summaryrefslogtreecommitdiff
path: root/macros/latex/contrib/cweb/cwebbase.tex
blob: dff1d3494b9a6e28b96974bb0e3f06fe9d6b9675 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
% $StyleId: cwebbase.doc,v 1.4 1995/11/07 17:55:19 schrod Exp $
%----------------------------------------------------------------------
% Written by Joachim Schrod <schrod@iti.informatik.th-darmstadt.de>.
% Copyright conditions see below (GPL).

%
% LaTeX macros cwebbase
% support for incorporation of CWEAVE markup in LaTeX documents
%
% [LaTeX in MAKEPROG]
% (history at end)




%%%%
%%%%
%%%% These TeX macros were documented with the documentation system
%%%% MAKEPROG and automatically converted to the current form.
%%%% If you have MAKEPROG available you may transform it back to
%%%% the original input: Remove every occurence of three percents
%%%% and one optional blank from the beginning of a line and remove
%%%% every line which starts with four percents.  The following lex
%%%% program will do this:
%%%%
%%%%	%%
%%%%
%%%%	^%%%\ ?   ;
%%%%	^%%%%.*\n ;
%%%%
%%%%	If you just want to print the documentation you may fetch
%%%% the archive print-makeprog.tar.Z from ftp.th-darmstadt.de (directory
%%%% pub/tex/latex). It contains *all* used styles -- but beware, they
%%%% may not be in a documented form...
%%%%
%%%%
%%% \documentclass{progltx}

%%% \usepackage{cweb-doc}		% document-specific markup
%%% \usepackage{fullpage}


%%% \RCS $StyleRevision: 1.4 $
%%% \RCS $StyleDate: 1995/11/07 17:55:19 $


%%% \begin{document}


%%% \title{Typesetting \cweave{} Output}
%%% \author{%			% LaTeX does not discard unnecessary glue...
%%%     Joachim Schrod%
%%%     \thanks{%
%%% 	\protect\raggedright
%%% 	TU~Darmstadt, Computer Science Department, WG Systems Programming,
%%% 	Alexanderstr.~10, D-64283~Darmstadt, Germany.
%%% 	Email: \texttt{schrod@iti.informatik.th-darmstadt.de}.%
%%% 	}%
%%%     }
%%% \date{%
%%%     Revision \RCSStyleRevision\\
%%%     (as of \RCSStyleDate)%
%%%     }

%%% \maketitle


%%% % doesn't work with progltx yet
%%% %\tableofcontents




%%% \chap Introduction.

%%% This code shall interpret and render the tag set output by \cweave{}.
%%% That's realized in this module as it's needed in multiple places --
%%% when we want to use \LaTeX{} markup in \cweb{}, when we want to
%%% incorporate \cweb{} pieces in \LaTeX{} documents (as examples, or in
%%% textbooks), and when we want to incorporate whole \cweb{} documents in
%%% a \LaTeX{} document. (That's by far the most difficult task.)

%%% As \cweb{} operates in modes, for obvious reasons our implementation
%%% is based on a finite state automaton (FSA). The whole tag set is
%%% separated into subsets that are valid in certain states.

%%% The complete set of tags does not obey any name conventions, some tags
%%% are even in conflict with standard \LaTeX{} tags (e.g.\ |\\|). Most
%%% tags are used to render code pieces. With the introduction of states
%%% we can restrict the visibilities of these tags to those states where
%%% we typeset program parts. The needed functionality will be available
%%% in cseqs of our namespace, we'll map that to \cweave{}'s names in a
%%% group.

%%% Then we have to implement the rendering of chunks and sections, that's
%%% (more or less) a straight-forward task. Actually, we only set up the
%%% basic code pieces, the integration of section divisions in the actual
%%% \LaTeX{} document is a task of higher-level code; that integration
%%% depends heavily on the type of document. That also implies that we
%%% won't implement front or back matter material (document markup, table
%%% of contents, index, list of refinements, etc.)\ in this module.


%%% \sect Let's identify this module.

%%% \noindent The code below is explained in the implementation
%%% documentation of the \pkg{rcs} package.

%%% \beginprog
\begingroup
    \def\RCSFile#1#2 $#3: #4 #5\endRCS $#6: #7 #8\endRCS{%
	\def\date{#4}\def\id{v#7}%
	\ProvidesFile{#1}[\date\space\id\space #2]%
	}

  \RCSFile{cwebbase.tex}{CWEAVE tags for LaTeX markup}
  $StyleDate: 1995/11/07 17:55:19 $: 9999/00/00 \endRCS
  $StyleRevision: 1.4 $: 0.0 \endRCS
\endgroup
%%% \endprog


%%% \sect This module is part of the supported \textsl{cweb-sty} bundle.
%%% Send bug reports, comments and repairs.

%%% It assumes that the canonical setup of my modules has happened and can
%%% be used herein. (In particular, the underscore is a letter.)


%%% \sect This module reserves the namespace |cwbb|. It assumes that
%%% |\cweb_warning| is bound either to |\PackageWarningNoLine| or
%%% |\ClassWarningNoLine|, depending on the environment.

%%% \beginprog
\ifx \cwbb@loaded\undefined
    \def\cwbb@loaded{$StyleRevision: 1.4 $}
\else
    \cweb_warning{cwebbase}%
        {Some other package already uses namespace `cwbb'}
\fi
%%% \endprog


%%% \sect This is freely distributable software; you can redistribute it
%%% and/or modify it under the terms of the GNU General Public License as
%%% published by the Free Software Foundation; either version~2 of the
%%% License, or (at your option) any later version.

%%% This software is distributed in the hope that it will be useful, but
%%% \textbf{without any warranty}; without even the implied warranty of
%%% \textbf{merchantability} or \textbf{fitness for a particular purpose}.
%%% See the GNU General Public License for more details.

%%% You should have received a copy of the GNU General Public License in
%%% the file |License| along with this package; if not, write to the Free
%%% Software Foundation, Inc., 675~Mass Ave, Cambridge, MA~02139,~USA.




%%% % ------------------------------------------------------------
%%% %
%%% % subdocument: The interface between CWEAVE and TeX
%%% %

%%% \input{cweave-spec}

%%% %
%%% % ------------------------------------------------------------




%%% \chap Interface of \pkg{cwebbase} module.

%%% \pkg{cwebbase} implements most of \cweave{} tags. It sets up tons of
%%% redefinable cseqs, all of them start with |Cweb|.
%%% %
%%% \begin{fixme}
%%%   How do I document them? We need a configuration document anyhow. Do
%%%   I present here just a list of those cseqs realized in this module
%%%   and refer to the config docs? Need to wait until I have it, then
%%%   we'll see how the stuff here will be added.
%%% \end{fixme}


%%% \sect This module supports both hierarchic and flat section
%%% structuring. You have to bind |\cweb@structure| to 0 (hierarchic) or 1
%%% (flat) before importing it, to show which structure is wanted.


%%% \sect This module supports suppression of different document parts.
%%% You have to define flags |@cweb@suppress@|\<foo>|@| with
%%% $\mbox{\<foo>} \in \{ |changehints|, |unchanged|, |format| \}$ to
%%% denote if suppression shall happen.


%%% \sect The counter |secnumdepth| must be~11 if chunks shall be
%%% numbered. That counter doesn't need to be set when this module is
%%% imported, it can be set locally during typesetting \cweb{} code.

%%% The definitions for flat structure assume that |secnumdepth| is~11
%%% during typesetting.


%%% \sect |\part|, |\chapter| (if available), and |\section| are used for
%%% \cweb{} section divisions (those with the lowest ranks). These
%%% commands will change the respective counters globally and will
%%% add toc lines and marks. You might want to prevent that by redefining
%%% |\addcontentsline| and |\mark|.

%%% Table of contents macros are not defined for the section divisions
%%% introduced by this module. You must either prevent that they are
%%% written to the |AUX| file at all; or you must define them.
%%% %
%%% \begin{fixme}
%%%   Add explanation of introduced section divisions.
%%% \end{fixme}

%%% For flat structure, all section divisions are of type `|chunk|', like
%%% chunks. The section names start with `|\cwbbRank{|\<rank>|}|'. By
%%% default, |\cwbbRank| takes one argument and expands to the empty list
%%% -- this can be redefined for the table of contents.


%%% \sect The document end, as explained in the \cweave{} interface specs
%%% (chunk~\ref{spec:sec:docend}), is not implemented. That must be done
%%% by the user of this module -- if the back matter is to be processed at
%%% all.




%%% \chap Processing states.

%%% We have to typeset five different categories of material:
%%% Documentation, program pieces---embedded within the documentation and
%%% as large chunks, \TeX{} material within program pieces (i.e., comments
%%% and refinement names), and cross reference information. Since we need
%%% a complete other environment for the program pieces than for the rest
%%% we design ``states'' where we switch to appropriately.
%%% %
%%% \begin{enumerate}

%%% \item A chunk starts in the \textsl{documentation state}.

%%% \item |\B| switches to \textsl{program state}. This can happen in
%%% documentation and program state.

%%% \item While we process the argument of |\PB| we're in \textsl{restricted
%%% program state}; |\PB| may appear in documentation and in \TeX{}
%%% state. Since \TeX{} state can be switched on within restricted
%%% program state, |\PB| can appear within the argument of itself.

%%% \item In the arguments of |\C|, |\SHC|, and |\X| we switch to
%%% \textsl{\TeX{} state}. All these cseqs appear only in (restricted)
%%% program state, their official names are actually different. I.e., only
%%% in (restricted) program state these cseq are bound to the meaning
%%% described here.

%%% \item Cross reference information are attached to most chunks with
%%% refinements. This information is processed in \textsl{CR state}.
%%% After CR state material
%%% comes always the next chunk or the document end, i.e., material in
%%% documentation state.

