summaryrefslogtreecommitdiff
path: root/src/ptx.c
blob: 9c35596fc05160fc717a7e1abea0860268318c50 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
/* Permuted index for GNU, with keywords in their context.
   Copyright (C) 1990, 1991, 1993, 1998-2006 Free Software Foundation, Inc.
   François Pinard <pinard@iro.umontreal.ca>, 1988.

   This program is free 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, or (at your option)
   any later version.

   This program is distributed in the hope that it will be useful, but
   WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or 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
   along with this program; if not, write to the Free Software Foundation,
   Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.

   François Pinard <pinard@iro.umontreal.ca> */

#include <config.h>

#include <stdio.h>
#include <getopt.h>
#include <sys/types.h>
#include "system.h"
#include "argmatch.h"
#include "diacrit.h"
#include "error.h"
#include "quote.h"
#include "quotearg.h"
#include "regex.h"
#include "xstrtol.h"

/* The official name of this program (e.g., no `g' prefix).  */
#define PROGRAM_NAME "ptx"

/* Note to translator: Please translate "F. Pinard" to "François
   Pinard" if "ç" (c-with-cedilla) is available in the
   translation's character set and encoding.  */
#define AUTHORS _("F. Pinard")

/* Number of possible characters in a byte.  */
#define CHAR_SET_SIZE 256

#define ISODIGIT(C) ((C) >= '0' && (C) <= '7')
#define HEXTOBIN(C) ((C) >= 'a' && (C) <= 'f' ? (C)-'a'+10 \
		     : (C) >= 'A' && (C) <= 'F' ? (C)-'A'+10 : (C)-'0')
#define OCTTOBIN(C) ((C) - '0')

/* Debugging the memory allocator.  */

#if WITH_DMALLOC
# define MALLOC_FUNC_CHECK 1
# include <dmalloc.h>
#endif

/* Global definitions.  */

/* FIXME: There are many unchecked integer overflows in this file,
   that will cause this command to misbehave given large inputs or
   options.  Many of the "int" values below should be "size_t" or
   something else like that.  */

/* Reallocation step when swallowing non regular files.  The value is not
   the actual reallocation step, but its base two logarithm.  */
#define SWALLOW_REALLOC_LOG 12

/* Imported from "regex.c".  */
#define Sword 1

/* The name this program was run with. */
char *program_name;

/* Program options.  */

enum Format
{
  UNKNOWN_FORMAT,		/* output format still unknown */
  DUMB_FORMAT,			/* output for a dumb terminal */
  ROFF_FORMAT,			/* output for `troff' or `nroff' */
  TEX_FORMAT			/* output for `TeX' or `LaTeX' */
};

static bool gnu_extensions = true;	/* trigger all GNU extensions */
static bool auto_reference = false;	/* refs are `file_name:line_number:' */
static bool input_reference = false;	/* refs at beginning of input lines */
static bool right_reference = false;	/* output refs after right context  */
static int line_width = 72;	/* output line width in characters */
static int gap_size = 3;	/* number of spaces between output fields */
static const char *truncation_string = "/";
				/* string used to mark line truncations */
static const char *macro_name = "xx";	/* macro name for roff or TeX output */
static enum Format output_format = UNKNOWN_FORMAT;
				/* output format */

static bool ignore_case = false;	/* fold lower to upper for sorting */
static const char *break_file = NULL;	/* name of the `Break characters' file */
static const char *only_file = NULL;	/* name of the `Only words' file */
static const char *ignore_file = NULL;	/* name of the `Ignore words' file */

/* Options that use regular expressions.  */
struct regex_data
{
  /* The original regular expression, as a string.  */
  char const *string;

  /* The compiled regular expression, and its fastmap.  */
  struct re_pattern_buffer pattern;
  char fastmap[UCHAR_MAX + 1];
};

static struct regex_data context_regex;	/* end of context */
static struct regex_data word_regex;	/* keyword */

/* A BLOCK delimit a region in memory of arbitrary size, like the copy of a
   whole file.  A WORD is something smaller, its length should fit in a
   short integer.  A WORD_TABLE may contain several WORDs.  */

typedef struct
  {
    char *start;		/* pointer to beginning of region */
    char *end;			/* pointer to end + 1 of region */
  }
BLOCK;

typedef struct
  {
    char *start;		/* pointer to beginning of region */
    short int size;		/* length of the region */
  }
WORD;

typedef struct
  {
    WORD *start;		/* array of WORDs */
    size_t alloc;		/* allocated length */
    size_t length;		/* number of used entries */
  }
WORD_TABLE;

/* Pattern description tables.  */

/* For each character, provide its folded equivalent.  */
static unsigned char folded_chars[CHAR_SET_SIZE];

/* End of context pattern register indices.  */
static struct re_registers context_regs;

/* Keyword pattern register indices.  */
static struct re_registers word_regs;

/* A word characters fastmap is used only when no word regexp has been
   provided.  A word is then made up of a sequence of one or more characters
   allowed by the fastmap.  Contains !0 if character allowed in word.  Not
   only this is faster in most cases, but it simplifies the implementation
   of the Break files.  */
static char word_fastmap[CHAR_SET_SIZE];

/* Maximum length of any word read.  */
static int maximum_word_length;

/* Maximum width of any reference used.  */
static int reference_max_width;

/* Ignore and Only word tables.  */

static WORD_TABLE ignore_table;	/* table of words to ignore */
static WORD_TABLE only_table;		/* table of words to select */

/* Source text table, and scanning macros.  */

static int number_input_files;	/* number of text input files */
static int total_line_count;	/* total number of lines seen so far */
static const char **input_file_name;	/* array of text input file names */
static int *file_line_count;	/* array of `total_line_count' values at end */

static BLOCK text_buffer;	/* file to study */

/* SKIP_NON_WHITE used only for getting or skipping the reference.  */

#define SKIP_NON_WHITE(cursor, limit) \
  while (cursor < limit && ! isspace (to_uchar (*cursor)))		\
    cursor++

#define SKIP_WHITE(cursor, limit) \
  while (cursor < limit && isspace (to_uchar (*cursor)))		\
    cursor++

#define SKIP_WHITE_BACKWARDS(cursor, start) \
  while (cursor > start && isspace (to_uchar (cursor[-1])))		\
    cursor--

#define SKIP_SOMETHING(cursor, limit) \
  if (word_regex.string)						\
    {									\
      regoff_t count;							\
      count = re_match (&word_regex.pattern, cursor, limit - cursor, 0, NULL); \
      if (count == -2)							\
        matcher_error ();						\
      cursor += count == -1 ? 1 : count;				\
    }									\
  else if (word_fastmap[to_uchar (*cursor)])				\
    while (cursor < limit && word_fastmap[to_uchar (*cursor)])		\
      cursor++;								\
  else									\
    cursor++

/* Occurrences table.

   The `keyword' pointer provides the central word, which is surrounded
   by a left context and a right context.  The `keyword' and `length'
   field allow full 8-bit characters keys, even including NULs.  At other
   places in this program, the name `keyafter' refers to the keyword
   followed by its right context.

   The left context does not extend, towards the beginning of the file,
   further than a distance given by the `left' value.  This value is
   relative to the keyword beginning, it is usually negative.  This
   insures that, except for white space, we will never have to backward
   scan the source text, when it is time to generate the final output
   lines.

   The right context, indirectly attainable through the keyword end, does
   not extend, towards the end of the file, further than a distance given
   by the `right' value.  This value is relative to the keyword
   beginning, it is usually positive.

