code/trunk/pcreposix.c

/*************************************************
*      Perl-Compatible Regular Expressions       *
*************************************************/

/* PCRE is a library of functions to support regular expressions whose syntax
and semantics are as close as possible to those of the Perl 5 language.

                       Written by Philip Hazel
           Copyright (c) 1997-2012 University of Cambridge

-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:

    * Redistributions of source code must retain the above copyright notice,
      this list of conditions and the following disclaimer.

    * Redistributions in binary form must reproduce the above copyright
      notice, this list of conditions and the following disclaimer in the
      documentation and/or other materials provided with the distribution.

    * Neither the name of the University of Cambridge nor the names of its
      contributors may be used to endorse or promote products derived from
      this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
POSSIBILITY OF SUCH DAMAGE.
-----------------------------------------------------------------------------
*/


/* This module is a wrapper that provides a POSIX API to the underlying PCRE
functions. */


#ifdef HAVE_CONFIG_H
#include "config.h"
#endif


/* Ensure that the PCREPOSIX_EXP_xxx macros are set appropriately for
compiling these functions. This must come before including pcreposix.h, where
they are set for an application (using these functions) if they have not
previously been set. */

#if defined(_WIN32) && !defined(PCRE_STATIC)
#  define PCREPOSIX_EXP_DECL extern __declspec(dllexport)
#  define PCREPOSIX_EXP_DEFN __declspec(dllexport)
#endif

/* We include pcre.h before pcre_internal.h so that the PCRE library functions
are declared as "import" for Windows by defining PCRE_EXP_DECL as "import".
This is needed even though pcre_internal.h itself includes pcre.h, because it
does so after it has set PCRE_EXP_DECL to "export" if it is not already set. */

#include "pcre.h"
#include "pcre_internal.h"
#include "pcreposix.h"


