summaryrefslogtreecommitdiff
path: root/Master/texmf-dist/source/latex/tools/bm.dtx
blob: cf1da612630d565b4e182d8629cb14c1f1b95fae (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
% \iffalse meta-comment
%
% Copyright 1993-2014
%
% The LaTeX3 Project and any individual authors listed elsewhere
% in this file.
%
% This file is part of the Standard LaTeX `Tools Bundle'.
% -------------------------------------------------------
%
% It may be distributed and/or modified under the
% conditions of the LaTeX Project Public License, either version 1.3c
% of this license or (at your option) any later version.
% The latest version of this license is in
%    http://www.latex-project.org/lppl.txt
% and version 1.3c or later is part of all distributions of LaTeX
% version 2005/12/01 or later.
%
% The list of all files belonging to the LaTeX `Tools Bundle' is
% given in the file `manifest.txt'.
%
% \fi
%
% \iffalse
%% Copyright 1996 1997 1998 1999 2002 2003 2004 David Carlisle Frank Mittelbach
%%
%% Development of this package was commissioned by Y&Y Inc.
%% http://www.yandy.com
%
%
%<*dtx>
          \ProvidesFile{bm.dtx}
%</dtx>
%<package>\NeedsTeXFormat{LaTeX2e}
%<package>\ProvidesPackage{bm}
%<driver>\ProvidesFile{bm.drv}
% \fi
%         \ProvidesFile{bm.dtx}
          [2014/10/28 v1.1c Bold Symbol Support (DPC/FMi)]
%
% \iffalse
%<*driver>
\documentclass{ltxdoc}
\usepackage{bm}
\begin{document}
\DocInput{bm.dtx}
\end{document}
%</driver>
% \fi
%
% \CheckSum{944}
%
% \GetFileInfo{bm.dtx}
%
% \title{The \textsf{bm} package\thanks{This file
%         has version number \fileversion, last
%         revised \filedate.}\ \thanks{Development of this package
%         was commissioned by Y\&Y.}}
%
% \date{\filedate}
% \author{David Carlisle with support by Frank Mittelbach}
%  \MaintainedByLaTeXTeam{tools}
%  \maketitle
%
% \section{Introduction}
%
% This package defines commands to access bold math
% symbols. The basic command is |\bm| which may be used to make the
% math expression in its argument be typeset using bold fonts.
%
% The syntax of |\bm| is:\\
% |\bm|\marg{math expression}\\
% So |$\alpha \not= \bm{\alpha}$| produces $\alpha \not= \bm{\alpha}$.
%
% |\bm| goes to some trouble to preserve the spacing, so that for
% instance |\bm<| is a bold $\bm<$ but with the correct |\mathrel|
% spacing that \TeX\ gives to $<$. The calculations that \TeX\ needs to
% do for |\bm| can be quite involved and so a definition form is
% provided.
%
% |\DeclareBoldMathCommand|%^^A
%       \oarg{math version}\marg{cmd}\marg{math expression}
%
% Defines |\cmd| to be the bold form of the math expression.
% The \meta{math version} defaults to `bold' (i.e., |\boldmath|).
%
% For relatively simple expressions, the resulting definitions are very
% efficient, for instance after:\\
% |\DeclareBoldMathCommand\balpha{\alpha}|\\
% |\balpha| is a single `mathchardef' token producing a bold alpha,
% and so is just as fast to execute as |\alpha|.
%
% The above command is mainly intended for use in packages.
% For occasional use in \LaTeX\ documents, and for compatibility
% with the plain \TeX\ support for the mathtime fonts, a `user-level'
% version, |\bmdefine| is provided that is equivalent to:\\
% |\DeclareBoldMathCommand[bold]|.
%
% If there is a `heavy' math version defined (usually accessed by a
% user-command |\heavymath|) then a similar command |\hm|
% is defined which access these `ultra bold' fonts. Currently this is
% probably only useful with the `mathtime plus' font collection.
% Definitions of commands that use these fonts may be made by
% specifying the optional argument `heavy' to |\DeclareBoldMathCommand|.
% Again an abbreviation, |\hmdefine|, is provided, equivalent to:\\
% |\DeclareBoldMathCommand[heavy]|.
%
% The command names (but not the implementation) are taken from Michael
% Spivak's macros to support the mathtime fonts for plain \TeX. In those
% original macros, the syntax for |\bmdefine| was
% |\bmdefine\balpha{\bm\alpha}| (with a nested |\bm|). This syntax also
% works with this package.
%
% \section{Font allocation}
% In order to access bold fonts in the simplest and quickest possible
% manner, the package normally allocates symbol fonts for bold
% (and possibly heavy) fonts into the `normal' math version.
% By default it allocates at most four fonts for |\bm| and at most
% three fonts for |\hm|. This means that if the mathtime plus font set
% is being used, seven additional symbol fonts will be used, in addition
% to the basic four that \LaTeX\ already declares. The mathtime
% package also declares an extra symbol font, bringing the total to
% twelve. The maximum number of symbol \emph{and} math alphabet fonts
% that can be used in a math version is sixteen. So the above allocation
% scheme does not leave room for many extra math symbols (such as the
% AMS symbols) or math alphabets (such as |\mathit|).
%
% Before loading the \textsf{bm} package you may define |\bmmax|
% and |\hmmax| to be suitable values, for instance you may want
% to set |\newcommand\hmmax{0}| if you will not be using |\hm|
% much, but you do have a heavy math version defined.
%
% Even if |\bmmax| is set to zero, |\bm| will still access the correct
% bold fonts (by accessing the fonts via |\boldmath|) but this method
% is slower, and does not work with delimiters. Delimiters can only be
% made bold if the bold font has been allocated.
%
% Conversely if you have a non standard font set that makes available
% extra math delimiters and accents in bold and medium weights you may
% want to \emph{increase} |\bmmax| so that fonts are allocated for
% your font set.
%
% \section{Features}
% In most cases this package should work in a fairly self explanatory
% way, but there are some things that might not be obvious.
%
% \subsection{Interaction with Math Alphabet Commands}
%
% As mentioned above, |\bm| goes to some trouble to try to make a
% command that is just like its argument, but using a bold font.
% This does not always produce the effect that you might expect.
%\begin{verbatim}
% $1 g \bm{g}$
% $2 \mathrm{g \bm{g}}$
% $3 {g} \bm{{g}}$
% $4 \mathrm{{g} \bm{{g}}}$
% $5 \mathrm{g} \bm{\mathrm{g}}$
%\end{verbatim}
% produces the following:
% \begin{flushleft}
% $1 g \bm{g}$
% $2 \mathrm{g \bm{g}}$
% $3 {g} \bm{{g}}$
% $4 \mathrm{{g} \bm{{g}}}$
% $5 \mathrm{g} \bm{\mathrm{g}}$
% \end{flushleft}
% In math mode `g' is effectively a command that produces the letter `g'
% from the `letters' alphabet, unless a Math Alphabet command is in
% effect, in which case the `g' comes from the specified alphabet.
% |\bm{g}| makes an equivalent command, but which defaults to a bold
% letter alphabet. So in the first example |\bm{g}| is bold math italic,
% but in the second example the |\mathrm| applies to both |g| and
% |\bm{g}| in the same way, and so they are both roman.
%
% |\bm| only inspects the `top level' definition of a command, for more
% complicated expressions, and anything inside a |{ }| group, |\bm|
% forces bold fonts by essentially the same (slow) technique used by the
% AMS |\boldsymbol| command (but |\bm| still takes more care of the
% spacing). So the third example produces identical output to the first
% (but \TeX\ takes more time producing it).
