summaryrefslogtreecommitdiff
path: root/macros/latex-dev/required/l3kernel/l3tl-analysis.dtx
blob: 4df9ee68a3d3fa35d20f7cf025daeea794c6575c (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
% \iffalse meta-comment
%
%% File: l3tl-analysis.dtx
%
% Copyright (C) 2011-2024 The LaTeX 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
%
%    https://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 development version of the bundle can be found at
%
%    https://github.com/latex3/latex3
%
% for those people who are interested.
%
%<*driver>
\documentclass[full,kernel]{l3doc}
\begin{document}
  \DocInput{\jobname.dtx}
\end{document}
%</driver>
% \fi
%
%
% \title{^^A
%   The \pkg{l3tl-analysis} module\\ Analysing token lists^^A
% }
%
% \author{^^A
%  The \LaTeX{} Project\thanks
%    {^^A
%      E-mail:
%        \href{mailto:latex-team@latex-project.org}
%          {latex-team@latex-project.org}^^A
%    }^^A
% }
%
% \date{Released 2024-05-27}
%
% \maketitle
%
% \begin{documentation}
%
% This module provides functions that are particularly useful in the
% \pkg{l3regex} module for mapping through a token list one \meta{token}
% at a time (including begin-group/end-group tokens).  For
% \cs{tl_analysis_map_inline:Nn} or \cs{tl_analysis_map_inline:nn}, the
% token list is given as an argument; the analogous function
% \cs{peek_analysis_map_inline:n} documented in \pkg{l3token} finds
% tokens in the input stream instead.  In both cases the user provides
% \meta{inline code} that receives three arguments for each
% \meta{token}:
% \begin{itemize}
%   \item \meta{tokens}, which both \texttt{o}-expand and
%     \texttt{e}/\texttt{x}-expand to the \meta{token}. The detailed form of
%     \meta{tokens} may change in later releases.
%   \item \meta{char code}, a decimal representation of the character
%     code of the \meta{token}, $-1$ if it is a control sequence.
%   \item \meta{catcode}, a capital hexadecimal digit which denotes the
%     category code of the \meta{token} (0:~control sequence,
%     1:~begin-group, 2:~end-group, 3:~math shift, 4:~alignment tab,
%     6:~parameter, 7:~superscript, 8:~subscript, A:~space, B:~letter,
%     C:~other, D:~active).  This can be converted to an integer by
%     writing |"|\meta{catcode}.
% \end{itemize}
% In addition, there is a debugging function \cs{tl_analysis_show:n},
% very similar to the \cs[no-index]{ShowTokens} macro from the \pkg{ted} package.
%
% \begin{function}[added = 2021-05-11]
%   {
%     \tl_analysis_show:N, \tl_analysis_show:n,
%     \tl_analysis_log:N, \tl_analysis_log:n
%   }
%   \begin{syntax}
%     \cs{tl_analysis_show:n} \Arg{token list}
%     \cs{tl_analysis_log:n} \Arg{token list}
%   \end{syntax}
%   Displays to the terminal (or log) the detailed decomposition of the
%   \meta{token list} into tokens, showing the category code of each
%   character token, the meaning of control sequences and active
%   characters, and the value of registers.
% \end{function}
%
% \begin{function}[added = 2018-04-09, updated = 2022-03-26]
%   {\tl_analysis_map_inline:nn, \tl_analysis_map_inline:Nn}
%   \begin{syntax}
%     \cs{tl_analysis_map_inline:nn} \Arg{token list} \Arg{inline function}
%   \end{syntax}
%   Applies the \meta{inline function} to each individual \meta{token}
%   in the \meta{token list}. The \meta{inline function} receives three
%   arguments as explained above.  As all other mappings the mapping is
%   done at the current group level, \emph{i.e.}~any local assignments
%   made by the \meta{inline function} remain in effect after the loop.
% \end{function}
%
% \end{documentation}
%
% \begin{implementation}
%
% \section{\pkg{l3tl-analysis} implementation}
%
%    \begin{macrocode}
%<@@=tl>
%    \end{macrocode}
%
% \subsection{Internal functions}
%
% \begin{variable}{\s_@@}
%   The format used to store token lists internally uses the scan mark
%   \cs{s_@@} as a delimiter.
% \end{variable}
%
% \subsection{Internal format}
%
% The task of the \pkg{l3tl-analysis} module is to convert token lists
% to an internal format which allows us to extract all the relevant
% information about individual tokens (category code, character code),
% as well as reconstruct the token list quickly. This internal format is
% used in \pkg{l3regex} where we need to support arbitrary tokens, and
% it is used in conversion functions in \pkg{l3str-convert}, where we wish to
% support clusters of characters instead of single tokens.
%
% We thus need a way to encode any \meta{token} (even begin-group and
% end-group character tokens) in a way amenable to manipulating tokens
% individually. The best we can do is to find \meta{tokens} which both
% \texttt{o}-expand and \texttt{e}/\texttt{x}-expand to the given
% \meta{token}. Collecting more information about the category code and
% character code is also useful for regular expressions, since most
% regexes are catcode-agnostic. The internal format thus takes the form
% of a succession of items of the form
% \begin{quote}
%   \meta{tokens} \cs{s_@@} \meta{catcode} \meta{char code} \cs{s_@@}
% \end{quote}
% The \meta{tokens} \texttt{o}- \emph{and} \texttt{e}/\texttt{x}-expand to the
% original token in the token list or to the cluster of tokens
% corresponding to one Unicode character in the given encoding (for
% \pkg{l3str-convert}). The \meta{catcode} is given as a single hexadecimal
% digit, $0$ for control sequences. The \meta{char code} is given as a
% decimal number, $-1$ for control sequences.
%
% Using delimited arguments lets us build the \meta{tokens}
% progressively when doing an encoding conversion in \pkg{l3str-convert}. On the
% other hand, the delimiter \cs{s_@@} may not appear unbraced in
% \meta{tokens}. This is not a problem because we are careful to wrap
% control sequences in braces (as an argument to \cs{exp_not:n}) when
% converting from a general token list to the internal format.
%
% The current rule for converting a \meta{token} to a balanced set of
% \meta{tokens} which both \texttt{o}-expands and \texttt{e}/\texttt{x}-expands to
% it is the following.
% \begin{itemize}
%   \item A control sequence |\cs| becomes |\exp_not:n { \cs }|
%     \cs{s_@@} $0$ $-1$ \cs{s_@@}.
%   \item A begin-group character |{| becomes \cs{exp_after:wN} |{|
%     \cs{if_false:} |}| \cs{fi:} \cs{s_@@} $1$ \meta{char code}
%     \cs{s_@@}.
%   \item An end-group character |}| becomes \cs{if_false:} |{| \cs{fi:}
%     |}| \cs{s_@@} $2$ \meta{char code} \cs{s_@@}.
%   \item A character with any other category code becomes
%     \cs{exp_not:n} \Arg{character} \cs{s_@@} \meta{hex catcode}
%     \meta{char code} \cs{s_@@}.
% \end{itemize}
% In contrast, for \cs{peek_analysis_map_inline:n} we must allow for an
% input stream containing \tn{outer} macros, so that wrapping all
% control sequences in \cs{exp_not:n} is unsafe.  Instead, we write the
% more elaborate \cs{__kernel_exp_not:w} \cs{exp_after:wN} |{|
% \cs{exp_not:N} |\cs| |}|.  (On the other hand we make a better effort
% by avoiding \cs{exp_not:n} for characters other than active and macro
% parameters.)
%
%    \begin{macrocode}
%<*package>
%    \end{macrocode}
%
% \subsection{Variables and helper functions}
%
% \begin{variable}{\s_@@}
%   The scan mark \cs{s_@@} is used as a delimiter in the internal
%   format. This is more practical than using a quark, because we would
%   then need to control expansion much more carefully: compare
%   \cs{int_value:w} |`#1| \cs{s_@@} with \cs{int_value:w} |`#1|
%   \cs{exp_stop_f:} \cs{exp_not:N} \cs{q_mark} to extract a character
%   code followed by the delimiter in an \texttt{e}-expansion.
%    \begin{macrocode}
\scan_new:N \s_@@
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}
%   {\l_@@_analysis_token, \l_@@_analysis_char_token}
%   The tokens in the token list are probed with the \TeX{} primitive
%   \tn{futurelet}. We use \cs{l_@@_analysis_token} in that
%   construction. In some cases, we convert the following token to a
%   string before probing it: then the token variable used is
%   \cs{l_@@_analysis_char_token}.
%    \begin{macrocode}
\cs_new_eq:NN \l_@@_analysis_token ?
\cs_new_eq:NN \l_@@_analysis_char_token ?
