summaryrefslogtreecommitdiffstats
path: root/src/xget-lex.c
blob: 08b31581869e3b5529df91aed4214e64d668b968 (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
/* GNU gettext - internationalization aids
   Copyright (C) 1995-1998, 2000, 2001 Free Software Foundation, Inc.

   This file was written by Peter Miller <millerp@canb.auug.org.au>

   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., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */

#ifdef HAVE_CONFIG_H
# include "config.h"
#endif

#include <ctype.h>
#include <errno.h>
#include <stdio.h>
#include <stdlib.h>

#include "dir-list.h"
#include "error.h"
#include "progname.h"
#include "system.h"
#include "libgettext.h"
#include "hash.h"
#include "str-list.h"
#include "xget-lex.h"

#ifndef errno
extern int errno;
#endif

#define _(s) gettext(s)


/* The ANSI C standard defines several phases of translation:

   1. Terminate line by \n, regardless of the external representation
      of a text line.  Stdio does this for us.

   2. Convert trigraphs to their single character equivalents.

   3. Concatenate each line ending in backslash (\) with the following
      line.

   4. Replace each comment with a space character.

   5. Parse each resulting logical line as preprocessing tokens a
      white space.

   6. Recognize and carry out directives (it also expands macros on
      non-directive lines, which we do not do here).

   7. Replaces escape sequences within character strings with their
      single character equivalents (we do this in step 5, because we
      don't have to worry about the #include argument).

   8. Concatenates adjacent string literals to form single string
      literals (because we don't expand macros, there are a few things
      we will miss).

   9. Converts the remaining preprocessing tokens to C tokens and
      discards any white space from the translation unit.

   This lexer implements the above, and presents the scanner (in
   xgettext.c) with a stream of C tokens.  The comments are
   accumulated in a buffer, and given to xgettext when asked for.  */

enum token_type_ty
{
  token_type_character_constant,
  token_type_eof,
  token_type_eoln,
  token_type_hash,
  token_type_lparen,
  token_type_rparen,
  token_type_comma,
  token_type_name,
  token_type_number,
  token_type_string_literal,
  token_type_symbol,
  token_type_white_space
};
typedef enum token_type_ty token_type_ty;

typedef struct token_ty token_ty;
struct token_ty
{
  token_type_ty type;
  char *string;
  long number;
  int line_number;
};


static const char *file_name;
static char *logical_file_name;
static int line_number;
static FILE *fp;
static int trigraphs;
static string_list_ty *comment;
static hash_table keywords;
static int default_keywords = 1;

/* These are for tracking whether comments count as immediately before
   keyword.  */
static int last_comment_line = -1;
static int last_non_comment_line = -1;
static int newline_count = 0;


/* Prototypes for local functions.  */
static int phase1_getc PARAMS ((void));
static void phase1_ungetc PARAMS ((int __c));
static int phase2_getc PARAMS ((void));
static void phase2_ungetc PARAMS ((int __c));
static int phase3_getc PARAMS ((void));
static void phase3_ungetc PARAMS ((int __c));
static int phase4_getc PARAMS ((void));
static void phase4_ungetc PARAMS ((int __c));
static int phase7_getc PARAMS ((void));
static void phase7_ungetc PARAMS ((int __c));
static void phase5_get PARAMS ((token_ty *__tp));
static void phase5_unget PARAMS ((token_ty *__tp));
static void phaseX_get PARAMS ((token_ty *__tp));
static void phase6_get PARAMS ((token_ty *__tp));
static void phase6_unget PARAMS ((token_ty *__tp));
static void phase8_get PARAMS ((token_ty *__tp));



void
xgettext_lex_open (fn)
     const char *fn;
{
  char *new_name;

  if (strcmp (fn, "-") == 0)
    {
      new_name = xstrdup (_("standard input"));
      logical_file_name = xstrdup (new_name);
      fp = stdin;
    }
  else if (IS_ABSOLUTE_PATH (fn))
    {
      new_name = xstrdup (fn);
      fp = fopen (fn, "r");
      if (fp == NULL)
	error (EXIT_FAILURE, errno, _("\
error while opening \"%s\" for reading"), fn);
      logical_file_name = xstrdup (new_name);
    }
  else
    {
      int j;

