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 |
|
*buffer++ = utf8_table2[i] | (cvalue & utf8_table3[i]); |
77 |
|
cvalue >>= 6 - i; |
78 |
|
for (j = 0; j < i; j++) |
79 |
|
{ |
80 |
|
*buffer++ = 0x80 | (cvalue & 0x3f); |
81 |
|
cvalue >>= 6; |
82 |
|
} |
83 |
|
return i + 1; |
84 |
|
} |
85 |
|
|
86 |
|
|
87 |
|
/************************************************* |
88 |
|
* Convert UTF-8 string to value * |
89 |
|
*************************************************/ |
90 |
|
|
91 |
|
/* This function takes one or more bytes that represents a UTF-8 character, |
92 |
|
and returns the value of the character. |
93 |
|
|
94 |
|
Argument: |
95 |
|
buffer a pointer to the byte vector |
96 |
|
vptr a pointer to an int to receive the value |
97 |
|
|
98 |
|
Returns: > 0 => the number of bytes consumed |
99 |
|
-6 to 0 => malformed UTF-8 character at offset = (-return) |
100 |
|
*/ |
101 |
|
|
102 |
|
int |
103 |
|
utf82ord(unsigned char *buffer, int *vptr) |
104 |
|
{ |
105 |
|
int c = *buffer++; |
106 |
|
int d = c; |
107 |
|
int i, j, s; |
108 |
|
|
109 |
|
for (i = -1; i < 6; i++) /* i is number of additional bytes */ |
110 |
|
{ |
111 |
|
if ((d & 0x80) == 0) break; |
112 |
|
d <<= 1; |
113 |
|
} |
114 |
|
|
115 |
|
if (i == -1) { *vptr = c; return 1; } /* ascii character */ |
116 |
|
if (i == 0 || i == 6) return 0; /* invalid UTF-8 */ |
117 |
|
|
118 |
|
/* i now has a value in the range 1-5 */ |
119 |
|
|
120 |
|
d = c & utf8_table3[i]; |
121 |
|
s = 6 - i; |
122 |
|
|
123 |
|
for (j = 0; j < i; j++) |
124 |
|
{ |
125 |
|
c = *buffer++; |
126 |
|
if ((c & 0xc0) != 0x80) return -(j+1); |
127 |
|
d |= (c & 0x3f) << s; |
128 |
|
s += 6; |
129 |
|
} |
130 |
|
|
131 |
|
/* Check that encoding was the correct unique one */ |
132 |
|
|
133 |
|
for (j = 0; j < sizeof(utf8_table1)/sizeof(int); j++) |
134 |
|
if (d <= utf8_table1[j]) break; |
135 |
|
if (j != i) return -(i+1); |
136 |
|
|
137 |
|
/* Valid value */ |
138 |
|
|
139 |
|
*vptr = d; |
140 |
|
return i+1; |
141 |
|
} |
142 |
|
|
143 |
|
|
144 |
|
|
145 |
|
|
146 |
|
|
147 |
|
|
156 |
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
157 |
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
"*", "*?", "+", "+?", "?", "??", "{", "{", "{", |
158 |
"*", "*?", "+", "+?", "?", "??", "{", "{", |
"*", "*?", "+", "+?", "?", "??", "{", "{", |
159 |
"class", "Ref", |
"class", "Ref", "Recurse", |
160 |
"Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", |
"Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", |
161 |
"AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cref", |
"AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cref", |
162 |
"Brazero", "Braminzero", "Bra" |
"Brazero", "Braminzero", "Branumber", "Bra" |
163 |
}; |
}; |
164 |
|
|
165 |
|
|
178 |
|
|
179 |
if (*code >= OP_BRA) |
if (*code >= OP_BRA) |
180 |
{ |
{ |
181 |
fprintf(outfile, "%3d Bra %d", (code[1] << 8) + code[2], *code - OP_BRA); |
if (*code - OP_BRA > EXTRACT_BASIC_MAX) |
182 |
|
fprintf(outfile, "%3d Bra extra", (code[1] << 8) + code[2]); |
183 |
|
else |
184 |
|
fprintf(outfile, "%3d Bra %d", (code[1] << 8) + code[2], *code - OP_BRA); |
185 |
code += 2; |
code += 2; |
186 |
} |
} |
187 |
|
|
197 |
code++; |
code++; |
198 |
break; |
break; |
199 |
|
|
|
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; |
|
|
|
|
