12 |
/* Use the internal info for displaying the results of pcre_study(). */ |
/* Use the internal info for displaying the results of pcre_study(). */ |
13 |
|
|
14 |
#include "internal.h" |
#include "internal.h" |
15 |
|
|
16 |
|
/* It is possible to compile this test program without including support for |
17 |
|
testing the POSIX interface, though this is not available via the standard |
18 |
|
Makefile. */ |
19 |
|
|
20 |
|
#if !defined NOPOSIX |
21 |
#include "pcreposix.h" |
#include "pcreposix.h" |
22 |
|
#endif |
23 |
|
|
24 |
#ifndef CLOCKS_PER_SEC |
#ifndef CLOCKS_PER_SEC |
25 |
#ifdef CLK_TCK |
#ifdef CLK_TCK |
29 |
#endif |
#endif |
30 |
#endif |
#endif |
31 |
|
|
32 |
#define LOOPREPEAT 10000 |
#define LOOPREPEAT 20000 |
33 |
|
|
34 |
|
|
35 |
static FILE *outfile; |
static FILE *outfile; |
36 |
static int log_store = 0; |
static int log_store = 0; |
37 |
|
static size_t gotten_store; |
38 |
|
|
39 |
|
|
40 |
|
|
41 |
|
static int utf8_table1[] = { |
42 |
|
0x0000007f, 0x000007ff, 0x0000ffff, 0x001fffff, 0x03ffffff, 0x7fffffff}; |
43 |
|
|
44 |
|
static int utf8_table2[] = { |
45 |
|
0, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc}; |
46 |
|
|
47 |
|
static int utf8_table3[] = { |
48 |
|
0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01}; |
49 |
|
|
50 |
|
|
51 |
|
/************************************************* |
52 |
|
* Convert character value to UTF-8 * |
53 |
|
*************************************************/ |
54 |
|
|
55 |
|
/* This function takes an integer value in the range 0 - 0x7fffffff |
56 |
|
and encodes it as a UTF-8 character in 0 to 6 bytes. |
57 |
|
|
58 |
|
Arguments: |
59 |
|
cvalue the character value |
60 |
|
buffer pointer to buffer for result - at least 6 bytes long |
61 |
|
|
62 |
|
Returns: number of characters placed in the buffer |
63 |
|
-1 if input character is negative |
64 |
|
0 if input character is positive but too big (only when |
65 |
|
int is longer than 32 bits) |
66 |
|
*/ |
67 |
|
|
68 |
|
static int |
69 |
|
ord2utf8(int cvalue, unsigned char *buffer) |
70 |
|
{ |
71 |
|
register int i, j; |
72 |
|
for (i = 0; i < sizeof(utf8_table1)/sizeof(int); i++) |
73 |
|
if (cvalue <= utf8_table1[i]) break; |
74 |
|
if (i >= sizeof(utf8_table1)/sizeof(int)) return 0; |
75 |
|
if (cvalue < 0) return -1; |
76 |
|
|
77 |
|
buffer += i; |
78 |
|
for (j = i; j > 0; j--) |
79 |
|
{ |
80 |
|
*buffer-- = 0x80 | (cvalue & 0x3f); |
81 |
|
cvalue >>= 6; |
82 |
|
} |
83 |
|
*buffer = utf8_table2[i] | cvalue; |
84 |
|
return i + 1; |
85 |
|
} |
86 |
|
|
87 |
|
|
88 |
|
/************************************************* |
89 |
|
* Convert UTF-8 string to value * |
90 |
|
*************************************************/ |
91 |
|
|
92 |
|
/* This function takes one or more bytes that represents a UTF-8 character, |
93 |
|
and returns the value of the character. |
94 |
|
|
95 |
|
Argument: |
96 |
|
buffer a pointer to the byte vector |
97 |
|
vptr a pointer to an int to receive the value |
98 |
|
|
99 |
|
Returns: > 0 => the number of bytes consumed |
100 |
|
-6 to 0 => malformed UTF-8 character at offset = (-return) |
101 |
|
*/ |
102 |
|
|
103 |
|
int |
104 |
|
utf82ord(unsigned char *buffer, int *vptr) |
105 |
|
{ |
106 |
|
int c = *buffer++; |
107 |
|
int d = c; |
108 |
|
int i, j, s; |
109 |
|
|
110 |
|
for (i = -1; i < 6; i++) /* i is number of additional bytes */ |
111 |
|
{ |
112 |
|
if ((d & 0x80) == 0) break; |
113 |
|
d <<= 1; |
114 |
|
} |
115 |
|
|
116 |
|
if (i == -1) { *vptr = c; return 1; } /* ascii character */ |
117 |
|
if (i == 0 || i == 6) return 0; /* invalid UTF-8 */ |
118 |
|
|
119 |
|
/* i now has a value in the range 1-5 */ |
120 |
|
|
121 |
|
s = 6*i; |
122 |
|
d = (c & utf8_table3[i]) << s; |
123 |
|
|
124 |
|
for (j = 0; j < i; j++) |
125 |
|
{ |
126 |
|
c = *buffer++; |
127 |
|
if ((c & 0xc0) != 0x80) return -(j+1); |
128 |
|
s -= 6; |
129 |
|
d |= (c & 0x3f) << s; |
130 |
|
} |
131 |
|
|
132 |
|
/* Check that encoding was the correct unique one */ |
133 |
|
|
134 |
|
for (j = 0; j < sizeof(utf8_table1)/sizeof(int); j++) |
135 |
|
if (d <= utf8_table1[j]) break; |
136 |
|
if (j != i) return -(i+1); |
137 |
|
|
138 |
|
/* Valid value */ |
139 |
|
|
140 |
|
*vptr = d; |
141 |
|
return i+1; |
142 |
|
} |
143 |
|
|
144 |
|
|
145 |
|
|
146 |
|
|
147 |
|
|
148 |
|
|
157 |
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
158 |
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
159 |
