42 |
supporting internal functions that are not used by other modules. */ |
supporting internal functions that are not used by other modules. */ |
43 |
|
|
44 |
|
|
45 |
|
#ifdef HAVE_CONFIG_H |
46 |
|
#include <config.h> |
47 |
|
#endif |
48 |
|
|
49 |
#define NLBLOCK cd /* Block containing newline information */ |
#define NLBLOCK cd /* Block containing newline information */ |
50 |
#define PSSTART start_pattern /* Field containing processed string start */ |
#define PSSTART start_pattern /* Field containing processed string start */ |
51 |
#define PSEND end_pattern /* Field containing processed string end */ |
#define PSEND end_pattern /* Field containing processed string end */ |
52 |
|
|
|
|
|
53 |
#include "pcre_internal.h" |
#include "pcre_internal.h" |
54 |
|
|
55 |
|
|
61 |
#endif |
#endif |
62 |
|
|
63 |
|
|
64 |
|
/* Macro for setting individual bits in class bitmaps. */ |
65 |
|
|
66 |
|
#define SETBIT(a,b) a[b/8] |= (1 << (b%8)) |
67 |
|
|
68 |
|
/* Maximum length value to check against when making sure that the integer that |
69 |
|
holds the compiled pattern length does not overflow. We make it a bit less than |
70 |
|
INT_MAX to allow for adding in group terminating bytes, so that we don't have |
71 |
|
to check them every time. */ |
72 |
|
|
73 |
|
#define OFLOW_MAX (INT_MAX - 20) |
74 |
|
|
75 |
|
|
76 |
/************************************************* |
/************************************************* |
77 |
* Code parameters and static tables * |
* Code parameters and static tables * |
78 |
*************************************************/ |
*************************************************/ |
102 |
0, 0, 0, 0, 0, 0, 0, 0, /* 0 - 7 */ |
0, 0, 0, 0, 0, 0, 0, 0, /* 0 - 7 */ |
103 |
0, 0, ':', ';', '<', '=', '>', '?', /* 8 - ? */ |
0, 0, ':', ';', '<', '=', '>', '?', /* 8 - ? */ |
104 |
'@', -ESC_A, -ESC_B, -ESC_C, -ESC_D, -ESC_E, 0, -ESC_G, /* @ - G */ |
'@', -ESC_A, -ESC_B, -ESC_C, -ESC_D, -ESC_E, 0, -ESC_G, /* @ - G */ |
105 |
0, 0, 0, 0, 0, 0, 0, 0, /* H - O */ |
-ESC_H, 0, 0, -ESC_K, 0, 0, 0, 0, /* H - O */ |
106 |
-ESC_P, -ESC_Q, -ESC_R, -ESC_S, 0, 0, 0, -ESC_W, /* P - W */ |
-ESC_P, -ESC_Q, -ESC_R, -ESC_S, 0, 0, -ESC_V, -ESC_W, /* P - W */ |
107 |
-ESC_X, 0, -ESC_Z, '[', '\\', ']', '^', '_', /* X - _ */ |
-ESC_X, 0, -ESC_Z, '[', '\\', ']', '^', '_', /* X - _ */ |
108 |
'`', 7, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0, /* ` - g */ |
'`', 7, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0, /* ` - g */ |
109 |
0, 0, 0, -ESC_k, 0, 0, ESC_n, 0, /* h - o */ |
-ESC_h, 0, 0, -ESC_k, 0, 0, ESC_n, 0, /* h - o */ |
110 |
-ESC_p, 0, ESC_r, -ESC_s, ESC_tee, 0, 0, -ESC_w, /* p - w */ |
-ESC_p, 0, ESC_r, -ESC_s, ESC_tee, 0, -ESC_v, -ESC_w, /* p - w */ |
111 |
0, 0, -ESC_z /* x - z */ |
0, 0, -ESC_z /* x - z */ |
112 |
}; |
}; |
113 |
|
|
121 |
/* 70 */ 0, 0, 0, 0, 0, 0, 0, 0, |
/* 70 */ 0, 0, 0, 0, 0, 0, 0, 0, |
122 |
/* 78 */ 0, '`', ':', '#', '@', '\'', '=', '"', |
/* 78 */ 0, '`', ':', '#', '@', '\'', '=', '"', |
123 |
/* 80 */ 0, 7, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0, |
/* 80 */ 0, 7, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0, |
124 |
/* 88 */ 0, 0, 0, '{', 0, 0, 0, 0, |
/* 88 */-ESC_h, 0, 0, '{', 0, 0, 0, 0, |
125 |
/* 90 */ 0, 0, -ESC_k, 'l', 0, ESC_n, 0, -ESC_p, |
/* 90 */ 0, 0, -ESC_k, 'l', 0, ESC_n, 0, -ESC_p, |
126 |
/* 98 */ 0, ESC_r, 0, '}', 0, 0, 0, 0, |
/* 98 */ 0, ESC_r, 0, '}', 0, 0, 0, 0, |
127 |
/* A0 */ 0, '~', -ESC_s, ESC_tee, 0, 0, -ESC_w, 0, |
/* A0 */ 0, '~', -ESC_s, ESC_tee, 0,-ESC_v, -ESC_w, 0, |
128 |
/* A8 */ 0,-ESC_z, 0, 0, 0, '[', 0, 0, |
/* A8 */ 0,-ESC_z, 0, 0, 0, '[', 0, 0, |
129 |
/* B0 */ 0, 0, 0, 0, 0, 0, 0, 0, |
/* B0 */ 0, 0, 0, 0, 0, 0, 0, 0, |
130 |
/* B8 */ 0, 0, 0, 0, 0, ']', '=', '-', |
/* B8 */ 0, 0, 0, 0, 0, ']', '=', '-', |
131 |
/* C0 */ '{',-ESC_A, -ESC_B, -ESC_C, -ESC_D,-ESC_E, 0, -ESC_G, |
/* C0 */ '{',-ESC_A, -ESC_B, -ESC_C, -ESC_D,-ESC_E, 0, -ESC_G, |
132 |
/* C8 */ 0, 0, 0, 0, 0, 0, 0, 0, |
/* C8 */-ESC_H, 0, 0, 0, 0, 0, 0, 0, |
133 |
/* D0 */ '}', 0, 0, 0, 0, 0, 0, -ESC_P, |
/* D0 */ '}', 0, -ESC_K, 0, 0, 0, 0, -ESC_P, |
134 |
/* D8 */-ESC_Q,-ESC_R, 0, 0, 0, 0, 0, 0, |
/* D8 */-ESC_Q,-ESC_R, 0, 0, 0, 0, 0, 0, |
135 |
/* E0 */ '\\', 0, -ESC_S, 0, 0, 0, -ESC_W, -ESC_X, |
/* E0 */ '\\', 0, -ESC_S, 0, 0,-ESC_V, -ESC_W, -ESC_X, |
136 |
/* E8 */ 0,-ESC_Z, 0, 0, 0, 0, 0, 0, |
/* E8 */ 0,-ESC_Z, 0, 0, 0, 0, 0, 0, |
137 |
/* F0 */ 0, 0, 0, 0, 0, 0, 0, 0, |
/* F0 */ 0, 0, 0, 0, 0, 0, 0, 0, |
138 |
/* F8 */ 0, 0, 0, 0, 0, 0, 0, 0 |
/* F8 */ 0, 0, 0, 0, 0, 0, 0, 0 |
140 |
#endif |
#endif |
141 |
|
|
142 |
|
|
143 |
|
/* Table of special "verbs" like (*PRUNE) */ |
144 |
|
|
145 |
|
typedef struct verbitem { |
146 |
|
const char *name; |
147 |
|
int len; |
148 |
|
int op; |
149 |
|
} verbitem; |
150 |
|
|
151 |
|
static verbitem verbs[] = { |
152 |
|
{ "ACCEPT", 6, OP_ACCEPT }, |
153 |
|
{ "COMMIT", 6, OP_COMMIT }, |
154 |
|
{ "F", 1, OP_FAIL }, |
155 |
|
{ "FAIL", 4, OP_FAIL }, |
156 |
|
{ "PRUNE", 5, OP_PRUNE }, |
157 |
|
{ "SKIP", 4, OP_SKIP }, |
158 |
|
{ "THEN", 4, OP_THEN } |
159 |
|
}; |
160 |
|
|
161 |
|
static int verbcount = sizeof(verbs)/sizeof(verbitem); |
162 |
|
|
163 |
|
|
164 |
/* Tables of names of POSIX character classes and their lengths. The list is |
/* Tables of names of POSIX character classes and their lengths. The list is |
165 |
terminated by a zero length entry. The first three must be alpha, lower, upper, |
terminated by a zero length entry. The first three must be alpha, lower, upper, |
166 |
as this is assumed for handling case independence. */ |
as this is assumed for handling case independence. */ |
234 |
"missing ) after comment", |
"missing ) after comment", |
235 |
"parentheses nested too deeply", /** DEAD **/ |
"parentheses nested too deeply", /** DEAD **/ |
236 |
/* 20 */ |
/* 20 */ |
237 |
"regular expression too large", |
"regular expression is too large", |
238 |
"failed to get memory", |
"failed to get memory", |
239 |
"unmatched parentheses", |
"unmatched parentheses", |
240 |
"internal error: code overflow", |
"internal error: code overflow", |
244 |
"malformed number or name after (?(", |
"malformed number or name after (?(", |
245 |
"conditional group contains more than two branches", |
"conditional group contains more than two branches", |
246 |
"assertion expected after (?(", |
"assertion expected after (?(", |
247 |
"(?R or (?digits must be followed by )", |
"(?R or (?[+-]digits must be followed by )", |
248 |
/* 30 */ |
/* 30 */ |
249 |
"unknown POSIX class name", |
"unknown POSIX class name", |
250 |
"POSIX collating elements are not supported", |
"POSIX collating elements are not supported", |
270 |
"subpattern name is too long (maximum " XSTRING(MAX_NAME_SIZE) " characters)", |
"subpattern name is too long (maximum " XSTRING(MAX_NAME_SIZE) " characters)", |
271 |
"too many named subpatterns (maximum " XSTRING(MAX_NAME_COUNT) ")", |
"too many named subpatterns (maximum " XSTRING(MAX_NAME_COUNT) ")", |
272 |
/* 50 */ |
/* 50 */ |
273 |
"repeated subpattern is too long", |
