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

Contents of /code/trunk/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1075 - (show annotations)
Tue Oct 16 15:54:51 2012 UTC (7 years ago) by chpe
File MIME type: text/plain
File size: 105952 byte(s)
Error occurred while calculating annotation data.
pcre32: compile: Fix signed/unsigned mismatch in GET_UCD
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 #if !defined COMPILE_PCRE16 && ! defined COMPILE_PCRE32
57 #define COMPILE_PCRE8
58 #endif
59
60 /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
61 "configure" script ensures this, but not everybody uses "configure". */
62
63 #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
64 #define SUPPORT_UTF 1
65 #endif
66
67 /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
68 reasons with existing code. */
69
70 #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
71 #define SUPPORT_UTF 1
72 #endif
73
74 /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
75 Until then we define it if SUPPORT_UTF is defined. */
76
77 #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
78 #define SUPPORT_UTF8 1
79 #endif
80
81 /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
82 script prevents both being selected, but not everybody uses "configure". */
83
84 #if defined EBCDIC && defined SUPPORT_UTF
85 #error The use of both EBCDIC and SUPPORT_UTF is not supported.
86 #endif
87
88 /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
89 inline, and there are *still* stupid compilers about that don't like indented
90 pre-processor statements, or at least there were when I first wrote this. After
91 all, it had only been about 10 years then...
92
93 It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so
94 be absolutely sure we get our version. */
95
96 #undef DPRINTF
97 #ifdef PCRE_DEBUG
98 #define DPRINTF(p) printf p
99 #else
100 #define DPRINTF(p) /* Nothing */
101 #endif
102
103
104 /* Standard C headers plus the external interface definition. The only time
105 setjmp and stdarg are used is when NO_RECURSE is set. */
106
107 #include <ctype.h>
108 #include <limits.h>
109 #include <stddef.h>
110 #include <stdio.h>
111 #include <stdlib.h>
112 #include <string.h>
113
114 /* When compiling a DLL for Windows, the exported symbols have to be declared
115 using some MS magic. I found some useful information on this web page:
116 http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
117 information there, using __declspec(dllexport) without "extern" we have a
118 definition; with "extern" we have a declaration. The settings here override the
119 setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL,
120 which is all that is needed for applications (they just import the symbols). We
121 use:
122
123 PCRE_EXP_DECL for declarations
124 PCRE_EXP_DEFN for definitions of exported functions
125 PCRE_EXP_DATA_DEFN for definitions of exported variables
126
127 The reason for the two DEFN macros is that in non-Windows environments, one
128 does not want to have "extern" before variable definitions because it leads to
129 compiler warnings. So we distinguish between functions and variables. In
130 Windows, the two should always be the same.
131
132 The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest,
133 which is an application, but needs to import this file in order to "peek" at
134 internals, can #include pcre.h first to get an application's-eye view.
135
136 In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon,
137 special-purpose environments) might want to stick other stuff in front of
138 exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and
139 PCRE_EXP_DATA_DEFN only if they are not already set. */
140
141 #ifndef PCRE_EXP_DECL
142 # ifdef _WIN32
143 # ifndef PCRE_STATIC
144 # define PCRE_EXP_DECL extern __declspec(dllexport)
145 # define PCRE_EXP_DEFN __declspec(dllexport)
146 # define PCRE_EXP_DATA_DEFN __declspec(dllexport)
147 # else
148 # define PCRE_EXP_DECL extern
149 # define PCRE_EXP_DEFN
150 # define PCRE_EXP_DATA_DEFN
151 # endif
152 # else
153 # ifdef __cplusplus
154 # define PCRE_EXP_DECL extern "C"
155 # else
156 # define PCRE_EXP_DECL extern
157 # endif
158 # ifndef PCRE_EXP_DEFN
159 # define PCRE_EXP_DEFN PCRE_EXP_DECL
160 # endif
161 # ifndef PCRE_EXP_DATA_DEFN
162 # define PCRE_EXP_DATA_DEFN
163 # endif
164 # endif
165 #endif
166
167 /* When compiling with the MSVC compiler, it is sometimes necessary to include
168 a "calling convention" before exported function names. (This is secondhand
169 information; I know nothing about MSVC myself). For example, something like
170
171 void __cdecl function(....)
172
173 might be needed. In order so make this easy, all the exported functions have
174 PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
175 set, we ensure here that it has no effect. */
176
177 #ifndef PCRE_CALL_CONVENTION
178 #define PCRE_CALL_CONVENTION
179 #endif
180
181 /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
182 cannot determine these outside the compilation (e.g. by running a program as
183 part of "configure") because PCRE is often cross-compiled for use on other
184 systems. Instead we make use of the maximum sizes that are available at
185 preprocessor time in standard C environments. */
186
187 typedef unsigned char pcre_uint8;
188
189 #if USHRT_MAX == 65535
190 typedef unsigned short pcre_uint16;
191 typedef short pcre_int16;
192 #elif UINT_MAX == 65535
193 typedef unsigned int pcre_uint16;
194 typedef int pcre_int16;
195 #else
196 #error Cannot determine a type for 16-bit unsigned integers
197 #endif
198
199 #if UINT_MAX == 4294967295
200 typedef unsigned int pcre_uint32;
201 typedef int pcre_int32;
202 #elif ULONG_MAX == 4294967295
203 typedef unsigned long int pcre_uint32;
204 typedef long int pcre_int32;
205 #else
206 #error Cannot determine a type for 32-bit unsigned integers
207 #endif
208
209 /* When checking for integer overflow in pcre_compile(), we need to handle
210 large integers. If a 64-bit integer type is available, we can use that.
211 Otherwise we have to cast to double, which of course requires floating point
212 arithmetic. Handle this by defining a macro for the appropriate type. If
213 stdint.h is available, include it; it may define INT64_MAX. Systems that do not
214 have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
215 by "configure". */
216
217 #if HAVE_STDINT_H
218 #include <stdint.h>
219 #elif HAVE_INTTYPES_H
220 #include <inttypes.h>
221 #endif
222
223 #if defined INT64_MAX || defined int64_t
224 #define INT64_OR_DOUBLE int64_t
225 #else
226 #define INT64_OR_DOUBLE double
227 #endif
228
229 /* All character handling must be done as unsigned characters. Otherwise there
230 are problems with top-bit-set characters and functions such as isspace().
231 However, we leave the interface to the outside world as char * or short *,
232 because that should make things easier for callers. This character type is
233 called pcre_uchar.
234
235 The IN_UCHARS macro multiply its argument with the byte size of the current
236 pcre_uchar type. Useful for memcpy and such operations, whose require the
237 byte size of their input/output buffers.
238
239 The MAX_255 macro checks whether its pcre_uchar input is less than 256.
240
241 The TABLE_GET macro is designed for accessing elements of tables whose contain
242 exactly 256 items. When the character is able to contain more than 256
243 items, some check is needed before accessing these tables.
244 */
245
246 #if defined COMPILE_PCRE8
247
248 typedef unsigned char pcre_uchar;
249 #define IN_UCHARS(x) (x)
250 #define MAX_255(c) 1
251 #define TABLE_GET(c, table, default) ((table)[c])
252
253 #elif defined COMPILE_PCRE16
254
255 #if USHRT_MAX != 65535
256 /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
257 pcre.h(.in) and disable (comment out) this message. */
258 #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
259 #endif
260
261 typedef pcre_uint16 pcre_uchar;
262 #define UCHAR_SHIFT (1)
263 #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
264 #define MAX_255(c) ((c) <= 255u)
265 #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
266
267 #elif defined COMPILE_PCRE32
268
269 typedef pcre_uint32 pcre_uchar;
270 #define UCHAR_SHIFT (2)
271 #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
272 #define MAX_255(c) ((c) <= 255u)
273 #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
274
275 /* Assert that pcre_uchar32 is a 32-bit type */
276 typedef int __assert_pcre_uchar32_size[sizeof(pcre_uchar) == 4 ? 1 : -1];
277
278 #else
279 #error Unsupported compiling mode
280 #endif /* COMPILE_PCRE[8|16|32] */
281
282 /* This is an unsigned int value that no character can ever have. UTF-8
283 characters only go up to 0x7fffffff (though Unicode doesn't go beyond
284 0x0010ffff). */
285
286 #define NOTACHAR 0xffffffff
287
288 /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
289 "any" and "anycrlf" at present). The following macros are used to package up
290 testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
291 modules to indicate in which datablock the parameters exist, and what the
292 start/end of string field names are. */
293
294 #define NLTYPE_FIXED 0 /* Newline is a fixed length string */
295 #define NLTYPE_ANY 1 /* Newline is any Unicode line ending */
296 #define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */
297
298 /* This macro checks for a newline at the given position */
299
300 #define IS_NEWLINE(p) \
301 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
302 ((p) < NLBLOCK->PSEND && \
303 PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
304 &(NLBLOCK->nllen), utf)) \
305 : \
306 ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
307 (p)[0] == NLBLOCK->nl[0] && \
308 (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \
309 ) \
310 )
311
312 /* This macro checks for a newline immediately preceding the given position */
313
314 #define WAS_NEWLINE(p) \
315 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
316 ((p) > NLBLOCK->PSSTART && \
317 PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
318 &(NLBLOCK->nllen), utf)) \
319 : \
320 ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
321 (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \
322 (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \
323 ) \
324 )
325
326 /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
327 with a custom type. This makes it possible, for example, to allow pcre_exec()
328 to process subject strings that are discontinuous by using a smart pointer
329 class. It must always be possible to inspect all of the subject string in
330 pcre_exec() because of the way it backtracks. Two macros are required in the
331 normal case, for sign-unspecified and unsigned char pointers. The former is
332 used for the external interface and appears in pcre.h, which is why its name
333 must begin with PCRE_. */
334
335 #ifdef CUSTOM_SUBJECT_PTR
336 #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
337 #else
338 #define PCRE_PUCHAR const pcre_uchar *
339 #endif
340
341 /* Include the public PCRE header and the definitions of UCP character property
342 values. */
343
344 #include "pcre.h"
345 #include "ucp.h"
346
347 /* When compiling for use with the Virtual Pascal compiler, these functions
348 need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
349 option on the command line. */
350
351 #ifdef VPCOMPAT
352 #define strlen(s) _strlen(s)
353 #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
354 #define memcmp(s,c,n) _memcmp(s,c,n)
355 #define memcpy(d,s,n) _memcpy(d,s,n)
356 #define memmove(d,s,n) _memmove(d,s,n)
357 #define memset(s,c,n) _memset(s,c,n)
358 #else /* VPCOMPAT */
359
360 /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
361 define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
362 is set. Otherwise, include an emulating function for those systems that have
363 neither (there some non-Unix environments where this is the case). */
364
365 #ifndef HAVE_MEMMOVE
366 #undef memmove /* some systems may have a macro */
367 #ifdef HAVE_BCOPY
368 #define memmove(a, b, c) bcopy(b, a, c)
369 #else /* HAVE_BCOPY */
370 static void *
371 pcre_memmove(void *d, const void *s, size_t n)
372 {
373 size_t i;
374 unsigned char *dest = (unsigned char *)d;
375 const unsigned char *src = (const unsigned char *)s;
376 if (dest > src)
377 {
378 dest += n;
379 src += n;
380 for (i = 0; i < n; ++i) *(--dest) = *(--src);
381 return (void *)dest;
382 }
383 else
384 {
385 for (i = 0; i < n; ++i) *dest++ = *src++;
386 return (void *)(dest - n);
387 }
388 }
389 #define memmove(a, b, c) pcre_memmove(a, b, c)
390 #endif /* not HAVE_BCOPY */
391 #endif /* not HAVE_MEMMOVE */
392 #endif /* not VPCOMPAT */
393
394
395 /* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
396 in big-endian order) by default. These are used, for example, to link from the
397 start of a subpattern to its alternatives and its end. The use of 2 bytes per
398 offset limits the size of the compiled regex to around 64K, which is big enough
399 for almost everybody. However, I received a request for an even bigger limit.