"*", "*?", "+", "+?", "?", "??", "{", "{", |
"*", "*?", "+", "+?", "?", "??", "{", "{", |
160 |
"class", "Ref", |
"class", "Ref", "Recurse", |
161 |
"Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", |
"Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", |
162 |
"AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cref", |
"AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cref", |
163 |
"Brazero", "Braminzero", "Bra" |
"Brazero", "Braminzero", "Branumber", "Bra" |
164 |
}; |
}; |
165 |
|
|
166 |
|
|
167 |
static void print_internals(pcre *re, FILE *outfile) |
static void print_internals(pcre *re) |
168 |
{ |
{ |
169 |
unsigned char *code = ((real_pcre *)re)->code; |
unsigned char *code = ((real_pcre *)re)->code; |
170 |
|
|
179 |
|
|
180 |
if (*code >= OP_BRA) |
if (*code >= OP_BRA) |
181 |
{ |
{ |
182 |
fprintf(outfile, "%3d Bra %d", (code[1] << 8) + code[2], *code - OP_BRA); |
if (*code - OP_BRA > EXTRACT_BASIC_MAX) |
183 |
|
fprintf(outfile, "%3d Bra extra", (code[1] << 8) + code[2]); |
184 |
|
else |
185 |
|
fprintf(outfile, "%3d Bra %d", (code[1] << 8) + code[2], *code - OP_BRA); |
186 |
code += 2; |
code += 2; |
187 |
} |
} |
188 |
|
|
198 |
code++; |
code++; |
199 |
break; |
break; |
200 |
|
|
|
case OP_COND: |
|
|
fprintf(outfile, "%3d Cond", (code[1] << 8) + code[2]); |
|
|
code += 2; |
|
|
break; |
|
|
|
|
|
case OP_CREF: |
|
|
fprintf(outfile, " %.2d %s", code[1], OP_names[*code]); |
|
|
code++; |
|
|
break; |
|
|
|
|
201 |
case OP_CHARS: |
case OP_CHARS: |
202 |
charlength = *(++code); |
charlength = *(++code); |
203 |
fprintf(outfile, "%3d ", charlength); |
fprintf(outfile, "%3d ", charlength); |
215 |
case OP_ASSERTBACK: |
case OP_ASSERTBACK: |
216 |
case OP_ASSERTBACK_NOT: |
case OP_ASSERTBACK_NOT: |
217 |
case OP_ONCE: |
case OP_ONCE: |
218 |
fprintf(outfile, "%3d %s", (code[1] << 8) + code[2], OP_names[*code]); |
case OP_COND: |
219 |
code += 2; |
case OP_BRANUMBER: |
|
break; |
|
|
|
|
220 |
case OP_REVERSE: |
case OP_REVERSE: |
221 |
|
case OP_CREF: |
222 |
fprintf(outfile, "%3d %s", (code[1] << 8) + code[2], OP_names[*code]); |
fprintf(outfile, "%3d %s", (code[1] << 8) + code[2], OP_names[*code]); |
223 |
code += 2; |
code += 2; |
224 |
break; |
break; |
291 |
break; |
break; |
292 |
|
|
293 |
case OP_REF: |
case OP_REF: |
294 |
fprintf(outfile, " \\%d", *(++code)); |
fprintf(outfile, " \\%d", (code[1] << 8) | code[2]); |
295 |
code++; |
code += 3; |
296 |
goto CLASS_REF_REPEAT; |
goto CLASS_REF_REPEAT; |
297 |
|
|
298 |
case OP_CLASS: |
case OP_CLASS: |
365 |
|
|
366 |
|
|
367 |
|
|
368 |
/* Character string printing function. */ |
/* Character string printing function. A "normal" and a UTF-8 version. */ |
369 |
|
|
370 |
static void pchars(unsigned char *p, int length) |
static void pchars(unsigned char *p, int length, int utf8) |
371 |
{ |
{ |
372 |
int c; |
int c; |
373 |
while (length-- > 0) |
while (length-- > 0) |
374 |
|
{ |
375 |
|
if (utf8) |
376 |
|
{ |
377 |
|
int rc = utf82ord(p, &c); |
378 |
|
if (rc > 0) |
379 |
|
{ |
380 |
|
length -= rc - 1; |
381 |
|
p += rc; |
382 |
|
if (c < 256 && isprint(c)) fprintf(outfile, "%c", c); |
383 |
|
else fprintf(outfile, "\\x{%02x}", c); |
384 |
|
continue; |
385 |
|
} |
386 |
|
} |
387 |
|
|
388 |
|
/* Not UTF-8, or malformed UTF-8 */ |
389 |
|
|
390 |
if (isprint(c = *(p++))) fprintf(outfile, "%c", c); |
if (isprint(c = *(p++))) fprintf(outfile, "%c", c); |
391 |
else fprintf(outfile, "\\x%02x", c); |
else fprintf(outfile, "\\x%02x", c); |
392 |
|
} |
393 |
} |
} |
394 |
|
|
395 |
|
|
399 |
|
|
400 |
static void *new_malloc(size_t size) |
static void *new_malloc(size_t size) |
401 |
{ |
{ |
402 |
if (log_store) fprintf(outfile, "Store size request: %d\n", (int)size); |
gotten_store = size; |
403 |
|
if (log_store) |
404 |
|
fprintf(outfile, "Memory allocation (code space): %d\n", |
405 |
|
(int)((int)size - offsetof(real_pcre, code[0]))); |
406 |
return malloc(size); |
return malloc(size); |
407 |
} |
} |
408 |
|
|
409 |
|
|
410 |
|
|
411 |
|
|
412 |
|
/* Get one piece of information from the pcre_fullinfo() function */ |
413 |
|
|
414 |
|
static void new_info(pcre *re, pcre_extra *study, int option, void *ptr) |
415 |
|
{ |
416 |
|
int rc; |
417 |
|
if ((rc = pcre_fullinfo(re, study, option, ptr)) < 0) |
418 |
|
fprintf(outfile, "Error %d from pcre_fullinfo(%d)\n", rc, option); |
419 |
|
} |
420 |
|
|
421 |
|
|
422 |
|
|
423 |
|
|
424 |
/* Read lines from named file or stdin and write to named file or stdout; lines |
