/[pcre]/code/trunk/pcretest.c
ViewVC logotype

Contents of /code/trunk/pcretest.c

Parent Directory Parent Directory | Revision Log Revision Log


Revision 426 - (show annotations) (download)
Wed Aug 26 15:38:32 2009 UTC (5 years, 3 months ago) by ph10
File MIME type: text/plain
File size: 73569 byte(s)
Remove restrictions on pcre_exec() partial matching.

1 /*************************************************
2 * PCRE testing program *
3 *************************************************/
4
5 /* This program was hacked up as a tester for PCRE. I really should have
6 written it more tidily in the first place. Will I ever learn? It has grown and
7 been extended and consequently is now rather, er, *very* untidy in places.
8
9 -----------------------------------------------------------------------------
10 Redistribution and use in source and binary forms, with or without
11 modification, are permitted provided that the following conditions are met:
12
13 * Redistributions of source code must retain the above copyright notice,
14 this list of conditions and the following disclaimer.
15
16 * 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
20 * Neither the name of the University of Cambridge nor the names of its
21 contributors may be used to endorse or promote products derived from
22 this software without specific prior written permission.
23
24 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
25 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
26 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27 ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
28 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
29 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
30 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
31 INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
32 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
33 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
34 POSSIBILITY OF SUCH DAMAGE.
35 -----------------------------------------------------------------------------
36 */
37
38
39 #ifdef HAVE_CONFIG_H
40 #include "config.h"
41 #endif
42
43 #include <ctype.h>
44 #include <stdio.h>
45 #include <string.h>
46 #include <stdlib.h>
47 #include <time.h>
48 #include <locale.h>
49 #include <errno.h>
50
51 #ifdef SUPPORT_LIBREADLINE
52 #ifdef HAVE_UNISTD_H
53 #include <unistd.h>
54 #endif
55 #include <readline/readline.h>
56 #include <readline/history.h>
57 #endif
58
59
60 /* A number of things vary for Windows builds. Originally, pcretest opened its
61 input and output without "b"; then I was told that "b" was needed in some
62 environments, so it was added for release 5.0 to both the input and output. (It
63 makes no difference on Unix-like systems.) Later I was told that it is wrong
64 for the input on Windows. I've now abstracted the modes into two macros that
65 are set here, to make it easier to fiddle with them, and removed "b" from the
66 input mode under Windows. */
67
68 #if defined(_WIN32) || defined(WIN32)
69 #include <io.h> /* For _setmode() */
70 #include <fcntl.h> /* For _O_BINARY */
71 #define INPUT_MODE "r"
72 #define OUTPUT_MODE "wb"
73
74 #ifndef isatty
75 #define isatty _isatty /* This is what Windows calls them, I'm told, */
76 #endif /* though in some environments they seem to */
77 /* be already defined, hence the #ifndefs. */
78 #ifndef fileno
79 #define fileno _fileno
80 #endif
81
82 #else
83 #include <sys/time.h> /* These two includes are needed */
84 #include <sys/resource.h> /* for setrlimit(). */
85 #define INPUT_MODE "rb"
86 #define OUTPUT_MODE "wb"
87 #endif
88
89
90 /* We have to include pcre_internal.h because we need the internal info for
91 displaying the results of pcre_study() and we also need to know about the
92 internal macros, structures, and other internal data values; pcretest has
93 "inside information" compared to a program that strictly follows the PCRE API.
94
95 Although pcre_internal.h does itself include pcre.h, we explicitly include it
96 here before pcre_internal.h so that the PCRE_EXP_xxx macros get set
97 appropriately for an application, not for building PCRE. */
98
99 #include "pcre.h"
100 #include "pcre_internal.h"
101
102 /* We need access to some of the data tables that PCRE uses. So as not to have
103 to keep two copies, we include the source file here, changing the names of the
104 external symbols to prevent clashes. */
105
106 #define _pcre_ucp_gentype ucp_gentype
107 #define _pcre_utf8_table1 utf8_table1
108 #define _pcre_utf8_table1_size utf8_table1_size
109 #define _pcre_utf8_table2 utf8_table2
110 #define _pcre_utf8_table3 utf8_table3
111 #define _pcre_utf8_table4 utf8_table4
112 #define _pcre_utt utt
113 #define _pcre_utt_size utt_size
114 #define _pcre_utt_names utt_names
115 #define _pcre_OP_lengths OP_lengths
116
117 #include "pcre_tables.c"
118
119 /* We also need the pcre_printint() function for printing out compiled
120 patterns. This function is in a separate file so that it can be included in
121 pcre_compile.c when that module is compiled with debugging enabled.
122
123 The definition of the macro PRINTABLE, which determines whether to print an
124 output character as-is or as a hex value when showing compiled patterns, is
125 contained in this file. We uses it here also, in cases when the locale has not
126 been explicitly changed, so as to get consistent output from systems that
127 differ in their output from isprint() even in the "C" locale. */
128
129 #include "pcre_printint.src"
130
131 #define PRINTHEX(c) (locale_set? isprint(c) : PRINTABLE(c))
132
133
134 /* It is possible to compile this test program without including support for
135 testing the POSIX interface, though this is not available via the standard
136 Makefile. */
137
138 #if !defined NOPOSIX
139 #include "pcreposix.h"
140 #endif
141
142 /* It is also possible, for the benefit of the version currently imported into
143 Exim, to build pcretest without support for UTF8 (define NOUTF8), without the
144 interface to the DFA matcher (NODFA), and without the doublecheck of the old
145 "info" function (define NOINFOCHECK). In fact, we automatically cut out the
146 UTF8 support if PCRE is built without it. */
147
148 #ifndef SUPPORT_UTF8
149 #ifndef NOUTF8
150 #define NOUTF8
151 #endif
152 #endif
153
154
155 /* Other parameters */
156
157 #ifndef CLOCKS_PER_SEC
158 #ifdef CLK_TCK
159 #define CLOCKS_PER_SEC CLK_TCK
160 #else
161 #define CLOCKS_PER_SEC 100
162 #endif
163 #endif
164
165 /* This is the default loop count for timing. */
166
167 #define LOOPREPEAT 500000
168
169 /* Static variables */
170
171 static FILE *outfile;
172 static int log_store = 0;
173 static int callout_count;
174 static int callout_extra;
175 static int callout_fail_count;
176 static int callout_fail_id;
177 static int debug_lengths;
178 static int first_callout;
179 static int locale_set = 0;
180 static int show_malloc;
181 static int use_utf8;
182 static size_t gotten_store;
183
184 /* The buffers grow automatically if very long input lines are encountered. */
185
186 static int buffer_size = 50000;
187 static uschar *buffer = NULL;
188 static uschar *dbuffer = NULL;
189 static uschar *pbuffer = NULL;
190
191
192
193 /*************************************************
194 * Read or extend an input line *
195 *************************************************/
196
197 /* Input lines are read into buffer, but both patterns and data lines can be
198 continued over multiple input lines. In addition, if the buffer fills up, we
199 want to automatically expand it so as to be able to handle extremely large
200 lines that are needed for certain stress tests. When the input buffer is
201 expanded, the other two buffers must also be expanded likewise, and the
202 contents of pbuffer, which are a copy of the input for callouts, must be
203 preserved (for when expansion happens for a data line). This is not the most
204 optimal way of handling this, but hey, this is just a test program!
205
206 Arguments:
207 f the file to read
208 start where in buffer to start (this *must* be within buffer)
209 prompt for stdin or readline()
210
211 Returns: pointer to the start of new data
212 could be a copy of start, or could be moved
213 NULL if no data read and EOF reached
214 */
215
216 static uschar *
217 extend_inputline(FILE *f, uschar *start, const char *prompt)
218 {
219 uschar *here = start;
220
221 for (;;)
222 {
223 int rlen = buffer_size - (here - buffer);
224
225 if (rlen > 1000)
226 {
227 int dlen;
228
229 /* If libreadline support is required, use readline() to read a line if the
230 input is a terminal. Note that readline() removes the trailing newline, so
231 we must put it back again, to be compatible with fgets(). */
232
233 #ifdef SUPPORT_LIBREADLINE
234 if (isatty(fileno(f)))
235 {
236 size_t len;
237 char *s = readline(prompt);
238 if (s == NULL) return (here == start)? NULL : start;
239 len = strlen(s);
240 if (len > 0) add_history(s);
241 if (len > rlen - 1) len = rlen - 1;
242 memcpy(here, s, len);
243 here[len] = '\n';
244 here[len+1] = 0;
245 free(s);
246 }
247 else
248 #endif
249
250 /* Read the next line by normal means, prompting if the file is stdin. */
251
252 {
253 if (f == stdin) printf(prompt);
254 if (fgets((char *)here, rlen, f) == NULL)
255 return (here == start)? NULL : start;
256 }
257
258 dlen = (int)strlen((char *)here);
259 if (dlen > 0 && here[dlen - 1] == '\n') return start;
260 here += dlen;
261 }
262
263 else
264 {
265 int new_buffer_size = 2*buffer_size;
266 uschar *new_buffer = (unsigned char *)malloc(new_buffer_size);
267 uschar *new_dbuffer = (unsigned char *)malloc(new_buffer_size);
268 uschar *new_pbuffer = (unsigned char *)malloc(new_buffer_size);
269
270 if (new_buffer == NULL || new_dbuffer == NULL || new_pbuffer == NULL)
271 {
272 fprintf(stderr, "pcretest: malloc(%d) failed\n", new_buffer_size);
273 exit(1);
274 }
275
276 memcpy(new_buffer, buffer, buffer_size);
277 memcpy(new_pbuffer, pbuffer, buffer_size);
278
279 buffer_size = new_buffer_size;
280
281 start = new_buffer + (start - buffer);
282 here = new_buffer + (here - buffer);
283
284 free(buffer);
285 free(dbuffer);
286 free(pbuffer);
287
288 buffer = new_buffer;
289 dbuffer = new_dbuffer;
290 pbuffer = new_pbuffer;
291 }
292 }
293
294 return NULL; /* Control never gets here */
295 }
296
297
298
299
300
301
302
303 /*************************************************
304 * Read number from string *
305 *************************************************/
306
307 /* We don't use strtoul() because SunOS4 doesn't have it. Rather than mess
308 around with conditional compilation, just do the job by hand. It is only used
309 for unpicking arguments, so just keep it simple.