%
% In the fourth example the |\mathrm{\bm{g}}| is essentially
% equivalent to |\mathrm{\mbox{\boldmath$g$}}|. Currently math alphabet
% settings are not passed down to `nested' math lists, and so in this
% example, the |\mathrm| has no effect, and a bold math italic $\bm g$
% is obtained.
%
% Similarly the last example is equivalent to
% |$\mbox{\boldmath$\mathrm{g}$}}| and so in this case, one obtains a
% bold roman \textbf{g}.
%
% \subsection{Delimiters}
% \TeX\ can treat character tokens in two\footnote
%  {Well more than two really.}
% ways. If there is a preceding
% |\left| or |\right| it can treat them as a delimiter, otherwise it can
% treat them as a standard character. For example |\left<\right>|
% produces $\left<\right>$, which is totally different from |<>|, which
% produces $<>$.
%
% \TeX\ can only do this for character tokens. Commands such as
% |\langle| do not act in this way. This means that |\bm| has to decide
% whether to treat a character as a delimiter or not. The rule it uses
% is, it makes a delimiter command for a character if the previous
% token in the argument was |\left| or |\right|. So |\left\bm{<}| does
% not work, but |\bm{\left<}| does.
%
% \subsection{Command Arguments}
%
% Normally if a command takes arguments the full command, including
% any arguments, should be included in |\bm|.
%
% So |\bm{\overbrace{abc}}| (producing \smash{$\bm{\overbrace{abc}}$})
% not
% |\bm{\overbrace}{abc}|. If you do not include all the arguments you
% will typically get the error message:\\
% |Runaway argument?|\\
% |! Forbidden control sequence found while scanning use of|
%                                                      \texttt{\ldots}
%
% However commands defined in terms of the \TeX\ accent and
% radical primitives \emph{may} be used without their arguments.
% So |\bm{\hat}{a}| produces $\bm{\hat}{a}$, a bold accent over a
% non-bold $a$ (compare $\hat{a}$)
%  whereas |\bm{\hat{a}}| makes both the $a$ and the accent bold,
% $\bm{\hat{a}}$.
% Similarly, although the \LaTeX\ command |\sqrt| must be used with its
% arguments, |\sqrtsign| may be used as in |\bm\sqrtsign{abc}| to
% produce $\bm\sqrtsign{abc}$ rather than  $\sqrtsign{abc}$ or
%  $\bm{\sqrtsign{abc}}$
%
% If you really need to make a command with arguments use bold fonts
% without making all of the arguments bold, you can explicitly
% reset the math version in the argument, eg:\\
% $\begin{array}{ccc}
% |\sqrt{xyz}|&|\bm{\sqrt{xyz}}|&|\bm{\sqrt{\mbox{\unboldmath$xyz$}}}|\\
% \sqrt{xyz}& \bm{\sqrt{xyz}}&  \bm{\sqrt{\mbox{\unboldmath$xyz$}}}
% \end{array}$
%
% \subsection{Bold fonts}
% This package interrogates the font allocations of the bold and heavy
% math versions, to determine which bold fonts are available.
% This means that it is best to load the package \emph{after}
% any packages that define new symbol fonts, or (like the
% \textsf{mathtime} package) completely change the symbol font
% allocations.
%
% If no bold font appears to be available for a particular symbol,
% |\bm| will use `poor man's bold' that is, overprinting the same
% character in slightly offset
% positions to give an appearance of boldness.
%
% In the standard Computer Modern font set, there is no bold
% `large symbols' font. In the `mathptm' and (standard) mathtime
% font sets there are no bold math fonts. In the `mathtime plus'
% font set there are suitable fonts for bold and heavy math setting,
% and so |\bm| and |\hm| work well. Similarly in the basic Lucida
% New Math font set there are no bold math fonts, so |\bm| will
% use `poor man's bold. However if the Lucida Expert set is used,
% Then |\bm| will detect, and use the bold math fonts that are
% available.
%
% As discussed above, one may set |\bmmax| higher or lower than its
% default value of four to control the font allocation system. Finer
% control may be gained by explicitly declaring bold symbol fonts.
% Suppose you have a symbol font `xyz' that is available in medium and
% bold weights, then you would declare this to \LaTeX\ via:\\
% |\DeclareSymbolFont{extras}  {OMS}{xyz}{m}{n}|\\
% |\SetSymbolFont{extras}{bold}{OMS}{xyz}{bx}{n}|\\
% At this point the symbols will be available in the normal math
% version, and their bold variants in |\boldmath|. If you also
% declare:\\
% |\DeclareSymbolFont{boldextras}{OMS}{xyz}{bx}{n}|\\
% That is, declare a symbol font whose name is formed by prefixing
% `bold' (or `heavy') to an existing symbol font, then |\bm| (or
% |\hm|) will use this font directly, rather then accessing the
% `extras' symbol font via |\boldmath|.
%
% \subsection{Strange failures}
% In order to get the correct spacing, |\bm| has to `investigate' the
% definition of the commands in its argument. It is possible that
% some strange constructions could `confuse' this investigation.
% If this happens then \LaTeX\ will almost certainly stop with a strange
% error. This should not happen with any of the math symbols
% defined in the base \LaTeX\ or AMS distributions, or any commands
% defined in terms of those symbols using normal \LaTeX\ math
% constructs. However if some command does fail to work inside |\bm|
% you should always be able to surround it with an extra set of braces
% |\bm{{\cmd}}| rather than |\bm{\cmd}|. |\bm| will not then attempt
% to set the correct spacing, so you may need to set it explicitly,
% for instance, for a relation, |\bm{\mathrel{\cmd}}|.
%
% \subsection{AMS package \textsf{amsbsy}}
% The |\bm| command shares some functionality with the |\boldsymbol|
% command from the AMS \LaTeX\ collection. To aid in moving documents
% between these two packages, this package defines |\boldsymbol| and
% |\heavysymbol| as alternative names for |\bm| and |\hm|.
%
% \StopEventually{}
%
% \section{Implementation}
%
% \changes{v0.01}{1996/12/01}
%      {Initial DPC attempt,
%       remove all assumptions about mathtime encoding}
% \changes{v0.02}{1996/12/02}
%      {Add \cs{mathchoice} stuff}
% \changes{v0.03}{1996/12/12}
%      {Completely reimplement (again). Add \cs{bmdefine}.}
% \changes{v0.04}{1996/12/12}
%      {Add \cs{hm} and support for special active mathcode.}
% \changes{v0.05}{1996/12/12}
%      {Assorted fixes}
% \changes{v0.99}{1997/01/16}
%      {First public version}
% \changes{v1.0a}{1997/02/14}
%      {First ctan version (same as 0.99b)}
% \changes{v1.0b}{1997/04/14}
%      {Add to tools bundle.}
% \changes{v1.0g}{1999/07/05}
%      {minor doc changes latex/3058}
%
%
%
% The commands |\bm| and |\hm| work by defining a number of additional
% symbol fonts corresponding to the standard ones
%  `operators', `letters', `symbols', and `largesymbols'.
% The names for these symbols fonts are produced by prefixing the usual
% name with `bold' or `heavy'.
%
% For maximum flexibility we get the font definitions by looking in the
% corresponding math versions, i.e., into |\mv@bold| and if defined into
% |\mv@heavy|.
%
%    \begin{macrocode}
%<*package>
%    \end{macrocode}
% \changes{v0.09}{1996/12/19}
%      {Always define \cs{bm} even if no bold math}
%
% \begin{macro}{\bm@table}
% \changes{v0.10}{1997/01/04}
%      {Macro added}
% \changes{v0.12}{1997/01/10}
%      {Ensure do not allocate too many math group slots.}
% \changes{v0.99a}{1997/01/17}
%      {Let \cs{hm} use font allocated for \cs{bm}}
% \changes{v0.99b}{1997/01/19}
%      {Restore NFSS internals}
% \begin{macro}{\bm@boldtable}
% \begin{macro}{\bm@heavytable}
% The table, |\bm@table|, (which is locally |\let| to either the bold
% or heavy version) defines, for each \meta{math group} (\meta{fam}),
% the `offset' to the bold version of the specified symbol font.