/* Read lines from named file or stdin and write to named file or stdout; lines |
425 |
consist of a regular expression, in delimiters and optionally followed by |
consist of a regular expression, in delimiters and optionally followed by |
426 |
options, followed by a set of test data, terminated by an empty line. */ |
options, followed by a set of test data, terminated by an empty line. */ |
433 |
int op = 1; |
int op = 1; |
434 |
int timeit = 0; |
int timeit = 0; |
435 |
int showinfo = 0; |
int showinfo = 0; |
436 |
|
int showstore = 0; |
437 |
|
int size_offsets = 45; |
438 |
|
int size_offsets_max; |
439 |
|
int *offsets; |
440 |
|
#if !defined NOPOSIX |
441 |
int posix = 0; |
int posix = 0; |
442 |
|
#endif |
443 |
int debug = 0; |
int debug = 0; |
444 |
int done = 0; |
int done = 0; |
445 |
unsigned char buffer[30000]; |
unsigned char buffer[30000]; |
453 |
|
|
454 |
while (argc > 1 && argv[op][0] == '-') |
while (argc > 1 && argv[op][0] == '-') |
455 |
{ |
{ |
456 |
if (strcmp(argv[op], "-s") == 0) log_store = 1; |
char *endptr; |
457 |
|
|
458 |
|
if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0) |
459 |
|
showstore = 1; |
460 |
else if (strcmp(argv[op], "-t") == 0) timeit = 1; |
else if (strcmp(argv[op], "-t") == 0) timeit = 1; |
461 |
else if (strcmp(argv[op], "-i") == 0) showinfo = 1; |
else if (strcmp(argv[op], "-i") == 0) showinfo = 1; |
462 |
else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1; |
else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1; |
463 |
|
else if (strcmp(argv[op], "-o") == 0 && argc > 2 && |
464 |
|
((size_offsets = (int)strtoul(argv[op+1], &endptr, 10)), *endptr == 0)) |
465 |
|
{ |
466 |
|
op++; |
467 |
|
argc--; |
468 |
|
} |
469 |
|
#if !defined NOPOSIX |
470 |
else if (strcmp(argv[op], "-p") == 0) posix = 1; |
else if (strcmp(argv[op], "-p") == 0) posix = 1; |
471 |
|
#endif |
472 |
else |
else |
473 |
{ |
{ |
474 |
printf("*** Unknown option %s\n", argv[op]); |
printf("** Unknown or malformed option %s\n", argv[op]); |
475 |
printf("Usage: pcretest [-d] [-i] [-p] [-s] [-t] [<input> [<output>]]\n"); |
printf("Usage: pcretest [-d] [-i] [-o <n>] [-p] [-s] [-t] [<input> [<output>]]\n"); |
476 |
printf(" -d debug: show compiled code; implies -i\n" |
printf(" -d debug: show compiled code; implies -i\n" |
477 |
" -i show information about compiled pattern\n" |
" -i show information about compiled pattern\n" |
478 |
" -p use POSIX interface\n" |
" -o <n> set size of offsets vector to <n>\n"); |
479 |
" -s output store information\n" |
#if !defined NOPOSIX |
480 |
" -t time compilation and execution\n"); |
printf(" -p use POSIX interface\n"); |
481 |
|
#endif |
482 |
|
printf(" -s output store information\n" |
483 |
|
" -t time compilation and execution\n"); |
484 |
return 1; |
return 1; |
485 |
} |
} |
486 |
op++; |
op++; |
487 |
argc--; |
argc--; |
488 |
} |
} |
489 |
|
|
490 |
|
/* Get the store for the offsets vector, and remember what it was */ |
491 |
|
|
492 |
|
size_offsets_max = size_offsets; |
493 |
|
offsets = malloc(size_offsets_max * sizeof(int)); |
494 |
|
if (offsets == NULL) |
495 |
|
{ |
496 |
|
printf("** Failed to get %d bytes of memory for offsets vector\n", |
497 |
|
size_offsets_max * sizeof(int)); |
498 |
|
return 1; |
499 |
|
} |
500 |
|
|
501 |
/* Sort out the input and output files */ |
/* Sort out the input and output files */ |
502 |
|
|
503 |
if (argc > 1) |
if (argc > 1) |
534 |
{ |
{ |
535 |
pcre *re = NULL; |
pcre *re = NULL; |
536 |
pcre_extra *extra = NULL; |
pcre_extra *extra = NULL; |
537 |
|
|
538 |
|
#if !defined NOPOSIX /* There are still compilers that require no indent */ |
539 |
regex_t preg; |
regex_t preg; |
540 |
|
int do_posix = 0; |
541 |
|
#endif |
542 |
|
|
543 |
const char *error; |
const char *error; |
544 |
unsigned char *p, *pp, *ppp; |
unsigned char *p, *pp, *ppp; |
545 |
unsigned const char *tables = NULL; |
const unsigned char *tables = NULL; |
546 |
int do_study = 0; |
int do_study = 0; |
547 |
int do_debug = debug; |
int do_debug = debug; |
548 |
|
int do_G = 0; |
549 |
|
int do_g = 0; |
550 |
int do_showinfo = showinfo; |
int do_showinfo = showinfo; |
551 |
int do_posix = 0; |
int do_showrest = 0; |
552 |
|
int utf8 = 0; |
553 |
int erroroffset, len, delimiter; |
int erroroffset, len, delimiter; |
554 |
|
|
555 |
if (infile == stdin) printf(" re> "); |
if (infile == stdin) printf(" re> "); |
565 |
|
|
566 |
delimiter = *p++; |
delimiter = *p++; |
567 |
|
|
568 |