310
311 Arguments:
312 str string to be converted
313 endptr where to put the end pointer
314
315 Returns: the unsigned long
316 */
317
318 static int
319 get_value(unsigned char *str, unsigned char **endptr)
320 {
321 int result = 0;
322 while(*str != 0 && isspace(*str)) str++;
323 while (isdigit(*str)) result = result * 10 + (int)(*str++ - '0');
324 *endptr = str;
325 return(result);
326 }
327
328
329
330
331 /*************************************************
332 * Convert UTF-8 string to value *
333 *************************************************/
334
335 /* This function takes one or more bytes that represents a UTF-8 character,
336 and returns the value of the character.
337
338 Argument:
339 utf8bytes a pointer to the byte vector
340 vptr a pointer to an int to receive the value
341
342 Returns: > 0 => the number of bytes consumed
343 -6 to 0 => malformed UTF-8 character at offset = (-return)
344 */
345
346 #if !defined NOUTF8
347
348 static int
349 utf82ord(unsigned char *utf8bytes, int *vptr)
350 {
351 int c = *utf8bytes++;
352 int d = c;
353 int i, j, s;
354
355 for (i = -1; i < 6; i++) /* i is number of additional bytes */
356 {
357 if ((d & 0x80) == 0) break;
358 d <<= 1;
359 }
360
361 if (i == -1) { *vptr = c; return 1; } /* ascii character */
362 if (i == 0 || i == 6) return 0; /* invalid UTF-8 */
363
364 /* i now has a value in the range 1-5 */
365
366 s = 6*i;
367 d = (c & utf8_table3[i]) << s;
368
369 for (j = 0; j < i; j++)
370 {
371 c = *utf8bytes++;
372 if ((c & 0xc0) != 0x80) return -(j+1);
373 s -= 6;
374 d |= (c & 0x3f) << s;
375 }
376
377 /* Check that encoding was the correct unique one */
378
379 for (j = 0; j < utf8_table1_size; j++)
380 if (d <= utf8_table1[j]) break;
381 if (j != i) return -(i+1);
382
383 /* Valid value */
384
385 *vptr = d;
386 return i+1;
387 }
388
389 #endif
390
391
392
393 /*************************************************
394 * Convert character value to UTF-8 *
395 *************************************************/
396
397 /* This function takes an integer value in the range 0 - 0x7fffffff
398 and encodes it as a UTF-8 character in 0 to 6 bytes.
399
400 Arguments:
401 cvalue the character value
402 utf8bytes pointer to buffer for result - at least 6 bytes long
403
404 Returns: number of characters placed in the buffer
405 */
406
407 #if !defined NOUTF8
408
409 static int
410 ord2utf8(int cvalue, uschar *utf8bytes)
411 {
412 register int i, j;
413 for (i = 0; i < utf8_table1_size; i++)
414 if (cvalue <= utf8_table1[i]) break;
415 utf8bytes += i;
416 for (j = i; j > 0; j--)
417 {
418 *utf8bytes-- = 0x80 | (cvalue & 0x3f);
419 cvalue >>= 6;
420 }
421 *utf8bytes = utf8_table2[i] | cvalue;
422 return i + 1;
423 }
424
425 #endif
426
427
428
429 /*************************************************
430 * Print character string *
431 *************************************************/
432
433 /* Character string printing function. Must handle UTF-8 strings in utf8
434 mode. Yields number of characters printed. If handed a NULL file, just counts
435 chars without printing. */
436
437 static int pchars(unsigned char *p, int length, FILE *f)
438 {
439 int c = 0;
440 int yield = 0;
441
442 while (length-- > 0)
443 {
444 #if !defined NOUTF8
445 if (use_utf8)
446 {
447 int rc = utf82ord(p, &c);
448
449 if (rc > 0 && rc <= length + 1) /* Mustn't run over the end */
450 {
451 length -= rc - 1;
452 p += rc;
453 if (PRINTHEX(c))
454 {
455 if (f != NULL) fprintf(f, "%c", c);
456 yield++;
457 }
458 else
459 {
460 int n = 4;
461 if (f != NULL) fprintf(f, "\\x{%02x}", c);
462 yield += (n <= 0x000000ff)? 2 :
463 (n <= 0x00000fff)? 3 :
464 (n <= 0x0000ffff)? 4 :
465 (n <= 0x000fffff)? 5 : 6;
466 }
467 continue;
468 }
469 }
470 #endif
471
472 /* Not UTF-8, or malformed UTF-8 */
473
474 c = *p++;
475 if (PRINTHEX(c))
476 {
477 if (f != NULL) fprintf(f, "%c", c);
478 yield++;
479 }
480 else
481 {
482 if (f != NULL) fprintf(f, "\\x%02x", c);
483 yield += 4;
484 }
485 }
486
487 return yield;
488 }
489
490
491
492 /*************************************************
493 * Callout function *
494 *************************************************/
495
496 /* Called from PCRE as a result of the (?C) item. We print out where we are in
497 the match. Yield zero unless more callouts than the fail count, or the callout
498 data is not zero. */
499
500 static int callout(pcre_callout_block *cb)
501 {
502 FILE *f = (first_callout | callout_extra)? outfile : NULL;
503 int i, pre_start, post_start, subject_length;
504
505 if (callout_extra)
506 {
507 fprintf(f, "Callout %d: last capture = %d\n",
508 cb->callout_number, cb->capture_last);
509
510 for (i = 0; i < cb->capture_top * 2; i += 2)
511 {
512 if (cb->offset_vector[i] < 0)
513 fprintf(f, "%2d: <unset>\n", i/2);
514 else
515 {
516 fprintf(f, "%2d: ", i/2);
517 (void)pchars((unsigned char *)cb->subject + cb->offset_vector[i],
518 cb->offset_vector[i+1] - cb->offset_vector[i], f);
519 fprintf(f, "\n");
520 }
521 }
522 }
523
524 /* Re-print the subject in canonical form, the first time or if giving full
525 datails. On subsequent calls in the same match, we use pchars just to find the
526 printed lengths of the substrings. */
527
528 if (f != NULL) fprintf(f, "--->");
529
530 pre_start = pchars((unsigned char *)cb->subject, cb->start_match, f);
531 post_start = pchars((unsigned char *)(cb->subject + cb->start_match),
532 cb->current_position - cb->start_match, f);
533
534 subject_length = pchars((unsigned char *)cb->subject, cb->subject_length, NULL);
535
536 (void)pchars((unsigned char *)(cb->subject + cb->current_position),
537 cb->subject_length - cb->current_position, f);
538
539 if (f != NULL) fprintf(f, "\n");
540
541 /* Always print appropriate indicators, with callout number if not already
542 shown. For automatic callouts, show the pattern offset. */
543
544 if (cb->callout_number == 255)
545 {
546 fprintf(outfile, "%+3d ", cb->pattern_position);
547 if (cb->pattern_position > 99) fprintf(outfile, "\n ");
548 }
549 else
550 {
551 if (callout_extra) fprintf(outfile, " ");
552 else fprintf(outfile, "%3d ", cb->callout_number);
553 }
554
555 for (i = 0; i < pre_start; i++) fprintf(outfile, " ");
556 fprintf(outfile, "^");
557
558 if (post_start > 0)
559 {
560 for (i = 0; i < post_start - 1; i++) fprintf(outfile, " ");
561 fprintf(outfile, "^");
562 }
563
564 for (i = 0; i < subject_length - pre_start - post_start + 4; i++)
565 fprintf(outfile, " ");
566
567 fprintf(outfile, "%.*s", (cb->next_item_length == 0)? 1 : cb->next_item_length,
568 pbuffer + cb->pattern_position);
569
570 fprintf(outfile, "\n");
571 first_callout = 0;
572
573 if (cb->callout_data != NULL)
574 {
575 int callout_data = *((int *)(cb->callout_data));
576 if (callout_data != 0)
577 {
578 fprintf(outfile, "Callout data = %d\n", callout_data);
579 return callout_data;
580 }
581 }
582
583 return (cb->callout_number != callout_fail_id)? 0 :
584 (++callout_count >= callout_fail_count)? 1 : 0;
585 }
586
587
588 /*************************************************
589 * Local malloc functions *
590 *************************************************/
591
592 /* Alternative malloc function, to test functionality and show the size of the
593 compiled re. */
594
595 static void *new_malloc(size_t size)
596 {
597 void *block = malloc(size);
598 gotten_store = size;
599 if (show_malloc)
600 fprintf(outfile, "malloc %3d %p\n", (int)size, block);
601 return block;
602 }
603
604 static void new_free(void *block)
605 {
606 if (show_malloc)
607 fprintf(outfile, "free %p\n", block);
608 free(block);
609 }
610
611
612 /* For recursion malloc/free, to test stacking calls */
613
614 static void *stack_malloc(size_t size)
615 {
616 void *block = malloc(size);
617 if (show_malloc)
618 fprintf(outfile, "stack_malloc %3d %p\n", (int)size, block);
619 return block;
620 }
621
622 static void stack_free(void *block)
623 {
624 if (show_malloc)
625 fprintf(outfile, "stack_free %p\n", block);
626 free(block);
627 }
628
629
630 /*************************************************
631 * Call pcre_fullinfo() *
632 *************************************************/
633
634 /* Get one piece of information from the pcre_fullinfo() function */
635
636 static void new_info(pcre *re, pcre_extra *study, int option, void *ptr)
637 {
638 int rc;
639 if ((rc = pcre_fullinfo(re, study, option, ptr)) < 0)
640 fprintf(outfile, "Error %d from pcre_fullinfo(%d)\n", rc, option);
641 }
642
643
644
645 /*************************************************
646 * Byte flipping function *
647 *************************************************/
648
649 static unsigned long int
650 byteflip(unsigned long int value, int n)
651 {
652 if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);
653 return ((value & 0x000000ff) << 24) |
654 ((value & 0x0000ff00) << 8) |
655 ((value & 0x00ff0000) >> 8) |
656 ((value & 0xff000000) >> 24);
657 }
658
659
660
661
662 /*************************************************
663 * Check match or recursion limit *
664 *************************************************/
665
666 static int
667 check_match_limit(pcre *re, pcre_extra *extra, uschar *bptr, int len,
668 int start_offset, int options, int *use_offsets, int use_size_offsets,
669 int flag, unsigned long int *limit, int errnumber, const char *msg)
670 {
671 int count;
672 int min = 0;
673 int mid = 64;
674 int max = -1;
675
676 extra->flags |= flag;
677
678 for (;;)
679 {
680 *limit = mid;
681
682 count = pcre_exec(re, extra, (char *)bptr, len, start_offset, options,
683 use_offsets, use_size_offsets);
684
685 if (count == errnumber)
686 {
687 /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
688 min = mid;
689 mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2;
690 }
691
692 else if (count >= 0 || count == PCRE_ERROR_NOMATCH ||
693 count == PCRE_ERROR_PARTIAL)
694 {
695 if (mid == min + 1)
696 {
697 fprintf(outfile, "Minimum %s limit = %d\n", msg, mid);
698 break;
699 }
700 /* fprintf(outfile, "Testing %s limit = %d\n", msg, mid); */
701 max = mid;
702 mid = (min + mid)/2;
703 }
704 else break; /* Some other error */
705 }
706
707 extra->flags &= ~flag;
708 return count;
709 }
710
711
712
713 /*************************************************
714 * Case-independent strncmp() function *
715 *************************************************/
716
717 /*
718 Arguments:
719 s first string
720 t second string
721 n number of characters to compare
722
723 Returns: < 0, = 0, or > 0, according to the comparison
724 */
725
726 static int
727 strncmpic(uschar *s, uschar *t, int n)
728 {
729 while (n--)
730 {
731 int c = tolower(*s++) - tolower(*t++);
732 if (c) return c;
733 }
734 return 0;
735 }
736
737
738
739 /*************************************************
740 * Check newline indicator *
741 *************************************************/
742
743 /* This is used both at compile and run-time to check for <xxx> escapes, where
744 xxx is LF, CR, CRLF, ANYCRLF, or ANY. Print a message and return 0 if there is
745 no match.
