaries-extra}
\end{codeenv}
I've used the \styopt{nostyles} package option to suppress loading
the default style packages, since I'm not using them. If you inspect
the \ext{log} file, you may notice that \sty{glossary-tree} is still
loaded. This is because it's required by \sty{glossary-bookindex}
as the \glostyle{bookindex} style is based on the \glostyle{index}
style provided by \sty{glossary-tree}.
With this style I need to use the \longarg{group} switch to instruct
the sort method to automatically create the letter groups.

The \glostyle{bookindex} style doesn't show the \field{description}
field (which means I don't need the \catattr{glossdescfont}
attribute) and, since the \abbrstyle{long-only-short-only} style sets
the \field{name} to the short form by default, only the short form
will show in the glossary. I'd rather it was just the long form.
This could simply be done using \csopt{replicate-fields} to copy the
\field{long} field to the \field{name} field:
\begin{codeenv}
\csopt[\field{long}=\field{name}]{replicate-fields}
\end{codeenv}
Again, I want to consider the possibility of adding other types of
abbreviations and this might not be appropriate for them (for
example, I might want some abbreviations with the long form followed
by the short form in parentheses). Another
approach is to redefine \ics{glsxtrbookindexname} which is used by
the \glostyle{bookindex} style to display the name. This takes the
entry's label as the argument. The default definition is:
\begin{codeenv}
\cs{newcommand}*\marg{\cs{glsxtrbookindexname}}[1]\marg{\cs{glossentryname}\marg{\idx{param}1}}
\end{codeenv}
This can be changed to test for the entry's category:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrbookindexname}}[1]\marg{\comment{}
 \ics{glsifcategory}\marg{\idx{param}1}\marg{bacteria}
 \marg{\cs{glossentrynameother}\marg{\idx{param}1}\marg{\field{long}}}\comment{}
 \marg{\cs{glossentryname}\marg{\idx{param}1}}\comment{}
}
\end{codeenv}
Note that I've used \ics{glossentrynameother} here rather than
\ics{glsentrylong}. This ensures that it follows the same formatting
as \ics{glossentryname} (so it will use \ics{glsnamefont} or the
\catattr{glossnamefont} attribute, the \catattr{glossname}
attribute, and the \idx{postnamehook}, if set). In this case it
picks up the \catattr{glossnamefont} attribute, which is used
instead of \cs{glsnamefont}.

If the \field{sort} field is missing for abbreviation styles, the
fallback value is the \field{short} field (not the \field{name}
field). In this case it would be better to fallback on the
\field{long} field instead, which can be done with the
\csopt{abbreviation-sort-fallback} option:
\begin{codeenv}
\csopt[\field{long}]{abbreviation-sort-fallback}
\end{codeenv}
If I do add other types of abbreviations, they will all be sorted
according to the \field{long} form, but at least this way I can have
some \meta{long} (\meta{short}) names as well.

The complete code is listed below.  The document build is:
\begin{verbatim}
pdflatex sample-bacteria
bib2gls --group sample-bacteria
pdflatex sample-bacteria
\end{verbatim}
This simple example only references entries on the first page so all
entries just have~1 in the number list. The complete document is
shown in \figureref{fig:sample-bacteria.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-bacteria.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-bacteria.pdf}}%
}
{\caption{\filefmt{sample-bacteria.pdf}}}
{fig:sample-bacteria.pdf}
\end{figure}

\filesection{sample-units1.tex}

This example uses the \exfile{baseunits.bib} and
\exfile{derivedunits.bib} files. The aim here is to have a glossary
in two blocks: base units and derived units. This can be achieved by
first loading \exfile{baseunits.bib} with \csopt{group} set to the
desired group title (\qt{Base Units} in this case) and then load
\exfile{derivedunits.bib} with the \csopt{group} set to the desired
title (\qt{Derived Units} in this case). Remember that the
\field{group} field needs to be used as a label. If the group title
contains any problematic characters or commands, then it's better to
use labels:
\begin{codeenv}
\csopt[baseunits]{group}
\end{codeenv}
for the first \idx{resourceset} and
\begin{codeenv}
\csopt[derivedunits]{group}
\end{codeenv}
for the second, and then set the group titles:
\begin{codeenv}
\gls{glsxtrsetgrouptitle}\marg{baseunits}\marg{Base Units}
\gls{glsxtrsetgrouptitle}\marg{derivedunits}\marg{Derived Units}
\end{codeenv}
I've used this method to make it easier to adapt to other languages
that may need extended characters in the group titles.
The \csopt{group} option requires the \longarg{group} switch to
ensure that the \field{group} field is correctly assigned.

The \exfile{baseunits.bib} file use a custom entry type
\atentryfmt{unit}, which must be aliased otherwise \bibgls\
will ignore the entries. I decided to use \atentry{symbol} for
semantic reasons:
\begin{codeenv}
\csopt[unit=symbol]{entry-type-aliases}
\end{codeenv}
Similarly for the custom \atentryfmt{measurement} entry type in
\exfile{derivedunits.bib}:
\begin{codeenv}
\csopt[measurement=symbol]{entry-type-aliases}
\end{codeenv}
Remember that \atentry{symbol} uses the label as the default sort fallback,
so I've changed it to use \field{name} instead:
\begin{codeenv}
\csopt[\field{name}]{symbol-sort-fallback}
\end{codeenv}
An alternative approach would be to alias \atentryfmt{unit}
and \atentryfmt{measurement} to \atentry{entry} instead.

Since there's no \csopt{type} set, all entries end up in the main
glossary, but since there are two resource commands the glossary
ends up with sorted blocks.

The document doesn't include any commands like \cs{gls}, so
I've use \csopt[all]{selection} to select all entries in the \ext{bib}
files. There won't be any number lists since there are no records.
I need a glossary style that shows the \field{symbol} field so I've
used \glostyle{mcolindexgroup}. Again I've suppressed the automatic
loading of the default styles with \styopt{nostyles} and used
\styopt[mcols]{stylemods} to load \isty{glossary-mcols} and patch the
styles. Note that although I've used \styopt{nostyles}, the
\sty{glossary-tree} style is loaded as it's required by
\sty{glossary-mcols}.

As with the previous example, the custom fields need to be aliased:
\begin{codeenv}
\csopt[
  \fieldfmt{unitname}=\field{name},
  \fieldfmt{unitsymbol}=\field{symbol},
  \fieldfmt{measurement}=\field{description}
]{field-aliases}
\end{codeenv}

The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-units1
bib2gls --group sample-units1
pdflatex sample-units1
\end{verbatim}
The complete document is shown in \figureref{fig:sample-units1.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-units1.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-units1.pdf}}%
}
{\caption{\filefmt{sample-units1.pdf}}}
{fig:sample-units1.pdf}
\end{figure}

\filesection{sample-units2.tex}

This example is provided for comparison with
\exfile{sample-units1.tex}. Instead of having a single glossary with
sorted blocks this example has two glossaries:
\begin{codeenv}
\ics{newglossary*}\marg{baseunits}\marg{Base Units}
\cs{newglossary*}\marg{derivedunits}\marg{Derived Units}
\end{codeenv}
I've used the \styopt{section} package option to use \ics{section*} for the
glossary titles. This overrides the default \ics{chapter*} which is
used with book or report type of classes. I've also used the
\styopt{nomain} option to suppress the creation of the main glossary
as I want to define my own glossary types instead.

As before the custom entry types need to be aliased:
\begin{codeenv}
\csopt[unit=symbol]{entry-type-aliases}
\end{codeenv}
for the first \idx{resourceset} and
\begin{codeenv}
\csopt[measurement=symbol]{entry-type-aliases}
\end{codeenv}
for the second. Similarly for the custom entry fields:
\begin{codeenv}
\csopt[
  \fieldfmt{unitname}=\field{name},
  \fieldfmt{unitsymbol}=\field{symbol},
  \fieldfmt{measurement}=\field{description}
]{field-aliases}
\end{codeenv}
The \longarg{group} switch is needed to ensure that the
\field{group} field is automatically assigned by the sort method.

The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-units2
bib2gls --group sample-units2
pdflatex sample-units2
\end{verbatim}
The complete document is shown in \figureref{fig:sample-units2.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-units2.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-units2.pdf}}%
}
{\caption{\filefmt{sample-units2.pdf}}}
{fig:sample-units2.pdf}
\end{figure}

\filesection{sample-units3.tex}

This is another example that uses the \exfile{baseunits.bib} and
\exfile{derivedunits.bib} files. As before the custom fields need to be aliased:
\begin{codeenv}
\csopt[
  \fieldfmt{unitname}=\field{name},
  \fieldfmt{unitsymbol}=\field{symbol},
  \fieldfmt{measurement}=\field{description}
]{field-aliases}
\end{codeenv}
This time I want two glossaries containing all the units (base and
derived) where the first glossary is ordered by name and the second
is ordered by symbol. This can be done with a single resource
command that instructs \bibgls\ to make the custom \atentryfmt{unit}
and \atentryfmt{measurement} entry types behave like
\atentry{dualsymbol}:
\begin{codeenv}
\csopt[
  unit=dualsymbol,
  measurement=dualsymbol
]{entry-type-aliases}
\end{codeenv}
This causes the \field{name} and \field{symbol} fields to be swapped in the
dual list. Remember that the fallback for the \field{sort} field is the label
for the symbol entry types so I need \csopt[\field{name}]{symbol-sort-fallback}
to fallback on \field{name} field instead. (Alternative, I could just sort by
the \field{name} field instead using \csopt[\field{name}]{sort-field}.)

The primary entries can still be sorted according to the default
locale collator, but the dual entries need a sort method that's
better suited to symbols. Fortunately, \bibgls\ has some (very
limited) support for \isty{siunitx} and is able to interpret
the \ics{si} commands in the sample \ext{bib} files. Since 
\idxpl{SIunit} are a mix of letters and numbers I've used one of the
letter-number methods listed in
\tableref{tab:sortoptionsletternumber}.

I've decided to define a custom style for the first glossary. Since
it's based on the \glostyle{long3col-booktabs} style I need to 
load \isty{glossary-longbooktabs}, which can conveniently be done with
the \styopt{stylemods} option. This uses \env{longtable} (provided
by \isty{longtable}, which is automatically loaded) which means
an extra \LaTeX\ call is required in the build process to ensure the
column widths are correct. Again I'm using \styopt{nostyles} to
suppress the automatic loading of the default styles, however
\sty{glossary-tree} will be loaded as it's listed in the value of
\styopt{stylemods} and \isty{glossary-long} will be loaded as it's
required by \sty{glossary-longbooktabs}. I can't use my custom style
in the \styopt{style} package option as it hasn't been defined at
that point. The default \glostyle{list} style is now unavailable
since \styopt{nostyles} has prevented it from being defined, so 
I've used \styopt[alttree]{style} to ensure there's a valid default
style.

Since my custom style is based on one of the long styles, I need to
set the length register \ics{glsdescwidth} to adjust the width of
the description column:
\begin{codeenv}
\cmd{setlength}\marg{\cs{glsdescwidth}}\marg{.4\cmd{hsize}}
\end{codeenv}
The \glostyle{long3col-booktabs} style sets up a three column
\env{longtable} so I just need to adjust the table header (to rename
the column headers) and the way each row is formatted:
\begin{codeenv}
\ics{newglossarystyle}\marg{units}\comment{style name}
\marg{\comment{base it on \glostyle{long3col-booktabs}}
  \ics{setglossarystyle}\marg{long3col-booktabs}\comment{}
  \cs{renewcommand}*\marg{\ics{glossaryheader}}\marg{\comment{}
    \ics{toprule}
    \ics{bfseries} Name \idx{colsep}
    \cs{bfseries} Measurement \idx{colsep}
    \cs{bfseries} Symbol
    \ics{tabularnewline}\ics{midrule}\ics{endhead}
    \ics{bottomrule}\ics{endfoot}}\comment{}
\comment{main entries:}
  \cs{renewcommand}\marg{\ics{glossentry}}[2]\marg{\comment{}
    \ics{glsentryitem}\marg{\idx{param}\idx{param}1}\ics{glstarget}\marg{\idx{param}\idx{param}1}\marg{\cs{glossentryname}\marg{\idx{param}\idx{param}1}} \idx{colsep}
    \cs{glossentrydesc}\marg{\idx{param}\idx{param}1}\ics{glspostdescription} \idx{colsep}
    \cs{glossentrysymbol}\marg{\idx{param}\idx{param}1}\cs{tabularnewline}
  }\comment{}
}
\end{codeenv}
There are no sub-entries in this document so I haven't bothered to
redefine \ics{subglossentry}. (The tabular styles aren't appropriate
for hierarchical glossaries.) This puts the symbol into the third
column (rather than the location list, which is ignored).
This style supports the letter group separator (although it doesn't
title the groups), so if I want this I need to use the
\longarg{group} switch.

I also need to make sure I've defined a glossary for the dual
entries:
\begin{codeenv}
\cs{newglossary*}\marg{units}\marg{Units of Measurement (by SI unit)}
\end{codeenv}
and specify the glossary types for the primary and dual entries:
\begin{codeenv}
\csopt[main]{type},
\csopt[units]{dual-type}
\end{codeenv}
The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-units3
bib2gls --group sample-units3
pdflatex sample-units3
pdflatex sample-units3
\end{verbatim}
The two pages of the document are shown in
\figureref{fig:sample-units3.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-units3.tex}

\begin{figure}
\figcontents
{\pdftwocol{../examples/sample-units3.pdf}{2}}
{\caption{\filefmt{sample-units3.pdf}}}
{fig:sample-units3.pdf}
\end{figure}

\filesection{sample-media.tex}

This example uses the sample files \exfile{books.bib},
\exfile{films.bib}, \exfile{no-interpret-preamble.bib} and
\exfile{interpret-preamble.bib}.  The aim is to produce a combined
list of books and films in a single glossary. The films are based on
some of the books so some of the entries have the same name. The
default setting for identical sort values is
\csopt[id]{identical-sort-action}, which means that the ordering for
the duplicate names is based on the entry labels. This can lead to
the odd effect of sometimes having the film listed first
(\code{film.thebigsleep} comes before \code{thebigsleep}) and
sometimes having the book listed first (\code{brightonrock} comes
before \code{film.brightonrock}).

One possible solution would be to also assign prefixes for the book
labels, but \csopt{label-prefix} is applied to all primary entries
for the given resource set and can't be applied selectively, so this
would require editing the \exfile{books.bib} file.

A more consistent approach would be to fallback on the category.
This means that the \field{category} field needs to be set. There
are two simple ways to achieve this: use \csopt[same as base]{category}
(which sets the \field{category} to \code{books} for entries in 
\exfile{books.bib} and to \code{films} for entries in
\exfile{films.bib}) or alias the custom \fieldfmt{identifier} field to
\field{category}. I've chosen the latter method and also provided
aliases for the custom \fieldfmt{year} and \fieldfmt{cast} fields:
\begin{codeenv}
\csopt[\fieldfmt{identifier}=\field{category},\fieldfmt{year}=\field{user1},\fieldfmt{cast}=\field{user2}]{field-aliases},
\csopt[\field{category}]{identical-sort-action}
\end{codeenv}
This ensures that books always come before films with the same
title. An oddity is the film \qt{Whisky Galore!}\ which is one
character different from the book \qt{Whisky Galore} but the default
locale collator ignores punctuation so the two titles are considered
identical by the collator (but not by \csopt[non-unique]{sort-suffix}).
If a letter comparison was used instead, they would no longer be 
considered identical, but in this case the film would
still be placed after the book since the film title is longer.

Since I've set the \field{category} I can provide semantic
formatting commands (as for \exfile{sample-bacteria.tex}):
\begin{codeenv}
\cs{newcommand}*\marg{\cmd{bookfont}}[1]\marg{\cs{emph}\marg{\idx{param}1}}
\cs{newcommand}*\marg{\cmd{filmfont}}[1]\marg{\ics{textsf}\marg{\ics{em} \idx{param}1}}
\cs{glssetcategoryattribute}\marg{book}\marg{\catattr{textformat}}\marg{bookfont}
\cs{glssetcategoryattribute}\marg{book}\marg{\catattr{glossnamefont}}\marg{bookfont}
\cs{glssetcategoryattribute}\marg{film}\marg{\catattr{textformat}}\marg{filmfont}
\cs{glssetcategoryattribute}\marg{film}\marg{\catattr{glossnamefont}}\marg{filmfont}
\end{codeenv}
I've given films a slightly different format to make them easier to
distinguish from books of the same name.