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_@@_peek_code_tl}
%   Holds some code to be run once the next token has been fully
%   analysed in \cs{peek_analysis_map_inline:n}.
%    \begin{macrocode}
\tl_new:N \l_@@_peek_code_tl
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\c_@@_peek_catcodes_tl}
%   A token list containing the character number~$32$ (space) with all
%   possible category codes except $1$ and $2$ (begin-group and
%   end-group).  Why $32$?  Because some \LuaTeX{} versions only allow
%   creation of catcode~$10$ (space) tokens with this character code, so
%   that we decided to make \cs{char_generate:nn} refuse to create such
%   weird spaces as well.  We do not include the macro parameter case
%   (catcode~$6$) because it cannot be used as a macro delimiter.
%    \begin{macrocode}
\group_begin:
\char_set_active_eq:NN \  \scan_stop:
\tl_const:Ne \c_@@_peek_catcodes_tl
  {
    \char_generate:nn { 32 } { 3 }   3
    \char_generate:nn { 32 } { 4 }   4
    \char_generate:nn { 32 } { 7 }   7
    \char_generate:nn { 32 } { 8 }   8
    \c_space_tl                     \token_to_str:N A
    \char_generate:nn { 32 } { 11 } \token_to_str:N B
    \char_generate:nn { 32 } { 12 } \token_to_str:N C
    \char_generate:nn { 32 } { 13 } \token_to_str:N D
  }
\group_end:
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_@@_analysis_normal_int}
%   The number of normal (\texttt{N}-type argument) tokens since the
%   last special token.
%    \begin{macrocode}
\int_new:N \l_@@_analysis_normal_int
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_@@_analysis_index_int}
%   During the first pass, this is the index in the array being built.
%   During the second pass, it is equal to the maximum index in the
%   array from the first pass.
%    \begin{macrocode}
\int_new:N \l_@@_analysis_index_int
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_@@_analysis_nesting_int}
%   Nesting depth of explicit begin-group and end-group characters
%   during the first pass. This lets us detect the end of the token list
%   without a reserved end-marker.
%    \begin{macrocode}
\int_new:N \l_@@_analysis_nesting_int
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\l_@@_analysis_type_int}
%   When encountering special characters, we record their \enquote{type}
%   in this integer.
%    \begin{macrocode}
\int_new:N \l_@@_analysis_type_int
%    \end{macrocode}
% \end{variable}
%
% \begin{variable}{\g_@@_analysis_result_tl}
%   The result of the conversion is stored in this token list, with a
%   succession of items of the form
%   \begin{quote}
%     \meta{tokens} \cs{s_@@} \meta{catcode} \meta{char code} \cs{s_@@}
%   \end{quote}
%    \begin{macrocode}
\tl_new:N \g_@@_analysis_result_tl
%    \end{macrocode}
% \end{variable}
%
% \begin{macro}[EXP]{\@@_analysis_extract_charcode:}
% \begin{macro}[EXP]{\@@_analysis_extract_charcode_aux:w}
%   Extracting the character code from the meaning of
%   \cs{l_@@_analysis_token}. This has no error checking, and should
%   only be assumed to work for begin-group and end-group character
%   tokens. It produces a number in the form |`|\meta{char}.
%    \begin{macrocode}
\cs_new:Npn \@@_analysis_extract_charcode:
  {
    \exp_after:wN \@@_analysis_extract_charcode_aux:w
      \token_to_meaning:N \l_@@_analysis_token
  }
\cs_new:Npn \@@_analysis_extract_charcode_aux:w #1 ~ #2 ~ { ` }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[EXP]{\@@_analysis_cs_space_count:NN}
% \begin{macro}[EXP]{\@@_analysis_cs_space_count:w}
% \begin{macro}[EXP]{\@@_analysis_cs_space_count_end:w}
%   Counts the number of spaces in the string representation of its
%   second argument, as well as the number of characters following the
%   last space in that representation, and feeds the two numbers as
%   semicolon-delimited arguments to the first argument. When this
%   function is used, the escape character is printable and non-space.
%    \begin{macrocode}
\cs_new:Npn \@@_analysis_cs_space_count:NN #1 #2
  {
    \exp_after:wN #1
    \int_value:w \int_eval:w 0
      \exp_after:wN \@@_analysis_cs_space_count:w
        \token_to_str:N #2
        \fi: \@@_analysis_cs_space_count_end:w ; ~ !
  }
\cs_new:Npn \@@_analysis_cs_space_count:w #1 ~
  {
    \if_false: #1 #1 \fi:
    + 1
    \@@_analysis_cs_space_count:w
  }
\cs_new:Npn \@@_analysis_cs_space_count_end:w ; #1 \fi: #2 !
  { \exp_after:wN ; \int_value:w \str_count_ignore_spaces:n {#1} ; }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsection{Plan of attack}
%
% Our goal is to produce a token list of the form roughly
% \begin{quote}
%   \meta{token 1} \cs{s_@@} \meta{catcode 1} \meta{char code 1} \cs{s_@@} \\
%   \meta{token 2} \cs{s_@@} \meta{catcode 2} \meta{char code 2} \cs{s_@@} \\
%   \ldots{}
%   \meta{token N} \cs{s_@@} \meta{catcode N} \meta{char code N} \cs{s_@@}
% \end{quote}
% Most but not all tokens can be grabbed as an undelimited
% (\texttt{N}-type) argument by \TeX{}. The plan is to have a two pass
% system. In the first pass, locate special tokens, and store them in
% various \tn{toks} registers. In the second pass, which is done within
% an \texttt{e}-expanding assignment, normal tokens are taken in as
% \texttt{N}-type arguments, and special tokens are retrieved from the
% \tn{toks} registers, and removed from the input stream by some means.
% The whole process takes linear time, because we avoid building the
% result one item at a time.
%
% We make the escape character printable (backslash, but this later
% oscillates between slash and backslash): this allows us to
% distinguish characters from control sequences.
%
% A token has two characteristics: its \tn{meaning}, and what it looks
% like for \TeX{} when it is in scanning mode (\emph{e.g.}, when
% capturing parameters for a macro). For our purposes, we distinguish
% the following meanings:
% \begin{itemize}
%   \item begin-group token (category code $1$), either space (character
%     code $32$), or non-space;
%   \item end-group token (category code $2$), either space (character
%     code $32$), or non-space;
%   \item space token (category code $10$, character code $32$);
%   \item anything else (then the token is always an \texttt{N}-type
%     argument).
% \end{itemize}
% The token itself can \enquote{look like} one of the following
% \begin{itemize}
%   \item a non-active character, in which case its meaning is
%     automatically that associated to its character code and category
%     code, we call it \enquote{true} character;
%   \item an active character;
%   \item a control sequence.
% \end{itemize}
% The only tokens which are not valid \texttt{N}-type arguments are true
% begin-group characters, true end-group characters, and true spaces.
% We detect those characters by scanning ahead with \tn{futurelet},
% then distinguishing true characters from control sequences set equal
% to them using the \tn{string} representation.
%
% The second pass is a simple exercise in expandable loops.
%
% \begin{macro}{\@@_analysis:n}
%   Everything is done within a group, and all definitions are
%   local. We use \cs{group_align_safe_begin/end:} to avoid problems in
%   case \cs{@@_analysis:n} is used within an alignment and its argument
%   contains alignment tab tokens.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_analysis:n #1
  {
    \group_begin:
      \group_align_safe_begin:
        \@@_analysis_a:n {#1}
        \@@_analysis_b:n {#1}
      \group_align_safe_end:
    \group_end:
  }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Disabling active characters}
%
% \begin{macro}{\@@_analysis_disable:n}
%   Active characters can cause problems later on in the processing, so
%   we provide a way to disable them, by setting them to
%   \texttt{undefined}. Since Unicode contains too many characters to
%   loop over all of them, we instead do this whenever we encounter a
%   character.  For \pTeX{} and \upTeX{} we skip characters beyond
%   $[0,255]$ because \tn{lccode} only allows those values.
%    \begin{macrocode}
\group_begin:
  \char_set_catcode_active:N \^^@
  \cs_new_protected:Npn \@@_analysis_disable:n #1
    {
      \tex_lccode:D 0 = #1 \exp_stop_f:
      \tex_lowercase:D { \tex_let:D ^^@ } \tex_undefined:D
    }
  \bool_lazy_or:nnT
    { \sys_if_engine_ptex_p: }
    { \sys_if_engine_uptex_p: }
    {
      \cs_gset_protected:Npn \@@_analysis_disable:n #1
        {
          \if_int_compare:w 256 > #1 \exp_stop_f:
            \tex_lccode:D 0 = #1 \exp_stop_f:
            \tex_lowercase:D { \tex_let:D ^^@ } \tex_undefined:D
          \fi:
        }
    }
\group_end:
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_analysis_disable_char:N}
%   Similar to \cs{@@_analysis_disable:n}, but it receives a normal
%   character token, tests if that token is active (by turning it into
%   a space: the active space has been undefined at this point), and
%   if so, disables it.  Even if the character is active and set equal
%   to a primitive conditional, nothing blows up.
