summaryrefslogtreecommitdiff
path: root/src/regress/lib/libc/regex
diff options
context:
space:
mode:
Diffstat (limited to '')
-rw-r--r--src/regress/lib/libc/regex/Makefile27
-rw-r--r--src/regress/lib/libc/regex/debug.c208
-rw-r--r--src/regress/lib/libc/regex/debug.ih17
-rw-r--r--src/regress/lib/libc/regex/main.c516
-rw-r--r--src/regress/lib/libc/regex/main.ih22
-rw-r--r--src/regress/lib/libc/regex/split.c317
-rw-r--r--src/regress/lib/libc/regex/t_exhaust.c183
-rw-r--r--src/regress/lib/libc/regex/tests612
8 files changed, 0 insertions, 1902 deletions
diff --git a/src/regress/lib/libc/regex/Makefile b/src/regress/lib/libc/regex/Makefile
deleted file mode 100644
index 5b9a080209..0000000000
--- a/src/regress/lib/libc/regex/Makefile
+++ /dev/null
@@ -1,27 +0,0 @@
1# $OpenBSD: Makefile,v 1.11 2017/04/16 16:01:05 kettenis Exp $
2# $NetBSD: Makefile,v 1.2 1995/02/16 19:38:45 cgd Exp $
3
4PROG= re
5SRCS= main.c split.c debug.c regcomp.c regerror.c regexec.c regfree.c
6.PATH: ${.CURDIR}/../../../../lib/libc/regex
7CLEANFILES += t_exhaust
8
9CFLAGS+= -I${.CURDIR}/../../../../lib/libc/regex -DREDEBUG -DPOSIX_MISTAKE
10CFLAGS+= -D'DEF_WEAK(x)=asm("")'
11
12TESTS= ${.CURDIR}/tests
13
14REGRESS_TARGETS = do-reg do-reg-long do-reg-backref
15# cannot run with large limits
16#REGRESS_TARGETS += do-t_exhaust
17
18do-reg: ${PROG}
19 ./re < ${TESTS}
20do-reg-long: ${PROG}
21 ./re -el < ${TESTS}
22do-reg-backref: ${PROG}
23 ./re -er < ${TESTS}
24#do-t_exhaust: t_exhaust
25# ./t_exhaust
26
27.include <bsd.regress.mk>
diff --git a/src/regress/lib/libc/regex/debug.c b/src/regress/lib/libc/regex/debug.c
deleted file mode 100644
index 7044a2a185..0000000000
--- a/src/regress/lib/libc/regex/debug.c
+++ /dev/null
@@ -1,208 +0,0 @@
1/* $OpenBSD: debug.c,v 1.5 2020/12/31 17:20:19 millert Exp $ */
2/* $NetBSD: debug.c,v 1.2 1995/04/20 22:39:42 cgd Exp $ */
3
4#include <stdio.h>
5#include <string.h>
6#include <ctype.h>
7#include <limits.h>
8#include <stdlib.h>
9#include <sys/types.h>
10#include <regex.h>
11
12#include "utils.h"
13#include "regex2.h"
14#include "debug.ih"
15
16/*
17 - regprint - print a regexp for debugging
18 == void regprint(regex_t *r, FILE *d);
19 */
20void
21regprint(r, d)
22regex_t *r;
23FILE *d;
24{
25 register struct re_guts *g = r->re_g;
26 register int i;
27 register int c;
28 register int last;
29
30 fprintf(d, "%ld states", (long)g->nstates);
31 fprintf(d, ", first %ld last %ld", (long)g->firststate,
32 (long)g->laststate);
33 if (g->iflags&USEBOL)
34 fprintf(d, ", USEBOL");
35 if (g->iflags&USEEOL)
36 fprintf(d, ", USEEOL");
37 if (g->iflags&BAD)
38 fprintf(d, ", BAD");
39 if (g->nsub > 0)
40 fprintf(d, ", nsub=%ld", (long)g->nsub);
41 if (g->must != NULL)
42 fprintf(d, ", must(%ld) `%*s'", (long)g->mlen, (int)g->mlen,
43 g->must);
44 if (g->backrefs)
45 fprintf(d, ", backrefs");
46 if (g->nplus > 0)
47 fprintf(d, ", nplus %ld", (long)g->nplus);
48 fprintf(d, "\n");
49 s_print(g, d);
50}
51
52/*
53 - s_print - print the strip for debugging
54 == static void s_print(register struct re_guts *g, FILE *d);
55 */
56static void
57s_print(g, d)
58register struct re_guts *g;
59FILE *d;
60{
61 register sop *s;
62 register cset *cs;
63 register int i;
64 register int done = 0;
65 register sop opnd;
66 register int col = 0;
67 register int last;
68 register sopno offset = 2;
69# define GAP() { if (offset % 5 == 0) { \
70 if (col > 40) { \
71 fprintf(d, "\n\t"); \
72 col = 0; \
73 } else { \
74 fprintf(d, " "); \
75 col++; \
76 } \
77 } else \
78 col++; \
79 offset++; \
80 }
81
82 if (OP(g->strip[0]) != OEND)
83 fprintf(d, "missing initial OEND!\n");
84 for (s = &g->strip[1]; !done; s++) {
85 opnd = OPND(*s);
86 switch (OP(*s)) {
87 case OEND:
88 fprintf(d, "\n");
89 done = 1;
90 break;
91 case OCHAR:
92 if (strchr("\\|()^$.[+*?{}!<> ", (char)opnd) != NULL)
93 fprintf(d, "\\%c", (char)opnd);
94 else
95 fprintf(d, "%s", regchar((char)opnd));
96 break;
97 case OBOL:
98 fprintf(d, "^");
99 break;
100 case OEOL:
101 fprintf(d, "$");
102 break;
103 case OBOW:
104 fprintf(d, "\\{");
105 break;
106 case OEOW:
107 fprintf(d, "\\}");
108 break;
109 case OANY:
110 fprintf(d, ".");
111 break;
112 case OANYOF:
113 fprintf(d, "[(%ld)", (long)opnd);
114 cs = &g->sets[opnd];
115 last = -1;
116 for (i = 0; i < g->csetsize+1; i++) /* +1 flushes */
117 if (CHIN(cs, i) && i < g->csetsize) {
118 if (last < 0) {
119 fprintf(d, "%s", regchar(i));
120 last = i;
121 }
122 } else {
123 if (last >= 0) {
124 if (last != i-1)
125 fprintf(d, "-%s",
126 regchar(i-1));
127 last = -1;
128 }
129 }
130 fprintf(d, "]");
131 break;
132 case OBACK_:
133 fprintf(d, "(\\<%ld>", (long)opnd);
134 break;
135 case O_BACK:
136 fprintf(d, "<%ld>\\)", (long)opnd);
137 break;
138 case OPLUS_:
139 fprintf(d, "(+");
140 if (OP(*(s+opnd)) != O_PLUS)
141 fprintf(d, "<%ld>", (long)opnd);
142 break;
143 case O_PLUS:
144 if (OP(*(s-opnd)) != OPLUS_)
145 fprintf(d, "<%ld>", (long)opnd);
146 fprintf(d, "+)");
147 break;
148 case OQUEST_:
149 fprintf(d, "(?");
150 if (OP(*(s+opnd)) != O_QUEST)
151 fprintf(d, "<%ld>", (long)opnd);
152 break;
153 case O_QUEST:
154 if (OP(*(s-opnd)) != OQUEST_)
155 fprintf(d, "<%ld>", (long)opnd);
156 fprintf(d, "?)");
157 break;
158 case OLPAREN:
159 fprintf(d, "((<%ld>", (long)opnd);
160 break;
161 case ORPAREN:
162 fprintf(d, "<%ld>))", (long)opnd);
163 break;
164 case OCH_:
165 fprintf(d, "<");
166 if (OP(*(s+opnd)) != OOR2)
167 fprintf(d, "<%ld>", (long)opnd);
168 break;
169 case OOR1:
170 if (OP(*(s-opnd)) != OOR1 && OP(*(s-opnd)) != OCH_)
171 fprintf(d, "<%ld>", (long)opnd);
172 fprintf(d, "|");
173 break;
174 case OOR2:
175 fprintf(d, "|");
176 if (OP(*(s+opnd)) != OOR2 && OP(*(s+opnd)) != O_CH)
177 fprintf(d, "<%ld>", (long)opnd);
178 break;
179 case O_CH:
180 if (OP(*(s-opnd)) != OOR1)
181 fprintf(d, "<%ld>", (long)opnd);
182 fprintf(d, ">");
183 break;
184 default:
185 fprintf(d, "!%ld(%ld)!", (long)OP(*s), (long)opnd);
186 break;
187 }
188 if (!done)
189 GAP();
190 }
191}
192
193/*
194 - regchar - make a character printable
195 == static char *regchar(int ch);
196 */
197static char * /* -> representation */
198regchar(ch)
199int ch;
200{
201 static char buf[10];
202
203 if (isprint(ch) || ch == ' ')
204 snprintf(buf, sizeof buf, "%c", ch);
205 else
206 snprintf(buf, sizeof buf, "\\%o", ch);
207 return(buf);
208}
diff --git a/src/regress/lib/libc/regex/debug.ih b/src/regress/lib/libc/regex/debug.ih
deleted file mode 100644
index 9eb313af23..0000000000
--- a/src/regress/lib/libc/regex/debug.ih
+++ /dev/null
@@ -1,17 +0,0 @@
1/* $OpenBSD: debug.ih,v 1.3 2002/02/16 21:27:32 millert Exp $ */
2/* $NetBSD: debug.ih,v 1.2 1995/04/20 22:39:47 cgd Exp $ */
3
4/* ========= begin header generated by ./mkh ========= */
5#ifdef __cplusplus
6extern "C" {
7#endif
8
9/* === debug.c === */
10void regprint(regex_t *r, FILE *d);
11static void s_print(register struct re_guts *g, FILE *d);
12static char *regchar(int ch);
13
14#ifdef __cplusplus
15}
16#endif
17/* ========= end header generated by ./mkh ========= */
