| 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 |
|
|
| 69 |
/************************************************* |
/************************************************* |
| 70 |
* Code parameters and static tables * |
* Code parameters and static tables * |
| 71 |
*************************************************/ |
*************************************************/ |
| 95 |
0, 0, 0, 0, 0, 0, 0, 0, /* 0 - 7 */ |
0, 0, 0, 0, 0, 0, 0, 0, /* 0 - 7 */ |
| 96 |
0, 0, ':', ';', '<', '=', '>', '?', /* 8 - ? */ |
0, 0, ':', ';', '<', '=', '>', '?', /* 8 - ? */ |
| 97 |
'@', -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 */ |
| 98 |
0, 0, 0, 0, 0, 0, 0, 0, /* H - O */ |
-ESC_H, 0, 0, -ESC_K, 0, 0, 0, 0, /* H - O */ |
| 99 |
-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 */ |
| 100 |
-ESC_X, 0, -ESC_Z, '[', '\\', ']', '^', '_', /* X - _ */ |
-ESC_X, 0, -ESC_Z, '[', '\\', ']', '^', '_', /* X - _ */ |
| 101 |
'`', 7, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0, /* ` - g */ |
'`', 7, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0, /* ` - g */ |
| 102 |
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 */ |
| 103 |
-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 */ |
| 104 |
0, 0, -ESC_z /* x - z */ |
0, 0, -ESC_z /* x - z */ |
| 105 |
}; |
}; |
| 106 |
|
|
| 114 |
/* 70 */ 0, 0, 0, 0, 0, 0, 0, 0, |
/* 70 */ 0, 0, 0, 0, 0, 0, 0, 0, |
| 115 |
/* 78 */ 0, '`', ':', '#', '@', '\'', '=', '"', |
/* 78 */ 0, '`', ':', '#', '@', '\'', '=', '"', |
| 116 |
/* 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, |
| 117 |
/* 88 */ 0, 0, 0, '{', 0, 0, 0, 0, |
/* 88 */-ESC_h, 0, 0, '{', 0, 0, 0, 0, |
| 118 |
/* 90 */ 0, 0, -ESC_k, 'l', 0, ESC_n, 0, -ESC_p, |
/* 90 */ 0, 0, -ESC_k, 'l', 0, ESC_n, 0, -ESC_p, |
| 119 |
/* 98 */ 0, ESC_r, 0, '}', 0, 0, 0, 0, |
/* 98 */ 0, ESC_r, 0, '}', 0, 0, 0, 0, |
| 120 |
/* A0 */ 0, '~', -ESC_s, ESC_tee, 0, 0, -ESC_w, 0, |
/* A0 */ 0, '~', -ESC_s, ESC_tee, 0,-ESC_v, -ESC_w, 0, |
| 121 |
/* A8 */ 0,-ESC_z, 0, 0, 0, '[', 0, 0, |
/* A8 */ 0,-ESC_z, 0, 0, 0, '[', 0, 0, |
| 122 |
/* B0 */ 0, 0, 0, 0, 0, 0, 0, 0, |
/* B0 */ 0, 0, 0, 0, 0, 0, 0, 0, |
| 123 |
/* B8 */ 0, 0, 0, 0, 0, ']', '=', '-', |
/* B8 */ 0, 0, 0, 0, 0, ']', '=', '-', |
| 124 |
/* 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, |
| 125 |
/* C8 */ 0, 0, 0, 0, 0, 0, 0, 0, |
/* C8 */-ESC_H, 0, 0, 0, 0, 0, 0, 0, |
| 126 |
/* D0 */ '}', 0, 0, 0, 0, 0, 0, -ESC_P, |
/* D0 */ '}', 0, -ESC_K, 0, 0, 0, 0, -ESC_P, |
| 127 |
/* D8 */-ESC_Q,-ESC_R, 0, 0, 0, 0, 0, 0, |
/* D8 */-ESC_Q,-ESC_R, 0, 0, 0, 0, 0, 0, |
| 128 |
/* E0 */ '\\', 0, -ESC_S, 0, 0, 0, -ESC_W, -ESC_X, |
/* E0 */ '\\', 0, -ESC_S, 0, 0,-ESC_V, -ESC_W, -ESC_X, |
| 129 |
/* E8 */ 0,-ESC_Z, 0, 0, 0, 0, 0, 0, |
/* E8 */ 0,-ESC_Z, 0, 0, 0, 0, 0, 0, |
| 130 |
/* F0 */ 0, 0, 0, 0, 0, 0, 0, 0, |
/* F0 */ 0, 0, 0, 0, 0, 0, 0, 0, |
| 131 |
/* F8 */ 0, 0, 0, 0, 0, 0, 0, 0 |
/* F8 */ 0, 0, 0, 0, 0, 0, 0, 0 |
| 216 |
"malformed number or name after (?(", |
"malformed number or name after (?(", |
| 217 |
"conditional group contains more than two branches", |
"conditional group contains more than two branches", |
| 218 |
"assertion expected after (?(", |
"assertion expected after (?(", |
| 219 |
"(?R or (?digits must be followed by )", |
"(?R or (?[+-]digits must be followed by )", |
| 220 |
/* 30 */ |
/* 30 */ |
| 221 |
"unknown POSIX class name", |
"unknown POSIX class name", |
| 222 |
"POSIX collating elements are not supported", |
"POSIX collating elements are not supported", |
| 250 |
/* 55 */ |
/* 55 */ |
| 251 |
"repeating a DEFINE group is not allowed", |
"repeating a DEFINE group is not allowed", |
| 252 |
"inconsistent NEWLINE options", |
"inconsistent NEWLINE options", |
| 253 |
"\\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", |
| 254 |
|
"(?+ or (?- or (?(+ or (?(- must be followed by a non-zero number" |
| 255 |
}; |
}; |
| 256 |
|
|
| 257 |
|
|
| 382 |
/* Definition to allow mutual recursion */ |
/* Definition to allow mutual recursion */ |
| 383 |
|
|
| 384 |
static BOOL |
static BOOL |
| 385 |
compile_regex(int, int, uschar **, const uschar **, int *, BOOL, int, int *, |
compile_regex(int, int, uschar **, const uschar **, int *, BOOL, BOOL, int, |
| 386 |
int *, branch_chain *, compile_data *, int *); |
int *, int *, branch_chain *, compile_data *, int *); |
| 387 |
|
|
| 388 |
|
|
| 389 |
|
|
| 461 |
|
|
| 462 |
/* \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 |
| 463 |
is an absolute backreference. If negative, it is a relative backreference. |
is an absolute backreference. If negative, it is a relative backreference. |
| 464 |
This is a Perl 5.10 feature. */ |
This is a Perl 5.10 feature. Perl 5.10 also supports \g{name} as a |
| 465 |
|
