summaryrefslogtreecommitdiff
path: root/usr.bin/unifdef/unifdef.c
blob: c96f11f7eea8a52f48e466efe62bed44134712d7 (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
/*	$OpenBSD: unifdef.c,v 1.15 2012/03/04 04:05:15 fgsch Exp $	*/
/*
 * Copyright (c) 2002, 2003 Tony Finch <dot@dotat.at>
 * Copyright (c) 1985, 1993
 *	The Regents of the University of California.  All rights reserved.
 *
 * This code is derived from software contributed to Berkeley by
 * Dave Yost. Support for #if and #elif was added by Tony Finch.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 */

/*
 * unifdef - remove ifdef'ed lines
 *
 *  Wishlist:
 *      provide an option which will append the name of the
 *        appropriate symbol after #else's and #endif's
 *      provide an option which will check symbols after
 *        #else's and #endif's to see that they match their
 *        corresponding #ifdef or #ifndef
 *      generate #line directives in place of deleted code
 *
 *   The first two items above require better buffer handling, which would
 *     also make it possible to handle all "dodgy" directives correctly.
 */

#include <ctype.h>
#include <err.h>
#include <stdarg.h>
#include <stdbool.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>

/* types of input lines: */
typedef enum {
	LT_TRUEI,		/* a true #if with ignore flag */
	LT_FALSEI,		/* a false #if with ignore flag */
	LT_IF,			/* an unknown #if */
	LT_TRUE,		/* a true #if */
	LT_FALSE,		/* a false #if */
	LT_ELIF,		/* an unknown #elif */
	LT_ELTRUE,		/* a true #elif */
	LT_ELFALSE,		/* a false #elif */
	LT_ELSE,		/* #else */
	LT_ENDIF,		/* #endif */
	LT_DODGY,		/* flag: directive is not on one line */
	LT_DODGY_LAST = LT_DODGY + LT_ENDIF,
	LT_PLAIN,		/* ordinary line */
	LT_EOF,			/* end of file */
	LT_COUNT
} Linetype;

static char const * const linetype_name[] = {
	"TRUEI", "FALSEI", "IF", "TRUE", "FALSE",
	"ELIF", "ELTRUE", "ELFALSE", "ELSE", "ENDIF",
	"DODGY TRUEI", "DODGY FALSEI",
	"DODGY IF", "DODGY TRUE", "DODGY FALSE",
	"DODGY ELIF", "DODGY ELTRUE", "DODGY ELFALSE",
	"DODGY ELSE", "DODGY ENDIF",
	"PLAIN", "EOF"
};

/* state of #if processing */
typedef enum {
	IS_OUTSIDE,
	IS_FALSE_PREFIX,	/* false #if followed by false #elifs */
	IS_TRUE_PREFIX,		/* first non-false #(el)if is true */
	IS_PASS_MIDDLE,		/* first non-false #(el)if is unknown */
	IS_FALSE_MIDDLE,	/* a false #elif after a pass state */
	IS_TRUE_MIDDLE,		/* a true #elif after a pass state */
	IS_PASS_ELSE,		/* an else after a pass state */
	IS_FALSE_ELSE,		/* an else after a true state */
	IS_TRUE_ELSE,		/* an else after only false states */
	IS_FALSE_TRAILER,	/* #elifs after a true are false */
	IS_COUNT
} Ifstate;

static char const * const ifstate_name[] = {
	"OUTSIDE", "FALSE_PREFIX", "TRUE_PREFIX",
	"PASS_MIDDLE", "FALSE_MIDDLE", "TRUE_MIDDLE",
	"PASS_ELSE", "FALSE_ELSE", "TRUE_ELSE",
	"FALSE_TRAILER"
};

/* state of comment parser */
typedef enum {
	NO_COMMENT = false,	/* outside a comment */
	C_COMMENT,		/* in a comment like this one */
	CXX_COMMENT,		/* between // and end of line */
	STARTING_COMMENT,	/* just after slash-backslash-newline */
	FINISHING_COMMENT	/* star-backslash-newline in a C comment */
} Comment_state;

static char const * const comment_name[] = {
	"NO", "C", "CXX", "STARTING", "FINISHING"
};

