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

Contents of /code/trunk/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1151 - (show annotations)
Sun Oct 21 16:53:46 2012 UTC (6 years, 11 months ago) by chpe
File MIME type: text/plain
File size: 109241 byte(s)
valgrind: Add --support-valgrind configure option

When --enable-valgrind is passed, PCRE will annotate some memory regions
as unaddressable, aiding in finding invalid memory accesses when running
under valgrind. This is useful mostly for debugging PCRE itself, and thus
probably should not be enabled in release builds.
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) 1997-2012 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_", "_pcre16_" or "_pcre32_"
44 depending on the PRIV macro. */
45
46 #ifndef PCRE_INTERNAL_H
47 #define PCRE_INTERNAL_H
48
49 /* Define PCRE_DEBUG to get debugging output on stdout. */
50
51 #if 0
52 #define PCRE_DEBUG
53 #endif
54
55 /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
56
57 #if !defined COMPILE_PCRE16 && !defined COMPILE_PCRE32
58 #define COMPILE_PCRE8
59 #endif
60
61 /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
62 "configure" script ensures this, but not everybody uses "configure". */
63
64 #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
65 #define SUPPORT_UTF 1
66 #endif
67
68 /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
69 reasons with existing code. */
70
71 #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
72 #define SUPPORT_UTF 1
73 #endif
74
75 /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
76 Until then we define it if SUPPORT_UTF is defined. */
77
78 #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
79 #define SUPPORT_UTF8 1
80 #endif
81
82 /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
83 script prevents both being selected, but not everybody uses "configure". */
84
85 #if defined EBCDIC && defined SUPPORT_UTF
86 #error The use of both EBCDIC and SUPPORT_UTF is not supported.
87 #endif
88
89 /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
90 inline, and there are *still* stupid compilers about that don't like indented
91 pre-processor statements, or at least there were when I first wrote this. After
92 all, it had only been about 10 years then...
93
94 It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so
95 be absolutely sure we get our version. */
96
97 #undef DPRINTF
98 #ifdef PCRE_DEBUG
99 #define DPRINTF(p) printf p
100 #else
101 #define DPRINTF(p) /* Nothing */
102 #endif
103
104
105 /* Standard C headers plus the external interface definition. The only time
106 setjmp and stdarg are used is when NO_RECURSE is set. */
107
108 #include <ctype.h>
109 #include <limits.h>
110 #include <stddef.h>
111 #include <stdio.h>
112 #include <stdlib.h>
113 #include <string.h>
114
115 /* Valgrind (memcheck) support */
116
117 #ifdef SUPPORT_VALGRIND
118 #include <valgrind/memcheck.h>
119 #endif
120
121 /* When compiling a DLL for Windows, the exported symbols have to be declared
122 using some MS magic. I found some useful information on this web page:
123 http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
124 information there, using __declspec(dllexport) without "extern" we have a
125 definition; with "extern" we have a declaration. The settings here override the
126 setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL,
127 which is all that is needed for applications (they just import the symbols). We
128 use:
129
130 PCRE_EXP_DECL for declarations
131 PCRE_EXP_DEFN for definitions of exported functions
132 PCRE_EXP_DATA_DEFN for definitions of exported variables
133
134 The reason for the two DEFN macros is that in non-Windows environments, one
135 does not want to have "extern" before variable definitions because it leads to
136 compiler warnings. So we distinguish between functions and variables. In
137 Windows, the two should always be the same.
138
139 The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest,
140 which is an application, but needs to import this file in order to "peek" at
141 internals, can #include pcre.h first to get an application's-eye view.
142
143 In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon,
144 special-purpose environments) might want to stick other stuff in front of
145 exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and
146 PCRE_EXP_DATA_DEFN only if they are not already set. */
147
148 #ifndef PCRE_EXP_DECL
149 # ifdef _WIN32
150 # ifndef PCRE_STATIC
151 # define PCRE_EXP_DECL extern __declspec(dllexport)
152 # define PCRE_EXP_DEFN __declspec(dllexport)
153 # define PCRE_EXP_DATA_DEFN __declspec(dllexport)
154 # else
155 # define PCRE_EXP_DECL extern
156 # define PCRE_EXP_DEFN
157 # define PCRE_EXP_DATA_DEFN
158 # endif
159 # else
160 # ifdef __cplusplus
161 # define PCRE_EXP_DECL extern "C"
162 # else
163 # define PCRE_EXP_DECL extern
164 # endif
165 # ifndef PCRE_EXP_DEFN
166 # define PCRE_EXP_DEFN PCRE_EXP_DECL
167 # endif
168 # ifndef PCRE_EXP_DATA_DEFN
169 # define PCRE_EXP_DATA_DEFN
170 # endif
171 # endif
172 #endif
173
174 /* When compiling with the MSVC compiler, it is sometimes necessary to include
175 a "calling convention" before exported function names. (This is secondhand
176 information; I know nothing about MSVC myself). For example, something like
177
178 void __cdecl function(....)
179
180 might be needed. In order so make this easy, all the exported functions have
181 PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
182 set, we ensure here that it has no effect. */
183
184 #ifndef PCRE_CALL_CONVENTION
185 #define PCRE_CALL_CONVENTION
186 #endif
187
188 /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
189 cannot determine these outside the compilation (e.g. by running a program as
190 part of "configure") because PCRE is often cross-compiled for use on other
191 systems. Instead we make use of the maximum sizes that are available at
192 preprocessor time in standard C environments. */
193
194 typedef unsigned char pcre_uint8;
195
196 #if USHRT_MAX == 65535
197 typedef unsigned short pcre_uint16;
198 typedef short pcre_int16;
199 #elif UINT_MAX == 65535
200 typedef unsigned int pcre_uint16;
201 typedef int pcre_int16;
202 #else
203 #error Cannot determine a type for 16-bit unsigned integers
204 #endif
205
206 #if UINT_MAX == 4294967295
207 typedef unsigned int pcre_uint32;
208 typedef int pcre_int32;
209 #elif ULONG_MAX == 4294967295
210 typedef unsigned long int pcre_uint32;
211 typedef long int pcre_int32;
212 #else
213 #error Cannot determine a type for 32-bit unsigned integers
214 #endif
215
216 /* When checking for integer overflow in pcre_compile(), we need to handle
217 large integers. If a 64-bit integer type is available, we can use that.
218 Otherwise we have to cast to double, which of course requires floating point
219 arithmetic. Handle this by defining a macro for the appropriate type. If
220 stdint.h is available, include it; it may define INT64_MAX. Systems that do not
221 have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
222 by "configure". */
223
224 #if HAVE_STDINT_H
225 #include <stdint.h>
226 #elif HAVE_INTTYPES_H
227 #include <inttypes.h>
228 #endif
229
230 #if defined INT64_MAX || defined int64_t
231 #define INT64_OR_DOUBLE int64_t
232 #else
233 #define INT64_OR_DOUBLE double
234 #endif
235
236 /* All character handling must be done as unsigned characters. Otherwise there
237 are problems with top-bit-set characters and functions such as isspace().
238 However, we leave the interface to the outside world as char * or short *,
239 because that should make things easier for callers. This character type is
240 called pcre_uchar.
241
242 The IN_UCHARS macro multiply its argument with the byte size of the current
243 pcre_uchar type. Useful for memcpy and such operations, whose require the
244 byte size of their input/output buffers.
245
246 The MAX_255 macro checks whether its pcre_uchar input is less than 256.
247
248 The TABLE_GET macro is designed for accessing elements of tables whose contain
249 exactly 256 items. When the character is able to contain more than 256
250 items, some check is needed before accessing these tables.
251 */
252
253 #if defined COMPILE_PCRE8
254
255 typedef unsigned char pcre_uchar;
256 #define IN_UCHARS(x) (x)
257 #define MAX_255(c) 1
258 #define TABLE_GET(c, table, default) ((table)[c])
259
260 #elif defined COMPILE_PCRE16
261
262 #if USHRT_MAX != 65535
263 /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
264 pcre.h(.in) and disable (comment out) this message. */
265 #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
266 #endif
267
268 typedef pcre_uint16 pcre_uchar;
269 #define UCHAR_SHIFT (1)
270 #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
271 #define MAX_255(c) ((c) <= 255u)
272 #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
273
274 #elif defined COMPILE_PCRE32
275
276 typedef pcre_uint32 pcre_uchar;
277 #define UCHAR_SHIFT (2)
278 #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
279 #define MAX_255(c) ((c) <= 255u)
280 #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
281
282 /* Assert that pcre_uchar32 is a 32-bit type */
283 typedef int __assert_pcre_uchar32_size[sizeof(pcre_uchar) == 4 ? 1 : -1];
284
285 #else
286 #error Unsupported compiling mode
287 #endif /* COMPILE_PCRE[8|16|32] */
288
289 /* This is an unsigned int value that no character can ever have. UTF-8
290 characters only go up to 0x7fffffff (though Unicode doesn't go beyond
291 0x0010ffff). */
292
293 #define NOTACHAR 0xffffffff
294
295 /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
296 "any" and "anycrlf" at present). The following macros are used to package up
297 testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
298 modules to indicate in which datablock the parameters exist, and what the
299 start/end of string field names are. */
300
301 #define NLTYPE_FIXED 0 /* Newline is a fixed length string */
302 #define NLTYPE_ANY 1 /* Newline is any Unicode line ending */
303 #define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */
304
305 /* This macro checks for a newline at the given position */
306
307 #define IS_NEWLINE(p) \
308 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
309 ((p) < NLBLOCK->PSEND && \
310 PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
311 &(NLBLOCK->nllen), utf)) \
312 : \
313 ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
314 RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
315 (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1]) \
316 ) \
317 )
318
319 /* This macro checks for a newline immediately preceding the given position */
320
321 #define WAS_NEWLINE(p) \
322 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
323 ((p) > NLBLOCK->PSSTART && \
324 PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
325 &(NLBLOCK->nllen), utf)) \
326 : \
327 ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
328 RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] && \
329 (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
330 ) \
331 )
332
333 /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
334 with a custom type. This makes it possible, for example, to allow pcre_exec()
335 to process subject strings that are discontinuous by using a smart pointer
336 class. It must always be possible to inspect all of the subject string in
337 pcre_exec() because of the way it backtracks. Two macros are required in the
338 normal case, for sign-unspecified and unsigned char pointers. The former is
339 used for the external interface and appears in pcre.h, which is why its name
340 must begin with PCRE_. */
341
342 #ifdef CUSTOM_SUBJECT_PTR
343 #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
344 #else
345 #define PCRE_PUCHAR const pcre_uchar *
346 #endif
347
348 /* Include the public PCRE header and the definitions of UCP character property
349 values. */
350
351 #include "pcre.h"
352 #include "ucp.h"
353
354 /* When compiling for use with the Virtual Pascal compiler, these functions
355 need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
356 option on the command line. */
357
358 #ifdef VPCOMPAT
359 #define strlen(s) _strlen(s)
360 #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
361 #define memcmp(s,c,n) _memcmp(s,c,n)
362 #define memcpy(d,s,n) _memcpy(d,s,n)
363 #define memmove(d,s,n) _memmove(d,s,n)
364 #define memset(s,c,n) _memset(s,c,n)
365 #else /* VPCOMPAT */
366
367 /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
368 define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
369 is set. Otherwise, include an emulating function for those systems that have
370 neither (there some non-Unix environments where this is the case). */
371
372 #ifndef HAVE_MEMMOVE
373 #undef memmove /* some systems may have a macro */
374 #ifdef HAVE_BCOPY
375 #define memmove(a, b, c) bcopy(b, a, c)
376 #else /* HAVE_BCOPY */
377 static void *
378 pcre_memmove(void *d, const void *s, size_t n)
379 {
380 size_t i;
381 unsigned char *dest = (unsigned char *)d;
382 const unsigned char *src = (const unsigned char *)s;
383 if (dest > src)
384 {
385 dest += n;
386 src += n;
387 for (i = 0; i < n; ++i) *(--dest) = *(--src);
388 return (void *)dest;
389 }
390 else
391 {
392 for (i = 0; i < n; ++i) *dest++ = *src++;
393 return (void *)(dest - n);
394 }
395 }
396 #define memmove(a, b, c) pcre_memmove(a, b, c)
397 #endif /* not HAVE_BCOPY */
398 #endif /* not HAVE_MEMMOVE */
399 #endif /* not VPCOMPAT */
400
401
402 /* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
403 in big-endian order) by default. These are used, for example, to link from the
404 start of a subpattern to its alternatives and its end. The use of 2 bytes per
405 offset limits the size of the compiled regex to around 64K, which is big enough
406 for almost everybody. However, I received a request for an even bigger limit.
407 For this reason, and also to make the code easier to maintain, the storing and
408 loading of offsets from the byte string is now handled by the macros that are
409 defined here.