200 |
case OP_CHARS: |
case OP_CHARS: |
201 |
charlength = *(++code); |
charlength = *(++code); |
202 |
fprintf(outfile, "%3d ", charlength); |
fprintf(outfile, "%3d ", charlength); |
214 |
case OP_ASSERTBACK: |
case OP_ASSERTBACK: |
215 |
case OP_ASSERTBACK_NOT: |
case OP_ASSERTBACK_NOT: |
216 |
case OP_ONCE: |
case OP_ONCE: |
217 |
fprintf(outfile, "%3d %s", (code[1] << 8) + code[2], OP_names[*code]); |
case OP_COND: |
218 |
code += 2; |
case OP_BRANUMBER: |
|
break; |
|
|
|
|
219 |
case OP_REVERSE: |
case OP_REVERSE: |
220 |
|
case OP_CREF: |
221 |
fprintf(outfile, "%3d %s", (code[1] << 8) + code[2], OP_names[*code]); |
fprintf(outfile, "%3d %s", (code[1] << 8) + code[2], OP_names[*code]); |
222 |
code += 2; |
code += 2; |
223 |
break; |
break; |
290 |
break; |
break; |
291 |
|
|
292 |
case OP_REF: |
case OP_REF: |
293 |
fprintf(outfile, " \\%d", *(++code)); |
fprintf(outfile, " \\%d", (code[1] << 8) | code[2]); |
294 |
code++; |
code += 3; |
295 |
goto CLASS_REF_REPEAT; |
goto CLASS_REF_REPEAT; |
296 |
|
|
297 |
case OP_CLASS: |
case OP_CLASS: |
364 |
|
|
365 |
|
|
366 |
|
|
367 |
/* Character string printing function. */ |
/* Character string printing function. A "normal" and a UTF-8 version. */ |
368 |
|
|
369 |
static void pchars(unsigned char *p, int length) |
static void pchars(unsigned char *p, int length, int utf8) |
370 |
{ |
{ |
371 |
int c; |
int c; |
372 |
while (length-- > 0) |
while (length-- > 0) |
373 |
|
{ |
374 |
|
if (utf8) |
375 |
|
{ |
376 |
|
int rc = utf82ord(p, &c); |
377 |
|
if (rc > 0) |
378 |
|
{ |
379 |
|
length -= rc - 1; |
380 |
|
p += rc; |
381 |
|
if (c < 256 && isprint(c)) fprintf(outfile, "%c", c); |
382 |
|
else fprintf(outfile, "\\x{%02x}", c); |
383 |
|
continue; |
384 |
|
} |
385 |
|
} |
386 |
|
|
387 |
|
/* Not UTF-8, or malformed UTF-8 */ |
388 |
|
|
389 |
if (isprint(c = *(p++))) fprintf(outfile, "%c", c); |
if (isprint(c = *(p++))) fprintf(outfile, "%c", c); |
390 |
else fprintf(outfile, "\\x%02x", c); |
else fprintf(outfile, "\\x%02x", c); |
391 |
|
} |
392 |
} |
} |
393 |
|
|
394 |
|
|
398 |
|
|
399 |
static void *new_malloc(size_t size) |
static void *new_malloc(size_t size) |
400 |
{ |
{ |
401 |
|
gotten_store = size; |
402 |
if (log_store) |
if (log_store) |
403 |
fprintf(outfile, "Memory allocation (code space): %d\n", |
fprintf(outfile, "Memory allocation (code space): %d\n", |
404 |
(int)((int)size - offsetof(real_pcre, code[0]))); |
(int)((int)size - offsetof(real_pcre, code[0]))); |
407 |
|
|
408 |
|
|
409 |
|
|
410 |
|
|
411 |
|
/* Get one piece of information from the pcre_fullinfo() function */ |
412 |
|
|
413 |
|
static void new_info(pcre *re, pcre_extra *study, int option, void *ptr) |
414 |
|
{ |
415 |
|
int rc; |
416 |
|
if ((rc = pcre_fullinfo(re, study, option, ptr)) < 0) |
417 |
|
fprintf(outfile, "Error %d from pcre_fullinfo(%d)\n", rc, option); |
418 |
|
} |
419 |
|
|
420 |
|
|
421 |
|
|
422 |
|
|
423 |
/* 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 |
424 |
consist of a regular expression, in delimiters and optionally followed by |
consist of a regular expression, in delimiters and optionally followed by |
425 |
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 timeit = 0; |
int timeit = 0; |
434 |
int showinfo = 0; |
int showinfo = 0; |
435 |
int showstore = 0; |
int showstore = 0; |