diff --git a/src/regress/lib/libc/regex/main.c b/src/regress/lib/libc/regex/main.c
deleted file mode 100644
index 4501e86b87..0000000000
--- a/src/regress/lib/libc/regex/main.c
+++ /dev/null
@@ -1,516 +0,0 @@
1/* $OpenBSD: main.c,v 1.12 2022/12/04 23:50:46 cheloha Exp $ */
2/* $NetBSD: main.c,v 1.2 1995/04/20 22:39:51 cgd Exp $ */
3
4#include <stdio.h>
5#include <stdlib.h>
6#include <string.h>
7#include <sys/types.h>
8#include <regex.h>
9#include <assert.h>
10#include <unistd.h>
11
12#include "main.ih"
13
14char *progname;
15int debug = 0;
16int line = 0;
17int status = 0;
18
19int copts = REG_EXTENDED;
20int eopts = 0;
21regoff_t startoff = 0;
22regoff_t endoff = 0;
23
24
25extern int split(char *, char *[], int, char *);
26extern void regprint(regex_t *, FILE *);
27
28/*
29 - main - do the simple case, hand off to regress() for regression
30 */
31int
32main(int argc, char *argv[])
33
34{
35 regex_t re;
36# define NS 10
37 regmatch_t subs[NS];
38 char erbuf[100];
39 int err;
40 size_t len;
41 int c;
42 int errflg = 0;
43 register int i;
44
45 progname = argv[0];
46
47 while ((c = getopt(argc, argv, "c:E:e:S:x")) != -1)
48 switch (c) {
49 case 'c': /* compile options */
50 copts = options('c', optarg);
51 break;
52 case 'E': /* end offset */
53 endoff = (regoff_t)atoi(optarg);
54 break;
55 case 'e': /* execute options */
56 eopts = options('e', optarg);
57 break;
58 case 'S': /* start offset */
59 startoff = (regoff_t)atoi(optarg);
60 break;
61 case 'x': /* Debugging. */
62 debug++;
63 break;
64 default:
65 errflg++;
66 break;
67 }
68 if (errflg) {
69 fprintf(stderr, "usage: %s ", progname);
70 fprintf(stderr, "[-x] [-c copt] [-E endoff] [-e eopt] [-S startoff] [re]\n");
71 exit(2);
72 }
73
74 if (optind >= argc) {
75 regress(stdin);
76 exit(status);
77 }
78
79 err = regcomp(&re, argv[optind++], copts);
80 if (err) {
81 len = regerror(err, &re, erbuf, sizeof(erbuf));
82 fprintf(stderr, "error %s, %zu/%zu `%s'\n",
83 eprint(err), len, sizeof(erbuf), erbuf);
84 exit(status);
85 }
86 regprint(&re, stdout);
87
88 if (optind >= argc) {
89 regfree(&re);
90 exit(status);
91 }
92
93 if (eopts&REG_STARTEND) {
94 subs[0].rm_so = startoff;
95 subs[0].rm_eo = strlen(argv[optind]) - endoff;
96 }
97 err = regexec(&re, argv[optind], (size_t)NS, subs, eopts);
98 if (err) {
99 len = regerror(err, &re, erbuf, sizeof(erbuf));
100 fprintf(stderr, "error %s, %zu/%zu `%s'\n",
101 eprint(err), len, sizeof(erbuf), erbuf);
102 exit(status);
103 }
104 if (!(copts&REG_NOSUB)) {
105 len = (size_t)(subs[0].rm_eo - subs[0].rm_so);
106 if (subs[0].rm_so != -1) {
107 if (len != 0)
108 printf("match `%.*s'\n", (int)len,
109 argv[optind] + subs[0].rm_so);
110 else
111 printf("match `'@%.1s\n",
112 argv[optind] + subs[0].rm_so);
113 }
114 for (i = 1; i < NS; i++)
115 if (subs[i].rm_so != -1)
116 printf("(%d) `%.*s'\n", i,
117 (int)(subs[i].rm_eo - subs[i].rm_so),
118 argv[optind] + subs[i].rm_so);
119 }
120 exit(status);
121}
122
123/*
124 - regress - main loop of regression test
125 == void regress(FILE *in);
126 */
127void
128regress(in)
129FILE *in;
130{
131 char inbuf[1000];
132# define MAXF 10
133 char *f[MAXF];
134 int nf;
135 int i;
136 char erbuf[100];
137 size_t ne;
138 char *badpat = "invalid regular expression";
139# define SHORT 10
140 char *bpname = "REG_BADPAT";
141 regex_t re;
142
143 while (fgets(inbuf, sizeof(inbuf), in) != NULL) {
144 line++;
145 if (inbuf[0] == '#' || inbuf[0] == '\n')
146 continue; /* NOTE CONTINUE */
147 inbuf[strcspn(inbuf, "\n")] = '\0'; /* get rid of stupid \n */
148 if (debug)
149 fprintf(stdout, "%d:\n", line);
150 nf = split(inbuf, f, MAXF, "\t\t");
151 if (nf < 3) {
152 fprintf(stderr, "bad input, line %d\n", line);
153 exit(1);
154 }
155 for (i = 0; i < nf; i++)
156 if (strcmp(f[i], "\"\"") == 0)
157 f[i] = "";
158 if (nf <= 3)
159 f[3] = NULL;
160 if (nf <= 4)
161 f[4] = NULL;
162 try(f[0], f[1], f[2], f[3], f[4], options('c', f[1]));
163 if (opt('&', f[1])) /* try with either type of RE */
164 try(f[0], f[1], f[2], f[3], f[4],
165 options('c', f[1]) &~ REG_EXTENDED);
166 }
167
168 ne = regerror(REG_BADPAT, (regex_t *)NULL, erbuf, sizeof(erbuf));
169 if (strcmp(erbuf, badpat) != 0 || ne != strlen(badpat)+1) {
170 fprintf(stderr, "end: regerror() test gave `%s' not `%s'\n",
171 erbuf, badpat);
172 status = 1;
173 }
174 ne = regerror(REG_BADPAT, (regex_t *)NULL, erbuf, (size_t)SHORT);
175 if (strncmp(erbuf, badpat, SHORT-1) != 0 || erbuf[SHORT-1] != '\0' ||
176 ne != strlen(badpat)+1) {
177 fprintf(stderr, "end: regerror() short test gave `%s' not `%.*s'\n",
178 erbuf, SHORT-1, badpat);
179 status = 1;
180 }
181 ne = regerror(REG_ITOA|REG_BADPAT, (regex_t *)NULL, erbuf, sizeof(erbuf));
182 if (strcmp(erbuf, bpname) != 0 || ne != strlen(bpname)+1) {
183 fprintf(stderr, "end: regerror() ITOA test gave `%s' not `%s'\n",
184 erbuf, bpname);
185 status = 1;
186 }
187 re.re_endp = bpname;
188 ne = regerror(REG_ATOI, &re, erbuf, sizeof(erbuf));
189 if (atoi(erbuf) != (int)REG_BADPAT) {
190 fprintf(stderr, "end: regerror() ATOI test gave `%s' not `%ld'\n",
191 erbuf, (long)REG_BADPAT);
192 status = 1;
193 } else if (ne != strlen(erbuf)+1) {
194 fprintf(stderr, "end: regerror() ATOI test len(`%s') = %ld\n",
195 erbuf, (long)REG_BADPAT);
196 status = 1;
197 }
198}
199
200/*
201 - try - try it, and report on problems
202 == void try(char *f0, char *f1, char *f2, char *f3, char *f4, int opts);
203 */
204void
205try(f0, f1, f2, f3, f4, opts)
206char *f0;
207char *f1;
208char *f2;
209char *f3;
210char *f4;
211int opts; /* may not match f1 */
212{
213 regex_t re;
214# define NSUBS 10
215 regmatch_t subs[NSUBS];
216# define NSHOULD 15
217 char *should[NSHOULD];
218 int nshould;
219 char erbuf[100];
220 int err;
221 int len;
222 char *type = (opts & REG_EXTENDED) ? "ERE" : "BRE";
223 register int i;
224 char *grump;
225 char f0copy[1000];
226 char f2copy[1000];
227
228 strlcpy(f0copy, f0, sizeof f0copy);
229 re.re_endp = (opts&REG_PEND) ? f0copy + strlen(f0copy) : NULL;
230 fixstr(f0copy);
231 err = regcomp(&re, f0copy, opts);
232 if (err != 0 && (!opt('C', f1) || err != efind(f2))) {
233 /* unexpected error or wrong error */
234 len = regerror(err, &re, erbuf, sizeof(erbuf));
235 fprintf(stderr, "%d: %s error %s, %d/%zu `%s'\n",
236 line, type, eprint(err), len,
237 sizeof(erbuf), erbuf);
238 status = 1;
239 } else if (err == 0 && opt('C', f1)) {
240 /* unexpected success */
241 fprintf(stderr, "%d: %s should have given REG_%s\n",
242 line, type, f2);
243 status = 1;
244 err = 1; /* so we won't try regexec */
245 }
246
247 if (err != 0) {
248 regfree(&re);
249 return;
250 }
251
252 strlcpy(f2copy, f2, sizeof f2copy);
253 fixstr(f2copy);
254
255 if (options('e', f1)&REG_STARTEND) {
256 if (strchr(f2, '(') == NULL || strchr(f2, ')') == NULL)
257 fprintf(stderr, "%d: bad STARTEND syntax\n", line);
258 subs[0].rm_so = strchr(f2, '(') - f2 + 1;
259 subs[0].rm_eo = strchr(f2, ')') - f2;
260 /* the preceding character is relevant with REG_NOTBOL */
261 f2copy[subs[0].rm_so - 1] = subs[0].rm_so > 1 ?