/* state of preprocessor line parser */
typedef enum {
	LS_START,		/* only space and comments on this line */
	LS_HASH,		/* only space, comments, and a hash */
	LS_DIRTY		/* this line can't be a preprocessor line */
} Line_state;

static char const * const linestate_name[] = {
	"START", "HASH", "DIRTY"
};

/*
 * Minimum translation limits from ISO/IEC 9899:1999 5.2.4.1
 */
#define	MAXDEPTH        64			/* maximum #if nesting */
#define	MAXLINE         4096			/* maximum length of line */
#define	MAXSYMS         4096			/* maximum number of symbols */

/*
 * Sometimes when editing a keyword the replacement text is longer, so
 * we leave some space at the end of the tline buffer to accommodate this.
 */
#define	EDITSLOP        10

/*
 * Globals.
 */

static bool             complement;		/* -c: do the complement */
static bool             debugging;		/* -d: debugging reports */
static bool             iocccok;		/* -e: fewer IOCCC errors */
static bool             killconsts;		/* -k: eval constant #ifs */
static bool             lnblank;		/* -l: blank deleted lines */
static bool             symlist;		/* -s: output symbol list */
static bool             text;			/* -t: this is a text file */

static const char      *symname[MAXSYMS];	/* symbol name */
static const char      *value[MAXSYMS];		/* -Dsym=value */
static bool             ignore[MAXSYMS];	/* -iDsym or -iUsym */
static int              nsyms;			/* number of symbols */

static FILE            *input;			/* input file pointer */
static const char      *filename;		/* input file name */
static int              linenum;		/* current line number */

static char             tline[MAXLINE+EDITSLOP];/* input buffer plus space */
static char            *keyword;		/* used for editing #elif's */

static Comment_state    incomment;		/* comment parser state */
static Line_state       linestate;		/* #if line parser state */
static Ifstate          ifstate[MAXDEPTH];	/* #if processor state */
static bool             ignoring[MAXDEPTH];	/* ignore comments state */
static int              stifline[MAXDEPTH];	/* start of current #if */
static int              depth;			/* current #if nesting */
static bool             keepthis;		/* don't delete constant #if */

static int              exitstat;		/* program exit status */

static void             addsym(bool, bool, char *);
static void             debug(const char *, ...);
static void             error(const char *);
static int              findsym(const char *);
static void             flushline(bool);
static Linetype         get_line(void);
static Linetype         ifeval(const char **);
static void             ignoreoff(void);
static void             ignoreon(void);
static void             keywordedit(const char *);
static void             nest(void);
static void             process(void);
static const char      *skipcomment(const char *);
static const char      *skipsym(const char *);
static void             state(Ifstate);
static int              strlcmp(const char *, const char *, size_t);
static void             usage(void);

#define endsym(c) (!isalpha((unsigned char)c) && !isdigit((unsigned char)c) && c != '_')

/*
 * The main program.
 */
int
main(int argc, char *argv[])
{
	int opt;

	while ((opt = getopt(argc, argv, "i:D:U:I:cdeklst")) != -1)
		switch (opt) {
		case 'i': /* treat stuff controlled by these symbols as text */
			/*
			 * For strict backwards-compatibility the U or D
			 * should be immediately after the -i but it doesn't
			 * matter much if we relax that requirement.
			 */
			opt = *optarg++;
			if (opt == 'D')
				addsym(true, true, optarg);
			else if (opt == 'U')
				addsym(true, false, optarg);
			else
				usage();
			break;
		case 'D': /* define a symbol */
			addsym(false, true, optarg);
			break;
		case 'U': /* undef a symbol */
			addsym(false, false, optarg);
			break;
		case 'I':
			/* no-op for compatibility with cpp */
			break;
		case 'c': /* treat -D as -U and vice versa */
			complement = true;
			break;
		case 'd':
			debugging = true;
			break;
		case 'e': /* fewer errors from dodgy lines */
			iocccok = true;
			break;
		case 'k': /* process constant #ifs */
			killconsts = true;
			break;
		case 'l': /* blank deleted lines instead of omitting them */
			lnblank = true;
			break;
		case 's': /* only output list of symbols that control #ifs */
			symlist = true;
			break;
		case 't': /* don't parse C comments */
			text = true;
			break;
		default:
			usage();
		}
	argc -= optind;
	argv += optind;
	if (nsyms == 0 && !symlist) {
		warnx("must -D or -U at least one symbol");
		usage();
	}
	if (argc > 1) {
		errx(2, "can only do one file");
	} else if (argc == 1 && strcmp(*argv, "-") != 0) {
		filename = *argv;
		if ((input = fopen(filename, "r")) != NULL) {
			process();
			(void) fclose(input);
		} else
			err(2, "can't open %s", *argv);
	} else {
		filename = "[stdin]";
		input = stdin;
		process();
	}