   When automatic references are used, the `reference' value is the
   overall line number in all input files read so far, in this case, it
   is of type (int).  When input references are used, the `reference'
   value indicates the distance between the keyword beginning and the
   start of the reference field, it is of type (DELTA) and usually
   negative.  */

typedef short int DELTA;	/* to hold displacement within one context */

typedef struct
  {
    WORD key;			/* description of the keyword */
    DELTA left;			/* distance to left context start */
    DELTA right;		/* distance to right context end */
    int reference;		/* reference descriptor */
  }
OCCURS;

/* The various OCCURS tables are indexed by the language.  But the time
   being, there is no such multiple language support.  */

static OCCURS *occurs_table[1];	/* all words retained from the read text */
static size_t occurs_alloc[1];	/* allocated size of occurs_table */
static size_t number_of_occurs[1]; /* number of used slots in occurs_table */


/* Communication among output routines.  */

/* Indicate if special output processing is requested for each character.  */
static char edited_flag[CHAR_SET_SIZE];

static int half_line_width;	/* half of line width, reference excluded */
static int before_max_width;	/* maximum width of before field */
static int keyafter_max_width;	/* maximum width of keyword-and-after field */
static int truncation_string_length;/* length of string used to flag truncation */

/* When context is limited by lines, wraparound may happen on final output:
   the `head' pointer gives access to some supplementary left context which
   will be seen at the end of the output line, the `tail' pointer gives
   access to some supplementary right context which will be seen at the
   beginning of the output line. */

static BLOCK tail;		/* tail field */
static int tail_truncation;	/* flag truncation after the tail field */

static BLOCK before;		/* before field */
static int before_truncation;	/* flag truncation before the before field */

static BLOCK keyafter;		/* keyword-and-after field */
static int keyafter_truncation;	/* flag truncation after the keyafter field */

static BLOCK head;		/* head field */
static int head_truncation;	/* flag truncation before the head field */

static BLOCK reference;		/* reference field for input reference mode */

/* Miscellaneous routines.  */

/* Diagnose an error in the regular expression matcher.  Then exit.  */

static void ATTRIBUTE_NORETURN
matcher_error (void)
{
  error (0, errno, _("error in regular expression matcher"));
  exit (EXIT_FAILURE);
}

/*------------------------------------------------------.
| Duplicate string STRING, while evaluating \-escapes.  |
`------------------------------------------------------*/

/* Loosely adapted from GNU sh-utils printf.c code.  */

static char *
copy_unescaped_string (const char *string)
{
  char *result;			/* allocated result */
  char *cursor;			/* cursor in result */
  int value;			/* value of \nnn escape */
  int length;			/* length of \nnn escape */

  result = xmalloc (strlen (string) + 1);
  cursor = result;

  while (*string)
    if (*string == '\\')
      {
	string++;
	switch (*string)
	  {
	  case 'x':		/* \xhhh escape, 3 chars maximum */
	    value = 0;
	    for (length = 0, string++;
		 length < 3 && isxdigit (to_uchar (*string));
		 length++, string++)
	      value = value * 16 + HEXTOBIN (*string);
	    if (length == 0)
	      {
		*cursor++ = '\\';
		*cursor++ = 'x';
	      }
	    else
	      *cursor++ = value;
	    break;

	  case '0':		/* \0ooo escape, 3 chars maximum */
	    value = 0;
	    for (length = 0, string++;
		 length < 3 && ISODIGIT (*string);
		 length++, string++)
	      value = value * 8 + OCTTOBIN (*string);
	    *cursor++ = value;
	    break;

	  case 'a':		/* alert */
#if __STDC__
	    *cursor++ = '\a';
#else
	    *cursor++ = 7;
#endif
	    string++;
	    break;

	  case 'b':		/* backspace */
	    *cursor++ = '\b';
	    string++;
	    break;

	  case 'c':		/* cancel the rest of the output */
	    while (*string)
	      string++;
	    break;

	  case 'f':		/* form feed */
	    *cursor++ = '\f';
	    string++;
	    break;

	  case 'n':		/* new line */
	    *cursor++ = '\n';
	    string++;
	    break;

	  case 'r':		/* carriage return */
	    *cursor++ = '\r';
	    string++;
	    break;

	  case 't':		/* horizontal tab */
	    *cursor++ = '\t';
	    string++;
	    break;

	  case 'v':		/* vertical tab */
#if __STDC__
	    *cursor++ = '\v';
#else
	    *cursor++ = 11;
#endif
	    string++;
	    break;

	  default:
	    *cursor++ = '\\';
	    *cursor++ = *string++;
	    break;
	  }
      }
    else
      *cursor++ = *string++;

  *cursor = '\0';
  return result;
}

/*--------------------------------------------------------------------------.
| Compile the regex represented by REGEX, diagnose and abort if any error.  |
`--------------------------------------------------------------------------*/

static void
compile_regex (struct regex_data *regex)
{
  struct re_pattern_buffer *pattern = &regex->pattern;
  char const *string = regex->string;
  char const *message;

  pattern->buffer = NULL;
  pattern->allocated = 0;
  pattern->fastmap = regex->fastmap;
  pattern->translate = ignore_case ? folded_chars : NULL;

  message = re_compile_pattern (string, strlen (string), pattern);
  if (message)
    error (EXIT_FAILURE, 0, _("%s (for regexp %s)"), message, quote (string));

  /* The fastmap should be compiled before `re_match'.  The following
     call is not mandatory, because `re_search' is always called sooner,
     and it compiles the fastmap if this has not been done yet.  */

  re_compile_fastmap (pattern);
}

/*------------------------------------------------------------------------.
| This will initialize various tables for pattern match and compiles some |
| regexps.								  |
`------------------------------------------------------------------------*/

static void
initialize_regex (void)
{
  int character;		/* character value */

  /* Initialize the case folding table.  */

  if (ignore_case)
    for (character = 0; character < CHAR_SET_SIZE; character++)
      folded_chars[character] = toupper (character);

  /* Unless the user already provided a description of the end of line or
     end of sentence sequence, select an end of line sequence to compile.
     If the user provided an empty definition, thus disabling end of line
     or sentence feature, make it NULL to speed up tests.  If GNU
     extensions are enabled, use end of sentence like in GNU emacs.  If
     disabled, use end of lines.  */

  if (context_regex.string)
    {
      if (!*context_regex.string)
	context_regex.string = NULL;
    }
  else if (gnu_extensions & !input_reference)
    context_regex.string = "[.?!][]\"')}]*\\($\\|\t\\|  \\)[ \t\n]*";
  else
    context_regex.string = "\n";

  if (context_regex.string)
    compile_regex (&context_regex);

  /* If the user has already provided a non-empty regexp to describe
     words, compile it.  Else, unless this has already been done through
     a user provided Break character file, construct a fastmap of
     characters that may appear in a word.  If GNU extensions enabled,
     include only letters of the underlying character set.  If disabled,
     include almost everything, even punctuations; stop only on white
     space.  */

  if (word_regex.string)
    compile_regex (&word_regex);
  else if (!break_file)
    {
      if (gnu_extensions)
	{

	  /* Simulate \w+.  */

	  for (character = 0; character < CHAR_SET_SIZE; character++)
	    word_fastmap[character] = !! isalpha (character);
	}
      else
	{

	  /* Simulate [^ \t\n]+.  */

	  memset (word_fastmap, 1, CHAR_SET_SIZE);
	  word_fastmap[' '] = 0;
	  word_fastmap['\t'] = 0;
	  word_fastmap['\n'] = 0;
	}
    }
}

/*------------------------------------------------------------------------.
| This routine will attempt to swallow a whole file name FILE_NAME into a |
| contiguous region of memory and return a description of it into BLOCK.  |
| Standard input is assumed whenever FILE_NAME is NULL, empty or "-".	  |
|									  |
| Previously, in some cases, white space compression was attempted while  |
| inputting text.  This was defeating some regexps like default end of	  |
| sentence, which checks for two consecutive spaces.  If white space	  |
| compression is ever reinstated, it should be in output routines.	  |
`------------------------------------------------------------------------*/

static void
swallow_file_in_memory (const char *file_name, BLOCK *block)
{
  int file_handle;		/* file descriptor number */
  struct stat stat_block;	/* stat block for file */
  size_t allocated_length;	/* allocated length of memory buffer */
  size_t used_length;		/* used length in memory buffer */
  int read_length;		/* number of character gotten on last read */

  /* As special cases, a file name which is NULL or "-" indicates standard
     input, which is already opened.  In all other cases, open the file from
     its name.  */
  bool using_stdin = !file_name || !*file_name || STREQ (file_name, "-");
  if (using_stdin)
    file_handle = STDIN_FILENO;
  else
    if ((file_handle = open (file_name, O_RDONLY)) < 0)
      error (EXIT_FAILURE, errno, "%s", file_name);