436 |
|
int size_offsets = 45; |
437 |
|
int size_offsets_max; |
438 |
|
int *offsets; |
439 |
|
#if !defined NOPOSIX |
440 |
int posix = 0; |
int posix = 0; |
441 |
|
#endif |
442 |
int debug = 0; |
int debug = 0; |
443 |
int done = 0; |
int done = 0; |
444 |
unsigned char buffer[30000]; |
unsigned char buffer[30000]; |
452 |
|
|
453 |
while (argc > 1 && argv[op][0] == '-') |
while (argc > 1 && argv[op][0] == '-') |
454 |
{ |
{ |
455 |
|
char *endptr; |
456 |
|
|
457 |
if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0) |
if (strcmp(argv[op], "-s") == 0 || strcmp(argv[op], "-m") == 0) |
458 |
showstore = 1; |
showstore = 1; |
459 |
else if (strcmp(argv[op], "-t") == 0) timeit = 1; |
else if (strcmp(argv[op], "-t") == 0) timeit = 1; |
460 |
else if (strcmp(argv[op], "-i") == 0) showinfo = 1; |
else if (strcmp(argv[op], "-i") == 0) showinfo = 1; |
461 |
else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1; |
else if (strcmp(argv[op], "-d") == 0) showinfo = debug = 1; |
462 |
|
else if (strcmp(argv[op], "-o") == 0 && argc > 2 && |
463 |
|
((size_offsets = strtoul(argv[op+1], &endptr, 10)), *endptr == 0)) |
464 |
|
{ |
465 |
|
op++; |
466 |
|
argc--; |
467 |
|
} |
468 |
|
#if !defined NOPOSIX |
469 |
else if (strcmp(argv[op], "-p") == 0) posix = 1; |
else if (strcmp(argv[op], "-p") == 0) posix = 1; |
470 |
|
#endif |
471 |
else |
else |
472 |
{ |
{ |
473 |
printf("*** Unknown option %s\n", argv[op]); |
printf("** Unknown or malformed option %s\n", argv[op]); |
474 |
printf("Usage: pcretest [-d] [-i] [-p] [-s] [-t] [<input> [<output>]]\n"); |
printf("Usage: pcretest [-d] [-i] [-o <n>] [-p] [-s] [-t] [<input> [<output>]]\n"); |
475 |
printf(" -d debug: show compiled code; implies -i\n" |
printf(" -d debug: show compiled code; implies -i\n" |
476 |
" -i show information about compiled pattern\n" |
" -i show information about compiled pattern\n" |
477 |
" -p use POSIX interface\n" |
" -o <n> set size of offsets vector to <n>\n"); |
478 |
" -s output store information\n" |
#if !defined NOPOSIX |
479 |
" -t time compilation and execution\n"); |
printf(" -p use POSIX interface\n"); |
480 |
|
#endif |
481 |
|
printf(" -s output store information\n" |
482 |
|
" -t time compilation and execution\n"); |
483 |
return 1; |
return 1; |
484 |
} |
} |
485 |
op++; |
op++; |
486 |
argc--; |
argc--; |
487 |
} |
} |
488 |
|
|
489 |
|
/* Get the store for the offsets vector, and remember what it was */ |
490 |
|
|
491 |
|
size_offsets_max = size_offsets; |
492 |
|
offsets = malloc(size_offsets_max * sizeof(int)); |
493 |
|
if (offsets == NULL) |
494 |
|
{ |
495 |
|
printf("** Failed to get %d bytes of memory for offsets vector\n", |
496 |
|
size_offsets_max * sizeof(int)); |
497 |
|
return 1; |
498 |
|
} |
499 |
|
|
500 |
/* Sort out the input and output files */ |
/* Sort out the input and output files */ |
501 |
|
|
502 |
if (argc > 1) |
if (argc > 1) |
536 |
|
|
537 |
#if !defined NOPOSIX /* There are still compilers that require no indent */ |
#if !defined NOPOSIX /* There are still compilers that require no indent */ |
538 |
regex_t preg; |
regex_t preg; |
539 |
|
int do_posix = 0; |
540 |
#endif |
#endif |
541 |
|
|
542 |
const char *error; |
const char *error; |
543 |
unsigned char *p, *pp, *ppp; |
unsigned char *p, *pp, *ppp; |
544 |
unsigned const char *tables = NULL; |