      for (j = 0; ; ++j)
	{
	  const char *dir = dir_list_nth (j);

	  if (dir == NULL)
	    error (EXIT_FAILURE, ENOENT, _("\
error while opening \"%s\" for reading"), fn);

	  new_name = concatenated_pathname (dir, fn, NULL);

	  fp = fopen (new_name, "r");
	  if (fp != NULL)
	    break;

	  if (errno != ENOENT)
	    error (EXIT_FAILURE, errno, _("\
error while opening \"%s\" for reading"), new_name);
	  free (new_name);
	}

      /* Note that the NEW_NAME variable contains the actual file name
	 and the logical file name is what is reported by xgettext.  In
	 this case NEW_NAME is set to the file which was found along the
	 directory search path, and LOGICAL_FILE_NAME is is set to the
	 file name which was searched for.  */
      logical_file_name = xstrdup (fn);
    }

  file_name = new_name;
  line_number = 1;
}


void
xgettext_lex_close ()
{
  if (fp != stdin)
    fclose (fp);
  free ((char *) file_name);
  free (logical_file_name);
  fp = NULL;
  file_name = NULL;
  logical_file_name = NULL;
  line_number = 0;
}


/* 1. Terminate line by \n, regardless of the external representation of
   a text line.  Stdio does this for us, we just need to check that
   there are no I/O errors, and cope with potentially 2 characters of
   pushback, not just the one that ungetc can cope with.  */

/* Maximum used guaranteed to be < 4.  */
static unsigned char phase1_pushback[4];
static int phase1_pushback_length;


static int
phase1_getc ()
{
  int c;

  if (phase1_pushback_length)
    {
      c = phase1_pushback[--phase1_pushback_length];
      if (c == '\n')
	++line_number;
      return c;
    }
  while (1)
    {
      c = getc (fp);
      switch (c)
	{
	case EOF:
	  if (ferror (fp))
	    {
	    bomb:
	      error (EXIT_FAILURE, errno, _("\
error while reading \"%s\""), file_name);
	    }
	  return EOF;

	case '\n':
	  ++line_number;
	  return '\n';

	case '\\':
	  c = getc (fp);
	  if (c == EOF)
	    {
	      if (ferror (fp))
		goto bomb;
	      return '\\';
	    }
	  if (c != '\n')
	    {
	      ungetc (c, fp);
	      return '\\';
	    }
	  ++line_number;
	  break;

	default:
	  return c;
	}
    }
}


static void
phase1_ungetc (c)
     int c;
{
  switch (c)
    {
    case EOF:
      break;

    case '\n':
      --line_number;
      /* FALLTHROUGH */

    default:
      phase1_pushback[phase1_pushback_length++] = c;
      break;
    }
}


/* 2. Convert trigraphs to their single character equivalents.  Most
   sane human beings vomit copiously at the mention of trigraphs, which
   is why they are an option.  */

/* Maximum used guaranteed to be < 4.  */
static unsigned char phase2_pushback[4];
static int phase2_pushback_length;


static int
phase2_getc ()
{
  int c;

  if (phase2_pushback_length)
    return phase2_pushback[--phase2_pushback_length];
  if (!trigraphs)
    return phase1_getc ();

  c = phase1_getc ();
  if (c != '?')
    return c;
  c = phase1_getc ();
  if (c != '?')
    {
      phase1_ungetc (c);
      return '?';
    }
  c = phase1_getc ();
  switch (c)
    {
    case '(':
      return '[';
    case '/':
      return '\\';
    case ')':
      return ']';
    case '\'':
      return '^';
    case '<':
      return '{';
    case '!':
      return '|';
    case '>':
      return '}';
    case '-':
      return '~';
    case '#':
      return '=';
    }
  phase1_ungetc (c);
  phase1_ungetc ('?');
  return '?';
}


static void
phase2_ungetc (c)
     int c;
{
  if (c != EOF)
    phase2_pushback[phase2_pushback_length++] = c;
}