/* Table to translate PCRE compile time error codes into POSIX error codes. */

static const int eint[] = {
  0,           /* no error */
  REG_EESCAPE, /* \ at end of pattern */
  REG_EESCAPE, /* \c at end of pattern */
  REG_EESCAPE, /* unrecognized character follows \ */
  REG_BADBR,   /* numbers out of order in {} quantifier */
  /* 5 */
  REG_BADBR,   /* number too big in {} quantifier */
  REG_EBRACK,  /* missing terminating ] for character class */
  REG_ECTYPE,  /* invalid escape sequence in character class */
  REG_ERANGE,  /* range out of order in character class */
  REG_BADRPT,  /* nothing to repeat */
  /* 10 */
  REG_BADRPT,  /* operand of unlimited repeat could match the empty string */
  REG_ASSERT,  /* internal error: unexpected repeat */
  REG_BADPAT,  /* unrecognized character after (? */
  REG_BADPAT,  /* POSIX named classes are supported only within a class */
  REG_EPAREN,  /* missing ) */
  /* 15 */
  REG_ESUBREG, /* reference to non-existent subpattern */
  REG_INVARG,  /* erroffset passed as NULL */
  REG_INVARG,  /* unknown option bit(s) set */
  REG_EPAREN,  /* missing ) after comment */
  REG_ESIZE,   /* parentheses nested too deeply */
  /* 20 */
  REG_ESIZE,   /* regular expression too large */
  REG_ESPACE,  /* failed to get memory */
  REG_EPAREN,  /* unmatched parentheses */
  REG_ASSERT,  /* internal error: code overflow */
  REG_BADPAT,  /* unrecognized character after (?< */
  /* 25 */
  REG_BADPAT,  /* lookbehind assertion is not fixed length */
  REG_BADPAT,  /* malformed number or name after (?( */
  REG_BADPAT,  /* conditional group contains more than two branches */
  REG_BADPAT,  /* assertion expected after (?( */
  REG_BADPAT,  /* (?R or (?[+-]digits must be followed by ) */
  /* 30 */
  REG_ECTYPE,  /* unknown POSIX class name */
  REG_BADPAT,  /* POSIX collating elements are not supported */
  REG_INVARG,  /* this version of PCRE is not compiled with PCRE_UTF8 support */
  REG_BADPAT,  /* spare error */
  REG_BADPAT,  /* character value in \x{...} sequence is too large */
  /* 35 */
  REG_BADPAT,  /* invalid condition (?(0) */
  REG_BADPAT,  /* \C not allowed in lookbehind assertion */
  REG_EESCAPE, /* PCRE does not support \L, \l, \N, \U, or \u */
  REG_BADPAT,  /* number after (?C is > 255 */
  REG_BADPAT,  /* closing ) for (?C expected */
  /* 40 */
  REG_BADPAT,  /* recursive call could loop indefinitely */
  REG_BADPAT,  /* unrecognized character after (?P */
  REG_BADPAT,  /* syntax error in subpattern name (missing terminator) */
  REG_BADPAT,  /* two named subpatterns have the same name */
  REG_BADPAT,  /* invalid UTF-8 string */
  /* 45 */
  REG_BADPAT,  /* support for \P, \p, and \X has not been compiled */
  REG_BADPAT,  /* malformed \P or \p sequence */
  REG_BADPAT,  /* unknown property name after \P or \p */
  REG_BADPAT,  /* subpattern name is too long (maximum 32 characters) */
  REG_BADPAT,  /* too many named subpatterns (maximum 10,000) */
  /* 50 */
  REG_BADPAT,  /* repeated subpattern is too long */
  REG_BADPAT,  /* octal value is greater than \377 (not in UTF-8 mode) */
  REG_BADPAT,  /* internal error: overran compiling workspace */
  REG_BADPAT,  /* internal error: previously-checked referenced subpattern not found */
  REG_BADPAT,  /* DEFINE group contains more than one branch */
  /* 55 */
  REG_BADPAT,  /* repeating a DEFINE group is not allowed */
  REG_INVARG,  /* inconsistent NEWLINE options */
  REG_BADPAT,  /* \g is not followed followed by an (optionally braced) non-zero number */
  REG_BADPAT,  /* a numbered reference must not be zero */
  REG_BADPAT,  /* an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT) */
  /* 60 */
  REG_BADPAT,  /* (*VERB) not recognized */
  REG_BADPAT,  /* number is too big */
  REG_BADPAT,  /* subpattern name expected */
  REG_BADPAT,  /* digit expected after (?+ */
  REG_BADPAT,  /* ] is an invalid data character in JavaScript compatibility mode */
  /* 65 */
  REG_BADPAT,  /* different names for subpatterns of the same number are not allowed */
  REG_BADPAT,  /* (*MARK) must have an argument */
  REG_INVARG,  /* this version of PCRE is not compiled with PCRE_UCP support */
  REG_BADPAT,  /* \c must be followed by an ASCII character */
  REG_BADPAT,  /* \k is not followed by a braced, angle-bracketed, or quoted name */
  /* 70 */
  REG_BADPAT,  /* internal error: unknown opcode in find_fixedlength() */
  REG_BADPAT,  /* \N is not supported in a class */
  REG_BADPAT,  /* too many forward references */
  REG_BADPAT,  /* disallowed UTF-8/16/32 code point (>= 0xd800 && <= 0xdfff) */
  REG_BADPAT,  /* invalid UTF-16 string (should not occur) */
  /* 75 */
  REG_BADPAT,  /* overlong MARK name */
  REG_BADPAT,  /* character value in \u.... sequence is too large */
  REG_BADPAT,  /* invalid UTF-32 string (should not occur) */
  REG_BADPAT   /* setting UTF is disabled by the application */
};

/* Table of texts corresponding to POSIX error codes */

static const char *const pstring[] = {
  "",                                /* Dummy for value 0 */
  "internal error",                  /* REG_ASSERT */
  "invalid repeat counts in {}",     /* BADBR      */
  "pattern error",                   /* BADPAT     */
  "? * + invalid",                   /* BADRPT     */
  "unbalanced {}",                   /* EBRACE     */
  "unbalanced []",                   /* EBRACK     */
  "collation error - not relevant",  /* ECOLLATE   */
  "bad class",                       /* ECTYPE     */
  "bad escape sequence",             /* EESCAPE    */
  "empty expression",                /* EMPTY      */
  "unbalanced ()",                   /* EPAREN     */
  "bad range inside []",             /* ERANGE     */
  "expression too big",              /* ESIZE      */
  "failed to get memory",            /* ESPACE     */
  "bad back reference",              /* ESUBREG    */
  "bad argument",                    /* INVARG     */
  "match failed"                     /* NOMATCH    */
};


/*************************************************
*          Translate error code to string        *
*************************************************/

PCREPOSIX_EXP_DEFN size_t PCRE_CALL_CONVENTION
regerror(int errcode, const regex_t *preg, char *errbuf, size_t errbuf_size)
{
const char *message, *addmessage;
size_t length, addlength;

message = (errcode >= (int)(sizeof(pstring)/sizeof(char *)))?
  "unknown error code" : pstring[errcode];
length = strlen(message) + 1;

addmessage = " at offset ";
addlength = (preg != NULL && (int)preg->re_erroffset != -1)?
  strlen(addmessage) + 6 : 0;

if (errbuf_size > 0)
  {
  if (addlength > 0 && errbuf_size >= length + addlength)
    sprintf(errbuf, "%s%s%-6d", message, addmessage, (int)preg->re_erroffset);
  else
    {
    strncpy(errbuf, message, errbuf_size - 1);
    errbuf[errbuf_size-1] = 0;
    }
  }

return length + addlength;
}


/*************************************************
*           Free store held by a regex           *
*************************************************/

