9007 import AT&T regex test suite
[unleashed.git] / usr / src / test / libc-tests / tests / regex / testregex.c
blob0dd7b646803b2aa74c2f69091ef6d007368d9538
1 #pragma prototyped noticed
3 /*
4 * regex(3) test harness
6 * build: cc -o testregex testregex.c
7 * help: testregex --man
8 * note: REG_* features are detected by #ifdef; if REG_* are enums
9 * then supply #define REG_foo REG_foo for each enum REG_foo
11 * Glenn Fowler <glenn.s.fowler@gmail.com>
12 * AT&T Research
14 * PLEASE: publish your tests so everyone can benefit
16 * The following license covers testregex.c and all associated test data.
18 * Permission is hereby granted, free of charge, to any person obtaining a
19 * copy of THIS SOFTWARE FILE (the "Software"), to deal in the Software
20 * without restriction, including without limitation the rights to use,
21 * copy, modify, merge, publish, distribute, and/or sell copies of the
22 * Software, and to permit persons to whom the Software is furnished to do
23 * so, subject to the following disclaimer:
25 * THIS SOFTWARE IS PROVIDED BY AT&T ``AS IS'' AND ANY EXPRESS OR IMPLIED
26 * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
27 * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
28 * IN NO EVENT SHALL AT&T BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
29 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
30 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
31 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
32 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
33 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
34 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
37 static const char id[] = "\n@(#)$Id: testregex (AT&T Research) 2010-06-10 $\0\n";
39 #if _PACKAGE_ast
40 #include <ast.h>
41 #else
42 #include <sys/types.h>
43 #endif
45 #include <stdio.h>
46 #include <regex.h>
47 #include <ctype.h>
48 #include <setjmp.h>
49 #include <signal.h>
50 #include <string.h>
51 #include <unistd.h>
53 #ifdef __STDC__
54 #include <stdlib.h>
55 #include <locale.h>
56 #endif
58 #ifndef RE_DUP_MAX
59 #define RE_DUP_MAX 32767
60 #endif
62 #if !_PACKAGE_ast
63 #undef REG_DISCIPLINE
64 #endif
66 #ifndef REG_DELIMITED
67 #undef _REG_subcomp
68 #endif
70 #define TEST_ARE 0x00000001
71 #define TEST_BRE 0x00000002
72 #define TEST_ERE 0x00000004
73 #define TEST_KRE 0x00000008
74 #define TEST_LRE 0x00000010
75 #define TEST_SRE 0x00000020
77 #define TEST_EXPAND 0x00000100
78 #define TEST_LENIENT 0x00000200
80 #define TEST_QUERY 0x00000400
81 #define TEST_SUB 0x00000800
82 #define TEST_UNSPECIFIED 0x00001000
83 #define TEST_VERIFY 0x00002000
84 #define TEST_AND 0x00004000
85 #define TEST_OR 0x00008000
87 #define TEST_DELIMIT 0x00010000
88 #define TEST_OK 0x00020000
89 #define TEST_SAME 0x00040000
91 #define TEST_ACTUAL 0x00100000
92 #define TEST_BASELINE 0x00200000
93 #define TEST_FAIL 0x00400000
94 #define TEST_PASS 0x00800000
95 #define TEST_SUMMARY 0x01000000
97 #define TEST_IGNORE_ERROR 0x02000000
98 #define TEST_IGNORE_OVER 0x04000000
99 #define TEST_IGNORE_POSITION 0x08000000
101 #define TEST_CATCH 0x10000000
102 #define TEST_VERBOSE 0x20000000
104 #define TEST_DECOMP 0x40000000
106 #define TEST_GLOBAL (TEST_ACTUAL|TEST_AND|TEST_BASELINE|TEST_CATCH|TEST_FAIL|TEST_IGNORE_ERROR|TEST_IGNORE_OVER|TEST_IGNORE_POSITION|TEST_OR|TEST_PASS|TEST_SUMMARY|TEST_VERBOSE)
108 #ifdef REG_DISCIPLINE
111 #include <stk.h>
113 typedef struct Disc_s
115 regdisc_t disc;
116 int ordinal;
117 Sfio_t* sp;
118 } Disc_t;
120 static void*
121 compf(const regex_t* re, const char* xstr, size_t xlen, regdisc_t* disc)
123 Disc_t* dp = (Disc_t*)disc;
125 return (void*)((char*)0 + ++dp->ordinal);
128 static int
129 execf(const regex_t* re, void* data, const char* xstr, size_t xlen, const char* sstr, size_t slen, char** snxt, regdisc_t* disc)
131 Disc_t* dp = (Disc_t*)disc;
133 sfprintf(dp->sp, "{%-.*s}(%lu:%d)", xlen, xstr, (char*)data - (char*)0, slen);
134 return atoi(xstr);
137 static void*
138 resizef(void* handle, void* data, size_t size)
140 if (!size)
141 return 0;
142 return stkalloc((Sfio_t*)handle, size);
145 #endif
147 #ifndef NiL
148 #ifdef __STDC__
149 #define NiL 0
150 #else
151 #define NiL (char*)0
152 #endif
153 #endif
155 #define H(x) do{if(html)fprintf(stderr,x);}while(0)
156 #define T(x) fprintf(stderr,x)
158 static void
159 help(int html)
161 H("<!DOCTYPE HTML PUBLIC \"-//IETF//DTD HTML//EN\">\n");
162 H("<HTML>\n");
163 H("<HEAD>\n");
164 H("<TITLE>testregex man document</TITLE>\n");
165 H("</HEAD>\n");
166 H("<BODY bgcolor=white>\n");
167 H("<PRE>\n");
168 T("NAME\n");
169 T(" testregex - regex(3) test harness\n");
170 T("\n");
171 T("SYNOPSIS\n");
172 T(" testregex [ options ]\n");
173 T("\n");
174 T("DESCRIPTION\n");
175 T(" testregex reads regex(3) test specifications, one per line, from the\n");
176 T(" standard input and writes one output line for each failed test. A\n");
177 T(" summary line is written after all tests are done. Each successful\n");
178 T(" test is run again with REG_NOSUB. Unsupported features are noted\n");
179 T(" before the first test, and tests requiring these features are\n");
180 T(" silently ignored.\n");
181 T("\n");
182 T("OPTIONS\n");
183 T(" -c catch signals and non-terminating calls\n");
184 T(" -e ignore error return mismatches\n");
185 T(" -h list help on standard error\n");
186 T(" -n do not repeat successful tests with regnexec()\n");
187 T(" -o ignore match[] overrun errors\n");
188 T(" -p ignore negative position mismatches\n");
189 T(" -s use stack instead of malloc\n");
190 T(" -x do not repeat successful tests with REG_NOSUB\n");
191 T(" -v list each test line\n");
192 T(" -A list failed test lines with actual answers\n");
193 T(" -B list all test lines with actual answers\n");
194 T(" -F list failed test lines\n");
195 T(" -P list passed test lines\n");
196 T(" -S output one summary line\n");
197 T("\n");
198 T("INPUT FORMAT\n");
199 T(" Input lines may be blank, a comment beginning with #, or a test\n");
200 T(" specification. A specification is five fields separated by one\n");
201 T(" or more tabs. NULL denotes the empty string and NIL denotes the\n");
202 T(" 0 pointer.\n");
203 T("\n");
204 T(" Field 1: the regex(3) flags to apply, one character per REG_feature\n");
205 T(" flag. The test is skipped if REG_feature is not supported by the\n");
206 T(" implementation. If the first character is not [BEASKLP] then the\n");
207 T(" specification is a global control line. One or more of [BEASKLP] may be\n");
208 T(" specified; the test will be repeated for each mode.\n");
209 T("\n");
210 T(" B basic BRE (grep, ed, sed)\n");
211 T(" E REG_EXTENDED ERE (egrep)\n");