  /* If the file is a plain, regular file, allocate the memory buffer all at
     once and swallow the file in one blow.  In other cases, read the file
     repeatedly in smaller chunks until we have it all, reallocating memory
     once in a while, as we go.  */

  if (fstat (file_handle, &stat_block) < 0)
    error (EXIT_FAILURE, errno, "%s", file_name);

  if (S_ISREG (stat_block.st_mode))
    {
      size_t in_memory_size;

      block->start = xmalloc ((size_t) stat_block.st_size);

      if ((in_memory_size = read (file_handle,
				  block->start, (size_t) stat_block.st_size))
	  != stat_block.st_size)
	{
#if MSDOS
	  /* On MSDOS, in memory size may be smaller than the file
	     size, because of end of line conversions.  But it can
	     never be smaller than half the file size, because the
	     minimum is when all lines are empty and terminated by
	     CR+LF.  */
	  if (in_memory_size != (size_t)-1
	      && in_memory_size >= stat_block.st_size / 2)
	    block->start = xrealloc (block->start, in_memory_size);
	  else
#endif /* not MSDOS */

	    error (EXIT_FAILURE, errno, "%s", file_name);
	}
      block->end = block->start + in_memory_size;
    }
  else
    {
      block->start = xmalloc ((size_t) 1 << SWALLOW_REALLOC_LOG);
      used_length = 0;
      allocated_length = (1 << SWALLOW_REALLOC_LOG);

      while (read_length = read (file_handle,
				 block->start + used_length,
				 allocated_length - used_length),
	     read_length > 0)
	{
	  used_length += read_length;
	  if (used_length == allocated_length)
	    {
	      allocated_length += (1 << SWALLOW_REALLOC_LOG);
	      block->start
		= xrealloc (block->start, allocated_length);
	    }
	}

      if (read_length < 0)
	error (EXIT_FAILURE, errno, "%s", file_name);

      block->end = block->start + used_length;
    }

  /* Close the file, but only if it was not the standard input.  */

  if (! using_stdin && close (file_handle) != 0)
    error (EXIT_FAILURE, errno, "%s", file_name);
}

/* Sort and search routines.  */

/*--------------------------------------------------------------------------.
| Compare two words, FIRST and SECOND, and return 0 if they are identical.  |
| Return less than 0 if the first word goes before the second; return	    |
| greater than 0 if the first word goes after the second.		    |
|									    |
| If a word is indeed a prefix of the other, the shorter should go first.   |
`--------------------------------------------------------------------------*/

static int
compare_words (const void *void_first, const void *void_second)
{
#define first ((const WORD *) void_first)
#define second ((const WORD *) void_second)
  int length;			/* minimum of two lengths */
  int counter;			/* cursor in words */
  int value;			/* value of comparison */

  length = first->size < second->size ? first->size : second->size;

  if (ignore_case)
    {
      for (counter = 0; counter < length; counter++)
	{
	  value = (folded_chars [to_uchar (first->start[counter])]
		   - folded_chars [to_uchar (second->start[counter])]);
	  if (value != 0)
	    return value;
	}
    }
  else
    {
      for (counter = 0; counter < length; counter++)
	{
	  value = (to_uchar (first->start[counter])
		   - to_uchar (second->start[counter]));
	  if (value != 0)
	    return value;
	}
    }

  return first->size - second->size;
#undef first
#undef second
}

/*-----------------------------------------------------------------------.
| Decides which of two OCCURS, FIRST or SECOND, should lexicographically |
| go first.  In case of a tie, preserve the original order through a	 |
| pointer comparison.							 |
`-----------------------------------------------------------------------*/

static int
compare_occurs (const void *void_first, const void *void_second)
{
#define first ((const OCCURS *) void_first)
#define second ((const OCCURS *) void_second)
  int value;

  value = compare_words (&first->key, &second->key);
  return value == 0 ? first->key.start - second->key.start : value;
#undef first
#undef second
}

/*------------------------------------------------------------.
| Return !0 if WORD appears in TABLE.  Uses a binary search.  |
`------------------------------------------------------------*/

static int
search_table (WORD *word, WORD_TABLE *table)
{
  int lowest;			/* current lowest possible index */
  int highest;			/* current highest possible index */
  int middle;			/* current middle index */
  int value;			/* value from last comparison */

  lowest = 0;
  highest = table->length - 1;
  while (lowest <= highest)
    {
      middle = (lowest + highest) / 2;
      value = compare_words (word, table->start + middle);
      if (value < 0)
	highest = middle - 1;
      else if (value > 0)
	lowest = middle + 1;
      else
	return 1;
    }
  return 0;
}

/*---------------------------------------------------------------------.
| Sort the whole occurs table in memory.  Presumably, `qsort' does not |
| take intermediate copies or table elements, so the sort will be      |
| stabilized throughout the comparison routine.			       |
`---------------------------------------------------------------------*/

static void
sort_found_occurs (void)
{

  /* Only one language for the time being.  */

  qsort (occurs_table[0], number_of_occurs[0], sizeof **occurs_table,
	 compare_occurs);
}

/* Parameter files reading routines.  */

/*----------------------------------------------------------------------.
| Read a file named FILE_NAME, containing a set of break characters.    |
| Build a content to the array word_fastmap in which all characters are |
| allowed except those found in the file.  Characters may be repeated.  |
`----------------------------------------------------------------------*/

static void
digest_break_file (const char *file_name)
{
  BLOCK file_contents;		/* to receive a copy of the file */
  char *cursor;			/* cursor in file copy */

  swallow_file_in_memory (file_name, &file_contents);

  /* Make the fastmap and record the file contents in it.  */

  memset (word_fastmap, 1, CHAR_SET_SIZE);
  for (cursor = file_contents.start; cursor < file_contents.end; cursor++)
    word_fastmap[to_uchar (*cursor)] = 0;

  if (!gnu_extensions)
    {

      /* If GNU extensions are enabled, the only way to avoid newline as
	 a break character is to write all the break characters in the
	 file with no newline at all, not even at the end of the file.
	 If disabled, spaces, tabs and newlines are always considered as
	 break characters even if not included in the break file.  */

      word_fastmap[' '] = 0;
      word_fastmap['\t'] = 0;
      word_fastmap['\n'] = 0;
    }

  /* Return the space of the file, which is no more required.  */

  free (file_contents.start);
}

/*-----------------------------------------------------------------------.
| Read a file named FILE_NAME, containing one word per line, then	 |
| construct in TABLE a table of WORD descriptors for them.  The routine	 |
| swallows the whole file in memory; this is at the expense of space	 |
| needed for newlines, which are useless; however, the reading is fast.	 |
`-----------------------------------------------------------------------*/

static void
digest_word_file (const char *file_name, WORD_TABLE *table)
{
  BLOCK file_contents;		/* to receive a copy of the file */
  char *cursor;			/* cursor in file copy */
  char *word_start;		/* start of the current word */

  swallow_file_in_memory (file_name, &file_contents);

  table->start = NULL;
  table->alloc = 0;
  table->length = 0;

  /* Read the whole file.  */

  cursor = file_contents.start;
  while (cursor < file_contents.end)
    {

      /* Read one line, and save the word in contains.  */

      word_start = cursor;
      while (cursor < file_contents.end && *cursor != '\n')
	cursor++;

      /* Record the word in table if it is not empty.  */

      if (cursor > word_start)
	{
	  if (table->length == table->alloc)
	    {
	      if ((SIZE_MAX / sizeof *table->start - 1) / 2 < table->alloc)
		xalloc_die ();
	      table->alloc = table->alloc * 2 + 1;
	      table->start = xrealloc (table->start,
				       table->alloc * sizeof *table->start);
	    }

	  table->start[table->length].start = word_start;
	  table->start[table->length].size = cursor - word_start;
	  table->length++;
	}