%   Again, in \pTeX{} and \upTeX{} we skip characters beyond $[0,255]$,
%   which cannot be active anyways.
%    \begin{macrocode}
\group_begin:
  \char_set_catcode_active:N \^^@
  \cs_new_protected:Npn \@@_analysis_disable_char:N #1
    {
      \tex_lccode:D `#1 = 32 \exp_stop_f:
      \tex_lowercase:D { \if_meaning:w #1 } \tex_undefined:D
        \tex_let:D #1 \tex_undefined:D
      \fi:
    }
  \bool_lazy_or:nnT
    { \sys_if_engine_ptex_p: }
    { \sys_if_engine_uptex_p: }
    {
      \cs_gset_protected:Npn \@@_analysis_disable_char:N #1
        {
          \if_int_compare:w 256 > `#1 \exp_stop_f:
            \tex_lccode:D `#1 = 32 \exp_stop_f:
            \tex_lowercase:D { \if_meaning:w #1 } \tex_undefined:D
              \tex_let:D #1 \tex_undefined:D
            \fi:
          \fi:
        }
    }
\group_end:
%    \end{macrocode}
% \end{macro}
%
% \subsection{First pass}
%
% The goal of this pass is to detect special (non-\texttt{N}-type) tokens,
% and count how many \texttt{N}-type tokens lie between special tokens.
% Also, we wish to store some representation of each special token
% in a \tn{toks} register.
%
% We have $11$ types of tokens:
% \begin{itemize}
% \item[1.] a true non-space begin-group character;
% \item[2.] a true space begin-group character;
% \item[3.] a true non-space end-group character;
% \item[4.] a true space end-group character;
% \item[5.] a true space blank space character;
% \item[6.] an active character;
% \item[7.] any other true character;
% \item[8.] a control sequence equal to a begin-group token (category code $1$);
% \item[9.] a control sequence equal to an end-group token (category code $2$);
% \item[10.] a control sequence equal to a space token
%   (character code $32$, category code $10$);
% \item[11.] any other control sequence.
% \end{itemize}
% Our first tool is \tn{futurelet}. This cannot distinguish
% case $8$ from $1$ or $2$, nor case $9$ from $3$ or $4$,
% nor case $10$ from case $5$. Those cases are later distinguished
% by applying the \tn{string} primitive to the following token,
% after possibly changing the escape character to ensure that
% a control sequence's string representation cannot be mistaken
% for the true character.
%
% In cases $6$, $7$, and $11$, the following token is a valid
% \texttt{N}-type argument, so we grab it and distinguish the case
% of a character from a control sequence: in the latter case,
% \cs{str_tail:n} \Arg{token} is non-empty, because the
% escape character is printable.
%
% \begin{macro}{\@@_analysis_a:n}
%   We read tokens one by one using \tn{futurelet}.
%   While performing the loop, we keep track of the number of
%   true begin-group characters minus the number of
%   true end-group characters in \cs{l_@@_analysis_nesting_int}.
%   This reaches $-1$ when we read the closing brace.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_analysis_a:n #1
  {
    \@@_analysis_disable:n { 32 }
    \int_set:Nn \tex_escapechar:D { 92 }
    \int_zero:N \l_@@_analysis_normal_int
    \int_zero:N \l_@@_analysis_index_int
    \int_zero:N \l_@@_analysis_nesting_int
    \if_false: { \fi: \@@_analysis_a_loop:w #1 }
    \int_decr:N \l_@@_analysis_index_int
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_analysis_a_loop:w}
%   Read one character and check its type.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_analysis_a_loop:w
  { \tex_futurelet:D \l_@@_analysis_token \@@_analysis_a_type:w }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_analysis_a_type:w}
%   At this point, \cs{l_@@_analysis_token} holds the meaning
%   of the following token. We store in \cs{l_@@_analysis_type_int}
%   information about the meaning of the token ahead:
%   \begin{itemize}
%   \item 0 space token;
%   \item 1 begin-group token;
%   \item -1 end-group token;
%   \item 2 other.
%   \end{itemize}
%   The values $0$, $1$, $-1$ correspond to how much a true such
%   character changes the nesting level ($2$ is used only here,
%   and is irrelevant later). Then call the auxiliary for each case.
%   Note that nesting conditionals here is safe because we only skip
%   over \cs{l_@@_analysis_token} if it matches with one of the
%   character tokens (hence is not a primitive conditional).
%    \begin{macrocode}
\cs_new_protected:Npn \@@_analysis_a_type:w
  {
    \l_@@_analysis_type_int =
      \if_meaning:w \l_@@_analysis_token \c_space_token
        0
      \else:
        \if_catcode:w \exp_not:N \l_@@_analysis_token \c_group_begin_token
          1
        \else:
          \if_catcode:w \exp_not:N \l_@@_analysis_token \c_group_end_token
            - 1
          \else:
            2
          \fi:
        \fi:
      \fi:
      \exp_stop_f:
    \if_case:w \l_@@_analysis_type_int
         \exp_after:wN \@@_analysis_a_space:w
    \or: \exp_after:wN \@@_analysis_a_bgroup:w
    \or: \exp_after:wN \@@_analysis_a_safe:N
    \else: \exp_after:wN \@@_analysis_a_egroup:w
    \fi:
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_analysis_a_space:w}
% \begin{macro}{\@@_analysis_a_space_test:w}
%   In this branch, the following token's meaning is a blank space.
%   Apply \tn{string} to that token: a true blank space gives a space, a
%   control sequence gives a result starting with the escape character,
%   an active character gives something else than a space since we
%   disabled the space.  We grab as \cs{l_@@_analysis_char_token} the first
%   character of the string representation then test it in
%   \cs{@@_analysis_a_space_test:w}.
%   Also, since \cs{@@_analysis_a_store:} expects the special token to be
%   stored in the relevant \tn{toks} register, we do that. The extra
%   \cs{exp_not:n} is unnecessary of course, but it makes the treatment
%   of all tokens more homogeneous.
%   If we discover that the next token was actually a control sequence
%   or an active character
%   instead of a true space, then we step the counter of normal tokens.
%   We now have in front of us the whole string representation of
%   the control sequence, including potential spaces; those will appear
%   to be true spaces later in this pass. Hence, all other branches of
%   the code in this first pass need to consider the string representation,
%   so that the second pass does not need to test the meaning of tokens,
%   only strings.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_analysis_a_space:w
  {
    \tex_afterassignment:D \@@_analysis_a_space_test:w
    \exp_after:wN \cs_set_eq:NN
    \exp_after:wN \l_@@_analysis_char_token
    \token_to_str:N
  }
\cs_new_protected:Npn \@@_analysis_a_space_test:w
  {
    \if_meaning:w \l_@@_analysis_char_token \c_space_token
      \tex_toks:D \l_@@_analysis_index_int { \exp_not:n { ~ } }
      \@@_analysis_a_store:
    \else:
      \int_incr:N \l_@@_analysis_normal_int
    \fi:
    \@@_analysis_a_loop:w
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\@@_analysis_a_bgroup:w, \@@_analysis_a_egroup:w}
% \begin{macro}
%   {\@@_analysis_a_group:nw, \@@_analysis_a_group_aux:w, \@@_analysis_a_group_auxii:w, \@@_analysis_a_group_test:w}
%   The token is most likely a true character token with catcode $1$ or
%   $2$, but it might be a control sequence, or an active character.
%   Optimizing for the first case, we store in a toks register some code
%   that expands to that token.  Since we will turn what follows into
%   a string, we make sure the escape character is different from the
%   current character code (by switching between solidus and backslash).
%   To detect the special case of an active character let to the catcode
%   $1$ or~$2$ character with the same character code, we disable the
%   active character with that character code and re-test: if the
%   following token has become undefined we can in fact safely grab it.
%   We are finally ready to turn what follows to a string and test it.
%   This is one place where we need \cs{l_@@_analysis_char_token} to be a
%   separate control sequence from \cs{l_@@_analysis_token}, to compare them.