if (isalnum(delimiter)) |
if (isalnum(delimiter) || delimiter == '\\') |
569 |
{ |
{ |
570 |
fprintf(outfile, "** Delimiter must not be alphameric\n"); |
fprintf(outfile, "** Delimiter must not be alphameric or \\\n"); |
571 |
goto SKIP_DATA; |
goto SKIP_DATA; |
572 |
} |
} |
573 |
|
|
575 |
|
|
576 |
for(;;) |
for(;;) |
577 |
{ |
{ |
578 |
while (*pp != 0 && *pp != delimiter) pp++; |
while (*pp != 0) |
579 |
|
{ |
580 |
|
if (*pp == '\\' && pp[1] != 0) pp++; |
581 |
|
else if (*pp == delimiter) break; |
582 |
|
pp++; |
583 |
|
} |
584 |
if (*pp != 0) break; |
if (*pp != 0) break; |
585 |
|
|
586 |
len = sizeof(buffer) - (pp - buffer); |
len = sizeof(buffer) - (pp - buffer); |
600 |
if (infile != stdin) fprintf(outfile, "%s", (char *)pp); |
if (infile != stdin) fprintf(outfile, "%s", (char *)pp); |
601 |
} |
} |
602 |
|
|
603 |
|
/* If the first character after the delimiter is backslash, make |
604 |
|
the pattern end with backslash. This is purely to provide a way |
605 |
|
of testing for the error message when a pattern ends with backslash. */ |
606 |
|
|
607 |
|
if (pp[1] == '\\') *pp++ = '\\'; |
608 |
|
|
609 |
/* Terminate the pattern at the delimiter */ |
/* Terminate the pattern at the delimiter */ |
610 |
|
|
611 |
*pp++ = 0; |
*pp++ = 0; |
614 |
|
|
615 |
options = 0; |
options = 0; |
616 |
study_options = 0; |
study_options = 0; |
617 |
|
log_store = showstore; /* default from command line */ |
618 |
|
|
619 |
while (*pp != 0) |
while (*pp != 0) |
620 |
{ |
{ |
621 |
switch (*pp++) |
switch (*pp++) |
622 |
{ |
{ |
623 |
|
case 'g': do_g = 1; break; |
624 |
case 'i': options |= PCRE_CASELESS; break; |
case 'i': options |= PCRE_CASELESS; break; |
625 |
case 'm': options |= PCRE_MULTILINE; break; |
case 'm': options |= PCRE_MULTILINE; break; |
626 |
case 's': options |= PCRE_DOTALL; break; |
case 's': options |= PCRE_DOTALL; break; |
627 |
case 'x': options |= PCRE_EXTENDED; break; |
case 'x': options |= PCRE_EXTENDED; break; |
628 |
|
|
629 |
|
case '+': do_showrest = 1; break; |
630 |
case 'A': options |= PCRE_ANCHORED; break; |
case 'A': options |= PCRE_ANCHORED; break; |
631 |
case 'D': do_debug = do_showinfo = 1; break; |
case 'D': do_debug = do_showinfo = 1; break; |
632 |
case 'E': options |= PCRE_DOLLAR_ENDONLY; break; |
case 'E': options |= PCRE_DOLLAR_ENDONLY; break; |
633 |
|
case 'G': do_G = 1; break; |
634 |
case 'I': do_showinfo = 1; break; |
case 'I': do_showinfo = 1; break; |
635 |
|
case 'M': log_store = 1; break; |
636 |
|
|
637 |
|
#if !defined NOPOSIX |
638 |
case 'P': do_posix = 1; break; |
case 'P': do_posix = 1; break; |
639 |
|
#endif |
640 |
|
|
641 |
case 'S': do_study = 1; break; |
case 'S': do_study = 1; break; |
642 |
case 'U': options |= PCRE_UNGREEDY; break; |
case 'U': options |= PCRE_UNGREEDY; break; |
643 |
case 'X': options |= PCRE_EXTRA; break; |
case 'X': options |= PCRE_EXTRA; break; |
644 |
|
case '8': options |= PCRE_UTF8; utf8 = 1; break; |
645 |
|
|
646 |
case 'L': |
case 'L': |
647 |
ppp = pp; |
ppp = pp; |
667 |
timing, showing, or debugging options, nor the ability to pass over |
timing, showing, or debugging options, nor the ability to pass over |
668 |
local character tables. */ |
local character tables. */ |
669 |
|
|
670 |
|
#if !defined NOPOSIX |
671 |
if (posix || do_posix) |
if (posix || do_posix) |
672 |
{ |
{ |
673 |
int rc; |
int rc; |
690 |
/* Handle compiling via the native interface */ |
/* Handle compiling via the native interface */ |
691 |
|
|
692 |
else |
else |
693 |
|
#endif /* !defined NOPOSIX */ |
694 |
|
|
695 |
{ |
{ |
696 |
if (timeit) |
if (timeit) |
697 |
{ |
{ |
704 |
if (re != NULL) free(re); |
if (re != NULL) free(re); |
705 |
} |
} |
706 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
707 |
fprintf(outfile, "Compile time %.2f milliseconds\n", |
fprintf(outfile, "Compile time %.3f milliseconds\n", |
708 |
((double)time_taken)/(4 * CLOCKS_PER_SEC)); |
((double)time_taken * 1000.0) / |
709 |
|
((double)LOOPREPEAT * (double)CLOCKS_PER_SEC)); |
710 |
} |
} |
711 |
|
|
712 |
re = pcre_compile((char *)p, options, &error, &erroroffset, tables); |
re = pcre_compile((char *)p, options, &error, &erroroffset, tables); |
736 |
goto CONTINUE; |
goto CONTINUE; |
737 |
} |
} |
738 |
|
|
739 |
/* Compilation succeeded; print data if required */ |
/* Compilation succeeded; print data if required. There are now two |
740 |
|
info-returning functions. The old one has a limited interface and |
741 |
|
