/[pcre]/code/trunk/pcre_internal.h
ViewVC logotype

Contents of /code/trunk/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log


Revision 149 - (show annotations) (download)
Mon Apr 16 15:28:08 2007 UTC (7 years, 7 months ago) by ph10
File MIME type: text/plain
File size: 45160 byte(s)
Add PCRE_NEWLINE_ANYCRLF.

1 /*************************************************
2 * Perl-Compatible Regular Expressions *
3 *************************************************/
4
5
6 /* PCRE is a library of functions to support regular expressions whose syntax
7 and semantics are as close as possible to those of the Perl 5 language.
8
9 Written by Philip Hazel
10 Copyright (c) 1997-2007 University of Cambridge
11
12 -----------------------------------------------------------------------------
13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions are met:
15
16 * Redistributions of source code must retain the above copyright notice,
17 this list of conditions and the following disclaimer.
18
19 * Redistributions in binary form must reproduce the above copyright
20 notice, this list of conditions and the following disclaimer in the
21 documentation and/or other materials provided with the distribution.
22
23 * Neither the name of the University of Cambridge nor the names of its
24 contributors may be used to endorse or promote products derived from
25 this software without specific prior written permission.
26
27 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
28 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
29 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
30 ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
31 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
34 INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
35 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
36 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
37 POSSIBILITY OF SUCH DAMAGE.
38 -----------------------------------------------------------------------------
39 */
40
41 /* This header contains definitions that are shared between the different
42 modules, but which are not relevant to the exported API. This includes some
43 functions whose names all begin with "_pcre_". */
44
45 #ifndef PCRE_INTERNAL_H
46 #define PCRE_INTERNAL_H
47
48 /* Define DEBUG to get debugging output on stdout. */
49
50 #if 0
51 #define DEBUG
52 #endif
53
54 /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
55 inline, and there are *still* stupid compilers about that don't like indented
56 pre-processor statements, or at least there were when I first wrote this. After
57 all, it had only been about 10 years then...
58
59 It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so
60 be absolutely sure we get our version. */
61
62 #undef DPRINTF
63 #ifdef DEBUG
64 #define DPRINTF(p) printf p
65 #else
66 #define DPRINTF(p) /* Nothing */
67 #endif
68
69
70 /* Get the definitions provided by running "configure" */
71
72 #include "config.h"
73
74 /* Standard C headers plus the external interface definition. The only time
75 setjmp and stdarg are used is when NO_RECURSE is set. */
76
77 #include <ctype.h>
78 #include <limits.h>
79 #include <setjmp.h>
80 #include <stdarg.h>
81 #include <stddef.h>
82 #include <stdio.h>
83 #include <stdlib.h>
84 #include <string.h>
85
86 /* When compiling a DLL for Windows, the exported symbols have to be declared
87 using some MS magic. I found some useful information on this web page:
88 http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
89 information there, using __declspec(dllesport) without "extern" we have a
90 definition; with "extern" we have a declaration. The settings here override the
91 setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL,
92 which is all that is needed for applications, which import the symbols. We use:
93
94 PCRE_EXP_DECL for declarations
95 PCRE_EXP_DEFN for definitions of exported functions
96 PCRE_EXP_DATA_DEFN for definitions of exported variables
97
98 The reason for the two DEFN macros is that in non-Windows environments, one
99 does not want to have "extern" before variable definitions because it leads to
100 compiler warnings. So we distinguish between functions and variables. In
101 Windows, the two should always be the same.
102
103 The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest,
104 which is an application, but needs to import this file in order to "peek" at
105 internals, can #include pcre.h first, can get an application's-eye view.
106
107 In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon,
108 special-purpose environments) might want to stick other stuff in front of
109 exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and
110 PCRE_EXP_DATA_DEFN only if they are not already set. */
111
112 #ifndef PCRE_EXP_DECL
113 # ifdef _WIN32
114 # ifdef DLL_EXPORT
115 # define PCRE_EXP_DECL extern __declspec(dllexport)
116 # define PCRE_EXP_DEFN __declspec(dllexport)
117 # define PCRE_EXP_DATA_DEFN __declspec(dllexport)
118 # else
119 # define PCRE_EXP_DECL extern
120 # define PCRE_EXP_DEFN
121 # define PCRE_EXP_DATA_DEFN
122 # endif
123 #
124 # else
125 # ifdef __cplusplus
126 # define PCRE_EXP_DECL extern "C"
127 # else
128 # define PCRE_EXP_DECL extern
129 # endif
130 # ifndef PCRE_EXP_DEFN
131 # define PCRE_EXP_DEFN PCRE_EXP_DECL
132 # endif
133 # ifndef PCRE_EXP_DATA_DEFN
134 # define PCRE_EXP_DATA_DEFN
135 # endif
136 # endif
137 #endif
138
139 /* We need to have types that specify unsigned 16-bit and 32-bit integers. We
140 cannot determine these outside the compilation (e.g. by running a program as
141 part of "configure") because PCRE is often cross-compiled for use on other
142 systems. Instead we make use of the maximum sizes that are available at
143 preprocessor time in standard C environments. */
144
145 #if USHRT_MAX == 65535
146 typedef unsigned short pcre_uint16;
147 #elif UINT_MAX == 65535
148 typedef unsigned int pcre_uint16;
149 #else
150 #error Cannot determine a type for 16-bit unsigned integers
151 #endif
152
153 #if UINT_MAX == 4294967295
154 typedef unsigned int pcre_uint32;
155 #elif ULONG_MAX == 4294967295
156 typedef unsigned long int pcre_uint32;
157 #else
158 #error Cannot determine a type for 32-bit unsigned integers
159 #endif
160
161 /* All character handling must be done as unsigned characters. Otherwise there
162 are problems with top-bit-set characters and functions such as isspace().