PCREPOSIX_EXP_DEFN void PCRE_CALL_CONVENTION
regfree(regex_t *preg)
{
(PUBL(free))(preg->re_pcre);
}


/*************************************************
*            Compile a regular expression        *
*************************************************/

/*
Arguments:
  preg        points to a structure for recording the compiled expression
  pattern     the pattern to compile
  cflags      compilation flags

Returns:      0 on success
              various non-zero codes on failure
*/

PCREPOSIX_EXP_DEFN int PCRE_CALL_CONVENTION
regcomp(regex_t *preg, const char *pattern, int cflags)
{
const char *errorptr;
int erroffset;
int errorcode;
int options = 0;
int re_nsub = 0;

if ((cflags & REG_ICASE) != 0)    options |= PCRE_CASELESS;
if ((cflags & REG_NEWLINE) != 0)  options |= PCRE_MULTILINE;
if ((cflags & REG_DOTALL) != 0)   options |= PCRE_DOTALL;
if ((cflags & REG_NOSUB) != 0)    options |= PCRE_NO_AUTO_CAPTURE;
if ((cflags & REG_UTF8) != 0)     options |= PCRE_UTF8;
if ((cflags & REG_UCP) != 0)      options |= PCRE_UCP;
if ((cflags & REG_UNGREEDY) != 0) options |= PCRE_UNGREEDY;

preg->re_pcre = pcre_compile2(pattern, options, &errorcode, &errorptr,
  &erroffset, NULL);
preg->re_erroffset = erroffset;

/* Safety: if the error code is too big for the translation vector (which
should not happen, but we all make mistakes), return REG_BADPAT. */

if (preg->re_pcre == NULL)
  {
  return (errorcode < (int)(sizeof(eint)/sizeof(const int)))?
    eint[errorcode] : REG_BADPAT;
  }

(void)pcre_fullinfo((const pcre *)preg->re_pcre, NULL, PCRE_INFO_CAPTURECOUNT,
  &re_nsub);
preg->re_nsub = (size_t)re_nsub;
return 0;
}


/*************************************************
*              Match a regular expression        *
*************************************************/

/* Unfortunately, PCRE requires 3 ints of working space for each captured
substring, so we have to get and release working store instead of just using
the POSIX structures as was done in earlier releases when PCRE needed only 2
ints. However, if the number of possible capturing brackets is small, use a
block of store on the stack, to reduce the use of malloc/free. The threshold is
in a macro that can be changed at configure time.

If REG_NOSUB was specified at compile time, the PCRE_NO_AUTO_CAPTURE flag will
be set. When this is the case, the nmatch and pmatch arguments are ignored, and
the only result is yes/no/error. */

PCREPOSIX_EXP_DEFN int PCRE_CALL_CONVENTION
regexec(const regex_t *preg, const char *string, size_t nmatch,
  regmatch_t pmatch[], int eflags)
{
int rc, so, eo;
int options = 0;
int *ovector = NULL;
int small_ovector[POSIX_MALLOC_THRESHOLD * 3];
BOOL allocated_ovector = FALSE;
BOOL nosub =
  (REAL_PCRE_OPTIONS((const pcre *)preg->re_pcre) & PCRE_NO_AUTO_CAPTURE) != 0;

if ((eflags & REG_NOTBOL) != 0) options |= PCRE_NOTBOL;
if ((eflags & REG_NOTEOL) != 0) options |= PCRE_NOTEOL;
if ((eflags & REG_NOTEMPTY) != 0) options |= PCRE_NOTEMPTY;

((regex_t *)preg)->re_erroffset = (size_t)(-1);  /* Only has meaning after compile */

/* When no string data is being returned, or no vector has been passed in which
to put it, ensure that nmatch is zero. Otherwise, ensure the vector for holding
the return data is large enough. */

if (nosub || pmatch == NULL) nmatch = 0;

else if (nmatch > 0)
  {
  if (nmatch <= POSIX_MALLOC_THRESHOLD)
    {
    ovector = &(small_ovector[0]);
    }
  else
    {
    if (nmatch > INT_MAX/(sizeof(int) * 3)) return REG_ESPACE;
    ovector = (int *)malloc(sizeof(int) * nmatch * 3);
    if (ovector == NULL) return REG_ESPACE;
    allocated_ovector = TRUE;
    }
  }