%    \begin{macrocode}
\group_begin:
  \char_set_catcode_group_begin:N \^^@ % {
  \cs_new_protected:Npn \@@_analysis_a_bgroup:w
    { \@@_analysis_a_group:nw { \exp_after:wN ^^@ \if_false: } \fi: } }
  \char_set_catcode_group_end:N \^^@
  \cs_new_protected:Npn \@@_analysis_a_egroup:w
    { \@@_analysis_a_group:nw { \if_false: { \fi: ^^@ } } % }
\group_end:
\cs_new_protected:Npn \@@_analysis_a_group:nw #1
  {
    \tex_lccode:D 0 = \@@_analysis_extract_charcode: \scan_stop:
    \tex_lowercase:D { \tex_toks:D \l_@@_analysis_index_int {#1} }
    \if_int_compare:w \tex_lccode:D 0 = \tex_escapechar:D
      \int_set:Nn \tex_escapechar:D { 139 - \tex_escapechar:D }
    \fi:
    \@@_analysis_disable:n { \tex_lccode:D 0 }
    \tex_futurelet:D \l_@@_analysis_token \@@_analysis_a_group_aux:w
  }
\cs_new_protected:Npn \@@_analysis_a_group_aux:w
  {
    \if_meaning:w \l_@@_analysis_token \tex_undefined:D
      \exp_after:wN \@@_analysis_a_safe:N
    \else:
      \exp_after:wN \@@_analysis_a_group_auxii:w
    \fi:
  }
\cs_new_protected:Npn \@@_analysis_a_group_auxii:w
  {
    \tex_afterassignment:D \@@_analysis_a_group_test:w
    \exp_after:wN \cs_set_eq:NN
    \exp_after:wN \l_@@_analysis_char_token
    \token_to_str:N
  }
\cs_new_protected:Npn \@@_analysis_a_group_test:w
  {
    \if_charcode:w \l_@@_analysis_token \l_@@_analysis_char_token
      \@@_analysis_a_store:
    \else:
      \int_incr:N \l_@@_analysis_normal_int
    \fi:
    \@@_analysis_a_loop:w
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}{\@@_analysis_a_store:}
%   This function is called each time we meet a special token;
%   at this point, the \tn{toks} register \cs{l_@@_analysis_index_int}
%   holds a token list which expands to the given special token.
%   Also, the value of \cs{l_@@_analysis_type_int} indicates which case
%   we are in:
%   \begin{itemize}
%   \item -1 end-group character;
%   \item 0 space character;
%   \item 1 begin-group character.
%   \end{itemize}
%   We need to distinguish further the case of a space character
%   (code $32$) from other character codes, because those
%   behave differently in the second pass. Namely, after testing
%   the \tn{lccode} of $0$ (which holds the present character code)
%   we change the cases above to
%   \begin{itemize}
%   \item -2 space end-group character;
%   \item -1 non-space end-group character;
%   \item 0 space blank space character;
%   \item 1 non-space begin-group character;
%   \item 2 space begin-group character.
%   \end{itemize}
%   This has the property that non-space characters correspond to odd
%   values of \cs{l_@@_analysis_type_int}.  The number of normal tokens until
%   here and the type of special token are packed into a \tn{skip}
%   register.  Finally, we check whether we reached the last closing
%   brace, in which case we stop by disabling the looping function
%   (locally).
%    \begin{macrocode}
\cs_new_protected:Npn \@@_analysis_a_store:
  {
    \tex_advance:D \l_@@_analysis_nesting_int \l_@@_analysis_type_int
    \if_int_compare:w \tex_lccode:D 0 = `\ \exp_stop_f:
      \tex_advance:D \l_@@_analysis_type_int \l_@@_analysis_type_int
    \fi:
    \tex_skip:D \l_@@_analysis_index_int
      = \l_@@_analysis_normal_int sp
         plus \l_@@_analysis_type_int sp \scan_stop:
    \int_incr:N \l_@@_analysis_index_int
    \int_zero:N \l_@@_analysis_normal_int
    \if_int_compare:w \l_@@_analysis_nesting_int = - \c_one_int
      \cs_set_eq:NN \@@_analysis_a_loop:w \scan_stop:
    \fi:
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\@@_analysis_a_safe:N}
% \begin{macro}{\@@_analysis_a_cs:ww}
%   This should be the simplest case: since the upcoming token is safe,
%   we can simply grab it in a second pass. If the token is a single
%   character (including space), the \cs{if_charcode:w} test yields
%   true; we disable a potentially active character (that could
%   otherwise masquerade as the true character in the next pass) and we
%   count one \enquote{normal} token. On the other
%   hand, if the token is a control sequence, we should replace it by
%   its string representation for compatibility with other code
%   branches. Instead of slowly looping through the characters with
%   the main code, we use the knowledge of how the second pass works:
%   if the control sequence name contains no space, count that token
%   as a number of normal tokens equal to its string length. If the
%   control sequence contains spaces, they should be registered as
%   special characters by increasing \cs{l_@@_analysis_index_int}
%   (no need to carefully count character between each space), and
%   all characters after the last space should be counted in the
%   following sequence of \enquote{normal} tokens.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_analysis_a_safe:N #1
  {
    \if_charcode:w
        \scan_stop:
        \exp_after:wN \use_none:n \token_to_str:N #1 \prg_do_nothing:
        \scan_stop:
      \exp_after:wN \use_i:nn
    \else:
      \exp_after:wN \use_ii:nn
    \fi:
      {
        \@@_analysis_disable_char:N #1
        \int_incr:N \l_@@_analysis_normal_int
      }
      { \@@_analysis_cs_space_count:NN \@@_analysis_a_cs:ww #1 }
    \@@_analysis_a_loop:w
  }
\cs_new_protected:Npn \@@_analysis_a_cs:ww #1; #2;
  {
    \if_int_compare:w #1 > \c_zero_int
      \tex_skip:D \l_@@_analysis_index_int
        = \int_eval:n { \l_@@_analysis_normal_int + 1 } sp \exp_stop_f:
      \tex_advance:D \l_@@_analysis_index_int #1 \exp_stop_f:
    \else:
      \tex_advance:D
    \fi:
    \l_@@_analysis_normal_int #2 \exp_stop_f:
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \subsection{Second pass}
%
% The second pass is an exercise in expandable loops.
% All the necessary information is stored in \tn{skip}
% and \tn{toks} registers.
%
% \begin{macro}{\@@_analysis_b:n}
% \begin{macro}[EXP]{\@@_analysis_b_loop:w}
%   Start the loop with the index $0$. No need for an end-marker:
%   the loop stops by itself when the last index is read.
%   We repeatedly oscillate between reading long stretches
%   of normal tokens, and reading special tokens.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_analysis_b:n #1
  {
    \__kernel_tl_gset:Nx \g_@@_analysis_result_tl
      {
        \@@_analysis_b_loop:w 0; #1
        \prg_break_point:
      }
  }
\cs_new:Npn \@@_analysis_b_loop:w #1;
  {
    \exp_after:wN \@@_analysis_b_normals:ww
      \int_value:w \tex_skip:D #1 ; #1 ;
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[EXP]{\@@_analysis_b_normals:ww}
% \begin{macro}[EXP]{\@@_analysis_b_normal:wwN}
%   The first argument is the number of normal tokens which remain
%   to be read, and the second argument is the index in the array
%   produced in the first step.
%   A character's string representation is always one character long,
%   while a control sequence is always longer (we have set the escape
%   character to a printable value). In both cases, we leave
%   \cs{exp_not:n} \Arg{token} \cs{s_@@} in the input stream
%   (after \texttt{e}-expansion). Here, \cs{exp_not:n} is used
%   rather than \cs{exp_not:N} because |#3| could be
%   a macro parameter character or could be \cs{s_@@}
%   (which must be hidden behind braces in the result).
%    \begin{macrocode}
\cs_new:Npn \@@_analysis_b_normals:ww #1;
  {
    \if_int_compare:w #1 = \c_zero_int
      \@@_analysis_b_special:w
    \fi:
    \@@_analysis_b_normal:wwN #1;
  }
\cs_new:Npn \@@_analysis_b_normal:wwN #1; #2; #3
  {
    \exp_not:n { \exp_not:n { #3 } } \s_@@
    \if_charcode:w
        \scan_stop:
        \exp_after:wN \use_none:n \token_to_str:N #3 \prg_do_nothing:
        \scan_stop:
      \exp_after:wN \@@_analysis_b_char:Nn
      \exp_after:wN \@@_analysis_b_char_aux:nww
    \else:
      \exp_after:wN \@@_analysis_b_cs:Nww
    \fi:
    #3 #1; #2;
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[EXP]{\@@_analysis_b_char:Nn, \@@_analysis_b_char_aux:nww}
%   This function is called here with arguments
%   \cs{@@_analysis_b_char_aux:nww} and a normal character, while in the
%   peek analysis code it is called with \cs{use_none:n} and possibly a
%   space character, which is why the function has signature |Nn|.