746
747 Arguments:
748 p points after the leading '<'
749 f file for error message
750
751 Returns: appropriate PCRE_NEWLINE_xxx flags, or 0
752 */
753
754 static int
755 check_newline(uschar *p, FILE *f)
756 {
757 if (strncmpic(p, (uschar *)"cr>", 3) == 0) return PCRE_NEWLINE_CR;
758 if (strncmpic(p, (uschar *)"lf>", 3) == 0) return PCRE_NEWLINE_LF;
759 if (strncmpic(p, (uschar *)"crlf>", 5) == 0) return PCRE_NEWLINE_CRLF;
760 if (strncmpic(p, (uschar *)"anycrlf>", 8) == 0) return PCRE_NEWLINE_ANYCRLF;
761 if (strncmpic(p, (uschar *)"any>", 4) == 0) return PCRE_NEWLINE_ANY;
762 if (strncmpic(p, (uschar *)"bsr_anycrlf>", 12) == 0) return PCRE_BSR_ANYCRLF;
763 if (strncmpic(p, (uschar *)"bsr_unicode>", 12) == 0) return PCRE_BSR_UNICODE;
764 fprintf(f, "Unknown newline type at: <%s\n", p);
765 return 0;
766 }
767
768
769
770 /*************************************************
771 * Usage function *
772 *************************************************/
773
774 static void
775 usage(void)
776 {
777 printf("Usage: pcretest [options] [<input file> [<output file>]]\n\n");
778 printf("Input and output default to stdin and stdout.\n");
779 #ifdef SUPPORT_LIBREADLINE
780 printf("If input is a terminal, readline() is used to read from it.\n");
781 #else
782 printf("This version of pcretest is not linked with readline().\n");
783 #endif
784 printf("\nOptions:\n");
785 printf(" -b show compiled code (bytecode)\n");
786 printf(" -C show PCRE compile-time options and exit\n");
787 printf(" -d debug: show compiled code and information (-b and -i)\n");
788 #if !defined NODFA
789 printf(" -dfa force DFA matching for all subjects\n");
790 #endif
791 printf(" -help show usage information\n");
792 printf(" -i show information about compiled patterns\n"
793 " -M find MATCH_LIMIT minimum for each subject\n"
794 " -m output memory used information\n"
795 " -o <n> set size of offsets vector to <n>\n");
796 #if !defined NOPOSIX
797 printf(" -p use POSIX interface\n");
798 #endif
799 printf(" -q quiet: do not output PCRE version number at start\n");
800 printf(" -S <n> set stack size to <n> megabytes\n");
801 printf(" -s output store (memory) used information\n"
802 " -t time compilation and execution\n");
803 printf(" -t <n> time compilation and execution, repeating <n> times\n");
804 printf(" -tm time execution (matching) only\n");
805 printf(" -tm <n> time execution (matching) only, repeating <n> times\n");
806 }
807
808
809
810 /*************************************************
811 * Main Program *
812 *************************************************/
813
814 /* Read lines from named file or stdin and write to named file or stdout; lines
815 consist of a regular expression, in delimiters and optionally followed by
816 options, followed by a set of test data, terminated by an empty line. */
817
818 int main(int argc, char **argv)
819 {
820 FILE *infile = stdin;
821 int options = 0;
822 int study_options = 0;
823 int default_find_match_limit = FALSE;
824 int op = 1;
825 int timeit = 0;
826 int timeitm = 0;
827 int showinfo = 0;
828 int showstore = 0;
829 int quiet = 0;
830 int size_offsets = 45;
831 int size_offsets_max;
832 int *offsets = NULL;
833 #if !defined NOPOSIX
834 int posix = 0;
835 #endif
836 int debug = 0;
837 int done = 0;
838 int all_use_dfa = 0;
839 int yield = 0;
840 int stack_size;
841
842 /* These vectors store, end-to-end, a list of captured substring names. Assume
843 that 1024 is plenty long enough for the few names we'll be testing. */
844
845 uschar copynames[1024];
846 uschar getnames[1024];
847
848 uschar *copynamesptr;
849 uschar *getnamesptr;
850
851 /* Get buffers from malloc() so that Electric Fence will check their misuse
852 when I am debugging. They grow automatically when very long lines are read. */
853
854 buffer = (unsigned char *)malloc(buffer_size);
855 dbuffer = (unsigned char *)malloc(buffer_size);
856 pbuffer = (unsigned char *)malloc(buffer_size);
857
858 /* The outfile variable is static so that new_malloc can use it. */
859
860 outfile = stdout;
861
862 /* The following _setmode() stuff is some Windows magic that tells its runtime
863 library to translate CRLF into a single LF character. At least, that's what
864 I've been told: never having used Windows I take this all on trust. Originally
865 it set 0x8000, but then I was advised that _O_BINARY was better. */
866
867 #if defined(_WIN32) || defined(WIN32)
868 _setmode( _fileno( stdout ), _O_BINARY );
869 #endif
870
871 /* Scan options */
872
873 while (argc > 1 && argv[op][0] == '-')
874 {
875 unsigned char *endptr;
876
877 if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0)
878 showstore = 1;
879 else if (strcmp(argv[op], "-q") == 0) quiet = 1;
880 else if (strcmp(argv[op], "-b") == 0) debug = 1;
881 else if (strcmp(argv[op], "-i") == 0) showinfo = 1;
882 else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1;
883 else if (strcmp(argv[op], "-M") == 0) default_find_match_limit = TRUE;
884 #if !defined NODFA
885 else if (strcmp(argv[op], "-dfa") == 0) all_use_dfa = 1;
886 #endif
887 else if (strcmp(argv[op], "-o") == 0 && argc > 2 &&
888 ((size_offsets = get_value((unsigned char *)argv[op+1], &endptr)),
889 *endptr == 0))
890 {
891 op++;
892 argc--;
893 }
894 else if (strcmp(argv[op], "-t") == 0 || strcmp(argv[op], "-tm") == 0)
895 {
896 int both = argv[op][2] == 0;
897 int temp;
898 if (argc > 2 && (temp = get_value((unsigned char *)argv[op+1], &endptr),
899 *endptr == 0))
900 {
901 timeitm = temp;
902 op++;
903 argc--;
904 }
905 else timeitm = LOOPREPEAT;
906 if (both) timeit = timeitm;
907 }
908 else if (strcmp(argv[op], "-S") == 0 && argc > 2 &&
909 ((stack_size = get_value((unsigned char *)argv[op+1], &endptr)),
910 *endptr == 0))
911 {
912 #if defined(_WIN32) || defined(WIN32)
913 printf("PCRE: -S not supported on this OS\n");
914 exit(1);
915 #else
916 int rc;
917 struct rlimit rlim;
918 getrlimit(RLIMIT_STACK, &rlim);
919 rlim.rlim_cur = stack_size * 1024 * 1024;
920 rc = setrlimit(RLIMIT_STACK, &rlim);
921 if (rc != 0)
922 {
923 printf("PCRE: setrlimit() failed with error %d\n", rc);
924 exit(1);
925 }
926 op++;
927 argc--;
928 #endif
929 }
930 #if !defined NOPOSIX
931 else if (strcmp(argv[op], "-p") == 0) posix = 1;
932 #endif
933 else if (strcmp(argv[op], "-C") == 0)
934 {
935 int rc;
936 unsigned long int lrc;
937 printf("PCRE version %s\n", pcre_version());
938 printf("Compiled with\n");
939 (void)pcre_config(PCRE_CONFIG_UTF8, &rc);
940 printf(" %sUTF-8 support\n", rc? "" : "No ");
941 (void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc);
942 printf(" %sUnicode properties support\n", rc? "" : "No ");
943 (void)pcre_config(PCRE_CONFIG_NEWLINE, &rc);
944 /* Note that these values are always the ASCII values, even
945 in EBCDIC environments. CR is 13 and NL is 10. */
946 printf(" Newline sequence is %s\n", (rc == 13)? "CR" :
947 (rc == 10)? "LF" : (rc == (13<<8 | 10))? "CRLF" :
948 (rc == -2)? "ANYCRLF" :
949 (rc == -1)? "ANY" : "???");
950 (void)pcre_config(PCRE_CONFIG_BSR, &rc);
951 printf(" \\R matches %s\n", rc? "CR, LF, or CRLF only" :
952 "all Unicode newlines");
953 (void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc);
954 printf(" Internal link size = %d\n", rc);
955 (void)pcre_config(PCRE_CONFIG_POSIX_MALLOC_THRESHOLD, &rc);
956 printf(" POSIX malloc threshold = %d\n", rc);
957 (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &lrc);
958 printf(" Default match limit = %ld\n", lrc);
959 (void)pcre_config(PCRE_CONFIG_MATCH_LIMIT_RECURSION, &lrc);
960 printf(" Default recursion depth limit = %ld\n", lrc);
961 (void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc);
962 printf(" Match recursion uses %s\n", rc? "stack" : "heap");
963 goto EXIT;
964 }
965 else if (strcmp(argv[op], "-help") == 0 ||
966 strcmp(argv[op], "--help") == 0)
967 {
968 usage();
969 goto EXIT;
970 }
971 else
972 {
973 printf("** Unknown or malformed option %s\n", argv[op]);
974 usage();
975 yield = 1;
976 goto EXIT;
977 }
978 op++;
979 argc--;
980 }
981
982 /* Get the store for the offsets vector, and remember what it was */
983
984 size_offsets_max = size_offsets;
985 offsets = (int *)malloc(size_offsets_max * sizeof(int));
986 if (offsets == NULL)
987 {
988 printf("** Failed to get %d bytes of memory for offsets vector\n",
989 (int)(size_offsets_max * sizeof(int)));
990 yield = 1;
991 goto EXIT;
992 }
993
994 /* Sort out the input and output files */
995
996 if (argc > 1)
997 {
998 infile = fopen(argv[op], INPUT_MODE);
999 if (infile == NULL)
1000 {
1001 printf("** Failed to open %s\n", argv[op]);
1002 yield = 1;
1003 goto EXIT;
1004 }
1005 }
1006
1007 if (argc > 2)
1008 {
1009 outfile = fopen(argv[op+1], OUTPUT_MODE);
1010 if (outfile == NULL)
1011 {
1012 printf("** Failed to open %s\n", argv[op+1]);
1013 yield = 1;
1014 goto EXIT;
1015 }
1016 }
1017
1018 /* Set alternative malloc function */
1019
1020 pcre_malloc = new_malloc;
1021 pcre_free = new_free;
1022 pcre_stack_malloc = stack_malloc;