/* REG_STARTEND is a BSD extension, to allow for non-NUL-terminated strings.
The man page from OS X says "REG_STARTEND affects only the location of the
string, not how it is matched". That is why the "so" value is used to bump the
start location rather than being passed as a PCRE "starting offset". */

if ((eflags & REG_STARTEND) != 0)
  {
  so = pmatch[0].rm_so;
  eo = pmatch[0].rm_eo;
  }
else
  {
  so = 0;
  eo = (int)strlen(string);
  }

rc = pcre_exec((const pcre *)preg->re_pcre, NULL, string + so, (eo - so),
  0, options, ovector, (int)(nmatch * 3));

if (rc == 0) rc = (int)nmatch;    /* All captured slots were filled in */

/* Successful match */

if (rc >= 0)
  {
  size_t i;
  if (!nosub)
    {
    for (i = 0; i < (size_t)rc; i++)
      {
      pmatch[i].rm_so = ovector[i*2];
      pmatch[i].rm_eo = ovector[i*2+1];
      }
    if (allocated_ovector) free(ovector);
    for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;
    }
  return 0;
  }

/* Unsuccessful match */

if (allocated_ovector) free(ovector);
switch(rc)
  {
/* ========================================================================== */
  /* These cases are never obeyed. This is a fudge that causes a compile-time
  error if the vector eint, which is indexed by compile-time error number, is
  not the correct length. It seems to be the only way to do such a check at
  compile time, as the sizeof() operator does not work in the C preprocessor.
  As all the PCRE_ERROR_xxx values are negative, we can use 0 and 1. */

  case 0:
  case (sizeof(eint)/sizeof(int) == ERRCOUNT):
  return REG_ASSERT;
/* ========================================================================== */

  case PCRE_ERROR_NOMATCH: return REG_NOMATCH;
  case PCRE_ERROR_NULL: return REG_INVARG;
  case PCRE_ERROR_BADOPTION: return REG_INVARG;
  case PCRE_ERROR_BADMAGIC: return REG_INVARG;
  case PCRE_ERROR_UNKNOWN_NODE: return REG_ASSERT;
  case PCRE_ERROR_NOMEMORY: return REG_ESPACE;
  case PCRE_ERROR_MATCHLIMIT: return REG_ESPACE;
  case PCRE_ERROR_BADUTF8: return REG_INVARG;
  case PCRE_ERROR_BADUTF8_OFFSET: return REG_INVARG;
  case PCRE_ERROR_BADMODE: return REG_INVARG;
  default: return REG_ASSERT;
  }
}