212 T(" A REG_AUGMENTED ARE (egrep with negation)\n");
213 T(" S REG_SHELL SRE (sh glob)\n");
214 T(" K REG_SHELL|REG_AUGMENTED KRE (ksh glob)\n");
215 T(" L REG_LITERAL LRE (fgrep)\n");
216 T("\n");
217 T(" a REG_LEFT|REG_RIGHT implicit ^...$\n");
218 T(" b REG_NOTBOL lhs does not match ^\n");
219 T(" c REG_COMMENT ignore space and #...\\n\n");
220 T(" d REG_SHELL_DOT explicit leading . match\n");
221 T(" e REG_NOTEOL rhs does not match $\n");
222 T(" f REG_MULTIPLE multiple \\n separated patterns\n");
223 T(" g FNM_LEADING_DIR testfnmatch only -- match until /\n");
224 T(" h REG_MULTIREF multiple digit backref\n");
225 T(" i REG_ICASE ignore case\n");
226 T(" j REG_SPAN . matches \\n\n");
227 T(" k REG_ESCAPE \\ to ecape [...] delimiter\n");
228 T(" l REG_LEFT implicit ^...\n");
229 T(" m REG_MINIMAL minimal match\n");
230 T(" n REG_NEWLINE explicit \\n match\n");
231 T(" o REG_ENCLOSED (|&) magic inside [@|&](...)\n");
232 T(" p REG_SHELL_PATH explicit / match\n");
233 T(" q REG_DELIMITED delimited pattern\n");
234 T(" r REG_RIGHT implicit ...$\n");
235 T(" s REG_SHELL_ESCAPED \\ not special\n");
236 T(" t REG_MUSTDELIM all delimiters must be specified\n");
237 T(" u standard unspecified behavior -- errors not counted\n");
238 T(" v REG_CLASS_ESCAPE \\ special inside [...]\n");
239 T(" w REG_NOSUB no subexpression match array\n");
240 T(" x REG_LENIENT let some errors slide\n");
241 T(" y REG_LEFT regexec() implicit ^...\n");
242 T(" z REG_NULL NULL subexpressions ok\n");
243 T(" $ expand C \\c escapes in fields 2 and 3\n");
244 T(" / field 2 is a regsubcomp() expression\n");
245 T(" = field 3 is a regdecomp() expression\n");
246 T("\n");
247 T(" Field 1 control lines:\n");
248 T("\n");
249 T(" C set LC_COLLATE and LC_CTYPE to locale in field 2\n");
250 T("\n");
251 T(" ?test ... output field 5 if passed and != EXPECTED, silent otherwise\n");
252 T(" &test ... output field 5 if current and previous passed\n");
253 T(" |test ... output field 5 if current passed and previous failed\n");
254 T(" ; ... output field 2 if previous failed\n");
255 T(" {test ... skip if failed until }\n");
256 T(" } end of skip\n");
257 T("\n");
258 T(" : comment comment copied as output NOTE\n");
259 T(" :comment:test :comment: ignored\n");
260 T(" N[OTE] comment comment copied as output NOTE\n");
261 T(" T[EST] comment comment\n");
262 T("\n");
263 T(" number use number for nmatch (20 by default)\n");
264 T("\n");
265 T(" Field 2: the regular expression pattern; SAME uses the pattern from\n");
266 T(" the previous specification. RE_DUP_MAX inside {...} expands to the\n");
267 T(" value from <limits.h>.\n");
268 T("\n");
269 T(" Field 3: the string to match. X...{RE_DUP_MAX} expands to RE_DUP_MAX\n");
270 T(" copies of X.\n");
271 T("\n");
272 T(" Field 4: the test outcome. This is either one of the posix error\n");
273 T(" codes (with REG_ omitted) or the match array, a list of (m,n)\n");
274 T(" entries with m and n being first and last+1 positions in the\n");
275 T(" field 3 string, or NULL if REG_NOSUB is in effect and success\n");
276 T(" is expected. BADPAT is acceptable in place of any regcomp(3)\n");
277 T(" error code. The match[] array is initialized to (-2,-2) before\n");
278 T(" each test. All array elements from 0 to nmatch-1 must be specified\n");
279 T(" in the outcome. Unspecified endpoints (offset -1) are denoted by ?.\n");
280 T(" Unset endpoints (offset -2) are denoted by X. {x}(o:n) denotes a\n");
281 T(" matched (?{...}) expression, where x is the text enclosed by {...},\n");
282 T(" o is the expression ordinal counting from 1, and n is the length of\n");
283 T(" the unmatched portion of the subject string. If x starts with a\n");
284 T(" number then that is the return value of re_execf(), otherwise 0 is\n");
285 T(" returned. RE_DUP_MAX[-+]N expands to the <limits.h> value -+N.\n");
286 T("\n");
287 T(" Field 5: optional comment appended to the report.\n");
288 T("\n");
289 T("CAVEAT\n");
290 T(" If a regex implementation misbehaves with memory then all bets are off.\n");
291 T("\n");
292 T("CONTRIBUTORS\n");
293 T(" Glenn Fowler glenn.s.fowler@gmail.com (ksh strmatch, regex extensions)\n");
294 T(" David Korn dgkorn@gmail.com (ksh glob matcher)\n");
295 T(" Doug McIlroy mcilroy@dartmouth.edu (ast regex/testre in C++)\n");
296 T(" Tom Lord lord@regexps.com (rx tests)\n");
297 T(" Henry Spencer henry@zoo.toronto.edu (original public regex)\n");
298 T(" Andrew Hume andrew@research.att.com (gre tests)\n");
299 T(" John Maddock John_Maddock@compuserve.com (regex++ tests)\n");
300 T(" Philip Hazel ph10@cam.ac.uk (pcre tests)\n");
301 T(" Ville Laurikari vl@iki.fi (libtre tests)\n");
302 H("</PRE>\n");
303 H("</BODY>\n");
304 H("</HTML>\n");
307 #ifndef elementsof
308 #define elementsof(x) (sizeof(x)/sizeof(x[0]))
309 #endif
311 #ifndef streq
312 #define streq(a,b) (*(a)==*(b)&&!strcmp(a,b))
313 #endif
315 #define HUNG 2
316 #define NOTEST (~0)
318 #ifndef REG_TEST_DEFAULT
319 #define REG_TEST_DEFAULT 0
320 #endif
322 #ifndef REG_EXEC_DEFAULT
323 #define REG_EXEC_DEFAULT 0
324 #endif
326 static const char* unsupported[] =
328 "BASIC",
329 #ifndef REG_EXTENDED
330 "EXTENDED",
331 #endif
332 #ifndef REG_AUGMENTED
333 "AUGMENTED",
334 #endif
335 #ifndef REG_SHELL
336 "SHELL",
337 #endif
339 #ifndef REG_CLASS_ESCAPE
340 "CLASS_ESCAPE",
341 #endif
342 #ifndef REG_COMMENT
343 "COMMENT",
344 #endif
345 #ifndef REG_DELIMITED
346 "DELIMITED",
347 #endif
348 #ifndef REG_DISCIPLINE
349 "DISCIPLINE",
350 #endif
351 #ifndef REG_ESCAPE
352 "ESCAPE",
353 #endif
354 #ifndef REG_ICASE
355 "ICASE",
356 #endif
357 #ifndef REG_LEFT
358 "LEFT",
359 #endif
360 #ifndef REG_LENIENT
361 "LENIENT",
362 #endif
363 #ifndef REG_LITERAL
364 "LITERAL",
365 #endif
366 #ifndef REG_MINIMAL
367 "MINIMAL",
368 #endif
369 #ifndef REG_MULTIPLE
370 "MULTIPLE",
371 #endif
372 #ifndef REG_MULTIREF
373 "MULTIREF",
374 #endif
375 #ifndef REG_MUSTDELIM
376 "MUSTDELIM",
377 #endif
378 #ifndef REG_NEWLINE
379 "NEWLINE",
380 #endif
381 #ifndef REG_NOTBOL
382 "NOTBOL",
383 #endif
384 #ifndef REG_NOTEOL
385 "NOTEOL",
386 #endif
387 #ifndef REG_NULL
388 "NULL",
389 #endif
390 #ifndef REG_RIGHT
391 "RIGHT",
392 #endif
393 #ifndef REG_SHELL_DOT
394 "SHELL_DOT",
395 #endif
396 #ifndef REG_SHELL_ESCAPED
397 "SHELL_ESCAPED",
398 #endif
399 #ifndef REG_SHELL_GROUP
400 "SHELL_GROUP",
401 #endif
402 #ifndef REG_SHELL_PATH
403 "SHELL_PATH",
404 #endif
405 #ifndef REG_SPAN
406 "SPAN",
407 #endif
408 #if REG_NOSUB & REG_TEST_DEFAULT
409 "SUBMATCH",
410 #endif
411 #if !_REG_nexec
412 "regnexec",
413 #endif
414 #if !_REG_subcomp
415 "regsubcomp",
416 #endif
417 #if !_REG_decomp