% If there is no bold symbol font defined, the offset will be
% set to zero if there is a bold font assigned to this slot in the
% bold math version, or $-1$ if the font in the bold math version
% is the same as the one in the normal math version. In this case
% a `poor man's bold' system of overprinting is used to achieve
% boldness where this is possible.
%
% The settings are made at the time this package is read, and so
% it is best to load this package late, after any font loading packages
% have been loaded. Symbol fonts loaded after this package will get the
% offset of zero, so they will still be made bold by |\bm| as long as an
% appropriate font is declared for the bold math version.
%
% |\bm@boldtable| and |\bm@heavytable| are set up using very similar
% code, which is temporarily defined to |\bm|, to save wasting a csname.
% Similarly |\bm@pmb|\ldots\ (which will be defined later) are used
% as scratch macros.
%
% The general plan. Run through the fonts allocated to the normal math
% version. Ignore \meta{math alphabet} allocations\footnote{For now?}
% but for each math symbol font, look in the math version specified by
% |#1| (bold or heavy). If the font there is different, then allocate
% a new symbol font in the normal math version to access that bold font
% and place the numerical difference between the allocations of the bold
% and normal font into the table being built (|\bm@boldtable|, if |#1|
% is bold). If the symbol allocation is already greater than |\bmmax|
% do not allocate a new symbol font, but rather set the offset in the
% table to zero. |\bm| will detect this, and use |\boldmath| on its
% argument in this case, so the bold font will be accessed but more
% slowly than using a direct access to a bold font allocated into the
% normal math version. If the font allocated in the bold math version is
% the same as the font in the normal math version, set the offset to
% $-1$, which is a flag value that causes |\bm| to use `poor man's bold'
% overprinting three copies of the symbol, offset slightly to give an
% appearance of boldness.
%
% Fonts containing delimiters and math accents \emph{must} be allocated
% into the normal math version if they are to be used with |\bm|. (In
% these cases |\bm| will produce the normal weight symbol, rather than
% using |\boldmath| or poor man's bold.)
%
%    \begin{macrocode}
\def\bm#1#2{%
%    \end{macrocode}
% This code can not work inside a group, as that would affect any symbol
% font allocations, so instead use some scratch macros to save and
% restore the definitions of commands we need to change locally.
%    \begin{macrocode}
  \let\bm@pmb\install@mathalphabet
  \let\bm@pmb@\getanddefine@fonts
  \let\bm@pmb@@\or
  \edef\bm@general{\f@encoding/\f@family/\f@series/\f@shape/\f@size}%
%    \end{macrocode}
%
% |#2| specifies the maximum number of fonts to allocate
% (either |\bmmax| or |\hmmax|). First check against |\count18|
% that there are that many slots left, and if not reduce accordingly.
% Put the resulting value in |\@tempcnta|.
%    \begin{macrocode}
  \@tempcnta#2%
  \count@-\count18%
  \advance\count@-\@tempcnta
  \advance\count@15\relax
  \ifnum\count@<\z@
    \advance\@tempcnta\count@
  \fi
%    \end{macrocode}
% Make |\or| non-expandable, so we can build an |\ifcase| bit-by-bit
% in a sequence of |\edef|s.
%    \begin{macrocode}
  \let\or\relax
%    \end{macrocode}
%
% Initialise the table (to |\@gobble| to remove the first |\or|).
%    \begin{macrocode}
  \expandafter\let\csname bm@#1table\endcsname\@gobble
%    \end{macrocode}
%
% Helper macro that adds the next entry to the table being built.
%    \begin{macrocode}
  \def\bm@define##1{%
    \expandafter\xdef\csname bm@#1table\endcsname{%
      \csname bm@#1table\endcsname\or##1}}%
%    \end{macrocode}
%
% Each symbol font is recorded in the math version list by a sequence
% such as:\\
% | \getanddefine@fonts \symsymbols \OMS/cmsy/m/n|\\
% Where the first argument is a chardef token carrying the number
% allocated (to symbols, in this example), and the second argument is a
% csname whose \emph{name} denotes the font used.
% So locally redefine |\getanddefine@fonts| to compare |#2| with
% the name in the appropriate slot in the bold math version.
%    \begin{macrocode}
  \def\getanddefine@fonts##1##2{%
    \def\@tempa{##2}%
%    \end{macrocode}
%
%    \begin{macrocode}
    \def\@tempb####1##1####2####3\@nil{\def\@tempb{####2}}%
%    \end{macrocode}
%
%    \begin{macrocode}
    \expandafter\expandafter\expandafter
      \@tempb\csname mv@#1\endcsname\@nil
%    \end{macrocode}
%
% Now |\@tempa| and  |\@tempb| contain the names of the fonts allocated
% to this slot in the two math versions.
%    \begin{macrocode}
    \ifx\@tempa\@tempb
%    \end{macrocode}
% If they are the same, set this offset to $-1$, as a flag to use
% poor man's bold.
%    \begin{macrocode}
      \bm@define\m@ne
    \else
%    \end{macrocode}
% Else make a new name by adjoining |#1| to the name of the symbol font
% eg, |\symboldsymbols| to match |\symsymbols|. If that font has already
% been allocated, or if |\@tempcnta| is positive so we can allocate a
% new slot for this font, then the table will be
% set with the offset between the two fonts. otherwise set the offset to
% zero (so |\boldmath| will be used to access the font).
%    \begin{macrocode}
      \edef\@tempa{sym#1\expandafter\@gobblefour\string##1}%
      \ifnum\@tempcnta<%
           \expandafter\ifx\csname\@tempa\endcsname\relax
             \@ne
           \else
             \m@ne
           \fi
%    \end{macrocode}
%
%    \begin{macrocode}
        \bm@define\z@
      \else
%    \end{macrocode}
%
% If the font is not yet allocated, allocate it now, using
% an internal hack into |\DeclareMathSymbolFont|.
%
% However before allocating it look in the bold math version
% to see if it is the same, and if so use that. For example
% with Mathtime the `operators' font in the `heavy' math version
% is different from that in `normal', but it is the same as
% the font in `bold' (Times bold). So rather than allocate
% |\symheavyoperators| just set it equal to |\symboldoperators|.
%    \begin{macrocode}
        \expandafter\ifx\csname\@tempa\endcsname\relax
          \begingroup
          \escapechar\m@ne
          \edef\@tempb{\endgroup
            \noexpand\split@name
            \expandafter\string\@tempb}%
          \@tempb/\@nil
%    \end{macrocode}
%
%    \begin{macrocode}
         \expandafter\ifx
           \csname symbold\expandafter\@gobblefour\string##1\endcsname
           \relax
%    \end{macrocode}
%  If no font has been allocated for |\bm| yet, then allocate it now.
%    \begin{macrocode}
            \expandafter\new@mathgroup\csname\@tempa\endcsname
            \expandafter\new@symbolfont\csname\@tempa\endcsname
                       \f@encoding\f@family\f@series\f@shape
%    \end{macrocode}
% Reduce by one the number of fonts we can still allocate.
%    \begin{macrocode}
            \advance\@tempcnta\m@ne
%    \end{macrocode}
%
%    \begin{macrocode}
         \else
%    \end{macrocode}
% Else do a similar look into the bold mathgroup.
% Use |\bm@expand| as a scratch macro to save on string space.
