/[pcre]/code/branches/pcre16/pcre_internal.h
ViewVC logotype

Contents of /code/branches/pcre16/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log


Revision 759 - (show annotations) (download)
Mon Nov 21 18:57:34 2011 UTC (2 years, 11 months ago) by zherczeg
File MIME type: text/plain
File size: 85082 byte(s)
remove double pcre_uint8 declaration and fix some string literals
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) 1997-2011 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 PCRE_DEBUG to get debugging output on stdout. */
49
50 #if 0
51 #define PCRE_DEBUG
52 #endif
53
54 /* We do not support both EBCDIC and UTF-8/16 at the same time. The "configure"
55 script prevents both being selected, but not everybody uses "configure". */
56
57 #if defined EBCDIC && (defined SUPPORT_UTF8 || defined SUPPORT_UTF16)
58 #error The use of both EBCDIC and SUPPORT_UTF8/16 is not supported.
59 #endif
60
61 /* If SUPPORT_UCP is defined, SUPPORT_UTF8 must also be defined. The
62 "configure" script ensures this, but not everybody uses "configure". */
63
64 #if defined SUPPORT_UCP && !defined SUPPORT_UTF8
65 #define SUPPORT_UTF8 1
66 #endif
67
68 /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
69 inline, and there are *still* stupid compilers about that don't like indented
70 pre-processor statements, or at least there were when I first wrote this. After
71 all, it had only been about 10 years then...
72
73 It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so
74 be absolutely sure we get our version. */
75
76 #undef DPRINTF
77 #ifdef PCRE_DEBUG
78 #define DPRINTF(p) printf p
79 #else
80 #define DPRINTF(p) /* Nothing */
81 #endif
82
83
84 /* Standard C headers plus the external interface definition. The only time
85 setjmp and stdarg are used is when NO_RECURSE is set. */
86
87 #include <ctype.h>
88 #include <limits.h>
89 #include <stddef.h>
90 #include <stdio.h>
91 #include <stdlib.h>
92 #include <string.h>
93
94 /* When compiling a DLL for Windows, the exported symbols have to be declared
95 using some MS magic. I found some useful information on this web page:
96 http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
97 information there, using __declspec(dllexport) without "extern" we have a
98 definition; with "extern" we have a declaration. The settings here override the
99 setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL,
100 which is all that is needed for applications (they just import the symbols). We
101 use:
102
103 PCRE_EXP_DECL for declarations
104 PCRE_EXP_DEFN for definitions of exported functions
105 PCRE_EXP_DATA_DEFN for definitions of exported variables
106
107 The reason for the two DEFN macros is that in non-Windows environments, one
108 does not want to have "extern" before variable definitions because it leads to
109 compiler warnings. So we distinguish between functions and variables. In
110 Windows, the two should always be the same.
111
112 The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest,
113 which is an application, but needs to import this file in order to "peek" at
114 internals, can #include pcre.h first to get an application's-eye view.
115
116 In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon,
117 special-purpose environments) might want to stick other stuff in front of
118 exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and
119 PCRE_EXP_DATA_DEFN only if they are not already set. */
120
121 #ifndef PCRE_EXP_DECL
122 # ifdef _WIN32
123 # ifndef PCRE_STATIC
124 # define PCRE_EXP_DECL extern __declspec(dllexport)
125 # define PCRE_EXP_DEFN __declspec(dllexport)
126 # define PCRE_EXP_DATA_DEFN __declspec(dllexport)
127 # else
128 # define PCRE_EXP_DECL extern
129 # define PCRE_EXP_DEFN
130 # define PCRE_EXP_DATA_DEFN
131 # endif
132 # else
133 # ifdef __cplusplus
134 # define PCRE_EXP_DECL extern "C"
135 # else
136 # define PCRE_EXP_DECL extern
137 # endif
138 # ifndef PCRE_EXP_DEFN
139 # define PCRE_EXP_DEFN PCRE_EXP_DECL
140 # endif
141 # ifndef PCRE_EXP_DATA_DEFN
142 # define PCRE_EXP_DATA_DEFN
143 # endif
144 # endif
145 #endif
146
147 /* When compiling with the MSVC compiler, it is sometimes necessary to include
148 a "calling convention" before exported function names. (This is secondhand
149 information; I know nothing about MSVC myself). For example, something like
150
151 void __cdecl function(....)
152
153 might be needed. In order so make this easy, all the exported functions have
154 PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
155 set, we ensure here that it has no effect. */
156
157 #ifndef PCRE_CALL_CONVENTION
158 #define PCRE_CALL_CONVENTION
159 #endif
160
161 /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
162 cannot determine these outside the compilation (e.g. by running a program as
163 part of "configure") because PCRE is often cross-compiled for use on other
164 systems. Instead we make use of the maximum sizes that are available at
165 preprocessor time in standard C environments. */
166
167 typedef unsigned char pcre_uint8;
168
169 #if USHRT_MAX == 65535
170 typedef unsigned short pcre_uint16;
171 typedef short pcre_int16;
172 #elif UINT_MAX == 65535
173 typedef unsigned int pcre_uint16;
174 typedef int pcre_int16;
175 #else
176 #error Cannot determine a type for 16-bit unsigned integers
177 #endif
178
179 #if UINT_MAX == 4294967295
180 typedef unsigned int pcre_uint32;
181 typedef int pcre_int32;
182 #elif ULONG_MAX == 4294967295
183 typedef unsigned long int pcre_uint32;
184 typedef long int pcre_int32;
185 #else
186 #error Cannot determine a type for 32-bit unsigned integers
187 #endif
188
189 /* When checking for integer overflow in pcre_compile(), we need to handle
190 large integers. If a 64-bit integer type is available, we can use that.
191 Otherwise we have to cast to double, which of course requires floating point
192 arithmetic. Handle this by defining a macro for the appropriate type. If
193 stdint.h is available, include it; it may define INT64_MAX. Systems that do not
194 have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
195 by "configure". */
196
197 #if HAVE_STDINT_H
198 #include <stdint.h>
199 #elif HAVE_INTTYPES_H
200 #include <inttypes.h>
201 #endif
202
203 #if defined INT64_MAX || defined int64_t
204 #define INT64_OR_DOUBLE int64_t
205 #else
206 #define INT64_OR_DOUBLE double
207 #endif
208
209 /* All character handling must be done as unsigned characters. Otherwise there
210 are problems with top-bit-set characters and functions such as isspace().
211 However, we leave the interface to the outside world as char * or short *,
212 because that should make things easier for callers. We define a short type
213 for the current character representation (either 8 or 16 bit) to save lots
214 of typing. I tried "uchar", but it causes problems on Digital Unix, where
215 it is defined in sys/types, so use "uschar" instead. */
216
217 #ifndef COMPILE_PCRE16
218 typedef unsigned char pcre_uchar;
219 #else
220 #if USHRT_MAX != 65535
221 /* This is a warning message. Change PCRE_SCHAR16 to a 16 bit data type in
222 pcre.h(.in) and disable (comment out) this message. */
223 #error Warning: PCRE_SCHAR16 is not a 16 bit data type.
224 #endif
225 typedef pcre_uint16 uschar;
226 #endif
227
228 /* This is an unsigned int value that no character can ever have. UTF-8
229 characters only go up to 0x7fffffff (though Unicode doesn't go beyond
230 0x0010ffff). */
231
232 #define NOTACHAR 0xffffffff
233
234 /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
235 "any" and "anycrlf" at present). The following macros are used to package up
236 testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
237 modules to indicate in which datablock the parameters exist, and what the
238 start/end of string field names are. */
239
240 #define NLTYPE_FIXED 0 /* Newline is a fixed length string */
241 #define NLTYPE_ANY 1 /* Newline is any Unicode line ending */
242 #define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */
243
244 /* This macro checks for a newline at the given position */
245
246 #define IS_NEWLINE(p) \
247 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
248 ((p) < NLBLOCK->PSEND && \
249 _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\
250 utf8)) \
251 : \
252 ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
253 (p)[0] == NLBLOCK->nl[0] && \
254 (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \
255 ) \
256 )
257
258 /* This macro checks for a newline immediately preceding the given position */
259
260 #define WAS_NEWLINE(p) \
261 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
262 ((p) > NLBLOCK->PSSTART && \
263 _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
264 &(NLBLOCK->nllen), utf8)) \
265 : \
266 ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
267 (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \
268 (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \
269 ) \
270 )
271
272 /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
273 with a custom type. This makes it possible, for example, to allow pcre_exec()
274 to process subject strings that are discontinuous by using a smart pointer
275 class. It must always be possible to inspect all of the subject string in
276 pcre_exec() because of the way it backtracks. Two macros are required in the
277 normal case, for sign-unspecified and unsigned char pointers. The former is
278 used for the external interface and appears in pcre.h, which is why its name
279 must begin with PCRE_. */
280
281 #ifdef CUSTOM_SUBJECT_PTR
282 #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
283 #else
284 #define PCRE_PUCHAR const pcre_uchar *
285
286 /* PCRE_SPTR is defined in pcre.h. */
287 #define USPTR const uschar *
288 #endif
289
290 /* Include the public PCRE header and the definitions of UCP character property
291 values. */
292
293 #include "pcre.h"
294 #include "ucp.h"
295
296 /* When compiling for use with the Virtual Pascal compiler, these functions
297 need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
298 option on the command line. */
299
300 #ifdef VPCOMPAT
301 #define strlen(s) _strlen(s)
302 #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
303 #define memcmp(s,c,n) _memcmp(s,c,n)
304 #define memcpy(d,s,n) _memcpy(d,s,n)
305 #define memmove(d,s,n) _memmove(d,s,n)
306 #define memset(s,c,n) _memset(s,c,n)
307 #else /* VPCOMPAT */
308
309 /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
310 define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
311 is set. Otherwise, include an emulating function for those systems that have
312 neither (there some non-Unix environments where this is the case). */
313
314 #ifndef HAVE_MEMMOVE
315 #undef memmove /* some systems may have a macro */
316 #ifdef HAVE_BCOPY
317 #define memmove(a, b, c) bcopy(b, a, c)
318 #else /* HAVE_BCOPY */
319 static void *
320 pcre_memmove(void *d, const void *s, size_t n)
321 {
322 size_t i;
323 unsigned char *dest = (unsigned char *)d;
324 const unsigned char *src = (const unsigned char *)s;
325 if (dest > src)
326 {
327 dest += n;
328 src += n;
329 for (i = 0; i < n; ++i) *(--dest) = *(--src);
330 return (void *)dest;
331 }
332 else
333 {
334 for (i = 0; i < n; ++i) *dest++ = *src++;
335 return (void *)(dest - n);
336 }
337 }
338 #define memmove(a, b, c) pcre_memmove(a, b, c)
339 #endif /* not HAVE_BCOPY */
340 #endif /* not HAVE_MEMMOVE */
341 #endif /* not VPCOMPAT */
342
343
344 /* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
345 in big-endian order) by default. These are used, for example, to link from the
346 start of a subpattern to its alternatives and its end. The use of 2 bytes per
347 offset limits the size of the compiled regex to around 64K, which is big enough
348 for almost everybody. However, I received a request for an even bigger limit.