      /* This test allows for an incomplete line at end of file.  */

      if (cursor < file_contents.end)
	cursor++;
    }

  /* Finally, sort all the words read.  */

  qsort (table->start, table->length, sizeof table->start[0], compare_words);
}

/* Keyword recognition and selection.  */

/*----------------------------------------------------------------------.
| For each keyword in the source text, constructs an OCCURS structure.  |
`----------------------------------------------------------------------*/

static void
find_occurs_in_text (void)
{
  char *cursor;			/* for scanning the source text */
  char *scan;			/* for scanning the source text also */
  char *line_start;		/* start of the current input line */
  char *line_scan;		/* newlines scanned until this point */
  int reference_length;		/* length of reference in input mode */
  WORD possible_key;		/* possible key, to ease searches */
  OCCURS *occurs_cursor;	/* current OCCURS under construction */

  char *context_start;		/* start of left context */
  char *context_end;		/* end of right context */
  char *word_start;		/* start of word */
  char *word_end;		/* end of word */
  char *next_context_start;	/* next start of left context */

  /* reference_length is always used within `if (input_reference)'.
     However, GNU C diagnoses that it may be used uninitialized.  The
     following assignment is merely to shut it up.  */

  reference_length = 0;

  /* Tracking where lines start is helpful for reference processing.  In
     auto reference mode, this allows counting lines.  In input reference
     mode, this permits finding the beginning of the references.

     The first line begins with the file, skip immediately this very first
     reference in input reference mode, to help further rejection any word
     found inside it.  Also, unconditionally assigning these variable has
     the happy effect of shutting up lint.  */

  line_start = text_buffer.start;
  line_scan = line_start;
  if (input_reference)
    {
      SKIP_NON_WHITE (line_scan, text_buffer.end);
      reference_length = line_scan - line_start;
      SKIP_WHITE (line_scan, text_buffer.end);
    }

  /* Process the whole buffer, one line or one sentence at a time.  */

  for (cursor = text_buffer.start;
       cursor < text_buffer.end;
       cursor = next_context_start)
    {

      /* `context_start' gets initialized before the processing of each
	 line, or once for the whole buffer if no end of line or sentence
	 sequence separator.  */

      context_start = cursor;

      /* If a end of line or end of sentence sequence is defined and
	 non-empty, `next_context_start' will be recomputed to be the end of
	 each line or sentence, before each one is processed.  If no such
	 sequence, then `next_context_start' is set at the end of the whole
	 buffer, which is then considered to be a single line or sentence.
	 This test also accounts for the case of an incomplete line or
	 sentence at the end of the buffer.  */

      next_context_start = text_buffer.end;
      if (context_regex.string)
	switch (re_search (&context_regex.pattern, cursor,
			   text_buffer.end - cursor,
			   0, text_buffer.end - cursor, &context_regs))
	  {
	  case -2:
	    matcher_error ();

	  case -1:
	    break;

	  default:
	    next_context_start = cursor + context_regs.end[0];
	    break;
	  }

      /* Include the separator into the right context, but not any suffix
	 white space in this separator; this insures it will be seen in
	 output and will not take more space than necessary.  */

      context_end = next_context_start;
      SKIP_WHITE_BACKWARDS (context_end, context_start);

      /* Read and process a single input line or sentence, one word at a
	 time.  */

      while (1)
	{
	  if (word_regex.string)

	    /* If a word regexp has been compiled, use it to skip at the
	       beginning of the next word.  If there is no such word, exit
	       the loop.  */

	    {
	      regoff_t r = re_search (&word_regex.pattern, cursor,
				      context_end - cursor,
				      0, context_end - cursor, &word_regs);
	      if (r == -2)
		matcher_error ();
	      if (r == -1)
		break;
	      word_start = cursor + word_regs.start[0];
	      word_end = cursor + word_regs.end[0];
	    }
	  else

	    /* Avoid re_search and use the fastmap to skip to the
	       beginning of the next word.  If there is no more word in
	       the buffer, exit the loop.  */

	    {
	      scan = cursor;
	      while (scan < context_end
		     && !word_fastmap[to_uchar (*scan)])
		scan++;

	      if (scan == context_end)
		break;

	      word_start = scan;

	      while (scan < context_end
		     && word_fastmap[to_uchar (*scan)])
		scan++;

	      word_end = scan;
	    }

	  /* Skip right to the beginning of the found word.  */

	  cursor = word_start;

	  /* Skip any zero length word.  Just advance a single position,
	     then go fetch the next word.  */

	  if (word_end == word_start)
	    {
	      cursor++;
	      continue;
	    }

	  /* This is a genuine, non empty word, so save it as a possible
	     key.  Then skip over it.  Also, maintain the maximum length of
	     all words read so far.  It is mandatory to take the maximum
	     length of all words in the file, without considering if they
	     are actually kept or rejected, because backward jumps at output
	     generation time may fall in *any* word.  */

	  possible_key.start = cursor;
	  possible_key.size = word_end - word_start;
	  cursor += possible_key.size;

	  if (possible_key.size > maximum_word_length)
	    maximum_word_length = possible_key.size;

	  /* In input reference mode, update `line_start' from its previous
	     value.  Count the lines just in case auto reference mode is
	     also selected. If it happens that the word just matched is
	     indeed part of a reference; just ignore it.  */

	  if (input_reference)
	    {
	      while (line_scan < possible_key.start)
		if (*line_scan == '\n')
		  {
		    total_line_count++;
		    line_scan++;
		    line_start = line_scan;
		    SKIP_NON_WHITE (line_scan, text_buffer.end);
		    reference_length = line_scan - line_start;
		  }
		else
		  line_scan++;
	      if (line_scan > possible_key.start)
		continue;
	    }

	  /* Ignore the word if an `Ignore words' table exists and if it is
	     part of it.  Also ignore the word if an `Only words' table and
	     if it is *not* part of it.

	     It is allowed that both tables be used at once, even if this
	     may look strange for now.  Just ignore a word that would appear
	     in both.  If regexps are eventually implemented for these
	     tables, the Ignore table could then reject words that would
	     have been previously accepted by the Only table.  */

	  if (ignore_file && search_table (&possible_key, &ignore_table))
	    continue;
	  if (only_file && !search_table (&possible_key, &only_table))
	    continue;

	  /* A non-empty word has been found.  First of all, insure
	     proper allocation of the next OCCURS, and make a pointer to
	     where it will be constructed.  */

	  if (number_of_occurs[0] == occurs_alloc[0])
	    {
	      if ((SIZE_MAX / sizeof *occurs_table[0] - 1) / 2
		  < occurs_alloc[0])
		xalloc_die ();
	      occurs_alloc[0] = occurs_alloc[0] * 2 + 1;
	      occurs_table[0] = xrealloc (occurs_table[0],
					  occurs_alloc[0] * sizeof *occurs_table[0]);
	    }

	  occurs_cursor = occurs_table[0] + number_of_occurs[0];

	  /* Define the refence field, if any.  */

	  if (auto_reference)
	    {

	      /* While auto referencing, update `line_start' from its
		 previous value, counting lines as we go.  If input
		 referencing at the same time, `line_start' has been
		 advanced earlier, and the following loop is never really
		 executed.  */

	      while (line_scan < possible_key.start)
		if (*line_scan == '\n')
		  {
		    total_line_count++;
		    line_scan++;
		    line_start = line_scan;
		    SKIP_NON_WHITE (line_scan, text_buffer.end);
		  }
		else
		  line_scan++;

	      occurs_cursor->reference = total_line_count;
	    }
	  else if (input_reference)
	    {

	      /* If only input referencing, `line_start' has been computed
		 earlier to detect the case the word matched would be part
		 of the reference.  The reference position is simply the
		 value of `line_start'.  */

	      occurs_cursor->reference
		= (DELTA) (line_start - possible_key.start);
	      if (reference_length > reference_max_width)
		reference_max_width = reference_length;
	    }

	  /* Exclude the reference from the context in simple cases.  */

	  if (input_reference && line_start == context_start)
	    {
	      SKIP_NON_WHITE (context_start, context_end);
	      SKIP_WHITE (context_start, context_end);
	    }

	  /* Completes the OCCURS structure.  */

	  occurs_cursor->key = possible_key;
	  occurs_cursor->left = context_start - possible_key.start;
	  occurs_cursor->right = context_end - possible_key.start;

	  number_of_occurs[0]++;
	}
    }
}

/* Formatting and actual output - service routines.  */

/*-----------------------------------------.
| Prints some NUMBER of spaces on stdout.  |
`-----------------------------------------*/

static void
print_spaces (int number)
{
  int counter;

  for (counter = number; counter > 0; counter--)
    putchar (' ');
}

/*-------------------------------------.
| Prints the field provided by FIELD.  |
`-------------------------------------*/

static void
print_field (BLOCK field)
{
  char *cursor;			/* Cursor in field to print */
  int base;			/* Base character, without diacritic */
  int diacritic;		/* Diacritic code for the character */

  /* Whitespace is not really compressed.  Instead, each white space
     character (tab, vt, ht etc.) is printed as one single space.  */

  for (cursor = field.start; cursor < field.end; cursor++)
    {
      unsigned char character = *cursor;
      if (edited_flag[character])
	{

	  /* First check if this is a diacriticized character.