reference to a named group. This is part of Perl's movement towards a |
| 466 |
|
unified syntax for back references. As this is synonymous with \k{name}, we |
| 467 |
|
fudge it up by pretending it really was \k. */ |
| 468 |
|
|
| 469 |
case 'g': |
case 'g': |
| 470 |
if (ptr[1] == '{') |
if (ptr[1] == '{') |
| 471 |
{ |
{ |
| 472 |
|
const uschar *p; |
| 473 |
|
for (p = ptr+2; *p != 0 && *p != '}'; p++) |
| 474 |
|
if (*p != '-' && (digitab[*p] & ctype_digit) == 0) break; |
| 475 |
|
if (*p != 0 && *p != '}') |
| 476 |
|
{ |
| 477 |
|
c = -ESC_k; |
| 478 |
|
break; |
| 479 |
|
} |
| 480 |
braced = TRUE; |
braced = TRUE; |
| 481 |
ptr++; |
ptr++; |
| 482 |
} |
} |
| 704 |
*negptr = TRUE; |
*negptr = TRUE; |
| 705 |
ptr++; |
ptr++; |
| 706 |
} |
} |
| 707 |
for (i = 0; i < sizeof(name) - 1; i++) |
for (i = 0; i < (int)sizeof(name) - 1; i++) |
| 708 |
{ |
{ |
| 709 |
c = *(++ptr); |
c = *(++ptr); |
| 710 |
if (c == 0) goto ERROR_RETURN; |
if (c == 0) goto ERROR_RETURN; |
| 1266 |
else |
else |
| 1267 |
{ |
{ |
| 1268 |
code += _pcre_OP_lengths[c]; |
code += _pcre_OP_lengths[c]; |
| 1269 |
|
#ifdef SUPPORT_UTF8 |
| 1270 |
if (utf8) switch(c) |
if (utf8) switch(c) |
| 1271 |
{ |
{ |
| 1272 |
case OP_CHAR: |
case OP_CHAR: |
| 1287 |
if (code[-1] >= 0xc0) code += _pcre_utf8_table4[code[-1] & 0x3f]; |
if (code[-1] >= 0xc0) code += _pcre_utf8_table4[code[-1] & 0x3f]; |
| 1288 |
break; |
break; |
| 1289 |
} |
} |
| 1290 |
|
#endif |
| 1291 |
} |
} |
| 1292 |
} |
} |
| 1293 |
} |
} |
| 1331 |
else |
else |
| 1332 |
{ |
{ |
| 1333 |
code += _pcre_OP_lengths[c]; |
code += _pcre_OP_lengths[c]; |
| 1334 |
|
#ifdef SUPPORT_UTF8 |
| 1335 |
if (utf8) switch(c) |
if (utf8) switch(c) |
| 1336 |
{ |
{ |
| 1337 |
case OP_CHAR: |
case OP_CHAR: |
| 1352 |
if (code[-1] >= 0xc0) code += _pcre_utf8_table4[code[-1] & 0x3f]; |
if (code[-1] >= 0xc0) code += _pcre_utf8_table4[code[-1] & 0x3f]; |
| 1353 |
break; |
break; |
| 1354 |
} |
} |
| 1355 |
|
#endif |
| 1356 |
} |
} |
| 1357 |
} |
} |
| 1358 |
} |
} |
| 1390 |
|
|
| 1391 |
c = *code; |
c = *code; |
| 1392 |
|
|
| 1393 |
|
/* Groups with zero repeats can of course be empty; skip them. */ |
| 1394 |
|
|
| 1395 |
|
if (c == OP_BRAZERO || c == OP_BRAMINZERO) |
| 1396 |
|
{ |
| 1397 |
|
code += _pcre_OP_lengths[c]; |
| 1398 |
|
do code += GET(code, 1); while (*code == OP_ALT); |
| 1399 |
|
c = *code; |
| 1400 |
|
continue; |
| 1401 |
|
} |
| 1402 |
|
|
| 1403 |
|
/* For other groups, scan the branches. */ |
| 1404 |
|
|
| 1405 |
if (c == OP_BRA || c == OP_CBRA || c == OP_ONCE) |
if (c == OP_BRA || c == OP_CBRA || c == OP_ONCE) |
| 1406 |
{ |
{ |
| 1407 |
BOOL empty_branch; |
BOOL empty_branch; |
| 1418 |
} |
} |
| 1419 |
while (*code == OP_ALT); |
while (*code == OP_ALT); |
| 1420 |
if (!empty_branch) return FALSE; /* All branches are non-empty */ |
if (!empty_branch) return FALSE; /* All branches are non-empty */ |
| 1421 |
|
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]; |
|
| 1422 |
continue; |
continue; |
| 1423 |
} |
} |
| 1424 |
|
|
| 1952 |
case OP_NOT_WORDCHAR: |
case OP_NOT_WORDCHAR: |
| 1953 |
return next <= 127 && (cd->ctypes[next] & ctype_word) != 0; |
return next <= 127 && (cd->ctypes[next] & ctype_word) != 0; |
| 1954 |
|
|
| 1955 |
|
case OP_HSPACE: |
| 1956 |
|
case OP_NOT_HSPACE: |
| 1957 |
|
switch(next) |
| 1958 |
|
{ |
| 1959 |
|
case 0x09: |
| 1960 |
|
case 0x20: |
| 1961 |
|
case 0xa0: |
| 1962 |
|
case 0x1680: |
| 1963 |
|
case 0x180e: |
| 1964 |
|
case 0x2000: |
| 1965 |
|
case 0x2001: |
| 1966 |
|
case 0x2002: |
| 1967 |
|
case 0x2003: |
| 1968 |
|
case 0x2004: |
| 1969 |
|
case 0x2005: |
| 1970 |
|
case 0x2006: |
| 1971 |
|
case 0x2007: |
| 1972 |
|
case 0x2008: |
| 1973 |
|
case 0x2009: |
| 1974 |
|
case 0x200A: |
| 1975 |
|
case 0x202f: |
| 1976 |
|
case 0x205f: |
| 1977 |
|
case 0x3000: |
| 1978 |
|
return op_code != OP_HSPACE; |
| 1979 |
|
default: |
| 1980 |
|
return op_code == OP_HSPACE; |
| 1981 |
|
} |
| 1982 |
|
|
| 1983 |
|
case OP_VSPACE: |
| 1984 |
|
case OP_NOT_VSPACE: |
| 1985 |
|
switch(next) |
| 1986 |
|
{ |
| 1987 |
|
case 0x0a: |
| 1988 |
|
case 0x0b: |
| 1989 |
|
case 0x0c: |
| 1990 |
|
case 0x0d: |
| 1991 |
|
case 0x85: |
| 1992 |
|
case 0x2028: |
| 1993 |
|
case 0x2029: |
| 1994 |
|
return op_code != OP_VSPACE; |
| 1995 |
|
default: |
| 1996 |
|
return op_code == OP_VSPACE; |
| 1997 |
|
} |
| 1998 |
|
|
| 1999 |
default: |
default: |
| 2000 |
return FALSE; |
return FALSE; |
| 2001 |
} |
} |
| 2030 |
case ESC_W: |
case ESC_W: |
| 2031 |
return item <= 127 && (cd->ctypes[item] & ctype_word) != 0; |
return item <= 127 && (cd->ctypes[item] & ctype_word) != 0; |
| 2032 |
|
|
| 2033 |
|
case ESC_h: |
| 2034 |
|
case ESC_H: |
| 2035 |
|
switch(item) |
| 2036 |
|
{ |
| 2037 |
|