163 However, we leave the interface to the outside world as char *, because that
164 should make things easier for callers. We define a short type for unsigned char
165 to save lots of typing. I tried "uchar", but it causes problems on Digital
166 Unix, where it is defined in sys/types, so use "uschar" instead. */
167
168 typedef unsigned char uschar;
169
170 /* This is an unsigned int value that no character can ever have. UTF-8
171 characters only go up to 0x7fffffff (though Unicode doesn't go beyond
172 0x0010ffff). */
173
174 #define NOTACHAR 0xffffffff
175
176 /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
177 "any" and "anycrlf" at present). The following macros are used to package up
178 testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
179 modules to indicate in which datablock the parameters exist, and what the
180 start/end of string field names are. */
181
182 #define NLTYPE_FIXED 0 /* Newline is a fixed length string */
183 #define NLTYPE_ANY 1 /* Newline is any Unicode line ending */
184 #define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */
185
186 /* This macro checks for a newline at the given position */
187
188 #define IS_NEWLINE(p) \
189 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
190 ((p) < NLBLOCK->PSEND && \
191 _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\
192 utf8)) \
193 : \
194 ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
195 (p)[0] == NLBLOCK->nl[0] && \
196 (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \
197 ) \
198 )
199
200 /* This macro checks for a newline immediately preceding the given position */
201
202 #define WAS_NEWLINE(p) \
203 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
204 ((p) > NLBLOCK->PSSTART && \
205 _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
206 &(NLBLOCK->nllen), utf8)) \
207 : \
208 ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
209 (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \
210 (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \
211 ) \
212 )
213
214 /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
215 with a custom type. This makes it possible, for example, to allow pcre_exec()
216 to process subject strings that are discontinuous by using a smart pointer
217 class. It must always be possible to inspect all of the subject string in
218 pcre_exec() because of the way it backtracks. Two macros are required in the
219 normal case, for sign-unspecified and unsigned char pointers. The former is
220 used for the external interface and appears in pcre.h, which is why its name
221 must begin with PCRE_. */
222
223 #ifdef CUSTOM_SUBJECT_PTR
224 #define PCRE_SPTR CUSTOM_SUBJECT_PTR
225 #define USPTR CUSTOM_SUBJECT_PTR
226 #else
227 #define PCRE_SPTR const char *
228 #define USPTR const unsigned char *
229 #endif
230
231
232
233 /* Include the public PCRE header and the definitions of UCP character property
234 values. */
235
236 #include <pcre.h>
237 #include "ucp.h"
238
239 /* When compiling for use with the Virtual Pascal compiler, these functions
240 need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
241 option on the command line. */
242
243 #ifdef VPCOMPAT
244 #define strlen(s) _strlen(s)
245 #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
246 #define memcmp(s,c,n) _memcmp(s,c,n)
247 #define memcpy(d,s,n) _memcpy(d,s,n)
248 #define memmove(d,s,n) _memmove(d,s,n)
249 #define memset(s,c,n) _memset(s,c,n)
250 #else /* VPCOMPAT */
251
252 /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
253 define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
254 is set. Otherwise, include an emulating function for those systems that have
255 neither (there some non-Unix environments where this is the case). */
256
257 #ifndef HAVE_MEMMOVE
258 #undef memmove /* some systems may have a macro */
259 #ifdef HAVE_BCOPY
260 #define memmove(a, b, c) bcopy(b, a, c)
261 #else /* HAVE_BCOPY */
262 static void *
263 pcre_memmove(void *d, const void *s, size_t n)
264 {
265 size_t i;
266 unsigned char *dest = (unsigned char *)d;
267 const unsigned char *src = (const unsigned char *)s;
268 if (dest > src)
269 {
270 dest += n;
271 src += n;
272 for (i = 0; i < n; ++i) *(--dest) = *(--src);
273 return (void *)dest;
274 }
275 else
276 {
277 for (i = 0; i < n; ++i) *dest++ = *src++;
278 return (void *)(dest - n);
279 }
280 }
281 #define memmove(a, b, c) pcre_memmove(a, b, c)
282 #endif /* not HAVE_BCOPY */
283 #endif /* not HAVE_MEMMOVE */
284 #endif /* not VPCOMPAT */
285
286
287 /* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
288 in big-endian order) by default. These are used, for example, to link from the
289 start of a subpattern to its alternatives and its end. The use of 2 bytes per
290 offset limits the size of the compiled regex to around 64K, which is big enough
291 for almost everybody. However, I received a request for an even bigger limit.
292 For this reason, and also to make the code easier to maintain, the storing and
293 loading of offsets from the byte string is now handled by the macros that are
294 defined here.