262 f2copy[subs[0].rm_so - 2] : 'X';
263 }
264 err = regexec(&re, f2copy, NSUBS, subs, options('e', f1));
265
266 if (err != 0 && (f3 != NULL || err != REG_NOMATCH)) {
267 /* unexpected error or wrong error */
268 len = regerror(err, &re, erbuf, sizeof(erbuf));
269 fprintf(stderr, "%d: %s exec error %s, %d/%zu `%s'\n",
270 line, type, eprint(err), len,
271 sizeof(erbuf), erbuf);
272 status = 1;
273 } else if (err != 0) {
274 /* nothing more to check */
275 } else if (f3 == NULL) {
276 /* unexpected success */
277 fprintf(stderr, "%d: %s exec should have failed\n",
278 line, type);
279 status = 1;
280 err = 1; /* just on principle */
281 } else if (opts&REG_NOSUB) {
282 /* nothing more to check */
283 } else if ((grump = check(f2, subs[0], f3)) != NULL) {
284 fprintf(stderr, "%d: %s %s\n", line, type, grump);
285 status = 1;
286 err = 1;
287 }
288
289 if (err != 0 || f4 == NULL) {
290 regfree(&re);
291 return;
292 }
293
294 for (i = 1; i < NSHOULD; i++)
295 should[i] = NULL;
296 nshould = split(f4, should+1, NSHOULD-1, ",");
297 if (nshould == 0) {
298 nshould = 1;
299 should[1] = "";
300 }
301 for (i = 1; i < NSUBS; i++) {
302 grump = check(f2, subs[i], should[i]);
303 if (grump != NULL) {
304 fprintf(stderr, "%d: %s $%d %s\n", line,
305 type, i, grump);
306 status = 1;
307 err = 1;
308 }
309 }
310
311 regfree(&re);
312}
313
314/*
315 - options - pick options out of a regression-test string
316 == int options(int type, char *s);
317 */
318int
319options(type, s)
320int type; /* 'c' compile, 'e' exec */
321char *s;
322{
323 register char *p;
324 register int o = (type == 'c') ? copts : eopts;
325 register char *legal = (type == 'c') ? "bisnmp" : "^$#tl";
326
327 for (p = s; *p != '\0'; p++)
328 if (strchr(legal, *p) != NULL)
329 switch (*p) {
330 case 'b':
331 o &= ~REG_EXTENDED;
332 break;
333 case 'i':
334 o |= REG_ICASE;
335 break;
336 case 's':
337 o |= REG_NOSUB;
338 break;
339 case 'n':
340 o |= REG_NEWLINE;
341 break;
342 case 'm':
343 o &= ~REG_EXTENDED;
344 o |= REG_NOSPEC;
345 break;
346 case 'p':
347 o |= REG_PEND;
348 break;
349 case '^':
350 o |= REG_NOTBOL;
351 break;
352 case '$':
353 o |= REG_NOTEOL;
354 break;
355 case '#':
356 o |= REG_STARTEND;
357 break;
358 case 't': /* trace */
359 o |= REG_TRACE;
360 break;
361 case 'l': /* force long representation */
362 o |= REG_LARGE;
363 break;
364 case 'r': /* force backref use */
365 o |= REG_BACKR;
366 break;
367 }
368 return(o);
369}
370
371/*
372 - opt - is a particular option in a regression string?
373 == int opt(int c, char *s);
374 */
375int /* predicate */
376opt(c, s)
377int c;
378char *s;
379{
380 return(strchr(s, c) != NULL);
381}
382
383/*
384 - fixstr - transform magic characters in strings
385 == void fixstr(register char *p);
386 */
387void
388fixstr(p)
389register char *p;
390{
391 if (p == NULL)
392 return;
393
394 for (; *p != '\0'; p++)
395 if (*p == 'N')
396 *p = '\n';
397 else if (*p == 'T')
398 *p = '\t';
399 else if (*p == 'S')
400 *p = ' ';
401 else if (*p == 'Z')
402 *p = '\0';
403}
404
405/*
406 - check - check a substring match
407 == char *check(char *str, regmatch_t sub, char *should);
408 */
409char * /* NULL or complaint */
410check(str, sub, should)
411char *str;
412regmatch_t sub;
413char *should;
414{
415 register int len;
416 register int shlen;
417 register char *p;
418 static char grump[500];
419 register char *at = NULL;
420
421 if (should != NULL && strcmp(should, "-") == 0)
422 should = NULL;
423 if (should != NULL && should[0] == '@') {
424 at = should + 1;
425 should = "";
426 }
427
428 /* check rm_so and rm_eo for consistency */
429 if (sub.rm_so > sub.rm_eo || (sub.rm_so == -1 && sub.rm_eo != -1) ||
430 (sub.rm_so != -1 && sub.rm_eo == -1) ||
431 (sub.rm_so != -1 && sub.rm_so < 0) ||
432 (sub.rm_eo != -1 && sub.rm_eo < 0) ) {
433 snprintf(grump, sizeof grump,
434 "start %ld end %ld", (long)sub.rm_so,
435 (long)sub.rm_eo);
436 return(grump);
437 }
438
439 /* check for no match */
440 if (sub.rm_so == -1 && should == NULL)
441 return(NULL);
442 if (sub.rm_so == -1)
443 return("did not match");
444
445 /* check for in range */
446 if (sub.rm_eo > strlen(str)) {
447 snprintf(grump, sizeof grump,
448 "start %ld end %ld, past end of string",
449 (long)sub.rm_so, (long)sub.rm_eo);
450 return(grump);
451 }
452
453 len = (int)(sub.rm_eo - sub.rm_so);
454 p = str + sub.rm_so;
455
456 /* check for not supposed to match */
457 if (should == NULL) {
458 snprintf(grump, sizeof grump, "matched `%.*s'", len, p);
459 return(grump);
460 }
461
462 /* check for wrong match */
463 shlen = (int)strlen(should);
464 if (len != shlen || strncmp(p, should, (size_t)shlen) != 0) {
465 snprintf(grump, sizeof grump, "matched `%.*s' instead", len, p);
466 return(grump);
467 }
468 if (shlen > 0)
469 return(NULL);
470
471 /* check null match in right place */
472 if (at == NULL)
473 return(NULL);
474 shlen = strlen(at);
475 if (shlen == 0)
476 shlen = 1; /* force check for end-of-string */
477 if (strncmp(p, at, shlen) != 0) {
478 snprintf(grump, sizeof grump, "matched null at `%.20s'", p);
479 return(grump);
480 }
481 return(NULL);
482}
483
484/*
485 - eprint - convert error number to name
486 == static char *eprint(int err);
487 */
488static char *
489eprint(err)
490int err;
491{
492 static char epbuf[100];
493 size_t len;
494
495 len = regerror(REG_ITOA|err, (regex_t *)NULL, epbuf, sizeof(epbuf));
496 assert(len <= sizeof(epbuf));
497 return(epbuf);
498}
499
500/*
501 - efind - convert error name to number
502 == static int efind(char *name);
503 */
504static int
505efind(name)
506char *name;
507{
508 static char efbuf[100];
509 regex_t re;
510
511 snprintf(efbuf, sizeof efbuf, "REG_%s", name);
512 assert(strlen(efbuf) < sizeof(efbuf));
513 re.re_endp = efbuf;
514 (void) regerror(REG_ATOI, &re, efbuf, sizeof(efbuf));
515 return(atoi(efbuf));
516}
diff --git a/src/regress/lib/libc/regex/main.ih b/src/regress/lib/libc/regex/main.ih
deleted file mode 100644
index 0860e26333..0000000000
--- a/src/regress/lib/libc/regex/main.ih
+++ /dev/null
@@ -1,22 +0,0 @@
1/* $OpenBSD: main.ih,v 1.3 2002/02/16 21:27:32 millert Exp $ */
2/* $NetBSD: main.ih,v 1.2 1995/04/20 22:39:55 cgd Exp $ */
3
4/* ========= begin header generated by ./mkh ========= */
5#ifdef __cplusplus
6extern "C" {
7#endif
8
9/* === main.c === */
10void regress(FILE *in);
11void try(char *f0, char *f1, char *f2, char *f3, char *f4, int opts);
12int options(int type, char *s);
13int opt(int c, char *s);
14void fixstr(register char *p);
15char *check(char *str, regmatch_t sub, char *should);
16static char *eprint(int err);