349 For this reason, and also to make the code easier to maintain, the storing and
350 loading of offsets from the byte string is now handled by the macros that are
351 defined here.
352
353 The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
354 the config.h file, but can be overridden by using -D on the command line. This
355 is automated on Unix systems via the "configure" command. */
356
357 #if LINK_SIZE == 2
358
359 #define PUT(a,n,d) \
360 (a[n] = (d) >> 8), \
361 (a[(n)+1] = (d) & 255)
362
363 #define GET(a,n) \
364 (((a)[n] << 8) | (a)[(n)+1])
365
366 #define MAX_PATTERN_SIZE (1 << 16)
367
368
369 #elif LINK_SIZE == 3
370
371 #define PUT(a,n,d) \
372 (a[n] = (d) >> 16), \
373 (a[(n)+1] = (d) >> 8), \
374 (a[(n)+2] = (d) & 255)
375
376 #define GET(a,n) \
377 (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
378
379 #define MAX_PATTERN_SIZE (1 << 24)
380
381
382 #elif LINK_SIZE == 4
383
384 #define PUT(a,n,d) \
385 (a[n] = (d) >> 24), \
386 (a[(n)+1] = (d) >> 16), \
387 (a[(n)+2] = (d) >> 8), \
388 (a[(n)+3] = (d) & 255)
389
390 #define GET(a,n) \
391 (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
392
393 #define MAX_PATTERN_SIZE (1 << 30) /* Keep it positive */
394
395
396 #else
397 #error LINK_SIZE must be either 2, 3, or 4
398 #endif
399
400
401 /* Convenience macro defined in terms of the others */
402
403 #define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
404
405
406 /* PCRE uses some other 2-byte quantities that do not change when the size of
407 offsets changes. There are used for repeat counts and for other things such as
408 capturing parenthesis numbers in back references. */
409
410 #define PUT2(a,n,d) \
411 a[n] = (d) >> 8; \
412 a[(n)+1] = (d) & 255
413
414 #define GET2(a,n) \
415 (((a)[n] << 8) | (a)[(n)+1])
416
417 #define PUT2INC(a,n,d) PUT2(a,n,d), a += 2
418
419
420 /* When UTF-8 encoding is being used, a character is no longer just a single
421 byte. The macros for character handling generate simple sequences when used in
422 byte-mode, and more complicated ones for UTF-8 characters. GETCHARLENTEST is
423 not used when UTF-8 is not supported, so it is not defined, and BACKCHAR should
424 never be called in byte mode. To make sure they can never even appear when
425 UTF-8 support is omitted, we don't even define them. */
426
427 #ifndef SUPPORT_UTF8
428 #define GETCHAR(c, eptr) c = *eptr;
429 #define GETCHARTEST(c, eptr) c = *eptr;
430 #define GETCHARINC(c, eptr) c = *eptr++;
431 #define GETCHARINCTEST(c, eptr) c = *eptr++;
432 #define GETCHARLEN(c, eptr, len) c = *eptr;
433 /* #define GETCHARLENTEST(c, eptr, len) */
434 /* #define BACKCHAR(eptr) */
435
436 #else /* SUPPORT_UTF8 */
437
438 /* These macros were originally written in the form of loops that used data
439 from the tables whose names start with _pcre_utf8_table. They were rewritten by
440 a user so as not to use loops, because in some environments this gives a
441 significant performance advantage, and it seems never to do any harm. */
442
443 /* Base macro to pick up the remaining bytes of a UTF-8 character, not
444 advancing the pointer. */
445
446 #define GETUTF8(c, eptr) \
447 { \
448 if ((c & 0x20) == 0) \
449 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
450 else if ((c & 0x10) == 0) \
451 c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
452 else if ((c & 0x08) == 0) \
453 c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
454 ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
455 else if ((c & 0x04) == 0) \
456 c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
457 ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
458 (eptr[4] & 0x3f); \
459 else \
460 c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
461 ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
462 ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
463 }
464
465 /* Get the next UTF-8 character, not advancing the pointer. This is called when
466 we know we are in UTF-8 mode. */
467
468 #define GETCHAR(c, eptr) \
469 c = *eptr; \
470 if (c >= 0xc0) GETUTF8(c, eptr);
471
472 /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
473 pointer. */
474
475 #define GETCHARTEST(c, eptr) \
476 c = *eptr; \
477 if (utf8 && c >= 0xc0) GETUTF8(c, eptr);
478
479 /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
480 the pointer. */
481
482 #define GETUTF8INC(c, eptr) \
483 { \
484 if ((c & 0x20) == 0) \
485 c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
486 else if ((c & 0x10) == 0) \
487 { \
488 c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
489 eptr += 2; \
490 } \
491 else if ((c & 0x08) == 0) \
492 { \
493 c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
494 ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
495 eptr += 3; \
496 } \
497 else if ((c & 0x04) == 0) \
498 { \
499 c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
500 ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
501 (eptr[3] & 0x3f); \
502 eptr += 4; \
503 } \
504 else \
505 { \
506 c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
507 ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
508 ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
509 eptr += 5; \
510 } \
511 }
512
513 /* Get the next UTF-8 character, advancing the pointer. This is called when we
514 know we are in UTF-8 mode. */
515
516 #define GETCHARINC(c, eptr) \
517 c = *eptr++; \
518 if (c >= 0xc0) GETUTF8INC(c, eptr);
519
520 /* Get the next character, testing for UTF-8 mode, and advancing the pointer.
521 This is called when we don't know if we are in UTF-8 mode. */
522
523 #define GETCHARINCTEST(c, eptr) \
524 c = *eptr++; \
525 if (utf8 && c >= 0xc0) GETUTF8INC(c, eptr);
526
527 /* Base macro to pick up the remaining bytes of a UTF-8 character, not
528 advancing the pointer, incrementing the length. */
529
530 #define GETUTF8LEN(c, eptr, len) \
531 { \
532 if ((c & 0x20) == 0) \
533 { \
534 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
535 len++; \
536 } \
537 else if ((c & 0x10) == 0) \
538 { \
539 c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
540 len += 2; \
541 } \
542 else if ((c & 0x08) == 0) \
543 {\
544 c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
545 ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
546 len += 3; \
547 } \
548 else if ((c & 0x04) == 0) \
549 { \
550 c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
551 ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
552 (eptr[4] & 0x3f); \
553 len += 4; \
554 } \
555 else \
556 {\
557 c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
558 ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
559 ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
560 len += 5; \
561 } \
562 }
563
564 /* Get the next UTF-8 character, not advancing the pointer, incrementing length
565 if there are extra bytes. This is called when we know we are in UTF-8 mode. */
566
567 #define GETCHARLEN(c, eptr, len) \
568 c = *eptr; \
569 if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
570
571 /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
572 pointer, incrementing length if there are extra bytes. This is called when we
573 do not know if we are in UTF-8 mode. */
574
575 #define GETCHARLENTEST(c, eptr, len) \
576 c = *eptr; \
577 if (utf8 && c >= 0xc0) GETUTF8LEN(c, eptr, len);
578
579 /* If the pointer is not at the start of a character, move it back until
580 it is. This is called only in UTF-8 mode - we don't put a test within the macro
581 because almost all calls are already within a block of UTF-8 only code. */
582
583 #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
584
585 #endif /* SUPPORT_UTF8 */
586
587
588 /* In case there is no definition of offsetof() provided - though any proper
589 Standard C system should have one. */
590
591 #ifndef offsetof
592 #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))
593 #endif
594
595
596 /* Private flags containing information about the compiled regex. They used to
597 live at the top end of the options word, but that got almost full, so now they
598 are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
599 the restrictions on partial matching have been lifted. It remains for backwards
600 compatibility. */
601
602 #define PCRE_NOPARTIAL 0x0001 /* can't use partial with this regex */
603 #define PCRE_FIRSTSET 0x0002 /* first_byte is set */
604 #define PCRE_REQCHSET 0x0004 /* req_byte is set */
605 #define PCRE_STARTLINE 0x0008 /* start after \n for multiline */
606 #define PCRE_JCHANGED 0x0010 /* j option used in regex */
607 #define PCRE_HASCRORLF 0x0020 /* explicit \r or \n in pattern */
608 #define PCRE_HASTHEN 0x0040 /* pattern contains (*THEN) */
609
610 /* Flags for the "extra" block produced by pcre_study(). */
611
612 #define PCRE_STUDY_MAPPED 0x0001 /* a map of starting chars exists */
613 #define PCRE_STUDY_MINLEN 0x0002 /* a minimum length field exists */
614
615 /* Masks for identifying the public options that are permitted at compile
616 time, run time, or study time, respectively. */
617
618 #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
619 PCRE_NEWLINE_ANYCRLF)
620
621 #define PUBLIC_COMPILE_OPTIONS \
622 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
623 PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
624 PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
625 PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
626 PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
627
628 #define PUBLIC_EXEC_OPTIONS \
629 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
630 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
631 PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
632
633 #define PUBLIC_DFA_EXEC_OPTIONS \
634 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
635 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
636 PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
637 PCRE_NO_START_OPTIMIZE)
638
639 #define PUBLIC_STUDY_OPTIONS \
640 PCRE_STUDY_JIT_COMPILE
641
642 /* Magic number to provide a small check against being handed junk. Also used
643 to detect whether a pattern was compiled on a host of different endianness. */
644
645 #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
646
647 /* Negative values for the firstchar and reqchar variables */
648
649 #define REQ_UNSET (-2)
650 #define REQ_NONE (-1)
651
652 /* The maximum remaining length of subject we are prepared to search for a
653 req_byte match. */
654
655 #define REQ_BYTE_MAX 1000
656
657 /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a
658 variable-length repeat, or a anything other than literal characters. */
659
660 #define REQ_CASELESS 0x0100 /* indicates caselessness */
661 #define REQ_VARY 0x0200 /* reqbyte followed non-literal item */
662
663 /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
664 environments where these macros are defined elsewhere. Unfortunately, there
665 is no way to do the same for the typedef. */
666
667 typedef int BOOL;
668
669 #ifndef FALSE
670 #define FALSE 0
671 #define TRUE 1
672 #endif
673
674 /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
675 character constants like '*' because the compiler would emit their EBCDIC code,
676 which is different from their ASCII/UTF-8 code. Instead we define macros for
677 the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
678 is enabled. When UTF-8 support is not enabled, the definitions use character
679 literals. Both character and string versions of each character are needed, and
680 there are some longer strings as well.