case 0x09: |
| 2038 |
|
case 0x20: |
| 2039 |
|
case 0xa0: |
| 2040 |
|
case 0x1680: |
| 2041 |
|
case 0x180e: |
| 2042 |
|
case 0x2000: |
| 2043 |
|
case 0x2001: |
| 2044 |
|
case 0x2002: |
| 2045 |
|
case 0x2003: |
| 2046 |
|
case 0x2004: |
| 2047 |
|
case 0x2005: |
| 2048 |
|
case 0x2006: |
| 2049 |
|
case 0x2007: |
| 2050 |
|
case 0x2008: |
| 2051 |
|
case 0x2009: |
| 2052 |
|
case 0x200A: |
| 2053 |
|
case 0x202f: |
| 2054 |
|
case 0x205f: |
| 2055 |
|
case 0x3000: |
| 2056 |
|
return -next != ESC_h; |
| 2057 |
|
default: |
| 2058 |
|
return -next == ESC_h; |
| 2059 |
|
} |
| 2060 |
|
|
| 2061 |
|
case ESC_v: |
| 2062 |
|
case ESC_V: |
| 2063 |
|
switch(item) |
| 2064 |
|
{ |
| 2065 |
|
case 0x0a: |
| 2066 |
|
case 0x0b: |
| 2067 |
|
case 0x0c: |
| 2068 |
|
case 0x0d: |
| 2069 |
|
case 0x85: |
| 2070 |
|
case 0x2028: |
| 2071 |
|
case 0x2029: |
| 2072 |
|
return -next != ESC_v; |
| 2073 |
|
default: |
| 2074 |
|
return -next == ESC_v; |
| 2075 |
|
} |
| 2076 |
|
|
| 2077 |
default: |
default: |
| 2078 |
return FALSE; |
return FALSE; |
| 2079 |
} |
} |
| 2080 |
|
|
| 2081 |
case OP_DIGIT: |
case OP_DIGIT: |
| 2082 |
return next == -ESC_D || next == -ESC_s || next == -ESC_W; |
return next == -ESC_D || next == -ESC_s || next == -ESC_W || |
| 2083 |
|
next == -ESC_h || next == -ESC_v; |
| 2084 |
|
|
| 2085 |
case OP_NOT_DIGIT: |
case OP_NOT_DIGIT: |
| 2086 |
return next == -ESC_d; |
return next == -ESC_d; |
| 2089 |
return next == -ESC_S || next == -ESC_d || next == -ESC_w; |
return next == -ESC_S || next == -ESC_d || next == -ESC_w; |
| 2090 |
|
|
| 2091 |
case OP_NOT_WHITESPACE: |
case OP_NOT_WHITESPACE: |
| 2092 |
return next == -ESC_s; |
return next == -ESC_s || next == -ESC_h || next == -ESC_v; |
| 2093 |
|
|
| 2094 |
|
case OP_HSPACE: |
| 2095 |
|
return next == -ESC_S || next == -ESC_H || next == -ESC_d || next == -ESC_w; |
| 2096 |
|
|
| 2097 |
|
case OP_NOT_HSPACE: |
| 2098 |
|
return next == -ESC_h; |
| 2099 |
|
|
| 2100 |
|
/* Can't have \S in here because VT matches \S (Perl anomaly) */ |
| 2101 |
|
case OP_VSPACE: |
| 2102 |
|
return next == -ESC_V || next == -ESC_d || next == -ESC_w; |
| 2103 |
|
|
| 2104 |
|
case OP_NOT_VSPACE: |
| 2105 |
|
return next == -ESC_v; |
| 2106 |
|
|
| 2107 |
case OP_WORDCHAR: |
case OP_WORDCHAR: |
| 2108 |
return next == -ESC_W || next == -ESC_s; |
return next == -ESC_W || next == -ESC_s || next == -ESC_h || next == -ESC_v; |
| 2109 |
|
|
| 2110 |
case OP_NOT_WORDCHAR: |
case OP_NOT_WORDCHAR: |
| 2111 |
return next == -ESC_w || next == -ESC_d; |
return next == -ESC_w || next == -ESC_d; |
| 2220 |
BOOL possessive_quantifier; |
BOOL possessive_quantifier; |
| 2221 |
BOOL is_quantifier; |
BOOL is_quantifier; |
| 2222 |
BOOL is_recurse; |
BOOL is_recurse; |
| 2223 |
|
BOOL reset_bracount; |
| 2224 |
int class_charcount; |
int class_charcount; |
| 2225 |
int class_lastchar; |
int class_lastchar; |
| 2226 |
int newoptions; |
int newoptions; |
| 2227 |
int recno; |
int recno; |
| 2228 |
|
int refsign; |
| 2229 |
int skipbytes; |
int skipbytes; |
| 2230 |
int subreqbyte; |
int subreqbyte; |
| 2231 |
int subfirstbyte; |
int subfirstbyte; |
| 2650 |
else if (c == -ESC_d || c == -ESC_D || c == -ESC_w || |
else if (c == -ESC_d || c == -ESC_D || c == -ESC_w || |
| 2651 |
c == -ESC_W || c == -ESC_s || c == -ESC_S) continue; |
c == -ESC_W || c == -ESC_s || c == -ESC_S) continue; |
| 2652 |
|
|
| 2653 |
|
/* We need to deal with \H, \h, \V, and \v in both phases because |
| 2654 |
|
they use extra memory. */ |
| 2655 |
|
|
| 2656 |
|
if (-c == ESC_h) |
| 2657 |
|
{ |
| 2658 |
|
SETBIT(classbits, 0x09); /* VT */ |
| 2659 |
|
SETBIT(classbits, 0x20); /* SPACE */ |
| 2660 |
|
SETBIT(classbits, 0xa0); /* NSBP */ |
| 2661 |
|
#ifdef SUPPORT_UTF8 |
| 2662 |
|
if (utf8) |
| 2663 |
|
{ |
| 2664 |
|
class_utf8 = TRUE; |
| 2665 |
|
*class_utf8data++ = XCL_SINGLE; |
| 2666 |
|
class_utf8data += _pcre_ord2utf8(0x1680, class_utf8data); |
| 2667 |
|
*class_utf8data++ = XCL_SINGLE; |
| 2668 |
|
class_utf8data += _pcre_ord2utf8(0x180e, class_utf8data); |
| 2669 |
|
*class_utf8data++ = XCL_RANGE; |
| 2670 |
|
class_utf8data += _pcre_ord2utf8(0x2000, class_utf8data); |
| 2671 |
|
class_utf8data += _pcre_ord2utf8(0x200A, class_utf8data); |
| 2672 |
|
*class_utf8data++ = XCL_SINGLE; |
| 2673 |
|
class_utf8data += _pcre_ord2utf8(0x202f, class_utf8data); |
| 2674 |
|
*class_utf8data++ = XCL_SINGLE; |
| 2675 |
|
class_utf8data += _pcre_ord2utf8(0x205f, class_utf8data); |
| 2676 |
|
*class_utf8data++ = XCL_SINGLE; |
| 2677 |
|
class_utf8data += _pcre_ord2utf8(0x3000, class_utf8data); |
| 2678 |
|
} |
| 2679 |
|
#endif |
| 2680 |
|
continue; |
| 2681 |
|
} |
| 2682 |
|
|
| 2683 |
|
if (-c == ESC_H) |
| 2684 |
|
{ |
| 2685 |
|
for (c = 0; c < 32; c++) |