returns only limited data. Check that it agrees with the newer one. */ |
742 |
|
|
743 |
if (do_showinfo) |
if (do_showinfo) |
744 |
{ |
{ |
745 |
int first_char, count; |
unsigned long int get_options; |
746 |
|
int old_first_char, old_options, old_count; |
747 |
if (do_debug) print_internals(re, outfile); |
int count, backrefmax, first_char, need_char; |
748 |
|
size_t size; |
749 |
|
|
750 |
|
if (do_debug) print_internals(re); |
751 |
|
|
752 |
|
new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options); |
753 |
|
new_info(re, NULL, PCRE_INFO_SIZE, &size); |
754 |
|
new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count); |
755 |
|
new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax); |
756 |
|
new_info(re, NULL, PCRE_INFO_FIRSTCHAR, &first_char); |
757 |
|
new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char); |
758 |
|
|
759 |
count = pcre_info(re, &options, &first_char); |
old_count = pcre_info(re, &old_options, &old_first_char); |
760 |
if (count < 0) fprintf(outfile, |
if (count < 0) fprintf(outfile, |
761 |
"Error %d while reading info\n", count); |
"Error %d from pcre_info()\n", count); |
762 |
else |
else |
763 |
{ |
{ |
764 |
fprintf(outfile, "Identifying subpattern count = %d\n", count); |
if (old_count != count) fprintf(outfile, |
765 |
if (options == 0) fprintf(outfile, "No options\n"); |
"Count disagreement: pcre_fullinfo=%d pcre_info=%d\n", count, |
766 |
else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s\n", |
old_count); |
767 |
((options & PCRE_ANCHORED) != 0)? " anchored" : "", |
|
768 |
((options & PCRE_CASELESS) != 0)? " caseless" : "", |
if (old_first_char != first_char) fprintf(outfile, |
769 |
((options & PCRE_EXTENDED) != 0)? " extended" : "", |
"First char disagreement: pcre_fullinfo=%d pcre_info=%d\n", |
770 |
((options & PCRE_MULTILINE) != 0)? " multiline" : "", |
first_char, old_first_char); |
771 |
((options & PCRE_DOTALL) != 0)? " dotall" : "", |
|
772 |
((options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
if (old_options != (int)get_options) fprintf(outfile, |
773 |
((options & PCRE_EXTRA) != 0)? " extra" : "", |
"Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n", |
774 |
((options & PCRE_UNGREEDY) != 0)? " ungreedy" : ""); |
get_options, old_options); |
775 |
if (first_char == -1) |
} |
776 |
{ |
|
777 |
fprintf(outfile, "First char at start or follows \\n\n"); |
if (size != gotten_store) fprintf(outfile, |
778 |
} |
"Size disagreement: pcre_fullinfo=%d call to malloc for %d\n", |
779 |
else if (first_char < 0) |
size, gotten_store); |
780 |
{ |
|
781 |
fprintf(outfile, "No first char\n"); |
fprintf(outfile, "Capturing subpattern count = %d\n", count); |
782 |
} |
if (backrefmax > 0) |
783 |
|
fprintf(outfile, "Max back reference = %d\n", backrefmax); |
784 |
|
if (get_options == 0) fprintf(outfile, "No options\n"); |
785 |
|
else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s\n", |
786 |
|
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
787 |
|
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
788 |
|
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
789 |
|
((get_options & PCRE_MULTILINE) != 0)? " multiline" : "", |
790 |
|
((get_options & PCRE_DOTALL) != 0)? " dotall" : "", |
791 |
|
((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
792 |
|
((get_options & PCRE_EXTRA) != 0)? " extra" : "", |
793 |
|
((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "", |
794 |
|
((get_options & PCRE_UTF8) != 0)? " utf8" : ""); |
795 |
|
|
796 |
|
if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0) |
797 |
|
fprintf(outfile, "Case state changes\n"); |
798 |
|
|
799 |
|
if (first_char == -1) |
800 |
|
{ |
801 |
|
fprintf(outfile, "First char at start or follows \\n\n"); |
802 |
|
} |
803 |
|
else if (first_char < 0) |
804 |
|
{ |
805 |
|
fprintf(outfile, "No first char\n"); |
806 |
|
} |
807 |
|
else |
808 |
|
{ |
809 |
|
if (isprint(first_char)) |
810 |
|
fprintf(outfile, "First char = \'%c\'\n", first_char); |
811 |
else |
else |
812 |
{ |
fprintf(outfile, "First char = %d\n", first_char); |
813 |
if (isprint(first_char)) |
} |
814 |
fprintf(outfile, "First char = \'%c\'\n", first_char); |
|
815 |
else |
if (need_char < 0) |
816 |
fprintf(outfile, "First char = %d\n", first_char); |
{ |
817 |
} |
fprintf(outfile, "No need char\n"); |
818 |
|
} |
819 |
|
else |
820 |
|
{ |
821 |
|
if (isprint(need_char)) |
822 |
|
fprintf(outfile, "Need char = \'%c\'\n", need_char); |
823 |
|
else |
824 |