681
682 This means that, on EBCDIC platforms, the PCRE library can handle either
683 EBCDIC, or UTF-8, but not both. To support both in the same compiled library
684 would need different lookups depending on whether PCRE_UTF8 was set or not.
685 This would make it impossible to use characters in switch/case statements,
686 which would reduce performance. For a theoretical use (which nobody has asked
687 for) in a minority area (EBCDIC platforms), this is not sensible. Any
688 application that did need both could compile two versions of the library, using
689 macros to give the functions distinct names. */
690
691 #ifndef SUPPORT_UTF8
692
693 /* UTF-8 support is not enabled; use the platform-dependent character literals
694 so that PCRE works on both ASCII and EBCDIC platforms, in non-UTF-mode only. */
695
696 #define CHAR_HT '\t'
697 #define CHAR_VT '\v'
698 #define CHAR_FF '\f'
699 #define CHAR_CR '\r'
700 #define CHAR_NL '\n'
701 #define CHAR_BS '\b'
702 #define CHAR_BEL '\a'
703 #ifdef EBCDIC
704 #define CHAR_ESC '\047'
705 #define CHAR_DEL '\007'
706 #else
707 #define CHAR_ESC '\033'
708 #define CHAR_DEL '\177'
709 #endif
710
711 #define CHAR_SPACE ' '
712 #define CHAR_EXCLAMATION_MARK '!'
713 #define CHAR_QUOTATION_MARK '"'
714 #define CHAR_NUMBER_SIGN '#'
715 #define CHAR_DOLLAR_SIGN '$'
716 #define CHAR_PERCENT_SIGN '%'
717 #define CHAR_AMPERSAND '&'
718 #define CHAR_APOSTROPHE '\''
719 #define CHAR_LEFT_PARENTHESIS '('
720 #define CHAR_RIGHT_PARENTHESIS ')'
721 #define CHAR_ASTERISK '*'
722 #define CHAR_PLUS '+'
723 #define CHAR_COMMA ','
724 #define CHAR_MINUS '-'
725 #define CHAR_DOT '.'
726 #define CHAR_SLASH '/'
727 #define CHAR_0 '0'
728 #define CHAR_1 '1'
729 #define CHAR_2 '2'
730 #define CHAR_3 '3'
731 #define CHAR_4 '4'
732 #define CHAR_5 '5'
733 #define CHAR_6 '6'
734 #define CHAR_7 '7'
735 #define CHAR_8 '8'
736 #define CHAR_9 '9'
737 #define CHAR_COLON ':'
738 #define CHAR_SEMICOLON ';'
739 #define CHAR_LESS_THAN_SIGN '<'
740 #define CHAR_EQUALS_SIGN '='
741 #define CHAR_GREATER_THAN_SIGN '>'
742 #define CHAR_QUESTION_MARK '?'
743 #define CHAR_COMMERCIAL_AT '@'
744 #define CHAR_A 'A'
745 #define CHAR_B 'B'
746 #define CHAR_C 'C'
747 #define CHAR_D 'D'
748 #define CHAR_E 'E'
749 #define CHAR_F 'F'
750 #define CHAR_G 'G'
751 #define CHAR_H 'H'
752 #define CHAR_I 'I'
753 #define CHAR_J 'J'
754 #define CHAR_K 'K'
755 #define CHAR_L 'L'
756 #define CHAR_M 'M'
757 #define CHAR_N 'N'
758 #define CHAR_O 'O'
759 #define CHAR_P 'P'
760 #define CHAR_Q 'Q'
761 #define CHAR_R 'R'
762 #define CHAR_S 'S'
763 #define CHAR_T 'T'
764 #define CHAR_U 'U'
765 #define CHAR_V 'V'
766 #define CHAR_W 'W'
767 #define CHAR_X 'X'
768 #define CHAR_Y 'Y'
769 #define CHAR_Z 'Z'
770 #define CHAR_LEFT_SQUARE_BRACKET '['
771 #define CHAR_BACKSLASH '\\'
772 #define CHAR_RIGHT_SQUARE_BRACKET ']'
773 #define CHAR_CIRCUMFLEX_ACCENT '^'
774 #define CHAR_UNDERSCORE '_'
775 #define CHAR_GRAVE_ACCENT '`'
776 #define CHAR_a 'a'
777 #define CHAR_b 'b'
778 #define CHAR_c 'c'
779 #define CHAR_d 'd'
780 #define CHAR_e 'e'
781 #define CHAR_f 'f'
782 #define CHAR_g 'g'
783 #define CHAR_h 'h'
784 #define CHAR_i 'i'
785 #define CHAR_j 'j'
786 #define CHAR_k 'k'
787 #define CHAR_l 'l'
788 #define CHAR_m 'm'
789 #define CHAR_n 'n'
790 #define CHAR_o 'o'
791 #define CHAR_p 'p'
792 #define CHAR_q 'q'
793 #define CHAR_r 'r'
794 #define CHAR_s 's'
795 #define CHAR_t 't'
796 #define CHAR_u 'u'
797 #define CHAR_v 'v'
798 #define CHAR_w 'w'
799 #define CHAR_x 'x'
800 #define CHAR_y 'y'
801 #define CHAR_z 'z'
802 #define CHAR_LEFT_CURLY_BRACKET '{'
803 #define CHAR_VERTICAL_LINE '|'
804 #define CHAR_RIGHT_CURLY_BRACKET '}'
805 #define CHAR_TILDE '~'
806
807 #define STR_HT "\t"
808 #define STR_VT "\v"
809 #define STR_FF "\f"
810 #define STR_CR "\r"
811 #define STR_NL "\n"
812 #define STR_BS "\b"
813 #define STR_BEL "\a"
814 #ifdef EBCDIC
815 #define STR_ESC "\047"
816 #define STR_DEL "\007"
817 #else
818 #define STR_ESC "\033"
819 #define STR_DEL "\177"
820 #endif
821
822 #define STR_SPACE " "
823 #define STR_EXCLAMATION_MARK "!"
824 #define STR_QUOTATION_MARK "\""
825 #define STR_NUMBER_SIGN "#"
826 #define STR_DOLLAR_SIGN "$"
827 #define STR_PERCENT_SIGN "%"
828 #define STR_AMPERSAND "&"
829 #define STR_APOSTROPHE "'"
830 #define STR_LEFT_PARENTHESIS "("
831 #define STR_RIGHT_PARENTHESIS ")"
832 #define STR_ASTERISK "*"
833 #define STR_PLUS "+"
834 #define STR_COMMA ","
835 #define STR_MINUS "-"
836 #define STR_DOT "."
837 #define STR_SLASH "/"
838 #define STR_0 "0"
839 #define STR_1 "1"
840 #define STR_2 "2"
841 #define STR_3 "3"
842 #define STR_4 "4"
843 #define STR_5 "5"
844 #define STR_6 "6"
845 #define STR_7 "7"
846 #define STR_8 "8"
847 #define STR_9 "9"
848 #define STR_COLON ":"
849 #define STR_SEMICOLON ";"
850 #define STR_LESS_THAN_SIGN "<"
851 #define STR_EQUALS_SIGN "="
852 #define STR_GREATER_THAN_SIGN ">"
853 #define STR_QUESTION_MARK "?"