| 2686 |
|
{ |
| 2687 |
|
int x = 0xff; |
| 2688 |
|
switch (c) |
| 2689 |
|
{ |
| 2690 |
|
case 0x09/8: x ^= 1 << (0x09%8); break; |
| 2691 |
|
case 0x20/8: x ^= 1 << (0x20%8); break; |
| 2692 |
|
case 0xa0/8: x ^= 1 << (0xa0%8); break; |
| 2693 |
|
default: break; |
| 2694 |
|
} |
| 2695 |
|
classbits[c] |= x; |
| 2696 |
|
} |
| 2697 |
|
|
| 2698 |
|
#ifdef SUPPORT_UTF8 |
| 2699 |
|
if (utf8) |
| 2700 |
|
{ |
| 2701 |
|
class_utf8 = TRUE; |
| 2702 |
|
*class_utf8data++ = XCL_RANGE; |
| 2703 |
|
class_utf8data += _pcre_ord2utf8(0x0100, class_utf8data); |
| 2704 |
|
class_utf8data += _pcre_ord2utf8(0x167f, class_utf8data); |
| 2705 |
|
*class_utf8data++ = XCL_RANGE; |
| 2706 |
|
class_utf8data += _pcre_ord2utf8(0x1681, class_utf8data); |
| 2707 |
|
class_utf8data += _pcre_ord2utf8(0x180d, class_utf8data); |
| 2708 |
|
*class_utf8data++ = XCL_RANGE; |
| 2709 |
|
class_utf8data += _pcre_ord2utf8(0x180f, class_utf8data); |
| 2710 |
|
class_utf8data += _pcre_ord2utf8(0x1fff, class_utf8data); |
| 2711 |
|
*class_utf8data++ = XCL_RANGE; |
| 2712 |
|
class_utf8data += _pcre_ord2utf8(0x200B, class_utf8data); |
| 2713 |
|
class_utf8data += _pcre_ord2utf8(0x202e, class_utf8data); |
| 2714 |
|
*class_utf8data++ = XCL_RANGE; |
| 2715 |
|
class_utf8data += _pcre_ord2utf8(0x2030, class_utf8data); |
| 2716 |
|
class_utf8data += _pcre_ord2utf8(0x205e, class_utf8data); |
| 2717 |
|
*class_utf8data++ = XCL_RANGE; |
| 2718 |
|
class_utf8data += _pcre_ord2utf8(0x2060, class_utf8data); |
| 2719 |
|
class_utf8data += _pcre_ord2utf8(0x2fff, class_utf8data); |
| 2720 |
|
*class_utf8data++ = XCL_RANGE; |
| 2721 |
|
class_utf8data += _pcre_ord2utf8(0x3001, class_utf8data); |
| 2722 |
|
class_utf8data += _pcre_ord2utf8(0x7fffffff, class_utf8data); |
| 2723 |
|
} |
| 2724 |
|
#endif |
| 2725 |
|
continue; |
| 2726 |
|
} |
| 2727 |
|
|
| 2728 |
|
if (-c == ESC_v) |
| 2729 |
|
{ |
| 2730 |
|
SETBIT(classbits, 0x0a); /* LF */ |
| 2731 |
|
SETBIT(classbits, 0x0b); /* VT */ |
| 2732 |
|
SETBIT(classbits, 0x0c); /* FF */ |
| 2733 |
|
SETBIT(classbits, 0x0d); /* CR */ |
| 2734 |
|
SETBIT(classbits, 0x85); /* NEL */ |
| 2735 |
|
#ifdef SUPPORT_UTF8 |
| 2736 |
|
if (utf8) |
| 2737 |
|
{ |
| 2738 |
|
class_utf8 = TRUE; |
| 2739 |
|
*class_utf8data++ = XCL_RANGE; |
| 2740 |
|
class_utf8data += _pcre_ord2utf8(0x2028, class_utf8data); |
| 2741 |
|
class_utf8data += _pcre_ord2utf8(0x2029, class_utf8data); |
| 2742 |
|
} |
| 2743 |
|
#endif |
| 2744 |
|
continue; |
| 2745 |
|
} |
| 2746 |
|
|
| 2747 |
|
if (-c == ESC_V) |
| 2748 |
|
{ |
| 2749 |
|
for (c = 0; c < 32; c++) |
| 2750 |
|
{ |
| 2751 |
|
int x = 0xff; |
| 2752 |
|
switch (c) |
| 2753 |
|
{ |
| 2754 |
|
case 0x0a/8: x ^= 1 << (0x0a%8); |
| 2755 |
|
x ^= 1 << (0x0b%8); |
| 2756 |
|
x ^= 1 << (0x0c%8); |
| 2757 |
|
x ^= 1 << (0x0d%8); |
| 2758 |
|
break; |
| 2759 |
|
case 0x85/8: x ^= 1 << (0x85%8); break; |
| 2760 |
|
default: break; |
| 2761 |
|
} |
| 2762 |
|
classbits[c] |= x; |
| 2763 |
|
} |
| 2764 |
|
|
| 2765 |
|
#ifdef SUPPORT_UTF8 |
| 2766 |
|
if (utf8) |
| 2767 |
|
{ |
| 2768 |
|
class_utf8 = TRUE; |
| 2769 |
|
*class_utf8data++ = XCL_RANGE; |
| 2770 |
|
class_utf8data += _pcre_ord2utf8(0x0100, class_utf8data); |
| 2771 |
|
class_utf8data += _pcre_ord2utf8(0x2027, class_utf8data); |
| 2772 |
|
*class_utf8data++ = XCL_RANGE; |
| 2773 |
|
class_utf8data += _pcre_ord2utf8(0x2029, class_utf8data); |
| 2774 |
|
class_utf8data += _pcre_ord2utf8(0x7fffffff, class_utf8data); |
| 2775 |
|
} |
| 2776 |
|
#endif |
| 2777 |
|
continue; |
| 2778 |
|
} |
| 2779 |
|
|
| 2780 |
/* We need to deal with \P and \p in both phases. */ |
/* We need to deal with \P and \p in both phases. */ |
| 2781 |
|
|
| 2782 |
#ifdef SUPPORT_UCP |
#ifdef SUPPORT_UCP |
| 2917 |
unsigned int origd = d; |
unsigned int origd = d; |
| 2918 |
while (get_othercase_range(&cc, origd, &occ, &ocd)) |
while (get_othercase_range(&cc, origd, &occ, &ocd)) |
| 2919 |
{ |
{ |
| 2920 |
if (occ >= c && ocd <= d) continue; /* Skip embedded ranges */ |
if (occ >= (unsigned int)c && |
| 2921 |
|
ocd <= (unsigned int)d) |
| 2922 |
|
continue; /* Skip embedded ranges */ |
| 2923 |
|
|
| 2924 |
if (occ < c && ocd >= c - 1) /* Extend the basic range */ |
if (occ < (unsigned int)c && |
| 2925 |
|
ocd >= (unsigned int)c - 1) /* Extend the basic range */ |
| 2926 |
{ /* if there is overlap, */ |
{ /* if there is overlap, */ |
| 2927 |
c = occ; /* noting that if occ < c */ |
c = occ; /* noting that if occ < c */ |
| 2928 |
continue; /* we can't have ocd > d */ |
continue; /* we can't have ocd > d */ |
| 2929 |
} /* because a subrange is */ |
} /* because a subrange is */ |
| 2930 |
if (ocd > d && occ <= d + 1) /* always shorter than */ |
if (ocd > (unsigned int)d && |