%%% \end{enumerate}
%%% %
%%% This FSA is illustrated by the diagram in figure~\ref{fig:fsa}.

%%% \begin{figure}

%%% \begin{center}
%%% \chardef\\=`\\
%%% \input{cweb-fsa.latex}
%%% \end{center}

%%% \caption{The processing state's FSA. The automaton starts and ends in
%%% the ``documentation'' state.}
%%% \label{fig:fsa}

%%% \end{figure}

%%% |\cwbb@documentation| will switch to documentation state,
%%% |\cwbb@program| to program state, |\cwbb@Rprogram| to restricted
%%% program state, |\cwbb@tex| to \TeX{} state, and |\cwbb@CR| to CR state.
%%% If we're already in a state, the switch to this state shall be a
%%% permissible null operation.


%%% \sect The basic difference between these states can be named with
%%% two parameters: (1)~The cseq bindings in effect and (2)~the layout
%%% parameters used for paragraph makeup.

%%% In (restricted) program state and in CR state the text is output
%%% under the control of \cweave{}, and tagged by \cweave{}. The used
%%% tags are from a global namespace and should only be in effect during
%%% these states. We call this tag set the \textsl{\cweave{} bindings}. In the
%%% other two states the tags are largely defined by the user, the tag
%%% set is called the \textsl{user bindings}. The switch to another binding
%%% is always done locally, i.e., if we switch from documentation to
%%% restricted program state within a group we don't have to bother about
%%% the restauration of the user binding; it will be done automatically
%%% by \TeX{} at the end of the group. Nevertheless we must be able to
%%% switch from the \cweave{} bindings back to the user bindings which
%%% were in effect when we activated the \cweave{} bindings. This is
%%% needed for the \TeX{} state which is always activated within
%%% (restricted) program state.

%%% The parameters for \textsl{program layout} are really special ones
%%% since they need to support the indentation which shows the program
%%% structure. These parameters are used in program and in \TeX{} state.
%%% The \textsl{document layout} parameters established by the user are
%%% used in the other three states.

%%% The following table shall summarize this. $C$~denotes \cweave{}
%%% bindings, $U$~user bindings, $P$~program layout, and $D$~document
%%% layout. If an entry is empty, its value is not changed on entry in
%%% this state.
%%% %
%%% \begin{center}
%%%   \DeleteShortVerb\|
%%% \begin{tabular}{r|c|c}
%%% \multicolumn1{c|}{\textsc{State}} & \textsc{Binding} & \textsc{Layout} \\[1pt]
%%% \hline
%%% documentation & $U$ & $D$ \\
%%% program & $C$ & $P$ \\
%%% restricted program & $C$ & \\
%%% \TeX{} & $U$ & \\
%%% CR & $C$ & $D$ \\
%%% \end{tabular}
%%%   \MakeShortVerb\|
%%% \end{center}


%%% \sect Since the user bindings and the document layout is defined
%%% initially, we don't have to do anything if it's requested. Only if we
%%% change it, i.e., within |\cwbb@cweave_bindings| and
%%% |\cwbb@program_layout|, we redefine |\cwbb@user_bindings| and
%%% |\cwbb@doc_layout|. If they are eventually executed, they shall
%%% rebind themselves back to |\relax|. This way we can switch to
%%% documentation state as often as we want.

%%% \beginprog
\let\cwbb@user_bindings=\relax
\let\cwbb@doc_layout=\relax

\def\cwbb@documentation{%
    \cwbb@user_bindings
    \cwbb@doc_layout
    }
\def\cwbb@Rprogram{\cwbb@cweave_bindings}

\def\cwbb@program{%
    \cwbb@cweave_bindings
    \cwbb@program_layout
    }
\def\cwbb@tex{\cwbb@user_bindings}

\def\cwbb@CR{%
    \cwbb@cweave_bindings
    \cwbb@doc_layout
    }
%%% \endprog




%%% \chap Saving and restoring control sequences.

%%% We have a lot of cseqs which are defined within the namespace of this
%%% module and which will be used with other names. This usage is in a
%%% controlled environment, namely neither in documentation nor in \TeX{}
%%% state. (I.e., the text processed is tagged by \cweave{}, not by humans;
%%% therefore we have a precise specification of the cseqs we have to
%%% accept.) We cannot work with groups where a cseq is just redefined
%%% and \TeX{} takes care for establishing the old binding again; when we
%%% switch from program state to \TeX{} state all the bindings which were
%%% in effect before the program state get active, i.e., in the
%%% documentation state, must be in effect again. We cannot simply
%%% consider the \TeX{} state as something parallel to program state, it
%%% must be a hierarchical relationship: In the program state values are
%%% set up which must be available after switching back from \TeX{} to
%%% program state.

%%% We save the current binding of a cseq in another cseq, but only if
%%% there exists a binding currently. This is done to save valuable
%%% \TeX{} main memory. (Why don't all those people who talk about
%%% extending \TeX{} not just implement dynamic memory management for the
%%% existing \TeX{}? It would still be \TeX{} and would \emph{really} be
%%% a worthy activity. But not enough fun, I assume\,\dots) Actually, one
%%% can assume that nearly no cseq bindings must be saved at all---the
%%% used names are strange enough. The bindings of |\foo|, i.e., of the
%%% cseq with the name \fbox{\ttfamily foo}, is saved as the cseq with the
%%% name \fbox{\ttfamily cweb\char `\\ foo}, i.e., as
%%% |\csname cweb\string\foo \endcsname|.


%%% \sect The save process is not done statically, but by the macro
%%% |\cwbb@rebind| which interprets a list of tuples $({\it old\_name},
%%% {\it new\_name})$, terminated by the tuple $(|\stop|, |\stop|)$.
%%% Eventually it constructs two new lists, |\cwbb@to_restore| with the
%%% cseqs which had a binding, and |\cwbb@undefined| with the names which
%%% didn't have one.

%%% The saving is actually done by |\cwbb@save_binding|, |\cwbb@rebind|
%%% is responsible for the initialization, |\cwbb@do_rebind| for the
%%% effective rebinding and the tail recursion on the list.

%%% \begin{fixme}
%%%   We could pull the |\next| assignment in the |\else| branch out of
%%%   the loop to get a better performance. Should measure if this is of
%%%   interest.
%%% \end{fixme}

%%% \beginprog
\newtoks\cwbb@undefined
\newtoks\cwbb@to_restore

\def\cwbb@rebind{%
    \cwbb@undefined{}%
    \cwbb@to_restore{}%
    \cwbb@do_rebind
    }
\def\cwbb@do_rebind#1#2{%
    \ifx #1\stop
	\let\next\relax
    \else
	\cwbb@save_binding #2%
	\let #2=#1%
	\let\next\cwbb@do_rebind
    \fi
    \next
    }
%%% \endprog


%%% \sect If the cseq to be saved is undefined, it may just be added to
%%% the ``undefined list.'' Otherwise its binding is saved and it's added
%%% to the ``to-be-restored list.''

%%% \TeX{}nical note: The cseq-name for the saved binding must be created
%%% before the |\let| is executed.

%%% \beginprog
\def\cwbb@save_binding#1{%
    \ifx #1\undefined
	\cwbb@undefined \expandafter{\the\cwbb@undefined #1}%
    \else
	\expandafter\let \csname cweb\string#1\endcsname =#1%
	\cwbb@to_restore \expandafter{\the\cwbb@to_restore #1}%
    \fi
    }
%%% \endprog


%%% \sect The restoration of rebound cseqs is a two-tied activity: All
%%% previously undefined cseqs must be made undefined again, and all saved
%%% cseqs must be restored. Actually, we don't need to reset the two token
%%% lists, but we do it to save space.

%%% Both |\cwbb@undefine| and |\cwbb@restore_binding| iterate over a list
%%% of cseqs terminated by |\stop|.

%%% \begin{fixme}
%%%   And here the |\next| assignment could be prepended to the loop as
%%%   well.
%%% \end{fixme}

%%% \beginprog
\def\cwbb@restore_bindings{%
    \expandafter\cwbb@undefine \the\cwbb@undefined \stop
    \cwbb@undefined{}%
    \expandafter\cwbb@restore_binding \the\cwbb@to_restore \stop
    \cwbb@to_restore{}%
    }

\def\cwbb@undefine#1{%
    \ifx #1\stop
	\let\next\relax
    \else
	\let#1\undefined
	\let\next\cwbb@undefine
    \fi
    \next
    }
%%% \endprog


%%% \sect \TeX{}nical note: As in |\cwbb@save_binding|, the cseq-name for
%%% the saved binding must be created before the |\let|s are executed.

%%% \begin{fixme}
%%%   Another |\next| assignment.
%%% \end{fixme}

%%% \beginprog
\def\cwbb@restore_binding#1{%
    \ifx #1\stop
	\let\next\relax
    \else
	\expandafter\let \expandafter#1\csname cweb\string#1\endcsname
	\expandafter\let \csname cweb\string#1\endcsname \undefined
	\let\next\cwbb@restore_binding
    \fi
    \next
    }
%%% \endprog




%%% \chap Chunks.

%%% All chunks are numbered subsequently. The chunk number is supplied as
%%% part of the markup. In addition, the `changed' tag |\*| may be
%%% appended to the chunk number.

%%% By default, chunks have approximately two picas vertical space in
%%% front.

%%% If we use an hierarchic structure, we output the chunk number at the
%%% refinement. (It's only needed to identify the refinement for cross
%%% references.) But chunks are important structural units and thus should
%%% be marked by typographic means that's more visible than the space
%%% mentioned above, we add a `chunk start marker'.

%%% For a flat structure, we output the chunk number as the chunk start
%%% marker, that is the traditional \WEB{} layout.

%%% In any case, the documentation part shall be run in after the chunk
%%% start marker.