854 #define STR_COMMERCIAL_AT "@"
855 #define STR_A "A"
856 #define STR_B "B"
857 #define STR_C "C"
858 #define STR_D "D"
859 #define STR_E "E"
860 #define STR_F "F"
861 #define STR_G "G"
862 #define STR_H "H"
863 #define STR_I "I"
864 #define STR_J "J"
865 #define STR_K "K"
866 #define STR_L "L"
867 #define STR_M "M"
868 #define STR_N "N"
869 #define STR_O "O"
870 #define STR_P "P"
871 #define STR_Q "Q"
872 #define STR_R "R"
873 #define STR_S "S"
874 #define STR_T "T"
875 #define STR_U "U"
876 #define STR_V "V"
877 #define STR_W "W"
878 #define STR_X "X"
879 #define STR_Y "Y"
880 #define STR_Z "Z"
881 #define STR_LEFT_SQUARE_BRACKET "["
882 #define STR_BACKSLASH "\\"
883 #define STR_RIGHT_SQUARE_BRACKET "]"
884 #define STR_CIRCUMFLEX_ACCENT "^"
885 #define STR_UNDERSCORE "_"
886 #define STR_GRAVE_ACCENT "`"
887 #define STR_a "a"
888 #define STR_b "b"
889 #define STR_c "c"
890 #define STR_d "d"
891 #define STR_e "e"
892 #define STR_f "f"
893 #define STR_g "g"
894 #define STR_h "h"
895 #define STR_i "i"
896 #define STR_j "j"
897 #define STR_k "k"
898 #define STR_l "l"
899 #define STR_m "m"
900 #define STR_n "n"
901 #define STR_o "o"
902 #define STR_p "p"
903 #define STR_q "q"
904 #define STR_r "r"
905 #define STR_s "s"
906 #define STR_t "t"
907 #define STR_u "u"
908 #define STR_v "v"
909 #define STR_w "w"
910 #define STR_x "x"
911 #define STR_y "y"
912 #define STR_z "z"
913 #define STR_LEFT_CURLY_BRACKET "{"
914 #define STR_VERTICAL_LINE "|"
915 #define STR_RIGHT_CURLY_BRACKET "}"
916 #define STR_TILDE "~"
917
918 #define STRING_ACCEPT0 "ACCEPT\0"
919 #define STRING_COMMIT0 "COMMIT\0"
920 #define STRING_F0 "F\0"
921 #define STRING_FAIL0 "FAIL\0"
922 #define STRING_MARK0 "MARK\0"
923 #define STRING_PRUNE0 "PRUNE\0"
924 #define STRING_SKIP0 "SKIP\0"
925 #define STRING_THEN "THEN"
926
927 #define STRING_alpha0 "alpha\0"
928 #define STRING_lower0 "lower\0"
929 #define STRING_upper0 "upper\0"
930 #define STRING_alnum0 "alnum\0"
931 #define STRING_ascii0 "ascii\0"
932 #define STRING_blank0 "blank\0"
933 #define STRING_cntrl0 "cntrl\0"
934 #define STRING_digit0 "digit\0"
935 #define STRING_graph0 "graph\0"
936 #define STRING_print0 "print\0"
937 #define STRING_punct0 "punct\0"
938 #define STRING_space0 "space\0"
939 #define STRING_word0 "word\0"
940 #define STRING_xdigit "xdigit"
941
942 #define STRING_DEFINE "DEFINE"
943
944 #define STRING_CR_RIGHTPAR "CR)"
945 #define STRING_LF_RIGHTPAR "LF)"
946 #define STRING_CRLF_RIGHTPAR "CRLF)"
947 #define STRING_ANY_RIGHTPAR "ANY)"
948 #define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)"
949 #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"
950 #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"
951 #define STRING_UTF8_RIGHTPAR "UTF8)"
952 #define STRING_UCP_RIGHTPAR "UCP)"
953 #define STRING_NO_START_OPT_RIGHTPAR "NO_START_OPT)"
954
955 #else /* SUPPORT_UTF8 */
956
957 /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
958 works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
959 only. */
960
961 #define CHAR_HT '\011'
962 #define CHAR_VT '\013'
963 #define CHAR_FF '\014'
964 #define CHAR_CR '\015'
965 #define CHAR_NL '\012'
966 #define CHAR_BS '\010'
967 #define CHAR_BEL '\007'
968 #define CHAR_ESC '\033'
969 #define CHAR_DEL '\177'
970
971 #define CHAR_SPACE '\040'
972 #define CHAR_EXCLAMATION_MARK '\041'
973 #define CHAR_QUOTATION_MARK '\042'
974 #define CHAR_NUMBER_SIGN '\043'
975 #define CHAR_DOLLAR_SIGN '\044'
976 #define CHAR_PERCENT_SIGN '\045'
977 #define CHAR_AMPERSAND '\046'
978 #define CHAR_APOSTROPHE '\047'
979 #define CHAR_LEFT_PARENTHESIS '\050'
980 #define CHAR_RIGHT_PARENTHESIS '\051'
981 #define CHAR_ASTERISK '\052'
982 #define CHAR_PLUS '\053'
983 #define CHAR_COMMA '\054'
984 #define CHAR_MINUS '\055'
985 #define CHAR_DOT '\056'
986 #define CHAR_SLASH '\057'
987 #define CHAR_0 '\060'
988 #define CHAR_1 '\061'
989 #define CHAR_2 '\062'
990 #define CHAR_3 '\063'
991 #define CHAR_4 '\064'
992 #define CHAR_5 '\065'
993 #define CHAR_6 '\066'
994 #define CHAR_7 '\067'
995 #define CHAR_8 '\070'
996 #define CHAR_9 '\071'
997 #define CHAR_COLON '\072'
998 #define CHAR_SEMICOLON '\073'
999 #define CHAR_LESS_THAN_SIGN '\074'
1000 #define CHAR_EQUALS_SIGN '\075'
1001 #define CHAR_GREATER_THAN_SIGN '\076'
1002 #define CHAR_QUESTION_MARK '\077'
1003 #define CHAR_COMMERCIAL_AT '\100'
1004 #define CHAR_A '\101'
1005 #define CHAR_B '\102'
1006 #define CHAR_C '\103'
1007 #define CHAR_D '\104'
1008 #define CHAR_E '\105'
1009 #define CHAR_F '\106'
1010 #define CHAR_G '\107'
1011 #define CHAR_H '\110'
1012 #define CHAR_I '\111'
1013 #define CHAR_J '\112'
1014 #define CHAR_K '\113'
1015 #define CHAR_L '\114'
1016 #define CHAR_M '\115'
1017 #define CHAR_N '\116'
1018 #define CHAR_O '\117'
1019 #define CHAR_P '\120'
1020 #define CHAR_Q '\121'
1021 #define CHAR_R '\122'
1022 #define CHAR_S '\123'
1023 #define CHAR_T '\124'
1024 #define CHAR_U '\125'
1025 #define CHAR_V '\126'
1026 #define CHAR_W '\127'
1027 #define CHAR_X '\130'
1028 #define CHAR_Y '\131'
1029 #define CHAR_Z '\132'
1030 #define CHAR_LEFT_SQUARE_BRACKET '\133'
1031 #define CHAR_BACKSLASH '\134'
1032 #define CHAR_RIGHT_SQUARE_BRACKET '\135'
1033 #define CHAR_CIRCUMFLEX_ACCENT '\136'
1034 #define CHAR_UNDERSCORE '\137'
1035 #define CHAR_GRAVE_ACCENT '\140'
1036 #define CHAR_a '\141'
1037 #define CHAR_b '\142'
1038 #define CHAR_c '\143'
1039 #define CHAR_d '\144'
1040 #define CHAR_e '\145'
1041 #define CHAR_f '\146'
1042 #define CHAR_g '\147'
1043 #define CHAR_h '\150'
1044 #define CHAR_i '\151'
1045 #define CHAR_j '\152'
1046 #define CHAR_k '\153'
1047 #define CHAR_l '\154'
1048 #define CHAR_m '\155'
1049 #define CHAR_n '\156'
1050 #define CHAR_o '\157'
1051 #define CHAR_p '\160'
1052 #define CHAR_q '\161'
1053 #define CHAR_r '\162'
1054 #define CHAR_s '\163'
1055 #define CHAR_t '\164'
1056 #define CHAR_u '\165'
1057 #define CHAR_v '\166'
1058 #define CHAR_w '\167'
1059 #define CHAR_x '\170'
1060 #define CHAR_y '\171'
1061 #define CHAR_z '\172'
1062 #define CHAR_LEFT_CURLY_BRACKET '\173'
1063 #define CHAR_VERTICAL_LINE '\174'
1064 #define CHAR_RIGHT_CURLY_BRACKET '\175'
1065 #define CHAR_TILDE '\176'
1066
1067 #define STR_HT "\011"
1068 #define STR_VT "\013"
1069 #define STR_FF "\014"
1070 #define STR_CR "\015"
1071 #define STR_NL "\012"
1072 #define STR_BS "\010"
1073 #define STR_BEL "\007"
1074 #define STR_ESC "\033"
1075 #define STR_DEL "\177"
1076
1077 #define STR_SPACE "\040"
1078 #define STR_EXCLAMATION_MARK "\041"
1079 #define STR_QUOTATION_MARK "\042"
1080 #define STR_NUMBER_SIGN "\043"
1081 #define STR_DOLLAR_SIGN "\044"
1082 #define STR_PERCENT_SIGN "\045"
1083 #define STR_AMPERSAND "\046"
1084 #define STR_APOSTROPHE "\047"
1085 #define STR_LEFT_PARENTHESIS "\050"
1086 #define STR_RIGHT_PARENTHESIS "\051"
1087 #define STR_ASTERISK "\052"
1088 #define STR_PLUS "\053"
1089 #define STR_COMMA "\054"
1090 #define STR_MINUS "\055"
1091 #define STR_DOT "\056"
1092 #define STR_SLASH "\057"
1093 #define STR_0 "\060"
1094 #define STR_1 "\061"
1095 #define STR_2 "\062"
1096 #define STR_3 "\063"
1097 #define STR_4 "\064"
1098 #define STR_5 "\065"
1099 #define STR_6 "\066"
1100 #define STR_7 "\067"
1101 #define STR_8 "\070"
1102 #define STR_9 "\071"
1103 #define STR_COLON "\072"
1104 #define STR_SEMICOLON "\073"
1105 #define STR_LESS_THAN_SIGN "\074"
1106 #define STR_EQUALS_SIGN "\075"
1107 #define STR_GREATER_THAN_SIGN "\076"
1108 #define STR_QUESTION_MARK "\077"
1109 #define STR_COMMERCIAL_AT "\100"
1110 #define STR_A "\101"
1111 #define STR_B "\102"
1112 #define STR_C "\103"
1113 #define STR_D "\104"
1114 #define STR_E "\105"
1115 #define STR_F "\106"
1116 #define STR_G "\107"
1117 #define STR_H "\110"
1118 #define STR_I "\111"
1119 #define STR_J "\112"
1120 #define STR_K "\113"
1121 #define STR_L "\114"
1122 #define STR_M "\115"
1123 #define STR_N "\116"
1124 #define STR_O "\117"
1125 #define STR_P "\120"
1126 #define STR_Q "\121"
1127 #define STR_R "\122"
1128 #define STR_S "\123"
1129 #define STR_T "\124"
1130 #define STR_U "\125"
1131 #define STR_V "\126"
1132 #define STR_W "\127"
1133 #define STR_X "\130"
1134 #define STR_Y "\131"
1135 #define STR_Z "\132"
1136 #define STR_LEFT_SQUARE_BRACKET "\133"
1137 #define STR_BACKSLASH "\134"
1138 #define STR_RIGHT_SQUARE_BRACKET "\135"
1139 #define STR_CIRCUMFLEX_ACCENT "\136"
1140 #define STR_UNDERSCORE "\137"
1141 #define STR_GRAVE_ACCENT "\140"
1142 #define STR_a "\141"
1143 #define STR_b "\142"
1144 #define STR_c "\143"
1145 #define STR_d "\144"
1146 #define STR_e "\145"
1147 #define STR_f "\146"
1148 #define STR_g "\147"
1149 #define STR_h "\150"
1150 #define STR_i "\151"
1151 #define STR_j "\152"
1152 #define STR_k "\153"
1153 #define STR_l "\154"
1154 #define STR_m "\155"
1155 #define STR_n "\156"
1156 #define STR_o "\157"
1157 #define STR_p "\160"
1158 #define STR_q "\161"
1159 #define STR_r "\162"
1160 #define STR_s "\163"
1161 #define STR_t "\164"
1162 #define STR_u "\165"
1163 #define STR_v "\166"
1164 #define STR_w "\167"
1165 #define STR_x "\170"
1166 #define STR_y "\171"
1167 #define STR_z "\172"
1168 #define STR_LEFT_CURLY_BRACKET "\173"
