summaryrefslogtreecommitdiff
path: root/Master/texmf-dist/source/latex/l3kernel/l3candidates.dtx
blob: d3ef2708afc8fc22da6db18cf15d9c39766ff113 (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
% \iffalse meta-comment
%
%% File: l3candidates.dtx Copyright (C) 2012-2016 The LaTeX3 Project
%%
%% It may be distributed and/or modified under the conditions of the
%% LaTeX Project Public License (LPPL), either version 1.3c of this
%% license or (at your option) any later version.  The latest version
%% of this license is in the file
%%
%%    http://www.latex-project.org/lppl.txt
%%
%% This file is part of the "l3kernel bundle" (The Work in LPPL)
%% and all files in that bundle must be distributed together.
%%
%% The released version of this bundle is available from CTAN.
%%
%% -----------------------------------------------------------------------
%%
%% The development version of the bundle can be found at
%%
%%    http://www.latex-project.org/svnroot/experimental/trunk/
%%
%% for those people who are interested.
%%
%%%%%%%%%%%
%% NOTE: %%
%%%%%%%%%%%
%%
%%   Snapshots taken from the repository represent work in progress and may
%%   not work or may contain conflicting material!  We therefore ask
%%   people _not_ to put them into distributions, archives, etc. without
%%   prior consultation with the LaTeX Project Team.
%%
%% -----------------------------------------------------------------------
%%
%
%<*driver>
\documentclass[full]{l3doc}
%</driver>
%<*driver|package>
\GetIdInfo$Id: l3candidates.dtx 6688 2016-08-19 19:21:48Z bruno $
  {L3 Experimental additions to l3kernel}
%</driver|package>
%<*driver>
\begin{document}
  \DocInput{\jobname.dtx}
\end{document}
%</driver>
% \fi
%
% \title{^^A
%   The \textsf{l3candidates} package\\ Experimental additions to
%   \pkg{l3kernel}^^A
%   \thanks{This file describes v\ExplFileVersion,
%     last revised \ExplFileDate.}^^A
% }
%
% \author{^^A
%  The \LaTeX3 Project\thanks
%    {^^A
%      E-mail:
%        \href{mailto:latex-team@latex-project.org}
%          {latex-team@latex-project.org}^^A
%    }^^A
% }
%
% \date{Released \ExplFileDate}
%
% \maketitle
%
% \begin{documentation}
%
% \section{Important notice}
%
% This module provides a space in which functions can be added to
% \pkg{l3kernel} (\pkg{expl3}) while still being experimental.
% \begin{quote}
%  \bfseries
% As such, the
% functions here may not remain in their current form, or indeed at all,
% in \pkg{l3kernel} in the future.
% \end{quote}
%  In contrast to the material in
% \pkg{l3experimental}, the functions here are all \emph{small} additions to
% the kernel. We encourage programmers to test them out and report back on
% the \texttt{LaTeX-L} mailing list.
%
% \medskip
%
%   Thus, if you intend to use any of these functions from the candidate module in a public package
%  offered to others for productive use (e.g., being placed on CTAN) please consider the following points carefully:
% \begin{itemize}
% \item Be prepared that your public packages might require updating when such functions
%        are being finalized.
% \item Consider informing us that you use a particular function in your public package, e.g., by
%         discussing this on the \texttt{LaTeX-L}
%    mailing list. This way it becomes easier to coordinate any updates necessary without issues
%    for the users of your package.
% \item Discussing and understanding use cases for a particular addition or concept also helps to
%         ensure that we provide the right interfaces in the final version so please give us feedback
%         if you consider a certain candidate function useful (or not).
% \end{itemize}
% We only add functions in this space if we consider them being serious candidates for a final inclusion
% into the kernel. However, real use sometimes leads to better ideas, so functions from this module are
% \textbf{not necessarily stable} and we may have to adjust them!
%
% \section{Additions to \pkg{l3basics}}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-03]{\cs_log:N, \cs_log:c}
%   \begin{syntax}
%     \cs{cs_log:N} \meta{control sequence}
%   \end{syntax}
%   Writes the definition of the \meta{control sequence} in the log
%   file.  See also \cs{cs_show:N} which displays the result in the
%   terminal.
% \end{function}
%
% \begin{function}[updated = 2015-08-03]
%   {\__kernel_register_log:N, \__kernel_register_log:c}
%   \begin{syntax}
%     \cs{__kernel_register_log:N} \meta{register}
%   \end{syntax}
%   Used to write the contents of a \TeX{} register to the log file in a
%   form similar to \cs{__kernel_register_show:N}.
% \end{function}
%
% \section{Additions to \pkg{l3box}}
%
% \subsection{Affine transformations}
%
% Affine transformations are changes which (informally) preserve straight
% lines. Simple translations are affine transformations, but are better handled
% in \TeX{} by doing the translation first, then inserting an unmodified box.
% On the other hand, rotation and resizing of boxed material can best be
% handled by modifying boxes. These transformations are described here.
%
% \begin{function}{\box_resize:Nnn, \box_resize:cnn}
%   \begin{syntax}
%     \cs{box_resize:Nnn} \meta{box} \Arg{x-size} \Arg{y-size}
%   \end{syntax}
%   Resize the \meta{box} to \meta{x-size} horizontally and \meta{y-size}
%   vertically (both of the sizes are dimension expressions).
%   The \meta{y-size} is the vertical size (height plus depth) of
%   the box. The updated \meta{box} will be an hbox, irrespective of the nature
%   of the \meta{box}  before the resizing is applied. Negative sizes will
%   cause the material in the \meta{box} to be reversed in direction, but the
%   reference point of the \meta{box} will be unchanged.
%   Thus negative $y$-sizes will result in a box a depth dependent on the
%   height of the original box a height dependent on the depth.
%   The resizing applies within the current \TeX{} group level.
% \end{function}
%
% \begin{function}
%   {\box_resize_to_ht_plus_dp:Nn, \box_resize_to_ht_plus_dp:cn}
%   \begin{syntax}
%     \cs{box_resize_to_ht_plus_dp:Nn} \meta{box} \Arg{y-size}
%   \end{syntax}
%   Resize the \meta{box} to \meta{y-size} vertically, scaling the horizontal
%   size by the same amount (\meta{y-size} is a dimension expression).
%   The \meta{y-size} is the vertical size (height plus depth) of
%   the box.
%   The updated \meta{box} will be an hbox, irrespective of the nature
%   of the \meta{box}  before the resizing is applied. A negative size will
%   cause the material in the \meta{box} to be reversed in direction, but the
%   reference point of the \meta{box} will be unchanged.
%   Thus negative $y$-sizes will result in a box with depth dependent on the
%   height of the original box and height dependent on the depth of the original.
%   The resizing applies within the current \TeX{} group level.
% \end{function}
%
% \begin{function}
%   {\box_resize_to_ht:Nn, \box_resize_to_ht:cn}
%   \begin{syntax}
%     \cs{box_resize_to_ht:Nn} \meta{box} \Arg{y-size}
%   \end{syntax}
%   Resize the \meta{box} to \meta{y-size} vertically, scaling the horizontal
%   size by the same amount (\meta{y-size} is a dimension expression).
%   The \meta{y-size} is the height only, not including depth, of
%   the box.
%   The updated \meta{box} will be an hbox, irrespective of the nature
%   of the \meta{box}  before the resizing is applied.
%   A negative size will
%   cause the material in the \meta{box} to be reversed in direction, but the
%   reference point of the \meta{box} will be unchanged.
%   Thus negative $y$-sizes will result in a box with depth dependent on the
%   height of the original box and height dependent on the depth of the original.
%   The resizing applies within the current \TeX{} group level.
% \end{function}
%
% \begin{function}{\box_resize_to_wd:Nn, \box_resize_to_wd:cn}
%   \begin{syntax}
%     \cs{box_resize_to_wd:Nn} \meta{box} \Arg{x-size}
%   \end{syntax}
%   Resize the \meta{box} to \meta{x-size} horizontally, scaling the vertical
%   size by the same amount (\meta{x-size} is a dimension expression).
%   The updated \meta{box} will be an hbox, irrespective of the nature
%   of the \meta{box}  before the resizing is applied. A negative size will
%   cause the material in the \meta{box} to be reversed in direction, but the
%   reference point of the \meta{box} will be unchanged.
%   Thus negative $y$-sizes will result in a box a depth dependent on the
%   height of the original box a height dependent on the depth.
%   The resizing applies within the current \TeX{} group level.
% \end{function}
%
% \begin{function}[added = 2014-07-03]
%   {\box_resize_to_wd_and_ht:Nnn, \box_resize_to_wd_and_ht:cnn}
%   \begin{syntax}
%     \cs{box_resize_to_wd_and_ht:Nnn} \meta{box} \Arg{x-size} \Arg{y-size}
%   \end{syntax}
%   Resize the \meta{box} to a \emph{height} of
%   \meta{x-size} horizontally and \meta{y-size}
%   vertically (both of the sizes are dimension expressions).
%   The \meta{y-size} is the \emph{height} of the box, ignoring any depth.
%   The updated \meta{box} will be an hbox, irrespective of the nature
%   of the \meta{box}  before the resizing is applied. Negative sizes will
%   cause the material in the \meta{box} to be reversed in direction, but the
%   reference point of the \meta{box} will be unchanged.
% \end{function}
%
% \begin{function}{\box_rotate:Nn, \box_rotate:cn}
%   \begin{syntax}
%     \cs{box_rotate:Nn} \meta{box} \Arg{angle}
%   \end{syntax}
%   Rotates the \meta{box} by \meta{angle} (in degrees) anti-clockwise about
%   its reference point. The reference point of the updated box will be moved
%   horizontally such that it is at the left side of the smallest rectangle
%   enclosing the rotated material.
%   The updated \meta{box} will be an hbox, irrespective of the nature
%   of the \meta{box} before the rotation is applied. The rotation applies
%   within the current \TeX{} group level.
% \end{function}
%
% \begin{function}{\box_scale:Nnn, \box_scale:cnn}
%   \begin{syntax}
%     \cs{box_scale:Nnn} \meta{box} \Arg{x-scale} \Arg{y-scale}
%   \end{syntax}
%   Scales the \meta{box} by factors \meta{x-scale} and \meta{y-scale} in
%   the horizontal and vertical directions, respectively (both scales are
%   integer expressions). The updated \meta{box} will be an hbox, irrespective
%   of the nature of the \meta{box} before the scaling is applied. Negative
%   scalings will cause the material in the \meta{box} to be reversed in
%   direction, but the reference point of the \meta{box} will be unchanged.
%   Thus negative $y$-scales will result in a box a depth dependent on the
%   height of the original box a height dependent on the depth.
%   The resizing applies within the current \TeX{} group level.
% \end{function}
%
% \subsection{Viewing part of a box}
%
% \begin{function}{\box_clip:N, \box_clip:c}
%   \begin{syntax}
%     \cs{box_clip:N} \meta{box}
%   \end{syntax}
%   Clips the \meta{box} in the output so that only material inside the
%   bounding box is displayed in the output. The updated \meta{box} will be an
%   hbox, irrespective of the nature of the \meta{box} before the clipping is
%   applied. The clipping applies within the current \TeX{} group level.
%
%   \textbf{These functions require the \LaTeX3 native drivers: they will
%   not work with the \LaTeXe{} \pkg{graphics} drivers!}
%
%   \begin{texnote}
%     Clipping is implemented by the driver, and as such the full content of
%     the box is placed in the output file. Thus clipping does not remove
%     any information from the raw output, and hidden material can therefore
%     be viewed by direct examination of the file.
%   \end{texnote}
% \end{function}
%
% \begin{function}{\box_trim:Nnnnn, \box_trim:cnnnn}
%   \begin{syntax}
%     \cs{box_trim:Nnnnn} \meta{box} \Arg{left} \Arg{bottom} \Arg{right} \Arg{top}
%   \end{syntax}
%   Adjusts the bounding box of the \meta{box} \meta{left} is removed from
%   the left-hand edge of the bounding box, \meta{right} from the right-hand
%   edge and so fourth. All adjustments are \meta{dimension expressions}.
%   Material output of the bounding box will still be displayed in the output
%   unless \cs{box_clip:N} is subsequently applied.
%   The updated \meta{box} will be an
%   hbox, irrespective of the nature of the \meta{box} before the trim
%   operation is applied. The adjustment applies within the current \TeX{}
%   group level. The behavior of the operation where the trims requested is
%   greater than the size of the box is undefined.
% \end{function}
%
% \begin{function}{\box_viewport:Nnnnn, \box_viewport:cnnnn}
%   \begin{syntax}
%     \cs{box_viewport:Nnnnn} \meta{box} \Arg{llx} \Arg{lly} \Arg{urx} \Arg{ury}
%   \end{syntax}
%   Adjusts the bounding box of the \meta{box} such that it has lower-left
%   co-ordinates (\meta{llx}, \meta{lly}) and upper-right co-ordinates
%   (\meta{urx}, \meta{ury}). All four co-ordinate positions are
%   \meta{dimension expressions}. Material output of the bounding box will
%   still be displayed in the output unless \cs{box_clip:N} is
%   subsequently applied.
%   The updated \meta{box} will be an
%   hbox, irrespective of the nature of the \meta{box} before the viewport
%   operation is applied. The adjustment applies within the current \TeX{}
%   group level.
% \end{function}
%
% \section{Additions to \pkg{l3clist}}
%
% \begin{function}[added = 2014-08-22]{\clist_log:N, \clist_log:c}
%   \begin{syntax}
%     \cs{clist_log:N} \meta{comma list}
%   \end{syntax}
%   Writes the entries in the \meta{comma list} in the log file.  See
%   also \cs{clist_show:N} which displays the result in the terminal.
% \end{function}
%
% \begin{function}[added = 2014-08-22]{\clist_log:n}
%   \begin{syntax}
%     \cs{clist_log:n} \Arg{tokens}
%   \end{syntax}
%   Writes the entries in the comma list in the log file.  See also
%   \cs{clist_show:n} which displays the result in the terminal.
% \end{function}
%
% \section{Additions to \pkg{l3coffins}}
%
% \begin{function}{\coffin_resize:Nnn, \coffin_resize:cnn}
%   \begin{syntax}
%     \cs{coffin_resize:Nnn} \meta{coffin} \Arg{width} \Arg{total-height}
%   \end{syntax}
%   Resized the \meta{coffin} to \meta{width} and \meta{total-height},
%   both of which should be given as dimension expressions.
% \end{function}
%
% \begin{function}{\coffin_rotate:Nn, \coffin_rotate:cn}
%   \begin{syntax}
%     \cs{coffin_rotate:Nn} \meta{coffin} \Arg{angle}
%   \end{syntax}
%   Rotates the \meta{coffin} by the given \meta{angle} (given in
%   degrees counter-clockwise). This process will rotate both the
%   coffin content and poles. Multiple rotations will not result in
%   the bounding box of the coffin growing unnecessarily.
% \end{function}
%
% \begin{function}{\coffin_scale:Nnn, \coffin_scale:cnn}
%   \begin{syntax}
%     \cs{coffin_scale:Nnn} \meta{coffin} \Arg{x-scale} \Arg{y-scale}
%   \end{syntax}
%   Scales the \meta{coffin} by a factors \meta{x-scale} and
%   \meta{y-scale} in the horizontal and vertical directions,
%   respectively. The two scale factors should be given as real numbers.
% \end{function}
%
% \begin{function}[added = 2014-08-22]
%   {\coffin_log_structure:N, \coffin_log_structure:c}
%   \begin{syntax}
%     \cs{coffin_log_structure:N} \meta{coffin}
%   \end{syntax}
%   This function writes the structural information about the
%   \meta{coffin} in the log file. The width, height and depth of the
%   typeset material are given, along with the location of all of the
%   poles of the coffin. See also \cs{coffin_show_structure:N} which
%   displays the result in the terminal.
% \end{function}
%
% \section{Additions to \pkg{l3file}}
%
% \begin{function}[TF, added = 2014-07-02]{\file_if_exist_input:n}
%   \begin{syntax}
%     \cs{file_if_exist_input:n} \Arg{file name}
%     \cs{file_if_exist_input:nTF} \Arg{file name} \Arg{true code} \Arg{false code}
%   \end{syntax}
%   Searches for \meta{file name} using the current \TeX{} search
%   path and the additional paths controlled by
%   \cs{file_path_include:n}). If found, inserts the \meta{true code} then
%   reads in the file as additional \LaTeX{} source as described for
%   \cs{file_input:n}. Note that \cs{file_if_exist_input:n} does not raise
%   an error if the file is not found, in contrast to \cs{file_input:n}.
% \end{function}
%
% \begin{function}[added = 2012-02-11]{\ior_map_inline:Nn}
%   \begin{syntax}
%     \cs{ior_map_inline:Nn} \meta{stream} \Arg{inline function}
%   \end{syntax}
%   Applies the \meta{inline function} to \meta{lines} obtained by
%   reading one or more lines (until an equal number of left and right
%   braces are found) from the \meta{stream}. The \meta{inline function}
%   should consist of code which will receive the \meta{line} as |#1|.
%   Note that \TeX{} removes trailing space and tab characters
%   (character codes 32 and 9) from every line upon input.  \TeX{} also
%   ignores any trailing new-line marker from the file it reads.
% \end{function}
%
% \begin{function}[added = 2012-02-11]{\ior_str_map_inline:Nn}
%   \begin{syntax}
%     \cs{ior_str_map_inline:Nn} \Arg{stream} \Arg{inline function}
%   \end{syntax}
%   Applies the \meta{inline function} to every \meta{line}
%   in the \meta{stream}. The material is read from the \meta{stream}
%   as a series of tokens with category code $12$ (other), with the
%   exception of space characters which are given category code $10$
%   (space). The \meta{inline function} should consist of code which
%   will receive the \meta{line} as |#1|.
%   Note that \TeX{} removes trailing space and tab characters
%   (character codes 32 and 9) from every line upon input.  \TeX{} also
%   ignores any trailing new-line marker from the file it reads.
% \end{function}
%
% \begin{function}[added = 2012-06-29]{\ior_map_break:}
%   \begin{syntax}
%     \cs{ior_map_break:}
%   \end{syntax}
%   Used to terminate a \cs[no-index]{ior_map_\ldots} function before all
%   lines from the \meta{stream} have been processed. This will
%   normally take place within a conditional statement, for example
%   \begin{verbatim}
%     \ior_map_inline:Nn \l_my_ior
%       {
%         \str_if_eq:nnTF { #1 } { bingo }
%           { \ior_map_break: }
%           {
%             % Do something useful
%           }
%       }
%   \end{verbatim}
%   Use outside of a \cs[no-index]{ior_map_\ldots} scenario will lead to low
%   level \TeX{} errors.
%   \begin{texnote}
%     When the mapping is broken, additional tokens may be inserted by the
%     internal macro \cs{__prg_break_point:Nn} before further items are taken
%     from the input stream. This will depend on the design of the mapping
%     function.
%   \end{texnote}
% \end{function}
%
% \begin{function}[added = 2012-06-29]{\ior_map_break:n}
%   \begin{syntax}
%     \cs{ior_map_break:n} \Arg{tokens}
%   \end{syntax}
%   Used to terminate a \cs[no-index]{ior_map_\ldots} function before all
%   lines in the \meta{stream} have been processed, inserting
%   the \meta{tokens} after the mapping has ended. This will
%   normally take place within a conditional statement, for example
%   \begin{verbatim}
%     \ior_map_inline:Nn \l_my_ior
%       {
%         \str_if_eq:nnTF { #1 } { bingo }
%           { \ior_map_break:n { <tokens> } }
%           {
%             % Do something useful
%           }
%       }
%   \end{verbatim}
%   Use outside of a \cs[no-index]{ior_map_\ldots} scenario will lead to low
%   level \TeX{} errors.
%   \begin{texnote}
%     When the mapping is broken, additional tokens may be inserted by the
%     internal macro \cs{__prg_break_point:Nn} before the \meta{tokens} are
%     inserted into the input stream.
%     This will depend on the design of the mapping function.
%   \end{texnote}
% \end{function}
%
% \begin{function}[added = 2014-08-22]
%   {\ior_log_streams:, \iow_log_streams:}
%   \begin{syntax}
%     \cs{ior_log_streams:}
%     \cs{iow_log_streams:}
%   \end{syntax}
%   Writes in the log file a list of the file names associated with each
%   open stream: intended for tracking down problems.
% \end{function}
%
% \section{Additions to \pkg{l3fp}}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-07]
%   {\fp_log:N, \fp_log:c, \fp_log:n}
%   \begin{syntax}
%     \cs{fp_log:N} \meta{fp~var}
%     \cs{fp_log:n} \Arg{floating point expression}
%   \end{syntax}
%   Evaluates the \meta{floating point expression} and writes the
%   result in the log file.
% \end{function}
%
% \section{Additions to \pkg{l3int}}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-03]{\int_log:N, \int_log:c}
%   \begin{syntax}
%     \cs{int_log:N} \meta{integer}
%   \end{syntax}
%   Writes the value of the \meta{integer} in the log file.
% \end{function}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-07]{\int_log:n}
%   \begin{syntax}
%     \cs{int_log:n} \Arg{integer expression}
%   \end{syntax}
%   Writes the result of evaluating the \meta{integer expression}
%   in the log file.
% \end{function}
%
% \section{Additions to \pkg{l3keys}}
%
% \begin{function}[added = 2014-08-22]{\keys_log:nn}
%   \begin{syntax}
%     \cs{keys_log:nn} \Arg{module} \Arg{key}
%   \end{syntax}
%   Writes in the log file the function which is used to actually
%   implement a \meta{key} for a \meta{module}.
% \end{function}
%
% \section{Additions to \pkg{l3msg}}
%
% In very rare cases it may be necessary to produce errors in an
% expansion-only context.  The functions in this section should only be
% used if there is no alternative approach using \cs{msg_error:nnnnnn}
% or other non-expandable commands from the previous section.  Despite
% having a similar interface as non-expandable messages, expandable
% errors must be handled internally very differently from normal error
% messages, as none of the tools to print to the terminal or the log
% file are expandable.  As a result, the message text and arguments are
% not expanded, and messages must be very short (with default settings,
% they are truncated after approximately 50 characters).  It is
% advisable to ensure that the message is understandable even when
% truncated.  Another particularity of expandable messages is that they
% cannot be redirected or turned off by the user.