%   If the normal token we grab is a character, leave
%   \meta{catcode} \meta{charcode} followed by \cs{s_@@}
%   in the input stream, and call \cs{@@_analysis_b_normals:ww}
%   with its first argument decremented.
%    \begin{macrocode}
\cs_new:Npe \@@_analysis_b_char:Nn #1#2
  {
    \exp_not:N \if_meaning:w #2 \exp_not:N \tex_undefined:D
      \token_to_str:N D \exp_not:N \else:
    \exp_not:N \if_catcode:w #2 \c_catcode_other_token
      \token_to_str:N C \exp_not:N \else:
    \exp_not:N \if_catcode:w #2 \c_catcode_letter_token
      \token_to_str:N B \exp_not:N \else:
    \exp_not:N \if_catcode:w #2 \c_math_toggle_token      3
      \exp_not:N \else:
    \exp_not:N \if_catcode:w #2 \c_alignment_token        4
      \exp_not:N \else:
    \exp_not:N \if_catcode:w #2 \c_math_superscript_token 7
      \exp_not:N \else:
    \exp_not:N \if_catcode:w #2 \c_math_subscript_token   8
      \exp_not:N \else:
    \exp_not:N \if_catcode:w #2 \c_space_token
      \token_to_str:N A \exp_not:N \else:
      6
    \exp_not:n { \fi: \fi: \fi: \fi: \fi: \fi: \fi: \fi: }
    #1 {#2}
  }
\cs_new:Npn \@@_analysis_b_char_aux:nww #1
  {
    \int_value:w `#1 \s_@@
    \exp_after:wN \@@_analysis_b_normals:ww
      \int_value:w \int_eval:w - 1 +
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[EXP]{\@@_analysis_b_cs:Nww}
% \begin{macro}[EXP]{\@@_analysis_b_cs_test:ww}
%   If the token we grab is a control sequence, leave
%   |0 -1| (as category code and character code) in the input stream,
%   followed by \cs{s_@@},
%   and call \cs{@@_analysis_b_normals:ww} with updated arguments.
%    \begin{macrocode}
\cs_new:Npn \@@_analysis_b_cs:Nww #1
  {
    0 -1 \s_@@
    \@@_analysis_cs_space_count:NN \@@_analysis_b_cs_test:ww #1
  }
\cs_new:Npn \@@_analysis_b_cs_test:ww #1 ; #2 ; #3 ; #4 ;
  {
    \exp_after:wN \@@_analysis_b_normals:ww
    \int_value:w \int_eval:w
    \if_int_compare:w #1 = \c_zero_int
      #3
    \else:
      \tex_skip:D \int_eval:n { #4 + #1 } \exp_stop_f:
    \fi:
    - #2
    \exp_after:wN ;
    \int_value:w \int_eval:n { #4 + #1 } ;
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
%
% \begin{macro}[EXP]{\@@_analysis_b_special:w}
% \begin{macro}[EXP]{\@@_analysis_b_special_char:wN}
% \begin{macro}[EXP]{\@@_analysis_b_special_space:w}
%   Here, |#1| is the current index in the array built in the first pass.
%   Check now whether we reached the end (we shouldn't keep the trailing
%   end-group character that marked the end of the token list in the
%   first pass).
%   Unpack the \tn{toks} register: when \texttt{e}/\texttt{x}-expanding again,
%   we will get the special token.
%   Then leave the category code in the input stream, followed by
%   the character code, and call \cs{@@_analysis_b_loop:w} with the next index.
%    \begin{macrocode}
\group_begin:
  \char_set_catcode_other:N A
  \cs_new:Npn \@@_analysis_b_special:w
      \fi: \@@_analysis_b_normal:wwN 0 ; #1 ;
    {
      \fi:
      \if_int_compare:w #1 = \l_@@_analysis_index_int
        \exp_after:wN \prg_break:
      \fi:
      \tex_the:D \tex_toks:D #1 \s_@@
      \if_case:w \tex_gluestretch:D \tex_skip:D #1 \exp_stop_f:
             \token_to_str:N A
      \or:   1
      \or:   1
      \else: 2
      \fi:
      \if_int_odd:w \tex_gluestretch:D \tex_skip:D #1 \exp_stop_f:
        \exp_after:wN \@@_analysis_b_special_char:wN \int_value:w
      \else:
        \exp_after:wN \@@_analysis_b_special_space:w \int_value:w
      \fi:
      \int_eval:n { 1 + #1 } \exp_after:wN ;
      \token_to_str:N
    }
\group_end:
\cs_new:Npn \@@_analysis_b_special_char:wN #1 ; #2
  {
    \int_value:w `#2 \s_@@
    \@@_analysis_b_loop:w #1 ;
  }
\cs_new:Npn \@@_analysis_b_special_space:w #1 ; ~
  {
    32 \s_@@
    \@@_analysis_b_loop:w #1 ;
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsection{Mapping through the analysis}
%
% \begin{macro}{\tl_analysis_map_inline:Nn, \tl_analysis_map_inline:nn}
% \begin{macro}{\@@_analysis_map:Nn}
% \begin{macro}{\@@_analysis_map:NwNw}
%   First obtain the analysis of the token list into
%   \cs{g_@@_analysis_result_tl}. To allow nested mappings, increase the
%   nesting depth \cs{g__kernel_prg_map_int} (shared between all
%   modules), then define the payload macro, which runs the user code
%   and has a name specific to that nesting depth. The looping macro
%   grabs the \meta{tokens}, \meta{catcode} and \meta{char code}; it
%   checks for the end of the loop with \cs{use_none:n} |##2|, normally
%   empty, but which becomes \cs{tl_map_break:} at the end; it then
%   calls the payload macro with the arguments in the correct order
%   (this is the reason why we cannot directly use the same macro for
%   looping and payload), and loops by calling itself. When the loop
%   ends, remember to decrease the nesting depth.
%    \begin{macrocode}
\cs_new_protected:Npn \tl_analysis_map_inline:Nn #1
  { \exp_args:No \tl_analysis_map_inline:nn #1 }
\cs_new_protected:Npn \tl_analysis_map_inline:nn #1
  {
    \@@_analysis:n {#1}
    \int_gincr:N \g__kernel_prg_map_int
    \exp_args:Nc \@@_analysis_map:Nn
      { @@_analysis_map_inline_ \int_use:N \g__kernel_prg_map_int :wNw }
  }
\cs_new_protected:Npn \@@_analysis_map:Nn #1#2
  {
    \cs_gset_protected:Npn #1 ##1##2##3 {#2}
    \exp_after:wN \@@_analysis_map:NwNw \exp_after:wN #1
      \g_@@_analysis_result_tl
      \s_@@ { ? \tl_map_break: } \s_@@
    \prg_break_point:Nn \tl_map_break:
      { \int_gdecr:N \g__kernel_prg_map_int }
  }
\cs_new_protected:Npn \@@_analysis_map:NwNw #1 #2 \s_@@ #3 #4 \s_@@
  {
    \use_none:n #3
    #1 {#2} {#4} {#3}
    \@@_analysis_map:NwNw #1
  }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsection{Showing the results}
%
% \begin{macro}{\tl_analysis_show:N, \tl_analysis_log:N, \@@_analysis_show:NNN}
%   Add to \cs{@@_analysis:n} a third pass to display tokens to the terminal.
%   If the token list variable is not defined, throw the same error
%   as \cs{tl_show:N} by simply calling that function.
%    \begin{macrocode}
\cs_new_protected:Npn \tl_analysis_show:N
  { \@@_analysis_show:NNN \msg_show:nneeee \tl_show:N }
\cs_new_protected:Npn \tl_analysis_log:N
  { \@@_analysis_show:NNN \msg_log:nneeee \tl_log:N }
\cs_new_protected:Npn \@@_analysis_show:NNN #1#2#3
  {
    \tl_if_exist:NTF #3
      {
        \exp_args:No \@@_analysis:n {#3}
        #1 { tl } { show-analysis }
          { \token_to_str:N #3 } { \@@_analysis_show: } { } { }
      }
      { #2 #3 }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}{\tl_analysis_show:n, \tl_analysis_log:n, \@@_analysis_show:Nn}
%   No existence test needed here.