1169 #define STR_VERTICAL_LINE "\174"
1170 #define STR_RIGHT_CURLY_BRACKET "\175"
1171 #define STR_TILDE "\176"
1172
1173 #define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1174 #define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1175 #define STRING_F0 STR_F "\0"
1176 #define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0"
1177 #define STRING_MARK0 STR_M STR_A STR_R STR_K "\0"
1178 #define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0"
1179 #define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0"
1180 #define STRING_THEN STR_T STR_H STR_E STR_N
1181
1182 #define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0"
1183 #define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0"
1184 #define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0"
1185 #define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0"
1186 #define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0"
1187 #define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0"
1188 #define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0"
1189 #define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0"
1190 #define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0"
1191 #define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0"
1192 #define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0"
1193 #define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0"
1194 #define STRING_word0 STR_w STR_o STR_r STR_d "\0"
1195 #define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t
1196
1197 #define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E
1198
1199 #define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS
1200 #define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS
1201 #define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1202 #define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1203 #define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1204 #define STRING_BSR_ANYCRLF_RIGHTPAR STR_B STR_S STR_R STR_UNDERSCORE STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1205 #define STRING_BSR_UNICODE_RIGHTPAR STR_B STR_S STR_R STR_UNDERSCORE STR_U STR_N STR_I STR_C STR_O STR_D STR_E STR_RIGHT_PARENTHESIS
1206 #define STRING_UTF8_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1207 #define STRING_UCP_RIGHTPAR STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1208 #define STRING_NO_START_OPT_RIGHTPAR STR_N STR_O STR_UNDERSCORE STR_S STR_T STR_A STR_R STR_T STR_UNDERSCORE STR_O STR_P STR_T STR_RIGHT_PARENTHESIS
1209
1210 #endif /* SUPPORT_UTF8 */
1211
1212 /* Escape items that are just an encoding of a particular data value. */
1213
1214 #ifndef ESC_e
1215 #define ESC_e CHAR_ESC
1216 #endif
1217
1218 #ifndef ESC_f
1219 #define ESC_f CHAR_FF
1220 #endif
1221
1222 #ifndef ESC_n
1223 #define ESC_n CHAR_NL
1224 #endif
1225
1226 #ifndef ESC_r
1227 #define ESC_r CHAR_CR
1228 #endif
1229
1230 /* We can't officially use ESC_t because it is a POSIX reserved identifier
1231 (presumably because of all the others like size_t). */
1232
1233 #ifndef ESC_tee
1234 #define ESC_tee CHAR_HT
1235 #endif
1236
1237 /* Codes for different types of Unicode property */
1238
1239 #define PT_ANY 0 /* Any property - matches all chars */
1240 #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
1241 #define PT_GC 2 /* Specified general characteristic (e.g. L) */
1242 #define PT_PC 3 /* Specified particular characteristic (e.g. Lu) */
1243 #define PT_SC 4 /* Script (e.g. Han) */
1244 #define PT_ALNUM 5 /* Alphanumeric - the union of L and N */
1245 #define PT_SPACE 6 /* Perl space - Z plus 9,10,12,13 */
1246 #define PT_PXSPACE 7 /* POSIX space - Z plus 9,10,11,12,13 */
1247 #define PT_WORD 8 /* Word - L plus N plus underscore */
1248
1249 /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1250 contain UTF-8 characters with values greater than 255. */
1251
1252 #define XCL_NOT 0x01 /* Flag: this is a negative class */
1253 #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
1254
1255 #define XCL_END 0 /* Marks end of individual items */
1256 #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
1257 #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
1258 #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
1259 #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
1260
1261 /* These are escaped items that aren't just an encoding of a particular data
1262 value such as \n. They must have non-zero values, as check_escape() returns
1263 their negation. Also, they must appear in the same order as in the opcode
1264 definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1265 corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1266 used for [^] in JavaScript compatibility mode, and for \C in non-utf8 mode. In
1267 non-DOTALL mode, "." behaves like \N.
1268
1269 The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1270 when PCRE_UCP is set, when replacement of \d etc by \p sequences is required.
1271 They must be contiguous, and remain in order so that the replacements can be
1272 looked up from a table.
1273
1274 The final escape must be ESC_REF as subsequent values are used for
1275 backreferences (\1, \2, \3, etc). There are two tests in the code for an escape
1276 greater than ESC_b and less than ESC_Z to detect the types that may be
1277 repeated. These are the types that consume characters. If any new escapes are
1278 put in between that don't consume a character, that code will have to change.
1279 */
1280
1281 enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1282 ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1283 ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1284 ESC_E, ESC_Q, ESC_g, ESC_k,
1285 ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu,
1286 ESC_REF };
1287
1288 /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1289 OP_EOD must correspond in order to the list of escapes immediately above.
1290
1291 *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1292 that follow must also be updated to match. There are also tables called
1293 "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1294
1295 enum {
1296 OP_END, /* 0 End of pattern */
1297
1298 /* Values corresponding to backslashed metacharacters */
1299
1300 OP_SOD, /* 1 Start of data: \A */
1301 OP_SOM, /* 2 Start of match (subject + offset): \G */
1302 OP_SET_SOM, /* 3 Set start of match (\K) */
1303 OP_NOT_WORD_BOUNDARY, /* 4 \B */
1304 OP_WORD_BOUNDARY, /* 5 \b */
1305 OP_NOT_DIGIT, /* 6 \D */
1306 OP_DIGIT, /* 7 \d */
1307 OP_NOT_WHITESPACE, /* 8 \S */
1308 OP_WHITESPACE, /* 9 \s */
1309 OP_NOT_WORDCHAR, /* 10 \W */
1310 OP_WORDCHAR, /* 11 \w */
1311
1312 OP_ANY, /* 12 Match any character except newline */
1313 OP_ALLANY, /* 13 Match any character */
1314 OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1315 OP_NOTPROP, /* 15 \P (not Unicode property) */
1316 OP_PROP, /* 16 \p (Unicode property) */
1317 OP_ANYNL, /* 17 \R (any newline sequence) */
1318 OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */
1319 OP_HSPACE, /* 19 \h (horizontal whitespace) */
1320 OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */
1321 OP_VSPACE, /* 21 \v (vertical whitespace) */
1322 OP_EXTUNI, /* 22 \X (extended Unicode sequence */
1323 OP_EODN, /* 23 End of data or \n at end of data: \Z. */
1324 OP_EOD, /* 24 End of data: \z */
1325
1326 OP_CIRC, /* 25 Start of line - not multiline */
1327 OP_CIRCM, /* 26 Start of line - multiline */
1328 OP_DOLL, /* 27 End of line - not multiline */
1329 OP_DOLLM, /* 28 End of line - multiline */
1330 OP_CHAR, /* 29 Match one character, casefully */
1331 OP_CHARI, /* 30 Match one character, caselessly */
1332 OP_NOT, /* 31 Match one character, not the given one, casefully */
1333 OP_NOTI, /* 32 Match one character, not the given one, caselessly */
1334
1335 /* The following sets of 13 opcodes must always be kept in step because
1336 the offset from the first one is used to generate the others. */
1337
1338 /**** Single characters, caseful, must precede the caseless ones ****/
1339
1340 OP_STAR, /* 33 The maximizing and minimizing versions of */
1341 OP_MINSTAR, /* 34 these six opcodes must come in pairs, with */
1342 OP_PLUS, /* 35 the minimizing one second. */
1343 OP_MINPLUS, /* 36 */
1344 OP_QUERY, /* 37 */
1345 OP_MINQUERY, /* 38 */
1346
1347 OP_UPTO, /* 39 From 0 to n matches of one character, caseful*/
1348 OP_MINUPTO, /* 40 */
1349 OP_EXACT, /* 41 Exactly n matches */
1350
1351 OP_POSSTAR, /* 42 Possessified star, caseful */
1352 OP_POSPLUS, /* 43 Possessified plus, caseful */
1353 OP_POSQUERY, /* 44 Posesssified query, caseful */
1354 OP_POSUPTO, /* 45 Possessified upto, caseful */
1355
1356 /**** Single characters, caseless, must follow the caseful ones */
1357
1358 OP_STARI, /* 46 */
1359 OP_MINSTARI, /* 47 */
1360 OP_PLUSI, /* 48 */
1361 OP_MINPLUSI, /* 49 */
1362 OP_QUERYI, /* 50 */
1363 OP_MINQUERYI, /* 51 */
1364
1365 OP_UPTOI, /* 52 From 0 to n matches of one character, caseless */
1366 OP_MINUPTOI, /* 53 */
1367 OP_EXACTI, /* 54 */
1368
1369 OP_POSSTARI, /* 55 Possessified star, caseless */
