4 |
|
|
5 |
/* This program was hacked up as a tester for PCRE. I really should have |
/* 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 |
written it more tidily in the first place. Will I ever learn? It has grown and |
7 |
been extended and consequently is now rather untidy in places. */ |
been extended and consequently is now rather 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 |
#include <ctype.h> |
#include <ctype.h> |
40 |
#include <stdio.h> |
#include <stdio.h> |
42 |
#include <stdlib.h> |
#include <stdlib.h> |
43 |
#include <time.h> |
#include <time.h> |
44 |
#include <locale.h> |
#include <locale.h> |
45 |
|
#include <errno.h> |
46 |
|
|
47 |
/* We need the internal info for displaying the results of pcre_study(). Also |
/* We need the internal info for displaying the results of pcre_study(). Also |
48 |
for getting the opcodes for showing compiled code. */ |
for getting the opcodes for showing compiled code. */ |
66 |
#endif |
#endif |
67 |
#endif |
#endif |
68 |
|
|
69 |
#define LOOPREPEAT 50000 |
#define LOOPREPEAT 500000 |
70 |
|
|
71 |
|
#define BUFFER_SIZE 30000 |
72 |
|
#define PBUFFER_SIZE BUFFER_SIZE |
73 |
|
#define DBUFFER_SIZE BUFFER_SIZE |
74 |
|
|
75 |
|
|
76 |
static FILE *outfile; |
static FILE *outfile; |
80 |
static int callout_fail_count; |
static int callout_fail_count; |
81 |
static int callout_fail_id; |
static int callout_fail_id; |
82 |
static int first_callout; |
static int first_callout; |
83 |
static int utf8; |
static int show_malloc; |
84 |
|
static int use_utf8; |
85 |
static size_t gotten_store; |
static size_t gotten_store; |
86 |
|
|
87 |
|
static uschar *pbuffer = NULL; |
88 |
|
|
89 |
|
|
90 |
static int utf8_table1[] = { |
static const int utf8_table1[] = { |
91 |
0x0000007f, 0x000007ff, 0x0000ffff, 0x001fffff, 0x03ffffff, 0x7fffffff}; |
0x0000007f, 0x000007ff, 0x0000ffff, 0x001fffff, 0x03ffffff, 0x7fffffff}; |
92 |
|
|
93 |
static int utf8_table2[] = { |
static const int utf8_table2[] = { |
94 |
0, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc}; |
0, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc}; |
95 |
|
|
96 |
static int utf8_table3[] = { |
static const int utf8_table3[] = { |
97 |
0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01}; |
0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01}; |
98 |
|
|
99 |
|
|
105 |
/* The code for doing this is held in a separate file that is also included in |
/* The code for doing this is held in a separate file that is also included in |
106 |
pcre.c when it is compiled with the debug switch. It defines a function called |
pcre.c when it is compiled with the debug switch. It defines a function called |
107 |
print_internals(), which uses a table of opcode lengths defined by the macro |
print_internals(), which uses a table of opcode lengths defined by the macro |
108 |
OP_LENGTHS, whose name must be OP_lengths. */ |
OP_LENGTHS, whose name must be OP_lengths. It also uses a table that translates |
109 |
|
Unicode property names to numbers; this is kept in a separate file. */ |
110 |
|
|
111 |
static uschar OP_lengths[] = { OP_LENGTHS }; |
static uschar OP_lengths[] = { OP_LENGTHS }; |
112 |
|
|
113 |
|
#include "ucp.h" |
114 |
|
#include "ucptypetable.c" |
115 |
#include "printint.c" |
#include "printint.c" |
116 |
|
|
117 |
|
|
195 |
-6 to 0 => malformed UTF-8 character at offset = (-return) |
-6 to 0 => malformed UTF-8 character at offset = (-return) |
196 |
*/ |
*/ |
197 |
|
|
198 |
int |
static int |
199 |
utf82ord(unsigned char *buffer, int *vptr) |
utf82ord(unsigned char *buffer, int *vptr) |
200 |
{ |
{ |
201 |
int c = *buffer++; |
int c = *buffer++; |
253 |
|
|
254 |
while (length-- > 0) |
while (length-- > 0) |
255 |
{ |
{ |
256 |
if (utf8) |
if (use_utf8) |
257 |
{ |
{ |
258 |
int rc = utf82ord(p, &c); |
int rc = utf82ord(p, &c); |
259 |
|
|
306 |
static int callout(pcre_callout_block *cb) |
static int callout(pcre_callout_block *cb) |
307 |
{ |
{ |
308 |
FILE *f = (first_callout | callout_extra)? outfile : NULL; |
FILE *f = (first_callout | callout_extra)? outfile : NULL; |
309 |
int i, pre_start, post_start; |
int i, pre_start, post_start, subject_length; |
310 |
|
|
311 |
if (callout_extra) |
if (callout_extra) |
312 |
{ |
{ |
|
int i; |
|
313 |
fprintf(f, "Callout %d: last capture = %d\n", |
fprintf(f, "Callout %d: last capture = %d\n", |
314 |
cb->callout_number, cb->capture_last); |
cb->callout_number, cb->capture_last); |
315 |
|
|
337 |
post_start = pchars((unsigned char *)(cb->subject + cb->start_match), |
post_start = pchars((unsigned char *)(cb->subject + cb->start_match), |
338 |
cb->current_position - cb->start_match, f); |
cb->current_position - cb->start_match, f); |
339 |
|
|
340 |
|
subject_length = pchars((unsigned char *)cb->subject, cb->subject_length, NULL); |
341 |
|
|
342 |
(void)pchars((unsigned char *)(cb->subject + cb->current_position), |
(void)pchars((unsigned char *)(cb->subject + cb->current_position), |
343 |
cb->subject_length - cb->current_position, f); |
cb->subject_length - cb->current_position, f); |
344 |
|
|
345 |
if (f != NULL) fprintf(f, "\n"); |
if (f != NULL) fprintf(f, "\n"); |
346 |
|
|
347 |
/* Always print appropriate indicators, with callout number if not already |
/* Always print appropriate indicators, with callout number if not already |
348 |
shown */ |
shown. For automatic callouts, show the pattern offset. */ |
349 |
|
|
350 |
if (callout_extra) fprintf(outfile, " "); |