/* 3. Concatenate each line ending in backslash (\) with the following
   line.  Basically, all you need to do is elide "\\\n" sequences from
   the input.  */

/* Maximum used guaranteed to be < 4.  */
static unsigned char phase3_pushback[4];
static int phase3_pushback_length;


static int
phase3_getc ()
{
  if (phase3_pushback_length)
    return phase3_pushback[--phase3_pushback_length];
  for (;;)
    {
      int c = phase2_getc ();
      if (c != '\\')
	return c;
      c = phase2_getc ();
      if (c != '\n')
	{
	  phase2_ungetc (c);
	  return '\\';
	}
    }
}


static void
phase3_ungetc (c)
     int c;
{
  if (c != EOF)
    phase3_pushback[phase3_pushback_length++] = c;
}


/* 4. Replace each comment that is not inside a character constant or
   string literal with a space character.  We need to remember the
   comment for later, because it may be attached to a keyword string.
   We also optionally understand C++ comments.  */

static int
phase4_getc ()
{
  static char *buffer;
  static size_t bufmax;
  size_t buflen;
  int c;
  int state;

  c = phase3_getc ();
  if (c != '/')
    return c;
  c = phase3_getc ();
  switch (c)
    {
    default:
      phase3_ungetc (c);
      return '/';

    case '*':
      /* C comment.  */
      buflen = 0;
      state = 0;
      if (comment == NULL)
	comment = string_list_alloc ();
      while (1)
	{
	  c = phase3_getc ();
	  if (c == EOF)
	    break;
	  /* We skip all leading white space, but not EOLs.  */
	  if (buflen == 0 && isspace (c) && c != '\n')
	    continue;
	  if (buflen >= bufmax)
	    {
	      bufmax += 100;
	      buffer = xrealloc (buffer, bufmax);
	    }
	  buffer[buflen++] = c;
	  switch (c)
	    {
	    case '\n':
	      --buflen;
	      while (buflen >= 1 && (buffer[buflen - 1] == ' '
				     || buffer[buflen - 1] == '\t'))
		--buflen;
	      buffer[buflen] = 0;
	      string_list_append (comment, buffer);
	      buflen = 0;
	      state = 0;
	      continue;

	    case '*':
	      state = 1;
	      continue;

	    case '/':
	      if (state == 1)
		{
		  buflen -= 2;
		  while (buflen >= 1 && (buffer[buflen - 1] == ' '
					 || buffer[buflen - 1] == '\t'))
		    --buflen;
		  buffer[buflen] = 0;
		  string_list_append (comment, buffer);
		  break;
		}
	      /* FALLTHROUGH */

	    default:
	      state = 0;
	      continue;
	    }
	  break;
	}
      last_comment_line = newline_count;
      return ' ';

    case '/':
      /* C++ or ISO C 99 comment.  */
      buflen = 0;
      while (1)
	{
	  c = phase3_getc ();
	  if (c == '\n' || c == EOF)
	    break;
	  if (buflen >= bufmax)
	    {
	      bufmax += 100;
	      buffer = xrealloc (buffer, bufmax);
	    }
	  buffer[buflen++] = c;
	}
      if (buflen >= bufmax)
	{
	  bufmax += 100;
	  buffer = xrealloc (buffer, bufmax);
	}
      buffer[buflen] = 0;
      if (comment == NULL)
	comment = string_list_alloc ();
      string_list_append (comment, buffer);
      last_comment_line = newline_count;
      return '\n';
    }
}


static void
phase4_ungetc (c)
     int c;
{
  phase3_ungetc (c);
}


/* 7. Replace escape sequences within character strings with their
   single character equivalents.  This is called from phase 5, because
   we don't have to worry about the #include argument.  There are
   pathological cases which could bite us (like the DOS directory
   separator), but just pretend it can't happen.  */

#define P7_QUOTES (1000 + '"')
#define P7_QUOTE (1000 + '\'')
#define P7_NEWLINE (1000 + '\n')

static int
phase7_getc ()
{
  int c, n, j;

  /* Use phase 3, because phase 4 elides comments.  */
  c = phase3_getc ();

  /* Return a magic newline indicator, so that we can distinguish
     between the user requesting a newline in the string (e.g. using
     "\n" or "\15") from the user failing to terminate the string or
     character constant.  The ANSI C standard says: 3.1.3.4 Character
     Constants contain ``any character except single quote, backslash or
     newline; or an escape sequence'' and 3.1.4 String Literals contain
     ``any character except double quote, backslash or newline; or an
     escape sequence''.