/* End of pcreposix.c */
1	nigel	3	/*************************************************
2			* Perl-Compatible Regular Expressions *
3			*************************************************/
4
5	nigel	77	/* PCRE is a library of functions to support regular expressions whose syntax
6			and semantics are as close as possible to those of the Perl 5 language.
7	nigel	3
8	nigel	77	Written by Philip Hazel
9	ph10	836	Copyright (c) 1997-2012 University of Cambridge
10	nigel	3
11			-----------------------------------------------------------------------------
12	nigel	75	Redistribution and use in source and binary forms, with or without
13			modification, are permitted provided that the following conditions are met:
14	nigel	3
15	nigel	75	* Redistributions of source code must retain the above copyright notice,
16			this list of conditions and the following disclaimer.
17	nigel	3
18	nigel	75	* Redistributions in binary form must reproduce the above copyright
19			notice, this list of conditions and the following disclaimer in the
20			documentation and/or other materials provided with the distribution.
21	nigel	3
22	nigel	75	* Neither the name of the University of Cambridge nor the names of its
23			contributors may be used to endorse or promote products derived from
24			this software without specific prior written permission.
25	nigel	29
26	nigel	75	THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
27			AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28			IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29			ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
30			LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
31			CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
32			SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
33			INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
34			CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
35			ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
36			POSSIBILITY OF SUCH DAMAGE.
37	nigel	3	-----------------------------------------------------------------------------
38			*/
39
40	nigel	77
41			/* This module is a wrapper that provides a POSIX API to the underlying PCRE
42			functions. */
43
44
45	ph10	200	#ifdef HAVE_CONFIG_H
46	ph10	236	#include "config.h"
47	ph10	200	#endif
48	ph10	199
49
50	ph10	200	/* Ensure that the PCREPOSIX_EXP_xxx macros are set appropriately for
51	ph10	199	compiling these functions. This must come before including pcreposix.h, where
52			they are set for an application (using these functions) if they have not
53			previously been set. */
54
55			#if defined(_WIN32) && !defined(PCRE_STATIC)
56			# define PCREPOSIX_EXP_DECL extern __declspec(dllexport)
57			# define PCREPOSIX_EXP_DEFN __declspec(dllexport)
58			#endif
59
60	ph10	545	/* We include pcre.h before pcre_internal.h so that the PCRE library functions
61	ph10	536	are declared as "import" for Windows by defining PCRE_EXP_DECL as "import".
62	ph10	545	This is needed even though pcre_internal.h itself includes pcre.h, because it
63	ph10	536	does so after it has set PCRE_EXP_DECL to "export" if it is not already set. */
64
65			#include "pcre.h"
66	nigel	77	#include "pcre_internal.h"
67	nigel	3	#include "pcreposix.h"
68
69
70	nigel	77	/* Table to translate PCRE compile time error codes into POSIX error codes. */
71	nigel	3
72	nigel	73	static const int eint[] = {
73	nigel	77	0, /* no error */
74			REG_EESCAPE, /* \ at end of pattern */
75			REG_EESCAPE, /* \c at end of pattern */
76			REG_EESCAPE, /* unrecognized character follows \ */
77			REG_BADBR, /* numbers out of order in {} quantifier */
78	ph10	461	/* 5 */
79	nigel	77	REG_BADBR, /* number too big in {} quantifier */
80			REG_EBRACK, /* missing terminating ] for character class */
81			REG_ECTYPE, /* invalid escape sequence in character class */
82			REG_ERANGE, /* range out of order in character class */
83			REG_BADRPT, /* nothing to repeat */
84	ph10	461	/* 10 */
85	nigel	77	REG_BADRPT, /* operand of unlimited repeat could match the empty string */
86			REG_ASSERT, /* internal error: unexpected repeat */
87			REG_BADPAT, /* unrecognized character after (? */
88			REG_BADPAT, /* POSIX named classes are supported only within a class */
89			REG_EPAREN, /* missing ) */
90	ph10	461	/* 15 */
91	nigel	77	REG_ESUBREG, /* reference to non-existent subpattern */
92			REG_INVARG, /* erroffset passed as NULL */
93			REG_INVARG, /* unknown option bit(s) set */
94			REG_EPAREN, /* missing ) after comment */
95			REG_ESIZE, /* parentheses nested too deeply */
96	ph10	461	/* 20 */