1023 pcre_stack_free = stack_free;
1024
1025 /* Heading line unless quiet, then prompt for first regex if stdin */
1026
1027 if (!quiet) fprintf(outfile, "PCRE version %s\n\n", pcre_version());
1028
1029 /* Main loop */
1030
1031 while (!done)
1032 {
1033 pcre *re = NULL;
1034 pcre_extra *extra = NULL;
1035
1036 #if !defined NOPOSIX /* There are still compilers that require no indent */
1037 regex_t preg;
1038 int do_posix = 0;
1039 #endif
1040
1041 const char *error;
1042 unsigned char *p, *pp, *ppp;
1043 unsigned char *to_file = NULL;
1044 const unsigned char *tables = NULL;
1045 unsigned long int true_size, true_study_size = 0;
1046 size_t size, regex_gotten_store;
1047 int do_study = 0;
1048 int do_debug = debug;
1049 int do_G = 0;
1050 int do_g = 0;
1051 int do_showinfo = showinfo;
1052 int do_showrest = 0;
1053 int do_flip = 0;
1054 int erroroffset, len, delimiter, poffset;
1055
1056 use_utf8 = 0;
1057 debug_lengths = 1;
1058
1059 if (extend_inputline(infile, buffer, " re> ") == NULL) break;
1060 if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1061 fflush(outfile);
1062
1063 p = buffer;
1064 while (isspace(*p)) p++;
1065 if (*p == 0) continue;
1066
1067 /* See if the pattern is to be loaded pre-compiled from a file. */
1068
1069 if (*p == '<' && strchr((char *)(p+1), '<') == NULL)
1070 {
1071 unsigned long int magic, get_options;
1072 uschar sbuf[8];
1073 FILE *f;
1074
1075 p++;
1076 pp = p + (int)strlen((char *)p);
1077 while (isspace(pp[-1])) pp--;
1078 *pp = 0;
1079
1080 f = fopen((char *)p, "rb");
1081 if (f == NULL)
1082 {
1083 fprintf(outfile, "Failed to open %s: %s\n", p, strerror(errno));
1084 continue;
1085 }
1086
1087 if (fread(sbuf, 1, 8, f) != 8) goto FAIL_READ;
1088
1089 true_size =
1090 (sbuf[0] << 24) | (sbuf[1] << 16) | (sbuf[2] << 8) | sbuf[3];
1091 true_study_size =
1092 (sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7];
1093
1094 re = (real_pcre *)new_malloc(true_size);
1095 regex_gotten_store = gotten_store;
1096
1097 if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ;
1098
1099 magic = ((real_pcre *)re)->magic_number;
1100 if (magic != MAGIC_NUMBER)
1101 {
1102 if (byteflip(magic, sizeof(magic)) == MAGIC_NUMBER)
1103 {
1104 do_flip = 1;
1105 }
1106 else
1107 {
1108 fprintf(outfile, "Data in %s is not a compiled PCRE regex\n", p);
1109 fclose(f);
1110 continue;
1111 }
1112 }
1113
1114 fprintf(outfile, "Compiled regex%s loaded from %s\n",
1115 do_flip? " (byte-inverted)" : "", p);
1116
1117 /* Need to know if UTF-8 for printing data strings */
1118
1119 new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1120 use_utf8 = (get_options & PCRE_UTF8) != 0;
1121
1122 /* Now see if there is any following study data */
1123
1124 if (true_study_size != 0)
1125 {
1126 pcre_study_data *psd;
1127
1128 extra = (pcre_extra *)new_malloc(sizeof(pcre_extra) + true_study_size);
1129 extra->flags = PCRE_EXTRA_STUDY_DATA;
1130
1131 psd = (pcre_study_data *)(((char *)extra) + sizeof(pcre_extra));
1132 extra->study_data = psd;
1133
1134 if (fread(psd, 1, true_study_size, f) != true_study_size)
1135 {
1136 FAIL_READ:
1137 fprintf(outfile, "Failed to read data from %s\n", p);
1138 if (extra != NULL) new_free(extra);
1139 if (re != NULL) new_free(re);
1140 fclose(f);
1141 continue;
1142 }
1143 fprintf(outfile, "Study data loaded from %s\n", p);
1144 do_study = 1; /* To get the data output if requested */
1145 }
1146 else fprintf(outfile, "No study data\n");
1147
1148 fclose(f);
1149 goto SHOW_INFO;
1150 }
1151
1152 /* In-line pattern (the usual case). Get the delimiter and seek the end of
1153 the pattern; if is isn't complete, read more. */
1154
1155 delimiter = *p++;
1156
1157 if (isalnum(delimiter) || delimiter == '\\')
1158 {
1159 fprintf(outfile, "** Delimiter must not be alphanumeric or \\\n");
1160 goto SKIP_DATA;
1161 }
1162
1163 pp = p;
1164 poffset = p - buffer;
1165
1166 for(;;)
1167 {
1168 while (*pp != 0)
1169 {
1170 if (*pp == '\\' && pp[1] != 0) pp++;
1171 else if (*pp == delimiter) break;
1172 pp++;
1173 }
1174 if (*pp != 0) break;
1175 if ((pp = extend_inputline(infile, pp, " > ")) == NULL)
1176 {
1177 fprintf(outfile, "** Unexpected EOF\n");
1178 done = 1;
1179 goto CONTINUE;
1180 }
1181 if (infile != stdin) fprintf(outfile, "%s", (char *)pp);
1182 }
1183
1184 /* The buffer may have moved while being extended; reset the start of data
1185 pointer to the correct relative point in the buffer. */
1186
1187 p = buffer + poffset;
1188
1189 /* If the first character after the delimiter is backslash, make
1190 the pattern end with backslash. This is purely to provide a way
1191 of testing for the error message when a pattern ends with backslash. */
1192
1193 if (pp[1] == '\\') *pp++ = '\\';
1194
1195 /* Terminate the pattern at the delimiter, and save a copy of the pattern
1196 for callouts. */
1197
1198 *pp++ = 0;
1199 strcpy((char *)pbuffer, (char *)p);
1200
1201 /* Look for options after final delimiter */
1202
1203 options = 0;
1204 study_options = 0;
1205 log_store = showstore; /* default from command line */
1206
1207 while (*pp != 0)
1208 {
1209 switch (*pp++)
1210 {
1211 case 'f': options |= PCRE_FIRSTLINE; break;
1212 case 'g': do_g = 1; break;
1213 case 'i': options |= PCRE_CASELESS; break;
1214 case 'm': options |= PCRE_MULTILINE; break;
1215 case 's': options |= PCRE_DOTALL; break;
1216 case 'x': options |= PCRE_EXTENDED; break;
1217
1218 case '+': do_showrest = 1; break;
1219 case 'A': options |= PCRE_ANCHORED; break;
1220 case 'B': do_debug = 1; break;
1221 case 'C': options |= PCRE_AUTO_CALLOUT; break;
1222 case 'D': do_debug = do_showinfo = 1; break;
1223 case 'E': options |= PCRE_DOLLAR_ENDONLY; break;
1224 case 'F': do_flip = 1; break;
1225 case 'G': do_G = 1; break;
1226 case 'I': do_showinfo = 1; break;
1227 case 'J': options |= PCRE_DUPNAMES; break;
1228 case 'M': log_store = 1; break;
1229 case 'N': options |= PCRE_NO_AUTO_CAPTURE; break;
1230
1231 #if !defined NOPOSIX
1232 case 'P': do_posix = 1; break;
1233 #endif
1234
1235 case 'S': do_study = 1; break;
1236 case 'U': options |= PCRE_UNGREEDY; break;
1237 case 'X': options |= PCRE_EXTRA; break;
1238 case 'Z': debug_lengths = 0; break;
1239 case '8': options |= PCRE_UTF8; use_utf8 = 1; break;
1240 case '?': options |= PCRE_NO_UTF8_CHECK; break;
1241
1242 case 'L':
1243 ppp = pp;
1244 /* The '\r' test here is so that it works on Windows. */
1245 /* The '0' test is just in case this is an unterminated line. */
1246 while (*ppp != 0 && *ppp != '\n' && *ppp != '\r' && *ppp != ' ') ppp++;
1247 *ppp = 0;
1248 if (setlocale(LC_CTYPE, (const char *)pp) == NULL)
1249 {
1250 fprintf(outfile, "** Failed to set locale \"%s\"\n", pp);
1251 goto SKIP_DATA;
1252 }
1253 locale_set = 1;
1254 tables = pcre_maketables();
1255 pp = ppp;
1256 break;
1257
1258 case '>':
1259 to_file = pp;
1260 while (*pp != 0) pp++;
1261 while (isspace(pp[-1])) pp--;
1262 *pp = 0;
1263 break;
1264
1265 case '<':
1266 {
1267 if (strncmp((char *)pp, "JS>", 3) == 0)
1268 {
1269 options |= PCRE_JAVASCRIPT_COMPAT;
1270 pp += 3;
1271 }
1272 else
1273 {
1274 int x = check_newline(pp, outfile);
1275 if (x == 0) goto SKIP_DATA;
1276 options |= x;
1277 while (*pp++ != '>');
1278 }
1279 }
1280 break;
1281
1282 case '\r': /* So that it works in Windows */
1283 case '\n':
1284 case ' ':
1285 break;
1286
1287 default:
1288 fprintf(outfile, "** Unknown option '%c'\n", pp[-1]);
1289 goto SKIP_DATA;
1290 }
1291 }
1292
1293 /* Handle compiling via the POSIX interface, which doesn't support the
1294 timing, showing, or debugging options, nor the ability to pass over
1295 local character tables. */
1296
1297 #if !defined NOPOSIX
1298 if (posix || do_posix)
1299 {
1300 int rc;
1301 int cflags = 0;
1302
1303 if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE;
1304 if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE;
1305 if ((options & PCRE_DOTALL) != 0) cflags |= REG_DOTALL;
1306 if ((options & PCRE_NO_AUTO_CAPTURE) != 0) cflags |= REG_NOSUB;
1307 if ((options & PCRE_UTF8) != 0) cflags |= REG_UTF8;
1308
1309 rc = regcomp(&preg, (char *)p, cflags);
1310
1311 /* Compilation failed; go back for another re, skipping to blank line
1312 if non-interactive. */
1313
1314 if (rc != 0)
1315 {
1316 (void)regerror(rc, &preg, (char *)buffer, buffer_size);
1317 fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer);
1318 goto SKIP_DATA;
1319 }
1320 }
1321
1322 /* Handle compiling via the native interface */
1323
1324 else
1325 #endif /* !defined NOPOSIX */
1326
1327 {
1328 unsigned long int get_options;
1329
1330 if (timeit > 0)
1331 {
1332 register int i;
1333 clock_t time_taken;
1334 clock_t start_time = clock();
1335 for (i = 0; i < timeit; i++)
1336 {
1337 re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1338 if (re != NULL) free(re);
1339 }
1340 time_taken = clock() - start_time;
1341 fprintf(outfile, "Compile time %.4f milliseconds\n",
1342 (((double)time_taken * 1000.0) / (double)timeit) /
1343 (double)CLOCKS_PER_SEC);
1344 }
1345