Both \exfile{books.bib} and \exfile{films.bib} had the custom
\fieldfmt{year} field, indicating the year of first publication
or release, which I've assigned to the \field{user1} field. 
I can define \idxpl{postnamehook} for each category
to append the year in brackets after the name is displayed in the
glossary:
\begin{codeenv}
\cs{newcommand}*\marg{\postnamehook{book}}\marg{\comment{}
 \cs{ifglshasfield}\marg{\field{user1}}\marg{\cs{glscurrententrylabel}}\comment{}
 \marg{\cs{space}(published \cs{glscurrentfieldvalue})}\comment{}
 \marg{}\comment{}
}
\strut
\cs{newcommand}*\marg{\postnamehook{film}}\marg{\comment{}
 \cs{ifglshasfield}\marg{\field{user1}}\marg{\cs{glscurrententrylabel}}\comment{}
 \marg{\cs{space}(released \cs{glscurrentfieldvalue})}\comment{}
 \marg{}\comment{}
}
\end{codeenv}
As with the post-description hook, if you have at least 
\styfmt{glossaries-extra} v1.31, it's better to use:
\begin{codeenv}
\ics{glsdefpostname}\margm{category}
\end{codeenv}
instead of:
\begin{codeenv}
\cs{newcommand}\marg{\cs{glsxtrpostnamecategory}}
\end{codeenv}
as it can guard against accidental misspelling of the \code{glsxtrpostname}
part of the command name.

I've assigned the \fieldfmt{cast} field to the \field{user2} field,
and since this field uses \BibTeX's contributor markup I need to 
convert this to a form that's easier to customize:
\begin{codeenv}
\csopt[\field{user2}]{bibtex-contributor-fields}
\end{codeenv}
I'm not sorting by this field and it would look better in the
document to list the forenames before the surname so I've also done:
\begin{codeenv}
\csopt[forenames]{contributor-order}
\end{codeenv}
Since I have at least version 2.28 of \sty{datatool-base} installed, 
the list will be formatted using \ics{DTLformatlist}. If I want an 
Oxford comma, I need to redefine \ics{DTLlistformatoxford} in the document:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{DTLlistformatoxford}}\marg{,}
\end{codeenv}
If I want to change \qt{\&} to \qt{and} I also need to redefine
\ics{DTLandname}:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{DTLandname}}\marg{and}
\end{codeenv}
If \ics{DTLformatlist} isn't defined (\sty{datatool-base} v2.27 or
earlier), the cast list will look a little odd as it uses a comma
separator between all elements of this list, including the final
pair (so there's no final \& or \qt{and}).

I've provided a \idx{postdescriptionhook}
\ics{glsxtrpostdesccategory} to append the cast list:
\begin{codeenv}
\cs{newcommand}*\marg{\postdeschook{film}}\marg{\comment{}
 \cs{ifglshasfield}\marg{\field{user2}}\marg{\cs{glscurrententrylabel}}\comment{}
 \marg{\comment{}
   \cs{glsxtrrestorepostpunc} \comment{requires glossaries-extra v1.23+}
   \cs{cs.space}featuring \cs{glscurrentfieldvalue}
 }\comment{}
 \marg{}\comment{}
}
\end{codeenv}
This uses \ics{glsxtrrestorepostpunc} to restore the
post-description punctuation if it was suppressed with
\ics{glsxtrnopostpunc}. This means that if I decide not to include
the \field{user2} field then the post-description punctuation will
be revert back to being suppressed for entries containing
\cs{glsxtrnopostpunc} in the \field{description} field.

I haven't referenced any of the entries in the main body of the
document, so I've used \csopt[all]{selection} to select all entries.
This means that there are no number lists on the first document
build (\LaTeX+\bibgls+\LaTeX) but the next build would show locations for 
the books that have been referenced by the film entries. Since this
looks a bit odd, I've added \csopt[false]{save-locations} to
prevent \bibgls\ from saving the locations.

I've used a style that shows letter group headings so I need to use
the \longarg{group} switch.
The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-media
bib2gls --group sample-media
pdflatex sample-media
\end{verbatim}
The four pages of the document are shown in
\figureref{fig:sample-media.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-media.tex}

\begin{figure}
\figcontents
{\pdftwocol{../examples/sample-media.pdf}{4}}
{\caption{\filefmt{sample-media.pdf}}}
{fig:sample-media.pdf}
\end{figure}

\filesection{sample-people.tex}

This example uses the files \exfile{people.bib},
 \exfile{no-interpret-preamble.bib} and \exfile{interpret-preamble.bib}. The
aim here is to have a list of people ordered alphabetically by
surname with a brief description, the same list ordered by date of
birth and an index of all the people without their details but with
a number list indicating where that person was mentioned in the
document. The first two lists shouldn't include aliases but the
index should. Not all the entries defined in \exfile{people.bib}
are included in the document. Those that aren't either explicitly
referenced or aliased are filtered by the \csopt{selection} criteria.
I've used a style that shows letter group headings so I need to use
the \longarg{group} switch.

Since this is just an example document all the \cs{gls} commands
only occur on page~1, which means that each number list is just
\qt{1}. A real document would have the references scattered about.
The aliases haven't actually been referenced anywhere in the
document.

The \fieldfmt{born}, \fieldfmt{died} and \fieldfmt{othername} fields will be
ignored by default since they don't correspond to recognised keys,
so the keys either need to be defined or the fields need to be
mapped to existing keys. In this case I've decided to map them to
the \field{user1}, \field{user2} and \field{user3} fields using
\csopt{field-aliases}:
\begin{codeenv}
\csopt[\fieldfmt{born}=\field{user1},\fieldfmt{died}=\field{user2},\fieldfmt{othername}=\field{user3}]{field-aliases}
\end{codeenv}
Although the aliases haven't been referenced in the document, I've
taken into account the possibility that they might later be added.
To prevent them from showing in the first two lists I've filtered
them out. This is easy to do since the aliases are all defined using
\atentry{index} whereas the remaining (non-aliased) entries are
defined using \atentry{entry} so \csopt{match} can be used to only
select entries defined with \atentry{entry}: 
\begin{codeenv}
\csopt[entrytype=entry]{match}
\end{codeenv}

I'd like the first use of \ics{gls} to display the full name, except
for the entry that has the \field{first} field set. The remaining
entries only have \field{text} set to a shortened version of the
name so they need to have the \field{name} field copied to the
\field{first} field using \csopt{replicate-fields}:
\begin{codeenv}
\csopt[\field{name}=\marg{\field{first}}]{replicate-fields}
\end{codeenv}

I'd like the first use to show the other name in parentheses where
provided. The simplest way to achieve this is by defining the
\idx{postlinkhook} \ics{glsxtrpostlinkcategory}. If the
\field{category} field isn't specified it will default to
\code{general} (for entries defined with \atentry{entry}), so I
could just define \csfmt{glsxtrpostlinkgeneral} but to allow for
the possibility of extending the document to incorporate other types
of entries I decided to set the \field{category} to \code{people}
through the use of the \csopt{category} option:
\begin{codeenv}
\csopt[people]{category}
\end{codeenv}

This means that I now need to define a command called
\csfmt{glsxtrpostlinkpeople} that will be used after instances of
\cs{gls} etc where the entry has the \field{category} set to
\code{people}. This first tests if that was the first use of the
entry with \ics{glsxtrifwasfirstuse} and then tests if the
\field{user3} field is set. If so, it does a space followed by that
field's value in parentheses. The entry's label can be obtained from
\ics{glslabel}:
\begin{codeenv}
\cs{newcommand}*\marg{\postlinkhook{people}}\marg{\comment{}
 \cs{glsxtrifwasfirstuse}
 \marg{\comment{}
   \cs{ifglshasfield}\marg{\field{user3}}\marg{\cs{glslabel}}\comment{}
   \marg{\cs{space}(\cs{glscurrentfieldvalue})}\comment{}
   \marg{}\comment{}
 }\comment{}
 \marg{}\comment{}
}
\end{codeenv}

I'd also like to do something similar after the name when the entry
is displayed in the glossary. This means defining the \idx{postnamehook}
\ics{glsxtrpostnamecategory}, in this case
\csfmt{glsxtrpostnamepeople}. The entry's label is referenced
with \ics{glscurrententrylabel}:
\begin{codeenv}
\cs{newcommand}*\marg{\postnamehook{people}}\marg{\comment{}
 \cs{ifglshasfield}\marg{\field{user3}}\marg{\cs{glscurrententrylabel}}\comment{}
 \marg{\cs{space}(\cs{glscurrentfieldvalue})}\comment{}
 \marg{}\comment{}
}
\end{codeenv}
(A different command is used since \cs{gls} may occur in the
description, which would interfere with the current entry label if
they shared the same command to reference the label.)

The \idx{postdescriptionhook} can be used to append the birth and
death dates. Although all the entries that have been selected from \exfile{people.bib} have a 
\fieldfmt{died} field, I've added a check for the corresponding
\field{user3} field in case new references are added for people who are
still alive:
\begin{codeenv}
\cs{newcommand}*\marg{\postdeschook{people}}\marg{\comment{}
 \cs{ifglshasfield}\marg{\field{user1}}\marg{\cs{glscurrententrylabel}}
 \marg{\comment{born}
   \cs{space}(\cs{glscurrentfieldvalue}\ics{comma}-{}-\cs{comma}\comment{}
     \cs{ifglshasfield}\marg{user2}\marg{\cs{glscurrententrylabel}}
     \marg{\comment{died}
        \cs{glscurrentfieldvalue}
     }\comment{}
     \marg{}\comment{}
   )\comment{}
 }\comment{}
 \marg{}\comment{}
}
\end{codeenv}

The first list is quite straight-forward and can be created with:
\begin{codeenv}
\gls!{GlsXtrLoadResources}\oarg{
  \csopt[people]{src},
  \csopt[entrytype=entry]{match},
  \csopt[people]{category},
  \csopt[\field{name}=\marg{\field{first}}]{replicate-fields},
  \csopt[\fieldfmt{born}=\field{user1},\fieldfmt{died}=\field{user2},\fieldfmt{othername}=\field{user3}]{field-aliases}
}
\end{codeenv}
I have used the \csopt{sort} option and there's no document language, 
so \bibgls\ will sort according to my locale. The custom
commands \gls{sortname} and \gls{sortvonname} ensure that the
entries are all sorted alphabetically according to the surnames.

The second list can easily be created by adding the \csopt{secondary}
option:
\begin{codeenv}
\csopt[date:\field{user1}:bybirth]{secondary}
\end{codeenv}
This sorts according to the \field{user1} field (which was
originally the \fieldfmt{birth} field).
Note that different locales have different default date formats.
There may also be a difference in the default date format depending
on the Java \idx{localeprovider}. For example, if you switch from
using the \idx{JRE} to using the \idx{CLDR} you may find a change in
the default format. In case the format provided in the \ext{bib}
file isn't recognised, the required format can be set with:
\begin{codeenv}
  \csopt[d MMM YYYY G]{secondary-date-sort-format}
\end{codeenv}

I've changed the date group headings by redefining \gls{bibglsdategroup} and
\gls{bibglsdategrouptitle}, which means that the grouping in the
\code{bybirth} glossary will be in the form \meta{year} \meta{era}:
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglsdategroup}}[7]\marg{\idx{param}1\idx{param}4\idx{param}7}
\cs{newcommand}\marg{\gls{bibglsdategrouptitle}}[7]\marg{\cs{number}\idx{param}1\ \idx{param}4}
\end{codeenv}
I've also defined the \code{bybirth} glossary and supplied a title:
\begin{codeenv}
\cs{newglossary*}\marg{bybirth}\marg{People (Ordered by Birth)}
\end{codeenv}

The first two glossaries have entries with fairly long names (especially those
with the \idx{postnamehook}), so the best style is the
\glostyle{altlistgroup}. The \sty{glossaries-extra-stylemods}
package patches this style to discourage page breaks occurring after
group headings, so I've also used the \styopt{stylemods} 
option to automatically load that package. I'd like to use the
\glostyle{bookindex} style for the index, which is provided by 
\sty{glossary-bookindex}, so I need:
\begin{codeenv}
\styopt[list,bookindex]{stylemods}
\end{codeenv}
This ensures that \isty{glossary-list} and \sty{glossary-bookindex}
are loaded and patches the list styles.

The first two glossaries would look better with a terminating 
\idx{full-stop}, so I've used the \styopt{postdot} package option. 
(The \glostyle{bookindex} style doesn't use the \field{description}
field and therefore doesn't use the \idx{postdescriptionhook}.)
The index glossary type can be defined with the \styopt{index}
package option. I've set the default style to \glostyle{altlistgroup}
but this can locally be changed to \glostyle{bookindex} when I
display the index. The \styopt{record} option is needed to use
\bibgls, so the \styfmt{glossaries-extra} package is loaded with:
\begin{codeenv}
\cmd{usepackage}[\styopt{record},\comment{using bib2gls}
 \styopt{index},\comment{create index glossary}
 \styopt{postdot},\comment{dot after descriptions}
\comment{load glossary-list.sty and glossary-bookindex.sty and patch:}
 \styopt[list,bookindex]{stylemods},
 \styopt[altlistgroup]{style}]\marg{glossaries-extra}
\end{codeenv}

The index needs to include all the entries that have already been
defined but also needs to include the aliased entries. This means
that existing entries simply need their label copied to the
\code{index} glossary but the other entries need to be defined
so this requires setting the \csopt{action} option:
\begin{codeenv}
\csopt[define or copy]{action}
\end{codeenv}
I would also like to have groups in the index (which the
\glostyle{bookindex} style supports) so I need to specify a field
in which to save the group information using
\csopt{copy-action-group-field}:
\begin{codeenv}
\csopt[indexgroup]{copy-action-group-field}
\end{codeenv}
I need to remember to redefine \ics{glsxtrgroupfield} to this value before
displaying the index:
\begin{codeenv}
\cs{renewcommand}\marg{\cs{glsxtrgroupfield}}\marg{indexgroup}
\end{codeenv}

The aliased entries won't be selected by default since they haven't
been used in the document, so I need to change the selection
criteria with \csopt{selection}:
\begin{codeenv}
\csopt[recorded and deps and see]{selection}
\end{codeenv}

In the index, I'd like the surnames first. This can be done by
redefining the custom commands used in the \field{name} fields.
There's a slight complication here. These commands aren't defined on
the first \LaTeX\ run as their definitions are written to the
\ext{glstex} file by \bibgls, so I can't use \cs{renewcommand} (although 
I could use \cs{glsrenewcommand}).
Instead I've provided some custom commands:
\begin{codeenv}
\cs{newcommand}*\marg{\cmd{swaptwo}}[2]\marg{\idx{param}2, \idx{param}1}
\cs{newcommand}*\marg{\cmd{swapthree}}[3]\marg{\idx{param}2 \idx{param}3, \idx{param}1}
\end{codeenv}
Now I just need to make an assignment using \ics{let}:
\begin{codeenv}
\cs{let}\gls{sortname}\cmd{swaptwo}
\cs{let}\gls{sortart}\cmd{swaptwo}
\cs{let}\gls{sortvonname}\cmd{swapthree}
\end{codeenv}
This doesn't perform any check to determine if the commands are
already defined so there won't be a problem on the first run.

The first two glossaries shouldn't have number lists:
\begin{codeenv*}
\cs{printunsrtglossary}\oarg{\printglossopt[People (Alphabetical)]{title},\printglossopt{nonumberlist}}
\cs{printunsrtglossary}\oarg{\printglossopt[bybirth]{type},\printglossopt[false]{target},\printglossopt{nonumberlist}}
\end{codeenv*}
I'd like to use \sty{hyperref} but I have to switch off the
hypertargets for the second glossary otherwise I'll end up with
duplicate targets. This is done with \code{\printglossopt[false]{target}}. 
All references using \cs{gls} etc will link to the first glossary.

I could also do this for the index but the cross-references in the
aliased entries will link to the first glossary rather than the
relevant entry in the index. The simplest way to fix this is to
redefine \ics{glolinkprefix} to provide a different target:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glolinkprefix}}\marg{idx:}
\end{codeenv}
These redefinitions need to be done before the index. I've decided
to use the starred \ics{printunsrtglossary*} to localise these
changes, although that's not needed for this document since the
index comes right at the end:
\begin{codeenv*}
\cs{printunsrtglossary*}
 \oarg{\printglossopt[index]{type},\printglossopt[\glostyle{bookindex}]{style}}
 \marg{\comment{}
   \cs{let}\gls!{sortname}\cmd{swaptwo}
   \cs{let}\gls!{sortart}\cmd{swaptwo}
   \cs{let}\gls!{sortvonname}\cmd{swapthree}
   \cs{renewcommand}\marg{\cs{glsxtrgroupfield}}\marg{indexgroup}\comment{}
   \cs{renewcommand}*\marg{\cs{glolinkprefix}}\marg{idx:}\comment{}
 }
\end{codeenv*}
The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-people
bib2gls --group --break-space sample-people
pdflatex sample-people
\end{verbatim}
The four pages of the document are shown in \figureref{fig:sample-people.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-people.tex}

\begin{figure}
\figcontents
{\pdftwocol{../examples/sample-people.pdf}{4}}
{\caption{\filefmt{sample-people.pdf}}}
{fig:sample-people.pdf}
\end{figure}

\filesection{sample-authors.tex}

This example uses the files \exfile{people.bib},
\exfile{books.bib}, \exfile{no-interpret-preamble.bib} and 
\exfile{interpret-preamble2.bib}. The aim is to reference the books
in \exfile{books.bib} and have them listed by author. This means
finding a way of assigning each book entry a \field{parent} field that
contains the label identifying the relevant author in \exfile{people.bib}.
I've used a style that shows letter group headings so I need to use
the \longarg{group} switch.