| 2931 |
|
occ <= (unsigned int)d + 1) /* always shorter than */ |
| 2932 |
{ /* the basic range. */ |
{ /* the basic range. */ |
| 2933 |
d = ocd; |
d = ocd; |
| 2934 |
continue; |
continue; |
| 3826 |
skipbytes = 0; |
skipbytes = 0; |
| 3827 |
bravalue = OP_CBRA; |
bravalue = OP_CBRA; |
| 3828 |
save_hwm = cd->hwm; |
save_hwm = cd->hwm; |
| 3829 |
|
reset_bracount = FALSE; |
| 3830 |
|
|
| 3831 |
if (*(++ptr) == '?') |
if (*(++ptr) == '?') |
| 3832 |
{ |
{ |
| 3849 |
|
|
| 3850 |
|
|
| 3851 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
| 3852 |
|
case '|': /* Reset capture count for each branch */ |
| 3853 |
|
reset_bracount = TRUE; |
| 3854 |
|
/* Fall through */ |
| 3855 |
|
|
| 3856 |
|
/* ------------------------------------------------------------ */ |
| 3857 |
case ':': /* Non-capturing bracket */ |
case ':': /* Non-capturing bracket */ |
| 3858 |
bravalue = OP_BRA; |
bravalue = OP_BRA; |
| 3859 |
ptr++; |
ptr++; |
| 3889 |
|
|
| 3890 |
code[1+LINK_SIZE] = OP_CREF; |
code[1+LINK_SIZE] = OP_CREF; |
| 3891 |
skipbytes = 3; |
skipbytes = 3; |
| 3892 |
|
refsign = -1; |
| 3893 |
|
|
| 3894 |
/* Check for a test for recursion in a named group. */ |
/* Check for a test for recursion in a named group. */ |
| 3895 |
|
|
| 3913 |
terminator = '\''; |
terminator = '\''; |
| 3914 |
ptr++; |
ptr++; |
| 3915 |
} |
} |
| 3916 |
else terminator = 0; |
else |
| 3917 |
|
{ |
| 3918 |
|
terminator = 0; |
| 3919 |
|
if (ptr[1] == '-' || ptr[1] == '+') refsign = *(++ptr); |
| 3920 |
|
} |
| 3921 |
|
|
| 3922 |
/* 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 */ |
| 3923 |
|
|
| 3953 |
if (lengthptr != NULL) break; |
if (lengthptr != NULL) break; |
| 3954 |
|
|
| 3955 |
/* 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 |
| 3956 |
reference. */ |
reference. If the string started with "+" or "-" we require the rest to |
| 3957 |
|
be digits, in which case recno will be set. */ |
| 3958 |
|
|
| 3959 |
|
if (refsign > 0) |
| 3960 |
|
{ |
| 3961 |
|
if (recno <= 0) |
| 3962 |
|
{ |
| 3963 |
|
*errorcodeptr = ERR58; |
| 3964 |
|
goto FAILED; |
| 3965 |
|
} |
| 3966 |
|
if (refsign == '-') |
| 3967 |
|
{ |
| 3968 |
|
recno = cd->bracount - recno + 1; |
| 3969 |
|
if (recno <= 0) |
| 3970 |
|
{ |
| 3971 |
|
*errorcodeptr = ERR15; |
| 3972 |
|
goto FAILED; |
| 3973 |
|
} |
| 3974 |
|
} |
| 3975 |
|
else recno += cd->bracount; |
| 3976 |
|
PUT2(code, 2+LINK_SIZE, recno); |
| 3977 |
|
break; |
| 3978 |
|
} |
| 3979 |
|
|
| 3980 |
|
/* Otherwise (did not start with "+" or "-"), start by looking for the |
| 3981 |
|
name. */ |
| 3982 |
|
|
| 3983 |
slot = cd->name_table; |
slot = cd->name_table; |
| 3984 |
for (i = 0; i < cd->names_found; i++) |
for (i = 0; i < cd->names_found; i++) |
| 4297 |
|
|
| 4298 |
|
|
| 4299 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
| 4300 |
|
case '-': case '+': |
| 4301 |
case '0': case '1': case '2': case '3': case '4': /* Recursion or */ |
case '0': case '1': case '2': case '3': case '4': /* Recursion or */ |
| 4302 |
case '5': case '6': case '7': case '8': case '9': /* subroutine */ |
case '5': case '6': case '7': case '8': case '9': /* subroutine */ |
| 4303 |
{ |
{ |
| 4304 |
const uschar *called; |
const uschar *called; |
| 4305 |
|
|
| 4306 |
|
if ((refsign = *ptr) == '+') ptr++; |
| 4307 |
|
else if (refsign == '-') |
| 4308 |
|
{ |
| 4309 |
|
if ((digitab[ptr[1]] & ctype_digit) == 0) |
| 4310 |
|
goto OTHER_CHAR_AFTER_QUERY; |
| 4311 |
|
ptr++; |
| 4312 |
|
} |
| 4313 |
|
|
| 4314 |
recno = 0; |
recno = 0; |
| 4315 |
while((digitab[*ptr] & ctype_digit) != 0) |
while((digitab[*ptr] & ctype_digit) != 0) |
| 4316 |
recno = recno * 10 + *ptr++ - '0'; |
recno = recno * 10 + *ptr++ - '0'; |
| 4317 |
|
|
| 4318 |
if (*ptr != ')') |
if (*ptr != ')') |
| 4319 |
{ |
{ |
| 4320 |
*errorcodeptr = ERR29; |
*errorcodeptr = ERR29; |
| 4321 |
goto FAILED; |
goto FAILED; |
| 4322 |
} |
} |
| 4323 |
|
|
| 4324 |
|
if (refsign == '-') |
| 4325 |
|
{ |
| 4326 |
|
if (recno == 0) |
| 4327 |
|
{ |
| 4328 |
|
*errorcodeptr = ERR58; |
| 4329 |
|
goto FAILED; |
| 4330 |
|
} |
| 4331 |
|
recno = cd->bracount - recno + 1; |
| 4332 |
|
if (recno <= 0) |
| 4333 |
|
{ |
| 4334 |
|
*errorcodeptr = ERR15; |
| 4335 |
|
goto FAILED; |
| 4336 |
|
} |
| 4337 |
|
} |
| 4338 |
|
else if (refsign == '+') |
| 4339 |
|
{ |
| 4340 |
|
if (recno == 0) |
| 4341 |
|
{ |
| 4342 |
|
*errorcodeptr = ERR58; |
| 4343 |
|
goto FAILED; |
| 4344 |
|
} |
| 4345 |
|
recno += cd->bracount; |
| 4346 |
|
} |
| 4347 |
|
|
| 4348 |
/* Come here from code above that handles a named recursion */ |
/* Come here from code above that handles a named recursion */ |
| 4349 |
|
|
| 4350 |
HANDLE_RECURSION: |
HANDLE_RECURSION: |
| 4417 |
|
|
| 4418 |