418 "redecomp",
419 #endif
423 #ifndef REG_CLASS_ESCAPE
424 #define REG_CLASS_ESCAPE NOTEST
425 #endif
426 #ifndef REG_COMMENT
427 #define REG_COMMENT NOTEST
428 #endif
429 #ifndef REG_DELIMITED
430 #define REG_DELIMITED NOTEST
431 #endif
432 #ifndef REG_ESCAPE
433 #define REG_ESCAPE NOTEST
434 #endif
435 #ifndef REG_ICASE
436 #define REG_ICASE NOTEST
437 #endif
438 #ifndef REG_LEFT
439 #define REG_LEFT NOTEST
440 #endif
441 #ifndef REG_LENIENT
442 #define REG_LENIENT 0
443 #endif
444 #ifndef REG_MINIMAL
445 #define REG_MINIMAL NOTEST
446 #endif
447 #ifndef REG_MULTIPLE
448 #define REG_MULTIPLE NOTEST
449 #endif
450 #ifndef REG_MULTIREF
451 #define REG_MULTIREF NOTEST
452 #endif
453 #ifndef REG_MUSTDELIM
454 #define REG_MUSTDELIM NOTEST
455 #endif
456 #ifndef REG_NEWLINE
457 #define REG_NEWLINE NOTEST
458 #endif
459 #ifndef REG_NOTBOL
460 #define REG_NOTBOL NOTEST
461 #endif
462 #ifndef REG_NOTEOL
463 #define REG_NOTEOL NOTEST
464 #endif
465 #ifndef REG_NULL
466 #define REG_NULL NOTEST
467 #endif
468 #ifndef REG_RIGHT
469 #define REG_RIGHT NOTEST
470 #endif
471 #ifndef REG_SHELL_DOT
472 #define REG_SHELL_DOT NOTEST
473 #endif
474 #ifndef REG_SHELL_ESCAPED
475 #define REG_SHELL_ESCAPED NOTEST
476 #endif
477 #ifndef REG_SHELL_GROUP
478 #define REG_SHELL_GROUP NOTEST
479 #endif
480 #ifndef REG_SHELL_PATH
481 #define REG_SHELL_PATH NOTEST
482 #endif
483 #ifndef REG_SPAN
484 #define REG_SPAN NOTEST
485 #endif
487 #define REG_UNKNOWN (-1)
489 #ifndef REG_ENEWLINE
490 #define REG_ENEWLINE (REG_UNKNOWN-1)
491 #endif
492 #ifndef REG_ENULL
493 #ifndef REG_EMPTY
494 #define REG_ENULL (REG_UNKNOWN-2)
495 #else
496 #define REG_ENULL REG_EMPTY
497 #endif
498 #endif
499 #ifndef REG_ECOUNT
500 #define REG_ECOUNT (REG_UNKNOWN-3)
501 #endif
502 #ifndef REG_BADESC
503 #define REG_BADESC (REG_UNKNOWN-4)
504 #endif
505 #ifndef REG_EMEM
506 #define REG_EMEM (REG_UNKNOWN-5)
507 #endif
508 #ifndef REG_EHUNG
509 #define REG_EHUNG (REG_UNKNOWN-6)
510 #endif
511 #ifndef REG_EBUS
512 #define REG_EBUS (REG_UNKNOWN-7)
513 #endif
514 #ifndef REG_EFAULT
515 #define REG_EFAULT (REG_UNKNOWN-8)
516 #endif
517 #ifndef REG_EFLAGS
518 #define REG_EFLAGS (REG_UNKNOWN-9)
519 #endif
520 #ifndef REG_EDELIM
521 #define REG_EDELIM (REG_UNKNOWN-9)
522 #endif
524 static const struct { int code; char* name; } codes[] =
526 REG_UNKNOWN, "UNKNOWN",
527 REG_NOMATCH, "NOMATCH",
528 REG_BADPAT, "BADPAT",
529 REG_ECOLLATE, "ECOLLATE",
530 REG_ECTYPE, "ECTYPE",
531 REG_EESCAPE, "EESCAPE",
532 REG_ESUBREG, "ESUBREG",
533 REG_EBRACK, "EBRACK",
534 REG_EPAREN, "EPAREN",
535 REG_EBRACE, "EBRACE",
536 REG_BADBR, "BADBR",
537 REG_ERANGE, "ERANGE",
538 REG_ESPACE, "ESPACE",
539 REG_BADRPT, "BADRPT",
540 REG_ENEWLINE, "ENEWLINE",
541 REG_ENULL, "ENULL",
542 REG_ECOUNT, "ECOUNT",
543 REG_BADESC, "BADESC",
544 REG_EMEM, "EMEM",
545 REG_EHUNG, "EHUNG",
546 REG_EBUS, "EBUS",
547 REG_EFAULT, "EFAULT",
548 REG_EFLAGS, "EFLAGS",
549 REG_EDELIM, "EDELIM",
552 static struct
554 regmatch_t NOMATCH;
555 int errors;
556 int extracted;
557 int ignored;
558 int lineno;
559 int passed;
560 int signals;
561 int unspecified;
562 int verify;
563 int warnings;
564 char* file;
565 char* stack;
566 char* which;
567 jmp_buf gotcha;
568 #ifdef REG_DISCIPLINE
569 Disc_t disc;
570 #endif
571 } state;
573 static void
574 quote(char* s, int len, unsigned long test)
576 unsigned char* u = (unsigned char*)s;
577 unsigned char* e;
578 int c;
579 #ifdef MB_CUR_MAX
580 int w;
581 #endif
583 if (!u)
584 printf("NIL");
585 else if (!*u && len <= 1)
586 printf("NULL");
587 else if (test & TEST_EXPAND)
589 if (len < 0)
590 len = strlen((char*)u);
591 e = u + len;
592 if (test & TEST_DELIMIT)
593 printf("\"");
594 while (u < e)
595 switch (c = *u++)
597 case '\\':
598 printf("\\\\");
599 break;
600 case '"':
601 if (test & TEST_DELIMIT)
602 printf("\\\"");
603 else
604 printf("\"");
605 break;
606 case '\a':
607 printf("\\a");
608 break;
609 case '\b':
610 printf("\\b");
611 break;
612 case 033:
613 printf("\\e");
614 break;
615 case '\f':
616 printf("\\f");
617 break;
618 case '\n':
619 printf("\\n");
620 break;
621 case '\r':
622 printf("\\r");
623 break;
624 case '\t':
625 printf("\\t");
626 break;
627 case '\v':
628 printf("\\v");
629 break;
630 default:
631 #ifdef MB_CUR_MAX
632 s = (char*)u - 1;
633 if ((w = mblen(s, (char*)e - s)) > 1)
635 u += w - 1;
636 fwrite(s, 1, w, stdout);
638 else
639 #endif
640 if (!iscntrl(c) && isprint(c))
641 putchar(c);
642 else
643 printf("\\x%02x", c);
644 break;
646 if (test & TEST_DELIMIT)
647 printf("\"");
649 else
650 printf("%s", s);
653 static void
654 report(char* comment, char* fun, char* re, char* s, int len, char* msg, int flags, unsigned long test)
656 if (state.file)
657 printf("%s:", state.file);
658 printf("%d:", state.lineno);
659 if (re)
661 printf(" ");
662 quote(re, -1, test|TEST_DELIMIT);
663 if (s)
665 printf(" versus ");
666 quote(s, len, test|TEST_DELIMIT);
669 if (test & TEST_UNSPECIFIED)
671 state.unspecified++;
672 printf(" unspecified behavior");
674 else
675 state.errors++;
676 if (state.which)
677 printf(" %s", state.which);
678 if (flags & REG_NOSUB)
679 printf(" NOSUB");
680 if (fun)
681 printf(" %s", fun);
682 if (comment[strlen(comment)-1] == '\n')
683 printf(" %s", comment);
684 else
686 printf(" %s: ", comment);
687 if (msg)
688 printf("%s: ", msg);
692 static void
693 error(regex_t* preg, int code)
695 char* msg;
696 char buf[256];
698 switch (code)
700 case REG_EBUS:
701 msg = "bus error";
702 break;
703 case REG_EFAULT:
704 msg = "memory fault";
705 break;
706 case REG_EHUNG:
707 msg = "did not terminate";
708 break;
709 default:
710 regerror(code, preg, msg = buf, sizeof buf);
711 break;
713 printf("%s\n", msg);
716 static void
717 bad(char* comment, char* re, char* s, int len, unsigned long test)
719 printf("bad test case ");
720 report(comment, NiL, re, s, len, NiL, 0, test);
721 exit(1);
724 static int
725 escape(char* s)
727 char* b;
728 char* t;
729 char* q;
730 char* e;
731 int c;
733 for (b = t = s; *t = *s; s++, t++)
734 if (*s == '\\')
735 switch (*++s)
737 case '\\':
738 break;
739 case 'a':
740 *t = '\a';
741 break;
742 case 'b':
743 *t = '\b';
744 break;
745 case 'c':
746 if (*t = *++s)
747 *t &= 037;
748 else
749 s--;
750 break;
751 case 'e':
752 case 'E':
753 *t = 033;
754 break;
755 case 'f':
756 *t = '\f';
757 break;
758 case 'n':
759 *t = '\n';
760 break;
761 case 'r':
762 *t = '\r';
763 break;
764 case 's':
765 *t = ' ';
766 break;
767 case 't':
768 *t = '\t';
769 break;
770 case 'v':
771 *t = '\v';
772 break;
773 case 'u':
774 case 'x':
775 c = 0;
776 q = c == 'u' ? (s + 5) : (char*)0;
777 e = s + 1;
778 while (!e || !q || s < q)