%
% \begin{function}[EXP, added = 2015-08-06]
%   {
%     \msg_expandable_error:nnnnnn ,
%     \msg_expandable_error:nnnnn  ,
%     \msg_expandable_error:nnnn   ,
%     \msg_expandable_error:nnn    ,
%     \msg_expandable_error:nn     ,
%     \msg_expandable_error:nnffff ,
%     \msg_expandable_error:nnfff  ,
%     \msg_expandable_error:nnff   ,
%     \msg_expandable_error:nnf    ,
%   }
%   \begin{syntax}
%     \cs{msg_expandable_error:nnnnnn} \Arg{module} \Arg{message} \Arg{arg one} \Arg{arg two} \Arg{arg three} \Arg{arg four}
%   \end{syntax}
%   Issues an \enquote{Undefined error} message from \TeX{} itself
%   using the undefined control sequence \cs{::error} then prints
%   \enquote{! \meta{module}: }\meta{error message}, which should be
%   short.  With default settings, anything beyond approximately $60$
%   characters long (or bytes in some engines) is cropped.  A leading
%   space might be removed as well.
% \end{function}
%
% \section{Additions to \pkg{l3prg}}
%
% Minimal (lazy) evaluation can be obtained using the conditionals
% \cs{bool_lazy_all:nTF}, \cs{bool_lazy_and:nnTF}, \cs{bool_lazy_any:nTF}, or
% \cs{bool_lazy_or:nnTF}, which only evaluate their boolean expression
% arguments when they are needed to determine the resulting truth
% value.  For example, when evaluating the boolean expression
% \begin{verbatim}
%     \bool_lazy_and_p:nn
%       {
%         \bool_lazy_any_p:n
%           {
%             { \int_compare_p:n { 2 = 3 } }
%             { \int_compare_p:n { 4 <= 4 } }
%             { \int_compare_p:n { 1 = \error } } % is skipped
%           }
%       }
%       { ! \int_compare_p:n { 2 = 4 } }
% \end{verbatim}
% the line marked with |is skipped| is not expanded because the result
% of \cs{bool_lazy_any_p:n} is known once the second boolean expression is
% found to be logically \texttt{true}.  On the other hand, the last
% line is expanded because its logical value is needed to determine the
% result of \cs{bool_lazy_and_p:nn}.
%   
% \begin{function}[EXP, pTF, added = 2015-11-15]{\bool_lazy_all:n}
%   \begin{syntax}
%     \cs{bool_lazy_all_p:n} \{ \Arg{boolexpr_1} \Arg{boolexpr_2} $\cdots$ \Arg{boolexpr_N} \}
%     \cs{bool_lazy_all:nTF} \{ \Arg{boolexpr_1} \Arg{boolexpr_2} $\cdots$ \Arg{boolexpr_N} \} \Arg{true code} \Arg{false code}
%   \end{syntax}
%   Implements the \enquote{And} operation on the \meta{boolean
%   expressions}, hence is \texttt{true} if all of them are
%   \texttt{true} and \texttt{false} if any of them is \texttt{false}.
%   Contrarily to the infix operator |&&|, only the \meta{boolean
%   expressions} which are needed to determine the result of
%   \cs{bool_lazy_all:nTF} will be evaluated.  See also \cs{bool_lazy_and:nnTF}
%   when there are only two \meta{boolean expressions}.
% \end{function}
%
% \begin{function}[EXP, pTF, added = 2015-11-15]{\bool_lazy_and:nn}
%   \begin{syntax}
%     \cs{bool_lazy_and_p:nn} \Arg{boolexpr_1} \Arg{boolexpr_2}
%     \cs{bool_lazy_and:nnTF} \Arg{boolexpr_1} \Arg{boolexpr_2} \Arg{true code} \Arg{false code}
%   \end{syntax}
%   Implements the \enquote{And} operation between two boolean
%   expressions, hence is \texttt{true} if both are \texttt{true}.
%   Contrarily to the infix operator |&&|, the \meta{boolexpr_2} will
%   only be evaluated if it is needed to determine the result of
%   \cs{bool_lazy_and:nnTF}.  See also \cs{bool_lazy_all:nTF} when there are more
%   than two \meta{boolean expressions}.
% \end{function}
%
% \begin{function}[EXP, pTF, added = 2015-11-15]{\bool_lazy_any:n}
%   \begin{syntax}
%     \cs{bool_lazy_any_p:n} \{ \Arg{boolexpr_1} \Arg{boolexpr_2} $\cdots$ \Arg{boolexpr_N} \}
%     \cs{bool_lazy_any:nTF} \{ \Arg{boolexpr_1} \Arg{boolexpr_2} $\cdots$ \Arg{boolexpr_N} \} \Arg{true code} \Arg{false code}
%   \end{syntax}
%   Implements the \enquote{Or} operation on the \meta{boolean
%   expressions}, hence is \texttt{true} if any of them is
%   \texttt{true} and \texttt{false} if all of them are \texttt{false}.
%   Contrarily to the infix operator \verb"||", only the \meta{boolean
%   expressions} which are needed to determine the result of
%   \cs{bool_lazy_any:nTF} will be evaluated.  See also \cs{bool_lazy_or:nnTF}
%   when there are only two \meta{boolean expressions}.
% \end{function}
%
% \begin{function}[EXP, pTF, added = 2015-11-15]{\bool_lazy_or:nn}
%   \begin{syntax}
%     \cs{bool_lazy_or_p:nn} \Arg{boolexpr_1} \Arg{boolexpr_2}
%     \cs{bool_lazy_or:nnTF} \Arg{boolexpr_1} \Arg{boolexpr_2} \Arg{true code} \Arg{false code}
%   \end{syntax}
%   Implements the \enquote{Or} operation between two boolean
%   expressions, hence is \texttt{true} if either one is \texttt{true}.
%   Contrarily to the infix operator \verb"||", the \meta{boolexpr_2}
%   will only be evaluated if it is needed to determine the result of
%   \cs{bool_lazy_or:nnTF}.  See also \cs{bool_lazy_any:nTF} when there are more
%   than two \meta{boolean expressions}.
% \end{function}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-03]{\bool_log:N, \bool_log:c}
%   \begin{syntax}
%     \cs{bool_log:N} \meta{boolean}
%   \end{syntax}
%   Writes the logical truth of the \meta{boolean} in the log file.
% \end{function}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-07]{\bool_log:n}
%   \begin{syntax}
%     \cs{bool_log:n} \Arg{boolean expression}
%   \end{syntax}
%   Writes the logical truth of the \meta{boolean expression} in the log
%   file.
% \end{function}
%
% \section{Additions to \pkg{l3prop}}
%
% \begin{function}[EXP]{\prop_count:N, \prop_count:c}
%   \begin{syntax}
%     \cs{prop_count:N} \meta{property list}
%   \end{syntax}
%   Leaves the number of key--value pairs in the \meta{property list} in
%   the input stream as an \meta{integer denotation}.
% \end{function}
%
% \begin{function}[rEXP]
%   {\prop_map_tokens:Nn, \prop_map_tokens:cn}
%   \begin{syntax}
%     \cs{prop_map_tokens:Nn} \meta{property list} \Arg{code}
%   \end{syntax}
%   Analogue of \cs{prop_map_function:NN} which maps several tokens
%   instead of a single function.  The \meta{code} receives each
%   key--value pair in the \meta{property list} as two trailing brace
%   groups. For instance,
%   \begin{verbatim}
%     \prop_map_tokens:Nn \l_my_prop { \str_if_eq:nnT { mykey } }
%   \end{verbatim}
%   will expand to the value corresponding to \texttt{mykey}: for each
%   pair in |\l_my_prop| the function \cs{str_if_eq:nnT} receives
%   \texttt{mykey}, the \meta{key} and the \meta{value} as its three
%   arguments.  For that specific task, \cs{prop_item:Nn} is faster.
% \end{function}
%
% \begin{function}[added = 2014-08-12]{\prop_log:N, \prop_log:c}
%   \begin{syntax}
%     \cs{prop_log:N} \meta{property list}
%   \end{syntax}
%   Writes the entries in the \meta{property list} in the log file.
% \end{function}
%
% \section{Additions to \pkg{l3seq}}
%
% \begin{function}[rEXP]
%   {
%     \seq_mapthread_function:NNN, \seq_mapthread_function:NcN,
%     \seq_mapthread_function:cNN, \seq_mapthread_function:ccN
%   }
%   \begin{syntax}
%     \cs{seq_mapthread_function:NNN} \meta{seq_1} \meta{seq_2} \meta{function}
%   \end{syntax}
%   Applies \meta{function} to every pair of items
%   \meta{seq_1-item}--\meta{seq_2-item} from the two sequences, returning
%   items from both sequences from left to right.   The \meta{function} will
%   receive two \texttt{n}-type arguments for each iteration. The  mapping
%   will terminate when
%   the end of either sequence is reached (\emph{i.e.}~whichever sequence has
%   fewer items determines how many iterations
%   occur).
% \end{function}
%
% \begin{function}{\seq_set_filter:NNn, \seq_gset_filter:NNn}
%   \begin{syntax}
%     \cs{seq_set_filter:NNn} \meta{sequence_1} \meta{sequence_2} \Arg{inline boolexpr}
%   \end{syntax}
%   Evaluates the \meta{inline boolexpr} for every \meta{item} stored
%   within the \meta{sequence_2}. The \meta{inline boolexpr} will
%   receive the \meta{item} as |#1|. The sequence of all \meta{items}
%   for which the \meta{inline boolexpr} evaluated to \texttt{true}
%   is assigned to \meta{sequence_1}.
%   \begin{texnote}
%     Contrarily to other mapping functions, \cs{seq_map_break:} cannot
%     be used in this function, and will lead to low-level \TeX{} errors.
%   \end{texnote}
% \end{function}
%
% \begin{function}[added = 2011-12-22]
%   {\seq_set_map:NNn, \seq_gset_map:NNn}
%   \begin{syntax}
%     \cs{seq_set_map:NNn} \meta{sequence_1} \meta{sequence_2} \Arg{inline function}
%   \end{syntax}
%   Applies \meta{inline function} to every \meta{item} stored
%   within the \meta{sequence_2}. The \meta{inline function} should
%   consist of code which will receive the \meta{item} as |#1|.
%   The sequence resulting from \texttt{x}-expanding
%   \meta{inline function} applied to each \meta{item}
%   is assigned to \meta{sequence_1}. As such, the code
%   in \meta{inline function} should be expandable.
%   \begin{texnote}
%     Contrarily to other mapping functions, \cs{seq_map_break:} cannot
%     be used in this function, and will lead to low-level \TeX{} errors.
%   \end{texnote}
% \end{function}
%
% \begin{function}[added = 2014-08-12]{\seq_log:N, \seq_log:c}
%   \begin{syntax}
%     \cs{seq_log:N} \meta{sequence}
%   \end{syntax}
%   Writes the entries in the \meta{sequence} in the log file.
% \end{function}
%
% \section{Additions to \pkg{l3skip}}
%
% \begin{function}{\skip_split_finite_else_action:nnNN}
%   \begin{syntax}
%     \cs{skip_split_finite_else_action:nnNN} \Arg{skipexpr} \Arg{action}
%     ~~\meta{dimen_1} \meta{dimen_2}
%   \end{syntax}
%   Checks if the \meta{skipexpr} contains finite glue. If it does then it
%   assigns
%   \meta{dimen_1} the stretch component and \meta{dimen_2} the shrink
%   component. If
%   it contains infinite glue set \meta{dimen_1} and \meta{dimen_2} to $0$\,pt
%   and place |#2| into the input stream: this is usually an error or
%   warning message of some sort.
% \end{function}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-03]{\dim_log:N, \dim_log:c}
%   \begin{syntax}
%     \cs{dim_log:N} \meta{dimension}
%    \end{syntax}
%   Writes the value of the \meta{dimension} in the log file.
% \end{function}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-07]{\dim_log:n}
%   \begin{syntax}
%     \cs{dim_log:n} \Arg{dimension expression}
%    \end{syntax}
%   Writes the result of evaluating the \meta{dimension expression}
%   in the log file.
% \end{function}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-03]{\skip_log:N, \skip_log:c}
%   \begin{syntax}
%     \cs{skip_log:N} \meta{skip}
%    \end{syntax}
%   Writes the value of the \meta{skip} in the log file.
% \end{function}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-07]{\skip_log:n}
%   \begin{syntax}
%     \cs{skip_log:n} \Arg{skip expression}
%    \end{syntax}
%   Writes the result of evaluating the \meta{skip expression}
%   in the log file.
% \end{function}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-03]{\muskip_log:N, \muskip_log:c}
%   \begin{syntax}
%     \cs{muskip_log:N} \meta{muskip}
%    \end{syntax}
%   Writes the value of the \meta{muskip} in the log file.
% \end{function}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-07]{\muskip_log:n}
%   \begin{syntax}
%     \cs{muskip_log:n} \Arg{muskip expression}
%    \end{syntax}
%   Writes the result of evaluating the \meta{muskip expression}
%   in the log file.
% \end{function}
%
% \section{Additions to \pkg{l3tl}}
%
% \begin{function}[EXP,pTF]{\tl_if_single_token:n}
%   \begin{syntax}
%   \cs{tl_if_single_token_p:n} \Arg{token list}
%   \cs{tl_if_single_token:nTF} \Arg{token list} \Arg{true code} \Arg{false code}
%   \end{syntax}
%   Tests if the token list consists of exactly one token, \emph{i.e.}~is
%   either a single space character or a single \enquote{normal} token.
%   Token groups (|{|\ldots|}|) are not single tokens.
% \end{function}
%
% \begin{function}[EXP]{\tl_reverse_tokens:n}
%   \begin{syntax}
%     \cs{tl_reverse_tokens:n} \Arg{tokens}
%   \end{syntax}
%   This function, which works directly on \TeX{} tokens, reverses
%   the order of the \meta{tokens}: the first will be the last and
%   the last will become first. Spaces are preserved. The reversal
%   also operates within brace groups, but the braces themselves
%   are not exchanged, as this would lead to an unbalanced token
%   list. For instance, \cs{tl_reverse_tokens:n} |{a~{b()}}|
%   leaves |{)(b}~a| in the input stream. This function requires
%   two steps of expansion.
%   \begin{texnote}
%     The result is returned within the \tn{unexpanded}
%     primitive (\cs{exp_not:n}), which means that the token
%     list will not expand further when appearing in an \texttt{x}-type
%     argument expansion.
%   \end{texnote}
% \end{function}
%
% \begin{function}[EXP]{\tl_count_tokens:n}
%   \begin{syntax}
%     \cs{tl_count_tokens:n} \Arg{tokens}
%   \end{syntax}
%   Counts the number of \TeX{} tokens in the \meta{tokens} and leaves
%   this information in the input stream. Every token, including spaces and
%   braces, contributes one to the total; thus for instance, the token count of
%   |a~{bc}| is $6$.
%   This function requires three expansions,
%   giving an \meta{integer denotation}.
% \end{function}
%
% \begin{function}[EXP, added = 2014-06-30, updated = 2016-01-12]
%   {
%     \tl_lower_case:n,  \tl_upper_case:n,  \tl_mixed_case:n,
%     \tl_lower_case:nn, \tl_upper_case:nn, \tl_mixed_case:nn
%   }
%   \begin{syntax}
%     \cs{tl_upper_case:n}  \Arg{tokens}
%     \cs{tl_upper_case:nn} \Arg{language} \Arg{tokens}
%   \end{syntax}
%   These functions are intended to be applied to input which may be
%   regarded broadly as \enquote{text}. They traverse the \meta{tokens} and
%   change the case of characters as discussed below. The character code of
%   the characters replaced may be arbitrary: the replacement characters will
%   have standard document-level category codes ($11$ for letters, $12$ for
%   letter-like characters which can also be case-changed).  Begin-group and
%   end-group characters in the \meta{tokens} are normalized and become |{|
%   and |}|, respectively.
%
%   Importantly, notice that these functions are intended for working with
%   user text for typesetting. For case changing programmatic data see the
%   \pkg{l3str} module and discussion there of \cs{str_lower_case:n},
%   \cs{str_upper_case:n} and \cs{str_fold_case:n}.
% \end{function}
%
% The functions perform expansion on the input in most cases. In particular,
% input in the form of token lists or expandable functions will be expanded
% \emph{unless} it falls within one of the special handling classes described
% below. This expansion approach means that in general the result of case
% changing will match the \enquote{natural} outcome expected from a
% \enquote{functional} approach to case modification. For example
% \begin{verbatim}
%   \tl_set:Nn \l_tmpa_tl { hello }
%   \tl_upper_case:n { \l_tmpa_tl \c_space_tl world }
% \end{verbatim}
% will produce
% \begin{verbatim}
%   HELLO WORLD
% \end{verbatim}
% The expansion approach taken means that in package mode any \LaTeXe{}
% \enquote{robust} commands which may appear in the input should be converted
% to engine-protected versions using for example the \tn{robustify} command
% from the \pkg{etoolbox} package.
%
% \begin{variable}{\l_tl_case_change_math_tl}
%   Case changing will not take place within math mode material so for example
%   \begin{verbatim}
%     \tl_upper_case:n { Some~text~$y = mx + c$~with~{Braces} }
%   \end{verbatim}
%   will become
%   \begin{verbatim}
%     SOME TEXT $y = mx + c$ WITH {BRACES}
%   \end{verbatim}
%   Material inside math mode is left entirely unchanged: in particular, no
%   expansion is undertaken.
%
%   Detection of math mode is controlled by the list of tokens in
%   \cs{l_tl_case_change_math_tl}, which should be in open--close pairs. In
%   package mode the standard settings is
%   \begin{verbatim}
%     $ $ \( \)
%   \end{verbatim}
%
%   Note that while expansion occurs when searching the text it does not
%   apply to math mode material (which should be unaffected by case changing).
%   As such, whilst the opening token for math mode may be \enquote{hidden}
%   inside a command/macro, the closing one cannot be as this is being
%   searched for in math mode. Typically, in the types of \enquote{text}
%   the case changing functions are intended to apply to this should not be
%   an issue.
% \end{variable}
%
% \begin{variable}{\l_tl_case_change_exclude_tl}
%   Case changing can be prevented by using any command on the list
%   \cs{l_tl_case_change_exclude_tl}. Each entry should be a function
%   to be followed by one argument: the latter will be preserved as-is
%   with no expansion. Thus for example following
%   \begin{verbatim}
%     \tl_put_right:Nn \l_tl_case_change_exclude_tl { \NoChangeCase }
%   \end{verbatim}
%   the input
%   \begin{verbatim}
%     \tl_upper_case:n
%       { Some~text~$y = mx + c$~with~\NoChangeCase {Protection} }
%   \end{verbatim}
%   will result in
%   \begin{verbatim}
%     SOME TEXT $y = mx + c$ WITH \NoChangeCase {Protection}
%   \end{verbatim}
%   Notice that the case changing mapping preserves the inclusion of
%   the escape functions: it is left to other code to provide suitable
%   definitions (typically equivalent to \cs{use:n}). In particular, the
%   result of case changing is returned protected by \cs{exp_not:n}.
%
%   When used with \LaTeXe{} the commands |\cite|, |\ensuremath|, |\label|
%   and |\ref| are automatically included in the list for exclusion from
%   case changing.
% \end{variable}
%
% \begin{variable}{\l_tl_case_change_accents_tl}
%   This list specifies accent commands which should be left unexpanded
%   in the output. This allows for example
%   \begin{verbatim}
%     \tl_upper_case:n { \" { a } }
%   \end{verbatim}
%   to yield
%   \begin{verbatim}
%     \" { A }
%   \end{verbatim}
%   irrespective of the expandability of |\"|.
%
%   The standard contents of this variable is |\"|, |\'|, |\.|, |\^|, |\`|,
%   |\~|, |\c|, |\H|, |\k|, |\r|, |\t|, |\u| and |\v|.
% \end{variable}
%
% \enquote{Mixed} case conversion may be regarded informally as converting the
% first character of the \meta{tokens} to upper case and the rest to lower
% case. However, the process is more complex than this as there are some
% situations where a single lower case character maps to a special form, for
% example \texttt{ij} in Dutch which becomes \texttt{IJ}. As such,
% \cs[index=tl_mixed_case:n]{tl_mixed_case:n(n)}
% implement a more sophisticated mapping which accounts
% for this and for modifying accents on the first letter. Spaces at the start
% of the \meta{tokens} are ignored when finding the first \enquote{letter} for
% conversion.
% \begin{verbatim}
%   \tl_mixed_case:n { hello~WORLD }   % => "Hello world"
%   \tl_mixed_case:n { ~hello~WORLD }  % => " Hello world"
%   \tl_mixed_case:n { {hello}~WORLD } % => "{Hello} world"
% \end{verbatim}
% When finding the first \enquote{letter} for this process, any content in
% math mode or covered by \cs{l_tl_case_change_exclude_tl} is ignored.
%
% (Note that the Unicode Consortium describe this as \enquote{title case}, but
% that in English title case applies on a word-by-word basis. The
% \enquote{mixed} case implemented here is a lower level concept needed for
% both \enquote{title} and \enquote{sentence} casing of text.)
%
% \begin{variable}{\l_tl_mixed_case_ignore_tl}
%   The list of characters to ignore when searching for the first
%   \enquote{letter} in mixed-casing is determined by
%   \cs{l_tl_mixed_change_ignore_tl}. This has the standard setting
%   \begin{verbatim}
%     ( [ { ` -
%   \end{verbatim}
%   where comparisons are made on a character basis.
% \end{variable}
%
%   As is generally true for \pkg{expl3}, these functions are designed to
%   work with Unicode input only. As such, UTF-8 input is assumed for
%   \emph{all} engines. When used with \XeTeX{} or \LuaTeX{} a full range of
%   Unicode transformations are enabled. Specifically, the standard mappings
%   here follow those defined by the \href{http://www.unicode.org}^^A
%   {Unicode Consortium} in \texttt{UnicodeData.txt} and
%   \texttt{SpecialCasing.txt}. In the case of $8$-bit engines, mappings
%   are provided for characters which can be represented in output typeset
%   using the |T1| font encoding. Thus for example |ä| can be case-changed
%   using \pdfTeX{}.  For \pTeX{} only the ASCII range is covered as the
%   engine treats input outside of this range as east Asian.
%
% Context-sensitive mappings are enabled: language-dependent cases are
% discussed below. Context detection will expand input but treats any
% unexpandable control sequences as \enquote{failures} to match a context.
%
%   Language-sensitive conversions are enabled using the \meta{language}
%   argument, and follow Unicode Consortium guidelines. Currently, the
%   languages recognised for special handling are as follows.
%   \begin{itemize}
%     \item Azeri and Turkish (\texttt{az} and \texttt{tr}).
%       The case pairs I/i-dotless and I-dot/i are activated for these
%       languages. The combining dot mark is removed when lower
%       casing I-dot and introduced when upper casing i-dotless.
%     \item German (\texttt{de-alt}).
%       An alternative mapping for German in which the lower case
%       \emph{Eszett} maps to a \emph{gro\ss{}es Eszett}.
%     \item Lithuanian (\texttt{lt}).