%    \begin{macrocode}
\cs_new_protected:Npn \tl_analysis_show:n
  { \@@_analysis_show:Nn \msg_show:nneeee }
\cs_new_protected:Npn \tl_analysis_log:n
  { \@@_analysis_show:Nn \msg_log:nneeee }
\cs_new_protected:Npn \@@_analysis_show:Nn #1#2
  {
    \@@_analysis:n {#2}
    #1 { tl } { show-analysis } { } { \@@_analysis_show: } { } { }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[rEXP]{\@@_analysis_show:, \@@_analysis_show_loop:wNw}
%   Here, |#1| \texttt{o}- and \texttt{e}/\texttt{x}-expands to the token;
%   |#2| is the category code (one uppercase hexadecimal digit),
%   $0$ for control sequences;
%   |#3| is the character code, which we ignore.
%   In the cases of control sequences and active characters,
%   the meaning may overflow one line, and we want to truncate
%   it. Those cases are thus separated out.
%    \begin{macrocode}
\cs_new:Npn \@@_analysis_show:
  {
    \exp_after:wN \@@_analysis_show_loop:wNw \g_@@_analysis_result_tl
    \s_@@ { ? \prg_break: } \s_@@
    \prg_break_point:
  }
\cs_new:Npn \@@_analysis_show_loop:wNw #1 \s_@@ #2 #3 \s_@@
  {
    \use_none:n #2
    \iow_newline: > \use:nn { ~ } { ~ }
    \if_int_compare:w "#2 = \c_zero_int
      \exp_after:wN \@@_analysis_show_cs:n
    \else:
      \if_int_compare:w "#2 = 13 \exp_stop_f:
        \exp_after:wN \exp_after:wN
        \exp_after:wN \@@_analysis_show_active:n
      \else:
        \exp_after:wN \exp_after:wN
        \exp_after:wN \@@_analysis_show_normal:n
      \fi:
    \fi:
    {#1}
    \@@_analysis_show_loop:wNw
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[rEXP]{\@@_analysis_show_normal:n}
%   Non-active characters are a simple matter of printing
%   the character, and its meaning. Our test suite checks that
%   begin-group and end-group characters do not mess up
%   \TeX{}'s alignment status.
%    \begin{macrocode}
\cs_new:Npn \@@_analysis_show_normal:n #1
  {
    \exp_after:wN \token_to_str:N #1 ~
    ( \exp_after:wN \token_to_meaning:N #1 )
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[EXP]{\@@_analysis_show_value:N}
%   This expands to the value of |#1| if it has any.
%    \begin{macrocode}
\cs_new:Npn \@@_analysis_show_value:N #1
  {
    \token_if_expandable:NF #1
      {
        \token_if_chardef:NTF       #1 \prg_break: { }
        \token_if_mathchardef:NTF   #1 \prg_break: { }
        \token_if_dim_register:NTF  #1 \prg_break: { }
        \token_if_int_register:NTF  #1 \prg_break: { }
        \token_if_skip_register:NTF #1 \prg_break: { }
        \token_if_toks_register:NTF #1 \prg_break: { }
        \use_none:nnn
        \prg_break_point:
        \use:n { \exp_after:wN = \tex_the:D #1 }
      }
  }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}[rEXP]{\@@_analysis_show_cs:n}
% \begin{macro}[rEXP]{\@@_analysis_show_active:n}
% \begin{macro}[rEXP]{\@@_analysis_show_long:nn}
% \begin{macro}[rEXP]{\@@_analysis_show_long_aux:nnnn}
%   Control sequences and active characters are printed in the same way,
%   making sure not to go beyond the \cs{l_iow_line_count_int}. In case
%   of an overflow, we replace the last characters by
%   \cs{c_@@_analysis_show_etc_str}.
%    \begin{macrocode}
\cs_new:Npn \@@_analysis_show_cs:n #1
  { \exp_args:No \@@_analysis_show_long:nn {#1} { control~sequence= } }
\cs_new:Npn \@@_analysis_show_active:n #1
  { \exp_args:No \@@_analysis_show_long:nn {#1} { active~character= } }
\cs_new:Npn \@@_analysis_show_long:nn #1
  {
    \@@_analysis_show_long_aux:oofn
      { \token_to_str:N #1 }
      { \token_to_meaning:N #1 }
      { \@@_analysis_show_value:N #1 }
  }
\cs_new:Npn \@@_analysis_show_long_aux:nnnn #1#2#3#4
  {
    \int_compare:nNnTF
      { \str_count:n { #1 ~ ( #4 #2 #3 ) } }
      > { \l_iow_line_count_int - 3 }
      {
        \str_range:nnn { #1 ~ ( #4 #2 #3 ) } { 1 }
          {
            \l_iow_line_count_int - 3
            - \str_count:N \c_@@_analysis_show_etc_str
          }
        \c_@@_analysis_show_etc_str
      }
      { #1 ~ ( #4 #2 #3 ) }
  }
\cs_generate_variant:Nn \@@_analysis_show_long_aux:nnnn { oof }
%    \end{macrocode}
% \end{macro}
% \end{macro}
% \end{macro}
% \end{macro}
%
% \subsection{Peeking ahead}
%
% \begin{macro}[EXP]{\peek_analysis_map_break:, \peek_analysis_map_break:n}
%   The break statements use the general \cs{prg_map_break:Nn}.
%    \begin{macrocode}
\cs_new:Npn \peek_analysis_map_break:
  { \prg_map_break:Nn \peek_analysis_map_break: { } }
\cs_new:Npn \peek_analysis_map_break:n
  { \prg_map_break:Nn \peek_analysis_map_break: }
%    \end{macrocode}
% \end{macro}
%
% \begin{variable}{\l_@@_peek_charcode_int}
%    \begin{macrocode}
\int_new:N \l_@@_peek_charcode_int
%    \end{macrocode}
% \end{variable}
%
% \begin{macro}{\@@_analysis_char_arg:Nw, \@@_analysis_char_arg_aux:Nw}
%   After a call to \tn{futurelet} \cs{l_@@_analysis_token} followed by
%   a stringified character token (either explicit space or catcode
%   other character), grab the argument and pass it to |#1|.  We only
%   need to do anything in the case of a space.
%    \begin{macrocode}
\cs_new:Npn \@@_analysis_char_arg:Nw
  {
    \if_meaning:w \l_@@_analysis_token \c_space_token
      \exp_after:wN \@@_analysis_char_arg_aux:Nw
    \fi:
  }
\cs_new:Npn \@@_analysis_char_arg_aux:Nw #1 ~ { #1 { ~ } }
%    \end{macrocode}
% \end{macro}
%
% \begin{macro}
%   {
%     \peek_analysis_map_inline:n,
%     \@@_peek_analysis_loop:NNn, \@@_peek_analysis_test:,
%     \@@_peek_analysis_exp:N, \@@_peek_analysis_exp_aux:N,
%     \@@_peek_analysis_nonexp:N, \@@_peek_analysis_cs:N,
%     \@@_peek_analysis_char:N, \@@_peek_analysis_char:w,
%     \@@_peek_analysis_special:, \@@_peek_analysis_retest:,
%     \@@_peek_analysis_str:,
%     \@@_peek_analysis_str:w, \@@_peek_analysis_str:n,
%     \@@_peek_analysis_active_str:n, \@@_peek_analysis_explicit:n,
%     \@@_peek_analysis_escape:, \@@_peek_analysis_collect:w,
%     \@@_peek_analysis_collect:n, \@@_peek_analysis_collect_loop:,
%     \@@_peek_analysis_collect_test:, \@@_peek_analysis_collect_end:NNNN
%   }
%   Save the user's code in a control sequence that is suitable for
%   nested maps.  We may wish to pass to this function an \tn{outer}
%   control sequence or active character; for this we will undefine
%   any expandable token (testing if it is \tn{outer} is much slower)
%   within a group, closed immediately after the function reads its
%   arguments to avoid affecting the user's code or even our peek code
%   (there is no risk of undefining \cs{group_end:} itself since that is
%   not expandable).  This user's code function also
%   calls the loop auxiliary, and includes the trailing
%   \cs{prg_break_point:Nn} for when the user wants to stop the loop.
%   The loop auxiliary must remove that break point because it must look
%   at the input stream.
%    \begin{macrocode}
\cs_new_protected:Npn \peek_analysis_map_inline:n #1
  {
    \group_align_safe_begin:
    \int_gincr:N \g__kernel_prg_map_int
    \cs_set_protected:cpn
      { @@_analysis_map_ \int_use:N \g__kernel_prg_map_int :nnN }
      ##1##2##3
      {
        \group_end:
        #1
        \@@_peek_analysis_loop:NNn
          \prg_break_point:Nn \peek_analysis_map_break:
            {
              \int_gdecr:N \g__kernel_prg_map_int
              \group_align_safe_end:
            }
      }
    \@@_peek_analysis_loop:NNn ? ? ?