780 switch (*++s)
782 case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
783 c = (c << 4) + *s - 'a' + 10;
784 continue;
785 case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
786 c = (c << 4) + *s - 'A' + 10;
787 continue;
788 case '0': case '1': case '2': case '3': case '4':
789 case '5': case '6': case '7': case '8': case '9':
790 c = (c << 4) + *s - '0';
791 continue;
792 case '{':
793 case '[':
794 if (s != e)
796 s--;
797 break;
799 e = 0;
800 continue;
801 case '}':
802 case ']':
803 if (e)
804 s--;
805 break;
806 default:
807 s--;
808 break;
810 break;
812 *t = c;
813 break;
814 case '0': case '1': case '2': case '3':
815 case '4': case '5': case '6': case '7':
816 c = *s - '0';
817 q = s + 2;
818 while (s < q)
820 switch (*++s)
822 case '0': case '1': case '2': case '3':
823 case '4': case '5': case '6': case '7':
824 c = (c << 3) + *s - '0';
825 break;
826 default:
827 q = --s;
828 break;
831 *t = c;
832 break;
833 default:
834 *(s + 1) = 0;
835 bad("invalid C \\ escape\n", s - 1, NiL, 0, 0);
837 return t - b;
840 static void
841 matchoffprint(int off)
843 switch (off)
845 case -2:
846 printf("X");
847 break;
848 case -1:
849 printf("?");
850 break;
851 default:
852 printf("%d", off);
853 break;
857 static void
858 matchprint(regmatch_t* match, int nmatch, int nsub, char* ans, unsigned long test)
860 int i;
862 for (; nmatch > nsub + 1; nmatch--)
863 if ((match[nmatch-1].rm_so != -1 || match[nmatch-1].rm_eo != -1) && (!(test & TEST_IGNORE_POSITION) || match[nmatch-1].rm_so >= 0 && match[nmatch-1].rm_eo >= 0))
864 break;
865 for (i = 0; i < nmatch; i++)
867 printf("(");
868 matchoffprint(match[i].rm_so);
869 printf(",");
870 matchoffprint(match[i].rm_eo);
871 printf(")");
873 if (!(test & (TEST_ACTUAL|TEST_BASELINE)))
875 if (ans)
876 printf(" expected: %s", ans);
877 printf("\n");
881 static int
882 matchcheck(regmatch_t* match, int nmatch, int nsub, char* ans, char* re, char* s, int len, int flags, unsigned long test)
884 char* p;
885 int i;
886 int m;
887 int n;
889 if (streq(ans, "OK"))
890 return test & (TEST_BASELINE|TEST_PASS|TEST_VERIFY);
891 for (i = 0, p = ans; i < nmatch && *p; i++)
893 if (*p == '{')
895 #ifdef REG_DISCIPLINE
896 char* x;
898 if (!(x = sfstruse(state.disc.sp)))
899 bad("out of space [discipline string]\n", NiL, NiL, 0, 0);
900 if (strcmp(p, x))
902 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
903 return 0;
904 report("callout failed", NiL, re, s, len, NiL, flags, test);
905 quote(p, -1, test);
906 printf(" expected, ");
907 quote(x, -1, test);
908 printf(" returned\n");
910 #endif
911 break;
913 if (*p++ != '(')
914 bad("improper answer\n", re, s, -1, test);
915 if (*p == '?')
917 m = -1;
918 p++;
920 else if (*p == 'R' && !memcmp(p, "RE_DUP_MAX", 10))
922 m = RE_DUP_MAX;
923 p += 10;
924 if (*p == '+' || *p == '-')
925 m += strtol(p, &p, 10);
927 else
928 m = strtol(p, &p, 10);
929 if (*p++ != ',')
930 bad("improper answer\n", re, s, -1, test);
931 if (*p == '?')
933 n = -1;
934 p++;
936 else if (*p == 'R' && !memcmp(p, "RE_DUP_MAX", 10))
938 n = RE_DUP_MAX;
939 p += 10;
940 if (*p == '+' || *p == '-')
941 n += strtol(p, &p, 10);
943 else
944 n = strtol(p, &p, 10);
945 if (*p++ != ')')
946 bad("improper answer\n", re, s, -1, test);
947 if (m!=match[i].rm_so || n!=match[i].rm_eo)
949 if (!(test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY)))
951 report("failed: match was", NiL, re, s, len, NiL, flags, test);
952 matchprint(match, nmatch, nsub, ans, test);
954 return 0;
957 for (; i < nmatch; i++)
959 if (match[i].rm_so!=-1 || match[i].rm_eo!=-1)
961 if (!(test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_VERIFY)))
963 if ((test & TEST_IGNORE_POSITION) && (match[i].rm_so<0 || match[i].rm_eo<0))
965 state.ignored++;
966 return 0;
968 if (!(test & TEST_SUMMARY))
970 report("failed: match was", NiL, re, s, len, NiL, flags, test);
971 matchprint(match, nmatch, nsub, ans, test);
974 return 0;
977 if (!(test & TEST_IGNORE_OVER) && match[nmatch].rm_so != state.NOMATCH.rm_so)
979 if (!(test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY)))
981 report("failed: overran match array", NiL, re, s, len, NiL, flags, test);
982 matchprint(match, nmatch + 1, nsub, NiL, test);
984 return 0;
986 return 1;
989 static void
990 sigunblock(int s)
992 #ifdef SIG_SETMASK
993 int op;
994 sigset_t mask;
996 sigemptyset(&mask);
997 if (s)
999 sigaddset(&mask, s);
1000 op = SIG_UNBLOCK;
1002 else op = SIG_SETMASK;
1003 sigprocmask(op, &mask, NiL);
1004 #else
1005 #ifdef sigmask
1006 sigsetmask(s ? (sigsetmask(0L) & ~sigmask(s)) : 0L);
1007 #endif
1008 #endif
1011 static void
1012 gotcha(int sig)
1014 int ret;
1016 signal(sig, gotcha);
1017 alarm(0);
1018 state.signals++;
1019 switch (sig)
1021 case SIGALRM:
1022 ret = REG_EHUNG;
1023 break;
1024 case SIGBUS:
1025 ret = REG_EBUS;
1026 break;
1027 default:
1028 ret = REG_EFAULT;
1029 break;
1031 sigunblock(sig);
1032 longjmp(state.gotcha, ret);
1035 static char*
1036 get_line(FILE* fp)
1038 static char buf[32 * 1024];
1040 register char* s = buf;
1041 register char* e = &buf[sizeof(buf)];
1042 register char* b;
1044 for (;;)
1046 if (!(b = fgets(s, e - s, fp)))
1047 return 0;
1048 state.lineno++;
1049 s += strlen(s);
1050 if (s == b || *--s != '\n' || s == b || *(s - 1) != '\\')
1052 *s = 0;
1053 break;
1055 s--;
1057 return buf;
1060 static unsigned long
1061 note(unsigned long level, char* msg, unsigned long skip, unsigned long test)
1063 if (!(test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_SUMMARY)) && !skip)
1065 printf("NOTE\t");
1066 if (msg)
1067 printf("%s: ", msg);
1068 printf("skipping lines %d", state.lineno);
1070 return skip | level;
1073 #define TABS(n) &ts[7-((n)&7)]
1075 static char ts[] = "\t\t\t\t\t\t\t";
1077 static unsigned long
1078 extract(int* tabs, char* spec, char* re, char* s, char* ans, char* msg, char* accept, regmatch_t* match, int nmatch, int nsub, unsigned long skip, unsigned long level, unsigned long test)
1080 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_OK|TEST_PASS|TEST_SUMMARY))
1082 state.extracted = 1;
1083 if (test & TEST_OK)
1085 state.passed++;
1086 if ((test & TEST_VERIFY) && !(test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_SUMMARY)))
1088 if (msg && strcmp(msg, "EXPECTED"))
1089 printf("NOTE\t%s\n", msg);
1090 return skip;
1092 test &= ~(TEST_PASS|TEST_QUERY);
1094 if (test & (TEST_QUERY|TEST_VERIFY))
1096 if (test & TEST_BASELINE)
1097 test &= ~(TEST_BASELINE|TEST_PASS);
1098 else
1099 test |= TEST_PASS;
1100 skip |= level;
1102 if (!(test & TEST_OK))
1104 if (test & TEST_UNSPECIFIED)
1105 state.unspecified++;
1106 else
1107 state.errors++;