400 For this reason, and also to make the code easier to maintain, the storing and
401 loading of offsets from the byte string is now handled by the macros that are
402 defined here.
403
404 The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
405 the config.h file, but can be overridden by using -D on the command line. This
406 is automated on Unix systems via the "configure" command. */
407
408 #if defined COMPILE_PCRE8
409
410 #if LINK_SIZE == 2
411
412 #define PUT(a,n,d) \
413 (a[n] = (d) >> 8), \
414 (a[(n)+1] = (d) & 255)
415
416 #define GET(a,n) \
417 (((a)[n] << 8) | (a)[(n)+1])
418
419 #define MAX_PATTERN_SIZE (1 << 16)
420
421
422 #elif LINK_SIZE == 3
423
424 #define PUT(a,n,d) \
425 (a[n] = (d) >> 16), \
426 (a[(n)+1] = (d) >> 8), \
427 (a[(n)+2] = (d) & 255)
428
429 #define GET(a,n) \
430 (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
431
432 #define MAX_PATTERN_SIZE (1 << 24)
433
434
435 #elif LINK_SIZE == 4
436
437 #define PUT(a,n,d) \
438 (a[n] = (d) >> 24), \
439 (a[(n)+1] = (d) >> 16), \
440 (a[(n)+2] = (d) >> 8), \
441 (a[(n)+3] = (d) & 255)
442
443 #define GET(a,n) \
444 (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
445
446 /* Keep it positive */
447 #define MAX_PATTERN_SIZE (1 << 30)
448
449 #else
450 #error LINK_SIZE must be either 2, 3, or 4
451 #endif
452
453 #elif defined COMPILE_PCRE16
454
455 #if LINK_SIZE == 2
456
457 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
458 #undef LINK_SIZE
459 #define LINK_SIZE 1
460
461 #define PUT(a,n,d) \
462 (a[n] = (d))
463
464 #define GET(a,n) \
465 (a[n])
466
467 #define MAX_PATTERN_SIZE (1 << 16)
468
469 #elif LINK_SIZE == 3 || LINK_SIZE == 4
470
471 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
472 #undef LINK_SIZE
473 #define LINK_SIZE 2
474
475 #define PUT(a,n,d) \
476 (a[n] = (d) >> 16), \
477 (a[(n)+1] = (d) & 65535)
478
479 #define GET(a,n) \
480 (((a)[n] << 16) | (a)[(n)+1])
481
482 /* Keep it positive */
483 #define MAX_PATTERN_SIZE (1 << 30)
484
485 #else
486 #error LINK_SIZE must be either 2, 3, or 4
487 #endif
488
489 #elif defined COMPILE_PCRE32
490
491 /* Only supported LINK_SIZE is 4 */
492 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
493 #undef LINK_SIZE
494 #define LINK_SIZE 1
495
496 #define PUT(a,n,d) \
497 (a[n] = (d))
498
499 #define GET(a,n) \
500 (a[n])
501
502 /* Keep it positive */
503 #define MAX_PATTERN_SIZE (1 << 30)
504
505 #else
506 #error Unsupported compiling mode
507 #endif /* COMPILE_PCRE[8|16|32] */
508
509 /* Convenience macro defined in terms of the others */
510
511 #define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
512
513
514 /* PCRE uses some other 2-byte quantities that do not change when the size of
515 offsets changes. There are used for repeat counts and for other things such as
516 capturing parenthesis numbers in back references. */
517
518 #if defined COMPILE_PCRE8
519
520 #define IMM2_SIZE 2
521
522 #define PUT2(a,n,d) \
523 a[n] = (d) >> 8; \
524 a[(n)+1] = (d) & 255
525
526 #define GET2(a,n) \
527 (((a)[n] << 8) | (a)[(n)+1])
528
529 #elif defined COMPILE_PCRE16
530
531 #define IMM2_SIZE 1
532
533 #define PUT2(a,n,d) \
534 a[n] = d
535
536 #define GET2(a,n) \
537 a[n]
538
539 #elif defined COMPILE_PCRE32
540
541 #define IMM2_SIZE 1
542
543 #define PUT2(a,n,d) \
544 a[n] = d
545
546 #define GET2(a,n) \
547 a[n]
548
549 #else
550 #error Unsupported compiling mode
551 #endif /* COMPILE_PCRE[8|16|32] */
552
553 #define PUT2INC(a,n,d) PUT2(a,n,d), a += IMM2_SIZE
554
555 /* The maximum length of a MARK name is currently one data unit; it may be
556 changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
557
558 #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
559 #define MAX_MARK ((1u << 16) - 1)
560 #else
561 #define MAX_MARK ((1u << 8) - 1)
562 #endif
563
564 /* When UTF encoding is being used, a character is no longer just a single
565 byte. The macros for character handling generate simple sequences when used in
566 character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
567 and other macros are not used when UTF is not supported, so they are not
568 defined. To make sure they can never even appear when UTF support is omitted,
569 we don't even define them. */
570
571 #ifndef SUPPORT_UTF
572
573 /* #define MAX_VALUE_FOR_SINGLE_CHAR */
574 /* #define HAS_EXTRALEN(c) */
575 /* #define GET_EXTRALEN(c) */
576 /* #define NOT_FIRSTCHAR(c) */
577 #define GETCHAR(c, eptr) c = *eptr;
578 #define GETCHARTEST(c, eptr) c = *eptr;
579 #define GETCHARINC(c, eptr) c = *eptr++;
580 #define GETCHARINCTEST(c, eptr) c = *eptr++;
581 #define GETCHARLEN(c, eptr, len) c = *eptr;
582 /* #define GETCHARLENTEST(c, eptr, len) */
583 /* #define BACKCHAR(eptr) */
584 /* #define FORWARDCHAR(eptr) */
585 /* #define ACROSSCHAR(condition, eptr, action) */
586
587 #else /* SUPPORT_UTF */
588
589 #if defined COMPILE_PCRE8
590
591 /* These macros were originally written in the form of loops that used data
592 from the tables whose names start with PRIV(utf8_table). They were rewritten by
593 a user so as not to use loops, because in some environments this gives a
594 significant performance advantage, and it seems never to do any harm. */
595
596 /* Tells the biggest code point which can be encoded as a single character. */
597
598 #define MAX_VALUE_FOR_SINGLE_CHAR 127
599
600 /* Tests whether the code point needs extra characters to decode. */
601
602 #define HAS_EXTRALEN(c) ((c) >= 0xc0)
603
604 /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
605 Otherwise it has an undefined behaviour. */
606
607 #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
608
609 /* Returns TRUE, if the given character is not the first character
610 of a UTF sequence. */
611
612 #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
613
614 /* Base macro to pick up the remaining bytes of a UTF-8 character, not
615 advancing the pointer. */
616
617 #define GETUTF8(c, eptr) \
618 { \
619 if ((c & 0x20) == 0) \
620 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
621 else if ((c & 0x10) == 0) \
622 c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
623 else if ((c & 0x08) == 0) \
624 c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
625 ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
626 else if ((c & 0x04) == 0) \
627 c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
628 ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
629 (eptr[4] & 0x3f); \
630 else \
631 c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
632 ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
633 ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
634 }
635
636 /* Get the next UTF-8 character, not advancing the pointer. This is called when
637 we know we are in UTF-8 mode. */
638
639 #define GETCHAR(c, eptr) \
640 c = *eptr; \
641 if (c >= 0xc0) GETUTF8(c, eptr);
642
643 /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
644 pointer. */
645
646 #define GETCHARTEST(c, eptr) \
647 c = *eptr; \
648 if (utf && c >= 0xc0) GETUTF8(c, eptr);
649
650 /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
651 the pointer. */
652
653 #define GETUTF8INC(c, eptr) \
654 { \
655 if ((c & 0x20) == 0) \
656 c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
657 else if ((c & 0x10) == 0) \
658 { \
659 c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
660 eptr += 2; \
661 } \
662 else if ((c & 0x08) == 0) \
663 { \
664 c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
665 ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
666 eptr += 3; \
667 } \
668 else if ((c & 0x04) == 0) \
669 { \
670 c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
671 ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
672 (eptr[3] & 0x3f); \
673 eptr += 4; \
674 } \
675 else \
676 { \
677 c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
678 ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
679 ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
680 eptr += 5; \
681 } \
682 }
683
684 /* Get the next UTF-8 character, advancing the pointer. This is called when we
685 know we are in UTF-8 mode. */
686
687 #define GETCHARINC(c, eptr) \
688 c = *eptr++; \
689 if (c >= 0xc0) GETUTF8INC(c, eptr);
690
691 /* Get the next character, testing for UTF-8 mode, and advancing the pointer.
692 This is called when we don't know if we are in UTF-8 mode. */
693
694 #define GETCHARINCTEST(c, eptr) \
695 c = *eptr++; \
696 if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
697
698 /* Base macro to pick up the remaining bytes of a UTF-8 character, not
699 advancing the pointer, incrementing the length. */
700
701 #define GETUTF8LEN(c, eptr, len) \
702 { \
703 if ((c & 0x20) == 0) \
704 { \
705 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
706 len++; \
707 } \
708 else if ((c & 0x10) == 0) \
709 { \
710 c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
711 len += 2; \
712 } \
713 else if ((c & 0x08) == 0) \
714 {\
715 c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
716 ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
717 len += 3; \
718 } \
719 else if ((c & 0x04) == 0) \
720 { \
721 c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
722 ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
723 (eptr[4] & 0x3f); \
724 len += 4; \
725 } \
726 else \
727 {\
728 c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
729 ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
730 ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
731 len += 5; \
732 } \
733 }
734
735 /* Get the next UTF-8 character, not advancing the pointer, incrementing length
736 if there are extra bytes. This is called when we know we are in UTF-8 mode. */
737
738 #define GETCHARLEN(c, eptr, len) \
739 c = *eptr; \
740 if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
741
742 /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
743 pointer, incrementing length if there are extra bytes. This is called when we
744 do not know if we are in UTF-8 mode. */
745
746 #define GETCHARLENTEST(c, eptr, len) \
747 c = *eptr; \
748 if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
749
750 /* If the pointer is not at the start of a character, move it back until
751 it is. This is called only in UTF-8 mode - we don't put a test within the macro
752 because almost all calls are already within a block of UTF-8 only code. */
753
754 #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
755
756 /* Same as above, just in the other direction. */
757 #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
758
759 /* Same as above, but it allows a fully customizable form. */
760 #define ACROSSCHAR(condition, eptr, action) \
761 while((condition) && ((eptr) & 0xc0) == 0x80) action
762
763 #elif defined COMPILE_PCRE16
764
765 /* Tells the biggest code point which can be encoded as a single character. */
766
767 #define MAX_VALUE_FOR_SINGLE_CHAR 65535
768
769 /* Tests whether the code point needs extra characters to decode. */
770
771 #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
772
773 /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
774 Otherwise it has an undefined behaviour. */
775
776 #define GET_EXTRALEN(c) 1
777
778 /* Returns TRUE, if the given character is not the first character
779 of a UTF sequence. */
780
781 #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
782
783 /* Base macro to pick up the low surrogate of a UTF-16 character, not
784 advancing the pointer. */
785
786 #define GETUTF16(c, eptr) \
787 { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
788
789 /* Get the next UTF-16 character, not advancing the pointer. This is called when
790 we know we are in UTF-16 mode. */
791
792 #define GETCHAR(c, eptr) \
793 c = *eptr; \
794 if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
795
796 /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
797 pointer. */
798
799 #define GETCHARTEST(c, eptr) \
800 c = *eptr; \
801 if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
802
803 /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
804 the pointer. */
805
806 #define GETUTF16INC(c, eptr) \
807 { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
808
809 /* Get the next UTF-16 character, advancing the pointer. This is called when we
810 know we are in UTF-16 mode. */
811
812 #define GETCHARINC(c, eptr) \
813 c = *eptr++; \
814 if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
815
816 /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
817 This is called when we don't know if we are in UTF-16 mode. */
818
819 #define GETCHARINCTEST(c, eptr) \
820 c = *eptr++; \
821 if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
822
823 /* Base macro to pick up the low surrogate of a UTF-16 character, not
824 advancing the pointer, incrementing the length. */
825
826 #define GETUTF16LEN(c, eptr, len) \
827 { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
828
829 /* Get the next UTF-16 character, not advancing the pointer, incrementing
830 length if there is a low surrogate. This is called when we know we are in
831 UTF-16 mode. */
832
833 #define GETCHARLEN(c, eptr, len) \
834 c = *eptr; \
835 if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
836
837 /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
838 pointer, incrementing length if there is a low surrogate. This is called when
839 we do not know if we are in UTF-16 mode. */
840
841 #define GETCHARLENTEST(c, eptr, len) \
842 c = *eptr; \
843 if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
844
845 /* If the pointer is not at the start of a character, move it back until
846 it is. This is called only in UTF-16 mode - we don't put a test within the
847 macro because almost all calls are already within a block of UTF-16 only
848 code. */
849
850 #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
851
852 /* Same as above, just in the other direction. */
853 #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
854
855 /* Same as above, but it allows a fully customizable form. */
856 #define ACROSSCHAR(condition, eptr, action) \
857 if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
858
859 #elif defined COMPILE_PCRE32
860
861 /* These are unnecessary for the 32-bit library */
862 #undef MAX_VALUE_FOR_SINGLE_CHAR
863 #undef HAS_EXTRALEN
864 #undef GET_EXTRALEN
865 #undef NOT_FIRSTCHAR
866
867 /* Get the next UTF-32 character, not advancing the pointer. This is called when
868 we know we are in UTF-32 mode. */
869
870 #define GETCHAR(c, eptr) \
871 c = *eptr;
872
873 /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
874 pointer. */
875
876 #define GETCHARTEST(c, eptr) \
877 c = *eptr;
878
879 /* Get the next UTF-32 character, advancing the pointer. This is called when we
880 know we are in UTF-32 mode. */
881
882 #define GETCHARINC(c, eptr) \
883 c = *eptr++;
884
885 /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
886 This is called when we don't know if we are in UTF-32 mode. */
887
888 #define GETCHARINCTEST(c, eptr) \
889 c = *eptr++;
890
891 /* Get the next UTF-32 character, not advancing the pointer, not incrementing
892 length (since all UTF-32 is of length 1). This is called when we know we are in
893 UTF-32 mode. */
894
895 #define GETCHARLEN(c, eptr, len) \
896 c = *eptr;
897
898 /* Get the next UTF-832character, testing for UTF-32 mode, not advancing the
899 pointer, not incrementing the length (since all UTF-32 is of length 1).