To recap, each author is defined in \exfile{people.bib} in the form:
\begin{codeenv}
\atentry{entry}\marg{dickens,
  \field{name}=\marg{\gls{sortname}\marg{Charles}\marg{Dickens}},
  \field{text}=\marg{Dickens},
  \field{description}=\marg{English writer and social critic},
  \fieldfmt{born}=\marg{7\idx{nbspchar}February 1812 AD},
  \fieldfmt{died}=\marg{9\idx{nbspchar}June 1870 AD},
  \fieldfmt{identifier}=\marg{person}
}
\end{codeenv}
and each book is defined in \exfile{books.bib} in the form:
\begin{codeenv}
\atentry{entry}\marg{bleakhouse,
  \field{name}=\marg{Bleak House},
  \field{description}=\marg{novel by Charles Dickens},
  \fieldfmt{identifier}=\marg{book},
  \fieldfmt{author}=\marg{\gls{sortmediacreator}\marg{Charles}\marg{Dickens}},
  \fieldfmt{year}=\marg{1852}
}
\end{codeenv}
There's a field here (the custom \fieldfmt{author} field) that contains
the author's name, and this can be aliased to the \field{parent}
field with \csopt{field-aliases}:
\begin{codeenv}
\csopt[\fieldfmt{author}=\field{parent}]{field-aliases}
\end{codeenv}
but the author's label in the \exfile{people.bib}
file is just the \idx!{lowercase} surname.

Remember from \sectionref{sec:texparserlib} that the interpreter
will be used on the \field{parent} field if the value contains
\idx{escchar} or \idx{bgroupchar} or \idx{egroupchar} and
\csopt[true]{interpret-label-fields}.  This means that with this
field alias and the interpreter on, \bibgls\ will attempt to
interpret the field contents. So all that's needed is to ensure that
\bibgls\ is given a definition of \gls{sortmediacreator} that
ignores the first argument and converts the second argument to lower
case. This definition is available in
\exfile{interpret-preamble2.bib} but, since this file uses
\cs{renewcommand} rather than \cs{providecommand},
\csopt[false]{write-preamble} is required to prevent \LaTeX\ from
picking up this definition.

As with the \exfile{sample-people.tex} example, I need to copy
the \field{name} field to the \field{first} field if that field is
missing using \csopt{replicate-fields}:
\begin{codeenv}
\csopt[\field{name}=\marg{\field{first}}]{replicate-fields}
\end{codeenv}
and I also want to provide a semantic command to format the book
title, so the field aliases also need to convert the custom
\fieldfmt{identifier} field to \field{category}:
\begin{codeenv}
\csopt[\fieldfmt{identifier}=\field{category},\fieldfmt{author}=\field{parent}]{field-aliases}
\end{codeenv}
so that the document can set the \catattr{textformat} and
\catattr{glossnamefont} attributes:
\begin{codeenv}
\cs{newcommand}*\marg{\cmd{bookfont}}[1]\marg{\cs{emph}\marg{\idx{param}1}}
\cs{glssetcategoryattribute}\marg{book}\marg{\catattr{textformat}}\marg{bookfont}
\cs{glssetcategoryattribute}\marg{book}\marg{\catattr{glossnamefont}}\marg{bookfont}
\end{codeenv}
As with \exfile{sample-media.tex}, the terminating question mark at
the end of some of the \field{name} fields can cause an awkward
situation if \cs{gls} is used at the end of a sentence. This can be
dealt with by getting \bibgls\ to make a note of the fields that
end with sentence-terminating punctuation through the use of
the \csopt{check-end-punctuation} option. In this example, the
\field{name}, \field{text} and \field{first} fields are the same for
all the books, so it's sufficient just to check the \field{name}
field:
\begin{codeenv}
\csopt[\field{name}]{check-end-punctuation}
\end{codeenv}
With \sty{glossaries-extra} v1.23+ it's easy to hook into the
\idx{postlinkhook} to check if 
\fielddisp{fieldendpunc}{nameendpunc} exists:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrifcustomdiscardperiod}}[2]\marg{\comment{}
 \cs{GlsXtrIfFieldUndef}\marg{\fielddisp{fieldendpunc}{nameendpunc}}\marg{\cs{glslabel}}\marg{\idx{param}2}\marg{\idx{param}1}\comment{}
}
\end{codeenv}
This will now cause the full stops following:
\begin{codeenv}
\cs{gls}\marg{whydidnttheyaskevans}.
\end{codeenv}
and
\begin{codeenv}
\cs{gls}\marg{doandroidsdreamofelectricsheep}.
\end{codeenv}
to be discarded.

The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-authors
bib2gls --group sample-authors
pdflatex sample-authors
\end{verbatim}
The resulting document is shown in \figureref{fig:sample-authors.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-authors.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-authors.pdf}}%
}
{\caption{\filefmt{sample-authors.pdf}}}
{fig:sample-authors.pdf}
\end{figure}

\filesection{sample-citations.tex}

This example uses the \BibTeX\ file \exfile{citations.bib} to create
a document that has both a bibliography created by \BibTeX\ and
glossaries created by \bibgls\ listing the authors and the titles.
There are no glossary reference commands, such as \cs{gls}, but
\bibgls\ can be run with \longarg{cite-as-record} to treat the
\ics{citation} commands (written to the \ext{aux} file by \ics{cite})
as ignored records. Since \cs{cite} doesn't record the page number,
there are no associated locations.

The \code{main} glossary isn't required, so I've used
\styopt{nomain} to suppress its creation. I want to use both the
\glostyle{altlist} and \glostyle{indexgroup} styles but none of the
other styles, so I've used \styopt{nostyles} to prevent the
automatic loading of the default style packages and 
\styopt{stylemods} to load the \sty{glossary-tree} and
\sty{glossary-list} packages and patch the styles. A \idx{full-stop}
is automatically placed after the descriptions with
\styopt{postdot}.
\begin{codeenv}
\cs{usepackage}\oarg{\styopt{record},\comment{using bib2gls}
\styopt{nomain},\comment{don't define main glossary}
\styopt{postdot},\comment{full stop after descriptions}
\styopt{nostyles},\comment{don't load default styles}
\comment{load glossary-tree and glossary-list and patch styles:}
\styopt[tree,list]{stylemods}
}\marg{glossaries-extra}
\end{codeenv}
Next I need to create the glossaries for the list of authors and
list of titles:
\begin{codeenv}
\cs{newglossary*}\marg{contributors}\marg{Authors}
\cs{newglossary*}\marg{titles}\marg{Titles}
\end{codeenv}
The simplest way of assigning the authors to the \code{contributors}
glossary and the titles to the \code{titles} glossary is to use:
\begin{codeenv}
\csopt[contributors]{type}
\end{codeenv}
in the \idx{resourceset} and provide a
modified version of \gls{bibglsnewbibtexentry} that assigns
\field{type} after the options:
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglsnewbibtexentry}}[4]\marg{\comment{}
 \gls{longnewglossaryentry}*\marg{\idx{param}1}\marg{\field{name}=\marg{\idx{param}3},\idx{param}2,\field{type}=\marg{titles}}\marg{\idx{param}4}\comment{}
}
\end{codeenv}
The standard \BibTeX\ entry types need aliasing to
\atentry{bibtexentry}:
\begin{codeenv}
\csopt[\ics{GlsXtrBibTeXEntryAliases}]{entry-type-aliases}
\end{codeenv}
and the \fieldfmt{title} field is aliased to \field{name}:
\begin{codeenv}
\csopt[\fieldfmt{title}=\field{name}]{field-aliases}
\end{codeenv}
(The other fields aren't required for the glossary lists.)
The \field{category} is set to the original entry type:
\begin{codeenv}
\csopt[same as original entry]{category}
\end{codeenv}
So, for example, an entry that's provided in the \ext{bib} file with
\atentryfmt{article} has the \field{category} field set to
\code{article}. (Compare this with \csopt[same as entry]{category}
which would set the \field{category} to \code{bibtexentry}.)
The spawned entries are all defined using \atentry{contributor} and
aren't aliased so both the entry type and the original entry type
are \code{contributor}.

In order to list the titles according to category, I've use this as
the sort field:
\begin{codeenv}
\csopt[\field{category}]{sort-field}
\end{codeenv}
and setting the sort suffix to the \field{name} field sub-sorts
the \atentry{bibtexentry} types according to the title 
(which was aliased to the \field{name}) and the
\atentry{contributor} types according to the author:
\begin{codeenv}
\csopt[\field{name}]{sort-suffix}
\end{codeenv}

Next the groups identified by the labels \code{article} and \code{book} 
are assigned titles.
\begin{codeenv}
\gls{glsxtrsetgrouptitle}\marg{article}\marg{Articles}
\gls{glsxtrsetgrouptitle}\marg{book}\marg{Books}
\end{codeenv}
The \field{group} field is actually set to the associated letter by the default
\csopt{sort} method. The desired labels are stored in the \field{category} 
field. Since the entries are sorted by category, then they are naturally 
in those sub-blocks, which means that the group titles can be set by locally
redefining \ics{glsxtrgroupfield} to \code{category}:
\begin{codeenv}
\cs{printunsrtglossary*}[\printglossopt[titles]{type},\printglossopt[indexgroup]{style}]
\marg{\comment{}
 \cs{renewcommand}\marg{\cs{glsxtrgroupfield}}\marg{\field{category}}\comment{}
 \cs{renewcommand}\marg{\cs{glstreenamefmt}}[1]\marg{\cs{emph}\marg{\idx{param}1}}\comment{}
 \cs{renewcommand}\marg{\cs{glstreegroupheaderfmt}}[1]\marg{\ics{textbf}\marg{\idx{param}1}}\comment{}
}
\end{codeenv}
This again contradicts the advice given in
\sectionref{sec:logicaldivisions} as I'm sorting by the
\field{group} label. (Technically it's sorting by the
\field{category} label but this is being used as the group.) In this
case it's not a problem as the labels closely match the titles and
the sorting options ensure that the groups aren't broken up.

There's no \field{description} field set for these entries, but the
\idx{postdescriptionhook} can still be used to append information.
In this case, I've appended a cross-reference to the bibliography.
Since the bibliography entry and the glossary term both have the
same label, the citation can easily be obtained with
\code{\ics{cite}\marg{\cs{glscurrententrylabel}}}:
\begin{codeenv}
\cs{newcommand}\marg{\postdeschook{article}}\marg{\cs{cite}\marg{\cs{glscurrententrylabel}}}
\cs{newcommand}\marg{\postdeschook{book}}\marg{\cs{cite}\marg{\cs{glscurrententrylabel}}}
\end{codeenv}
Note that this needs to be done for each \BibTeX\ entry type, but in
this case the \ext{bib} file only contains \atentryfmt{article} and
\atentryfmt{book} entries. (Similarly for the group titles above.)

The list of contributors can simply be displayed with:
\begin{codeenv}
\cs{printunsrtglossary}[\printglossopt[contributors]{type},\printglossopt[altlist]{style}]
\end{codeenv}
This will only list the names as there's no description, but again
the \idx{postdescriptionhook} can be used, in this case for the \code{contributor}
category. The hook iterates
over the internal list provided by the \field{bibtexentry} field.
This allows the titles to be listed as well:
\begin{codeenv}
\cs{newcommand}\marg{\postdeschook{contributor}}\marg{\comment{}
  \ics{glsxtrifhasfield}\marg{\field{bibtexentry}}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{\comment{}
    \cs{glsxtrfieldforlistloop}
    \marg{\cs{glscurrententrylabel}}\marg{\field{bibtexentry}}\comment{}
    \marg{\csfmt{contributorhandler}}\comment{}
  }\comment{}
  \marg{\ics{par} No titles.}\comment{}
}
\end{codeenv}
The \idx{handler} macro displays the name of the associated
\atentry{bibtexentry} term and the citation:
\begin{codeenv}
\cs{newcommand}\marg{\csfmt{contributorhandler}}[1]\marg{\cs{par}\ics{glsentryname}\marg{\idx{param}1} \cs{cite}\marg{\idx{param}1}}
\end{codeenv}

The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-citations
bib2gls --cite-as-record sample-citations
bibtex sample-citations
pdflatex sample-citations
pdflatex sample-citations
\end{verbatim}
The resulting document is shown in \figureref{fig:sample-citations.pdf}.

\lstinputlisting[firstline=7]{../examples/sample-citations.tex}

\begin{figure}
\figcontents
{\pdftwocol{../examples/sample-citations.pdf}{2}}
{\caption{\filefmt{sample-citations.pdf}}}
{fig:sample-citations.pdf}
\end{figure}

\filesection{sample-msymbols.tex}

This example uses \exfile{bigmathsymbols.bib},
\exfile{mathsrelations.bib}, \exfile{binaryoperators.bib},
\exfile{unaryoperators.bib} and \exfile{mathgreek.bib}. The \isty{stix}
package is required for some of the commands used in
\exfile{bigmathsymbols.bib}, so that must be loaded in the document.

I'm using the \glostyle{mcolalttree} style for this document,
which means that the \sty{glossary-mcols} package is required and
the styles need patching, which can be done with 
the \styopt{stylemods} package option:
\begin{codeenv}
\cs{usepackage}\oarg{\styopt{record},\comment{using bib2gls}
 \styopt{nostyles},\comment{don't load default styles}
 \styopt{postdot},\comment{append a dot after descriptions}
 \styopt[mcols]{stylemods},\comment{load glossary-mcols.sty and patch}
 \styopt[mcolalttree]{style}}\marg{glossaries-extra}
\end{codeenv}

I'm not referencing any of the entries in the document as I'm just
generating a complete list of all the defined symbols. This means I
need to tell \bibgls\ to select all entries and don't bother saving
the \field{location} field:
\begin{codeenv}
\csopt[false]{save-locations},
\csopt[all]{selection}
\end{codeenv}
Since I'm using a style that's based on \glostyle{alttree} I need to
find the widest \field{name}, which can be done with \csopt{set-widest}.

The simplest way of dividing the glossary into logical blocks is to
sort according to the category, but first I need to use \csopt{field-aliases} 
to convert the custom \fieldfmt{identifier} field to \field{category}:
\begin{codeenv}
\csopt[\fieldfmt{identifier}=\field{category}]{field-aliases}
\end{codeenv}
and sort by the \field{category} field:
\begin{codeenv}
\csopt[\field{category}]{sort-field}
\end{codeenv}
Since this will cause identical sort values, I need to provide a
a way of ordering these identical values. Here I've decided 
to fallback on the \field{description} field:
\begin{codeenv}
\csopt[\field{description}]{identical-sort-action}
\end{codeenv}
This means that entries will be ordered by \field{category} and 
then \field{description}, which naturally creates blocks of symbol
types in the glossary. This only uses a simple case-sensitive string
comparison which is fine for English, but for another language it
would be better to use \csopt{sort-suffix} as in the
\exfile{sample-textsymbols.tex} file.

Remember that I want a small vertical gap between each logical
block. These need the \field{group} field which, with the default
locale sort, is obtained from the first letter of the sort value.
In this case the sort value is obtained from the \field{category}
field, and as each category happens to start with a different
letter, this means I get the desired effect. However, in the event
that I add more entries with a new category that happens to start
with the same letter as an existing category, it's better to provide
a more future-proof method, so I've set the \field{group} field to 
fetch its value from the \field{category} field:
\begin{codeenv}
\csopt[\field{category}=\field{group}]{replicate-fields}
\end{codeenv}
(Since the \csopt{field-aliases} option is always performed before
\csopt{replicate-fields}, the \field{category} field will already
have been set and is available for replicating.)

This means that I'm essentially sorting by the \field{group} labels,
which this manual has warned against doing. In this case, it's an
acceptable break from that rule as I've used options that ensure the
groups aren't broken up during sorting and I'm not concerned with the group
titles. A method such as that used in \exfile{sample-textsymbols2.tex}
would end up with titled blocks, which I don't want here. By using
resource options such as \csopt{field-aliases} and
\csopt{replicate-fields} I can avoid the warning that's triggered
with the default \longarg{warn-non-bib-fields}.