"repeated subpattern is too long", /** DEAD **/ |
274 |
"octal value is greater than \\377 (not in UTF-8 mode)", |
"octal value is greater than \\377 (not in UTF-8 mode)", |
275 |
"internal error: overran compiling workspace", |
"internal error: overran compiling workspace", |
276 |
"internal error: previously-checked referenced subpattern not found", |
"internal error: previously-checked referenced subpattern not found", |
278 |
/* 55 */ |
/* 55 */ |
279 |
"repeating a DEFINE group is not allowed", |
"repeating a DEFINE group is not allowed", |
280 |
"inconsistent NEWLINE options", |
"inconsistent NEWLINE options", |
281 |
"\\g is not followed by an (optionally braced) non-zero number" |
"\\g is not followed by a braced name or an optionally braced non-zero number", |
282 |
|
"(?+ or (?- or (?(+ or (?(- must be followed by a non-zero number", |
283 |
|
"(*VERB) with an argument is not supported", |
284 |
|
/* 60 */ |
285 |
|
"(*VERB) not recognized", |
286 |
|
"number is too big" |
287 |
}; |
}; |
288 |
|
|
289 |
|
|
414 |
/* Definition to allow mutual recursion */ |
/* Definition to allow mutual recursion */ |
415 |
|
|
416 |
static BOOL |
static BOOL |
417 |
compile_regex(int, int, uschar **, const uschar **, int *, BOOL, int, int *, |
compile_regex(int, int, uschar **, const uschar **, int *, BOOL, BOOL, int, |
418 |
int *, branch_chain *, compile_data *, int *); |
int *, int *, branch_chain *, compile_data *, int *); |
419 |
|
|
420 |
|
|
421 |
|
|
440 |
|
|
441 |
Returns: zero or positive => a data character |
Returns: zero or positive => a data character |
442 |
negative => a special escape sequence |
negative => a special escape sequence |
443 |
on error, errorptr is set |
on error, errorcodeptr is set |
444 |
*/ |
*/ |
445 |
|
|
446 |
static int |
static int |
493 |
|
|
494 |
/* \g must be followed by a number, either plain or braced. If positive, it |
/* \g must be followed by a number, either plain or braced. If positive, it |
495 |
is an absolute backreference. If negative, it is a relative backreference. |
is an absolute backreference. If negative, it is a relative backreference. |
496 |
This is a Perl 5.10 feature. */ |
This is a Perl 5.10 feature. Perl 5.10 also supports \g{name} as a |
497 |
|
reference to a named group. This is part of Perl's movement towards a |
498 |
|
unified syntax for back references. As this is synonymous with \k{name}, we |
499 |
|
fudge it up by pretending it really was \k. */ |
500 |
|
|
501 |
case 'g': |
case 'g': |
502 |
if (ptr[1] == '{') |
if (ptr[1] == '{') |
503 |
{ |
{ |
504 |
|
const uschar *p; |
505 |
|
for (p = ptr+2; *p != 0 && *p != '}'; p++) |
506 |
|
if (*p != '-' && (digitab[*p] & ctype_digit) == 0) break; |
507 |
|
if (*p != 0 && *p != '}') |
508 |
|
{ |
509 |
|
c = -ESC_k; |
510 |
|
break; |
511 |
|
} |
512 |
braced = TRUE; |
braced = TRUE; |
513 |
ptr++; |
ptr++; |
514 |
} |
} |
525 |
while ((digitab[ptr[1]] & ctype_digit) != 0) |
while ((digitab[ptr[1]] & ctype_digit) != 0) |
526 |
c = c * 10 + *(++ptr) - '0'; |
c = c * 10 + *(++ptr) - '0'; |
527 |
|
|
528 |
|
if (c < 0) |
529 |
|
{ |
530 |
|
*errorcodeptr = ERR61; |
531 |
|
break; |
532 |
|
} |
533 |
|
|
534 |
if (c == 0 || (braced && *(++ptr) != '}')) |
if (c == 0 || (braced && *(++ptr) != '}')) |
535 |
{ |
{ |
536 |
*errorcodeptr = ERR57; |
*errorcodeptr = ERR57; |
537 |
return 0; |
break; |
538 |
} |
} |
539 |
|
|
540 |
if (negated) |
if (negated) |
542 |
if (c > bracount) |
if (c > bracount) |
543 |
{ |
{ |
544 |
*errorcodeptr = ERR15; |
*errorcodeptr = ERR15; |
545 |
return 0; |
break; |
546 |
} |
} |
547 |
c = bracount - (c - 1); |
c = bracount - (c - 1); |
548 |
} |
} |
571 |
c -= '0'; |
c -= '0'; |
572 |
while ((digitab[ptr[1]] & ctype_digit) != 0) |
while ((digitab[ptr[1]] & ctype_digit) != 0) |
573 |
c = c * 10 + *(++ptr) - '0'; |
c = c * 10 + *(++ptr) - '0'; |
574 |
|
if (c < 0) |
575 |
|
{ |
576 |
|
*errorcodeptr = ERR61; |
577 |
|
break; |
578 |
|
} |
579 |
if (c < 10 || c <= bracount) |
if (c < 10 || c <= bracount) |
580 |
{ |
{ |
581 |
c = -(ESC_REF + c); |
c = -(ESC_REF + c); |
671 |
if (c == 0) |
if (c == 0) |
672 |
{ |
{ |
673 |
*errorcodeptr = ERR2; |
*errorcodeptr = ERR2; |
674 |
return 0; |
break; |
675 |
} |
} |
676 |
|
|
677 |
#ifndef EBCDIC /* ASCII coding */ |
#ifndef EBCDIC /* ASCII coding */ |
747 |
*negptr = TRUE; |
*negptr = TRUE; |
748 |
ptr++; |
ptr++; |
749 |
} |
} |
750 |
for (i = 0; i < sizeof(name) - 1; i++) |
for (i = 0; i < (int)sizeof(name) - 1; i++) |
751 |
{ |
{ |
752 |
c = *(++ptr); |
c = *(++ptr); |
753 |
if (c == 0) goto ERROR_RETURN; |
if (c == 0) goto ERROR_RETURN; |
950 |
{ |
{ |
951 |
while (*(++ptr) != ']') |
while (*(++ptr) != ']') |
952 |
{ |
{ |
953 |
|
if (*ptr == 0) return -1; |
954 |
if (*ptr == '\\') |
if (*ptr == '\\') |
955 |
{ |
{ |
956 |
if (*(++ptr) == 0) return -1; |
if (*(++ptr) == 0) return -1; |
978 |
/* An opening parens must now be a real metacharacter */ |
/* An opening parens must now be a real metacharacter */ |
979 |
|
|
980 |
if (*ptr != '(') continue; |
if (*ptr != '(') continue; |
981 |
if (ptr[1] != '?') |
if (ptr[1] != '?' && ptr[1] != '*') |
982 |
{ |
{ |
983 |
count++; |
count++; |
984 |
if (name == NULL && count == lorn) return count; |
if (name == NULL && count == lorn) return count; |
1106 |
{ |
{ |
1107 |
int d; |
int d; |
1108 |
register int op = *cc; |
register int op = *cc; |
|
|
|
1109 |
switch (op) |
switch (op) |
1110 |
{ |
{ |
1111 |
case OP_CBRA: |
case OP_CBRA: |
1194 |
|
|
1195 |
case OP_TYPEEXACT: |
case OP_TYPEEXACT: |
1196 |
branchlength += GET2(cc,1); |
branchlength += GET2(cc,1); |
1197 |
|
if (cc[3] == OP_PROP || cc[3] == OP_NOTPROP) cc += 2; |
1198 |
cc += 4; |
cc += 4; |
1199 |
break; |
break; |
1200 |
|
|
1303 |
code += _pcre_OP_lengths[c]; |
code += _pcre_OP_lengths[c]; |
1304 |
} |
} |
1305 |
|
|
1306 |
/* In UTF-8 mode, opcodes that are followed by a character may be followed by |
/* Otherwise, we can get the item's length from the table, except that for |
1307 |
a multi-byte character. The length in the table is a minimum, so we have to |
repeated character types, we have to test for \p and \P, which have an extra |
1308 |
arrange to skip the extra bytes. */ |
two bytes of parameters. */ |
1309 |
|
|
1310 |
else |
else |
1311 |
{ |
{ |
1312 |
|
switch(c) |
1313 |
|
{ |
1314 |
|
case OP_TYPESTAR: |
1315 |
|
case OP_TYPEMINSTAR: |
1316 |
|
case OP_TYPEPLUS: |
1317 |
|
case OP_TYPEMINPLUS: |
1318 |
|
case OP_TYPEQUERY: |
1319 |
|
case OP_TYPEMINQUERY: |
1320 |
|
case OP_TYPEUPTO: |
1321 |
|
case OP_TYPEMINUPTO: |
1322 |
|
case OP_TYPEEXACT: |
1323 |
|
case OP_TYPEPOSSTAR: |
1324 |
|
case OP_TYPEPOSPLUS: |
1325 |
|
case OP_TYPEPOSQUERY: |
1326 |
|
case OP_TYPEPOSUPTO: |
1327 |
|
if (code[1] == OP_PROP || code[1] == OP_NOTPROP) code += 2; |
1328 |
|
break; |
1329 |
|
} |
1330 |
|
|
1331 |
|
/* Add in the fixed length from the table */ |
1332 |
|
|
1333 |
code += _pcre_OP_lengths[c]; |
code += _pcre_OP_lengths[c]; |
1334 |
|
|
1335 |
|
/* In UTF-8 mode, opcodes that are followed by a character may be followed by |
1336 |
|
a multi-byte character. The length in the table is a minimum, so we have to |
1337 |
|
arrange to skip the extra bytes. */ |
1338 |
|
|
1339 |
#ifdef SUPPORT_UTF8 |
#ifdef SUPPORT_UTF8 |
1340 |
if (utf8) switch(c) |
if (utf8) switch(c) |
1341 |
{ |
{ |
1393 |
|
|
1394 |
if (c == OP_XCLASS) code += GET(code, 1); |
if (c == OP_XCLASS) code += GET(code, 1); |
1395 |
|
|
1396 |
/* Otherwise, we get the item's length from the table. In UTF-8 mode, opcodes |
/* Otherwise, we can get the item's length from the table, except that for |
1397 |
that are followed by a character may be followed by a multi-byte character. |
repeated character types, we have to test for \p and \P, which have an extra |
1398 |
The length in the table is a minimum, so we have to arrange to skip the extra |
two bytes of parameters. */ |
|
bytes. */ |
|
1399 |
|
|
1400 |
else |
else |
1401 |
{ |
{ |
1402 |
|
switch(c) |
1403 |
|
{ |
1404 |
|
case OP_TYPESTAR: |
1405 |
|
case OP_TYPEMINSTAR: |
1406 |
|
case OP_TYPEPLUS: |
1407 |
|
case OP_TYPEMINPLUS: |
1408 |
|
case OP_TYPEQUERY: |
1409 |
|
case OP_TYPEMINQUERY: |
1410 |
|
case OP_TYPEUPTO: |
1411 |
|
case OP_TYPEMINUPTO: |
1412 |
|
case OP_TYPEEXACT: |
1413 |
|
case OP_TYPEPOSSTAR: |
1414 |
|
case OP_TYPEPOSPLUS: |
1415 |
|
case OP_TYPEPOSQUERY: |
1416 |
|
case OP_TYPEPOSUPTO: |
1417 |
|
if (code[1] == OP_PROP || code[1] == OP_NOTPROP) code += 2; |
1418 |
|
break; |
1419 |
|
} |
1420 |
|
|
1421 |
|
/* Add in the fixed length from the table */ |
1422 |
|
|
1423 |
code += _pcre_OP_lengths[c]; |
code += _pcre_OP_lengths[c]; |
1424 |
|
|
1425 |
|
/* In UTF-8 mode, opcodes that are followed by a character may be followed |
1426 |
|
by a multi-byte character. The length in the table is a minimum, so we have |
1427 |
|
to arrange to skip the extra bytes. */ |
1428 |
|
|
1429 |
#ifdef SUPPORT_UTF8 |
#ifdef SUPPORT_UTF8 |
1430 |
if (utf8) switch(c) |
if (utf8) switch(c) |
1431 |
{ |
{ |
1485 |
|
|
1486 |
c = *code; |
c = *code; |
1487 |
|
|
1488 |
if (c == OP_BRA || c == OP_CBRA || c == OP_ONCE) |
/* Groups with zero repeats can of course be empty; skip them. */ |
1489 |
|
|
1490 |
|
if (c == OP_BRAZERO || c == OP_BRAMINZERO) |
1491 |
|
{ |
1492 |
|
code += _pcre_OP_lengths[c]; |
1493 |
|
do code += GET(code, 1); while (*code == OP_ALT); |
1494 |
|
c = *code; |
1495 |
|
continue; |
1496 |
|
} |
1497 |
|
|
1498 |
|
/* For other groups, scan the branches. */ |
1499 |
|
|
1500 |
|
if (c == OP_BRA || c == OP_CBRA || c == OP_ONCE || c == OP_COND) |
1501 |
{ |
{ |
1502 |
BOOL empty_branch; |
BOOL empty_branch; |
1503 |
if (GET(code, 1) == 0) return TRUE; /* Hit unclosed bracket */ |
if (GET(code, 1) == 0) return TRUE; /* Hit unclosed bracket */ |
1513 |
} |
} |
1514 |
while (*code == OP_ALT); |
while (*code == OP_ALT); |
1515 |
if (!empty_branch) return FALSE; /* All branches are non-empty */ |
if (!empty_branch) return FALSE; /* All branches are non-empty */ |
1516 |
|
c = *code; |
|
/* Move past the KET and fudge things so that the increment in the "for" |
|
|
above has no effect. */ |
|
|
|
|
|
c = OP_END; |
|
|
code += 1 + LINK_SIZE - _pcre_OP_lengths[c]; |
|
1517 |
continue; |
continue; |
1518 |
} |
} |
1519 |
|
|
1521 |
|
|
1522 |
switch (c) |
switch (c) |
1523 |
{ |
{ |
1524 |
/* Check for quantifiers after a class */ |
/* Check for quantifiers after a class. XCLASS is used for classes that |
1525 |
|
cannot be represented just by a bit map. This includes negated single |
1526 |
|
high-valued characters. The length in _pcre_OP_lengths[] is zero; the |
1527 |
|
actual length is stored in the compiled code, so we must update "code" |
1528 |
|
here. */ |
1529 |
|
|
1530 |
#ifdef SUPPORT_UTF8 |
#ifdef SUPPORT_UTF8 |
1531 |
case OP_XCLASS: |
case OP_XCLASS: |
1532 |
ccode = code + GET(code, 1); |
ccode = code += GET(code, 1); |
1533 |
goto CHECK_CLASS_REPEAT; |
goto CHECK_CLASS_REPEAT; |
1534 |
#endif |
#endif |
1535 |
|
|
2051 |
case OP_NOT_WORDCHAR: |
case OP_NOT_WORDCHAR: |
2052 |
return next <= 127 && (cd->ctypes[next] & ctype_word) != 0; |
return next <= 127 && (cd->ctypes[next] & ctype_word) != 0; |
2053 |
|
|
2054 |
|
case OP_HSPACE: |
2055 |
|
case OP_NOT_HSPACE: |
2056 |
|
switch(next) |
2057 |
|
{ |
2058 |
|
case 0x09: |
2059 |
|
case 0x20: |
2060 |
|
case 0xa0: |
2061 |
|
case 0x1680: |
2062 |
|
case 0x180e: |
2063 |
|
case 0x2000: |
2064 |
|
case 0x2001: |
2065 |
|
case 0x2002: |
2066 |
|
case 0x2003: |
2067 |
|
case 0x2004: |
2068 |
|
case 0x2005: |
2069 |
|
case 0x2006: |
2070 |
|
case 0x2007: |
2071 |
|
case 0x2008: |
2072 |
|
case 0x2009: |
2073 |
|
case 0x200A: |
2074 |
|
case 0x202f: |
2075 |
|
case 0x205f: |
2076 |
|
case 0x3000: |
2077 |
|
return op_code != OP_HSPACE; |
2078 |
|
default: |
2079 |
|
return op_code == OP_HSPACE; |
2080 |
|
} |
2081 |
|
|
2082 |
|
case OP_VSPACE: |
2083 |
|
case OP_NOT_VSPACE: |
2084 |
|
switch(next) |
2085 |
|
{ |
2086 |
|
case 0x0a: |
2087 |
|
case 0x0b: |
2088 |
|
case 0x0c: |
2089 |
|
case 0x0d: |
2090 |
|
case 0x85: |
2091 |
|
case 0x2028: |
2092 |
|
case 0x2029: |
2093 |
|
return op_code != OP_VSPACE; |
2094 |
|
default: |
2095 |
|
return op_code == OP_VSPACE; |
2096 |
|
} |
2097 |
|
|
2098 |
default: |
default: |
2099 |
return FALSE; |
return FALSE; |
2100 |
} |
} |
2129 |
case ESC_W: |
case ESC_W: |
2130 |
return item <= 127 && (cd->ctypes[item] & ctype_word) != 0; |
return item <= 127 && (cd->ctypes[item] & ctype_word) != 0; |
2131 |
|
|
2132 |
|
case ESC_h: |
2133 |
|
case ESC_H: |
2134 |
|
switch(item) |
2135 |
|
{ |
2136 |
|
case 0x09: |
2137 |
|
case 0x20: |
2138 |
|
case 0xa0: |
2139 |
|
case 0x1680: |
2140 |
|
case 0x180e: |
2141 |
|
case 0x2000: |
2142 |
|
case 0x2001: |
2143 |
|
case 0x2002: |
2144 |
|
case 0x2003: |
2145 |
|
case 0x2004: |
2146 |
|
case 0x2005: |
2147 |
|
case 0x2006: |
2148 |
|
case 0x2007: |
2149 |
|
case 0x2008: |
2150 |
|
case 0x2009: |
2151 |
|
case 0x200A: |
2152 |
|
case 0x202f: |
2153 |
|
case 0x205f: |
2154 |
|
case 0x3000: |
2155 |
|
return -next != ESC_h; |
2156 |
|
default: |
2157 |
|
return -next == ESC_h; |
2158 |
|
} |
2159 |
|
|
2160 |
|
case ESC_v: |
2161 |
|
case ESC_V: |
2162 |
|
switch(item) |
2163 |
|
{ |
2164 |
|
case 0x0a: |
2165 |
|
case 0x0b: |
2166 |
|
case 0x0c: |
2167 |
|
case 0x0d: |
2168 |
|
case 0x85: |
2169 |
|
case 0x2028: |
2170 |
|
case 0x2029: |
2171 |
|
return -next != ESC_v; |
2172 |
|
default: |
2173 |
|
return -next == ESC_v; |
2174 |
|
} |
2175 |
|
|
2176 |
default: |
default: |
2177 |
return FALSE; |
return FALSE; |
2178 |
} |
} |
2179 |
|
|
2180 |
case OP_DIGIT: |
case OP_DIGIT: |
2181 |
return next == -ESC_D || next == -ESC_s || next == -ESC_W; |
return next == -ESC_D || next == -ESC_s || next == -ESC_W || |
2182 |
|
next == -ESC_h || next == -ESC_v; |
2183 |
|
|
2184 |
case OP_NOT_DIGIT: |
case OP_NOT_DIGIT: |
2185 |
return next == -ESC_d; |
return next == -ESC_d; |
2188 |
return next == -ESC_S || next == -ESC_d || next == -ESC_w; |
return next == -ESC_S || next == -ESC_d || next == -ESC_w; |
2189 |
|
|
2190 |
case OP_NOT_WHITESPACE: |
case OP_NOT_WHITESPACE: |
2191 |
return next == -ESC_s; |
return next == -ESC_s || next == -ESC_h || next == -ESC_v; |
2192 |
|
|
2193 |
|
case OP_HSPACE: |
2194 |
|
return next == -ESC_S || next == -ESC_H || next == -ESC_d || next == -ESC_w; |
2195 |
|
|
2196 |
|
case OP_NOT_HSPACE: |
2197 |
|
return next == -ESC_h; |
2198 |
|
|
2199 |
|
/* Can't have \S in here because VT matches \S (Perl anomaly) */ |
2200 |
|
case OP_VSPACE: |
2201 |
|
return next == -ESC_V || next == -ESC_d || next == -ESC_w; |
2202 |
|
|
2203 |
|
case OP_NOT_VSPACE: |
2204 |
|
return next == -ESC_v; |
2205 |
|
|
2206 |
case OP_WORDCHAR: |
case OP_WORDCHAR: |
2207 |
return next == -ESC_W || next == -ESC_s; |
return next == -ESC_W || next == -ESC_s || next == -ESC_h || next == -ESC_v; |
2208 |
|
|
2209 |
case OP_NOT_WORDCHAR: |
case OP_NOT_WORDCHAR: |
2210 |
return next == -ESC_w || next == -ESC_d; |
return next == -ESC_w || next == -ESC_d; |
2319 |
BOOL possessive_quantifier; |
BOOL possessive_quantifier; |
2320 |
BOOL is_quantifier; |
BOOL is_quantifier; |
2321 |
BOOL is_recurse; |
BOOL is_recurse; |
2322 |
|
BOOL reset_bracount; |
2323 |
int class_charcount; |
int class_charcount; |
2324 |
int class_lastchar; |
int class_lastchar; |
2325 |
int newoptions; |
int newoptions; |
2326 |
int recno; |
int recno; |
2327 |
|
int refsign; |
2328 |
int skipbytes; |
int skipbytes; |
2329 |
int subreqbyte; |
int subreqbyte; |
2330 |
int subfirstbyte; |
int subfirstbyte; |
2357 |
*/ |
*/ |
2358 |
|
|
2359 |
if (code < last_code) code = last_code; |
if (code < last_code) code = last_code; |
2360 |
|
|
2361 |
|
/* Paranoid check for integer overflow */ |
2362 |
|
|
2363 |
|
if (OFLOW_MAX - *lengthptr < code - last_code) |
2364 |
|
{ |
2365 |
|
*errorcodeptr = ERR20; |
2366 |
|
goto FAILED; |
2367 |
|
} |
2368 |
|
|
2369 |
*lengthptr += code - last_code; |
*lengthptr += code - last_code; |
2370 |
DPRINTF(("length=%d added %d c=%c\n", *lengthptr, code - last_code, c)); |
DPRINTF(("length=%d added %d c=%c\n", *lengthptr, code - last_code, c)); |
2371 |
|
|
2478 |
*ptrptr = ptr; |
*ptrptr = ptr; |
2479 |
if (lengthptr != NULL) |
if (lengthptr != NULL) |
2480 |
{ |
{ |
2481 |
|
if (OFLOW_MAX - *lengthptr < code - last_code) |
2482 |
|
{ |
2483 |
|
*errorcodeptr = ERR20; |
2484 |
|
goto FAILED; |
2485 |
|
} |
2486 |
*lengthptr += code - last_code; /* To include callout length */ |
*lengthptr += code - last_code; /* To include callout length */ |
2487 |
DPRINTF((">> end branch\n")); |
DPRINTF((">> end branch\n")); |
2488 |
} |
} |
2545 |
goto FAILED; |
goto FAILED; |
2546 |
} |
} |
2547 |
|
|
2548 |
/* If the first character is '^', set the negation flag and skip it. */ |
/* If the first character is '^', set the negation flag and skip it. Also, |
2549 |
|
if the first few characters (either before or after ^) are \Q\E or \E we |
2550 |
|
skip them too. This makes for compatibility with Perl. */ |
2551 |
|
|
2552 |
if ((c = *(++ptr)) == '^') |
negate_class = FALSE; |
2553 |
|
for (;;) |
2554 |
{ |
{ |
|
negate_class = TRUE; |
|
2555 |
c = *(++ptr); |
c = *(++ptr); |
2556 |
} |
if (c == '\\') |
2557 |
else |
{ |
2558 |
{ |
if (ptr[1] == 'E') ptr++; |
2559 |
negate_class = FALSE; |
else if (strncmp((const char *)ptr+1, "Q\\E", 3) == 0) ptr += 3; |
2560 |
|
else break; |
2561 |
|
} |
2562 |
|
else if (!negate_class && c == '^') |
2563 |
|
negate_class = TRUE; |
2564 |
|
else break; |
2565 |
} |
} |
2566 |
|
|
2567 |
/* Keep a count of chars with values < 256 so that we can optimize the case |
/* Keep a count of chars with values < 256 so that we can optimize the case |
2702 |
of the specials, which just set a flag. The sequence \b is a special |
of the specials, which just set a flag. The sequence \b is a special |
2703 |
case. Inside a class (and only there) it is treated as backspace. |
case. Inside a class (and only there) it is treated as backspace. |
2704 |
Elsewhere it marks a word boundary. Other escapes have preset maps ready |
Elsewhere it marks a word boundary. Other escapes have preset maps ready |
2705 |
to or into the one we are building. We assume they have more than one |
to 'or' into the one we are building. We assume they have more than one |
2706 |
character in them, so set class_charcount bigger than one. */ |
character in them, so set class_charcount bigger than one. */ |
2707 |
|
|
2708 |
if (c == '\\') |
if (c == '\\') |
2722 |
else inescq = TRUE; |
else inescq = TRUE; |
2723 |
continue; |
continue; |
2724 |
} |
} |
2725 |
|
else if (-c == ESC_E) continue; /* Ignore orphan \E */ |
2726 |
|
|
2727 |
if (c < 0) |
if (c < 0) |
2728 |
{ |
{ |
2771 |
else if (c == -ESC_d || c == -ESC_D || c == -ESC_w || |
else if (c == -ESC_d || c == -ESC_D || c == -ESC_w || |
2772 |
c == -ESC_W || c == -ESC_s || c == -ESC_S) continue; |
c == -ESC_W || c == -ESC_s || c == -ESC_S) continue; |
2773 |
|
|
2774 |
|
/* We need to deal with \H, \h, \V, and \v in both phases because |
2775 |
|
they use extra memory. */ |
2776 |
|
|
2777 |
|
if (-c == ESC_h) |
2778 |
|
{ |
2779 |
|
SETBIT(classbits, 0x09); /* VT */ |
2780 |
|
SETBIT(classbits, 0x20); /* SPACE */ |
2781 |
|
SETBIT(classbits, 0xa0); /* NSBP */ |
2782 |
|
#ifdef SUPPORT_UTF8 |
2783 |
|
if (utf8) |
2784 |
|
{ |
2785 |
|
class_utf8 = TRUE; |
2786 |
|
*class_utf8data++ = XCL_SINGLE; |
2787 |
|
class_utf8data += _pcre_ord2utf8(0x1680, class_utf8data); |
2788 |
|
*class_utf8data++ = XCL_SINGLE; |
2789 |
|
class_utf8data += _pcre_ord2utf8(0x180e, class_utf8data); |
2790 |
|
*class_utf8data++ = XCL_RANGE; |
2791 |
|
class_utf8data += _pcre_ord2utf8(0x2000, class_utf8data); |
2792 |
|
class_utf8data += _pcre_ord2utf8(0x200A, class_utf8data); |
2793 |
|
*class_utf8data++ = XCL_SINGLE; |
2794 |
|
class_utf8data += _pcre_ord2utf8(0x202f, class_utf8data); |
2795 |
|
*class_utf8data++ = XCL_SINGLE; |
2796 |
|
class_utf8data += _pcre_ord2utf8(0x205f, class_utf8data); |
2797 |
|
*class_utf8data++ = XCL_SINGLE; |
2798 |
|
class_utf8data += _pcre_ord2utf8(0x3000, class_utf8data); |
2799 |
|
} |
2800 |
|
#endif |
2801 |
|
continue; |
2802 |
|
} |
2803 |
|
|
2804 |
|
if (-c == ESC_H) |
2805 |
|
{ |
2806 |
|
for (c = 0; c < 32; c++) |
2807 |
|
{ |
2808 |
|
int x = 0xff; |
2809 |
|
switch (c) |
2810 |
|
{ |
2811 |
|
case 0x09/8: x ^= 1 << (0x09%8); break; |
2812 |
|
case 0x20/8: x ^= 1 << (0x20%8); break; |
2813 |
|
case 0xa0/8: x ^= 1 << (0xa0%8); break; |
2814 |
|
default: break; |
2815 |
|
} |
2816 |
|
classbits[c] |= x; |
2817 |
|
} |
2818 |
|
|
2819 |
|
#ifdef SUPPORT_UTF8 |
2820 |
|
if (utf8) |
2821 |
|
{ |
2822 |
|
class_utf8 = TRUE; |
2823 |
|
*class_utf8data++ = XCL_RANGE; |
2824 |
|
class_utf8data += _pcre_ord2utf8(0x0100, class_utf8data); |
2825 |
|
class_utf8data += _pcre_ord2utf8(0x167f, class_utf8data); |
2826 |
|
*class_utf8data++ = XCL_RANGE; |
2827 |
|
class_utf8data += _pcre_ord2utf8(0x1681, class_utf8data); |
2828 |
|
class_utf8data += _pcre_ord2utf8(0x180d, class_utf8data); |
2829 |
|
*class_utf8data++ = XCL_RANGE; |
2830 |
|
class_utf8data += _pcre_ord2utf8(0x180f, class_utf8data); |
2831 |
|
class_utf8data += _pcre_ord2utf8(0x1fff, class_utf8data); |
2832 |
|
*class_utf8data++ = XCL_RANGE; |
2833 |
|
class_utf8data += _pcre_ord2utf8(0x200B, class_utf8data); |
2834 |
|
class_utf8data += _pcre_ord2utf8(0x202e, class_utf8data); |
2835 |
|
*class_utf8data++ = XCL_RANGE; |
2836 |
|
class_utf8data += _pcre_ord2utf8(0x2030, class_utf8data); |
2837 |
|
class_utf8data += _pcre_ord2utf8(0x205e, class_utf8data); |
2838 |
|
*class_utf8data++ = XCL_RANGE; |
2839 |
|
class_utf8data += _pcre_ord2utf8(0x2060, class_utf8data); |
2840 |
|
class_utf8data += _pcre_ord2utf8(0x2fff, class_utf8data); |
2841 |
|
*class_utf8data++ = XCL_RANGE; |
2842 |
|
class_utf8data += _pcre_ord2utf8(0x3001, class_utf8data); |
2843 |
|
class_utf8data += _pcre_ord2utf8(0x7fffffff, class_utf8data); |
2844 |
|
} |
2845 |
|
#endif |
2846 |
|
continue; |
2847 |
|
} |
2848 |
|
|
2849 |
|
if (-c == ESC_v) |
2850 |
|
{ |
2851 |
|
SETBIT(classbits, 0x0a); /* LF */ |
2852 |
|
SETBIT(classbits, 0x0b); /* VT */ |
2853 |
|
SETBIT(classbits, 0x0c); /* FF */ |
2854 |
|
SETBIT(classbits, 0x0d); /* CR */ |
2855 |
|
SETBIT(classbits, 0x85); /* NEL */ |
2856 |
|
#ifdef SUPPORT_UTF8 |
2857 |
|
if (utf8) |
2858 |
|
{ |
2859 |
|
class_utf8 = TRUE; |
2860 |
|
*class_utf8data++ = XCL_RANGE; |
2861 |
|
class_utf8data += _pcre_ord2utf8(0x2028, class_utf8data); |
2862 |
|
class_utf8data += _pcre_ord2utf8(0x2029, class_utf8data); |
2863 |
|
} |
2864 |
|
#endif |
2865 |
|
continue; |
2866 |
|
} |
2867 |
|
|
2868 |
|
if (-c == ESC_V) |
2869 |
|
{ |
2870 |
|
for (c = 0; c < 32; c++) |
2871 |
|
{ |
2872 |
|
int x = 0xff; |
2873 |
|
switch (c) |
2874 |
|
{ |
2875 |
|
case 0x0a/8: x ^= 1 << (0x0a%8); |
2876 |
|
x ^= 1 << (0x0b%8); |
2877 |
|
x ^= 1 << (0x0c%8); |
2878 |
|
x ^= 1 << (0x0d%8); |
2879 |
|
break; |
2880 |
|
case 0x85/8: x ^= 1 << (0x85%8); break; |
2881 |
|
default: break; |
2882 |
|
} |
2883 |
|
classbits[c] |= x; |
2884 |
|
} |
2885 |
|
|
2886 |
|
#ifdef SUPPORT_UTF8 |
2887 |
|
if (utf8) |
2888 |
|
{ |
2889 |
|
class_utf8 = TRUE; |
2890 |
|
*class_utf8data++ = XCL_RANGE; |
2891 |
|
class_utf8data += _pcre_ord2utf8(0x0100, class_utf8data); |
2892 |
|
class_utf8data += _pcre_ord2utf8(0x2027, class_utf8data); |
2893 |
|
*class_utf8data++ = XCL_RANGE; |
2894 |
|
class_utf8data += _pcre_ord2utf8(0x2029, class_utf8data); |
2895 |
|
class_utf8data += _pcre_ord2utf8(0x7fffffff, class_utf8data); |
2896 |
|
} |
2897 |
|
#endif |
2898 |
|
continue; |
2899 |
|
} |
2900 |
|
|
2901 |
/* We need to deal with \P and \p in both phases. */ |
/* We need to deal with \P and \p in both phases. */ |
2902 |
|
|
2903 |
#ifdef SUPPORT_UCP |
#ifdef SUPPORT_UCP |
3038 |
unsigned int origd = d; |
unsigned int origd = d; |
3039 |
while (get_othercase_range(&cc, origd, &occ, &ocd)) |
while (get_othercase_range(&cc, origd, &occ, &ocd)) |
3040 |
{ |
{ |
3041 |
if (occ >= c && ocd <= d) continue; /* Skip embedded ranges */ |
if (occ >= (unsigned int)c && |
3042 |
|
ocd <= (unsigned int)d) |
3043 |
|
continue; /* Skip embedded ranges */ |
3044 |
|
|
3045 |
if (occ < c && ocd >= c - 1) /* Extend the basic range */ |
if (occ < (unsigned int)c && |
3046 |
|
ocd >= (unsigned int)c - 1) /* Extend the basic range */ |
3047 |
{ /* if there is overlap, */ |
{ /* if there is overlap, */ |
3048 |
c = occ; /* noting that if occ < c */ |
c = occ; /* noting that if occ < c */ |
3049 |
continue; /* we can't have ocd > d */ |
continue; /* we can't have ocd > d */ |
3050 |
} /* because a subrange is */ |
} /* because a subrange is */ |
3051 |
if (ocd > d && occ <= d + 1) /* always shorter than */ |
if (ocd > (unsigned int)d && |
3052 |
|
occ <= (unsigned int)d + 1) /* always shorter than */ |
3053 |
{ /* the basic range. */ |
{ /* the basic range. */ |
3054 |
d = ocd; |
d = ocd; |
3055 |
continue; |
continue; |
3645 |
goto FAILED; |
goto FAILED; |
3646 |
} |
} |
3647 |
|
|
|
/* This is a paranoid check to stop integer overflow later on */ |
|
|
|
|
|
if (len > MAX_DUPLENGTH) |
|
|
{ |
|
|
*errorcodeptr = ERR50; |
|
|
goto FAILED; |
|
|
} |
|
|
|
|
3648 |
/* If the maximum repeat count is unlimited, find the end of the bracket |
/* If the maximum repeat count is unlimited, find the end of the bracket |
3649 |
by scanning through from the start, and compute the offset back to it |
by scanning through from the start, and compute the offset back to it |
3650 |
from the current code pointer. There may be an OP_OPT setting following |
from the current code pointer. There may be an OP_OPT setting following |
3733 |
if (repeat_min > 1) |
if (repeat_min > 1) |
3734 |
{ |
{ |
3735 |
/* In the pre-compile phase, we don't actually do the replication. We |
/* In the pre-compile phase, we don't actually do the replication. We |
3736 |
just adjust the length as if we had. */ |
just adjust the length as if we had. Do some paranoid checks for |
3737 |
|
potential integer overflow. */ |
3738 |
|
|
3739 |
if (lengthptr != NULL) |
if (lengthptr != NULL) |
3740 |
*lengthptr += (repeat_min - 1)*length_prevgroup; |
{ |
3741 |
|
int delta = (repeat_min - 1)*length_prevgroup; |
3742 |
|
if ((double)(repeat_min - 1)*(double)length_prevgroup > |
3743 |
|
(double)INT_MAX || |
3744 |
|
OFLOW_MAX - *lengthptr < delta) |
3745 |
|
{ |
3746 |
|
*errorcodeptr = ERR20; |
3747 |
|
goto FAILED; |
3748 |
|
} |
3749 |
|
*lengthptr += delta; |
3750 |
|
} |
3751 |
|
|
3752 |
/* This is compiling for real */ |
/* This is compiling for real */ |
3753 |
|
|
3785 |
/* In the pre-compile phase, we don't actually do the replication. We |
/* In the pre-compile phase, we don't actually do the replication. We |
3786 |
just adjust the length as if we had. For each repetition we must add 1 |
just adjust the length as if we had. For each repetition we must add 1 |
3787 |
to the length for BRAZERO and for all but the last repetition we must |
to the length for BRAZERO and for all but the last repetition we must |
3788 |
add 2 + 2*LINKSIZE to allow for the nesting that occurs. */ |
add 2 + 2*LINKSIZE to allow for the nesting that occurs. Do some |
3789 |
|
paranoid checks to avoid integer overflow. */ |
3790 |
|
|
3791 |
if (lengthptr != NULL && repeat_max > 0) |
if (lengthptr != NULL && repeat_max > 0) |
3792 |
*lengthptr += repeat_max * (length_prevgroup + 1 + 2 + 2*LINK_SIZE) - |
{ |
3793 |
2 - 2*LINK_SIZE; /* Last one doesn't nest */ |
int delta = repeat_max * (length_prevgroup + 1 + 2 + 2*LINK_SIZE) - |
3794 |
|
2 - 2*LINK_SIZE; /* Last one doesn't nest */ |
3795 |
|
if ((double)repeat_max * |
3796 |
|
(double)(length_prevgroup + 1 + 2 + 2*LINK_SIZE) |
3797 |
|
> (double)INT_MAX || |
3798 |
|
OFLOW_MAX - *lengthptr < delta) |
3799 |
|
{ |
3800 |
|
*errorcodeptr = ERR20; |
3801 |
|
goto FAILED; |
3802 |
|
} |
3803 |
|
*lengthptr += delta; |
3804 |
|
} |
3805 |
|
|
3806 |
/* This is compiling for real */ |
/* This is compiling for real */ |
3807 |
|
|
3953 |
/* ===================================================================*/ |
/* ===================================================================*/ |
3954 |
/* Start of nested parenthesized sub-expression, or comment or lookahead or |
/* Start of nested parenthesized sub-expression, or comment or lookahead or |
3955 |
lookbehind or option setting or condition or all the other extended |
lookbehind or option setting or condition or all the other extended |
3956 |
parenthesis forms. First deal with the specials; all are introduced by ?, |
parenthesis forms. */ |
|
and the appearance of any of them means that this is not a capturing |
|
|
group. */ |
|
3957 |
|
|
3958 |
case '(': |
case '(': |
3959 |
newoptions = options; |
newoptions = options; |
3960 |
skipbytes = 0; |
skipbytes = 0; |
3961 |
bravalue = OP_CBRA; |
bravalue = OP_CBRA; |
3962 |
save_hwm = cd->hwm; |
save_hwm = cd->hwm; |
3963 |
|
reset_bracount = FALSE; |
3964 |
|
|
3965 |
if (*(++ptr) == '?') |
/* First deal with various "verbs" that can be introduced by '*'. */ |
3966 |
|
|
3967 |
|
if (*(++ptr) == '*' && (cd->ctypes[ptr[1]] & ctype_letter) != 0) |
3968 |
|
{ |
3969 |
|
int i, namelen; |
3970 |
|
const uschar *name = ++ptr; |
3971 |
|
previous = NULL; |
3972 |
|
while ((cd->ctypes[*++ptr] & ctype_letter) != 0); |
3973 |
|
if (*ptr == ':') |
3974 |
|
{ |
3975 |
|
*errorcodeptr = ERR59; /* Not supported */ |
3976 |
|
goto FAILED; |
3977 |
|
} |
3978 |
|
if (*ptr != ')') |
3979 |
|
{ |
3980 |
|
*errorcodeptr = ERR60; |
3981 |
|
goto FAILED; |
3982 |
|
} |
3983 |
|
namelen = ptr - name; |
3984 |
|
for (i = 0; i < verbcount; i++) |
3985 |
|
{ |
3986 |
|
if (namelen == verbs[i].len && |
3987 |
|
strncmp((char *)name, verbs[i].name, namelen) == 0) |
3988 |
|
{ |
3989 |
|
*code = verbs[i].op; |
3990 |
|
if (*code++ == OP_ACCEPT) cd->had_accept = TRUE; |
3991 |
|
break; |
3992 |
|
} |
3993 |
|
} |
3994 |
|
if (i < verbcount) continue; |
3995 |
|
*errorcodeptr = ERR60; |
3996 |
|
goto FAILED; |
3997 |
|
} |
3998 |
|
|
3999 |
|
/* Deal with the extended parentheses; all are introduced by '?', and the |
4000 |
|
appearance of any of them means that this is not a capturing group. */ |
4001 |
|
|
4002 |
|
else if (*ptr == '?') |
4003 |
{ |
{ |
4004 |
int i, set, unset, namelen; |
int i, set, unset, namelen; |
4005 |
int *optset; |
int *optset; |
4020 |
|
|
4021 |
|
|
4022 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
4023 |
|
case '|': /* Reset capture count for each branch */ |
4024 |
|
reset_bracount = TRUE; |
4025 |
|
/* Fall through */ |
4026 |
|
|
4027 |
|
/* ------------------------------------------------------------ */ |
4028 |
case ':': /* Non-capturing bracket */ |
case ':': /* Non-capturing bracket */ |
4029 |
bravalue = OP_BRA; |
bravalue = OP_BRA; |
4030 |
ptr++; |
ptr++; |
4060 |
|
|
4061 |
code[1+LINK_SIZE] = OP_CREF; |
code[1+LINK_SIZE] = OP_CREF; |
4062 |
skipbytes = 3; |
skipbytes = 3; |
4063 |
|
refsign = -1; |
4064 |
|
|
4065 |
/* Check for a test for recursion in a named group. */ |
/* Check for a test for recursion in a named group. */ |
4066 |
|
|
4084 |
terminator = '\''; |
terminator = '\''; |
4085 |
ptr++; |
ptr++; |
4086 |
} |
} |
4087 |
else terminator = 0; |
else |
4088 |
|
{ |
4089 |
|
terminator = 0; |
4090 |
|
if (ptr[1] == '-' || ptr[1] == '+') refsign = *(++ptr); |
4091 |
|
} |
4092 |
|
|
4093 |
/* We now expect to read a name; any thing else is an error */ |
/* We now expect to read a name; any thing else is an error */ |
4094 |
|
|
4124 |
if (lengthptr != NULL) break; |
if (lengthptr != NULL) break; |
4125 |
|
|
4126 |
/* In the real compile we do the work of looking for the actual |
/* In the real compile we do the work of looking for the actual |
4127 |
reference. */ |
reference. If the string started with "+" or "-" we require the rest to |
4128 |
|
be digits, in which case recno will be set. */ |
4129 |
|
|
4130 |
|
if (refsign > 0) |
4131 |
|
{ |
4132 |
|
if (recno <= 0) |
4133 |
|
{ |
4134 |
|
*errorcodeptr = ERR58; |
4135 |
|
goto FAILED; |
4136 |
|
} |
4137 |
|
if (refsign == '-') |
4138 |
|
{ |
4139 |
|
recno = cd->bracount - recno + 1; |
4140 |
|
if (recno <= 0) |
4141 |
|
{ |
4142 |
|
*errorcodeptr = ERR15; |
4143 |
|
goto FAILED; |
4144 |
|
} |
4145 |
|
} |
4146 |
|
else recno += cd->bracount; |
4147 |
|
PUT2(code, 2+LINK_SIZE, recno); |
4148 |
|
break; |
4149 |
|
} |
4150 |
|
|
4151 |
|
/* Otherwise (did not start with "+" or "-"), start by looking for the |
4152 |
|
name. */ |
4153 |
|
|
4154 |
slot = cd->name_table; |
slot = cd->name_table; |
4155 |
for (i = 0; i < cd->names_found; i++) |
for (i = 0; i < cd->names_found; i++) |
4241 |
|
|
4242 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
4243 |
case '!': /* Negative lookahead */ |
case '!': /* Negative lookahead */ |
|
bravalue = OP_ASSERT_NOT; |
|
4244 |
ptr++; |
ptr++; |
4245 |
|
if (*ptr == ')') /* Optimize (?!) */ |
4246 |
|
{ |
4247 |
|
*code++ = OP_FAIL; |
4248 |
|
previous = NULL; |
4249 |
|
continue; |
4250 |
|
} |
4251 |
|
bravalue = OP_ASSERT_NOT; |
4252 |
break; |
break; |
4253 |
|
|
4254 |
|
|
4474 |
|
|
4475 |
|
|
4476 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
4477 |
|
case '-': case '+': |
4478 |
case '0': case '1': case '2': case '3': case '4': /* Recursion or */ |
case '0': case '1': case '2': case '3': case '4': /* Recursion or */ |
4479 |
case '5': case '6': case '7': case '8': case '9': /* subroutine */ |
case '5': case '6': case '7': case '8': case '9': /* subroutine */ |
4480 |
{ |
{ |
4481 |
const uschar *called; |
const uschar *called; |
4482 |
|
|
4483 |
|
if ((refsign = *ptr) == '+') ptr++; |
4484 |
|
else if (refsign == '-') |
4485 |
|
{ |
4486 |
|
if ((digitab[ptr[1]] & ctype_digit) == 0) |
4487 |
|
goto OTHER_CHAR_AFTER_QUERY; |
4488 |
|
ptr++; |
4489 |
|
} |
4490 |
|
|
4491 |
recno = 0; |
recno = 0; |
4492 |
while((digitab[*ptr] & ctype_digit) != 0) |
while((digitab[*ptr] & ctype_digit) != 0) |
4493 |
recno = recno * 10 + *ptr++ - '0'; |
recno = recno * 10 + *ptr++ - '0'; |
4494 |
|
|
4495 |
if (*ptr != ')') |
if (*ptr != ')') |
4496 |
{ |
{ |
4497 |
*errorcodeptr = ERR29; |
*errorcodeptr = ERR29; |
4498 |
goto FAILED; |
goto FAILED; |
4499 |
} |
} |
4500 |
|
|
4501 |
|
if (refsign == '-') |
4502 |
|
{ |
4503 |
|
if (recno == 0) |
4504 |
|
{ |
4505 |
|
*errorcodeptr = ERR58; |
4506 |
|
goto FAILED; |
4507 |
|
} |
4508 |
|
recno = cd->bracount - recno + 1; |
4509 |
|
if (recno <= 0) |
4510 |
|
{ |
4511 |
|
*errorcodeptr = ERR15; |
4512 |
|
goto FAILED; |
4513 |
|
} |
4514 |
|
} |
4515 |
|
else if (refsign == '+') |
4516 |
|
{ |
4517 |
|
if (recno == 0) |
4518 |
|
{ |
4519 |
|
*errorcodeptr = ERR58; |
4520 |
|
goto FAILED; |
4521 |
|
} |
4522 |
|
recno += cd->bracount; |
4523 |
|
} |
4524 |
|
|
4525 |
/* Come here from code above that handles a named recursion */ |
/* Come here from code above that handles a named recursion */ |
4526 |
|
|
4527 |
HANDLE_RECURSION: |
HANDLE_RECURSION: |
4594 |
|
|
4595 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
4596 |
default: /* Other characters: check option setting */ |
default: /* Other characters: check option setting */ |
4597 |
|
OTHER_CHAR_AFTER_QUERY: |
4598 |
set = unset = 0; |
set = unset = 0; |
4599 |
optset = &set; |
optset = &set; |
4600 |
|
|
4729 |
errorcodeptr, /* Where to put an error message */ |
errorcodeptr, /* Where to put an error message */ |
4730 |
(bravalue == OP_ASSERTBACK || |
(bravalue == OP_ASSERTBACK || |
4731 |
bravalue == OP_ASSERTBACK_NOT), /* TRUE if back assert */ |
bravalue == OP_ASSERTBACK_NOT), /* TRUE if back assert */ |
4732 |
|
reset_bracount, /* True if (?| group */ |
4733 |
skipbytes, /* Skip over bracket number */ |
skipbytes, /* Skip over bracket number */ |
4734 |
&subfirstbyte, /* For possible first char */ |
&subfirstbyte, /* For possible first char */ |
4735 |
&subreqbyte, /* For possible last char */ |
&subreqbyte, /* For possible last char */ |
4746 |
is on the bracket. */ |
is on the bracket. */ |
4747 |
|
|
4748 |
/* If this is a conditional bracket, check that there are no more than |
/* If this is a conditional bracket, check that there are no more than |
4749 |
two branches in the group, or just one if it's a DEFINE group. */ |
two branches in the group, or just one if it's a DEFINE group. We do this |
4750 |
|
in the real compile phase, not in the pre-pass, where the whole group may |
4751 |
|
not be available. */ |
4752 |
|
|
4753 |
if (bravalue == OP_COND) |
if (bravalue == OP_COND && lengthptr == NULL) |
4754 |
{ |
{ |
4755 |
uschar *tc = code; |
uschar *tc = code; |
4756 |
int condcount = 0; |
int condcount = 0; |
4797 |
goto FAILED; |
goto FAILED; |
4798 |
} |
} |
4799 |
|
|
4800 |
/* In the pre-compile phase, update the length by the length of the nested |
/* In the pre-compile phase, update the length by the length of the group, |
4801 |
group, less the brackets at either end. Then reduce the compiled code to |
less the brackets at either end. Then reduce the compiled code to just a |
4802 |
just the brackets so that it doesn't use much memory if it is duplicated by |
set of non-capturing brackets so that it doesn't use much memory if it is |
4803 |
a quantifier. */ |
duplicated by a quantifier.*/ |
4804 |
|
|
4805 |
if (lengthptr != NULL) |
if (lengthptr != NULL) |
4806 |
{ |
{ |
4807 |
|
if (OFLOW_MAX - *lengthptr < length_prevgroup - 2 - 2*LINK_SIZE) |
4808 |
|
{ |
4809 |
|
*errorcodeptr = ERR20; |
4810 |
|
goto FAILED; |
4811 |
|
} |
4812 |
*lengthptr += length_prevgroup - 2 - 2*LINK_SIZE; |
*lengthptr += length_prevgroup - 2 - 2*LINK_SIZE; |
4813 |
code++; |
*code++ = OP_BRA; |
4814 |
PUTINC(code, 0, 1 + LINK_SIZE); |
PUTINC(code, 0, 1 + LINK_SIZE); |
4815 |
*code++ = OP_KET; |
*code++ = OP_KET; |
4816 |
PUTINC(code, 0, 1 + LINK_SIZE); |
PUTINC(code, 0, 1 + LINK_SIZE); |
4817 |
|
break; /* No need to waste time with special character handling */ |
4818 |
} |
} |
4819 |
|
|
4820 |
/* Otherwise update the main code pointer to the end of the group. */ |
/* Otherwise update the main code pointer to the end of the group. */ |
4821 |
|
|
4822 |
else code = tempcode; |
code = tempcode; |
4823 |
|
|
4824 |
/* For a DEFINE group, required and first character settings are not |
/* For a DEFINE group, required and first character settings are not |
4825 |
relevant. */ |
relevant. */ |
4916 |
zerofirstbyte = firstbyte; |
zerofirstbyte = firstbyte; |
4917 |
zeroreqbyte = reqbyte; |
zeroreqbyte = reqbyte; |
4918 |
|
|
4919 |
/* \k<name> or \k'name' is a back reference by name (Perl syntax) */ |
/* \k<name> or \k'name' is a back reference by name (Perl syntax). |
4920 |
|
We also support \k{name} (.NET syntax) */ |
4921 |
|
|
4922 |
if (-c == ESC_k && (ptr[1] == '<' || ptr[1] == '\'')) |
if (-c == ESC_k && (ptr[1] == '<' || ptr[1] == '\'' || ptr[1] == '{')) |
4923 |
{ |
{ |
4924 |
is_recurse = FALSE; |
is_recurse = FALSE; |
4925 |
terminator = (*(++ptr) == '<')? '>' : '\''; |
terminator = (*(++ptr) == '<')? '>' : (*ptr == '\'')? '\'' : '}'; |
4926 |
goto NAMED_REF_OR_RECURSE; |
goto NAMED_REF_OR_RECURSE; |
4927 |
} |
} |
4928 |
|
|
5088 |
out the amount of memory needed, as well as during the real compile phase. The |
out the amount of memory needed, as well as during the real compile phase. The |
5089 |
value of lengthptr distinguishes the two phases. |
value of lengthptr distinguishes the two phases. |
5090 |
|
|
5091 |
Argument: |
Arguments: |
5092 |
options option bits, including any changes for this subpattern |
options option bits, including any changes for this subpattern |
5093 |
oldims previous settings of ims option bits |
oldims previous settings of ims option bits |
5094 |
codeptr -> the address of the current code pointer |
codeptr -> the address of the current code pointer |
5095 |
ptrptr -> the address of the current pattern pointer |
ptrptr -> the address of the current pattern pointer |
5096 |
errorcodeptr -> pointer to error code variable |
errorcodeptr -> pointer to error code variable |
5097 |
lookbehind TRUE if this is a lookbehind assertion |
lookbehind TRUE if this is a lookbehind assertion |
5098 |
|
reset_bracount TRUE to reset the count for each branch |
5099 |
skipbytes skip this many bytes at start (for brackets and OP_COND) |
skipbytes skip this many bytes at start (for brackets and OP_COND) |
5100 |
firstbyteptr place to put the first required character, or a negative number |
firstbyteptr place to put the first required character, or a negative number |
5101 |
reqbyteptr place to put the last required character, or a negative number |
reqbyteptr place to put the last required character, or a negative number |
5109 |
|
|
5110 |
static BOOL |
static BOOL |
5111 |
compile_regex(int options, int oldims, uschar **codeptr, const uschar **ptrptr, |
compile_regex(int options, int oldims, uschar **codeptr, const uschar **ptrptr, |
5112 |
int *errorcodeptr, BOOL lookbehind, int skipbytes, int *firstbyteptr, |
int *errorcodeptr, BOOL lookbehind, BOOL reset_bracount, int skipbytes, |
5113 |
int *reqbyteptr, branch_chain *bcptr, compile_data *cd, int *lengthptr) |
int *firstbyteptr, int *reqbyteptr, branch_chain *bcptr, compile_data *cd, |
5114 |
|
int *lengthptr) |
5115 |
{ |
{ |
5116 |
const uschar *ptr = *ptrptr; |
const uschar *ptr = *ptrptr; |
5117 |
uschar *code = *codeptr; |
uschar *code = *codeptr; |
5121 |
int firstbyte, reqbyte; |
int firstbyte, reqbyte; |
5122 |
int branchfirstbyte, branchreqbyte; |
int branchfirstbyte, branchreqbyte; |
5123 |
int length; |
int length; |
5124 |
|
int orig_bracount; |
5125 |
|
int max_bracount; |
5126 |
branch_chain bc; |
branch_chain bc; |
5127 |
|
|
5128 |
bc.outer = bcptr; |
bc.outer = bcptr; |
5151 |
|
|
5152 |
/* Loop for each alternative branch */ |
/* Loop for each alternative branch */ |
5153 |
|
|
5154 |
|
orig_bracount = max_bracount = cd->bracount; |
5155 |
for (;;) |
for (;;) |
5156 |
{ |
{ |
5157 |
|
/* For a (?| group, reset the capturing bracket count so that each branch |
5158 |
|
uses the same numbers. */ |
5159 |
|
|
5160 |
|
if (reset_bracount) cd->bracount = orig_bracount; |
5161 |
|
|
5162 |
/* Handle a change of ims options at the start of the branch */ |
/* Handle a change of ims options at the start of the branch */ |
5163 |
|
|
5164 |
if ((options & PCRE_IMS) != oldims) |
if ((options & PCRE_IMS) != oldims) |
5188 |
return FALSE; |
return FALSE; |
5189 |
} |
} |
5190 |
|
|
5191 |
|
/* Keep the highest bracket count in case (?| was used and some branch |
5192 |
|
has fewer than the rest. */ |
5193 |
|
|
5194 |
|
if (cd->bracount > max_bracount) max_bracount = cd->bracount; |
5195 |
|
|
5196 |
/* In the real compile phase, there is some post-processing to be done. */ |
/* In the real compile phase, there is some post-processing to be done. */ |
5197 |
|
|
5198 |
if (lengthptr == NULL) |
if (lengthptr == NULL) |
5256 |
} |
} |
5257 |
} |
} |
5258 |
|
|
5259 |
/* Reached end of expression, either ')' or end of pattern. Go back through |
/* Reached end of expression, either ')' or end of pattern. In the real |
5260 |
the alternative branches and reverse the chain of offsets, with the field in |
compile phase, go back through the alternative branches and reverse the chain |
5261 |
the BRA item now becoming an offset to the first alternative. If there are |
of offsets, with the field in the BRA item now becoming an offset to the |
5262 |
no alternatives, it points to the end of the group. The length in the |
first alternative. If there are no alternatives, it points to the end of the |
5263 |
terminating ket is always the length of the whole bracketed item. If any of |
group. The length in the terminating ket is always the length of the whole |
5264 |
the ims options were changed inside the group, compile a resetting op-code |
bracketed item. If any of the ims options were changed inside the group, |
5265 |
following, except at the very end of the pattern. Return leaving the pointer |
compile a resetting op-code following, except at the very end of the pattern. |
5266 |
at the terminating char. */ |
Return leaving the pointer at the terminating char. */ |
5267 |
|
|
5268 |
if (*ptr != '|') |
if (*ptr != '|') |
5269 |
{ |
{ |
5270 |
int branch_length = code - last_branch; |
if (lengthptr == NULL) |
|
do |
|
5271 |
{ |
{ |
5272 |
int prev_length = GET(last_branch, 1); |
int branch_length = code - last_branch; |
5273 |
PUT(last_branch, 1, branch_length); |
do |
5274 |
branch_length = prev_length; |
{ |
5275 |
last_branch -= branch_length; |
int prev_length = GET(last_branch, 1); |
5276 |
|
PUT(last_branch, 1, branch_length); |
5277 |
|
branch_length = prev_length; |
5278 |
|
last_branch -= branch_length; |
5279 |
|
} |
5280 |
|
while (branch_length > 0); |
5281 |
} |
} |
|
while (branch_length > 0); |
|
5282 |
|
|
5283 |
/* Fill in the ket */ |
/* Fill in the ket */ |
5284 |
|
|
5295 |
length += 2; |
length += 2; |
5296 |
} |
} |
5297 |
|
|
5298 |
|
/* Retain the highest bracket number, in case resetting was used. */ |
5299 |
|
|
5300 |
|
cd->bracount = max_bracount; |
5301 |
|
|
5302 |
/* Set values to pass back */ |
/* Set values to pass back */ |
5303 |
|
|
5304 |
*codeptr = code; |
*codeptr = code; |
5305 |
*ptrptr = ptr; |
*ptrptr = ptr; |
5306 |
*firstbyteptr = firstbyte; |
*firstbyteptr = firstbyte; |
5307 |
*reqbyteptr = reqbyte; |
*reqbyteptr = reqbyte; |
5308 |
if (lengthptr != NULL) *lengthptr += length; |
if (lengthptr != NULL) |
5309 |
|
{ |
5310 |
|
if (OFLOW_MAX - *lengthptr < length) |
5311 |
|
{ |
5312 |
|
*errorcodeptr = ERR20; |
5313 |
|
return FALSE; |
5314 |
|
} |
5315 |
|
*lengthptr += length; |
5316 |
|
} |
5317 |
return TRUE; |
return TRUE; |
5318 |
} |
} |
5319 |
|
|
5320 |
/* Another branch follows; insert an "or" node. Its length field points back |
/* Another branch follows. In the pre-compile phase, we can move the code |
5321 |
|
pointer back to where it was for the start of the first branch. (That is, |
5322 |
|
pretend that each branch is the only one.) |
5323 |
|
|
5324 |
|
In the real compile phase, insert an ALT node. Its length field points back |
5325 |
to the previous branch while the bracket remains open. At the end the chain |
to the previous branch while the bracket remains open. At the end the chain |
5326 |
is reversed. It's done like this so that the start of the bracket has a |
is reversed. It's done like this so that the start of the bracket has a |
5327 |
zero offset until it is closed, making it possible to detect recursion. */ |
zero offset until it is closed, making it possible to detect recursion. */ |
5328 |
|
|
5329 |
*code = OP_ALT; |
if (lengthptr != NULL) |
5330 |
PUT(code, 1, code - last_branch); |
{ |
5331 |
bc.current = last_branch = code; |
code = *codeptr + 1 + LINK_SIZE + skipbytes; |
5332 |
code += 1 + LINK_SIZE; |
length += 1 + LINK_SIZE; |
5333 |
|
} |
5334 |
|
else |
5335 |
|
{ |
5336 |
|
*code = OP_ALT; |
5337 |
|
PUT(code, 1, code - last_branch); |
5338 |
|
bc.current = last_branch = code; |
5339 |
|
code += 1 + LINK_SIZE; |
5340 |
|
} |
5341 |
|
|
5342 |
ptr++; |
ptr++; |
|
length += 1 + LINK_SIZE; |
|
5343 |
} |
} |
5344 |
/* Control never reaches here */ |
/* Control never reaches here */ |
5345 |
} |
} |
5781 |
code = cworkspace; |
code = cworkspace; |
5782 |
*code = OP_BRA; |
*code = OP_BRA; |
5783 |
(void)compile_regex(cd->external_options, cd->external_options & PCRE_IMS, |
(void)compile_regex(cd->external_options, cd->external_options & PCRE_IMS, |
5784 |
&code, &ptr, &errorcode, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, &length); |
&code, &ptr, &errorcode, FALSE, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, |
5785 |
|
&length); |
5786 |
if (errorcode != 0) goto PCRE_EARLY_ERROR_RETURN; |
if (errorcode != 0) goto PCRE_EARLY_ERROR_RETURN; |
5787 |
|
|
5788 |
DPRINTF(("end pre-compile: length=%d workspace=%d\n", length, |
DPRINTF(("end pre-compile: length=%d workspace=%d\n", length, |
5841 |
cd->hwm = cworkspace; |
cd->hwm = cworkspace; |
5842 |
cd->req_varyopt = 0; |
cd->req_varyopt = 0; |
5843 |
cd->nopartial = FALSE; |
cd->nopartial = FALSE; |
5844 |
|
cd->had_accept = FALSE; |
5845 |
|
|
5846 |
/* Set up a starting, non-extracting bracket, then compile the expression. On |
/* Set up a starting, non-extracting bracket, then compile the expression. On |
5847 |
error, errorcode will be set non-zero, so we don't need to look at the result |
error, errorcode will be set non-zero, so we don't need to look at the result |
5851 |
code = (uschar *)codestart; |
code = (uschar *)codestart; |
5852 |
*code = OP_BRA; |
*code = OP_BRA; |
5853 |
(void)compile_regex(re->options, re->options & PCRE_IMS, &code, &ptr, |
(void)compile_regex(re->options, re->options & PCRE_IMS, &code, &ptr, |
5854 |
&errorcode, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, NULL); |
&errorcode, FALSE, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, NULL); |
5855 |
re->top_bracket = cd->bracount; |
re->top_bracket = cd->bracount; |
5856 |
re->top_backref = cd->top_backref; |
re->top_backref = cd->top_backref; |
5857 |
|
|
5858 |
if (cd->nopartial) re->options |= PCRE_NOPARTIAL; |
if (cd->nopartial) re->options |= PCRE_NOPARTIAL; |
5859 |
|
if (cd->had_accept) reqbyte = -1; /* Must disable after (*ACCEPT) */ |
5860 |
|
|
5861 |
/* If not reached end of pattern on success, there's an excess bracket. */ |
/* If not reached end of pattern on success, there's an excess bracket. */ |
5862 |
|
|