const unsigned char *tables = NULL; |
545 |
int do_study = 0; |
int do_study = 0; |
546 |
int do_debug = debug; |
int do_debug = debug; |
547 |
int do_G = 0; |
int do_G = 0; |
548 |
int do_g = 0; |
int do_g = 0; |
549 |
int do_showinfo = showinfo; |
int do_showinfo = showinfo; |
550 |
int do_showrest = 0; |
int do_showrest = 0; |
551 |
int do_posix = 0; |
int utf8 = 0; |
552 |
int erroroffset, len, delimiter; |
int erroroffset, len, delimiter; |
553 |
|
|
554 |
if (infile == stdin) printf(" re> "); |
if (infile == stdin) printf(" re> "); |
640 |
case 'S': do_study = 1; break; |
case 'S': do_study = 1; break; |
641 |
case 'U': options |= PCRE_UNGREEDY; break; |
case 'U': options |= PCRE_UNGREEDY; break; |
642 |
case 'X': options |= PCRE_EXTRA; break; |
case 'X': options |= PCRE_EXTRA; break; |
643 |
|
case '8': options |= PCRE_UTF8; utf8 = 1; break; |
644 |
|
|
645 |
case 'L': |
case 'L': |
646 |
ppp = pp; |
ppp = pp; |
735 |
goto CONTINUE; |
goto CONTINUE; |
736 |
} |
} |
737 |
|
|
738 |
/* Compilation succeeded; print data if required */ |
/* Compilation succeeded; print data if required. There are now two |
739 |
|
info-returning functions. The old one has a limited interface and |
740 |
|
returns only limited data. Check that it agrees with the newer one. */ |
741 |
|
|
742 |
if (do_showinfo) |
if (do_showinfo) |
743 |
{ |
{ |
744 |
int first_char, count; |
unsigned long int get_options; |
745 |
|
int old_first_char, old_options, old_count; |
746 |
|
int count, backrefmax, first_char, need_char; |
747 |
|
size_t size; |
748 |
|
|
749 |
if (do_debug) print_internals(re); |
if (do_debug) print_internals(re); |
750 |
|
|
751 |
count = pcre_info(re, &options, &first_char); |
new_info(re, NULL, PCRE_INFO_OPTIONS, &get_options); |
752 |
|
new_info(re, NULL, PCRE_INFO_SIZE, &size); |
753 |
|
new_info(re, NULL, PCRE_INFO_CAPTURECOUNT, &count); |
754 |
|
new_info(re, NULL, PCRE_INFO_BACKREFMAX, &backrefmax); |
755 |
|
new_info(re, NULL, PCRE_INFO_FIRSTCHAR, &first_char); |
756 |
|
new_info(re, NULL, PCRE_INFO_LASTLITERAL, &need_char); |
757 |
|
|
758 |
|
old_count = pcre_info(re, &old_options, &old_first_char); |
759 |
if (count < 0) fprintf(outfile, |
if (count < 0) fprintf(outfile, |
760 |
"Error %d while reading info\n", count); |
"Error %d from pcre_info()\n", count); |
761 |
else |
else |
762 |
{ |
{ |
763 |
fprintf(outfile, "Identifying subpattern count = %d\n", count); |
if (old_count != count) fprintf(outfile, |
764 |
if (options == 0) fprintf(outfile, "No options\n"); |
"Count disagreement: pcre_fullinfo=%d pcre_info=%d\n", count, |
765 |
else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s\n", |
old_count); |
766 |
((options & PCRE_ANCHORED) != 0)? " anchored" : "", |
|
767 |
((options & PCRE_CASELESS) != 0)? " caseless" : "", |
if (old_first_char != first_char) fprintf(outfile, |
768 |
((options & PCRE_EXTENDED) != 0)? " extended" : "", |
"First char disagreement: pcre_fullinfo=%d pcre_info=%d\n", |
769 |
((options & PCRE_MULTILINE) != 0)? " multiline" : "", |
first_char, old_first_char); |
770 |
((options & PCRE_DOTALL) != 0)? " dotall" : "", |
|
771 |
((options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
if (old_options != (int)get_options) fprintf(outfile, |
772 |
((options & PCRE_EXTRA) != 0)? " extra" : "", |