     Most compilers give a fatal error in this case, however gcc is
     stupidly silent, even though this is a very common typo.  OK, so
     gcc --pedantic will tell me, but that gripes about too much other
     stuff.  Could I have a ``gcc -Wnewline-in-string'' option, or
     better yet a ``gcc -fno-newline-in-string'' option, please?  Gcc is
     also inconsistent between string literals and character constants:
     you may not embed newlines in character constants; try it, you get
     a useful diagnostic.  --PMiller  */
  if (c == '\n')
    return P7_NEWLINE;

  if (c == '"')
    return P7_QUOTES;
  if (c == '\'')
    return P7_QUOTE;
  if (c != '\\')
    return c;
  c = phase3_getc ();
  switch (c)
    {
    default:
      /* Unknown escape sequences really should be an error, but just
	 ignore them, and let the real compiler complain.  */
      phase3_ungetc (c);
      return '\\';

    case '"':
    case '\'':
    case '?':
    case '\\':
      return c;

      /* The \a and \v escapes were added by the ANSI C Standard.
	 Prior to the Standard, most compilers did not have them.
	 Because we need the same program on all platforms we don't
	 provide support for them here.

	 The gcc sources comment that \a is commonly available in
	 pre-ANSI compilers.  --PMiller  */

    case 'b':
      return '\b';

      /* The \e escape is preculiar to gcc, and assumes an ASCII
         character set (or superset).  We don't provide support for it
         here.  */

    case 'f':
      return '\f';
    case 'n':
      return '\n';
    case 'r':
      return '\r';
    case 't':
      return '\t';

    case 'x':
      c = phase3_getc ();
      switch (c)
	{
	default:
	  phase3_ungetc (c);
	  phase3_ungetc ('x');
	  return '\\';

	case '0': case '1': case '2': case '3': case '4':
	case '5': case '6': case '7': case '8': case '9':
	case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
	case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
	  break;
	}
      n = 0;
      for (;;)
	{
	  switch (c)
	    {
	    default:
	      phase3_ungetc (c);
	      return n;
	      break;

	    case '0': case '1': case '2': case '3': case '4':
	    case '5': case '6': case '7': case '8': case '9':
	      n = n * 16 + c - '0';
	      break;

	    case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
	      n = n * 16 + 10 + c - 'A';
	      break;

	    case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
	      n = n * 16 + 10 + c - 'a';
	      break;
	    }
	  c = phase3_getc ();
	}
      return n;

    case '0': case '1': case '2': case '3':
    case '4': case '5': case '6': case '7':
      n = 0;
      for (j = 0; j < 3; ++j)
	{
	  n = n * 8 + c - '0';
	  c = phase3_getc ();
	  switch (c)
	    {
	    default:
	      break;

	    case '0': case '1': case '2': case '3':
	    case '4': case '5': case '6': case '7':
	      continue;
	    }
	  break;
	}
      phase3_ungetc (c);
      return n;
    }
}


static void
phase7_ungetc (c)
     int c;
{
  phase3_ungetc (c);
}


/* 5. Parse each resulting logical line as preprocessing tokens and
   white space.  Preprocessing tokens and C tokens don't always match.  */