%       The lower case letters i and j should retain a dot above when the
%       accents grave, acute or tilde are present. This is implemented for
%       lower casing of the relevant upper case letters both when input as
%       single Unicode codepoints and when using combining accents. The
%       combining dot is removed when upper casing in these cases. Note that
%       \emph{only} the accents used in Lithuanian are covered: the behaviour
%       of other accents are not modified.
%     \item Dutch (\texttt{nl}).
%       Capitalisation of \texttt{ij} at the beginning of mixed cased
%       input produces \texttt{IJ} rather than \texttt{Ij}. The output
%       retains two separate letters, thus this transformation \emph{is}
%       available using \pdfTeX{}.
%   \end{itemize}
%
%   Creating additional context-sensitive mappings requires knowledge
%   of the underlying mapping implementation used here. The team are happy
%   to add these to the kernel where they are well-documented
%   (\emph{e.g.}~in Unicode Consortium or relevant government publications).
%
% \begin{function}[added = 2014-06-25]
%   {
%     \tl_set_from_file:Nnn,  \tl_set_from_file:cnn,
%     \tl_gset_from_file:Nnn, \tl_gset_from_file:cnn
%   }
%   \begin{syntax}
%     \cs{tl_set_from_file:Nnn} \meta{tl} \Arg{setup} \Arg{filename}
%   \end{syntax}
%   Defines \meta{tl} to the contents of \meta{filename}.
%   Category codes may need to be set appropriately via the \meta{setup}
%   argument.
% \end{function}
%
% \begin{function}[added = 2014-06-25]
%   {
%     \tl_set_from_file_x:Nnn,  \tl_set_from_file_x:cnn,
%     \tl_gset_from_file_x:Nnn, \tl_gset_from_file_x:cnn
%   }
%   \begin{syntax}
%     \cs{tl_set_from_file_x:Nnn} \meta{tl} \Arg{setup} \Arg{filename}
%   \end{syntax}
%   Defines \meta{tl} to the contents of \meta{filename}, expanding
%   the contents of the file as it is read. Category codes and other
%   definitions may need to be set appropriately via the \meta{setup}
%   argument.
% \end{function}
%
% \begin{function}[added = 2014-08-22, updated = 2015-08-01]{\tl_log:N, \tl_log:c}
%   \begin{syntax}
%     \cs{tl_log:N} \meta{tl~var}
%   \end{syntax}
%   Writes the content of the \meta{tl~var} in the log file.  See also
%   \cs{tl_show:N} which displays the result in the terminal.
% \end{function}
%
% \begin{function}[added = 2014-08-22]{\tl_log:n}
%   \begin{syntax}
%     \cs{tl_log:n} \meta{token list}
%   \end{syntax}
%   Writes the \meta{token list} in the log file.  See also
%   \cs{tl_show:n} which displays the result in the terminal.
% \end{function}
%
% \section{Additions to \pkg{l3tokens}}
%
% \begin{function}[TF, updated = 2012-12-20]{\peek_N_type:}
%   \begin{syntax}
%     \cs{peek_N_type:TF} \Arg{true code} \Arg{false code}
%   \end{syntax}
%   Tests if the next \meta{token} in the input stream can be safely
%   grabbed as an \texttt{N}-type argument. The test will be \meta{false}
%   if the next \meta{token} is either an explicit or implicit
%   begin-group or end-group token (with any character code), or
%   an explicit or implicit space character (with character code $32$
%   and category code $10$), or an outer token (never used in \LaTeX3)
%   and \meta{true} in all other cases.
%   Note that a \meta{true} result ensures that the next \meta{token} is
%   a valid \texttt{N}-type argument. However, if the next \meta{token}
%   is for instance \cs{c_space_token}, the test will take the
%   \meta{false} branch, even though the next \meta{token} is in fact
%   a valid \texttt{N}-type argument. The \meta{token} will be left
%   in the input stream after the \meta{true code} or \meta{false code}
%   (as appropriate to the result of the test).
% \end{function}
%
% \end{documentation}
%
% \begin{implementation}
%
% \section{\pkg{l3candidates} Implementation}
%
%    \begin{macrocode}
%<*initex|package>
%    \end{macrocode}
%
% \subsection{Additions to \pkg{l3basics}}
%
%    \begin{macrocode}
%<@@=cs>
%    \end{macrocode}
%
% \begin{macro}{\cs_log:N, \cs_log:c}
%   Use \cs{cs_show:N} or \cs{cs_show:c} after calling
%   \cs{__msg_log_next:} to redirect their output to the log file only.
%   Note that \cs{cs_log:c} is not just a variant of \cs{cs_log:N} as
%   the csname should be turned to a control sequence within a group
%   (see \cs{cs_show:c}).
%    \begin{macrocode}
\cs_new_protected:Npn \cs_log:N
  { \__msg_log_next: \cs_show:N }
\cs_new_protected:Npn \cs_log:c
  { \__msg_log_next: \cs_show:c }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[int]{\__kernel_register_log:N, \__kernel_register_log:c}
%   Redirect the output of \cs{__kernel_register_show:N} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \__kernel_register_log:N
  { \__msg_log_next: \__kernel_register_show:N }
\cs_generate_variant:Nn \__kernel_register_log:N { c }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3box}}
%
%    \begin{macrocode}
%<@@=box>
%    \end{macrocode}
%
% \subsection{Affine transformations}
%
% \begin{variable}{\l_@@_angle_fp}
%   When rotating boxes, the angle itself may be needed by the
%   engine-dependent code. This is done using the \pkg{fp} module so
%   that the value is tidied up properly.
%    \begin{macrocode}
\fp_new:N \l_@@_angle_fp
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_@@_cos_fp, \l_@@_sin_fp}
%   These are used to hold the calculated sine and cosine values while
%   carrying out a rotation.
%    \begin{macrocode}
\fp_new:N \l_@@_cos_fp
\fp_new:N \l_@@_sin_fp
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}
%   {\l_@@_top_dim, \l_@@_bottom_dim, \l_@@_left_dim, \l_@@_right_dim}
%   These are the positions of the four edges of a box before
%   manipulation.
%    \begin{macrocode}
\dim_new:N \l_@@_top_dim
\dim_new:N \l_@@_bottom_dim
\dim_new:N \l_@@_left_dim
\dim_new:N \l_@@_right_dim
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}
%  {
%    \l_@@_top_new_dim,  \l_@@_bottom_new_dim ,
%    \l_@@_left_new_dim, \l_@@_right_new_dim
%  }
%   These are the positions of the four edges of a box after
%   manipulation.
%    \begin{macrocode}
\dim_new:N \l_@@_top_new_dim
\dim_new:N \l_@@_bottom_new_dim
\dim_new:N \l_@@_left_new_dim
\dim_new:N \l_@@_right_new_dim
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_@@_internal_box}
%   Scratch space, but also needed by some parts of the driver.
%    \begin{macrocode}
\box_new:N \l_@@_internal_box
%    \end{macrocode}
% \end{variable}
%
% \begin{macro}{\box_rotate:Nn}
% \begin{macro}[aux]{\@@_rotate:N}
% \begin{macro}[aux]{\@@_rotate_x:nnN, \@@_rotate_y:nnN}
% \begin{macro}[aux]
%   {
%     \@@_rotate_quadrant_one:,   \@@_rotate_quadrant_two:,
%     \@@_rotate_quadrant_three:, \@@_rotate_quadrant_four:
%   }
%   Rotation of a box starts with working out the relevant sine and
%   cosine. The actual rotation is in an auxiliary to keep the flow slightly
%   clearer
%    \begin{macrocode}
\cs_new_protected:Npn \box_rotate:Nn #1#2
  {
    \hbox_set:Nn #1
      {
        \group_begin:
          \fp_set:Nn \l_@@_angle_fp {#2}
          \fp_set:Nn \l_@@_sin_fp { sind ( \l_@@_angle_fp ) }
          \fp_set:Nn \l_@@_cos_fp { cosd ( \l_@@_angle_fp ) }
          \@@_rotate:N #1
        \group_end:
    }
  }
%    \end{macrocode}
%   The edges of the box are then recorded: the left edge will
%   always be at zero. Rotation of the four edges then takes place: this is
%   most efficiently done on a quadrant by quadrant basis.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_rotate:N #1
  {
    \dim_set:Nn \l_@@_top_dim    {  \box_ht:N #1 }
    \dim_set:Nn \l_@@_bottom_dim { -\box_dp:N #1 }
    \dim_set:Nn \l_@@_right_dim  {  \box_wd:N #1 }
    \dim_zero:N \l_@@_left_dim
%    \end{macrocode}
%   The next step is to work out the $x$ and $y$ coordinates of vertices of
%   the rotated box in relation to its original coordinates. The box can be
%   visualized with vertices $B$, $C$, $D$ and $E$ is illustrated
%   (Figure~\ref{fig:l3candidates:rotation}). The vertex $O$ is the reference point
%   on the baseline, and in this implementation is also the centre of rotation.
%   \begin{figure}
%     \centering
%     \setlength{\unitlength}{3pt}^^A
%     \begin{picture}(34,36)(12,44)
%       \thicklines
%       \put(20,52){\dashbox{1}(20,21){}}
%       \put(20,80){\line(0,-1){36}}
%       \put(12,58){\line(1, 0){34}}
%       \put(41,59){A}
%       \put(40,74){B}
%       \put(21,74){C}
%       \put(21,49){D}
%       \put(40,49){E}
%       \put(21,59){O}
%     \end{picture}
%     \caption{Co-ordinates of a box prior to rotation.}
%     \label{fig:l3candidates:rotation}
%   \end{figure}
%   The formulae are, for a point $P$ and angle $\alpha$:
%   \[
%     \begin{array}{l}
%       P'_x = P_x - O_x \\
%       P'_y = P_y - O_y \\
%       P''_x =  ( P'_x \cos(\alpha)) - ( P'_y \sin(\alpha) ) \\
%       P''_y =  ( P'_x \sin(\alpha)) + ( P'_y \cos(\alpha) ) \\
%       P'''_x = P''_x + O_x + L_x \\
%       P'''_y = P''_y + O_y
%    \end{array}
%   \]
%   The \enquote{extra} horizontal translation $L_x$ at the end is calculated
%   so that the leftmost point of the resulting box has $x$-coordinate $0$.
%   This is desirable as \TeX{} boxes must have the reference point at
%   the left edge of the box. (As $O$ is always $(0,0)$, this part of the
%   calculation is omitted here.)
%    \begin{macrocode}
    \fp_compare:nNnTF \l_@@_sin_fp > \c_zero_fp
      {
        \fp_compare:nNnTF \l_@@_cos_fp > \c_zero_fp
          { \@@_rotate_quadrant_one: }
          { \@@_rotate_quadrant_two: }
      }
      {
        \fp_compare:nNnTF \l_@@_cos_fp < \c_zero_fp
          { \@@_rotate_quadrant_three: }
          { \@@_rotate_quadrant_four: }
      }
%    \end{macrocode}
%   The position of the box edges are now known, but the box at this
%   stage be misplaced relative to the current \TeX{} reference point. So the
%   content of the box is moved such that the reference point of the
%   rotated box will be in the same place as the original.
%    \begin{macrocode}
    \hbox_set:Nn \l_@@_internal_box { \box_use:N #1 }
    \hbox_set:Nn \l_@@_internal_box
      {
        \tex_kern:D -\l_@@_left_new_dim
        \hbox:n
          {
            \__driver_box_use_rotate:Nn
              \l_@@_internal_box
              \l_@@_angle_fp
          }
      }
%    \end{macrocode}
%   Tidy up the size of the box so that the material is actually inside
%   the bounding box. The result can then be used to reset the original
%   box.
%    \begin{macrocode}
    \box_set_ht:Nn \l_@@_internal_box {  \l_@@_top_new_dim }
    \box_set_dp:Nn \l_@@_internal_box { -\l_@@_bottom_new_dim }
    \box_set_wd:Nn \l_@@_internal_box
      { \l_@@_right_new_dim - \l_@@_left_new_dim }
    \box_use:N \l_@@_internal_box
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%   These functions take a general point $(|#1|, |#2|)$ and rotate its
%   location about the origin, using the previously-set sine and cosine
%   values. Each function gives only one component of the location of the
%   updated point. This is because for rotation of a box each step needs
%   only one value, and so performance is gained by avoiding working
%   out both $x'$ and $y'$ at the same time. Contrast this with
%   the equivalent function in the \pkg{l3coffins} module, where both parts
%   are needed.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_rotate_x:nnN #1#2#3
  {
    \dim_set:Nn #3
      {
        \fp_to_dim:n
          {
              \l_@@_cos_fp * \dim_to_fp:n {#1}
            - \l_@@_sin_fp * \dim_to_fp:n {#2}
          }
      }
  }
\cs_new_protected:Npn \@@_rotate_y:nnN #1#2#3
  {
    \dim_set:Nn #3
      {
        \fp_to_dim:n
          {
              \l_@@_sin_fp * \dim_to_fp:n {#1}
            + \l_@@_cos_fp * \dim_to_fp:n {#2}
          }
      }
  }
%    \end{macrocode}
%   Rotation of the edges is done using a different formula for each
%   quadrant. In every case, the top and bottom edges only need the
%   resulting $y$-values, whereas the left and right edges need the
%   $x$-values. Each case is a question of picking out which corner
%   ends up at with the maximum top, bottom, left and right value. Doing
%   this by hand means a lot less calculating and avoids lots of
%   comparisons.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_rotate_quadrant_one:
  {
    \@@_rotate_y:nnN \l_@@_right_dim \l_@@_top_dim
      \l_@@_top_new_dim
    \@@_rotate_y:nnN \l_@@_left_dim  \l_@@_bottom_dim
      \l_@@_bottom_new_dim
    \@@_rotate_x:nnN \l_@@_left_dim  \l_@@_top_dim
      \l_@@_left_new_dim
    \@@_rotate_x:nnN \l_@@_right_dim \l_@@_bottom_dim
      \l_@@_right_new_dim
  }
\cs_new_protected:Npn \@@_rotate_quadrant_two:
  {
    \@@_rotate_y:nnN \l_@@_right_dim \l_@@_bottom_dim
      \l_@@_top_new_dim
    \@@_rotate_y:nnN \l_@@_left_dim  \l_@@_top_dim
      \l_@@_bottom_new_dim
    \@@_rotate_x:nnN \l_@@_right_dim  \l_@@_top_dim
      \l_@@_left_new_dim
    \@@_rotate_x:nnN \l_@@_left_dim   \l_@@_bottom_dim
      \l_@@_right_new_dim
  }
\cs_new_protected:Npn \@@_rotate_quadrant_three:
  {
    \@@_rotate_y:nnN \l_@@_left_dim  \l_@@_bottom_dim
      \l_@@_top_new_dim
    \@@_rotate_y:nnN \l_@@_right_dim \l_@@_top_dim
      \l_@@_bottom_new_dim
    \@@_rotate_x:nnN \l_@@_right_dim \l_@@_bottom_dim
      \l_@@_left_new_dim
    \@@_rotate_x:nnN \l_@@_left_dim   \l_@@_top_dim
      \l_@@_right_new_dim
  }
\cs_new_protected:Npn \@@_rotate_quadrant_four:
  {
    \@@_rotate_y:nnN \l_@@_left_dim  \l_@@_top_dim
      \l_@@_top_new_dim
    \@@_rotate_y:nnN \l_@@_right_dim \l_@@_bottom_dim
      \l_@@_bottom_new_dim
    \@@_rotate_x:nnN \l_@@_left_dim  \l_@@_bottom_dim
      \l_@@_left_new_dim
    \@@_rotate_x:nnN \l_@@_right_dim \l_@@_top_dim
      \l_@@_right_new_dim
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{variable}{\l_@@_scale_x_fp, \l_@@_scale_y_fp}
%   Scaling is potentially-different in the two axes.
%    \begin{macrocode}
\fp_new:N \l_@@_scale_x_fp
\fp_new:N \l_@@_scale_y_fp
%    \end{macrocode}
% \end{variable}
%
% \begin{macro}{\box_resize:Nnn, \box_resize:cnn}
% \begin{macro}[aux]{\@@_resize_set_corners:N}
% \begin{macro}[aux]{\@@_resize:N}
% \begin{macro}[aux]{\@@_resize:NNN}
%   Resizing a box starts by working out the various dimensions of the
%   existing box.
%    \begin{macrocode}
\cs_new_protected:Npn \box_resize:Nnn #1#2#3
  {
    \hbox_set:Nn #1
      {
        \group_begin:
          \@@_resize_set_corners:N #1
%    \end{macrocode}
%   The $x$-scaling and resulting box size is easy enough to work
%   out: the dimension is that given as |#2|, and the scale is simply the
%   new width divided by the old one.
%    \begin{macrocode}
          \fp_set:Nn \l_@@_scale_x_fp
            { \dim_to_fp:n {#2} / \dim_to_fp:n { \l_@@_right_dim } }
%    \end{macrocode}
%   The $y$-scaling needs both the height and the depth of the current box.
%    \begin{macrocode}
          \fp_set:Nn \l_@@_scale_y_fp
            {
                \dim_to_fp:n {#3}
              / \dim_to_fp:n { \l_@@_top_dim - \l_@@_bottom_dim }
            }
%    \end{macrocode}
%   Hand off to the auxiliary which does the rest of the work.
%    \begin{macrocode}
          \@@_resize:N #1
        \group_end:
      }
  }
\cs_generate_variant:Nn \box_resize:Nnn { c }
\cs_new_protected:Npn \@@_resize_set_corners:N #1
  {
    \dim_set:Nn \l_@@_top_dim    {  \box_ht:N #1 }
    \dim_set:Nn \l_@@_bottom_dim { -\box_dp:N #1 }
    \dim_set:Nn \l_@@_right_dim  {  \box_wd:N #1 }
    \dim_zero:N \l_@@_left_dim
  }
%    \end{macrocode}
%   With at least one real scaling to do, the next phase is to find the new
%   edge co-ordinates. In the $x$~direction this is relatively easy: just
%   scale the right edge. In the $y$~direction, both dimensions have to be
%   scaled, and this again needs the absolute scale value.
%   Once that is all done, the common resize/rescale code can be employed.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_resize:N #1
  {
    \@@_resize:NNN \l_@@_right_new_dim
      \l_@@_scale_x_fp \l_@@_right_dim
    \@@_resize:NNN \l_@@_bottom_new_dim
      \l_@@_scale_y_fp \l_@@_bottom_dim
    \@@_resize:NNN \l_@@_top_new_dim
      \l_@@_scale_y_fp \l_@@_top_dim
    \@@_resize_common:N #1
  }
\cs_new_protected:Npn \@@_resize:NNN #1#2#3
  {
    \dim_set:Nn #1
      { \fp_to_dim:n { \fp_abs:n { #2 } * \dim_to_fp:n { #3 } } }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\box_resize_to_ht:Nn, \box_resize_to_ht:cn}
% \begin{macro}{\box_resize_to_ht_plus_dp:Nn, \box_resize_to_ht_plus_dp:cn}
% \begin{macro}{\box_resize_to_wd:Nn, \box_resize_to_wd:cn}
% \begin{macro}{\box_resize_to_wd_and_ht:Nnn, \box_resize_to_wd_and_ht:cnn}
%   Scaling to a (total) height or to a width is a simplified version of the main
%   resizing operation, with the scale simply copied between the two parts. The
%   internal auxiliary is called using the scaling value twice, as the sign for
%   both parts is needed (as this allows the same internal code to be used as
%   for the general case).
%    \begin{macrocode}
\cs_new_protected:Npn \box_resize_to_ht:Nn #1#2
  {
    \hbox_set:Nn #1
      {
        \group_begin:
          \@@_resize_set_corners:N #1
          \fp_set:Nn \l_@@_scale_y_fp
            {
                \dim_to_fp:n {#2}
              / \dim_to_fp:n { \l_@@_top_dim }
            }
          \fp_set_eq:NN \l_@@_scale_x_fp \l_@@_scale_y_fp
          \@@_resize:N #1
        \group_end:
      }
  }
\cs_generate_variant:Nn \box_resize_to_ht:Nn { c }
\cs_new_protected:Npn \box_resize_to_ht_plus_dp:Nn #1#2
  {
    \hbox_set:Nn #1
      {
        \group_begin:
          \@@_resize_set_corners:N #1
          \fp_set:Nn \l_@@_scale_y_fp
            {
                \dim_to_fp:n {#2}
              / \dim_to_fp:n { \l_@@_top_dim - \l_@@_bottom_dim }
            }
          \fp_set_eq:NN \l_@@_scale_x_fp \l_@@_scale_y_fp
          \@@_resize:N #1
        \group_end:
      }
  }
\cs_generate_variant:Nn \box_resize_to_ht_plus_dp:Nn { c }
\cs_new_protected:Npn \box_resize_to_wd:Nn #1#2
  {
    \hbox_set:Nn #1
      {
        \group_begin:
          \@@_resize_set_corners:N #1
          \fp_set:Nn \l_@@_scale_x_fp
            { \dim_to_fp:n {#2} / \dim_to_fp:n { \l_@@_right_dim } }
          \fp_set_eq:NN \l_@@_scale_y_fp \l_@@_scale_x_fp
          \@@_resize:N #1
        \group_end:
      }
  }
\cs_generate_variant:Nn \box_resize_to_wd:Nn { c }
\cs_new_protected:Npn \box_resize_to_wd_and_ht:Nnn #1#2#3
  {
    \hbox_set:Nn #1
      {
        \group_begin:
          \@@_resize_set_corners:N #1
          \fp_set:Nn \l_@@_scale_x_fp
            { \dim_to_fp:n {#2} / \dim_to_fp:n { \l_@@_right_dim } }
          \fp_set:Nn \l_@@_scale_y_fp
            {
                \dim_to_fp:n {#3}
              / \dim_to_fp:n { \l_@@_top_dim }
            }
          \@@_resize:N #1
        \group_end:
      }
  }
\cs_generate_variant:Nn \box_resize_to_wd_and_ht:Nnn { c }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\box_scale:Nnn, \box_scale:cnn}
%   When scaling a box, setting the scaling itself is easy enough. The
%   new dimensions are also relatively easy to find, allowing only for
%   the need to keep them positive in all cases. Once that is done then
%   after a check for the trivial scaling a hand-off can be made to the
%   common code. The dimension scaling operations are carried out using
%   the \TeX{} mechanism as it avoids needing to use too many \texttt{fp}
%   operations.