1346 re = pcre_compile((char *)p, options, &error, &erroroffset, tables);
1347
1348 /* Compilation failed; go back for another re, skipping to blank line
1349 if non-interactive. */
1350
1351 if (re == NULL)
1352 {
1353 fprintf(outfile, "Failed: %s at offset %d\n", error, erroroffset);
1354 SKIP_DATA:
1355 if (infile != stdin)
1356 {
1357 for (;;)
1358 {
1359 if (extend_inputline(infile, buffer, NULL) == NULL)
1360 {
1361 done = 1;
1362 goto CONTINUE;
1363 }
1364 len = (int)strlen((char *)buffer);
1365 while (len > 0 && isspace(buffer[len-1])) len--;
1366 if (len == 0) break;
1367 }
1368 fprintf(outfile, "\n");
1369 }
1370 goto CONTINUE;
1371 }
1372
1373 /* Compilation succeeded. It is now possible to set the UTF-8 option from
1374 within the regex; check for this so that we know how to process the data
1375 lines. */
1376
1377 new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options);
1378 if ((get_options & PCRE_UTF8) != 0) use_utf8 = 1;
1379
1380 /* Print information if required. There are now two info-returning
1381 functions. The old one has a limited interface and returns only limited
1382 data. Check that it agrees with the newer one. */
1383
1384 if (log_store)
1385 fprintf(outfile, "Memory allocation (code space): %d\n",
1386 (int)(gotten_store -
1387 sizeof(real_pcre) -
1388 ((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size));
1389
1390 /* Extract the size for possible writing before possibly flipping it,
1391 and remember the store that was got. */
1392
1393 true_size = ((real_pcre *)re)->size;
1394 regex_gotten_store = gotten_store;
1395
1396 /* If /S was present, study the regexp to generate additional info to
1397 help with the matching. */
1398
1399 if (do_study)
1400 {
1401 if (timeit > 0)
1402 {
1403 register int i;
1404 clock_t time_taken;
1405 clock_t start_time = clock();
1406 for (i = 0; i < timeit; i++)
1407 extra = pcre_study(re, study_options, &error);
1408 time_taken = clock() - start_time;
1409 if (extra != NULL) free(extra);
1410 fprintf(outfile, " Study time %.4f milliseconds\n",
1411 (((double)time_taken * 1000.0) / (double)timeit) /
1412 (double)CLOCKS_PER_SEC);
1413 }
1414 extra = pcre_study(re, study_options, &error);
1415 if (error != NULL)
1416 fprintf(outfile, "Failed to study: %s\n", error);
1417 else if (extra != NULL)
1418 true_study_size = ((pcre_study_data *)(extra->study_data))->size;
1419 }
1420
1421 /* If the 'F' option was present, we flip the bytes of all the integer
1422 fields in the regex data block and the study block. This is to make it
1423 possible to test PCRE's handling of byte-flipped patterns, e.g. those
1424 compiled on a different architecture. */
1425
1426 if (do_flip)
1427 {
1428 real_pcre *rre = (real_pcre *)re;
1429 rre->magic_number =
1430 byteflip(rre->magic_number, sizeof(rre->magic_number));
1431 rre->size = byteflip(rre->size, sizeof(rre->size));
1432 rre->options = byteflip(rre->options, sizeof(rre->options));
1433 rre->flags = (pcre_uint16)byteflip(rre->flags, sizeof(rre->flags));
1434 rre->top_bracket =
1435 (pcre_uint16)byteflip(rre->top_bracket, sizeof(rre->top_bracket));
1436 rre->top_backref =
1437 (pcre_uint16)byteflip(rre->top_backref, sizeof(rre->top_backref));
1438 rre->first_byte =
1439 (pcre_uint16)byteflip(rre->first_byte, sizeof(rre->first_byte));
1440 rre->req_byte =
1441 (pcre_uint16)byteflip(rre->req_byte, sizeof(rre->req_byte));
1442 rre->name_table_offset = (pcre_uint16)byteflip(rre->name_table_offset,
1443 sizeof(rre->name_table_offset));
1444 rre->name_entry_size = (pcre_uint16)byteflip(rre->name_entry_size,
1445 sizeof(rre->name_entry_size));
1446 rre->name_count = (pcre_uint16)byteflip(rre->name_count,
1447 sizeof(rre->name_count));
1448
1449 if (extra != NULL)
1450 {
1451 pcre_study_data *rsd = (pcre_study_data *)(extra->study_data);
1452 rsd->size = byteflip(rsd->size, sizeof(rsd->size));
1453 rsd->options = byteflip(rsd->options, sizeof(rsd->options));
1454 }
1455 }
1456
1457 /* Extract information from the compiled data if required */
1458
1459 SHOW_INFO:
1460
1461 if (do_debug)
1462 {
1463 fprintf(outfile, "------------------------------------------------------------------\n");
1464 pcre_printint(re, outfile, debug_lengths);
1465 }
1466
1467 /* We already have the options in get_options (see above) */
1468
1469 if (do_showinfo)
1470 {
1471 unsigned long int all_options;
1472 #if !defined NOINFOCHECK
1473 int old_first_char, old_options, old_count;
1474 #endif
1475 int count, backrefmax, first_char, need_char, okpartial, jchanged,
1476 hascrorlf;
1477 int nameentrysize, namecount;
1478 const uschar *nametable;
1479
1480 new_info(re, NULL, PCRE_INFO_SIZE, &size);
1481 new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count);
1482 new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax);
1483 new_info(re, NULL, PCRE_INFO_FIRSTBYTE, &first_char);
1484 new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char);
1485 new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize);
1486 new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount);
1487 new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable);
1488 new_info(re, NULL, PCRE_INFO_OKPARTIAL, &okpartial);
1489 new_info(re, NULL, PCRE_INFO_JCHANGED, &jchanged);
1490 new_info(re, NULL, PCRE_INFO_HASCRORLF, &hascrorlf);
1491
1492 #if !defined NOINFOCHECK
1493 old_count = pcre_info(re, &old_options, &old_first_char);
1494 if (count < 0) fprintf(outfile,
1495 "Error %d from pcre_info()\n", count);
1496 else
1497 {
1498 if (old_count != count) fprintf(outfile,
1499 "Count disagreement: pcre_fullinfo=%d pcre_info=%d\n", count,
1500 old_count);
1501
1502 if (old_first_char != first_char) fprintf(outfile,
1503 "First char disagreement: pcre_fullinfo=%d pcre_info=%d\n",
1504 first_char, old_first_char);
1505
1506 if (old_options != (int)get_options) fprintf(outfile,
1507 "Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n",
1508 get_options, old_options);
1509 }
1510 #endif
1511
1512 if (size != regex_gotten_store) fprintf(outfile,
1513 "Size disagreement: pcre_fullinfo=%d call to malloc for %d\n",
1514 (int)size, (int)regex_gotten_store);
1515
1516 fprintf(outfile, "Capturing subpattern count = %d\n", count);
1517 if (backrefmax > 0)
1518 fprintf(outfile, "Max back reference = %d\n", backrefmax);
1519
1520 if (namecount > 0)
1521 {
1522 fprintf(outfile, "Named capturing subpatterns:\n");
1523 while (namecount-- > 0)
1524 {
1525 fprintf(outfile, " %s %*s%3d\n", nametable + 2,
1526 nameentrysize - 3 - (int)strlen((char *)nametable + 2), "",
1527 GET2(nametable, 0));
1528 nametable += nameentrysize;
1529 }
1530 }
1531
1532 if (!okpartial) fprintf(outfile, "Partial matching not supported\n");
1533 if (hascrorlf) fprintf(outfile, "Contains explicit CR or LF match\n");
1534
1535 all_options = ((real_pcre *)re)->options;
1536 if (do_flip) all_options = byteflip(all_options, sizeof(all_options));
1537
1538 if (get_options == 0) fprintf(outfile, "No options\n");
1539 else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s\n",
1540 ((get_options & PCRE_ANCHORED) != 0)? " anchored" : "",
1541 ((get_options & PCRE_CASELESS) != 0)? " caseless" : "",
1542 ((get_options & PCRE_EXTENDED) != 0)? " extended" : "",
1543 ((get_options & PCRE_MULTILINE) != 0)? " multiline" : "",
1544 ((get_options & PCRE_FIRSTLINE) != 0)? " firstline" : "",
1545 ((get_options & PCRE_DOTALL) != 0)? " dotall" : "",
1546 ((get_options & PCRE_BSR_ANYCRLF) != 0)? " bsr_anycrlf" : "",
1547 ((get_options & PCRE_BSR_UNICODE) != 0)? " bsr_unicode" : "",
1548 ((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "",
1549 ((get_options & PCRE_EXTRA) != 0)? " extra" : "",
1550 ((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "",
1551 ((get_options & PCRE_NO_AUTO_CAPTURE) != 0)? " no_auto_capture" : "",
1552 ((get_options & PCRE_UTF8) != 0)? " utf8" : "",
1553 ((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : "",
1554 ((get_options & PCRE_DUPNAMES) != 0)? " dupnames" : "");
1555
1556 if (jchanged) fprintf(outfile, "Duplicate name status changes\n");
1557
1558 switch (get_options & PCRE_NEWLINE_BITS)
1559 {
1560 case PCRE_NEWLINE_CR:
1561 fprintf(outfile, "Forced newline sequence: CR\n");
1562 break;
1563
1564 case PCRE_NEWLINE_LF:
1565 fprintf(outfile, "Forced newline sequence: LF\n");
1566 break;
1567
1568 case PCRE_NEWLINE_CRLF:
1569 fprintf(outfile, "Forced newline sequence: CRLF\n");
1570 break;
1571
1572 case PCRE_NEWLINE_ANYCRLF:
1573 fprintf(outfile, "Forced newline sequence: ANYCRLF\n");
1574 break;
1575
1576 case PCRE_NEWLINE_ANY:
1577 fprintf(outfile, "Forced newline sequence: ANY\n");
1578 break;
1579
1580 default:
1581 break;
1582 }
1583
1584 if (first_char == -1)
1585 {
1586 fprintf(outfile, "First char at start or follows newline\n");
1587 }
1588 else if (first_char < 0)
1589 {
1590 fprintf(outfile, "No first char\n");
1591 }
1592 else
1593 {
1594 int ch = first_char & 255;
1595 const char *caseless = ((first_char & REQ_CASELESS) == 0)?