	     This works only for TeX.  I do not know how diacriticized
	     letters work with `roff'.  Please someone explain it to me!  */

	  diacritic = todiac (character);
	  if (diacritic != 0 && output_format == TEX_FORMAT)
	    {
	      base = tobase (character);
	      switch (diacritic)
		{

		case 1:		/* Latin diphthongs */
		  switch (base)
		    {
		    case 'o':
		      fputs ("\\oe{}", stdout);
		      break;

		    case 'O':
		      fputs ("\\OE{}", stdout);
		      break;

		    case 'a':
		      fputs ("\\ae{}", stdout);
		      break;

		    case 'A':
		      fputs ("\\AE{}", stdout);
		      break;

		    default:
		      putchar (' ');
		    }
		  break;

		case 2:		/* Acute accent */
		  printf ("\\'%s%c", (base == 'i' ? "\\" : ""), base);
		  break;

		case 3:		/* Grave accent */
		  printf ("\\`%s%c", (base == 'i' ? "\\" : ""), base);
		  break;

		case 4:		/* Circumflex accent */
		  printf ("\\^%s%c", (base == 'i' ? "\\" : ""), base);
		  break;

		case 5:		/* Diaeresis */
		  printf ("\\\"%s%c", (base == 'i' ? "\\" : ""), base);
		  break;

		case 6:		/* Tilde accent */
		  printf ("\\~%s%c", (base == 'i' ? "\\" : ""), base);
		  break;

		case 7:		/* Cedilla */
		  printf ("\\c{%c}", base);
		  break;

		case 8:		/* Small circle beneath */
		  switch (base)
		    {
		    case 'a':
		      fputs ("\\aa{}", stdout);
		      break;

		    case 'A':
		      fputs ("\\AA{}", stdout);
		      break;

		    default:
		      putchar (' ');
		    }
		  break;

		case 9:		/* Strike through */
		  switch (base)
		    {
		    case 'o':
		      fputs ("\\o{}", stdout);
		      break;

		    case 'O':
		      fputs ("\\O{}", stdout);
		      break;

		    default:
		      putchar (' ');
		    }
		  break;
		}
	    }
	  else

	    /* This is not a diacritic character, so handle cases which are
	       really specific to `roff' or TeX.  All white space processing
	       is done as the default case of this switch.  */

	    switch (character)
	      {
	      case '"':
		/* In roff output format, double any quote.  */
		putchar ('"');
		putchar ('"');
		break;

	      case '$':
	      case '%':
	      case '&':
	      case '#':
	      case '_':
		/* In TeX output format, precede these with a backslash.  */
		putchar ('\\');
		putchar (character);
		break;

	      case '{':
	      case '}':
		/* In TeX output format, precede these with a backslash and
		   force mathematical mode.  */
		printf ("$\\%c$", character);
		break;

	      case '\\':
		/* In TeX output mode, request production of a backslash.  */
		fputs ("\\backslash{}", stdout);
		break;

	      default:
		/* Any other flagged character produces a single space.  */
		putchar (' ');
	      }
	}
      else
	putchar (*cursor);
    }
}

/* Formatting and actual output - planning routines.  */

/*--------------------------------------------------------------------.
| From information collected from command line options and input file |
| readings, compute and fix some output parameter values.	      |
`--------------------------------------------------------------------*/

static void
fix_output_parameters (void)
{
  int file_index;		/* index in text input file arrays */
  int line_ordinal;		/* line ordinal value for reference */
  char ordinal_string[12];	/* edited line ordinal for reference */
  int reference_width;		/* width for the whole reference */
  int character;		/* character ordinal */
  const char *cursor;		/* cursor in some constant strings */

  /* In auto reference mode, the maximum width of this field is
     precomputed and subtracted from the overall line width.  Add one for
     the column which separate the file name from the line number.  */

  if (auto_reference)
    {
      reference_max_width = 0;
      for (file_index = 0; file_index < number_input_files; file_index++)
	{
	  line_ordinal = file_line_count[file_index] + 1;
	  if (file_index > 0)
	    line_ordinal -= file_line_count[file_index - 1];
	  sprintf (ordinal_string, "%d", line_ordinal);
	  reference_width = strlen (ordinal_string);
	  if (input_file_name[file_index])
	    reference_width += strlen (input_file_name[file_index]);
	  if (reference_width > reference_max_width)
	    reference_max_width = reference_width;
	}
      reference_max_width++;
      reference.start = xmalloc ((size_t) reference_max_width + 1);
    }

  /* If the reference appears to the left of the output line, reserve some
     space for it right away, including one gap size.  */

  if ((auto_reference | input_reference) & !right_reference)
    line_width -= reference_max_width + gap_size;

  /* The output lines, minimally, will contain from left to right a left
     context, a gap, and a keyword followed by the right context with no
     special intervening gap.  Half of the line width is dedicated to the
     left context and the gap, the other half is dedicated to the keyword
     and the right context; these values are computed once and for all here.
     There also are tail and head wrap around fields, used when the keyword
     is near the beginning or the end of the line, or when some long word
     cannot fit in, but leave place from wrapped around shorter words.  The
     maximum width of these fields are recomputed separately for each line,
     on a case by case basis.  It is worth noting that it cannot happen that
     both the tail and head fields are used at once.  */

  half_line_width = line_width / 2;
  before_max_width = half_line_width - gap_size;
  keyafter_max_width = half_line_width;

  /* If truncation_string is the empty string, make it NULL to speed up
     tests.  In this case, truncation_string_length will never get used, so
     there is no need to set it.  */

  if (truncation_string && *truncation_string)
    truncation_string_length = strlen (truncation_string);
  else
    truncation_string = NULL;

  if (gnu_extensions)
    {

      /* When flagging truncation at the left of the keyword, the
	 truncation mark goes at the beginning of the before field,
	 unless there is a head field, in which case the mark goes at the
	 left of the head field.  When flagging truncation at the right
	 of the keyword, the mark goes at the end of the keyafter field,
	 unless there is a tail field, in which case the mark goes at the
	 end of the tail field.  Only eight combination cases could arise
	 for truncation marks:

	 . None.
	 . One beginning the before field.
	 . One beginning the head field.
	 . One ending the keyafter field.
	 . One ending the tail field.
	 . One beginning the before field, another ending the keyafter field.
	 . One ending the tail field, another beginning the before field.
	 . One ending the keyafter field, another beginning the head field.