/* ------------------------------------------------------------ */ |
/* ------------------------------------------------------------ */ |
| 4419 |
default: /* Other characters: check option setting */ |
default: /* Other characters: check option setting */ |
| 4420 |
|
OTHER_CHAR_AFTER_QUERY: |
| 4421 |
set = unset = 0; |
set = unset = 0; |
| 4422 |
optset = &set; |
optset = &set; |
| 4423 |
|
|
| 4552 |
errorcodeptr, /* Where to put an error message */ |
errorcodeptr, /* Where to put an error message */ |
| 4553 |
(bravalue == OP_ASSERTBACK || |
(bravalue == OP_ASSERTBACK || |
| 4554 |
bravalue == OP_ASSERTBACK_NOT), /* TRUE if back assert */ |
bravalue == OP_ASSERTBACK_NOT), /* TRUE if back assert */ |
| 4555 |
|
reset_bracount, /* True if (?| group */ |
| 4556 |
skipbytes, /* Skip over bracket number */ |
skipbytes, /* Skip over bracket number */ |
| 4557 |
&subfirstbyte, /* For possible first char */ |
&subfirstbyte, /* For possible first char */ |
| 4558 |
&subreqbyte, /* For possible last char */ |
&subreqbyte, /* For possible last char */ |
| 4569 |
is on the bracket. */ |
is on the bracket. */ |
| 4570 |
|
|
| 4571 |
/* 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 |
| 4572 |
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 |
| 4573 |
|
in the real compile phase, not in the pre-pass, where the whole group may |
| 4574 |
|
not be available. */ |
| 4575 |
|
|
| 4576 |
if (bravalue == OP_COND) |
if (bravalue == OP_COND && lengthptr == NULL) |
| 4577 |
{ |
{ |
| 4578 |
uschar *tc = code; |
uschar *tc = code; |
| 4579 |
int condcount = 0; |
int condcount = 0; |
| 4733 |
zerofirstbyte = firstbyte; |
zerofirstbyte = firstbyte; |
| 4734 |
zeroreqbyte = reqbyte; |
zeroreqbyte = reqbyte; |
| 4735 |
|
|
| 4736 |
/* \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). |
| 4737 |
|
We also support \k{name} (.NET syntax) */ |
| 4738 |
|
|
| 4739 |
if (-c == ESC_k && (ptr[1] == '<' || ptr[1] == '\'')) |
if (-c == ESC_k && (ptr[1] == '<' || ptr[1] == '\'' || ptr[1] == '{')) |
| 4740 |
{ |
{ |
| 4741 |
is_recurse = FALSE; |
is_recurse = FALSE; |
| 4742 |
terminator = (*(++ptr) == '<')? '>' : '\''; |
terminator = (*(++ptr) == '<')? '>' : (*ptr == '\'')? '\'' : '}'; |
| 4743 |
goto NAMED_REF_OR_RECURSE; |
goto NAMED_REF_OR_RECURSE; |
| 4744 |
} |
} |
| 4745 |
|
|
| 4905 |
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 |
| 4906 |
value of lengthptr distinguishes the two phases. |
value of lengthptr distinguishes the two phases. |
| 4907 |
|
|
| 4908 |
Argument: |
Arguments: |
| 4909 |
options option bits, including any changes for this subpattern |
options option bits, including any changes for this subpattern |
| 4910 |
oldims previous settings of ims option bits |
oldims previous settings of ims option bits |
| 4911 |
codeptr -> the address of the current code pointer |
codeptr -> the address of the current code pointer |
| 4912 |
ptrptr -> the address of the current pattern pointer |
ptrptr -> the address of the current pattern pointer |
| 4913 |
errorcodeptr -> pointer to error code variable |
errorcodeptr -> pointer to error code variable |
| 4914 |
lookbehind TRUE if this is a lookbehind assertion |
lookbehind TRUE if this is a lookbehind assertion |
| 4915 |
|
reset_bracount TRUE to reset the count for each branch |
| 4916 |
skipbytes skip this many bytes at start (for brackets and OP_COND) |
skipbytes skip this many bytes at start (for brackets and OP_COND) |
| 4917 |
firstbyteptr place to put the first required character, or a negative number |
firstbyteptr place to put the first required character, or a negative number |
| 4918 |
reqbyteptr place to put the last required character, or a negative number |
reqbyteptr place to put the last required character, or a negative number |
| 4926 |
|
|
| 4927 |
static BOOL |
static BOOL |
| 4928 |
compile_regex(int options, int oldims, uschar **codeptr, const uschar **ptrptr, |
compile_regex(int options, int oldims, uschar **codeptr, const uschar **ptrptr, |
| 4929 |
int *errorcodeptr, BOOL lookbehind, int skipbytes, int *firstbyteptr, |
int *errorcodeptr, BOOL lookbehind, BOOL reset_bracount, int skipbytes, |
| 4930 |
int *reqbyteptr, branch_chain *bcptr, compile_data *cd, int *lengthptr) |
int *firstbyteptr, int *reqbyteptr, branch_chain *bcptr, compile_data *cd, |
| 4931 |
|
int *lengthptr) |
| 4932 |
{ |
{ |
| 4933 |
const uschar *ptr = *ptrptr; |
const uschar *ptr = *ptrptr; |
| 4934 |
uschar *code = *codeptr; |
uschar *code = *codeptr; |
| 4938 |
int firstbyte, reqbyte; |
int firstbyte, reqbyte; |
| 4939 |
int branchfirstbyte, branchreqbyte; |
int branchfirstbyte, branchreqbyte; |
| 4940 |
int length; |
int length; |
| 4941 |
|
int orig_bracount; |
| 4942 |
|
int max_bracount; |
| 4943 |