if (cb->callout_number == 255) |
351 |
else fprintf(outfile, "%3d ", cb->callout_number); |
{ |
352 |
|
fprintf(outfile, "%+3d ", cb->pattern_position); |
353 |
|
if (cb->pattern_position > 99) fprintf(outfile, "\n "); |
354 |
|
} |
355 |
|
else |
356 |
|
{ |
357 |
|
if (callout_extra) fprintf(outfile, " "); |
358 |
|
else fprintf(outfile, "%3d ", cb->callout_number); |
359 |
|
} |
360 |
|
|
361 |
for (i = 0; i < pre_start; i++) fprintf(outfile, " "); |
for (i = 0; i < pre_start; i++) fprintf(outfile, " "); |
362 |
fprintf(outfile, "^"); |
fprintf(outfile, "^"); |
367 |
fprintf(outfile, "^"); |
fprintf(outfile, "^"); |
368 |
} |
} |
369 |
|
|
370 |
fprintf(outfile, "\n"); |
for (i = 0; i < subject_length - pre_start - post_start + 4; i++) |
371 |
|
fprintf(outfile, " "); |
372 |
|
|
373 |
|
fprintf(outfile, "%.*s", (cb->next_item_length == 0)? 1 : cb->next_item_length, |
374 |
|
pbuffer + cb->pattern_position); |
375 |
|
|
376 |
|
fprintf(outfile, "\n"); |
377 |
first_callout = 0; |
first_callout = 0; |
378 |
|
|
379 |
if ((int)(cb->callout_data) != 0) |
if (cb->callout_data != NULL) |
380 |
{ |
{ |
381 |
fprintf(outfile, "Callout data = %d\n", (int)(cb->callout_data)); |
int callout_data = *((int *)(cb->callout_data)); |
382 |
return (int)(cb->callout_data); |
if (callout_data != 0) |
383 |
|
{ |
384 |
|
fprintf(outfile, "Callout data = %d\n", callout_data); |
385 |
|
return callout_data; |
386 |
|
} |
387 |
} |
} |
388 |
|
|
389 |
return (cb->callout_number != callout_fail_id)? 0 : |
return (cb->callout_number != callout_fail_id)? 0 : |
392 |
|
|
393 |
|
|
394 |
/************************************************* |
/************************************************* |
395 |
* Local malloc function * |
* Local malloc functions * |
396 |
*************************************************/ |
*************************************************/ |
397 |
|
|
398 |
/* Alternative malloc function, to test functionality and show the size of the |
/* Alternative malloc function, to test functionality and show the size of the |
400 |
|
|
401 |
static void *new_malloc(size_t size) |
static void *new_malloc(size_t size) |
402 |
{ |
{ |
403 |
|
void *block = malloc(size); |
404 |
gotten_store = size; |
gotten_store = size; |
405 |
return malloc(size); |
if (show_malloc) |
406 |
|
fprintf(outfile, "malloc %3d %p\n", size, block); |
407 |
|
return block; |
408 |
|
} |
409 |
|
|
410 |
|
static void new_free(void *block) |
411 |
|
{ |
412 |
|
if (show_malloc) |
413 |
|
fprintf(outfile, "free %p\n", block); |
414 |
|
free(block); |
415 |
} |
} |
416 |
|
|
417 |
|
|
418 |
|
/* For recursion malloc/free, to test stacking calls */ |
419 |
|
|
420 |
|
static void *stack_malloc(size_t size) |
421 |
|
{ |
422 |
|
void *block = malloc(size); |
423 |
|
if (show_malloc) |
424 |
|
fprintf(outfile, "stack_malloc %3d %p\n", size, block); |
425 |
|
return block; |
426 |
|
} |
427 |
|
|
428 |
|
static void stack_free(void *block) |
429 |
|
{ |
430 |
|
if (show_malloc) |
431 |
|
fprintf(outfile, "stack_free %p\n", block); |
432 |
|
free(block); |
433 |
|
} |
434 |
|
|
435 |
|
|
436 |
/************************************************* |
/************************************************* |
437 |
* Call pcre_fullinfo() * |
* Call pcre_fullinfo() * |
449 |
|
|
450 |
|
|
451 |
/************************************************* |
/************************************************* |
452 |
|
* Byte flipping function * |
453 |
|
*************************************************/ |
454 |
|
|
455 |
|
static long int |
456 |
|
byteflip(long int value, int n) |
457 |
|
{ |
458 |
|
if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8); |
459 |
|
return ((value & 0x000000ff) << 24) | |
460 |
|
((value & 0x0000ff00) << 8) | |
461 |
|
((value & 0x00ff0000) >> 8) | |
462 |
|
((value & 0xff000000) >> 24); |
463 |
|
} |
464 |
|
|
465 |
|
|
466 |
|
|
467 |
|
|
468 |
|
/************************************************* |
469 |
* Main Program * |
* Main Program * |
470 |
*************************************************/ |
*************************************************/ |
471 |
|
|
490 |
#endif |
#endif |
491 |
int debug = 0; |
int debug = 0; |
492 |
int done = 0; |
int done = 0; |
|
unsigned char buffer[30000]; |
|
|
unsigned char dbuffer[1024]; |
|
493 |
|
|
494 |
/* Static so that new_malloc can use it. */ |
unsigned char *buffer; |
495 |
|
unsigned char *dbuffer; |
496 |
|
|
497 |
|
/* Get buffers from malloc() so that Electric Fence will check their misuse |
498 |
|
when I am debugging. */ |
499 |
|
|
500 |
|
buffer = (unsigned char *)malloc(BUFFER_SIZE); |
501 |
|
dbuffer = (unsigned char *)malloc(DBUFFER_SIZE); |
502 |
|
pbuffer = (unsigned char *)malloc(PBUFFER_SIZE); |
503 |
|
|
504 |
|
/* The outfile variable is static so that new_malloc can use it. The _setmode() |
505 |
|
stuff is some magic that I don't understand, but which apparently does good |
506 |
|
things in Windows. It's related to line terminations. */ |
507 |
|
|
508 |
|
#if defined(_WIN32) || defined(WIN32) |
509 |
|
_setmode( _fileno( stdout ), 0x8000 ); |
510 |
|
#endif /* defined(_WIN32) || defined(WIN32) */ |
511 |
|
|
512 |
outfile = stdout; |
outfile = stdout; |
513 |
|
|
539 |
printf("Compiled with\n"); |
printf("Compiled with\n"); |
540 |
(void)pcre_config(PCRE_CONFIG_UTF8, &rc); |
(void)pcre_config(PCRE_CONFIG_UTF8, &rc); |
541 |
printf(" %sUTF-8 support\n", rc? "" : "No "); |
printf(" %sUTF-8 support\n", rc? "" : "No "); |