410
411 The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
412 the config.h file, but can be overridden by using -D on the command line. This
413 is automated on Unix systems via the "configure" command. */
414
415 #if defined COMPILE_PCRE8
416
417 #if LINK_SIZE == 2
418
419 #define PUT(a,n,d) \
420 (a[n] = (d) >> 8), \
421 (a[(n)+1] = (d) & 255)
422
423 #define GET(a,n) \
424 (((a)[n] << 8) | (a)[(n)+1])
425
426 #define MAX_PATTERN_SIZE (1 << 16)
427
428
429 #elif LINK_SIZE == 3
430
431 #define PUT(a,n,d) \
432 (a[n] = (d) >> 16), \
433 (a[(n)+1] = (d) >> 8), \
434 (a[(n)+2] = (d) & 255)
435
436 #define GET(a,n) \
437 (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
438
439 #define MAX_PATTERN_SIZE (1 << 24)
440
441
442 #elif LINK_SIZE == 4
443
444 #define PUT(a,n,d) \
445 (a[n] = (d) >> 24), \
446 (a[(n)+1] = (d) >> 16), \
447 (a[(n)+2] = (d) >> 8), \
448 (a[(n)+3] = (d) & 255)
449
450 #define GET(a,n) \
451 (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
452
453 /* Keep it positive */
454 #define MAX_PATTERN_SIZE (1 << 30)
455
456 #else
457 #error LINK_SIZE must be either 2, 3, or 4
458 #endif
459
460 #elif defined COMPILE_PCRE16
461
462 #if LINK_SIZE == 2
463
464 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
465 #undef LINK_SIZE
466 #define LINK_SIZE 1
467
468 #define PUT(a,n,d) \
469 (a[n] = (d))
470
471 #define GET(a,n) \
472 (a[n])
473
474 #define MAX_PATTERN_SIZE (1 << 16)
475
476 #elif LINK_SIZE == 3 || LINK_SIZE == 4
477
478 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
479 #undef LINK_SIZE
480 #define LINK_SIZE 2
481
482 #define PUT(a,n,d) \
483 (a[n] = (d) >> 16), \
484 (a[(n)+1] = (d) & 65535)
485
486 #define GET(a,n) \
487 (((a)[n] << 16) | (a)[(n)+1])
488
489 /* Keep it positive */
490 #define MAX_PATTERN_SIZE (1 << 30)
491
492 #else
493 #error LINK_SIZE must be either 2, 3, or 4
494 #endif
495
496 #elif defined COMPILE_PCRE32
497
498 /* Only supported LINK_SIZE is 4 */
499 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
500 #undef LINK_SIZE
501 #define LINK_SIZE 1
502
503 #define PUT(a,n,d) \
504 (a[n] = (d))
505
506 #define GET(a,n) \
507 (a[n])
508
509 /* Keep it positive */
510 #define MAX_PATTERN_SIZE (1 << 30)
511
512 #else
513 #error Unsupported compiling mode
514 #endif /* COMPILE_PCRE[8|16|32] */
515
516 /* Convenience macro defined in terms of the others */
517
518 #define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
519
520
521 /* PCRE uses some other 2-byte quantities that do not change when the size of
522 offsets changes. There are used for repeat counts and for other things such as
523 capturing parenthesis numbers in back references. */
524
525 #if defined COMPILE_PCRE8
526
527 #define IMM2_SIZE 2
528
529 #define PUT2(a,n,d) \
530 a[n] = (d) >> 8; \
531 a[(n)+1] = (d) & 255
532
533 /* For reasons that I do not understand, the expression in this GET2 macro is
534 treated by gcc as a signed expression, even when a is declared as unsigned. It
535 seems that any kind of arithmetic results in a signed value. */
536
537 #define GET2(a,n) \
538 (unsigned int)(((a)[n] << 8) | (a)[(n)+1])
539
540 #elif defined COMPILE_PCRE16
541
542 #define IMM2_SIZE 1
543
544 #define PUT2(a,n,d) \
545 a[n] = d
546
547 #define GET2(a,n) \
548 a[n]
549
550 #elif defined COMPILE_PCRE32
551
552 #define IMM2_SIZE 1
553
554 #define PUT2(a,n,d) \
555 a[n] = d
556
557 #define GET2(a,n) \
558 a[n]
559
560 #else
561 #error Unsupported compiling mode
562 #endif /* COMPILE_PCRE[8|16|32] */
563
564 #define PUT2INC(a,n,d) PUT2(a,n,d), a += IMM2_SIZE
565
566 /* The maximum length of a MARK name is currently one data unit; it may be
567 changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
568
569 #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
570 #define MAX_MARK ((1u << 16) - 1)
571 #else
572 #define MAX_MARK ((1u << 8) - 1)
573 #endif
574
575 /* When UTF encoding is being used, a character is no longer just a single
576 byte. The macros for character handling generate simple sequences when used in
577 character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
578 and other macros are not used when UTF is not supported, so they are not
579 defined. To make sure they can never even appear when UTF support is omitted,
580 we don't even define them. */
581
582 #ifndef SUPPORT_UTF
583
584 /* #define MAX_VALUE_FOR_SINGLE_CHAR */
585 /* #define HAS_EXTRALEN(c) */
586 /* #define GET_EXTRALEN(c) */
587 /* #define NOT_FIRSTCHAR(c) */
588 #define GETCHAR(c, eptr) c = *eptr;
589 #define GETCHARTEST(c, eptr) c = *eptr;
590 #define GETCHARINC(c, eptr) c = *eptr++;
591 #define GETCHARINCTEST(c, eptr) c = *eptr++;
592 #define GETCHARLEN(c, eptr, len) c = *eptr;
593 #define RAWUCHAR(eptr) (*(eptr))
594 #define RAWUCHARINC(eptr) (*(eptr)++)
595 #define RAWUCHARTEST(eptr) (*(eptr))
596 #define RAWUCHARINCTEST(eptr) (*(eptr)++)
597 /* #define GETCHARLENTEST(c, eptr, len) */
598 /* #define BACKCHAR(eptr) */
599 /* #define FORWARDCHAR(eptr) */
600 /* #define ACROSSCHAR(condition, eptr, action) */
601
602 #else /* SUPPORT_UTF */
603
604 /* Tests whether the code point needs extra characters to decode. */
605
606 #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
607
608 /* Base macro to pick up the remaining bytes of a UTF-8 character, not
609 advancing the pointer. */
610
611 #define GETUTF8(c, eptr) \
612 { \
613 if ((c & 0x20) == 0) \
614 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
615 else if ((c & 0x10) == 0) \
616 c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
617 else if ((c & 0x08) == 0) \
618 c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
619 ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
620 else if ((c & 0x04) == 0) \
621 c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
622 ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
623 (eptr[4] & 0x3f); \
624 else \
625 c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
626 ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
627 ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
628 }
629
630 /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
631 the pointer. */
632
633 #define GETUTF8INC(c, eptr) \
634 { \
635 if ((c & 0x20) == 0) \
636 c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
637 else if ((c & 0x10) == 0) \
638 { \
639 c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
640 eptr += 2; \
641 } \
642 else if ((c & 0x08) == 0) \
643 { \
644 c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
645 ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
646 eptr += 3; \
647 } \
648 else if ((c & 0x04) == 0) \
649 { \
650 c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
651 ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
652 (eptr[3] & 0x3f); \
653 eptr += 4; \
654 } \
655 else \
656 { \
657 c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
658 ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
659 ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
660 eptr += 5; \
661 } \
662 }
663
664 #if defined COMPILE_PCRE8
665
666 /* These macros were originally written in the form of loops that used data
667 from the tables whose names start with PRIV(utf8_table). They were rewritten by
668 a user so as not to use loops, because in some environments this gives a
669 significant performance advantage, and it seems never to do any harm. */
670
671 /* Tells the biggest code point which can be encoded as a single character. */
672
673 #define MAX_VALUE_FOR_SINGLE_CHAR 127
674
675 /* Tests whether the code point needs extra characters to decode. */
676
677 #define HAS_EXTRALEN(c) ((c) >= 0xc0)
678
679 /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
680 Otherwise it has an undefined behaviour. */
681
682 #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
683
684 /* Returns TRUE, if the given character is not the first character
685 of a UTF sequence. */
686
687 #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
688
689 /* Get the next UTF-8 character, not advancing the pointer. This is called when
690 we know we are in UTF-8 mode. */
691
692 #define GETCHAR(c, eptr) \
693 c = *eptr; \
694 if (c >= 0xc0) GETUTF8(c, eptr);
695
696 /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
697 pointer. */
698
699 #define GETCHARTEST(c, eptr) \
700 c = *eptr; \
701 if (utf && c >= 0xc0) GETUTF8(c, eptr);
702
703 /* Get the next UTF-8 character, advancing the pointer. This is called when we
704 know we are in UTF-8 mode. */
705
706 #define GETCHARINC(c, eptr) \
707 c = *eptr++; \
708 if (c >= 0xc0) GETUTF8INC(c, eptr);
709
710 /* Get the next character, testing for UTF-8 mode, and advancing the pointer.
711 This is called when we don't know if we are in UTF-8 mode. */
712
713 #define GETCHARINCTEST(c, eptr) \
714 c = *eptr++; \
715 if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
716
717 /* Base macro to pick up the remaining bytes of a UTF-8 character, not
718 advancing the pointer, incrementing the length. */
719
720 #define GETUTF8LEN(c, eptr, len) \
721 { \
722 if ((c & 0x20) == 0) \
723 { \
724 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
725 len++; \
726 } \
727 else if ((c & 0x10) == 0) \
728 { \
729 c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
730 len += 2; \
731 } \
732 else if ((c & 0x08) == 0) \
733 {\
734 c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
735 ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
736 len += 3; \
737 } \
738 else if ((c & 0x04) == 0) \
739 { \
740 c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
741 ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
742 (eptr[4] & 0x3f); \
743 len += 4; \
744 } \
745 else \
746 {\
747 c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
748 ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
749 ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
750 len += 5; \
751 } \
752 }
753
754 /* Get the next UTF-8 character, not advancing the pointer, incrementing length
755 if there are extra bytes. This is called when we know we are in UTF-8 mode. */
756
757 #define GETCHARLEN(c, eptr, len) \
758 c = *eptr; \
759 if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
760
761 /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
762 pointer, incrementing length if there are extra bytes. This is called when we
763 do not know if we are in UTF-8 mode. */
764
765 #define GETCHARLENTEST(c, eptr, len) \
766 c = *eptr; \
767 if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
768
769 /* Returns the next uchar, not advancing the pointer. This is called when
770 we know we are in UTF mode. */
771
772 #define RAWUCHAR(eptr) \
773 (*(eptr))
774
775 /* Returns the next uchar, advancing the pointer. This is called when
776 we know we are in UTF mode. */
777
778 #define RAWUCHARINC(eptr) \
779 (*(eptr)++)
780
781 /* Returns the next uchar, testing for UTF mode, and not advancing the
782 pointer. */
783
784 #define RAWUCHARTEST(eptr) \
785 (*(eptr))
786
787 /* Returns the next uchar, testing for UTF mode, advancing the
788 pointer. */
789
790 #define RAWUCHARINCTEST(eptr) \
791 (*(eptr)++)
792
793 /* If the pointer is not at the start of a character, move it back until
794 it is. This is called only in UTF-8 mode - we don't put a test within the macro
795 because almost all calls are already within a block of UTF-8 only code. */
796
797 #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
798
799 /* Same as above, just in the other direction. */
800 #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
801
802 /* Same as above, but it allows a fully customizable form. */
803 #define ACROSSCHAR(condition, eptr, action) \
804 while((condition) && ((eptr) & 0xc0) == 0x80) action
805
806 #elif defined COMPILE_PCRE16
807
808 /* Tells the biggest code point which can be encoded as a single character. */
809
810 #define MAX_VALUE_FOR_SINGLE_CHAR 65535
811
812 /* Tests whether the code point needs extra characters to decode. */
813
814 #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
815
816 /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
817 Otherwise it has an undefined behaviour. */
818
819 #define GET_EXTRALEN(c) 1
820
821 /* Returns TRUE, if the given character is not the first character
822 of a UTF sequence. */
823
824 #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
825
826 /* Base macro to pick up the low surrogate of a UTF-16 character, not
827 advancing the pointer. */
828
829 #define GETUTF16(c, eptr) \
830 { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
831
832 /* Get the next UTF-16 character, not advancing the pointer. This is called when
833 we know we are in UTF-16 mode. */
834
835 #define GETCHAR(c, eptr) \
836 c = *eptr; \
837 if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
838
839 /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
840 pointer. */
841
842 #define GETCHARTEST(c, eptr) \
843 c = *eptr; \
844 if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
845
846 /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
847 the pointer. */
848
849 #define GETUTF16INC(c, eptr) \
850 { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
851
852 /* Get the next UTF-16 character, advancing the pointer. This is called when we
853 know we are in UTF-16 mode. */
854
855 #define GETCHARINC(c, eptr) \
856 c = *eptr++; \
857 if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
858
859 /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
860 This is called when we don't know if we are in UTF-16 mode. */
861
862 #define GETCHARINCTEST(c, eptr) \
863 c = *eptr++; \
864 if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
865
866 /* Base macro to pick up the low surrogate of a UTF-16 character, not
867 advancing the pointer, incrementing the length. */
868
869 #define GETUTF16LEN(c, eptr, len) \
870 { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
871
872 /* Get the next UTF-16 character, not advancing the pointer, incrementing
873 length if there is a low surrogate. This is called when we know we are in
874 UTF-16 mode. */
875
876 #define GETCHARLEN(c, eptr, len) \
877 c = *eptr; \
878 if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
879
880 /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
881 pointer, incrementing length if there is a low surrogate. This is called when
882 we do not know if we are in UTF-16 mode. */
883
884 #define GETCHARLENTEST(c, eptr, len) \
885 c = *eptr; \
886 if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
887
888 /* Returns the next uchar, not advancing the pointer. This is called when
889 we know we are in UTF mode. */
890
891 #define RAWUCHAR(eptr) \
892 (*(eptr))
893
894 /* Returns the next uchar, advancing the pointer. This is called when
895 we know we are in UTF mode. */
896
897 #define RAWUCHARINC(eptr) \
898 (*(eptr)++)
899
900 /* Returns the next uchar, testing for UTF mode, and not advancing the
901 pointer. */
902
903 #define RAWUCHARTEST(eptr) \
904 (*(eptr))
905
906 /* Returns the next uchar, testing for UTF mode, advancing the
907 pointer. */
908
909 #define RAWUCHARINCTEST(eptr) \
910 (*(eptr)++)
911
912 /* If the pointer is not at the start of a character, move it back until
913 it is. This is called only in UTF-16 mode - we don't put a test within the
914 macro because almost all calls are already within a block of UTF-16 only
915 code. */
916
917 #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
918
919 /* Same as above, just in the other direction. */
920 #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
921
922 /* Same as above, but it allows a fully customizable form. */
923 #define ACROSSCHAR(condition, eptr, action) \
924 if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
925
926 #elif defined COMPILE_PCRE32
927
928 /* These are trivial for the 32-bit library, since all UTF-32 characters fit
929 into one pcre_uchar unit. */
930 #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
931 #define HAS_EXTRALEN(c) (0)
932 #define GET_EXTRALEN(c) (0)
933 #define NOT_FIRSTCHAR(c) (0)
934
935 #define UTF32_MASK (0x1fffffu)
936
937 /* Get the next UTF-32 character, not advancing the pointer. This is called when
938 we know we are in UTF-32 mode. */
939
940 #define GETCHAR(c, eptr) \
941 c = (*eptr) & UTF32_MASK;
942
943 /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
944 pointer. */
945
946 #define GETCHARTEST(c, eptr) \
947 c = *eptr; \
948 if (utf) c &= UTF32_MASK;
949
950 /* Get the next UTF-32 character, advancing the pointer. This is called when we
951 know we are in UTF-32 mode. */
952
953 #define GETCHARINC(c, eptr) \
954 c = (*eptr++) & UTF32_MASK;
955
956 /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
957 This is called when we don't know if we are in UTF-32 mode. */
958
959 #define GETCHARINCTEST(c, eptr) \
960 c = *eptr++; \
961 if (utf) c &= UTF32_MASK;
962
963 /* Get the next UTF-32 character, not advancing the pointer, not incrementing
964 length (since all UTF-32 is of length 1). This is called when we know we are in
965 UTF-32 mode. */
966
967 #define GETCHARLEN(c, eptr, len) \
968 GETCHAR(c, eptr)
969
970 /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
971 pointer, not incrementing the length (since all UTF-32 is of length 1).