  }
%    \end{macrocode}
%   The loop starts a group (closed by the user-code function defined
%   above) with a normalized escape character, and checks if the next
%   token is special or \texttt{N}-type (distinguishing expandable from
%   non-expandable tokens).  The test for nonexpandable tokens in
%   \cs{@@_peek_analysis_test:} must be done after the tests for
%   begin-group, end-group, and space tokens, in case \cs{l_peek_token}
%   is either \tn{outer} or is a primitive \TeX{} conditional, as such
%   tokens cannot be skipped over correctly by conditional code.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_peek_analysis_loop:NNn #1#2#3
  {
    \group_begin:
    \tl_set:Ne \l_@@_peek_code_tl
      {
        \exp_not:c
          { @@_analysis_map_ \int_use:N \g__kernel_prg_map_int :nnN }
      }
    \int_set:Nn \tex_escapechar:D { `\\ }
    \peek_after:Nw \@@_peek_analysis_test:
  }
\cs_new_protected:Npn \@@_peek_analysis_test:
  {
    \if_case:w
      \if_catcode:w \exp_not:N \l_peek_token {   \c_max_int \fi:
      \if_catcode:w \exp_not:N \l_peek_token }   \c_max_int \fi:
      \if_meaning:w \l_peek_token \c_space_token \c_max_int \fi:
      \exp_after:wN \if_meaning:w \exp_not:N \l_peek_token \l_peek_token
        \c_one_int
      \fi:
      \c_zero_int
      \exp_after:wN \exp_after:wN
      \exp_after:wN \@@_peek_analysis_exp:N
      \exp_after:wN \exp_not:N
    \or:
      \exp_after:wN \@@_peek_analysis_nonexp:N
    \else:
      \exp_after:wN \@@_peek_analysis_special:
    \fi:
  }
%    \end{macrocode}
%   Expandable tokens (which are automatically |N|-type) can be
%   \tn{outer} macros, hence the need for \cs{exp_after:wN} and
%   \cs{exp_not:N} in the code above, which allows the next function to
%   safely grab the token as an argument.  To allow the
%   possibly-\tn{outer} token~|#1| as an argument of the \meta{user's
%   function} (which is protected and stored in \cs{l_@@_peek_code_tl}),
%   we set it equal to a harmless macro.  This must be done at the very
%   last minute because |#1| may be some pretty important function such
%   as \cs{exp_after:wN}.  Using a primitive \cs{cs_set_nopar:Npe}
%   expansion (to avoid \tn{outer} problems) we set up to run the code
%   \tn{let} |#1| \meta{user's function} \meta{user's function} followed
%   by arguments involving~|#1|.  Regardless of~|#1| (including the
%   user's function itself), the user's function is run.  It always
%   starts with \cs{group_end:}, which has not been redefined since |#1|
%   started out as expandable, and which restores the definition of~|#1|.
%
%   Then we put the elaborate first argument
%   \cs{__kernel_exp_not:w} \cs{exp_after:wN} |{| \cs{exp_not:N} |#1| |}|:
%   indeed we cannot use \cs{exp_not:n} |{#1}| as this breaks for an
%   \tn{outer} macro and we cannot use \cs{exp_not:N} |#1|, as
%   \texttt{o}-expanding this yields a \enquote{notexpanded} token equal
%   to (a weird) \tn{relax}, which would have the wrong value for
%   primitive \TeX{} conditionals such as \cs{if_meaning:w}.
%
%   Then we must add |{-1}0| if the token is a
%   control sequence and \Arg{charcode}|D| otherwise.  Distinguishing
%   the two cases is easy: since we have made the escape character
%   printable, \cs{token_to_str:N} gives at least two characters for a
%   control sequence versus a single one for an active character
%   (possibly being a space, in which case the trailing brace group is
%   taken as the first argument of \cs{@@_peek_analysis_exp_aux:Nw}).
%   Importantly, |#1| could be an \tn{outer} token (as it is only set to
%   \cs{scan_stop:} at the last minute) but once we apply
%   \cs{token_to_str:N} we no longer need to worry about it.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_peek_analysis_exp:N #1
  {
    \cs_set_nopar:Npe \l_@@_peek_code_tl
      {
        \tex_let:D \exp_not:N #1 \l_@@_peek_code_tl
        \l_@@_peek_code_tl
        {
          \exp_not:n { \__kernel_exp_not:w \exp_after:wN }
            { \exp_not:N \exp_not:N \exp_not:N #1 }
        }
        \exp_after:wN \@@_peek_analysis_exp_aux:Nw
        \token_to_str:N #1 { } \s_@@
      }
    \l_@@_peek_code_tl
  }
\cs_new:Npe \@@_peek_analysis_exp_aux:Nw #1#2 \s_@@
  {
    \exp_not:N \if_meaning:w \scan_stop: #2 \scan_stop:
      { \exp_not:N \int_value:w `#1 ~ } \token_to_str:N D
    \exp_not:N \else:
      { -1 } 0
    \exp_not:N \fi:
  }
%    \end{macrocode}
%   For normal non-expandable tokens we must distinguish characters
%   (including active ones and macro parameter characters) from control
%   sequences (whose string representation is more than one character
%   because we made the escape character printable).  For a control
%   sequence call the user code with suitable arguments, wrapping |#1|
%   within \cs{exp_not:n} just in case it happens to be equal to a macro
%   parameter character.  We do not skip \cs{exp_not:n} when
%   unnecessary, because this auxiliary is also called in
%   \cs{@@_peek_analysis_retest:} where we have changed some control
%   sequences or active characters to \cs{scan_stop:} temporarily.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_peek_analysis_nonexp:N #1
  {
    \if_charcode:w
        \scan_stop:
        \exp_after:wN \use_none:n \token_to_str:N #1 \prg_do_nothing:
        \scan_stop:
      \exp_after:wN \@@_peek_analysis_char:N
    \else:
      \exp_after:wN \@@_peek_analysis_cs:N
    \fi:
    #1
  }
\cs_new_protected:Npn \@@_peek_analysis_cs:N #1
  { \l_@@_peek_code_tl { \exp_not:n {#1} } { -1 } 0 }
%    \end{macrocode}
%   For normal characters we must determine their catcode.  The main
%   difficulty is that the character may be an active character
%   masquerading as (i.e., set equal to) itself with a different
%   catcode.  Two approaches based on \tn{lowercase} can detect this.
%   One could make an active character with the same catcode as~|#1| and
%   change its definition before testing the catcode of~|#1|, but in
%   some Unicode engine this fills up the hash table uselessly.
%   Instead, we lowercase~|#1| itself, changing its character code
%   to~$32$, namely space (because \LuaTeX{} cannot turn catcode~$10$
%   characters to anything else than character code~$32$), then we apply
%   \cs{@@_analysis_b_char:Nn}, which detects active characters by
%   comparing them to \cs{tex_undefined:D}, and we must have undefined
%   the active space (locally) for this test to work.
%   To define \cs{@@_peek_analysis_char:N} itself we use an
%   |e|-expanding assignment to get the active space in the right place
%   after making it (just for this definition) unexpandable.
%   Finally \cs{@@_peek_analysis_char:w} receives the \meta{charcode},
%   \meta{user function}, \meta{catcode}, and \meta{token}, and places
%   the arguments in the correct order.  It keeps \cs{exp_not:n} for
%   macro parameter characters and active characters (the latter could
%   be macro parameter characters, and it seems more uniform to always
%   put \cs{exp_not:n}), and otherwise eliminates it by expanding once
%   with \cs{exp_args:NNNo}.