1109 if (test & (TEST_PASS|TEST_SUMMARY))
1110 return skip;
1111 test &= ~TEST_DELIMIT;
1112 printf("%s%s", spec, TABS(*tabs++));
1113 if ((test & (TEST_BASELINE|TEST_SAME)) == (TEST_BASELINE|TEST_SAME))
1114 printf("SAME");
1115 else
1116 quote(re, -1, test);
1117 printf("%s", TABS(*tabs++));
1118 quote(s, -1, test);
1119 printf("%s", TABS(*tabs++));
1120 if (!(test & (TEST_ACTUAL|TEST_BASELINE)) || !accept && !match)
1121 printf("%s", ans);
1122 else if (accept)
1123 printf("%s", accept);
1124 else
1125 matchprint(match, nmatch, nsub, NiL, test);
1126 if (msg)
1127 printf("%s%s", TABS(*tabs++), msg);
1128 putchar('\n');
1130 else if (test & TEST_QUERY)
1131 skip = note(level, msg, skip, test);
1132 else if (test & TEST_VERIFY)
1133 state.extracted = 1;
1134 return skip;
1137 static int
1138 catchfree(regex_t* preg, int flags, int* tabs, char* spec, char* re, char* s, char* ans, char* msg, char* accept, regmatch_t* match, int nmatch, int nsub, unsigned long skip, unsigned long level, unsigned long test)
1140 int eret;
1142 if (!(test & TEST_CATCH))
1144 regfree(preg);
1145 eret = 0;
1147 else if (!(eret = setjmp(state.gotcha)))
1149 alarm(HUNG);
1150 regfree(preg);
1151 alarm(0);
1153 else if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
1154 extract(tabs, spec, re, s, ans, msg, NiL, NiL, 0, 0, skip, level, test);
1155 else
1157 report("failed", "regfree", re, NiL, -1, msg, flags, test);
1158 error(preg, eret);
1160 return eret;
1163 static char*
1164 expand(char* os, char* ot)
1166 char* s = os;
1167 char* t;
1168 int n = 0;
1169 int r;
1170 long m;
1172 for (;;)
1174 switch (*s++)
1176 case 0:
1177 break;
1178 case '{':
1179 n++;
1180 continue;
1181 case '}':
1182 n--;
1183 continue;
1184 case 'R':
1185 if (n == 1 && !memcmp(s, "E_DUP_MAX", 9))
1187 s--;
1188 for (t = ot; os < s; *t++ = *os++);
1189 r = ((t - ot) >= 5 && t[-1] == '{' && t[-2] == '.' && t[-3] == '.' && t[-4] == '.') ? t[-5] : 0;
1190 os = ot;
1191 m = RE_DUP_MAX;
1192 if (*(s += 10) == '+' || *s == '-')
1193 m += strtol(s, &s, 10);
1194 if (r)
1196 t -= 5;
1197 while (m-- > 0)
1198 *t++ = r;
1199 while (*s && *s++ != '}');
1201 else
1202 t += snprintf(t, 32, "%ld", m);
1203 while (*t = *s++)
1204 t++;
1205 break;
1207 continue;
1208 default:
1209 continue;
1211 break;
1213 return os;
1217 main(int argc, char** argv)
1219 int flags;
1220 int cflags;
1221 int eflags;
1222 int nmatch;
1223 int nexec;
1224 int nstr;
1225 int cret;
1226 int eret;
1227 int nsub;
1228 int i;
1229 int j;
1230 int expected;
1231 int got;
1232 int locale;
1233 int subunitlen;
1234 int testno;
1235 unsigned long level;
1236 unsigned long skip;
1237 char* p;
1238 char* line;
1239 char* spec;
1240 char* re;
1241 char* s;
1242 char* ans;
1243 char* msg;
1244 char* fun;
1245 char* ppat;
1246 char* subunit;
1247 char* version;
1248 char* field[6];
1249 char* delim[6];
1250 FILE* fp;
1251 int tabs[6];
1252 char unit[64];
1253 regmatch_t match[100];
1254 regex_t preg;
1256 static char pat[32 * 1024];
1257 static char patbuf[32 * 1024];
1258 static char strbuf[32 * 1024];
1260 int nonosub = REG_NOSUB == 0;
1261 int nonexec = 0;
1263 unsigned long test = 0;
1265 static char* filter[] = { "-", 0 };
1267 state.NOMATCH.rm_so = state.NOMATCH.rm_eo = -2;
1268 p = unit;
1269 version = (char*)id + 10;
1270 while (p < &unit[sizeof(unit)-1] && (*p = *version++) && !isspace(*p))
1271 p++;
1272 *p = 0;
1273 while ((p = *++argv) && *p == '-')
1274 for (;;)
1276 switch (*++p)
1278 case 0:
1279 break;
1280 case 'c':
1281 test |= TEST_CATCH;
1282 continue;
1283 case 'e':
1284 test |= TEST_IGNORE_ERROR;
1285 continue;
1286 case 'h':
1287 case '?':
1288 help(0);
1289 return 2;
1290 case '-':
1291 help(p[1] == 'h');
1292 return 2;
1293 case 'n':
1294 nonexec = 1;
1295 continue;
1296 case 'o':
1297 test |= TEST_IGNORE_OVER;
1298 continue;
1299 case 'p':
1300 test |= TEST_IGNORE_POSITION;
1301 continue;
1302 case 's':
1303 #ifdef REG_DISCIPLINE
1304 if (!(state.stack = stkalloc(stkstd, 0)))
1305 fprintf(stderr, "%s: out of space [stack]", unit);
1306 state.disc.disc.re_resizef = resizef;
1307 state.disc.disc.re_resizehandle = (void*)stkstd;
1308 #endif
1309 continue;
1310 case 'x':
1311 nonosub = 1;
1312 continue;
1313 case 'v':
1314 test |= TEST_VERBOSE;
1315 continue;
1316 case 'A':
1317 test |= TEST_ACTUAL;
1318 continue;
1319 case 'B':
1320 test |= TEST_BASELINE;
1321 continue;
1322 case 'F':
1323 test |= TEST_FAIL;
1324 continue;
1325 case 'P':
1326 test |= TEST_PASS;
1327 continue;
1328 case 'S':
1329 test |= TEST_SUMMARY;
1330 continue;
1331 default:
1332 fprintf(stderr, "%s: %c: invalid option\n", unit, *p);
1333 return 2;
1335 break;
1337 if (!*argv)
1338 argv = filter;
1339 locale = 0;
1340 while (state.file = *argv++)
1342 if (streq(state.file, "-") || streq(state.file, "/dev/stdin") || streq(state.file, "/dev/fd/0"))
1344 state.file = 0;
1345 fp = stdin;
1347 else if (!(fp = fopen(state.file, "r")))
1349 fprintf(stderr, "%s: %s: cannot read\n", unit, state.file);
1350 return 2;
1352 testno = state.errors = state.ignored = state.lineno = state.passed =
1353 state.signals = state.unspecified = state.warnings = 0;
1354 skip = 0;
1355 level = 1;
1356 if (!(test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_SUMMARY)))
1358 printf("TEST\t%s ", unit);
1359 if (s = state.file)
1361 subunit = p = 0;
1362 for (;;)
1364 switch (*s++)
1366 case 0:
1367 break;
1368 case '/':
1369 subunit = s;
1370 continue;
1371 case '.':
1372 p = s - 1;
1373 continue;
1374 default:
1375 continue;
1377 break;
1379 if (!subunit)
1380 subunit = state.file;
1381 if (p < subunit)
1382 p = s - 1;
1383 subunitlen = p - subunit;
1384 printf("%-.*s ", subunitlen, subunit);
1386 else
1387 subunit = 0;
1388 for (s = version; *s && (*s != ' ' || *(s + 1) != '$'); s++)
1389 putchar(*s);
1390 if (test & TEST_CATCH)
1391 printf(", catch");
1392 if (test & TEST_IGNORE_ERROR)
1393 printf(", ignore error code mismatches");
1394 if (test & TEST_IGNORE_POSITION)
1395 printf(", ignore negative position mismatches");
1396 #ifdef REG_DISCIPLINE
1397 if (state.stack)
1398 printf(", stack");
1399 #endif
1400 if (test & TEST_VERBOSE)
1401 printf(", verbose");
1402 printf("\n");
1403 #ifdef REG_VERSIONID
1404 if (regerror(REG_VERSIONID, NiL, pat, sizeof(pat)) > 0)
1405 s = pat;
1406 else
1407 #endif
1408 #ifdef REG_TEST_VERSION
1409 s = REG_TEST_VERSION;
1410 #else
1411 s = "regex";
1412 #endif
1413 printf("NOTE\t%s\n", s);
1414 if (elementsof(unsupported) > 1)
1416 #if (REG_TEST_DEFAULT & (REG_AUGMENTED|REG_EXTENDED|REG_SHELL)) || !defined(REG_EXTENDED)
1417 i = 0;
1418 #else
1419 i = REG_EXTENDED != 0;
1420 #endif