900 This is called when we do not know if we are in UTF-32 mode. */
901
902 #define GETCHARLENTEST(c, eptr, len) \
903 c = *eptr;
904
905 /* If the pointer is not at the start of a character, move it back until
906 it is. This is called only in UTF-32 mode - we don't put a test within the
907 macro because almost all calls are already within a block of UTF-32 only
908 code. */
909
910 #define BACKCHAR(eptr) do { } while (0)
911
912 /* Same as above, just in the other direction. */
913 #define FORWARDCHAR(eptr) do { } while (0)
914
915 /* Same as above, but it allows a fully customizable form. */
916 #define ACROSSCHAR(condition, eptr, action) do { } while (0)
917
918 #else
919 #error Unsupported compiling mode
920 #endif /* COMPILE_PCRE[8|16|32] */
921
922 #endif /* SUPPORT_UTF */
923
924 /* Tests for Unicode horizontal and vertical whitespace characters must check a
925 number of different values. Using a switch statement for this generates the
926 fastest code (no loop, no memory access), and there are several places in the
927 interpreter code where this happens. In order to ensure that all the case lists
928 remain in step, we use macros so that there is only one place where the lists
929 are defined.
930
931 These values are also required as lists in pcre_compile.c when processing \h,
932 \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
933 macros that define the values are here so that all the definitions are
934 together. The lists must be in ascending character order, terminated by
935 NOTACHAR (which is 0xffffffff).
936
937 Any changes should ensure that the various macros are kept in step with each
938 other. NOTE: The values also appear in pcre_jit_compile.c. */
939
940 /* ------ ASCII/Unicode environments ------ */
941
942 #ifndef EBCDIC
943
944 #define HSPACE_LIST \
945 CHAR_HT, CHAR_SPACE, 0xa0, \
946 0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
947 0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
948 NOTACHAR
949
950 #define HSPACE_MULTIBYTE_CASES \
951 case 0x1680: /* OGHAM SPACE MARK */ \
952 case 0x180e: /* MONGOLIAN VOWEL SEPARATOR */ \
953 case 0x2000: /* EN QUAD */ \
954 case 0x2001: /* EM QUAD */ \
955 case 0x2002: /* EN SPACE */ \
956 case 0x2003: /* EM SPACE */ \
957 case 0x2004: /* THREE-PER-EM SPACE */ \
958 case 0x2005: /* FOUR-PER-EM SPACE */ \
959 case 0x2006: /* SIX-PER-EM SPACE */ \
960 case 0x2007: /* FIGURE SPACE */ \
961 case 0x2008: /* PUNCTUATION SPACE */ \
962 case 0x2009: /* THIN SPACE */ \
963 case 0x200A: /* HAIR SPACE */ \
964 case 0x202f: /* NARROW NO-BREAK SPACE */ \
965 case 0x205f: /* MEDIUM MATHEMATICAL SPACE */ \
966 case 0x3000 /* IDEOGRAPHIC SPACE */
967
968 #define HSPACE_BYTE_CASES \
969 case CHAR_HT: \
970 case CHAR_SPACE: \
971 case 0xa0 /* NBSP */
972
973 #define HSPACE_CASES \
974 HSPACE_BYTE_CASES: \
975 HSPACE_MULTIBYTE_CASES
976
977 #define VSPACE_LIST \
978 CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
979
980 #define VSPACE_MULTIBYTE_CASES \
981 case 0x2028: /* LINE SEPARATOR */ \
982 case 0x2029 /* PARAGRAPH SEPARATOR */
983
984 #define VSPACE_BYTE_CASES \
985 case CHAR_LF: \
986 case CHAR_VT: \
987 case CHAR_FF: \
988 case CHAR_CR: \
989 case CHAR_NEL
990
991 #define VSPACE_CASES \
992 VSPACE_BYTE_CASES: \
993 VSPACE_MULTIBYTE_CASES
994
995 /* ------ EBCDIC environments ------ */
996
997 #else
998 #define HSPACE_LIST CHAR_HT, CHAR_SPACE
999
1000 #define HSPACE_BYTE_CASES \
1001 case CHAR_HT: \
1002 case CHAR_SPACE
1003
1004 #define HSPACE_CASES HSPACE_BYTE_CASES
1005
1006 #ifdef EBCDIC_NL25
1007 #define VSPACE_LIST \
1008 CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1009 #else
1010 #define VSPACE_LIST \
1011 CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1012 #endif
1013
1014 #define VSPACE_BYTE_CASES \
1015 case CHAR_LF: \
1016 case CHAR_VT: \
1017 case CHAR_FF: \
1018 case CHAR_CR: \
1019 case CHAR_NEL
1020
1021 #define VSPACE_CASES VSPACE_BYTE_CASES
1022 #endif /* EBCDIC */
1023
1024 /* ------ End of whitespace macros ------ */
1025
1026
1027
1028 /* Private flags containing information about the compiled regex. They used to
1029 live at the top end of the options word, but that got almost full, so now they
1030 are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
1031 the restrictions on partial matching have been lifted. It remains for backwards
1032 compatibility. */
1033
1034 #define PCRE_MODE8 0x0001 /* compiled in 8 bit mode */
1035 #define PCRE_MODE16 0x0002 /* compiled in 16 bit mode */
1036 #define PCRE_MODE32 0x0004 /* compiled in 32 bit mode */
1037 #define PCRE_FIRSTSET 0x0010 /* first_char is set */
1038 #define PCRE_FCH_CASELESS 0x0020 /* caseless first char */
1039 #define PCRE_REQCHSET 0x0040 /* req_byte is set */
1040 #define PCRE_RCH_CASELESS 0x0080 /* caseless requested char */
1041 #define PCRE_STARTLINE 0x0100 /* start after \n for multiline */
1042 #define PCRE_NOPARTIAL 0x0200 /* can't use partial with this regex */
1043 #define PCRE_JCHANGED 0x0400 /* j option used in regex */
1044 #define PCRE_HASCRORLF 0x0800 /* explicit \r or \n in pattern */
1045 #define PCRE_HASTHEN 0x1000 /* pattern contains (*THEN) */
1046
1047 #if defined COMPILE_PCRE8
1048 #define PCRE_MODE PCRE_MODE8
1049 #elif defined COMPILE_PCRE16
1050 #define PCRE_MODE PCRE_MODE16
1051 #elif defined COMPILE_PCRE32
1052 #define PCRE_MODE PCRE_MODE32
1053 #endif
1054 #define PCRE_MODE_MASK (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1055
1056 /* Flags for the "extra" block produced by pcre_study(). */
1057
1058 #define PCRE_STUDY_MAPPED 0x0001 /* a map of starting chars exists */
1059 #define PCRE_STUDY_MINLEN 0x0002 /* a minimum length field exists */
1060
1061 /* Masks for identifying the public options that are permitted at compile
1062 time, run time, or study time, respectively. */
1063
1064 #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
1065 PCRE_NEWLINE_ANYCRLF)
1066
1067 #define PUBLIC_COMPILE_OPTIONS \
1068 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
1069 PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1070 PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
1071 PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1072 PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
1073
1074 #define PUBLIC_EXEC_OPTIONS \
1075 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1076 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
1077 PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
1078
1079 #define PUBLIC_DFA_EXEC_OPTIONS \
1080 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1081 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
1082 PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1083 PCRE_NO_START_OPTIMIZE)
1084
1085 #define PUBLIC_STUDY_OPTIONS \
1086 (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1087 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1088
1089 /* Magic number to provide a small check against being handed junk. */
1090
1091 #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
1092
1093 /* This variable is used to detect a loaded regular expression
1094 in different endianness. */
1095
1096 #define REVERSED_MAGIC_NUMBER 0x45524350UL /* 'ERCP' */
1097
1098 /* Negative values for the firstchar and reqchar variables */
1099
1100 #define REQ_UNSET (-2)
1101 #define REQ_NONE (-1)
1102
1103 /* The maximum remaining length of subject we are prepared to search for a
1104 req_byte match. */
1105
1106 #define REQ_BYTE_MAX 1000
1107
1108 /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1109 environments where these macros are defined elsewhere. Unfortunately, there
1110 is no way to do the same for the typedef. */
1111
1112 typedef int BOOL;
1113
1114 #ifndef FALSE
1115 #define FALSE 0
1116 #define TRUE 1
1117 #endif
1118
1119 /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
1120 character constants like '*' because the compiler would emit their EBCDIC code,
1121 which is different from their ASCII/UTF-8 code. Instead we define macros for
1122 the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
1123 is enabled. When UTF-8 support is not enabled, the definitions use character
1124 literals. Both character and string versions of each character are needed, and
1125 there are some longer strings as well.
1126
1127 This means that, on EBCDIC platforms, the PCRE library can handle either
1128 EBCDIC, or UTF-8, but not both. To support both in the same compiled library
1129 would need different lookups depending on whether PCRE_UTF8 was set or not.