	 So, there is at most two truncation marks, which could appear both
	 on the left side of the center of the output line, both on the
	 right side, or one on either side.  */

      before_max_width -= 2 * truncation_string_length;
      keyafter_max_width -= 2 * truncation_string_length;
    }
  else
    {

      /* I never figured out exactly how UNIX' ptx plans the output width
	 of its various fields.  If GNU extensions are disabled, do not
	 try computing the field widths correctly; instead, use the
	 following formula, which does not completely imitate UNIX' ptx,
	 but almost.  */

      keyafter_max_width -= 2 * truncation_string_length + 1;
    }

  /* Compute which characters need special output processing.  Initialize
     by flagging any white space character.  Some systems do not consider
     form feed as a space character, but we do.  */

  for (character = 0; character < CHAR_SET_SIZE; character++)
    edited_flag[character] = !! isspace (character);
  edited_flag['\f'] = 1;

  /* Complete the special character flagging according to selected output
     format.  */

  switch (output_format)
    {
    case UNKNOWN_FORMAT:
      /* Should never happen.  */

    case DUMB_FORMAT:
      break;

    case ROFF_FORMAT:

      /* `Quote' characters should be doubled.  */

      edited_flag['"'] = 1;
      break;

    case TEX_FORMAT:

      /* Various characters need special processing.  */

      for (cursor = "$%&#_{}\\"; *cursor; cursor++)
	edited_flag[to_uchar (*cursor)] = 1;

      /* Any character with 8th bit set will print to a single space, unless
	 it is diacriticized.  */

      for (character = 0200; character < CHAR_SET_SIZE; character++)
	edited_flag[character] = todiac (character) != 0;
      break;
    }
}

/*------------------------------------------------------------------.
| Compute the position and length of all the output fields, given a |
| pointer to some OCCURS.					    |
`------------------------------------------------------------------*/

static void
define_all_fields (OCCURS *occurs)
{
  int tail_max_width;		/* allowable width of tail field */
  int head_max_width;		/* allowable width of head field */
  char *cursor;			/* running cursor in source text */
  char *left_context_start;	/* start of left context */
  char *right_context_end;	/* end of right context */
  char *left_field_start;	/* conservative start for `head'/`before' */
  int file_index;		/* index in text input file arrays */
  const char *file_name;	/* file name for reference */
  int line_ordinal;		/* line ordinal for reference */

  /* Define `keyafter', start of left context and end of right context.
     `keyafter' starts at the saved position for keyword and extend to the
     right from the end of the keyword, eating separators or full words, but
     not beyond maximum allowed width for `keyafter' field or limit for the
     right context.  Suffix spaces will be removed afterwards.  */

  keyafter.start = occurs->key.start;
  keyafter.end = keyafter.start + occurs->key.size;
  left_context_start = keyafter.start + occurs->left;
  right_context_end = keyafter.start + occurs->right;

  cursor = keyafter.end;
  while (cursor < right_context_end
	 && cursor <= keyafter.start + keyafter_max_width)
    {
      keyafter.end = cursor;
      SKIP_SOMETHING (cursor, right_context_end);
    }
  if (cursor <= keyafter.start + keyafter_max_width)
    keyafter.end = cursor;

  keyafter_truncation = truncation_string && keyafter.end < right_context_end;

  SKIP_WHITE_BACKWARDS (keyafter.end, keyafter.start);

  /* When the left context is wide, it might take some time to catch up from
     the left context boundary to the beginning of the `head' or `before'
     fields.  So, in this case, to speed the catchup, we jump back from the
     keyword, using some secure distance, possibly falling in the middle of
     a word.  A secure backward jump would be at least half the maximum
     width of a line, plus the size of the longest word met in the whole
     input.  We conclude this backward jump by a skip forward of at least
     one word.  In this manner, we should not inadvertently accept only part
     of a word.  From the reached point, when it will be time to fix the
     beginning of `head' or `before' fields, we will skip forward words or
     delimiters until we get sufficiently near.  */

  if (-occurs->left > half_line_width + maximum_word_length)
    {
      left_field_start
	= keyafter.start - (half_line_width + maximum_word_length);
      SKIP_SOMETHING (left_field_start, keyafter.start);
    }
  else
    left_field_start = keyafter.start + occurs->left;

  /* `before' certainly ends at the keyword, but not including separating
     spaces.  It starts after than the saved value for the left context, by
     advancing it until it falls inside the maximum allowed width for the
     before field.  There will be no prefix spaces either.  `before' only
     advances by skipping single separators or whole words. */

  before.start = left_field_start;
  before.end = keyafter.start;
  SKIP_WHITE_BACKWARDS (before.end, before.start);

  while (before.start + before_max_width < before.end)
    SKIP_SOMETHING (before.start, before.end);

  if (truncation_string)
    {
      cursor = before.start;
      SKIP_WHITE_BACKWARDS (cursor, text_buffer.start);
      before_truncation = cursor > left_context_start;
    }
  else
    before_truncation = 0;

  SKIP_WHITE (before.start, text_buffer.end);

  /* The tail could not take more columns than what has been left in the
     left context field, and a gap is mandatory.  It starts after the
     right context, and does not contain prefixed spaces.  It ends at
     the end of line, the end of buffer or when the tail field is full,
     whichever comes first.  It cannot contain only part of a word, and
     has no suffixed spaces.  */

  tail_max_width
    = before_max_width - (before.end - before.start) - gap_size;

  if (tail_max_width > 0)
    {
      tail.start = keyafter.end;
      SKIP_WHITE (tail.start, text_buffer.end);

      tail.end = tail.start;
      cursor = tail.end;
      while (cursor < right_context_end
	     && cursor < tail.start + tail_max_width)
	{
	  tail.end = cursor;
	  SKIP_SOMETHING (cursor, right_context_end);
	}

      if (cursor < tail.start + tail_max_width)
	tail.end = cursor;

      if (tail.end > tail.start)
	{
	  keyafter_truncation = 0;
	  tail_truncation = truncation_string && tail.end < right_context_end;
	}
      else
	tail_truncation = 0;

      SKIP_WHITE_BACKWARDS (tail.end, tail.start);
    }
  else
    {

      /* No place left for a tail field.  */

      tail.start = NULL;
      tail.end = NULL;
      tail_truncation = 0;
    }

  /* `head' could not take more columns than what has been left in the right
     context field, and a gap is mandatory.  It ends before the left
     context, and does not contain suffixed spaces.  Its pointer is advanced
     until the head field has shrunk to its allowed width.  It cannot
     contain only part of a word, and has no suffixed spaces.  */

  head_max_width
    = keyafter_max_width - (keyafter.end - keyafter.start) - gap_size;

  if (head_max_width > 0)
    {
      head.end = before.start;
      SKIP_WHITE_BACKWARDS (head.end, text_buffer.start);

      head.start = left_field_start;
      while (head.start + head_max_width < head.end)
	SKIP_SOMETHING (head.start, head.end);

      if (head.end > head.start)
	{
	  before_truncation = 0;
	  head_truncation = (truncation_string
			     && head.start > left_context_start);
	}
      else
	head_truncation = 0;

      SKIP_WHITE (head.start, head.end);
    }
  else
    {

      /* No place left for a head field.  */

      head.start = NULL;
      head.end = NULL;
      head_truncation = 0;
    }

  if (auto_reference)
    {

      /* Construct the reference text in preallocated space from the file
	 name and the line number.  Find out in which file the reference
	 occurred.  Standard input yields an empty file name.  Insure line
	 numbers are one based, even if they are computed zero based.  */

      file_index = 0;
      while (file_line_count[file_index] < occurs->reference)
	file_index++;

      file_name = input_file_name[file_index];
      if (!file_name)
	file_name = "";

      line_ordinal = occurs->reference + 1;
      if (file_index > 0)
	line_ordinal -= file_line_count[file_index - 1];

      sprintf (reference.start, "%s:%d", file_name, line_ordinal);
      reference.end = reference.start + strlen (reference.start);
    }
  else if (input_reference)
    {

      /* Reference starts at saved position for reference and extends right
	 until some white space is met.  */

      reference.start = keyafter.start + (DELTA) occurs->reference;
      reference.end = reference.start;
      SKIP_NON_WHITE (reference.end, right_context_end);
    }
}

/* Formatting and actual output - control routines.  */

/*----------------------------------------------------------------------.
| Output the current output fields as one line for `troff' or `nroff'.  |
`----------------------------------------------------------------------*/

static void
output_one_roff_line (void)
{
  /* Output the `tail' field.  */

  printf (".%s \"", macro_name);
  print_field (tail);
  if (tail_truncation)
    fputs (truncation_string, stdout);
  putchar ('"');