%    \begin{macrocode}
           \def\bm@expand####1##1####2####3\@nil{\def\bm@expand{####2}}%
    \expandafter\expandafter\expandafter
           \bm@expand\csname mv@bold\endcsname\@nil
%    \end{macrocode}
%
%    \begin{macrocode}
           \ifx\bm@expand\@tempb
%    \end{macrocode}
% If the font just found (in heavy) is the same as the font in bold
% use the slot (in normal) previously allocated for the bold font.
% (That clear?)
%    \begin{macrocode}
             \expandafter\let\csname\@tempa\expandafter\endcsname
                \csname symbold\expandafter
                        \@gobblefour\string##1\endcsname
%    \end{macrocode}
%
%    \begin{macrocode}
           \else
%    \end{macrocode}
% Otherwise allocate a new slot for it.
%    \begin{macrocode}
             \expandafter\new@mathgroup\csname\@tempa\endcsname
             \expandafter\new@symbolfont\csname\@tempa\endcsname
                       \f@encoding\f@family\f@series\f@shape
             \advance\@tempcnta\m@ne
%    \end{macrocode}
%
%    \begin{macrocode}
           \fi
         \fi
%    \end{macrocode}
%
%    \begin{macrocode}
        \else
%    \end{macrocode}
% If the font has been allocated already, use the existing allocation.
%    \begin{macrocode}
          \PackageInfo{bm}%
            {Symbol font \@tempa\space already defined.\MessageBreak
             Not overwriting it}%
        \fi
%    \end{macrocode}
% Whether the font has just been allocated, or whether it was previously
% allocated, compute the offset and add it to the table.
%    \begin{macrocode}
        \count@\csname\@tempa\endcsname
        \advance\count@-##1%
        \bm@define{\the\count@\relax}%
      \fi
    \fi}%
%    \end{macrocode}
%
% The math version list also contains information about math alphabet
% commands, but we want to ignore those  here, so \ldots
%    \begin{macrocode}
  \let\install@mathalphabet\@gobbletwo
%    \end{macrocode}
%
% Having set up the local definitions, execute the list for the normal
% math version.
%    \begin{macrocode}
  \mv@normal
%    \end{macrocode}
%
% So now the offsets are all entered into the table, separated by
% |\or|.
% Finish off the definition by making this an |\ifcase|. Add a default
% value of zero, so that any symbol fonts declared later will also
% work, as long as a bold version is assigned to the bold math version.
%    \begin{macrocode}
  \expandafter\xdef\csname bm@#1table\endcsname{%
    \noexpand\ifcase\@tempcnta
      \csname bm@#1table\endcsname
    \noexpand\else
      \z@
    \noexpand\fi}%
%    \end{macrocode}
%
% Put things back as they were.
%    \begin{macrocode}
  \expandafter\split@name\bm@general\@nil
  \let\install@mathalphabet\bm@pmb
  \let\getanddefine@fonts\bm@pmb@
  \let\or\bm@pmb@@}
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\bmmax}
% To save declaring too many symbol fonts, do not auto-declare
% any more than |\bmmax| bold symbol fonts into the normal math
% version. Any bold fonts not so allocated will be accessed via
% |\boldmath| which is slower and doesn't work for delimiters
% and accents. It may be set in the preamble with |\newcommand|
% but use |\chardef| here for a slight efficiency gain.
%
% If this is set to a higher value before this
% package is loaded, keep that value.
%    \begin{macrocode}
\ifx\bmmax\@undefined
  \chardef\bmmax=4
\fi
%    \end{macrocode}
%
% If there is no bold math version, It is very easy to set up
% the table, no need to use all the tricky code above.
% Also, at the end of the package redefine the internal macro
% that |\bm| uses to call |\boldmath|, to use poor man's bold
% instead.
%    \begin{macrocode}
\ifx\mv@bold\@undefined
  \def\bm@boldtable{\m@ne}
  \AtEndOfPackage{%
    \def\bm@gr@up#1#2{%
      \bm@pmb{#2}}}
\else
%    \end{macrocode}
% Otherwise use the definition of |\bm| above to set up |\bm@boldtable|
% by comparing the fonts available in the normal and bold math versions.
%    \begin{macrocode}
  \bm{bold}\bmmax
%    \end{macrocode}
%
% \begin{macro}{\mathbf}
% As the bold font has been defined as a symbol font, make |\mathbf|
% access that rather than have it allocate a new math group for the
% same font. (Just in case there were no free slots wrap this
% in an extra test.)
%    \begin{macrocode}
  \@ifundefined{symboldoperators}
    {}
    {\DeclareSymbolFontAlphabet\mathbf{boldoperators}}
%    \end{macrocode}
% \end{macro}
%
%    \begin{macrocode}
\fi
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\hmmax}
%
% Same for heavy (but default to three this time (enough for mathtime
% plus, as no heavy operators font).
%    \begin{macrocode}
\ifx\hmmax\@undefined
  \chardef\hmmax=3
\fi
%    \end{macrocode}
%
% Similarly if there is a heavy math version, set up |\bm@heavytable|.
% (If there is no heavy math version, do nothing here, as |\hm| will be
% set to |\bm| later, once that is defined.)
%    \begin{macrocode}
\ifx\mv@heavy\@undefined
\else
  \bm{heavy}\hmmax
\fi
%    \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\bm@general}
% \changes{v1.0d}{1997/11/15}
%      {extra code to support prime lookahead and breqn}
% \changes{v1.0h}{2002/11/22}{Pass math version as third (new) argument}
%
% |\bm| is pretty much |\bmdefine\bm@command| followed by executing
% |\bm@command|. It would in principle be possible to execute the
% emboldened tokens directly, rather than building up a macro first,
% but (as I learned the hard way) it's difficult to do this in the midst
% of all these nested |\if| constructs.
% First extract the central bit of code for |\hm| |\bm| |\hmdefine| and
% |\bmdefine|. Note that in the case of the inline versions they take
% an argument and brace it, rather than relying on |\bm@general| to pick
% up the argument. This makes the code robust with respect to premature
% expansion.
%    \begin{macrocode}
\begingroup
\catcode`\'=\active
\@firstofone{\endgroup
\def\bm@general#1#2#3#4#5{%
  \begingroup
%    \end{macrocode}
% \changes{v1.0d}{1997/11/15}
%      {make nested \cs{bm} vanish completely, not leave a brace group}
% First locally disable |\bm| and |\hm|, as they would mess things
% up terribly, and the original Spivak versions used the syntax
% |\bmdefine\balpha{\bm\alpha}|.
%    \begin{macrocode}
    \let\bm\@firstofone
    \let\hm\@firstofone
%    \end{macrocode}
% Now initialise the commands used to save the tokens constructed.
%    \begin{macrocode}
    \global\let\bm@command\@empty
    \let\@let@token\@empty
%    \end{macrocode}
% As we want to expand the macros to look at their definition
% turn off protection. Otherwise the |\protect| will be carried over
% and apply to the wrong token, eg |{|.
%    \begin{macrocode}
    \let\protect\@empty
    \let\@typeset@protect\@empty
%    \end{macrocode}
% Set up either bold or heavy
%    \begin{macrocode}
    \def\bm@mathchoice{\bm@m@thchoice#1}%
    \def\bm@group{\bm@gr@up#1}%
    \let\bm@table#2%
%    \end{macrocode}
% Make sure |\left| and |\right| are really non expandable,
% and not |\ifx| equal to anything else.
% \changes{v1.0d}{1997/11/15}
%      {make sure \cs{left} is primitive}
%    \begin{macrocode}
    \let\left\holdinginserts
%    \end{macrocode}
% These three save on the number of |\ifx| tests below.
% \changes{v1.1a}{2003/09/01}{Forgotten to check for \cs{hskip} (pr/3572)}
%    \begin{macrocode}
    \let\right\left
    \let\mskip\mkern
    \let\hskip\kern
%    \end{macrocode}
% Definition of |'| locally modified so as not to use |\futurelet|
% in the look ahead, but to make the |\prime| available at the top level
% to be made bold, or heavy or whatever.