/* Maximum used guaranteed to be < 4.  */
static token_ty phase5_pushback[4];
static int phase5_pushback_length;


static void
phase5_get (tp)
     token_ty *tp;
{
  static char *buffer;
  static int bufmax;
  int bufpos;
  int c;

  if (phase5_pushback_length)
    {
      *tp = phase5_pushback[--phase5_pushback_length];
      return;
    }
  tp->string = 0;
  tp->number = 0;
  tp->line_number = line_number;
  c = phase4_getc ();
  switch (c)
    {
    case EOF:
      tp->type = token_type_eof;
      return;

    case '\n':
      tp->type = token_type_eoln;
      return;

    case ' ':
    case '\f':
    case '\t':
      for (;;)
	{
	  c = phase4_getc ();
	  switch (c)
	    {
	    case ' ':
	    case '\f':
	    case '\t':
	      continue;

	    default:
	      phase4_ungetc (c);
	      break;
	    }
	  break;
	}
      tp->type = token_type_white_space;
      return;

    case 'A': case 'B': case 'C': case 'D': case 'E': case 'F': case 'G':
    case 'H': case 'I': case 'J': case 'K': case 'L': case 'M': case 'N':
    case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
    case 'V': case 'W': case 'X': case 'Y': case 'Z':
    case '_':
    case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
    case 'h': case 'i': case 'j': case 'k': case 'l': case 'm': case 'n':
    case 'o': case 'p': case 'q': case 'r': case 's': case 't': case 'u':
    case 'v': case 'w': case 'x': case 'y': case 'z':
      bufpos = 0;
      for (;;)
	{
	  if (bufpos >= bufmax)
	    {
	      bufmax += 100;
	      buffer = xrealloc (buffer, bufmax);
	    }
	  buffer[bufpos++] = c;
	  c = phase4_getc ();
	  switch (c)
	    {
	    case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
	    case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
	    case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
	    case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
	    case 'Y': case 'Z':
	    case '_':
	    case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
	    case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
	    case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
	    case 's': case 't': case 'u': case 'v': case 'w': case 'x':
	    case 'y': case 'z':
	    case '0': case '1': case '2': case '3': case '4':
	    case '5': case '6': case '7': case '8': case '9':
	      continue;

	    default:
	      phase4_ungetc (c);
	      break;
	    }
	  break;
	}
      if (bufpos >= bufmax)
	{
	  bufmax += 100;
	  buffer = xrealloc (buffer, bufmax);
	}
      buffer[bufpos] = 0;
      tp->string = xstrdup (buffer);
      tp->type = token_type_name;
      return;

    case '.':
      c = phase4_getc ();
      phase4_ungetc (c);
      switch (c)
	{
	default:
	  tp->type = token_type_symbol;
	  return;

	case '0': case '1': case '2': case '3': case '4':
	case '5': case '6': case '7': case '8': case '9':
	  c = '.';
	  break;
	}
      /* FALLTHROUGH */

    case '0': case '1': case '2': case '3': case '4':
    case '5': case '6': case '7': case '8': case '9':
      /* The preprocessing number token is more "generous" than the C
	 number tokens.  This is mostly due to token pasting (another
	 thing we can ignore here).  */
      bufpos = 0;
      while (1)
	{
	  if (bufpos >= bufmax)
	    {
	      bufmax += 100;
	      buffer = xrealloc (buffer, bufmax);
	    }
	  buffer[bufpos++] = c;
	  c = phase4_getc ();
	  switch (c)
	    {
	    case 'e':
	    case 'E':
	      if (bufpos >= bufmax)
		{
		  bufmax += 100;
		  buffer = xrealloc (buffer, bufmax);
		}
	      buffer[bufpos++] = c;
	      c = phase4_getc ();
	      if (c != '+' || c != '-')
		{
		  phase4_ungetc (c);
		  break;
		}
	      continue;

	    case 'A': case 'B': case 'C': case 'D':           case 'F':
	    case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
	    case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
	    case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
	    case 'Y': case 'Z':
	    case 'a': case 'b': case 'c': case 'd':           case 'f':
	    case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
	    case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
	    case 's': case 't': case 'u': case 'v': case 'w': case 'x':
	    case 'y': case 'z':
	    case '0': case '1': case '2': case '3': case '4':
	    case '5': case '6': case '7': case '8': case '9':
	    case '.':
	      continue;