1421 for (got = 0; i < elementsof(unsupported) - 1; i++)
1423 if (!got)
1425 got = 1;
1426 printf("NOTE\tunsupported: %s", unsupported[i]);
1428 else
1429 printf(",%s", unsupported[i]);
1431 if (got)
1432 printf("\n");
1435 #ifdef REG_DISCIPLINE
1436 state.disc.disc.re_version = REG_VERSION;
1437 state.disc.disc.re_compf = compf;
1438 state.disc.disc.re_execf = execf;
1439 if (!(state.disc.sp = sfstropen()))
1440 bad("out of space [discipline string stream]\n", NiL, NiL, 0, 0);
1441 preg.re_disc = &state.disc.disc;
1442 #endif
1443 if (test & TEST_CATCH)
1445 signal(SIGALRM, gotcha);
1446 signal(SIGBUS, gotcha);
1447 signal(SIGSEGV, gotcha);
1449 while (p = get_line(fp))
1452 /* parse: */
1454 line = p;
1455 if (*p == ':' && !isspace(*(p + 1)))
1457 while (*++p && *p != ':');
1458 if (!*p++)
1460 if (test & TEST_BASELINE)
1461 printf("%s\n", line);
1462 continue;
1465 while (isspace(*p))
1466 p++;
1467 if (*p == 0 || *p == '#' || *p == 'T')
1469 if (test & TEST_BASELINE)
1470 printf("%s\n", line);
1471 continue;
1473 if (*p == ':' || *p == 'N')
1475 if (test & TEST_BASELINE)
1476 printf("%s\n", line);
1477 else if (!(test & (TEST_ACTUAL|TEST_FAIL|TEST_PASS|TEST_SUMMARY)))
1479 while (*++p && !isspace(*p));
1480 while (isspace(*p))
1481 p++;
1482 printf("NOTE %s\n", p);
1484 continue;
1486 j = 0;
1487 i = 0;
1488 field[i++] = p;
1489 for (;;)
1491 switch (*p++)
1493 case 0:
1494 p--;
1495 j = 0;
1496 goto checkfield;
1497 case '\t':
1498 *(delim[i] = p - 1) = 0;
1499 j = 1;
1500 checkfield:
1501 s = field[i - 1];
1502 if (streq(s, "NIL"))
1503 field[i - 1] = 0;
1504 else if (streq(s, "NULL"))
1505 *s = 0;
1506 while (*p == '\t')
1508 p++;
1509 j++;
1511 tabs[i - 1] = j;
1512 if (!*p)
1513 break;
1514 if (i >= elementsof(field))
1515 bad("too many fields\n", NiL, NiL, 0, 0);
1516 field[i++] = p;
1517 /*FALLTHROUGH*/
1518 default:
1519 continue;
1521 break;
1523 if (!(spec = field[0]))
1524 bad("NIL spec\n", NiL, NiL, 0, 0);
1526 /* interpret: */
1528 cflags = REG_TEST_DEFAULT;
1529 eflags = REG_EXEC_DEFAULT;
1530 test &= TEST_GLOBAL;
1531 state.extracted = 0;
1532 nmatch = 20;
1533 nsub = -1;
1534 for (p = spec; *p; p++)
1536 if (isdigit(*p))
1538 nmatch = strtol(p, &p, 10);
1539 if (nmatch >= elementsof(match))
1540 bad("nmatch must be < 100\n", NiL, NiL, 0, 0);
1541 p--;
1542 continue;
1544 switch (*p)
1546 case 'A':
1547 test |= TEST_ARE;
1548 continue;
1549 case 'B':
1550 test |= TEST_BRE;
1551 continue;
1552 case 'C':
1553 if (!(test & TEST_QUERY) && !(skip & level))
1554 bad("locale must be nested\n", NiL, NiL, 0, 0);
1555 test &= ~TEST_QUERY;
1556 if (locale)
1557 bad("locale nesting not supported\n", NiL, NiL, 0, 0);
1558 if (i != 2)
1559 bad("locale field expected\n", NiL, NiL, 0, 0);
1560 if (!(skip & level))
1562 #if defined(LC_COLLATE) && defined(LC_CTYPE)
1563 s = field[1];
1564 if (!s || streq(s, "POSIX"))
1565 s = "C";
1566 if ((ans = setlocale(LC_COLLATE, s)) && streq(ans, "POSIX"))
1567 ans = "C";
1568 if (!ans || !streq(ans, s) && streq(s, "C"))
1569 ans = 0;
1570 else if ((ans = setlocale(LC_CTYPE, s)) && streq(ans, "POSIX"))
1571 ans = "C";
1572 if (!ans || !streq(ans, s) && streq(s, "C"))
1573 skip = note(level, s, skip, test);
1574 else
1576 if (!(test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_SUMMARY)))
1577 printf("NOTE \"%s\" locale\n", s);
1578 locale = level;
1580 #else
1581 skip = note(level, skip, test, "locales not supported");
1582 #endif
1584 cflags = NOTEST;
1585 continue;
1586 case 'E':
1587 test |= TEST_ERE;
1588 continue;
1589 case 'K':
1590 test |= TEST_KRE;
1591 continue;
1592 case 'L':
1593 test |= TEST_LRE;
1594 continue;
1595 case 'S':
1596 test |= TEST_SRE;
1597 continue;
1599 case 'a':
1600 cflags |= REG_LEFT|REG_RIGHT;
1601 continue;
1602 case 'b':
1603 eflags |= REG_NOTBOL;
1604 continue;
1605 case 'c':
1606 cflags |= REG_COMMENT;
1607 continue;
1608 case 'd':
1609 cflags |= REG_SHELL_DOT;
1610 continue;
1611 case 'e':
1612 eflags |= REG_NOTEOL;
1613 continue;
1614 case 'f':
1615 cflags |= REG_MULTIPLE;
1616 continue;
1617 case 'g':
1618 cflags |= NOTEST;
1619 continue;
1620 case 'h':
1621 cflags |= REG_MULTIREF;
1622 continue;
1623 case 'i':
1624 cflags |= REG_ICASE;
1625 continue;
1626 case 'j':
1627 cflags |= REG_SPAN;
1628 continue;
1629 case 'k':
1630 cflags |= REG_ESCAPE;
1631 continue;
1632 case 'l':
1633 cflags |= REG_LEFT;
1634 continue;
1635 case 'm':
1636 cflags |= REG_MINIMAL;
1637 continue;
1638 case 'n':
1639 cflags |= REG_NEWLINE;
1640 continue;
1641 case 'o':
1642 cflags |= REG_SHELL_GROUP;
1643 continue;
1644 case 'p':
1645 cflags |= REG_SHELL_PATH;
1646 continue;
1647 case 'q':
1648 cflags |= REG_DELIMITED;
1649 continue;
1650 case 'r':
1651 cflags |= REG_RIGHT;
1652 continue;
1653 case 's':
1654 cflags |= REG_SHELL_ESCAPED;
1655 continue;
1656 case 't':
1657 cflags |= REG_MUSTDELIM;
1658 continue;
1659 case 'u':
1660 test |= TEST_UNSPECIFIED;
1661 continue;
1662 case 'v':
1663 cflags |= REG_CLASS_ESCAPE;
1664 continue;
1665 case 'w':
1666 cflags |= REG_NOSUB;
1667 continue;
1668 case 'x':
1669 if (REG_LENIENT)
1670 cflags |= REG_LENIENT;
1671 else
1672 test |= TEST_LENIENT;
1673 continue;
1674 case 'y':
1675 eflags |= REG_LEFT;
1676 continue;
1677 case 'z':
1678 cflags |= REG_NULL;
1679 continue;
1681 case '$':
1682 test |= TEST_EXPAND;
1683 continue;
1685 case '/':
1686 test |= TEST_SUB;
1687 continue;
1689 case '=':
1690 test |= TEST_DECOMP;
1691 continue;
1693 case '?':
1694 test |= TEST_VERIFY;
1695 test &= ~(TEST_AND|TEST_OR);
1696 state.verify = state.passed;
1697 continue;
1698 case '&':
1699 test |= TEST_VERIFY|TEST_AND;
1700 test &= ~TEST_OR;
1701 continue;
1702 case '|':
1703 test |= TEST_VERIFY|TEST_OR;
1704 test &= ~TEST_AND;
1705 continue;
1706 case ';':
1707 test |= TEST_OR;
1708 test &= ~TEST_AND;
1709 continue;
1711 case '{':
1712 level <<= 1;
1713 if (skip & (level >> 1))
1715 skip |= level;
1716 cflags = NOTEST;
1718 else
1720 skip &= ~level;
1721 test |= TEST_QUERY;
1723 continue;
1724 case '}':
1725 if (level == 1)
1726 bad("invalid {...} nesting\n", NiL, NiL, 0, 0);
1727 if ((skip & level) && !(skip & (level>>1)))
1729 if (!(test & (TEST_BASELINE|TEST_SUMMARY)))
1731 if (test & (TEST_ACTUAL|TEST_FAIL))
1732 printf("}\n");
1733 else if (!(test & TEST_PASS))
1734 printf("-%d\n", state.lineno);
1737 #if defined(LC_COLLATE) && defined(LC_CTYPE)
1738 else if (locale & level)
1740 locale = 0;
1741 if (!(skip & level))
1743 s = "C";
1744 setlocale(LC_COLLATE, s);
1745 setlocale(LC_CTYPE, s);
1746 if (!(test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_SUMMARY)))
1747 printf("NOTE \"%s\" locale\n", s);
1748 else if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_PASS))