	exit(exitstat);
}

static void
usage(void)
{
	fprintf(stderr,
	    "usage: unifdef [-ceklst] [-Dsym[=val]] [-Ipath] [-iDsym[=val]] "
	    "[-iUsym] [-Usym]\n"
	    "\t[file]\n");
	exit(2);
}

/*
 * A state transition function alters the global #if processing state
 * in a particular way. The table below is indexed by the current
 * processing state and the type of the current line. A NULL entry
 * indicates that processing is complete.
 *
 * Nesting is handled by keeping a stack of states; some transition
 * functions increase or decrease the depth. They also maintain the
 * ignore state on a stack. In some complicated cases they have to
 * alter the preprocessor directive, as follows.
 *
 * When we have processed a group that starts off with a known-false
 * #if/#elif sequence (which has therefore been deleted) followed by a
 * #elif that we don't understand and therefore must keep, we edit the
 * latter into a #if to keep the nesting correct.
 *
 * When we find a true #elif in a group, the following block will
 * always be kept and the rest of the sequence after the next #elif or
 * #else will be discarded. We edit the #elif into a #else and the
 * following directive to #endif since this has the desired behaviour.
 *
 * "Dodgy" directives are split across multiple lines, the most common
 * example being a multi-line comment hanging off the right of the
 * directive. We can handle them correctly only if there is no change
 * from printing to dropping (or vice versa) caused by that directive.
 * If the directive is the first of a group we have a choice between
 * failing with an error, or passing it through unchanged instead of
 * evaluating it. The latter is not the default to avoid questions from
 * users about unifdef unexpectedly leaving behind preprocessor directives.
 */
typedef void state_fn(void);

/* report an error */
static void
Eelif(void)
{
	error("Inappropriate #elif");
}

static void
Eelse(void)
{
	error("Inappropriate #else");
}

static void
Eendif(void)
{
	error("Inappropriate #endif");
}

static void
Eeof(void)
{
	error("Premature EOF");
}

static void
Eioccc(void)
{
	error("Obfuscated preprocessor control line");
}

/* plain line handling */
static void
print(void)
{
	flushline(true);
}

static void
drop(void)
{
	flushline(false);
}

/* output lacks group's start line */
static void
Strue(void)
{
	drop();
	ignoreoff();
	state(IS_TRUE_PREFIX);
}

static void
Sfalse(void)
{
	drop();
	ignoreoff();
	state(IS_FALSE_PREFIX);
}

static void
Selse(void)
{
	drop();
	state(IS_TRUE_ELSE);
}

/* print/pass this block */
static void
Pelif(void)
{
	print();
	ignoreoff();
	state(IS_PASS_MIDDLE);
}

static void
Pelse(void)
{
	print();
	state(IS_PASS_ELSE);
}

static void
Pendif(void)
{
	print();
	--depth;
}

/* discard this block */
static void
Dfalse(void)
{
	drop();
	ignoreoff();
	state(IS_FALSE_TRAILER);
}