|
fprintf(outfile, "Need char = %d\n", need_char); |
825 |
} |
} |
826 |
} |
} |
827 |
|
|
839 |
extra = pcre_study(re, study_options, &error); |
extra = pcre_study(re, study_options, &error); |
840 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
841 |
if (extra != NULL) free(extra); |
if (extra != NULL) free(extra); |
842 |
fprintf(outfile, " Study time %.2f milliseconds\n", |
fprintf(outfile, " Study time %.3f milliseconds\n", |
843 |
((double)time_taken)/(4 * CLOCKS_PER_SEC)); |
((double)time_taken * 1000.0)/ |
844 |
|
((double)LOOPREPEAT * (double)CLOCKS_PER_SEC)); |
845 |
} |
} |
846 |
|
|
847 |
extra = pcre_study(re, study_options, &error); |
extra = pcre_study(re, study_options, &error); |
850 |
else if (extra == NULL) |
else if (extra == NULL) |
851 |
fprintf(outfile, "Study returned NULL\n"); |
fprintf(outfile, "Study returned NULL\n"); |
852 |
|
|
|
/* This looks at internal information. A bit kludgy to do it this |
|
|
way, but it is useful for testing. */ |
|
|
|
|
853 |
else if (do_showinfo) |
else if (do_showinfo) |
854 |
{ |
{ |
855 |
real_pcre_extra *xx = (real_pcre_extra *)extra; |
uschar *start_bits = NULL; |
856 |
if ((xx->options & PCRE_STUDY_MAPPED) == 0) |
new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits); |
857 |
|
if (start_bits == NULL) |
858 |
fprintf(outfile, "No starting character set\n"); |
fprintf(outfile, "No starting character set\n"); |
859 |
else |
else |
860 |
{ |
{ |
863 |
fprintf(outfile, "Starting character set: "); |
fprintf(outfile, "Starting character set: "); |
864 |
for (i = 0; i < 256; i++) |
for (i = 0; i < 256; i++) |
865 |
{ |
{ |
866 |
if ((xx->start_bits[i/8] & (1<<(i%8))) != 0) |
if ((start_bits[i/8] & (1<<(i%8))) != 0) |
867 |
{ |
{ |
868 |
if (c > 75) |
if (c > 75) |
869 |
{ |
{ |
893 |
for (;;) |
for (;;) |
894 |
{ |
{ |
895 |
unsigned char *q; |
unsigned char *q; |
896 |
|
unsigned char *bptr = dbuffer; |
897 |
|
int *use_offsets = offsets; |
898 |
|
int use_size_offsets = size_offsets; |
899 |
int count, c; |
int count, c; |
900 |
int offsets[45]; |
int copystrings = 0; |
901 |
int size_offsets = sizeof(offsets)/sizeof(int); |
int getstrings = 0; |
902 |
|
int getlist = 0; |
903 |
|
int gmatched = 0; |
904 |
|
int start_offset = 0; |
905 |
|
int g_notempty = 0; |
906 |
|
|
907 |
options = 0; |
options = 0; |
908 |
|
|
909 |
if (infile == stdin) printf(" data> "); |
if (infile == stdin) printf("data> "); |
910 |
if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) |
if (fgets((char *)buffer, sizeof(buffer), infile) == NULL) |
911 |
{ |
{ |
912 |
done = 1; |
done = 1; |
946 |
break; |
break; |
947 |
|
|
948 |
case 'x': |
case 'x': |
949 |
|
|
950 |
|
/* Handle \x{..} specially - new Perl thing for utf8 */ |
951 |
|
|
952 |
|
if (*p == '{') |
953 |
|
{ |
954 |
|
unsigned char *pt = p; |
955 |
|
c = 0; |
956 |
|
while (isxdigit(*(++pt))) |
957 |
|
c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W'); |
958 |
|
if (*pt == '}') |
959 |
|
{ |
960 |
|
unsigned char buffer[8]; |
961 |
|
int ii, utn; |
962 |
|
utn = ord2utf8(c, buffer); |
963 |
|
for (ii = 0; ii < utn - 1; ii++) *q++ = buffer[ii]; |
964 |
|
c = buffer[ii]; /* Last byte */ |
965 |
|
p = pt + 1; |
966 |
|
break; |
967 |
|
} |
968 |
|
/* Not correct form; fall through */ |
969 |
|
} |
970 |
|
|
971 |
|
/* Ordinary \x */ |
972 |
|
|
973 |
c = 0; |
c = 0; |
974 |
while (i++ < 2 && isxdigit(*p)) |
while (i++ < 2 && isxdigit(*p)) |
975 |
{ |
{ |
990 |
options |= PCRE_NOTBOL; |
options |= PCRE_NOTBOL; |
991 |
continue; |
continue; |
992 |
|
|
993 |
|
case 'C': |
994 |
|
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
995 |
|
copystrings |= 1 << n; |
996 |
|
continue; |
997 |
|
|
998 |
|
case 'G': |
999 |
|
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
1000 |
|
getstrings |= 1 << n; |
1001 |
|
continue; |
1002 |
|
|
1003 |
|
case 'L': |
1004 |
|
getlist = 1; |
1005 |
|
continue; |
1006 |
|
|
1007 |
|
case 'N': |
1008 |
|
options |= PCRE_NOTEMPTY; |
1009 |
|
continue; |
1010 |
|
|
1011 |
case 'O': |
case 'O': |
1012 |
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
1013 |
if (n <= (int)(sizeof(offsets)/sizeof(int))) size_offsets = n; |
if (n > size_offsets_max) |
1014 |
|
{ |
1015 |
|
size_offsets_max = n; |
1016 |
|
free(offsets); |
1017 |
|
use_offsets = offsets = malloc(size_offsets_max * sizeof(int)); |
1018 |
|
if (offsets == NULL) |
1019 |
|
{ |
1020 |
|
printf("** Failed to get %d bytes of memory for offsets vector\n", |
1021 |