17static int efind(char *name);
18
19#ifdef __cplusplus
20}
21#endif
22/* ========= end header generated by ./mkh ========= */
diff --git a/src/regress/lib/libc/regex/split.c b/src/regress/lib/libc/regex/split.c
deleted file mode 100644
index fcd81a3503..0000000000
--- a/src/regress/lib/libc/regex/split.c
+++ /dev/null
@@ -1,317 +0,0 @@
1/* $OpenBSD: split.c,v 1.5 2007/09/09 23:25:12 chl Exp $ */
2/* $NetBSD: split.c,v 1.2 1995/04/20 22:39:57 cgd Exp $ */
3
4#include <stdio.h>
5#include <string.h>
6
7int split(char *string, char *fields[], int nfields, char *sep);
8
9/*
10 - split - divide a string into fields, like awk split()
11 = int split(char *string, char *fields[], int nfields, char *sep);
12 */
13int /* number of fields, including overflow */
14split(char *string, char *fields[], int nfields, char *sep)
15{
16 register char *p = string;
17 register char c; /* latest character */
18 register char sepc = sep[0];
19 register char sepc2;
20 register int fn;
21 register char **fp = fields;
22 register char *sepp;
23 register int trimtrail;
24
25 /* white space */
26 if (sepc == '\0') {
27 while ((c = *p++) == ' ' || c == '\t')
28 continue;
29 p--;
30 trimtrail = 1;
31 sep = " \t"; /* note, code below knows this is 2 long */
32 sepc = ' ';
33 } else
34 trimtrail = 0;
35 sepc2 = sep[1]; /* now we can safely pick this up */
36
37 /* catch empties */
38 if (*p == '\0')
39 return(0);
40
41 /* single separator */
42 if (sepc2 == '\0') {
43 fn = nfields;
44 for (;;) {
45 *fp++ = p;
46 fn--;
47 if (fn == 0)
48 break;
49 while ((c = *p++) != sepc)
50 if (c == '\0')
51 return(nfields - fn);
52 *(p-1) = '\0';
53 }
54 /* we have overflowed the fields vector -- just count them */
55 fn = nfields;
56 for (;;) {
57 while ((c = *p++) != sepc)
58 if (c == '\0')
59 return(fn);
60 fn++;
61 }
62 /* not reached */
63 }
64
65 /* two separators */
66 if (sep[2] == '\0') {
67 fn = nfields;
68 for (;;) {
69 *fp++ = p;
70 fn--;
71 while ((c = *p++) != sepc && c != sepc2)
72 if (c == '\0') {
73 if (trimtrail && **(fp-1) == '\0')
74 fn++;
75 return(nfields - fn);
76 }
77 if (fn == 0)
78 break;
79 *(p-1) = '\0';
80 while ((c = *p++) == sepc || c == sepc2)
81 continue;
82 p--;
83 }
84 /* we have overflowed the fields vector -- just count them */
85 fn = nfields;
86 while (c != '\0') {
87 while ((c = *p++) == sepc || c == sepc2)
88 continue;
89 p--;
90 fn++;
91 while ((c = *p++) != '\0' && c != sepc && c != sepc2)
92 continue;
93 }
94 /* might have to trim trailing white space */
95 if (trimtrail) {
96 p--;
97 while ((c = *--p) == sepc || c == sepc2)
98 continue;
99 p++;
100 if (*p != '\0') {
101 if (fn == nfields+1)
102 *p = '\0';
103 fn--;
104 }
105 }
106 return(fn);
107 }
108
109 /* n separators */
110 fn = 0;
111 for (;;) {
112 if (fn < nfields)
113 *fp++ = p;
114 fn++;
115 for (;;) {
116 c = *p++;
117 if (c == '\0')
118 return(fn);
119 sepp = sep;
120 while ((sepc = *sepp++) != '\0' && sepc != c)
121 continue;
122 if (sepc != '\0') /* it was a separator */
123 break;
124 }
125 if (fn < nfields)
126 *(p-1) = '\0';
127 for (;;) {
128 c = *p++;
129 sepp = sep;
130 while ((sepc = *sepp++) != '\0' && sepc != c)
131 continue;
132 if (sepc == '\0') /* it wasn't a separator */
133 break;
134 }
135 p--;
136 }
137
138 /* not reached */
139}
140
141#ifdef TEST_SPLIT
142
143
144/*
145 * test program
146 * pgm runs regression
147 * pgm sep splits stdin lines by sep
148 * pgm str sep splits str by sep
149 * pgm str sep n splits str by sep n times
150 */
151int
152main(argc, argv)
153int argc;
154char *argv[];
155{
156 char buf[512];
157 register int n;
158# define MNF 10
159 char *fields[MNF];
160
161 if (argc > 4)
162 for (n = atoi(argv[3]); n > 0; n--) {
163 (void) strlcpy(buf, argv[1], sizeof buf);
164 }
165 else if (argc > 3)
166 for (n = atoi(argv[3]); n > 0; n--) {
167 (void) strlcpy(buf, argv[1], sizeof buf);
168 (void) split(buf, fields, MNF, argv[2]);
169 }
170 else if (argc > 2)
171 dosplit(argv[1], argv[2]);
172 else if (argc > 1)
173 while (fgets(buf, sizeof(buf), stdin) != NULL) {
174 buf[strcspn(buf, "\n")] = '\0'; /* stomp newline */
175 dosplit(buf, argv[1]);
176 }
177 else
178 regress();
179
180 exit(0);
181}
182
183dosplit(string, seps)
184char *string;
185char *seps;
186{
187# define NF 5
188 char *fields[NF];
189 register int nf;
190
191 nf = split(string, fields, NF, seps);
192 print(nf, NF, fields);
193}
194
195print(nf, nfp, fields)
196int nf;
197int nfp;
198char *fields[];
199{
200 register int fn;
201 register int bound;
202
203 bound = (nf > nfp) ? nfp : nf;
204 printf("%d:\t", nf);
205 for (fn = 0; fn < bound; fn++)
206 printf("\"%s\"%s", fields[fn], (fn+1 < nf) ? ", " : "\n");
207}
208
209#define RNF 5 /* some table entries know this */
210struct {
211 char *str;
212 char *seps;
213 int nf;
214 char *fi[RNF];
215} tests[] = {
216 "", " ", 0, { "" },
217 " ", " ", 2, { "", "" },
218 "x", " ", 1, { "x" },
219 "xy", " ", 1, { "xy" },
220 "x y", " ", 2, { "x", "y" },
221 "abc def g ", " ", 5, { "abc", "def", "", "g", "" },
222 " a bcd", " ", 4, { "", "", "a", "bcd" },
223 "a b c d e f", " ", 6, { "a", "b", "c", "d", "e f" },
224 " a b c d ", " ", 6, { "", "a", "b", "c", "d " },
225
226 "", " _", 0, { "" },
227 " ", " _", 2, { "", "" },
228 "x", " _", 1, { "x" },
229 "x y", " _", 2, { "x", "y" },
230 "ab _ cd", " _", 2, { "ab", "cd" },
231 " a_b c ", " _", 5, { "", "a", "b", "c", "" },
232 "a b c_d e f", " _", 6, { "a", "b", "c", "d", "e f" },
233 " a b c d ", " _", 6, { "", "a", "b", "c", "d " },
234
235 "", " _~", 0, { "" },
236 " ", " _~", 2, { "", "" },
237 "x", " _~", 1, { "x" },
238 "x y", " _~", 2, { "x", "y" },
239 "ab _~ cd", " _~", 2, { "ab", "cd" },
240 " a_b c~", " _~", 5, { "", "a", "b", "c", "" },
241 "a b_c d~e f", " _~", 6, { "a", "b", "c", "d", "e f" },
242 "~a b c d ", " _~", 6, { "", "a", "b", "c", "d " },
243
244 "", " _~-", 0, { "" },
245 " ", " _~-", 2, { "", "" },
246 "x", " _~-", 1, { "x" },
247 "x y", " _~-", 2, { "x", "y" },
248 "ab _~- cd", " _~-", 2, { "ab", "cd" },
249 " a_b c~", " _~-", 5, { "", "a", "b", "c", "" },
250 "a b_c-d~e f", " _~-", 6, { "a", "b", "c", "d", "e f" },
251 "~a-b c d ", " _~-", 6, { "", "a", "b", "c", "d " },
252
253 "", " ", 0, { "" },
254 " ", " ", 2, { "", "" },
255 "x", " ", 1, { "x" },
256 "xy", " ", 1, { "xy" },
257 "x y", " ", 2, { "x", "y" },
258 "abc def g ", " ", 4, { "abc", "def", "g", "" },
259 " a bcd", " ", 3, { "", "a", "bcd" },
260 "a b c d e f", " ", 6, { "a", "b", "c", "d", "e f" },
261 " a b c d ", " ", 6, { "", "a", "b", "c", "d " },
262
263 "", "", 0, { "" },
264 " ", "", 0, { "" },
265 "x", "", 1, { "x" },