static void
Delif(void)
{
	drop();
	ignoreoff();
	state(IS_FALSE_MIDDLE);
}

static void
Delse(void)
{
	drop();
	state(IS_FALSE_ELSE);
}

static void
Dendif(void)
{
	drop();
	--depth;
}

/* first line of group */
static void
Fdrop(void)
{
	nest();
	Dfalse();
}

static void
Fpass(void)
{
	nest();
	Pelif();
}

static void
Ftrue(void)
{
	nest();
	Strue();
}

static void
Ffalse(void)
{
	nest();
	Sfalse();
}

/* variable pedantry for obfuscated lines */
static void
Oiffy(void)
{
	if (iocccok)
		Fpass();
	else
		Eioccc();
	ignoreon();
}

static void
Oif(void)
{
	if (iocccok)
		Fpass();
	else
		Eioccc();
}

static void
Oelif(void)
{
	if (iocccok)
		Pelif();
	else
		Eioccc();
}

/* ignore comments in this block */
static void
Idrop(void)
{
	Fdrop();
	ignoreon();
}

static void
Itrue(void)
{
	Ftrue();
	ignoreon();
}

static void
Ifalse(void)
{
	Ffalse();
	ignoreon();
}

/* edit this line */
static void
Mpass (void)
{
	strncpy(keyword, "if  ", 4);
	Pelif();
}

static void
Mtrue (void)
{
	keywordedit("else\n");
	state(IS_TRUE_MIDDLE);
}

static void
Melif (void)
{
	keywordedit("endif\n");
	state(IS_FALSE_TRAILER);
}

static void
Melse (void)
{
	keywordedit("endif\n");
	state(IS_FALSE_ELSE);
}

static state_fn * const trans_table[IS_COUNT][LT_COUNT] = {
/* IS_OUTSIDE */
{ Itrue, Ifalse,Fpass, Ftrue, Ffalse,Eelif, Eelif, Eelif, Eelse, Eendif,
  Oiffy, Oiffy, Fpass, Oif,   Oif,   Eelif, Eelif, Eelif, Eelse, Eendif,
  print, NULL },
/* IS_FALSE_PREFIX */
{ Idrop, Idrop, Fdrop, Fdrop, Fdrop, Mpass, Strue, Sfalse,Selse, Dendif,
  Idrop, Idrop, Fdrop, Fdrop, Fdrop, Mpass, Eioccc,Eioccc,Eioccc,Eioccc,
  drop,  Eeof },
/* IS_TRUE_PREFIX */
{ Itrue, Ifalse,Fpass, Ftrue, Ffalse,Dfalse,Dfalse,Dfalse,Delse, Dendif,
  Oiffy, Oiffy, Fpass, Oif,   Oif,   Eioccc,Eioccc,Eioccc,Eioccc,Eioccc,
  print, Eeof },
/* IS_PASS_MIDDLE */
{ Itrue, Ifalse,Fpass, Ftrue, Ffalse,Pelif, Mtrue, Delif, Pelse, Pendif,
  Oiffy, Oiffy, Fpass, Oif,   Oif,   Pelif, Oelif, Oelif, Pelse, Pendif,
  print, Eeof },
/* IS_FALSE_MIDDLE */
{ Idrop, Idrop, Fdrop, Fdrop, Fdrop, Pelif, Mtrue, Delif, Pelse, Pendif,
  Idrop, Idrop, Fdrop, Fdrop, Fdrop, Eioccc,Eioccc,Eioccc,Eioccc,Eioccc,
  drop,  Eeof },
/* IS_TRUE_MIDDLE */
{ Itrue, Ifalse,Fpass, Ftrue, Ffalse,Melif, Melif, Melif, Melse, Pendif,
  Oiffy, Oiffy, Fpass, Oif,   Oif,   Eioccc,Eioccc,Eioccc,Eioccc,Pendif,
  print, Eeof },
/* IS_PASS_ELSE */
{ Itrue, Ifalse,Fpass, Ftrue, Ffalse,Eelif, Eelif, Eelif, Eelse, Pendif,
  Oiffy, Oiffy, Fpass, Oif,   Oif,   Eelif, Eelif, Eelif, Eelse, Pendif,
  print, Eeof },
/* IS_FALSE_ELSE */
{ Idrop, Idrop, Fdrop, Fdrop, Fdrop, Eelif, Eelif, Eelif, Eelse, Dendif,
  Idrop, Idrop, Fdrop, Fdrop, Fdrop, Eelif, Eelif, Eelif, Eelse, Eioccc,
  drop,  Eeof },
/* IS_TRUE_ELSE */
{ Itrue, Ifalse,Fpass, Ftrue, Ffalse,Eelif, Eelif, Eelif, Eelse, Dendif,
  Oiffy, Oiffy, Fpass, Oif,   Oif,   Eelif, Eelif, Eelif, Eelse, Eioccc,
  print, Eeof },
/* IS_FALSE_TRAILER */
{ Idrop, Idrop, Fdrop, Fdrop, Fdrop, Dfalse,Dfalse,Dfalse,Delse, Dendif,
  Idrop, Idrop, Fdrop, Fdrop, Fdrop, Dfalse,Dfalse,Dfalse,Delse, Eioccc,
  drop,  Eeof }
/*TRUEI  FALSEI IF     TRUE   FALSE  ELIF   ELTRUE ELFALSE ELSE  ENDIF
  TRUEI  FALSEI IF     TRUE   FALSE  ELIF   ELTRUE ELFALSE ELSE  ENDIF (DODGY)
  PLAIN  EOF */
};