972 This is called when we do not know if we are in UTF-32 mode. */
973
974 #define GETCHARLENTEST(c, eptr, len) \
975 GETCHARTEST(c, eptr)
976
977 /* Returns the next uchar, not advancing the pointer. This is called when
978 we know we are in UTF mode. */
979
980 #define RAWUCHAR(eptr) \
981 (*(eptr) & UTF32_MASK)
982
983 /* Returns the next uchar, advancing the pointer. This is called when
984 we know we are in UTF mode. */
985
986 #define RAWUCHARINC(eptr) \
987 (*(eptr)++ & UTF32_MASK)
988
989 /* Returns the next uchar, testing for UTF mode, and not advancing the
990 pointer. */
991
992 #define RAWUCHARTEST(eptr) \
993 (utf ? (*(eptr) & UTF32_MASK) : *(eptr))
994
995 /* Returns the next uchar, testing for UTF mode, advancing the
996 pointer. */
997
998 #define RAWUCHARINCTEST(eptr) \
999 (utf ? (*(eptr)++ & UTF32_MASK) : *(eptr)++)
1000
1001 /* If the pointer is not at the start of a character, move it back until
1002 it is. This is called only in UTF-32 mode - we don't put a test within the
1003 macro because almost all calls are already within a block of UTF-32 only
1004 code.
1005 These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
1006
1007 #define BACKCHAR(eptr) do { } while (0)
1008
1009 /* Same as above, just in the other direction. */
1010 #define FORWARDCHAR(eptr) do { } while (0)
1011
1012 /* Same as above, but it allows a fully customizable form. */
1013 #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1014
1015 #else
1016 #error Unsupported compiling mode
1017 #endif /* COMPILE_PCRE[8|16|32] */
1018
1019 #endif /* SUPPORT_UTF */
1020
1021 /* Tests for Unicode horizontal and vertical whitespace characters must check a
1022 number of different values. Using a switch statement for this generates the
1023 fastest code (no loop, no memory access), and there are several places in the
1024 interpreter code where this happens. In order to ensure that all the case lists
1025 remain in step, we use macros so that there is only one place where the lists
1026 are defined.
1027
1028 These values are also required as lists in pcre_compile.c when processing \h,
1029 \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1030 macros that define the values are here so that all the definitions are
1031 together. The lists must be in ascending character order, terminated by
1032 NOTACHAR (which is 0xffffffff).
1033
1034 Any changes should ensure that the various macros are kept in step with each
1035 other. NOTE: The values also appear in pcre_jit_compile.c. */
1036
1037 /* ------ ASCII/Unicode environments ------ */
1038
1039 #ifndef EBCDIC
1040
1041 #define HSPACE_LIST \
1042 CHAR_HT, CHAR_SPACE, 0xa0, \
1043 0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1044 0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1045 NOTACHAR
1046
1047 #define HSPACE_MULTIBYTE_CASES \
1048 case 0x1680: /* OGHAM SPACE MARK */ \
1049 case 0x180e: /* MONGOLIAN VOWEL SEPARATOR */ \
1050 case 0x2000: /* EN QUAD */ \
1051 case 0x2001: /* EM QUAD */ \
1052 case 0x2002: /* EN SPACE */ \
1053 case 0x2003: /* EM SPACE */ \
1054 case 0x2004: /* THREE-PER-EM SPACE */ \
1055 case 0x2005: /* FOUR-PER-EM SPACE */ \
1056 case 0x2006: /* SIX-PER-EM SPACE */ \
1057 case 0x2007: /* FIGURE SPACE */ \
1058 case 0x2008: /* PUNCTUATION SPACE */ \
1059 case 0x2009: /* THIN SPACE */ \
1060 case 0x200A: /* HAIR SPACE */ \
1061 case 0x202f: /* NARROW NO-BREAK SPACE */ \
1062 case 0x205f: /* MEDIUM MATHEMATICAL SPACE */ \
1063 case 0x3000 /* IDEOGRAPHIC SPACE */
1064
1065 #define HSPACE_BYTE_CASES \
1066 case CHAR_HT: \
1067 case CHAR_SPACE: \
1068 case 0xa0 /* NBSP */
1069
1070 #define HSPACE_CASES \
1071 HSPACE_BYTE_CASES: \
1072 HSPACE_MULTIBYTE_CASES
1073
1074 #define VSPACE_LIST \
1075 CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1076
1077 #define VSPACE_MULTIBYTE_CASES \
1078 case 0x2028: /* LINE SEPARATOR */ \
1079 case 0x2029 /* PARAGRAPH SEPARATOR */
1080
1081 #define VSPACE_BYTE_CASES \
1082 case CHAR_LF: \
1083 case CHAR_VT: \
1084 case CHAR_FF: \
1085 case CHAR_CR: \
1086 case CHAR_NEL
1087
1088 #define VSPACE_CASES \
1089 VSPACE_BYTE_CASES: \
1090 VSPACE_MULTIBYTE_CASES
1091
1092 /* ------ EBCDIC environments ------ */
1093
1094 #else
1095 #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1096
1097 #define HSPACE_BYTE_CASES \
1098 case CHAR_HT: \
1099 case CHAR_SPACE
1100
1101 #define HSPACE_CASES HSPACE_BYTE_CASES
1102
1103 #ifdef EBCDIC_NL25
1104 #define VSPACE_LIST \
1105 CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1106 #else
1107 #define VSPACE_LIST \
1108 CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1109 #endif
1110
1111 #define VSPACE_BYTE_CASES \
1112 case CHAR_LF: \
1113 case CHAR_VT: \
1114 case CHAR_FF: \
1115 case CHAR_CR: \
1116 case CHAR_NEL
1117
1118 #define VSPACE_CASES VSPACE_BYTE_CASES
1119 #endif /* EBCDIC */
1120
1121 /* ------ End of whitespace macros ------ */
1122
1123
1124
1125 /* Private flags containing information about the compiled regex. They used to
1126 live at the top end of the options word, but that got almost full, so now they
1127 are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
1128 the restrictions on partial matching have been lifted. It remains for backwards
1129 compatibility. */
1130
1131 #define PCRE_MODE8 0x0001 /* compiled in 8 bit mode */
1132 #define PCRE_MODE16 0x0002 /* compiled in 16 bit mode */
1133 #define PCRE_MODE32 0x0004 /* compiled in 32 bit mode */
1134 #define PCRE_FIRSTSET 0x0010 /* first_char is set */
1135 #define PCRE_FCH_CASELESS 0x0020 /* caseless first char */
1136 #define PCRE_REQCHSET 0x0040 /* req_byte is set */
1137 #define PCRE_RCH_CASELESS 0x0080 /* caseless requested char */
1138 #define PCRE_STARTLINE 0x0100 /* start after \n for multiline */
1139 #define PCRE_NOPARTIAL 0x0200 /* can't use partial with this regex */
1140 #define PCRE_JCHANGED 0x0400 /* j option used in regex */
1141 #define PCRE_HASCRORLF 0x0800 /* explicit \r or \n in pattern */
1142 #define PCRE_HASTHEN 0x1000 /* pattern contains (*THEN) */
1143
1144 #if defined COMPILE_PCRE8
1145 #define PCRE_MODE PCRE_MODE8
1146 #elif defined COMPILE_PCRE16
1147 #define PCRE_MODE PCRE_MODE16
1148 #elif defined COMPILE_PCRE32
1149 #define PCRE_MODE PCRE_MODE32
1150 #endif
1151 #define PCRE_MODE_MASK (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1152
1153 /* Flags for the "extra" block produced by pcre_study(). */
1154
1155 #define PCRE_STUDY_MAPPED 0x0001 /* a map of starting chars exists */
1156 #define PCRE_STUDY_MINLEN 0x0002 /* a minimum length field exists */
1157
1158 /* Masks for identifying the public options that are permitted at compile
1159 time, run time, or study time, respectively. */
1160
1161 #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
1162 PCRE_NEWLINE_ANYCRLF)
1163
1164 #define PUBLIC_COMPILE_OPTIONS \
1165 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
1166 PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1167 PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
1168 PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1169 PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
1170
1171 #define PUBLIC_EXEC_OPTIONS \
1172 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1173 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
1174 PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
1175
1176 #define PUBLIC_DFA_EXEC_OPTIONS \
1177 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1178 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
1179 PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1180 PCRE_NO_START_OPTIMIZE)
1181
1182 #define PUBLIC_STUDY_OPTIONS \
1183 (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1184 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1185
1186 /* Magic number to provide a small check against being handed junk. */
1187
1188 #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
1189
1190 /* This variable is used to detect a loaded regular expression
1191 in different endianness. */
1192
1193 #define REVERSED_MAGIC_NUMBER 0x45524350UL /* 'ERCP' */
1194
1195 /* The maximum remaining length of subject we are prepared to search for a
1196 req_byte match. */
1197
1198 #define REQ_BYTE_MAX 1000
1199
1200 /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1201 environments where these macros are defined elsewhere. Unfortunately, there
1202 is no way to do the same for the typedef. */
1203
1204 typedef int BOOL;
1205
1206 #ifndef FALSE
1207 #define FALSE 0
1208 #define TRUE 1
1209 #endif
1210
1211 /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
1212 character constants like '*' because the compiler would emit their EBCDIC code,
1213 which is different from their ASCII/UTF-8 code. Instead we define macros for
1214 the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
1215 is enabled. When UTF-8 support is not enabled, the definitions use character
1216 literals. Both character and string versions of each character are needed, and
1217 there are some longer strings as well.
1218
1219 This means that, on EBCDIC platforms, the PCRE library can handle either
1220 EBCDIC, or UTF-8, but not both. To support both in the same compiled library
1221 would need different lookups depending on whether PCRE_UTF8 was set or not.