295
296 The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
297 the config.h file, but can be overridden by using -D on the command line. This
298 is automated on Unix systems via the "configure" command. */
299
300 #if LINK_SIZE == 2
301
302 #define PUT(a,n,d) \
303 (a[n] = (d) >> 8), \
304 (a[(n)+1] = (d) & 255)
305
306 #define GET(a,n) \
307 (((a)[n] << 8) | (a)[(n)+1])
308
309 #define MAX_PATTERN_SIZE (1 << 16)
310
311
312 #elif LINK_SIZE == 3
313
314 #define PUT(a,n,d) \
315 (a[n] = (d) >> 16), \
316 (a[(n)+1] = (d) >> 8), \
317 (a[(n)+2] = (d) & 255)
318
319 #define GET(a,n) \
320 (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
321
322 #define MAX_PATTERN_SIZE (1 << 24)
323
324
325 #elif LINK_SIZE == 4
326
327 #define PUT(a,n,d) \
328 (a[n] = (d) >> 24), \
329 (a[(n)+1] = (d) >> 16), \
330 (a[(n)+2] = (d) >> 8), \
331 (a[(n)+3] = (d) & 255)
332
333 #define GET(a,n) \
334 (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
335
336 #define MAX_PATTERN_SIZE (1 << 30) /* Keep it positive */
337
338
339 #else
340 #error LINK_SIZE must be either 2, 3, or 4
341 #endif
342
343
344 /* Convenience macro defined in terms of the others */
345
346 #define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
347
348
349 /* PCRE uses some other 2-byte quantities that do not change when the size of
350 offsets changes. There are used for repeat counts and for other things such as
351 capturing parenthesis numbers in back references. */
352
353 #define PUT2(a,n,d) \
354 a[n] = (d) >> 8; \
355 a[(n)+1] = (d) & 255
356
357 #define GET2(a,n) \
358 (((a)[n] << 8) | (a)[(n)+1])
359
360 #define PUT2INC(a,n,d) PUT2(a,n,d), a += 2
361
362
363 /* When UTF-8 encoding is being used, a character is no longer just a single
364 byte. The macros for character handling generate simple sequences when used in
365 byte-mode, and more complicated ones for UTF-8 characters. */
366
367 #ifndef SUPPORT_UTF8
368 #define GETCHAR(c, eptr) c = *eptr;
369 #define GETCHARTEST(c, eptr) c = *eptr;
370 #define GETCHARINC(c, eptr) c = *eptr++;
371 #define GETCHARINCTEST(c, eptr) c = *eptr++;
372 #define GETCHARLEN(c, eptr, len) c = *eptr;
373 #define BACKCHAR(eptr)
374
375 #else /* SUPPORT_UTF8 */
376
377 /* Get the next UTF-8 character, not advancing the pointer. This is called when
378 we know we are in UTF-8 mode. */
379
380 #define GETCHAR(c, eptr) \
381 c = *eptr; \
382 if (c >= 0xc0) \
383 { \
384 int gcii; \
385 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
386 int gcss = 6*gcaa; \
387 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
388 for (gcii = 1; gcii <= gcaa; gcii++) \
389 { \
390 gcss -= 6; \
391 c |= (eptr[gcii] & 0x3f) << gcss; \
392 } \
393 }
394
395 /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
396 pointer. */
397
398 #define GETCHARTEST(c, eptr) \
399 c = *eptr; \
400 if (utf8 && c >= 0xc0) \
401 { \
402 int gcii; \
403 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
404 int gcss = 6*gcaa; \
405 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
406 for (gcii = 1; gcii <= gcaa; gcii++) \
407 { \
408 gcss -= 6; \
409 c |= (eptr[gcii] & 0x3f) << gcss; \
410 } \
411 }
412
413 /* Get the next UTF-8 character, advancing the pointer. This is called when we
414 know we are in UTF-8 mode. */
415
416 #define GETCHARINC(c, eptr) \
417 c = *eptr++; \
418 if (c >= 0xc0) \
419 { \
420 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
421 int gcss = 6*gcaa; \
422 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
423 while (gcaa-- > 0) \
424 { \
425 gcss -= 6; \
426 c |= (*eptr++ & 0x3f) << gcss; \
427 } \
428 }
429
430 /* Get the next character, testing for UTF-8 mode, and advancing the pointer */
431
432 #define GETCHARINCTEST(c, eptr) \
433 c = *eptr++; \
434 if (utf8 && c >= 0xc0) \
435 { \
436 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
437 int gcss = 6*gcaa; \
438 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
439 while (gcaa-- > 0) \
440 { \
441 gcss -= 6; \
442 c |= (*eptr++ & 0x3f) << gcss; \
443 } \
444 }
445
446 /* Get the next UTF-8 character, not advancing the pointer, incrementing length
447 if there are extra bytes. This is called when we know we are in UTF-8 mode. */
448
449 #define GETCHARLEN(c, eptr, len) \
450 c = *eptr; \
451 if (c >= 0xc0) \
452 { \
453 int gcii; \
454 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
455 int gcss = 6*gcaa; \
456 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
457 for (gcii = 1; gcii <= gcaa; gcii++) \
458 { \
459 gcss -= 6; \
460 c |= (eptr[gcii] & 0x3f) << gcss; \
461 } \
462 len += gcaa; \
463 }
464
465 /* If the pointer is not at the start of a character, move it back until
466 it is. Called only in UTF-8 mode. */
467
468 #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--;
469
470 #endif
471
472
473 /* In case there is no definition of offsetof() provided - though any proper
474 Standard C system should have one. */
475
476 #ifndef offsetof
477 #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))
478 #endif
479
480
481 /* These are the public options that can change during matching. */
482
483 #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)
484
485 /* Private options flags start at the most significant end of the four bytes.