% |'| is locally active for this definition.
% \changes{v1.0d}{1997/11/15}
%      {prime code added}
%    \begin{macrocode}
    \let\bm@prime\copy
    \def'{\bm@prime\prime\relax}%
%    \end{macrocode}
% For optional argument commands. This expandable version of
% |\@ifnextchar| is not 100\% safe, but works for |\sqrt| unless
% you put something really strange in the arguments.
% \changes{v0.11}{1997/01/06}
%      {\cs{@ifnextchar} made safe.}
%    \begin{macrocode}
    \def\@ifnextchar##1##2##3##4{%
      \if##1##4%
        \expandafter\@firstoftwo
      \else
        \expandafter\@secondoftwo
      \fi
      {##2##4}{##3{##4}}}%
%    \end{macrocode}
% For Vladimir Volovich\ldots
% \changes{v1.0f}{1998/12/10}
%      {Errors and warnings made safe. tools/2917}
%    \begin{macrocode}
    \def\GenericWarning##1##2{%
      \unvcopy{\GenericWarning{##1}{##2}}}%
    \def\GenericError##1##2##3##4{%
      \unvcopy{\GenericError{##1}{##2}{##3}{##4}}}%
%    \end{macrocode}
% For AMS definitions.
%    \begin{macrocode}
    \let\DN@\copy
    \let\FN@\copy
    \let\next@\copy
    \global\let\bm@first\@empty
%    \end{macrocode}
%   For AMS version of |\sqrt|: don't expand just wrap in brace group
%   so that it can be made bold in a safe but slow way. Do the same for
%   internal accent command
% \changes{v1.1b}{2003/10/05}{AMS \cs{sqrt} not working}
%
% \changes{v1.1c}{2004/02/26}{\cs{accentV} made safe (pr/3625)}
% Code for AMS accent allows bm to be used (just) with accent
% but stops the nested accents stacking correctly, this can be
% corrected by using an extra brace group as usual.
% |\bm{{\hat{\hat{F}}}}|
%    \begin{macrocode}
    \ifx\uproot@\undefined\else
       \def\root##1\of##2{{\root##1\of{##2}}}%
    \fi
    \def\mathaccentV##1{\mathaccent"\accentclass@}%
%    \end{macrocode}
% For breqn definitions.
%    \begin{macrocode}
    \let\@ifnext\@ifnextchar
    \let\measure@lhs\copy
    \let \rel@break\copy
    \let \bin@break\copy
    \let \after@open\copy
    \let \after@close\copy
%    \end{macrocode}
% Make sure things like |\pounds| take the `math branch' even
% in |\bmdefine| (which is not executed in math mode).
%    \begin{macrocode}
    \let\ifmmode\iftrue
%    \end{macrocode}
%    We have to ensure that the math alphabets have definitions that
%    correspond the the ``bold'' math version we are going to switch
%    to. As these definitions are globally assigned when a math
%    version is changed it is likely that right now we have those of
%    the normal math version active. Argument |#3| holds either
%    |\mv@bold| or |\mv@heavy| and we execute that after redefining
%    |\install@mathalphabet| and |\getanddefine@fonts| suitably.
%    The definitions are reverted back to their original the moment
%    the scanning is done
% \changes{v1.0h}{2002/11/22}{Get math alphabets right (pr/3476)}
%    \begin{macrocode}
     \let\install@mathalphabet\def
     \let\getanddefine@fonts\@gobbletwo
     #3%
%    \end{macrocode}
% The last redefinition just makes |\mathit| type commands re-insert
% themselves (more or less) as if they are allowed to expand
% they die horribly if the expansions are put into |\mathchoice|
% and so executed more than once.
%    \begin{macrocode}
    \def\select@group##1##2##3##4{{%
      \protect##1{##4}}}%
    \def\use@mathgroup##1##2##3{{%
      \protect\use@mathgroup##1{##2}{##3}}}%
%    \end{macrocode}
%
% So now start looking at the argument.
% \changes{v0.10}{1997/01/04}
%      {Use \cs{bm@end}}
%    \begin{macrocode}
    \bm@expand#5\bm@end
  \endgroup
%    \end{macrocode}
%
% Finally outside the group either execute |\bm@command| (for |\bm|)
% or save its definition (for |\bmdefine|).
%    \begin{macrocode}
  #4}
%    \end{macrocode}
% End of the |\@firstofone| above, and the scope of the active |'|.
%    \begin{macrocode}
}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm}
% \changes{v1.0d}{1997/11/15}
%      {Make \cs{bm} grab its argument even when protected}
% Set up the bold (rather than heavy) version, and run |\bm@command|
% right at the end, to execute the emboldened argument.
% The argument is grabbed by the top level function, and explicitly
% braced, so that |\bm| works even if the braces are omitted round
% its argument in a `moving argument'.
% \changes{v1.0h}{2002/11/22}{Pass math version as third (new) argument}
%    \begin{macrocode}
\DeclareRobustCommand\bm{%
  \bm@general\boldmath\bm@boldtable\mv@bold\bm@command}
\protected@edef\bm#1{\bm{#1}}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\DeclareBoldMathCommand}
% \changes{v0.07}{1996/12/14}
%      {Macro added}
% \begin{macro}{\bm@declare}
% \changes{v0.07}{1996/12/14}
%      {Macro added}
% |DeclareBoldMathCommand|%^^A
%      \oarg{mathversion}\marg{command}\marg{math expression}\\
% looks like |\bm| except at the end the specified command is
% globally defined to be |\bm@command|.
% The \meta{mathversion} defaults to `bold'.
%    \begin{macrocode}
\def\DeclareBoldMathCommand{\@testopt\bm@declare{bold}}
%    \end{macrocode}
%
% \changes{v1.0h}{2002/11/22}{Pass math version as third (new) argument}
%    \begin{macrocode}
\def\bm@declare[#1]#2{%
  \expandafter\bm@general
        \csname #1math\expandafter\endcsname
        \csname bm@#1table\expandafter\endcsname
        \csname mv@#1\endcsname
        {\bm@define#2}}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\bmdefine}
% |\bmdefine| Shorthand for |\DeclareBoldMathCommand[bold]|.
%
% |\bm| is empty within the definition, so that either\\
% |\bmdefine\balpha{\bm\alpha}| or  |\bmdefine\balpha{\alpha}| \\
% may be used. (The former just for compatibility with the original
% version for plain \TeX).
%    \begin{macrocode}
\def\bmdefine{\DeclareBoldMathCommand[bold]}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\hm}
% \changes{v1.0d}{1997/11/15}
%      {Make \cs{hm} grab its argument even when protected}
% \begin{macro}{\hmdefine}
% Same again for |\hm|.
%    \begin{macrocode}
\ifx\mv@heavy\@undefined
%    \end{macrocode}
%
% If there is no heavy math version defined, let |\hm| be defined
% as |\bm|. Currently there is no warning given, perhaps there should
% be, or even an error?
%    \begin{macrocode}
  \let\hm\bm
  \let\heavymath\boldmath
  \let\bm@heavytable\bm@boldtable
%    \end{macrocode}
%
%    \begin{macrocode}
\else
%    \end{macrocode}
%
% Otherwise define |\hm| and |\hmdefine| in direct analogy with the
% above.