"Options disagreement: pcre_fullinfo=%ld pcre_info=%d\n", |
773 |
((options & PCRE_UNGREEDY) != 0)? " ungreedy" : ""); |
get_options, old_options); |
774 |
|
} |
|
if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0) |
|
|
fprintf(outfile, "Case state changes\n"); |
|
775 |
|
|
776 |
if (first_char == -1) |
if (size != gotten_store) fprintf(outfile, |
777 |
{ |
"Size disagreement: pcre_fullinfo=%d call to malloc for %d\n", |
778 |
fprintf(outfile, "First char at start or follows \\n\n"); |
size, gotten_store); |
779 |
} |
|
780 |
else if (first_char < 0) |
fprintf(outfile, "Capturing subpattern count = %d\n", count); |
781 |
{ |
if (backrefmax > 0) |
782 |
fprintf(outfile, "No first char\n"); |
fprintf(outfile, "Max back reference = %d\n", backrefmax); |
783 |
} |
if (get_options == 0) fprintf(outfile, "No options\n"); |
784 |
|
else fprintf(outfile, "Options:%s%s%s%s%s%s%s%s%s\n", |
785 |
|
((get_options & PCRE_ANCHORED) != 0)? " anchored" : "", |
786 |
|
((get_options & PCRE_CASELESS) != 0)? " caseless" : "", |
787 |
|
((get_options & PCRE_EXTENDED) != 0)? " extended" : "", |
788 |
|
((get_options & PCRE_MULTILINE) != 0)? " multiline" : "", |
789 |
|
((get_options & PCRE_DOTALL) != 0)? " dotall" : "", |
790 |
|
((get_options & PCRE_DOLLAR_ENDONLY) != 0)? " dollar_endonly" : "", |
791 |
|
((get_options & PCRE_EXTRA) != 0)? " extra" : "", |
792 |
|
((get_options & PCRE_UNGREEDY) != 0)? " ungreedy" : "", |
793 |
|
((get_options & PCRE_UTF8) != 0)? " utf8" : ""); |
794 |
|
|
795 |
|
if (((((real_pcre *)re)->options) & PCRE_ICHANGED) != 0) |
796 |
|
fprintf(outfile, "Case state changes\n"); |
797 |
|
|
798 |
|
if (first_char == -1) |
799 |
|
{ |
800 |
|
fprintf(outfile, "First char at start or follows \\n\n"); |
801 |
|
} |
802 |
|
else if (first_char < 0) |
803 |
|
{ |
804 |
|
fprintf(outfile, "No first char\n"); |
805 |
|
} |
806 |
|
else |
807 |
|
{ |
808 |
|
if (isprint(first_char)) |
809 |
|
fprintf(outfile, "First char = \'%c\'\n", first_char); |
810 |
else |
else |
811 |
{ |
fprintf(outfile, "First char = %d\n", first_char); |
812 |
if (isprint(first_char)) |
} |
|
fprintf(outfile, "First char = \'%c\'\n", first_char); |
|
|
else |
|
|
fprintf(outfile, "First char = %d\n", first_char); |
|
|
} |
|
813 |
|
|
814 |
if (((((real_pcre *)re)->options) & PCRE_REQCHSET) != 0) |
if (need_char < 0) |
815 |
{ |
{ |
816 |
int req_char = ((real_pcre *)re)->req_char; |
fprintf(outfile, "No need char\n"); |
817 |
if (isprint(req_char)) |
} |
818 |
fprintf(outfile, "Req char = \'%c\'\n", req_char); |
else |
819 |
else |
{ |
820 |
fprintf(outfile, "Req char = %d\n", req_char); |
if (isprint(need_char)) |
821 |
} |
fprintf(outfile, "Need char = \'%c\'\n", need_char); |
822 |
else fprintf(outfile, "No req char\n"); |
else |
823 |
|
fprintf(outfile, "Need char = %d\n", need_char); |
824 |
} |
} |
825 |
} |
} |
826 |
|
|
849 |
else if (extra == NULL) |
else if (extra == NULL) |
850 |
fprintf(outfile, "Study returned NULL\n"); |
fprintf(outfile, "Study returned NULL\n"); |
851 |
|
|
|
/* This looks at internal information. A bit kludgy to do it this |
|
|
way, but it is useful for testing. */ |
|
|
|
|
852 |
else if (do_showinfo) |
else if (do_showinfo) |
853 |
{ |
{ |
854 |
real_pcre_extra *xx = (real_pcre_extra *)extra; |
uschar *start_bits = NULL; |
855 |
if ((xx->options & PCRE_STUDY_MAPPED) == 0) |