266 "xy", "", 1, { "xy" },
267 "x y", "", 2, { "x", "y" },
268 "abc def g ", "", 3, { "abc", "def", "g" },
269 "\t a bcd", "", 2, { "a", "bcd" },
270 " a \tb\t c ", "", 3, { "a", "b", "c" },
271 "a b c d e ", "", 5, { "a", "b", "c", "d", "e" },
272 "a b\tc d e f", "", 6, { "a", "b", "c", "d", "e f" },
273 " a b c d e f ", "", 6, { "a", "b", "c", "d", "e f " },
274
275 NULL, NULL, 0, { NULL },
276};
277
278regress()
279{
280 char buf[512];
281 register int n;
282 char *fields[RNF+1];
283 register int nf;
284 register int i;
285 register int printit;
286 register char *f;
287
288 for (n = 0; tests[n].str != NULL; n++) {
289 (void) strlcpy(buf, tests[n].str, sizeof buf);
290 fields[RNF] = NULL;
291 nf = split(buf, fields, RNF, tests[n].seps);
292 printit = 0;
293 if (nf != tests[n].nf) {
294 printf("split `%s' by `%s' gave %d fields, not %d\n",
295 tests[n].str, tests[n].seps, nf, tests[n].nf);
296 printit = 1;
297 } else if (fields[RNF] != NULL) {
298 printf("split() went beyond array end\n");
299 printit = 1;
300 } else {
301 for (i = 0; i < nf && i < RNF; i++) {
302 f = fields[i];
303 if (f == NULL)
304 f = "(NULL)";
305 if (strcmp(f, tests[n].fi[i]) != 0) {
306 printf("split `%s' by `%s', field %d is `%s', not `%s'\n",
307 tests[n].str, tests[n].seps,
308 i, fields[i], tests[n].fi[i]);
309 printit = 1;
310 }
311 }
312 }
313 if (printit)
314 print(nf, RNF, fields);
315 }
316}
317#endif
diff --git a/src/regress/lib/libc/regex/t_exhaust.c b/src/regress/lib/libc/regex/t_exhaust.c
deleted file mode 100644
index f54bab18da..0000000000
--- a/src/regress/lib/libc/regex/t_exhaust.c
+++ /dev/null
@@ -1,183 +0,0 @@
1/* $OpenBSD: t_exhaust.c,v 1.4 2024/07/15 10:11:56 anton Exp $ */
2/* $NetBSD: t_exhaust.c,v 1.2 2011/10/21 00:41:34 christos Exp $ */
3
4/*-
5 * Copyright (c) 2011 The NetBSD Foundation, Inc.
6 * All rights reserved.
7 *
8 * This code is derived from software contributed to The NetBSD Foundation
9 * by Christos Zoulas.
10 *
11 * Redistribution and use in source and binary forms, with or without
12 * modification, are permitted provided that the following conditions
13 * are met:
14 * 1. Redistributions of source code must retain the above copyright
15 * notice, this list of conditions and the following disclaimer.
16 * 2. Redistributions in binary form must reproduce the above copyright
17 * notice, this list of conditions and the following disclaimer in the
18 * documentation and/or other materials provided with the distribution.
19 * 3. All advertising materials mentioning features or use of this software
20 * must display the following acknowledgement:
21 * This product includes software developed by the NetBSD
22 * Foundation, Inc. and its contributors.
23 * 4. Neither the name of The NetBSD Foundation nor the names of its
24 * contributors may be used to endorse or promote products derived
25 * from this software without specific prior written permission.
26 *
27 * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
28 * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
29 * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
30 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
31 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
34 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
35 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
36 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
37 * POSSIBILITY OF SUCH DAMAGE.
38 */
39
40#include <stdio.h>
41#include <regex.h>
42#include <string.h>
43#include <stdlib.h>
44#include <err.h>
45//#include <atf-c.h>
46
47
48static char *
49mkstr(const char *str, size_t len)
50{
51 size_t i, slen = strlen(str);
52 char *p = malloc(slen * len + 1);
53 if (p == NULL)
54 err(1, "malloc");
55 for (i = 0; i < len; i++)
56 strlcpy(&p[i * slen], str, slen * len + 1 - (i * slen));
57 return p;
58}
59
60static char *
61concat(const char *d, const char *s)
62{
63 size_t dlen = strlen(d);
64 size_t slen = strlen(s);
65 char *p = malloc(dlen + slen + 1);
66 strlcpy(p, d, dlen + slen + 1);
67 strlcat(p, s, dlen + slen + 1);
68 return p;
69}
70
71static char *
72p0(size_t len)
73{
74 char *d, *s1, *s2;
75 s1 = mkstr("\\(", len);
76 s2 = concat(s1, ")");
77 free(s1);
78 d = concat("(", s2);
79 free(s2);
80 return d;
81}
82
83static char *
84p1(size_t len)
85{
86 char *d, *s1, *s2, *s3;
87 s1 = mkstr("\\(", 60);
88 s2 = mkstr("(.*)", len);
89 s3 = concat(s1, s2);
90 free(s2);
91 free(s1);
92 s1 = concat(s3, ")");
93 free(s3);
94 d = concat("(", s1);
95 free(s1);
96 return d;
97}
98
99static char *
100ps(const char *m, const char *s, size_t len)
101{
102 char *d, *s1, *s2, *s3;
103 s1 = mkstr(m, len);
104 s2 = mkstr(s, len);
105 s3 = concat(s1, s2);
106 free(s2);
107 free(s1);
108 d = concat("(.?)", s3);
109 free(s3);
110 return d;
111}
112
113static char *
114p2(size_t len)
115{
116 return ps("((.*){0,255}", ")", len);
117}
118
119static char *
120p3(size_t len)
121{
122 return ps("(.\\{0,}", ")", len);
123}
124
125static char *
126p4(size_t len)
127{
128 return ps("((.*){1,255}", ")", len);
129}
130
131static char *
132p5(size_t len)
133{
134 return ps("(", "){1,100}", len);
135}
136
137static char *
138p6(size_t len)
139{
140 char *d, *s1, *s2;
141 s1 = mkstr("(?:(.*)|", len);
142 s2 = concat(s1, "(.*)");
143 free(s1);
144 s1 = mkstr(")", len);
145 d = concat(s2, s1);
146 free(s1);
147 free(s2);
148 return d;
149}
150
151static char *(*patterns[])(size_t) = {
152 p0,
153 p1,
154 p2,
155 p3,
156 p4,
157 p5,
158 p6,
159};
160
161int
162main(void)
163{
164 regex_t re;
165 int e, ret = 0;
166 size_t i;
167
168 for (i = 0; i < sizeof(patterns) / sizeof(patterns[0]); i++) {
169 char *d = (*patterns[i])(9999);
170 e = regcomp(&re, d, i == 6 ? REG_BASIC : REG_EXTENDED);
171 free(d);
172 if (e) {
173 if (e != REG_ESPACE) {
174 printf("regcomp returned %d for pattern %zu", e, i);
175 ret = 1;
176 }
177 continue;
178 }
179 (void)regexec(&re, "aaaaaaaa", 0, NULL, 0);
180 regfree(&re);
181 }
182 return ret;
183}
diff --git a/src/regress/lib/libc/regex/tests b/src/regress/lib/libc/regex/tests
deleted file mode 100644
index f3fd9d0e6b..0000000000
--- a/src/regress/lib/libc/regex/tests
+++ /dev/null
@@ -1,612 +0,0 @@
1# $OpenBSD: tests,v 1.10 2021/04/02 14:20:57 otto Exp $
2# $NetBSD: tests,v 1.5 1995/04/20 22:40:00 cgd Exp $
3
4# regular expression test set
5# Lines are at least three fields, separated by one or more tabs. "" stands
6# for an empty field. First field is an RE. Second field is flags. If
7# C flag given, regcomp() is expected to fail, and the third field is the
8# error name (minus the leading REG_).