% \changes{v1.0h}{2002/11/22}{Pass math version as third (new) argument}
%    \begin{macrocode}
  \DeclareRobustCommand\hm{%
    \bm@general\heavymath\bm@heavytable\mv@heavy\bm@command}
  \protected@edef\hm#1{\hm{#1}}
%    \end{macrocode}
%
%    \begin{macrocode}
  \def\hmdefine{\DeclareBoldMathCommand[heavy]}
%    \end{macrocode}
%
%    \begin{macrocode}
\fi
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\bm@end}
% \changes{v0.10}{1997/01/04}
%      {Macro added}
% Normally speaking |\outer| declarations should be avoided at all
% costs. (\LaTeX\ redefines all of plain \TeX's allocation macros
% to be non-outer.) However this is one place where it seems like a
% good idea. If a command taking an argument is put in |\bm| without
% its argument, then the |\@@end| terminating token would be taken as
%  the argument, and so the rest of the paragraph would be gobbled up
% and the \LaTeX\ would die horribly. So make the internal terminating
% token |\outer|. (The actual test for termination is made against
% |\@@end| not |\bm@end| as this macro will be expanded by the look-ahead
% system.)
%    \begin{macrocode}
\outer\def\bm@end{\@@end}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@expand}
% \changes{v0.11}{1997/01/06}
%      {Double up.}
% \begin{macro}{\bm@exp@nd}
% \changes{v0.11}{1997/01/06}
%      {Macro added}
% |\afterassignment| trick to fully expand the following tokens until
% the first non-expandable token is revealed. This may discard a space
% token (which is what \TeX\ is looking for) but that doesn't matter in
% math mode. The expansion lookahead is done twice in case any stray
% space tokens have crept in.\footnote{The need for this was noticed
% while testing \cs{sqrt}. The definition of \cs{root} inherited from
% plain \TeX\ has an anomalous space token, that is normally harmless
% (just wastes memory), but which killed earlier versions of this
%  package.}
%    \begin{macrocode}
\def\bm@expand{\afterassignment\bm@exp@nd\count@`\a}
%    \end{macrocode}
%
%    \begin{macrocode}
\def\bm@exp@nd{\afterassignment\bm@test\count@`\a}
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\bm@test}
% Normally we will grab the non-expandable token as a macro argument
% but better check it is not |{| first. Save the previous token
% so we can check later if it was |\left|, in which case use the delcode
% rather than the mathcode if the current token is a character.
%    \begin{macrocode}
\def\bm@test{%
  \let\bm@previous\@let@token
  \futurelet\@let@token\bm@test@}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@test@}
% If looking at a single token, switch to |\bm@test@token|, else if
% looking at a |{ }| group, grab the whole group with |\bm@group|.
% A |\bgroup| token will take the wrong branch here (currently not
% trapped).
%    \begin{macrocode}
\def\bm@test@{%
  \ifx\@let@token\bgroup
    \expandafter\bm@group
  \else
    \expandafter\bm@test@token
  \fi}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@gr@up}
% \changes{v1.0c}{1997/10/09}
%      {Extra brace around argument for \cs{over}}
% If faced with a group,
% If we are in math mode, stick it in a |\boldsymbol| like construct
% and then recurse on |\bm@expand|.
% Otherwise just use |\bfseries\boldmath|.
% The actual test is deferred till `run time'.
% Here and elsewhere could deal with the inner list with an inner call
% to |\bm|, but that doesn't seem to gain very much, and complicates the
% code quite a bit.
%
% |#1| is either |\boldmath| or |\heavymath|.
% Need to add an extra set of explicit braces around |#2| as otherwise
% the math style commands applied in |\mathchoice| might only apply
% to the first half of an |\over| construction.
%    \begin{macrocode}
\def\bm@gr@up#1#2{%
  \bm@add{{\bm@gr@@p#1{{#2}}}}}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@gr@@p}
% |#1| is either |\boldmath| or |\heavymath|.
%    \begin{macrocode}
\def\bm@gr@@p#1#2{%
  \ifmmode
    \bm@mchoice#1{#2}{#2}{#2}{#2}%
  \else
    \bfseries#1#2%
  \fi}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@test@token}
% If not facing a |{  }| group then test to see what we have.
% Basic idea: Trap |\mathchardef| tokens, character tokens, and calls to
% |\mathchar|, |\mathaccent|, etc, and change the \emph{math-group}
% (fam) to point at the bold version.
% Other things just copy straight over to the command being built.
% (Anything inside a |\mathop| or similar will end up being made bold
% as the |\mathop| will be copied over, but its argument will be made
% bold by the group code above.
%    \begin{macrocode}
\def\bm@test@token#1{%
  \let\bm@next\@empty
%    \end{macrocode}
%
% Stop here. Note that it is vital that the terminating token
% is non-expandable and defined, rather than the usual \LaTeX\
% terminators |\@nil| or |\@@|. (Worse still would be a `quark'
% like |docstrip|'s |\qStop|.)
%    \begin{macrocode}
  \ifx#1\@@end
%    \end{macrocode}
%
% |\bm@mathchoice| uses macro arguments, so need to make the tail
% recursion explicit here. All the other cases recurse by way of
% |\afterassignment| which means all the trailing |\fi| are eaten
% while making the assignment.
%    \begin{macrocode}
  \else\ifx#1\mathchoice
    \let\bm@next\bm@mathchoice
%    \end{macrocode}
%
% The main point: Find these expressions, and change the mathgroup.
%    \begin{macrocode}
  \else\ifx#1\mathchar
    \afterassignment\bm@mathchar\count@
  \else\ifx#1\mathaccent
    \afterassignment\bm@mathaccent\count@
  \else\ifx#1\delimiter
    \afterassignment\bm@delimiter\count@
  \else\ifx#1\radical
    \afterassignment\bm@radical\count@
%    \end{macrocode}
%
% Need to trap spaces otherwise digits will get turned to bold
% mathchars.
% \changes{v1.1a}{2003/09/01}{Forgotten to check for \cs{hskip} (pr/3572)}
%    \begin{macrocode}
  \else\ifx#1\mkern
    \bm@register#1{\muskip\z@}%
  \else\ifx#1\kern
    \bm@register#1\skip@
  \else\ifx#1\penalty
    \bm@register#1\count@
%    \end{macrocode}
%
% \changes{v1.0f}{1998/12/10}
%      {Add possibility to copy brace group unmodified. tools/2917}
% |\vcopy| is a flag to copy the next group unchanged to the
% result command.
%    \begin{macrocode}
  \else\ifx#1\unvcopy
    \let\bm@next\bm@add
%    \end{macrocode}
%
%    \begin{macrocode}
  \else\ifcat\noexpand#1\relax
%    \end{macrocode}
% Other command, look if it's a mathchardef token (otherwise just add
% it).
%    \begin{macrocode}
    \xdef\meaning@{\meaning#1}%
    \expandafter\bm@mchar@test\meaning@""\@nil#1%
%    \end{macrocode}
% \changes{v1.0d}{1997/11/15}
%      {New active \cs{mathcode} code.}
% Character token.
% If it is of catcode 11 or 12, get its mathcode.
% If that is |"8000| replace the token by its active version, and then
% let bm expansion look again at the character. Being really active
% this time, it will expand away (probably).
%
% If the previous token was |\left| or |\right|, get the delcode
% instead of the mathcode.
%    \begin{macrocode}
  \else\ifcat.\ifcat a#1.\else#1\fi
    \count@\mathcode`#1\relax
    \ifnum\count@=\mathcode`\'%
      \begingroup\uccode`\~`#1\uppercase{\endgroup
        \def\bm@next{\bm@expand~}}%
    \else
      \ifx\bm@previous\left
        \count@\delcode`#1\relax
        \bm@delimiter
      \else
        \bm@mathchar
      \fi
    \fi
  \else
%    \end{macrocode}
% And final possibility:
% a character token of catcode other than 11 or 12.