%    \begin{macrocode}
\group_begin:
\char_set_active_eq:NN \ \scan_stop:
\cs_new_protected:Npe \@@_peek_analysis_char:N #1
  {
    \cs_set_eq:NN
      \char_generate:nn { 32 } { 13 }
      \exp_not:N \tex_undefined:D
    \tex_lccode:D `#1 = 32 \exp_stop_f:
    \tex_lowercase:D
      {
        \tl_put_right:Ne \exp_not:N \l_@@_peek_code_tl
          { \exp_not:n { \@@_analysis_b_char:Nn \use_none:n } {#1} }
      }
    \exp_not:n
      {
        \exp_after:wN \@@_peek_analysis_char:w
        \int_value:w
      }
      `#1
    \exp_not:n { \exp_after:wN \s_@@ \l_@@_peek_code_tl }
    #1
  }
\group_end:
\cs_new_protected:Npn \@@_peek_analysis_char:w #1 \s_@@ #2#3#4
  {
    \if_charcode:w 6 #3
    \else:
      \if_charcode:w D #3
      \else:
        \exp_args:NNNo
      \fi:
    \fi:
    #2 { \exp_not:n {#4} } {#1} #3
  }
%    \end{macrocode}
%   For special characters the idea is to eventually act with
%   \cs{token_to_str:N}, then pick up one by one the characters of this
%   string representation until hitting the token that follows.  First
%   determine the character code of (the meaning of) the \meta{token}
%   (which we know is a special token), make sure the escape character
%   is different from it, normalize the meanings of two active
%   characters and the empty control sequence, and filter out these
%   cases in \cs{@@_peek_analysis_retest:}.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_peek_analysis_special:
  {
    \tex_let:D \l_@@_analysis_token = ~ \l_peek_token
    \int_set:Nn \l_@@_peek_charcode_int
      { \@@_analysis_extract_charcode: }
    \if_int_compare:w \l_@@_peek_charcode_int = \tex_escapechar:D
      \int_set:Nn \tex_escapechar:D { `\/ }
    \fi:
    \char_set_active_eq:nN { \l_@@_peek_charcode_int } \scan_stop:
    \char_set_active_eq:nN { \tex_escapechar:D } \scan_stop:
    \cs_set_eq:cN { } \scan_stop:
    \tex_futurelet:D \l_@@_analysis_token
    \@@_peek_analysis_retest:
  }
\cs_new_protected:Npn \@@_peek_analysis_retest:
  {
    \if_meaning:w \l_@@_analysis_token \scan_stop:
      \exp_after:wN \@@_peek_analysis_nonexp:N
    \else:
      \exp_after:wN \@@_peek_analysis_str:
    \fi:
  }
%    \end{macrocode}
%   At this point we know the meaning of the \meta{token} in the input
%   stream is \cs{l_peek_token}, either a space (32, 10) or a
%   begin-group or end-group token (catcode $1$ or~$2$), and we excluded
%   a few cases that would be difficult later (empty control sequence,
%   active character with the same character code as its meaning or as
%   the escape character).  The idea is to apply \cs{token_to_str:N} to
%   the \meta{token} then grab characters (of category code~$12$ except
%   for spaces that have category code~$10$) to reconstruct it.  In
%   earlier versions of the code we would peek at the \meta{next token}
%   that lies after \meta{token} in the input stream, which would help
%   us be more accurate in reconstructing the \meta{token} case in edge
%   cases (mentioned below), but this had the side-effect of tokenizing
%   the input stream (turning characters into tokens) farther ahead than
%   needed.
%
%   We hit the \meta{token} with \cs{token_to_str:N} and start grabbing
%   characters.  More
%   precisely, by looking at the first character in the string
%   representation of the \meta{token} we distinguish three cases:
%   a stringified control sequence starts with the escape character; for
%   an explicit character we find that same character; for an active
%   character we find anything else (we made sure to exclude the case of
%   an active character whose string representation coincides with the
%   other two cases).
%    \begin{macrocode}
\cs_new_protected:Npn \@@_peek_analysis_str:
  {
    \exp_after:wN \tex_futurelet:D
    \exp_after:wN \l_@@_analysis_token
    \exp_after:wN \@@_peek_analysis_str:w
    \token_to_str:N
  }
\cs_new_protected:Npn \@@_peek_analysis_str:w
  { \@@_analysis_char_arg:Nw \@@_peek_analysis_str:n }
\cs_new_protected:Npn \@@_peek_analysis_str:n #1
  {
    \int_case:nnF { `#1 }
      {
        { \l_@@_peek_charcode_int }
          { \@@_peek_analysis_explicit:n {#1} }
        { \tex_escapechar:D } { \@@_peek_analysis_escape: }
      }
      { \@@_peek_analysis_active_str:n {#1} }
  }
%    \end{macrocode}
%   When |#1| is a stringified active character we pass appropriate
%   arguments to the user's code; thankfully \cs{char_generate:nn}
%   can make active characters.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_peek_analysis_active_str:n #1
  {
    \tl_put_right:Ne \l_@@_peek_code_tl
      {
        { \char_generate:nn { `#1 } { 13 } }
        { \int_value:w `#1 }
        \token_to_str:N D
      }
    \l_@@_peek_code_tl
  }
%    \end{macrocode}
%   When |#1| matches the character we had extracted from the meaning of
%   \cs{l_peek_token}, the token was an explicit character, which can be
%   a standard space, or a begin-group or end-group character with some
%   character code.  In the latter two cases we call
%   \cs{char_generate:nn} with suitable arguments and put suitable
%   \cs{if_false:} \cs{fi:} constructions to make the result balanced
%   and such that \texttt{o}-expanding or \texttt{e}/\texttt{x}-expanding gives
%   back a single (unbalanced) begin-group or end-group character.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_peek_analysis_explicit:n #1
  {
    \tl_put_right:Ne \l_@@_peek_code_tl
      {
        \if_meaning:w \l_peek_token \c_space_token
          { ~ } { 32 } \token_to_str:N A
        \else:
          \if_catcode:w \l_peek_token \c_group_begin_token
            {
              \exp_not:N \exp_after:wN
              \char_generate:nn { `#1 } { 1 }
              \exp_not:N \if_false:
              \if_false: { \fi: }
              \exp_not:N \fi:
            }
            { \int_value:w `#1 }
            1
          \else:
            {
              \exp_not:N \if_false:
              { \if_false: } \fi:
              \exp_not:N \fi:
              \char_generate:nn { `#1 } { 2 }
            }
            { \int_value:w `#1 }
            2
          \fi:
        \fi:
      }
    \l_@@_peek_code_tl
  }
%    \end{macrocode}
%   Finally there is the case of a special token whose string
%   representation starts with an escape character, namely the token was
%   a control sequence.  In that case we could have grabbed the token
%   directly as an \texttt{N}-type argument, but of course we couldn't
%   know that until we had run all the various tests including
%   stringifying the token.  We are thus left with the hard work of
%   picking up one by one the characters in the csname (being careful
%   about spaces), until the constructed csname has the expected
%   meaning.  This fails if someone defines a token like
%   \cs[no-index]{bgroup@my} whose string representation starts the same
%   as another token with the same meaning being an implicit character
%   token of category code $1$, $2$, or $10$.
%    \begin{macrocode}
\cs_new_protected:Npn \@@_peek_analysis_escape:
  {
    \tl_clear:N \l_@@_internal_a_tl
    \tex_futurelet:D \l_@@_analysis_token
      \@@_peek_analysis_collect:w
  }
\cs_new_protected:Npn \@@_peek_analysis_collect:w
  { \@@_analysis_char_arg:Nw \@@_peek_analysis_collect:n }
\cs_new_protected:Npn \@@_peek_analysis_collect:n #1
  {
    \tl_put_right:Nn \l_@@_internal_a_tl {#1}
    \@@_peek_analysis_collect_loop:
  }
\cs_new_protected:Npn \@@_peek_analysis_collect_loop:
  {
    \exp_after:wN \if_meaning:w
      \cs:w
      \if_cs_exist:w \l_@@_internal_a_tl \cs_end:
        \l_@@_internal_a_tl
      \else:
        c_one % anything short
      \fi:
      \cs_end:
      \l_peek_token
      \@@_peek_analysis_collect_end:NNNN
    \fi:
    \tex_futurelet:D \l_@@_analysis_token
      \@@_peek_analysis_collect:w
  }
%    \end{macrocode}
%   As in all other cases, end by calling the user code with suitable
%   arguments (here |#1| is \cs{fi:}).
%    \begin{macrocode}
\cs_new_protected:Npn \@@_peek_analysis_collect_end:NNNN #1#2#3#4
  {
    #1
    \tl_put_right:Ne \l_@@_peek_code_tl
      {
        { \exp_not:N \exp_not:n { \exp_not:c { \l_@@_internal_a_tl } } }
        { -1 }
        0
      }
    \l_@@_peek_code_tl
  }
%    \end{macrocode}
% \end{macro}
%
% \subsection{Messages}
%
% \begin{variable}{\c_@@_analysis_show_etc_str}
%   When a control sequence (or active character)
%   and its meaning are too long to fit in one line
%   of the terminal, the end is replaced by this token list.
%    \begin{macrocode}
\tl_const:Ne \c_@@_analysis_show_etc_str % (
  { \token_to_str:N \ETC.) }
%    \end{macrocode}
% \end{variable}
%
%    \begin{macrocode}
\msg_new:nnn { tl } { show-analysis }
  {
    The~token~list~ \tl_if_empty:nF {#1} { #1 ~ }
    \tl_if_empty:nTF {#2}
      { is~empty }
      { contains~the~tokens: #2 }
  }
%    \end{macrocode}
%
%    \begin{macrocode}
%</package>
%    \end{macrocode}
%
% \end{implementation}
%
% \PrintIndex