/*
 * State machine utility functions
 */
static void
ignoreoff(void)
{
	ignoring[depth] = ignoring[depth-1];
}

static void
ignoreon(void)
{
	ignoring[depth] = true;
}

static void
keywordedit(const char *replacement)
{
	strlcpy(keyword, replacement, tline + sizeof(tline) - keyword);
	print();
}

static void
nest(void)
{
	depth += 1;
	if (depth >= MAXDEPTH)
		error("Too many levels of nesting");
	stifline[depth] = linenum;
}

static void
state(Ifstate is)
{
	ifstate[depth] = is;
}

/*
 * Write a line to the output or not, according to command line options.
 */
static void
flushline(bool keep)
{
	if (symlist)
		return;
	if (keep ^ complement)
		fputs(tline, stdout);
	else {
		if (lnblank)
			putc('\n', stdout);
		exitstat = 1;
	}
}

/*
 * The driver for the state machine.
 */
static void
process(void)
{
	Linetype lineval;
	state_fn *trans;

	for (;;) {
		linenum++;
		lineval = get_line();
		trans = trans_table[ifstate[depth]][lineval];
		if (trans == NULL)
			break;
		trans();
		debug("process %s -> %s depth %d",
		    linetype_name[lineval],
		    ifstate_name[ifstate[depth]], depth);
	}
	if (incomment)
		error("EOF in comment");
}

/*
 * Parse a line and determine its type. We keep the preprocessor line
 * parser state between calls in a global variable.
 */
static Linetype
get_line(void)
{
	const char *cp;
	int cursym;
	int kwlen;
	Linetype retval;
	Comment_state wascomment;