|
size_offsets_max * sizeof(int)); |
1022 |
|
return 1; |
1023 |
|
} |
1024 |
|
} |
1025 |
|
use_size_offsets = n; |
1026 |
|
if (n == 0) use_offsets = NULL; |
1027 |
continue; |
continue; |
1028 |
|
|
1029 |
case 'Z': |
case 'Z': |
1038 |
/* Handle matching via the POSIX interface, which does not |
/* Handle matching via the POSIX interface, which does not |
1039 |
support timing. */ |
support timing. */ |
1040 |
|
|
1041 |
|
#if !defined NOPOSIX |
1042 |
if (posix || do_posix) |
if (posix || do_posix) |
1043 |
{ |
{ |
1044 |
int rc; |
int rc; |
1045 |
int eflags = 0; |
int eflags = 0; |
1046 |
regmatch_t pmatch[30]; |
regmatch_t *pmatch = malloc(sizeof(regmatch_t) * use_size_offsets); |
1047 |
if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL; |
if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL; |
1048 |
if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL; |
if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL; |
1049 |
|
|
1050 |
rc = regexec(&preg, (char *)dbuffer, sizeof(pmatch)/sizeof(regmatch_t), |
rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags); |
|
pmatch, eflags); |
|
1051 |
|
|
1052 |
if (rc != 0) |
if (rc != 0) |
1053 |
{ |
{ |
1057 |
else |
else |
1058 |
{ |
{ |
1059 |
size_t i; |
size_t i; |
1060 |
for (i = 0; i < sizeof(pmatch)/sizeof(regmatch_t); i++) |
for (i = 0; i < use_size_offsets; i++) |
1061 |
{ |
{ |
1062 |
if (pmatch[i].rm_so >= 0) |
if (pmatch[i].rm_so >= 0) |
1063 |
{ |
{ |
1064 |
fprintf(outfile, "%2d: ", (int)i); |
fprintf(outfile, "%2d: ", (int)i); |
1065 |
pchars(dbuffer + pmatch[i].rm_so, |
pchars(dbuffer + pmatch[i].rm_so, |
1066 |
pmatch[i].rm_eo - pmatch[i].rm_so); |
pmatch[i].rm_eo - pmatch[i].rm_so, utf8); |
1067 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
1068 |
|
if (i == 0 && do_showrest) |
1069 |
|
{ |
1070 |
|
fprintf(outfile, " 0+ "); |
1071 |
|
pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo, utf8); |
1072 |
|
fprintf(outfile, "\n"); |
1073 |
|
} |
1074 |
} |
} |
1075 |
} |
} |
1076 |
} |
} |
1077 |
|
free(pmatch); |
1078 |
} |
} |
1079 |
|
|
1080 |
/* Handle matching via the native interface */ |
/* Handle matching via the native interface - repeats for /g and /G */ |
1081 |
|
|
1082 |
else |
else |
1083 |
|
#endif /* !defined NOPOSIX */ |
1084 |
|
|
1085 |
|
for (;; gmatched++) /* Loop for /g or /G */ |
1086 |
{ |
{ |
1087 |
if (timeit) |
if (timeit) |
1088 |
{ |
{ |
1089 |
register int i; |
register int i; |
1090 |
clock_t time_taken; |
clock_t time_taken; |
1091 |
clock_t start_time = clock(); |
clock_t start_time = clock(); |
1092 |
for (i = 0; i < 4000; i++) |
for (i = 0; i < LOOPREPEAT; i++) |
1093 |
count = pcre_exec(re, extra, (char *)dbuffer, len, options, offsets, |
count = pcre_exec(re, extra, (char *)bptr, len, |
1094 |
size_offsets); |
start_offset, options | g_notempty, use_offsets, use_size_offsets); |
1095 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
1096 |
fprintf(outfile, "Execute time %.2f milliseconds\n", |
fprintf(outfile, "Execute time %.3f milliseconds\n", |
1097 |
((double)time_taken)/(4 * CLOCKS_PER_SEC)); |
((double)time_taken * 1000.0)/ |
1098 |
|
((double)LOOPREPEAT * (double)CLOCKS_PER_SEC)); |
1099 |
} |
} |
1100 |
|
|
1101 |
count = pcre_exec(re, extra, (char *)dbuffer, len, options, offsets, |
count = pcre_exec(re, extra, (char *)bptr, len, |
1102 |
size_offsets); |
start_offset, options | g_notempty, use_offsets, use_size_offsets); |
1103 |
|
|
1104 |
if (count == 0) |
if (count == 0) |
1105 |
{ |
{ |
1106 |
fprintf(outfile, "Matched, but too many substrings\n"); |
fprintf(outfile, "Matched, but too many substrings\n"); |
1107 |
count = size_offsets/3; |
count = use_size_offsets/3; |
1108 |
} |
} |
1109 |
|
|
1110 |
|
/* Matched */ |
1111 |
|
|
1112 |
if (count >= 0) |
if (count >= 0) |
1113 |
{ |
{ |
1114 |
int i; |
int i; |
1115 |
count *= 2; |
for (i = 0; i < count * 2; i += 2) |
|
for (i = 0; i < count; i += 2) |
|
1116 |
{ |
{ |
1117 |
if (offsets[i] < 0) |
if (use_offsets[i] < 0) |
1118 |
fprintf(outfile, "%2d: <unset>\n", i/2); |
fprintf(outfile, "%2d: <unset>\n", i/2); |
1119 |
else |
else |
1120 |
{ |
{ |
1121 |
fprintf(outfile, "%2d: ", i/2); |
fprintf(outfile, "%2d: ", i/2); |
1122 |
pchars(dbuffer + offsets[i], offsets[i+1] - offsets[i]); |
pchars(bptr + use_offsets[i], use_offsets[i+1] - use_offsets[i], utf8); |
1123 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
1124 |
|
if (i == 0) |
1125 |
|
{ |
1126 |
|
if (do_showrest) |