The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-msymbols
bib2gls sample-msymbols
pdflatex sample-msymbols
\end{verbatim}
The resulting document is shown in \figureref{fig:sample-msymbols.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-msymbols.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-msymbols.pdf}}%
}
{\caption{\filefmt{sample-msymbols.pdf}}}
{fig:sample-msymbols.pdf}
\end{figure}

\filesection{sample-maths.tex}

This example uses \exfile{bigmathsymbols.bib} and \exfile{mathsobjects.bib}.
It has a fairly similar preamble to \exfile{sample-msymbols.tex},
but \exfile{no-interpret-preamble.bib} and
\exfile{interpret-preamble.bib} are now needed to provide
the \gls!{sortart} command:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[no-interpret-preamble]{src},
  \csopt[false]{interpret-preamble}
}
\end{codeenv}
There's also an extra custom field to alias:
\begin{codeenv}
\csopt[\fieldfmt{identifier}=\field{category},\fieldfmt{format}=\field{user1}]{field-aliases}
\end{codeenv}
I've aliased \fieldfmt{format} to \field{user1} since \gls!{glsxtrfmt}
defaults to that field. If I decided to use a different field I also
need to remember to redefine \ics{GlsXtrFmtField} to match.

As with \exfile{sample-msymbols.tex} I'm sorting by the
\field{category} label and this value is copied to the \field{group}
field, but again I don't have a hierarchical glossary as the logical
blocks don't have titles.

In this document I only want to select entries that have been
indexed, so I've omitted the \csopt{selection} option I used in the
\exfile{sample-msymbols.tex} example, however
I still don't want any number lists so I still have
\csopt[false]{save-locations}.

I want \gls!{glsxtrfmt} to index the term (which it doesn't by default) so
that means I need to redefine \ics{GlsXtrFmtDefaultOptions} 
to prevent it from using \code{noindex}:
\begin{codeenv}
\cs{renewcommand}\marg{\cs{GlsXtrFmtDefaultOptions}}\marg{}
\end{codeenv}
I've provided some convenient wrapper commands that use
\gls!{glsxtrfmt*} or the non-linking \gls!{glsxtrentryfmt} that
are in the form:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{set}}[2][]\marg{\gls{glsxtrfmt*}\oarg{\idx{param}1}\marg{set}\marg{\idx{param}2}}
\cs{newcommand}\marg{\cmd{nlset}}[1]\marg{\gls{glsxtrentryfmt}\marg{set}\marg{\idx{param}1}}
\end{codeenv}
The use of the starred form allows:
\begin{codeenv}
\cmd{[}\cmd{set}\marg{A} = \cs{gls}\marg{bigcup}[\idx{sbchar}\marg{i=1}\idx{spchar}n] \cmd{set}\marg{B}[\idx{sbchar}i] \cmd{]}
\end{codeenv}
which produces:
\[ \mathcal{A} = \bigcup_{i=1}^n \mathcal{B}_i \]
Note the difference if the optional arguments aren't used:
\begin{codeenv}
\cmd{[}\cmd{set}\marg{A} = \cs{gls}\marg{bigcup}\idx{sbchar}\marg{i=1}\idx{spchar}n \cmd{set}\marg{B}\idx{sbchar}i \cmd{]}
\end{codeenv}
This produces:
\[ \mathcal{A} = \bigcup{}_{i=1}^n \mathcal{B}{}_i \]

Be careful with the set cardinality example. You might be tempted to
nest \csfmt{set} within the argument of \csfmt{setcard} but this
results in nested hyperlinks. These are unpredictable and there's no
consistent handling of them between different PDF viewers. It can
also be confusing to the reader. If
$\lvert\mathcal{B}_1\cup\mathcal{B}_2\rvert$
shows up as what appears to be a single hyperlink, where would the 
reader expect the target? This is the reason for providing the
non-linking commands like \csfmt{nlset} and \csfmt{nlsetcard}.

The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-maths
bib2gls sample-maths
pdflatex sample-maths
\end{verbatim}
The resulting document is shown in \figureref{fig:sample-maths.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-maths.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-maths.pdf}}%
}
{\caption{\filefmt{sample-maths.pdf}}}
{fig:sample-maths.pdf}
\end{figure}

\filesection{sample-textsymbols.tex}

This example uses \exfile{miscsymbols.bib}. 
This requires both \isty{marvosym} and (with the \styoptfmt{weather} option)
\isty{ifsym}. Unfortunately both define the commands \csfmt{Sun}
and \csfmt{Lightning}, so these commands need to be undefined after
the first package is loaded and before the second.
Since I want the definitions provide by \sty{ifsym} I have to first
load \sty{marvosym}, then undefine the conflicting commands and then
load \sty{ifsym}:
\begin{codeenv}
\cs{usepackage}\marg{etoolbox}
\cs{usepackage}\marg{marvosym}
\cs{undef}\cmd{Sun}
\cs{undef}\cmd{Lightning}
\cs{usepackage}\oarg{\styoptfmt{weather}}\marg{ifsym}
\end{codeenv}
The \sty{etoolbox} package is also loaded as it provides
\ics{undef}. (An alternative is to modify the
\exfile{miscsymbols.bib} file so that it uses \sty{ifsym}'s more
generic \cs{textweathersymbol} command and omit the \styoptfmt{weather}
option when loading the package, but the method used here
demonstrates how to deal with such conflicts.)

The custom entry type \atentryfmt{icon} must be aliased for the
entries to be recognised:
\begin{codeenv}
\csopt[icon=symbol]{entry-type-aliases}
\end{codeenv}

Since none of the entries have a \field{name} or \field{description}
field, the custom fields \fieldfmt{icon} and
\fieldfmt{icondescription} need to be aliased to them.
The document uses the \glostyle{alttreegroup} style where the groups
are obtained from the \field{category}, which again I obtain from
the custom \fieldfmt{identifier} field using:
\begin{codeenv}
\csopt[
 \fieldfmt{identifier}=\field{category},
 \fieldfmt{icon}=\field{name},
 \fieldfmt{icondescription}=\field{description}]{field-aliases},
\csopt[\field{category}=\field{group}]{replicate-fields}
\end{codeenv}
The \field{group} field is just a label and an appropriate title needs to be
supplied for each group label:
\begin{codeenv}
\cs{glsxtrsetgrouptitle}\marg{information}\marg{Information}
\cs{glsxtrsetgrouptitle}\marg{mediacontrol}\marg{Media Controls}
\cs{glsxtrsetgrouptitle}\marg{weather}\marg{Weather Symbols}
\end{codeenv}
This also requires sorting first by \field{category} and then
fallback on another field. The most appropriate here is the
\field{description} field, but instead of using
\csopt{identical-sort-action}, I'm using \csopt{sort-suffix},
which works better with the default locale sort when the fallback
field consists of words or phrases.
\begin{codeenv}
\csopt[\field{category}]{sort-field},
\csopt[\field{description}]{sort-suffix},
\csopt[|]{sort-suffix-marker}
\end{codeenv}
Since I'm using one of the \glostyle{alttree} styles, I need to set
the widest name:
\begin{codeenv}
\csopt{set-widest}
\end{codeenv}
In this case, \bibgls\ won't be able to determine the widest name
since it doesn't recognise any of the commands, so it will have to
use the fallback command, which will use one of the commands
provided by the \sty{glossaries-extra-stylemods} package.

This is actually not the best method as \bibgls\ can't see the group
titles as they're in the document, so it's only able to sort by the
label. While this might work for English, it can become a problem
for other languages that use extended Latin or non-Latin characters
in their alphabet. A much better method is to treat this as a
hierarchical glossary with topic titles as the top-level entries.
This is covered in the next example \exfile{sample-textsymbols2.tex}.

The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-textsymbols
bib2gls sample-textsymbols
pdflatex sample-textsymbols
\end{verbatim}
The resulting document is shown in \figureref{fig:sample-textsymbols.pdf}.

\lstinputlisting[firstline=6]{../examples/sample-textsymbols.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-textsymbols.pdf}}%
}
{\caption{\filefmt{sample-textsymbols.pdf}}}
{fig:sample-textsymbols.pdf}
\end{figure}

\filesection{sample-textsymbols2.tex}

This example is a better approach than the 
\exfile{sample-textsymbols.tex} example above. As with the previous
example, this requires both \isty{marvosym} and \isty{ifsym} so the
same patch is applied to avoid conflict.

As before, the custom entry type \atentryfmt{icon} must be aliased for the
entries to be recognised:
\begin{codeenv}
\csopt[icon=symbol]{entry-type-aliases}
\end{codeenv}

The \exfile{topics.bib} file contains terms with
labels that match the custom \fieldfmt{identifier} fields used in
the \exfile{miscsymbols.bib} file. So both files are loaded and the
\fieldfmt{identifier} field is now aliased to \field{parent}. These
parent entries represent the topics and unlike the previous example
it's now possible to sort by the topic title (obtained from the
\field{name} field) instead of by the label.
\begin{codeenv}
\csopt[topics,miscsymbols]{src},
\csopt[
 \fieldfmt{identifier}=\field{parent},
 \fieldfmt{icon}=\field{name},
 \fieldfmt{icondescription}=\field{description}]{field-aliases},
\end{codeenv}
There's no \csopt{sort-field} option in this example. The default
\field{sort} field is used. Since it's not set for any of the
entries, the fallback value will be used. In the case of the topic
titles (\atentry{index} and \atentry{indexplural}), I want to sort
by the \field{name}, which is the default fallback if the
\field{sort} field is missing for the index entry types.

The default fallback for the \field{sort} field for \atentry{symbol}
entries is the label, but in this case I want to use the
\field{description} field:
\begin{codeenv}
  \csopt[description]{symbol-sort-fallback}
\end{codeenv}
The best styles for this kind of glossary are the topic styles
provided by \isty{glossary-topic}. This package was only added to
\sty{glossaries-extra} v1.40, so you need to make sure you have at
least that version installed.

In this case I've decided to use the \glostyle{topic} style. I can
use it with or without the \csopt{set-widest} option. 
As with the previous example, \bibgls\ won't be able to determine
the widest name since it doesn't recognise any of the commands
contained in the \field{name} fields, so it will have to use the
fallback method, which will use one of the commands provided by the
\sty{glossaries-extra-stylemods} package. The \code{tree} option is
needed to enable the appropriate commands:
\begin{codeenv}
\cmd{usepackage}\oarg{\styopt{record},
 \styopt{nostyles},
 \styopt{postdot},
 \styopt[tree,topic]{stylemods},
 \styopt[topic]{style}}\marg{glossaries-extra}
\end{codeenv}

The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-textsymbols2
bib2gls --group sample-textsymbols2
pdflatex sample-textsymbols2
\end{verbatim}
The resulting document is shown in \figureref{fig:sample-textsymbols2.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-textsymbols2.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-textsymbols2.pdf}}%
}
{\caption{\filefmt{sample-textsymbols2.pdf}}}
{fig:sample-textsymbols2.pdf}
\end{figure}

\filesection{sample-markuplanguages.tex}

This example uses \exfile{markuplanguages.bib}. Since the file
includes abbreviations, any commands that must be used before
abbreviations are defined need to go before
\gls!{GlsXtrLoadResources}. This includes the abbreviation style, 
which I've set to \abbrstyle{long-short-desc}:
\begin{codeenv}
\cs{setabbreviationstyle}\oarg{markuplanguage}\marg{\abbrstyle{long-short-desc}}
\end{codeenv}
This style sets the \field{name} field using
\ics{glsxtrlongshortdescname}, which defaults to the long form
followed by the short form in parentheses. I decided to switch this
round so that the short form is shown first, which conveniently
matches the default \csopt{abbreviation-sort-fallback}.
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrlongshortdescname}}\marg{\comment{}
  \cs{protect}\ics{glsabbrvfont}\marg{\ics{the}\ics{glsshorttok}}\cs{space}
  \ics{glsxtrparen}\marg{\ics{glslongfont}\marg{\cs{the}\ics{glslongtok}}}\comment{}
}
\end{codeenv}
(The long form is still shown before the short form on the first use
of \cs{gls} in the document. The switch in the above code only
affects how the term is displayed in the glossary.)

This redefinition must be done before the abbreviations are defined
as it's expanded when the \field{name} field is set. (Note the need
to protect commands that shouldn't be expanded.) If I decide not to change
the \field{name} format in this way, I would then need to use
\csopt[long]{abbreviation-sort-fallback}.

I also decided to make use of the custom command \csfmt{abbrvtag}
that marks up the letters in the \field{long} field used to obtain
the abbreviation. As with the abbreviation style, this must be
done before the abbreviations are defined:
\begin{codeenv}
\cs{GlsXtrEnableInitialTagging}\marg{markuplanguage}\marg{\cmd{abbrvtag}}
\end{codeenv}
If you accidentally place it after \gls{GlsXtrLoadResources}, you'll
encounter an error on the second \LaTeX\ run (but not the first).
This is because \ics{GlsXtrEnableInitialTagging} requires that the
supplied command (\csfmt{abbrvtag} in this case) be undefined. On
the first \LaTeX\ it's undefined, but on the second it picks up
the \atentry{preamble} definition, which is now in the resource file.

The tagging format is governed by \ics{glsxtrtagfont} which
underlines its argument by default. I've redefined it to also
convert the letter to \idx{uppercase}:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrtagfont}}[1]\marg{\ics{underline}\marg{\cs{MakeTextUppercase}\marg{\idx{param}1}}}
\end{codeenv}
Note that in the \code{mathml} case, the first tag consists of more
than one letter:
\begin{codeenv}
\field{long}=\marg{\cmd{abbrvtag}\marg{m\cs{NoCaseChange}\marg{ath}}ematical }\idx{stringconcat}markuplang
\end{codeenv}
Here \ics{NoCaseChange} prevents \ics{MakeTextUppercase}
from applying the case change.

The default \csopt{selection} criteria includes entries that have
been indexed and any cross-references. Some of the
\field{description} fields include \ics{glsxtrshort}, which \bibgls\ picks
up and the referenced entry is included in the dependency list.
However, I don't want any indexing
performed by commands occurring in the glossary. This can be dealt
with in one of two ways: either switch the format to \encap{glsignore}
or suppress the indexing by changing the default options with
\ics{GlsXtrSetDefaultGlsOpts}. In this case I decided to turn the
records into ignored records:
\begin{codeenv}
\ics{GlsXtrSetDefaultNumberFormat}\marg{\encap{glsignore}}
\end{codeenv}
This means that some of the entries won't have location lists, so
I've defined a \idx{postdescriptionhook} that inserts a \idx{full-stop} after the
\field{description} if there's no location otherwise it inserts a
comma:
\begin{codeenv}
\cs{newcommand}\marg{\postdeschook{markuplanguage}}\marg{\comment{}
  \cs{glsxtrifhasfield}\marg{\field{location}}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{,}\comment{}
  \marg{.}\comment{}
}
\end{codeenv}
I've used \csopt{loc-suffix} to append a \idx{full-stop} after the
location lists. This doesn't affect the entries that haven't been
indexed.

I decided to convert the first letter of the \field{name} field to
\idx{uppercase}. Since the \field{name} is implicitly set for
abbreviations based on the style, I've decided to implement this
through the \catattr{glossname} attribute rather than using
\csopt{name-case-change}:
\begin{codeenv}
\cs{glssetcategoryattribute}\marg{markuplanguage}\marg{\catattr{glossname}}\marg{firstuc}
\end{codeenv}
If this line causes an error when the glossary is displayed that goes 
away if it's commented out, make sure you have at least version 2.06
of \sty{mfirstuc}. For most of the entries, this doesn't make a
difference as they already start with a capital. It's only the 
\code{markdown} entry that's actually affected.

The description case change is dealt with by \bibgls\ instead:
\begin{codeenv}
\csopt[firstuc]{description-case-change}
\end{codeenv}
This works better than the \catattr{glossdesc} attribute as \bibgls\ can
convert commands like \ics{glstext} into \ics{Glstext} which
\ics{makefirstuc} can't do. (Although in this particular example,
there's no difference as both instances of \cs{glstext} already
produce \idx{uppercase} text.)

The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-markuplanguages
bib2gls --group sample-markuplanguages
pdflatex sample-markuplanguages
\end{verbatim}
The resulting document is shown in \figureref{fig:sample-markuplanguages.pdf}.

\lstinputlisting[firstline=5]{../examples/sample-markuplanguages.tex}

\begin{figure}
\figcontents
{%
 \frame{\includegraphics[height=.9\textheight]{../examples/sample-markuplanguages.pdf}}%
}
{\caption{\filefmt{sample-markuplanguages.pdf}}}
{fig:sample-markuplanguages.pdf}
\end{figure}


\filesection{sample-usergroups.tex}

This example uses \exfile{usergroups.bib}. This requires \XeLaTeX\
or \LuaLaTeX\ as the \ext{bib} file includes \idx{non-ASCII} labels.
The entries include fields in different languages, the main one
being English. If an entry has a non-English \field{name} or
\field{long} field, it also includes the custom field
\fieldfmt{translation} that provides an (approximate) translation.
If this field is present, the language is given by the first element
of the custom \fieldfmt{language} field.