%%% \beginprog
\newskip\CwebChunkPreSkip
	\CwebChunkPreSkip=2pc plus 1pc minus 6pt
%%% \endprog


%%% \sect In \LaTeX{}, structural divisions are best realized by
%%% |\@startsection|, we'll use it also for chunk starts. The chunk start
%%% marker is the section heading. For hierarchic structure, we use a
%%% traditional paragraph sign, used since medieval times to tag new text
%%% units. Our first test showed that the paragraph sign has typically an
%%% underlength, that doesn't look good at the start of a chunk. We raise
%%% the sign by its depth.

%%% For flat structure, we don't have text as the section heading,
%%% but we have to cope for a deficiency of \LaTeX{}: It always leaves a
%%% quad after the section number (by the way, this is not mentioned in
%%% the documentation---you have to look in the macros to figure this
%%% out). So we use a backskip of $\rm -1\,em$ as the chunk start marker
%%% then.

%%% Changed chunks are marked by a changeflag, by default a star lapping
%%% to the right. We need the information if that chunk is changed to
%%% decide if we shall output the changeflag. We demand from the chunk
%%% macros that they record that info in the flag |@cwbb_changed_chunk@|.

%%% Each structural division must be categorized, it belongs to a
%%% \textsl{type}. This type determines also the name of the counter, the
%%% mark handling for this category, and how this section tag is
%%% represented in the table of contents. The type of chunks is `|chunk|'
%%% (trivial, isn't it?).

%%% Furthermore, a structural division in \LaTeX{} has an associated
%%% \textsl{level}. That level is used to determine if that division get
%%% numbered in text or table of contents. For sections, we'll compute the
%%% level from the rank, it will not be larger than the rank.%
%%%     \footnote{Check the next section for a more precise definition of
%%% 	`rank' and `level'.}
%%% Since the highest section rank is~10, we may use 11 as chunk level.

%%% For flat structure, we output the number for all section headings,
%%% including chunks. For hierarchic structure we use the default maximum
%%% level as supplied by our document class (e.g., 3~for \cls{article},
%%% and 2~for \cls{report} or \cls{book}).

%%% As a last point, we have to care already for sections in flat
%%% structure. They will use the |chunk| category as well, we have to
%%% distinguish them in the table of contents. For that, we'll record the
%%% level in the text, it will get written to the table of contents.

%%% \beginprog
\newif\if@cwbb_changed_chunk@
\let\CwebChangeFlag=*
\if@cweb@suppress@changehints@
    \let\cwbb@change_flag=\relax
\else
    \let\cwbb@change_flag=\CwebChangeFlag
\fi
\def\cwbb@lap_change_flag{\rlap{\cwbb@change_flag}}

\ifcase \cweb@structure
    %% hierarchic
    \let\CwebChunkHeadingStyle\relax
    \newbox\cweb@chunk_marker
    \setbox\cweb@chunk_marker=\hbox{\P}	% tradition from medivial times
    \setbox\cweb@chunk_marker=\hbox{%
	\raise\dp\cweb@chunk_marker \box\cweb@chunk_marker
	}
    \def\CwebChunkHeadingMarker{\copy\cweb@chunk_marker}
    \def\CwebChunkHeading#1{%
	\CwebChunkHeadingMarker
	\if@cwbb_changed_chunk@ \cwbb@lap_change_flag \fi
	}
  \or
    %% flat
    \def\cwbbRank#1{}
    \def\CwebChunkHeadingStyle{\bfseries}
    \def\CwebChunkHeading#1{%
	\protect\cwbbRank{#1}%
	\kern -1em
	\if@cwbb_changed_chunk@ \cwbb@lap_change_flag \fi
	}
\fi
%%% \endprog


%%% \sect Chunks are tagged with |\M| and get passed the chunk number as
%%% argument. Before we evaluate |\@startsection|, we'll start the chunk.
%%% That's common code to be used by sections as well, as they also
%%% implicitely start a chunk.

%%% \beginprog
\def\M#1{%
    \cwbb@start_chunk{#1}%
    \@startsection
	{chunk}%		% category, == counter etc.
	{11}%			% level
	{\z@}%			% no indentation for heading
	{\CwebChunkPreSkip}%	% skip above heading
	{\m@ne em}%		% run-in heading, 1em distance to text
	{\normalfont\CwebChunkHeadingStyle}%	% layout
	{\CwebChunkHeading{11}}% % text, level might go to toc file
    }
%%% \endprog


%%% \sect But let's first declare how a chunk number is represented. The
%%% principle rendering is controlled by |\thechunk|, namely, we get
%%% arabic numbers.  This rendering is used both for creation of the
%%% section header (if in flat mode) and for cross references. I.e.,
%%% |\ref| expands into it.

%%% But for flat structure we need different expansions: In the section
%%% header a period shall be appended, a cross reference shall appear
%%% without such a period. We use two different macros: |\thechunk|
%%% creates the number with the period, |\cwbb@refchunk| only the number.
%%% The latter is used for a redefinition of |\@currentlabel|. Of course,
%%% for hierarchic structure, both are the same.

%%% \beginprog
\newcounter{chunk}

\def\cwbb@refchunk{\arabic{chunk}}

\ifcase \cweb@structure
    %% hierarchic
    \let\thechunk=\cwbb@refchunk
  \or
    %% flat
    \def\thechunk{\arabic{chunk}.}
\fi
%%% \endprog


%%% \sect Each chunk is started by |\cwbb@start_chunk|, which does the
%%% common actions. First of all, we switch to documentation state.

%%% Then we store the chunk number in the |chunk| counter. That implies
%%% that the chunk start must not step this counter! We'll take care for
%%% that later by redefining |\refstepcounter|. The actual number is
%%% passed as the argument, including an optional |\*| used to denote
%%% changed chunks. We'll redefine that cseq to update the flag
%%% |@cwbb_changed_chunk@|.

%%% Each chunk must be typeset with an open |\if| since it is finished by
%%% |\fi|. If we shall suppress the output of unchanged chunks, we have
%%% to use |\iffalse| if not |@cwbb_changed_chunk@|, otherwise |\iftrue|.
%%% The flag |@cwbb_print_chunk@| is set up to provide an appropriate
%%% test.

%%% When we typeset a chunk, we evaluate |\cwbb@init_print_chunk|. For
%%% hierarchic structure, that cseq will be used to initialize the
%%% typesetting of the chunk number at program parts.

%%% \beginprog
\if@cweb@suppress@unchanged@
    \def\if@cwbb_print_chunk@{\if@cwbb_changed_chunk@}
    \@gobble\fi				% close \if in case it's skipped
\else
    \let\if@cwbb_print_chunk@=\iftrue
    \@gobble\fi				% same here
\fi

\def\cwbb@start_chunk#1{%
    \cwbb@documentation
    \begingroup
        \def\*{\global\@cwbb_changed_chunk@true}%
        \global\@cwbb_changed_chunk@false
        \global \c@chunk #1\relax	% might expand \*
    \endgroup
    \if@cwbb_print_chunk@
        \cwbb@init_print_chunk
    }
%%% \endprog


%%% \sect |\@startsection| will step the section counter (i.e., |chunk|).
%%% Since we just assigned it in |\cwbb@start_chunk|, that would leave us
%%% with a counter that's one too high. And our output in cross references
%%% may be wrong, as a period is appended in flat structure.

%%% That damage all comes from one macro, |\refstepcounter|; let's
%%% advice it to do the Right Thing for |chunk| counters.

%%% \beginprog
\let\cwbb@refstepcounter=\refstepcounter

\def\cwbb@string@chunk{chunk}
\def\refstepcounter#1{%
    \def\@tempa{#1}%
    \ifx \@tempa\cwbb@string@chunk
        \protected@edef\@currentlabel{\cwbb@refchunk}%
    \else
    	\cwbb@refstepcounter{#1}%
    \fi
    }
%%% \endprog




%%% \chap Sections.

%%% Important sections are started on a new page, normal sections have the
%%% same space as chunks in front. The very first section does not
%%% automatically start on a new page since a title may be in front of it.

%%% Each section division has a \textsl{rank}, that denotes the section
%%% depth as seen by \cweave{}.

%%% For this depth we compute a \textsl{level}, used to denote the section
%%% depth for \LaTeX{}. This section depth is used to decide where
%%% numbering shall stop, or if that entry shall be presented in the table
%%% of contents, etc.

%%% These two numbers are not the same, due to the support of document
%%% classes with chapters.

%%% We consider a section as important if it's rank is smaller than~3,
%%% i.e., then we'll start a new page. This limit is controlled by the
%%% number |\CwebRankNoEject|. The rank is used as the section level
%%% and is therefore available to configure depth of typeset numbering and
%%% appearance in table of contents.

%%% \beginprog
\def\CwebRankNoEject{3}
%%% \endprog


%%% \sect Sections are started with the macro |\N|. It's first parameter
%%% is the section's rank. The second parameter is the chunk number, the
%%% third the title of the section. This last parameter must be terminated
%%% by a dot.

%%% Each section implicitly starts a chunk. As explained above, major
%%% sections may eject the page, and we add our actual section heading
%%% with |\CwebSection|.

%%% We also tell the user that we have reached the next section by telling
%%% him about the current chunk number. The message is output after the
%%% section header is set, it shall be on the correct page. Since the
%%% token list to be output is expanded at shipout time, we must take care
%%% for the immediate expansion ourselves.

%%% \beginprog
\def\N#1#2#3.{%
    \cwbb@start_chunk{#2}%
    \CwebEjectSection{#1}%
    \CwebSection{#1}{#3}%
    \begingroup
        \def\*{}%
        \edef\tmp{*#2}%
        \message \expandafter{\tmp}%
    \endgroup
    }
%%% \endprog


%%% \sect At the start of a section we usually eject the page if our rank
%%% is not too low. But if the user had a title above the first section,
%%% he won't want this page break. So we try to guess it. If we guess
%%% wrong, the user may either add |\newpage| himself or he may redefine
%%% |\CwebEjectSection| in a package.