542 |
|
(void)pcre_config(PCRE_CONFIG_UNICODE_PROPERTIES, &rc); |
543 |
|
printf(" %sUnicode properties support\n", rc? "" : "No "); |
544 |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
(void)pcre_config(PCRE_CONFIG_NEWLINE, &rc); |
545 |
printf(" Newline character is %s\n", (rc == '\r')? "CR" : "LF"); |
printf(" Newline character is %s\n", (rc == '\r')? "CR" : "LF"); |
546 |
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
(void)pcre_config(PCRE_CONFIG_LINK_SIZE, &rc); |
549 |
printf(" POSIX malloc threshold = %d\n", rc); |
printf(" POSIX malloc threshold = %d\n", rc); |
550 |
(void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc); |
(void)pcre_config(PCRE_CONFIG_MATCH_LIMIT, &rc); |
551 |
printf(" Default match limit = %d\n", rc); |
printf(" Default match limit = %d\n", rc); |
552 |
|
(void)pcre_config(PCRE_CONFIG_STACKRECURSE, &rc); |
553 |
|
printf(" Match recursion uses %s\n", rc? "stack" : "heap"); |
554 |
exit(0); |
exit(0); |
555 |
} |
} |
556 |
else |
else |
560 |
printf(" -C show PCRE compile-time options and exit\n"); |
printf(" -C show PCRE compile-time options and exit\n"); |
561 |
printf(" -d debug: show compiled code; implies -i\n" |
printf(" -d debug: show compiled code; implies -i\n" |
562 |
" -i show information about compiled pattern\n" |
" -i show information about compiled pattern\n" |
563 |
|
" -m output memory used information\n" |
564 |
" -o <n> set size of offsets vector to <n>\n"); |
" -o <n> set size of offsets vector to <n>\n"); |
565 |
#if !defined NOPOSIX |
#if !defined NOPOSIX |
566 |
printf(" -p use POSIX interface\n"); |
printf(" -p use POSIX interface\n"); |
567 |
#endif |
#endif |
568 |
printf(" -s output store information\n" |
printf(" -s output store (memory) used information\n" |
569 |
" -t time compilation and execution\n"); |
" -t time compilation and execution\n"); |
570 |
return 1; |
return 1; |
571 |
} |
} |
576 |
/* Get the store for the offsets vector, and remember what it was */ |
/* Get the store for the offsets vector, and remember what it was */ |
577 |
|
|
578 |
size_offsets_max = size_offsets; |
size_offsets_max = size_offsets; |
579 |
offsets = malloc(size_offsets_max * sizeof(int)); |
offsets = (int *)malloc(size_offsets_max * sizeof(int)); |
580 |
if (offsets == NULL) |
if (offsets == NULL) |
581 |
{ |
{ |
582 |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
588 |
|
|
589 |
if (argc > 1) |
if (argc > 1) |
590 |
{ |
{ |
591 |
infile = fopen(argv[op], "r"); |
infile = fopen(argv[op], "rb"); |
592 |
if (infile == NULL) |
if (infile == NULL) |
593 |
{ |
{ |
594 |
printf("** Failed to open %s\n", argv[op]); |
printf("** Failed to open %s\n", argv[op]); |
598 |
|
|
599 |
if (argc > 2) |
if (argc > 2) |
600 |
{ |
{ |
601 |
outfile = fopen(argv[op+1], "w"); |
outfile = fopen(argv[op+1], "wb"); |
602 |
if (outfile == NULL) |
if (outfile == NULL) |
603 |
{ |
{ |
604 |
printf("** Failed to open %s\n", argv[op+1]); |
printf("** Failed to open %s\n", argv[op+1]); |
609 |
/* Set alternative malloc function */ |
/* Set alternative malloc function */ |
610 |
|
|
611 |
pcre_malloc = new_malloc; |
pcre_malloc = new_malloc; |
612 |
|
pcre_free = new_free; |
613 |
|
pcre_stack_malloc = stack_malloc; |
614 |
|
pcre_stack_free = stack_free; |
615 |
|
|
616 |
/* Heading line, then prompt for first regex if stdin */ |
/* Heading line, then prompt for first regex if stdin */ |
617 |
|
|
631 |
|
|
632 |
const char *error; |
const char *error; |
633 |
unsigned char *p, *pp, *ppp; |
unsigned char *p, *pp, *ppp; |
634 |
|
unsigned char *to_file = NULL; |
635 |
const unsigned char *tables = NULL; |
const unsigned char *tables = NULL; |
636 |
|
unsigned long int true_size, true_study_size = 0; |
637 |
|
size_t size, regex_gotten_store; |
638 |
int do_study = 0; |
int do_study = 0; |
639 |
int do_debug = debug; |
int do_debug = debug; |
640 |
int do_G = 0; |
int do_G = 0; |
641 |
int do_g = 0; |
int do_g = 0; |
642 |
int do_showinfo = showinfo; |
int do_showinfo = showinfo; |
643 |
int do_showrest = 0; |
int do_showrest = 0; |
644 |
|
int do_flip = 0; |
645 |
int erroroffset, len, delimiter; |
int erroroffset, len, delimiter; |
646 |
|
|
647 |
utf8 = 0; |
use_utf8 = 0; |
648 |
|
|
649 |
if (infile == stdin) printf(" re> "); |
if (infile == stdin) printf(" re> "); |
650 |
if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) break; |
if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) break; |
651 |
if (infile != stdin) fprintf(outfile, "%s", (char *)buffer); |
if (infile != stdin) fprintf(outfile, "%s", (char *)buffer); |
652 |
fflush(outfile); |
fflush(outfile); |
653 |
|
|
655 |
while (isspace(*p)) p++; |
while (isspace(*p)) p++; |
656 |
if (*p == 0) continue; |
if (*p == 0) continue; |
657 |
|
|
658 |
/* Get the delimiter and seek the end of the pattern; if is isn't |
/* See if the pattern is to be loaded pre-compiled from a file. */ |
659 |
complete, read more. */ |
|
660 |
|
if (*p == '<' && strchr((char *)(p+1), '<') == NULL) |
661 |
|
{ |
662 |
|
unsigned long int magic; |
663 |
|
uschar sbuf[8]; |
664 |
|
FILE *f; |
665 |
|
|
666 |
|
p++; |
667 |
|
pp = p + (int)strlen((char *)p); |
668 |
|
while (isspace(pp[-1])) pp--; |
669 |
|
*pp = 0; |
670 |
|
|
671 |
|
f = fopen((char *)p, "rb"); |
672 |
|
if (f == NULL) |
673 |
|
{ |
674 |
|
fprintf(outfile, "Failed to open %s: %s\n", p, strerror(errno)); |
675 |
|
continue; |
676 |
|
} |
677 |
|
|
678 |
|
if (fread(sbuf, 1, 8, f) != 8) goto FAIL_READ; |