In this case, I'm providing keys for the custom
\fieldfmt{language} and \fieldfmt{translation} fields, and, for a bit
of variety from the other examples, I'm ignoring the custom
\fieldfmt{identifier} field. The custom keys are provided with
\ics{glsaddstoragekey}:
\begin{codeenv}
\cs{glsaddstoragekey}\marg{language}\marg{}\marg{\cmd{glsentrylanguage}}
\cs{glsaddstoragekey}\marg{translation}\marg{}\marg{\cmd{glsentrytranslation}}
\end{codeenv}
The \ext{bib} file includes abbreviations. Remember that the
abbreviation style must be set before the resource file is loaded:
\begin{codeenv}
\cs{setabbreviationstyle}\oarg{tug}\marg{\abbrstyle{long-short-user}}
\end{codeenv}
For this example, I'm explicitly setting the \field{category} field
to \code{tug}:
\begin{codeenv}
\csopt[tug]{category}
\end{codeenv}
Some of the fields end with a \idx{full-stop}. This isn't a problem with
the \field{long} field as the first use follows the long form with
the short form in parentheses, but it will be a problem on
subsequent use if the \field{short} field ends with a \idx{full-stop}.
This means I need to check for end-of-sentence punctuation
for the \field{short} field. It's also a good idea to do this for
the \field{name} field for the non-abbreviations.
\begin{codeenv}
\csopt[\field{name},\field{short}]{check-end-punctuation}
\end{codeenv}
It's now possible to discard a \idx{full-stop} that follows
\ics{gls}:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrifcustomdiscardperiod}}[2]\marg{\comment{}
 \ics{ifglshasshort}\marg{\cs{glslabel}}\comment{}
 \marg{\comment{}
   \cs{glsxtrifwasfirstuse}\marg{}\comment{}
   \marg{\comment{}
     \cs{GlsXtrIfFieldUndef}\marg{\fielddisp{fieldendpunc}{shortendpunc}}\marg{\cs{glslabel}}\marg{\idx{param}2}\marg{\idx{param}1}\comment{}
   }\comment{}
 }\comment{}
 \marg{\comment{}
   \cs{GlsXtrIfFieldUndef}\marg{\fielddisp{fieldendpunc}{nameendpunc}}\marg{\cs{glslabel}}\marg{\idx{param}2}\marg{\idx{param}1}\comment{}
 }\comment{}
}
\end{codeenv}
This first tests if the entry that's just been referenced has a
\field{short} field. If it has, then the next test is to check if
that was the first use for that entry. If it was, nothing is done.
If it wasn't, then \cs{GlsXtrIfFieldUndef} is used to determine
if \fielddisp{fieldendpunc}{shortendpunc} has been set. If it has
been set then the period discard function is performed. If the
entry doesn't have a \field{short} field, then the
\fielddisp{fieldendpunc}{nameendpunc} field needs checking instead.

Since the document requires \XeLaTeX\ or \LuaLaTeX\ and has some
\idx{non-ASCII} characters, it needs \sty{fontspec} and an
appropriate font. In this case I've chosen \qt{Linux Libertine O}.
If you don't have it installed, you'll need to change it.
\begin{verbatim}
\usepackage{fontspec}
\setmainfont{Linux Libertine O}
\end{verbatim}
Since it's a multilingual document I also need \isty{polyglossia}
with the main language set to \code{english}:
\begin{codeenv}
\cmd{usepackage}\marg{polyglossia}
\ics{setmainlanguage}\oarg{variant=uk}\marg{english}
\end{codeenv}
Now comes the difficult bit. The document needs to determine what
other languages need to be loaded. The \sty{tracklang} package
provides a convenient interface when dealing with language tags.
This is automatically loaded by \styfmt{glossaries} but I've loaded
it here explicitly as a reminder:
\begin{verbatim}
\usepackage{tracklang}
\end{verbatim}
Once the \hyperref[sec:resourcesets]{resource file} has been loaded, I need to
iterate over all the defined entries and check if the \fieldfmt{translation}
field has been set. If it has, then the first language tag in the
\fieldfmt{language} field will supply the language, but this needs to be
converted from the \idx{IETF} language tag to a language name recognised by
\isty{polyglossia}.

Iterating over all entries can be done with \ics{forglsentries}
but remember that no entries will be defined before \bibgls\ has
been run, so this does nothing on the first \LaTeX\ run.
\begin{codeenv}
\cs{forglsentries}\marg{\cmd{thislabel}}\marg{\comment{}
  \cs{glsxtrifhasfield}\marg{translation}\marg{\cmd{thislabel}}\comment{}
  \marg{\comment{}
    \comment{requires glossaries-extra v1.24}
    \cs{glsxtrforcsvfield}\marg{\cmd{thislabel}}\marg{language}\marg{\cmd{addfirstlang}}\comment{}
  }\comment{}
  \marg{}\comment{}
}
\end{codeenv}
Within the outer (\cs{forglsentries}) \idx{loop}, there's a check
for the \fieldfmt{translation} field using \ics{glsxtrifhasfield}.
If it's present, then the first element of the \fieldfmt{language}
field is required. The simplest way to get this is to use
\ics{glsxtrforcsvfield} which iterates over all elements of the
given field (\fieldfmt{language} in this case) and break out of the
\idx{loop} (with \ics{glsxtrendfor}) once the language has been found.

The \idx{handler} function (\csfmt{addfirstlang}) is defined so that it
adds the given language tag as a tracked language using
\ics{TrackLocale}. This command sets
\ics{TrackLangLastTrackedDialect} to the associated
(\sty{tracklang}) dialect label for convenience. This dialect label
can then be converted to the root language label using
\ics{TrackedLanguageFromDialect}. If this language is supported
by \isty{polyglossia}, then there should be a file called
\metafilefmt{gloss-}{language}{.ldf}.

Some of the entries use the same language, so it's necessary to
check if the language has already been defined before loading it.
There's also a problem in that the language file should not be
loaded in a scoped context, but both \ics{glsxtrforcsvfield}
and the unstarred \ics{glsxtrifhasfield} add implicit grouping.
To solve both problems, an internal \sty{etoolbox} list is 
defined:
\begin{verbatim}
\newcommand{\langlist}{}%
\end{verbatim}
and \ics{xifinlist} is used to first check if the language label is
already in the list before adding it. Since this part of the code is
scoped, the global \ics{listxadd} is used to add the language label
to the list.

Next the \field{useri} field is set to 
\encapdisp{textlanguage}{text\meta{language}}
which is the name of the control sequence used with
\sty{polyglossia} to switch language for a short block of text. This means that
\gls{glsxtrentryfmt}\margm{text} can be used to format \meta{text}
in the relevant language. Finally, \ics{glsxtrendfor} is used to
break out of the \idx{loop}.
\begin{codeenv}
\cs{newcommand}*\marg{\cmd{addfirstlang}}[1]\marg{\comment{}
  \cs{TrackLocale}\marg{\idx{param}1}\comment{}
  \ics{edef}\cmd{thislanguage}\marg{\comment{}
    \cs{TrackedLanguageFromDialect}\cs{TrackLangLastTrackedDialect}}\comment{}
  \ics{IfFileExists}\marg{gloss-\cmd{thislanguage}.ldf}\comment{}
  \marg{\comment{}
    \ics{xifinlist}\marg{\cmd{thislanguage}}\marg{\cmd{langlist}}\marg{}\comment{}
    \marg{\ics{listxadd}\marg{\cmd{langlist}}\marg{\cmd{thislanguage}}}\comment{}
    \ics{xGlsXtrSetField}\marg{\cmd{thislabel}}\marg{\field{useri}}\marg{\encapdisp{textlanguage}{text\cmd{thislanguage}}}\comment{}
    \ics{glsxtrendfor}
  }\comment{}
  \marg{}\comment{}
}
\end{codeenv}
Once the \ics{forglsentries} \idx{loop} has found the appropriate
languages, it's now necessary to iterate over the internal list
\csfmt{langlist} and set the language:
\begin{codeenv}
\ics{forlistloop}\marg{\ics{setotherlanguage}}\marg{\cmd{langlist}}
\end{codeenv}
The \abbrstyle{long-short-user} style now needs to be adjusted
to ensure that it picks up the appropriate language change.
By default this style checks the \field{useri} field, so this needs
to be changed to \fieldfmt{translation} by redefining
\ics{glsxtruserfield}:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtruserfield}}\marg{translation}
\end{codeenv}
The command that governs the format of the parenthetical material
(\ics{glsxtruserparen}) also needs adjusting. I've changed the
space before the parenthesis to \ics{cs.space} because some of the
long fields end with a \idx{full-stop} and this corrects the
spacing. The \fieldfmt{translation} field is in English, 
so this needs to be encapsulated with \icsdisp{textlanguage}{textenglish}
in case the surrounding text is in a different language.
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtruserparen}}[2]\marg{\comment{}
  \cs{cs.space}
  \cs{glsxtrparen}\marg{\idx{param}1\comment{}
  \cs{ifglshasfield}\marg{\cs{glsxtruserfield}}\marg{\idx{param}2}\marg{,
   \csdisp{textlanguage}{textenglish}\marg{\cs{glscurrentfieldvalue}}}\marg{}}\comment{}
}
\end{codeenv}
Next I've defined a convenient command for use in the \catattr{textformat}
attributes for the custom \code{tug} category:
\begin{codeenv}
\cs{newcommand}*\marg{\cmd{tugtextformat}}[1]\marg{\comment{}
  \cs{glsxtrentryfmt}\marg{\cs{glslabel}}\marg{\idx{param}1}\comment{}
}
\end{codeenv}
This uses \gls{glsxtrentryfmt} to encapsulate the given text
in the appropriate language command (if provided). When this is set
as the \catattr{textformat} attribute, it will be used instead of
\ics{glstextformat}, which means that the entry label can be
referenced with \cs{glslabel}.

There's a similar command for use in the \catattr{glossnamefont}
attribute. This is used in the glossary, so the label is referenced
with \cs{glscurrententrylabel}:
\begin{codeenv}
\cs{newcommand}*\marg{\cmd{tugnameformat}}[1]\marg{\comment{}
  \cs{glsxtrentryfmt}\marg{\cs{glscurrententrylabel}}\marg{\idx{param}1}\comment{}
}
\end{codeenv}
The attributes can now be set to the relevant control sequence name:
\begin{codeenv}
\cs{glssetcategoryattribute}\marg{tug}\marg{\catattr{textformat}}\marg{tugtextformat}
\cs{glssetcategoryattribute}\marg{tug}\marg{\catattr{glossnamefont}}\marg{tugnameformat}
\end{codeenv}

The document uses the \glostyle{bookindex} style, which is set in
the package options:
\begin{codeenv}
\cmd{usepackage}\oarg{\styopt{record},
 \styopt{nostyles},
 \styopt[bookindex]{stylemods},
 \styopt[bookindex]{style}
}\marg{glossaries-extra}
\end{codeenv}
The \glostyle{bookindex} style ignores the \field{description} field, so I've 
provided a \idx{postnamehook} to append it in parentheses (with the
translation, if provided):
\begin{codeenv}
\cs{newcommand}\marg{\postnamehook{tug}}\marg{\comment{}
  \ics{ifglshasdesc}\marg{\cs{glscurrententrylabel}}\comment{}
  \marg{\cs{cs.space}(\cs{glossentrydesc}\marg{\cs{glscurrententrylabel}}\comment{}
     \cs{glsxtrifhasfield}\marg{translation}\marg{\cs{glscurrententrylabel}}\comment{}
     \marg{, \csdisp{textlanguage}{textenglish}\marg{\cs{glscurrentfieldvalue}}}\comment{}
     \marg{}\comment{}
  )}\comment{}
  \marg{\comment{}
   \cs{glsxtrifhasfield}\marg{translation}\marg{\cs{glscurrententrylabel}}\comment{}
   \marg{\cs{cs.space}(\csdisp{textlanguage}{textenglish}\marg{\cs{glscurrentfieldvalue}})}\comment{}
   \marg{}\comment{}
  }\comment{}
}
\end{codeenv}
Remember that this hook is included within the \field{name} font
(provided by the \catattr{glossnamefont} attribute in this case)
so \csdisp{textlanguage}{textenglish} is again used to switch the language to
English for the translation.

The complete document code is listed below. The document build is:
\begin{verbatim}
xelatex sample-usergroups
bib2gls --group sample-usergroups
xelatex sample-usergroups
xelatex sample-usergroups
\end{verbatim}
The two pages of the document are shown in
\figureref{fig:sample-usergroups.pdf}. Since the entries have all
been referenced on page~1, the location lists are all simply~\qt{1}.

%\lstinputlisting[firstline=5]{../examples/sample-usergroups.tex}
\begin{lstlisting}[escapechar=|]
\documentclass{scrreprt}

\usepackage{fontspec}
\setmainfont{Linux Libertine O}

\usepackage{polyglossia}
\setmainlanguage[variant=uk]{english}
\usepackage{tracklang}
\usepackage{etoolbox}

\usepackage[record,% use bib2gls
 nostyles,% don't load default styles
 stylemods={bookindex},
 style={bookindex}
]{glossaries-extra}

\glsaddstoragekey{language}{}{\glsentrylanguage}
\glsaddstoragekey{translation}{}{\glsentrytranslation}

\setabbreviationstyle[tug]{long-short-user}

\GlsXtrLoadResources[
  src={usergroups}, % data in usergroups.bib
  check-end-punctuation={name,short},
  category=tug
]

\renewcommand*{\glsxtrifcustomdiscardperiod}[2]{%
 \ifglshasshort{\glslabel}%
 {%
   \glsxtrifwasfirstuse{}%
   {%
     \GlsXtrIfFieldUndef{shortendpunc}{\glslabel}{#2}{#1}%
   }%
 }%
 {%
   \GlsXtrIfFieldUndef{nameendpunc}{\glslabel}{#2}{#1}%
 }%
}

\newcommand{\langlist}{}%

\newcommand*{\addfirstlang}[1]{%
  \TrackLocale{#1}%
  \edef\thislanguage{%
    \TrackedLanguageFromDialect\TrackLangLastTrackedDialect}%
  \IfFileExists{gloss-\thislanguage.ldf}%
  {%
    \xifinlist{\thislanguage}{\langlist}{}%
    {\listxadd{\langlist}{\thislanguage}}%
    \xGlsXtrSetField{\thislabel}{useri}{text\thislanguage}%
    \glsxtrendfor
  }%
  {}%
}

\forglsentries{\thislabel}{%
  \glsxtrifhasfield{translation}{\thislabel}%
  {%
    % requires glossaries-extra v1.24
    \glsxtrforcsvfield{\thislabel}{language}{\addfirstlang}%
  }%
  {}%
}

\forlistloop{\setotherlanguage}{\langlist}

\renewcommand*{\glsxtruserfield}{translation}

\renewcommand*{\glsxtruserparen}[2]{%
  \  
  \glsxtrparen{#1%
  \ifglshasfield{\glsxtruserfield}{#2}{,
   \textenglish{\glscurrentfieldvalue}}{}}%
}

\newcommand*{\tugtextformat}[1]{%
  \glsxtrentryfmt{\glslabel}{#1}%
}

\newcommand*{\tugnameformat}[1]{%
  \glsxtrentryfmt{\glscurrententrylabel}{#1}%
}

\glssetcategoryattribute{tug}{textformat}{tugtextformat}
\glssetcategoryattribute{tug}{glossnamefont}{tugnameformat}

\newcommand{\glsxtrpostnametug}{%
  \ifglshasdesc{\glscurrententrylabel}%
  {\ (\glossentrydesc{\glscurrententrylabel}%
     \glsxtrifhasfield{translation}{\glscurrententrylabel}%
     {, \textenglish{\glscurrentfieldvalue}}%
     {}%
  )}%
  {%
   \glsxtrifhasfield{translation}{\glscurrententrylabel}%
   {\ (\textenglish{\glscurrentfieldvalue})}%
   {}%
  }%
}

\begin{document}
\chapter{Sample}
\section{First Use}
\gls{TUG}. \gls{bgTeX}. \gls{latex-br}. \gls{CTeX}.
\gls{CSTUG}. \gls{DANTE}. \gls{DKTUG}. \gls{EUG}.
\gls{CervanTeX}. \gls{TirantloTeX}. \gls{GUTenberg}.
\gls{UKTUG}. \gls{|\greekmono ɛϕτ|}. \gls{MaTeX}. \gls{ITALIC}.
\gls{ÍsTeX}. \gls{GuIT}. \gls{KTS}. \gls{LTVG}.
\gls{mxTeX}. \gls{NTG}. \gls{NTUG}. \gls{GUST}. \gls{GUTpt}.
\gls{VietTUG}. \gls{LUGSA}.