	if (fgets(tline, MAXLINE, input) == NULL)
		return (LT_EOF);
	retval = LT_PLAIN;
	wascomment = incomment;
	cp = skipcomment(tline);
	if (linestate == LS_START) {
		if (*cp == '#') {
			linestate = LS_HASH;
			cp = skipcomment(cp + 1);
		} else if (*cp != '\0')
			linestate = LS_DIRTY;
	}
	if (!incomment && linestate == LS_HASH) {
		keyword = tline + (cp - tline);
		cp = skipsym(cp);
		kwlen = cp - keyword;
		/* no way can we deal with a continuation inside a keyword */
		if (strncmp(cp, "\\\n", 2) == 0)
			Eioccc();
		if (strlcmp("ifdef", keyword, kwlen) == 0 ||
		    strlcmp("ifndef", keyword, kwlen) == 0) {
			cp = skipcomment(cp);
			if ((cursym = findsym(cp)) < 0)
				retval = LT_IF;
			else {
				retval = (keyword[2] == 'n')
				    ? LT_FALSE : LT_TRUE;
				if (value[cursym] == NULL)
					retval = (retval == LT_TRUE)
					    ? LT_FALSE : LT_TRUE;
				if (ignore[cursym])
					retval = (retval == LT_TRUE)
					    ? LT_TRUEI : LT_FALSEI;
			}
			cp = skipsym(cp);
		} else if (strlcmp("if", keyword, kwlen) == 0)
			retval = ifeval(&cp);
		else if (strlcmp("elif", keyword, kwlen) == 0)
			retval = ifeval(&cp) - LT_IF + LT_ELIF;
		else if (strlcmp("else", keyword, kwlen) == 0)
			retval = LT_ELSE;
		else if (strlcmp("endif", keyword, kwlen) == 0)
			retval = LT_ENDIF;
		else {
			linestate = LS_DIRTY;
			retval = LT_PLAIN;
		}
		cp = skipcomment(cp);
		if (*cp != '\0') {
			linestate = LS_DIRTY;
			if (retval == LT_TRUE || retval == LT_FALSE ||
			    retval == LT_TRUEI || retval == LT_FALSEI)
				retval = LT_IF;
			if (retval == LT_ELTRUE || retval == LT_ELFALSE)
				retval = LT_ELIF;
		}
		if (retval != LT_PLAIN && (wascomment || incomment)) {
			retval += LT_DODGY;
			if (incomment)
				linestate = LS_DIRTY;
		}
		/* skipcomment should have changed the state */
		if (linestate == LS_HASH)
			abort(); /* bug */
	}
	if (linestate == LS_DIRTY) {
		while (*cp != '\0')
			cp = skipcomment(cp + 1);
	}
	debug("parser %s comment %s line",
	    comment_name[incomment], linestate_name[linestate]);
	return (retval);
}

/*
 * These are the operators that are supported by the expression
 * evaluator. Note that if support for division is added then we also
 * need short-circuiting booleans because of divide-by-zero.
 */
static int
op_lt(int a, int b)
{
	return (a < b);
}

static int
op_gt(int a, int b)
{
	return (a > b);
}

static int
op_le(int a, int b)
{
	return (a <= b);
}

static int
op_ge(int a, int b)
{
	return (a >= b);
}

static int
op_eq(int a, int b)
{
	return (a == b);
}

static int
op_ne(int a, int b)
{
	return (a != b);
}

static int
op_or(int a, int b)
{
	return (a || b);
}

static int
op_and(int a, int b)
{
	return (a && b);
}

/*
 * An evaluation function takes three arguments, as follows: (1) a pointer to
 * an element of the precedence table which lists the operators at the current
 * level of precedence; (2) a pointer to an integer which will receive the
 * value of the expression; and (3) a pointer to a char* that points to the
 * expression to be evaluated and that is updated to the end of the expression
 * when evaluation is complete. The function returns LT_FALSE if the value of
 * the expression is zero, LT_TRUE if it is non-zero, or LT_IF if the
 * expression could not be evaluated.
 */
struct ops;

typedef Linetype eval_fn(const struct ops *, int *, const char **);

static eval_fn eval_table, eval_unary;

/*
 * The precedence table. Expressions involving binary operators are evaluated
 * in a table-driven way by eval_table. When it evaluates a subexpression it
 * calls the inner function with its first argument pointing to the next
 * element of the table. Innermost expressions have special non-table-driven
 * handling.
 */
static const struct ops {
	eval_fn *inner;
	struct op {
		const char *str;
		int (*fn)(int, int);
	} op[5];
} eval_ops[] = {
	{ eval_table, { { "||", op_or } } },
	{ eval_table, { { "&&", op_and } } },
	{ eval_table, { { "==", op_eq },
			{ "!=", op_ne } } },
	{ eval_unary, { { "<=", op_le },
			{ ">=", op_ge },
			{ "<", op_lt },
			{ ">", op_gt } } }
};