679 |
|
|
680 |
|
true_size = |
681 |
|
(sbuf[0] << 24) | (sbuf[1] << 16) | (sbuf[2] << 8) | sbuf[3]; |
682 |
|
true_study_size = |
683 |
|
(sbuf[4] << 24) | (sbuf[5] << 16) | (sbuf[6] << 8) | sbuf[7]; |
684 |
|
|
685 |
|
re = (real_pcre *)new_malloc(true_size); |
686 |
|
regex_gotten_store = gotten_store; |
687 |
|
|
688 |
|
if (fread(re, 1, true_size, f) != true_size) goto FAIL_READ; |
689 |
|
|
690 |
|
magic = ((real_pcre *)re)->magic_number; |
691 |
|
if (magic != MAGIC_NUMBER) |
692 |
|
{ |
693 |
|
if (byteflip(magic, sizeof(magic)) == MAGIC_NUMBER) |
694 |
|
{ |
695 |
|
do_flip = 1; |
696 |
|
} |
697 |
|
else |
698 |
|
{ |
699 |
|
fprintf(outfile, "Data in %s is not a compiled PCRE regex\n", p); |
700 |
|
fclose(f); |
701 |
|
continue; |
702 |
|
} |
703 |
|
} |
704 |
|
|
705 |
|
fprintf(outfile, "Compiled regex%s loaded from %s\n", |
706 |
|
do_flip? " (byte-inverted)" : "", p); |
707 |
|
|
708 |
|
/* Need to know if UTF-8 for printing data strings */ |
709 |
|
|
710 |
|
new_info(re, NULL, PCRE_INFO_OPTIONS, &options); |
711 |
|
use_utf8 = (options & PCRE_UTF8) != 0; |
712 |
|
|
713 |
|
/* Now see if there is any following study data */ |
714 |
|
|
715 |
|
if (true_study_size != 0) |
716 |
|
{ |
717 |
|
pcre_study_data *psd; |
718 |
|
|
719 |
|
extra = (pcre_extra *)new_malloc(sizeof(pcre_extra) + true_study_size); |
720 |
|
extra->flags = PCRE_EXTRA_STUDY_DATA; |
721 |
|
|
722 |
|
psd = (pcre_study_data *)(((char *)extra) + sizeof(pcre_extra)); |
723 |
|
extra->study_data = psd; |
724 |
|
|
725 |
|
if (fread(psd, 1, true_study_size, f) != true_study_size) |
726 |
|
{ |
727 |
|
FAIL_READ: |
728 |
|
fprintf(outfile, "Failed to read data from %s\n", p); |
729 |
|
if (extra != NULL) new_free(extra); |
730 |
|
if (re != NULL) new_free(re); |
731 |
|
fclose(f); |
732 |
|
continue; |
733 |
|
} |
734 |
|
fprintf(outfile, "Study data loaded from %s\n", p); |
735 |
|
do_study = 1; /* To get the data output if requested */ |
736 |
|
} |
737 |
|
else fprintf(outfile, "No study data\n"); |
738 |
|
|
739 |
|
fclose(f); |
740 |
|
goto SHOW_INFO; |
741 |
|
} |
742 |
|
|
743 |
|
/* In-line pattern (the usual case). Get the delimiter and seek the end of |
744 |
|
the pattern; if is isn't complete, read more. */ |
745 |
|
|
746 |
delimiter = *p++; |
delimiter = *p++; |
747 |
|
|
763 |
} |
} |
764 |
if (*pp != 0) break; |
if (*pp != 0) break; |
765 |
|
|
766 |
len = sizeof(buffer) - (pp - buffer); |
len = BUFFER_SIZE - (pp - buffer); |
767 |
if (len < 256) |
if (len < 256) |
768 |
{ |
{ |
769 |
fprintf(outfile, "** Expression too long - missing delimiter?\n"); |
fprintf(outfile, "** Expression too long - missing delimiter?\n"); |
786 |
|
|
787 |
if (pp[1] == '\\') *pp++ = '\\'; |
if (pp[1] == '\\') *pp++ = '\\'; |
788 |
|
|
789 |
/* Terminate the pattern at the delimiter */ |
/* Terminate the pattern at the delimiter, and save a copy of the pattern |
790 |
|
for callouts. */ |
791 |
|
|
792 |
*pp++ = 0; |
*pp++ = 0; |
793 |
|
strcpy((char *)pbuffer, (char *)p); |
794 |
|
|
795 |
/* Look for options after final delimiter */ |
/* Look for options after final delimiter */ |
796 |
|
|
810 |
|
|
811 |
case '+': do_showrest = 1; break; |
case '+': do_showrest = 1; break; |
812 |
case 'A': options |= PCRE_ANCHORED; break; |
case 'A': options |= PCRE_ANCHORED; break; |
813 |
|
case 'C': options |= PCRE_AUTO_CALLOUT; break; |
814 |
case 'D': do_debug = do_showinfo = 1; break; |
case 'D': do_debug = do_showinfo = 1; break; |
815 |
case 'E': options |= PCRE_DOLLAR_ENDONLY; break; |
case 'E': options |= PCRE_DOLLAR_ENDONLY; break; |
816 |
|
case 'F': do_flip = 1; break; |
817 |
case 'G': do_G = 1; break; |
case 'G': do_G = 1; break; |
818 |
case 'I': do_showinfo = 1; break; |
case 'I': do_showinfo = 1; break; |
819 |
case 'M': log_store = 1; break; |
case 'M': log_store = 1; break; |
826 |
case 'S': do_study = 1; break; |
case 'S': do_study = 1; break; |
827 |
case 'U': options |= PCRE_UNGREEDY; break; |
case 'U': options |= PCRE_UNGREEDY; break; |
828 |
case 'X': options |= PCRE_EXTRA; break; |
case 'X': options |= PCRE_EXTRA; break; |
829 |
case '8': options |= PCRE_UTF8; utf8 = 1; break; |
case '8': options |= PCRE_UTF8; use_utf8 = 1; break; |
830 |
|
case '?': options |= PCRE_NO_UTF8_CHECK; break; |
831 |
|
|
832 |
case 'L': |
case 'L': |
833 |
ppp = pp; |
ppp = pp; |
842 |
pp = ppp; |
pp = ppp; |
843 |
break; |
break; |
844 |
|
|
845 |
|
case '>': |
846 |
|
to_file = pp; |
847 |
|
while (*pp != 0) pp++; |
848 |
|
while (isspace(pp[-1])) pp--; |
849 |
|
*pp = 0; |
850 |
|
break; |
851 |
|
|
852 |
case '\n': case ' ': break; |
case '\n': case ' ': break; |
853 |
|
|
854 |
default: |
default: |
855 |
fprintf(outfile, "** Unknown option '%c'\n", pp[-1]); |
fprintf(outfile, "** Unknown option '%c'\n", pp[-1]); |
856 |
goto SKIP_DATA; |
goto SKIP_DATA; |
866 |
{ |
{ |
867 |
int rc; |
int rc; |
868 |
int cflags = 0; |
int cflags = 0; |
869 |
|
|
870 |
if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE; |
if ((options & PCRE_CASELESS) != 0) cflags |= REG_ICASE; |
871 |
if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE; |
if ((options & PCRE_MULTILINE) != 0) cflags |= REG_NEWLINE; |
872 |
rc = regcomp(&preg, (char *)p, cflags); |
rc = regcomp(&preg, (char *)p, cflags); |
876 |
|
|
877 |
if (rc != 0) |
if (rc != 0) |
878 |
{ |
{ |
879 |
(void)regerror(rc, &preg, (char *)buffer, sizeof(buffer)); |
(void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE); |