\section{Next Use}

\gls{TUG}. \gls{bgTeX}. \gls{latex-br}. \gls{CTeX}.
\gls{CSTUG}. \gls{DANTE}. \gls{DKTUG}. \gls{EUG}.
\gls{CervanTeX}. \gls{TirantloTeX}. \gls{GUTenberg}.
\gls{UKTUG}. \gls{|\greekmono ɛϕτ|}. \gls{MaTeX}. \gls{ITALIC}.
\gls{ÍsTeX}. \gls{GuIT}. \gls{KTS}. \gls{LTVG}.
\gls{mxTeX}. \gls{NTG}. \gls{NTUG}. \gls{GUST}. \gls{GUTpt}.
\gls{VietTUG}. \gls{LUGSA}.

\printunsrtglossaries
\end{document}
\end{lstlisting}

\begin{figure}
\figcontents
{\pdftwocol{../examples/sample-usergroups.pdf}{2}}
{\caption{\filefmt{sample-usergroups.pdf}}}
{fig:sample-usergroups.pdf}
\end{figure}

\filesection{sample-multi1.tex}

This example uses \exfile{bacteria.bib},
\exfile{markuplanguages.bib}, \exfile{vegetables.bib}, \exfile{minerals.bib},
\exfile{animals.bib}, \exfile{chemicalformula.bib},
\exfile{baseunits.bib} and \exfile{derivedunits.bib}.
Since there's one or more UTF-8 character, the document requires
UTF-8 support:
\begin{verbatim}
\usepackage[T1]{fontenc}
\usepackage[utf8]{inputenc}
\end{verbatim}

The aim of this example document is to have a separate glossary
(without number lists) for each type of data (bacteria, markup
languages, vegetables, minerals, animals, chemical formula, base
units and derived units) and also an
index listing all referenced entries with number lists as well as
aliased entries that haven't explicitly been used but the
cross-reference term as been indexed. This requires:
\begin{codeenv}
\csopt[recorded and deps and see]{selection}
\end{codeenv}
to ensure the aliased entries are selected.

Since I don't need the default \code{main} glossary (I'm providing
my own custom glossaries) I've used the \styopt{nomain} option to
suppress its automatic creation, but I do want the \code{index}
glossary so I've used the \styopt{index} package option. As with the
other examples, I've used \styopt{nostyles} to suppress the creation
of the default styles and used \styopt{stylemods} to load the
particular style packages that I need and use
\sty{glossaries-extra-stylemods} to patch them. The index needs to
be in an unnumbered chapter, which is the default for book-like
styles, but I want the other glossaries in unnumbered sections so
I've used the \styopt{section} option. I just need to remember to
switch this before displaying the index:
\begin{codeenv}
\cmd{usepackage}\oarg{\styopt{record},\comment{use bib2gls}
 \styopt{section},\comment{use \ics{section*} for glossary headings}
 \styopt{postdot},\comment{insert dot after descriptions in glossaries}
 \styopt{nomain},\comment{don't create 'main' glossary}
 \styopt{index},\comment{create 'index' glossary}
 \styopt{nostyles},\comment{don't load default styles}
\comment{load and patch required style packages:}
 \styopt[list,mcols,tree,bookindex]{stylemods}
}\marg{glossaries-extra}
\end{codeenv}
The remaining glossaries need defining:
\begin{codeenv}
\cs{newglossary*}\marg{bacteria}\marg{Bacteria}
\cs{newglossary*}\marg{markuplanguage}\marg{Markup Languages}
\cs{newglossary*}\marg{vegetable}\marg{Vegetables}
\cs{newglossary*}\marg{mineral}\marg{Minerals}
\cs{newglossary*}\marg{animal}\marg{Animals}
\cs{newglossary*}\marg{chemical}\marg{Chemical Formula}
\cs{newglossary*}\marg{baseunit}\marg{SI Units}
\cs{newglossary*}\marg{derivedunit}\marg{Derived Units}
\end{codeenv}
As with \exfile{sample-bacteria.tex} and \exfile{sample-markuplanguages.tex}
I need to set the abbreviation styles before the abbreviations are
defined:
\begin{codeenv}
\cs{setabbreviationstyle}\oarg{bacteria}\marg{\abbrstyle{long-only-short-only}}
\cs{setabbreviationstyle}\oarg{markuplanguage}\marg{\abbrstyle{long-short-desc}}
\end{codeenv}
Unlike the \exfile{sample-markuplanguages.tex} example, I'm not interested
in tagging the initials in this case, but I still want to change the
way the \field{name} field is set with the
\abbrstyle{long-short-desc} abbreviation style:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrlongshortdescname}}\marg{\comment{}
  \cs{protect}\cs{glsabbrvfont}\marg{\cs{the}\cs{glsshorttok}}\cs{space}
  \cs{glsxtrparen}\marg{\cs{glslongfont}\marg{\cs{the}\cs{glslongtok}}}\comment{}
}
\end{codeenv}
Remember that this also needs to be set before the abbreviations are
defined. The \catattr{textformat} and \catattr{glossnamefont}
attributes may be set after definition:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{bacteriafont}}[1]\marg{\cs{emph}\marg{\idx{param}1}}
\cs{glssetcategoryattribute}\marg{bacteria}\marg{\catattr{textformat}}\marg{bacteriafont}
\cs{glssetcategoryattribute}\marg{bacteria}\marg{\catattr{glossnamefont}}\marg{bacteriafont}
\end{codeenv}
The description font also needs to be set since this will contain
the long form:
\begin{codeenv}
\cs{glssetcategoryattribute}\marg{bacteria}\marg{\catattr{glossdescfont}}\marg{bacteriafont}
\end{codeenv}

The \code{markuplanguage} glossary contains descriptions and some
long names, so it's better suited to the \glostyle{altlist} style,
in which case the descriptions would look better if they started
with a capital letter:
\begin{codeenv}
\cs{glssetcategoryattribute}\marg{markuplanguage}\marg{\catattr{glossdesc}}\marg{firstuc}
\end{codeenv}
Remember that the \glostyle{altlist} style uses the
\env{description} environment, which is governed by the document
class (and may be modified by list-related packages). In this case,
one of the KOMA-Script classes is used, so the list items are
typeset in sans-serif.

There are various ways of dealing with the duplicated data in the
index, such as using the \csopt{secondary} option or having a
separate resource set with a copy \csopt{action}. In this case, I've
decided to use a dual entry system. Since the entries aren't defined
using any dual types, I've used \csopt{entry-type-aliases} to
make \bibgls\ treat them as though they were, and I also need to
alias the custom \atentryfmt{chemical}, \atentryfmt{unit} and 
\atentryfmt{measurement} entry types:
\begin{codeenv}
\csopt[
  abbreviation=dualindexabbreviation,
  entry=dualindexentry,
  symbol=dualindexsymbol,
  unit=dualindexsymbol,
  measurement=dualindexsymbol,
  chemical=dualindexsymbol
 ]{entry-type-aliases}
\end{codeenv}
Note that I haven't aliased the \atentry{index} types as I only want
these in the index and not replicated in a separate glossary.

The primary entries for the \atentry{dualindexabbreviation} type
ignore the short form. It would be useful to store it. This could be
done by copying the \field{short} field with
\csopt{replicate-fields}. For example,
\csopt[\field{short}=\field{symbol}]{replicate-fields}. However, this will cause the
\field{symbol} field to be set for both the primary and dual
entries, which will cause an unwanted duplication if the dual
entries are displayed using a glossary style that shows the
\field{symbol} field. Another field (such as \field{user1}) could be
used instead or \gls{bibglsnewdualindexabbreviation} could be
defined before \gls{GlsXtrLoadResources}:
\begin{codeenv}
\cs{newcommand}\marg{\gls{bibglsnewdualindexabbreviation}}[7]\marg{\comment{}
 \cs{longnewglossaryentry*}\marg{\idx{param}1}\marg{\comment{}
  \field{name}=\marg{\cs{protect}\gls{bibglsuselongfont}\marg{\idx{param}4}\marg{\cs{glscategory}\marg{\idx{param}2}}},\comment{}
  \field{symbol}=\marg{\cs{protect}\gls{bibglsuseabbrvfont}\marg{\idx{param}5}\marg{\cs{glscategory}\marg{\idx{param}2}}},\comment{}
  \field{category}=\marg{index},\idx{param}3}\marg{}\comment{}
}
\end{codeenv}
However, this will affect all \atentry{dualindexabbreviation} entry
types, but it's not necessary for the bacteria abbreviations.
Instead it's simpler to just keep a record of the dual label so that
the short form can be obtained from the dual entry:
\begin{codeenv}
\csopt{dual-field}
\end{codeenv}

By default, the \atentry{dualindexabbreviation} entry type falls back on the
\field{short} field if the \field{name} is omitted. In this case I
want it to fall back on the \field{long} field instead.
\begin{codeenv}
\csopt[long]{abbreviation-name-fallback}
\end{codeenv}
Remember that the sort fallback for abbreviations is still \field{short} (but can be
changed with \csopt{abbreviation-sort-fallback}), but I've changed
the sort fallback for symbols:
\begin{codeenv}
\csopt[name]{symbol-sort-fallback}
\end{codeenv}
I also need to alias the custom fields (especially for those in the
\exfile{chemicalformula.bib}, \exfile{baseunits.bib} and
\exfile{derivedunits.bib} files):
\begin{codeenv}
\csopt[
  \fieldfmt{identifier}=\field{category},
  \fieldfmt{formula}=\field{symbol},
  \fieldfmt{chemicalname}=\field{name},
  \fieldfmt{unitname}=\field{name},
  \fieldfmt{unitsymbol}=\field{symbol},
  \fieldfmt{measurement}=\field{description}
]{field-aliases}
\end{codeenv}
There's a slight problem here. This ensures that the entries defined
in \exfile{chemicalformula.bib} have a \field{name} and
\field{symbol} field, which are swapped round for the dual
(according to the default \csopt{dual-indexsymbol-map}) but these
entries don't have a \field{description} field. Since I'd like to
use the \glostyle{mcolalttreegroup} style, this will end up with
the odd appearance of the formula (stored in the \field{name} field for the
dual) followed by the chemical name (stored in the \field{symbol}
field for the dual) in parenthesis. This is the default \meta{name}
(\meta{symbol}) \meta{description} format for the style. I've fixed
this by locally redefining \ics{glsxtralttreeSymbolDescLocation} for
just that glossary:
\begin{codeenv}
\cs{printunsrtglossary*}\oarg{\printglossopt[chemical]{type},\printglossopt[mcolalttreegroup]{style}}
\marg{\comment{}
  \cs{renewcommand}\cs{glsxtralttreeSymbolDescLocation}[2]\marg{\comment{}
   \ics{glossentrysymbol}\marg{\idx{param}1}\cs{glspostdescription}\ics{glsxtrAltTreePar}
  }\comment{}
  \cs{renewcommand}*\marg{\cs{glstreenamefmt}}[1]\marg{\idx{param}1}\comment{}
  \cs{renewcommand}*\marg{\cs{glstreegroupheaderfmt}}[1]\marg{\cs{textbf}\marg{\idx{param}1}}\comment{}
}
\end{codeenv}
I've also redefined \cs{glstreenamefmt} to prevent the names
appearing in bold, which means I also need to redefine
\cs{glstreegroupheaderfmt} to keep the headers bold.

All the \atentryfmt{dualindex\meta{type}} entry types provide a primary
entry that behaves like \atentry{index}. The secondary behaves like
\atentryfmt{\meta{type}}. This means that the primaries are
conveniently gathered together with all the unaliased \atentry{index}
entries, so the primary entry type needs to be set to \code{index}:
\begin{codeenv}
\csopt[index]{type}
\end{codeenv}
The dual entry type depends on the entry's category. Since I've
defined my custom glossaries with a label that matches the
custom \fieldfmt{identifier} field, I can both alias this custom field
to the \field{category} field and also set \csopt{dual-type} so that
it matches the category:
\begin{codeenv}
\csopt[\fieldfmt{identifier}=\field{category}]{field-aliases},
\csopt[same as category]{dual-type}
\end{codeenv}
The primary entries (in the \code{index} glossary) need to be sorted
alphabetically, and since the document is in English I'm sorting
according to that language (identified by the language code
\code{en}), but I also want to make sure that all the primary
entries are sorted by the \field{name} field to avoid discrepancies
in the fallback value for the \field{sort} field:
\begin{codeenv}
\csopt[en]{sort},
\csopt[name]{sort-field}
\end{codeenv}
With \csopt[long]{abbreviation-name-fallback} now set, 
this means that \emph{Coxiella burnetii} comes after
\emph{Clostridium tetani} in the index. I haven't changed the sort
field for the dual entries, so in that case the
\csopt{abbreviation-sort-fallback} and \csopt{symbol-sort-fallback}
settings will be used with the duals.  This means that
\emph{C.\ burnetii} is between \emph{C.\ botulinum} and \emph{C.\
perfringens} rather than after \emph{C.\ tetani}.

I'd like to sort the dual entries according to a letter-number rule
(as for the above \exfile{sample-chemical.tex} and \exfile{sample-units3.tex}
examples)
but this would order \qt{b\'ilinite} after \qt{biotite} in the
\code{minerals} glossary, so instead I'm also using the English sort
rule for the duals, but with the numbers padded:
\begin{codeenv}
\csopt[en]{dual-sort},
\csopt[2]{dual-sort-number-pad},
\end{codeenv}
This method doesn't work as well as the method used in
\exfile{sample-chemical.tex} as it doesn't separate the capitals,
digits and \idx!{lowercase} characters in the way that can be achieved with the
letter-number methods. An improvement can be made by changing the
break-points. I could use \csopt[upper-upper]{dual-break-at} but
this would put \qt{seal} before \qt{sea lion} in the \code{animal}
glossary, so instead I've used:
\begin{codeenv}
\csopt[upper-upper-word]{dual-break-at}
\end{codeenv}
This now puts \qt{sea lion} before \qt{seal}. Unfortunately the word
break points will cause a break at the markers used to indicate
positive and negative numbers that are inserted with
\csopt{dual-sort-number-pad}, so these need to be changed to
something that won't cause them to be discarded:
\begin{codeenv}
\csopt[0]{dual-sort-pad-minus},
\csopt[1]{dual-sort-pad-plus}
\end{codeenv}

The document loads \sty{hyperref} which means that all the \cs{gls}
references will create hyperlinks. Since the primaries are in the
index, the default prefixes mean that, for example, 
\code{\cs{gls}\marg{svg}} links to the \qt{scalable vector graphics} item in
the index rather than to the abbreviation \qt{SVG} in the
\code{markuplanguage} glossary. There are two alternatives: change
\code{\cs{gls}\marg{svg}} to \code{\cs{gls}\marg{dual.svg}} or change the 
default prefixes, which is the more convenient approach and is the one used
here:
\begin{codeenv}
\csopt[idx.]{label-prefix},
\csopt[\empty]{dual-prefix}
\end{codeenv}
Now \code{\cs{gls}\marg{svg}} refers to the dual abbreviation \qt{SVG} and
\code{\cs{gls}\marg{idx.svg}} refers to the primary entry \qt{scalable vector
graphics}. Unfortunately this means that the records created with
\code{\cs{gls}\marg{svg}} now refer to the dual abbreviation and will end up
being displayed in the glossary instead of the index. This can be
fixed with:
\begin{codeenv}
\csopt[primary]{combine-dual-locations}
\end{codeenv}
Which transfers the dual entry locations to the corresponding
primary.

The other problem is the cross-references in the \field{description}
fields. Since the labels don't start with \qt{\idprefix{dual}} \bibgls\
will assume they refer to the primary entries, which means that
\qt{\idprefixfmt{idx}} (the value of \csopt{label-prefix}) will be
inserted. This means that they'll link to the index rather than the
glossary entry. It also means that the cross-references where the
dual is an abbreviation won't behave like an abbreviation as the
reference is to the primary (non-abbreviation) entry. This can be
fixed by setting \csopt{cs-label-prefix} to the same value as
\csopt{dual-prefix}:
\begin{codeenv}
\csopt[\empty]{cs-label-prefix}
\end{codeenv}
The index is displayed using the \glostyle{bookindex} style. This
doesn't show the description or symbol by default, but it would be
useful to include the symbol in parentheses after the name. This can
be done by redefining \ics{glsxtrbookindexname}:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrbookindexname}}[1]\marg{\comment{}
  \cs{glossentryname}\marg{\idx{param}1}\comment{}
  \cs{ifglshassymbol}\marg{\idx{param}1}\marg{\cs{space}(\cs{glossentrysymbol}\marg{\idx{param}1})}\marg{}\comment{}
}
\end{codeenv}
However the chemical forumlae look a little odd in parentheses
(especially those that contain parenthetical parts) but this can be
fixed by adding a category check:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrbookindexname}}[1]\marg{\comment{}
  \cs{glossentryname}\marg{\idx{param}1}\comment{}
  \cs{ifglshassymbol}\marg{\idx{param}1}\comment{}
  \marg{\comment{}
    \cs{glsifcategory}\marg{\idx{param}1}\marg{chemical}\comment{}
    \marg{, \cs{glossentrysymbol}\marg{\idx{param}1}}\comment{}
    \marg{\cs{space}(\cs{glossentrysymbol}\marg{\idx{param}1})}\comment{}
  }\comment{}
  \marg{}\comment{}
}
\end{codeenv}
Unfortunately \ics{glossentrysymbol} doesn't pick up the
\catattr{glossnamefont} attribute, so if the short form of the
abbreviations is saved in the \field{symbol} field, using one of the
methods discussed above, then the custom \csfmt{bacteriafont} won't
be applied. (As from \sty{glossaries-extra} version 1.42, there is
now the \catattr{glosssymbolfont} attribute that's used by
\ics{glossentrysymbol}.)