%    \begin{macrocode}
    \bm@add{#1}%
  \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi
  \bm@next}
%    \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\bm@define}
% \changes{v0.08}{1996/12/18}
%      {Defined commands now always robust}
% End code for |\bmdefine|. Define the given command name to the
% robust form of the accumulated code.
%
% If |\bm@command| is equal to |\@gtempa| then it is a macro whose
% expansion is a single call to |\mathchar|, so that can be optimised
% with a |\mathchardef|.
%    \begin{macrocode}
\def\bm@define#1{%
  \begingroup
    \ifx\bm@command\@gtempa
      \def\mathchar{\global\mathchardef#1}%
      \bm@command
    \else
%    \end{macrocode}
% Rather than simply |\let#1\bm@command|, make the defined command
% robust. |\bm@first| is normally empty, but might be something like
% |\DOTSI| which needs to be lifted to the top level, in front
% of any |\protect| because of the lookahead mechanism used
% in the \textsf{amsmath} package.
%    \begin{macrocode}
      \toks@\expandafter{\bm@command}%
      \xdef#1{\bm@first\noexpand\bm@protect\noexpand#1{\the\toks@}}%
    \fi
  \endgroup}
%    \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\bm@protect}
% \changes{v0.08}{1996/12/18}
%      {macro added}
% Commands defined by |\bmdefine| re-insert themselves
% if protection is enabled.
%    \begin{macrocode}
\def\bm@protect#1{%
  \ifx\protect\@typeset@protect
    \expandafter\@firstofone
  \else
    \protect#1\expandafter\@gobble
  \fi}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@mchoice}
% |\boldsymbol|, more or less.
% |#1| is either |\boldmath| or |\heavymath|.
%    \begin{macrocode}
\def\bm@mchoice#1#2#3#4#5{%
  \mathchoice{\hbox{#1$\displaystyle\m@th#2$}}%
             {\hbox{#1$\textstyle\m@th#3$}}%
             {\hbox{#1$\scriptstyle\m@th#4$}}%
             {\hbox{#1$\scriptscriptstyle\m@th#5$}}}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@m@thchoice}
% Action if you find a |\mathchoice|. Add the bold version to
% |\bm@command| then recurse.
%
% |#1| is either |\boldmath| or |\heavymath|.
%    \begin{macrocode}
\def\bm@m@thchoice#1#2#3#4#5{%
  \bm@add{\bm@mchoice#1{#2}{#3}{#4}{#5}}}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@register}
% \changes{v0.07}{1996/12/14}
%      {Macro added}
% Combined code for setting up |\bm@r@gister| with the correct
% register type.
%    \begin{macrocode}
\def\bm@register#1#2{%
  \def\@tempa{#1\the#2}%
  \afterassignment\bm@r@gister#2}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@r@gister}
% \changes{v0.06}{1996/12/12}
%      {Support \cs{mskip}}
% \changes{v0.07}{1996/12/14}
%      {Combine all register macros}
% |\mkern| itself would transfer to |\bm@command| without any special
% test, but any explicit dimension following would be converted to
% |\mathchar|. So trap this and grab the muskip as a muskip.
% This is used in |\iiint|. |\penalty| was needed for the
% AMS  version of |\colon|, and so do most of the others as well.
%    \begin{macrocode}
\def\bm@r@gister{%
  \bm@xadd{\@tempa\space}}
%    \end{macrocode}
% \end{macro}
%
%
% \begin{macro}{\bm@mathchar}
% \changes{v0.10}{1997/01/04}
%      {Modify to use \cs{boldmath} or \cs{pmb} if needed.}
% Change the family (math group) of a mathcode and then
% use the modified code with |\mathchar|. If there is no
% suitable bold font in the current math version, use the original
% unmodified mathcode, but switch to |\boldmath| (if there is a bold
% font there) or use `poor man's bold'. Note that these other
% possibilities are only possible here, not for the otherwise similar
% code for |\delimiter| or |\mathaccent|, as those commands
% must work with fonts from the same math version.
%
% Finally recurse down the list.
%    \begin{macrocode}
\def\bm@mathchar{%
%    \end{macrocode}
%
%    \begin{macrocode}
  \@tempcntb\count@
  \let\@tempa\bm@group
%    \end{macrocode}
%
% |\bm@changefam| will isolate the math group from the mathcode
% and look up the offset in the current table.
%    \begin{macrocode}
  \bm@changefam{}%
%    \end{macrocode}
%
% If the mathcode has changed, then just add the new |\mathchar|
% (saving |\@gtempa| allows |\bmdefine| to optimise this to a
%  mathchardef if it turns out to be the only symbol in the argument).
%    \begin{macrocode}
  \ifnum\count@>\@tempcntb
%    \end{macrocode}
%
%    \begin{macrocode}
    \ifx\bm@command\@empty
      \xdef\@gtempa{\mathchar\the\count@\space}%
    \fi
    \bm@xadd{\mathchar\the\count@\space}%
  \else
%    \end{macrocode}
%
% Otherwise grab the math class from the math code
% and add that (locally zapping |\bm@expand| as we don't want
% to recurse at this point).
%    \begin{macrocode}
    \begingroup
      \divide\count@"1000
      \let\bm@expand\relax
      \bm@xadd\bm@class
    \endgroup
%    \end{macrocode}
% |\@tempa| will be |\bm@group| (which applies |\boldmath| and
% |\mathchoice|) unless it was changed by |\bm@changefam| to |\bm@pmb|
% (which applies a `poor man's bold' construction in a |\mathchoice|).
%    \begin{macrocode}
    \edef\@tempb{%
      \noexpand\@tempa{\mathchar\the\count@\space}}%
    \@tempb
  \fi}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@pmb}
% \changes{v0.10}{1997/01/04}
%      {Macro added}
% Add a poor man's bold construction to the list being built.
%    \begin{macrocode}
\def\bm@pmb#1{%
  \bm@add{\bm@pmb@{#1}}}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@pmb@}
% \changes{v0.10}{1997/01/04}
%      {Macro added}
% |\pmb| variant. (See \TeX{}Book, or AMS \textsf{amsbsy} package).
% This one takes a bit more care to use smaller offsets in subscripts.
%    \begin{macrocode}
\def\bm@pmb@#1{{%
  \setbox\tw@\hbox{$\m@th\mkern.4mu$}%
  \mathchoice
    \bm@pmb@@\displaystyle\@empty{#1}%
    \bm@pmb@@\textstyle\@empty{#1}%
    \bm@pmb@@\scriptstyle\defaultscriptratio{#1}%
    \bm@pmb@@\scriptscriptstyle\defaultscriptscriptratio{#1}}}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@pmb@@}
% \changes{v0.10}{1997/01/04}
%      {Macro added}
% Helper macro. Box |#3| and set it three times in the style |#1|,
% offset by an amount reduced by the ratio specified in |#2|.
%    \begin{macrocode}
\def\bm@pmb@@#1#2#3{{%
  \setbox\z@\hbox{$\m@th#1#3$}%
  \dimen@#2\wd\tw@
  \rlap{\copy\z@}%
  \kern\dimen@
  \raise1.5\dimen@\rlap{\copy\z@}%
  \kern\dimen@
  \box\z@}}%
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@class}
% \changes{v0.10}{1997/01/04}
%      {Macro added}
% Convert a numeric math class back to a math class command.
% |\mathord| is omitted in class $0$ and $7$ to save space and so
% things work out right in constructions such as |x^a| where
% |x^\mathord{a}| would not work.
%    \begin{macrocode}
\def\bm@class{%
  \ifcase\count@
    \or
    \mathop\or
    \mathbin\or
    \mathrel\or
    \mathopen\or
    \mathclose\or
    \mathpunct\or
  \fi}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@add}
% \changes{v0.07}{1996/12/14}
%      {macro added}
% A version of |\g@addto@macro| that internally uses a |\begingroup|
% rather than a brace group\footnote{This bug is fixed in the \LaTeX\
% kernel of 1996/12/01}, to save creating a mathord.