880 |
fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer); |
fprintf(outfile, "Failed: POSIX code %d: %s\n", rc, buffer); |
881 |
goto SKIP_DATA; |
goto SKIP_DATA; |
882 |
} |
} |
917 |
{ |
{ |
918 |
for (;;) |
for (;;) |
919 |
{ |
{ |
920 |
if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) |
if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) |
921 |
{ |
{ |
922 |
done = 1; |
done = 1; |
923 |
goto CONTINUE; |
goto CONTINUE; |
941 |
sizeof(real_pcre) - |
sizeof(real_pcre) - |
942 |
((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size)); |
((real_pcre *)re)->name_count * ((real_pcre *)re)->name_entry_size)); |
943 |
|
|
944 |
|
/* Extract the size for possible writing before possibly flipping it, |
945 |
|
and remember the store that was got. */ |
946 |
|
|
947 |
|
true_size = ((real_pcre *)re)->size; |
948 |
|
regex_gotten_store = gotten_store; |
949 |
|
|
950 |
|
/* If /S was present, study the regexp to generate additional info to |
951 |
|
help with the matching. */ |
952 |
|
|
953 |
|
if (do_study) |
954 |
|
{ |
955 |
|
if (timeit) |
956 |
|
{ |
957 |
|
register int i; |
958 |
|
clock_t time_taken; |
959 |
|
clock_t start_time = clock(); |
960 |
|
for (i = 0; i < LOOPREPEAT; i++) |
961 |
|
extra = pcre_study(re, study_options, &error); |
962 |
|
time_taken = clock() - start_time; |
963 |
|
if (extra != NULL) free(extra); |
964 |
|
fprintf(outfile, " Study time %.3f milliseconds\n", |
965 |
|
(((double)time_taken * 1000.0) / (double)LOOPREPEAT) / |
966 |
|
(double)CLOCKS_PER_SEC); |
967 |
|
} |
968 |
|
extra = pcre_study(re, study_options, &error); |
969 |
|
if (error != NULL) |
970 |
|
fprintf(outfile, "Failed to study: %s\n", error); |
971 |
|
else if (extra != NULL) |
972 |
|
true_study_size = ((pcre_study_data *)(extra->study_data))->size; |
973 |
|
} |
974 |
|
|
975 |
|
/* If the 'F' option was present, we flip the bytes of all the integer |
976 |
|
fields in the regex data block and the study block. This is to make it |
977 |
|
possible to test PCRE's handling of byte-flipped patterns, e.g. those |
978 |
|
compiled on a different architecture. */ |
979 |
|
|
980 |
|
if (do_flip) |
981 |
|
{ |
982 |
|
real_pcre *rre = (real_pcre *)re; |
983 |
|
rre->magic_number = byteflip(rre->magic_number, sizeof(rre->magic_number)); |
984 |
|
rre->size = byteflip(rre->size, sizeof(rre->size)); |
985 |
|
rre->options = byteflip(rre->options, sizeof(rre->options)); |
986 |
|
rre->top_bracket = byteflip(rre->top_bracket, sizeof(rre->top_bracket)); |
987 |
|
rre->top_backref = byteflip(rre->top_backref, sizeof(rre->top_backref)); |
988 |
|
rre->first_byte = byteflip(rre->first_byte, sizeof(rre->first_byte)); |
989 |
|
rre->req_byte = byteflip(rre->req_byte, sizeof(rre->req_byte)); |
990 |
|
rre->name_table_offset = byteflip(rre->name_table_offset, |
991 |
|
sizeof(rre->name_table_offset)); |
992 |
|
rre->name_entry_size = byteflip(rre->name_entry_size, |
993 |
|
sizeof(rre->name_entry_size)); |
994 |
|
rre->name_count = byteflip(rre->name_count, sizeof(rre->name_count)); |
995 |
|
|
996 |
|
if (extra != NULL) |
997 |
|
{ |
998 |
|
pcre_study_data *rsd = (pcre_study_data *)(extra->study_data); |
999 |
|
rsd->size = byteflip(rsd->size, sizeof(rsd->size)); |
1000 |
|
rsd->options = byteflip(rsd->options, sizeof(rsd->options)); |
1001 |
|
} |
1002 |
|
} |
1003 |
|
|
1004 |
|
/* Extract information from the compiled data if required */ |
1005 |
|
|
1006 |
|
SHOW_INFO: |
1007 |
|
|
1008 |
if (do_showinfo) |
if (do_showinfo) |
1009 |
{ |
{ |
1010 |
unsigned long int get_options; |
unsigned long int get_options, all_options; |
1011 |
int old_first_char, old_options, old_count; |
int old_first_char, old_options, old_count; |
1012 |
int count, backrefmax, first_char, need_char; |
int count, backrefmax, first_char, need_char; |
1013 |
int nameentrysize, namecount; |
int nameentrysize, namecount; |
1014 |
const uschar *nametable; |
const uschar *nametable; |
|
size_t size; |
|
1015 |
|
|
1016 |
if (do_debug) |
if (do_debug) |
1017 |
{ |
{ |
1027 |
new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char); |
new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char); |
1028 |
new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize); |
new_info(re, NULL, PCRE_INFO_NAMEENTRYSIZE, &nameentrysize); |
1029 |
new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount); |
new_info(re, NULL, PCRE_INFO_NAMECOUNT, &namecount); |
1030 |
new_info(re, NULL, PCRE_INFO_NAMETABLE, &nametable); |
new_info(re, NULL, PCRE_INFO_NAMETABLE, (void *)&nametable); |
1031 |
|
|
1032 |
old_count = pcre_info(re, &old_options, &old_first_char); |
old_count = pcre_info(re, &old_options, &old_first_char); |
1033 |
if (count < 0) fprintf(outfile, |
if (count < 0) fprintf(outfile, |
1047 |
get_options, old_options); |
get_options, old_options); |
1048 |
} |
} |
1049 |
|
|
1050 |
if (size != gotten_store) fprintf(outfile, |
if (size != regex_gotten_store) fprintf(outfile, |
1051 |
"Size disagreement: pcre_fullinfo=%d call to malloc for %d\n", |
"Size disagreement: pcre_fullinfo=%d call to malloc for %d\n", |
1052 |
size, gotten_store); |
size, regex_gotten_store); |
1053 |
|
|
1054 |
fprintf(outfile, "Capturing subpattern count = %d\n", count); |
fprintf(outfile, "Capturing subpattern count = %d\n", count); |
1055 |
if (backrefmax > 0) |
if (backrefmax > 0) |
1067 |
} |
} |
1068 |
} |
} |
1069 |
|
|
1070 |
|
/* The NOPARTIAL bit is a private bit in the options, so we have |