1596 "" : " (caseless)";
1597 if (PRINTHEX(ch))
1598 fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless);
1599 else
1600 fprintf(outfile, "First char = %d%s\n", ch, caseless);
1601 }
1602
1603 if (need_char < 0)
1604 {
1605 fprintf(outfile, "No need char\n");
1606 }
1607 else
1608 {
1609 int ch = need_char & 255;
1610 const char *caseless = ((need_char & REQ_CASELESS) == 0)?
1611 "" : " (caseless)";
1612 if (PRINTHEX(ch))
1613 fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless);
1614 else
1615 fprintf(outfile, "Need char = %d%s\n", ch, caseless);
1616 }
1617
1618 /* Don't output study size; at present it is in any case a fixed
1619 value, but it varies, depending on the computer architecture, and
1620 so messes up the test suite. (And with the /F option, it might be
1621 flipped.) */
1622
1623 if (do_study)
1624 {
1625 if (extra == NULL)
1626 fprintf(outfile, "Study returned NULL\n");
1627 else
1628 {
1629 uschar *start_bits = NULL;
1630 new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits);
1631
1632 if (start_bits == NULL)
1633 fprintf(outfile, "No starting byte set\n");
1634 else
1635 {
1636 int i;
1637 int c = 24;
1638 fprintf(outfile, "Starting byte set: ");
1639 for (i = 0; i < 256; i++)
1640 {
1641 if ((start_bits[i/8] & (1<<(i&7))) != 0)
1642 {
1643 if (c > 75)
1644 {
1645 fprintf(outfile, "\n ");
1646 c = 2;
1647 }
1648 if (PRINTHEX(i) && i != ' ')
1649 {
1650 fprintf(outfile, "%c ", i);
1651 c += 2;
1652 }
1653 else
1654 {
1655 fprintf(outfile, "\\x%02x ", i);
1656 c += 5;
1657 }
1658 }
1659 }
1660 fprintf(outfile, "\n");
1661 }
1662 }
1663 }
1664 }
1665
1666 /* If the '>' option was present, we write out the regex to a file, and
1667 that is all. The first 8 bytes of the file are the regex length and then
1668 the study length, in big-endian order. */
1669
1670 if (to_file != NULL)
1671 {
1672 FILE *f = fopen((char *)to_file, "wb");
1673 if (f == NULL)
1674 {
1675 fprintf(outfile, "Unable to open %s: %s\n", to_file, strerror(errno));
1676 }
1677 else
1678 {
1679 uschar sbuf[8];
1680 sbuf[0] = (uschar)((true_size >> 24) & 255);
1681 sbuf[1] = (uschar)((true_size >> 16) & 255);
1682 sbuf[2] = (uschar)((true_size >> 8) & 255);
1683 sbuf[3] = (uschar)((true_size) & 255);
1684
1685 sbuf[4] = (uschar)((true_study_size >> 24) & 255);
1686 sbuf[5] = (uschar)((true_study_size >> 16) & 255);
1687 sbuf[6] = (uschar)((true_study_size >> 8) & 255);
1688 sbuf[7] = (uschar)((true_study_size) & 255);
1689
1690 if (fwrite(sbuf, 1, 8, f) < 8 ||
1691 fwrite(re, 1, true_size, f) < true_size)
1692 {
1693 fprintf(outfile, "Write error on %s: %s\n", to_file, strerror(errno));
1694 }
1695 else
1696 {
1697 fprintf(outfile, "Compiled regex written to %s\n", to_file);
1698 if (extra != NULL)
1699 {
1700 if (fwrite(extra->study_data, 1, true_study_size, f) <
1701 true_study_size)
1702 {
1703 fprintf(outfile, "Write error on %s: %s\n", to_file,
1704 strerror(errno));
1705 }
1706 else fprintf(outfile, "Study data written to %s\n", to_file);
1707
1708 }
1709 }
1710 fclose(f);
1711 }
1712
1713 new_free(re);
1714 if (extra != NULL) new_free(extra);
1715 if (tables != NULL) new_free((void *)tables);
1716 continue; /* With next regex */
1717 }
1718 } /* End of non-POSIX compile */
1719
1720 /* Read data lines and test them */
1721
1722 for (;;)
1723 {
1724 uschar *q;
1725 uschar *bptr;
1726 int *use_offsets = offsets;
1727 int use_size_offsets = size_offsets;
1728 int callout_data = 0;
1729 int callout_data_set = 0;
1730 int count, c;
1731 int copystrings = 0;
1732 int find_match_limit = default_find_match_limit;
1733 int getstrings = 0;
1734 int getlist = 0;
1735 int gmatched = 0;
1736 int start_offset = 0;
1737 int g_notempty = 0;
1738 int use_dfa = 0;
1739
1740 options = 0;
1741
1742 *copynames = 0;
1743 *getnames = 0;
1744
1745 copynamesptr = copynames;
1746 getnamesptr = getnames;
1747
1748 pcre_callout = callout;
1749 first_callout = 1;
1750 callout_extra = 0;
1751 callout_count = 0;
1752 callout_fail_count = 999999;
1753 callout_fail_id = -1;
1754 show_malloc = 0;
1755
1756 if (extra != NULL) extra->flags &=
1757 ~(PCRE_EXTRA_MATCH_LIMIT|PCRE_EXTRA_MATCH_LIMIT_RECURSION);
1758
1759 len = 0;
1760 for (;;)
1761 {
1762 if (extend_inputline(infile, buffer + len, "data> ") == NULL)
1763 {
1764 if (len > 0) break;
1765 done = 1;
1766 goto CONTINUE;
1767 }
1768 if (infile != stdin) fprintf(outfile, "%s", (char *)buffer);
1769 len = (int)strlen((char *)buffer);
1770 if (buffer[len-1] == '\n') break;
1771 }
1772
1773 while (len > 0 && isspace(buffer[len-1])) len--;
1774 buffer[len] = 0;
1775 if (len == 0) break;
1776
1777 p = buffer;
1778 while (isspace(*p)) p++;
1779
1780 bptr = q = dbuffer;
1781 while ((c = *p++) != 0)
1782 {
1783 int i = 0;
1784 int n = 0;
1785
1786 if (c == '\\') switch ((c = *p++))
1787 {
1788 case 'a': c = 7; break;
1789 case 'b': c = '\b'; break;
1790 case 'e': c = 27; break;
1791 case 'f': c = '\f'; break;
1792 case 'n': c = '\n'; break;
1793 case 'r': c = '\r'; break;
1794 case 't': c = '\t'; break;
1795 case 'v': c = '\v'; break;
1796
1797 case '0': case '1': case '2': case '3':
1798 case '4': case '5': case '6': case '7':
1799 c -= '0';
1800 while (i++ < 2 && isdigit(*p) && *p != '8' && *p != '9')
1801 c = c * 8 + *p++ - '0';
1802
1803 #if !defined NOUTF8
1804 if (use_utf8 && c > 255)
1805 {
1806 unsigned char buff8[8];
1807 int ii, utn;
1808 utn = ord2utf8(c, buff8);
1809 for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1810 c = buff8[ii]; /* Last byte */
1811 }
1812 #endif
1813 break;
1814
1815 case 'x':
1816
1817 /* Handle \x{..} specially - new Perl thing for utf8 */
1818
1819 #if !defined NOUTF8
1820 if (*p == '{')
1821 {
1822 unsigned char *pt = p;
1823 c = 0;
1824 while (isxdigit(*(++pt)))
1825 c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W');
1826 if (*pt == '}')
1827 {
1828 unsigned char buff8[8];
1829 int ii, utn;
1830 if (use_utf8)
1831 {
1832 utn = ord2utf8(c, buff8);
1833 for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
1834 c = buff8[ii]; /* Last byte */
1835 }
1836 else
1837 {
1838 if (c > 255)
1839 fprintf(outfile, "** Character \\x{%x} is greater than 255 and "
1840 "UTF-8 mode is not enabled.\n"
1841 "** Truncation will probably give the wrong result.\n", c);
1842 }
1843 p = pt + 1;
1844 break;
1845 }
1846 /* Not correct form; fall through */
1847 }
1848 #endif
1849
1850 /* Ordinary \x */
1851
1852 c = 0;
1853 while (i++ < 2 && isxdigit(*p))
1854 {
1855 c = c * 16 + tolower(*p) - ((isdigit(*p))? '0' : 'W');
1856 p++;
1857 }
1858 break;
1859
1860 case 0: /* \ followed by EOF allows for an empty line */
1861 p--;
1862 continue;
1863
1864 case '>':
1865 while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0';
1866 continue;
1867
1868 case 'A': /* Option setting */
1869 options |= PCRE_ANCHORED;
1870 continue;
1871
1872 case 'B':
1873 options |= PCRE_NOTBOL;
1874 continue;
1875
1876 case 'C':
1877 if (isdigit(*p)) /* Set copy string */
1878 {
1879 while(isdigit(*p)) n = n * 10 + *p++ - '0';
1880 copystrings |= 1 << n;
1881 }
1882 else if (isalnum(*p))
1883 {
1884 uschar *npp = copynamesptr;
1885 while (isalnum(*p)) *npp++ = *p++;
1886 *npp++ = 0;
1887 *npp = 0;
1888 n = pcre_get_stringnumber(re, (char *)copynamesptr);
1889 if (n < 0)
1890 fprintf(outfile, "no parentheses with name \"%s\"\n", copynamesptr);
1891 copynamesptr = npp;
1892 }
1893 else if (*p == '+')
1894 {
1895 callout_extra = 1;
1896 p++;
1897 }
1898 else if (*p == '-')
1899 {
1900 pcre_callout = NULL;
1901 p++;
1902 }
1903 else if (*p == '!')