	    default:
	      phase4_ungetc (c);
	      break;
	    }
	  break;
	}
      if (bufpos >= bufmax)
	{
	  bufmax += 100;
	  buffer = xrealloc (buffer, bufmax);
	}
      buffer[bufpos] = 0;
      tp->type = token_type_number;
      tp->number = atol (buffer);
      return;

    case '\'':
      /* We could worry about the 'L' before wide character constants,
	 but ignoring it has no effect unless one of the keywords is
	 "L".  Just pretend it won't happen.  Also, we don't need to
	 remember the character constant.  */
      while (1)
	{
	  c = phase7_getc ();
	  if (c == P7_NEWLINE)
	    {
	      error_with_progname = 0;
	      error (0, 0, _("%s:%d: warning: unterminated character constant"),
	        logical_file_name, line_number - 1);
	      error_with_progname = 1;
	      phase7_ungetc ('\n');
	      break;
	    }
	  if (c == EOF || c == P7_QUOTE)
	    break;
	}
      tp->type = token_type_character_constant;
      return;

    case '"':
      /* We could worry about the 'L' before wide string constants,
	 but since gettext's argument is not a wide character string,
	 let the compiler complain about the argument not matching the
	 prototype.  Just pretend it won't happen.  */
      bufpos = 0;
      while (1)
	{
	  c = phase7_getc ();
	  if (c == P7_NEWLINE)
	    {
	      error_with_progname = 0;
	      error (0, 0, _("%s:%d: warning: unterminated string literal"),
	        logical_file_name, line_number - 1);
	      error_with_progname = 1;
	      phase7_ungetc ('\n');
	      break;
	    }
	  if (c == EOF || c == P7_QUOTES)
	    break;
	  if (c == P7_QUOTE)
	    c = '\'';
	  if (bufpos >= bufmax)
	    {
	      bufmax += 100;
	      buffer = xrealloc (buffer, bufmax);
	    }
	  buffer[bufpos++] = c;
	}
      if (bufpos >= bufmax)
	{
	  bufmax += 100;
	  buffer = xrealloc (buffer, bufmax);
	}
      buffer[bufpos] = 0;
      tp->type = token_type_string_literal;
      tp->string = xstrdup (buffer);
      return;

    case '(':
      tp->type = token_type_lparen;
      return;

    case ')':
      tp->type = token_type_rparen;
      return;

    case ',':
      tp->type = token_type_comma;
      return;

    case '#':
      tp->type = token_type_hash;
      return;

    default:
      /* We could carefully recognize each of the 2 and 3 character
        operators, but it is not necessary, as we only need to recognize
        gettext invocations.  Don't bother.  */
      tp->type = token_type_symbol;
      return;
    }
}


static void
phase5_unget (tp)
     token_ty *tp;
{
  if (tp->type != token_type_eof)
    phase5_pushback[phase5_pushback_length++] = *tp;
}


/* X. Recognize a leading # symbol.  Leave leading hash as a hash, but
   turn hash in the middle of a line into a plain symbol token.  This
   makes the phase 6 easier.  */

static void
phaseX_get (tp)
     token_ty *tp;
{
  static int middle;	/* 0 at the beginning of a line, 1 otherwise.  */

  phase5_get (tp);

  if (tp->type == token_type_eoln || tp->type == token_type_eof)
    middle = 0;
  else
    {
      if (middle)
	{
	  /* Turn hash in the middle of a line into a plain symbol token.  */
	  if (tp->type == token_type_hash)
	    tp->type = token_type_symbol;
	}
      else
	{
	  /* When we see leading whitespace followed by a hash sign,
	     discard the leading white space token.  The hash is all
	     phase 6 is interested in.  */
	  if (tp->type == token_type_white_space)
	    {
	      token_ty next;

	      phase5_get (&next);
	      if (next.type == token_type_hash)
		*tp = next;
	      else
		phase5_unget (&next);
	    }
	  middle = 1;
	}
    }
}


/* 6. Recognize and carry out directives (it also expands macros on
   non-directive lines, which we do not do here).  The only directive
   we care about are the #line and #define directive.  We throw all the
   others away.  */