486 The public options defined in pcre.h start at the least significant end. Make
487 sure they don't overlap! The bits are getting a bit scarce now -- when we run
488 out, there is a dummy word in the structure that could be used for the private
489 bits. */
490
491 #define PCRE_NOPARTIAL 0x80000000 /* can't use partial with this regex */
492 #define PCRE_FIRSTSET 0x40000000 /* first_byte is set */
493 #define PCRE_REQCHSET 0x20000000 /* req_byte is set */
494 #define PCRE_STARTLINE 0x10000000 /* start after \n for multiline */
495 #define PCRE_JCHANGED 0x08000000 /* j option changes within regex */
496
497 /* Options for the "extra" block produced by pcre_study(). */
498
499 #define PCRE_STUDY_MAPPED 0x01 /* a map of starting chars exists */
500
501 /* Masks for identifying the public options that are permitted at compile
502 time, run time, or study time, respectively. */
503
504 #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
505 PCRE_NEWLINE_ANYCRLF)
506
507 #define PUBLIC_OPTIONS \
508 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
509 PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
510 PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
511 PCRE_DUPNAMES|PCRE_NEWLINE_BITS)
512
513 #define PUBLIC_EXEC_OPTIONS \
514 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \
515 PCRE_PARTIAL|PCRE_NEWLINE_BITS)
516
517 #define PUBLIC_DFA_EXEC_OPTIONS \
518 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \
519 PCRE_PARTIAL|PCRE_DFA_SHORTEST|PCRE_DFA_RESTART|PCRE_NEWLINE_BITS)
520
521 #define PUBLIC_STUDY_OPTIONS 0 /* None defined */
522
523 /* Magic number to provide a small check against being handed junk. Also used
524 to detect whether a pattern was compiled on a host of different endianness. */
525
526 #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
527
528 /* Negative values for the firstchar and reqchar variables */
529
530 #define REQ_UNSET (-2)
531 #define REQ_NONE (-1)
532
533 /* The maximum remaining length of subject we are prepared to search for a
534 req_byte match. */
535
536 #define REQ_BYTE_MAX 1000
537
538 /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a
539 variable-length repeat, or a anything other than literal characters. */
540
541 #define REQ_CASELESS 0x0100 /* indicates caselessness */
542 #define REQ_VARY 0x0200 /* reqbyte followed non-literal item */
543
544 /* Miscellaneous definitions */
545
546 typedef int BOOL;
547
548 #define FALSE 0
549 #define TRUE 1
550
551 /* Escape items that are just an encoding of a particular data value. */
552
553 #ifndef ESC_e
554 #define ESC_e 27
555 #endif
556
557 #ifndef ESC_f
558 #define ESC_f '\f'
559 #endif
560
561 #ifndef ESC_n
562 #define ESC_n '\n'
563 #endif
564
565 #ifndef ESC_r
566 #define ESC_r '\r'
567 #endif
568
569 /* We can't officially use ESC_t because it is a POSIX reserved identifier
570 (presumably because of all the others like size_t). */
571
572 #ifndef ESC_tee
573 #define ESC_tee '\t'
574 #endif
575
576 /* Codes for different types of Unicode property */
577
578 #define PT_ANY 0 /* Any property - matches all chars */
579 #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
580 #define PT_GC 2 /* General characteristic (e.g. L) */
581 #define PT_PC 3 /* Particular characteristic (e.g. Lu) */
582 #define PT_SC 4 /* Script (e.g. Han) */
583
584 /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
585 contain UTF-8 characters with values greater than 255. */
586
587 #define XCL_NOT 0x01 /* Flag: this is a negative class */
588 #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
589
590 #define XCL_END 0 /* Marks end of individual items */
591 #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
592 #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
593 #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
594 #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
595
596 /* These are escaped items that aren't just an encoding of a particular data
597 value such as \n. They must have non-zero values, as check_escape() returns
598 their negation. Also, they must appear in the same order as in the opcode
599 definitions below, up to ESC_z. There's a dummy for OP_ANY because it
600 corresponds to "." rather than an escape sequence. The final one must be
601 ESC_REF as subsequent values are used for backreferences (\1, \2, \3, etc).
602 There are two tests in the code for an escape greater than ESC_b and less than
603 ESC_Z to detect the types that may be repeated. These are the types that
604 consume characters. If any new escapes are put in between that don't consume a
605 character, that code will have to change. */
606
607 enum { ESC_A = 1, ESC_G, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, ESC_W,
608 ESC_w, ESC_dum1, ESC_C, ESC_P, ESC_p, ESC_R, ESC_X, ESC_Z, ESC_z,
609 ESC_E, ESC_Q, ESC_k, ESC_REF };
610
611
612 /* Opcode table: OP_BRA must be last, as all values >= it are used for brackets
613 that extract substrings. Starting from 1 (i.e. after OP_END), the values up to
614 OP_EOD must correspond in order to the list of escapes immediately above.
615
616 To keep stored, compiled patterns compatible, new opcodes should be added
617 immediately before OP_BRA, where (since release 7.0) a gap is left for this
618 purpose.