%%% \beginprog
\def\CwebEjectSection#1{%		% #1 == rank
    \gdef\CwebEjectSection##1{%
	\relax
        \ifnum ##1<\CwebRankNoEject\relax
            \newpage
        \fi
	}%
    }

\if@titlepage
    \CwebEjectSection{}
\fi
%%% \endprog


%%% \sect Now comes a heary piece. We have 11~section ranks, and need 11
%%% different section categories, section counters, etc. For flat
%%% structure, they shall all do the same; for hierarchic structure, they
%%% shall be really different.

%%% \noindent Let's have a look at the hierarchic structure first, it's
%%% more complex.
%%% %
%%% \begin{itemize}

%%% \item We associate section rank~0 (that's `|@**|') with document
%%%   parts, i.e., map it to |\part|. Many smaller programs won't use it.

%%% \item If we are used in a class that has a bound cseq |\chapter|, we
%%%   associate rank~1 (`|@*0|') with it and rank~2 (`|@*1|') with
%%%   |\section|. Otherwise we associate rank~1 with |\section|. These
%%%   tags (|\chapter| and/or |\section|) must be defined because many
%%%   macros that want to create section divisions will use them.

%%% \item All other section ranks are placed below |\section|.
%%%   |\subsection| and other \LaTeX{} style markup is not supported.

%%%   For hierarchic structure, we identify these sections with category
%%%   names `|section|$l$', where $l$ is the respective level, computed
%%%   from the rank. Those section names do not exist yet, we have to
%%%   create them.

%%% \end{itemize}

%%% Now we come to the point why we can't use the rank as the section
%%% level passed to \LaTeX{}: When we assume the level usage of \LaTeX{}'s
%%% standard classes, |\section| is \emph{always} on level~1, independent
%%% of the existence of |\chapter|. The latter is placed on level~0, its
%%% existence only influences the section level of |\part|, which is
%%% either $-1$ or~0.

%%% This implies that |\subsection| in a standard class is always on
%%% level~2, but our equivalent section division is either `|@*0|'
%%% (rank~1) or `|@*1|' (rank~2). I.e., when we have chapters, we use a
%%% level that's one lower than the rank, otherwise we use the rank.

%%% For flat structure, we use type |chunk|. This way we'll use the chunk
%%% numbers, they are already set up.

%%% For flat structure, we will map all ranks to the same section markup.
%%% This markup will also be accessible by |\section| or |\chapter|. They
%%% will be declared to be on a level that is the same as the rank.


%%% \sect The printed representation of a section divison number is simply
%%% the presentation from the next upper level, a full stop, and an arabic
%%% section number. When we compute the name of the previous level we must
%%% pay attention to our first section division created this way. This
%%% division must be placed below |section|, without any following number.

%%% Since registers are a valuable resource in \TeX{}, we have to take
%%% care not to allocate a superfluous one.

%%% \beginprog
\ifcase \cweb@structure
    %% hierarchic
    \newcounter{section2}[section]
    \@namedef{thesection2}{\@nameuse{thesection}.\arabic{section2}}
    \newcounter{section3}[section2]
    \@namedef{thesection3}{\@nameuse{thesection2}.\arabic{section3}}
    \newcounter{section4}[section3]
    \@namedef{thesection4}{\@nameuse{thesection3}.\arabic{section4}}
    \newcounter{section5}[section4]
    \@namedef{thesection5}{\@nameuse{thesection4}.\arabic{section5}}
    \newcounter{section6}[section5]
    \@namedef{thesection6}{\@nameuse{thesection5}.\arabic{section6}}
    \newcounter{section7}[section6]
    \@namedef{thesection7}{\@nameuse{thesection6}.\arabic{section7}}
    \newcounter{section8}[section7]
    \@namedef{thesection8}{\@nameuse{thesection7}.\arabic{section8}}
    \newcounter{section9}[section8]
    \@namedef{thesection9}{\@nameuse{thesection8}.\arabic{section9}}
    \ifx \chapter\undefined
        \newcounter{section10}[section9]
	\@namedef{thesection10}{\@nameuse{thesection9}.\arabic{section10}}
    \fi
\fi
%%% \endprog


%%% \sect Sections are typeset by |\CwebSection|; rank and title are the
%%% arguments.

%%% For hierarchic structure, we select the proper cseq according to rank:
%%% |\part|, |\chapter| if this is a structure with chapters, |\section|,
%%% and |\cwbb@section{|\<rank>|}| for other ranks. For the last cseq we
%%% also have to supply the section category, it's `|section|$l$' where
%%% $l$ is the rank~$r$ if we don't have chapters, $r-1$ otherwise.

%%% That category cannot be computed by |\cwbb@section| itself since we
%%% want to use that macro also for flat structure. But there all sections
%%% are basically chunks, we have to use the |chunk| category there. Since
%%% we want to differentiate them in the table of contents nevertheless,
%%% we record the current rank in the title.

%%% \beginprog
\ifcase \cweb@structure
    %% hierarchic
    \ifx \chapter\undefined
        \def\CwebSection#1#2{%
	    \ifcase #1
	        %% 0
	        \let\next\part
	      \or  %% 1
		\let\next\section
	    \else  %% level = rank
		\def\cwbb@sect_name{section#1}%
		\def\next{\cwbb@section{#1}}%
	    \fi
	    \next{#2}%
	    }
    \else
        \def\CwebSection#1#2{%
	    \ifcase #1
	        %% 0
	        \let\next\part
	      \or  %% 1
		\let\next\chapter
	      \or  %% 2
		\let\next\section
	    \else  %% level = rank - 1
		\count@=#1  \advance \count@\m@ne
		\edef\cwbb@sect_name{section\number\count@}%
		\edef\next{\noexpand\cwbb@section{\number\count@}}%
	    \fi
	    \next{#2}%
	    }
    \fi
  \or
    %% flat
    \def\CwebSection#1#2{%
	\def\cwbb@sect_name{chunk}%
	\cwbb@section{#1}{\protect\cwbbRank{#1}#2}%
	}
    \def\section{%
	\def\cwbb@sect_name{chunk}%
	\cwbb@section{1}%
	}
\fi
%%% \endprog


%%% \sect We use the same layout for all minor sections. That layout can
%%% be changed by ammended by defining |\CwebSect|$r$|Hook| to redefine
%%% |\CwebSectPreSkip| and/or |\CwebSectLayout| locally. (Of couse, it may
%%% be redefined once globally, too.) These are the values to be used for
%%% arguments 4 and~5 of |\@startsection|, pay attention to the skip sign,
%%% telling about indentation of the first paragraph. And don't forget
%%% that the skip is only used for sections on ranks below
%%% |\CwebRankNoEject|, the others start on a new page anyhow.

%%% The default definition sets the section title in bold and leaves the
%%% same space as before chunks. The first paragraph of the following
%%% documentation has no indentation.

%%% Other layout changes (indentation for heading, etc.)\ must be made by
%%% redefining |\cwbb@section|. Or---email me, if you want to do it. Then
%%% I might be enclined to introduce further possibilities.

%%% \beginprog
\newskip\CwebSectPreSkip
	\CwebSectPreSkip = -\CwebChunkPreSkip
\def\CwebSectLayout{\normalfont\bfseries}

\def\cwbb@section#1{%			% #2 (next token) is title
    \csname CwebSect#1Hook\endcsname	% \relax if undefined
    \expandafter\@startsection
	\expandafter{\cwbb@sect_name}%	% category; for counter, etc.
	{#1}%				% level is rank
	\z@				% no indentation for heading
	\CwebSectPreSkip		% skip above, maybe no next par indent
	\medskipamount			% skip below heading
        \CwebSectLayout
    }
%%% \endprog




%%% \chap Marks.

%%% \begin{fixme}
%%%   Since I don't know yet how to handle marks I ignore most of them for
%%%   the time being.

%%%   Have to change this later.
%%% \end{fixme}

%%% \beginprog
\let\chunkmark\@gobble

\count@=2
\@whilenum \count@<11 \do {
    \expandafter\let \csname section\number\count@ mark\endcsname \@gobble
    \advance \count@ by 1
    }
%%% \endprog




%%% \chap Typesetting programs.

%%% Program pieces come in two flavours: as argument of |\PB|, or as
%%% material after |\B|. In the former case we can use a group for the
%%% switch to the restricted program state, the group end will restore the
%%% previous state again. In the latter case we use |\B| for the switch to
%%% program state, the cross reference list or the next chunk will go to
%%% another state. Since we have to do more for the material after |\B|,
%%% we define this cseq later.

%%% \beginprog
\def\PB#1{%
    \begingroup
	\cwbb@Rprogram
	\leavevmode
	#1%
    \endgroup
    }
%%% \endprog


%%% \sect Note, that |\Y| cannot just be |\smallskip|, as in Plain
%%% \cweb{}. We must assert that the current paragraph is ended before the
%%% vertical glue is inserted, and that's not done by the \LaTeX{}
%%% definition of |\smallskip|. In addition we add some negative penalty,
%%% here is a good place for a page break. As the penalty value we use
%%% half the chunk break penalty---of course a chunk start is an even
%%% better place for a page break\,\dots

%%% It might be that |\Y| is the very first token in a chunk. Then the
%%% chunk start marker isn't set already since it is regarded as a run-in
%%% section heading. In that case we don't set the vertical skip but
%%% simply start the chunk with the program part.

%%% \TeX{}nical note: The flag |@noskipsec| may be used to test if we're
%%% immediately after a run-in section heading.