97	nigel	77	REG_ESIZE, /* regular expression too large */
98			REG_ESPACE, /* failed to get memory */
99	ph10	418	REG_EPAREN, /* unmatched parentheses */
100	nigel	77	REG_ASSERT, /* internal error: code overflow */
101			REG_BADPAT, /* unrecognized character after (?< */
102	ph10	461	/* 25 */
103	nigel	77	REG_BADPAT, /* lookbehind assertion is not fixed length */
104	nigel	91	REG_BADPAT, /* malformed number or name after (?( */
105	nigel	93	REG_BADPAT, /* conditional group contains more than two branches */
106	nigel	77	REG_BADPAT, /* assertion expected after (?( */
107	ph10	166	REG_BADPAT, /* (?R or (?[+-]digits must be followed by ) */
108	ph10	461	/* 30 */
109	nigel	77	REG_ECTYPE, /* unknown POSIX class name */
110			REG_BADPAT, /* POSIX collating elements are not supported */
111			REG_INVARG, /* this version of PCRE is not compiled with PCRE_UTF8 support */
112			REG_BADPAT, /* spare error */
113			REG_BADPAT, /* character value in \x{...} sequence is too large */
114	ph10	461	/* 35 */
115	nigel	77	REG_BADPAT, /* invalid condition (?(0) */
116			REG_BADPAT, /* \C not allowed in lookbehind assertion */
117			REG_EESCAPE, /* PCRE does not support \L, \l, \N, \U, or \u */
118			REG_BADPAT, /* number after (?C is > 255 */
119			REG_BADPAT, /* closing ) for (?C expected */
120	ph10	461	/* 40 */
121	nigel	77	REG_BADPAT, /* recursive call could loop indefinitely */
122			REG_BADPAT, /* unrecognized character after (?P */
123	nigel	93	REG_BADPAT, /* syntax error in subpattern name (missing terminator) */
124	nigel	91	REG_BADPAT, /* two named subpatterns have the same name */
125	nigel	77	REG_BADPAT, /* invalid UTF-8 string */
126	ph10	461	/* 45 */
127	nigel	77	REG_BADPAT, /* support for \P, \p, and \X has not been compiled */
128			REG_BADPAT, /* malformed \P or \p sequence */
129	nigel	91	REG_BADPAT, /* unknown property name after \P or \p */
130			REG_BADPAT, /* subpattern name is too long (maximum 32 characters) */
131			REG_BADPAT, /* too many named subpatterns (maximum 10,000) */
132	ph10	461	/* 50 */
133	nigel	91	REG_BADPAT, /* repeated subpattern is too long */
134	nigel	93	REG_BADPAT, /* octal value is greater than \377 (not in UTF-8 mode) */
135			REG_BADPAT, /* internal error: overran compiling workspace */
136			REG_BADPAT, /* internal error: previously-checked referenced subpattern not found */
137			REG_BADPAT, /* DEFINE group contains more than one branch */
138	ph10	461	/* 55 */
139	nigel	93	REG_BADPAT, /* repeating a DEFINE group is not allowed */
140			REG_INVARG, /* inconsistent NEWLINE options */
141	ph10	166	REG_BADPAT, /* \g is not followed followed by an (optionally braced) non-zero number */
142	ph10	418	REG_BADPAT, /* a numbered reference must not be zero */
143	ph10	510	REG_BADPAT, /* an argument is not allowed for (ACCEPT), (FAIL), or (COMMIT) /
144	ph10	418	/* 60 */
145	ph10	461	REG_BADPAT, /* (VERB) not recognized /
146	ph10	268	REG_BADPAT, /* number is too big */
147	ph10	286	REG_BADPAT, /* subpattern name expected */
148	ph10	336	REG_BADPAT, /* digit expected after (?+ */
149	ph10	457	REG_BADPAT, /* ] is an invalid data character in JavaScript compatibility mode */
150			/* 65 */
151	ph10	510	REG_BADPAT, /* different names for subpatterns of the same number are not allowed */
152			REG_BADPAT, /* (MARK) must have an argument /
153	ph10	518	REG_INVARG, /* this version of PCRE is not compiled with PCRE_UCP support */
154	ph10	579	REG_BADPAT, /* \c must be followed by an ASCII character */
155	ph10	654	REG_BADPAT, /* \k is not followed by a braced, angle-bracketed, or quoted name */
156	ph10	747	/* 70 */
157	ph10	836	REG_BADPAT, /* internal error: unknown opcode in find_fixedlength() */
158			REG_BADPAT, /* \N is not supported in a class */
159			REG_BADPAT, /* too many forward references */
160	chpe	1055	REG_BADPAT, /* disallowed UTF-8/16/32 code point (>= 0xd800 && <= 0xdfff) */
161	ph10	964	REG_BADPAT, /* invalid UTF-16 string (should not occur) */
162			/* 75 */
163	ph10	978	REG_BADPAT, /* overlong MARK name */
164	chpe	1055	REG_BADPAT, /* character value in \u.... sequence is too large */
165	ph10	1309	REG_BADPAT, /* invalid UTF-32 string (should not occur) */
166	ph10	1320	REG_BADPAT /* setting UTF is disabled by the application */
167	nigel	3	};
168
169			/* Table of texts corresponding to POSIX error codes */
170
171	nigel	73	static const char *const pstring[] = {
172	nigel	3	"", /* Dummy for value 0 */
173			"internal error", /* REG_ASSERT */
174			"invalid repeat counts in {}", /* BADBR */
175			"pattern error", /* BADPAT */
176			"? * + invalid", /* BADRPT */