1370 OP_POSPLUSI, /* 56 Possessified plus, caseless */
1371 OP_POSQUERYI, /* 57 Posesssified query, caseless */
1372 OP_POSUPTOI, /* 58 Possessified upto, caseless */
1373
1374 /**** The negated ones must follow the non-negated ones, and match them ****/
1375 /**** Negated single character, caseful; must precede the caseless ones ****/
1376
1377 OP_NOTSTAR, /* 59 The maximizing and minimizing versions of */
1378 OP_NOTMINSTAR, /* 60 these six opcodes must come in pairs, with */
1379 OP_NOTPLUS, /* 61 the minimizing one second. They must be in */
1380 OP_NOTMINPLUS, /* 62 exactly the same order as those above. */
1381 OP_NOTQUERY, /* 63 */
1382 OP_NOTMINQUERY, /* 64 */
1383
1384 OP_NOTUPTO, /* 65 From 0 to n matches, caseful */
1385 OP_NOTMINUPTO, /* 66 */
1386 OP_NOTEXACT, /* 67 Exactly n matches */
1387
1388 OP_NOTPOSSTAR, /* 68 Possessified versions, caseful */
1389 OP_NOTPOSPLUS, /* 69 */
1390 OP_NOTPOSQUERY, /* 70 */
1391 OP_NOTPOSUPTO, /* 71 */
1392
1393 /**** Negated single character, caseless; must follow the caseful ones ****/
1394
1395 OP_NOTSTARI, /* 72 */
1396 OP_NOTMINSTARI, /* 73 */
1397 OP_NOTPLUSI, /* 74 */
1398 OP_NOTMINPLUSI, /* 75 */
1399 OP_NOTQUERYI, /* 76 */
1400 OP_NOTMINQUERYI, /* 77 */
1401
1402 OP_NOTUPTOI, /* 78 From 0 to n matches, caseless */
1403 OP_NOTMINUPTOI, /* 79 */
1404 OP_NOTEXACTI, /* 80 Exactly n matches */
1405
1406 OP_NOTPOSSTARI, /* 81 Possessified versions, caseless */
1407 OP_NOTPOSPLUSI, /* 82 */
1408 OP_NOTPOSQUERYI, /* 83 */
1409 OP_NOTPOSUPTOI, /* 84 */
1410
1411 /**** Character types ****/
1412
1413 OP_TYPESTAR, /* 85 The maximizing and minimizing versions of */
1414 OP_TYPEMINSTAR, /* 86 these six opcodes must come in pairs, with */
1415 OP_TYPEPLUS, /* 87 the minimizing one second. These codes must */
1416 OP_TYPEMINPLUS, /* 88 be in exactly the same order as those above. */
1417 OP_TYPEQUERY, /* 89 */
1418 OP_TYPEMINQUERY, /* 90 */
1419
1420 OP_TYPEUPTO, /* 91 From 0 to n matches */
1421 OP_TYPEMINUPTO, /* 92 */
1422 OP_TYPEEXACT, /* 93 Exactly n matches */
1423
1424 OP_TYPEPOSSTAR, /* 94 Possessified versions */
1425 OP_TYPEPOSPLUS, /* 95 */
1426 OP_TYPEPOSQUERY, /* 96 */
1427 OP_TYPEPOSUPTO, /* 97 */
1428
1429 /* These are used for character classes and back references; only the
1430 first six are the same as the sets above. */
1431
1432 OP_CRSTAR, /* 98 The maximizing and minimizing versions of */
1433 OP_CRMINSTAR, /* 99 all these opcodes must come in pairs, with */
1434 OP_CRPLUS, /* 100 the minimizing one second. These codes must */
1435 OP_CRMINPLUS, /* 101 be in exactly the same order as those above. */
1436 OP_CRQUERY, /* 102 */
1437 OP_CRMINQUERY, /* 103 */
1438
1439 OP_CRRANGE, /* 104 These are different to the three sets above. */
1440 OP_CRMINRANGE, /* 105 */
1441
1442 /* End of quantifier opcodes */
1443
1444 OP_CLASS, /* 106 Match a character class, chars < 256 only */
1445 OP_NCLASS, /* 107 Same, but the bitmap was created from a negative
1446 class - the difference is relevant only when a
1447 UTF-8 character > 255 is encountered. */
1448 OP_XCLASS, /* 108 Extended class for handling UTF-8 chars within the
1449 class. This does both positive and negative. */
1450 OP_REF, /* 109 Match a back reference, casefully */
1451 OP_REFI, /* 110 Match a back reference, caselessly */
1452 OP_RECURSE, /* 111 Match a numbered subpattern (possibly recursive) */
1453 OP_CALLOUT, /* 112 Call out to external function if provided */
1454
1455 OP_ALT, /* 113 Start of alternation */
1456 OP_KET, /* 114 End of group that doesn't have an unbounded repeat */
1457 OP_KETRMAX, /* 115 These two must remain together and in this */
1458 OP_KETRMIN, /* 116 order. They are for groups the repeat for ever. */
1459 OP_KETRPOS, /* 117 Possessive unlimited repeat. */
1460
1461 /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
1462 asserts must remain in order. */
1463
1464 OP_REVERSE, /* 118 Move pointer back - used in lookbehind assertions */
1465 OP_ASSERT, /* 119 Positive lookahead */
1466 OP_ASSERT_NOT, /* 120 Negative lookahead */
1467 OP_ASSERTBACK, /* 121 Positive lookbehind */
1468 OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
1469
1470 /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
1471 after the assertions, with ONCE first, as there's a test for >= ONCE for a
1472 subpattern that isn't an assertion. The POS versions must immediately follow
1473 the non-POS versions in each case. */
1474
1475 OP_ONCE, /* 123 Atomic group, contains captures */
1476 OP_ONCE_NC, /* 124 Atomic group containing no captures */
1477 OP_BRA, /* 125 Start of non-capturing bracket */
1478 OP_BRAPOS, /* 126 Ditto, with unlimited, possessive repeat */
1479 OP_CBRA, /* 127 Start of capturing bracket */
1480 OP_CBRAPOS, /* 128 Ditto, with unlimited, possessive repeat */
1481 OP_COND, /* 129 Conditional group */
1482
1483 /* These five must follow the previous five, in the same order. There's a
1484 check for >= SBRA to distinguish the two sets. */
1485
1486 OP_SBRA, /* 130 Start of non-capturing bracket, check empty */
1487 OP_SBRAPOS, /* 131 Ditto, with unlimited, possessive repeat */
1488 OP_SCBRA, /* 132 Start of capturing bracket, check empty */
1489 OP_SCBRAPOS, /* 133 Ditto, with unlimited, possessive repeat */
1490 OP_SCOND, /* 134 Conditional group, check empty */
1491
1492 /* The next two pairs must (respectively) be kept together. */
1493
1494 OP_CREF, /* 135 Used to hold a capture number as condition */
1495 OP_NCREF, /* 136 Same, but generated by a name reference*/
1496 OP_RREF, /* 137 Used to hold a recursion number as condition */
1497 OP_NRREF, /* 138 Same, but generated by a name reference*/
1498 OP_DEF, /* 139 The DEFINE condition */
1499
1500 OP_BRAZERO, /* 140 These two must remain together and in this */
1501 OP_BRAMINZERO, /* 141 order. */
1502 OP_BRAPOSZERO, /* 142 */
1503
1504 /* These are backtracking control verbs */
1505
1506 OP_MARK, /* 143 always has an argument */
1507 OP_PRUNE, /* 144 */
1508 OP_PRUNE_ARG, /* 145 same, but with argument */
1509 OP_SKIP, /* 146 */
1510 OP_SKIP_ARG, /* 147 same, but with argument */
1511 OP_THEN, /* 148 */
1512 OP_THEN_ARG, /* 149 same, but with argument */
1513 OP_COMMIT, /* 150 */
1514
1515 /* These are forced failure and success verbs */
1516
1517 OP_FAIL, /* 151 */
1518 OP_ACCEPT, /* 152 */
1519 OP_ASSERT_ACCEPT, /* 153 Used inside assertions */
1520 OP_CLOSE, /* 154 Used before OP_ACCEPT to close open captures */
1521
1522 /* This is used to skip a subpattern with a {0} quantifier */
1523
1524 OP_SKIPZERO, /* 155 */
1525
1526 /* This is not an opcode, but is used to check that tables indexed by opcode
1527 are the correct length, in order to catch updating errors - there have been
1528 some in the past. */
1529
1530 OP_TABLE_LENGTH
1531 };
1532
1533 /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
1534 definitions that follow must also be updated to match. There are also tables
1535 called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1536
1537
1538 /* This macro defines textual names for all the opcodes. These are used only
1539 for debugging, and some of them are only partial names. The macro is referenced
1540 only in pcre_printint.c, which fills out the full names in many cases (and in
1541 some cases doesn't actually use these names at all). */
1542
1543 #define OP_NAME_LIST \
1544 "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \
1545 "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \
1546 "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \
1547 "extuni", "\\Z", "\\z", \
1548 "^", "^", "$", "$", "char", "chari", "not", "noti", \
1549 "*", "*?", "+", "+?", "?", "??", \
1550 "{", "{", "{", \
1551 "*+","++", "?+", "{", \
1552 "*", "*?", "+", "+?", "?", "??", \
1553 "{", "{", "{", \
1554 "*+","++", "?+", "{", \
1555 "*", "*?", "+", "+?", "?", "??", \
1556 "{", "{", "{", \
1557 "*+","++", "?+", "{", \
1558 "*", "*?", "+", "+?", "?", "??", \
1559 "{", "{", "{", \
1560 "*+","++", "?+", "{", \
1561 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1562 "*+","++", "?+", "{", \
1563 "*", "*?", "+", "+?", "?", "??", "{", "{", \
1564 "class", "nclass", "xclass", "Ref", "Refi", \
1565 "Recurse", "Callout", \
1566 "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos", \
1567 "Reverse", "Assert", "Assert not", "AssertB", "AssertB not", \
1568 "Once", "Once_NC", \
1569 "Bra", "BraPos", "CBra", "CBraPos", \
1570 "Cond", \
1571 "SBra", "SBraPos", "SCBra", "SCBraPos", \
1572 "SCond", \
1573 "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def", \
1574 "Brazero", "Braminzero", "Braposzero", \
1575 "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP", \
1576 "*THEN", "*THEN", "*COMMIT", "*FAIL", \
1577 "*ACCEPT", "*ASSERT_ACCEPT", \
1578 "Close", "Skip zero"
1579
1580
1581 /* This macro defines the length of fixed length operations in the compiled
1582 regex. The lengths are used when searching for specific things, and also in the
1583 debugging printing of a compiled regex. We use a macro so that it can be
1584 defined close to the definitions of the opcodes themselves.