1222 This would make it impossible to use characters in switch/case statements,
1223 which would reduce performance. For a theoretical use (which nobody has asked
1224 for) in a minority area (EBCDIC platforms), this is not sensible. Any
1225 application that did need both could compile two versions of the library, using
1226 macros to give the functions distinct names. */
1227
1228 #ifndef SUPPORT_UTF
1229
1230 /* UTF-8 support is not enabled; use the platform-dependent character literals
1231 so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1232 mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1233 characters, a common practice has been to use its NL (0x15) character as the
1234 line terminator in C-like processing environments. However, sometimes the LF
1235 (0x25) character is used instead, according to this Unicode document:
1236
1237 http://unicode.org/standard/reports/tr13/tr13-5.html
1238
1239 PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1240 instead. Whichever is *not* chosen is defined as NEL.
1241
1242 In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1243 same code point. */
1244
1245 #ifdef EBCDIC
1246
1247 #ifndef EBCDIC_NL25
1248 #define CHAR_NL '\x15'
1249 #define CHAR_NEL '\x25'
1250 #define STR_NL "\x15"
1251 #define STR_NEL "\x25"
1252 #else
1253 #define CHAR_NL '\x25'
1254 #define CHAR_NEL '\x15'
1255 #define STR_NL "\x25"
1256 #define STR_NEL "\x15"
1257 #endif
1258
1259 #define CHAR_LF CHAR_NL
1260 #define STR_LF STR_NL
1261
1262 #define CHAR_ESC '\047'
1263 #define CHAR_DEL '\007'
1264 #define STR_ESC "\047"
1265 #define STR_DEL "\007"
1266
1267 #else /* Not EBCDIC */
1268
1269 /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1270 compatibility. NEL is the Unicode newline character; make sure it is
1271 a positive value. */
1272
1273 #define CHAR_LF '\n'
1274 #define CHAR_NL CHAR_LF
1275 #define CHAR_NEL ((unsigned char)'\x85')
1276 #define CHAR_ESC '\033'
1277 #define CHAR_DEL '\177'
1278
1279 #define STR_LF "\n"
1280 #define STR_NL STR_LF
1281 #define STR_NEL "\x85"
1282 #define STR_ESC "\033"
1283 #define STR_DEL "\177"
1284
1285 #endif /* EBCDIC */
1286
1287 /* The remaining definitions work in both environments. */
1288
1289 #define CHAR_HT '\t'
1290 #define CHAR_VT '\v'
1291 #define CHAR_FF '\f'
1292 #define CHAR_CR '\r'
1293 #define CHAR_BS '\b'
1294 #define CHAR_BEL '\a'
1295
1296 #define CHAR_SPACE ' '
1297 #define CHAR_EXCLAMATION_MARK '!'
1298 #define CHAR_QUOTATION_MARK '"'
1299 #define CHAR_NUMBER_SIGN '#'
1300 #define CHAR_DOLLAR_SIGN '$'
1301 #define CHAR_PERCENT_SIGN '%'
1302 #define CHAR_AMPERSAND '&'
1303 #define CHAR_APOSTROPHE '\''
1304 #define CHAR_LEFT_PARENTHESIS '('
1305 #define CHAR_RIGHT_PARENTHESIS ')'
1306 #define CHAR_ASTERISK '*'
1307 #define CHAR_PLUS '+'
1308 #define CHAR_COMMA ','
1309 #define CHAR_MINUS '-'
1310 #define CHAR_DOT '.'
1311 #define CHAR_SLASH '/'
1312 #define CHAR_0 '0'
1313 #define CHAR_1 '1'
1314 #define CHAR_2 '2'
1315 #define CHAR_3 '3'
1316 #define CHAR_4 '4'
1317 #define CHAR_5 '5'
1318 #define CHAR_6 '6'
1319 #define CHAR_7 '7'
1320 #define CHAR_8 '8'
1321 #define CHAR_9 '9'
1322 #define CHAR_COLON ':'
1323 #define CHAR_SEMICOLON ';'
1324 #define CHAR_LESS_THAN_SIGN '<'
1325 #define CHAR_EQUALS_SIGN '='
1326 #define CHAR_GREATER_THAN_SIGN '>'
1327 #define CHAR_QUESTION_MARK '?'
1328 #define CHAR_COMMERCIAL_AT '@'
1329 #define CHAR_A 'A'
1330 #define CHAR_B 'B'
1331 #define CHAR_C 'C'
1332 #define CHAR_D 'D'
1333 #define CHAR_E 'E'
1334 #define CHAR_F 'F'
1335 #define CHAR_G 'G'
1336 #define CHAR_H 'H'
1337 #define CHAR_I 'I'
1338 #define CHAR_J 'J'
1339 #define CHAR_K 'K'
1340 #define CHAR_L 'L'
1341 #define CHAR_M 'M'
1342 #define CHAR_N 'N'
1343 #define CHAR_O 'O'
1344 #define CHAR_P 'P'
1345 #define CHAR_Q 'Q'
1346 #define CHAR_R 'R'
1347 #define CHAR_S 'S'
1348 #define CHAR_T 'T'
1349 #define CHAR_U 'U'
1350 #define CHAR_V 'V'
1351 #define CHAR_W 'W'
1352 #define CHAR_X 'X'
1353 #define CHAR_Y 'Y'
1354 #define CHAR_Z 'Z'
1355 #define CHAR_LEFT_SQUARE_BRACKET '['
1356 #define CHAR_BACKSLASH '\\'
1357 #define CHAR_RIGHT_SQUARE_BRACKET ']'
1358 #define CHAR_CIRCUMFLEX_ACCENT '^'
1359 #define CHAR_UNDERSCORE '_'
1360 #define CHAR_GRAVE_ACCENT '`'
1361 #define CHAR_a 'a'
1362 #define CHAR_b 'b'
1363 #define CHAR_c 'c'
1364 #define CHAR_d 'd'
1365 #define CHAR_e 'e'
1366 #define CHAR_f 'f'
1367 #define CHAR_g 'g'
1368 #define CHAR_h 'h'
1369 #define CHAR_i 'i'
1370 #define CHAR_j 'j'
1371 #define CHAR_k 'k'
1372 #define CHAR_l 'l'
1373 #define CHAR_m 'm'
1374 #define CHAR_n 'n'
1375 #define CHAR_o 'o'
1376 #define CHAR_p 'p'
1377 #define CHAR_q 'q'
1378 #define CHAR_r 'r'
1379 #define CHAR_s 's'
1380 #define CHAR_t 't'
1381 #define CHAR_u 'u'
1382 #define CHAR_v 'v'
1383 #define CHAR_w 'w'
1384 #define CHAR_x 'x'
1385 #define CHAR_y 'y'
1386 #define CHAR_z 'z'
1387 #define CHAR_LEFT_CURLY_BRACKET '{'
1388 #define CHAR_VERTICAL_LINE '|'
1389 #define CHAR_RIGHT_CURLY_BRACKET '}'
1390 #define CHAR_TILDE '~'
1391
1392 #define STR_HT "\t"
1393 #define STR_VT "\v"
1394 #define STR_FF "\f"
1395 #define STR_CR "\r"
1396 #define STR_BS "\b"
1397 #define STR_BEL "\a"
1398
1399 #define STR_SPACE " "
1400 #define STR_EXCLAMATION_MARK "!"
1401 #define STR_QUOTATION_MARK "\""
1402 #define STR_NUMBER_SIGN "#"
1403 #define STR_DOLLAR_SIGN "$"
1404 #define STR_PERCENT_SIGN "%"
1405 #define STR_AMPERSAND "&"
1406 #define STR_APOSTROPHE "'"
1407 #define STR_LEFT_PARENTHESIS "("
1408 #define STR_RIGHT_PARENTHESIS ")"
1409 #define STR_ASTERISK "*"
1410 #define STR_PLUS "+"
1411 #define STR_COMMA ","
1412 #define STR_MINUS "-"
1413 #define STR_DOT "."
1414 #define STR_SLASH "/"
1415 #define STR_0 "0"
1416 #define STR_1 "1"
1417 #define STR_2 "2"
1418 #define STR_3 "3"
1419 #define STR_4 "4"
1420 #define STR_5 "5"
1421 #define STR_6 "6"
1422 #define STR_7 "7"
1423 #define STR_8 "8"
1424 #define STR_9 "9"
1425 #define STR_COLON ":"
1426 #define STR_SEMICOLON ";"
1427 #define STR_LESS_THAN_SIGN "<"
1428 #define STR_EQUALS_SIGN "="
1429 #define STR_GREATER_THAN_SIGN ">"
1430 #define STR_QUESTION_MARK "?"
1431 #define STR_COMMERCIAL_AT "@"
1432 #define STR_A "A"
1433 #define STR_B "B"
1434 #define STR_C "C"
1435 #define STR_D "D"
1436 #define STR_E "E"
1437 #define STR_F "F"
1438 #define STR_G "G"
1439 #define STR_H "H"
1440 #define STR_I "I"
1441 #define STR_J "J"
1442 #define STR_K "K"
1443 #define STR_L "L"
1444 #define STR_M "M"
1445 #define STR_N "N"
1446 #define STR_O "O"
1447 #define STR_P "P"
1448 #define STR_Q "Q"
1449 #define STR_R "R"
1450 #define STR_S "S"
1451 #define STR_T "T"
1452 #define STR_U "U"
1453 #define STR_V "V"
1454 #define STR_W "W"
1455 #define STR_X "X"
1456 #define STR_Y "Y"
1457 #define STR_Z "Z"
1458 #define STR_LEFT_SQUARE_BRACKET "["
1459 #define STR_BACKSLASH "\\"
1460 #define STR_RIGHT_SQUARE_BRACKET "]"
1461 #define STR_CIRCUMFLEX_ACCENT "^"
1462 #define STR_UNDERSCORE "_"
1463 #define STR_GRAVE_ACCENT "`"
1464 #define STR_a "a"
1465 #define STR_b "b"
1466 #define STR_c "c"
1467 #define STR_d "d"
1468 #define STR_e "e"
1469 #define STR_f "f"
1470 #define STR_g "g"
1471 #define STR_h "h"
1472 #define STR_i "i"
1473 #define STR_j "j"
1474 #define STR_k "k"
1475 #define STR_l "l"
1476 #define STR_m "m"
1477 #define STR_n "n"
1478 #define STR_o "o"
1479 #define STR_p "p"
1480 #define STR_q "q"
1481 #define STR_r "r"
1482 #define STR_s "s"
1483 #define STR_t "t"
1484 #define STR_u "u"
1485 #define STR_v "v"
1486 #define STR_w "w"
1487 #define STR_x "x"
1488 #define STR_y "y"
1489 #define STR_z "z"
1490 #define STR_LEFT_CURLY_BRACKET "{"
1491 #define STR_VERTICAL_LINE "|"
1492 #define STR_RIGHT_CURLY_BRACKET "}"
1493 #define STR_TILDE "~"
1494
1495 #define STRING_ACCEPT0 "ACCEPT\0"
1496 #define STRING_COMMIT0 "COMMIT\0"
1497 #define STRING_F0 "F\0"
1498 #define STRING_FAIL0 "FAIL\0"
1499 #define STRING_MARK0 "MARK\0"
1500 #define STRING_PRUNE0 "PRUNE\0"
1501 #define STRING_SKIP0 "SKIP\0"
1502 #define STRING_THEN "THEN"
1503
1504 #define STRING_alpha0 "alpha\0"
1505 #define STRING_lower0 "lower\0"
1506 #define STRING_upper0 "upper\0"
1507 #define STRING_alnum0 "alnum\0"
1508 #define STRING_ascii0 "ascii\0"
1509 #define STRING_blank0 "blank\0"
1510 #define STRING_cntrl0 "cntrl\0"
1511 #define STRING_digit0 "digit\0"
1512 #define STRING_graph0 "graph\0"
1513 #define STRING_print0 "print\0"
1514 #define STRING_punct0 "punct\0"
1515 #define STRING_space0 "space\0"
1516 #define STRING_word0 "word\0"
1517 #define STRING_xdigit "xdigit"
1518
1519 #define STRING_DEFINE "DEFINE"
1520
1521 #define STRING_CR_RIGHTPAR "CR)"
1522 #define STRING_LF_RIGHTPAR "LF)"
1523 #define STRING_CRLF_RIGHTPAR "CRLF)"
1524 #define STRING_ANY_RIGHTPAR "ANY)"
1525 #define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)"
1526 #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"
1527 #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"
1528 #ifdef COMPILE_PCRE8
1529 #define STRING_UTF_RIGHTPAR "UTF8)"
1530 #endif
1531 #ifdef COMPILE_PCRE16
1532 #define STRING_UTF_RIGHTPAR "UTF16)"
1533 #endif
1534 #ifdef COMPILE_PCRE32
1535 #define STRING_UTF_RIGHTPAR "UTF32)"
1536 #endif
1537 #define STRING_UCP_RIGHTPAR "UCP)"
1538 #define STRING_NO_START_OPT_RIGHTPAR "NO_START_OPT)"
1539
1540 #else /* SUPPORT_UTF */
1541
1542 /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1543 works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
1544 only. */
1545
1546 #define CHAR_HT '\011'
1547 #define CHAR_VT '\013'
1548 #define CHAR_FF '\014'
1549 #define CHAR_CR '\015'
1550 #define CHAR_LF '\012'
1551 #define CHAR_NL CHAR_LF
1552 #define CHAR_NEL ((unsigned char)'\x85')
1553 #define CHAR_BS '\010'
1554 #define CHAR_BEL '\007'
1555 #define CHAR_ESC '\033'
1556 #define CHAR_DEL '\177'
1557
1558 #define CHAR_SPACE '\040'
1559 #define CHAR_EXCLAMATION_MARK '\041'
1560 #define CHAR_QUOTATION_MARK '\042'
1561 #define CHAR_NUMBER_SIGN '\043'
1562 #define CHAR_DOLLAR_SIGN '\044'
1563 #define CHAR_PERCENT_SIGN '\045'
1564 #define CHAR_AMPERSAND '\046'
1565 #define CHAR_APOSTROPHE '\047'
1566 #define CHAR_LEFT_PARENTHESIS '\050'
1567 #define CHAR_RIGHT_PARENTHESIS '\051'
1568 #define CHAR_ASTERISK '\052'
1569 #define CHAR_PLUS '\053'
1570 #define CHAR_COMMA '\054'
1571 #define CHAR_MINUS '\055'
1572 #define CHAR_DOT '\056'
1573 #define CHAR_SLASH '\057'
1574 #define CHAR_0 '\060'
1575 #define CHAR_1 '\061'
1576 #define CHAR_2 '\062'
1577 #define CHAR_3 '\063'
1578 #define CHAR_4 '\064'
1579 #define CHAR_5 '\065'
1580 #define CHAR_6 '\066'
1581 #define CHAR_7 '\067'
1582 #define CHAR_8 '\070'
1583 #define CHAR_9 '\071'
1584 #define CHAR_COLON '\072'
1585 #define CHAR_SEMICOLON '\073'
1586 #define CHAR_LESS_THAN_SIGN '\074'
1587 #define CHAR_EQUALS_SIGN '\075'
1588 #define CHAR_GREATER_THAN_SIGN '\076'