A simple solution is to use \ics{glossentrynameother} instead:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrbookindexname}}[1]\marg{\comment{}
  \cs{glossentryname}\marg{\idx{param}1}\comment{}
  \cs{ifglshassymbol}\marg{\idx{param}1}\comment{}
  \marg{\comment{}
    \cs{glsifcategory}\marg{\idx{param}1}\marg{chemical}\comment{}
    \marg{, \cs{glossentrysymbol}\marg{\idx{param}1}}\comment{}
    \marg{\cs{space}(\cs{glossentrynameother}\marg{\idx{param}1}\marg{\field{symbol}})}\comment{}
  }\comment{}
  \marg{}\comment{}
}
\end{codeenv}
However, since I decided not to store the short form in the
\field{symbol} field and just saved the dual entry label instead, I
need to lookup the short form from the dual entry:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrbookindexname}}[1]\marg{\comment{}
  \cs{glossentryname}\marg{\idx{param}1}\comment{}
  \cs{ifglshassymbol}\marg{\idx{param}1}\comment{}
  \marg{\comment{}
    \cs{glsifcategory}\marg{\idx{param}1}\marg{chemical}\comment{}
    \marg{, \ics{glossentrysymbol}\marg{\idx{param}1}}\comment{}
    \marg{\cs{space}(\ics{glossentrynameother}\marg{\idx{param}1}\marg{\field{symbol}})}\comment{}
  }\comment{}
  \marg{\comment{}
    \cs{glsifcategory}\marg{\idx{param}1}\marg{markuplanguage}\comment{}
    \marg{\comment{}
      \cs{glsxtrifhasfield}\marg{\field{short}}\marg{\cs{glsxtrusefield}\marg{\idx{param}1}\marg{\field{dual}}}\comment{}
      \marg{\cs{space}(\cs{glscurrentfieldvalue})}\comment{}
      \marg{}\comment{}
    }\comment{}
    \marg{}\comment{}
  }\comment{}
}
\end{codeenv}
Not all of the markup languages are abbreviations so this uses
\ics{glsxtrifhasfield} to check if the \field{short} field is set.
The dual entry's label is easily obtained because \csopt{dual-field}
has provided the \fieldfmt{dual} internal field and set it to the
corresponding label.

It's sometimes useful for the index to include a reference to the
term's definition. This can be done by making use of 
\idx{glsextrapostnamehook}, which can be redefined before the
glossaries to automatically record each entry:
\begin{codeenv}
\cs{renewcommand}\marg{\cs{glsextrapostnamehook}}[1]\marg{\cs{glsadd}\oarg{\glsaddopt[\encap{hyperbf}]{format}}\marg{\idx{param}1}}
\end{codeenv}
This needs to be redefined to ignore its argument before the index,
to avoid the redundant index record:
\begin{codeenv}
\cs{renewcommand}\marg{\cs{glsextrapostnamehook}}[1]\marg{}
\end{codeenv}
Remember that if any records are added within a glossary, an extra
\LaTeX\ and \bibgls\ call are required to ensure that the location list 
is correct, so the document build is:
\begin{verbatim}
pdflatex sample-multi1
bib2gls --group sample-multi1
pdflatex sample-multi1
bib2gls --group sample-multi1
pdflatex sample-multi1
\end{verbatim}
The complete document code is listed below. 
The resulting document is shown in \figureref{fig:sample-multi1.pdf}
and \figureref{fig:sample-multi1.pdf2}.

\lstinputlisting[firstline=5]{../examples/sample-multi1.tex}

\begin{figure}
\figcontents
{\pdftwocol{../examples/sample-multi1.pdf}{4}}
{\caption{\filefmt{sample-multi1.pdf} (pages~1 to~4)}}
{fig:sample-multi1.pdf}
\end{figure}

\begin{figure}
\figcontents
{\pdftwocol[5]{../examples/sample-multi1.pdf}{8}}
{\caption{\filefmt{sample-multi1.pdf} (pages~5 to 8)}}
{fig:sample-multi1.pdf2}
\end{figure}

\filesection{sample-multi2.tex}

This example is an alternative approach to
\exfile{sample-multi1.tex}. Instead of using dual entry types to
define entries that appear in both a glossary and the index, this
example makes use of \csopt{record-label-prefix} to reselect the
recorded entries for the index. This is more complicated but it
allows the entries that have natural word ordering to use a locale
sort method while the entries that are symbolic can use one of the
letter-number sort methods.

This document uses some additional \ext{bib} files to the previous
example, so it has extra glossaries, which all need to be defined:
\begin{codeenv}
\cs{newglossary*}\marg{bacteria}\marg{Bacteria}
\cs{newglossary*}\marg{markuplanguage}\marg{Markup Languages}
\cs{newglossary*}\marg{vegetable}\marg{Vegetables}
\cs{newglossary*}\marg{mineral}\marg{Minerals}
\cs{newglossary*}\marg{animal}\marg{Animals}
\cs{newglossary*}\marg{chemical}\marg{Chemical Formula}
\cs{newglossary*}\marg{baseunit}\marg{SI Units}
\cs{newglossary*}\marg{measurement}\marg{Measurements}
\cs{newglossary*}\marg{film}\marg{Films}
\cs{newglossary*}\marg{book}\marg{Books}
\cs{newglossary*}\marg{person}\marg{People}
\cs{newglossary*}\marg{mediacontrol}\marg{Media Control Symbols}
\cs{newglossary*}\marg{information}\marg{Information Symbols}
\cs{newglossary*}\marg{weather}\marg{Weather Symbols}
\end{codeenv}
Note that this is a total of 15~glossaries (including the
\code{index}). With the basic \ics{cs.makeglossaries} method, this would
require 16 write registers (including the write register used to
create the indexing style file), and a total of $15\times3 + 1 = 46$
associated files. (This doesn't include the standard \ext{aux} file
and the \ext{out} file created by \sty{hyperref}.) With \bibgls, no
additional write registers are required and the number of associated
\bibgls\ files is equal to the number of resource commands plus the
transcript file (in this example, $9+1=10$).

Since this document requires \exfile!{people.bib}, \exfile!{books.bib}
and \exfile!{films.bib} it also requires the files that
supply the definitions of the custom commands
(\exfile!{no-interpret-preamble.bib} and either
\exfile!{interpret-preamble.bib} or \exfile!{interpret-preamble2.bib})
to ensure the custom commands are provided both for the document and
for \bibgls's interpreter.

The first resource set to be loaded simply reads
\exfile{no-interpret-preamble.bib} with the preamble interpreter
switched off:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[no-interpret-preamble]{src},
  \csopt[false]{interpret-preamble}
}
\end{codeenv}
This ensures that \LaTeX\ can pick up the provided commands and
prevents them from being added to the interpreter.

The \exfile{people.bib} file is the next to be loaded with
\exfile{interpret-preamble.bib}. This is loaded separately from the
other resources as this needs the \field{name} field to be copied to
\field{first} (if not already set), as in the
\exfile{sample-people.tex} file. By having a separate resource set,
this setting doesn't affect the other entries. I've also converted
the date fields so that I can customise the format in the document.
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[interpret-preamble,people]{src},
 \csopt[
   \fieldfmt{identifier}=\field{category},
   \fieldfmt{born}=\field{user1},
   \fieldfmt{died}=\field{user2},
   \fieldfmt{othername}=\field{user3}
 ]{field-aliases},
 \csopt[\field{name}=\marg{\field{first}}]{replicate-fields},
 \csopt[person]{type},
 \csopt[false]{save-locations}
 \csopt[\field{user1},\field{user2}]{date-fields},
 \csopt[d MMM y G]{date-field-format}
}
\end{codeenv}
As with the \exfile{sample-people.tex} document, I need to use the
\longarg{break-space} switch to convert the \idx{nbspchar} to a
normal breakable space so that it matches the given format. I've
loaded the \sty{datetime2} package:\footnote{The
\styoptfmt{en-GB} option to \sty{datetime2} also requires that
\styfmt{datetime2-english} must be installed.}
\begin{verbatim}
\usepackage[en-GB]{datetime2}
\end{verbatim}
so that I can use \ics{DTMdisplaydate} to adjust the formatting:
\begin{codeenv}
\cs{newcommand}*\marg{\gls{bibglsdate}}[7]\marg{\cs{DTMdisplaydate}\marg{\idx{param}1}\marg{\idx{param}2}\marg{\idx{param}3}\marg{\idx{param}4}}
\end{codeenv}
This needs to go before the \idx{resourceset} is loaded. Note that
the \styoptfmt{en-GB} option identifies the document locale as
\code{en-GB} (since there are no language packages loaded).

Note that unlike \exfile{sample-people.tex} which had
\csopt[people]{category}, this document obtains the \field{category} field
from the custom \fieldfmt{identifier} field, which in this case 
has the value \code{person}. This means that the category hooks from
\exfile{sample-people.tex} need to be renamed to reflect the
different category label:
\begin{codeenv}
\cs{newcommand}*\marg{\postlinkhook{person}}{\comment{}
 \cs{glsxtrifwasfirstuse}
 \marg{\comment{}
   \cs{ifglshasfield}\marg{\field{user3}}\marg{\cs{glslabel}}\comment{}
   \marg{\cs{space}(\cs{glscurrentfieldvalue})}\comment{}
   \marg{}\comment{}
 }\comment{}
 \marg{}\comment{}
}
\strut
\cs{newcommand}*\marg{\postnamehook{person}}\marg{\comment{}
 \cs{ifglshasfield}\marg{\field{user3}}\marg{\cs{glscurrententrylabel}}\comment{}
 \marg{\cs{space}(\cs{glscurrentfieldvalue})}\comment{}
 \marg{}\comment{}
}
\strut
\cs{newcommand}*\marg{\postdeschook{person}}\marg{\comment{}
 \cs{ifglshasfield}\marg{\field{user1}}\marg{\cs{glscurrententrylabel}}
 \marg{\comment{born}
   \cs{space}(\cs{glscurrentfieldvalue}\cs{comma}-{}-\cs{comma}\comment{}
     \cs{ifglshasfield}\marg{\field{user2}}\marg{\cs{glscurrententrylabel}}
     \marg{\comment{died}
        \cs{glscurrentfieldvalue}
     }\comment{}
     \marg{}\comment{}
   )\comment{}
 }\comment{}
 \marg{}\comment{}
}
\end{codeenv}
The other \ext{bib} files that require locale sorting can now be
loaded, but remember that the abbreviation style settings must be
set first since this resource set includes abbreviations:
\begin{codeenv}
\cs{setabbreviationstyle}\oarg{bacteria}\marg{\abbrstyle{long-only-short-only}}
\cs{setabbreviationstyle}\oarg{markuplanguage}\marg{\abbrstyle{long-short-desc}}
\strut
\cs{renewcommand}*\marg{\cs{glsxtrlongshortdescname}}\marg{\comment{}
  \cs{protect}\cs{glsabbrvfont}\marg{\cs{the}\cs{glsshorttok}}\cs{space}
  \glsxtrparen{\glslongfont{\the\glslongtok}}%
}
\end{codeenv}
Now the resource set can be loaded:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[bacteria,markuplanguages,vegetables,
  minerals,animals,books,films]{src},
 \csopt[\fieldfmt{identifier}=\field{category}]{field-aliases},
 \csopt[same as category]{type},
 \csopt[false]{save-locations}
}
\end{codeenv}
The semantic markup command and attributes are as for
\exfile{sample-multi1.tex}:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{bacteriafont}}[1]\marg{\cs{emph}\marg{\idx{param}1}}
\cs{glssetcategoryattribute}\marg{bacteria}\marg{\catattr{textformat}}\marg{bacteriafont}
\cs{glssetcategoryattribute}\marg{bacteria}\marg{\catattr{glossnamefont}}\marg{bacteriafont}
\cs{glssetcategoryattribute}\marg{bacteria}\marg{\catattr{glossdescfont}}\marg{bacteriafont}
\cs{glssetcategoryattribute}\marg{markuplanguage}\marg{\catattr{glossdesc}}\marg{firstuc}
\end{codeenv}
Similarly for the books:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{bookfont}}[1]\marg{\cs{emph}\marg{\idx{param}1}}
\cs{glssetcategoryattribute}\marg{book}\marg{\catattr{textformat}}\marg{bookfont}
\cs{glssetcategoryattribute}\marg{book}\marg{\catattr{glossnamefont}}\marg{bookfont}
\end{codeenv}
(as for \exfile{sample-media.tex}) and for films:
\begin{codeenv}
\cs{newcommand}\marg{\cmd{filmfont}}[1]\marg{\cs{emph}\marg{\idx{param}1}}
\cs{glssetcategoryattribute}\marg{film}\marg{\catattr{textformat}}\marg{filmfont}
\cs{glssetcategoryattribute}\marg{film}\marg{\catattr{glossnamefont}}\marg{filmfont}
\end{codeenv}

Next come the chemical formulae:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[chemicalformula]{src},
 \csopt[chemical=symbol]{entry-type-aliases},
 \csopt[
   \fieldfmt{identifier}=\field{category},
   \fieldfmt{formula}=\field{name},
   \fieldfmt{chemicalname}=\field{description}
 ]{field-aliases},
 \csopt[chemical]{type},
 \csopt{set-widest},
 \csopt[letternumber-case]{sort},
 \csopt[\field{name}]{symbol-sort-fallback},
 \csopt[false]{save-locations}
}
\end{codeenv}
and the \idxpl{SIunit}, which are now combined into a single
glossary:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[baseunits,derivedunits]{src},
 \csopt[\fieldfmt{measurement}=\field{symbol},\fieldfmt{unit}=\field{symbol}]{entry-type-aliases},
 \csopt[
   \fieldfmt{unitname}=\field{description},
   \fieldfmt{unitsymbol}=\field{symbol},
   \fieldfmt{measurement}=\field{name}
 ]{field-aliases},
 \csopt[measurement]{category},
 \csopt[measurement]{type},
 \csopt{set-widest},
 \csopt[\field{name}]{symbol-sort-fallback},
 \csopt[false]{save-locations}
}
\end{codeenv}
Here the \field{name} field is obtained from the custom
\fieldfmt{measurement} field. Since this contains a word, the
default locale sort is appropriate. I've locally redefined
\ics{glsxtralttreeSymbolDescLocation} to place the symbol in
parentheses after the description:
\begin{codeenv}
\cs{printunsrtglossary*}\oarg{\printglossopt[measurement]{type},\printglossopt[alttree]{style},\printglossopt{nogroupskip}}
\marg{\comment{}
  \cs{renewcommand}\marg{\cs{glsxtralttreeSymbolDescLocation}}[2]\marg{\comment{}
    \cs{glossentrydesc}\marg{\idx{param}1}\comment{}
    \cs{ifglshassymbol}\marg{\idx{param}1}\marg{\cs{space}(\cs{glossentrysymbol}\marg{\idx{param}1})}\marg{}\comment{}
    \cs{glspostdescription}
    \cs{glsxtrAltTreePar}
  }\comment{}
}
\end{codeenv}

The base units are replicated in the \code{baseunit} glossary, this
time with the \field{name} field obtained from the custom
\fieldfmt{unitsymbol} field. This means that I need to find a way to 
prevent duplicate labels. The simplest method is to use
\csopt{duplicate-label-suffix}:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[baseunits]{src},
 \csopt[unit=symbol]{entry-type-aliases},
 \csopt[
   \fieldfmt{unitname}=\field{description},
   \fieldfmt{unitsymbol}=\field{name}
 ]{field-aliases},
 \csopt[measurement]{category},
 \csopt[baseunit]{type},
 \csopt[.copy]{duplicate-label-suffix},
 \csopt[\field{name}]{symbol-sort-fallback},
 \csopt[false]{save-locations}
}
\end{codeenv}
I can't use \csopt{set-widest} here as it won't pick up the modified
label and will instead use the label from the original entry.
Instead I've used \cs{glsFindWidestTopLevelName} to find it:
\begin{codeenv}
\cs{printunsrtglossary*}\oarg{\printglossopt[baseunit]{type},\printglossopt[alttree]{style},\printglossopt{nogroupskip}}
\marg{\comment{}
  \cs{glsFindWidestTopLevelName}\oarg{baseunit}\comment{}
}
\end{codeenv}