1749 printf("}\n");
1751 else if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL))
1752 printf("}\n");
1754 #endif
1755 level >>= 1;
1756 cflags = NOTEST;
1757 continue;
1759 default:
1760 bad("bad spec\n", spec, NiL, 0, test);
1761 break;
1764 break;
1766 if ((cflags|eflags) == NOTEST || (skip & level) && (test & TEST_BASELINE))
1768 if (test & TEST_BASELINE)
1770 while (i > 1)
1771 *delim[--i] = '\t';
1772 printf("%s\n", line);
1774 continue;
1776 if (test & TEST_OR)
1778 if (!(test & TEST_VERIFY))
1780 test &= ~TEST_OR;
1781 if (state.passed == state.verify && i > 1)
1782 printf("NOTE\t%s\n", field[1]);
1783 continue;
1785 else if (state.passed > state.verify)
1786 continue;
1788 else if (test & TEST_AND)
1790 if (state.passed == state.verify)
1791 continue;
1792 state.passed = state.verify;
1794 if (i < ((test & TEST_DECOMP) ? 3 : 4))
1795 bad("too few fields\n", NiL, NiL, 0, test);
1796 while (i < elementsof(field))
1797 field[i++] = 0;
1798 if (re = field[1])
1800 if (streq(re, "SAME"))
1802 re = ppat;
1803 test |= TEST_SAME;
1805 else
1807 if (test & TEST_EXPAND)
1808 escape(re);
1809 re = expand(re, patbuf);
1810 strcpy(ppat = pat, re);
1813 else
1814 ppat = 0;
1815 nstr = -1;
1816 if (s = field[2])
1818 s = expand(s, strbuf);
1819 if (test & TEST_EXPAND)
1821 nstr = escape(s);
1822 #if _REG_nexec
1823 if (nstr != strlen(s))
1824 nexec = nstr;
1825 #endif
1828 if (!(ans = field[(test & TEST_DECOMP) ? 2 : 3]))
1829 bad("NIL answer\n", NiL, NiL, 0, test);
1830 msg = field[4];
1831 fflush(stdout);
1832 if (test & TEST_SUB)
1833 #if _REG_subcomp
1834 cflags |= REG_DELIMITED;
1835 #else
1836 continue;
1837 #endif
1838 #if !_REG_decomp
1839 if (test & TEST_DECOMP)
1840 continue;
1841 #endif
1843 compile:
1845 if (state.extracted || (skip & level))
1846 continue;
1847 #if !(REG_TEST_DEFAULT & (REG_AUGMENTED|REG_EXTENDED|REG_SHELL))
1848 #ifdef REG_EXTENDED
1849 if (REG_EXTENDED != 0 && (test & TEST_BRE))
1850 #else
1851 if (test & TEST_BRE)
1852 #endif
1854 test &= ~TEST_BRE;
1855 flags = cflags;
1856 state.which = "BRE";
1858 else
1859 #endif
1860 #ifdef REG_EXTENDED
1861 if (test & TEST_ERE)
1863 test &= ~TEST_ERE;
1864 flags = cflags | REG_EXTENDED;
1865 state.which = "ERE";
1867 else
1868 #endif
1869 #ifdef REG_AUGMENTED
1870 if (test & TEST_ARE)
1872 test &= ~TEST_ARE;
1873 flags = cflags | REG_AUGMENTED;
1874 state.which = "ARE";
1876 else
1877 #endif
1878 #ifdef REG_LITERAL
1879 if (test & TEST_LRE)
1881 test &= ~TEST_LRE;
1882 flags = cflags | REG_LITERAL;
1883 state.which = "LRE";
1885 else
1886 #endif
1887 #ifdef REG_SHELL
1888 if (test & TEST_SRE)
1890 test &= ~TEST_SRE;
1891 flags = cflags | REG_SHELL;
1892 state.which = "SRE";
1894 else
1895 #ifdef REG_AUGMENTED
1896 if (test & TEST_KRE)
1898 test &= ~TEST_KRE;
1899 flags = cflags | REG_SHELL | REG_AUGMENTED;
1900 state.which = "KRE";
1902 else
1903 #endif
1904 #endif
1906 if (test & (TEST_BASELINE|TEST_PASS|TEST_VERIFY))
1907 extract(tabs, line, re, s, ans, msg, NiL, NiL, 0, 0, skip, level, test|TEST_OK);
1908 continue;
1910 if ((test & (TEST_QUERY|TEST_VERBOSE|TEST_VERIFY)) == TEST_VERBOSE)
1912 printf("test %-3d %s ", state.lineno, state.which);
1913 quote(re, -1, test|TEST_DELIMIT);
1914 printf(" ");
1915 quote(s, nstr, test|TEST_DELIMIT);
1916 printf("\n");
1919 nosub:
1920 fun = "regcomp";
1921 #if _REG_nexec
1922 if (nstr >= 0 && nstr != strlen(s))
1923 nexec = nstr;
1925 else
1926 #endif
1927 nexec = -1;
1928 if (state.extracted || (skip & level))
1929 continue;
1930 if (!(test & TEST_QUERY))
1931 testno++;
1932 #ifdef REG_DISCIPLINE
1933 if (state.stack)
1934 stkset(stkstd, state.stack, 0);
1935 flags |= REG_DISCIPLINE;
1936 state.disc.ordinal = 0;
1937 sfstrseek(state.disc.sp, 0, SEEK_SET);
1938 #endif
1939 if (!(test & TEST_CATCH))
1940 cret = regcomp(&preg, re, flags);
1941 else if (!(cret = setjmp(state.gotcha)))
1943 alarm(HUNG);
1944 cret = regcomp(&preg, re, flags);
1945 alarm(0);
1947 #if _REG_subcomp
1948 if (!cret && (test & TEST_SUB))
1950 fun = "regsubcomp";
1951 p = re + preg.re_npat;
1952 if (!(test & TEST_CATCH))
1953 cret = regsubcomp(&preg, p, NiL, 0, 0);
1954 else if (!(cret = setjmp(state.gotcha)))
1956 alarm(HUNG);
1957 cret = regsubcomp(&preg, p, NiL, 0, 0);
1958 alarm(0);
1960 if (!cret && *(p += preg.re_npat) && !(preg.re_sub->re_flags & REG_SUB_LAST))
1962 if (catchfree(&preg, flags, tabs, line, re, s, ans, msg, NiL, NiL, 0, 0, skip, level, test))
1963 continue;
1964 cret = REG_EFLAGS;
1967 #endif
1968 #if _REG_decomp
1969 if (!cret && (test & TEST_DECOMP))
1971 char buf[128];
1973 if ((j = nmatch) > sizeof(buf))
1974 j = sizeof(buf);
1975 fun = "regdecomp";
1976 p = re + preg.re_npat;
1977 if (!(test & TEST_CATCH))
1978 i = regdecomp(&preg, -1, buf, j);
1979 else if (!(cret = setjmp(state.gotcha)))
1981 alarm(HUNG);
1982 i = regdecomp(&preg, -1, buf, j);
1983 alarm(0);
1985 if (!cret)
1987 catchfree(&preg, flags, tabs, line, re, s, ans, msg, NiL, NiL, 0, 0, skip, level, test);
1988 if (i > j)
1990 if (i != (strlen(ans) + 1))
1992 report("failed", fun, re, s, nstr, msg, flags, test);
1993 printf(" %d byte buffer supplied, %d byte buffer required\n", j, i);
1996 else if (strcmp(buf, ans))
1998 report("failed", fun, re, s, nstr, msg, flags, test);
1999 quote(ans, -1, test|TEST_DELIMIT);
2000 printf(" expected, ");
2001 quote(buf, -1, test|TEST_DELIMIT);
2002 printf(" returned\n");
2004 continue;
2007 #endif
2008 if (!cret)
2010 if (!(flags & REG_NOSUB) && nsub < 0 && *ans == '(')
2012 for (p = ans; *p; p++)
2013 if (*p == '(')
2014 nsub++;
2015 else if (*p == '{')
2016 nsub--;
2017 if (nsub >= 0)
2019 if (test & TEST_IGNORE_OVER)
2021 if (nmatch > nsub)
2022 nmatch = nsub + 1;
2024 else if (nsub != preg.re_nsub)
2026 if (nsub > preg.re_nsub)
2028 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2029 skip = extract(tabs, line, re, s, ans, msg, "OK", NiL, 0, 0, skip, level, test|TEST_DELIMIT);
2030 else
2032 report("re_nsub incorrect", fun, re, NiL, -1, msg, flags, test);
2033 printf("at least %d expected, %d returned\n", nsub, preg.re_nsub);
2034 state.errors++;
2037 else
2038 nsub = preg.re_nsub;
2042 if (!(test & (TEST_DECOMP|TEST_SUB)) && *ans && *ans != '(' && !streq(ans, "OK") && !streq(ans, "NOMATCH"))
2044 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2045 skip = extract(tabs, line, re, s, ans, msg, "OK", NiL, 0, 0, skip, level, test|TEST_DELIMIT);
2046 else if (!(test & TEST_LENIENT))
2048 report("failed", fun, re, NiL, -1, msg, flags, test);
2049 printf("%s expected, OK returned\n", ans);
2051 catchfree(&preg, flags, tabs, line, re, s, ans, msg, NiL, NiL, 0, 0, skip, level, test);