1904 {
1905 callout_fail_id = 0;
1906 p++;
1907 while(isdigit(*p))
1908 callout_fail_id = callout_fail_id * 10 + *p++ - '0';
1909 callout_fail_count = 0;
1910 if (*p == '!')
1911 {
1912 p++;
1913 while(isdigit(*p))
1914 callout_fail_count = callout_fail_count * 10 + *p++ - '0';
1915 }
1916 }
1917 else if (*p == '*')
1918 {
1919 int sign = 1;
1920 callout_data = 0;
1921 if (*(++p) == '-') { sign = -1; p++; }
1922 while(isdigit(*p))
1923 callout_data = callout_data * 10 + *p++ - '0';
1924 callout_data *= sign;
1925 callout_data_set = 1;
1926 }
1927 continue;
1928
1929 #if !defined NODFA
1930 case 'D':
1931 #if !defined NOPOSIX
1932 if (posix || do_posix)
1933 printf("** Can't use dfa matching in POSIX mode: \\D ignored\n");
1934 else
1935 #endif
1936 use_dfa = 1;
1937 continue;
1938
1939 case 'F':
1940 options |= PCRE_DFA_SHORTEST;
1941 continue;
1942 #endif
1943
1944 case 'G':
1945 if (isdigit(*p))
1946 {
1947 while(isdigit(*p)) n = n * 10 + *p++ - '0';
1948 getstrings |= 1 << n;
1949 }
1950 else if (isalnum(*p))
1951 {
1952 uschar *npp = getnamesptr;
1953 while (isalnum(*p)) *npp++ = *p++;
1954 *npp++ = 0;
1955 *npp = 0;
1956 n = pcre_get_stringnumber(re, (char *)getnamesptr);
1957 if (n < 0)
1958 fprintf(outfile, "no parentheses with name \"%s\"\n", getnamesptr);
1959 getnamesptr = npp;
1960 }
1961 continue;
1962
1963 case 'L':
1964 getlist = 1;
1965 continue;
1966
1967 case 'M':
1968 find_match_limit = 1;
1969 continue;
1970
1971 case 'N':
1972 options |= PCRE_NOTEMPTY;
1973 continue;
1974
1975 case 'O':
1976 while(isdigit(*p)) n = n * 10 + *p++ - '0';
1977 if (n > size_offsets_max)
1978 {
1979 size_offsets_max = n;
1980 free(offsets);
1981 use_offsets = offsets = (int *)malloc(size_offsets_max * sizeof(int));
1982 if (offsets == NULL)
1983 {
1984 printf("** Failed to get %d bytes of memory for offsets vector\n",
1985 (int)(size_offsets_max * sizeof(int)));
1986 yield = 1;
1987 goto EXIT;
1988 }
1989 }
1990 use_size_offsets = n;
1991 if (n == 0) use_offsets = NULL; /* Ensures it can't write to it */
1992 continue;
1993
1994 case 'P':
1995 options |= PCRE_PARTIAL;
1996 continue;
1997
1998 case 'Q':
1999 while(isdigit(*p)) n = n * 10 + *p++ - '0';
2000 if (extra == NULL)
2001 {
2002 extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2003 extra->flags = 0;
2004 }
2005 extra->flags |= PCRE_EXTRA_MATCH_LIMIT_RECURSION;
2006 extra->match_limit_recursion = n;
2007 continue;
2008
2009 case 'q':
2010 while(isdigit(*p)) n = n * 10 + *p++ - '0';
2011 if (extra == NULL)
2012 {
2013 extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2014 extra->flags = 0;
2015 }
2016 extra->flags |= PCRE_EXTRA_MATCH_LIMIT;
2017 extra->match_limit = n;
2018 continue;
2019
2020 #if !defined NODFA
2021 case 'R':
2022 options |= PCRE_DFA_RESTART;
2023 continue;
2024 #endif
2025
2026 case 'S':
2027 show_malloc = 1;
2028 continue;
2029
2030 case 'Y':
2031 options |= PCRE_NO_START_OPTIMIZE;
2032 continue;
2033
2034 case 'Z':
2035 options |= PCRE_NOTEOL;
2036 continue;
2037
2038 case '?':
2039 options |= PCRE_NO_UTF8_CHECK;
2040 continue;
2041
2042 case '<':
2043 {
2044 int x = check_newline(p, outfile);
2045 if (x == 0) goto NEXT_DATA;
2046 options |= x;
2047 while (*p++ != '>');
2048 }
2049 continue;
2050 }
2051 *q++ = c;
2052 }
2053 *q = 0;
2054 len = q - dbuffer;
2055
2056 /* Move the data to the end of the buffer so that a read over the end of
2057 the buffer will be seen by valgrind, even if it doesn't cause a crash. If
2058 we are using the POSIX interface, we must include the terminating zero. */
2059
2060 #if !defined NOPOSIX
2061 if (posix || do_posix)
2062 {
2063 memmove(bptr + buffer_size - len - 1, bptr, len + 1);
2064 bptr += buffer_size - len - 1;
2065 }
2066 else
2067 #endif
2068 {
2069 memmove(bptr + buffer_size - len, bptr, len);
2070 bptr += buffer_size - len;
2071 }
2072
2073 if ((all_use_dfa || use_dfa) && find_match_limit)
2074 {
2075 printf("**Match limit not relevant for DFA matching: ignored\n");
2076 find_match_limit = 0;
2077 }
2078
2079 /* Handle matching via the POSIX interface, which does not
2080 support timing or playing with the match limit or callout data. */
2081
2082 #if !defined NOPOSIX
2083 if (posix || do_posix)
2084 {
2085 int rc;
2086 int eflags = 0;
2087 regmatch_t *pmatch = NULL;
2088 if (use_size_offsets > 0)
2089 pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets);
2090 if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL;
2091 if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL;
2092 if ((options & PCRE_NOTEMPTY) != 0) eflags |= REG_NOTEMPTY;
2093
2094 rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags);
2095
2096 if (rc != 0)
2097 {
2098 (void)regerror(rc, &preg, (char *)buffer, buffer_size);
2099 fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer);
2100 }
2101 else if ((((const pcre *)preg.re_pcre)->options & PCRE_NO_AUTO_CAPTURE)
2102 != 0)
2103 {
2104 fprintf(outfile, "Matched with REG_NOSUB\n");
2105 }
2106 else
2107 {
2108 size_t i;
2109 for (i = 0; i < (size_t)use_size_offsets; i++)
2110 {
2111 if (pmatch[i].rm_so >= 0)
2112 {
2113 fprintf(outfile, "%2d: ", (int)i);
2114 (void)pchars(dbuffer + pmatch[i].rm_so,
2115 pmatch[i].rm_eo - pmatch[i].rm_so, outfile);
2116 fprintf(outfile, "\n");
2117 if (i == 0 && do_showrest)
2118 {
2119 fprintf(outfile, " 0+ ");
2120 (void)pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo,
2121 outfile);
2122 fprintf(outfile, "\n");
2123 }
2124 }
2125 }
2126 }
2127 free(pmatch);
2128 }
2129
2130 /* Handle matching via the native interface - repeats for /g and /G */
2131
2132 else
2133 #endif /* !defined NOPOSIX */
2134
2135 for (;; gmatched++) /* Loop for /g or /G */
2136 {
2137 if (timeitm > 0)
2138 {
2139 register int i;
2140 clock_t time_taken;
2141 clock_t start_time = clock();
2142
2143 #if !defined NODFA
2144 if (all_use_dfa || use_dfa)
2145 {
2146 int workspace[1000];
2147 for (i = 0; i < timeitm; i++)
2148 count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
2149 options | g_notempty, use_offsets, use_size_offsets, workspace,
2150 sizeof(workspace)/sizeof(int));
2151 }
2152 else
2153 #endif
2154
2155 for (i = 0; i < timeitm; i++)
2156 count = pcre_exec(re, extra, (char *)bptr, len,
2157 start_offset, options | g_notempty, use_offsets, use_size_offsets);
2158
2159 time_taken = clock() - start_time;
2160 fprintf(outfile, "Execute time %.4f milliseconds\n",
2161 (((double)time_taken * 1000.0) / (double)timeitm) /
2162 (double)CLOCKS_PER_SEC);
2163 }
2164
2165 /* If find_match_limit is set, we want to do repeated matches with
2166 varying limits in order to find the minimum value for the match limit and
2167 for the recursion limit. */
2168
2169 if (find_match_limit)
2170 {
2171 if (extra == NULL)
2172 {
2173 extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2174 extra->flags = 0;
2175 }
2176
2177 (void)check_match_limit(re, extra, bptr, len, start_offset,
2178 options|g_notempty, use_offsets, use_size_offsets,
2179 PCRE_EXTRA_MATCH_LIMIT, &(extra->match_limit),
2180 PCRE_ERROR_MATCHLIMIT, "match()");
2181
2182 count = check_match_limit(re, extra, bptr, len, start_offset,
2183 options|g_notempty, use_offsets, use_size_offsets,
2184 PCRE_EXTRA_MATCH_LIMIT_RECURSION, &(extra->match_limit_recursion),
2185 PCRE_ERROR_RECURSIONLIMIT, "match() recursion");
2186 }
2187
2188 /* If callout_data is set, use the interface with additional data */
2189
2190 else if (callout_data_set)
2191 {
2192 if (extra == NULL)
2193 {
2194 extra = (pcre_extra *)malloc(sizeof(pcre_extra));
2195 extra->flags = 0;
2196 }
2197 extra->flags |= PCRE_EXTRA_CALLOUT_DATA;
2198 extra->callout_data = &callout_data;
2199 count = pcre_exec(re, extra, (char *)bptr, len, start_offset,
2200 options | g_notempty, use_offsets, use_size_offsets);
2201 extra->flags &= ~PCRE_EXTRA_CALLOUT_DATA;
2202 }
2203
2204 /* The normal case is just to do the match once, with the default
2205 value of match_limit. */
2206
2207 #if !defined NODFA
2208 else if (all_use_dfa || use_dfa)
2209 {
2210 int workspace[1000];
2211 count = pcre_dfa_exec(re, NULL, (char *)bptr, len, start_offset,
2212 options | g_notempty, use_offsets, use_size_offsets, workspace,
2213 sizeof(workspace)/sizeof(int));