9#
10# Otherwise it is expected to succeed, and the third field is the string to
11# try matching it against. If there is no fourth field, the match is
12# expected to fail. If there is a fourth field, it is the substring that
13# the RE is expected to match. If there is a fifth field, it is a comma-
14# separated list of what the subexpressions should match, with - indicating
15# no match for that one. In both the fourth and fifth fields, a (sub)field
16# starting with @ indicates that the (sub)expression is expected to match
17# a null string followed by the stuff after the @; this provides a way to
18# test where null strings match. The character `N' in REs and strings
19# is newline, `S' is space, `T' is tab, `Z' is NUL.
20#
21# The full list of flags:
22# - placeholder, does nothing
23# b RE is a BRE, not an ERE
24# & try it as both an ERE and a BRE
25# C regcomp() error expected, third field is error name
26# i REG_ICASE
27# m ("mundane") REG_NOSPEC
28# s REG_NOSUB (not really testable)
29# n REG_NEWLINE
30# ^ REG_NOTBOL
31# $ REG_NOTEOL
32# # REG_STARTEND (see below)
33# p REG_PEND
34#
35# For REG_STARTEND, the start/end offsets are those of the substring
36# enclosed in ().
37
38# basics
39a & a a
40abc & abc abc
41abc|de - abc abc
42a|b|c - abc a
43
44# parentheses and perversions thereof
45a(b)c - abc abc
46a\(b\)c b abc abc
47a( C EPAREN
48a( b a( a(
49a\( - a( a(
50a\( bC EPAREN
51a\(b bC EPAREN
52a(b C EPAREN
53a(b b a(b a(b
54# gag me with a right parenthesis -- 1003.2 goofed here (my fault, partly)
55a) - a) a)
56) - ) )
57# end gagging (in a just world, those *should* give EPAREN)
58a) b a) a)
59a\) bC EPAREN
60\) bC EPAREN
61a()b - ab ab
62a\(\)b b ab ab
63
64# anchoring and REG_NEWLINE
65^abc$ & abc abc
66^b & abc
67a^b - a^b
68a^b b a^b a^b
69a$b - a$b
70a$b b a$b a$b
71^ & abc @abc
72$ & abc @
73^$ & "" @
74$^ - "" @
75\($\)\(^\) b "" @
76# stop retching, those are legitimate (although disgusting)
77^^ - "" @
78$$ - "" @
79b$ & abNc
80b$ &n abNc b
81^b & aNbNc
82b$ & aNbNc
83^a &n aNb a
84^b &n abc
85^b$ &n aNbNc b
86^$ &n aNNb @Nb
87^$ n abc
88^$ n abcN @
89$^ n aNNb @Nb
90\($\)\(^\) bn aNNb @Nb
91^^ n^ aNNb @Nb
92$$ n aNNb @NN
93^a &^ a
94a$ &$ a
95^b &^ abc
96^b &^ aNb
97^a &^n aNb
98^b &^n abc
99^b &^n aNb b
100a$ &$n bNa
101b$ &$n bNa b
102a*(^b$)c* - b b
103a*\(^b$\)c* b b b
104
105# certain syntax errors and non-errors
106| C EMPTY
107| b | |
108* C BADRPT
109* b * *
110+ C BADRPT
111? C BADRPT
112"" &C EMPTY
113() - abc @abc
114\(\) b abc @abc
115a||b C EMPTY
116|ab C EMPTY
117ab| C EMPTY
118(|a)b C EMPTY
119(a|)b C EMPTY
120(*a) C BADRPT
121(+a) C BADRPT
122(?a) C BADRPT
123({1}a) C BADRPT
124\(\{1\}a\) bC BADRPT
125(a|*b) C BADRPT
126(a|+b) C BADRPT
127(a|?b) C BADRPT
128(a|{1}b) C BADRPT
129^* C BADRPT
130^* b * *
131^+ C BADRPT
132^? C BADRPT
133^{1} C BADRPT
134^\{1\} bC BADRPT
135
136# metacharacters, backslashes
137a.c & abc abc
138a[bc]d & abd abd
139a\*c & a*c a*c
140a\\b & a\b a\b
141a\\\*b & a\*b a\*b
142a\bc & abc abc
143a\ &C EESCAPE
144a\\bc & a\bc a\bc
145\{ bC BADRPT
146a\[b & a[b a[b
147a[b &C EBRACK
148# trailing $ is a peculiar special case for the BRE code
149a$ & a a
150a$ & a$
151a\$ & a
152a\$ & a$ a$
153a\\$ & a
154a\\$ & a$
155a\\$ & a\$
156a\\$ & a\ a\
157
158# back references, ugh
159a\(b\)\2c bC ESUBREG
160a\(b\1\)c bC ESUBREG
161a\(b*\)c\1d b abbcbbd abbcbbd bb
162a\(b*\)c\1d b abbcbd
163a\(b*\)c\1d b abbcbbbd
164^\(.\)\1 b abc
165a\([bc]\)\1d b abcdabbd abbd b
166a\(\([bc]\)\2\)*d b abbccd abbccd
167a\(\([bc]\)\2\)*d b abbcbd
168# actually, this next one probably ought to fail, but the spec is unclear
169a\(\(b\)*\2\)*d b abbbd abbbd
170# here is a case that no NFA implementation does right
171\(ab*\)[ab]*\1 b ababaaa ababaaa a
172# check out normal matching in the presence of back refs
173\(a\)\1bcd b aabcd aabcd
174\(a\)\1bc*d b aabcd aabcd
175\(a\)\1bc*d b aabd aabd
176\(a\)\1bc*d b aabcccd aabcccd
177\(a\)\1bc*[ce]d b aabcccd aabcccd
178^\(a\)\1b\(c\)*cd$ b aabcccd aabcccd
179\(b*\)\(a*\1\)* b ab a
180\([^_]*\)\(_*\1\)* b foo_foo_bar_bar_bar_baz foo_foo foo,_foo
181\([^_]*\)\(_*\1\)* b bar_bar_bar_baz bar_bar_bar bar,_bar
182\([^_]*\)\(_*\1\)* b foo_bar_baz foo foo
183\(.*\)\1 b "" ""
184\(.*\)\1 b a ""
185\(.*\)\1 b aa aa
186\(.*\)\1 b aaa aa
187\(.*\)\1 b aaaa aaaa
188\([^_]*\)\1 b "" ""
189\([^_]*\)\1 b a ""
190\([^_]*\)\1 b aa aa
191\([^_]*\)\1 b aaa aa
192\([^_]*\)\1 b aaaa aaaa
193foo\(.*\)bar\1 b foolbarl foolbarl l
194foo\(.*\)bar\1 b foobar foobar ""
195\(\(.\)b\)*\1 b aba
196\(\(.\)b\)*\1 b abba
197\(\(.\)b\)*\1 b abbba
198\(\(.\)b\)*\1 b abbbba bbbb bb,b
199\(\(.\)b\)*\1 b abbbbba abbbbb bb,b
200\(\(.\)b\)*\1 b abbbbbba abbbbb bb,b
201\(\(.\)b\)*\1 b abbbbbbbbbbbbbba abbbbbbbbbbbbb bb,b
202\(\(.\)b\)*\1 b abbbbbbbbbbbbbbba abbbbbbbbbbbbbbb bb,b
203# these used to segfault, buffer underflow in engine.c, backref()
204\(^a\)*\(b.\)\2 b^ sbxbxe bxbx -,bx
205\([[:<:]]a\)*\(b.\)\2 b^ sbxbxe bxbx -,bx
206
207# ordinary repetitions
208ab*c & abc abc
209ab+c - abc abc
210ab?c - abc abc
211a\(*\)b b a*b a*b
212a\(**\)b b ab ab
213a\(***\)b bC BADRPT
214*a b *a *a
215**a b a a
216***a bC BADRPT
217
218# the dreaded bounded repetitions
219{ & { {
220{abc & {abc {abc
221{1 C BADRPT
222{1} C BADRPT
223a{b & a{b a{b
224a{1}b - ab ab
225a\{1\}b b ab ab
226a{1,}b - ab ab
227a\{1,\}b b ab ab
228a{1,2}b - aab aab
229a\{1,2\}b b aab aab
230a{1 C EBRACE
231a\{1 bC EBRACE
232a{1a C EBRACE
233a\{1a bC EBRACE
234a{1a} C BADBR
235a\{1a\} bC BADBR
236a{,2} - a{,2} a{,2}
237a\{,2\} bC BADBR
238a{,} - a{,} a{,}
239a\{,\} bC BADBR
240a{1,x} C BADBR
241a\{1,x\} bC BADBR
242a{1,x C EBRACE
243a\{1,x bC EBRACE
244a{300} C BADBR
245a\{300\} bC BADBR
246a{1,0} C BADBR
247a\{1,0\} bC BADBR
248ab{0,0}c - abcac ac
249ab\{0,0\}c b abcac ac
250ab{0,1}c - abcac abc
251ab\{0,1\}c b abcac abc
252ab{0,3}c - abbcac abbc
253ab\{0,3\}c b abbcac abbc
254ab{1,1}c - acabc abc