/* Maximum used guaranteed to be < 4.  */
static token_ty phase6_pushback[4];
static int phase6_pushback_length;


static void
phase6_get (tp)
     token_ty *tp;
{
  static token_ty *buf;
  static int bufmax;
  int bufpos;
  int j;

  if (phase6_pushback_length)
    {
      *tp = phase6_pushback[--phase6_pushback_length];
      return;
    }
  while (1)
    {
      /* Get the next token.  If it is not a '#' at the beginning of a
	 line (ignoring whitespace), return immediately.  */
      phaseX_get (tp);
      if (tp->type != token_type_hash)
	return;

      /* Accumulate the rest of the directive in a buffer, until the
	 "define" keyword is seen or until end of line.  */
      bufpos = 0;
      while (1)
	{
	  phaseX_get (tp);
	  if (tp->type == token_type_eoln || tp->type == token_type_eof)
	    break;

	  /* Before the "define" keyword and inside other directives
	     white space is irrelevant.  So just throw it away.  */
	  if (tp->type != token_type_white_space)
	    {
	      /* If it is a #define directive, return immediately,
		 thus treating the body of the #define directive like
		 normal input.  */
	      if (bufpos == 0
		  && tp->type == token_type_name
		  && strcmp (tp->string, "define") == 0)
		return;

	      /* Accumulate.  */
	      if (bufpos >= bufmax)
		{
		  bufmax += 100;
		  buf = xrealloc (buf, bufmax * sizeof (buf[0]));
		}
	      buf[bufpos++] = *tp;
	    }
	}

      /* If it is a #line directive, with no macros to expand, act on
	 it.  Ignore all other directives.  */
      if (bufpos >= 3 && buf[0].type == token_type_name
	  && strcmp (buf[0].string, "line") == 0
	  && buf[1].type == token_type_number
	  && buf[2].type == token_type_string_literal)
	{
	  free (logical_file_name);
	  logical_file_name = xstrdup (buf[2].string);
	  line_number = buf[1].number;
	}
      if (bufpos >= 2 && buf[0].type == token_type_number
	  && buf[1].type == token_type_string_literal)
	{
	  free (logical_file_name);
	  logical_file_name = xstrdup (buf[1].string);
	  line_number = buf[0].number;
	}

      /* Release the storage held by the directive.  */
      for (j = 0; j < bufpos; ++j)
	{
	  switch (buf[j].type)
	    {
	    case token_type_name:
	    case token_type_string_literal:
	      free (buf[j].string);
	      break;

	    default:
	      break;
	    }
	}

      /* We must reset the selected comments.  */
      xgettext_lex_comment_reset ();
    }
}


static void
phase6_unget (tp)
     token_ty *tp;
{
  if (tp->type != token_type_eof)
    phase6_pushback[phase6_pushback_length++] = *tp;
}


/* 8. Concatenate adjacent string literals to form single string
   literals (because we don't expand macros, there are a few things we
   will miss).  */

static void
phase8_get (tp)
     token_ty *tp;
{
  phase6_get (tp);
  if (tp->type != token_type_string_literal)
    return;
  while (1)
    {
      token_ty tmp;
      size_t len;

      phase6_get (&tmp);
      if (tmp.type == token_type_white_space)
	continue;
      if (tmp.type == token_type_eoln)
	continue;
      if (tmp.type != token_type_string_literal)
	{
	  phase6_unget (&tmp);
	  return;
	}
      len = strlen (tp->string);
      tp->string = xrealloc (tp->string, len + strlen (tmp.string) + 1);
      strcpy (tp->string + len, tmp.string);
      free (tmp.string);
    }
}