619
620 *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
621 that follow must also be updated to match. There is also a table called
622 "coptable" in pcre_dfa_exec.c that must be updated. */
623
624 enum {
625 OP_END, /* 0 End of pattern */
626
627 /* Values corresponding to backslashed metacharacters */
628
629 OP_SOD, /* 1 Start of data: \A */
630 OP_SOM, /* 2 Start of match (subject + offset): \G */
631 OP_NOT_WORD_BOUNDARY, /* 3 \B */
632 OP_WORD_BOUNDARY, /* 4 \b */
633 OP_NOT_DIGIT, /* 5 \D */
634 OP_DIGIT, /* 6 \d */
635 OP_NOT_WHITESPACE, /* 7 \S */
636 OP_WHITESPACE, /* 8 \s */
637 OP_NOT_WORDCHAR, /* 9 \W */
638 OP_WORDCHAR, /* 10 \w */
639 OP_ANY, /* 11 Match any character */
640 OP_ANYBYTE, /* 12 Match any byte (\C); different to OP_ANY for UTF-8 */
641 OP_NOTPROP, /* 13 \P (not Unicode property) */
642 OP_PROP, /* 14 \p (Unicode property) */
643 OP_ANYNL, /* 15 \R (any newline sequence) */
644 OP_EXTUNI, /* 16 \X (extended Unicode sequence */
645 OP_EODN, /* 17 End of data or \n at end of data: \Z. */
646 OP_EOD, /* 18 End of data: \z */
647
648 OP_OPT, /* 19 Set runtime options */
649 OP_CIRC, /* 20 Start of line - varies with multiline switch */
650 OP_DOLL, /* 21 End of line - varies with multiline switch */
651 OP_CHAR, /* 22 Match one character, casefully */
652 OP_CHARNC, /* 23 Match one character, caselessly */
653 OP_NOT, /* 24 Match one character, not the following one */
654
655 OP_STAR, /* 25 The maximizing and minimizing versions of */
656 OP_MINSTAR, /* 26 these six opcodes must come in pairs, with */
657 OP_PLUS, /* 27 the minimizing one second. */
658 OP_MINPLUS, /* 28 This first set applies to single characters.*/
659 OP_QUERY, /* 29 */
660 OP_MINQUERY, /* 30 */
661
662 OP_UPTO, /* 31 From 0 to n matches */
663 OP_MINUPTO, /* 32 */
664 OP_EXACT, /* 33 Exactly n matches */
665
666 OP_POSSTAR, /* 34 Possessified star */
667 OP_POSPLUS, /* 35 Possessified plus */
668 OP_POSQUERY, /* 36 Posesssified query */
669 OP_POSUPTO, /* 37 Possessified upto */
670
671 OP_NOTSTAR, /* 38 The maximizing and minimizing versions of */
672 OP_NOTMINSTAR, /* 39 these six opcodes must come in pairs, with */
673 OP_NOTPLUS, /* 40 the minimizing one second. They must be in */
674 OP_NOTMINPLUS, /* 41 exactly the same order as those above. */
675 OP_NOTQUERY, /* 42 This set applies to "not" single characters. */
676 OP_NOTMINQUERY, /* 43 */
677
678 OP_NOTUPTO, /* 44 From 0 to n matches */
679 OP_NOTMINUPTO, /* 45 */
680 OP_NOTEXACT, /* 46 Exactly n matches */
681
682 OP_NOTPOSSTAR, /* 47 Possessified versions */
683 OP_NOTPOSPLUS, /* 48 */
684 OP_NOTPOSQUERY, /* 49 */
685 OP_NOTPOSUPTO, /* 50 */
686
687 OP_TYPESTAR, /* 51 The maximizing and minimizing versions of */
688 OP_TYPEMINSTAR, /* 52 these six opcodes must come in pairs, with */
689 OP_TYPEPLUS, /* 53 the minimizing one second. These codes must */
690 OP_TYPEMINPLUS, /* 54 be in exactly the same order as those above. */
691 OP_TYPEQUERY, /* 55 This set applies to character types such as \d */
692 OP_TYPEMINQUERY, /* 56 */
693
694 OP_TYPEUPTO, /* 57 From 0 to n matches */
695 OP_TYPEMINUPTO, /* 58 */
696 OP_TYPEEXACT, /* 59 Exactly n matches */
697
698 OP_TYPEPOSSTAR, /* 60 Possessified versions */
699 OP_TYPEPOSPLUS, /* 61 */
700 OP_TYPEPOSQUERY, /* 62 */
701 OP_TYPEPOSUPTO, /* 63 */
702
703 OP_CRSTAR, /* 64 The maximizing and minimizing versions of */
704 OP_CRMINSTAR, /* 65 all these opcodes must come in pairs, with */
705 OP_CRPLUS, /* 66 the minimizing one second. These codes must */
706 OP_CRMINPLUS, /* 67 be in exactly the same order as those above. */
707 OP_CRQUERY, /* 68 These are for character classes and back refs */
708 OP_CRMINQUERY, /* 69 */
709 OP_CRRANGE, /* 70 These are different to the three sets above. */
710 OP_CRMINRANGE, /* 71 */
711
712 OP_CLASS, /* 72 Match a character class, chars < 256 only */
713 OP_NCLASS, /* 73 Same, but the bitmap was created from a negative
714 class - the difference is relevant only when a UTF-8
715 character > 255 is encountered. */
716
717 OP_XCLASS, /* 74 Extended class for handling UTF-8 chars within the
718 class. This does both positive and negative. */
719
720 OP_REF, /* 75 Match a back reference */
721 OP_RECURSE, /* 76 Match a numbered subpattern (possibly recursive) */
722 OP_CALLOUT, /* 77 Call out to external function if provided */
723
724 OP_ALT, /* 78 Start of alternation */
725 OP_KET, /* 79 End of group that doesn't have an unbounded repeat */
726 OP_KETRMAX, /* 80 These two must remain together and in this */
727 OP_KETRMIN, /* 81 order. They are for groups the repeat for ever. */
728
729 /* The assertions must come before BRA, CBRA, ONCE, and COND.*/
730
731 OP_ASSERT, /* 82 Positive lookahead */
732 OP_ASSERT_NOT, /* 83 Negative lookahead */
733 OP_ASSERTBACK, /* 84 Positive lookbehind */
734 OP_ASSERTBACK_NOT, /* 85 Negative lookbehind */
735 OP_REVERSE, /* 86 Move pointer back - used in lookbehind assertions */
736
737 /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,
738 as there's a test for >= ONCE for a subpattern that isn't an assertion. */
739
740 OP_ONCE, /* 87 Atomic group */
741 OP_BRA, /* 88 Start of non-capturing bracket */
742 OP_CBRA, /* 89 Start of capturing bracket */
743 OP_COND, /* 90 Conditional group */
744