%    \begin{macrocode}
\cs_new_protected:Npn \box_scale:Nnn #1#2#3
  {
    \hbox_set:Nn #1
      {
        \group_begin:
          \fp_set:Nn \l_@@_scale_x_fp {#2}
          \fp_set:Nn \l_@@_scale_y_fp {#3}
          \dim_set:Nn \l_@@_top_dim    {  \box_ht:N #1 }
          \dim_set:Nn \l_@@_bottom_dim { -\box_dp:N #1 }
          \dim_set:Nn \l_@@_right_dim  {  \box_wd:N #1 }
          \dim_zero:N \l_@@_left_dim
          \dim_set:Nn \l_@@_top_new_dim
            { \fp_abs:n { \l_@@_scale_y_fp } \l_@@_top_dim }
          \dim_set:Nn \l_@@_bottom_new_dim
            { \fp_abs:n { \l_@@_scale_y_fp } \l_@@_bottom_dim }
          \dim_set:Nn \l_@@_right_new_dim
              { \fp_abs:n { \l_@@_scale_x_fp } \l_@@_right_dim }
           \@@_resize_common:N #1
        \group_end:
      }
  }
\cs_generate_variant:Nn \box_scale:Nnn { c }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[aux]{\@@_resize_common:N}
%   The main resize function places in input into a box which will start
%   of with zero width, and includes the handles for engine rescaling.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_resize_common:N #1
  {
    \hbox_set:Nn \l_@@_internal_box
      {
        \__driver_box_use_scale:Nnn
          #1
          \l_@@_scale_x_fp
          \l_@@_scale_y_fp
      }
%    \end{macrocode}
%   The new height and depth can be applied directly.
%    \begin{macrocode}
    \fp_compare:nNnTF \l_@@_scale_y_fp > \c_zero_fp
      {
        \box_set_ht:Nn \l_@@_internal_box { \l_@@_top_new_dim }
        \box_set_dp:Nn \l_@@_internal_box { -\l_@@_bottom_new_dim }
      }
      {
        \box_set_dp:Nn \l_@@_internal_box { \l_@@_top_new_dim }
        \box_set_ht:Nn \l_@@_internal_box { -\l_@@_bottom_new_dim }
      }
%    \end{macrocode}
%   Things are not quite as obvious for the width, as the reference point
%   needs to remain unchanged. For positive scaling factors resizing the
%   box is all that is needed. However, for case of a negative scaling
%   the material must be shifted such that the reference point ends up in
%   the right place.
%    \begin{macrocode}
    \fp_compare:nNnTF \l_@@_scale_x_fp < \c_zero_fp
      {
        \hbox_to_wd:nn { \l_@@_right_new_dim }
          {
            \tex_kern:D \l_@@_right_new_dim
            \box_use:N \l_@@_internal_box
            \tex_hss:D
          }
      }
      {
        \box_set_wd:Nn \l_@@_internal_box { \l_@@_right_new_dim }
        \hbox:n
          {
            \tex_kern:D \c_zero_dim
            \box_use:N \l_@@_internal_box
            \tex_hss:D
          }
      }
  }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Viewing part of a box}
%
% \begin{macro}{\box_clip:N, \box_clip:c}
%   A wrapper around the driver-dependent code.
%    \begin{macrocode}
\cs_new_protected:Npn \box_clip:N #1
  { \hbox_set:Nn #1 { \__driver_box_use_clip:N #1 } }
\cs_generate_variant:Nn \box_clip:N { c }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\box_trim:Nnnnn, \box_trim:cnnnn}
%   Trimming from the left- and right-hand edges of the box is easy: kern the
%   appropriate parts off each side.
%    \begin{macrocode}
\cs_new_protected:Npn \box_trim:Nnnnn #1#2#3#4#5
  {
    \hbox_set:Nn \l_@@_internal_box
      {
        \tex_kern:D -\__dim_eval:w #2 \__dim_eval_end:
        \box_use:N #1
        \tex_kern:D -\__dim_eval:w #4 \__dim_eval_end:
      }
%    \end{macrocode}
%   For the height and depth, there is a need to watch the baseline is
%   respected. Material always has to stay on the correct side, so trimming
%   has to check that there is enough material to trim. First, the bottom
%   edge. If there is enough depth, simply set the depth, or if not move
%   down so the result is zero depth. \cs{box_move_down:nn} is used in both
%   cases so the resulting box always contains a \tn{lower} primitive.
%   The internal box is used here as it allows safe use of \cs{box_set_dp:Nn}.
%    \begin{macrocode}
    \dim_compare:nNnTF { \box_dp:N #1 } > {#3}
      {
        \hbox_set:Nn \l_@@_internal_box
          {
            \box_move_down:nn \c_zero_dim
              { \box_use:N \l_@@_internal_box }
          }
        \box_set_dp:Nn \l_@@_internal_box { \box_dp:N #1 - (#3) }
      }
      {
        \hbox_set:Nn \l_@@_internal_box
          {
            \box_move_down:nn { #3 - \box_dp:N #1 }
              { \box_use:N \l_@@_internal_box }
          }
        \box_set_dp:Nn \l_@@_internal_box \c_zero_dim
      }
%    \end{macrocode}
%   Same thing, this time from the top of the box.
%    \begin{macrocode}
    \dim_compare:nNnTF { \box_ht:N \l_@@_internal_box } > {#5}
      {
        \hbox_set:Nn \l_@@_internal_box
          {
            \box_move_up:nn \c_zero_dim
              { \box_use:N \l_@@_internal_box }
          }
        \box_set_ht:Nn \l_@@_internal_box
          { \box_ht:N \l_@@_internal_box - (#5) }
      }
      {
        \hbox_set:Nn \l_@@_internal_box
          {
            \box_move_up:nn { #5 - \box_ht:N \l_@@_internal_box }
              { \box_use:N \l_@@_internal_box }
          }
        \box_set_ht:Nn \l_@@_internal_box \c_zero_dim
      }
    \box_set_eq:NN #1 \l_@@_internal_box
  }
\cs_generate_variant:Nn \box_trim:Nnnnn { c }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\box_viewport:Nnnnn, \box_viewport:cnnnn}
%   The same general logic as for the trim operation, but with absolute
%   dimensions. As a result, there are some things to watch out for in the
%   vertical direction.
%    \begin{macrocode}
\cs_new_protected:Npn \box_viewport:Nnnnn #1#2#3#4#5
  {
    \hbox_set:Nn \l_@@_internal_box
      {
        \tex_kern:D -\__dim_eval:w #2 \__dim_eval_end:
        \box_use:N #1
        \tex_kern:D \__dim_eval:w #4 - \box_wd:N #1 \__dim_eval_end:
      }
    \dim_compare:nNnTF {#3} < \c_zero_dim
      {
        \hbox_set:Nn \l_@@_internal_box
          {
            \box_move_down:nn \c_zero_dim
              { \box_use:N \l_@@_internal_box }
          }
        \box_set_dp:Nn \l_@@_internal_box { -\dim_eval:n {#3} }
      }
      {
        \hbox_set:Nn \l_@@_internal_box
          { \box_move_down:nn {#3} { \box_use:N \l_@@_internal_box } }
        \box_set_dp:Nn \l_@@_internal_box \c_zero_dim
      }
    \dim_compare:nNnTF {#5} > \c_zero_dim
      {
        \hbox_set:Nn \l_@@_internal_box
          {
            \box_move_up:nn \c_zero_dim
              { \box_use:N \l_@@_internal_box }
          }
        \box_set_ht:Nn \l_@@_internal_box
          {
            #5
            \dim_compare:nNnT {#3} > \c_zero_dim
              { - (#3) }
          }
      }
      {
        \hbox_set:Nn \l_@@_internal_box
          {
            \box_move_up:nn { -\dim_eval:n {#5} }
              { \box_use:N \l_@@_internal_box }
          }
        \box_set_ht:Nn \l_@@_internal_box \c_zero_dim
      }
    \box_set_eq:NN #1 \l_@@_internal_box
  }
\cs_generate_variant:Nn \box_viewport:Nnnnn { c }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3clist}}
%
%    \begin{macrocode}
%<@@=clist>
%    \end{macrocode}
%
% \begin{macro}{\clist_log:N, \clist_log:c, \clist_log:n}
%   Redirect output of \cs{clist_show:N} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \clist_log:N
  { \__msg_log_next: \clist_show:N }
\cs_new_protected:Npn \clist_log:n
  { \__msg_log_next: \clist_show:n }
\cs_generate_variant:Nn \clist_log:N { c }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3coffins}}
%
%    \begin{macrocode}
%<@@=coffin>
%    \end{macrocode}
%
% \subsection{Rotating coffins}
%
% \begin{variable}{\l_@@_sin_fp}
% \begin{variable}{\l_@@_cos_fp}
%   Used for rotations to get the sine and cosine values.
%    \begin{macrocode}
\fp_new:N \l_@@_sin_fp
\fp_new:N \l_@@_cos_fp
%    \end{macrocode}
% \end{variable}
% \end{variable}
%
% \begin{variable}{\l_@@_bounding_prop}
%   A property list for the bounding box of a coffin. This is only needed
%   during the rotation, so there is just the one.
%    \begin{macrocode}
\prop_new:N \l_@@_bounding_prop
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_@@_bounding_shift_dim}
%   The shift of the bounding box of a coffin from the real content.
%    \begin{macrocode}
\dim_new:N \l_@@_bounding_shift_dim
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_@@_left_corner_dim}
% \begin{variable}{\l_@@_right_corner_dim}
% \begin{variable}{\l_@@_bottom_corner_dim}
% \begin{variable}{\l_@@_top_corner_dim}
%   These are used to hold maxima for the various corner values: these
%   thus define the minimum size of the bounding box after rotation.
%    \begin{macrocode}
\dim_new:N \l_@@_left_corner_dim
\dim_new:N \l_@@_right_corner_dim
\dim_new:N \l_@@_bottom_corner_dim
\dim_new:N \l_@@_top_corner_dim
%    \end{macrocode}
% \end{variable}
% \end{variable}
% \end{variable}
% \end{variable}
%
% \begin{macro}{\coffin_rotate:Nn, \coffin_rotate:cn}
%   Rotating a coffin requires several steps which can be conveniently
%   run together. The sine and cosine of the angle in degrees are
%   computed.  This is then used to set \cs{l_@@_sin_fp} and
%   \cs{l_@@_cos_fp}, which are carried through unchanged for the rest
%   of the procedure.
%    \begin{macrocode}
\cs_new_protected:Npn \coffin_rotate:Nn #1#2
  {
    \fp_set:Nn \l_@@_sin_fp { sind ( #2 ) }
    \fp_set:Nn \l_@@_cos_fp { cosd ( #2 ) }
%    \end{macrocode}
%   The corners and poles of the coffin can now be rotated around the
%    origin. This is best achieved using mapping functions.
%    \begin{macrocode}
    \prop_map_inline:cn { l_@@_corners_ \__int_value:w #1 _prop }
      { \@@_rotate_corner:Nnnn #1 {##1} ##2 }
    \prop_map_inline:cn { l_@@_poles_ \__int_value:w #1 _prop }
      { \@@_rotate_pole:Nnnnnn #1 {##1} ##2 }
%    \end{macrocode}
%   The bounding box of the coffin needs to be rotated, and to do this
%   the corners have to be found first. They are then rotated in the same
%   way as the corners of the coffin material itself.
%    \begin{macrocode}
    \@@_set_bounding:N #1
    \prop_map_inline:Nn \l_@@_bounding_prop
      { \@@_rotate_bounding:nnn {##1} ##2 }
%    \end{macrocode}
%   At this stage, there needs to be a calculation to find where the
%   corners of the content and the box itself will end up.
%    \begin{macrocode}
    \@@_find_corner_maxima:N #1
    \@@_find_bounding_shift:
    \box_rotate:Nn #1 {#2}
%    \end{macrocode}
%   The correction of the box position itself takes place here. The idea
%   is that the bounding box for a coffin is tight up to the content, and
%   has the reference point at the bottom-left. The $x$-direction is
%   handled by moving the content by the difference in the positions of
%   the bounding box and the content left edge. The $y$-direction is
%   dealt with by moving the box down by any depth it has acquired. The
%   internal box is used here to allow for the next step.
%    \begin{macrocode}
    \hbox_set:Nn \l_@@_internal_box
      {
        \tex_kern:D
          \__dim_eval:w
            \l_@@_bounding_shift_dim - \l_@@_left_corner_dim
          \__dim_eval_end:
        \box_move_down:nn { \l_@@_bottom_corner_dim }
          { \box_use:N #1 }
      }
%    \end{macrocode}
%   If there have been any previous rotations then the size of the
%   bounding box will be bigger than the contents. This can be corrected
%   easily by setting the size of the box to the height and width of the
%   content. As this operation requires setting box dimensions and these
%   transcend grouping, the safe way to do this is to use the internal box
%   and to reset the result into the target box.
%    \begin{macrocode}
    \box_set_ht:Nn \l_@@_internal_box
      { \l_@@_top_corner_dim - \l_@@_bottom_corner_dim }
    \box_set_dp:Nn \l_@@_internal_box { 0 pt }
    \box_set_wd:Nn \l_@@_internal_box
      { \l_@@_right_corner_dim - \l_@@_left_corner_dim }
    \hbox_set:Nn #1 { \box_use:N \l_@@_internal_box }
%    \end{macrocode}
%   The final task is to move the poles and corners such that they are
%   back in alignment with the box reference point.
%    \begin{macrocode}
    \prop_map_inline:cn { l_@@_corners_ \__int_value:w #1 _prop }
      { \@@_shift_corner:Nnnn #1 {##1} ##2 }
    \prop_map_inline:cn { l_@@_poles_ \__int_value:w #1 _prop }
      { \@@_shift_pole:Nnnnnn #1 {##1} ##2 }
  }
\cs_generate_variant:Nn \coffin_rotate:Nn { c }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_set_bounding:N}
%   The bounding box corners for a coffin are easy enough to find: this
%   is the same code as for the corners of the material itself, but
%   using a dedicated property list.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_set_bounding:N #1
  {
    \prop_put:Nnx \l_@@_bounding_prop { tl }
      { { 0 pt } { \dim_eval:n { \box_ht:N #1 } } }
    \prop_put:Nnx \l_@@_bounding_prop { tr }
      { { \dim_eval:n { \box_wd:N #1 } } { \dim_eval:n { \box_ht:N #1 } } }
    \dim_set:Nn \l_@@_internal_dim { -\box_dp:N #1 }
    \prop_put:Nnx \l_@@_bounding_prop { bl }
      { { 0 pt } { \dim_use:N \l_@@_internal_dim } }
    \prop_put:Nnx \l_@@_bounding_prop { br }
      { { \dim_eval:n { \box_wd:N #1 } } { \dim_use:N \l_@@_internal_dim } }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_rotate_bounding:nnn}
% \begin{macro}{\@@_rotate_corner:Nnnn}
%   Rotating the position of the corner of the coffin is just a case
%   of treating this as a vector from the reference point. The same
%   treatment is used for the corners of the material itself and the
%   bounding box.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_rotate_bounding:nnn #1#2#3
  {
    \@@_rotate_vector:nnNN {#2} {#3} \l_@@_x_dim \l_@@_y_dim
    \prop_put:Nnx \l_@@_bounding_prop {#1}
      { { \dim_use:N \l_@@_x_dim } { \dim_use:N \l_@@_y_dim } }
  }
\cs_new_protected:Npn \@@_rotate_corner:Nnnn #1#2#3#4
  {
    \@@_rotate_vector:nnNN {#3} {#4} \l_@@_x_dim \l_@@_y_dim
    \prop_put:cnx { l_@@_corners_ \__int_value:w #1 _prop } {#2}
      { { \dim_use:N \l_@@_x_dim } { \dim_use:N \l_@@_y_dim } }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\@@_rotate_pole:Nnnnnn}
%   Rotating a single pole simply means shifting the co-ordinate of
%   the pole and its direction. The rotation here is about the bottom-left
%   corner of the coffin.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_rotate_pole:Nnnnnn #1#2#3#4#5#6
  {
    \@@_rotate_vector:nnNN {#3} {#4} \l_@@_x_dim \l_@@_y_dim
    \@@_rotate_vector:nnNN {#5} {#6}
      \l_@@_x_prime_dim \l_@@_y_prime_dim
    \@@_set_pole:Nnx #1 {#2}
      {
        { \dim_use:N \l_@@_x_dim } { \dim_use:N \l_@@_y_dim }
        { \dim_use:N \l_@@_x_prime_dim }
        { \dim_use:N \l_@@_y_prime_dim }
      }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_rotate_vector:nnNN}
%   A rotation function, which needs only an input vector (as dimensions)
%   and an output space. The values \cs{l_@@_cos_fp} and
%   \cs{l_@@_sin_fp} should previously have been set up correctly.
%   Working this way means that the floating point work is kept to a
%   minimum: for any given rotation the sin and cosine values do no
%   change, after all.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_rotate_vector:nnNN #1#2#3#4
  {
    \dim_set:Nn #3
      {
        \fp_to_dim:n
          {
              \dim_to_fp:n {#1} * \l_@@_cos_fp
            - \dim_to_fp:n {#2} * \l_@@_sin_fp
          }
      }
    \dim_set:Nn #4
      {
        \fp_to_dim:n
          {
              \dim_to_fp:n {#1} * \l_@@_sin_fp
            + \dim_to_fp:n {#2} * \l_@@_cos_fp
          }
      }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_find_corner_maxima:N}
% \begin{macro}[aux]{\@@_find_corner_maxima_aux:nn}
%   The idea here is to find the extremities of the content of the
%   coffin. This is done by looking for the smallest values for the bottom
%   and left corners, and the largest values for the top and right
%   corners. The values start at the maximum dimensions so that the
%   case where all are positive or all are negative works out correctly.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_find_corner_maxima:N #1
  {
    \dim_set:Nn \l_@@_top_corner_dim   { -\c_max_dim }
    \dim_set:Nn \l_@@_right_corner_dim { -\c_max_dim }
    \dim_set:Nn \l_@@_bottom_corner_dim { \c_max_dim }
    \dim_set:Nn \l_@@_left_corner_dim   { \c_max_dim }
    \prop_map_inline:cn { l_@@_corners_ \__int_value:w #1 _prop }
      { \@@_find_corner_maxima_aux:nn ##2 }
  }
\cs_new_protected:Npn \@@_find_corner_maxima_aux:nn #1#2
  {
    \dim_set:Nn \l_@@_left_corner_dim
     { \dim_min:nn { \l_@@_left_corner_dim } {#1} }
    \dim_set:Nn \l_@@_right_corner_dim
     { \dim_max:nn { \l_@@_right_corner_dim } {#1} }
    \dim_set:Nn \l_@@_bottom_corner_dim
     { \dim_min:nn { \l_@@_bottom_corner_dim } {#2} }
    \dim_set:Nn \l_@@_top_corner_dim
     { \dim_max:nn { \l_@@_top_corner_dim } {#2} }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\@@_find_bounding_shift:}
% \begin{macro}[aux]{\@@_find_bounding_shift_aux:nn}
%   The approach to finding the shift for the bounding box is similar to
%   that for the corners. However, there is only one value needed here and
%   a fixed input property list, so things are a bit clearer.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_find_bounding_shift:
  {
    \dim_set:Nn \l_@@_bounding_shift_dim { \c_max_dim }
    \prop_map_inline:Nn \l_@@_bounding_prop
      { \@@_find_bounding_shift_aux:nn ##2 }
  }
\cs_new_protected:Npn \@@_find_bounding_shift_aux:nn #1#2
  {
    \dim_set:Nn \l_@@_bounding_shift_dim
      { \dim_min:nn { \l_@@_bounding_shift_dim } {#1} }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\@@_shift_corner:Nnnn}
% \begin{macro}{\@@_shift_pole:Nnnnnn}
%   Shifting the corners and poles of a coffin means subtracting the
%   appropriate values from the $x$- and $y$-components. For
%   the poles, this means that the direction vector is unchanged.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_shift_corner:Nnnn #1#2#3#4
  {
    \prop_put:cnx { l_@@_corners_ \__int_value:w #1 _ prop } {#2}
      {
        { \dim_eval:n { #3 - \l_@@_left_corner_dim } }
        { \dim_eval:n { #4 - \l_@@_bottom_corner_dim } }
      }
  }
\cs_new_protected:Npn \@@_shift_pole:Nnnnnn #1#2#3#4#5#6
  {
    \prop_put:cnx { l_@@_poles_ \__int_value:w #1 _ prop } {#2}
      {
        { \dim_eval:n { #3 - \l_@@_left_corner_dim } }
        { \dim_eval:n { #4 - \l_@@_bottom_corner_dim } }
        {#5} {#6}
      }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsection{Resizing coffins}
%
% \begin{variable}{\l_@@_scale_x_fp}
% \begin{variable}{\l_@@_scale_y_fp}
%   Storage for the scaling factors in $x$ and $y$, respectively.
%    \begin{macrocode}
\fp_new:N \l_@@_scale_x_fp
\fp_new:N \l_@@_scale_y_fp
%    \end{macrocode}
% \end{variable}
% \end{variable}
%
% \begin{variable}{\l_@@_scaled_total_height_dim}
% \begin{variable}{\l_@@_scaled_width_dim}
%   When scaling, the values given have to be turned into absolute values.
%    \begin{macrocode}
\dim_new:N \l_@@_scaled_total_height_dim
\dim_new:N \l_@@_scaled_width_dim
%    \end{macrocode}
% \end{variable}
% \end{variable}
%
% \begin{macro}{\coffin_resize:Nnn, \coffin_resize:cnn}
%   Resizing a coffin begins by setting up the user-friendly names for
%   the dimensions of the coffin box. The new sizes are then turned into
%   scale factor. This is the same operation as takes place for the
%   underlying box, but that operation is grouped and so the same
%   calculation is done here.
%    \begin{macrocode}
\cs_new_protected:Npn \coffin_resize:Nnn #1#2#3
  {
    \fp_set:Nn \l_@@_scale_x_fp
      { \dim_to_fp:n {#2} / \dim_to_fp:n { \coffin_wd:N #1 } }
    \fp_set:Nn \l_@@_scale_y_fp
      {
          \dim_to_fp:n {#3}
        / \dim_to_fp:n { \coffin_ht:N #1 + \coffin_dp:N #1 }
      }
    \box_resize:Nnn #1 {#2} {#3}
    \@@_resize_common:Nnn #1 {#2} {#3}
  }
\cs_generate_variant:Nn \coffin_resize:Nnn { c }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_resize_common:Nnn}
%   The poles and corners of the coffin are scaled to the appropriate
%   places before actually resizing the underlying box.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_resize_common:Nnn #1#2#3
  {
    \prop_map_inline:cn { l_@@_corners_ \__int_value:w #1 _prop }
      { \@@_scale_corner:Nnnn #1 {##1} ##2 }
    \prop_map_inline:cn { l_@@_poles_ \__int_value:w #1 _prop }
      { \@@_scale_pole:Nnnnnn #1 {##1} ##2 }
%    \end{macrocode}
%   Negative $x$-scaling values will place the poles in the wrong
%   location: this is corrected here.