%%% \beginprog
\newcount\CwebProgPenalty
	\CwebProgPenalty=\@secpenalty
	\divide\CwebProgPenalty by 2
\def\Y{%
    \if@noskipsec
    \else
	\par
	\penalty\CwebProgPenalty
	\smallskip
    \fi
    }
%%% \endprog


%%% \sect Yep, let's unfold the ``official'' names of the cseqs used in
%%% program state.

%%% If we turn on the \cweave{} bindings, they might be in effect already,
%%% we don't need to establish them again. We can test
%%% |\cwbb@user_bindings| for this case, it will be redefined then.

%%% \beginprog
\def\cwbb@cweave_bindings{%
    \ifx \cwbb@user_bindings\relax
	\cwbb@rebind
	    % indentation and paragraph layout
	    \cwbb@incr_indent	\1%
	    \cwbb@decr_indent	\2%
	    \cwbb@expr_break	\3%
	    \cwbb@backup	\4%
	    \cwbb@opt_break	\5%
	    \cwbb@break		\6%
	    \cwbb@big_break	\7%
	    \cwbb@noindent	\8%
	    % C/C++ tokens
	    \CwebRel		\?%
	    \CwebAddress	\AND
	    \CwebComplement	\CM
	    \CwebScope		\DC
	    \CwebEquiv		\E
	    \CwebGe		\G
	    \CwebRightShift	\GG
	    \CwebNe		\I
	    \CwebAssign		\K
	    \CwebLeftShift	\LL
	    \CwebMod		\MOD
	    \CwebNull		\NULL
	    \CwebNot		\R
	    \CwebBinOr		\OR
	    \CwebMemberRef	\PA
	    \CwebThis		\this
	    \CwebOr		\V
	    \CwebAnd		\W
	    \CwebXor		\XOR
	    \CwebLe		\Z
	    \CwebPointer	\MG
	    \CwebPointerMemberRef \MGA
	    \CwebDecr		\MM
	    \CwebIncr		\PP
	    % more tokens
	    \CwebId		\\%
	    \CwebIdLetter	\|%
	    \CwebRes		\&%
	    \CwebString		\.%		%% ( ...Emacs...
	    \CwebStringBreak	\)%
	    \CwebNumber		\T
	    \CwebCombinedOp	\MRL
	    % goes to TeX state
	    \CwebComment	\C
	    \CwebCxxComment	\SHC
	    \CwebRefName	\X
	    % CWEB tokens
	    \CwebMacrosHere	\ATH
	    \CwebDefine		\D
	    \cwbb@format	\F
	    \CwebIdCat		\J
	    \CwebVerbString	\vb
	    % cross reference tags
	    \cwbb@change_flag	\*%
	    \CwebCRAlso		\A
	    \CwebCRsAlso	\As
	    \CwebCRCite		\Q
	    \CwebCRsCite	\Qs
	    \CwebCRUse		\U
	    \CwebCRsUse		\Us
	    \CwebCREt		\ET
	    \CwebCRsEt		\ETs
	    % finish the list
	    \stop\stop
	\def\cwbb@user_bindings{%
	    \cwbb@restore_bindings
	    \let\cwbb@user_bindings\relax
	    }%
    \fi
    }
%%% \endprog


%%% \sect Since most of the \cweave{} bindings are simple and tedious
%%% coding, we'll have a look at the program layout next. Between two
%%% paragraphs there must not be any skip, the skip used in document
%%% layout is saved in |\cwbb@save@parskip|. A few other layout
%%% parameters from the document layout must be saved as well.

%%% \beginprog
\newskip\cwbb@save@parskip
\newskip\cwbb@save@rightskip
\newcount\cwbb@save@sem_sfcode
\newcount\cwbb@save@pretolerance
\newcount\cwbb@save@hyphenpenalty
\newcount\cwbb@save@exhyphenpenalty

\def\cwbb@save_doc_layout{%
    \cwbb@save@parskip\parskip
    \cwbb@save@rightskip\rightskip
    \cwbb@save@sem_sfcode\sfcode`\;
    \cwbb@save@pretolerance\pretolerance
    \cwbb@save@hyphenpenalty\hyphenpenalty
    \cwbb@save@exhyphenpenalty\exhyphenpenalty
    }
\def\cwbb@restore_doc_layout{%
    \parskip\cwbb@save@parskip
    \rightskip\cwbb@save@rightskip
    \sfcode`\;=\cwbb@save@sem_sfcode
    \pretolerance\cwbb@save@pretolerance
    \hyphenpenalty\cwbb@save@hyphenpenalty
    \exhyphenpenalty\cwbb@save@exhyphenpenalty
    }
%%% \endprog


%%% \sect When |\cwbb@program_layout| is called for the very first time
%%% we have to set up some variables. This cannot be done in advance
%%% since they depend on values which might be changed by the user in the
%%% preamble. This setup is done in |\cwbb@layout_init|.

%%% If the program layout is already active, we must not switch it on
%%% another time. This can be tested by the current binding of
%%% |\cwbb@doc_layout|.

%%% \beginprog
\def\cwbb@program_layout{%
    \ifx \cwbb@doc_layout\relax
	\cwbb@layout_init
	\cwbb@save_doc_layout
	\def\cwbb@doc_layout{%
	    \cwbb@restore_doc_layout
	    \let\cwbb@doc_layout\relax
	    }%
	% set new values
	\parskip\z@skip
	\rightskip\z@ plus 100\p@ minus 10\p@
	\sfcode`\;=3000
	\pretolerance\@M
	\hyphenpenalty 1000	% strings can be broken this way
	\exhyphenpenalty\@M
    \fi
    }
%%% \endprog


%%% \sect The unit of the basic indentation is stored in
%%% |\CwebIndentUnit|. Continuation lines are indented two units further,
%%% i.e., the initial hanging indentation is 3~units. The current hanging
%%% indentation is kept in |\cwbb@indent|.

%%% We need undiscardable items which can be used as backspaces, of one
%%% and two units, respectively. This is done best by boxes. The
%%% initialization of the boxes must be postponed until the user had the
%%% chance to change |\CwebIndentUnit|, it's done in the program state
%%% initialization.

%%% \beginprog
\newdimen\CwebIndentUnit
	\CwebIndentUnit=1em
\newdimen\cwbb@indent

\newbox\cwbb@bak	% backspace one unit
\newbox\cwbb@bakk	% backspace two units

\def\cwbb@layout_init{%
    \global\setbox\cwbb@bak \hbox to -1\CwebIndentUnit{}%
    \global\setbox\cwbb@bakk \hbox to -2\CwebIndentUnit{}%
    \global\let\cwbb@layout_init\relax
    }
%%% \endprog


%%% \sect Now we can formulate how to start typesetting program pieces: We
%%% switch to program state, set the initial indentation, and add the
%%% basic indentation of the first code line. Then we might add a marker
%%% to this program part; for hierarchic structure we want to typeset the
%%% chunk number in the margin of the chunk's program part. That's not
%%% really possible, 'though: definitions and format requests are also
%%% introduced with |\B|, so the chunk number will happen to be at the
%%% start of the definition part if one exists.

%%% If the code line is the very first text to be typeset in this chunk,
%%% we don't add the basic indentation---we're already indented from the
%%% run-in chunk start marker. But we must not start our chunk with
%%% |\noindent|, the marker would be typeset in the left margin then. (I
%%% consider this as a \LaTeX{} error.) So we just emulate an empty
%%% documentation part, the marker is now set correctly. |\B| can test for
%%% this case: |@noskipsec| is still true then.

%%% \beginprog
\def\B{%
    \cwbb@program
    \cwbb@indent 3\CwebIndentUnit  \hangindent\cwbb@indent
    \ifvmode
	\if@noskipsec
	    \indent		% add an empty documentation part
	\else
	    \noindent \kern\CwebIndentUnit
	\fi
    \fi
    \CwebNumberProgramPart
    }
%%% \endprog


%%% \sect The chunk number shall be added at the first program part. We
%%% have a hook in the chunk start where we can set up
%%% |\CwebNumberProgramPart|. It will redefine itself at first call, to
%%% disable further output on more |\B| occurences in a chunk.

%%% The chunk number should always be in the left margin. If it's in the
%%% right margin, there would be too much horizontal space between the
%%% code start and the actual number. (The first code line is typically
%%% not very long.) \LaTeX{} does not allow to select the margin for
%%% marginpars, therefore we have to typeset the marginal note `by hand'.
%%% As we know that we're in horizontal mode when we're called, |\vadjust|
%%% is used to create the note. The standard \LaTeX{} dimen register for
%%% the distance between marginal notes and text matter is used, of
%%% course.

%%% Note that marginal notes may overwrite the chunk number. As it's very
%%% unlikely that somebody issues marginal notes in program mode (that's
%%% quite difficult and involves ingenious usage of the implementation of
%%% this macros), I won't care for that case.

%%% \begin{fixme}
%%%   The baseline of the marginal note shall be aligned to the baseline
%%%   of the preceding line. That does not happen by default. Currently, I
%%%   assume that the preceding line has a depth of appropriately the
%%%   depth of a~`g'. Experiments showed that this gives good results for
%%%   most \cweb{} documents. But nevertheless one should determine the
%%%   exact depth of the preceding line, to know how much |box0| should be
%%%   raised. If anybody knows how this can be done, please tell me. (I
%%%   know about the route over the output routine, implementing an own
%%%   marginal note category, but this seems too much effort.)
%%% \end{fixme}

%%% For flat structure, we don't add any program part marker and we
%%% don't need to initialize it.

%%% \beginprog
\ifcase \cweb@structure
    %% hierarchic
    \def\cwbb@marginal_chunkno{%
	\vadjust{%
	    \setbox\z@ \hbox to \z@{%
		\hss \vphantom{g}%
		\small \thechunk
		\hskip \marginparsep
		}%
	    \vbox to \z@{%
		\hsize\z@
		\vss
		\noindent \raise \dp\z@ \box\z@
		}%
	    }%
	}
    \def\cwbb@init_print_chunk{%
       \def\CwebNumberProgramPart{%
	   \cwbb@marginal_chunkno
	   \let\CwebNumberProgramPart\relax
	   }%
       }
  \or
    %% flat
    \let\cwbb@init_print_chunk\relax
    \let\CwebNumberProgramPart\relax
\fi
%%% \endprog


%%% \sect If a statement is finished, a new paragraph with the basic
%%% indentation has to be started.