branch_chain bc; |
branch_chain bc; |
| 4944 |
|
|
| 4945 |
bc.outer = bcptr; |
bc.outer = bcptr; |
| 4968 |
|
|
| 4969 |
/* Loop for each alternative branch */ |
/* Loop for each alternative branch */ |
| 4970 |
|
|
| 4971 |
|
orig_bracount = max_bracount = cd->bracount; |
| 4972 |
for (;;) |
for (;;) |
| 4973 |
{ |
{ |
| 4974 |
|
/* For a (?| group, reset the capturing bracket count so that each branch |
| 4975 |
|
uses the same numbers. */ |
| 4976 |
|
|
| 4977 |
|
if (reset_bracount) cd->bracount = orig_bracount; |
| 4978 |
|
|
| 4979 |
/* Handle a change of ims options at the start of the branch */ |
/* Handle a change of ims options at the start of the branch */ |
| 4980 |
|
|
| 4981 |
if ((options & PCRE_IMS) != oldims) |
if ((options & PCRE_IMS) != oldims) |
| 5005 |
return FALSE; |
return FALSE; |
| 5006 |
} |
} |
| 5007 |
|
|
| 5008 |
|
/* Keep the highest bracket count in case (?| was used and some branch |
| 5009 |
|
has fewer than the rest. */ |
| 5010 |
|
|
| 5011 |
|
if (cd->bracount > max_bracount) max_bracount = cd->bracount; |
| 5012 |
|
|
| 5013 |
/* 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. */ |
| 5014 |
|
|
| 5015 |
if (lengthptr == NULL) |
if (lengthptr == NULL) |
| 5073 |
} |
} |
| 5074 |
} |
} |
| 5075 |
|
|
| 5076 |
/* Reached end of expression, either ')' or end of pattern. Go back through |
/* Reached end of expression, either ')' or end of pattern. In the real |
| 5077 |
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 |
| 5078 |
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 |
| 5079 |
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 |
| 5080 |
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 |
| 5081 |
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, |
| 5082 |
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. |
| 5083 |
at the terminating char. */ |
Return leaving the pointer at the terminating char. */ |
| 5084 |
|
|
| 5085 |
if (*ptr != '|') |
if (*ptr != '|') |
| 5086 |
{ |
{ |
| 5087 |
int branch_length = code - last_branch; |
if (lengthptr == NULL) |
|
do |
|
| 5088 |
{ |
{ |
| 5089 |
int prev_length = GET(last_branch, 1); |
int branch_length = code - last_branch; |
| 5090 |
PUT(last_branch, 1, branch_length); |
do |
| 5091 |
branch_length = prev_length; |
{ |
| 5092 |
last_branch -= branch_length; |
int prev_length = GET(last_branch, 1); |
| 5093 |
|
PUT(last_branch, 1, branch_length); |
| 5094 |
|
branch_length = prev_length; |
| 5095 |
|
last_branch -= branch_length; |
| 5096 |
|
} |
| 5097 |
|
while (branch_length > 0); |
| 5098 |
} |
} |
|
while (branch_length > 0); |
|
| 5099 |
|
|
| 5100 |
/* Fill in the ket */ |
/* Fill in the ket */ |
| 5101 |
|
|
| 5112 |
length += 2; |
length += 2; |
| 5113 |
} |
} |
| 5114 |
|
|
| 5115 |
|
/* Retain the highest bracket number, in case resetting was used. */ |
| 5116 |
|
|
| 5117 |
|
cd->bracount = max_bracount; |
| 5118 |
|
|
| 5119 |
/* Set values to pass back */ |
/* Set values to pass back */ |
| 5120 |
|
|
| 5121 |
*codeptr = code; |
*codeptr = code; |
| 5126 |
return TRUE; |
return TRUE; |
| 5127 |
} |
} |
| 5128 |
|
|
| 5129 |
/* 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 |
| 5130 |
|
pointer back to where it was for the start of the first branch. (That is, |
| 5131 |
|
pretend that each branch is the only one.) |
| 5132 |
|
|
| 5133 |
|
In the real compile phase, insert an ALT node. Its length field points back |
| 5134 |
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 |
| 5135 |
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 |
| 5136 |
zero offset until it is closed, making it possible to detect recursion. */ |
zero offset until it is closed, making it possible to detect recursion. */ |
| 5137 |
|
|
| 5138 |
*code = OP_ALT; |
if (lengthptr != NULL) |
| 5139 |
PUT(code, 1, code - last_branch); |
{ |
| 5140 |
bc.current = last_branch = code; |
code = *codeptr + 1 + LINK_SIZE + skipbytes; |
| 5141 |
code += 1 + LINK_SIZE; |
length += 1 + LINK_SIZE; |
| 5142 |
|
} |
| 5143 |
|
else |
| 5144 |
|
{ |
| 5145 |
|
*code = OP_ALT; |
| 5146 |
|
PUT(code, 1, code - last_branch); |
| 5147 |
|
bc.current = last_branch = code; |
| 5148 |
|
code += 1 + LINK_SIZE; |
| 5149 |
|
} |
| 5150 |
|
|
| 5151 |
ptr++; |
ptr++; |
|
length += 1 + LINK_SIZE; |
|
| 5152 |
} |
} |
| 5153 |
/* Control never reaches here */ |
/* Control never reaches here */ |
| 5154 |
} |
} |
| 5415 |
with errorptr and erroroffset set |
with errorptr and erroroffset set |
| 5416 |
*/ |
*/ |
| 5417 |
|
|
| 5418 |
PCRE_DATA_SCOPE pcre * |
PCRE_EXP_DEFN pcre * |
| 5419 |
pcre_compile(const char *pattern, int options, const char **errorptr, |
pcre_compile(const char *pattern, int options, const char **errorptr, |
| 5420 |