1130 This would make it impossible to use characters in switch/case statements,
1131 which would reduce performance. For a theoretical use (which nobody has asked
1132 for) in a minority area (EBCDIC platforms), this is not sensible. Any
1133 application that did need both could compile two versions of the library, using
1134 macros to give the functions distinct names. */
1135
1136 #ifndef SUPPORT_UTF
1137
1138 /* UTF-8 support is not enabled; use the platform-dependent character literals
1139 so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1140 mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1141 characters, a common practice has been to use its NL (0x15) character as the
1142 line terminator in C-like processing environments. However, sometimes the LF
1143 (0x25) character is used instead, according to this Unicode document:
1144
1145 http://unicode.org/standard/reports/tr13/tr13-5.html
1146
1147 PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1148 instead. Whichever is *not* chosen is defined as NEL.
1149
1150 In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1151 same code point. */
1152
1153 #ifdef EBCDIC
1154
1155 #ifndef EBCDIC_NL25
1156 #define CHAR_NL '\x15'
1157 #define CHAR_NEL '\x25'
1158 #define STR_NL "\x15"
1159 #define STR_NEL "\x25"
1160 #else
1161 #define CHAR_NL '\x25'
1162 #define CHAR_NEL '\x15'
1163 #define STR_NL "\x25"
1164 #define STR_NEL "\x15"
1165 #endif
1166
1167 #define CHAR_LF CHAR_NL
1168 #define STR_LF STR_NL
1169
1170 #define CHAR_ESC '\047'
1171 #define CHAR_DEL '\007'
1172 #define STR_ESC "\047"
1173 #define STR_DEL "\007"
1174
1175 #else /* Not EBCDIC */
1176
1177 /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1178 compatibility. NEL is the Unicode newline character; make sure it is
1179 a positive value. */
1180
1181 #define CHAR_LF '\n'
1182 #define CHAR_NL CHAR_LF
1183 #define CHAR_NEL ((unsigned char)'\x85')
1184 #define CHAR_ESC '\033'
1185 #define CHAR_DEL '\177'
1186
1187 #define STR_LF "\n"
1188 #define STR_NL STR_LF
1189 #define STR_NEL "\x85"
1190 #define STR_ESC "\033"
1191 #define STR_DEL "\177"
1192
1193 #endif /* EBCDIC */
1194
1195 /* The remaining definitions work in both environments. */
1196
1197 #define CHAR_HT '\t'
1198 #define CHAR_VT '\v'
1199 #define CHAR_FF '\f'
1200 #define CHAR_CR '\r'
1201 #define CHAR_BS '\b'
1202 #define CHAR_BEL '\a'
1203
1204 #define CHAR_SPACE ' '
1205 #define CHAR_EXCLAMATION_MARK '!'
1206 #define CHAR_QUOTATION_MARK '"'
1207 #define CHAR_NUMBER_SIGN '#'
1208 #define CHAR_DOLLAR_SIGN '$'
1209 #define CHAR_PERCENT_SIGN '%'
1210 #define CHAR_AMPERSAND '&'
1211 #define CHAR_APOSTROPHE '\''
1212 #define CHAR_LEFT_PARENTHESIS '('
1213 #define CHAR_RIGHT_PARENTHESIS ')'
1214 #define CHAR_ASTERISK '*'
1215 #define CHAR_PLUS '+'
1216 #define CHAR_COMMA ','
1217 #define CHAR_MINUS '-'
1218 #define CHAR_DOT '.'
1219 #define CHAR_SLASH '/'
1220 #define CHAR_0 '0'
1221 #define CHAR_1 '1'
1222 #define CHAR_2 '2'
1223 #define CHAR_3 '3'
1224 #define CHAR_4 '4'
1225 #define CHAR_5 '5'
1226 #define CHAR_6 '6'
1227 #define CHAR_7 '7'
1228 #define CHAR_8 '8'
1229 #define CHAR_9 '9'
1230 #define CHAR_COLON ':'
1231 #define CHAR_SEMICOLON ';'
1232 #define CHAR_LESS_THAN_SIGN '<'
1233 #define CHAR_EQUALS_SIGN '='
1234 #define CHAR_GREATER_THAN_SIGN '>'
1235 #define CHAR_QUESTION_MARK '?'
1236 #define CHAR_COMMERCIAL_AT '@'
1237 #define CHAR_A 'A'
1238 #define CHAR_B 'B'
1239 #define CHAR_C 'C'
1240 #define CHAR_D 'D'
1241 #define CHAR_E 'E'
1242 #define CHAR_F 'F'
1243 #define CHAR_G 'G'
1244 #define CHAR_H 'H'
1245 #define CHAR_I 'I'
1246 #define CHAR_J 'J'
1247 #define CHAR_K 'K'
1248 #define CHAR_L 'L'
1249 #define CHAR_M 'M'
1250 #define CHAR_N 'N'
1251 #define CHAR_O 'O'
1252 #define CHAR_P 'P'
1253 #define CHAR_Q 'Q'
1254 #define CHAR_R 'R'
1255 #define CHAR_S 'S'
1256 #define CHAR_T 'T'
1257 #define CHAR_U 'U'
1258 #define CHAR_V 'V'
1259 #define CHAR_W 'W'
1260 #define CHAR_X 'X'
1261 #define CHAR_Y 'Y'
1262 #define CHAR_Z 'Z'
1263 #define CHAR_LEFT_SQUARE_BRACKET '['
1264 #define CHAR_BACKSLASH '\\'
1265 #define CHAR_RIGHT_SQUARE_BRACKET ']'
1266 #define CHAR_CIRCUMFLEX_ACCENT '^'
1267 #define CHAR_UNDERSCORE '_'
1268 #define CHAR_GRAVE_ACCENT '`'
1269 #define CHAR_a 'a'
1270 #define CHAR_b 'b'
1271 #define CHAR_c 'c'
1272 #define CHAR_d 'd'
1273 #define CHAR_e 'e'
1274 #define CHAR_f 'f'
1275 #define CHAR_g 'g'
1276 #define CHAR_h 'h'
1277 #define CHAR_i 'i'
1278 #define CHAR_j 'j'
1279 #define CHAR_k 'k'
1280 #define CHAR_l 'l'
1281 #define CHAR_m 'm'
1282 #define CHAR_n 'n'
1283 #define CHAR_o 'o'
1284 #define CHAR_p 'p'
1285 #define CHAR_q 'q'
1286 #define CHAR_r 'r'
1287 #define CHAR_s 's'
1288 #define CHAR_t 't'
1289 #define CHAR_u 'u'
1290 #define CHAR_v 'v'
1291 #define CHAR_w 'w'
1292 #define CHAR_x 'x'
1293 #define CHAR_y 'y'
1294 #define CHAR_z 'z'
1295 #define CHAR_LEFT_CURLY_BRACKET '{'
1296 #define CHAR_VERTICAL_LINE '|'
1297 #define CHAR_RIGHT_CURLY_BRACKET '}'
1298 #define CHAR_TILDE '~'
1299
1300 #define STR_HT "\t"
1301 #define STR_VT "\v"
1302 #define STR_FF "\f"
1303 #define STR_CR "\r"
1304 #define STR_BS "\b"
1305 #define STR_BEL "\a"
1306
1307 #define STR_SPACE " "
1308 #define STR_EXCLAMATION_MARK "!"
1309 #define STR_QUOTATION_MARK "\""
1310 #define STR_NUMBER_SIGN "#"
1311 #define STR_DOLLAR_SIGN "$"
1312 #define STR_PERCENT_SIGN "%"
1313 #define STR_AMPERSAND "&"
1314 #define STR_APOSTROPHE "'"
1315 #define STR_LEFT_PARENTHESIS "("
1316 #define STR_RIGHT_PARENTHESIS ")"
1317 #define STR_ASTERISK "*"
1318 #define STR_PLUS "+"
1319 #define STR_COMMA ","
1320 #define STR_MINUS "-"
1321 #define STR_DOT "."
1322 #define STR_SLASH "/"
1323 #define STR_0 "0"
1324 #define STR_1 "1"
1325 #define STR_2 "2"
1326 #define STR_3 "3"
1327 #define STR_4 "4"
1328 #define STR_5 "5"
1329 #define STR_6 "6"
1330 #define STR_7 "7"
1331 #define STR_8 "8"
1332 #define STR_9 "9"
1333 #define STR_COLON ":"
1334 #define STR_SEMICOLON ";"
1335 #define STR_LESS_THAN_SIGN "<"
1336 #define STR_EQUALS_SIGN "="
1337 #define STR_GREATER_THAN_SIGN ">"
1338 #define STR_QUESTION_MARK "?"