1589 #define CHAR_QUESTION_MARK '\077'
1590 #define CHAR_COMMERCIAL_AT '\100'
1591 #define CHAR_A '\101'
1592 #define CHAR_B '\102'
1593 #define CHAR_C '\103'
1594 #define CHAR_D '\104'
1595 #define CHAR_E '\105'
1596 #define CHAR_F '\106'
1597 #define CHAR_G '\107'
1598 #define CHAR_H '\110'
1599 #define CHAR_I '\111'
1600 #define CHAR_J '\112'
1601 #define CHAR_K '\113'
1602 #define CHAR_L '\114'
1603 #define CHAR_M '\115'
1604 #define CHAR_N '\116'
1605 #define CHAR_O '\117'
1606 #define CHAR_P '\120'
1607 #define CHAR_Q '\121'
1608 #define CHAR_R '\122'
1609 #define CHAR_S '\123'
1610 #define CHAR_T '\124'
1611 #define CHAR_U '\125'
1612 #define CHAR_V '\126'
1613 #define CHAR_W '\127'
1614 #define CHAR_X '\130'
1615 #define CHAR_Y '\131'
1616 #define CHAR_Z '\132'
1617 #define CHAR_LEFT_SQUARE_BRACKET '\133'
1618 #define CHAR_BACKSLASH '\134'
1619 #define CHAR_RIGHT_SQUARE_BRACKET '\135'
1620 #define CHAR_CIRCUMFLEX_ACCENT '\136'
1621 #define CHAR_UNDERSCORE '\137'
1622 #define CHAR_GRAVE_ACCENT '\140'
1623 #define CHAR_a '\141'
1624 #define CHAR_b '\142'
1625 #define CHAR_c '\143'
1626 #define CHAR_d '\144'
1627 #define CHAR_e '\145'
1628 #define CHAR_f '\146'
1629 #define CHAR_g '\147'
1630 #define CHAR_h '\150'
1631 #define CHAR_i '\151'
1632 #define CHAR_j '\152'
1633 #define CHAR_k '\153'
1634 #define CHAR_l '\154'
1635 #define CHAR_m '\155'
1636 #define CHAR_n '\156'
1637 #define CHAR_o '\157'
1638 #define CHAR_p '\160'
1639 #define CHAR_q '\161'
1640 #define CHAR_r '\162'
1641 #define CHAR_s '\163'
1642 #define CHAR_t '\164'
1643 #define CHAR_u '\165'
1644 #define CHAR_v '\166'
1645 #define CHAR_w '\167'
1646 #define CHAR_x '\170'
1647 #define CHAR_y '\171'
1648 #define CHAR_z '\172'
1649 #define CHAR_LEFT_CURLY_BRACKET '\173'
1650 #define CHAR_VERTICAL_LINE '\174'
1651 #define CHAR_RIGHT_CURLY_BRACKET '\175'
1652 #define CHAR_TILDE '\176'
1653
1654 #define STR_HT "\011"
1655 #define STR_VT "\013"
1656 #define STR_FF "\014"
1657 #define STR_CR "\015"
1658 #define STR_NL "\012"
1659 #define STR_BS "\010"
1660 #define STR_BEL "\007"
1661 #define STR_ESC "\033"
1662 #define STR_DEL "\177"
1663
1664 #define STR_SPACE "\040"
1665 #define STR_EXCLAMATION_MARK "\041"
1666 #define STR_QUOTATION_MARK "\042"
1667 #define STR_NUMBER_SIGN "\043"
1668 #define STR_DOLLAR_SIGN "\044"
1669 #define STR_PERCENT_SIGN "\045"
1670 #define STR_AMPERSAND "\046"
1671 #define STR_APOSTROPHE "\047"
1672 #define STR_LEFT_PARENTHESIS "\050"
1673 #define STR_RIGHT_PARENTHESIS "\051"
1674 #define STR_ASTERISK "\052"
1675 #define STR_PLUS "\053"
1676 #define STR_COMMA "\054"
1677 #define STR_MINUS "\055"
1678 #define STR_DOT "\056"
1679 #define STR_SLASH "\057"
1680 #define STR_0 "\060"
1681 #define STR_1 "\061"
1682 #define STR_2 "\062"
1683 #define STR_3 "\063"
1684 #define STR_4 "\064"
1685 #define STR_5 "\065"
1686 #define STR_6 "\066"
1687 #define STR_7 "\067"
1688 #define STR_8 "\070"
1689 #define STR_9 "\071"
1690 #define STR_COLON "\072"
1691 #define STR_SEMICOLON "\073"
1692 #define STR_LESS_THAN_SIGN "\074"
1693 #define STR_EQUALS_SIGN "\075"
1694 #define STR_GREATER_THAN_SIGN "\076"
1695 #define STR_QUESTION_MARK "\077"
1696 #define STR_COMMERCIAL_AT "\100"
1697 #define STR_A "\101"
1698 #define STR_B "\102"
1699 #define STR_C "\103"
1700 #define STR_D "\104"
1701 #define STR_E "\105"
1702 #define STR_F "\106"
1703 #define STR_G "\107"
1704 #define STR_H "\110"
1705 #define STR_I "\111"
1706 #define STR_J "\112"
1707 #define STR_K "\113"
1708 #define STR_L "\114"
1709 #define STR_M "\115"
1710 #define STR_N "\116"
1711 #define STR_O "\117"
1712 #define STR_P "\120"
1713 #define STR_Q "\121"
1714 #define STR_R "\122"
1715 #define STR_S "\123"
1716 #define STR_T "\124"
1717 #define STR_U "\125"
1718 #define STR_V "\126"
1719 #define STR_W "\127"
1720 #define STR_X "\130"
1721 #define STR_Y "\131"
1722 #define STR_Z "\132"
1723 #define STR_LEFT_SQUARE_BRACKET "\133"
1724 #define STR_BACKSLASH "\134"
1725 #define STR_RIGHT_SQUARE_BRACKET "\135"
1726 #define STR_CIRCUMFLEX_ACCENT "\136"
1727 #define STR_UNDERSCORE "\137"
1728 #define STR_GRAVE_ACCENT "\140"
1729 #define STR_a "\141"
1730 #define STR_b "\142"
1731 #define STR_c "\143"
1732 #define STR_d "\144"
1733 #define STR_e "\145"
1734 #define STR_f "\146"
1735 #define STR_g "\147"
1736 #define STR_h "\150"
1737 #define STR_i "\151"
1738 #define STR_j "\152"
1739 #define STR_k "\153"
1740 #define STR_l "\154"
1741 #define STR_m "\155"
1742 #define STR_n "\156"
1743 #define STR_o "\157"
1744 #define STR_p "\160"
1745 #define STR_q "\161"
1746 #define STR_r "\162"
1747 #define STR_s "\163"
1748 #define STR_t "\164"
1749 #define STR_u "\165"
1750 #define STR_v "\166"
1751 #define STR_w "\167"
1752 #define STR_x "\170"
1753 #define STR_y "\171"
1754 #define STR_z "\172"
1755 #define STR_LEFT_CURLY_BRACKET "\173"
1756 #define STR_VERTICAL_LINE "\174"
1757 #define STR_RIGHT_CURLY_BRACKET "\175"
1758 #define STR_TILDE "\176"
1759
1760 #define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1761 #define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1762 #define STRING_F0 STR_F "\0"
1763 #define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0"
1764 #define STRING_MARK0 STR_M STR_A STR_R STR_K "\0"
1765 #define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0"
1766 #define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0"
1767 #define STRING_THEN STR_T STR_H STR_E STR_N
1768
1769 #define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0"
1770 #define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0"
1771 #define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0"
1772 #define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0"
1773 #define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0"
1774 #define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0"
1775 #define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0"
1776 #define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0"
1777 #define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0"
1778 #define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0"
1779 #define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0"
1780 #define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0"
1781 #define STRING_word0 STR_w STR_o STR_r STR_d "\0"
1782 #define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t
1783
1784 #define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E
1785
1786 #define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS
1787 #define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS
1788 #define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1789 #define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1790 #define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1791 #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
1792 #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
1793 #ifdef COMPILE_PCRE8
1794 #define STRING_UTF_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1795 #endif
1796 #ifdef COMPILE_PCRE16
1797 #define STRING_UTF_RIGHTPAR STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1798 #endif
1799 #ifdef COMPILE_PCRE32
1800 #define STRING_UTF_RIGHTPAR STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1801 #endif
1802 #define STRING_UCP_RIGHTPAR STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1803 #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
1804
1805 #endif /* SUPPORT_UTF */
1806
1807 /* Escape items that are just an encoding of a particular data value. */
1808
1809 #ifndef ESC_e
1810 #define ESC_e CHAR_ESC
1811 #endif
1812
1813 #ifndef ESC_f
1814 #define ESC_f CHAR_FF
1815 #endif
1816
1817 #ifndef ESC_n
1818 #define ESC_n CHAR_LF
1819 #endif
1820
1821 #ifndef ESC_r
1822 #define ESC_r CHAR_CR
1823 #endif
1824
1825 /* We can't officially use ESC_t because it is a POSIX reserved identifier
1826 (presumably because of all the others like size_t). */
1827
1828 #ifndef ESC_tee
1829 #define ESC_tee CHAR_HT
1830 #endif
1831
1832 /* Codes for different types of Unicode property */
1833
1834 #define PT_ANY 0 /* Any property - matches all chars */
1835 #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
1836 #define PT_GC 2 /* Specified general characteristic (e.g. L) */
1837 #define PT_PC 3 /* Specified particular characteristic (e.g. Lu) */
1838 #define PT_SC 4 /* Script (e.g. Han) */
1839 #define PT_ALNUM 5 /* Alphanumeric - the union of L and N */
1840 #define PT_SPACE 6 /* Perl space - Z plus 9,10,12,13 */
1841 #define PT_PXSPACE 7 /* POSIX space - Z plus 9,10,11,12,13 */
1842 #define PT_WORD 8 /* Word - L plus N plus underscore */
1843 #define PT_CLIST 9 /* Pseudo-property: match character list */
1844
1845 /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1846 contain characters with values greater than 255. */
1847
1848 #define XCL_NOT 0x01 /* Flag: this is a negative class */
1849 #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
1850
1851 #define XCL_END 0 /* Marks end of individual items */
1852 #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
1853 #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
1854 #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
1855 #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
1856
1857 /* These are escaped items that aren't just an encoding of a particular data
1858 value such as \n. They must have non-zero values, as check_escape() returns
1859 0 for a data character. Also, they must appear in the same order as in the opcode
1860 definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1861 corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1862 used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1863 non-DOTALL mode, "." behaves like \N.
1864
1865 The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1866 when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1867 They must be contiguous, and remain in order so that the replacements can be
1868 looked up from a table.
1869
1870 Negative numbers are used to encode a backreference (\1, \2, \3, etc.) in
1871 check_escape(). There are two tests in the code for an escape
1872 greater than ESC_b and less than ESC_Z to detect the types that may be
1873 repeated. These are the types that consume characters. If any new escapes are
1874 put in between that don't consume a character, that code will have to change.
1875 */
1876
1877 enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1878 ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1879 ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1880 ESC_E, ESC_Q, ESC_g, ESC_k,
1881 ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1882
1883 /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1884 OP_EOD must correspond in order to the list of escapes immediately above.