745 /* These three must follow the previous three, in the same order. There's a
746 check for >= SBRA to distinguish the two sets. */
747
748 OP_SBRA, /* 91 Start of non-capturing bracket, check empty */
749 OP_SCBRA, /* 92 Start of capturing bracket, check empty */
750 OP_SCOND, /* 93 Conditional group, check empty */
751
752 OP_CREF, /* 94 Used to hold a capture number as condition */
753 OP_RREF, /* 95 Used to hold a recursion number as condition */
754 OP_DEF, /* 96 The DEFINE condition */
755
756 OP_BRAZERO, /* 97 These two must remain together and in this */
757 OP_BRAMINZERO /* 98 order. */
758 };
759
760
761 /* This macro defines textual names for all the opcodes. These are used only
762 for debugging. The macro is referenced only in pcre_printint.c. */
763
764 #define OP_NAME_LIST \
765 "End", "\\A", "\\G", "\\B", "\\b", "\\D", "\\d", \
766 "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte", \
767 "notprop", "prop", "anynl", "extuni", \
768 "\\Z", "\\z", \
769 "Opt", "^", "$", "char", "charnc", "not", \
770 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
771 "*+","++", "?+", "{", \
772 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
773 "*+","++", "?+", "{", \
774 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
775 "*+","++", "?+", "{", \
776 "*", "*?", "+", "+?", "?", "??", "{", "{", \
777 "class", "nclass", "xclass", "Ref", "Recurse", "Callout", \
778 "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", \
779 "AssertB", "AssertB not", "Reverse", \
780 "Once", "Bra 0", "Bra", "Cond", "SBra 0", "SBra", "SCond", \
781 "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero"
782
783
784 /* This macro defines the length of fixed length operations in the compiled
785 regex. The lengths are used when searching for specific things, and also in the
786 debugging printing of a compiled regex. We use a macro so that it can be
787 defined close to the definitions of the opcodes themselves.
788
789 As things have been extended, some of these are no longer fixed lenths, but are
790 minima instead. For example, the length of a single-character repeat may vary
791 in UTF-8 mode. The code that uses this table must know about such things. */
792
793 #define OP_LENGTHS \
794 1, /* End */ \
795 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* \A, \G, \B, \B, \D, \d, \S, \s, \W, \w */ \
796 1, 1, /* Any, Anybyte */ \
797 3, 3, 1, 1, /* NOTPROP, PROP, EXTUNI, ANYNL */ \
798 1, 1, 2, 1, 1, /* \Z, \z, Opt, ^, $ */ \
799 2, /* Char - the minimum length */ \
800 2, /* Charnc - the minimum length */ \
801 2, /* not */ \
802 /* Positive single-char repeats ** These are */ \
803 2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
804 4, 4, 4, /* upto, minupto, exact ** UTF-8 mode */ \
805 2, 2, 2, 4, /* *+, ++, ?+, upto+ */ \
806 /* Negative single-char repeats - only for chars < 256 */ \
807 2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
808 4, 4, 4, /* NOT upto, minupto, exact */ \
809 2, 2, 2, 4, /* Possessive *, +, ?, upto */ \
810 /* Positive type repeats */ \
811 2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
812 4, 4, 4, /* Type upto, minupto, exact */ \
813 2, 2, 2, 4, /* Possessive *+, ++, ?+, upto+ */ \
814 /* Character class & ref repeats */ \
815 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
816 5, 5, /* CRRANGE, CRMINRANGE */ \
817 33, /* CLASS */ \
818 33, /* NCLASS */ \
819 0, /* XCLASS - variable length */ \
820 3, /* REF */ \
821 1+LINK_SIZE, /* RECURSE */ \
822 2+2*LINK_SIZE, /* CALLOUT */ \
823 1+LINK_SIZE, /* Alt */ \
824 1+LINK_SIZE, /* Ket */ \
825 1+LINK_SIZE, /* KetRmax */ \
826 1+LINK_SIZE, /* KetRmin */ \
827 1+LINK_SIZE, /* Assert */ \
828 1+LINK_SIZE, /* Assert not */ \
829 1+LINK_SIZE, /* Assert behind */ \
830 1+LINK_SIZE, /* Assert behind not */ \
831 1+LINK_SIZE, /* Reverse */ \
832 1+LINK_SIZE, /* ONCE */ \
833 1+LINK_SIZE, /* BRA */ \
834 3+LINK_SIZE, /* CBRA */ \
835 1+LINK_SIZE, /* COND */ \
836 1+LINK_SIZE, /* SBRA */ \
837 3+LINK_SIZE, /* SCBRA */ \
838 1+LINK_SIZE, /* SCOND */ \
839 3, /* CREF */ \
840 3, /* RREF */ \
841 1, /* DEF */ \
842 1, 1, /* BRAZERO, BRAMINZERO */ \
843
844
845 /* A magic value for OP_RREF to indicate the "any recursion" condition. */
846
847 #define RREF_ANY 0xffff
848
849 /* Error code numbers. They are given names so that they can more easily be
850 tracked. */
851
852 enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
853 ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
854 ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
855 ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
856 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
857 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57 };
858
859 /* The real format of the start of the pcre block; the index of names and the
860 code vector run on as long as necessary after the end. We store an explicit
861 offset to the name table so that if a regex is compiled on one host, saved, and
862 then run on another where the size of pointers is different, all might still
863 be well. For the case of compiled-on-4 and run-on-8, we include an extra
864 pointer that is always NULL. For future-proofing, a few dummy fields were
865 originally included - even though you can never get this planning right - but
866 there is only one left now.