/* 9. Convert the remaining preprocessing tokens to C tokens and
   discards any white space from the translation unit.  */

void
xgettext_lex (tp)
     xgettext_token_ty *tp;
{
  while (1)
    {
      token_ty token;
      void *keyword_value;

      phase8_get (&token);
      switch (token.type)
	{
	case token_type_eof:
	  newline_count = 0;
	  last_comment_line = -1;
	  last_non_comment_line = -1;
	  tp->type = xgettext_token_type_eof;
	  return;

	case token_type_white_space:
	  break;

	case token_type_eoln:
	  /* We have to track the last occurrence of a string.  One
	     mode of xgettext allows to group an extracted message
	     with a comment for documentation.  The rule which states
	     which comment is assumed to be grouped with the message
	     says it should immediately precede it.  Our
	     interpretation: between the last line of the comment and
	     the line in which the keyword is found must be no line
	     with non-white space tokens.  */
	  ++newline_count;
	  if (last_non_comment_line > last_comment_line)
	    xgettext_lex_comment_reset ();
	  break;

	case token_type_name:
	  last_non_comment_line = newline_count;

	  if (default_keywords)
	    {
	      xgettext_lex_keyword ("gettext");
	      xgettext_lex_keyword ("dgettext:2");
	      xgettext_lex_keyword ("dcgettext:2");
	      xgettext_lex_keyword ("ngettext:1,2");
	      xgettext_lex_keyword ("dngettext:2,3");
	      xgettext_lex_keyword ("dcngettext:2,3");
	      xgettext_lex_keyword ("gettext_noop");
	      default_keywords = 0;
	    }

	  if (find_entry (&keywords, token.string, strlen (token.string) + 1,
			  &keyword_value)
	      == 0)
	    {
	      tp->type = xgettext_token_type_keyword;
	      tp->argnum1 = (int) (long) keyword_value & ((1 << 10) - 1);
	      tp->argnum2 = (int) (long) keyword_value >> 10;
	      tp->line_number = token.line_number;
	      tp->file_name = logical_file_name;
	    }
	  else
	    tp->type = xgettext_token_type_symbol;
	  free (token.string);
	  return;

	case token_type_lparen:
	  last_non_comment_line = newline_count;

	  tp->type = xgettext_token_type_lparen;
	  return;

	case token_type_rparen:
	  last_non_comment_line = newline_count;

	  tp->type = xgettext_token_type_rparen;
	  return;

	case token_type_comma:
	  last_non_comment_line = newline_count;

	  tp->type = xgettext_token_type_comma;
	  return;

	case token_type_string_literal:
	  last_non_comment_line = newline_count;

	  tp->type = xgettext_token_type_string_literal;
	  tp->string = token.string;
	  tp->line_number = token.line_number;
	  tp->file_name = logical_file_name;
	  return;

	default:
	  last_non_comment_line = newline_count;

	  tp->type = xgettext_token_type_symbol;
	  return;
	}
    }
}


void
xgettext_lex_keyword (name)
     const char *name;
{
  if (name == NULL)
    default_keywords = 0;
  else
    {
      int argnum1;
      int argnum2;
      size_t len;
      char *sp;

      if (keywords.table == NULL)
	init_hash (&keywords, 100);

      sp = strchr (name, ':');
      if (sp)
	{
	  /* Make a temporary copy of 'name' up to 'sp', because
	     insert_entry() expects a NUL terminated string.  */
	  char *name_copy;

	  len = sp - name;
	  name_copy = (char *) alloca (len + 1);
	  memcpy (name_copy, name, len);
	  name_copy[len] = '\0';
	  name = name_copy;

	  sp++;
	  argnum1 = strtol (sp, &sp, 10);
	  if (*sp == ',')
	    {
	      sp++;
	      argnum2 = strtol (sp, &sp, 10);
	    }
	  else
	    argnum2 = 0;
	}
      else
	{
	  len = strlen (name);

	  argnum1 = 1;
	  argnum2 = 0;
	}

      insert_entry (&keywords, name, len + 1,
		    (void *) (long) (argnum1 + (argnum2 << 10)));
    }
}


int
xgettext_any_keywords ()
{
  return (keywords.filled > 0) || default_keywords;
}


const char *
xgettext_lex_comment (n)
     size_t n;
{
  if (comment == NULL || n >= comment->nitems)
    return NULL;
  return comment->item[n];
}


void
xgettext_lex_comment_reset ()
{
  if (comment != NULL)
    {
      string_list_free (comment);
      comment = NULL;
    }
}


void
xgettext_lex_trigraphs ()
{
  trigraphs = 1;
}