int *erroroffset, const unsigned char *tables) |
int *erroroffset, const unsigned char *tables) |
| 5421 |
{ |
{ |
| 5423 |
} |
} |
| 5424 |
|
|
| 5425 |
|
|
| 5426 |
PCRE_DATA_SCOPE pcre * |
PCRE_EXP_DEFN pcre * |
| 5427 |
pcre_compile2(const char *pattern, int options, int *errorcodeptr, |
pcre_compile2(const char *pattern, int options, int *errorcodeptr, |
| 5428 |
const char **errorptr, int *erroroffset, const unsigned char *tables) |
const char **errorptr, int *erroroffset, const unsigned char *tables) |
| 5429 |
{ |
{ |
| 5472 |
if (erroroffset == NULL) |
if (erroroffset == NULL) |
| 5473 |
{ |
{ |
| 5474 |
errorcode = ERR16; |
errorcode = ERR16; |
| 5475 |
goto PCRE_EARLY_ERROR_RETURN; |
goto PCRE_EARLY_ERROR_RETURN2; |
| 5476 |
} |
} |
| 5477 |
|
|
| 5478 |
*erroroffset = 0; |
*erroroffset = 0; |
| 5485 |
(*erroroffset = _pcre_valid_utf8((uschar *)pattern, -1)) >= 0) |
(*erroroffset = _pcre_valid_utf8((uschar *)pattern, -1)) >= 0) |
| 5486 |
{ |
{ |
| 5487 |
errorcode = ERR44; |
errorcode = ERR44; |
| 5488 |
goto PCRE_UTF8_ERROR_RETURN; |
goto PCRE_EARLY_ERROR_RETURN2; |
| 5489 |
} |
} |
| 5490 |
#else |
#else |
| 5491 |
if ((options & PCRE_UTF8) != 0) |
if ((options & PCRE_UTF8) != 0) |
| 5510 |
cd->ctypes = tables + ctypes_offset; |
cd->ctypes = tables + ctypes_offset; |
| 5511 |
|
|
| 5512 |
/* Handle different types of newline. The three bits give seven cases. The |
/* Handle different types of newline. The three bits give seven cases. The |
| 5513 |
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 |
| 5514 |
|
"anycrlf". */ |
| 5515 |
|
|
| 5516 |
switch (options & (PCRE_NEWLINE_CRLF | PCRE_NEWLINE_ANY)) |
switch (options & (PCRE_NEWLINE_CRLF | PCRE_NEWLINE_ANY)) |
| 5517 |
{ |
{ |
| 5521 |
case PCRE_NEWLINE_CR+ |
case PCRE_NEWLINE_CR+ |
| 5522 |
PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break; |
PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break; |
| 5523 |
case PCRE_NEWLINE_ANY: newline = -1; break; |
case PCRE_NEWLINE_ANY: newline = -1; break; |
| 5524 |
|
case PCRE_NEWLINE_ANYCRLF: newline = -2; break; |
| 5525 |
default: errorcode = ERR56; goto PCRE_EARLY_ERROR_RETURN; |
default: errorcode = ERR56; goto PCRE_EARLY_ERROR_RETURN; |
| 5526 |
} |
} |
| 5527 |
|
|
| 5528 |
if (newline < 0) |
if (newline == -2) |
| 5529 |
|
{ |
| 5530 |
|
cd->nltype = NLTYPE_ANYCRLF; |
| 5531 |
|
} |
| 5532 |
|
else if (newline < 0) |
| 5533 |
{ |
{ |
| 5534 |
cd->nltype = NLTYPE_ANY; |
cd->nltype = NLTYPE_ANY; |
| 5535 |
} |
} |
| 5590 |
code = cworkspace; |
code = cworkspace; |
| 5591 |
*code = OP_BRA; |
*code = OP_BRA; |
| 5592 |
(void)compile_regex(cd->external_options, cd->external_options & PCRE_IMS, |
(void)compile_regex(cd->external_options, cd->external_options & PCRE_IMS, |
| 5593 |
&code, &ptr, &errorcode, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, &length); |
&code, &ptr, &errorcode, FALSE, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, |
| 5594 |
|
&length); |
| 5595 |
if (errorcode != 0) goto PCRE_EARLY_ERROR_RETURN; |
if (errorcode != 0) goto PCRE_EARLY_ERROR_RETURN; |
| 5596 |
|
|
| 5597 |
DPRINTF(("end pre-compile: length=%d workspace=%d\n", length, |
DPRINTF(("end pre-compile: length=%d workspace=%d\n", length, |
| 5659 |
code = (uschar *)codestart; |
code = (uschar *)codestart; |
| 5660 |
*code = OP_BRA; |
*code = OP_BRA; |
| 5661 |
(void)compile_regex(re->options, re->options & PCRE_IMS, &code, &ptr, |
(void)compile_regex(re->options, re->options & PCRE_IMS, &code, &ptr, |
| 5662 |
&errorcode, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, NULL); |
&errorcode, FALSE, FALSE, 0, &firstbyte, &reqbyte, NULL, cd, NULL); |
| 5663 |
re->top_bracket = cd->bracount; |
re->top_bracket = cd->bracount; |
| 5664 |
re->top_backref = cd->top_backref; |
re->top_backref = cd->top_backref; |
| 5665 |
|
|
| 5704 |
(pcre_free)(re); |
(pcre_free)(re); |
| 5705 |
PCRE_EARLY_ERROR_RETURN: |
PCRE_EARLY_ERROR_RETURN: |
| 5706 |
*erroroffset = ptr - (const uschar *)pattern; |
*erroroffset = ptr - (const uschar *)pattern; |
| 5707 |
#ifdef SUPPORT_UTF8 |
PCRE_EARLY_ERROR_RETURN2: |
|
PCRE_UTF8_ERROR_RETURN: |
|
|
#endif |
|
| 5708 |
*errorptr = error_texts[errorcode]; |
*errorptr = error_texts[errorcode]; |
| 5709 |
if (errorcodeptr != NULL) *errorcodeptr = errorcode; |
if (errorcodeptr != NULL) *errorcodeptr = errorcode; |
| 5710 |
return NULL; |
return NULL; |
| 5794 |
else printf("Req char = \\x%02x%s\n", ch, caseless); |
else printf("Req char = \\x%02x%s\n", ch, caseless); |
| 5795 |
} |
} |
| 5796 |
|
|
| 5797 |
pcre_printint(re, stdout); |
pcre_printint(re, stdout, TRUE); |
| 5798 |
|
|
| 5799 |
/* 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 |
| 5800 |
was compiled can be seen. */ |
was compiled can be seen. */ |