new_info(re, extra, PCRE_INFO_FIRSTTABLE, &start_bits); |
856 |
|
if (start_bits == NULL) |
857 |
fprintf(outfile, "No starting character set\n"); |
fprintf(outfile, "No starting character set\n"); |
858 |
else |
else |
859 |
{ |
{ |
862 |
fprintf(outfile, "Starting character set: "); |
fprintf(outfile, "Starting character set: "); |
863 |
for (i = 0; i < 256; i++) |
for (i = 0; i < 256; i++) |
864 |
{ |
{ |
865 |
if ((xx->start_bits[i/8] & (1<<(i%8))) != 0) |
if ((start_bits[i/8] & (1<<(i%8))) != 0) |
866 |
{ |
{ |
867 |
if (c > 75) |
if (c > 75) |
868 |
{ |
{ |
893 |
{ |
{ |
894 |
unsigned char *q; |
unsigned char *q; |
895 |
unsigned char *bptr = dbuffer; |
unsigned char *bptr = dbuffer; |
896 |
|
int use_size_offsets = size_offsets; |
897 |
int count, c; |
int count, c; |
898 |
int copystrings = 0; |
int copystrings = 0; |
899 |
int getstrings = 0; |
int getstrings = 0; |
900 |
int getlist = 0; |
int getlist = 0; |
901 |
|
int gmatched = 0; |
902 |
int start_offset = 0; |
int start_offset = 0; |
903 |
int offsets[45]; |
int g_notempty = 0; |
|
int size_offsets = sizeof(offsets)/sizeof(int); |
|
904 |
|
|
905 |
options = 0; |
options = 0; |
906 |
|
|
944 |
break; |
break; |
945 |
|
|
946 |
case 'x': |
case 'x': |
947 |
|
|
948 |
|
/* Handle \x{..} specially - new Perl thing for utf8 */ |
949 |
|
|
950 |
|
if (*p == '{') |
951 |
|
{ |
952 |
|
unsigned char *pt = p; |
953 |
|
c = 0; |
954 |
|
while (isxdigit(*(++pt))) |
955 |
|
c = c * 16 + tolower(*pt) - ((isdigit(*pt))? '0' : 'W'); |
956 |
|
if (*pt == '}') |
957 |
|
{ |
958 |
|
unsigned char buffer[8]; |
959 |
|
int ii, utn; |
960 |
|
utn = ord2utf8(c, buffer); |
961 |
|
for (ii = 0; ii < utn - 1; ii++) *q++ = buffer[ii]; |
962 |
|
c = buffer[ii]; /* Last byte */ |
963 |
|
p = pt + 1; |
964 |
|
break; |
965 |
|
} |
966 |
|
/* Not correct form; fall through */ |
967 |
|
} |
968 |
|
|
969 |
|
/* Ordinary \x */ |
970 |
|
|
971 |
c = 0; |
c = 0; |
972 |
while (i++ < 2 && isxdigit(*p)) |
while (i++ < 2 && isxdigit(*p)) |
973 |
{ |
{ |
1008 |
|
|
1009 |
case 'O': |
case 'O': |
1010 |
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
while(isdigit(*p)) n = n * 10 + *p++ - '0'; |
1011 |
if (n <= (int)(sizeof(offsets)/sizeof(int))) size_offsets = n; |
if (n > size_offsets_max) |
1012 |
|
{ |
1013 |
|
free(offsets); |
1014 |
|
size_offsets_max = n; |
1015 |
|
offsets = malloc(size_offsets_max * sizeof(int)); |
1016 |
|
if (offsets == NULL) |
1017 |
|
{ |
1018 |
|
printf("** Failed to get %d bytes of memory for offsets vector\n", |
1019 |
|
size_offsets_max * sizeof(int)); |
1020 |
|
return 1; |
1021 |
|
} |
1022 |
|
} |
1023 |
|
use_size_offsets = n; |
1024 |
continue; |
continue; |
1025 |
|
|
1026 |
case 'Z': |
case 'Z': |
1040 |
{ |
{ |
1041 |
int rc; |
int rc; |
1042 |
int eflags = 0; |
int eflags = 0; |
1043 |
regmatch_t pmatch[30]; |
regmatch_t *pmatch = malloc(sizeof(regmatch_t) * use_size_offsets); |
1044 |
if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL; |
if ((options & PCRE_NOTBOL) != 0) eflags |= REG_NOTBOL; |
1045 |
if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL; |
if ((options & PCRE_NOTEOL) != 0) eflags |= REG_NOTEOL; |
1046 |
|
|
1047 |
rc = regexec(&preg, (const char *)bptr, |
rc = regexec(&preg, (const char *)bptr, use_size_offsets, pmatch, eflags); |
|