1885
1886 *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1887 that follow must also be updated to match. There are also tables called
1888 "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1889
1890 enum {
1891 OP_END, /* 0 End of pattern */
1892
1893 /* Values corresponding to backslashed metacharacters */
1894
1895 OP_SOD, /* 1 Start of data: \A */
1896 OP_SOM, /* 2 Start of match (subject + offset): \G */
1897 OP_SET_SOM, /* 3 Set start of match (\K) */
1898 OP_NOT_WORD_BOUNDARY, /* 4 \B */
1899 OP_WORD_BOUNDARY, /* 5 \b */
1900 OP_NOT_DIGIT, /* 6 \D */
1901 OP_DIGIT, /* 7 \d */
1902 OP_NOT_WHITESPACE, /* 8 \S */
1903 OP_WHITESPACE, /* 9 \s */
1904 OP_NOT_WORDCHAR, /* 10 \W */
1905 OP_WORDCHAR, /* 11 \w */
1906
1907 OP_ANY, /* 12 Match any character except newline (\N) */
1908 OP_ALLANY, /* 13 Match any character */
1909 OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1910 OP_NOTPROP, /* 15 \P (not Unicode property) */
1911 OP_PROP, /* 16 \p (Unicode property) */
1912 OP_ANYNL, /* 17 \R (any newline sequence) */
1913 OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */
1914 OP_HSPACE, /* 19 \h (horizontal whitespace) */
1915 OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */
1916 OP_VSPACE, /* 21 \v (vertical whitespace) */
1917 OP_EXTUNI, /* 22 \X (extended Unicode sequence */
1918 OP_EODN, /* 23 End of data or \n at end of data (\Z) */
1919 OP_EOD, /* 24 End of data (\z) */
1920
1921 OP_CIRC, /* 25 Start of line - not multiline */
1922 OP_CIRCM, /* 26 Start of line - multiline */
1923 OP_DOLL, /* 27 End of line - not multiline */
1924 OP_DOLLM, /* 28 End of line - multiline */
1925 OP_CHAR, /* 29 Match one character, casefully */
1926 OP_CHARI, /* 30 Match one character, caselessly */
1927 OP_NOT, /* 31 Match one character, not the given one, casefully */
1928 OP_NOTI, /* 32 Match one character, not the given one, caselessly */
1929
1930 /* The following sets of 13 opcodes must always be kept in step because
1931 the offset from the first one is used to generate the others. */
1932
1933 /**** Single characters, caseful, must precede the caseless ones ****/
1934
1935 OP_STAR, /* 33 The maximizing and minimizing versions of */
1936 OP_MINSTAR, /* 34 these six opcodes must come in pairs, with */
1937 OP_PLUS, /* 35 the minimizing one second. */
1938 OP_MINPLUS, /* 36 */
1939 OP_QUERY, /* 37 */
1940 OP_MINQUERY, /* 38 */
1941
1942 OP_UPTO, /* 39 From 0 to n matches of one character, caseful*/
1943 OP_MINUPTO, /* 40 */
1944 OP_EXACT, /* 41 Exactly n matches */
1945
1946 OP_POSSTAR, /* 42 Possessified star, caseful */
1947 OP_POSPLUS, /* 43 Possessified plus, caseful */
1948 OP_POSQUERY, /* 44 Posesssified query, caseful */
1949 OP_POSUPTO, /* 45 Possessified upto, caseful */
1950
1951 /**** Single characters, caseless, must follow the caseful ones */
1952
1953 OP_STARI, /* 46 */
1954 OP_MINSTARI, /* 47 */
1955 OP_PLUSI, /* 48 */
1956 OP_MINPLUSI, /* 49 */
1957 OP_QUERYI, /* 50 */
1958 OP_MINQUERYI, /* 51 */
1959
1960 OP_UPTOI, /* 52 From 0 to n matches of one character, caseless */
1961 OP_MINUPTOI, /* 53 */
1962 OP_EXACTI, /* 54 */
1963
1964 OP_POSSTARI, /* 55 Possessified star, caseless */
1965 OP_POSPLUSI, /* 56 Possessified plus, caseless */
1966 OP_POSQUERYI, /* 57 Posesssified query, caseless */
1967 OP_POSUPTOI, /* 58 Possessified upto, caseless */
1968
1969 /**** The negated ones must follow the non-negated ones, and match them ****/
1970 /**** Negated single character, caseful; must precede the caseless ones ****/
1971
1972 OP_NOTSTAR, /* 59 The maximizing and minimizing versions of */
1973 OP_NOTMINSTAR, /* 60 these six opcodes must come in pairs, with */
1974 OP_NOTPLUS, /* 61 the minimizing one second. They must be in */
1975 OP_NOTMINPLUS, /* 62 exactly the same order as those above. */
1976 OP_NOTQUERY, /* 63 */
1977 OP_NOTMINQUERY, /* 64 */
1978
1979 OP_NOTUPTO, /* 65 From 0 to n matches, caseful */
1980 OP_NOTMINUPTO, /* 66 */
1981 OP_NOTEXACT, /* 67 Exactly n matches */
1982
1983 OP_NOTPOSSTAR, /* 68 Possessified versions, caseful */
1984 OP_NOTPOSPLUS, /* 69 */
1985 OP_NOTPOSQUERY, /* 70 */
1986 OP_NOTPOSUPTO, /* 71 */
1987
1988 /**** Negated single character, caseless; must follow the caseful ones ****/
1989
1990 OP_NOTSTARI, /* 72 */
1991 OP_NOTMINSTARI, /* 73 */
1992 OP_NOTPLUSI, /* 74 */
1993 OP_NOTMINPLUSI, /* 75 */
1994 OP_NOTQUERYI, /* 76 */
1995 OP_NOTMINQUERYI, /* 77 */
1996
1997 OP_NOTUPTOI, /* 78 From 0 to n matches, caseless */
1998 OP_NOTMINUPTOI, /* 79 */
1999 OP_NOTEXACTI, /* 80 Exactly n matches */
2000
2001 OP_NOTPOSSTARI, /* 81 Possessified versions, caseless */
2002 OP_NOTPOSPLUSI, /* 82 */
2003 OP_NOTPOSQUERYI, /* 83 */
2004 OP_NOTPOSUPTOI, /* 84 */
2005
2006 /**** Character types ****/
2007
2008 OP_TYPESTAR, /* 85 The maximizing and minimizing versions of */
2009 OP_TYPEMINSTAR, /* 86 these six opcodes must come in pairs, with */
2010 OP_TYPEPLUS, /* 87 the minimizing one second. These codes must */
2011 OP_TYPEMINPLUS, /* 88 be in exactly the same order as those above. */
2012 OP_TYPEQUERY, /* 89 */
2013 OP_TYPEMINQUERY, /* 90 */
2014
2015 OP_TYPEUPTO, /* 91 From 0 to n matches */
2016 OP_TYPEMINUPTO, /* 92 */
2017 OP_TYPEEXACT, /* 93 Exactly n matches */
2018
2019 OP_TYPEPOSSTAR, /* 94 Possessified versions */
2020 OP_TYPEPOSPLUS, /* 95 */
2021 OP_TYPEPOSQUERY, /* 96 */
2022 OP_TYPEPOSUPTO, /* 97 */
2023
2024 /* These are used for character classes and back references; only the
2025 first six are the same as the sets above. */
2026
2027 OP_CRSTAR, /* 98 The maximizing and minimizing versions of */
2028 OP_CRMINSTAR, /* 99 all these opcodes must come in pairs, with */
2029 OP_CRPLUS, /* 100 the minimizing one second. These codes must */
2030 OP_CRMINPLUS, /* 101 be in exactly the same order as those above. */
2031 OP_CRQUERY, /* 102 */
2032 OP_CRMINQUERY, /* 103 */
2033
2034 OP_CRRANGE, /* 104 These are different to the three sets above. */
2035 OP_CRMINRANGE, /* 105 */
2036
2037 /* End of quantifier opcodes */
2038
2039 OP_CLASS, /* 106 Match a character class, chars < 256 only */
2040 OP_NCLASS, /* 107 Same, but the bitmap was created from a negative
2041 class - the difference is relevant only when a
2042 character > 255 is encountered. */
2043 OP_XCLASS, /* 108 Extended class for handling > 255 chars within the
2044 class. This does both positive and negative. */
2045 OP_REF, /* 109 Match a back reference, casefully */
2046 OP_REFI, /* 110 Match a back reference, caselessly */
2047 OP_RECURSE, /* 111 Match a numbered subpattern (possibly recursive) */
2048 OP_CALLOUT, /* 112 Call out to external function if provided */
2049
2050 OP_ALT, /* 113 Start of alternation */
2051 OP_KET, /* 114 End of group that doesn't have an unbounded repeat */
2052 OP_KETRMAX, /* 115 These two must remain together and in this */
2053 OP_KETRMIN, /* 116 order. They are for groups the repeat for ever. */
2054 OP_KETRPOS, /* 117 Possessive unlimited repeat. */
2055
2056 /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
2057 asserts must remain in order. */
2058
2059 OP_REVERSE, /* 118 Move pointer back - used in lookbehind assertions */
2060 OP_ASSERT, /* 119 Positive lookahead */
2061 OP_ASSERT_NOT, /* 120 Negative lookahead */
2062 OP_ASSERTBACK, /* 121 Positive lookbehind */
2063 OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
2064
2065 /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
2066 after the assertions, with ONCE first, as there's a test for >= ONCE for a
2067 subpattern that isn't an assertion. The POS versions must immediately follow
2068 the non-POS versions in each case. */
2069
2070 OP_ONCE, /* 123 Atomic group, contains captures */
2071 OP_ONCE_NC, /* 124 Atomic group containing no captures */
2072 OP_BRA, /* 125 Start of non-capturing bracket */
2073 OP_BRAPOS, /* 126 Ditto, with unlimited, possessive repeat */
2074 OP_CBRA, /* 127 Start of capturing bracket */
2075 OP_CBRAPOS, /* 128 Ditto, with unlimited, possessive repeat */
2076 OP_COND, /* 129 Conditional group */
2077
2078 /* These five must follow the previous five, in the same order. There's a
2079 check for >= SBRA to distinguish the two sets. */
2080
2081 OP_SBRA, /* 130 Start of non-capturing bracket, check empty */
2082 OP_SBRAPOS, /* 131 Ditto, with unlimited, possessive repeat */
2083 OP_SCBRA, /* 132 Start of capturing bracket, check empty */
2084 OP_SCBRAPOS, /* 133 Ditto, with unlimited, possessive repeat */
2085 OP_SCOND, /* 134 Conditional group, check empty */
2086
2087 /* The next two pairs must (respectively) be kept together. */
2088
2089 OP_CREF, /* 135 Used to hold a capture number as condition */
2090 OP_NCREF, /* 136 Same, but generated by a name reference*/
2091 OP_RREF, /* 137 Used to hold a recursion number as condition */
2092 OP_NRREF, /* 138 Same, but generated by a name reference*/
2093 OP_DEF, /* 139 The DEFINE condition */
2094
2095 OP_BRAZERO, /* 140 These two must remain together and in this */
2096 OP_BRAMINZERO, /* 141 order. */
2097 OP_BRAPOSZERO, /* 142 */
2098
2099 /* These are backtracking control verbs */
2100
2101 OP_MARK, /* 143 always has an argument */
2102 OP_PRUNE, /* 144 */
2103 OP_PRUNE_ARG, /* 145 same, but with argument */
2104 OP_SKIP, /* 146 */
2105 OP_SKIP_ARG, /* 147 same, but with argument */
2106 OP_THEN, /* 148 */
2107 OP_THEN_ARG, /* 149 same, but with argument */
2108 OP_COMMIT, /* 150 */
2109
2110 /* These are forced failure and success verbs */
2111
2112 OP_FAIL, /* 151 */
2113 OP_ACCEPT, /* 152 */
2114 OP_ASSERT_ACCEPT, /* 153 Used inside assertions */
2115 OP_CLOSE, /* 154 Used before OP_ACCEPT to close open captures */
2116
2117 /* This is used to skip a subpattern with a {0} quantifier */
2118
2119 OP_SKIPZERO, /* 155 */
2120
2121 /* This is not an opcode, but is used to check that tables indexed by opcode
2122 are the correct length, in order to catch updating errors - there have been
2123 some in the past. */
2124
2125 OP_TABLE_LENGTH
2126 };
2127
2128 /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2129 definitions that follow must also be updated to match. There are also tables
2130 called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
2131
2132
2133 /* This macro defines textual names for all the opcodes. These are used only
2134 for debugging, and some of them are only partial names. The macro is referenced
2135 only in pcre_printint.c, which fills out the full names in many cases (and in
2136 some cases doesn't actually use these names at all). */
2137
2138 #define OP_NAME_LIST \
2139 "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \
2140 "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \
2141 "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \
2142 "extuni", "\\Z", "\\z", \
2143 "^", "^", "$", "$", "char", "chari", "not", "noti", \
2144 "*", "*?", "+", "+?", "?", "??", \
2145 "{", "{", "{", \
2146 "*+","++", "?+", "{", \
2147 "*", "*?", "+", "+?", "?", "??", \
2148 "{", "{", "{", \
2149 "*+","++", "?+", "{", \
2150 "*", "*?", "+", "+?", "?", "??", \
2151 "{", "{", "{", \
2152 "*+","++", "?+", "{", \
2153 "*", "*?", "+", "+?", "?", "??", \
2154 "{", "{", "{", \
2155 "*+","++", "?+", "{", \
2156 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
2157 "*+","++", "?+", "{", \
2158 "*", "*?", "+", "+?", "?", "??", "{", "{", \
2159 "class", "nclass", "xclass", "Ref", "Refi", \
2160 "Recurse", "Callout", \
2161 "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos", \
2162 "Reverse", "Assert", "Assert not", "AssertB", "AssertB not", \
2163 "Once", "Once_NC", \
2164 "Bra", "BraPos", "CBra", "CBraPos", \
2165 "Cond", \
2166 "SBra", "SBraPos", "SCBra", "SCBraPos", \
2167 "SCond", \
2168 "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def", \
2169 "Brazero", "Braminzero", "Braposzero", \
2170 "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP", \
2171 "*THEN", "*THEN", "*COMMIT", "*FAIL", \
2172 "*ACCEPT", "*ASSERT_ACCEPT", \
2173 "Close", "Skip zero"
2174
2175
2176 /* This macro defines the length of fixed length operations in the compiled
2177 regex. The lengths are used when searching for specific things, and also in the
2178 debugging printing of a compiled regex. We use a macro so that it can be
2179 defined close to the definitions of the opcodes themselves.