1071 |
|
to fish it out via out back door */ |
1072 |
|
|
1073 |
|
all_options = ((real_pcre *)re)->options; |
1074 |
|
if (do_flip) |
1075 |
|
{ |
1076 |
|
all_options = byteflip(all_options, sizeof(all_options)); |
1077 |
|
} |
1078 |
|
|
1079 |
|
if ((all_options & PCRE_NOPARTIAL) != 0) |
1080 |
|
fprintf(outfile, "Partial matching not supported\n"); |
1081 |
|
|
1082 |
if (get_options == 0) fprintf(outfile, "No options\n"); |
if (get_options == 0) fprintf(outfile, "No options\n"); |
1083 |
else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s\n", |
else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s%s\n", |
1084 |
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
1085 |
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
1086 |
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
1089 |
((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
1090 |
((get_options & PCRE_EXTRA) != 0)? " extra" : "", |
((get_options & PCRE_EXTRA) != 0)? " extra" : "", |
1091 |
((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "", |
((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "", |
1092 |
((get_options & PCRE_UTF8) != 0)? " utf8" : ""); |
((get_options & PCRE_UTF8) != 0)? " utf8" : "", |
1093 |
|
((get_options & PCRE_NO_UTF8_CHECK) != 0)? " no_utf8_check" : ""); |
1094 |
|
|
1095 |
if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0) |
if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0) |
1096 |
fprintf(outfile, "Case state changes\n"); |
fprintf(outfile, "Case state changes\n"); |
1106 |
else |
else |
1107 |
{ |
{ |
1108 |
int ch = first_char & 255; |
int ch = first_char & 255; |
1109 |
char *caseless = ((first_char & REQ_CASELESS) == 0)? |
const char *caseless = ((first_char & REQ_CASELESS) == 0)? |
1110 |
"" : " (caseless)"; |
"" : " (caseless)"; |
1111 |
if (isprint(ch)) |
if (isprint(ch)) |
1112 |
fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless); |
fprintf(outfile, "First char = \'%c\'%s\n", ch, caseless); |
1121 |
else |
else |
1122 |
{ |
{ |
1123 |
int ch = need_char & 255; |
int ch = need_char & 255; |
1124 |
char *caseless = ((need_char & REQ_CASELESS) == 0)? |
const char *caseless = ((need_char & REQ_CASELESS) == 0)? |
1125 |
"" : " (caseless)"; |
"" : " (caseless)"; |
1126 |
if (isprint(ch)) |
if (isprint(ch)) |
1127 |
fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless); |
fprintf(outfile, "Need char = \'%c\'%s\n", ch, caseless); |
1128 |
else |
else |
1129 |
fprintf(outfile, "Need char = %d%s\n", ch, caseless); |
fprintf(outfile, "Need char = %d%s\n", ch, caseless); |
1130 |
} |
} |
|
} |
|
1131 |
|
|
1132 |
/* If /S was present, study the regexp to generate additional info to |
/* Don't output study size; at present it is in any case a fixed |
1133 |
help with the matching. */ |
value, but it varies, depending on the computer architecture, and |
1134 |
|
so messes up the test suite. (And with the /F option, it might be |
1135 |
|
flipped.) */ |
1136 |
|
|
1137 |
if (do_study) |
if (do_study) |
|
{ |
|
|
if (timeit) |
|
1138 |
{ |
{ |
1139 |
register int i; |
if (extra == NULL) |
1140 |
clock_t time_taken; |
fprintf(outfile, "Study returned NULL\n"); |
1141 |
clock_t start_time = clock(); |
else |
1142 |
for (i = 0; i < LOOPREPEAT; i++) |
{ |
1143 |
extra = pcre_study(re, study_options, &error); |
uschar *start_bits = NULL; |
1144 |
time_taken = clock() - start_time; |
new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits); |
1145 |
if (extra != NULL) free(extra); |
|
1146 |
fprintf(outfile, " Study time %.3f milliseconds\n", |
if (start_bits == NULL) |
1147 |
(((double)time_taken * 1000.0) / (double)LOOPREPEAT) / |
fprintf(outfile, "No starting byte set\n"); |
1148 |
(double)CLOCKS_PER_SEC); |
else |
1149 |
|
{ |
1150 |
|
int i; |
1151 |
|
int c = 24; |
1152 |
|
fprintf(outfile, "Starting byte set: "); |
1153 |
|
for (i = 0; i < 256; i++) |
1154 |
|
{ |
1155 |
|
if ((start_bits[i/8] & (1<<(i&7))) != 0) |
1156 |
|
{ |
1157 |
|
if (c > 75) |
1158 |
|
{ |
1159 |
|
fprintf(outfile, "\n "); |
1160 |
|
c = 2; |
1161 |
|
} |
1162 |
|
if (isprint(i) && i != ' ') |
1163 |
|
{ |
1164 |
|
fprintf(outfile, "%c ", i); |
1165 |
|
c += 2; |
1166 |
|
} |
1167 |
|
else |
1168 |
|
{ |
1169 |
|
fprintf(outfile, "\\x%02x ", i); |
1170 |
|
c += 5; |
1171 |
|
} |
1172 |
|
} |
1173 |
|
} |
1174 |
|
fprintf(outfile, "\n"); |
1175 |
|
} |
1176 |
|
} |
1177 |
} |
} |
1178 |
|
} |
1179 |
|
|
1180 |
extra = pcre_study(re, study_options, &error); |
/* If the '>' option was present, we write out the regex to a file, and |
1181 |
if (error != NULL) |
that is all. The first 8 bytes of the file are the regex length and then |
1182 |
fprintf(outfile, "Failed to study: %s\n", error); |
the study length, in big-endian order. */ |
|
else if (extra == NULL) |
|
|
fprintf(outfile, "Study returned NULL\n"); |
|
1183 |
|
|
1184 |
else if (do_showinfo) |
if (to_file != NULL) |
1185 |
|
{ |
1186 |
|
FILE *f = fopen((char *)to_file, "wb"); |
1187 |
|
if (f == NULL) |
1188 |
{ |
{ |
1189 |
size_t size; |
fprintf(outfile, "Unable to open %s: %s\n", to_file, strerror(errno)); |
1190 |
uschar *start_bits = NULL; |
} |
1191 |
new_info(re, extra, PCRE_INFO_STUDYSIZE, &size); |
else |
1192 |
new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits); |
{ |
1193 |
fprintf(outfile, "Study size = %d\n", size); |
uschar sbuf[8]; |
1194 |
if (start_bits == NULL) |
sbuf[0] = (true_size >> 24) & 255; |
1195 |