255ab\{1,1\}c b acabc abc
256ab{1,3}c - acabc abc
257ab\{1,3\}c b acabc abc
258ab{2,2}c - abcabbc abbc
259ab\{2,2\}c b abcabbc abbc
260ab{2,4}c - abcabbc abbc
261ab\{2,4\}c b abcabbc abbc
262((a{1,10}){1,10}){1,10} - a a a,a
263
264# multiple repetitions
265a** &C BADRPT
266a++ C BADRPT
267a?? C BADRPT
268a*+ C BADRPT
269a*? C BADRPT
270a+* C BADRPT
271a+? C BADRPT
272a?* C BADRPT
273a?+ C BADRPT
274a{1}{1} C BADRPT
275a*{1} C BADRPT
276a+{1} C BADRPT
277a?{1} C BADRPT
278a{1}* C BADRPT
279a{1}+ C BADRPT
280a{1}? C BADRPT
281a*{b} - a{b} a{b}
282a\{1\}\{1\} bC BADRPT
283a*\{1\} bC BADRPT
284a\{1\}* bC BADRPT
285
286# brackets, and numerous perversions thereof
287a[b]c & abc abc
288a[ab]c & abc abc
289a[^ab]c & adc adc
290a[]b]c & a]c a]c
291a[[b]c & a[c a[c
292a[-b]c & a-c a-c
293a[^]b]c & adc adc
294a[^-b]c & adc adc
295a[b-]c & a-c a-c
296a[b &C EBRACK
297a[] &C EBRACK
298a[1-3]c & a2c a2c
299a[3-1]c &C ERANGE
300a[1-3-5]c &C ERANGE
301a[[.-.]--]c & a-c a-c
302a[1- &C ERANGE
303a[[. &C EBRACK
304a[[.x &C EBRACK
305a[[.x. &C EBRACK
306a[[.x.] &C EBRACK
307a[[.x.]] & ax ax
308a[[.x,.]] &C ECOLLATE
309a[[.one.]]b & a1b a1b
310a[[.notdef.]]b &C ECOLLATE
311a[[.].]]b & a]b a]b
312a[[:alpha:]]c & abc abc
313a[[:notdef:]]c &C ECTYPE
314a[[: &C EBRACK
315a[[:alpha &C EBRACK
316a[[:alpha:] &C EBRACK
317a[[:alpha,:] &C ECTYPE
318a[[:]:]]b &C ECTYPE
319a[[:-:]]b &C ECTYPE
320a[[:alph:]] &C ECTYPE
321a[[:alphabet:]] &C ECTYPE
322[[:alnum:]]+ - -%@a0X- a0X
323[[:alpha:]]+ - -%@aX0- aX
324[[:blank:]]+ - aSSTb SST
325[[:cntrl:]]+ - aNTb NT
326[[:digit:]]+ - a019b 019
327[[:graph:]]+ - Sa%bS a%b
328[[:lower:]]+ - AabC ab
329[[:print:]]+ - NaSbN aSb
330[[:punct:]]+ - S%-&T %-&
331[[:space:]]+ - aSNTb SNT
332[[:upper:]]+ - aBCd BC
333[[:xdigit:]]+ - p0f3Cq 0f3C
334a[[=b=]]c & abc abc
335a[[= &C EBRACK
336a[[=b &C EBRACK
337a[[=b= &C EBRACK
338a[[=b=] &C EBRACK
339a[[=b,=]] &C ECOLLATE
340a[[=one=]]b & a1b a1b
341
342# complexities
343a(((b)))c - abc abc
344a(b|(c))d - abd abd
345a(b*|c)d - abbd abbd
346# just gotta have one DFA-buster, of course
347a[ab]{20} - aaaaabaaaabaaaabaaaab aaaaabaaaabaaaabaaaab
348# and an inline expansion in case somebody gets tricky
349a[ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab] - aaaaabaaaabaaaabaaaab aaaaabaaaabaaaabaaaab
350# and in case somebody just slips in an NFA...
351a[ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab][ab](wee|week)(knights|night) - aaaaabaaaabaaaabaaaabweeknights aaaaabaaaabaaaabaaaabweeknights
352# fish for anomalies as the number of states passes 32
35312345678901234567890123456789 - a12345678901234567890123456789b 12345678901234567890123456789
354123456789012345678901234567890 - a123456789012345678901234567890b 123456789012345678901234567890
3551234567890123456789012345678901 - a1234567890123456789012345678901b 1234567890123456789012345678901
35612345678901234567890123456789012 - a12345678901234567890123456789012b 12345678901234567890123456789012
357123456789012345678901234567890123 - a123456789012345678901234567890123b 123456789012345678901234567890123
358# and one really big one, beyond any plausible word width
3591234567890123456789012345678901234567890123456789012345678901234567890 - a1234567890123456789012345678901234567890123456789012345678901234567890b 1234567890123456789012345678901234567890123456789012345678901234567890
360# fish for problems as brackets go past 8
361[ab][cd][ef][gh][ij][kl][mn] - xacegikmoq acegikm
362[ab][cd][ef][gh][ij][kl][mn][op] - xacegikmoq acegikmo
363[ab][cd][ef][gh][ij][kl][mn][op][qr] - xacegikmoqy acegikmoq
364[ab][cd][ef][gh][ij][kl][mn][op][q] - xacegikmoqy acegikmoq
365
366# subtleties of matching
367abc & xabcy abc
368a\(b\)?c\1d b acd
369aBc i Abc Abc
370a[Bc]*d i abBCcd abBCcd
3710[[:upper:]]1 &i 0a1 0a1
3720[[:lower:]]1 &i 0A1 0A1
373a[^b]c &i abc
374a[^b]c &i aBc
375a[^b]c &i adc adc
376[a]b[c] - abc abc
377[a]b[a] - aba aba
378[abc]b[abc] - abc abc
379[abc]b[abd] - abd abd
380a(b?c)+d - accd accd
381(wee|week)(knights|night) - weeknights weeknights
382(we|wee|week|frob)(knights|night|day) - weeknights weeknights
383a[bc]d - xyzaaabcaababdacd abd
384a[ab]c - aaabc abc
385abc s abc abc
386a* & b @b
387
388# Let's have some fun -- try to match a C comment.
389# first the obvious, which looks okay at first glance...
390/\*.*\*/ - /*x*/ /*x*/
391# but...
392/\*.*\*/ - /*x*/y/*z*/ /*x*/y/*z*/
393# okay, we must not match */ inside; try to do that...
394/\*([^*]|\*[^/])*\*/ - /*x*/ /*x*/
395/\*([^*]|\*[^/])*\*/ - /*x*/y/*z*/ /*x*/
396# but...
397/\*([^*]|\*[^/])*\*/ - /*x**/y/*z*/ /*x**/y/*z*/
398# and a still fancier version, which does it right (I think)...
399/\*([^*]|\*+[^*/])*\*+/ - /*x*/ /*x*/
400/\*([^*]|\*+[^*/])*\*+/ - /*x*/y/*z*/ /*x*/
401/\*([^*]|\*+[^*/])*\*+/ - /*x**/y/*z*/ /*x**/
402/\*([^*]|\*+[^*/])*\*+/ - /*x****/y/*z*/ /*x****/
403/\*([^*]|\*+[^*/])*\*+/ - /*x**x*/y/*z*/ /*x**x*/
404/\*([^*]|\*+[^*/])*\*+/ - /*x***x/y/*z*/ /*x***x/y/*z*/
405
406# subexpressions
407a(b)(c)d - abcd abcd b,c
408a(((b)))c - abc abc b,b,b
409a(b|(c))d - abd abd b,-
410a(b*|c|e)d - abbd abbd bb
411a(b*|c|e)d - acd acd c
412a(b*|c|e)d - ad ad @d
413a(b?)c - abc abc b
414a(b?)c - ac ac @c
415a(b+)c - abc abc b
416a(b+)c - abbbc abbbc bbb
417a(b*)c - ac ac @c
418(a|ab)(bc([de]+)f|cde) - abcdef abcdef a,bcdef,de
419# the regression tester only asks for 9 subexpressions
420a(b)(c)(d)(e)(f)(g)(h)(i)(j)k - abcdefghijk abcdefghijk b,c,d,e,f,g,h,i,j
421a(b)(c)(d)(e)(f)(g)(h)(i)(j)(k)l - abcdefghijkl abcdefghijkl b,c,d,e,f,g,h,i,j,k
422a([bc]?)c - abc abc b
423a([bc]?)c - ac ac @c
424a([bc]+)c - abc abc b
425a([bc]+)c - abcc abcc bc
426a([bc]+)bc - abcbc abcbc bc
427a(bb+|b)b - abb abb b
428a(bbb+|bb+|b)b - abb abb b
429a(bbb+|bb+|b)b - abbb abbb bb
430a(bbb+|bb+|b)bb - abbb abbb b
431(.*).* - abcdef abcdef abcdef
432(a*)* - bc @b @b
433
434# do we get the right subexpression when it is used more than once?