%    \begin{macrocode}
    \fp_compare:nNnT \l_@@_scale_x_fp < \c_zero_fp
      {
        \prop_map_inline:cn { l_@@_corners_ \__int_value:w #1 _prop }
          { \@@_x_shift_corner:Nnnn #1 {##1} ##2 }
        \prop_map_inline:cn { l_@@_poles_ \__int_value:w #1 _prop }
          { \@@_x_shift_pole:Nnnnnn #1 {##1} ##2 }
      }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\coffin_scale:Nnn, \coffin_scale:cnn}
%   For scaling, the opposite calculation is done to find the new
%   dimensions for the coffin. Only the total height is needed, as this
%   is the shift required for corners and poles. The scaling is done
%   the \TeX{} way as this works properly with floating point values
%   without needing to use the \texttt{fp} module.
%    \begin{macrocode}
\cs_new_protected:Npn \coffin_scale:Nnn #1#2#3
  {
    \fp_set:Nn \l_@@_scale_x_fp {#2}
    \fp_set:Nn \l_@@_scale_y_fp {#3}
    \box_scale:Nnn #1 { \l_@@_scale_x_fp } { \l_@@_scale_y_fp }
    \dim_set:Nn \l_@@_internal_dim
      { \coffin_ht:N #1 + \coffin_dp:N #1 }
    \dim_set:Nn \l_@@_scaled_total_height_dim
      { \fp_abs:n { \l_@@_scale_y_fp } \l_@@_internal_dim }
    \dim_set:Nn \l_@@_scaled_width_dim
      { -\fp_abs:n { \l_@@_scale_x_fp  } \coffin_wd:N #1 }
    \@@_resize_common:Nnn #1
      { \l_@@_scaled_width_dim } { \l_@@_scaled_total_height_dim }
  }
\cs_generate_variant:Nn \coffin_scale:Nnn { c }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_scale_vector:nnNN}
%   This functions scales a vector from the origin using the pre-set scale
%   factors in $x$ and $y$. This is a much less complex operation
%   than rotation, and as a result the code is a lot clearer.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_scale_vector:nnNN #1#2#3#4
  {
    \dim_set:Nn #3
      { \fp_to_dim:n { \dim_to_fp:n {#1} * \l_@@_scale_x_fp } }
    \dim_set:Nn #4
      { \fp_to_dim:n { \dim_to_fp:n {#2} * \l_@@_scale_y_fp } }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_scale_corner:Nnnn}
% \begin{macro}{\@@_scale_pole:Nnnnnn}
%   Scaling both corners and poles is a simple calculation using the
%   preceding vector scaling.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_scale_corner:Nnnn #1#2#3#4
  {
    \@@_scale_vector:nnNN {#3} {#4} \l_@@_x_dim \l_@@_y_dim
    \prop_put:cnx { l_@@_corners_ \__int_value:w #1 _prop } {#2}
      { { \dim_use:N \l_@@_x_dim } { \dim_use:N \l_@@_y_dim } }
  }
\cs_new_protected:Npn \@@_scale_pole:Nnnnnn #1#2#3#4#5#6
  {
    \@@_scale_vector:nnNN {#3} {#4} \l_@@_x_dim \l_@@_y_dim
    \@@_set_pole:Nnx #1 {#2}
      {
        { \dim_use:N \l_@@_x_dim } { \dim_use:N \l_@@_y_dim }
        {#5} {#6}
      }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\@@_x_shift_corner:Nnnn}
% \begin{macro}{\@@_x_shift_pole:Nnnnnn}
%   These functions correct for the $x$ displacement that takes
%   place with a negative horizontal scaling.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_x_shift_corner:Nnnn #1#2#3#4
  {
    \prop_put:cnx { l_@@_corners_ \__int_value:w #1 _prop } {#2}
      {
        { \dim_eval:n { #3 + \box_wd:N #1 } } {#4}
      }
  }
\cs_new_protected:Npn \@@_x_shift_pole:Nnnnnn #1#2#3#4#5#6
  {
    \prop_put:cnx { l_@@_poles_ \__int_value:w #1 _prop } {#2}
      {
        { \dim_eval:n #3 + \box_wd:N #1 } {#4}
        {#5} {#6}
      }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsection{Coffin diagnostics}
%
% \begin{macro}{\coffin_log_structure:N, \coffin_log_structure:c}
%   Redirect output of \cs{coffin_show_structure:N} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \coffin_log_structure:N
  { \__msg_log_next: \coffin_show_structure:N }
\cs_generate_variant:Nn \coffin_log_structure:N { c }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3file}}
%
%    \begin{macrocode}
%<@@=file>
%    \end{macrocode}
%
% \begin{macro}[TF]{\file_if_exist_input:n}
%   Input of a file with a test for existence cannot be done the usual
%   way as the tokens to insert are in an odd place.
%    \begin{macrocode}
\cs_new_protected:Npn \file_if_exist_input:n #1
  {
    \file_if_exist:nT {#1}
      { \@@_input:V \l_@@_internal_name_tl }
  }
\cs_new_protected:Npn \file_if_exist_input:nT #1#2
  {
    \file_if_exist:nT {#1}
      {
        #2
        \@@_input:V \l_@@_internal_name_tl
      }
  }
\cs_new_protected:Npn \file_if_exist_input:nF #1
  {
    \file_if_exist:nTF {#1}
      { \@@_input:V \l_@@_internal_name_tl }
  }
\cs_new_protected:Npn \file_if_exist_input:nTF #1#2
  {
    \file_if_exist:nTF {#1}
      {
        #2
        \@@_input:V \l_@@_internal_name_tl
      }
  }
%    \end{macrocode}
% \end{macro}
%
%    \begin{macrocode}
%<@@=ior>
%    \end{macrocode}
%
% \begin{macro}[EXP]{\ior_map_break:, \ior_map_break:n}
%   Usual map breaking functions.  Those are not yet in \pkg{l3kernel}
%   proper since the mapping below is the first of its kind.
%    \begin{macrocode}
\cs_new:Npn \ior_map_break:
  { \__prg_map_break:Nn \ior_map_break: { } }
\cs_new:Npn \ior_map_break:n
  { \__prg_map_break:Nn \ior_map_break: }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\ior_map_inline:Nn, \ior_str_map_inline:Nn}
% \begin{macro}[aux]{\@@_map_inline:NNn}
% \begin{macro}[aux]{\@@_map_inline:NNNn}
% \begin{macro}[aux]{\@@_map_inline_loop:NNN}
% \begin{variable}{\l_@@_internal_tl}
%   Mapping to an input stream can be done on either a token or a string
%   basis, hence the set up. Within that, there is a check to avoid reading
%   past the end of a file, hence the two applications of \cs{ior_if_eof:N}.
%   This mapping cannot be nested as the stream has only one \enquote{current
%   line}.
%    \begin{macrocode}
\cs_new_protected:Npn \ior_map_inline:Nn
  { \@@_map_inline:NNn \ior_get:NN }
\cs_new_protected:Npn \ior_str_map_inline:Nn
  { \@@_map_inline:NNn \ior_get_str:NN }
\cs_new_protected:Npn \@@_map_inline:NNn
  {
    \int_gincr:N \g__prg_map_int
    \exp_args:Nc \@@_map_inline:NNNn
      { __prg_map_ \int_use:N \g__prg_map_int :n }
  }
\cs_new_protected:Npn \@@_map_inline:NNNn #1#2#3#4
  {
    \cs_set:Npn #1 ##1 {#4}
    \ior_if_eof:NF #3 { \@@_map_inline_loop:NNN #1#2#3 }
    \__prg_break_point:Nn \ior_map_break:
      { \int_gdecr:N \g__prg_map_int }
  }
\cs_new_protected:Npn \@@_map_inline_loop:NNN #1#2#3
  {
    #2 #3 \l_@@_internal_tl
    \ior_if_eof:NF #3
      {
        \exp_args:No #1 \l_@@_internal_tl
        \@@_map_inline_loop:NNN #1#2#3
      }
  }
\tl_new:N  \l_@@_internal_tl
%    \end{macrocode}
% \end{variable}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\ior_log_streams:}
%   Redirect output of \cs{ior_list_streams:} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \ior_log_streams:
  { \__msg_log_next: \ior_list_streams: }
%    \end{macrocode}
% \end{macro}
%
%    \begin{macrocode}
%<@@=iow>
%    \end{macrocode}
%
% \begin{macro}{\iow_log_streams:}
%   Redirect output of \cs{iow_list_streams:} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \iow_log_streams:
  { \__msg_log_next: \iow_list_streams: }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3fp-assign}}
%
%    \begin{macrocode}
%<@@=fp>
%    \end{macrocode}
%
% \begin{macro}{\fp_log:N, \fp_log:c, \fp_log:n}
%   Redirect output of \cs{fp_show:N} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \fp_log:N
  { \__msg_log_next: \fp_show:N }
\cs_new_protected:Npn \fp_log:n
  { \__msg_log_next: \fp_show:n }
\cs_generate_variant:Nn \fp_log:N { c }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3int}}
%
% \begin{macro}{\int_log:N, \int_log:c}
%   Redirect output of \cs{int_show:N} to the log.  This is not just a
%   copy of \cs{__kernel_register_log:N} because of subtleties
%   involving \tn{currentgrouplevel} and \tn{currentgrouptype}.  See
%   \cs{int_show:N} for details.
%    \begin{macrocode}
\cs_new_protected:Npn \int_log:N
  { \__msg_log_next: \int_show:N }
\cs_generate_variant:Nn \int_log:N { c }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\int_log:n}
%   Redirect output of \cs{int_show:n} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \int_log:n
  { \__msg_log_next: \int_show:n }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3keys}}
%
%    \begin{macrocode}
%<@@=keys>
%    \end{macrocode}
%
% \begin{macro}{\keys_log:nn}
%   Redirect output of \cs{keys_show:nn} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \keys_log:nn
  { \__msg_log_next: \keys_show:nn }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3msg}}
%
%    \begin{macrocode}
%<@@=msg>
%    \end{macrocode}
%
% \begin{macro}[EXP]
%   {
%     \msg_expandable_error:nnnnnn ,
%     \msg_expandable_error:nnnnn  ,
%     \msg_expandable_error:nnnn   ,
%     \msg_expandable_error:nnn    ,
%     \msg_expandable_error:nn     ,
%     \msg_expandable_error:nnffff ,
%     \msg_expandable_error:nnfff  ,
%     \msg_expandable_error:nnff   ,
%     \msg_expandable_error:nnf
%   }
% \begin{macro}[aux]{\__msg_expandable_error_module:nn}
%   Pass to an auxiliary the message to display and the module name
%    \begin{macrocode}
\cs_new:Npn \msg_expandable_error:nnnnnn #1#2#3#4#5#6
  {
    \exp_args:Nf \@@_expandable_error_module:nn
      {
        \exp_args:Nf \tl_to_str:n
          { \use:c { \c_@@_text_prefix_tl #1 / #2 } {#3} {#4} {#5} {#6} }
      }
      {#1}
  }
\cs_new:Npn \msg_expandable_error:nnnnn #1#2#3#4#5
  { \msg_expandable_error:nnnnnn {#1} {#2} {#3} {#4} {#5} { } }
\cs_new:Npn \msg_expandable_error:nnnn #1#2#3#4
  { \msg_expandable_error:nnnnnn {#1} {#2} {#3} {#4} { } { } }
\cs_new:Npn \msg_expandable_error:nnn #1#2#3
  { \msg_expandable_error:nnnnnn {#1} {#2} {#3} { } { } { } }
\cs_new:Npn \msg_expandable_error:nn #1#2
  { \msg_expandable_error:nnnnnn {#1} {#2} { } { } { } { } }
\cs_generate_variant:Nn \msg_expandable_error:nnnnnn { nnffff }
\cs_generate_variant:Nn \msg_expandable_error:nnnnn  { nnfff }
\cs_generate_variant:Nn \msg_expandable_error:nnnn   { nnff }
\cs_generate_variant:Nn \msg_expandable_error:nnn    { nnf }
\cs_new:Npn \@@_expandable_error_module:nn #1#2
  {
    \exp_after:wN \exp_after:wN
    \exp_after:wN \use_none_delimit_by_q_stop:w
    \use:n { \::error ! ~ #2 : ~ #1 } \q_stop
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsection{Additions to \pkg{l3prg}}
%
%    \begin{macrocode}
%<@@=bool>
%    \end{macrocode}
%
% \begin{macro}[pTF]{\bool_lazy_all:n}
% \begin{macro}[aux]{\@@_lazy_all:n}
%   Go through the list of expressions, stopping whenever an expression
%   is \texttt{false}.  If the end is reached without finding any
%   \texttt{false} expression, then the result is \texttt{true}.
%    \begin{macrocode}
\prg_new_conditional:Npnn \bool_lazy_all:n #1 { p , T , F , TF }
  { \@@_lazy_all:n #1 \q_recursion_tail \q_recursion_stop }
\cs_new:Npn \@@_lazy_all:n #1
  {
    \quark_if_recursion_tail_stop_do:nn {#1} { \prg_return_true: }
    \bool_if:nF {#1}
      { \use_i_delimit_by_q_recursion_stop:nw { \prg_return_false: } }
    \@@_lazy_all:n
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[pTF]{\bool_lazy_and:nn}
%   Only evaluate the second expression if the first is \texttt{true}.
%    \begin{macrocode}
\prg_new_conditional:Npnn \bool_lazy_and:nn #1#2 { p , T , F , TF }
  {
    \bool_if:nTF {#1}
      { \bool_if:nTF {#2} { \prg_return_true: } { \prg_return_false: } }
      { \prg_return_false: }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[pTF]{\bool_lazy_any:n}
% \begin{macro}[aux]{\@@_lazy_any:n}
%   Go through the list of expressions, stopping whenever an expression
%   is \texttt{true}.  If the end is reached without finding any
%   \texttt{true} expression, then the result is \texttt{false}.
%    \begin{macrocode}
\prg_new_conditional:Npnn \bool_lazy_any:n #1 { p , T , F , TF }
  { \@@_lazy_any:n #1 \q_recursion_tail \q_recursion_stop }
\cs_new:Npn \@@_lazy_any:n #1
  {
    \quark_if_recursion_tail_stop_do:nn {#1} { \prg_return_false: }
    \bool_if:nT {#1}
      { \use_i_delimit_by_q_recursion_stop:nw { \prg_return_true: } }
    \@@_lazy_any:n
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[pTF]{\bool_lazy_or:nn}
%   Only evaluate the second expression if the first is \texttt{false}.
%    \begin{macrocode}
\prg_new_conditional:Npnn \bool_lazy_or:nn #1#2 { p , T , F , TF }
  {
    \bool_if:nTF {#1}
      { \prg_return_true: }
      { \bool_if:nTF {#2} { \prg_return_true: } { \prg_return_false: } }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bool_log:N, \bool_log:c, \bool_log:n}
%   Redirect output of \cs{bool_show:N} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \bool_log:N
  { \__msg_log_next: \bool_show:N }
\cs_new_protected:Npn \bool_log:n
  { \__msg_log_next: \bool_show:n }
\cs_generate_variant:Nn \bool_log:N { c }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3prop}}
%
%    \begin{macrocode}
%<@@=prop>
%    \end{macrocode}
%
% \begin{macro}[EXP]{\prop_count:N, \prop_count:c}
% \begin{macro}[aux,EXP]{\@@_count:nn}
%   Counting the key--value pairs in a property list is done using the
%   same approach as for other count functions: turn each entry into a
%   \texttt{+1} then use integer evaluation to actually do the
%   mathematics.
%    \begin{macrocode}
\cs_new:Npn \prop_count:N #1
  {
    \int_eval:n
      {
        \c_zero
        \prop_map_function:NN #1 \@@_count:nn
      }
  }
\cs_new:Npn \@@_count:nn #1#2 { + \c_one }
\cs_generate_variant:Nn \prop_count:N { c }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[rEXP]{\prop_map_tokens:Nn, \prop_map_tokens:cn}
% \begin{macro}[aux]{\@@_map_tokens:nwwn}
%   The mapping is very similar to \cs{prop_map_function:NN}.  It grabs
%   one key--value pair at a time, and stops when reaching the marker
%   key \cs{q_recursion_tail}, which cannot appear in normal keys since
%   those are strings.  The odd construction |\use:n {#1}| allows |#1|
%   to contain any token without interfering with \cs{prop_map_break:}.
%   Argument |#2| of \cs{@@_map_tokens:nwwn} is \cs{s_@@} the first
%   time, and is otherwise empty.
%    \begin{macrocode}
\cs_new:Npn \prop_map_tokens:Nn #1#2
  {
    \exp_last_unbraced:Nno \@@_map_tokens:nwwn {#2} #1
      \@@_pair:wn \q_recursion_tail \s_@@ { }
    \__prg_break_point:Nn \prop_map_break: { }
  }
\cs_new:Npn \@@_map_tokens:nwwn #1#2 \@@_pair:wn #3 \s_@@ #4
  {
    \if_meaning:w \q_recursion_tail #3
      \exp_after:wN \prop_map_break:
    \fi:
    \use:n {#1} {#3} {#4}
    \@@_map_tokens:nwwn {#1}
  }
\cs_generate_variant:Nn \prop_map_tokens:Nn { c }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\prop_log:N, \prop_log:c}
%   Redirect output of \cs{prop_show:N} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \prop_log:N
  { \__msg_log_next: \prop_show:N }
\cs_generate_variant:Nn \prop_log:N { c }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3seq}}
%
%    \begin{macrocode}
%<@@=seq>
%    \end{macrocode}
%
% \begin{macro}
%   {
%     \seq_mapthread_function:NNN, \seq_mapthread_function:NcN,
%     \seq_mapthread_function:cNN, \seq_mapthread_function:ccN
%   }
% \begin{macro}[aux]
%   {
%     \@@_mapthread_function:wNN, \@@_mapthread_function:wNw,
%     \@@_mapthread_function:Nnnwnn
%   }
%   The idea is to first expand both sequences, adding the
%   usual |{ ? \__prg_break: } { }| to the end of each one.  This is
%   most conveniently done in two steps using an auxiliary function.
%   The mapping then throws away the first tokens of |#2| and |#5|,
%   which for items in the sequences will both be \cs{s_@@}
%   \cs{@@_item:n}.  The function to be mapped will then be applied to
%   the two entries.  When the code hits the end of one of the
%   sequences, the break material will stop the entire loop and tidy up.
%   This avoids needing to find the count of the two sequences, or
%   worrying about which is longer.
%    \begin{macrocode}
\cs_new:Npn \seq_mapthread_function:NNN #1#2#3
  { \exp_after:wN \@@_mapthread_function:wNN #2 \q_stop #1 #3 }
\cs_new:Npn \@@_mapthread_function:wNN \s_@@ #1 \q_stop #2#3
  {
    \exp_after:wN \@@_mapthread_function:wNw #2 \q_stop #3
      #1 { ? \__prg_break: } { }
    \__prg_break_point:
  }
\cs_new:Npn \@@_mapthread_function:wNw \s_@@ #1 \q_stop #2
  {
    \@@_mapthread_function:Nnnwnn #2
      #1 { ? \__prg_break: } { }
    \q_stop
  }
\cs_new:Npn \@@_mapthread_function:Nnnwnn #1#2#3#4 \q_stop #5#6
  {
    \use_none:n #2
    \use_none:n #5
    #1 {#3} {#6}
    \@@_mapthread_function:Nnnwnn #1 #4 \q_stop
  }
\cs_generate_variant:Nn \seq_mapthread_function:NNN {     Nc }
\cs_generate_variant:Nn \seq_mapthread_function:NNN { c , cc }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\seq_set_filter:NNn, \seq_gset_filter:NNn}
% \begin{macro}[aux]{\@@_set_filter:NNNn}
%   Similar to \cs{seq_map_inline:Nn}, without a
%   \cs{__prg_break_point:} because the user's code
%   is performed within the evaluation of a boolean expression,
%   and skipping out of that would break horribly.
%   The \cs{@@_wrap_item:n} function inserts the relevant
%   \cs{@@_item:n} without expansion in the input stream,
%   hence in the \texttt{x}-expanding assignment.
%    \begin{macrocode}
\cs_new_protected:Npn \seq_set_filter:NNn
  { \@@_set_filter:NNNn \tl_set:Nx }
\cs_new_protected:Npn \seq_gset_filter:NNn
  { \@@_set_filter:NNNn \tl_gset:Nx }
\cs_new_protected:Npn \@@_set_filter:NNNn #1#2#3#4
  {
    \@@_push_item_def:n { \bool_if:nT {#4} { \@@_wrap_item:n {##1} } }
    #1 #2 { #3 }
    \@@_pop_item_def:
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\seq_set_map:NNn, \seq_gset_map:NNn}
% \begin{macro}[aux]{\@@_set_map:NNNn}
%   Very similar to \cs{seq_set_filter:NNn}. We could actually
%   merge the two within a single function, but it would have weird
%   semantics.
%    \begin{macrocode}
\cs_new_protected:Npn \seq_set_map:NNn
  { \@@_set_map:NNNn \tl_set:Nx }
\cs_new_protected:Npn \seq_gset_map:NNn
  { \@@_set_map:NNNn \tl_gset:Nx }
\cs_new_protected:Npn \@@_set_map:NNNn #1#2#3#4
  {
    \@@_push_item_def:n { \exp_not:N \@@_item:n {#4} }
    #1 #2 { #3 }
    \@@_pop_item_def:
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\seq_log:N, \seq_log:c}
%   Redirect output of \cs{seq_show:N} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \seq_log:N
  { \__msg_log_next: \seq_show:N }
\cs_generate_variant:Nn \seq_log:N { c }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3skip}}
%
%    \begin{macrocode}
%<@@=skip>
%    \end{macrocode}
%
% \begin{macro}{\skip_split_finite_else_action:nnNN}
%   This macro is useful when performing error checking in certain
%   circumstances. If the \meta{skip} register holds finite glue it sets
%   |#3| and |#4| to the stretch and shrink component, resp. If it holds
%   infinite glue set |#3| and |#4| to zero and issue the special action
%   |#2| which is probably an error message.
%   Assignments are local.
%    \begin{macrocode}
\cs_new:Npn \skip_split_finite_else_action:nnNN #1#2#3#4
  {
    \skip_if_finite:nTF {#1}
      {
        #3 = \etex_gluestretch:D #1 \scan_stop:
        #4 = \etex_glueshrink:D  #1 \scan_stop:
      }
      {
        #3 = \c_zero_skip
        #4 = \c_zero_skip
        #2
      }
  }
%    \end{macrocode}
%  \end{macro}
%
% \begin{macro}{\dim_log:N, \dim_log:c, \dim_log:n}
%   Diagnostics.  Redirect output of \cs{dim_show:n} to the log.