fprintf(outfile, "No starting character set\n"); |
sbuf[1] = (true_size >> 16) & 255; |
1196 |
|
sbuf[2] = (true_size >> 8) & 255; |
1197 |
|
sbuf[3] = (true_size) & 255; |
1198 |
|
|
1199 |
|
sbuf[4] = (true_study_size >> 24) & 255; |
1200 |
|
sbuf[5] = (true_study_size >> 16) & 255; |
1201 |
|
sbuf[6] = (true_study_size >> 8) & 255; |
1202 |
|
sbuf[7] = (true_study_size) & 255; |
1203 |
|
|
1204 |
|
if (fwrite(sbuf, 1, 8, f) < 8 || |
1205 |
|
fwrite(re, 1, true_size, f) < true_size) |
1206 |
|
{ |
1207 |
|
fprintf(outfile, "Write error on %s: %s\n", to_file, strerror(errno)); |
1208 |
|
} |
1209 |
else |
else |
1210 |
{ |
{ |
1211 |
int i; |
fprintf(outfile, "Compiled regex written to %s\n", to_file); |
1212 |
int c = 24; |
if (extra != NULL) |
|
fprintf(outfile, "Starting character set: "); |
|
|
for (i = 0; i < 256; i++) |
|
1213 |
{ |
{ |
1214 |
if ((start_bits[i/8] & (1<<(i%8))) != 0) |
if (fwrite(extra->study_data, 1, true_study_size, f) < |
1215 |
|
true_study_size) |
1216 |
{ |
{ |
1217 |
if (c > 75) |
fprintf(outfile, "Write error on %s: %s\n", to_file, |
1218 |
{ |
strerror(errno)); |
|
fprintf(outfile, "\n "); |
|
|
c = 2; |
|
|
} |
|
|
if (isprint(i) && i != ' ') |
|
|
{ |
|
|
fprintf(outfile, "%c ", i); |
|
|
c += 2; |
|
|
} |
|
|
else |
|
|
{ |
|
|
fprintf(outfile, "\\x%02x ", i); |
|
|
c += 5; |
|
|
} |
|
1219 |
} |
} |
1220 |
|
else fprintf(outfile, "Study data written to %s\n", to_file); |
1221 |
} |
} |
|
fprintf(outfile, "\n"); |
|
1222 |
} |
} |
1223 |
|
fclose(f); |
1224 |
} |
} |
1225 |
|
continue; /* With next regex */ |
1226 |
} |
} |
1227 |
} |
} /* End of non-POSIX compile */ |
1228 |
|
|
1229 |
/* Read data lines and test them */ |
/* Read data lines and test them */ |
1230 |
|
|
1253 |
callout_count = 0; |
callout_count = 0; |
1254 |
callout_fail_count = 999999; |
callout_fail_count = 999999; |
1255 |
callout_fail_id = -1; |
callout_fail_id = -1; |
1256 |
|
show_malloc = 0; |
1257 |
|
|
1258 |
if (infile == stdin) printf("data> "); |
if (infile == stdin) printf("data> "); |
1259 |
if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) |
if (fgets((char *)buffer, BUFFER_SIZE, infile) == NULL) |
1260 |
{ |
{ |
1261 |
done = 1; |
done = 1; |
1262 |
goto CONTINUE; |
goto CONTINUE; |
1307 |
c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W'); |
c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W'); |
1308 |
if (*pt == '}') |
if (*pt == '}') |
1309 |
{ |
{ |
1310 |
unsigned char buffer[8]; |
unsigned char buff8[8]; |
1311 |
int ii, utn; |
int ii, utn; |
1312 |
utn = ord2utf8(c, buffer); |
utn = ord2utf8(c, buff8); |
1313 |
for (ii = 0; ii < utn - 1; ii++) *q++ = buffer[ii]; |
for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii]; |
1314 |
c = buffer[ii]; /* Last byte */ |
c = buff8[ii]; /* Last byte */ |
1315 |
p = pt + 1; |
p = pt + 1; |
1316 |
break; |
break; |
1317 |
} |
} |
1328 |
} |
} |
1329 |
break; |
break; |
1330 |
|
|
1331 |
case 0: /* Allows for an empty line */ |
case 0: /* \ followed by EOF allows for an empty line */ |
1332 |
p--; |
p--; |
1333 |
continue; |
continue; |
1334 |
|
|
1335 |
|
case '>': |
1336 |
|
while(isdigit(*p)) start_offset = start_offset * 10 + *p++ - '0'; |
1337 |
|
continue; |
1338 |
|
|
1339 |
case 'A': /* Option setting */ |
case 'A': /* Option setting */ |
1340 |
options |= PCRE_ANCHORED; |
options |= PCRE_ANCHORED; |
1341 |
continue; |
continue; |
1353 |
else if (isalnum(*p)) |
else if (isalnum(*p)) |
1354 |
{ |
{ |
1355 |
uschar name[256]; |
uschar name[256]; |
1356 |
uschar *pp = name; |
uschar *npp = name; |
1357 |
while (isalnum(*p)) *pp++ = *p++; |
while (isalnum(*p)) *npp++ = *p++; |
1358 |
*pp = 0; |
*npp = 0; |
1359 |
n = pcre_get_stringnumber(re, (char *)name); |
n = pcre_get_stringnumber(re, (char *)name); |
1360 |
if (n < 0) |
if (n < 0) |
1361 |
fprintf(outfile, "no parentheses with name \"%s\"\n", name); |
fprintf(outfile, "no parentheses with name \"%s\"\n", name); |
1406 |
else if (isalnum(*p)) |
else if (isalnum(*p)) |
1407 |
{ |
{ |
1408 |
uschar name[256]; |
uschar name[256]; |
1409 |
uschar *pp = name; |
uschar *npp = name; |
1410 |
while (isalnum(*p)) *pp++ = *p++; |
while (isalnum(*p)) *npp++ = *p++; |
1411 |
*pp = 0; |
*npp = 0; |
1412 |
n = pcre_get_stringnumber(re, (char *)name); |
n = pcre_get_stringnumber(re, (char *)name); |
1413 |
if (n < 0) |
if (n < 0) |
1414 |
fprintf(outfile, "no parentheses with name \"%s\"\n", name); |
fprintf(outfile, "no parentheses with name \"%s\"\n", name); |
1434 |
{ |
{ |
1435 |
size_offsets_max = n; |
size_offsets_max = n; |
1436 |
free(offsets); |
free(offsets); |
1437 |
use_offsets = offsets = malloc(size_offsets_max * sizeof(int)); |
use_offsets = offsets = (int *)malloc(size_offsets_max * sizeof(int)); |
1438 |
if (offsets == NULL) |
if (offsets == NULL) |
1439 |
{ |
{ |
1440 |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
printf("** Failed to get %d bytes of memory for offsets vector\n", |
1446 |
if (n == 0) use_offsets = NULL; /* Ensures it can't write to it */ |
if (n == 0) use_offsets = NULL; /* Ensures it can't write to it */ |
1447 |
continue; |
continue; |
1448 |
|
|
1449 |
|
case 'P': |
1450 |
|
options |= PCRE_PARTIAL; |
1451 |
|
continue; |
1452 |
|
|
1453 |
|
case 'S': |
1454 |
|
show_malloc = 1; |
1455 |
|
continue; |
1456 |
|
|
1457 |
case 'Z': |
case 'Z': |
1458 |
options |= PCRE_NOTEOL; |
options |= PCRE_NOTEOL; |
1459 |
continue; |
continue; |
1460 |
|
|
1461 |
|
case '?': |
1462 |
|