1339 #define STR_COMMERCIAL_AT "@"
1340 #define STR_A "A"
1341 #define STR_B "B"
1342 #define STR_C "C"
1343 #define STR_D "D"
1344 #define STR_E "E"
1345 #define STR_F "F"
1346 #define STR_G "G"
1347 #define STR_H "H"
1348 #define STR_I "I"
1349 #define STR_J "J"
1350 #define STR_K "K"
1351 #define STR_L "L"
1352 #define STR_M "M"
1353 #define STR_N "N"
1354 #define STR_O "O"
1355 #define STR_P "P"
1356 #define STR_Q "Q"
1357 #define STR_R "R"
1358 #define STR_S "S"
1359 #define STR_T "T"
1360 #define STR_U "U"
1361 #define STR_V "V"
1362 #define STR_W "W"
1363 #define STR_X "X"
1364 #define STR_Y "Y"
1365 #define STR_Z "Z"
1366 #define STR_LEFT_SQUARE_BRACKET "["
1367 #define STR_BACKSLASH "\\"
1368 #define STR_RIGHT_SQUARE_BRACKET "]"
1369 #define STR_CIRCUMFLEX_ACCENT "^"
1370 #define STR_UNDERSCORE "_"
1371 #define STR_GRAVE_ACCENT "`"
1372 #define STR_a "a"
1373 #define STR_b "b"
1374 #define STR_c "c"
1375 #define STR_d "d"
1376 #define STR_e "e"
1377 #define STR_f "f"
1378 #define STR_g "g"
1379 #define STR_h "h"
1380 #define STR_i "i"
1381 #define STR_j "j"
1382 #define STR_k "k"
1383 #define STR_l "l"
1384 #define STR_m "m"
1385 #define STR_n "n"
1386 #define STR_o "o"
1387 #define STR_p "p"
1388 #define STR_q "q"
1389 #define STR_r "r"
1390 #define STR_s "s"
1391 #define STR_t "t"
1392 #define STR_u "u"
1393 #define STR_v "v"
1394 #define STR_w "w"
1395 #define STR_x "x"
1396 #define STR_y "y"
1397 #define STR_z "z"
1398 #define STR_LEFT_CURLY_BRACKET "{"
1399 #define STR_VERTICAL_LINE "|"
1400 #define STR_RIGHT_CURLY_BRACKET "}"
1401 #define STR_TILDE "~"
1402
1403 #define STRING_ACCEPT0 "ACCEPT\0"
1404 #define STRING_COMMIT0 "COMMIT\0"
1405 #define STRING_F0 "F\0"
1406 #define STRING_FAIL0 "FAIL\0"
1407 #define STRING_MARK0 "MARK\0"
1408 #define STRING_PRUNE0 "PRUNE\0"
1409 #define STRING_SKIP0 "SKIP\0"
1410 #define STRING_THEN "THEN"
1411
1412 #define STRING_alpha0 "alpha\0"
1413 #define STRING_lower0 "lower\0"
1414 #define STRING_upper0 "upper\0"
1415 #define STRING_alnum0 "alnum\0"
1416 #define STRING_ascii0 "ascii\0"
1417 #define STRING_blank0 "blank\0"
1418 #define STRING_cntrl0 "cntrl\0"
1419 #define STRING_digit0 "digit\0"
1420 #define STRING_graph0 "graph\0"
1421 #define STRING_print0 "print\0"
1422 #define STRING_punct0 "punct\0"
1423 #define STRING_space0 "space\0"
1424 #define STRING_word0 "word\0"
1425 #define STRING_xdigit "xdigit"
1426
1427 #define STRING_DEFINE "DEFINE"
1428
1429 #define STRING_CR_RIGHTPAR "CR)"
1430 #define STRING_LF_RIGHTPAR "LF)"
1431 #define STRING_CRLF_RIGHTPAR "CRLF)"
1432 #define STRING_ANY_RIGHTPAR "ANY)"
1433 #define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)"
1434 #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"
1435 #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"
1436 #ifdef COMPILE_PCRE8
1437 #define STRING_UTF_RIGHTPAR "UTF8)"
1438 #endif
1439 #ifdef COMPILE_PCRE16
1440 #define STRING_UTF_RIGHTPAR "UTF16)"
1441 #endif
1442 #ifdef COMPILE_PCRE32
1443 #define STRING_UTF_RIGHTPAR "UTF32)"
1444 #endif
1445 #define STRING_UCP_RIGHTPAR "UCP)"
1446 #define STRING_NO_START_OPT_RIGHTPAR "NO_START_OPT)"
1447
1448 #else /* SUPPORT_UTF */
1449
1450 /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1451 works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
1452 only. */
1453
1454 #define CHAR_HT '\011'
1455 #define CHAR_VT '\013'
1456 #define CHAR_FF '\014'
1457 #define CHAR_CR '\015'
1458 #define CHAR_LF '\012'
1459 #define CHAR_NL CHAR_LF
1460 #define CHAR_NEL ((unsigned char)'\x85')
1461 #define CHAR_BS '\010'
1462 #define CHAR_BEL '\007'
1463 #define CHAR_ESC '\033'
1464 #define CHAR_DEL '\177'
1465
1466 #define CHAR_SPACE '\040'
1467 #define CHAR_EXCLAMATION_MARK '\041'
1468 #define CHAR_QUOTATION_MARK '\042'
1469 #define CHAR_NUMBER_SIGN '\043'
1470 #define CHAR_DOLLAR_SIGN '\044'
1471 #define CHAR_PERCENT_SIGN '\045'
1472 #define CHAR_AMPERSAND '\046'
1473 #define CHAR_APOSTROPHE '\047'
1474 #define CHAR_LEFT_PARENTHESIS '\050'
1475 #define CHAR_RIGHT_PARENTHESIS '\051'
1476 #define CHAR_ASTERISK '\052'
1477 #define CHAR_PLUS '\053'
1478 #define CHAR_COMMA '\054'
1479 #define CHAR_MINUS '\055'
1480 #define CHAR_DOT '\056'
1481 #define CHAR_SLASH '\057'
1482 #define CHAR_0 '\060'
1483 #define CHAR_1 '\061'
1484 #define CHAR_2 '\062'
1485 #define CHAR_3 '\063'
1486 #define CHAR_4 '\064'
1487 #define CHAR_5 '\065'
1488 #define CHAR_6 '\066'
1489 #define CHAR_7 '\067'
1490 #define CHAR_8 '\070'
1491 #define CHAR_9 '\071'
1492 #define CHAR_COLON '\072'
1493 #define CHAR_SEMICOLON '\073'
1494 #define CHAR_LESS_THAN_SIGN '\074'
1495 #define CHAR_EQUALS_SIGN '\075'
1496 #define CHAR_GREATER_THAN_SIGN '\076'
1497 #define CHAR_QUESTION_MARK '\077'
1498 #define CHAR_COMMERCIAL_AT '\100'
1499 #define CHAR_A '\101'
1500 #define CHAR_B '\102'
1501 #define CHAR_C '\103'
1502 #define CHAR_D '\104'
1503 #define CHAR_E '\105'
1504 #define CHAR_F '\106'
1505 #define CHAR_G '\107'
1506 #define CHAR_H '\110'
1507 #define CHAR_I '\111'
1508 #define CHAR_J '\112'
1509 #define CHAR_K '\113'
1510 #define CHAR_L '\114'
1511 #define CHAR_M '\115'
1512 #define CHAR_N '\116'
1513 #define CHAR_O '\117'
1514 #define CHAR_P '\120'
1515 #define CHAR_Q '\121'
1516 #define CHAR_R '\122'
1517 #define CHAR_S '\123'
1518 #define CHAR_T '\124'
1519 #define CHAR_U '\125'
1520 #define CHAR_V '\126'
1521 #define CHAR_W '\127'
1522 #define CHAR_X '\130'
1523 #define CHAR_Y '\131'
1524 #define CHAR_Z '\132'
1525 #define CHAR_LEFT_SQUARE_BRACKET '\133'
1526 #define CHAR_BACKSLASH '\134'
1527 #define CHAR_RIGHT_SQUARE_BRACKET '\135'
1528 #define CHAR_CIRCUMFLEX_ACCENT '\136'
1529 #define CHAR_UNDERSCORE '\137'
1530 #define CHAR_GRAVE_ACCENT '\140'
1531 #define CHAR_a '\141'
1532 #define CHAR_b '\142'
1533 #define CHAR_c '\143'
1534 #define CHAR_d '\144'
1535 #define CHAR_e '\145'
1536 #define CHAR_f '\146'
1537 #define CHAR_g '\147'
1538 #define CHAR_h '\150'
1539 #define CHAR_i '\151'
1540 #define CHAR_j '\152'
1541 #define CHAR_k '\153'
1542 #define CHAR_l '\154'
1543 #define CHAR_m '\155'
1544 #define CHAR_n '\156'
1545 #define CHAR_o '\157'
1546 #define CHAR_p '\160'
1547 #define CHAR_q '\161'
1548 #define CHAR_r '\162'
1549 #define CHAR_s '\163'
1550 #define CHAR_t '\164'
1551 #define CHAR_u '\165'
1552 #define CHAR_v '\166'
1553 #define CHAR_w '\167'
1554 #define CHAR_x '\170'
1555 #define CHAR_y '\171'
1556 #define CHAR_z '\172'
1557 #define CHAR_LEFT_CURLY_BRACKET '\173'
1558 #define CHAR_VERTICAL_LINE '\174'
1559 #define CHAR_RIGHT_CURLY_BRACKET '\175'
1560 #define CHAR_TILDE '\176'
1561
1562 #define STR_HT "\011"
1563 #define STR_VT "\013"
1564 #define STR_FF "\014"
1565 #define STR_CR "\015"
1566 #define STR_NL "\012"
1567 #define STR_BS "\010"
1568 #define STR_BEL "\007"
1569 #define STR_ESC "\033"
1570 #define STR_DEL "\177"
1571
1572 #define STR_SPACE "\040"
1573 #define STR_EXCLAMATION_MARK "\041"
1574 #define STR_QUOTATION_MARK "\042"
1575 #define STR_NUMBER_SIGN "\043"
1576 #define STR_DOLLAR_SIGN "\044"
1577 #define STR_PERCENT_SIGN "\045"
1578 #define STR_AMPERSAND "\046"
1579 #define STR_APOSTROPHE "\047"
1580 #define STR_LEFT_PARENTHESIS "\050"
1581 #define STR_RIGHT_PARENTHESIS "\051"
1582 #define STR_ASTERISK "\052"
1583 #define STR_PLUS "\053"
1584 #define STR_COMMA "\054"
1585 #define STR_MINUS "\055"
1586 #define STR_DOT "\056"
1587 #define STR_SLASH "\057"
1588 #define STR_0 "\060"
1589 #define STR_1 "\061"
1590 #define STR_2 "\062"
1591 #define STR_3 "\063"
1592 #define STR_4 "\064"
1593 #define STR_5 "\065"
1594 #define STR_6 "\066"
1595 #define STR_7 "\067"
1596 #define STR_8 "\070"
1597 #define STR_9 "\071"
1598 #define STR_COLON "\072"
1599 #define STR_SEMICOLON "\073"
1600 #define STR_LESS_THAN_SIGN "\074"
1601 #define STR_EQUALS_SIGN "\075"
1602 #define STR_GREATER_THAN_SIGN "\076"
1603 #define STR_QUESTION_MARK "\077"
1604 #define STR_COMMERCIAL_AT "\100"
1605 #define STR_A "\101"
1606 #define STR_B "\102"
1607 #define STR_C "\103"
1608 #define STR_D "\104"
1609 #define STR_E "\105"
1610 #define STR_F "\106"
1611 #define STR_G "\107"
1612 #define STR_H "\110"
1613 #define STR_I "\111"
1614 #define STR_J "\112"
1615 #define STR_K "\113"
1616 #define STR_L "\114"
1617 #define STR_M "\115"
1618 #define STR_N "\116"
1619 #define STR_O "\117"
1620 #define STR_P "\120"
1621 #define STR_Q "\121"
1622 #define STR_R "\122"
1623 #define STR_S "\123"
1624 #define STR_T "\124"
1625 #define STR_U "\125"
1626 #define STR_V "\126"
1627 #define STR_W "\127"
1628 #define STR_X "\130"
1629 #define STR_Y "\131"
1630 #define STR_Z "\132"
1631 #define STR_LEFT_SQUARE_BRACKET "\133"
1632 #define STR_BACKSLASH "\134"
1633 #define STR_RIGHT_SQUARE_BRACKET "\135"
1634 #define STR_CIRCUMFLEX_ACCENT "\136"
1635 #define STR_UNDERSCORE "\137"
1636 #define STR_GRAVE_ACCENT "\140"
1637 #define STR_a "\141"
1638 #define STR_b "\142"
1639 #define STR_c "\143"
1640 #define STR_d "\144"
1641 #define STR_e "\145"
1642 #define STR_f "\146"
1643 #define STR_g "\147"
1644 #define STR_h "\150"
1645 #define STR_i "\151"
1646 #define STR_j "\152"
1647 #define STR_k "\153"
1648 #define STR_l "\154"
1649 #define STR_m "\155"
1650 #define STR_n "\156"
1651 #define STR_o "\157"
1652 #define STR_p "\160"
1653 #define STR_q "\161"
1654 #define STR_r "\162"
1655 #define STR_s "\163"
1656 #define STR_t "\164"
1657 #define STR_u "\165"
1658 #define STR_v "\166"
1659 #define STR_w "\167"
1660 #define STR_x "\170"
1661 #define STR_y "\171"
1662 #define STR_z "\172"
1663 #define STR_LEFT_CURLY_BRACKET "\173"
1664 #define STR_VERTICAL_LINE "\174"
1665 #define STR_RIGHT_CURLY_BRACKET "\175"
1666 #define STR_TILDE "\176"
1667
1668 #define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1669 #define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1670 #define STRING_F0 STR_F "\0"
1671 #define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0"
1672 #define STRING_MARK0 STR_M STR_A STR_R STR_K "\0"
1673 #define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0"
1674 #define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0"
1675 #define STRING_THEN STR_T STR_H STR_E STR_N
1676
1677 #define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0"
1678 #define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0"
1679 #define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0"
1680 #define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0"
1681 #define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0"
1682 #define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0"
1683 #define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0"
1684 #define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0"
1685 #define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0"
1686 #define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0"
1687 #define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0"
1688 #define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0"
1689 #define STRING_word0 STR_w STR_o STR_r STR_d "\0"
1690 #define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t
1691
1692 #define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E
1693
1694 #define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS
1695 #define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS
1696 #define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1697 #define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1698 #define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1699 #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
1700 #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
1701 #ifdef COMPILE_PCRE8
1702 #define STRING_UTF_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1703 #endif
1704 #ifdef COMPILE_PCRE16
1705 #define STRING_UTF_RIGHTPAR STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1706 #endif
1707 #ifdef COMPILE_PCRE32
1708 #define STRING_UTF_RIGHTPAR STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1709 #endif
1710 #define STRING_UCP_RIGHTPAR STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1711 #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
1712
1713 #endif /* SUPPORT_UTF */
1714
1715 /* Escape items that are just an encoding of a particular data value. */
1716
1717 #ifndef ESC_e
1718 #define ESC_e CHAR_ESC
1719 #endif
1720
1721 #ifndef ESC_f
1722 #define ESC_f CHAR_FF
1723 #endif
1724
1725 #ifndef ESC_n
1726 #define ESC_n CHAR_LF
1727 #endif
1728
1729 #ifndef ESC_r
1730 #define ESC_r CHAR_CR
1731 #endif
1732
1733 /* We can't officially use ESC_t because it is a POSIX reserved identifier
1734 (presumably because of all the others like size_t). */
1735
1736 #ifndef ESC_tee
1737 #define ESC_tee CHAR_HT
1738 #endif
1739
1740 /* Codes for different types of Unicode property */
1741
1742 #define PT_ANY 0 /* Any property - matches all chars */
1743 #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
1744 #define PT_GC 2 /* Specified general characteristic (e.g. L) */
1745 #define PT_PC 3 /* Specified particular characteristic (e.g. Lu) */
1746 #define PT_SC 4 /* Script (e.g. Han) */
1747 #define PT_ALNUM 5 /* Alphanumeric - the union of L and N */
1748 #define PT_SPACE 6 /* Perl space - Z plus 9,10,12,13 */
1749 #define PT_PXSPACE 7 /* POSIX space - Z plus 9,10,11,12,13 */
1750 #define PT_WORD 8 /* Word - L plus N plus underscore */
1751 #define PT_CLIST 9 /* Pseudo-property: match character list */
1752
1753 /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1754 contain characters with values greater than 255. */
1755
1756 #define XCL_NOT 0x01 /* Flag: this is a negative class */
1757 #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
1758
1759 #define XCL_END 0 /* Marks end of individual items */
1760 #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
1761 #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
1762 #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
1763 #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
1764
1765 /* These are escaped items that aren't just an encoding of a particular data
1766 value such as \n. They must have non-zero values, as check_escape() returns
1767 0 for a data character. Also, they must appear in the same order as in the opcode
1768 definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1769 corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1770 used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1771 non-DOTALL mode, "." behaves like \N.