%%% An optional statement break is implemented by a low penalty which
%%% will be selected if the line has to be broken. We assume that the
%%% hanging indentation of the new line is already set correctly, and
%%% have to backup two units to get the basic indentation. It might be
%%% that the line break is not chosen by \TeX{}, we compensate the
%%% backspace for this case. This compensation is discarded at the start
%%% of a new line.

%%% \beginprog
\def\cwbb@break{%	% forced break, between statements
    \ifmmode\else
	\endgraf	% in LaTeX it isn't sure what \par is really...
	\noindent
	\hangindent\cwbb@indent  \kern\cwbb@indent
	\copy\cwbb@bakk	% go back to basic indentation
	\ignorespaces
    \fi
    }
\def\cwbb@opt_break{%	% optional break between statements
    \hfil \penalty\m@ne
    \hfilneg  \kern .5em  \kern 2\CwebIndentUnit   % discarded on line break
    \copy\cwbb@bakk
    \ignorespaces
    }
\def\cwbb@big_break{%	% forced break and a little extra space
    \Y
    \cwbb@break
    }
\def\cwbb@expr_break#1{%	% break with penalty #1 * 10
    \hfil \penalty#10
    \hfilneg		% discarded on line break
    }
%%% \endprog


%%% \sect When we increment the indentation, we must not forget to set
%%% the hanging indentation immediately since |\cwbb@opt_break| relies on
%%% the new value. If a continuation line is needed it will also be
%%% indented one unit more, which is ok since it should be distinguishable
%%% from the next line.
%%% %
%%% \begin{fixme}
%%%   Shouldn't the hanging indentation be set anew at the end of
%%%   |\cwbb@decr_indent| as well?
%%% \end{fixme}

%%% \beginprog
\def\cwbb@incr_indent{%
    \global\advance\cwbb@indent\CwebIndentUnit
    \hangindent\cwbb@indent
    }
\def\cwbb@decr_indent{\global\advance\cwbb@indent -\CwebIndentUnit}
\def\cwbb@backup{\copy\cwbb@bak}
\def\cwbb@noindent{%				% no indentation
    \hskip -\cwbb@indent \hskip 2\CwebIndentUnit
    }
%%% \endprog




%%% \chap Program (C or \C++) tokens.

%%% Since the user might want to change the (somewhat unusual) way some
%%% of the operators are typeset, we supply names for them.

%%% \beginprog
\let\CwebAnd=\land		% logical and, &&
\let\CwebEquiv=\equiv		% equiv sign, for ==
\let\CwebGe=\ge			% greater or equal
\let\CwebLe=\le			% less or equal
\let\CwebNe=\ne			% unequal, !=
\let\CwebNull=\Lambda		% NULL pointer
\let\CwebNot=\lnot		% logical not, !
\let\CwebOr=\lor		% logical or, ||
\let\CwebXor=\oplus		% bitwise exclusive or, ^
%%% \endprog


%%% \sect Some symbols have to be shifted around, to save computation time
%%% we put them in boxes.

%%% \begin{fixme}
%%%   It's not sure if the following symbols still look good in other font
%%%   families than Computer Modern. The movements might need to get
%%%   adapted when more experience is available. They were chosen from
%%%   Plain \cweb{}, where these were explicit movements measured in
%%%   points. I changed the dimension to quads, to achieve at least the
%%%   support for different sizes in the Computer Modern family.
%%% \end{fixme}

%%% \TeX{}nical note: The pointer symbol must not be declared as a math
%%% symbol (e.g., with |\DeclareMathSymbol|). Then \TeX{} would insert an
%%% italic correction behind it, according to rule~17 from appendix~G.
%%% (Please send email if you know of an easy way to prevent that
%%% addition.) So I use the symbol as an accent, but then I have to supply
%%% a nucleus for this math list. That nucleus also determines the total
%%% width of this construct. As the vector accent has a width of
%%% approximately 5\,pt in |cmmi10|, I'm going to use 0.5\,em. (That
%%% heuristic is as good as any other I can think of.)

%%% \beginprog
%% pointer to struct component (`->'), use symbol of \vec accent
\newbox\cwbb@pointer
\setbox\cwbb@pointer=\hbox{%
	    \kern -.2em
	    \lower .3em \hbox{$\vec{\kern .5em}$}%
	    \kern .1em
	    }

\newbox\cwbb@decr		% decrement, --
\setbox\cwbb@decr=\hbox{%
	    \kern .05em
	    \raise .1em \hbox{$\scriptstyle {-}\kern -.1em{-}$}%
	    \kern .05em
	    }
\newbox\cwbb@incr		% increment, ++
\setbox\cwbb@incr=\hbox{%
	    \kern .05em
	    \raise .1em \hbox{$\scriptstyle {+}\kern -.1em{+}$}%
	    \kern .05em
	    }
%%% \endprog


%%% \sect Although the following symbols are typeset like an ``ordinary
%%% \C++ programmer'' would expect them, we provide own module names
%%% nevertheless. They can now be changed as well, i.e., orthogonality is
%%% enhanced. (And we can use |\cwbb@rebind| for assigning them to their
%%% names while we switch to program state\,\dots).

%%% Note: The comment at |\CwebMod| says ``modulo/remainder'', as this is
%%% not defined by the C standard, i.e., implementation-dependent.

%%% \beginprog
\mathchardef\CwebAddress="2026	% `&', as binary op
\let\CwebAssign==		% assignment
\let\CwebBinOr=\mid		% binary or
\def\CwebComplement{{\sim}}	% `~', as ordinary symbol
\def\CwebDecr{\copy\cwbb@decr}	% decrement
\def\CwebIncr{\copy\cwbb@incr}	% increment
\let\CwebLeftShift=\ll		% left shift, <<
\def\CwebMod{\mathbin{\hbox{\footnotesize\rm\%}}}	% modulo/remainder, %
\def\CwebMemberRef{\mathbin{.*}}	% ptr to member (on object)
\def\CwebPointer{\copy\cwbb@pointer}
\def\CwebPointerMemberRef{\mathbin{\CwebPointer*}} % ptr to member (on ptr)
\def\CwebRel{\mathrel?}		% relation operator
\let\CwebRightShift=\gg		% right shift, >>
\def\CwebScope{\kern.1em{::}\kern.1em}	% scope resolution

\def\CwebThis{\CwebRes{this}}	% reserved identifier `this'
%%% \endprog


%%% \sect The special identifier |\TeX| remains. In math mode it shall be
%%% the identifier, in other modes the usual logo. We use the definition
%%% supplied by DEK (or \textsc{Silvio Levy}?)\ from Plain \cweb{}.

%%% \beginprog
\def\TeX{%
    {%
	\ifmmode\it\fi
	\mbox{T\kern-.1667em\lower.424ex\hbox{E}\hskip-.125emX}%
    }}
%%% \endprog


%%% \sect Some tokens don't have constant names, the name is supplied as
%%% the argument.

%%% Identifiers are typeset in italics, reserved words and type names in
%%% boldface, and strings in typewriter. For an underscore in the bold
%%% words we use a line that's 50\,\% thicker as the usual line thickness.
%%% -- But what is the usual line thickness. As Knuth uses 0.4\,pt for a
%%% 10\,pt Computer Modern font, we'll use 0.06\,em for a CM font in an
%%% arbitrary size. If that line thickness is still OK for other fonts is
%%% doubtful, as it should fit to the stem width. (CM is a rather light
%%% font, after all.) But since there is no font dimension that gives a
%%% hint to an appropriate line thickness, it's our best try.

%%% \beginprog
\def\CwebCombinedOp#1{\mathrel{\let\K==#1}}  % e.g., += operator
\def\CwebId#1{\hbox{\it#1\/\kern.05em}} % identifier, more than one char
\def\CwebIdLetter#1{\hbox{$#1$}}	% identifier, one letter
\def\CwebRes#1{%			% reserverd words and type names
    \hbox{\bf
	\def\_{\kern.04em\vbox{\hrule width.3em height.06em}\kern.08em}%
	#1\/\kern .05em
	}%
    }
%%% \endprog


%%% \sect In Plain \cweb{} strings are typeset in the font |cmtex10|,
%%% which is a typewriter font with extended ASCII characters. This font
%%% does not have the usual accents and can therefore not used for
%%% typesetting national characters. (Very often they are input in some
%%% 8-bit encoding, the respective character code is made active and is
%%% substituted by a cseq which expands to the correct glyph. Of course,
%%% there are better ways, but that's the reality we have to cope with.)
%%% Instead of this special font we use the standard typewriter font.
%%% This has the further advantage that |cmtex10| may not be introduced
%%% to NFSS\@.

%%% \beginprog
\def\CwebString{%
    \hbox \bgroup
        \tt
	\cwbb@string_setup
	\cwbb@string
    }
\def\cwbb@string#1{%
	#1\kern .05em
	\egroup
    }
%%% \endprog


%%% \sect Within strings certain cseqs have a special meaning; this is
%%% introduced by |\cwbb@string_setup|. The cseqs within strings are
%%% mostly special \TeX{} characters that are to be typeset by cseqs with
%%% the same name.

%%% But a problem remains: active characters. They will still expand, and
%%% will mess up the string visualization. This is a particular problem
%%% with |german.sty| where the double quote is active. (Of course, that's
%%% a problem -- I'm German. :--) When one typesets strings, a double
%%% quote shall be typeset as double quotes, and must not expand to the
%%% special meaning. Remember that double quotes happen to appear quite
%%% often in strings. As the string argument is already tokenized, one has
%%% to redefine the binding of $(\textit{\textup{symbol}}\ \textbf{.}\
%%% |#\"|)$.