/*
 * Function for evaluating the innermost parts of expressions,
 * viz. !expr (expr) defined(symbol) symbol number
 * We reset the keepthis flag when we find a non-constant subexpression.
 */
static Linetype
eval_unary(const struct ops *ops, int *valp, const char **cpp)
{
	const char *cp;
	char *ep;
	int sym;

	cp = skipcomment(*cpp);
	if (*cp == '!') {
		debug("eval%d !", ops - eval_ops);
		cp++;
		if (eval_unary(ops, valp, &cp) == LT_IF)
			return (LT_IF);
		*valp = !*valp;
	} else if (*cp == '(') {
		cp++;
		debug("eval%d (", ops - eval_ops);
		if (eval_table(eval_ops, valp, &cp) == LT_IF)
			return (LT_IF);
		cp = skipcomment(cp);
		if (*cp++ != ')')
			return (LT_IF);
	} else if (isdigit((unsigned char)*cp)) {
		debug("eval%d number", ops - eval_ops);
		*valp = strtol(cp, &ep, 0);
		cp = skipsym(cp);
	} else if (strncmp(cp, "defined", 7) == 0 && endsym(cp[7])) {
		cp = skipcomment(cp+7);
		debug("eval%d defined", ops - eval_ops);
		if (*cp++ != '(')
			return (LT_IF);
		cp = skipcomment(cp);
		sym = findsym(cp);
		if (sym < 0 && !symlist)
			return (LT_IF);
		*valp = (value[sym] != NULL);
		cp = skipsym(cp);
		cp = skipcomment(cp);
		if (*cp++ != ')')
			return (LT_IF);
		keepthis = false;
	} else if (!endsym(*cp)) {
		debug("eval%d symbol", ops - eval_ops);
		sym = findsym(cp);
		if (sym < 0 && !symlist)
			return (LT_IF);
		if (value[sym] == NULL)
			*valp = 0;
		else {
			*valp = strtol(value[sym], &ep, 0);
			if (*ep != '\0' || ep == value[sym])
				return (LT_IF);
		}
		cp = skipsym(cp);
		keepthis = false;
	} else
		return (LT_IF);

	*cpp = cp;
	debug("eval%d = %d", ops - eval_ops, *valp);
	return (*valp ? LT_TRUE : LT_FALSE);
}

/*
 * Table-driven evaluation of binary operators.
 */
static Linetype
eval_table(const struct ops *ops, int *valp, const char **cpp)
{
	const struct op *op;
	const char *cp;
	int val;

	debug("eval%d", ops - eval_ops);
	cp = *cpp;
	if (ops->inner(ops+1, valp, &cp) == LT_IF)
		return (LT_IF);
	for (;;) {
		cp = skipcomment(cp);
		for (op = ops->op; op->str != NULL; op++)
			if (strncmp(cp, op->str, strlen(op->str)) == 0)
				break;
		if (op->str == NULL)
			break;
		cp += strlen(op->str);
		debug("eval%d %s", ops - eval_ops, op->str);
		if (ops->inner(ops+1, &val, &cp) == LT_IF)
			return (LT_IF);
		*valp = op->fn(*valp, val);
	}

	*cpp = cp;
	debug("eval%d = %d", ops - eval_ops, *valp);
	return (*valp ? LT_TRUE : LT_FALSE);
}

/*
 * Evaluate the expression on a #if or #elif line. If we can work out
 * the result we return LT_TRUE or LT_FALSE accordingly, otherwise we
 * return just a generic LT_IF.
 */
static Linetype
ifeval(const char **cpp)
{
	int ret;
	int val;

	debug("eval %s", *cpp);
	keepthis = killconsts ? false : true;
	ret = eval_table(eval_ops, &val, cpp);
	return (keepthis ? LT_IF : ret);
}