2180
2181 As things have been extended, some of these are no longer fixed lenths, but are
2182 minima instead. For example, the length of a single-character repeat may vary
2183 in UTF-8 mode. The code that uses this table must know about such things. */
2184
2185 #define OP_LENGTHS \
2186 1, /* End */ \
2187 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \
2188 1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \
2189 1, 1, 1, /* Any, AllAny, Anybyte */ \
2190 3, 3, /* \P, \p */ \
2191 1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \
2192 1, /* \X */ \
2193 1, 1, 1, 1, 1, 1, /* \Z, \z, ^, ^M, $, $M */ \
2194 2, /* Char - the minimum length */ \
2195 2, /* Chari - the minimum length */ \
2196 2, /* not */ \
2197 2, /* noti */ \
2198 /* Positive single-char repeats ** These are */ \
2199 2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
2200 2+IMM2_SIZE, 2+IMM2_SIZE, /* upto, minupto ** mode */ \
2201 2+IMM2_SIZE, /* exact */ \
2202 2, 2, 2, 2+IMM2_SIZE, /* *+, ++, ?+, upto+ */ \
2203 2, 2, 2, 2, 2, 2, /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8 */ \
2204 2+IMM2_SIZE, 2+IMM2_SIZE, /* upto I, minupto I */ \
2205 2+IMM2_SIZE, /* exact I */ \
2206 2, 2, 2, 2+IMM2_SIZE, /* *+I, ++I, ?+I, upto+I */ \
2207 /* Negative single-char repeats - only for chars < 256 */ \
2208 2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
2209 2+IMM2_SIZE, 2+IMM2_SIZE, /* NOT upto, minupto */ \
2210 2+IMM2_SIZE, /* NOT exact */ \
2211 2, 2, 2, 2+IMM2_SIZE, /* Possessive NOT *, +, ?, upto */ \
2212 2, 2, 2, 2, 2, 2, /* NOT *I, *?I, +I, +?I, ?I, ??I */ \
2213 2+IMM2_SIZE, 2+IMM2_SIZE, /* NOT upto I, minupto I */ \
2214 2+IMM2_SIZE, /* NOT exact I */ \
2215 2, 2, 2, 2+IMM2_SIZE, /* Possessive NOT *I, +I, ?I, upto I */ \
2216 /* Positive type repeats */ \
2217 2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
2218 2+IMM2_SIZE, 2+IMM2_SIZE, /* Type upto, minupto */ \
2219 2+IMM2_SIZE, /* Type exact */ \
2220 2, 2, 2, 2+IMM2_SIZE, /* Possessive *+, ++, ?+, upto+ */ \
2221 /* Character class & ref repeats */ \
2222 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
2223 1+2*IMM2_SIZE, 1+2*IMM2_SIZE, /* CRRANGE, CRMINRANGE */ \
2224 1+(32/sizeof(pcre_uchar)), /* CLASS */ \
2225 1+(32/sizeof(pcre_uchar)), /* NCLASS */ \
2226 0, /* XCLASS - variable length */ \
2227 1+IMM2_SIZE, /* REF */ \
2228 1+IMM2_SIZE, /* REFI */ \
2229 1+LINK_SIZE, /* RECURSE */ \
2230 2+2*LINK_SIZE, /* CALLOUT */ \
2231 1+LINK_SIZE, /* Alt */ \
2232 1+LINK_SIZE, /* Ket */ \
2233 1+LINK_SIZE, /* KetRmax */ \
2234 1+LINK_SIZE, /* KetRmin */ \
2235 1+LINK_SIZE, /* KetRpos */ \
2236 1+LINK_SIZE, /* Reverse */ \
2237 1+LINK_SIZE, /* Assert */ \
2238 1+LINK_SIZE, /* Assert not */ \
2239 1+LINK_SIZE, /* Assert behind */ \
2240 1+LINK_SIZE, /* Assert behind not */ \
2241 1+LINK_SIZE, /* ONCE */ \
2242 1+LINK_SIZE, /* ONCE_NC */ \
2243 1+LINK_SIZE, /* BRA */ \
2244 1+LINK_SIZE, /* BRAPOS */ \
2245 1+LINK_SIZE+IMM2_SIZE, /* CBRA */ \
2246 1+LINK_SIZE+IMM2_SIZE, /* CBRAPOS */ \
2247 1+LINK_SIZE, /* COND */ \
2248 1+LINK_SIZE, /* SBRA */ \
2249 1+LINK_SIZE, /* SBRAPOS */ \
2250 1+LINK_SIZE+IMM2_SIZE, /* SCBRA */ \
2251 1+LINK_SIZE+IMM2_SIZE, /* SCBRAPOS */ \
2252 1+LINK_SIZE, /* SCOND */ \
2253 1+IMM2_SIZE, 1+IMM2_SIZE, /* CREF, NCREF */ \
2254 1+IMM2_SIZE, 1+IMM2_SIZE, /* RREF, NRREF */ \
2255 1, /* DEF */ \
2256 1, 1, 1, /* BRAZERO, BRAMINZERO, BRAPOSZERO */ \
2257 3, 1, 3, /* MARK, PRUNE, PRUNE_ARG */ \
2258 1, 3, /* SKIP, SKIP_ARG */ \
2259 1, 3, /* THEN, THEN_ARG */ \
2260 1, 1, 1, 1, /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT */ \
2261 1+IMM2_SIZE, 1 /* CLOSE, SKIPZERO */
2262
2263 /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
2264 condition. */
2265
2266 #define RREF_ANY 0xffff
2267
2268 /* Compile time error code numbers. They are given names so that they can more
2269 easily be tracked. When a new number is added, the table called eint in
2270 pcreposix.c must be updated. */
2271
2272 enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
2273 ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
2274 ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
2275 ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2276 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2277 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2278 ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2279 ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2280
2281 /* JIT compiling modes. The function list is indexed by them. */
2282 enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2283 JIT_NUMBER_OF_COMPILE_MODES };
2284
2285 /* The real format of the start of the pcre block; the index of names and the
2286 code vector run on as long as necessary after the end. We store an explicit
2287 offset to the name table so that if a regex is compiled on one host, saved, and
2288 then run on another where the size of pointers is different, all might still
2289 be well. For the case of compiled-on-4 and run-on-8, we include an extra
2290 pointer that is always NULL. For future-proofing, a few dummy fields were
2291 originally included - even though you can never get this planning right - but
2292 there is only one left now.
2293
2294 NOTE NOTE NOTE:
2295 Because people can now save and re-use compiled patterns, any additions to this
2296 structure should be made at the end, and something earlier (e.g. a new
2297 flag in the options or one of the dummy fields) should indicate that the new
2298 fields are present. Currently PCRE always sets the dummy fields to zero.
2299 NOTE NOTE NOTE
2300 */
2301
2302 #if defined COMPILE_PCRE8
2303 #define REAL_PCRE real_pcre
2304 #elif defined COMPILE_PCRE16
2305 #define REAL_PCRE real_pcre16
2306 #elif defined COMPILE_PCRE32
2307 #define REAL_PCRE real_pcre32
2308 #endif
2309
2310 /* It is necessary to fork the struct for 32 bit, since it needs to use
2311 * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2312 * typedef since pcretest needs access to the struct of the 8-, 16-
2313 * and 32-bit variants. */
2314
2315 typedef struct real_pcre8_or_16 {
2316 pcre_uint32 magic_number;
2317 pcre_uint32 size; /* Total that was malloced */
2318 pcre_uint32 options; /* Public options */
2319 pcre_uint16 flags; /* Private flags */
2320 pcre_uint16 max_lookbehind; /* Longest lookbehind (characters) */
2321 pcre_uint16 top_bracket; /* Highest numbered group */
2322 pcre_uint16 top_backref; /* Highest numbered back reference */
2323 pcre_uint16 first_char; /* Starting character */
2324 pcre_uint16 req_char; /* This character must be seen */
2325 pcre_uint16 name_table_offset; /* Offset to name table that follows */
2326 pcre_uint16 name_entry_size; /* Size of any name items */
2327 pcre_uint16 name_count; /* Number of name items */
2328 pcre_uint16 ref_count; /* Reference count */
2329 const pcre_uint8 *tables; /* Pointer to tables or NULL for std */
2330 const pcre_uint8 *nullpad; /* NULL padding */
2331 } real_pcre8_or_16;
2332
2333 typedef struct real_pcre8_or_16 real_pcre;
2334 typedef struct real_pcre8_or_16 real_pcre16;
2335
2336 typedef struct real_pcre32 {
2337 pcre_uint32 magic_number;
2338 pcre_uint32 size; /* Total that was malloced */
2339 pcre_uint32 options; /* Public options */
2340 pcre_uint16 flags; /* Private flags */
2341 pcre_uint16 max_lookbehind; /* Longest lookbehind (characters) */
2342 pcre_uint16 top_bracket; /* Highest numbered group */
2343 pcre_uint16 top_backref; /* Highest numbered back reference */
2344 pcre_uint32 first_char; /* Starting character */
2345 pcre_uint32 req_char; /* This character must be seen */
2346 pcre_uint16 name_table_offset; /* Offset to name table that follows */
2347 pcre_uint16 name_entry_size; /* Size of any name items */
2348 pcre_uint16 name_count; /* Number of name items */
2349 pcre_uint16 ref_count; /* Reference count */
2350 pcre_uint16 dummy1; /* for later expansion */
2351 pcre_uint16 dummy2; /* for later expansion */
2352 const pcre_uint8 *tables; /* Pointer to tables or NULL for std */
2353 void *nullpad; /* for later expansion */
2354 } real_pcre32;
2355
2356 /* Assert that the size of REAL_PCRE is divisible by 8 */
2357 typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2358
2359 /* Needed in pcretest to access some fields in the real_pcre* structures
2360 * directly. They're unified for 8/16/32 bits since the structs only differ
2361 * after these fields; if that ever changes, need to fork those defines into
2362 * 8/16 and 32 bit versions. */
2363 #define REAL_PCRE_MAGIC(re) (((REAL_PCRE*)re)->magic_number)
2364 #define REAL_PCRE_SIZE(re) (((REAL_PCRE*)re)->size)
2365 #define REAL_PCRE_OPTIONS(re) (((REAL_PCRE*)re)->options)
2366 #define REAL_PCRE_FLAGS(re) (((REAL_PCRE*)re)->flags)
2367
2368 /* The format of the block used to store data from pcre_study(). The same
2369 remark (see NOTE above) about extending this structure applies. */
2370
2371 typedef struct pcre_study_data {
2372 pcre_uint32 size; /* Total that was malloced */
2373 pcre_uint32 flags; /* Private flags */
2374 pcre_uint8 start_bits[32]; /* Starting char bits */
2375 pcre_uint32 minlength; /* Minimum subject length */
2376 } pcre_study_data;
2377
2378 /* Structure for building a chain of open capturing subpatterns during
2379 compiling, so that instructions to close them can be compiled when (*ACCEPT) is
2380 encountered. This is also used to identify subpatterns that contain recursive
2381 back references to themselves, so that they can be made atomic. */
2382
2383 typedef struct open_capitem {
2384 struct open_capitem *next; /* Chain link */
2385 pcre_uint16 number; /* Capture number */
2386 pcre_uint16 flag; /* Set TRUE if recursive back ref */
2387 } open_capitem;
2388
2389 /* Structure for passing "static" information around between the functions
2390 doing the compiling, so that they are thread-safe. */
2391
2392 typedef struct compile_data {
2393 const pcre_uint8 *lcc; /* Points to lower casing table */
2394 const pcre_uint8 *fcc; /* Points to case-flipping table */
2395 const pcre_uint8 *cbits; /* Points to character type table */
2396 const pcre_uint8 *ctypes; /* Points to table of type maps */
2397 const pcre_uchar *start_workspace;/* The start of working space */
2398 const pcre_uchar *start_code; /* The start of the compiled code */
2399 const pcre_uchar *start_pattern; /* The start of the pattern */
2400 const pcre_uchar *end_pattern; /* The end of the pattern */
2401 open_capitem *open_caps; /* Chain of open capture items */
2402 pcre_uchar *hwm; /* High watermark of workspace */
2403 pcre_uchar *name_table; /* The name/number table */
2404 int names_found; /* Number of entries so far */
2405 int name_entry_size; /* Size of each entry */
2406 int workspace_size; /* Size of workspace */
2407 unsigned int bracount; /* Count of capturing parens as we compile */
2408 int final_bracount; /* Saved value after first pass */
2409 int max_lookbehind; /* Maximum lookbehind (characters) */
2410 int top_backref; /* Maximum back reference */
2411 unsigned int backref_map; /* Bitmap of low back refs */
2412 int assert_depth; /* Depth of nested assertions */
2413 int external_options; /* External (initial) options */
2414 int external_flags; /* External flag bits to be set */
2415 int req_varyopt; /* "After variable item" flag for reqbyte */
2416 BOOL had_accept; /* (*ACCEPT) encountered */
2417 BOOL had_pruneorskip; /* (*PRUNE) or (*SKIP) encountered */
2418 BOOL check_lookbehind; /* Lookbehinds need later checking */
2419 int nltype; /* Newline type */
2420 int nllen; /* Newline string length */
2421 pcre_uchar nl[4]; /* Newline string when fixed length */
2422 } compile_data;
2423
2424 /* Structure for maintaining a chain of pointers to the currently incomplete
2425 branches, for testing for left recursion while compiling. */
2426
2427 typedef struct branch_chain {
2428 struct branch_chain *outer;
2429 pcre_uchar *current_branch;
2430 } branch_chain;
2431
2432 /* Structure for items in a linked list that represents an explicit recursive
2433 call within the pattern; used by pcre_exec(). */
2434
2435 typedef struct recursion_info {
2436 struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2437 unsigned int group_num; /* Number of group that was called */
2438 int *offset_save; /* Pointer to start of saved offsets */
2439 int saved_max; /* Number of saved offsets */
2440 PCRE_PUCHAR subject_position; /* Position at start of recursion */
2441 } recursion_info;
2442