  /* Output the `before' field.  */

  fputs (" \"", stdout);
  if (before_truncation)
    fputs (truncation_string, stdout);
  print_field (before);
  putchar ('"');

  /* Output the `keyafter' field.  */

  fputs (" \"", stdout);
  print_field (keyafter);
  if (keyafter_truncation)
    fputs (truncation_string, stdout);
  putchar ('"');

  /* Output the `head' field.  */

  fputs (" \"", stdout);
  if (head_truncation)
    fputs (truncation_string, stdout);
  print_field (head);
  putchar ('"');

  /* Conditionally output the `reference' field.  */

  if (auto_reference | input_reference)
    {
      fputs (" \"", stdout);
      print_field (reference);
      putchar ('"');
    }

  putchar ('\n');
}

/*---------------------------------------------------------.
| Output the current output fields as one line for `TeX'.  |
`---------------------------------------------------------*/

static void
output_one_tex_line (void)
{
  BLOCK key;			/* key field, isolated */
  BLOCK after;			/* after field, isolated */
  char *cursor;			/* running cursor in source text */

  printf ("\\%s ", macro_name);
  putchar ('{');
  print_field (tail);
  fputs ("}{", stdout);
  print_field (before);
  fputs ("}{", stdout);
  key.start = keyafter.start;
  after.end = keyafter.end;
  cursor = keyafter.start;
  SKIP_SOMETHING (cursor, keyafter.end);
  key.end = cursor;
  after.start = cursor;
  print_field (key);
  fputs ("}{", stdout);
  print_field (after);
  fputs ("}{", stdout);
  print_field (head);
  putchar ('}');
  if (auto_reference | input_reference)
    {
      putchar ('{');
      print_field (reference);
      putchar ('}');
    }
  putchar ('\n');
}

/*-------------------------------------------------------------------.
| Output the current output fields as one line for a dumb terminal.  |
`-------------------------------------------------------------------*/

static void
output_one_dumb_line (void)
{
  if (!right_reference)
    {
      if (auto_reference)
	{

	  /* Output the `reference' field, in such a way that GNU emacs
	     next-error will handle it.  The ending colon is taken from the
	     gap which follows.  */

	  print_field (reference);
	  putchar (':');
	  print_spaces (reference_max_width
			+ gap_size
			- (reference.end - reference.start)
			- 1);
	}
      else
	{

	  /* Output the `reference' field and its following gap.  */

	  print_field (reference);
	  print_spaces (reference_max_width
			+ gap_size
			- (reference.end - reference.start));
	}
    }

  if (tail.start < tail.end)
    {
      /* Output the `tail' field.  */

      print_field (tail);
      if (tail_truncation)
	fputs (truncation_string, stdout);

      print_spaces (half_line_width - gap_size
		    - (before.end - before.start)
		    - (before_truncation ? truncation_string_length : 0)
		    - (tail.end - tail.start)
		    - (tail_truncation ? truncation_string_length : 0));
    }
  else
    print_spaces (half_line_width - gap_size
		  - (before.end - before.start)
		  - (before_truncation ? truncation_string_length : 0));

  /* Output the `before' field.  */

  if (before_truncation)
    fputs (truncation_string, stdout);
  print_field (before);

  print_spaces (gap_size);

  /* Output the `keyafter' field.  */

  print_field (keyafter);
  if (keyafter_truncation)
    fputs (truncation_string, stdout);

  if (head.start < head.end)
    {
      /* Output the `head' field.  */

      print_spaces (half_line_width
		    - (keyafter.end - keyafter.start)
		    - (keyafter_truncation ? truncation_string_length : 0)
		    - (head.end - head.start)
		    - (head_truncation ? truncation_string_length : 0));
      if (head_truncation)
	fputs (truncation_string, stdout);
      print_field (head);
    }
  else

    if ((auto_reference | input_reference) & right_reference)
      print_spaces (half_line_width
		    - (keyafter.end - keyafter.start)
		    - (keyafter_truncation ? truncation_string_length : 0));

  if ((auto_reference | input_reference) & right_reference)
    {
      /* Output the `reference' field.  */

      print_spaces (gap_size);
      print_field (reference);
    }

  putchar ('\n');
}

/*------------------------------------------------------------------------.
| Scan the whole occurs table and, for each entry, output one line in the |
| appropriate format.							  |
`------------------------------------------------------------------------*/

static void
generate_all_output (void)
{
  size_t occurs_index;		/* index of keyword entry being processed */
  OCCURS *occurs_cursor;	/* current keyword entry being processed */

  /* The following assignments are useful to provide default values in case
     line contexts or references are not used, in which case these variables
     would never be computed.  */

  tail.start = NULL;
  tail.end = NULL;
  tail_truncation = 0;

  head.start = NULL;
  head.end = NULL;
  head_truncation = 0;

  /* Loop over all keyword occurrences.  */

  occurs_cursor = occurs_table[0];

  for (occurs_index = 0; occurs_index < number_of_occurs[0]; occurs_index++)
    {
      /* Compute the exact size of every field and whenever truncation flags
	 are present or not.  */

      define_all_fields (occurs_cursor);

      /* Produce one output line according to selected format.  */

      switch (output_format)
	{
	case UNKNOWN_FORMAT:
	  /* Should never happen.  */

	case DUMB_FORMAT:
	  output_one_dumb_line ();
	  break;

	case ROFF_FORMAT:
	  output_one_roff_line ();
	  break;

	case TEX_FORMAT:
	  output_one_tex_line ();
	  break;
	}

      /* Advance the cursor into the occurs table.  */

      occurs_cursor++;
    }
}

/* Option decoding and main program.  */

/*------------------------------------------------------.
| Print program identification and options, then exit.  |
`------------------------------------------------------*/

void
usage (int status)
{
  if (status != EXIT_SUCCESS)
    fprintf (stderr, _("Try `%s --help' for more information.\n"),
	     program_name);
  else
    {
      printf (_("\
Usage: %s [OPTION]... [INPUT]...   (without -G)\n\
  or:  %s -G [OPTION]... [INPUT [OUTPUT]]\n"),
	      program_name, program_name);
      fputs (_("\
Output a permuted index, including context, of the words in the input files.\n\
\n\
"), stdout);
      fputs (_("\
Mandatory arguments to long options are mandatory for short options too.\n\
"), stdout);
      fputs (_("\
  -A, --auto-reference           output automatically generated references\n\
  -G, --traditional              behave more like System V `ptx'\n\
  -F, --flag-truncation=STRING   use STRING for flagging line truncations\n\
"), stdout);
      fputs (_("\
  -M, --macro-name=STRING        macro name to use instead of `xx'\n\
  -O, --format=roff              generate output as roff directives\n\
  -R, --right-side-refs          put references at right, not counted in -w\n\
  -S, --sentence-regexp=REGEXP   for end of lines or end of sentences\n\
  -T, --format=tex               generate output as TeX directives\n\
"), stdout);
      fputs (_("\
  -W, --word-regexp=REGEXP       use REGEXP to match each keyword\n\
  -b, --break-file=FILE          word break characters in this FILE\n\
  -f, --ignore-case              fold lower case to upper case for sorting\n\
  -g, --gap-size=NUMBER          gap size in columns between output fields\n\
  -i, --ignore-file=FILE         read ignore word list from FILE\n\
  -o, --only-file=FILE           read only word list from this FILE\n\
"), stdout);
      fputs (_("\
  -r, --references               first field of each line is a reference\n\
  -t, --typeset-mode               - not implemented -\n\
  -w, --width=NUMBER             output width in columns, reference excluded\n\
"), stdout);
      fputs (HELP_OPTION_DESCRIPTION, stdout);
      fputs (VERSION_OPTION_DESCRIPTION, stdout);
      fputs (_("\
\n\
With no FILE or if FILE is -, read Standard Input.  `-F /' by default.\n\
"), stdout);
      printf (_("\nReport bugs to <%s>.\n"), PACKAGE_BUGREPORT);
    }
  exit (status);
}

/*----------------------------------------------------------------------.
| Main program.  Decode ARGC arguments passed through the ARGV array of |
| strings, then launch execution.				        |
`----------------------------------------------------------------------*/