options |= PCRE_NO_UTF8_CHECK; |
1463 |
|
continue; |
1464 |
} |
} |
1465 |
*q++ = c; |
*q++ = c; |
1466 |
} |
} |
1477 |
int eflags = 0; |
int eflags = 0; |
1478 |
regmatch_t *pmatch = NULL; |
regmatch_t *pmatch = NULL; |
1479 |
if (use_size_offsets > 0) |
if (use_size_offsets > 0) |
1480 |
pmatch = malloc(sizeof(regmatch_t) * use_size_offsets); |
pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * use_size_offsets); |
1481 |
if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL; |
if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL; |
1482 |
if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL; |
if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL; |
1483 |
|
|
1485 |
|
|
1486 |
if (rc != 0) |
if (rc != 0) |
1487 |
{ |
{ |
1488 |
(void)regerror(rc, &preg, (char *)buffer, sizeof(buffer)); |
(void)regerror(rc, &preg, (char *)buffer, BUFFER_SIZE); |
1489 |
fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer); |
fprintf(outfile, "No match: POSIX code %d: %s\n", rc, buffer); |
1490 |
} |
} |
1491 |
else |
else |
1544 |
|
|
1545 |
if (extra == NULL) |
if (extra == NULL) |
1546 |
{ |
{ |
1547 |
extra = malloc(sizeof(pcre_extra)); |
extra = (pcre_extra *)malloc(sizeof(pcre_extra)); |
1548 |
extra->flags = 0; |
extra->flags = 0; |
1549 |
} |
} |
1550 |
extra->flags |= PCRE_EXTRA_MATCH_LIMIT; |
extra->flags |= PCRE_EXTRA_MATCH_LIMIT; |
1560 |
min = mid; |
min = mid; |
1561 |
mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2; |
mid = (mid == max - 1)? max : (max > 0)? (min + max)/2 : mid*2; |
1562 |
} |
} |
1563 |
else if (count >= 0 || count == PCRE_ERROR_NOMATCH) |
else if (count >= 0 || count == PCRE_ERROR_NOMATCH || |
1564 |
|
count == PCRE_ERROR_PARTIAL) |
1565 |
{ |
{ |
1566 |
if (mid == min + 1) |
if (mid == min + 1) |
1567 |
{ |
{ |
1584 |
{ |
{ |
1585 |
if (extra == NULL) |
if (extra == NULL) |
1586 |
{ |
{ |
1587 |
extra = malloc(sizeof(pcre_extra)); |
extra = (pcre_extra *)malloc(sizeof(pcre_extra)); |
1588 |
extra->flags = 0; |
extra->flags = 0; |
1589 |
} |
} |
1590 |
extra->flags |= PCRE_EXTRA_CALLOUT_DATA; |
extra->flags |= PCRE_EXTRA_CALLOUT_DATA; |
1591 |
extra->callout_data = (void *)callout_data; |
extra->callout_data = &callout_data; |
1592 |
count = pcre_exec(re, extra, (char *)bptr, len, start_offset, |
count = pcre_exec(re, extra, (char *)bptr, len, start_offset, |
1593 |
options | g_notempty, use_offsets, use_size_offsets); |
options | g_notempty, use_offsets, use_size_offsets); |
1594 |
extra->flags &= ~PCRE_EXTRA_CALLOUT_DATA; |
extra->flags &= ~PCRE_EXTRA_CALLOUT_DATA; |
1597 |
/* The normal case is just to do the match once, with the default |
/* The normal case is just to do the match once, with the default |
1598 |
value of match_limit. */ |
value of match_limit. */ |
1599 |
|
|
1600 |
else count = pcre_exec(re, extra, (char *)bptr, len, |
else |
1601 |
start_offset, options | g_notempty, use_offsets, use_size_offsets); |
{ |
1602 |
|
count = pcre_exec(re, extra, (char *)bptr, len, |
1603 |
|
start_offset, options | g_notempty, use_offsets, use_size_offsets); |
1604 |
|
} |
1605 |
|
|
1606 |
if (count == 0) |
if (count == 0) |
1607 |
{ |
{ |
1688 |
} |
} |
1689 |
} |
} |
1690 |
|
|
1691 |
|
/* There was a partial match */ |
1692 |
|
|
1693 |
|
else if (count == PCRE_ERROR_PARTIAL) |
1694 |
|
{ |
1695 |
|
fprintf(outfile, "Partial match\n"); |
1696 |
|
break; /* Out of the /g loop */ |
1697 |
|
} |
1698 |
|
|
1699 |
/* Failed to match. If this is a /g or /G loop and we previously set |
/* Failed to match. If this is a /g or /G loop and we previously set |
1700 |
g_notempty after a null match, this is not necessarily the end. |
g_notempty after a null match, this is not necessarily the end. |
1701 |
We want to advance the start offset, and continue. Fudge the offset |
We want to advance the start offset, and continue. In the case of UTF-8 |
1702 |
values to achieve this. We won't be at the end of the string - that |
matching, the advance must be one character, not one byte. Fudge the |
1703 |
was checked before setting g_notempty. */ |
offset values to achieve this. We won't be at the end of the string - |
1704 |
|
that was checked before setting g_notempty. */ |
1705 |
|
|
1706 |
else |
else |
1707 |
{ |
{ |
1708 |
if (g_notempty != 0) |
if (g_notempty != 0) |
1709 |
{ |
{ |
1710 |
|
int onechar = 1; |
1711 |
use_offsets[0] = start_offset; |
use_offsets[0] = start_offset; |
1712 |
use_offsets[1] = start_offset + 1; |
if (use_utf8) |
1713 |
|
{ |
1714 |
|
while (start_offset + onechar < len) |
1715 |
|
{ |
1716 |
|
int tb = bptr[start_offset+onechar]; |
1717 |
|
if (tb <= 127) break; |
1718 |
|
tb &= 0xc0; |
1719 |
|
if (tb != 0 && tb != 0xc0) onechar++; |
1720 |
|
} |
1721 |
|
} |
1722 |
|
use_offsets[1] = start_offset + onechar; |
1723 |
} |
} |
1724 |
else |
else |
1725 |
{ |
{ |
1726 |
if (gmatched == 0) /* Error if no previous matches */ |
if (count == PCRE_ERROR_NOMATCH) |
1727 |
{ |
{ |
1728 |
if (count == -1) fprintf(outfile, "No match\n"); |
if (gmatched == 0) fprintf(outfile, "No match\n"); |
|
else fprintf(outfile, "Error %d\n", count); |
|
1729 |
} |
} |
1730 |
|
else fprintf(outfile, "Error %d\n", count); |
1731 |
break; /* Out of the /g loop */ |
break; /* Out of the /g loop */ |
1732 |
} |
} |
1733 |
} |
} |
1779 |
} |
} |
1780 |
} |
} |
1781 |
|
|
1782 |
fprintf(outfile, "\n"); |
if (infile == stdin) fprintf(outfile, "\n"); |
1783 |
return 0; |
return 0; |
1784 |
} |
} |
1785 |
|
|