sizeof(pmatch)/sizeof(regmatch_t), pmatch, eflags); |
|
1048 |
|
|
1049 |
if (rc != 0) |
if (rc != 0) |
1050 |
{ |
{ |
1054 |
else |
else |
1055 |
{ |
{ |
1056 |
size_t i; |
size_t i; |
1057 |
for (i = 0; i < sizeof(pmatch)/sizeof(regmatch_t); i++) |
for (i = 0; i < use_size_offsets; i++) |
1058 |
{ |
{ |
1059 |
if (pmatch[i].rm_so >= 0) |
if (pmatch[i].rm_so >= 0) |
1060 |
{ |
{ |
1061 |
fprintf(outfile, "%2d: ", (int)i); |
fprintf(outfile, "%2d: ", (int)i); |
1062 |
pchars(dbuffer + pmatch[i].rm_so, |
pchars(dbuffer + pmatch[i].rm_so, |
1063 |
pmatch[i].rm_eo - pmatch[i].rm_so); |
pmatch[i].rm_eo - pmatch[i].rm_so, utf8); |
1064 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
1065 |
if (i == 0 && do_showrest) |
if (i == 0 && do_showrest) |
1066 |
{ |
{ |
1067 |
fprintf(outfile, " 0+ "); |
fprintf(outfile, " 0+ "); |
1068 |
pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo); |
pchars(dbuffer + pmatch[i].rm_eo, len - pmatch[i].rm_eo, utf8); |
1069 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
1070 |
} |
} |
1071 |
} |
} |
1072 |
} |
} |
1073 |
} |
} |
1074 |
|
free(pmatch); |
1075 |
} |
} |
1076 |
|
|
1077 |
/* Handle matching via the native interface - repeats for /g and /G */ |
/* Handle matching via the native interface - repeats for /g and /G */ |
1079 |
else |
else |
1080 |
#endif /* !defined NOPOSIX */ |
#endif /* !defined NOPOSIX */ |
1081 |
|
|
1082 |
for (;;) |
for (;; gmatched++) /* Loop for /g or /G */ |
1083 |
{ |
{ |
1084 |
if (timeit) |
if (timeit) |
1085 |
{ |
{ |
1088 |
clock_t start_time = clock(); |
clock_t start_time = clock(); |
1089 |
for (i = 0; i < LOOPREPEAT; i++) |
for (i = 0; i < LOOPREPEAT; i++) |
1090 |
count = pcre_exec(re, extra, (char *)bptr, len, |
count = pcre_exec(re, extra, (char *)bptr, len, |
1091 |
(do_g? start_offset : 0), options, offsets, size_offsets); |
start_offset, options | g_notempty, offsets, use_size_offsets); |
1092 |
time_taken = clock() - start_time; |
time_taken = clock() - start_time; |
1093 |
fprintf(outfile, "Execute time %.3f milliseconds\n", |
fprintf(outfile, "Execute time %.3f milliseconds\n", |
1094 |
((double)time_taken * 1000.0)/ |
((double)time_taken * 1000.0)/ |
1096 |
} |
} |
1097 |
|
|
1098 |
count = pcre_exec(re, extra, (char *)bptr, len, |
count = pcre_exec(re, extra, (char *)bptr, len, |
1099 |
(do_g? start_offset : 0), options, offsets, size_offsets); |
start_offset, options | g_notempty, offsets, use_size_offsets); |
1100 |
|
|
1101 |
if (count == 0) |
if (count == 0) |
1102 |
{ |
{ |
1103 |
fprintf(outfile, "Matched, but too many substrings\n"); |
fprintf(outfile, "Matched, but too many substrings\n"); |
1104 |
count = size_offsets/3; |
count = use_size_offsets/3; |
1105 |
} |
} |
1106 |
|
|
1107 |
|
/* Matched */ |
1108 |
|
|
1109 |
if (count >= 0) |
if (count >= 0) |
1110 |
{ |
{ |
1111 |
int i; |
int i; |
1116 |
else |
else |
1117 |
{ |
{ |
1118 |
fprintf(outfile, "%2d: ", i/2); |
fprintf(outfile, "%2d: ", i/2); |
1119 |
pchars(bptr + offsets[i], offsets[i+1] - offsets[i]); |
pchars(bptr + offsets[i], offsets[i+1] - offsets[i], utf8); |
1120 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
1121 |
if (i == 0) |
if (i == 0) |
1122 |
{ |
{ |
|
start_offset = offsets[1]; |
|
1123 |
if (do_showrest) |
if (do_showrest) |
1124 |
{ |
{ |
1125 |
fprintf(outfile, " 0+ "); |
fprintf(outfile, " 0+ "); |