177			"unbalanced {}", /* EBRACE */
178			"unbalanced []", /* EBRACK */
179			"collation error - not relevant", /* ECOLLATE */
180			"bad class", /* ECTYPE */
181			"bad escape sequence", /* EESCAPE */
182			"empty expression", /* EMPTY */
183			"unbalanced ()", /* EPAREN */
184			"bad range inside []", /* ERANGE */
185			"expression too big", /* ESIZE */
186			"failed to get memory", /* ESPACE */
187			"bad back reference", /* ESUBREG */
188			"bad argument", /* INVARG */
189			"match failed" /* NOMATCH */
190			};
191
192
193
194
195			/*************************************************
196			* Translate error code to string *
197			*************************************************/
198
199	ph10	359	PCREPOSIX_EXP_DEFN size_t PCRE_CALL_CONVENTION
200	nigel	3	regerror(int errcode, const regex_t preg, char errbuf, size_t errbuf_size)
201			{
202	nigel	7	const char message, addmessage;
203			size_t length, addlength;
204	nigel	3
205	nigel	7	message = (errcode >= (int)(sizeof(pstring)/sizeof(char *)))?
206	nigel	3	"unknown error code" : pstring[errcode];
207	nigel	7	length = strlen(message) + 1;
208	nigel	3
209	nigel	7	addmessage = " at offset ";
210			addlength = (preg != NULL && (int)preg->re_erroffset != -1)?
211			strlen(addmessage) + 6 : 0;
212	nigel	3
213			if (errbuf_size > 0)
214			{
215	nigel	7	if (addlength > 0 && errbuf_size >= length + addlength)
216	nigel	23	sprintf(errbuf, "%s%s%-6d", message, addmessage, (int)preg->re_erroffset);
217	nigel	3	else
218			{
219			strncpy(errbuf, message, errbuf_size - 1);
220			errbuf[errbuf_size-1] = 0;
221			}
222			}
223
224	nigel	7	return length + addlength;
225	nigel	3	}
226
227
228
229
230			/*************************************************
231			* Free store held by a regex *
232			*************************************************/
233
234	ph10	359	PCREPOSIX_EXP_DEFN void PCRE_CALL_CONVENTION
235	nigel	3	regfree(regex_t *preg)
236			{
237	ph10	836	(PUBL(free))(preg->re_pcre);
238	nigel	3	}
239
240
241
242
243			/*************************************************
244			* Compile a regular expression *
245			*************************************************/
246
247			/*
248			Arguments:
249			preg points to a structure for recording the compiled expression
250			pattern the pattern to compile
251			cflags compilation flags
252
253			Returns: 0 on success
254			various non-zero codes on failure
255			*/
256
257	ph10	359	PCREPOSIX_EXP_DEFN int PCRE_CALL_CONVENTION
258	nigel	3	regcomp(regex_t preg, const char pattern, int cflags)
259			{
260	nigel	7	const char *errorptr;
261	nigel	3	int erroffset;
262	nigel	77	int errorcode;
263	nigel	3	int options = 0;
264	ph10	1024	int re_nsub = 0;
265	nigel	3
266	ph10	432	if ((cflags & REG_ICASE) != 0) options \|= PCRE_CASELESS;
267			if ((cflags & REG_NEWLINE) != 0) options \|= PCRE_MULTILINE;
268			if ((cflags & REG_DOTALL) != 0) options \|= PCRE_DOTALL;
269			if ((cflags & REG_NOSUB) != 0) options \|= PCRE_NO_AUTO_CAPTURE;
270			if ((cflags & REG_UTF8) != 0) options \|= PCRE_UTF8;
271	ph10	518	if ((cflags & REG_UCP) != 0) options \|= PCRE_UCP;
272	ph10	432	if ((cflags & REG_UNGREEDY) != 0) options \|= PCRE_UNGREEDY;
273	nigel	3
274	nigel	77	preg->re_pcre = pcre_compile2(pattern, options, &errorcode, &errorptr,
275			&erroffset, NULL);
276	nigel	3	preg->re_erroffset = erroffset;
277
278	ph10	461	/* Safety: if the error code is too big for the translation vector (which
279	ph10	418	should not happen, but we all make mistakes), return REG_BADPAT. */
280	nigel	3
281	ph10	461	if (preg->re_pcre == NULL)
282	ph10	418	{
283	ph10	904	return (errorcode < (int)(sizeof(eint)/sizeof(const int)))?
284	ph10	418	eint[errorcode] : REG_BADPAT;
285	ph10	461	}
286	ph10	418
287	ph10	836	(void)pcre_fullinfo((const pcre *)preg->re_pcre, NULL, PCRE_INFO_CAPTURECOUNT,
288	ph10	1024	&re_nsub);
289	ph10	1221	preg->re_nsub = (size_t)re_nsub;
290	nigel	3	return 0;
291			}
292
293
294
295
296			/*************************************************
297			* Match a regular expression *
298			*************************************************/
299
300	nigel	41	/* Unfortunately, PCRE requires 3 ints of working space for each captured
301			substring, so we have to get and release working store instead of just using
302			the POSIX structures as was done in earlier releases when PCRE needed only 2
303	nigel	63	ints. However, if the number of possible capturing brackets is small, use a
304			block of store on the stack, to reduce the use of malloc/free. The threshold is
305	nigel	87	in a macro that can be changed at configure time.