867
868 NOTE NOTE NOTE:
869 Because people can now save and re-use compiled patterns, any additions to this
870 structure should be made at the end, and something earlier (e.g. a new
871 flag in the options or one of the dummy fields) should indicate that the new
872 fields are present. Currently PCRE always sets the dummy fields to zero.
873 NOTE NOTE NOTE:
874 */
875
876 typedef struct real_pcre {
877 pcre_uint32 magic_number;
878 pcre_uint32 size; /* Total that was malloced */
879 pcre_uint32 options;
880 pcre_uint32 dummy1; /* For future use, maybe */
881
882 pcre_uint16 top_bracket;
883 pcre_uint16 top_backref;
884 pcre_uint16 first_byte;
885 pcre_uint16 req_byte;
886 pcre_uint16 name_table_offset; /* Offset to name table that follows */
887 pcre_uint16 name_entry_size; /* Size of any name items */
888 pcre_uint16 name_count; /* Number of name items */
889 pcre_uint16 ref_count; /* Reference count */
890
891 const unsigned char *tables; /* Pointer to tables or NULL for std */
892 const unsigned char *nullpad; /* NULL padding */
893 } real_pcre;
894
895 /* The format of the block used to store data from pcre_study(). The same
896 remark (see NOTE above) about extending this structure applies. */
897
898 typedef struct pcre_study_data {
899 pcre_uint32 size; /* Total that was malloced */
900 pcre_uint32 options;
901 uschar start_bits[32];
902 } pcre_study_data;
903
904 /* Structure for passing "static" information around between the functions
905 doing the compiling, so that they are thread-safe. */
906
907 typedef struct compile_data {
908 const uschar *lcc; /* Points to lower casing table */
909 const uschar *fcc; /* Points to case-flipping table */
910 const uschar *cbits; /* Points to character type table */
911 const uschar *ctypes; /* Points to table of type maps */
912 const uschar *start_workspace;/* The start of working space */
913 const uschar *start_code; /* The start of the compiled code */
914 const uschar *start_pattern; /* The start of the pattern */
915 const uschar *end_pattern; /* The end of the pattern */
916 uschar *hwm; /* High watermark of workspace */
917 uschar *name_table; /* The name/number table */
918 int names_found; /* Number of entries so far */
919 int name_entry_size; /* Size of each entry */
920 int bracount; /* Count of capturing parens */
921 int top_backref; /* Maximum back reference */
922 unsigned int backref_map; /* Bitmap of low back refs */
923 int external_options; /* External (initial) options */
924 int req_varyopt; /* "After variable item" flag for reqbyte */
925 BOOL nopartial; /* Set TRUE if partial won't work */
926 int nltype; /* Newline type */
927 int nllen; /* Newline string length */
928 uschar nl[4]; /* Newline string when fixed length */
929 } compile_data;
930
931 /* Structure for maintaining a chain of pointers to the currently incomplete
932 branches, for testing for left recursion. */
933
934 typedef struct branch_chain {
935 struct branch_chain *outer;
936 uschar *current;
937 } branch_chain;
938
939 /* Structure for items in a linked list that represents an explicit recursive
940 call within the pattern. */
941
942 typedef struct recursion_info {
943 struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
944 int group_num; /* Number of group that was called */
945 const uschar *after_call; /* "Return value": points after the call in the expr */
946 USPTR save_start; /* Old value of md->start_match */
947 int *offset_save; /* Pointer to start of saved offsets */
948 int saved_max; /* Number of saved offsets */
949 } recursion_info;
950
951 /* When compiling in a mode that doesn't use recursive calls to match(),
952 a structure is used to remember local variables on the heap. It is defined in
953 pcre_exec.c, close to the match() function, so that it is easy to keep it in
954 step with any changes of local variable. However, the pointer to the current
955 frame must be saved in some "static" place over a longjmp(). We declare the
956 structure here so that we can put a pointer in the match_data structure. NOTE:
957 This isn't used for a "normal" compilation of pcre. */
958
959 struct heapframe;
960
961 /* Structure for building a chain of data for holding the values of the subject
962 pointer at the start of each subpattern, so as to detect when an empty string
963 has been matched by a subpattern - to break infinite loops. */
964
965 typedef struct eptrblock {
966 struct eptrblock *epb_prev;
967 USPTR epb_saved_eptr;
968 } eptrblock;
969
970
971 /* Structure for passing "static" information around between the functions
972 doing traditional NFA matching, so that they are thread-safe. */
973
974 typedef struct match_data {
975 unsigned long int match_call_count; /* As it says */
976 unsigned long int match_limit; /* As it says */
977 unsigned long int match_limit_recursion; /* As it says */
978 int *offset_vector; /* Offset vector */
979 int offset_end; /* One past the end */
980 int offset_max; /* The maximum usable for return data */
981 int nltype; /* Newline type */
982 int nllen; /* Newline string length */
983 uschar nl[4]; /* Newline string when fixed */