%    \begin{macrocode}
\cs_new_eq:NN \dim_log:N \__kernel_register_log:N
\cs_new_eq:NN \dim_log:c \__kernel_register_log:c
\cs_new_protected:Npn \dim_log:n
  { \__msg_log_next: \dim_show:n }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\skip_log:N, \skip_log:c, \skip_log:n}
%   Diagnostics.  Redirect output of \cs{skip_show:n} to the log.
%    \begin{macrocode}
\cs_new_eq:NN \skip_log:N \__kernel_register_log:N
\cs_new_eq:NN \skip_log:c \__kernel_register_log:c
\cs_new_protected:Npn \skip_log:n
  { \__msg_log_next: \skip_show:n }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\muskip_log:N, \muskip_log:c, \muskip_log:n}
%   Diagnostics.  Redirect output of \cs{muskip_show:n} to the log.
%    \begin{macrocode}
\cs_new_eq:NN \muskip_log:N \__kernel_register_log:N
\cs_new_eq:NN \muskip_log:c \__kernel_register_log:c
\cs_new_protected:Npn \muskip_log:n
  { \__msg_log_next: \muskip_show:n }
%    \end{macrocode}
% \end{macro}
%
%  \subsection{Additions to \pkg{l3tl}}
%
%    \begin{macrocode}
%<@@=tl>
%    \end{macrocode}
%
% \begin{macro}[EXP,pTF]{\tl_if_single_token:n}
%   There are four cases: empty token list, token list starting with a
%   normal token, with a brace group, or with a space token.  If the
%   token list starts with a normal token, remove it and check for
%   emptiness.  For the next case, an empty token list is not a single
%   token.  Finally, we have a non-empty token list starting with a
%   space or a brace group.  Applying \texttt{f}-expansion yields an
%   empty result if and only if the token list is a single space.
%    \begin{macrocode}
\prg_new_conditional:Npnn \tl_if_single_token:n #1 { p , T , F , TF }
  {
    \tl_if_head_is_N_type:nTF {#1}
      { \@@_if_empty_return:o { \use_none:n #1 } }
      {
        \tl_if_empty:nTF {#1}
          { \prg_return_false: }
          { \@@_if_empty_return:o { \exp:w \exp_end_continue_f:w #1 } }
      }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[EXP]{\tl_reverse_tokens:n}
% \begin{macro}[EXP,aux]{\@@_reverse_group:nn}
%   The same as \cs{tl_reverse:n} but with recursion within brace groups.
%    \begin{macrocode}
\cs_new:Npn \tl_reverse_tokens:n #1
  {
    \etex_unexpanded:D \exp_after:wN
      {
        \exp:w
        \@@_act:NNNnn
          \@@_reverse_normal:nN
          \@@_reverse_group:nn
          \@@_reverse_space:n
          { }
          {#1}
      }
  }
\cs_new:Npn \@@_reverse_group:nn #1
  {
    \@@_act_group_recurse:Nnn
      \@@_act_reverse_output:n
      { \tl_reverse_tokens:n }
  }
%    \end{macrocode}
% \end{macro}
% \begin{macro}[EXP,aux]{\@@_act_group_recurse:Nnn}
%   In many applications of \cs{@@_act:NNNnn}, we need to recursively
%   apply some transformation within brace groups, then output. In this
%   code, |#1| is the output function, |#2| is the transformation,
%   which should expand in two steps, and |#3| is the group.
%    \begin{macrocode}
\cs_new:Npn \@@_act_group_recurse:Nnn #1#2#3
  {
    \exp_args:Nf #1
      { \exp_after:wN \exp_after:wN \exp_after:wN { #2 {#3} } }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[EXP]{\tl_count_tokens:n}
% \begin{macro}[EXP,aux]{\@@_act_count_normal:nN,
%     \@@_act_count_group:nn, \@@_act_count_space:n}
%   The token count is computed through an \cs{int_eval:n} construction.
%   Each \texttt{1+} is output to the \emph{left}, into the integer
%   expression, and the sum is ended by the \cs{exp_end:} inserted by
%   \cs{@@_act_end:wn} (which is technically implemented as  \cs{c_zero}).
%   Somewhat a hack!
%    \begin{macrocode}
\cs_new:Npn \tl_count_tokens:n #1
  {
    \int_eval:n
      {
        \@@_act:NNNnn
          \@@_act_count_normal:nN
          \@@_act_count_group:nn
          \@@_act_count_space:n
          { }
          {#1}
      }
  }
\cs_new:Npn \@@_act_count_normal:nN #1 #2 { 1 + }
\cs_new:Npn \@@_act_count_space:n #1 { 1 + }
\cs_new:Npn \@@_act_count_group:nn #1 #2
  { 2 + \tl_count_tokens:n {#2} + }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}
%   {
%     \tl_set_from_file:Nnn,  \tl_set_from_file:cnn,
%     \tl_gset_from_file:Nnn, \tl_gset_from_file:cnn
%   }
% \begin{macro}[aux]{\@@_set_from_file:NNnn}
% \begin{macro}[aux]{\@@_from_file_do:w}
%   The approach here is similar to that for doing a rescan, and so the same
%   internals can be reused. Thus the plan is to insert a pair of tokens of
%   the same charcode but different catcodes after the file has been read.
%   This plus \cs{exp_not:N} allows the primitive to be used to carry out
%   a set operation.
%    \begin{macrocode}
\cs_new_protected:Npn \tl_set_from_file:Nnn
  { \@@_set_from_file:NNnn \tl_set:Nn }
\cs_new_protected:Npn \tl_gset_from_file:Nnn
  { \@@_set_from_file:NNnn \tl_gset:Nn }
\cs_generate_variant:Nn \tl_set_from_file:Nnn  { c }
\cs_generate_variant:Nn \tl_gset_from_file:Nnn { c }
\cs_new_protected:Npn \@@_set_from_file:NNnn #1#2#3#4
  {
    \__file_if_exist:nT {#4}
      {
        \group_begin:
          \exp_args:No \etex_everyeof:D
            { \c_@@_rescan_marker_tl \exp_not:N }
          #3 \scan_stop:
          \exp_after:wN \@@_from_file_do:w
          \exp_after:wN \prg_do_nothing:
            \tex_input:D \l__file_internal_name_tl \scan_stop:
        \exp_args:NNNo \group_end:
        #1 #2 \l_@@_internal_a_tl
      }
  }
\exp_args:Nno \use:nn
  { \cs_set_protected:Npn \@@_from_file_do:w #1 }
  { \c_@@_rescan_marker_tl }
  { \tl_set:No \l_@@_internal_a_tl {#1} }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}
%   {
%     \tl_set_from_file_x:Nnn,  \tl_set_from_file_x:cnn,
%     \tl_gset_from_file_x:Nnn, \tl_gset_from_file_x:cnn
%   }
% \begin{macro}[aux]{\@@_set_from_file_x:NNnn}
%   When reading a file and allowing expansion of the content, the set up
%   only needs to prevent \TeX{} complaining about the end of the file. That
%   is done simply, with a group then used to trap the definition needed.
%   Once the business is done using some scratch space, the tokens can be
%   transferred to the real target.
%    \begin{macrocode}
\cs_new_protected:Npn \tl_set_from_file_x:Nnn
  { \@@_set_from_file_x:NNnn \tl_set:Nn }
\cs_new_protected:Npn \tl_gset_from_file_x:Nnn
  { \@@_set_from_file_x:NNnn \tl_gset:Nn }
\cs_generate_variant:Nn \tl_set_from_file_x:Nnn  { c }
\cs_generate_variant:Nn \tl_gset_from_file_x:Nnn { c }
\cs_new_protected:Npn \@@_set_from_file_x:NNnn #1#2#3#4
  {
    \__file_if_exist:nT {#4}
      {
        \group_begin:
          \etex_everyeof:D { \exp_not:N }
          #3 \scan_stop:
          \tl_set:Nx \l_@@_internal_a_tl
            { \tex_input:D \l__file_internal_name_tl \c_space_token }
        \exp_args:NNNo \group_end:
        #1 #2 \l_@@_internal_a_tl
      }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsubsection{Unicode case changing}
%
% The mechanisms needed for case changing are somewhat involved, particularly
% to allow for all of the special cases. These functions also require the
% appropriate data extracted from the Unicode documentation (either manually
% or automatically).
%
% \begin{macro}[EXP, documented-as=\tl_if_head_eq_catcode:nNTF]
%   {\tl_if_head_eq_catcode:oNTF}
%   Extra variants.
%    \begin{macrocode}
\cs_generate_variant:Nn \tl_if_head_eq_catcode:nNTF { o }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[EXP]{\tl_lower_case:n, \tl_upper_case:n, \tl_mixed_case:n}
% \begin{macro}[EXP]{\tl_lower_case:nn, \tl_upper_case:nn, \tl_mixed_case:nn}
%   The user level functions here are all wrappers around the internal
%   functions for case changing. Note that \cs{tl_mixed_case:nn} could be
%   done without an internal, but this way the logic is slightly clearer as
%   everything essentially follows the same path.
%    \begin{macrocode}
\cs_new:Npn \tl_lower_case:n { \@@_change_case:nnn { lower } { } }
\cs_new:Npn \tl_upper_case:n { \@@_change_case:nnn { upper } { } }
\cs_new:Npn \tl_mixed_case:n { \@@_mixed_case:nn { } }
\cs_new:Npn \tl_lower_case:nn { \@@_change_case:nnn { lower } }
\cs_new:Npn \tl_upper_case:nn { \@@_change_case:nnn { upper } }
\cs_new:Npn \tl_mixed_case:nn { \@@_mixed_case:nn }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[aux, EXP]{\@@_change_case:nnn}
% \begin{macro}[aux, EXP]{\@@_change_case_aux:nnn}
% \begin{macro}[aux, EXP]{\@@_change_case_loop:wnn}
% \begin{macro}[aux, EXP]
%   {
%     \@@_change_case_output:nwn ,
%     \@@_change_case_output:Vwn ,
%     \@@_change_case_output:own ,
%     \@@_change_case_output:vwn ,
%     \@@_change_case_output:fwn ,
%   }
% \begin{macro}[aux, EXP]{\@@_change_case_end:wn}
% \begin{macro}[aux, EXP]{\@@_change_case_group:nwnn}
% \begin{macro}[aux, EXP]{\@@_change_case_space:wnn}
% \begin{macro}[aux, EXP]{\@@_change_case_N_type:Nwnn}
% \begin{macro}[aux, EXP]{\@@_change_case_N_type:NNNnnn}
% \begin{macro}[aux, EXP]{\@@_change_case_math:NNNnnn}
% \begin{macro}[aux, EXP]{\@@_change_case_math_loop:wNNnn}
% \begin{macro}[aux, EXP]{\@@_change_case_math:NwNNnn}
% \begin{macro}[aux, EXP]{\@@_change_case_math_group:nwNNnn}
% \begin{macro}[aux, EXP]{\@@_change_case_math_space:wNNnn}
% \begin{macro}[aux, EXP]{\@@_change_case_N_type:Nnnn}
% \begin{macro}[aux, EXP]{\@@_change_case_char:Nnn}
% \begin{macro}[aux, EXP]{\@@_change_case_char:nN}
% \begin{macro}[aux, EXP]
%   {\@@_change_case_char_auxi:nN, \@@_change_case_char_auxii:nN}
% \begin{macro}[aux]
%   {\@@_lookup_lower:N, \@@_lookup_upper:N, \@@_lookup_title:N}
% \begin{macro}[aux, EXP]{\@@_change_case_char_UTFviii:nNN}
% \begin{macro}[aux, EXP]{\@@_change_case_char_UTFviii:nNNN}
% \begin{macro}[aux, EXP]{\@@_change_case_char_UTFviii:nNNNN}
% \begin{macro}[aux, EXP]{\@@_change_case_char_UTFviii:nn}
% \begin{macro}[aux, EXP]{\@@_change_case_cs_letterlike:Nnn}
% \begin{macro}[aux, EXP]{\@@_change_case_cs_accents:NN}
% \begin{macro}[aux, EXP]{\@@_change_case_cs:N}
% \begin{macro}[aux, EXP]{\@@_change_case_cs:NN}
% \begin{macro}[aux, EXP]{\@@_change_case_cs:NNn}
% \begin{macro}[aux, EXP]{\@@_change_case_protect:wNN}
% \begin{macro}[aux, EXP]{\@@_change_case_if_expandable:NTF}
% \begin{macro}[aux, EXP]{\@@_change_case_cs_expand:Nnw}
% \begin{macro}[aux, EXP]{\@@_change_case_cs_expand:NN}
%   The mechanism for the core conversion of case is based on the idea that
%   we can use a loop to grab the entire token list plus a quark: the latter is
%   used as an end marker and to avoid any brace stripping. Depending on the
%   nature of the first item in the grabbed argument, it can either processed
%   as a single token, treated as a group or treated as a space. These
%   different cases all work by re-reading |#1| in the appropriate way, hence
%   the repetition of |#1 \q_recursion_stop|.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case:nnn #1#2#3
  {
    \etex_unexpanded:D \exp_after:wN
      {
        \exp:w
        \@@_change_case_aux:nnn {#1} {#2} {#3}
      }
  }
\cs_new:Npn \@@_change_case_aux:nnn #1#2#3
  {
    \group_align_safe_begin:
    \@@_change_case_loop:wnn
      #3 \q_recursion_tail \q_recursion_stop {#1} {#2}
    \@@_change_case_result:n { }
  }
\cs_new:Npn \@@_change_case_loop:wnn #1 \q_recursion_stop
  {
    \tl_if_head_is_N_type:nTF {#1}
      { \@@_change_case_N_type:Nwnn }
      {
        \tl_if_head_is_group:nTF {#1}
          { \@@_change_case_group:nwnn }
          { \@@_change_case_space:wnn }
      }
    #1 \q_recursion_stop
  }
%    \end{macrocode}
%   Earlier versions of the code where only \texttt{x}-type expandable rather
%   than \texttt{f}-type: this causes issues with nesting and so the slight
%   performance hit is taken for a better outcome in usability terms. Setting
%   up for \texttt{f}-type expandability has two requirements: a marker
%   token after the main loop (see above) and a mechanism to \enquote{load}
%   and finalise the result. That is handled in the code below, which includes
%   the necessary material to end the \cs{exp:w} expansion.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_output:nwn #1#2 \@@_change_case_result:n #3
  { #2 \@@_change_case_result:n { #3 #1 } }
\cs_generate_variant:Nn \@@_change_case_output:nwn { V , o , v , f }
\cs_new:Npn \@@_change_case_end:wn #1 \@@_change_case_result:n #2
  {
    \group_align_safe_end:
    \exp_end:
    #2
  }
%    \end{macrocode}
%   Handling for the cases where the current argument is a brace group or
%   a space is relatively easy. For the brace case, the routine works
%   recursively, using the expandability of the mechanism to ensure that the
%   result is finalised before storage. For the space case it is simply a
%   question of removing the space in the input and storing it in the output.
%   In both cases, and indeed for the \texttt{N}-type grabber, after removing
%   the current item from the input \cs{@@_change_case_loop:wnn} is inserted
%   in front of the remaining tokens.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_group:nwnn #1#2 \q_recursion_stop #3#4
  {
    \@@_change_case_output:own
      {
        \exp_after:wN
          {
            \exp:w
            \@@_change_case_aux:nnn {#3} {#4} {#1}
          }
      }
    \@@_change_case_loop:wnn #2 \q_recursion_stop {#3} {#4}
  }
\exp_last_unbraced:NNo \cs_new:Npn \@@_change_case_space:wnn \c_space_tl
  {
    \@@_change_case_output:nwn { ~ }
    \@@_change_case_loop:wnn
  }
%    \end{macrocode}
%   For \texttt{N}-type arguments there are several stages to the approach.
%   First, a simply check for the end-of-input marker, which if found triggers
%   the final clean up and output step. Assuming that is not the case, the
%   first check is for math-mode escaping: this test can encompass control
%   sequences or other \texttt{N}-type tokens so is handled up front.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_N_type:Nwnn #1#2 \q_recursion_stop
  {
    \quark_if_recursion_tail_stop_do:Nn #1
      { \@@_change_case_end:wn }
    \exp_after:wN \@@_change_case_N_type:NNNnnn
      \exp_after:wN #1 \l_tl_case_change_math_tl
      \q_recursion_tail ? \q_recursion_stop {#2}
  }
%    \end{macrocode}
%   Looking for math mode escape first requires a loop over the possible
%   token pairs to see if the current input (|#1|) matches an open-math case
%   (|#2|). If if does then this test loop is ended and a new input-gathering
%   one is begun. The latter simply transfers material from the input to the
%   output without any expansion, testing each \texttt{N}-type token to see
%   if it matches the close-math case required. If that is the situation then
%   the \enquote{math loop} stops and resumes the main loop: as that might
%   be either the standard case-changing one or the mixed-case alternative,
%   it is not hard-coded into the math loop but is rather passed as argument
%   |#3| to \cs{@@_change_case_math:NNNnnn}. If no close-math token is found
%   then the final clean-up will be forced (\emph{i.e.}~there is no assumption
%   of \enquote{well-behaved} code in terms of math mode).
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_N_type:NNNnnn #1#2#3
  {
    \quark_if_recursion_tail_stop_do:Nn #2
     { \@@_change_case_N_type:Nnnn #1 }
    \token_if_eq_meaning:NNTF #1 #2
      {
        \use_i_delimit_by_q_recursion_stop:nw
           {
             \@@_change_case_math:NNNnnn
               #1 #3 \@@_change_case_loop:wnn
           }
      }
      { \@@_change_case_N_type:NNNnnn #1 }
  }
\cs_new:Npn \@@_change_case_math:NNNnnn #1#2#3#4
  {
    \@@_change_case_output:nwn {#1}
    \@@_change_case_math_loop:wNNnn #4 \q_recursion_stop #2 #3
  }
\cs_new:Npn \@@_change_case_math_loop:wNNnn #1 \q_recursion_stop
  {
    \tl_if_head_is_N_type:nTF {#1}
      { \@@_change_case_math:NwNNnn }
      {
        \tl_if_head_is_group:nTF {#1}
          { \@@_change_case_math_group:nwNNnn }
          { \@@_change_case_math_space:wNNnn }
      }
    #1 \q_recursion_stop
  }
\cs_new:Npn \@@_change_case_math:NwNNnn #1#2 \q_recursion_stop #3#4
  {
    \token_if_eq_meaning:NNTF \q_recursion_tail #1
      { \@@_change_case_end:wn }
      {
        \@@_change_case_output:nwn {#1}
        \token_if_eq_meaning:NNTF #1 #3
          { #4 #2 \q_recursion_stop }
          { \@@_change_case_math_loop:wNNnn #2 \q_recursion_stop #3#4 }
      }
  }
\cs_new:Npn \@@_change_case_math_group:nwNNnn #1#2 \q_recursion_stop
  {
    \@@_change_case_output:nwn { {#1} }
    \@@_change_case_math_loop:wNNnn #2 \q_recursion_stop
  }
\exp_last_unbraced:NNo
  \cs_new:Npn \@@_change_case_math_space:wNNnn \c_space_tl
  {
    \@@_change_case_output:nwn { ~ }
    \@@_change_case_math_loop:wNNnn
  }
%    \end{macrocode}
%   Once potential math-mode cases are filtered out the next stage is to
%   test if the token grabbed is a control sequence: they cannot be used in
%   the lookup table and also may require expansion. At this stage the loop
%   code starting \cs{@@_change_case_loop:wnn} is inserted: all subsequent
%   steps in the code which need a look-ahead are coded to rely on this and
%   thus have \texttt{w}-type arguments if they may do a look-ahead.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_N_type:Nnnn #1#2#3#4
  {
    \token_if_cs:NTF #1
      { \@@_change_case_cs_letterlike:Nnn #1 {#3} { } }
      { \@@_change_case_char:Nnn #1 {#3} {#4} }
    \@@_change_case_loop:wnn #2 \q_recursion_stop {#3} {#4}
  }
%    \end{macrocode}
%   For character tokens there are some special cases to deal with then
%   the majority of changes are covered by using the \TeX{} data as a lookup
%   along with expandable character generation. This avoids needing a very
%   large number of macros or (as seen in earlier versions) a somewhat tricky
%   split of the characters into various blocks. Notice that the special case
%   code may do a look-ahead so requires a final \texttt{w}-type argument
%   whereas the core lookup table does not and also guarantees an output so
%   \texttt{f}-type expansion may be used to obtain the case-changed result.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_char:Nnn #1#2#3
  {
    \cs_if_exist_use:cF { @@_change_case_ #2 _ #3 :Nnw }
      { \use_ii:nn }
        #1
        {
          \use:c { @@_change_case_ #2 _ sigma:Nnw } #1
            { \@@_change_case_char:nN {#2} #1 }
        }
  }
%    \end{macrocode}
%   For Unicode engines we can handle all characters directly. However, for
%   the $8$-bit engines the aim is to deal with (a subset of) Unicode (UTF-8)
%   input. They deal with that by making the upper half of the range active,
%   so we look for that and if found work out how many UTF-8 octets there
%   are to deal with. Those can then be grabbed to reconstruct the full
%   Unicode character, which is then used in a lookup. (As will become
%   obvious below, there is no intention here of covering all of Unicode.)