%%% \begin{fixme}
%%%   Perhaps one might use a loop over all characters except the braces,
%%%   and set their catcode to other. That would give us enough security.

%%%   For now, we handle only the double quote.
%%% \end{fixme}

%%% Every once in a while we have discretionary breaks in a string,
%%% denoted by |\CwebStringBreak|. This break is shown with the C
%%% convention of escaped newlines.

%%% \beginprog
\def\cwbb@string_setup{%
    \chardef\ =`\  %		% <-- two spaces !
    \chardef\&=`\&
    \chardef\\=`\\
    \chardef\^=`\^
    \chardef\_=`\_
    \chardef\{=`\{
    \chardef\}=`\}
    \chardef\~=`\~
    \catcode`\"=\CatOther
    }

\def\CwebStringBreak{\discretionary{\hbox{\tt\char`\\}}{}{}}
%%% \endprog


%%% \sect Numbers are typeset in different ways. We use the definition of
%%% Plain \cweb{}.
%%% %
%%% \begin{fixme}
%%%   Should add a specification of the possible input and an explanation
%%%   of the macros below. In particular, that the closing brace after
%%%   |\aftergroup| is used much later is probably not grokked by most
%%%   \TeX{} programmers.
%%% \end{fixme}

%%% \beginprog
\def\cwbb@oct{\hbox{$^\circ$\kern-.1em\it\aftergroup\?\aftergroup}}
\def\cwbb@hex{\hbox{$^{\scriptscriptstyle\#}$\tt\aftergroup}}

\def\CwebNumber#1{%		% octal, hex, or decimal constant
    \hbox{%
	$%
	    \def\?{\kern.2em}%
	    \def\$##1{\egroup\sb{\,\rm##1}\bgroup}% suffix to constant
	    \def\_{\cdot 10^{\aftergroup}}% power of ten (via dirty trick)
	    \let\~\cwbb@oct \let\^\cwbb@hex
	    {#1}%
	$}%
    }
%%% \endprog


%%% \sect Comments are typeset in \TeX{} state. We add a hook, the user
%%% shall be able to change the layout (e.g., he might want another font).
%%% %
%%% \begin{fixme}
%%%   Currently \C++ comments are typeset like C comments. This is
%%%   horrible in usual circumstances, i.e., when complete blocks of text
%%%   are prefixed with |//|. We should simply catenate all these text and
%%%   typeset it as one paragraph, each line prefixed by |//|. But then we
%%%   have to implement an |\everyline| first, and since that's not so
%%%   easy we postpone it\,\dots
%%% \end{fixme}
%%% %
%%% In front of a C comment there is an optional stmt break, with 2\,quad
%%% in front if the line is not broken and 1.5\,quad if the line is broken.

%%% \beginprog
\def\CwebComment#1{%
    \5%				% 0.5em will be discarded on line break
    \hskip 1.5em
    $/\ast\,$%
    {\cwbb@tex
	\CwebCommentHook
	#1%
    }%
    $\,\ast/$%
    }
\let\CwebCommentHook\relax
\let\CwebCxxComment\CwebComment
%%% \endprog





%%% \chap \cweb{} tokens.

%%% We distinguish three categories of \cweb{} tokens: (1) Those which
%%% output constant text, (2) those which have attributes to be displayed
%%% in a special way, and (3) those which start a new structure element,
%%% namely |@d| and~|@f|. Let's consider them in this order.


%%% \sect \cweb{} tokens which expand in a constant string are the
%%% identifier catenation operator~(`|@&|') and the macro placement
%%% directive~(`|@h|').

%%% \beginprog
\def\CwebIdCat{\CwebString{@\&}}
\def\CwebMacrosHere{%
    \begingroup
	\def\CwebRefNumber##1{}%
	\CwebRefName :Preprocessor definitions\X
    \endgroup
    }
%%% \endprog


%%% \sect Verbatim program strings, i.e., strings passed verbatim by
%%% \ctangle{}~(`|@=|') are typeset like normal strings, but within a box.
%%% We use 2\,pt as the separating distance, this is set locally.

%%% \beginprog
\def\CwebVerbString#1{{\fboxsep\tw@\p@ \fbox{\CwebString{#1}}}}
%%% \endprog


%%% \sect The refinement names are typeset in angles, this has a long
%%% tradition. We burry the typesetting of the chunk number (which is
%%% the first parameter) in a macro call; the user may change this to
%%% achieve special effects.
%%% %
%%% \begin{fixme}
%%%   One could think about handling special values of the chunk numbers
%%%   differently. E.g., an empty argument is not typeset at all, a~0
%%%   triggers a marginal note about a missing definition, etc.
%%% \end{fixme}

%%% A refinement name may be typeset both in math and horizontal mode.
%%% The name itself is typeset in horizontal mode, of course; for the
%%% angles we need math mode. Therefore we assert at the start of the
%%% macro that we're not in math mode any more. At the end we switch back
%%% to math mode if we've started in it. This conditional switch from and
%%% back to math mode is done by |\cwbb@toggle_text|.

%%% \TeX{}nical note: If we're in math mode |\cwbb@toggle_text| must be
%%% defined globally, as it will turn off math mode and the definition
%%% would be un-made then. The second invocation of |\cwbb@toggle_text|
%%% would be undefined then.

%%% A refinement may also consist of the file name the expansion of this
%%% refinement shall be written to.
%%% %
%%% \label{sec:refname-dot}
%%% %
%%% This file name is tagged with |\.|, it shall be typeset like a string.
%%% But the user shall be able to use the dot accent in the refinement
%%% name as well. We check if the text consists solely of the tag and its
%%% argument; in this case we substitute |\.| with |\CwebString|.
%%% Otherwise we leave it as it is.  Then a refinement name may not
%%% consist of a single dot-accented expression---well, that's highly
%%% unlikely. (Nevertheless I'll document it in the user's manual\,\dots)

%%% \begin{fixme}
%%%   One should be able to use |\EnsureMath| here. But the refinement
%%%   name must not be set in math mode and must not be set in an hbox;
%%%   the latter would prevent line breaks.
%%% \end{fixme}

%%% \beginprog
\def\CwebRefName#1:#2\X{%
    \ifmmode  \gdef\cwbb@toggle_text{\null$\null}%
    \else  \let\cwbb@toggle_text\relax
    \fi
    \cwbb@toggle_text
    $\langle\,${\cwbb@tex \cwbb@check_dot{#2}\CwebRefNumber{#1}}$\,\rangle$%
    \cwbb@toggle_text
    }
\def\CwebRefNumber#1{%
    {\reset@font \footnotesize \kern .5em#1}%
    }
%%% \endprog


%%% \sect The next macro, |\cwbb@check_dot| is a hassle to implement. We
%%% must check if it's argument consists solely of a `|\.|'~tag, together
%%% with the argument to this tag. Then the `|\.|'-argument shall be
%%% typeset like a string, |\.| itself must be ignored.

%%% The check for |\.| must not evaluate the parameter. Especially |\PB|
%%% must not be evaluated within, it would lead to havoc with the
%%% redefinition of all bindings, etc. (At least that's the empirical
%%% result---it happened; although I don't know where the problems are.
%%% Anyhow, it wasn't in the specs of the rebinding process.) All other
%%% cseqs which shouldn't be used in moving arguments may cause problems
%%% here, too. Note that |\protect| is no solution here, the user often is
%%% not aware of the presence of \cweave{}-generated tags. As an example,
%%% consider the refinement \verb"@< local variables of |foo| @>|" which
%%% will lead to a call with the argument `|local variables of \PB{\\{foo}}|'.
%%% |\\| must not be evaluated here.

%%% I use an approach where I hope that it will work---but to be honest,
%%% I'm not sure. I want to define a macro |\next| which shall have an
%%% empty expansion iff |#1| was a `|\.|'~tag with its argument. Empty
%%% refinement names are not allowed so we can assume that |#1| consists
%%% of at least one token. I evaluate the first token of |#1| before the
%%% |\def| is done. With an appropriate binding a |\.| will disappear,
%%% perhaps leaving an empty token list. I also take care that |\PB| is
%%% not evaluated then, by temporary rebinding it to |\relax|. Thereby I
%%% have taken care for all cseqs \cweave{} might have introduced at the
%%% very front of |#1|. Other cseqs are user-tags. If they cause
%%% problems, the user might circumstance them by adding |\protect|.

%%% \beginprog
\def\cwbb@check_dot#1{%
    \begingroup
	\let\.\@gobble
	\let\PB\relax
	\expandafter\def \expandafter\next \expandafter{#1}%
	\ifx \next\empty
	    \gdef\next##1{\CwebString}%
	\else
	    \global\let\next\relax
	\fi
    \endgroup
    \next #1%
    }
%%% \endprog


%%% \sect \cweb{} macro definitions (done by~|@d|) and format directives are
%%% basically program text. When they are started we are already in
%%% program state, i.e., |\B| has appeared in front of it. But since we have
%%% an introductionary identifier (either |#define| or |format|) to set at
%%% the very front we increase the indentation by three units.

%%% The output of format directives may be suppressed.  Then we have to
%%% gather both identifiers that will appear after |\F|.  Both of them
%%% consist of one group-token (a cseq and the identifier name), after
%%% each identifier one other token separates them. I.e., we have to
%%% gobble six group-tokens. As the very last token is |\par| if a format
%%% directive follows directly on another one, the definition must be
%%% long.

%%% \TeX{}nical note: |\cwbb@format| must expand to |\CwebFormat|, not to
%%% its expansion. Otherwise a redefinition of |\CwebFormat| in a package
%%% would not be used.