/* Long options equivalences.  */
static const struct option long_options[] =
{
  {"auto-reference", no_argument, NULL, 'A'},
  {"break-file", required_argument, NULL, 'b'},
  {"copyright", no_argument, NULL, 'C'}, /* Deprecated, remove in 2007.  */
  {"flag-truncation", required_argument, NULL, 'F'},
  {"ignore-case", no_argument, NULL, 'f'},
  {"gap-size", required_argument, NULL, 'g'},
  {"ignore-file", required_argument, NULL, 'i'},
  {"macro-name", required_argument, NULL, 'M'},
  {"only-file", required_argument, NULL, 'o'},
  {"references", no_argument, NULL, 'r'},
  {"right-side-refs", no_argument, NULL, 'R'},
  {"format", required_argument, NULL, 10},
  {"sentence-regexp", required_argument, NULL, 'S'},
  {"traditional", no_argument, NULL, 'G'},
  {"typeset-mode", no_argument, NULL, 't'},
  {"width", required_argument, NULL, 'w'},
  {"word-regexp", required_argument, NULL, 'W'},
  {GETOPT_HELP_OPTION_DECL},
  {GETOPT_VERSION_OPTION_DECL},
  {NULL, 0, NULL, 0},
};

static char const* const format_args[] =
{
  "roff", "tex", NULL
};

static enum Format const format_vals[] =
{
  ROFF_FORMAT, TEX_FORMAT
};

int
main (int argc, char **argv)
{
  int optchar;			/* argument character */
  int file_index;		/* index in text input file arrays */

  /* Decode program options.  */

  initialize_main (&argc, &argv);
  program_name = argv[0];
  setlocale (LC_ALL, "");
  bindtextdomain (PACKAGE, LOCALEDIR);
  textdomain (PACKAGE);

  atexit (close_stdout);

#if HAVE_SETCHRCLASS
  setchrclass (NULL);
#endif

  while (optchar = getopt_long (argc, argv, "ACF:GM:ORS:TW:b:i:fg:o:trw:",
				long_options, NULL),
	 optchar != EOF)
    {
      switch (optchar)
	{
	default:
	  usage (EXIT_FAILURE);

	case 'G':
	  gnu_extensions = false;
	  break;

	case 'b':
	  break_file = optarg;
	  break;

	case 'f':
	  ignore_case = true;
	  break;

	case 'g':
	  {
	    unsigned long int tmp_ulong;
	    if (xstrtoul (optarg, NULL, 0, &tmp_ulong, NULL) != LONGINT_OK
		|| ! (0 < tmp_ulong && tmp_ulong <= INT_MAX))
	      error (EXIT_FAILURE, 0, _("invalid gap width: %s"),
		     quotearg (optarg));
	    gap_size = tmp_ulong;
	    break;
	  }

	case 'i':
	  ignore_file = optarg;
	  break;

	case 'o':
	  only_file = optarg;
	  break;

	case 'r':
	  input_reference = true;
	  break;

	case 't':
	  /* Yet to understand...  */
	  break;

	case 'w':
	  {
	    unsigned long int tmp_ulong;
	    if (xstrtoul (optarg, NULL, 0, &tmp_ulong, NULL) != LONGINT_OK
		|| ! (0 < tmp_ulong && tmp_ulong <= INT_MAX))
	      error (EXIT_FAILURE, 0, _("invalid line width: %s"),
		     quotearg (optarg));
	    line_width = tmp_ulong;
	    break;
	  }

	case 'A':
	  auto_reference = true;
	  break;

	case 'F':
	  truncation_string = copy_unescaped_string (optarg);
	  break;

	case 'M':
	  macro_name = optarg;
	  break;

	case 'O':
	  output_format = ROFF_FORMAT;
	  break;

	case 'R':
	  right_reference = true;
	  break;

	case 'S':
	  context_regex.string = copy_unescaped_string (optarg);
	  break;

	case 'T':
	  output_format = TEX_FORMAT;
	  break;

	case 'W':
	  word_regex.string = copy_unescaped_string (optarg);
	  if (!*word_regex.string)
	    word_regex.string = NULL;
	  break;

	case 10:
	  output_format = XARGMATCH ("--format", optarg,
				     format_args, format_vals);
	case_GETOPT_HELP_CHAR;

	case 'C':
	  error (0, 0, _("\
the --copyright option is deprecated; use --version instead"));
          /* fallthrough */

	case_GETOPT_VERSION_CHAR (PROGRAM_NAME, AUTHORS);
	}
    }

  /* Process remaining arguments.  If GNU extensions are enabled, process
     all arguments as input parameters.  If disabled, accept at most two
     arguments, the second of which is an output parameter.  */

  if (optind == argc)
    {

      /* No more argument simply means: read standard input.  */

      input_file_name = xmalloc (sizeof *input_file_name);
      file_line_count = xmalloc (sizeof *file_line_count);
      number_input_files = 1;
      input_file_name[0] = NULL;
    }
  else if (gnu_extensions)
    {
      number_input_files = argc - optind;
      input_file_name = xmalloc (number_input_files * sizeof *input_file_name);
      file_line_count = xmalloc (number_input_files * sizeof *file_line_count);

      for (file_index = 0; file_index < number_input_files; file_index++)
	{
	  input_file_name[file_index] = argv[optind];
	  if (!*argv[optind] || STREQ (argv[optind], "-"))
	    input_file_name[0] = NULL;
	  else
	    input_file_name[0] = argv[optind];
	  optind++;
	}
    }
  else
    {

      /* There is one necessary input file.  */

      number_input_files = 1;
      input_file_name = xmalloc (sizeof *input_file_name);
      file_line_count = xmalloc (sizeof *file_line_count);
      if (!*argv[optind] || STREQ (argv[optind], "-"))
	input_file_name[0] = NULL;
      else
	input_file_name[0] = argv[optind];
      optind++;

      /* Redirect standard output, only if requested.  */

      if (optind < argc)
	{
	  if (! freopen (argv[optind], "w", stdout))
	    error (EXIT_FAILURE, errno, "%s", argv[optind]);
	  optind++;
	}

      /* Diagnose any other argument as an error.  */

      if (optind < argc)
	{
	  error (0, 0, _("extra operand %s"), quote (argv[optind]));
	  usage (EXIT_FAILURE);
	}
    }

  /* If the output format has not been explicitly selected, choose dumb
     terminal format if GNU extensions are enabled, else `roff' format.  */

  if (output_format == UNKNOWN_FORMAT)
    output_format = gnu_extensions ? DUMB_FORMAT : ROFF_FORMAT;

  /* Initialize the main tables.  */

  initialize_regex ();

  /* Read `Break character' file, if any.  */

  if (break_file)
    digest_break_file (break_file);

  /* Read `Ignore words' file and `Only words' files, if any.  If any of
     these files is empty, reset the name of the file to NULL, to avoid
     unnecessary calls to search_table. */

  if (ignore_file)
    {
      digest_word_file (ignore_file, &ignore_table);
      if (ignore_table.length == 0)
	ignore_file = NULL;
    }

  if (only_file)
    {
      digest_word_file (only_file, &only_table);
      if (only_table.length == 0)
	only_file = NULL;
    }

  /* Prepare to study all the input files.  */

  number_of_occurs[0] = 0;
  total_line_count = 0;
  maximum_word_length = 0;
  reference_max_width = 0;

  for (file_index = 0; file_index < number_input_files; file_index++)
    {

      /* Read the file in core, than study it.  */

      swallow_file_in_memory (input_file_name[file_index], &text_buffer);
      find_occurs_in_text ();

      /* Maintain for each file how many lines has been read so far when its
	 end is reached.  Incrementing the count first is a simple kludge to
	 handle a possible incomplete line at end of file.  */

      total_line_count++;
      file_line_count[file_index] = total_line_count;
    }

  /* Do the output process phase.  */

  sort_found_occurs ();
  fix_output_parameters ();
  generate_all_output ();

  /* All done.  */

  exit (EXIT_SUCCESS);
}