2214 if (count == 0)
2215 {
2216 fprintf(outfile, "Matched, but too many subsidiary matches\n");
2217 count = use_size_offsets/2;
2218 }
2219 }
2220 #endif
2221
2222 else
2223 {
2224 count = pcre_exec(re, extra, (char *)bptr, len,
2225 start_offset, options | g_notempty, use_offsets, use_size_offsets);
2226 if (count == 0)
2227 {
2228 fprintf(outfile, "Matched, but too many substrings\n");
2229 count = use_size_offsets/3;
2230 }
2231 }
2232
2233 /* Matched */
2234
2235 if (count >= 0)
2236 {
2237 int i, maxcount;
2238
2239 #if !defined NODFA
2240 if (all_use_dfa || use_dfa) maxcount = use_size_offsets/2; else
2241 #endif
2242 maxcount = use_size_offsets/3;
2243
2244 /* This is a check against a lunatic return value. */
2245
2246 if (count > maxcount)
2247 {
2248 fprintf(outfile,
2249 "** PCRE error: returned count %d is too big for offset size %d\n",
2250 count, use_size_offsets);
2251 count = use_size_offsets/3;
2252 if (do_g || do_G)
2253 {
2254 fprintf(outfile, "** /%c loop abandoned\n", do_g? 'g' : 'G');
2255 do_g = do_G = FALSE; /* Break g/G loop */
2256 }
2257 }
2258
2259 for (i = 0; i < count * 2; i += 2)
2260 {
2261 if (use_offsets[i] < 0)
2262 fprintf(outfile, "%2d: <unset>\n", i/2);
2263 else
2264 {
2265 fprintf(outfile, "%2d: ", i/2);
2266 (void)pchars(bptr + use_offsets[i],
2267 use_offsets[i+1] - use_offsets[i], outfile);
2268 fprintf(outfile, "\n");
2269 if (i == 0)
2270 {
2271 if (do_showrest)
2272 {
2273 fprintf(outfile, " 0+ ");
2274 (void)pchars(bptr + use_offsets[i+1], len - use_offsets[i+1],
2275 outfile);
2276 fprintf(outfile, "\n");
2277 }
2278 }
2279 }
2280 }
2281
2282 for (i = 0; i < 32; i++)
2283 {
2284 if ((copystrings & (1 << i)) != 0)
2285 {
2286 char copybuffer[256];
2287 int rc = pcre_copy_substring((char *)bptr, use_offsets, count,
2288 i, copybuffer, sizeof(copybuffer));
2289 if (rc < 0)
2290 fprintf(outfile, "copy substring %d failed %d\n", i, rc);
2291 else
2292 fprintf(outfile, "%2dC %s (%d)\n", i, copybuffer, rc);
2293 }
2294 }
2295
2296 for (copynamesptr = copynames;
2297 *copynamesptr != 0;
2298 copynamesptr += (int)strlen((char*)copynamesptr) + 1)
2299 {
2300 char copybuffer[256];
2301 int rc = pcre_copy_named_substring(re, (char *)bptr, use_offsets,
2302 count, (char *)copynamesptr, copybuffer, sizeof(copybuffer));
2303 if (rc < 0)
2304 fprintf(outfile, "copy substring %s failed %d\n", copynamesptr, rc);
2305 else
2306 fprintf(outfile, " C %s (%d) %s\n", copybuffer, rc, copynamesptr);
2307 }
2308
2309 for (i = 0; i < 32; i++)
2310 {
2311 if ((getstrings & (1 << i)) != 0)
2312 {
2313 const char *substring;
2314 int rc = pcre_get_substring((char *)bptr, use_offsets, count,
2315 i, &substring);
2316 if (rc < 0)
2317 fprintf(outfile, "get substring %d failed %d\n", i, rc);
2318 else
2319 {
2320 fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc);
2321 pcre_free_substring(substring);
2322 }
2323 }
2324 }
2325
2326 for (getnamesptr = getnames;
2327 *getnamesptr != 0;
2328 getnamesptr += (int)strlen((char*)getnamesptr) + 1)
2329 {
2330 const char *substring;
2331 int rc = pcre_get_named_substring(re, (char *)bptr, use_offsets,
2332 count, (char *)getnamesptr, &substring);
2333 if (rc < 0)
2334 fprintf(outfile, "copy substring %s failed %d\n", getnamesptr, rc);
2335 else
2336 {
2337 fprintf(outfile, " G %s (%d) %s\n", substring, rc, getnamesptr);
2338 pcre_free_substring(substring);
2339 }
2340 }
2341
2342 if (getlist)
2343 {
2344 const char **stringlist;
2345 int rc = pcre_get_substring_list((char *)bptr, use_offsets, count,
2346 &stringlist);
2347 if (rc < 0)
2348 fprintf(outfile, "get substring list failed %d\n", rc);
2349 else
2350 {
2351 for (i = 0; i < count; i++)
2352 fprintf(outfile, "%2dL %s\n", i, stringlist[i]);
2353 if (stringlist[i] != NULL)
2354 fprintf(outfile, "string list not terminated by NULL\n");
2355 /* free((void *)stringlist); */
2356 pcre_free_substring_list(stringlist);
2357 }
2358 }
2359 }
2360
2361 /* There was a partial match */
2362
2363 else if (count == PCRE_ERROR_PARTIAL)
2364 {
2365 fprintf(outfile, "Partial match");
2366 if (use_size_offsets > 1)
2367 {
2368 fprintf(outfile, ": ");
2369 pchars(bptr + use_offsets[0], use_offsets[1] - use_offsets[0],
2370 outfile);
2371 }
2372 fprintf(outfile, "\n");
2373 break; /* Out of the /g loop */
2374 }
2375
2376 /* Failed to match. If this is a /g or /G loop and we previously set
2377 g_notempty after a null match, this is not necessarily the end. We want
2378 to advance the start offset, and continue. We won't be at the end of the
2379 string - that was checked before setting g_notempty.
2380
2381 Complication arises in the case when the newline option is "any" or
2382 "anycrlf". If the previous match was at the end of a line terminated by
2383 CRLF, an advance of one character just passes the \r, whereas we should
2384 prefer the longer newline sequence, as does the code in pcre_exec().
2385 Fudge the offset value to achieve this.
2386
2387 Otherwise, in the case of UTF-8 matching, the advance must be one
2388 character, not one byte. */
2389
2390 else
2391 {
2392 if (g_notempty != 0)
2393 {
2394 int onechar = 1;
2395 unsigned int obits = ((real_pcre *)re)->options;
2396 use_offsets[0] = start_offset;
2397 if ((obits & PCRE_NEWLINE_BITS) == 0)
2398 {
2399 int d;
2400 (void)pcre_config(PCRE_CONFIG_NEWLINE, &d);
2401 /* Note that these values are always the ASCII ones, even in
2402 EBCDIC environments. CR = 13, NL = 10. */
2403 obits = (d == 13)? PCRE_NEWLINE_CR :
2404 (d == 10)? PCRE_NEWLINE_LF :
2405 (d == (13<<8 | 10))? PCRE_NEWLINE_CRLF :
2406 (d == -2)? PCRE_NEWLINE_ANYCRLF :
2407 (d == -1)? PCRE_NEWLINE_ANY : 0;
2408 }
2409 if (((obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANY ||
2410 (obits & PCRE_NEWLINE_BITS) == PCRE_NEWLINE_ANYCRLF)
2411 &&
2412 start_offset < len - 1 &&
2413 bptr[start_offset] == '\r' &&
2414 bptr[start_offset+1] == '\n')
2415 onechar++;
2416 else if (use_utf8)
2417 {
2418 while (start_offset + onechar < len)
2419 {
2420 int tb = bptr[start_offset+onechar];
2421 if (tb <= 127) break;
2422 tb &= 0xc0;
2423 if (tb != 0 && tb != 0xc0) onechar++;
2424 }
2425 }
2426 use_offsets[1] = start_offset + onechar;
2427 }
2428 else
2429 {
2430 if (count == PCRE_ERROR_NOMATCH)
2431 {
2432 if (gmatched == 0) fprintf(outfile, "No match\n");
2433 }
2434 else fprintf(outfile, "Error %d\n", count);
2435 break; /* Out of the /g loop */
2436 }
2437 }
2438
2439 /* If not /g or /G we are done */
2440
2441 if (!do_g && !do_G) break;
2442
2443 /* If we have matched an empty string, first check to see if we are at
2444 the end of the subject. If so, the /g loop is over. Otherwise, mimic
2445 what Perl's /g options does. This turns out to be rather cunning. First
2446 we set PCRE_NOTEMPTY and PCRE_ANCHORED and try the match again at the
2447 same point. If this fails (picked up above) we advance to the next
2448 character. */
2449
2450 g_notempty = 0;
2451
2452 if (use_offsets[0] == use_offsets[1])
2453 {
2454 if (use_offsets[0] == len) break;
2455 g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED;
2456 }
2457
2458 /* For /g, update the start offset, leaving the rest alone */
2459
2460 if (do_g) start_offset = use_offsets[1];
2461
2462 /* For /G, update the pointer and length */
2463
2464 else
2465 {
2466 bptr += use_offsets[1];
2467 len -= use_offsets[1];
2468 }
2469 } /* End of loop for /g and /G */
2470
2471 NEXT_DATA: continue;
2472 } /* End of loop for data lines */
2473
2474 CONTINUE:
2475
2476 #if !defined NOPOSIX
2477 if (posix || do_posix) regfree(&preg);
2478 #endif
2479
2480 if (re != NULL) new_free(re);
2481 if (extra != NULL) new_free(extra);
2482 if (tables != NULL)
2483 {
2484 new_free((void *)tables);
2485 setlocale(LC_CTYPE, "C");
2486 locale_set = 0;
2487 }
2488 }
2489
2490 if (infile == stdin) fprintf(outfile, "\n");
2491
2492 EXIT:
2493
2494 if (infile != NULL && infile != stdin) fclose(infile);
2495 if (outfile != NULL && outfile != stdout) fclose(outfile);
2496
2497 free(buffer);
2498 free(dbuffer);
2499 free(pbuffer);
2500 free(offsets);
2501
2502 return yield;
2503 }
2504
2505 /* End of pcretest.c */

Properties

Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12