1772
1773 The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1774 when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1775 They must be contiguous, and remain in order so that the replacements can be
1776 looked up from a table.
1777
1778 Negative numbers are used to encode a backreference (\1, \2, \3, etc.) in
1779 check_escape(). There are two tests in the code for an escape
1780 greater than ESC_b and less than ESC_Z to detect the types that may be
1781 repeated. These are the types that consume characters. If any new escapes are
1782 put in between that don't consume a character, that code will have to change.
1783 */
1784
1785 enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1786 ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1787 ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1788 ESC_E, ESC_Q, ESC_g, ESC_k,
1789 ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1790
1791 /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1792 OP_EOD must correspond in order to the list of escapes immediately above.
1793
1794 *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1795 that follow must also be updated to match. There are also tables called
1796 "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1797
1798 enum {
1799 OP_END, /* 0 End of pattern */
1800
1801 /* Values corresponding to backslashed metacharacters */
1802
1803 OP_SOD, /* 1 Start of data: \A */
1804 OP_SOM, /* 2 Start of match (subject + offset): \G */
1805 OP_SET_SOM, /* 3 Set start of match (\K) */
1806 OP_NOT_WORD_BOUNDARY, /* 4 \B */
1807 OP_WORD_BOUNDARY, /* 5 \b */
1808 OP_NOT_DIGIT, /* 6 \D */
1809 OP_DIGIT, /* 7 \d */
1810 OP_NOT_WHITESPACE, /* 8 \S */
1811 OP_WHITESPACE, /* 9 \s */
1812 OP_NOT_WORDCHAR, /* 10 \W */
1813 OP_WORDCHAR, /* 11 \w */
1814
1815 OP_ANY, /* 12 Match any character except newline (\N) */
1816 OP_ALLANY, /* 13 Match any character */
1817 OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1818 OP_NOTPROP, /* 15 \P (not Unicode property) */
1819 OP_PROP, /* 16 \p (Unicode property) */
1820 OP_ANYNL, /* 17 \R (any newline sequence) */
1821 OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */
1822 OP_HSPACE, /* 19 \h (horizontal whitespace) */
1823 OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */
1824 OP_VSPACE, /* 21 \v (vertical whitespace) */
1825 OP_EXTUNI, /* 22 \X (extended Unicode sequence */
1826 OP_EODN, /* 23 End of data or \n at end of data (\Z) */
1827 OP_EOD, /* 24 End of data (\z) */
1828
1829 OP_CIRC, /* 25 Start of line - not multiline */
1830 OP_CIRCM, /* 26 Start of line - multiline */
1831 OP_DOLL, /* 27 End of line - not multiline */
1832 OP_DOLLM, /* 28 End of line - multiline */
1833 OP_CHAR, /* 29 Match one character, casefully */
1834 OP_CHARI, /* 30 Match one character, caselessly */
1835 OP_NOT, /* 31 Match one character, not the given one, casefully */
1836 OP_NOTI, /* 32 Match one character, not the given one, caselessly */
1837
1838 /* The following sets of 13 opcodes must always be kept in step because
1839 the offset from the first one is used to generate the others. */
1840
1841 /**** Single characters, caseful, must precede the caseless ones ****/
1842
1843 OP_STAR, /* 33 The maximizing and minimizing versions of */
1844 OP_MINSTAR, /* 34 these six opcodes must come in pairs, with */
1845 OP_PLUS, /* 35 the minimizing one second. */
1846 OP_MINPLUS, /* 36 */
1847 OP_QUERY, /* 37 */
1848 OP_MINQUERY, /* 38 */
1849
1850 OP_UPTO, /* 39 From 0 to n matches of one character, caseful*/
1851 OP_MINUPTO, /* 40 */
1852 OP_EXACT, /* 41 Exactly n matches */
1853
1854 OP_POSSTAR, /* 42 Possessified star, caseful */
1855 OP_POSPLUS, /* 43 Possessified plus, caseful */
1856 OP_POSQUERY, /* 44 Posesssified query, caseful */
1857 OP_POSUPTO, /* 45 Possessified upto, caseful */
1858
1859 /**** Single characters, caseless, must follow the caseful ones */
1860
1861 OP_STARI, /* 46 */
1862 OP_MINSTARI, /* 47 */
1863 OP_PLUSI, /* 48 */
1864 OP_MINPLUSI, /* 49 */
1865 OP_QUERYI, /* 50 */
1866 OP_MINQUERYI, /* 51 */
1867
1868 OP_UPTOI, /* 52 From 0 to n matches of one character, caseless */
1869 OP_MINUPTOI, /* 53 */
1870 OP_EXACTI, /* 54 */
1871
1872 OP_POSSTARI, /* 55 Possessified star, caseless */
1873 OP_POSPLUSI, /* 56 Possessified plus, caseless */
1874 OP_POSQUERYI, /* 57 Posesssified query, caseless */
1875 OP_POSUPTOI, /* 58 Possessified upto, caseless */
1876
1877 /**** The negated ones must follow the non-negated ones, and match them ****/
1878 /**** Negated single character, caseful; must precede the caseless ones ****/
1879
1880 OP_NOTSTAR, /* 59 The maximizing and minimizing versions of */
1881 OP_NOTMINSTAR, /* 60 these six opcodes must come in pairs, with */
1882 OP_NOTPLUS, /* 61 the minimizing one second. They must be in */
1883 OP_NOTMINPLUS, /* 62 exactly the same order as those above. */
1884 OP_NOTQUERY, /* 63 */
1885 OP_NOTMINQUERY, /* 64 */
1886
1887 OP_NOTUPTO, /* 65 From 0 to n matches, caseful */
1888 OP_NOTMINUPTO, /* 66 */
1889 OP_NOTEXACT, /* 67 Exactly n matches */
1890
1891 OP_NOTPOSSTAR, /* 68 Possessified versions, caseful */
1892 OP_NOTPOSPLUS, /* 69 */
1893 OP_NOTPOSQUERY, /* 70 */
1894 OP_NOTPOSUPTO, /* 71 */
1895
1896 /**** Negated single character, caseless; must follow the caseful ones ****/
1897
1898 OP_NOTSTARI, /* 72 */
1899 OP_NOTMINSTARI, /* 73 */
1900 OP_NOTPLUSI, /* 74 */
1901 OP_NOTMINPLUSI, /* 75 */
1902 OP_NOTQUERYI, /* 76 */
1903 OP_NOTMINQUERYI, /* 77 */
1904
1905 OP_NOTUPTOI, /* 78 From 0 to n matches, caseless */
1906 OP_NOTMINUPTOI, /* 79 */
1907 OP_NOTEXACTI, /* 80 Exactly n matches */
1908
1909 OP_NOTPOSSTARI, /* 81 Possessified versions, caseless */
1910 OP_NOTPOSPLUSI, /* 82 */
1911 OP_NOTPOSQUERYI, /* 83 */
1912 OP_NOTPOSUPTOI, /* 84 */
1913
1914 /**** Character types ****/
1915
1916 OP_TYPESTAR, /* 85 The maximizing and minimizing versions of */
1917 OP_TYPEMINSTAR, /* 86 these six opcodes must come in pairs, with */
1918 OP_TYPEPLUS, /* 87 the minimizing one second. These codes must */
1919 OP_TYPEMINPLUS, /* 88 be in exactly the same order as those above. */
1920 OP_TYPEQUERY, /* 89 */
1921 OP_TYPEMINQUERY, /* 90 */
1922
1923 OP_TYPEUPTO, /* 91 From 0 to n matches */
1924 OP_TYPEMINUPTO, /* 92 */
1925 OP_TYPEEXACT, /* 93 Exactly n matches */
1926
1927 OP_TYPEPOSSTAR, /* 94 Possessified versions */
1928 OP_TYPEPOSPLUS, /* 95 */
1929 OP_TYPEPOSQUERY, /* 96 */
1930 OP_TYPEPOSUPTO, /* 97 */
1931
1932 /* These are used for character classes and back references; only the
1933 first six are the same as the sets above. */
1934
1935 OP_CRSTAR, /* 98 The maximizing and minimizing versions of */
1936 OP_CRMINSTAR, /* 99 all these opcodes must come in pairs, with */
1937 OP_CRPLUS, /* 100 the minimizing one second. These codes must */
1938 OP_CRMINPLUS, /* 101 be in exactly the same order as those above. */
1939 OP_CRQUERY, /* 102 */
1940 OP_CRMINQUERY, /* 103 */
1941
1942 OP_CRRANGE, /* 104 These are different to the three sets above. */
1943 OP_CRMINRANGE, /* 105 */
1944
1945 /* End of quantifier opcodes */
1946
1947 OP_CLASS, /* 106 Match a character class, chars < 256 only */
1948 OP_NCLASS, /* 107 Same, but the bitmap was created from a negative
1949 class - the difference is relevant only when a
1950 character > 255 is encountered. */
1951 OP_XCLASS, /* 108 Extended class for handling > 255 chars within the
1952 class. This does both positive and negative. */
1953 OP_REF, /* 109 Match a back reference, casefully */
1954 OP_REFI, /* 110 Match a back reference, caselessly */
1955 OP_RECURSE, /* 111 Match a numbered subpattern (possibly recursive) */
1956 OP_CALLOUT, /* 112 Call out to external function if provided */
1957
1958 OP_ALT, /* 113 Start of alternation */
1959 OP_KET, /* 114 End of group that doesn't have an unbounded repeat */
1960 OP_KETRMAX, /* 115 These two must remain together and in this */
1961 OP_KETRMIN, /* 116 order. They are for groups the repeat for ever. */
1962 OP_KETRPOS, /* 117 Possessive unlimited repeat. */
1963
1964 /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
1965 asserts must remain in order. */
1966
1967 OP_REVERSE, /* 118 Move pointer back - used in lookbehind assertions */
1968 OP_ASSERT, /* 119 Positive lookahead */
1969 OP_ASSERT_NOT, /* 120 Negative lookahead */
1970 OP_ASSERTBACK, /* 121 Positive lookbehind */
1971 OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
1972
1973 /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
1974 after the assertions, with ONCE first, as there's a test for >= ONCE for a
1975 subpattern that isn't an assertion. The POS versions must immediately follow
1976 the non-POS versions in each case. */
1977
1978 OP_ONCE, /* 123 Atomic group, contains captures */
1979 OP_ONCE_NC, /* 124 Atomic group containing no captures */
1980 OP_BRA, /* 125 Start of non-capturing bracket */
1981 OP_BRAPOS, /* 126 Ditto, with unlimited, possessive repeat */
1982 OP_CBRA, /* 127 Start of capturing bracket */
1983 OP_CBRAPOS, /* 128 Ditto, with unlimited, possessive repeat */
1984 OP_COND, /* 129 Conditional group */
1985
1986 /* These five must follow the previous five, in the same order. There's a
1987 check for >= SBRA to distinguish the two sets. */
1988
1989 OP_SBRA, /* 130 Start of non-capturing bracket, check empty */
1990 OP_SBRAPOS, /* 131 Ditto, with unlimited, possessive repeat */
1991 OP_SCBRA, /* 132 Start of capturing bracket, check empty */