%%% \beginprog
\def\CwebDefine{\cwbb@macro{\#define}}
\def\CwebFormat{\cwbb@macro{format}}
\if@cweb@suppress@format@
    \long\def\cwbb@format#1#2#3#4#5#6{}
\else
    \def\cwbb@format{\CwebFormat}
\fi

\def\cwbb@macro#1{%
    \global\advance\cwbb@indent \tw@\CwebIndentUnit
    \cwbb@incr_indent
    \CwebRes{#1 }%		% <-- blank!
    }
%%% \endprog




%%% \chap Refinement cross references and changes.

%%% At the end of chunks with refinements, \cweave{} may output cross
%%% reference information: In which chunks additional definitions for
%%% this refinement are found, where this refinement is used, and where it
%%% is cited. This cross reference information is always introduced by a
%%% tag followed by a non-empty list of chunk numbers. Different tags
%%% are used for lists with only one element and for lists with more than
%%% one element, this way the introductionary text may be adapted.

%%% Within a list of $n$~chunk numbers the first $n-1$~numbers are
%%% separated by commas with one following blank. The last two numbers
%%% are separated by |\ET| if~$n=2$, and by |\ETs| if~$n>2$. If a chunk
%%% is changed by the changefile, a changeflag~(`|\*|') is appended to
%%% its number. The list is eventually terminated by a full stop.


%%% \sect We separate the cross reference information by a smallskip from
%%% the refinement or from a previous cross reference information. The
%%% information itself is typeset in a smaller font, as it is auxilliary,
%%% inserted stuff. The number list has a hanging indentation of
%%% |\CwebNumberListHangindent|. But beware: This isn't a dimension
%%% register, it's a macro. This way one can use ems as dimensions.

%%% The actual size change is represented by a macro. We will redefine it
%%% when we need an other size, e.g., in the lists at the end of the
%%% document.

%%% We must assert that we're in CR state while we set a cross reference.

%%% The first parameter of a cross reference information unit is the
%%% introductionary text, the second is the number list. The parameters
%%% must be evaluated in a group---local parameter changes therein must
%%% not influence the environment.

%%% \beginprog
\def\CwebNumberListHangindent{2em}
\let\CwebCRSize\footnotesize
\def\CwebCrossRef#1#2.{%
    \par\smallskip
    \cwbb@CR
    \begingroup
	\reset@font \CwebCRSize
	\noindent \hangindent\CwebNumberListHangindent\relax
	#1~#2.\par
    \endgroup
    }
%%% \endprog


%%% \sect Well, let's define all introducing tags, which in fact start the
%%% cross reference information unit. The number list is gathered by
%%% |\CwebCrossRef|.

%%% \beginprog
\def\CwebCRAlso{\CwebCrossRef{See also chunk}}
\def\CwebCRsAlso{\CwebCrossRef{See also chunks}}

\def\CwebCRCite{\CwebCrossRef{This code is cited in chunk}}
\def\CwebCRsCite{\CwebCrossRef{This code is cited in chunks}}

\def\CwebCRUse{\CwebCrossRef{This code is used in chunk}}
\def\CwebCRsUse{\CwebCrossRef{This code is used in chunks}}

\def\CwebCREt{ and~}
\def\CwebCRsEt{, and~}
%%% \endprog




%%% \chap Bells and whistles.

%%% \begin{fixme}
%%%   Just copied from Plain \cweb{}. It doesn't work
%%%   anyhow---|\char"|\<xy> usually doesn't typeset the correct
%%%   character. That's because this character is most probably an active
%%%   character; at least that's typical for the way \TeX{} systems are
%%%   used in Europe. I have to think about the implementation.
%%% \end{fixme}

%%% \beginprog
\def\ATL{\par\noindent\bgroup\catcode`\_=12 \postATL} % print @l in limbo
\def\postATL#1 #2 {\bf letter \\{\uppercase{\char"#1}}
   tangles as \tentex "#2"\egroup\par}
\def\noATL#1 #2 {}
\def\noatl{\let\ATL=\noATL} % suppress output from @l
%%% \endprog



%%% \chap The end.

%%% Well, after all we're finished with this module

%%% \beginprog
\endinput
%%% \endprog


%%% 
%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

%%% \vskip \PltxPreSectSkip

%%% \rcsLogIntro{After the actual revision log of this module, we give an
%%%     extract of log messages of the \cls{cweb} class. This module was
%%%     extracted from revision~2.8 of that class, we present to you the
%%%     log entries that concern the code of the new module.}

%%% \begin{rcslog}
%%% $StyleLog: cwebbase.doc,v $
%%% \Revision 1.4 1995/11/07 17:55:19 schrod
%%%     \LaTeX{} \cweb{} should work at least with the last two \LaTeX{}
%%% versions; make it work with the previous-to-last one, version
%%% \mbox{$\langle$1994/12/01$\rangle$}. For that, one has to install the
%%% \mbox{$\langle$1995/06/01$\rangle$} (non-outer) definition of
%%% |\newif|, and |\hb@xt@| must not be used.\\
%%% Problem reported by Laurent Desnogues
%%% \path|<laurent.desnogues@aiguemarine.unice.fr>| and somebody else (XXX
%%% -- add name).

%%% \Revision 1.3 1995/11/07 14:13:12 schrod
%%% Reset section numbers when higher-level section begins.\\
%%% Error reported by Bronne Louis \path|<bronne@montefiore.ulg.ac.be>|.

%%% \Revision 1.2 1995/11/06 10:58:47 schrod
%%%     Support active double quotes in strings. That happens when
%%% |german.sty| is used. Problems with other active characters may still
%%% appear.

%%% \Revision 1.1 1995/09/12 23:02:22 schrod
%%% Moved all code that does the actual typesetting of \cweave{} tags and
%%% is therefore also needed for a |cweb| environment. It's now an own
%%% module named \pkg{cwebbase}.

%%% Chunk markers and chunk marginal numbers don't hang below the
%%% baseline any more.


%%% \item[] \noindent\hrulefill\break
%%%   \textbf{Extract of the revision log for |cweb.doc|:}

%%%   \vskip 2ex

%%% \Revision 2.5 1995/08/29 02:07:26 schrod
%%% Discard dependencies on 10\,pt fonts.

%%% Support suppression of format directives.

%%% \Revision 2.4 1995/08/27 19:31:43 schrod
%%% Make configuration of change flag easier.

%%% Put section title of changed chunks list in marks, and tell the user
%%% that it got typeset.

%%% \Revision 2.3 1995/08/27 17:24:46 schrod
%%% Make usage of baseclass with chapters work.

%%% \Revision 2.2 1995/08/27 13:26:22 schrod
%%% Add possibility to suppress change hints.

%%% \Revision 2.1 1995/08/25 19:11:18 schrod
%%% Hierarchic structures are supported now, in addition to the flat
%%% structure of the beta-test version. One can choose with an option. For
%%% that step, the terminology was cleaned up, too: Chunks are not named
%%% sections any more. (That change involved reimplementation of almost
%%% all the structure and toc stuff.)

%%% The chunk number supplied by \cweave{} is used now, not some computed
%%% number. Change flags are printed, too.

%%% One can suppress output of unchanged sections.

%%% \Revision 1.12  1993/08/10  14:15:43  schrod
%%% New page on main section only if group level $<$ |\cwebSecNoEject|.
%%% Default for the latter is 3.

%%% \Revision 1.11  1993/08/10  11:21:07  schrod
%%% Reference to section number does not render a period after the
%%% number any more.

%%% \Revision 1.10  1993/08/09  20:08:20  schrod
%%% |\cweb@cweave_bindings| is now a no-op if \cweave{} bindings are in
%%% effect already.\\
%%% (Problem reported by Michael M\"uller \path|<mimu@mpi-sb.mpg.de>|.)

%%% \Revision 1.9  1993/08/09  18:05:28  schrod
%%% Left shift operator wasn't defined correctly.\\
%%% (Problem reported by Michael M\"uller \path|<mimu@mpi-sb.mpg.de>|.)

%%% \Revision 1.6  1993/06/15  08:49:23  schrod
%%% |\cweb@check_dot| must not evaluate its argument in an |\edef|, this
%%% causes problems if a |\PB| is within. Now I try hard not to evaluate any
%%% tokens outside of my control.

%%% Can use |\@defpar| for an empty line in |\cweb@has_entries|, don't
%%% need an own macro.

%%% \Revision 1.4  1993/06/14  15:54:18  schrod
%%% Add FSA diagram about processing states. CR state is also switched
%%% to from \TeX{} state (that happens with |\ch| at the document end).

%%% \Revision 1.3  1993/05/13  17:51:21  schrod
%%% Refinements may also be filenames (`|@(|'). Then the complete name
%%% consists of a |\.| macro call, which is handled now.\\
%%% (Problem reported by Michael M\"uller \path|<mimu@mpi-sb.mpg.de>|.)

%%% Made the detection of `|@.|' index entries more robust.

%%% \Revision 1.2  1993/05/12  18:28:59  schrod
%%% Adapted to recent changes of \cweave{} (of April 93):

%%% Main sections have a group level, represented in the table of
%%% contents. This changed the complete implementation of section tags.

%%% New C token cseqs: |\Z| and |\MRL|, implemented as |\CwebLe| and
%%% |\CwebCombinedOp|.

%%% \Revision 1.1  1993/04/09  15:00:37  schrod
%%% Initial revision

%%% \end{rcslog}



%%% \end{document}


%%% 
%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%% Local Variables:
%%% mode: LaTeX
%%% TeX-brace-indent-level: 4
%%% indent-tabs-mode: t
%%% TeX-auto-untabify: nil
%%% TeX-auto-regexp-list: LaTeX-auto-regexp-list
%%% compile-command: "make cwebbase.tex"
%%% End: