| 6 |
and semantics are as close as possible to those of the Perl 5 language. |
and semantics are as close as possible to those of the Perl 5 language. |
| 7 |
|
|
| 8 |
Written by Philip Hazel |
Written by Philip Hazel |
| 9 |
Copyright (c) 1997-2006 University of Cambridge |
Copyright (c) 1997-2007 University of Cambridge |
| 10 |
|
|
| 11 |
----------------------------------------------------------------------------- |
----------------------------------------------------------------------------- |
| 12 |
Redistribution and use in source and binary forms, with or without |
Redistribution and use in source and binary forms, with or without |
| 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 |
}; |
}; |
| 287 |
|
|
| 288 |
|
|
| 413 |
/* Definition to allow mutual recursion */ |
/* Definition to allow mutual recursion */ |
| 414 |
|
|
| 415 |
static BOOL |
static BOOL |
| 416 |
compile_regex(int, int, uschar **, const uschar **, int *, BOOL, int, int *, |
compile_regex(int, int, uschar **, const uschar **, int *, BOOL, BOOL, int, |
| 417 |
int *, branch_chain *, compile_data *, int *); |
int *, int *, branch_chain *, compile_data *, int *); |
| 418 |
|
|
| 419 |
|
|
| 420 |
|
|
| 492 |
|
|
| 493 |
/* \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 |
| 494 |
is an absolute backreference. If negative, it is a relative backreference. |
is an absolute backreference. If negative, it is a relative backreference. |
| 495 |
This is a Perl 5.10 feature. */ |
This is a Perl 5.10 feature. Perl 5.10 also supports \g{name} as a |
| 496 |
|
reference to a named group. This is part of Perl's movement towards a |
| 497 |
|
unified syntax for back references. As this is synonymous with \k{name}, we |
| 498 |
|
fudge it up by pretending it really was \k. */ |
| 499 |
|
|
| 500 |
case 'g': |
case 'g': |
| 501 |
if (ptr[1] == '{') |
if (ptr[1] == '{') |
| 502 |
{ |
{ |
| 503 |
|
const uschar *p; |
| 504 |
|
for (p = ptr+2; *p != 0 && *p != '}'; p++) |
| 505 |
|
if (*p != '-' && (digitab[*p] & ctype_digit) == 0) break; |
| 506 |
|
if (*p != 0 && *p != '}') |
| 507 |
|
{ |
| 508 |
|
c = -ESC_k; |
| 509 |
|
break; |
| 510 |
|
} |
| 511 |
braced = TRUE; |
braced = TRUE; |
| 512 |
ptr++; |
ptr++; |
| 513 |
} |
} |
| 735 |
*negptr = TRUE; |
*negptr = TRUE; |
| 736 |
ptr++; |
ptr++; |
| 737 |
} |
} |
| 738 |
for (i = 0; i < sizeof(name) - 1; i++) |
for (i = 0; i < (int)sizeof(name) - 1; i++) |
| 739 |
{ |
{ |
| 740 |
c = *(++ptr); |
c = *(++ptr); |
| 741 |
if (c == 0) goto ERROR_RETURN; |
if (c == 0) goto ERROR_RETURN; |
| 965 |
/* An opening parens must now be a real metacharacter */ |
/* An opening parens must now be a real metacharacter */ |
| 966 |
|
|
| 967 |
if (*ptr != '(') continue; |
if (*ptr != '(') continue; |
| 968 |
if (ptr[1] != '?') |
if (ptr[1] != '?' && ptr[1] != '*') |
| 969 |
{ |
{ |
| 970 |
count++; |
count++; |
| 971 |
if (name == NULL && count == lorn) return count; |
if (name == NULL && count == lorn) return count; |
| 1318 |
if (code[-1] >= 0xc0) code += _pcre_utf8_table4[code[-1] & 0x3f]; |
if (code[-1] >= 0xc0) code += _pcre_utf8_table4[code[-1] & 0x3f]; |
| 1319 |
break; |
break; |
| 1320 |
} |
} |
| 1321 |
#endif |
#endif |
| 1322 |
} |
} |
| 1323 |
} |
} |
| 1324 |
} |
} |
| 1383 |
if (code[-1] >= 0xc0) code += _pcre_utf8_table4[code[-1] & 0x3f]; |
if (code[-1] >= 0xc0) code += _pcre_utf8_table4[code[-1] & 0x3f]; |
| 1384 |
break; |
break; |
| 1385 |
} |
} |
| 1386 |
#endif |
#endif |
| 1387 |
} |
} |
| 1388 |
} |
} |
| 1389 |
} |
} |
| 1421 |
|
|
| 1422 |
c = *code; |
c = *code; |
| 1423 |
|
|
| 1424 |
if (c == OP_BRA || c == OP_CBRA || c == OP_ONCE) |
/* Groups with zero repeats can of course be empty; skip them. */ |
| 1425 |
|
|
| 1426 |
|
if (c == OP_BRAZERO || c == OP_BRAMINZERO) |
| 1427 |
|
{ |
| 1428 |
|
code += _pcre_OP_lengths[c]; |
| 1429 |
|
do code += GET(code, 1); while (*code == OP_ALT); |
| 1430 |
|
c = *code; |
| 1431 |
|
continue; |
| 1432 |
|
} |
| 1433 |
|
|
| 1434 |
|
/* For other groups, scan the branches. */ |
| 1435 |
|
|
| 1436 |
|
if (c == OP_BRA || c == OP_CBRA || c == OP_ONCE || c == OP_COND) |
| 1437 |
{ |
{ |
| 1438 |
BOOL empty_branch; |
BOOL empty_branch; |
| 1439 |
if (GET(code, 1) == 0) return TRUE; /* Hit unclosed bracket */ |
if (GET(code, 1) == 0) return TRUE; /* Hit unclosed bracket */ |
| 1449 |
} |
} |
| 1450 |
while (*code == OP_ALT); |
while (*code == OP_ALT); |
| 1451 |
if (!empty_branch) return FALSE; /* All branches are non-empty */ |
if (!empty_branch) return FALSE; /* All branches are non-empty */ |
| 1452 |
|
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]; |
|
| 1453 |
continue; |
continue; |
| 1454 |
} |
} |
| 1455 |
|
|
| 1983 |
case OP_NOT_WORDCHAR: |
case OP_NOT_WORDCHAR: |
| 1984 |
return next <= 127 && (cd->ctypes[next] & ctype_word) != 0; |
return next <= 127 && (cd->ctypes[next] & ctype_word) != 0; |
| 1985 |
|
|
| 1986 |
|
case OP_HSPACE: |
| 1987 |
|
case OP_NOT_HSPACE: |
| 1988 |
|
switch(next) |
| 1989 |
|
{ |
| 1990 |
|
case 0x09: |
| 1991 |
|
case 0x20: |
| 1992 |
|
case 0xa0: |
| 1993 |
|
case 0x1680: |
| 1994 |
|
case 0x180e: |
| 1995 |
|
case 0x2000: |
| 1996 |
|
case 0x2001: |
| 1997 |
|
case 0x2002: |
| 1998 |
|
case 0x2003: |
| 1999 |
|
case 0x2004: |
| 2000 |
|
case 0x2005: |
| 2001 |
|
case 0x2006: |
| 2002 |
|
case 0x2007: |
| 2003 |
|
case 0x2008: |
| 2004 |
|
case 0x2009: |
| 2005 |
|
case 0x200A: |
| 2006 |
|
case 0x202f: |
| 2007 |
|
case 0x205f: |
| 2008 |
|
case 0x3000: |
| 2009 |
|
return op_code != OP_HSPACE; |
| 2010 |
|
default: |
| 2011 |
|
return op_code == OP_HSPACE; |
| 2012 |
|
} |
| 2013 |
|
|
| 2014 |
|
case OP_VSPACE: |
| 2015 |
|
case OP_NOT_VSPACE: |
| 2016 |
|
switch(next) |
| 2017 |
|
{ |
| 2018 |
|
case 0x0a: |
| 2019 |
|
case 0x0b: |
| 2020 |
|
case 0x0c: |
| 2021 |
|
case 0x0d: |
| 2022 |
|
case 0x85: |
| 2023 |
|
case 0x2028: |
| 2024 |
|
case 0x2029: |
| 2025 |
|
return op_code != OP_VSPACE; |
| 2026 |
|
default: |
| 2027 |
|
return op_code == OP_VSPACE; |
| 2028 |
|
} |
| 2029 |
|
|
| 2030 |
default: |
default: |
| 2031 |
return FALSE; |
return FALSE; |
| 2032 |
} |
} |
| 2061 |
case ESC_W: |
case ESC_W: |
| 2062 |
return item <= 127 && (cd->ctypes[item] & ctype_word) != 0; |
return item <= 127 && (cd->ctypes[item] & ctype_word) != 0; |
| 2063 |
|
|
| 2064 |
|
case ESC_h: |
| 2065 |
|
case ESC_H: |
| 2066 |
|
switch(item) |
| 2067 |
|
{ |
| 2068 |
|
case 0x09: |
| 2069 |
|
case 0x20: |
| 2070 |
|
case 0xa0: |
| 2071 |
|
case 0x1680: |
| 2072 |
|
case 0x180e: |
| 2073 |
|
case 0x2000: |
| 2074 |
|
case 0x2001: |
| 2075 |
|
case 0x2002: |
| 2076 |
|
case 0x2003: |
| 2077 |
|
case 0x2004: |
| 2078 |
|
case 0x2005: |
| 2079 |
|
case 0x2006: |
| 2080 |
|
case 0x2007: |
| 2081 |
|
case 0x2008: |
| 2082 |
|
case 0x2009: |
| 2083 |
|
case 0x200A: |
| 2084 |
|
case 0x202f: |
| 2085 |
|
case 0x205f: |
| 2086 |
|
case 0x3000: |
| 2087 |
|
return -next != ESC_h; |
| 2088 |
|
default: |
| 2089 |
|
return -next == ESC_h; |
| 2090 |
|
} |
| 2091 |
|
|
| 2092 |
|
case ESC_v: |
| 2093 |
|
case ESC_V: |
| 2094 |
|
switch(item) |
| 2095 |
|
{ |
| 2096 |
|
case 0x0a: |
| 2097 |
|
case 0x0b: |
| 2098 |
|
case 0x0c: |
| 2099 |
|
case 0x0d: |
| 2100 |
|
case 0x85: |
| 2101 |
|
case 0x2028: |
| 2102 |
|
case 0x2029: |
| 2103 |
|
return -next != ESC_v; |
| 2104 |
|
default: |
| 2105 |
|
return -next == ESC_v; |
| 2106 |
|
} |
| 2107 |
|
|
| 2108 |
default: |
default: |
| 2109 |
return FALSE; |
return FALSE; |
| 2110 |
} |
} |
| 2111 |
|
|
| 2112 |
case OP_DIGIT: |
case OP_DIGIT: |
| 2113 |
return next == -ESC_D || next == -ESC_s || next == -ESC_W; |
return next == -ESC_D || next == -ESC_s || next == -ESC_W || |
| 2114 |
|
next == -ESC_h || next == -ESC_v; |
| 2115 |
|
|
| 2116 |
case OP_NOT_DIGIT: |
case OP_NOT_DIGIT: |
| 2117 |
return next == -ESC_d; |
return next == -ESC_d; |
| 2120 |
return next == -ESC_S || next == -ESC_d || next == -ESC_w; |
return next == -ESC_S || next == -ESC_d || next == -ESC_w; |
| 2121 |
|
|
| 2122 |
case OP_NOT_WHITESPACE: |
case OP_NOT_WHITESPACE: |
| 2123 |
return next == -ESC_s; |
return next == -ESC_s || next == -ESC_h || next == -ESC_v; |
| 2124 |
|
|
| 2125 |
|
case OP_HSPACE: |
| 2126 |
|
return next == -ESC_S || next == -ESC_H || next == -ESC_d || next == -ESC_w; |
| 2127 |
|
|
| 2128 |
|
case OP_NOT_HSPACE: |
| 2129 |
|
return next == -ESC_h; |
| 2130 |
|
|
| 2131 |
|
/* Can't have \S in here because VT matches \S (Perl anomaly) */ |
| 2132 |
|
case OP_VSPACE: |
| 2133 |
|
return next == -ESC_V || next == -ESC_d || next == -ESC_w; |
| 2134 |
|
|
| 2135 |
|
case OP_NOT_VSPACE: |
| 2136 |
|
return next == -ESC_v; |
| 2137 |
|
|
| 2138 |
case OP_WORDCHAR: |
case OP_WORDCHAR: |
| 2139 |
return next == -ESC_W || next == -ESC_s; |
return next == -ESC_W || next == -ESC_s || next == -ESC_h || next == -ESC_v; |
| 2140 |
|
|
| 2141 |
case OP_NOT_WORDCHAR: |
case OP_NOT_WORDCHAR: |
| 2142 |
return next == -ESC_w || next == -ESC_d; |
return next == -ESC_w || next == -ESC_d; |
| 2251 |
BOOL possessive_quantifier; |
BOOL possessive_quantifier; |
| 2252 |
BOOL is_quantifier; |
BOOL is_quantifier; |
| 2253 |
BOOL is_recurse; |
BOOL is_recurse; |
| 2254 |
|
BOOL reset_bracount; |
| 2255 |
int class_charcount; |
int class_charcount; |
| 2256 |
int class_lastchar; |
int class_lastchar; |
| 2257 |
int newoptions; |
int newoptions; |
| 2258 |
int recno; |
int recno; |
| 2259 |
|
int refsign; |
| 2260 |
int skipbytes; |
int skipbytes; |
| 2261 |
int subreqbyte; |
int subreqbyte; |
| 2262 |
int subfirstbyte; |
int subfirstbyte; |
| 2289 |
*/ |
*/ |
| 2290 |
|
|
| 2291 |
if (code < last_code) code = last_code; |
if (code < last_code) code = last_code; |
| 2292 |
|
|
| 2293 |
|
/* Paranoid check for integer overflow */ |
| 2294 |
|
|
| 2295 |
|
if (OFLOW_MAX - *lengthptr < code - last_code) |
| 2296 |
|
{ |
| 2297 |
|
*errorcodeptr = ERR20; |
| 2298 |
|
goto FAILED; |
| 2299 |
|
} |
| 2300 |
|
|
| 2301 |
*lengthptr += code - last_code; |
*lengthptr += code - last_code; |
| 2302 |
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)); |
| 2303 |
|
|
| 2410 |
*ptrptr = ptr; |
*ptrptr = ptr; |
| 2411 |
if (lengthptr != NULL) |
if (lengthptr != NULL) |
| 2412 |
{ |
{ |
| 2413 |
|
if (OFLOW_MAX - *lengthptr < code - last_code) |
| 2414 |
|
{ |
| 2415 |
|
*errorcodeptr = ERR20; |
| 2416 |
|
goto FAILED; |
| 2417 |
|
} |
| 2418 |
*lengthptr += code - last_code; /* To include callout length */ |
*lengthptr += code - last_code; /* To include callout length */ |
| 2419 |
DPRINTF((">> end branch\n")); |
DPRINTF((">> end branch\n")); |
| 2420 |
} |
} |
| 2477 |
goto FAILED; |
goto FAILED; |
| 2478 |
} |
} |
| 2479 |
|
|
| 2480 |
/* 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, |
| 2481 |
|
if the first few characters (either before or after ^) are \Q\E or \E we |
| 2482 |
|
skip them too. This makes for compatibility with Perl. */ |
| 2483 |
|
|
| 2484 |
if ((c = *(++ptr)) == '^') |
negate_class = FALSE; |
| 2485 |
|
for (;;) |
| 2486 |
{ |
{ |
|
negate_class = TRUE; |
|
| 2487 |
c = *(++ptr); |
c = *(++ptr); |
| 2488 |
} |
if (c == '\\') |
| 2489 |
else |
{ |
| 2490 |
{ |
if (ptr[1] == 'E') ptr++; |
| 2491 |
negate_class = FALSE; |
else if (strncmp((const char *)ptr+1, "Q\\E", 3) == 0) ptr += 3; |
| 2492 |
|
else break; |
| 2493 |
|
} |
| 2494 |
|
else if (!negate_class && c == '^') |
| 2495 |
|
negate_class = TRUE; |
| 2496 |
|
else break; |
| 2497 |
} |
} |
| 2498 |
|
|
| 2499 |
/* 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 |
| 2634 |
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 |
| 2635 |
case. Inside a class (and only there) it is treated as backspace. |
case. Inside a class (and only there) it is treated as backspace. |
| 2636 |
Elsewhere it marks a word boundary. Other escapes have preset maps ready |
Elsewhere it marks a word boundary. Other escapes have preset maps ready |
| 2637 |
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 |
| 2638 |
character in them, so set class_charcount bigger than one. */ |
character in them, so set class_charcount bigger than one. */ |
| 2639 |
|
|
| 2640 |
if (c == '\\') |
if (c == '\\') |
| 2702 |
else if (c == -ESC_d || c == -ESC_D || c == -ESC_w || |
else if (c == -ESC_d || c == -ESC_D || c == -ESC_w || |
| 2703 |
c == -ESC_W || c == -ESC_s || c == -ESC_S) continue; |
c == -ESC_W || c == -ESC_s || c == -ESC_S) continue; |
| 2704 |
|
|
| 2705 |
|
/* We need to deal with \H, \h, \V, and \v in both phases because |
| 2706 |
|
they use extra memory. */ |
| 2707 |
|
|
| 2708 |
|
if (-c == ESC_h) |
| 2709 |
|
{ |
| 2710 |
|
SETBIT(classbits, 0x09); /* VT */ |
| 2711 |
|
SETBIT(classbits, 0x20); /* SPACE */ |
| 2712 |
|
SETBIT(classbits, 0xa0); /* NSBP */ |
| 2713 |
|
#ifdef SUPPORT_UTF8 |
| 2714 |
|
if (utf8) |
| 2715 |
|
{ |
| 2716 |
|
class_utf8 = TRUE; |
| 2717 |
|
*class_utf8data++ = XCL_SINGLE; |
| 2718 |
|
class_utf8data += _pcre_ord2utf8(0x1680, class_utf8data); |
| 2719 |
|
*class_utf8data++ = XCL_SINGLE; |
| 2720 |
|
class_utf8data += _pcre_ord2utf8(0x180e, class_utf8data); |
| 2721 |
|
*class_utf8data++ = XCL_RANGE; |
| 2722 |
|
class_utf8data += _pcre_ord2utf8(0x2000, class_utf8data); |
| 2723 |
|
class_utf8data += _pcre_ord2utf8(0x200A, class_utf8data); |
| 2724 |
|
*class_utf8data++ = XCL_SINGLE; |
| 2725 |
|
class_utf8data += _pcre_ord2utf8(0x202f, class_utf8data); |
| 2726 |
|
*class_utf8data++ = XCL_SINGLE; |
| 2727 |
|
class_utf8data += _pcre_ord2utf8(0x205f, class_utf8data); |
| 2728 |
|
*class_utf8data++ = XCL_SINGLE; |
| 2729 |
|
class_utf8data += _pcre_ord2utf8(0x3000, class_utf8data); |
| 2730 |
|
} |
| 2731 |
|
#endif |
| 2732 |
|
continue; |
| 2733 |
|
} |
| 2734 |
|
|
| 2735 |
|
if (-c == ESC_H) |
| 2736 |
|
{ |
| 2737 |
|
for (c = 0; c < 32; c++) |
| 2738 |
|
{ |
| 2739 |
|
int x = 0xff; |
| 2740 |
|
switch (c) |
| 2741 |
|
{ |
| 2742 |
|
case 0x09/8: x ^= 1 << (0x09%8); break; |
| 2743 |
|
case 0x20/8: x ^= 1 << (0x20%8); break; |
| 2744 |
|
case 0xa0/8: x ^= 1 << (0xa0%8); break; |
| 2745 |
|
default: break; |
| 2746 |
|
} |
| 2747 |
|
classbits[c] |= x; |
| 2748 |
|
} |
| 2749 |
|
|
| 2750 |
|
#ifdef SUPPORT_UTF8 |
| 2751 |
|
if (utf8) |
| 2752 |
|
{ |
| 2753 |
|
class_utf8 = TRUE; |
| 2754 |
|
*class_utf8data++ = XCL_RANGE; |
| 2755 |
|
class_utf8data += _pcre_ord2utf8(0x0100, class_utf8data); |
| 2756 |
|
class_utf8data += _pcre_ord2utf8(0x167f, class_utf8data); |
| 2757 |
|
*class_utf8data++ = XCL_RANGE; |
| 2758 |
|
class_utf8data += _pcre_ord2utf8(0x1681, class_utf8data); |
| 2759 |
|
class_utf8data += _pcre_ord2utf8(0x180d, class_utf8data); |
| 2760 |
|
*class_utf8data++ = XCL_RANGE; |
| 2761 |
|
class_utf8data += _pcre_ord2utf8(0x180f, class_utf8data); |
| 2762 |
|
class_utf8data += _pcre_ord2utf8(0x1fff, class_utf8data); |
| 2763 |
|
*class_utf8data++ = XCL_RANGE; |
| 2764 |
|
class_utf8data += _pcre_ord2utf8(0x200B, class_utf8data); |
| 2765 |
|
class_utf8data += _pcre_ord2utf8(0x202e, class_utf8data); |
| 2766 |
|
*class_utf8data++ = XCL_RANGE; |
| 2767 |
|
class_utf8data += _pcre_ord2utf8(0x2030, class_utf8data); |
| 2768 |
|
class_utf8data += _pcre_ord2utf8(0x205e, class_utf8data); |
| 2769 |
|
*class_utf8data++ = XCL_RANGE; |
| 2770 |
|
class_utf8data += _pcre_ord2utf8(0x2060, class_utf8data); |
| 2771 |
|
class_utf8data += _pcre_ord2utf8(0x2fff, class_utf8data); |
| 2772 |
|
*class_utf8data++ = XCL_RANGE; |
| 2773 |
|
class_utf8data += _pcre_ord2utf8(0x3001, class_utf8data); |
| 2774 |
|
class_utf8data += _pcre_ord2utf8(0x7fffffff, class_utf8data); |
| 2775 |
|
} |
| 2776 |
|
#endif |
| 2777 |
|
continue; |
| 2778 |
|
} |
| 2779 |
|
|
| 2780 |
|
if (-c == ESC_v) |
| 2781 |
|
{ |
| 2782 |
|
SETBIT(classbits, 0x0a); /* LF */ |
| 2783 |
|
SETBIT(classbits, 0x0b); /* VT */ |
| 2784 |
|
SETBIT(classbits, 0x0c); /* FF */ |
| 2785 |
|
SETBIT(classbits, 0x0d); /* CR */ |
| 2786 |
|
SETBIT(classbits, 0x85); /* NEL */ |
| 2787 |
|
#ifdef SUPPORT_UTF8 |
| 2788 |
|
if (utf8) |
| 2789 |
|
{ |
| 2790 |
|
class_utf8 = TRUE; |
| 2791 |
|
*class_utf8data++ = XCL_RANGE; |
| 2792 |
|
class_utf8data += _pcre_ord2utf8(0x2028, class_utf8data); |
| 2793 |
|
class_utf8data += _pcre_ord2utf8(0x2029, class_utf8data); |
| 2794 |
|
} |
| 2795 |
|
#endif |
| 2796 |
|
continue; |
| 2797 |
|
} |
| 2798 |
|
|
| 2799 |
|
if (-c == ESC_V) |
| 2800 |
|
{ |
| 2801 |
|
for (c = 0; c < 32; c++) |
| 2802 |
|
{ |
| 2803 |
|
int x = 0xff; |
| 2804 |
|
switch (c) |
| 2805 |
|
{ |
| 2806 |
|
case 0x0a/8: x ^= 1 << (0x0a%8); |
| 2807 |
|
x ^= 1 << (0x0b%8); |
| 2808 |
|
x ^= 1 << (0x0c%8); |
| 2809 |
|
x ^= 1 << (0x0d%8); |
| 2810 |
|
break; |
| 2811 |
|
case 0x85/8: x ^= 1 << (0x85%8); break; |
| 2812 |
|
default: break; |
| 2813 |
|
} |
| 2814 |
|
classbits[c] |= x; |
| 2815 |
|
} |
| 2816 |
|
|
| 2817 |
|
#ifdef SUPPORT_UTF8 |
| 2818 |
|
if (utf8) |
| 2819 |
|
{ |
| 2820 |
|
class_utf8 = TRUE; |
| 2821 |
|
*class_utf8data++ = XCL_RANGE; |
| 2822 |
|
class_utf8data += _pcre_ord2utf8(0x0100, class_utf8data); |
| 2823 |
|
class_utf8data += _pcre_ord2utf8(0x2027, class_utf8data); |
| 2824 |
|
*class_utf8data++ = XCL_RANGE; |
| 2825 |
|
class_utf8data += _pcre_ord2utf8(0x2029, class_utf8data); |
| 2826 |
|
class_utf8data += _pcre_ord2utf8(0x7fffffff, class_utf8data); |
| 2827 |
|
} |
| 2828 |
|
#endif |
| 2829 |
|
continue; |
| 2830 |
|
} |
| 2831 |
|
|
| 2832 |
/* We need to deal with \P and \p in both phases. */ |
/* We need to deal with \P and \p in both phases. */ |
| 2833 |
|
|
| 2834 |
#ifdef SUPPORT_UCP |
#ifdef SUPPORT_UCP |
| 2969 |
unsigned int origd = d; |
unsigned int origd = d; |
| 2970 |
while (get_othercase_range(&cc, origd, &occ, &ocd)) |
while (get_othercase_range(&cc, origd, &occ, &ocd)) |
| 2971 |
{ |
{ |
| 2972 |
if (occ >= c && ocd <= d) continue; /* Skip embedded ranges */ |
if (occ >= (unsigned int)c && |
| 2973 |
|
ocd <= (unsigned int)d) |
| 2974 |
|
continue; /* Skip embedded ranges */ |
| 2975 |
|
|
| 2976 |
if (occ < c && ocd >= c - 1) /* Extend the basic range */ |
if (occ < (unsigned int)c && |
| 2977 |
|
ocd >= (unsigned int)c - 1) /* Extend the basic range */ |
| 2978 |
{ /* if there is overlap, */ |
{ /* if there is overlap, */ |
| 2979 |
c = occ; /* noting that if occ < c */ |
c = occ; /* noting that if occ < c */ |
| 2980 |
continue; /* we can't have ocd > d */ |
continue; /* we can't have ocd > d */ |
| 2981 |
} /* because a subrange is */ |
} /* because a subrange is */ |
| 2982 |
if (ocd > d && occ <= d + 1) /* always shorter than */ |
if (ocd > (unsigned int)d && |
| 2983 |
|
occ <= (unsigned int)d + 1) /* always shorter than */ |
| 2984 |
{ /* the basic range. */ |
{ /* the basic range. */ |
| 2985 |
d = ocd; |
d = ocd; |
| 2986 |
continue; |
continue; |
| 3576 |
goto FAILED; |
goto FAILED; |
| 3577 |
} |
} |
| 3578 |
|
|
|
/* This is a paranoid check to stop integer overflow later on */ |
|
|
|
|
|
if (len > MAX_DUPLENGTH) |
|
|
{ |
|
|
*errorcodeptr = ERR50; |
|
|
goto FAILED; |
|
|
} |
|
|
|
|
| 3579 |
/* 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 |
| 3580 |
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 |
| 3581 |
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 |
| 3664 |
if (repeat_min > 1) |
if (repeat_min > 1) |
| 3665 |
{ |
{ |
| 3666 |
/* 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 |
| 3667 |
just adjust the length as if we had. */ |
just adjust the length as if we had. Do some paranoid checks for |
| 3668 |
|
potential integer overflow. */ |
| 3669 |
|
|
| 3670 |
if (lengthptr != NULL) |
if (lengthptr != NULL) |
| 3671 |
*lengthptr += (repeat_min - 1)*length_prevgroup; |
{ |
| 3672 |
|
int delta = (repeat_min - 1)*length_prevgroup; |
| 3673 |
|
if ((double)(repeat_min - 1)*(double)length_prevgroup > |
| 3674 |
|
(double)INT_MAX || |
| 3675 |
|
OFLOW_MAX - *lengthptr < delta) |
| 3676 |
|
{ |
| 3677 |
|
*errorcodeptr = ERR20; |
| 3678 |
|
goto FAILED; |
| 3679 |
|
} |
| 3680 |
|
*lengthptr += delta; |
| 3681 |
|
} |
| 3682 |
|
|
| 3683 |
/* This is compiling for real */ |
/* This is compiling for real */ |
| 3684 |
|
|
| 3716 |
/* 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 |
| 3717 |
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 |
| 3718 |
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 |
| 3719 |
add 2 + 2*LINKSIZE to allow for the nesting that occurs. */ |
add 2 + 2*LINKSIZE to allow for the nesting that occurs. Do some |
| 3720 |
|
paranoid checks to avoid integer overflow. */ |
| 3721 |
|
|
| 3722 |
if (lengthptr != NULL && repeat_max > 0) |
if (lengthptr != NULL && repeat_max > 0) |
| 3723 |
*lengthptr += repeat_max * (length_prevgroup + 1 + 2 + 2*LINK_SIZE) - |
{ |
| 3724 |
2 - 2*LINK_SIZE; /* Last one doesn't nest */ |
int delta = repeat_max * (length_prevgroup + 1 + 2 + 2*LINK_SIZE) - |
| 3725 |
|
2 - 2*LINK_SIZE; /* Last one doesn't nest */ |
| 3726 |
|
if ((double)repeat_max * |
| 3727 |
|
(double)(length_prevgroup + 1 + 2 + 2*LINK_SIZE) |
| 3728 |
|
> (double)INT_MAX || |
| 3729 |
|
OFLOW_MAX - *lengthptr < delta) |
| 3730 |
|
{ |
| 3731 |
|
*errorcodeptr = ERR20; |
| 3732 |
|
goto FAILED; |
| 3733 |
|
} |
| 3734 |
|
*lengthptr += delta; |
| 3735 |
|
} |
| 3736 |
|
|
| 3737 |
/* This is compiling for real */ |
/* This is compiling for real */ |
| 3738 |
|
|
| 3884 |
/* ===================================================================*/ |
/* ===================================================================*/ |
| 3885 |
/* Start of nested parenthesized sub-expression, or comment or lookahead or |
/* Start of nested parenthesized sub-expression, or comment or lookahead or |
| 3886 |
lookbehind or option setting or condition or all the other extended |
lookbehind or option setting or condition or all the other extended |
| 3887 |
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. */ |
|
| 3888 |
|
|
| 3889 |
case '(': |
case '(': |
| 3890 |
newoptions = options; |
newoptions = options; |
| 3891 |
skipbytes = 0; |
skipbytes = 0; |
| 3892 |
bravalue = OP_CBRA; |
bravalue = OP_CBRA; |
| 3893 |
save_hwm = cd->hwm; |
save_hwm = cd->hwm; |
| 3894 |
|
reset_bracount = FALSE; |
| 3895 |
|
|
| 3896 |
|
/* First deal with various "verbs" that can be introduced by '*'. */ |
| 3897 |
|
|
| 3898 |
|
if (*(++ptr) == '*' && (cd->ctypes[ptr[1]] & ctype_letter) != 0) |
| 3899 |
|
{ |
| 3900 |
|
int i, namelen; |
| 3901 |
|
const uschar *name = ++ptr; |
| 3902 |
|
previous = NULL; |
| 3903 |
|
while ((cd->ctypes[*++ptr] & ctype_letter) != 0); |
| 3904 |
|
if (*ptr == ':') |
| 3905 |
|
{ |
| 3906 |
|
*errorcodeptr = ERR59; /* Not supported */ |
| 3907 |
|
goto FAILED; |
| 3908 |
|
} |
| 3909 |
|
if (*ptr != ')') |
| 3910 |
|
{ |
| 3911 |
|
*errorcodeptr = ERR60; |
| 3912 |
|
goto FAILED; |
| 3913 |
|
} |
| 3914 |
|
namelen = ptr - name; |
| 3915 |
|
for (i = 0; i < verbcount; i++) |
| 3916 |
|
{ |
| 3917 |
|
if (namelen == verbs[i].len && |
| 3918 |
|
strncmp((char *)name, verbs[i].name, namelen) == 0) |
| 3919 |
|
{ |
| 3920 |
|
*code = verbs[i].op; |
| 3921 |
|
if (*code++ == OP_ACCEPT) cd->had_accept = TRUE; |
| 3922 |
|
break; |
| 3923 |
|
} |
| 3924 |
|
} |
| 3925 |
|
if (i < verbcount) continue; |
| 3926 |
|
*errorcodeptr = ERR60; |
| 3927 |
|
goto FAILED; |
| 3928 |
|
} |
| 3929 |
|
|
| 3930 |
|
/* Deal with the extended parentheses; all are introduced by '?', and the |
| 3931 |
|
appearance of any of them means that this is not a capturing group. */ |
| 3932 |
|
|
| 3933 |
if (*(++ptr) == '?') |
else if (*ptr == '?') |
| 3934 |
{ |
{ |
| 3935 |
int i, set, unset, namelen; |
int i, set, unset, namelen; |
| 3936 |
int *optset; |
int *optset; |
| 3951 |
|
|
| 3952 |
|
|
| 3953 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
| 3954 |
|
case '|': /* Reset capture count for each branch */ |
| 3955 |
|
reset_bracount = TRUE; |
| 3956 |
|
/* Fall through */ |
| 3957 |
|
|
| 3958 |
|
/* ------------------------------------------------------------ */ |
| 3959 |
case ':': /* Non-capturing bracket */ |
case ':': /* Non-capturing bracket */ |
| 3960 |
bravalue = OP_BRA; |
bravalue = OP_BRA; |
| 3961 |
ptr++; |
ptr++; |
| 3991 |
|
|
| 3992 |
code[1+LINK_SIZE] = OP_CREF; |
code[1+LINK_SIZE] = OP_CREF; |
| 3993 |
skipbytes = 3; |
skipbytes = 3; |
| 3994 |
|
refsign = -1; |
| 3995 |
|
|
| 3996 |
/* Check for a test for recursion in a named group. */ |
/* Check for a test for recursion in a named group. */ |
| 3997 |
|
|
| 4015 |
terminator = '\''; |
terminator = '\''; |
| 4016 |
ptr++; |
ptr++; |
| 4017 |
} |
} |
| 4018 |
else terminator = 0; |
else |
| 4019 |
|
{ |
| 4020 |
|
terminator = 0; |
| 4021 |
|
if (ptr[1] == '-' || ptr[1] == '+') refsign = *(++ptr); |
| 4022 |
|
} |
| 4023 |
|
|
| 4024 |
/* 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 */ |
| 4025 |
|
|
| 4055 |
if (lengthptr != NULL) break; |
if (lengthptr != NULL) break; |
| 4056 |
|
|
| 4057 |
/* 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 |
| 4058 |
reference. */ |
reference. If the string started with "+" or "-" we require the rest to |
| 4059 |
|
be digits, in which case recno will be set. */ |
| 4060 |
|
|
| 4061 |
|
if (refsign > 0) |
| 4062 |
|
{ |
| 4063 |
|
if (recno <= 0) |
| 4064 |
|
{ |
| 4065 |
|
*errorcodeptr = ERR58; |
| 4066 |
|
goto FAILED; |
| 4067 |
|
} |
| 4068 |
|
if (refsign == '-') |
| 4069 |
|
{ |
| 4070 |
|
recno = cd->bracount - recno + 1; |
| 4071 |
|
if (recno <= 0) |
| 4072 |
|
{ |
| 4073 |
|
*errorcodeptr = ERR15; |
| 4074 |
|
goto FAILED; |
| 4075 |
|
} |
| 4076 |
|
} |
| 4077 |
|
else recno += cd->bracount; |
| 4078 |
|
PUT2(code, 2+LINK_SIZE, recno); |
| 4079 |
|
break; |
| 4080 |
|
} |
| 4081 |
|
|
| 4082 |
|
/* Otherwise (did not start with "+" or "-"), start by looking for the |
| 4083 |
|
name. */ |
| 4084 |
|
|
| 4085 |
slot = cd->name_table; |
slot = cd->name_table; |
| 4086 |
for (i = 0; i < cd->names_found; i++) |
for (i = 0; i < cd->names_found; i++) |
| 4172 |
|
|
| 4173 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
| 4174 |
case '!': /* Negative lookahead */ |
case '!': /* Negative lookahead */ |
|
bravalue = OP_ASSERT_NOT; |
|
| 4175 |
ptr++; |
ptr++; |
| 4176 |
|
if (*ptr == ')') /* Optimize (?!) */ |
| 4177 |
|
{ |
| 4178 |
|
*code++ = OP_FAIL; |
| 4179 |
|
previous = NULL; |
| 4180 |
|
continue; |
| 4181 |
|
} |
| 4182 |
|
bravalue = OP_ASSERT_NOT; |
| 4183 |
break; |
break; |
| 4184 |
|
|
| 4185 |
|
|
| 4405 |
|
|
| 4406 |
|
|
| 4407 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
| 4408 |
|
case '-': case '+': |
| 4409 |
case '0': case '1': case '2': case '3': case '4': /* Recursion or */ |
case '0': case '1': case '2': case '3': case '4': /* Recursion or */ |
| 4410 |
case '5': case '6': case '7': case '8': case '9': /* subroutine */ |
case '5': case '6': case '7': case '8': case '9': /* subroutine */ |
| 4411 |
{ |
{ |
| 4412 |
const uschar *called; |
const uschar *called; |
| 4413 |
|
|
| 4414 |
|
if ((refsign = *ptr) == '+') ptr++; |
| 4415 |
|
else if (refsign == '-') |
| 4416 |
|
{ |
| 4417 |
|
if ((digitab[ptr[1]] & ctype_digit) == 0) |
| 4418 |
|
goto OTHER_CHAR_AFTER_QUERY; |
| 4419 |
|
ptr++; |
| 4420 |
|
} |
| 4421 |
|
|
| 4422 |
recno = 0; |
recno = 0; |
| 4423 |
while((digitab[*ptr] & ctype_digit) != 0) |
while((digitab[*ptr] & ctype_digit) != 0) |
| 4424 |
recno = recno * 10 + *ptr++ - '0'; |
recno = recno * 10 + *ptr++ - '0'; |
| 4425 |
|
|
| 4426 |
if (*ptr != ')') |
if (*ptr != ')') |
| 4427 |
{ |
{ |
| 4428 |
*errorcodeptr = ERR29; |
*errorcodeptr = ERR29; |
| 4429 |
goto FAILED; |
goto FAILED; |
| 4430 |
} |
} |
| 4431 |
|
|
| 4432 |
|
if (refsign == '-') |
| 4433 |
|
{ |
| 4434 |
|
if (recno == 0) |
| 4435 |
|
{ |
| 4436 |
|
*errorcodeptr = ERR58; |
| 4437 |
|
goto FAILED; |
| 4438 |
|
} |
| 4439 |
|
recno = cd->bracount - recno + 1; |
| 4440 |
|
if (recno <= 0) |
| 4441 |
|
{ |
| 4442 |
|
*errorcodeptr = ERR15; |
| 4443 |
|
goto FAILED; |
| 4444 |
|
} |
| 4445 |
|
} |
| 4446 |
|
else if (refsign == '+') |
| 4447 |
|
{ |
| 4448 |
|
if (recno == 0) |
| 4449 |
|
{ |
| 4450 |
|
*errorcodeptr = ERR58; |
| 4451 |
|
goto FAILED; |
| 4452 |
|
} |
| 4453 |
|
recno += cd->bracount; |
| 4454 |
|
} |
| 4455 |
|
|
| 4456 |
/* Come here from code above that handles a named recursion */ |
/* Come here from code above that handles a named recursion */ |
| 4457 |
|
|
| 4458 |
HANDLE_RECURSION: |
HANDLE_RECURSION: |
| 4525 |
|
|
| 4526 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
| 4527 |
default: /* Other characters: check option setting */ |
default: /* Other characters: check option setting */ |
| 4528 |
|
OTHER_CHAR_AFTER_QUERY: |
| 4529 |
set = unset = 0; |
set = unset = 0; |
| 4530 |
optset = &set; |
optset = &set; |
| 4531 |
|
|
| 4660 |
errorcodeptr, /* Where to put an error message */ |
errorcodeptr, /* Where to put an error message */ |
| 4661 |
(bravalue == OP_ASSERTBACK || |
(bravalue == OP_ASSERTBACK || |
| 4662 |
bravalue == OP_ASSERTBACK_NOT), /* TRUE if back assert */ |
bravalue == OP_ASSERTBACK_NOT), /* TRUE if back assert */ |
| 4663 |
|
reset_bracount, /* True if (?| group */ |
| 4664 |
skipbytes, /* Skip over bracket number */ |
skipbytes, /* Skip over bracket number */ |
| 4665 |
&subfirstbyte, /* For possible first char */ |
&subfirstbyte, /* For possible first char */ |
| 4666 |
&subreqbyte, /* For possible last char */ |
&subreqbyte, /* For possible last char */ |
| 4677 |
is on the bracket. */ |
is on the bracket. */ |
| 4678 |
|
|
| 4679 |
/* 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 |
| 4680 |
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 |
| 4681 |
|
in the real compile phase, not in the pre-pass, where the whole group may |
| 4682 |
|
not be available. */ |
| 4683 |
|
|
| 4684 |
if (bravalue == OP_COND) |
if (bravalue == OP_COND && lengthptr == NULL) |
| 4685 |
{ |
{ |
| 4686 |
uschar *tc = code; |
uschar *tc = code; |
| 4687 |
int condcount = 0; |
int condcount = 0; |
| 4735 |
|
|
| 4736 |
if (lengthptr != NULL) |
if (lengthptr != NULL) |
| 4737 |
{ |
{ |
| 4738 |
|
if (OFLOW_MAX - *lengthptr < length_prevgroup - 2 - 2*LINK_SIZE) |
| 4739 |
|
{ |
| 4740 |
|
*errorcodeptr = ERR20; |
| 4741 |
|
goto FAILED; |
| 4742 |
|
} |
| 4743 |
*lengthptr += length_prevgroup - 2 - 2*LINK_SIZE; |
*lengthptr += length_prevgroup - 2 - 2*LINK_SIZE; |
| 4744 |
code++; |
code++; |
| 4745 |
PUTINC(code, 0, 1 + LINK_SIZE); |
PUTINC(code, 0, 1 + LINK_SIZE); |
| 4846 |
zerofirstbyte = firstbyte; |
zerofirstbyte = firstbyte; |
| 4847 |
zeroreqbyte = reqbyte; |
zeroreqbyte = reqbyte; |
| 4848 |
|
|
| 4849 |
/* \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). |
| 4850 |
|
We also support \k{name} (.NET syntax) */ |
| 4851 |
|
|
| 4852 |
if (-c == ESC_k && (ptr[1] == '<' || ptr[1] == '\'')) |
if (-c == ESC_k && (ptr[1] == '<' || ptr[1] == '\'' || ptr[1] == '{')) |
| 4853 |
{ |
{ |
| 4854 |
is_recurse = FALSE; |
is_recurse = FALSE; |
| 4855 |
terminator = (*(++ptr) == '<')? '>' : '\''; |
terminator = (*(++ptr) == '<')? '>' : (*ptr == '\'')? '\'' : '}'; |
| 4856 |
goto NAMED_REF_OR_RECURSE; |
goto NAMED_REF_OR_RECURSE; |
| 4857 |
} |
} |
| 4858 |
|
|
| 5018 |
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 |
| 5019 |
value of lengthptr distinguishes the two phases. |
value of lengthptr distinguishes the two phases. |
| 5020 |
|
|
| 5021 |
Argument: |
Arguments: |
| 5022 |
options option bits, including any changes for this subpattern |
options option bits, including any changes for this subpattern |
| 5023 |
oldims previous settings of ims option bits |
oldims previous settings of ims option bits |
| 5024 |
codeptr -> the address of the current code pointer |
codeptr -> the address of the current code pointer |
| 5025 |
ptrptr -> the address of the current pattern pointer |
ptrptr -> the address of the current pattern pointer |
| 5026 |
errorcodeptr -> pointer to error code variable |
errorcodeptr -> pointer to error code variable |
| 5027 |
lookbehind TRUE if this is a lookbehind assertion |
lookbehind TRUE if this is a lookbehind assertion |
| 5028 |
|
reset_bracount TRUE to reset the count for each branch |
| 5029 |
skipbytes skip this many bytes at start (for brackets and OP_COND) |
skipbytes skip this many bytes at start (for brackets and OP_COND) |
| 5030 |
firstbyteptr place to put the first required character, or a negative number |
firstbyteptr place to put the first required character, or a negative number |
| 5031 |
reqbyteptr place to put the last required character, or a negative number |
reqbyteptr place to put the last required character, or a negative number |
| 5039 |
|
|
| 5040 |
static BOOL |
static BOOL |
| 5041 |
compile_regex(int options, int oldims, uschar **codeptr, const uschar **ptrptr, |
compile_regex(int options, int oldims, uschar **codeptr, const uschar **ptrptr, |
| 5042 |
int *errorcodeptr, BOOL lookbehind, int skipbytes, int *firstbyteptr, |
int *errorcodeptr, BOOL lookbehind, BOOL reset_bracount, int skipbytes, |
| 5043 |
int *reqbyteptr, branch_chain *bcptr, compile_data *cd, int *lengthptr) |
int *firstbyteptr, int *reqbyteptr, branch_chain *bcptr, compile_data *cd, |
| 5044 |
|
int *lengthptr) |
| 5045 |
{ |
{ |
| 5046 |
const uschar *ptr = *ptrptr; |
const uschar *ptr = *ptrptr; |
| 5047 |
uschar *code = *codeptr; |
uschar *code = *codeptr; |
| 5051 |
int firstbyte, reqbyte; |
int firstbyte, reqbyte; |
| 5052 |
int branchfirstbyte, branchreqbyte; |
int branchfirstbyte, branchreqbyte; |
| 5053 |
int length; |
int length; |
| 5054 |
|
int orig_bracount; |
| 5055 |
|
int max_bracount; |
| 5056 |
branch_chain bc; |
branch_chain bc; |
| 5057 |
|
|
| 5058 |
bc.outer = bcptr; |
bc.outer = bcptr; |
| 5081 |
|
|
| 5082 |
/* Loop for each alternative branch */ |
/* Loop for each alternative branch */ |
| 5083 |
|
|
| 5084 |
|
orig_bracount = max_bracount = cd->bracount; |
| 5085 |
for (;;) |
for (;;) |
| 5086 |
{ |
{ |
| 5087 |
|
/* For a (?| group, reset the capturing bracket count so that each branch |
| 5088 |
|
uses the same numbers. */ |
| 5089 |
|
|
| 5090 |
|
if (reset_bracount) cd->bracount = orig_bracount; |
| 5091 |
|
|
| 5092 |
/* Handle a change of ims options at the start of the branch */ |
/* Handle a change of ims options at the start of the branch */ |
| 5093 |
|
|
| 5094 |
if ((options & PCRE_IMS) != oldims) |
if ((options & PCRE_IMS) != oldims) |
| 5118 |
return FALSE; |
return FALSE; |
| 5119 |
} |
} |
| 5120 |
|
|
| 5121 |
|
/* Keep the highest bracket count in case (?| was used and some branch |
| 5122 |
|
has fewer than the rest. */ |
| 5123 |
|
|
| 5124 |
|
if (cd->bracount > max_bracount) max_bracount = cd->bracount; |
| 5125 |
|
|
| 5126 |
/* 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. */ |
| 5127 |
|
|
| 5128 |
if (lengthptr == NULL) |
if (lengthptr == NULL) |
| 5186 |
} |
} |
| 5187 |
} |
} |
| 5188 |
|
|
| 5189 |
/* Reached end of expression, either ')' or end of pattern. Go back through |
/* Reached end of expression, either ')' or end of pattern. In the real |
| 5190 |
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 |
| 5191 |
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 |
| 5192 |
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 |
| 5193 |
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 |
| 5194 |
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, |
| 5195 |
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. |
| 5196 |
at the terminating char. */ |
Return leaving the pointer at the terminating char. */ |
| 5197 |
|
|
| 5198 |
if (*ptr != '|') |
if (*ptr != '|') |
| 5199 |
{ |
{ |
| 5200 |
int branch_length = code - last_branch; |
if (lengthptr == NULL) |
|
do |
|
| 5201 |
{ |
{ |
| 5202 |
int prev_length = GET(last_branch, 1); |
int branch_length = code - last_branch; |
| 5203 |
PUT(last_branch, 1, branch_length); |
do |
| 5204 |
branch_length = prev_length; |
{ |
| 5205 |
last_branch -= branch_length; |
int prev_length = GET(last_branch, 1); |
| 5206 |
|
PUT(last_branch, 1, branch_length); |
| 5207 |
|
branch_length = prev_length; |
| 5208 |
|
last_branch -= branch_length; |
| 5209 |
|
} |
| 5210 |
|
while (branch_length > 0); |
| 5211 |
} |
} |
|
while (branch_length > 0); |
|
| 5212 |
|
|
| 5213 |
/* Fill in the ket */ |
/* Fill in the ket */ |
| 5214 |
|
|
| 5225 |
length += 2; |
length += 2; |
| 5226 |
} |
} |
| 5227 |
|
|
| 5228 |
|
/* Retain the highest bracket number, in case resetting was used. */ |
| 5229 |
|
|
| 5230 |
|
cd->bracount = max_bracount; |
| 5231 |
|
|
| 5232 |
/* Set values to pass back */ |
/* Set values to pass back */ |
| 5233 |
|
|
| 5234 |
*codeptr = code; |
*codeptr = code; |
| 5235 |
*ptrptr = ptr; |
*ptrptr = ptr; |
| 5236 |
*firstbyteptr = firstbyte; |
*firstbyteptr = firstbyte; |
| 5237 |
*reqbyteptr = reqbyte; |
*reqbyteptr = reqbyte; |
| 5238 |
if (lengthptr != NULL) *lengthptr += length; |
if (lengthptr != NULL) |
| 5239 |
|
{ |
| 5240 |
|
if (OFLOW_MAX - *lengthptr < length) |
| 5241 |
|
{ |
| 5242 |
|
*errorcodeptr = ERR20; |
| 5243 |
|
return FALSE; |
| 5244 |
|
} |
| 5245 |
|
*lengthptr += length; |
| 5246 |
|
} |
| 5247 |
return TRUE; |
return TRUE; |
| 5248 |
} |
} |
| 5249 |
|
|
| 5250 |
/* 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 |
| 5251 |
|
pointer back to where it was for the start of the first branch. (That is, |
| 5252 |
|
pretend that each branch is the only one.) |
| 5253 |
|
|
| 5254 |
|
In the real compile phase, insert an ALT node. Its length field points back |
| 5255 |
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 |
| 5256 |
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 |
| 5257 |
zero offset until it is closed, making it possible to detect recursion. */ |
zero offset until it is closed, making it possible to detect recursion. */ |
| 5258 |
|
|
| 5259 |
*code = OP_ALT; |
if (lengthptr != NULL) |
| 5260 |
PUT(code, 1, code - last_branch); |
{ |
| 5261 |
bc.current = last_branch = code; |
code = *codeptr + 1 + LINK_SIZE + skipbytes; |
| 5262 |
code += 1 + LINK_SIZE; |
length += 1 + LINK_SIZE; |
| 5263 |
|
} |
| 5264 |
|
else |
| 5265 |
|
{ |
| 5266 |
|
*code = OP_ALT; |
| 5267 |
|
PUT(code, 1, code - last_branch); |
| 5268 |
|
bc.current = last_branch = code; |
| 5269 |
|
code += 1 + LINK_SIZE; |
| 5270 |
|
} |
| 5271 |
|
|
| 5272 |
ptr++; |
ptr++; |
|
length += 1 + LINK_SIZE; |
|
| 5273 |
} |
} |
| 5274 |
/* Control never reaches here */ |
/* Control never reaches here */ |
| 5275 |
} |
} |
| 5536 |
with errorptr and erroroffset set |
with errorptr and erroroffset set |
| 5537 |
*/ |
*/ |
| 5538 |
|
|
| 5539 |
PCRE_DATA_SCOPE pcre * |
PCRE_EXP_DEFN pcre * |
| 5540 |
pcre_compile(const char *pattern, int options, const char **errorptr, |
pcre_compile(const char *pattern, int options, const char **errorptr, |
| 5541 |
int *erroroffset, const unsigned char *tables) |
int *erroroffset, const unsigned char *tables) |
| 5542 |
{ |
{ |
| 5544 |
} |
} |
| 5545 |
|
|
| 5546 |
|
|
| 5547 |
PCRE_DATA_SCOPE pcre * |
PCRE_EXP_DEFN pcre * |
| 5548 |
pcre_compile2(const char *pattern, int options, int *errorcodeptr, |
pcre_compile2(const char *pattern, int options, int *errorcodeptr, |
| 5549 |
const char **errorptr, int *erroroffset, const unsigned char *tables) |
const char **errorptr, int *erroroffset, const unsigned char *tables) |
| 5550 |
{ |
{ |
| 5593 |
if (erroroffset == NULL) |
if (erroroffset == NULL) |
| 5594 |
{ |
{ |
| 5595 |
errorcode = ERR16; |
errorcode = ERR16; |
| 5596 |
goto PCRE_EARLY_ERROR_RETURN; |
goto PCRE_EARLY_ERROR_RETURN2; |
| 5597 |
} |
} |
| 5598 |
|
|
| 5599 |
*erroroffset = 0; |
*erroroffset = 0; |
| 5606 |
(*erroroffset = _pcre_valid_utf8((uschar *)pattern, -1)) >= 0) |
(*erroroffset = _pcre_valid_utf8((uschar *)pattern, -1)) >= 0) |
| 5607 |
{ |
{ |
| 5608 |
errorcode = ERR44; |
errorcode = ERR44; |
| 5609 |
goto PCRE_UTF8_ERROR_RETURN; |
goto PCRE_EARLY_ERROR_RETURN2; |
| 5610 |
} |
} |
| 5611 |
#else |
#else |
| 5612 |
if ((options & PCRE_UTF8) != 0) |
if ((options & PCRE_UTF8) != 0) |
| 5631 |
cd->ctypes = tables + ctypes_offset; |
cd->ctypes = tables + ctypes_offset; |
| 5632 |
|
|
| 5633 |
/* Handle different types of newline. The three bits give seven cases. The |
/* Handle different types of newline. The three bits give seven cases. The |
| 5634 |
current code allows for fixed one- or two-byte sequences, plus "any". */ |
current code allows for fixed one- or two-byte sequences, plus "any" and |
| 5635 |
|
"anycrlf". */ |
| 5636 |
|
|
| 5637 |
switch (options & (PCRE_NEWLINE_CRLF | PCRE_NEWLINE_ANY)) |
switch (options & (PCRE_NEWLINE_CRLF | PCRE_NEWLINE_ANY)) |
| 5638 |
{ |
{ |
| 5642 |
case PCRE_NEWLINE_CR+ |
case PCRE_NEWLINE_CR+ |
| 5643 |
PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break; |
PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break; |
| 5644 |
case PCRE_NEWLINE_ANY: newline = -1; break; |
case PCRE_NEWLINE_ANY: newline = -1; break; |
| 5645 |
|
case PCRE_NEWLINE_ANYCRLF: newline = -2; break; |
| 5646 |
default: errorcode = ERR56; goto PCRE_EARLY_ERROR_RETURN; |
default: errorcode = ERR56; goto PCRE_EARLY_ERROR_RETURN; |
| 5647 |
} |
} |
| 5648 |
|
|
| 5649 |
if (newline < 0) |
if (newline == -2) |
| 5650 |
|
{ |
| 5651 |
|
cd->nltype = NLTYPE_ANYCRLF; |
| 5652 |
|
} |
| 5653 |
|
else if (newline < 0) |
| 5654 |
{ |
{ |
| 5655 |
cd->nltype = NLTYPE_ANY; |
cd->nltype = NLTYPE_ANY; |
| 5656 |
} |
} |
| 5711 |
code = cworkspace; |
code = cworkspace; |
| 5712 |
*code = OP_BRA; |
*code = OP_BRA; |
| 5713 |
(void)compile_regex(cd->external_options, cd->external_options & PCRE_IMS, |
(void)compile_regex(cd->external_options, cd->external_options & PCRE_IMS, |
| 5714 |
&code, &ptr, &errorcode, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, &length); |
&code, &ptr, &errorcode, FALSE, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, |
| 5715 |
|
&length); |
| 5716 |
if (errorcode != 0) goto PCRE_EARLY_ERROR_RETURN; |
if (errorcode != 0) goto PCRE_EARLY_ERROR_RETURN; |
| 5717 |
|
|
| 5718 |
DPRINTF(("end pre-compile: length=%d workspace=%d\n", length, |
DPRINTF(("end pre-compile: length=%d workspace=%d\n", length, |
| 5771 |
cd->hwm = cworkspace; |
cd->hwm = cworkspace; |
| 5772 |
cd->req_varyopt = 0; |
cd->req_varyopt = 0; |
| 5773 |
cd->nopartial = FALSE; |
cd->nopartial = FALSE; |
| 5774 |
|
cd->had_accept = FALSE; |
| 5775 |
|
|
| 5776 |
/* Set up a starting, non-extracting bracket, then compile the expression. On |
/* Set up a starting, non-extracting bracket, then compile the expression. On |
| 5777 |
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 |
| 5781 |
code = (uschar *)codestart; |
code = (uschar *)codestart; |
| 5782 |
*code = OP_BRA; |
*code = OP_BRA; |
| 5783 |
(void)compile_regex(re->options, re->options & PCRE_IMS, &code, &ptr, |
(void)compile_regex(re->options, re->options & PCRE_IMS, &code, &ptr, |
| 5784 |
&errorcode, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, NULL); |
&errorcode, FALSE, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, NULL); |
| 5785 |
re->top_bracket = cd->bracount; |
re->top_bracket = cd->bracount; |
| 5786 |
re->top_backref = cd->top_backref; |
re->top_backref = cd->top_backref; |
| 5787 |
|
|
| 5788 |
if (cd->nopartial) re->options |= PCRE_NOPARTIAL; |
if (cd->nopartial) re->options |= PCRE_NOPARTIAL; |
| 5789 |
|
if (cd->had_accept) reqbyte = -1; /* Must disable after (*ACCEPT) */ |
| 5790 |
|
|
| 5791 |
/* 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. */ |
| 5792 |
|
|
| 5827 |
(pcre_free)(re); |
(pcre_free)(re); |
| 5828 |
PCRE_EARLY_ERROR_RETURN: |
PCRE_EARLY_ERROR_RETURN: |
| 5829 |
*erroroffset = ptr - (const uschar *)pattern; |
*erroroffset = ptr - (const uschar *)pattern; |
| 5830 |
#ifdef SUPPORT_UTF8 |
PCRE_EARLY_ERROR_RETURN2: |
|
PCRE_UTF8_ERROR_RETURN: |
|
|
#endif |
|
| 5831 |
*errorptr = error_texts[errorcode]; |
*errorptr = error_texts[errorcode]; |
| 5832 |
if (errorcodeptr != NULL) *errorcodeptr = errorcode; |
if (errorcodeptr != NULL) *errorcodeptr = errorcode; |
| 5833 |
return NULL; |
return NULL; |
| 5917 |
else printf("Req char = \\x%02x%s\n", ch, caseless); |
else printf("Req char = \\x%02x%s\n", ch, caseless); |
| 5918 |
} |
} |
| 5919 |
|
|
| 5920 |
pcre_printint(re, stdout); |
pcre_printint(re, stdout, TRUE); |
| 5921 |
|
|
| 5922 |
/* This check is done here in the debugging case so that the code that |
/* This check is done here in the debugging case so that the code that |
| 5923 |
was compiled can be seen. */ |
was compiled can be seen. */ |