1992 OP_SCBRAPOS, /* 133 Ditto, with unlimited, possessive repeat */
1993 OP_SCOND, /* 134 Conditional group, check empty */
1994
1995 /* The next two pairs must (respectively) be kept together. */
1996
1997 OP_CREF, /* 135 Used to hold a capture number as condition */
1998 OP_NCREF, /* 136 Same, but generated by a name reference*/
1999 OP_RREF, /* 137 Used to hold a recursion number as condition */
2000 OP_NRREF, /* 138 Same, but generated by a name reference*/
2001 OP_DEF, /* 139 The DEFINE condition */
2002
2003 OP_BRAZERO, /* 140 These two must remain together and in this */
2004 OP_BRAMINZERO, /* 141 order. */
2005 OP_BRAPOSZERO, /* 142 */
2006
2007 /* These are backtracking control verbs */
2008
2009 OP_MARK, /* 143 always has an argument */
2010 OP_PRUNE, /* 144 */
2011 OP_PRUNE_ARG, /* 145 same, but with argument */
2012 OP_SKIP, /* 146 */
2013 OP_SKIP_ARG, /* 147 same, but with argument */
2014 OP_THEN, /* 148 */
2015 OP_THEN_ARG, /* 149 same, but with argument */
2016 OP_COMMIT, /* 150 */
2017
2018 /* These are forced failure and success verbs */
2019
2020 OP_FAIL, /* 151 */
2021 OP_ACCEPT, /* 152 */
2022 OP_ASSERT_ACCEPT, /* 153 Used inside assertions */
2023 OP_CLOSE, /* 154 Used before OP_ACCEPT to close open captures */
2024
2025 /* This is used to skip a subpattern with a {0} quantifier */
2026
2027 OP_SKIPZERO, /* 155 */
2028
2029 /* This is not an opcode, but is used to check that tables indexed by opcode
2030 are the correct length, in order to catch updating errors - there have been
2031 some in the past. */
2032
2033 OP_TABLE_LENGTH
2034 };
2035
2036 /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2037 definitions that follow must also be updated to match. There are also tables
2038 called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
2039
2040
2041 /* This macro defines textual names for all the opcodes. These are used only
2042 for debugging, and some of them are only partial names. The macro is referenced
2043 only in pcre_printint.c, which fills out the full names in many cases (and in
2044 some cases doesn't actually use these names at all). */
2045
2046 #define OP_NAME_LIST \
2047 "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \
2048 "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \
2049 "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \
2050 "extuni", "\\Z", "\\z", \
2051 "^", "^", "$", "$", "char", "chari", "not", "noti", \
2052 "*", "*?", "+", "+?", "?", "??", \
2053 "{", "{", "{", \
2054 "*+","++", "?+", "{", \
2055 "*", "*?", "+", "+?", "?", "??", \
2056 "{", "{", "{", \
2057 "*+","++", "?+", "{", \
2058 "*", "*?", "+", "+?", "?", "??", \
2059 "{", "{", "{", \
2060 "*+","++", "?+", "{", \
2061 "*", "*?", "+", "+?", "?", "??", \
2062 "{", "{", "{", \
2063 "*+","++", "?+", "{", \
2064 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
2065 "*+","++", "?+", "{", \
2066 "*", "*?", "+", "+?", "?", "??", "{", "{", \
2067 "class", "nclass", "xclass", "Ref", "Refi", \
2068 "Recurse", "Callout", \
2069 "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos", \
2070 "Reverse", "Assert", "Assert not", "AssertB", "AssertB not", \
2071 "Once", "Once_NC", \
2072 "Bra", "BraPos", "CBra", "CBraPos", \
2073 "Cond", \
2074 "SBra", "SBraPos", "SCBra", "SCBraPos", \
2075 "SCond", \
2076 "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def", \
2077 "Brazero", "Braminzero", "Braposzero", \
2078 "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP", \
2079 "*THEN", "*THEN", "*COMMIT", "*FAIL", \
2080 "*ACCEPT", "*ASSERT_ACCEPT", \
2081 "Close", "Skip zero"
2082
2083
2084 /* This macro defines the length of fixed length operations in the compiled
2085 regex. The lengths are used when searching for specific things, and also in the
2086 debugging printing of a compiled regex. We use a macro so that it can be
2087 defined close to the definitions of the opcodes themselves.
2088
2089 As things have been extended, some of these are no longer fixed lenths, but are
2090 minima instead. For example, the length of a single-character repeat may vary
2091 in UTF-8 mode. The code that uses this table must know about such things. */
2092
2093 #define OP_LENGTHS \
2094 1, /* End */ \
2095 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \
2096 1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \
2097 1, 1, 1, /* Any, AllAny, Anybyte */ \
2098 3, 3, /* \P, \p */ \
2099 1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \
2100 1, /* \X */ \
2101 1, 1, 1, 1, 1, 1, /* \Z, \z, ^, ^M, $, $M */ \
2102 2, /* Char - the minimum length */ \
2103 2, /* Chari - the minimum length */ \
2104 2, /* not */ \
2105 2, /* noti */ \
2106 /* Positive single-char repeats ** These are */ \
2107 2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
2108 2+IMM2_SIZE, 2+IMM2_SIZE, /* upto, minupto ** mode */ \
2109 2+IMM2_SIZE, /* exact */ \
2110 2, 2, 2, 2+IMM2_SIZE, /* *+, ++, ?+, upto+ */ \
2111 2, 2, 2, 2, 2, 2, /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8 */ \
2112 2+IMM2_SIZE, 2+IMM2_SIZE, /* upto I, minupto I */ \
2113 2+IMM2_SIZE, /* exact I */ \
2114 2, 2, 2, 2+IMM2_SIZE, /* *+I, ++I, ?+I, upto+I */ \
2115 /* Negative single-char repeats - only for chars < 256 */ \
2116 2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
2117 2+IMM2_SIZE, 2+IMM2_SIZE, /* NOT upto, minupto */ \
2118 2+IMM2_SIZE, /* NOT exact */ \
2119 2, 2, 2, 2+IMM2_SIZE, /* Possessive NOT *, +, ?, upto */ \
2120 2, 2, 2, 2, 2, 2, /* NOT *I, *?I, +I, +?I, ?I, ??I */ \
2121 2+IMM2_SIZE, 2+IMM2_SIZE, /* NOT upto I, minupto I */ \
2122 2+IMM2_SIZE, /* NOT exact I */ \
2123 2, 2, 2, 2+IMM2_SIZE, /* Possessive NOT *I, +I, ?I, upto I */ \
2124 /* Positive type repeats */ \
2125 2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
2126 2+IMM2_SIZE, 2+IMM2_SIZE, /* Type upto, minupto */ \
2127 2+IMM2_SIZE, /* Type exact */ \
2128 2, 2, 2, 2+IMM2_SIZE, /* Possessive *+, ++, ?+, upto+ */ \
2129 /* Character class & ref repeats */ \
2130 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
2131 1+2*IMM2_SIZE, 1+2*IMM2_SIZE, /* CRRANGE, CRMINRANGE */ \
2132 1+(32/sizeof(pcre_uchar)), /* CLASS */ \
2133 1+(32/sizeof(pcre_uchar)), /* NCLASS */ \
2134 0, /* XCLASS - variable length */ \
2135 1+IMM2_SIZE, /* REF */ \
2136 1+IMM2_SIZE, /* REFI */ \
2137 1+LINK_SIZE, /* RECURSE */ \
2138 2+2*LINK_SIZE, /* CALLOUT */ \
2139 1+LINK_SIZE, /* Alt */ \
2140 1+LINK_SIZE, /* Ket */ \
2141 1+LINK_SIZE, /* KetRmax */ \
2142 1+LINK_SIZE, /* KetRmin */ \
2143 1+LINK_SIZE, /* KetRpos */ \
2144 1+LINK_SIZE, /* Reverse */ \
2145 1+LINK_SIZE, /* Assert */ \
2146 1+LINK_SIZE, /* Assert not */ \
2147 1+LINK_SIZE, /* Assert behind */ \
2148 1+LINK_SIZE, /* Assert behind not */ \
2149 1+LINK_SIZE, /* ONCE */ \
2150 1+LINK_SIZE, /* ONCE_NC */ \
2151 1+LINK_SIZE, /* BRA */ \
2152 1+LINK_SIZE, /* BRAPOS */ \
2153 1+LINK_SIZE+IMM2_SIZE, /* CBRA */ \
2154 1+LINK_SIZE+IMM2_SIZE, /* CBRAPOS */ \
2155 1+LINK_SIZE, /* COND */ \
2156 1+LINK_SIZE, /* SBRA */ \
2157 1+LINK_SIZE, /* SBRAPOS */ \
2158 1+LINK_SIZE+IMM2_SIZE, /* SCBRA */ \
2159 1+LINK_SIZE+IMM2_SIZE, /* SCBRAPOS */ \
2160 1+LINK_SIZE, /* SCOND */ \
2161 1+IMM2_SIZE, 1+IMM2_SIZE, /* CREF, NCREF */ \
2162 1+IMM2_SIZE, 1+IMM2_SIZE, /* RREF, NRREF */ \
2163 1, /* DEF */ \
2164 1, 1, 1, /* BRAZERO, BRAMINZERO, BRAPOSZERO */ \
2165 3, 1, 3, /* MARK, PRUNE, PRUNE_ARG */ \
2166 1, 3, /* SKIP, SKIP_ARG */ \
2167 1, 3, /* THEN, THEN_ARG */ \
2168 1, 1, 1, 1, /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT */ \
2169 1+IMM2_SIZE, 1 /* CLOSE, SKIPZERO */
2170
2171 /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
2172 condition. */
2173
2174 #define RREF_ANY 0xffff
2175
2176 /* Compile time error code numbers. They are given names so that they can more
2177 easily be tracked. When a new number is added, the table called eint in
2178 pcreposix.c must be updated. */
2179
2180 enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
2181 ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
2182 ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
2183 ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2184 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2185 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2186 ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2187 ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2188
2189 /* JIT compiling modes. The function list is indexed by them. */
2190 enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2191 JIT_NUMBER_OF_COMPILE_MODES };
2192
2193 /* The real format of the start of the pcre block; the index of names and the
2194 code vector run on as long as necessary after the end. We store an explicit
2195 offset to the name table so that if a regex is compiled on one host, saved, and
2196 then run on another where the size of pointers is different, all might still
2197 be well. For the case of compiled-on-4 and run-on-8, we include an extra
2198 pointer that is always NULL. For future-proofing, a few dummy fields were
2199 originally included - even though you can never get this planning right - but
2200 there is only one left now.
2201
2202 NOTE NOTE NOTE:
2203 Because people can now save and re-use compiled patterns, any additions to this
2204 structure should be made at the end, and something earlier (e.g. a new
2205 flag in the options or one of the dummy fields) should indicate that the new
2206 fields are present. Currently PCRE always sets the dummy fields to zero.