984 const uschar *lcc; /* Points to lower casing table */
985 const uschar *ctypes; /* Points to table of type maps */
986 BOOL offset_overflow; /* Set if too many extractions */
987 BOOL notbol; /* NOTBOL flag */
988 BOOL noteol; /* NOTEOL flag */
989 BOOL utf8; /* UTF8 flag */
990 BOOL endonly; /* Dollar not before final \n */
991 BOOL notempty; /* Empty string match not wanted */
992 BOOL partial; /* PARTIAL flag */
993 BOOL hitend; /* Hit the end of the subject at some point */
994 const uschar *start_code; /* For use when recursing */
995 USPTR start_subject; /* Start of the subject string */
996 USPTR end_subject; /* End of the subject string */
997 USPTR start_match; /* Start of this match attempt */
998 USPTR end_match_ptr; /* Subject position at end match */
999 int end_offset_top; /* Highwater mark at end of match */
1000 int capture_last; /* Most recent capture number */
1001 int start_offset; /* The start offset value */
1002 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
1003 int eptrn; /* Next free eptrblock */
1004 recursion_info *recursive; /* Linked list of recursion data */
1005 void *callout_data; /* To pass back to callouts */
1006 struct heapframe *thisframe; /* Used only when compiling for no recursion */
1007 } match_data;
1008
1009 /* A similar structure is used for the same purpose by the DFA matching
1010 functions. */
1011
1012 typedef struct dfa_match_data {
1013 const uschar *start_code; /* Start of the compiled pattern */
1014 const uschar *start_subject; /* Start of the subject string */
1015 const uschar *end_subject; /* End of subject string */
1016 const uschar *tables; /* Character tables */
1017 int moptions; /* Match options */
1018 int poptions; /* Pattern options */
1019 int nltype; /* Newline type */
1020 int nllen; /* Newline string length */
1021 uschar nl[4]; /* Newline string when fixed */
1022 void *callout_data; /* To pass back to callouts */
1023 } dfa_match_data;
1024
1025 /* Bit definitions for entries in the pcre_ctypes table. */
1026
1027 #define ctype_space 0x01
1028 #define ctype_letter 0x02
1029 #define ctype_digit 0x04
1030 #define ctype_xdigit 0x08
1031 #define ctype_word 0x10 /* alphameric or '_' */
1032 #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
1033
1034 /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
1035 of bits for a class map. Some classes are built by combining these tables. */
1036
1037 #define cbit_space 0 /* [:space:] or \s */
1038 #define cbit_xdigit 32 /* [:xdigit:] */
1039 #define cbit_digit 64 /* [:digit:] or \d */
1040 #define cbit_upper 96 /* [:upper:] */
1041 #define cbit_lower 128 /* [:lower:] */
1042 #define cbit_word 160 /* [:word:] or \w */
1043 #define cbit_graph 192 /* [:graph:] */
1044 #define cbit_print 224 /* [:print:] */
1045 #define cbit_punct 256 /* [:punct:] */
1046 #define cbit_cntrl 288 /* [:cntrl:] */
1047 #define cbit_length 320 /* Length of the cbits table */
1048
1049 /* Offsets of the various tables from the base tables pointer, and
1050 total length. */
1051
1052 #define lcc_offset 0
1053 #define fcc_offset 256
1054 #define cbits_offset 512
1055 #define ctypes_offset (cbits_offset + cbit_length)
1056 #define tables_length (ctypes_offset + 256)
1057
1058 /* Layout of the UCP type table that translates property names into types and
1059 codes. */
1060
1061 typedef struct {
1062 const char *name;
1063 pcre_uint16 type;
1064 pcre_uint16 value;
1065 } ucp_type_table;
1066
1067
1068 /* Internal shared data tables. These are tables that are used by more than one
1069 of the exported public functions. They have to be "external" in the C sense,
1070 but are not part of the PCRE public API. The data for these tables is in the
1071 pcre_tables.c module. */
1072
1073 extern const int _pcre_utf8_table1[];
1074 extern const int _pcre_utf8_table2[];
1075 extern const int _pcre_utf8_table3[];
1076 extern const uschar _pcre_utf8_table4[];
1077
1078 extern const int _pcre_utf8_table1_size;
1079
1080 extern const ucp_type_table _pcre_utt[];
1081 extern const int _pcre_utt_size;
1082
1083 extern const uschar _pcre_default_tables[];
1084
1085 extern const uschar _pcre_OP_lengths[];
1086
1087
1088 /* Internal shared functions. These are functions that are used by more than
1089 one of the exported public functions. They have to be "external" in the C
1090 sense, but are not part of the PCRE public API. */
1091
1092 extern BOOL _pcre_is_newline(const uschar *, int, const uschar *,
1093 int *, BOOL);
1094 extern int _pcre_ord2utf8(int, uschar *);
1095 extern real_pcre *_pcre_try_flipped(const real_pcre *, real_pcre *,
1096 const pcre_study_data *, pcre_study_data *);
1097 extern int _pcre_ucp_findprop(const unsigned int, int *, int *);
1098 extern unsigned int _pcre_ucp_othercase(const unsigned int);
1099 extern int _pcre_valid_utf8(const uschar *, int);
1100 extern BOOL _pcre_was_newline(const uschar *, int, const uschar *,
1101 int *, BOOL);
1102 extern BOOL _pcre_xclass(int, const uschar *);
1103
1104 #endif
1105
1106 /* End of pcre_internal.h */

Properties

Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12