/*
 * Skip over comments and stop at the next character position that is
 * not whitespace. Between calls we keep the comment state in the
 * global variable incomment, and we also adjust the global variable
 * linestate when we see a newline.
 * XXX: doesn't cope with the buffer splitting inside a state transition.
 */
static const char *
skipcomment(const char *cp)
{
	if (text || ignoring[depth]) {
		for (; isspace((unsigned char)*cp); cp++)
			if (*cp == '\n')
				linestate = LS_START;
		return (cp);
	}
	while (*cp != '\0')
		if (strncmp(cp, "\\\n", 2) == 0)
			cp += 2;
		else switch (incomment) {
		case NO_COMMENT:
			if (strncmp(cp, "/\\\n", 3) == 0) {
				incomment = STARTING_COMMENT;
				cp += 3;
			} else if (strncmp(cp, "/*", 2) == 0) {
				incomment = C_COMMENT;
				cp += 2;
			} else if (strncmp(cp, "//", 2) == 0) {
				incomment = CXX_COMMENT;
				cp += 2;
			} else if (strncmp(cp, "\n", 1) == 0) {
				linestate = LS_START;
				cp += 1;
			} else if (strchr(" \t", *cp) != NULL) {
				cp += 1;
			} else
				return (cp);
			continue;
		case CXX_COMMENT:
			if (strncmp(cp, "\n", 1) == 0) {
				incomment = NO_COMMENT;
				linestate = LS_START;
			}
			cp += 1;
			continue;
		case C_COMMENT:
			if (strncmp(cp, "*\\\n", 3) == 0) {
				incomment = FINISHING_COMMENT;
				cp += 3;
			} else if (strncmp(cp, "*/", 2) == 0) {
				incomment = NO_COMMENT;
				cp += 2;
			} else
				cp += 1;
			continue;
		case STARTING_COMMENT:
			if (*cp == '*') {
				incomment = C_COMMENT;
				cp += 1;
			} else if (*cp == '/') {
				incomment = CXX_COMMENT;
				cp += 1;
			} else {
				incomment = NO_COMMENT;
				linestate = LS_DIRTY;
			}
			continue;
		case FINISHING_COMMENT:
			if (*cp == '/') {
				incomment = NO_COMMENT;
				cp += 1;
			} else
				incomment = C_COMMENT;
			continue;
		default:
			/* bug */
			abort();
		}
	return (cp);
}

/*
 * Skip over an identifier.
 */
static const char *
skipsym(const char *cp)
{
	while (!endsym(*cp))
		++cp;
	return (cp);
}

/*
 * Look for the symbol in the symbol table. If it is found, we return
 * the symbol table index, else we return -1.
 */
static int
findsym(const char *str)
{
	const char *cp;
	int symind;

	cp = skipsym(str);
	if (cp == str)
		return (-1);
	if (symlist)
		printf("%.*s\n", (int)(cp-str), str);
	for (symind = 0; symind < nsyms; ++symind) {
		if (strlcmp(symname[symind], str, cp-str) == 0) {
			debug("findsym %s %s", symname[symind],
			    value[symind] ? value[symind] : "");
			return (symind);
		}
	}
	return (-1);
}

/*
 * Add a symbol to the symbol table.
 */
static void
addsym(bool ignorethis, bool definethis, char *sym)
{
	int symind;
	char *val;

	symind = findsym(sym);
	if (symind < 0) {
		if (nsyms >= MAXSYMS)
			errx(2, "too many symbols");
		symind = nsyms++;
	}
	symname[symind] = sym;
	ignore[symind] = ignorethis;
	val = sym + (skipsym(sym) - sym);
	if (definethis) {
		if (*val == '=') {
			value[symind] = val+1;
			*val = '\0';
		} else if (*val == '\0')
			value[symind] = "";
		else
			usage();
	} else {
		if (*val != '\0')
			usage();
		value[symind] = NULL;
	}
}

/*
 * Compare s with n characters of t.
 * The same as strncmp() except that it checks that s[n] == '\0'.
 */
static int
strlcmp(const char *s, const char *t, size_t n)
{
	while (n-- && *t != '\0')
		if (*s != *t)
			return ((unsigned char)*s - (unsigned char)*t);
		else
			++s, ++t;
	return ((unsigned char)*s);
}

/*
 * Diagnostics.
 */
static void
debug(const char *msg, ...)
{
	va_list ap;

	if (debugging) {
		va_start(ap, msg);
		vwarnx(msg, ap);
		va_end(ap);
	}
}

static void
error(const char *msg)
{
	if (depth == 0)
		warnx("%s: %d: %s", filename, linenum, msg);
	else
		warnx("%s: %d: %s (#if line %d depth %d)",
		    filename, linenum, msg, stifline[depth], depth);
	errx(2, "output may be truncated");
}