2443 /* A similar structure for pcre_dfa_exec(). */
2444
2445 typedef struct dfa_recursion_info {
2446 struct dfa_recursion_info *prevrec;
2447 int group_num;
2448 PCRE_PUCHAR subject_position;
2449 } dfa_recursion_info;
2450
2451 /* Structure for building a chain of data for holding the values of the subject
2452 pointer at the start of each subpattern, so as to detect when an empty string
2453 has been matched by a subpattern - to break infinite loops; used by
2454 pcre_exec(). */
2455
2456 typedef struct eptrblock {
2457 struct eptrblock *epb_prev;
2458 PCRE_PUCHAR epb_saved_eptr;
2459 } eptrblock;
2460
2461
2462 /* Structure for passing "static" information around between the functions
2463 doing traditional NFA matching, so that they are thread-safe. */
2464
2465 typedef struct match_data {
2466 unsigned long int match_call_count; /* As it says */
2467 unsigned long int match_limit; /* As it says */
2468 unsigned long int match_limit_recursion; /* As it says */
2469 int *offset_vector; /* Offset vector */
2470 int offset_end; /* One past the end */
2471 int offset_max; /* The maximum usable for return data */
2472 int nltype; /* Newline type */
2473 int nllen; /* Newline string length */
2474 int name_count; /* Number of names in name table */
2475 int name_entry_size; /* Size of entry in names table */
2476 pcre_uchar *name_table; /* Table of names */
2477 pcre_uchar nl[4]; /* Newline string when fixed */
2478 const pcre_uint8 *lcc; /* Points to lower casing table */
2479 const pcre_uint8 *fcc; /* Points to case-flipping table */
2480 const pcre_uint8 *ctypes; /* Points to table of type maps */
2481 BOOL offset_overflow; /* Set if too many extractions */
2482 BOOL notbol; /* NOTBOL flag */
2483 BOOL noteol; /* NOTEOL flag */
2484 BOOL utf; /* UTF-8 / UTF-16 flag */
2485 BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */
2486 BOOL use_ucp; /* PCRE_UCP flag */
2487 BOOL endonly; /* Dollar not before final \n */
2488 BOOL notempty; /* Empty string match not wanted */
2489 BOOL notempty_atstart; /* Empty string match at start not wanted */
2490 BOOL hitend; /* Hit the end of the subject at some point */
2491 BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
2492 BOOL hasthen; /* Pattern contains (*THEN) */
2493 BOOL ignore_skip_arg; /* For re-run when SKIP name not found */
2494 const pcre_uchar *start_code; /* For use when recursing */
2495 PCRE_PUCHAR start_subject; /* Start of the subject string */
2496 PCRE_PUCHAR end_subject; /* End of the subject string */
2497 PCRE_PUCHAR start_match_ptr; /* Start of matched string */
2498 PCRE_PUCHAR end_match_ptr; /* Subject position at end match */
2499 PCRE_PUCHAR start_used_ptr; /* Earliest consulted character */
2500 int partial; /* PARTIAL options */
2501 int end_offset_top; /* Highwater mark at end of match */
2502 int capture_last; /* Most recent capture number */
2503 int start_offset; /* The start offset value */
2504 int match_function_type; /* Set for certain special calls of MATCH() */
2505 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
2506 int eptrn; /* Next free eptrblock */
2507 recursion_info *recursive; /* Linked list of recursion data */
2508 void *callout_data; /* To pass back to callouts */
2509 const pcre_uchar *mark; /* Mark pointer to pass back on success */
2510 const pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2511 const pcre_uchar *once_target; /* Where to back up to for atomic groups */
2512 #ifdef NO_RECURSE
2513 void *match_frames_base; /* For remembering malloc'd frames */
2514 #endif
2515 } match_data;
2516
2517 /* A similar structure is used for the same purpose by the DFA matching
2518 functions. */
2519
2520 typedef struct dfa_match_data {
2521 const pcre_uchar *start_code; /* Start of the compiled pattern */
2522 const pcre_uchar *start_subject ; /* Start of the subject string */
2523 const pcre_uchar *end_subject; /* End of subject string */
2524 const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2525 const pcre_uint8 *tables; /* Character tables */
2526 int start_offset; /* The start offset value */
2527 int moptions; /* Match options */
2528 int poptions; /* Pattern options */
2529 int nltype; /* Newline type */
2530 int nllen; /* Newline string length */
2531 pcre_uchar nl[4]; /* Newline string when fixed */
2532 void *callout_data; /* To pass back to callouts */
2533 dfa_recursion_info *recursive; /* Linked list of recursion data */
2534 } dfa_match_data;
2535
2536 /* Bit definitions for entries in the pcre_ctypes table. */
2537
2538 #define ctype_space 0x01
2539 #define ctype_letter 0x02
2540 #define ctype_digit 0x04
2541 #define ctype_xdigit 0x08
2542 #define ctype_word 0x10 /* alphanumeric or '_' */
2543 #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
2544
2545 /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
2546 of bits for a class map. Some classes are built by combining these tables. */
2547
2548 #define cbit_space 0 /* [:space:] or \s */
2549 #define cbit_xdigit 32 /* [:xdigit:] */
2550 #define cbit_digit 64 /* [:digit:] or \d */
2551 #define cbit_upper 96 /* [:upper:] */
2552 #define cbit_lower 128 /* [:lower:] */
2553 #define cbit_word 160 /* [:word:] or \w */
2554 #define cbit_graph 192 /* [:graph:] */
2555 #define cbit_print 224 /* [:print:] */
2556 #define cbit_punct 256 /* [:punct:] */
2557 #define cbit_cntrl 288 /* [:cntrl:] */
2558 #define cbit_length 320 /* Length of the cbits table */
2559
2560 /* Offsets of the various tables from the base tables pointer, and
2561 total length. */
2562
2563 #define lcc_offset 0
2564 #define fcc_offset 256
2565 #define cbits_offset 512
2566 #define ctypes_offset (cbits_offset + cbit_length)
2567 #define tables_length (ctypes_offset + 256)
2568
2569 /* Internal function and data prefixes. */
2570
2571 #if defined COMPILE_PCRE8
2572 #ifndef PUBL
2573 #define PUBL(name) pcre_##name
2574 #endif
2575 #ifndef PRIV
2576 #define PRIV(name) _pcre_##name
2577 #endif
2578 #elif defined COMPILE_PCRE16
2579 #ifndef PUBL
2580 #define PUBL(name) pcre16_##name
2581 #endif
2582 #ifndef PRIV
2583 #define PRIV(name) _pcre16_##name
2584 #endif
2585 #elif defined COMPILE_PCRE32
2586 #ifndef PUBL
2587 #define PUBL(name) pcre32_##name
2588 #endif
2589 #ifndef PRIV
2590 #define PRIV(name) _pcre32_##name
2591 #endif
2592 #else
2593 #error Unsupported compiling mode
2594 #endif /* COMPILE_PCRE[8|16|32] */
2595
2596 /* Layout of the UCP type table that translates property names into types and
2597 codes. Each entry used to point directly to a name, but to reduce the number of
2598 relocations in shared libraries, it now has an offset into a single string
2599 instead. */
2600
2601 typedef struct {
2602 pcre_uint16 name_offset;
2603 pcre_uint16 type;
2604 pcre_uint16 value;
2605 } ucp_type_table;
2606
2607
2608 /* Internal shared data tables. These are tables that are used by more than one
2609 of the exported public functions. They have to be "external" in the C sense,
2610 but are not part of the PCRE public API. The data for these tables is in the
2611 pcre_tables.c module. */
2612
2613 #ifdef COMPILE_PCRE8
2614 extern const int PRIV(utf8_table1)[];
2615 extern const int PRIV(utf8_table1_size);
2616 extern const int PRIV(utf8_table2)[];
2617 extern const int PRIV(utf8_table3)[];
2618 extern const pcre_uint8 PRIV(utf8_table4)[];
2619 #endif /* COMPILE_PCRE8 */
2620
2621 extern const char PRIV(utt_names)[];
2622 extern const ucp_type_table PRIV(utt)[];
2623 extern const int PRIV(utt_size);
2624
2625 extern const pcre_uint8 PRIV(OP_lengths)[];
2626 extern const pcre_uint8 PRIV(default_tables)[];
2627
2628 extern const pcre_uint32 PRIV(hspace_list)[];
2629 extern const pcre_uint32 PRIV(vspace_list)[];
2630
2631
2632 /* Internal shared functions. These are functions that are used by more than
2633 one of the exported public functions. They have to be "external" in the C
2634 sense, but are not part of the PCRE public API. */
2635
2636 /* String comparison functions. */
2637 #if defined COMPILE_PCRE8
2638
2639 #define STRCMP_UC_UC(str1, str2) \
2640 strcmp((char *)(str1), (char *)(str2))
2641 #define STRCMP_UC_C8(str1, str2) \
2642 strcmp((char *)(str1), (str2))
2643 #define STRNCMP_UC_UC(str1, str2, num) \
2644 strncmp((char *)(str1), (char *)(str2), (num))
2645 #define STRNCMP_UC_C8(str1, str2, num) \
2646 strncmp((char *)(str1), (str2), (num))
2647 #define STRLEN_UC(str) strlen((const char *)str)
2648
2649 #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2650
2651 extern int PRIV(strcmp_uc_uc)(const pcre_uchar *,
2652 const pcre_uchar *);
2653 extern int PRIV(strcmp_uc_c8)(const pcre_uchar *,
2654 const char *);
2655 extern int PRIV(strncmp_uc_uc)(const pcre_uchar *,
2656 const pcre_uchar *, unsigned int num);
2657 extern int PRIV(strncmp_uc_c8)(const pcre_uchar *,
2658 const char *, unsigned int num);
2659 extern unsigned int PRIV(strlen_uc)(const pcre_uchar *str);
2660
2661 #define STRCMP_UC_UC(str1, str2) \
2662 PRIV(strcmp_uc_uc)((str1), (str2))
2663 #define STRCMP_UC_C8(str1, str2) \
2664 PRIV(strcmp_uc_c8)((str1), (str2))
2665 #define STRNCMP_UC_UC(str1, str2, num) \
2666 PRIV(strncmp_uc_uc)((str1), (str2), (num))
2667 #define STRNCMP_UC_C8(str1, str2, num) \
2668 PRIV(strncmp_uc_c8)((str1), (str2), (num))
2669 #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2670
2671 #endif /* COMPILE_PCRE[8|16|32] */
2672
2673 #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2674
2675 #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2676 #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2677
2678 #elif defined COMPILE_PCRE32
2679
2680 extern int PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2681 const pcre_uchar *);
2682 extern int PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2683 const char *);
2684
2685 #define STRCMP_UC_UC_TEST(str1, str2) \
2686 (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2687 #define STRCMP_UC_C8_TEST(str1, str2) \
2688 (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2689
2690 #endif /* COMPILE_PCRE[8|16|32] */
2691
2692 extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2693 extern BOOL PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2694 int *, BOOL);
2695 extern unsigned int PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2696 extern int PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2697 extern BOOL PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2698 int *, BOOL);
2699 extern BOOL PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2700
2701 #ifdef SUPPORT_JIT
2702 extern void PRIV(jit_compile)(const REAL_PCRE *,
2703 PUBL(extra) *, int);
2704 extern int PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2705 const pcre_uchar *, int, int, int, int *, int);
2706 extern void PRIV(jit_free)(void *);
2707 extern int PRIV(jit_get_size)(void *);
2708 extern const char* PRIV(jit_get_target)(void);
2709 #endif
2710
2711 /* Unicode character database (UCD) */
2712
2713 typedef struct {
2714 pcre_uint8 script; /* ucp_Arabic, etc. */
2715 pcre_uint8 chartype; /* ucp_Cc, etc. (general categories) */
2716 pcre_uint8 gbprop; /* ucp_gbControl, etc. (grapheme break property) */
2717 pcre_uint8 caseset; /* offset to multichar other cases or zero */
2718 pcre_int32 other_case; /* offset to other case, or zero if none */
2719 } ucd_record;
2720
2721 extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2722 extern const ucd_record PRIV(ucd_records)[];
2723 extern const pcre_uint8 PRIV(ucd_stage1)[];
2724 extern const pcre_uint16 PRIV(ucd_stage2)[];
2725 extern const pcre_uint32 PRIV(ucp_gentype)[];
2726 extern const pcre_uint32 PRIV(ucp_gbtable)[];
2727 #ifdef SUPPORT_JIT
2728 extern const int PRIV(ucp_typerange)[];
2729 #endif
2730
2731 #ifdef SUPPORT_UCP
2732 /* UCD access macros */
2733
2734 #define UCD_BLOCK_SIZE 128
2735 #define GET_UCD(ch) (PRIV(ucd_records) + \
2736 PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2737 UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2738
2739 #define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype
2740 #define UCD_SCRIPT(ch) GET_UCD(ch)->script
2741 #define UCD_CATEGORY(ch) PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2742 #define UCD_GRAPHBREAK(ch) GET_UCD(ch)->gbprop
2743 #define UCD_CASESET(ch) GET_UCD(ch)->caseset
2744 #define UCD_OTHERCASE(ch) ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2745
2746 #endif /* SUPPORT_UCP */
2747
2748 #endif
2749
2750 /* End of pcre_internal.h */

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5