%    \begin{macrocode}
\cs_if_exist:NTF \utex_char:D
  {
    \cs_new:Npn \@@_change_case_char:nN #1#2
      { \@@_change_case_char_auxi:nN {#1} #2 }
  }
  {
    \cs_new:Npn \@@_change_case_char:nN #1#2
      {
        \int_compare:nNnTF { `#2 } > { "80 }
          {
            \int_compare:nNnTF { `#2 } < { "E0 }
              { \@@_change_case_char_UTFviii:nNNN {#1} #2 }
              {
                \int_compare:nNnTF { `#2 } < { "F0 }
                  { \@@_change_case_char_UTFviii:nNNNN {#1} #2 }
                  { \@@_change_case_char_UTFviii:nNNNNN {#1} #2 }
              }
          }
          { \@@_change_case_char_auxi:nN {#1} #2 }
       }
  }
\cs_new:Npn \@@_change_case_char_auxi:nN #1#2
  {
    \@@_change_case_output:fwn
      {
        \cs_if_exist_use:cF { c__unicode_ #1 _ \token_to_str:N #2 _tl }
          { \@@_change_case_char_auxii:nN {#1} #2 }
      }
  }
\cs_if_exist:NTF \utex_char:D
  {
    \cs_new:Npn \@@_change_case_char_auxii:nN #1#2
      {
        \int_compare:nNnTF { \use:c { @@_lookup_ #1 :N } #2 } = { 0 }
          { \exp_stop_f: #2 }
          {
            \char_generate:nn
              { \use:c { @@_lookup_ #1 :N } #2 }
              { \char_value_catcode:n { \use:c { @@_lookup_ #1 :N } #2 } }
          }
      }
    \cs_new_protected:Npn \@@_lookup_lower:N #1 { \tex_lccode:D `#1 }
    \cs_new_protected:Npn \@@_lookup_upper:N #1 { \tex_uccode:D `#1 }
    \cs_new_eq:NN \@@_lookup_title:N \@@_lookup_upper:N
  }
  {
    \cs_new:Npn \@@_change_case_char_auxii:nN #1#2 { \exp_stop_f: #2 }
    \cs_new:Npn \@@_change_case_char_UTFviii:nNNN #1#2#3#4
      { \@@_change_case_char_UTFviii:nnN {#1} {#2#4} #3 }
    \cs_new:Npn \@@_change_case_char_UTFviii:nNNNN #1#2#3#4#5
      { \@@_change_case_char_UTFviii:nnN {#1} {#2#4#5} #3 }
    \cs_new:Npn \@@_change_case_char_UTFviii:nNNNNN #1#2#3#4#5#6
      { \@@_change_case_char_UTFviii:nnN {#1} {#2#4#5#6} #3 }
    \cs_new:Npn \@@_change_case_char_UTFviii:nnN #1#2#3
      {
        \cs_if_exist:cTF { c__unicode_ #1 _ \tl_to_str:n {#2} _tl }
          {
            \@@_change_case_output:vwn
              { c__unicode_ #1 _ \tl_to_str:n {#2} _tl }
          }
          { \@@_change_case_output:nwn {#2} }
        #3
      }
  }
%    \end{macrocode}
%   Before dealing with general control sequences there are the special
%   ones to deal with. Letter-like control sequences are a simple look-up,
%   while for accents the loop is much as done elsewhere. Notice that
%   we have a no-op test to make sure there is no unexpected expansion of
%   letter-like input. The third argument here is needed for mixed casing,
%   where it if there is a hit there has to be a change-of-path.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_cs_letterlike:Nnn #1#2#3
  {
    \cs_if_exist:cTF { c_@@_change_case_ #2 _ \token_to_str:N #1 _tl }
      {
        \@@_change_case_output:vwn
          { c_@@_change_case_ #2 _ \token_to_str:N #1 _tl } 
        #3
      }
      {
        \cs_if_exist:cTF
          {
            c_@@_change_case_
            \str_if_eq:nnTF {#2} { lower } { upper } { lower }
            _ \token_to_str:N #1 _tl 
          }
          {
            \@@_change_case_output:nwn {#1}
            #3
          }
          {
            \exp_after:wN \@@_change_case_cs_accents:NN
              \exp_after:wN #1 \l_tl_case_change_accents_tl
              \q_recursion_tail \q_recursion_stop
          }
      }
  }
\cs_new:Npn \@@_change_case_cs_accents:NN #1#2
  {
    \quark_if_recursion_tail_stop_do:Nn #2
      { \@@_change_case_cs:N #1 }
    \str_if_eq:nnTF {#1} {#2}
      {
        \use_i_delimit_by_q_recursion_stop:nw
          { \@@_change_case_output:nwn {#1} }
      }
      { \@@_change_case_cs_accents:NN #1 }
  }
%    \end{macrocode}
%   To deal with a control sequence there is first a need to test if it is
%   on the list which indicate that case changing should be skipped. That's
%   done using a loop as for the other special cases. If a hit is found then
%   the argument is grabbed: that comes \emph{after} the loop function which
%   is therefore rearranged.  In a \LaTeXe{} context, \tn{protect} needs
%   to be treated specially, to prevent expansion of the next token but
%   output it without braces.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_cs:N #1
  {
%<*package>
    \str_if_eq:nnTF {#1} { \protect } { \@@_change_case_protect:wNN }
%</package>
    \exp_after:wN \@@_change_case_cs:NN
      \exp_after:wN #1 \l_tl_case_change_exclude_tl
      \q_recursion_tail \q_recursion_stop
  }
\cs_new:Npn \@@_change_case_cs:NN #1#2
  {
    \quark_if_recursion_tail_stop_do:Nn #2
      {
        \@@_change_case_cs_expand:Nnw #1
          { \@@_change_case_output:nwn {#1} }
      }
    \str_if_eq:nnTF {#1} {#2}
      {
        \use_i_delimit_by_q_recursion_stop:nw
          { \@@_change_case_cs:NNn #1 }
      }
      { \@@_change_case_cs:NN #1 }
  }
\cs_new:Npn \@@_change_case_cs:NNn #1#2#3
  {
    \@@_change_case_output:nwn { #1 {#3} }
    #2
  }
%<*package>
\cs_new:Npn \@@_change_case_protect:wNN #1 \q_recursion_stop #2 #3
  { \@@_change_case_output:nwn { \protect #3 } #2 }
%</package>
%    \end{macrocode}
%   When a control sequence is not on the exclude list the other test if
%   to see if it is expandable. Once again, if there is a hit then the loop
%   function is grabbed as part of the clean-up and reinserted before the
%   now expanded material. The test for expandablity has to check for
%   end-of-recursion as it is needed by the look-ahead code which might hit
%   the end of the input. The test is done in two parts as \cs{bool_if:nTF}
%   will choke if |#1| is |(|!
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_if_expandable:NTF #1
  {
    \token_if_expandable:NTF #1
      {
        \bool_if:nTF
          {
               \token_if_protected_macro_p:N      #1
            || \token_if_protected_long_macro_p:N #1
            || \token_if_eq_meaning_p:NN \q_recursion_tail #1
          }
          { \use_ii:nn }
          { \use_i:nn }
      }
      { \use_ii:nn }
  }
\cs_new:Npn \@@_change_case_cs_expand:Nnw #1#2
  {
    \@@_change_case_if_expandable:NTF #1
      { \@@_change_case_cs_expand:NN #1 }
      { #2 }
  }
\cs_new:Npn \@@_change_case_cs_expand:NN #1#2
  { \exp_after:wN #2 #1 }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}[aux, EXP]{\@@_change_case_lower_sigma:Nnw}
% \begin{macro}[aux, EXP]{\@@_change_case_lower_sigma:w}
% \begin{macro}[aux, EXP]{\@@_change_case_lower_sigma:Nw}
% \begin{macro}[aux, EXP]{\@@_change_case_upper_sigma:Nnw}
%   If the current char is an upper case sigma, the a check is made on the next
%   item in the input.  If it is \texttt{N}-type and not a control sequence
%   then there is a look-ahead phase.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_lower_sigma:Nnw #1#2#3#4 \q_recursion_stop
  {
    \int_compare:nNnTF { `#1 } = { "03A3 }
      {
        \@@_change_case_output:fwn
          { \@@_change_case_lower_sigma:w #4 \q_recursion_stop }
      }
      {#2}
    #3 #4 \q_recursion_stop
  }
\cs_new:Npn \@@_change_case_lower_sigma:w #1 \q_recursion_stop
  {
    \tl_if_head_is_N_type:nTF {#1}
      { \@@_change_case_lower_sigma:Nw #1 \q_recursion_stop }
      { \c__unicode_final_sigma_tl }
  }
\cs_new:Npn \@@_change_case_lower_sigma:Nw #1#2 \q_recursion_stop
  {
    \@@_change_case_if_expandable:NTF #1
      {
        \exp_after:wN \@@_change_case_lower_sigma:w #1
          #2 \q_recursion_stop
      }
      {
        \token_if_letter:NTF #1
          { \c__unicode_std_sigma_tl }
          { \c__unicode_final_sigma_tl }
      }
  }
%    \end{macrocode}
%   Simply skip to the final step for upper casing.
%    \begin{macrocode}
\cs_new_eq:NN \@@_change_case_upper_sigma:Nnw \use_ii:nn
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}[aux, EXP]{\@@_change_case_lower_tr:Nnw}
% \begin{macro}[aux, EXP]{\@@_change_case_lower_tr_auxi:Nw}
% \begin{macro}[aux, EXP]{\@@_change_case_lower_tr_auxii:Nw}
% \begin{macro}[aux, EXP]{\@@_change_case_upper_tr:Nnw}
% \begin{macro}[aux, EXP]{\@@_change_case_lower_az:Nnw}
% \begin{macro}[aux, EXP]{\@@_change_case_upper_az:Nnw}
%   The Turkic languages need special treatment for dotted-i and dotless-i.
%   The lower casing rule can be expressed in terms of searching first for
%   either a dotless-I or a dotted-I. In the latter case the mapping is
%   easy, but in the former there is a second stage search.
%    \begin{macrocode}
\cs_if_exist:NTF \utex_char:D
  {
    \cs_new:Npn \@@_change_case_lower_tr:Nnw #1#2
      {
        \int_compare:nNnTF { `#1 } = { "0049 }
          { \@@_change_case_lower_tr_auxi:Nw }
          {
            \int_compare:nNnTF { `#1 } = { "0130 }
              { \@@_change_case_output:nwn { i } }
              {#2}
          }
      }
%    \end{macrocode}
%   After a dotless-I there may be a dot-above character. If there is then
%   a dotted-i should be produced, otherwise output a dotless-i. When the
%   combination is found both the dotless-I and the dot-above char have to
%   be removed from the input, which is done by the \cs{use_i:nn}
%   (it grabs \cs{@@_change_case_loop:wn} and the dot-above char and
%   discards the latter).
%    \begin{macrocode}
    \cs_new:Npn \@@_change_case_lower_tr_auxi:Nw #1#2 \q_recursion_stop
      {
        \tl_if_head_is_N_type:nTF {#2}
          { \@@_change_case_lower_tr_auxii:Nw #2 \q_recursion_stop }
          { \@@_change_case_output:Vwn \c__unicode_dotless_i_tl }
        #1 #2 \q_recursion_stop
      }
    \cs_new:Npn \@@_change_case_lower_tr_auxii:Nw #1#2 \q_recursion_stop
      {
        \@@_change_case_if_expandable:NTF #1
          {
            \exp_after:wN \@@_change_case_lower_tr_auxi:Nw #1
              #2 \q_recursion_stop
          }
          {
            \bool_if:nTF
              {
                   \token_if_cs_p:N #1
                || ! ( \int_compare_p:nNn { `#1 } = { "0307 } )
              }
              { \@@_change_case_output:Vwn \c__unicode_dotless_i_tl }
              {
                \@@_change_case_output:nwn { i }
                \use_i:nn
              }
          }
      }
  }
%    \end{macrocode}
%   For $8$-bit engines, dot-above is not available so there is a simple
%   test for an upper-case I. Then we can look for the UTF-8 representation of
%   an upper case dotted-I without the combining char. If it's not there,
%   preserve the UTF-8 sequence as-is.
%    \begin{macrocode}
  {
    \cs_new:Npn \@@_change_case_lower_tr:Nnw #1#2
      {
        \int_compare:nNnTF { `#1 } = { "0049 }
          { \@@_change_case_output:Vwn \c__unicode_dotless_i_tl }
          {
            \int_compare:nNnTF { `#1 } = { 196 }
              { \@@_change_case_lower_tr_auxi:Nw #1 {#2} }
              {#2}
          }
      }
    \cs_new:Npn \@@_change_case_lower_tr_auxi:Nw #1#2#3#4
      {
        \int_compare:nNnTF { `#4 } = { 176 }
          {
            \@@_change_case_output:nwn { i }
            #3
          }
          {
            #2
            #3 #4
          }
      }
  }
%    \end{macrocode}
%   Upper casing is easier: just one exception with no context.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_upper_tr:Nnw #1#2
  {
    \int_compare:nNnTF { `#1 } = { "0069 }
      { \@@_change_case_output:Vwn \c__unicode_dotted_I_tl }
      {#2}
  }
%    \end{macrocode}
%   Straight copies.
%    \begin{macrocode}
\cs_new_eq:NN \@@_change_case_lower_az:Nnw \@@_change_case_lower_tr:Nnw
\cs_new_eq:NN \@@_change_case_upper_az:Nnw \@@_change_case_upper_tr:Nnw
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}[aux, EXP]{\@@_change_case_lower_lt:Nnw}
% \begin{macro}[aux, EXP]{\@@_change_case_lower_lt:nNnw}
% \begin{macro}[aux, EXP]{\@@_change_case_lower_lt:nnw}
% \begin{macro}[aux, EXP]{\@@_change_case_lower_lt:Nw}
% \begin{macro}[aux, EXP]{\@@_change_case_lower_lt:NNw}
% \begin{macro}[aux, EXP]{\@@_change_case_upper_lt:Nnw}
% \begin{macro}[aux, EXP]{\@@_change_case_upper_lt:nnw}
% \begin{macro}[aux, EXP]{\@@_change_case_upper_lt:Nw}
% \begin{macro}[aux, EXP]{\@@_change_case_upper_lt:NNw}
%   For  Lithuanian, the issue to be dealt with is dots over lower case
%   letters: these should be present if there is another accent. That means
%   that there is some work to do when lower casing I and J. The first step
%   is a simple match attempt: \cs{c_@@_accents_lt_tl} contains
%   accented upper case letters which should gain a dot-above char in their
%   lower case form. This is done using \texttt{f}-type expansion so only one
%   pass is needed to find if it works or not. If there was no hit, the second
%   stage is to check for I, J and I-ogonek, and if the current char is a
%   match to look for a following accent.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_lower_lt:Nnw #1
  {
    \exp_args:Nf \@@_change_case_lower_lt:nNnw
      { \str_case:nVF #1 \c__unicode_accents_lt_tl \exp_stop_f: }
      #1
  }
\cs_new:Npn \@@_change_case_lower_lt:nNnw #1#2
  {
    \tl_if_blank:nTF {#1}
      {
        \exp_args:Nf \@@_change_case_lower_lt:nnw
          {
            \int_case:nnF {`#2}
              {
                { "0049 } i
                { "004A } j
                { "012E } \c__unicode_i_ogonek_tl
              }
              \exp_stop_f:
          }
      }
      {
        \@@_change_case_output:nwn {#1}
        \use_none:n
      }
  }
\cs_new:Npn \@@_change_case_lower_lt:nnw #1#2
  {
    \tl_if_blank:nTF {#1}
      {#2}
      {
        \@@_change_case_output:nwn {#1}
        \@@_change_case_lower_lt:Nw
      }
  }
%    \end{macrocode}
%   Grab the next char and see if it is one of the accents used in Lithuanian:
%   if it is, add the dot-above char into the output.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_lower_lt:Nw #1#2 \q_recursion_stop
  {
    \tl_if_head_is_N_type:nT {#2}
      { \@@_change_case_lower_lt:NNw }
    #1 #2 \q_recursion_stop
  }
\cs_new:Npn \@@_change_case_lower_lt:NNw #1#2#3 \q_recursion_stop
  {
    \@@_change_case_if_expandable:NTF #2
      {
        \exp_after:wN \@@_change_case_lower_lt:Nw \exp_after:wN #1 #2
          #3 \q_recursion_stop
      }
      {
        \bool_if:nT
          {
            ! \token_if_cs_p:N #2
            &&
              (
                   \int_compare_p:nNn { `#2 } = { "0300 }
                || \int_compare_p:nNn { `#2 } = { "0301 }
                || \int_compare_p:nNn { `#2 } = { "0303 }
              )
          }
          { \@@_change_case_output:Vwn \c__unicode_dot_above_tl }
        #1 #2#3 \q_recursion_stop
      }
  }
%    \end{macrocode}
%   For upper casing, the test required is for a dot-above char after an I,
%   J or I-ogonek. First a test for the appropriate letter, and if found a
%   look-ahead and potentially one token dropped.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_upper_lt:Nnw #1
  {
    \exp_args:Nf \@@_change_case_upper_lt:nnw
      {
        \int_case:nnF {`#1}
          {
            { "0069 } I
            { "006A } J
            { "012F } \c__unicode_I_ogonek_tl
          }
          \exp_stop_f:
      }
  }
\cs_new:Npn \@@_change_case_upper_lt:nnw #1#2
  {
    \tl_if_blank:nTF {#1}
      {#2}
      {
        \@@_change_case_output:nwn {#1}
        \@@_change_case_upper_lt:Nw
      }
  }
\cs_new:Npn \@@_change_case_upper_lt:Nw #1#2 \q_recursion_stop
  {
    \tl_if_head_is_N_type:nT {#2}
      { \@@_change_case_upper_lt:NNw }
    #1 #2 \q_recursion_stop
  }
\cs_new:Npn \@@_change_case_upper_lt:NNw #1#2#3 \q_recursion_stop
  {
    \@@_change_case_if_expandable:NTF #2
      {
        \exp_after:wN \@@_change_case_upper_lt:Nw \exp_after:wN #1 #2
          #3 \q_recursion_stop
      }
      {
        \bool_if:nTF
          {
               ! \token_if_cs_p:N #2
            && \int_compare_p:nNn { `#2 } = { "0307 }
          }
          { #1 }
          { #1 #2 }
        #3 \q_recursion_stop
      }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\@@_change_case_upper_de-alt:Nnw}
%   A simple alternative version for German.
%    \begin{macrocode}
\cs_new:cpn { @@_change_case_upper_de-alt:Nnw } #1#2
  {
    \int_compare:nNnTF { `#1 } = { 223 }
      { \@@_change_case_output:Vwn \c__unicode_upper_Eszett_tl }
      {#2}
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[EXP, int]{\__unicode_codepoint_to_UTFviii:n}
% \begin{macro}[EXP, aux]{\__unicode_codepoint_to_UTFviii_auxi:n}
% \begin{macro}[EXP, aux]{\__unicode_codepoint_to_UTFviii_auxii:Nnn}
% \begin{macro}[EXP, aux]{\__unicode_codepoint_to_UTFviii_auxiii:n}
%   This code will convert a codepoint into the correct UTF-8 representation.
%   As there are a variable number of octets, the result starts with the
%   numeral |1|--|4| to indicate the nature of the returned value. Note that
%   this code will cover the full range even though at this stage it is not
%   required here. Also note that longer-term this is likely to need a public
%   interface and/or moving to \pkg{l3str} (see experimental string
%   conversions). In terms of the algorithm itself, see
%   \url{https://en.wikipedia.org/wiki/UTF-8} for the octet pattern.
%    \begin{macrocode}
\cs_new:Npn \__unicode_codepoint_to_UTFviii:n #1
  {
    \exp_args:Nf \__unicode_codepoint_to_UTFviii_auxi:n
      { \int_eval:n {#1} }
  }
\cs_new:Npn \__unicode_codepoint_to_UTFviii_auxi:n #1
  {
    \if_int_compare:w #1 > "80 ~
      \if_int_compare:w #1 < "800 ~
        2
        \__unicode_codepoint_to_UTFviii_auxii:Nnn C {#1} { 64 }
        \__unicode_codepoint_to_UTFviii_auxiii:n {#1}
      \else:
        \if_int_compare:w #1 < "10000 ~
          3
          \__unicode_codepoint_to_UTFviii_auxii:Nnn E {#1} { 64 * 64 }
          \__unicode_codepoint_to_UTFviii_auxiii:n {#1}
          \__unicode_codepoint_to_UTFviii_auxiii:n
            { \int_div_truncate:nn {#1} { 64 } }
        \else:
          4
          \__unicode_codepoint_to_UTFviii_auxii:Nnn F
            {#1} { 64 * 64 * 64 }
          \__unicode_codepoint_to_UTFviii_auxiii:n
            { \int_div_truncate:nn {#1} { 64 * 64 } }
          \__unicode_codepoint_to_UTFviii_auxiii:n
            { \int_div_truncate:nn {#1} { 64 } }
          \__unicode_codepoint_to_UTFviii_auxiii:n {#1}

        \fi:
      \fi:
    \else:
      1 {#1}
    \fi:
  }
\cs_new:Npn \__unicode_codepoint_to_UTFviii_auxii:Nnn #1#2#3
  { { \int_eval:n { "#10 + \int_div_truncate:nn {#2} {#3} } } }
\cs_new:Npn \__unicode_codepoint_to_UTFviii_auxiii:n #1
  { { \int_eval:n { \int_mod:nn {#1} { 64 } + 128 } } }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{variable}
%   {
%     \c__unicode_std_sigma_tl    ,
%     \c__unicode_final_sigma_tl  ,
%     \c__unicode_accents_lt_tl   ,
%     \c__unicode_dot_above_tl    ,
%     \c__unicode_upper_Eszett_tl
%   }
%   The above needs various special token lists containg pre-formed characters.
%   This set are only available in Unicode engines, with no-op definitions
%   for $8$-bit use.
%    \begin{macrocode}
\cs_if_exist:NTF \utex_char:D
  {
    \tl_const:Nx \c__unicode_std_sigma_tl    { \utex_char:D "03C3 ~ }
    \tl_const:Nx \c__unicode_final_sigma_tl  { \utex_char:D "03C2 ~ }
    \tl_const:Nx \c__unicode_accents_lt_tl
      {
        \utex_char:D "00CC ~
          { \utex_char:D "0069 ~ \utex_char:D "0307 ~ \utex_char:D "0300 ~ }
        \utex_char:D "00CD ~
          { \utex_char:D "0069 ~ \utex_char:D "0307 ~ \utex_char:D "0301 ~ }
        \utex_char:D "0128 ~
          { \utex_char:D "0069 ~ \utex_char:D "0307 ~ \utex_char:D "0303 ~ }
      }
    \tl_const:Nx \c__unicode_dot_above_tl    { \utex_char:D "0307 ~ }
    \tl_const:Nx \c__unicode_upper_Eszett_tl { \utex_char:D "1E9E ~ }
  }
  {
      \tl_const:Nn \c__unicode_std_sigma_tl    { }
      \tl_const:Nn \c__unicode_final_sigma_tl  { }
      \tl_const:Nn \c__unicode_accents_lt_tl   { }
      \tl_const:Nn \c__unicode_dot_above_tl    { }
      \tl_const:Nn \c__unicode_upper_Eszett_tl { }
  }
%    \end{macrocode}
% \end{variable}
% \begin{variable}
%   {
%     \c__unicode_dotless_i_tl    ,
%     \c__unicode_dotted_I_tl     ,
%     \c__unicode_i_ogonek_tl     ,
%     \c__unicode_I_ogonek_tl     ,
%   }
%  For cases where there is an $8$-bit option in the |T1| font set up,
%  a variant is provided in both cases.