1126 |
pchars(bptr + offsets[i+1], len - offsets[i+1]); |
pchars(bptr + offsets[i+1], len - offsets[i+1], utf8); |
1127 |
fprintf(outfile, "\n"); |
fprintf(outfile, "\n"); |
1128 |
} |
} |
1129 |
} |
} |
1156 |
else |
else |
1157 |
{ |
{ |
1158 |
fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc); |
fprintf(outfile, "%2dG %s (%d)\n", i, substring, rc); |
1159 |
free((void *)substring); |
/* free((void *)substring); */ |
1160 |
|
pcre_free_substring(substring); |
1161 |
} |
} |
1162 |
} |
} |
1163 |
} |
} |
1175 |
fprintf(outfile, "%2dL %s\n", i, stringlist[i]); |
fprintf(outfile, "%2dL %s\n", i, stringlist[i]); |
1176 |
if (stringlist[i] != NULL) |
if (stringlist[i] != NULL) |
1177 |
fprintf(outfile, "string list not terminated by NULL\n"); |
fprintf(outfile, "string list not terminated by NULL\n"); |
1178 |
free((void *)stringlist); |
/* free((void *)stringlist); */ |
1179 |
|
pcre_free_substring_list(stringlist); |
1180 |
} |
} |
1181 |
} |
} |
|
|
|
1182 |
} |
} |
1183 |
|
|
1184 |
|
/* Failed to match. If this is a /g or /G loop and we previously set |
1185 |
|
g_notempty after a null match, this is not necessarily the end. |
1186 |
|
We want to advance the start offset, and continue. Fudge the offset |
1187 |
|
values to achieve this. We won't be at the end of the string - that |
1188 |
|
was checked before setting g_notempty. */ |
1189 |
|
|
1190 |
else |
else |
1191 |
{ |
{ |
1192 |
if (start_offset == 0) |
if (g_notempty != 0) |
1193 |
{ |
{ |
1194 |
if (count == -1) fprintf(outfile, "No match\n"); |
offsets[0] = start_offset; |
1195 |
else fprintf(outfile, "Error %d\n", count); |
offsets[1] = start_offset + 1; |
1196 |
} |
} |
1197 |
start_offset = -1; |
else |
1198 |
|
{ |
1199 |
|
if (gmatched == 0) /* Error if no previous matches */ |
1200 |
|
{ |
1201 |
|
if (count == -1) fprintf(outfile, "No match\n"); |
1202 |
|
else fprintf(outfile, "Error %d\n", count); |
1203 |
|
} |
1204 |
|
break; /* Out of the /g loop */ |
1205 |
|
} |
1206 |
|
} |
1207 |
|
|
1208 |
|
/* If not /g or /G we are done */ |
1209 |
|
|
1210 |
|
if (!do_g && !do_G) break; |
1211 |
|
|
1212 |
|
/* If we have matched an empty string, first check to see if we are at |
1213 |
|
the end of the subject. If so, the /g loop is over. Otherwise, mimic |
1214 |
|
what Perl's /g options does. This turns out to be rather cunning. First |
1215 |
|
we set PCRE_NOTEMPTY and PCRE_ANCHORED and try the match again at the |
1216 |
|
same point. If this fails (picked up above) we advance to the next |
1217 |
|
character. */ |
1218 |
|
|
1219 |
|
g_notempty = 0; |
1220 |
|
if (offsets[0] == offsets[1]) |
1221 |
|
{ |
1222 |
|
if (offsets[0] == len) break; |
1223 |
|
g_notempty = PCRE_NOTEMPTY | PCRE_ANCHORED; |
1224 |
} |
} |
1225 |
|
|
1226 |
if ((!do_g && !do_G) || start_offset <= 0) break; |
/* For /g, update the start offset, leaving the rest alone */ |
1227 |
if (do_G) |
|
1228 |
|
if (do_g) start_offset = offsets[1]; |
1229 |
|
|
1230 |
|
/* For /G, update the pointer and length */ |
1231 |
|
|
1232 |
|
else |
1233 |
{ |
{ |
1234 |
bptr += start_offset; |
bptr += offsets[1]; |
1235 |
len -= start_offset; |
len -= offsets[1]; |
1236 |
} |
} |
1237 |
} |
} /* End of loop for /g and /G */ |
1238 |
} |
} /* End of loop for data lines */ |
1239 |
|
|
1240 |
CONTINUE: |
CONTINUE: |
1241 |
|
|