435a(b|c)*d - ad ad -
436a(b|c)*d - abcd abcd c
437a(b|c)+d - abd abd b
438a(b|c)+d - abcd abcd c
439a(b|c?)+d - ad ad @d
440a(b|c?)+d - abcd abcd @d
441a(b|c){0,0}d - ad ad -
442a(b|c){0,1}d - ad ad -
443a(b|c){0,1}d - abd abd b
444a(b|c){0,2}d - ad ad -
445a(b|c){0,2}d - abcd abcd c
446a(b|c){0,}d - ad ad -
447a(b|c){0,}d - abcd abcd c
448a(b|c){1,1}d - abd abd b
449a(b|c){1,1}d - acd acd c
450a(b|c){1,2}d - abd abd b
451a(b|c){1,2}d - abcd abcd c
452a(b|c){1,}d - abd abd b
453a(b|c){1,}d - abcd abcd c
454a(b|c){2,2}d - acbd acbd b
455a(b|c){2,2}d - abcd abcd c
456a(b|c){2,4}d - abcd abcd c
457a(b|c){2,4}d - abcbd abcbd b
458a(b|c){2,4}d - abcbcd abcbcd c
459a(b|c){2,}d - abcd abcd c
460a(b|c){2,}d - abcbd abcbd b
461a(b+|((c)*))+d - abd abd @d,@d,-
462a(b+|((c)*))+d - abcd abcd @d,@d,-
463
464# check out the STARTEND option
465^[abc] &# a(b)c b
466^[abc] &# a(xb)c
467^[abc] &# aN(b)c b
468^[abc] &n# a(b)c b
469^[abc] &n# a(xb)c
470^[abc] &n# aN(b)c b
471^[abc] &^# a(b)c
472^[abc] &^# a(xb)c
473^[abc] &^# aN(b)c
474^[abc] &n^# a(b)c
475^[abc] &n^# a(xb)c
476^[abc] &n^# aN(b)c b
477[abc] &# a(d)c
478[abc] &# a(bc)d b
479[abc] &# a(dc)d c
480. &# a()c
481b.*c &# b(bc)c bc
482b.* &# b(bc)c bc
483.*c &# b(bc)c bc
484
485# plain strings, with the NOSPEC flag
486abc m abc abc
487abc m xabcy abc
488abc m xyz
489a*b m aba*b a*b
490a*b m ab
491"" mC EMPTY
492
493# cases involving NULs
494aZb & a a
495aZb &p a
496aZb &p# (aZb) aZb
497aZ*b &p# (ab) ab
498a.b &# (aZb) aZb
499a.* &# (aZb)c aZb
500
501# word boundaries (ick)
502[[:<:]]a & a a
503\<a & a a
504[[:<:]]a & ba
505\<a & ba
506[[:<:]]a & -a a
507\<a & -a a
508[[:<:]]a & Na a
509\<a & Na a
510[[:<:]]a &n a a
511\<a &n a a
512[[:<:]]a &n ba
513\<a &n ba
514[[:<:]]a &n -a a
515\<a &n -a a
516[[:<:]]a &n Na a
517\<a &n Na a
518[[:<:]]a &^ a
519\<a &^ a
520[[:<:]]a &^ ba
521\<a &^ ba
522[[:<:]]a &^ -a a
523\<a &^ -a a
524[[:<:]]a &^ Na a
525\<a &^ Na a
526[[:<:]]a &n^ a
527\<a &n^ a
528[[:<:]]a &n^ ba
529\<a &n^ ba
530[[:<:]]a &n^ -a a
531\<a &n^ -a a
532[[:<:]]a &n^ Na a
533\<a &n^ Na a
534[[:<:]]b &# a(b)c b
535\<b &# a(b)c b
536[[:<:]]b &# a(xb)c
537\<b &# a(xb)c
538[[:<:]]b &# -(b)c b
539\<b &# -(b)c b
540[[:<:]]b &# aN(b)c b
541\<b &# aN(b)c b
542[[:<:]]b &n# a(b)c b
543\<b &n# a(b)c b
544[[:<:]]b &n# a(xb)c
545\<b &n# a(xb)c
546[[:<:]]b &n# -(b)c b
547\<b &n# -(b)c b
548[[:<:]]b &n# aN(b)c b
549\<b &n# aN(b)c b
550[[:<:]]b &^# a(b)c
551\<b &^# a(b)c
552[[:<:]]b &^# a(xb)c
553\<b &^# a(xb)c
554[[:<:]]b &^# -(b)c b
555\<b &^# -(b)c b
556[[:<:]]b &^# aN(b)c b
557\<b &^# aN(b)c b
558[[:<:]]b &n^# a(b)c
559\<b &n^# a(b)c
560[[:<:]]b &n^# a(xb)c
561\<b &n^# a(xb)c
562[[:<:]]b &n^# -(b)c b
563\<b &n^# -(b)c b
564[[:<:]]b &n^# aN(b)c b
565\<b &n^# aN(b)c b
566a[[:>:]] & a a
567a\> & a a
568a[[:>:]] & ab
569a\> & ab
570a[[:>:]] & a- a
571a\> & a- a
572[[:<:]]a.c[[:>:]] & axcd-dayc-dazce-abc abc
573\<a.c\> & axcd-dayc-dazce-abc abc
574[[:<:]]a.c[[:>:]] & axcd-dayc-dazce-abc-q abc
575\<a.c\> & axcd-dayc-dazce-abc-q abc
576[[:<:]]a.c[[:>:]] & axc-dayc-dazce-abc axc
577\<a.c\> & axc-dayc-dazce-abc axc
578[[:<:]]b.c[[:>:]] & a_bxc-byc_d-bzc-q bzc
579\<b.c\> & a_bxc-byc_d-bzc-q bzc
580[[:<:]].x..[[:>:]] & y_xa_-_xb_y-_xc_-axdc _xc_
581\<.x..\> & y_xa_-_xb_y-_xc_-axdc _xc_
582[[:<:]]a_b[[:>:]] & x_a_b
583\<a_b\> & x_a_b
584
585# past problems, and suspected problems
586(A[1])|(A[2])|(A[3])|(A[4])|(A[5])|(A[6])|(A[7])|(A[8])|(A[9])|(A[A]) - A1 A1
587abcdefghijklmnop i abcdefghijklmnop abcdefghijklmnop
588abcdefghijklmnopqrstuv i abcdefghijklmnopqrstuv abcdefghijklmnopqrstuv
589(ALAK)|(ALT[AB])|(CC[123]1)|(CM[123]1)|(GAMC)|(LC[23][EO ])|(SEM[1234])|(SL[ES][12])|(SLWW)|(SLF )|(SLDT)|(VWH[12])|(WH[34][EW])|(WP1[ESN]) - CC11 CC11
590CC[13]1|a{21}[23][EO][123][Es][12]a{15}aa[34][EW]aaaaaaa[X]a - CC11 CC11
591Char \([a-z0-9_]*\)\[.* b Char xyz[k Char xyz[k xyz
592a?b - ab ab
593-\{0,1\}[0-9]*$ b -5 -5
594
595# FreeBSD PR 130504
596(.|())(b) - ab ab
597(()|.)(b) - ab ab
598
599# Some BRE cases where \{0,\} makes a backref go wrong, as reported by Michael Paoli
600Y*\(x\)\1 b YYxx YYxx
601Y\{2,\}\(x\)\1 b YYxx YYxx
602# Fails currently
603#Y\{0,\}\(x\)\1 b YYxx YYxx
604Y\{0,\}\(x\) b YYxx YYx
605Y\{2,\}x\{1,\} b YYxx YYxx
606Y\{2,\}x\{0,\}z b YYxxz YYxxz
607Y\{0,\}x\{0,\}z b YYxxz YYxxz
608Y\{2,\}\(xy\)\1 b YYxyxy YYxyxy
609# Fails currently
610#Y\{0,\}\(xy\)\1 b YYxyxy YYxyxy
611Y*\(xy\)\1 b YYxyxy YYxyxy
612Y\{0,\}\(xy\)xy b YYxyxy YYxyxy
generated by cgit v1.2.3-55-g6feb (git 2.39.1) at 2025-04-22 02:46:36 +0000