1585
1586 As things have been extended, some of these are no longer fixed lenths, but are
1587 minima instead. For example, the length of a single-character repeat may vary
1588 in UTF-8 mode. The code that uses this table must know about such things. */
1589
1590 #define OP_LENGTHS \
1591 1, /* End */ \
1592 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \
1593 1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \
1594 1, 1, 1, /* Any, AllAny, Anybyte */ \
1595 3, 3, /* \P, \p */ \
1596 1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \
1597 1, /* \X */ \
1598 1, 1, 1, 1, 1, 1, /* \Z, \z, ^, ^M, $, $M */ \
1599 2, /* Char - the minimum length */ \
1600 2, /* Chari - the minimum length */ \
1601 2, /* not */ \
1602 2, /* noti */ \
1603 /* Positive single-char repeats ** These are */ \
1604 2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
1605 4, 4, 4, /* upto, minupto, exact ** mode */ \
1606 2, 2, 2, 4, /* *+, ++, ?+, upto+ */ \
1607 2, 2, 2, 2, 2, 2, /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8 */ \
1608 4, 4, 4, /* upto I, minupto I, exact I */ \
1609 2, 2, 2, 4, /* *+I, ++I, ?+I, upto+I */ \
1610 /* Negative single-char repeats - only for chars < 256 */ \
1611 2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
1612 4, 4, 4, /* NOT upto, minupto, exact */ \
1613 2, 2, 2, 4, /* Possessive NOT *, +, ?, upto */ \
1614 2, 2, 2, 2, 2, 2, /* NOT *I, *?I, +I, +?I, ?I, ??I */ \
1615 4, 4, 4, /* NOT upto I, minupto I, exact I */ \
1616 2, 2, 2, 4, /* Possessive NOT *I, +I, ?I, upto I */ \
1617 /* Positive type repeats */ \
1618 2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
1619 4, 4, 4, /* Type upto, minupto, exact */ \
1620 2, 2, 2, 4, /* Possessive *+, ++, ?+, upto+ */ \
1621 /* Character class & ref repeats */ \
1622 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
1623 5, 5, /* CRRANGE, CRMINRANGE */ \
1624 33, /* CLASS */ \
1625 33, /* NCLASS */ \
1626 0, /* XCLASS - variable length */ \
1627 3, /* REF */ \
1628 3, /* REFI */ \
1629 1+LINK_SIZE, /* RECURSE */ \
1630 2+2*LINK_SIZE, /* CALLOUT */ \
1631 1+LINK_SIZE, /* Alt */ \
1632 1+LINK_SIZE, /* Ket */ \
1633 1+LINK_SIZE, /* KetRmax */ \
1634 1+LINK_SIZE, /* KetRmin */ \
1635 1+LINK_SIZE, /* KetRpos */ \
1636 1+LINK_SIZE, /* Reverse */ \
1637 1+LINK_SIZE, /* Assert */ \
1638 1+LINK_SIZE, /* Assert not */ \
1639 1+LINK_SIZE, /* Assert behind */ \
1640 1+LINK_SIZE, /* Assert behind not */ \
1641 1+LINK_SIZE, /* ONCE */ \
1642 1+LINK_SIZE, /* ONCE_NC */ \
1643 1+LINK_SIZE, /* BRA */ \
1644 1+LINK_SIZE, /* BRAPOS */ \
1645 3+LINK_SIZE, /* CBRA */ \
1646 3+LINK_SIZE, /* CBRAPOS */ \
1647 1+LINK_SIZE, /* COND */ \
1648 1+LINK_SIZE, /* SBRA */ \
1649 1+LINK_SIZE, /* SBRAPOS */ \
1650 3+LINK_SIZE, /* SCBRA */ \
1651 3+LINK_SIZE, /* SCBRAPOS */ \
1652 1+LINK_SIZE, /* SCOND */ \
1653 3, 3, /* CREF, NCREF */ \
1654 3, 3, /* RREF, NRREF */ \
1655 1, /* DEF */ \
1656 1, 1, 1, /* BRAZERO, BRAMINZERO, BRAPOSZERO */ \
1657 3, 1, 3, /* MARK, PRUNE, PRUNE_ARG */ \
1658 1, 3, /* SKIP, SKIP_ARG */ \
1659 1, 3, /* THEN, THEN_ARG */ \
1660 1, 1, 1, 1, /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT */ \
1661 3, 1 /* CLOSE, SKIPZERO */
1662
1663 /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
1664 condition. */
1665
1666 #define RREF_ANY 0xffff
1667
1668 /* Compile time error code numbers. They are given names so that they can more
1669 easily be tracked. When a new number is added, the table called eint in
1670 pcreposix.c must be updated. */
1671
1672 enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
1673 ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
1674 ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
1675 ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
1676 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
1677 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
1678 ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
1679 ERR70, ERRCOUNT };
1680
1681 /* The real format of the start of the pcre block; the index of names and the
1682 code vector run on as long as necessary after the end. We store an explicit
1683 offset to the name table so that if a regex is compiled on one host, saved, and
1684 then run on another where the size of pointers is different, all might still
1685 be well. For the case of compiled-on-4 and run-on-8, we include an extra
1686 pointer that is always NULL. For future-proofing, a few dummy fields were
1687 originally included - even though you can never get this planning right - but
1688 there is only one left now.
1689
1690 NOTE NOTE NOTE:
1691 Because people can now save and re-use compiled patterns, any additions to this
1692 structure should be made at the end, and something earlier (e.g. a new
1693 flag in the options or one of the dummy fields) should indicate that the new
1694 fields are present. Currently PCRE always sets the dummy fields to zero.
1695 NOTE NOTE NOTE
1696 */
1697
1698 typedef struct real_pcre {
1699 pcre_uint32 magic_number;
1700 pcre_uint32 size; /* Total that was malloced */
1701 pcre_uint32 options; /* Public options */
1702 pcre_uint16 flags; /* Private flags */
1703 pcre_uint16 dummy1; /* For future use */
1704 pcre_uint16 top_bracket;
1705 pcre_uint16 top_backref;
1706 pcre_uint16 first_byte;
1707 pcre_uint16 req_byte;
1708 pcre_uint16 name_table_offset; /* Offset to name table that follows */
1709 pcre_uint16 name_entry_size; /* Size of any name items */
1710 pcre_uint16 name_count; /* Number of name items */
1711 pcre_uint16 ref_count; /* Reference count */
1712
1713 const pcre_uint8 *tables; /* Pointer to tables or NULL for std */
1714 const pcre_uint8 *nullpad; /* NULL padding */
1715 } real_pcre;
1716
1717 /* The format of the block used to store data from pcre_study(). The same
1718 remark (see NOTE above) about extending this structure applies. */
1719
1720 typedef struct pcre_study_data {
1721 pcre_uint32 size; /* Total that was malloced */
1722 pcre_uint32 flags; /* Private flags */
1723 pcre_uint8 start_bits[32]; /* Starting char bits */
1724 pcre_uint32 minlength; /* Minimum subject length */
1725 } pcre_study_data;
1726
1727 /* Structure for building a chain of open capturing subpatterns during
1728 compiling, so that instructions to close them can be compiled when (*ACCEPT) is
1729 encountered. This is also used to identify subpatterns that contain recursive
1730 back references to themselves, so that they can be made atomic. */
1731
1732 typedef struct open_capitem {
1733 struct open_capitem *next; /* Chain link */
1734 pcre_uint16 number; /* Capture number */
1735 pcre_uint16 flag; /* Set TRUE if recursive back ref */
1736 } open_capitem;
1737
1738 /* Structure for passing "static" information around between the functions
1739 doing the compiling, so that they are thread-safe. */
1740
1741 typedef struct compile_data {
1742 const pcre_uint8 *lcc; /* Points to lower casing table */
1743 const pcre_uint8 *fcc; /* Points to case-flipping table */
1744 const pcre_uint8 *cbits; /* Points to character type table */
1745 const pcre_uint8 *ctypes; /* Points to table of type maps */
1746 const pcre_uchar *start_workspace;/* The start of working space */
1747 const pcre_uchar *start_code; /* The start of the compiled code */
1748 const pcre_uchar *start_pattern; /* The start of the pattern */
1749 const pcre_uchar *end_pattern; /* The end of the pattern */
1750 open_capitem *open_caps; /* Chain of open capture items */
1751 pcre_uchar *hwm; /* High watermark of workspace */
1752 pcre_uchar *name_table; /* The name/number table */
1753 int names_found; /* Number of entries so far */
1754 int name_entry_size; /* Size of each entry */
1755 int bracount; /* Count of capturing parens as we compile */
1756 int final_bracount; /* Saved value after first pass */
1757 int top_backref; /* Maximum back reference */
1758 unsigned int backref_map; /* Bitmap of low back refs */
1759 int assert_depth; /* Depth of nested assertions */
1760 int external_options; /* External (initial) options */
1761 int external_flags; /* External flag bits to be set */
1762 int req_varyopt; /* "After variable item" flag for reqbyte */
1763 BOOL had_accept; /* (*ACCEPT) encountered */
1764 BOOL check_lookbehind; /* Lookbehinds need later checking */
1765 int nltype; /* Newline type */
1766 int nllen; /* Newline string length */
1767 pcre_uchar nl[4]; /* Newline string when fixed length */
1768 } compile_data;
1769
1770 /* Structure for maintaining a chain of pointers to the currently incomplete
1771 branches, for testing for left recursion while compiling. */
1772
1773 typedef struct branch_chain {
1774 struct branch_chain *outer;
1775 pcre_uchar *current_branch;
1776 } branch_chain;
1777
1778 /* Structure for items in a linked list that represents an explicit recursive
1779 call within the pattern; used by pcre_exec(). */
1780
1781 typedef struct recursion_info {
1782 struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
1783 int group_num; /* Number of group that was called */