1127 |
|
{ |
1128 |
|
fprintf(outfile, " 0+ "); |
1129 |
|
pchars(bptr + use_offsets[i+1], len - use_offsets[i+1], utf8); |
1130 |
|
fprintf(outfile, "\n"); |
1131 |
|
} |
1132 |
|
} |
1133 |
|
} |
1134 |
|
} |
1135 |
|
|
1136 |
|
for (i = 0; i < 32; i++) |
1137 |
|
{ |
1138 |
|
if ((copystrings & (1 << i)) != 0) |
1139 |
|
{ |
1140 |
|
char copybuffer[16]; |
1141 |
|
int rc = pcre_copy_substring((char *)bptr, use_offsets, count, |
1142 |
|
i, copybuffer, sizeof(copybuffer)); |
1143 |
|
if (rc < 0) |
1144 |
|
fprintf(outfile, "copy substring %d failed %d\n", i, rc); |
1145 |
|
else |
1146 |
|
fprintf(outfile, "%2dC %s (%d)\n", i, copybuffer, rc); |
1147 |
|
} |
1148 |
|
} |
1149 |
|
|
1150 |
|
for (i = 0; i < 32; i++) |
1151 |
|
{ |
1152 |
|
if ((getstrings & (1 << i)) != 0) |
1153 |
|
{ |
1154 |
|
const char *substring; |
1155 |
|
int rc = pcre_get_substring((char *)bptr, use_offsets, count, |
1156 |
|
i, &substring); |
1157 |
|
if (rc < 0) |
1158 |
|
fprintf(outfile, "get substring %d failed %d\n", i, rc); |
1159 |
|
else |
1160 |
|
{ |
1161 |
|
fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc); |
1162 |
|
/* free((void *)substring); */ |
1163 |
|
pcre_free_substring(substring); |
1164 |
|
} |
1165 |
|
} |
1166 |
|
} |
1167 |
|
|
1168 |
|
if (getlist) |
1169 |
|
{ |
1170 |
|
const char **stringlist; |
1171 |
|
int rc = pcre_get_substring_list((char *)bptr, use_offsets, count, |
1172 |
|
&stringlist); |
1173 |
|
if (rc < 0) |
1174 |
|
fprintf(outfile, "get substring list failed %d\n", rc); |
1175 |
|
else |
1176 |
|
{ |
1177 |
|
for (i = 0; i < count; i++) |
1178 |
|
fprintf(outfile, "%2dL %s\n", i, stringlist[i]); |
1179 |
|
if (stringlist[i] != NULL) |
1180 |
|
fprintf(outfile, "string list not terminated by NULL\n"); |
1181 |
|
/* free((void *)stringlist); */ |
1182 |
|
pcre_free_substring_list(stringlist); |
1183 |
} |
} |
1184 |
} |
} |
1185 |
} |
} |
1186 |
|
|
1187 |
|
/* Failed to match. If this is a /g or /G loop and we previously set |
1188 |
|
g_notempty after a null match, this is not necessarily the end. |
1189 |
|
We want to advance the start offset, and continue. Fudge the offset |
1190 |
|
values to achieve this. We won't be at the end of the string - that |
1191 |
|
was checked before setting g_notempty. */ |
1192 |
|
|
1193 |
else |
else |
1194 |
{ |
{ |
1195 |
if (count == -1) fprintf(outfile, "No match\n"); |
if (g_notempty != 0) |
1196 |
else fprintf(outfile, "Error %d\n", count); |
{ |
1197 |
|
use_offsets[0] = start_offset; |
1198 |
|
use_offsets[1] = start_offset + 1; |
1199 |
|
} |
1200 |
|
else |
1201 |
|
{ |
1202 |
|
if (gmatched == 0) /* Error if no previous matches */ |
1203 |
|
{ |
1204 |
|
if (count == -1) fprintf(outfile, "No match\n"); |
1205 |
|
else fprintf(outfile, "Error %d\n", count); |
1206 |
|
} |
1207 |
|
break; /* Out of the /g loop */ |
1208 |
|
} |
1209 |
} |
} |
1210 |
} |
|
1211 |
} |
/* If not /g or /G we are done */ |
1212 |
|
|
1213 |
|
if (!do_g && !do_G) break; |
1214 |
|
|
1215 |
|
/* If we have matched an empty string, first check to see if we are at |
1216 |
|
the end of the subject. If so, the /g loop is over. Otherwise, mimic |
1217 |
|
what Perl's /g options does. This turns out to be rather cunning. First |
1218 |
|
we set PCRE_NOTEMPTY and PCRE_ANCHORED and try the match again at the |
1219 |
|
same point. If this fails (picked up above) we advance to the next |
1220 |
|
character. */ |
1221 |
|
|
1222 |
|
g_notempty = 0; |
1223 |
|
if (use_offsets[0] == use_offsets[1]) |
1224 |
|
{ |
1225 |
|
if (use_offsets[0] == len) break; |
1226 |
|
g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED; |
1227 |
|
} |
1228 |
|
|
1229 |
|
/* For /g, update the start offset, leaving the rest alone */ |
1230 |
|
|
1231 |
|
if (do_g) start_offset = use_offsets[1]; |
1232 |
|
|
1233 |
|
/* For /G, update the pointer and length */ |
1234 |
|
|
1235 |
|
else |
1236 |
|
{ |
1237 |
|
bptr += use_offsets[1]; |
1238 |
|
len -= use_offsets[1]; |
1239 |
|
} |
1240 |
|
} /* End of loop for /g and /G */ |
1241 |
|
} /* End of loop for data lines */ |
1242 |
|
|
1243 |
CONTINUE: |
CONTINUE: |
1244 |
|
|
1245 |
|
#if !defined NOPOSIX |
1246 |
if (posix || do_posix) regfree(&preg); |
if (posix || do_posix) regfree(&preg); |
1247 |
|
#endif |
1248 |
|
|
1249 |
if (re != NULL) free(re); |
if (re != NULL) free(re); |
1250 |
if (extra != NULL) free(extra); |
if (extra != NULL) free(extra); |
1251 |
if (tables != NULL) |
if (tables != NULL) |