%    \begin{macrocode}
\group_begin:
  \cs_if_exist:NTF \utex_char:D
    {
      \cs_set_protected:Npn \@@_tmp:w #1#2
        { \tl_const:Nx #1 { \utex_char:D "#2 ~ } }
    }
    {
      \cs_set_protected:Npn \@@_tmp:w #1#2
        {
          \group_begin:
            \cs_set_protected:Npn \@@_tmp:w ##1##2##3
              {
                \tl_const:Nx #1
                  {
                    \exp_after:wN \exp_after:wN \exp_after:wN
                      \exp_not:N \__char_generate:nn {##2} { 13 }
                    \exp_after:wN \exp_after:wN \exp_after:wN
                      \exp_not:N \__char_generate:nn {##3} { 13 }
                  }
              }
            \tl_set:Nx \l_@@_internal_a_tl
              { \__unicode_codepoint_to_UTFviii:n {"#2} }
            \exp_after:wN \@@_tmp:w \l_@@_internal_a_tl
          \group_end:
        }
    }
  \@@_tmp:w \c__unicode_dotless_i_tl { 0131 }
  \@@_tmp:w \c__unicode_dotted_I_tl  { 0130 }
  \@@_tmp:w \c__unicode_i_ogonek_tl  { 012F }
  \@@_tmp:w \c__unicode_I_ogonek_tl  { 012E }  
\group_end:
%    \end{macrocode}
% \end{variable}
%
% For $8$-bit engines we now need to define the case-change data for
% the multi-octet mappings. These need a list of what code points are
% doable in |T1| so the list is hard coded (there's no saving in loading
% the mappings dynamically). All of the straight-forward ones have two
% octets, so that is taken as read.
%    \begin{macrocode}
\group_begin:
  \bool_if:nT
    {
      \sys_if_engine_pdftex_p: || \sys_if_engine_uptex_p:
    }
    {
      \cs_set_protected:Npn \@@_loop:nn #1#2
        {
          \quark_if_recursion_tail_stop:n {#1}
          \tl_set:Nx \l_@@_internal_a_tl
            {
              \__unicode_codepoint_to_UTFviii:n {"#1}
              \__unicode_codepoint_to_UTFviii:n {"#2}
            }
          \exp_after:wN \@@_tmp:w \l_@@_internal_a_tl
          \@@_loop:nn
        }
      \cs_set_protected:Npn \@@_tmp:w #1#2#3#4#5#6
        {
          \tl_const:cx
            {
              c__unicode_lower_
              \char_generate:nn {#2} { 12 }
              \char_generate:nn {#3} { 12 }
              _tl
            }
            {
              \exp_after:wN \exp_after:wN \exp_after:wN
                \exp_not:N \__char_generate:nn {#5} { 13 }
              \exp_after:wN \exp_after:wN \exp_after:wN
                \exp_not:N \__char_generate:nn {#6} { 13 }
            }
          \tl_const:cx
            {
              c__unicode_upper_
              \char_generate:nn {#5} { 12 }
              \char_generate:nn {#6} { 12 }
              _tl
            }
            {
              \exp_after:wN \exp_after:wN \exp_after:wN
                \exp_not:N \__char_generate:nn {#2} { 13 }
              \exp_after:wN \exp_after:wN \exp_after:wN
                \exp_not:N \__char_generate:nn {#3} { 13 }
            }
        }
      \@@_loop:nn
        { 00C0 } { 00E0 }
        { 00C2 } { 00E2 }
        { 00C3 } { 00E3 }
        { 00C4 } { 00E4 }
        { 00C5 } { 00E5 }
        { 00C6 } { 00E6 }
        { 00C7 } { 00E7 }
        { 00C8 } { 00E8 }
        { 00C9 } { 00E9 }
        { 00CA } { 00EA }
        { 00CB } { 00EB }
        { 00CC } { 00EC }
        { 00CD } { 00ED }
        { 00CE } { 00EE }
        { 00CF } { 00EF }
        { 00D0 } { 00F0 }
        { 00D1 } { 00F1 }
        { 00D2 } { 00F2 }
        { 00D3 } { 00F3 }
        { 00D4 } { 00F4 }
        { 00D5 } { 00F5 }
        { 00D6 } { 00F6 }
        { 00D8 } { 00F8 }
        { 00D9 } { 00F9 }
        { 00DA } { 00FA }
        { 00DB } { 00FB }
        { 00DC } { 00FC }
        { 00DD } { 00FD }
        { 00DE } { 00FE }
        { 0100 } { 0101 }
        { 0102 } { 0103 }
        { 0104 } { 0105 }
        { 0106 } { 0107 }
        { 0108 } { 0109 }
        { 010A } { 010B }
        { 010C } { 010D }
        { 010E } { 010F }
        { 0110 } { 0111 }
        { 0112 } { 0113 }
        { 0114 } { 0115 }
        { 0116 } { 0117 }
        { 0118 } { 0119 }
        { 011A } { 011B }
        { 011C } { 011D }
        { 011E } { 011F }
        { 0120 } { 0121 }
        { 0122 } { 0123 }
        { 0124 } { 0125 }
        { 0128 } { 0129 }
        { 012A } { 012B }
        { 012C } { 012D }
        { 012E } { 012F }
        { 0132 } { 0133 }
        { 0134 } { 0135 }
        { 0136 } { 0137 }
        { 0139 } { 013A }
        { 013B } { 013C }
        { 013E } { 013F }
        { 0141 } { 0142 }
        { 0143 } { 0144 }
        { 0145 } { 0146 }
        { 0147 } { 0148 }
        { 014A } { 014B }
        { 014C } { 014D }
        { 014E } { 014F }
        { 0150 } { 0151 }
        { 0152 } { 0153 }
        { 0154 } { 0155 }
        { 0156 } { 0157 }
        { 0158 } { 0159 }
        { 015A } { 015B }
        { 015C } { 015D }
        { 015E } { 015F }
        { 0160 } { 0161 }
        { 0162 } { 0163 }
        { 0164 } { 0165 }
        { 0168 } { 0169 }
        { 016A } { 016B }
        { 016C } { 016D }
        { 016E } { 016F }
        { 0170 } { 0171 }
        { 0172 } { 0173 }
        { 0174 } { 0175 }
        { 0176 } { 0177 }
        { 0178 } { 00FF }
        { 0179 } { 017A }
        { 017B } { 017C }
        { 017D } { 017E }
        { 01CD } { 01CE }
        { 01CF } { 01D0 }
        { 01D1 } { 01D2 }
        { 01D3 } { 01D4 }
        { 01E2 } { 01E3 }
        { 01E6 } { 01E7 }
        { 01E8 } { 01E9 }
        { 01EA } { 01EB }
        { 01F4 } { 01F5 }
        { 0218 } { 0219 }
        { 021A } { 021B }
        \q_recursion_tail ?
        \q_recursion_stop
      \cs_set_protected:Npn \@@_tmp:w #1#2#3
        {
          \group_begin:
            \cs_set_protected:Npn \@@_tmp:w ##1##2##3
              {
                \tl_const:cx
                  {
                    c__unicode_ #3 _
                    \char_generate:nn {##2} { 12 }
                    \char_generate:nn {##3} { 12 }
                    _tl
                  }
                    {#2}
              }
            \tl_set:Nx \l_@@_internal_a_tl
              { \__unicode_codepoint_to_UTFviii:n { "#1 } }
            \exp_after:wN \@@_tmp:w \l_@@_internal_a_tl
          \group_end:
        }
      \@@_tmp:w { 00DF } { SS } { upper }
      \@@_tmp:w { 00DF } { Ss } { title }
      \@@_tmp:w { 0131 } { I }  { upper }
    }
  \group_end:
%    \end{macrocode}
%
% The (fixed) look-up mappings for letter-like control sequences.
%    \begin{macrocode}
\group_begin:
  \cs_set_protected:Npn \@@_change_case_setup:NN #1#2
    {
      \quark_if_recursion_tail_stop:N #1
      \tl_const:cn { c_@@_change_case_lower_ \token_to_str:N #1 _tl } { #2 }
      \tl_const:cn { c_@@_change_case_upper_ \token_to_str:N #2 _tl } { #1 }
      \@@_change_case_setup:NN
    }
  \@@_change_case_setup:NN
  \AA \aa
  \AE \ae
  \DH \dh
  \DJ \dj
  \IJ \ij
  \L  \l
  \NG \ng
  \O  \o
  \OE \oe
  \SS \ss
  \TH \th
  \q_recursion_tail ?
  \q_recursion_stop
  \tl_const:cn { c_@@_change_case_upper_ \token_to_str:N \i _tl } { I }
  \tl_const:cn { c_@@_change_case_upper_ \token_to_str:N \j _tl } { J }
\group_end:
%    \end{macrocode}
%
% \begin{variable}{\l_tl_case_change_accents_tl}
%   A list of accents to leave alone.
%    \begin{macrocode}
\tl_new:N \l_tl_case_change_accents_tl
\tl_set:Nn \l_tl_case_change_accents_tl
  { \" \' \. \^ \` \~ \c \H \k \r \t \u \v }
%    \end{macrocode}
% \end{variable}
%
% \begin{macro}[aux, EXP]{\@@_mixed_case:nn}
% \begin{macro}[aux, EXP]{\@@_mixed_case_aux:nn}
% \begin{macro}[aux, EXP]{\@@_mixed_case_loop:wn}
% \begin{macro}[aux, EXP]{\@@_mixed_case_group:nwn}
% \begin{macro}[aux, EXP]{\@@_mixed_case_space:wn}
% \begin{macro}[aux, EXP]{\@@_mixed_case_N_type:Nwn}
% \begin{macro}[aux, EXP]{\@@_mixed_case_N_type:NNNnn}
% \begin{macro}[aux, EXP]{\@@_mixed_case_N_type:Nnn}
% \begin{macro}[aux, EXP]{\@@_mixed_case_letterlike:Nw}
% \begin{macro}[aux, EXP]{\@@_mixed_case_char:N}
% \begin{macro}[aux, EXP]{\@@_mixed_case_skip:N}
% \begin{macro}[aux, EXP]{\@@_mixed_case_skip:NN}
% \begin{macro}[aux, EXP]{\@@_mixed_case_skip_tidy:Nwn}
% \begin{macro}[aux, EXP]{\@@_mixed_case_char:nN}
%   Mixed (title) casing requires some custom handling of the case changing
%   of the first letter in the input followed by a switch to the normal
%   lower casing routine. That could be covered by passing a set of functions
%   to generic routines, but at the cost of making the process rather opaque.
%   Instead, the approach taken here is to use a dedicated set of functions
%   which keep the different loop requirements clearly separate.
%
%   The main loop looks for the first \enquote{real} char in the input
%   (skipping any pre-letter chars). Once one is found, it is case changed to
%   upper case but first checking that there is not an entry in the exceptions
%   list. Note that simply grabbing the first token in the input is no good
%   here: it can't handle pre-letter tokens or any special treatment of the
%   first letter found (\emph{e.g.}~words starting with \texttt{i} in
%   Turkish). Spaces at the start of the input are passed through without
%   counting as being the \enquote{start} of the first word, while a brace
%   group is assumed to be contain the first char with everything after the
%   brace therefore lower cased.
%    \begin{macrocode}
\cs_new:Npn \@@_mixed_case:nn #1#2
  {
    \etex_unexpanded:D \exp_after:wN
      {
        \exp:w
        \@@_mixed_case_aux:nn {#1} {#2}
      }
  }
\cs_new:Npn \@@_mixed_case_aux:nn #1#2
  {
    \group_align_safe_begin:
    \@@_mixed_case_loop:wn
      #2 \q_recursion_tail \q_recursion_stop {#1}
    \@@_change_case_result:n { }
  }
\cs_new:Npn \@@_mixed_case_loop:wn #1 \q_recursion_stop
  {
    \tl_if_head_is_N_type:nTF {#1}
      { \@@_mixed_case_N_type:Nwn }
      {
        \tl_if_head_is_group:nTF {#1}
          { \@@_mixed_case_group:nwn }
          { \@@_mixed_case_space:wn }
      }
    #1 \q_recursion_stop
  }
\cs_new:Npn \@@_mixed_case_group:nwn #1#2 \q_recursion_stop #3
  {
    \@@_change_case_output:own
      {
        \exp_after:wN
          {
            \exp:w
            \@@_mixed_case_aux:nn {#3} {#1}
          }
      }
    \@@_change_case_loop:wnn #2 \q_recursion_stop { lower } {#3}
  }
\exp_last_unbraced:NNo \cs_new:Npn \@@_mixed_case_space:wn \c_space_tl
  {
    \@@_change_case_output:nwn { ~ }
    \@@_mixed_case_loop:wn
  }
\cs_new:Npn \@@_mixed_case_N_type:Nwn #1#2 \q_recursion_stop
  {
    \quark_if_recursion_tail_stop_do:Nn #1
      { \@@_change_case_end:wn }
    \exp_after:wN \@@_mixed_case_N_type:NNNnn
      \exp_after:wN #1 \l_tl_case_change_math_tl
      \q_recursion_tail ? \q_recursion_stop {#2}
  }
\cs_new:Npn \@@_mixed_case_N_type:NNNnn #1#2#3
  {
    \quark_if_recursion_tail_stop_do:Nn #2
      { \@@_mixed_case_N_type:Nnn #1 }
    \token_if_eq_meaning:NNTF #1 #2
      {
        \use_i_delimit_by_q_recursion_stop:nw
          {
            \@@_change_case_math:NNNnnn
              #1 #3 \@@_mixed_case_loop:wn
          }
      }
      { \@@_mixed_case_N_type:NNNnn #1 }
  }
%    \end{macrocode}
%   The business end of the loop is here: there is first a need to deal
%   with any control sequence cases before looking for characters to skip.
%   If there is a hit for a letter-like control sequence, switch to lower
%   casing.
%    \begin{macrocode}
\cs_new:Npn \@@_mixed_case_N_type:Nnn #1#2#3
  {
    \token_if_cs:NTF #1
      {
        \@@_change_case_cs_letterlike:Nnn #1 { upper }
          { \@@_mixed_case_letterlike:Nw }
        \@@_mixed_case_loop:wn #2 \q_recursion_stop {#3}
      }
      {
        \@@_mixed_case_char:Nn #1 {#3}
        \@@_change_case_loop:wnn #2 \q_recursion_stop { lower } {#3}
      }
  }
\cs_new:Npn \@@_mixed_case_letterlike:Nw #1#2 \q_recursion_stop
  { \@@_change_case_loop:wnn #2 \q_recursion_stop { lower } }
%    \end{macrocode}
%   As detailed above, handling a mixed case char means first looking for
%   exceptions then treating as an upper cased letter, but with a list of
%   tokens to skip over too.
%    \begin{macrocode}
\cs_new:Npn \@@_mixed_case_char:Nn #1#2
  {
    \cs_if_exist_use:cF { @@_change_case_mixed_ #2 :Nnw }
      {
        \cs_if_exist_use:cF { @@_change_case_upper_ #2 :Nnw }
          { \use_ii:nn }
      }
        #1
        { \@@_mixed_case_skip:N #1 }
  }
\cs_new:Npn \@@_mixed_case_skip:N #1
  {
    \exp_after:wN \@@_mixed_case_skip:NN
      \exp_after:wN #1 \l_tl_mixed_case_ignore_tl
      \q_recursion_tail \q_recursion_stop
  }
\cs_new:Npn \@@_mixed_case_skip:NN #1#2
  {
    \quark_if_recursion_tail_stop_do:nn {#2}
      { \@@_mixed_case_char:N #1 }
    \int_compare:nNnT { `#1 }  = { `#2 }
      {
        \use_i_delimit_by_q_recursion_stop:nw
          {
            \@@_change_case_output:nwn {#1}
            \@@_mixed_case_skip_tidy:Nwn
          }
      }
    \@@_mixed_case_skip:NN #1
  }
\cs_new:Npn \@@_mixed_case_skip_tidy:Nwn #1#2 \q_recursion_stop #3
  {
    \@@_mixed_case_loop:wn #2 \q_recursion_stop
  }
\cs_new:Npn \@@_mixed_case_char:N #1
  {
    \cs_if_exist:cTF { c__unicode_title_  #1 _tl }
      {
        \@@_change_case_output:fwn
          { \tl_use:c { c__unicode_title_ #1 _tl } }
      }
      { \@@_change_case_char:nN { upper } #1 }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}[aux, EXP]{\@@_change_case_mixed_nl:Nnw}
% \begin{macro}[aux, EXP]{\@@_change_case_mixed_nl:Nw}
% \begin{macro}[aux, EXP]{\@@_change_case_mixed_nl:NNw}
%   For Dutch, there is a single look-ahead test for \texttt{ij} when
%   title casing. If the appropriate letters are found, produce \texttt{IJ}
%   and gobble the \texttt{j}/\texttt{J}.
%    \begin{macrocode}
\cs_new:Npn \@@_change_case_mixed_nl:Nnw #1
  {
    \bool_if:nTF
      {
           \int_compare_p:nNn { `#1 } = { `i }
        || \int_compare_p:nNn { `#1 } = { `I }
      }
      {
        \@@_change_case_output:nwn { I }
        \@@_change_case_mixed_nl:Nw
      }
  }
\cs_new:Npn \@@_change_case_mixed_nl:Nw #1#2 \q_recursion_stop
  {
    \tl_if_head_is_N_type:nT {#2}
      { \@@_change_case_mixed_nl:NNw }
    #1 #2 \q_recursion_stop
  }
\cs_new:Npn \@@_change_case_mixed_nl:NNw #1#2#3 \q_recursion_stop
  {
    \@@_change_case_if_expandable:NTF #2
      {
        \exp_after:wN \@@_change_case_mixed_nl:Nw \exp_after:wN #1 #2
          #3 \q_recursion_stop
      }
      {
        \bool_if:nTF
          {
            ! ( \token_if_cs_p:N #2 )
            &&
              (
                   \int_compare_p:nNn { `#2 } = { `j }
                || \int_compare_p:nNn { `#2 } = { `J }
              )
          }
          {
            \@@_change_case_output:nwn { J }
            #1
          }
          { #1 #2 }
        #3 \q_recursion_stop
      }
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{variable}{\l_tl_case_change_math_tl}
%   The list of token pairs which are treated as math mode and so
%   not case changed.
%    \begin{macrocode}
\tl_new:N \l_tl_case_change_math_tl
%<*package>
\tl_set:Nn \l_tl_case_change_math_tl
  { $ $ \( \) }
%</package>
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_tl_case_change_exclude_tl}
%   The list of commands for which an argument is not case changed.
%    \begin{macrocode}
\tl_new:N \l_tl_case_change_exclude_tl
%<*package>
\tl_set:Nn \l_tl_case_change_exclude_tl
  { \cite \ensuremath \label \ref }
%</package>
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_tl_mixed_case_ignore_tl}
%   Characters to skip over when finding the first letter in a word to be
%   mixed cased.
%    \begin{macrocode}
\tl_new:N \l_tl_mixed_case_ignore_tl
\tl_set:Nx \l_tl_mixed_case_ignore_tl
  {
    ( % )
    [ % ]
    \cs_to_str:N \{ % \}
    `
    -
  }
%    \end{macrocode}
% \end{variable}
%
%
% \begin{macro}{\tl_log:N, \tl_log:c}
%   Redirect output of \cs{tl_show:N} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \tl_log:N
  { \__msg_log_next: \tl_show:N }
\cs_generate_variant:Nn \tl_log:N { c }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\tl_log:n}
%   Redirect output of \cs{tl_show:n} to the log.
%    \begin{macrocode}
\cs_new_protected:Npn \tl_log:n
  { \__msg_log_next: \tl_show:n }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Additions to \pkg{l3tokens}}
%
%    \begin{macrocode}
%<@@=peek>
%    \end{macrocode}
%
% \begin{macro}[TF]{\peek_N_type:}
% \begin{macro}[aux]
%   {\@@_execute_branches_N_type:, \@@_N_type:w, \@@_N_type_aux:nnw}
%   All tokens are \texttt{N}-type tokens, except in four cases:
%   begin-group tokens, end-group tokens, space tokens with character
%   code~$32$, and outer tokens.  Since \cs{l_peek_token} might be
%   outer, we cannot use the convenient \cs{bool_if:nTF} function, and
%   must resort to the old trick of using \tn{ifodd} to expand a set of
%   tests.  The \texttt{false} branch of this test is taken if the token
%   is one of the first three kinds of non-\texttt{N}-type tokens
%   (explicit or implicit), thus we call \cs{@@_false:w}.  In the
%   \texttt{true} branch, we must detect outer tokens, without impacting
%   performance too much for non-outer tokens.  The first filter is to
%   search for \texttt{outer} in the \tn{meaning} of \cs{l_peek_token}.
%   If that is absent, \cs{use_none_delimit_by_q_stop:w} cleans up, and
%   we call \cs{@@_true:w}.  Otherwise, the token can be a non-outer
%   macro or a primitive mark whose parameter or replacement text
%   contains \texttt{outer}, it can be the primitive \tn{outer}, or it
%   can be an outer token.  Macros and marks would have \texttt{ma} in
%   the part before the first occurrence of \texttt{outer}; the meaning
%   of \tn{outer} has nothing after \texttt{outer}, contrarily to outer
%   macros; and that covers all cases, calling \cs{@@_true:w} or
%   \cs{@@_false:w} as appropriate.  Here, there is no \meta{search
%     token}, so we feed a dummy \cs{scan_stop:} to the
%   \cs{@@_token_generic:NNTF} function.
%    \begin{macrocode}
\group_begin:
  \cs_set_protected:Npn \@@_tmp:w #1 \q_stop
    {
      \cs_new_protected:Npn \@@_execute_branches_N_type:
        {
          \if_int_odd:w
              \if_catcode:w \exp_not:N \l_peek_token {   \c_two \fi:
              \if_catcode:w \exp_not:N \l_peek_token }   \c_two \fi:
              \if_meaning:w \l_peek_token \c_space_token \c_two \fi:
              \c_one
            \exp_after:wN \@@_N_type:w
              \token_to_meaning:N \l_peek_token
              \q_mark \@@_N_type_aux:nnw
              #1 \q_mark \use_none_delimit_by_q_stop:w
              \q_stop
            \exp_after:wN \@@_true:w
          \else:
            \exp_after:wN \@@_false:w
          \fi:
        }
      \cs_new_protected:Npn \@@_N_type:w ##1 #1 ##2 \q_mark ##3
        { ##3 {##1} {##2} }
    }
  \exp_after:wN \@@_tmp:w \tl_to_str:n { outer } \q_stop
\group_end:
\cs_new_protected:Npn \@@_N_type_aux:nnw #1 #2 #3 \fi:
  {
    \fi:
    \tl_if_in:noTF {#1} { \tl_to_str:n {ma} }
      { \@@_true:w }
      { \tl_if_empty:nTF {#2} { \@@_true:w } { \@@_false:w } }
  }
\cs_new_protected:Npn \peek_N_type:TF
  { \@@_token_generic:NNTF \@@_execute_branches_N_type: \scan_stop: }
\cs_new_protected:Npn \peek_N_type:T
  { \@@_token_generic:NNT \@@_execute_branches_N_type: \scan_stop: }
\cs_new_protected:Npn \peek_N_type:F
  { \@@_token_generic:NNF \@@_execute_branches_N_type: \scan_stop: }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
%    \begin{macrocode}
%</initex|package>
%    \end{macrocode}
%
% \end{implementation}
%
% \PrintIndex