%
% As need to redefine it anyway, save some tokens by making
% it specific to |\bm@command|, and to execute |\bm@expand|
% to continue the loop.
%    \begin{macrocode}
\def\bm@add#1{%
  \begingroup
    \toks@\expandafter{\bm@command#1}%
    \xdef\bm@command{\the\toks@}%
  \endgroup
  \bm@expand}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@xadd}
% An |\xdef| version of |\bm@add|.
%    \begin{macrocode}
\def\bm@xadd#1{%
  \begingroup
    \toks@\expandafter{\bm@command}%
    \xdef\bm@command{\the\toks@#1}%
  \endgroup
  \bm@expand}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@mathaccent}
% \changes{v1.0e}{1997/11/21}
%      {Use bm on argument group, as spotted on c.t.t.}
% |\mathaccent| version of |\bm@mathchar|.
%    \begin{macrocode}
\def\bm@mathaccent{%
 \bm@changefam{}%
%    \end{macrocode}
% The next four lines were added a v1.0e. Without them |\bm{\hat{A}}|
% makes the accent bold using |\bm| but the group |{A}| is made bold
% via a |\mathchoice| construction as for any other group, as |\bm|
% does not attempt to parse inside brace groups. While that produces
% something acceptable for lower case letters, it produces
% $\bm{\hat{{A}}}$ which is not too good. The braces may simply be
% omitted:
% |\bm{\hat A}| would work, producing $\bm{\hat A}$, however I did not
% want to document such a restriction, so now modify bm so that such
% brace groups are handled gracefully.
%
% It would be possible to locally make mathaccents take an argument
% during the bm look-ahead, so the brace groups would then vanish
% during expansion, however I would then need to explicitly skip past
% \meta{filler} and also make sure that the end of parse token
% was not gobbled in marginal cases like |$\bm\hat$|.
%
% So instead do the following which gets rid of \meta{filler}
% with a redefinition of |\relax|, and just locally changes
% |\bm@group| so that instead of doing a |\mathchoice| it simply adds
% |\bgroup| and |\egroup| around the tokens, and lets bm modify the
% tokens of the `argument'. This means that |\bm{\hat{A}}| now produces
%\begin{verbatim}
% \mathaccent 29790 \bgroup \mathchar 30017 \egroup
%\end{verbatim}
% The inner math list is a single mathchar, and so \TeX\ will not box
% it, and the math accent will correctly position, taking into account
% the skewchar information.
%
% As the normal bm lookahead is used, it is automatic that the parse
% will end without trying to go past |\bm@end|.
%
% One disadvantage is that the group will mean that |\bm@previous|
% will not be correctly updated. However that is only used for
% delimiter checking, so can not matter here.
%    \begin{macrocode}
 \begingroup
 \def\bm@group##1{\endgroup\bm@xadd{\bgroup}##1\egroup}%
 \def\bm@test@token{\endgroup\bm@test@token}%
 \let\relax\@empty
%    \end{macrocode}
%
%    \begin{macrocode}
 \bm@xadd{\mathaccent\the\count@\space}}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@delimiter}
% \changes{v1.0d}{1997/11/15}
%      {Support \cs{left}. null delimiter.}
% Change both families (math groups) of a delcode and then
% use the modified code with  |\delimiter|. Don't change code `0'
% as that denotes a null delimiter.
%    \begin{macrocode}
\def\bm@delimiter{%
  \ifnum\count@>\z@
    \bm@changefam{}%
    \bm@changefam{000}%
  \fi
  \bm@xadd{\delimiter\the\count@\space}}%
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@radical}
% Same for |\radical|.
%    \begin{macrocode}
\def\bm@radical{%
 \bm@changefam{}%
 \bm@changefam{000}%
 \bm@xadd{\radical\the\count@\space}}%
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@mchar@}
% Catcode 12 |\mathchar|, for |\ifx| tests.
%    \begin{macrocode}
\edef\bm@mchar@{\meaning\mathchar}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@mchar@test}
% Test if the |\meaning| starts with |\mathchar|. If it does, grab the
% value into |\count@| and call |\bm@mathchar|, else just copy the
% command into the accumulated tokens.
% |#1|, |#2|, |#3| are all |\meaning| produced tokens, or `dummy tokens'
% added at the time this is called. |#4| is the original token, in case
%  decide not to use the |\meaning|.
%    \begin{macrocode}
\def\bm@mchar@test#1"#2"#3\@nil#4{%
  \xdef\meaning@{#1}%
  \ifx\meaning@\bm@mchar@
    \count@"#2\relax
    \bm@mathchar
  \else
%    \end{macrocode}
% Some other command: copy it straight over. If it is the first thing
% added, and it is a |\relax| token, save it in |\bm@first| for use
% in |\bm@define|.
%    \begin{macrocode}
    \ifx\bm@previous\@empty
      \ifx\relax#4%
        \gdef\bm@first{#4}%
      \fi
    \fi
    \bm@add{#4}%
  \fi}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@changefam}
% \changes{v0.10}{1997/01/04}
%      {Rewrite for new \cs{bm@table} system}
% Pull out one specified hex digit and passes
% it to |\bm@modify| to change. argument is empty normally but |000| to
% access the second math group in a delimiter code.
%    \begin{macrocode}
\def\bm@changefam#1{%
  \@tempcnta\count@
  \divide\@tempcnta"1000#1 %
  \multiply\@tempcnta"1000#1 %
  \advance\@tempcnta-\count@
  \divide\@tempcnta-"100#1 %
%    \end{macrocode}
% Having isolated the required math group (fam), look up
% the offset in the current table.
%    \begin{macrocode}
  \@tempcnta\bm@table
%    \end{macrocode}
% If the offset is $-1$, keep |\count@| unchanged, but set |\@tempa|
% to use poor man's bold. Otherwise increment |\count@| to change the
% math group specified.
%    \begin{macrocode}
  \ifnum\@tempcnta=\m@ne
    \let\@tempa\bm@pmb
  \else
    \multiply\@tempcnta"100#1 %
    \advance\count@\@tempcnta
  \fi}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\bm@prime}
% \changes{v1.0d}{1997/11/15}
%      {Macro added}
% Support |'|. Earlier versions did not make the prime bold in a'.
%
% |\bm{a''}| will now produce (with the normal encodings)
%\begin{verbatim}
%      \mathchar 30049
%      \bm@prime \mathchar 1584 \relax
%      \bm@prime \mathchar 1584 \relax
%\end{verbatim}
% So |\bm@prime| does essentially the same as the active definition of
% |'|, which is to start a superscript group then keep adding |\prime|
% for each |'| (or |\bm@prime|) following. Here modified to grab a
% |\relax| delimited argument and use that instead of |\prime|.
% |\bm@prime| is locally |\let| to |'| so the |\ifx| tests in |\pr@m@s|
% don't need changing.
%    \begin{macrocode}
\def\bm@prime{^\bgroup
  \let\bm@prime'%
  \def\prim@s##1\relax{##1\futurelet\@let@token\pr@m@s}%
  \prim@s}
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\boldsymbol}
% \changes{v1.0b}{1997/04/14}
%      {Macro added}
% \begin{macro}{\heavysymbol}
% Finally, to ease conversion of documents between this package and
% the \textsf{amsbsy} package:
%    \begin{macrocode}
\let\boldsymbol\bm
%    \end{macrocode}
%
%    \begin{macrocode}
\let\heavysymbol\hm
%    \end{macrocode}
% \end{macro}
% \end{macro}
%    \begin{macrocode}
%</package>
%    \end{macrocode}
%
% \Finale
%