The text symbols from \exfile{miscsymbols.bib} are all loaded in a
single \idx{resourceset}, where the \field{type} field can be
obtained from the \field{category}, which in turns is obtained from
the custom \fieldfmt{identifier} field. Since \bibgls\ doesn't
recognise any of the symbol commands, I'm sorting according to the
\field{description} field. (Even if \bibgls\ could determine a
Unicode value for each of the symbols, sorting by the description
makes more sense in this case.)
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
 \csopt[miscsymbols]{src},
 \csopt[
   \fieldfmt{identifier}=\field{category},
   \fieldfmt{icon}=\field{name},
   \fieldfmt{icondescription}=\field{description}
 ]{field-aliases},
 \csopt[icon=symbol]{entry-type-aliases},
 \csopt[same as category]{type},
 \csopt[\field{description}]{sort-field},
 \csopt[false]{save-locations},
 \csopt{set-widest}
}
\end{codeenv}

Finally, all recorded and cross-referenced terms are needed for the
index. This includes entries that have already been defined in the
earlier \idxpl{resourceset} (so a guard against duplicates is
necessary) but it also includes entries from the \exfile{terms.bib}
file that haven't yet been dealt with. I'd like the index to start
with a \idx{symbolgroup} containing the icons from
\exfile{miscsymbols.bib}. This needs to be dealt with separately from
the rest of the index to keep them together in a single group:
\begin{codeenv}
\gls{GlsXtrLoadResources}\oarg{
  \csopt[miscsymbols]{src},
  \csopt[recorded no deps]{selection},
  \csopt[.copy]{duplicate-label-suffix},
  \csopt[icon=index]{entry-type-aliases},
  \csopt[
   \fieldfmt{identifier}=\field{category},
   \fieldfmt{icondescription}=\field{symbol},
   \fieldfmt{icon}=\field{name}
  ]{field-aliases},
  \csopt[index]{type},
  \csopt[\field{symbol}]{sort-field},
  \csopt[glssymbols]{group}
}
\end{codeenv}
Since I know that there are no parents or cross-references in this
set of entries I've used \csopt[recorded no deps]{selection} to skip
the dependency checks. In this \idx{resourceset}, the \field{name}
field has the symbol command (obtained from the custom
\fieldfmt{icon} field), and the \field{symbol} field has the symbol
description (obtained from the custom \fieldfmt{icondescription}
field), which is used as the sort field. I've set the \field{group}
label to \code{glssymbols}, which keeps all these entries in a
single group and the title will be obtained from
\ics{glssymbolsgroupname}.

Before loading the final \idx{resourceset} \ics{glsxtrlongshortdescname} needs 
to be changed
so that the abbreviations using the \abbrstyle{long-short-desc}
style (that is, the abbreviations with the \field{category} set to 
\code{markuplanguage}) have the \field{name} field set to
\meta{long} (\meta{short}):
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrlongshortdescname}}\marg{\comment{}
  \cs{protect}\cs{glslongfont}\marg{\cs{the}\cs{glslongtok}}\cs{space}
  \cs{glsxtrparen}\marg{\cs{glsabbrvfont}\marg{\cs{the}\cs{glsshorttok}}}\comment{}
}
\end{codeenv}
The \abbrstyle{long-only-short-only} style has a similar command,
but it was only introduced to \sty{glossaries-extra} version 1.25:
\begin{codeenv}
\cs{renewcommand}*\marg{\ics{glsxtronlyname}}{\comment{}
  \cs{protect}\cs{glsabbrvonlyfont}\marg{\cs{the}\cs{glslongtok}}\comment{}
}
\end{codeenv}

The abbreviations all need to be sorted according to the long form:
\begin{codeenv}
\csopt[\field{long}]{abbreviation-sort-fallback}
\end{codeenv}
The custom entry types and fields again need to be aliased
\begin{codeenv}
\csopt[
 chemical=index,
 measurement=entry,
 unit=dualentry,
 icon=index
]{entry-type-aliases},
\csopt[
   \fieldfmt{identifier}=\field{category},
   \fieldfmt{formula}=\field{symbol},
   \fieldfmt{chemicalname}=\field{name},
   \fieldfmt{unitname}=\field{description},
   \fieldfmt{unitsymbol}=\field{symbol},
   \fieldfmt{measurement}=\field{name},
   \fieldfmt{icon}=\field{symbol},
   \fieldfmt{icondescription}=\field{name}
 ]{field-aliases}
\end{codeenv}
The chemical formulae and icons are now defined using
\atentry{index} with the \field{name} field set to a word form
(chemical name and icon description). This means
they're appropriate for alphabetical sorting. (Both \atentry{entry}
and \atentry{symbol} require the \field{description} field, which is
why I've aliased \atentryfmt{chemical} and \atentryfmt{icon}
to \atentry{index} here.) The custom \atentryfmt{measurement} entry
type has a \field{description} field (obtained from
\fieldfmt{unitname}), so that's aliased to \atentry{entry} as again
the \field{name} field is suitable for alphabetical sorting.

I've aliased \atentryfmt{unit} to \atentry{dualentry}
rather than \atentry{symbol} as I want both the unit name and the
measurement in the index and I've combined their location lists:
\begin{codeenv}
\csopt[both]{combine-dual-locations}
\end{codeenv}
Both primary and dual entries need to go in the \code{index}
glossary:
\begin{codeenv}
\csopt[index]{type},
\csopt[index]{dual-type}
\end{codeenv}

All \ext{bib} files used in the previous \idxpl{resourceset} are needed
as well as the \exfile{terms.bib} file:
\begin{codeenv}
\csopt[terms,bacteria,markuplanguages,vegetables,minerals,
  animals,chemicalformula,baseunits,derivedunits,people,
  films,books,miscsymbols]{src}
\end{codeenv}
but this time I also want to select entries that haven't been
recorded but have a cross-reference to a recorded entry:
\begin{codeenv}
\csopt{selection}[recorded and deps and see]
\end{codeenv}
Again it's necessary to provide a way to avoid duplicate entry
labels, which can be done with
\begin{codeenv}
\csopt[.copy]{duplicate-label-suffix},
\end{codeenv}
as above.  However, this will cause the cross-references (from the
\field{alias} fields) to link to the glossary rather than the index.
This may or may not confuse the reader. For consistency, it may be
more suitable to have the cross-reference in the index link to the
aliased entry in the index rather than in the glossary. I've
therefore instead used:
\begin{codeenv}
 \csopt[idx.]{label-prefix},
 \csopt[idx.]{record-label-prefix},
\end{codeenv}
This means that the entries defined in \exfile{terms.bib} need to be
referenced with this prefix.

All instances of \cs{gls} will link to the original entry, so all
entries except for those in the \exfile{terms.bib} file will link to
the relevant glossary. Those in the \exfile{terms.bib} file will
link to the index. It's possible to disable the hyperlinks for those
entries, but the reader may find it useful to jump to the index to
look up other locations for that entry in the document.

To deal with the identical book and film titles, I'm again using the
\field{category} to resolve identical sort values:
\begin{codeenv}
\csopt[\field{category}]{identical-sort-action}
\end{codeenv}
For the people who have a \field{first} field, I've decided that
this would be more appropriate for the index as it's more compact
than the \field{name}, so here I've done the reverse to earlier and copied the
\field{first} field (if supplied) into the \field{name} field, but
since the \field{name} field is already provided the override 
setting needs to be on:
\begin{codeenv}
\csopt{replicate-override},
\csopt[\field{first}=\field{name}]{replicate-fields}
\end{codeenv}
As with \exfile{sample-people.tex} I've provided some custom
commands to make it easier to locally redefine \gls{sortname} and
\gls{sortvonname}:
\begin{codeenv}
\cs{newcommand}*\marg{\cmd{swaptwo}}[2]\marg{\idx{param}2, \idx{param}1}
\cs{newcommand}*\marg{\cmd{swapthree}}[3]\marg{\idx{param}2 \idx{param}3, \idx{param}1}
\end{codeenv}
I've redefined \ics{glsxtrbookindexname} in a similar manner to
\exfile{sample-multi1.tex} but it has some modifications:
\begin{codeenv}
\cs{renewcommand}*\marg{\cs{glsxtrbookindexname}}[1]\marg{\comment{}
  \cs{glossentryname}\marg{\idx{param}1}\comment{}
  \cs{ifglshassymbol}\marg{\idx{param}1}\comment{}
  \marg{\comment{}
    \cs{glsifcategory}\marg{\idx{param}1}\marg{chemical}\comment{}
    \marg{, \cs{glossentrysymbol}\marg{\idx{param}1}}\comment{}
    \marg{\cs{space}(\cs{glossentrynameother}\marg{\idx{param}1}\marg{\field{symbol}})}\comment{}
  }\comment{}
  \marg{\comment{}
    \cs{glsifcategory}\marg{\idx{param}1}\marg{film}\comment{}
    \marg{\cs{cs.space}(film)}\comment{}
    \marg{}\comment{}
  }\comment{}
}
\end{codeenv}
This appends \qt{(film)} to film names. I've chosen this method
rather than using the \idx{postnamehook} as I only want this in the
index and not in the list of films.

For some of the entries that are referenced in the document, I've
appended information in parentheses:
\begin{codeenv}
\cs{gls}\marg{Al2SO43} (\ics{glsdesc}\marg{Al2SO43})
\end{codeenv}
This is all right for odd instances, but if this always needs to be done 
on first use, then it's better to use the \idx{postlinkhook}, which
is what I've done for the icons for comparison:
\begin{codeenv}
\cs{newcommand}*\marg{\postlinkhook{mediacontrol}}\marg{\comment{}
 \ics{glsxtrpostlinkAddDescOnFirstUse}
}
\strut
\cs{newcommand}*\marg{\postlinkhook{information}}\marg{\comment{}
 \cs{glsxtrpostlinkAddDescOnFirstUse}
}
\strut
\cs{newcommand}*\marg{\postlinkhook{weather}}\marg{\comment{}
 \cs{glsxtrpostlinkAddDescOnFirstUse}
}
\end{codeenv}

I've also provided some custom commands to make it easier to reference
entries without worrying about the prefixes:
\begin{codeenv}
\cs{newcommand}\marg{\csfmt{unit}}\marg{\ics{glssymbol}}
\cs{newcommand}\marg{\csfmt{measurement}}\marg{\cs{gls}}
\gls{glsxtrnewgls}\marg{film.}\marg{\csfmt{film}}
\end{codeenv}

As with \exfile{sample-multi1.tex}, it would be useful to include
the page where the entries are defined in their corresponding lists.
Again this can be done by redefining the general purpose
non-category post-name hook \ics{glsextrapostnamehook}:
\begin{codeenv}
\cs{newcommand}*\marg{\cs{glsextrapostnamehook}}[1]\marg{\comment{}
  \cs{glsadd}\oarg{\glsaddopt[\encap{hyperbf}]{format}}\marg{\idx{param}1}\comment{}
}
\end{codeenv}
This needs resetting before the index, since it's redundant to
record an entry in the index. This will require an extra
\bibgls+\LaTeX\ system call as this code can't be performed until
the glossaries have been created.

The complete document code is listed below. The document build is:
\begin{verbatim}
pdflatex sample-multi2
bib2gls --group --break-space sample-multi2
pdflatex sample-multi2
bib2gls --group --break-space sample-multi2
pdflatex sample-multi2
\end{verbatim}
The resulting document is shown in \figureref{fig:sample-multi2.pdf},
\figureref{fig:sample-multi2.pdf2} and \figureref{fig:sample-multi2.pdf3}.

\lstinputlisting[firstline=5]{../examples/sample-multi2.tex}

\begin{figure}
\figcontents
{\pdftwocol{../examples/sample-multi2.pdf}{4}}
{\caption{\filefmt{sample-multi2.pdf} (pages~1 to~4)}}
{fig:sample-multi2.pdf}
\end{figure}

\begin{figure}
\figcontents
{\pdftwocol[5]{../examples/sample-multi2.pdf}{8}}
{\caption{\filefmt{sample-multi2.pdf} (pages~5 to 8)}}
{fig:sample-multi2.pdf2}
\end{figure}

\begin{figure}
\figcontents
{\pdftwocol[9]{../examples/sample-multi2.pdf}{12}}
{\caption{\filefmt{sample-multi2.pdf} (pages~9 and 12)}}
{fig:sample-multi2.pdf3}
\end{figure}

\setupglossaries{indexonlyfirst}
\printstyoptsummary
\printcommandsummary
%\printcommandtopic

\bibliographystyle{plain}
\bibliography{bib2gls-cite}

\renewcommand*{\glsxtrbookindexsubname}[1]{%
  \glsifcategory{#1}{command}%
  {%
    \glsxtrifhasfield{useri}{#1}%
    {\glossentrynameother{#1}{useri}}%
    {\glossentryname{#1}}%
  }%
  {\glossentryname{#1}}%
}
\newcommand{\locationfont}{\small}
\newcommand{\widecrossreffont}{\footnotesize}
\newcommand{\ifwidecrossref}[3]{%
  \dimen0=0pt\relax
  \glsxtrifhasfield*{see}{#1}%
  {\settowidth{\dimen0}{\locationfont\glsxtrusesee{#1}}}%
  {%
    \glsxtrifhasfield*{seealso}{#1}%
    {\settowidth{\dimen0}{\locationfont\glsxtruseseealso{#1}}}%
    {%
      \glsxtrifhasfield*{alias}{#1}%
      {\settowidth{\dimen0}{\locationfont\glsxtrusealias{#1}}}%
      {}%
    }%
  }%
  \ifdim\dimen0>0.4\linewidth #2\else#3\fi
}
\newcommand{\precrossref}{}
\newcommand{\indexdotfill}{\,\cleaders\hbox to .44em{\hss\textcolor{lightgray}{.}\hss}\hfill\,}
\renewcommand*{\glsxtrbookindexprelocation}[1]{%
  \glsxtrifhasfield*{location}{#1}%
  {\nobreak\indexdotfill
   \ifwidecrossref{#1}%
   {\def\precrossref{\widecrossreffont\glstreesubitem}}%
   {\def\precrossref{, }}%
  }%
  {%
    \ifwidecrossref{#1}{\nobreak\def\precrossref{\widecrossreffont\glstreesubitem}}%
    {\def\precrossref{\nobreak\indexdotfill}}%
    \glsxtrprelocation
  }%
}
\renewcommand{\glsxtrbookindexlocation}[2]{%
  \begingroup
  \locationfont
  #2%
  \glsxtrifhasfield{see}{#1}{\precrossref\glsxtrusesee{#1}}%
  {%
    \glsxtrifhasfield{seealso}{#1}%
    {\precrossref\glsxtruseseealso{#1}}%
    {%
      \glsxtrifhasfield{alias}{#1}{\precrossref\glsxtrusealias{#1}}{}%
    }%
  }%
  \endgroup
  \let\precrossref\empty
}
% With \printcommandtopic missing a lot of spawned entries will
% appear in the index without a record. Without the topic list they
% need to be skipped
\renewcommand{\printunsrtglossaryentryprocesshook}[1]{%
 \glsxtrifhasfield*{progenitor}{#1}%
 {%\typeout{SKIPPING [has progenitor] #1}
   \printunsrtglossaryskipentry
 }%
 {%
   \GlsXtrIfFieldEqStr*{originalentrytype}{#1}{topic}%
   {%
     \glsxtrifhasfield*{location}{#1}%
     {}%
     {%
       \ifglshasdesc{#1}{}%
       {%
         %\typeout{SKIPPING [no location] #1}
         \printunsrtglossaryskipentry
       }%
     }%
   }%
   {%
     \GlsXtrIfFieldValueInCsvList*{#1}{category}%
     {abbreviationstyle,packageoption,glossarystyle,package}%
     {%
       \glsxtrifhasfield*{location}{#1}{}%
       {%\typeout{SKIPPING [topic \glscategory{#1} no location] #1}%
        \printunsrtglossaryskipentry
       }%
     }%
     {}%
   }%
 }%
}
\renewcommand*{\unicodecategoryfmt}[1]{\textsf{#1}}
\glsrenewcommand*{\bibglspassim}{}
\newcommand{\glsxtrpostnamecommand}{%
  \glsxtrifhasfield{userii}{\glscurrententrylabel}{ (\glscurrentfieldvalue)}{}%
}
\printunsrtglossary[type=index]
\end{document}