1784 int *offset_save; /* Pointer to start of saved offsets */
1785 int saved_max; /* Number of saved offsets */
1786 PCRE_PUCHAR subject_position; /* Position at start of recursion */
1787 } recursion_info;
1788
1789 /* A similar structure for pcre_dfa_exec(). */
1790
1791 typedef struct dfa_recursion_info {
1792 struct dfa_recursion_info *prevrec;
1793 int group_num;
1794 PCRE_PUCHAR subject_position;
1795 } dfa_recursion_info;
1796
1797 /* Structure for building a chain of data for holding the values of the subject
1798 pointer at the start of each subpattern, so as to detect when an empty string
1799 has been matched by a subpattern - to break infinite loops; used by
1800 pcre_exec(). */
1801
1802 typedef struct eptrblock {
1803 struct eptrblock *epb_prev;
1804 PCRE_PUCHAR epb_saved_eptr;
1805 } eptrblock;
1806
1807
1808 /* Structure for passing "static" information around between the functions
1809 doing traditional NFA matching, so that they are thread-safe. */
1810
1811 typedef struct match_data {
1812 unsigned long int match_call_count; /* As it says */
1813 unsigned long int match_limit; /* As it says */
1814 unsigned long int match_limit_recursion; /* As it says */
1815 int *offset_vector; /* Offset vector */
1816 int offset_end; /* One past the end */
1817 int offset_max; /* The maximum usable for return data */
1818 int nltype; /* Newline type */
1819 int nllen; /* Newline string length */
1820 int name_count; /* Number of names in name table */
1821 int name_entry_size; /* Size of entry in names table */
1822 pcre_uchar *name_table; /* Table of names */
1823 pcre_uchar nl[4]; /* Newline string when fixed */
1824 const pcre_uint8 *lcc; /* Points to lower casing table */
1825 const pcre_uint8 *ctypes; /* Points to table of type maps */
1826 BOOL offset_overflow; /* Set if too many extractions */
1827 BOOL notbol; /* NOTBOL flag */
1828 BOOL noteol; /* NOTEOL flag */
1829 BOOL utf8; /* UTF8 flag */
1830 BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */
1831 BOOL use_ucp; /* PCRE_UCP flag */
1832 BOOL endonly; /* Dollar not before final \n */
1833 BOOL notempty; /* Empty string match not wanted */
1834 BOOL notempty_atstart; /* Empty string match at start not wanted */
1835 BOOL hitend; /* Hit the end of the subject at some point */
1836 BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
1837 BOOL hasthen; /* Pattern contains (*THEN) */
1838 const pcre_uchar *start_code; /* For use when recursing */
1839 PCRE_PUCHAR start_subject; /* Start of the subject string */
1840 PCRE_PUCHAR end_subject; /* End of the subject string */
1841 PCRE_PUCHAR start_match_ptr; /* Start of matched string */
1842 PCRE_PUCHAR end_match_ptr; /* Subject position at end match */
1843 PCRE_PUCHAR start_used_ptr; /* Earliest consulted character */
1844 int partial; /* PARTIAL options */
1845 int end_offset_top; /* Highwater mark at end of match */
1846 int capture_last; /* Most recent capture number */
1847 int start_offset; /* The start offset value */
1848 int match_function_type; /* Set for certain special calls of MATCH() */
1849 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
1850 int eptrn; /* Next free eptrblock */
1851 recursion_info *recursive; /* Linked list of recursion data */
1852 void *callout_data; /* To pass back to callouts */
1853 const pcre_uchar *mark; /* Mark pointer to pass back */
1854 const pcre_uchar *once_target; /* Where to back up to for atomic groups */
1855 } match_data;
1856
1857 /* A similar structure is used for the same purpose by the DFA matching
1858 functions. */
1859
1860 typedef struct dfa_match_data {
1861 const pcre_uchar *start_code; /* Start of the compiled pattern */
1862 const pcre_uchar *start_subject ; /* Start of the subject string */
1863 const pcre_uchar *end_subject; /* End of subject string */
1864 const pcre_uchar *start_used_ptr; /* Earliest consulted character */
1865 const pcre_uint8 *tables; /* Character tables */
1866 int start_offset; /* The start offset value */
1867 int moptions; /* Match options */
1868 int poptions; /* Pattern options */
1869 int nltype; /* Newline type */
1870 int nllen; /* Newline string length */
1871 pcre_uchar nl[4]; /* Newline string when fixed */
1872 void *callout_data; /* To pass back to callouts */
1873 dfa_recursion_info *recursive; /* Linked list of recursion data */
1874 } dfa_match_data;
1875
1876 /* Bit definitions for entries in the pcre_ctypes table. */
1877
1878 #define ctype_space 0x01
1879 #define ctype_letter 0x02
1880 #define ctype_digit 0x04
1881 #define ctype_xdigit 0x08
1882 #define ctype_word 0x10 /* alphanumeric or '_' */
1883 #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
1884
1885 /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
1886 of bits for a class map. Some classes are built by combining these tables. */
1887
1888 #define cbit_space 0 /* [:space:] or \s */
1889 #define cbit_xdigit 32 /* [:xdigit:] */
1890 #define cbit_digit 64 /* [:digit:] or \d */
1891 #define cbit_upper 96 /* [:upper:] */
1892 #define cbit_lower 128 /* [:lower:] */
1893 #define cbit_word 160 /* [:word:] or \w */
1894 #define cbit_graph 192 /* [:graph:] */
1895 #define cbit_print 224 /* [:print:] */
1896 #define cbit_punct 256 /* [:punct:] */
1897 #define cbit_cntrl 288 /* [:cntrl:] */
1898 #define cbit_length 320 /* Length of the cbits table */
1899
1900 /* Offsets of the various tables from the base tables pointer, and
1901 total length. */
1902
1903 #define lcc_offset 0
1904 #define fcc_offset 256
1905 #define cbits_offset 512
1906 #define ctypes_offset (cbits_offset + cbit_length)
1907 #define tables_length (ctypes_offset + 256)
1908
1909 /* Layout of the UCP type table that translates property names into types and
1910 codes. Each entry used to point directly to a name, but to reduce the number of
1911 relocations in shared libraries, it now has an offset into a single string
1912 instead. */
1913
1914 typedef struct {
1915 pcre_uint16 name_offset;
1916 pcre_uint16 type;
1917 pcre_uint16 value;
1918 } ucp_type_table;
1919
1920
1921 /* Internal shared data tables. These are tables that are used by more than one
1922 of the exported public functions. They have to be "external" in the C sense,
1923 but are not part of the PCRE public API. The data for these tables is in the
1924 pcre_tables.c module. */
1925
1926 extern const int _pcre_utf8_table1[];
1927 extern const int _pcre_utf8_table2[];
1928 extern const int _pcre_utf8_table3[];
1929 extern const pcre_uint8 _pcre_utf8_table4[];
1930
1931 extern const int _pcre_utf8_table1_size;
1932
1933 extern const char _pcre_utt_names[];
1934 extern const ucp_type_table _pcre_utt[];
1935 extern const int _pcre_utt_size;
1936
1937 extern const pcre_uint8 _pcre_default_tables[];
1938
1939 extern const pcre_uint8 _pcre_OP_lengths[];
1940
1941
1942 /* Internal shared functions. These are functions that are used by more than
1943 one of the exported public functions. They have to be "external" in the C
1944 sense, but are not part of the PCRE public API. */
1945
1946 extern const pcre_uchar *_pcre_find_bracket(const pcre_uchar *, BOOL, int);
1947 extern BOOL _pcre_is_newline(PCRE_PUCHAR, int, PCRE_PUCHAR,
1948 int *, BOOL);
1949 extern int _pcre_ord2utf8(int, pcre_uint8 *);
1950 extern real_pcre *_pcre_try_flipped(const real_pcre *, real_pcre *,
1951 const pcre_study_data *, pcre_study_data *);
1952 #ifndef COMPILE_PCRE16
1953 extern int _pcre_valid_utf8(PCRE_PUCHAR, int, int *);
1954 #else
1955 extern int _pcre16_valid_utf16(PCRE_PUCHAR, int, int *);
1956 #endif
1957 extern BOOL _pcre_was_newline(PCRE_PUCHAR, int, PCRE_PUCHAR,
1958 int *, BOOL);
1959 extern BOOL _pcre_xclass(int, const pcre_uchar *);
1960
1961 #ifdef SUPPORT_JIT
1962 extern void _pcre_jit_compile(const real_pcre *, pcre_extra *);
1963 extern int _pcre_jit_exec(const real_pcre *, void *, PCRE_SPTR,
1964 int, int, int, int, int *, int);
1965 extern void _pcre_jit_free(void *);
1966 #endif
1967
1968 /* Unicode character database (UCD) */
1969
1970 typedef struct {
1971 pcre_uint8 script;
1972 pcre_uint8 chartype;
1973 pcre_int32 other_case;
1974 } ucd_record;
1975
1976 extern const ucd_record _pcre_ucd_records[];
1977 extern const pcre_uint8 _pcre_ucd_stage1[];
1978 extern const pcre_uint16 _pcre_ucd_stage2[];
1979 extern const int _pcre_ucp_gentype[];
1980 #ifdef SUPPORT_JIT
1981 extern const int _pcre_ucp_typerange[];
1982 #endif
1983
1984 /* UCD access macros */
1985
1986 #define UCD_BLOCK_SIZE 128
1987 #define GET_UCD(ch) (_pcre_ucd_records + \
1988 _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \
1989 UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])
1990
1991 #define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype
1992 #define UCD_SCRIPT(ch) GET_UCD(ch)->script
1993 #define UCD_CATEGORY(ch) _pcre_ucp_gentype[UCD_CHARTYPE(ch)]
1994 #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)
1995
1996 #endif
1997
1998 /* 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