306	nigel	41
307	nigel	87	If REG_NOSUB was specified at compile time, the PCRE_NO_AUTO_CAPTURE flag will
308			be set. When this is the case, the nmatch and pmatch arguments are ignored, and
309			the only result is yes/no/error. */
310
311	ph10	359	PCREPOSIX_EXP_DEFN int PCRE_CALL_CONVENTION
312	nigel	69	regexec(const regex_t preg, const char string, size_t nmatch,
313	nigel	3	regmatch_t pmatch[], int eflags)
314			{
315	ph10	332	int rc, so, eo;
316	nigel	3	int options = 0;
317	nigel	41	int *ovector = NULL;
318	nigel	63	int small_ovector[POSIX_MALLOC_THRESHOLD * 3];
319			BOOL allocated_ovector = FALSE;
320	nigel	87	BOOL nosub =
321	chpe	1055	(REAL_PCRE_OPTIONS((const pcre *)preg->re_pcre) & PCRE_NO_AUTO_CAPTURE) != 0;
322	nigel	3
323			if ((eflags & REG_NOTBOL) != 0) options \|= PCRE_NOTBOL;
324			if ((eflags & REG_NOTEOL) != 0) options \|= PCRE_NOTEOL;
325	ph10	388	if ((eflags & REG_NOTEMPTY) != 0) options \|= PCRE_NOTEMPTY;
326	nigel	3
327	nigel	69	((regex_t )preg)->re_erroffset = (size_t)(-1); / Only has meaning after compile */
328	nigel	3
329	ph10	461	/* When no string data is being returned, or no vector has been passed in which
330	ph10	432	to put it, ensure that nmatch is zero. Otherwise, ensure the vector for holding
331			the return data is large enough. */
332	nigel	87
333	ph10	432	if (nosub \|\| pmatch == NULL) nmatch = 0;
334	nigel	87
335			else if (nmatch > 0)
336	nigel	41	{
337	nigel	63	if (nmatch <= POSIX_MALLOC_THRESHOLD)
338			{
339			ovector = &(small_ovector[0]);
340			}
341			else
342			{
343	nigel	87	if (nmatch > INT_MAX/(sizeof(int) * 3)) return REG_ESPACE;
344	nigel	63	ovector = (int )malloc(sizeof(int) nmatch * 3);
345			if (ovector == NULL) return REG_ESPACE;
346			allocated_ovector = TRUE;
347			}
348	nigel	41	}
349
350	ph10	332	/* REG_STARTEND is a BSD extension, to allow for non-NUL-terminated strings.
351			The man page from OS X says "REG_STARTEND affects only the location of the
352	ph10	345	string, not how it is matched". That is why the "so" value is used to bump the
353	ph10	332	start location rather than being passed as a PCRE "starting offset". */
354
355	ph10	345	if ((eflags & REG_STARTEND) != 0)
356	ph10	332	{
357			so = pmatch[0].rm_so;
358			eo = pmatch[0].rm_eo;
359	ph10	345	}
360			else
361	ph10	332	{
362			so = 0;
363	ph10	530	eo = (int)strlen(string);
364	ph10	332	}
365
366			rc = pcre_exec((const pcre *)preg->re_pcre, NULL, string + so, (eo - so),
367	ph10	530	0, options, ovector, (int)(nmatch * 3));
368	nigel	3
369	ph10	530	if (rc == 0) rc = (int)nmatch; /* All captured slots were filled in */
370	nigel	3
371	ph10	499	/* Successful match */
372
373	nigel	41	if (rc >= 0)
374	nigel	3	{
375	nigel	7	size_t i;
376	nigel	87	if (!nosub)
377	nigel	41	{
378	nigel	87	for (i = 0; i < (size_t)rc; i++)
379			{
380			pmatch[i].rm_so = ovector[i*2];
381			pmatch[i].rm_eo = ovector[i*2+1];
382			}
383			if (allocated_ovector) free(ovector);
384			for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;
385	nigel	41	}
386	nigel	3	return 0;
387			}
388
389	ph10	499	/* Unsuccessful match */
390
391			if (allocated_ovector) free(ovector);
392			switch(rc)
393	nigel	3	{
394	ph10	499	/* ========================================================================== */
395			/* These cases are never obeyed. This is a fudge that causes a compile-time
396			error if the vector eint, which is indexed by compile-time error number, is
397			not the correct length. It seems to be the only way to do such a check at
398			compile time, as the sizeof() operator does not work in the C preprocessor.
399	ph10	507	As all the PCRE_ERROR_xxx values are negative, we can use 0 and 1. */
400	ph10	499
401			case 0:
402			case (sizeof(eint)/sizeof(int) == ERRCOUNT):
403			return REG_ASSERT;
404			/* ========================================================================== */
405
406			case PCRE_ERROR_NOMATCH: return REG_NOMATCH;
407			case PCRE_ERROR_NULL: return REG_INVARG;
408			case PCRE_ERROR_BADOPTION: return REG_INVARG;
409			case PCRE_ERROR_BADMAGIC: return REG_INVARG;
410			case PCRE_ERROR_UNKNOWN_NODE: return REG_ASSERT;
411			case PCRE_ERROR_NOMEMORY: return REG_ESPACE;
412			case PCRE_ERROR_MATCHLIMIT: return REG_ESPACE;
413			case PCRE_ERROR_BADUTF8: return REG_INVARG;
414			case PCRE_ERROR_BADUTF8_OFFSET: return REG_INVARG;
415	ph10	836	case PCRE_ERROR_BADMODE: return REG_INVARG;
416	ph10	499	default: return REG_ASSERT;
417	nigel	3	}
418			}
419
420			/* End of pcreposix.c */
Name	Value
svn:eol-style	native
svn:keywords	"Author Date Id Revision Url"