2052 continue;
2055 else
2057 if (test & TEST_LENIENT)
2058 /* we'll let it go this time */;
2059 else if (!*ans || ans[0]=='(' || cret == REG_BADPAT && streq(ans, "NOMATCH"))
2061 got = 0;
2062 for (i = 1; i < elementsof(codes); i++)
2063 if (cret==codes[i].code)
2064 got = i;
2065 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2066 skip = extract(tabs, line, re, s, ans, msg, codes[got].name, NiL, 0, 0, skip, level, test|TEST_DELIMIT);
2067 else
2069 report("failed", fun, re, NiL, -1, msg, flags, test);
2070 printf("%s returned: ", codes[got].name);
2071 error(&preg, cret);
2074 else
2076 expected = got = 0;
2077 for (i = 1; i < elementsof(codes); i++)
2079 if (streq(ans, codes[i].name))
2080 expected = i;
2081 if (cret==codes[i].code)
2082 got = i;
2084 if (!expected)
2086 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2087 skip = extract(tabs, line, re, s, ans, msg, codes[got].name, NiL, 0, 0, skip, level, test|TEST_DELIMIT);
2088 else
2090 report("failed: invalid error code", NiL, re, NiL, -1, msg, flags, test);
2091 printf("%s expected, %s returned\n", ans, codes[got].name);
2094 else if (cret != codes[expected].code && cret != REG_BADPAT)
2096 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2097 skip = extract(tabs, line, re, s, ans, msg, codes[got].name, NiL, 0, 0, skip, level, test|TEST_DELIMIT);
2098 else if (test & TEST_IGNORE_ERROR)
2099 state.ignored++;
2100 else
2102 report("should fail and did", fun, re, NiL, -1, msg, flags, test);
2103 printf("%s expected, %s returned: ", ans, codes[got].name);
2104 state.errors--;
2105 state.warnings++;
2106 error(&preg, cret);
2110 goto compile;
2113 #if _REG_nexec
2114 execute:
2115 if (nexec >= 0)
2116 fun = "regnexec";
2117 else
2118 #endif
2119 fun = "regexec";
2121 for (i = 0; i < elementsof(match); i++)
2122 match[i] = state.NOMATCH;
2124 #if _REG_nexec
2125 if (nexec >= 0)
2127 eret = regnexec(&preg, s, nexec, nmatch, match, eflags);
2128 s[nexec] = 0;
2130 else
2131 #endif
2133 if (!(test & TEST_CATCH))
2134 eret = regexec(&preg, s, nmatch, match, eflags);
2135 else if (!(eret = setjmp(state.gotcha)))
2137 alarm(HUNG);
2138 eret = regexec(&preg, s, nmatch, match, eflags);
2139 alarm(0);
2142 #if _REG_subcomp
2143 if ((test & TEST_SUB) && !eret)
2145 fun = "regsubexec";
2146 if (!(test & TEST_CATCH))
2147 eret = regsubexec(&preg, s, nmatch, match);
2148 else if (!(eret = setjmp(state.gotcha)))
2150 alarm(HUNG);
2151 eret = regsubexec(&preg, s, nmatch, match);
2152 alarm(0);
2155 #endif
2156 if (flags & REG_NOSUB)
2158 if (eret)
2160 if (eret != REG_NOMATCH || !streq(ans, "NOMATCH"))
2162 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2163 skip = extract(tabs, line, re, s, ans, msg, "NOMATCH", NiL, 0, 0, skip, level, test|TEST_DELIMIT);
2164 else
2166 report("REG_NOSUB failed", fun, re, s, nstr, msg, flags, test);
2167 error(&preg, eret);
2171 else if (streq(ans, "NOMATCH"))
2173 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2174 skip = extract(tabs, line, re, s, ans, msg, NiL, match, nmatch, nsub, skip, level, test|TEST_DELIMIT);
2175 else
2177 report("should fail and didn't", fun, re, s, nstr, msg, flags, test);
2178 error(&preg, eret);
2182 else if (eret)
2184 if (eret != REG_NOMATCH || !streq(ans, "NOMATCH"))
2186 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2187 skip = extract(tabs, line, re, s, ans, msg, "NOMATCH", NiL, 0, nsub, skip, level, test|TEST_DELIMIT);
2188 else
2190 report("failed", fun, re, s, nstr, msg, flags, test);
2191 if (eret != REG_NOMATCH)
2192 error(&preg, eret);
2193 else if (*ans)
2194 printf("expected: %s\n", ans);
2195 else
2196 printf("\n");
2200 else if (streq(ans, "NOMATCH"))
2202 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2203 skip = extract(tabs, line, re, s, ans, msg, NiL, match, nmatch, nsub, skip, level, test|TEST_DELIMIT);
2204 else
2206 report("should fail and didn't", fun, re, s, nstr, msg, flags, test);
2207 matchprint(match, nmatch, nsub, NiL, test);
2210 #if _REG_subcomp
2211 else if (test & TEST_SUB)
2213 p = preg.re_sub->re_buf;
2214 if (strcmp(p, ans))
2216 report("failed", fun, re, s, nstr, msg, flags, test);
2217 quote(ans, -1, test|TEST_DELIMIT);
2218 printf(" expected, ");
2219 quote(p, -1, test|TEST_DELIMIT);
2220 printf(" returned\n");
2223 #endif
2224 else if (!*ans)
2226 if (match[0].rm_so != state.NOMATCH.rm_so)
2228 if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2229 skip = extract(tabs, line, re, s, ans, msg, NiL, NiL, 0, 0, skip, level, test);
2230 else
2232 report("failed: no match but match array assigned", NiL, re, s, nstr, msg, flags, test);
2233 matchprint(match, nmatch, nsub, NiL, test);
2237 else if (matchcheck(match, nmatch, nsub, ans, re, s, nstr, flags, test))
2239 #if _REG_nexec
2240 if (nexec < 0 && !nonexec)
2242 nexec = nstr >= 0 ? nstr : strlen(s);
2243 s[nexec] = '\n';
2244 testno++;
2245 goto execute;
2247 #endif
2248 if (!(test & (TEST_DECOMP|TEST_SUB|TEST_VERIFY)) && !nonosub)
2250 if (catchfree(&preg, flags, tabs, line, re, s, ans, msg, NiL, NiL, 0, 0, skip, level, test))
2251 continue;
2252 flags |= REG_NOSUB;
2253 goto nosub;
2255 if (test & (TEST_BASELINE|TEST_PASS|TEST_VERIFY))
2256 skip = extract(tabs, line, re, s, ans, msg, NiL, match, nmatch, nsub, skip, level, test|TEST_OK);
2258 else if (test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS|TEST_QUERY|TEST_SUMMARY|TEST_VERIFY))
2259 skip = extract(tabs, line, re, s, ans, msg, NiL, match, nmatch, nsub, skip, level, test|TEST_DELIMIT);
2260 if (catchfree(&preg, flags, tabs, line, re, s, ans, msg, NiL, NiL, 0, 0, skip, level, test))
2261 continue;
2262 goto compile;
2264 if (test & TEST_SUMMARY)
2265 printf("tests=%-4d errors=%-4d warnings=%-2d ignored=%-2d unspecified=%-2d signals=%d\n", testno, state.errors, state.warnings, state.ignored, state.unspecified, state.signals);
2266 else if (!(test & (TEST_ACTUAL|TEST_BASELINE|TEST_FAIL|TEST_PASS)))
2268 printf("TEST\t%s", unit);
2269 if (subunit)
2270 printf(" %-.*s", subunitlen, subunit);
2271 printf(", %d test%s", testno, testno == 1 ? "" : "s");
2272 if (state.ignored)
2273 printf(", %d ignored mismatche%s", state.ignored, state.ignored == 1 ? "" : "s");
2274 if (state.warnings)
2275 printf(", %d warning%s", state.warnings, state.warnings == 1 ? "" : "s");
2276 if (state.unspecified)
2277 printf(", %d unspecified difference%s", state.unspecified, state.unspecified == 1 ? "" : "s");
2278 if (state.signals)
2279 printf(", %d signal%s", state.signals, state.signals == 1 ? "" : "s");
2280 printf(", %d error%s\n", state.errors, state.errors == 1 ? "" : "s");
2282 if (fp != stdin)
2283 fclose(fp);
2285 return 0;