2207 NOTE NOTE NOTE
2208 */
2209
2210 #if defined COMPILE_PCRE8
2211 #define REAL_PCRE real_pcre
2212 #elif defined COMPILE_PCRE16
2213 #define REAL_PCRE real_pcre16
2214 #elif defined COMPILE_PCRE32
2215 #define REAL_PCRE real_pcre32
2216 #endif
2217
2218 /* It is necessary to fork the struct for 32 bit, since it needs to use
2219 * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2220 * typedef since pcretest needs access to the struct of the 8-, 16-
2221 * and 32-bit variants. */
2222
2223 typedef struct real_pcre8_or_16 {
2224 pcre_uint32 magic_number;
2225 pcre_uint32 size; /* Total that was malloced */
2226 pcre_uint32 options; /* Public options */
2227 pcre_uint16 flags; /* Private flags */
2228 pcre_uint16 max_lookbehind; /* Longest lookbehind (characters) */
2229 pcre_uint16 top_bracket; /* Highest numbered group */
2230 pcre_uint16 top_backref; /* Highest numbered back reference */
2231 pcre_uint16 first_char; /* Starting character */
2232 pcre_uint16 req_char; /* This character must be seen */
2233 pcre_uint16 name_table_offset; /* Offset to name table that follows */
2234 pcre_uint16 name_entry_size; /* Size of any name items */
2235 pcre_uint16 name_count; /* Number of name items */
2236 pcre_uint16 ref_count; /* Reference count */
2237 const pcre_uint8 *tables; /* Pointer to tables or NULL for std */
2238 const pcre_uint8 *nullpad; /* NULL padding */
2239 } real_pcre8_or_16;
2240
2241 typedef struct real_pcre8_or_16 real_pcre;
2242 typedef struct real_pcre8_or_16 real_pcre16;
2243
2244 typedef struct real_pcre32 {
2245 pcre_uint32 magic_number;
2246 pcre_uint32 size; /* Total that was malloced */
2247 pcre_uint32 options; /* Public options */
2248 pcre_uint16 flags; /* Private flags */
2249 pcre_uint16 max_lookbehind; /* Longest lookbehind (characters) */
2250 pcre_uint16 top_bracket; /* Highest numbered group */
2251 pcre_uint16 top_backref; /* Highest numbered back reference */
2252 pcre_uint32 first_char; /* Starting character */
2253 pcre_uint32 req_char; /* This character must be seen */
2254 pcre_uint16 name_table_offset; /* Offset to name table that follows */
2255 pcre_uint16 name_entry_size; /* Size of any name items */
2256 pcre_uint16 name_count; /* Number of name items */
2257 pcre_uint16 ref_count; /* Reference count */
2258 pcre_uint16 dummy1; /* for later expansion */
2259 pcre_uint16 dummy2; /* for later expansion */
2260 const pcre_uint8 *tables; /* Pointer to tables or NULL for std */
2261 void *nullpad; /* for later expansion */
2262 } real_pcre32;
2263
2264 /* Assert that the size of REAL_PCRE is divisible by 8 */
2265 typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2266
2267 /* Needed in pcretest to access some fields in the real_pcre* structures
2268 * directly. They're unified for 8/16/32 bits since the structs only differ
2269 * after these fields; if that ever changes, need to fork those defines into
2270 * 8/16 and 32 bit versions. */
2271 #define REAL_PCRE_MAGIC(re) (((REAL_PCRE*)re)->magic_number)
2272 #define REAL_PCRE_SIZE(re) (((REAL_PCRE*)re)->size)
2273 #define REAL_PCRE_OPTIONS(re) (((REAL_PCRE*)re)->options)
2274 #define REAL_PCRE_FLAGS(re) (((REAL_PCRE*)re)->flags)
2275
2276 /* The format of the block used to store data from pcre_study(). The same
2277 remark (see NOTE above) about extending this structure applies. */
2278
2279 typedef struct pcre_study_data {
2280 pcre_uint32 size; /* Total that was malloced */
2281 pcre_uint32 flags; /* Private flags */
2282 pcre_uint8 start_bits[32]; /* Starting char bits */
2283 pcre_uint32 minlength; /* Minimum subject length */
2284 } pcre_study_data;
2285
2286 /* Structure for building a chain of open capturing subpatterns during
2287 compiling, so that instructions to close them can be compiled when (*ACCEPT) is
2288 encountered. This is also used to identify subpatterns that contain recursive
2289 back references to themselves, so that they can be made atomic. */
2290
2291 typedef struct open_capitem {
2292 struct open_capitem *next; /* Chain link */
2293 pcre_uint16 number; /* Capture number */
2294 pcre_uint16 flag; /* Set TRUE if recursive back ref */
2295 } open_capitem;
2296
2297 /* Structure for passing "static" information around between the functions
2298 doing the compiling, so that they are thread-safe. */
2299
2300 typedef struct compile_data {
2301 const pcre_uint8 *lcc; /* Points to lower casing table */
2302 const pcre_uint8 *fcc; /* Points to case-flipping table */
2303 const pcre_uint8 *cbits; /* Points to character type table */
2304 const pcre_uint8 *ctypes; /* Points to table of type maps */
2305 const pcre_uchar *start_workspace;/* The start of working space */
2306 const pcre_uchar *start_code; /* The start of the compiled code */
2307 const pcre_uchar *start_pattern; /* The start of the pattern */
2308 const pcre_uchar *end_pattern; /* The end of the pattern */
2309 open_capitem *open_caps; /* Chain of open capture items */
2310 pcre_uchar *hwm; /* High watermark of workspace */
2311 pcre_uchar *name_table; /* The name/number table */
2312 int names_found; /* Number of entries so far */
2313 int name_entry_size; /* Size of each entry */
2314 int workspace_size; /* Size of workspace */
2315 int bracount; /* Count of capturing parens as we compile */
2316 int final_bracount; /* Saved value after first pass */
2317 int max_lookbehind; /* Maximum lookbehind (characters) */
2318 int top_backref; /* Maximum back reference */
2319 unsigned int backref_map; /* Bitmap of low back refs */
2320 int assert_depth; /* Depth of nested assertions */
2321 int external_options; /* External (initial) options */
2322 int external_flags; /* External flag bits to be set */
2323 int req_varyopt; /* "After variable item" flag for reqbyte */
2324 BOOL had_accept; /* (*ACCEPT) encountered */
2325 BOOL had_pruneorskip; /* (*PRUNE) or (*SKIP) encountered */
2326 BOOL check_lookbehind; /* Lookbehinds need later checking */
2327 int nltype; /* Newline type */
2328 int nllen; /* Newline string length */
2329 pcre_uchar nl[4]; /* Newline string when fixed length */
2330 } compile_data;
2331
2332 /* Structure for maintaining a chain of pointers to the currently incomplete
2333 branches, for testing for left recursion while compiling. */
2334
2335 typedef struct branch_chain {
2336 struct branch_chain *outer;
2337 pcre_uchar *current_branch;
2338 } branch_chain;
2339
2340 /* Structure for items in a linked list that represents an explicit recursive
2341 call within the pattern; used by pcre_exec(). */
2342
2343 typedef struct recursion_info {
2344 struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2345 int group_num; /* Number of group that was called */
2346 int *offset_save; /* Pointer to start of saved offsets */
2347 int saved_max; /* Number of saved offsets */
2348 PCRE_PUCHAR subject_position; /* Position at start of recursion */
2349 } recursion_info;
2350
2351 /* A similar structure for pcre_dfa_exec(). */
2352
2353 typedef struct dfa_recursion_info {
2354 struct dfa_recursion_info *prevrec;
2355 int group_num;
2356 PCRE_PUCHAR subject_position;
2357 } dfa_recursion_info;
2358
2359 /* Structure for building a chain of data for holding the values of the subject
2360 pointer at the start of each subpattern, so as to detect when an empty string
2361 has been matched by a subpattern - to break infinite loops; used by
2362 pcre_exec(). */
2363
2364 typedef struct eptrblock {
2365 struct eptrblock *epb_prev;
2366 PCRE_PUCHAR epb_saved_eptr;
2367 } eptrblock;
2368
2369
2370 /* Structure for passing "static" information around between the functions
2371 doing traditional NFA matching, so that they are thread-safe. */
2372
2373 typedef struct match_data {
2374 unsigned long int match_call_count; /* As it says */
2375 unsigned long int match_limit; /* As it says */
2376 unsigned long int match_limit_recursion; /* As it says */
2377 int *offset_vector; /* Offset vector */
2378 int offset_end; /* One past the end */
2379 int offset_max; /* The maximum usable for return data */
2380 int nltype; /* Newline type */
2381 int nllen; /* Newline string length */
2382 int name_count; /* Number of names in name table */
2383 int name_entry_size; /* Size of entry in names table */
2384 pcre_uchar *name_table; /* Table of names */
2385 pcre_uchar nl[4]; /* Newline string when fixed */
2386 const pcre_uint8 *lcc; /* Points to lower casing table */
2387 const pcre_uint8 *fcc; /* Points to case-flipping table */
2388 const pcre_uint8 *ctypes; /* Points to table of type maps */
2389 BOOL offset_overflow; /* Set if too many extractions */
2390 BOOL notbol; /* NOTBOL flag */
2391 BOOL noteol; /* NOTEOL flag */
2392 BOOL utf; /* UTF-8 / UTF-16 flag */
2393 BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */
2394 BOOL use_ucp; /* PCRE_UCP flag */
2395 BOOL endonly; /* Dollar not before final \n */
2396 BOOL notempty; /* Empty string match not wanted */
2397 BOOL notempty_atstart; /* Empty string match at start not wanted */
2398 BOOL hitend; /* Hit the end of the subject at some point */
2399 BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
2400 BOOL hasthen; /* Pattern contains (*THEN) */
2401 BOOL ignore_skip_arg; /* For re-run when SKIP name not found */
2402 const pcre_uchar *start_code; /* For use when recursing */
2403 PCRE_PUCHAR start_subject; /* Start of the subject string */
2404 PCRE_PUCHAR end_subject; /* End of the subject string */
2405 PCRE_PUCHAR start_match_ptr; /* Start of matched string */
2406 PCRE_PUCHAR end_match_ptr; /* Subject position at end match */
2407 PCRE_PUCHAR start_used_ptr; /* Earliest consulted character */
2408 int partial; /* PARTIAL options */
2409 int end_offset_top; /* Highwater mark at end of match */
2410 int capture_last; /* Most recent capture number */
2411 int start_offset; /* The start offset value */
2412 int match_function_type; /* Set for certain special calls of MATCH() */
2413 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
2414 int eptrn; /* Next free eptrblock */
2415 recursion_info *recursive; /* Linked list of recursion data */
2416 void *callout_data; /* To pass back to callouts */
2417 const pcre_uchar *mark; /* Mark pointer to pass back on success */
2418 const pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2419 const pcre_uchar *once_target; /* Where to back up to for atomic groups */
2420 #ifdef NO_RECURSE
2421 void *match_frames_base; /* For remembering malloc'd frames */
2422 #endif
2423 } match_data;
2424
2425 /* A similar structure is used for the same purpose by the DFA matching
2426 functions. */
2427
2428 typedef struct dfa_match_data {
2429 const pcre_uchar *start_code; /* Start of the compiled pattern */
2430 const pcre_uchar *start_subject ; /* Start of the subject string */
2431 const pcre_uchar *end_subject; /* End of subject string */
2432 const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2433 const pcre_uint8 *tables; /* Character tables */
2434 int start_offset; /* The start offset value */
2435 int moptions; /* Match options */
2436 int poptions; /* Pattern options */
2437 int nltype; /* Newline type */
2438 int nllen; /* Newline string length */
2439 pcre_uchar nl[4]; /* Newline string when fixed */
2440 void *callout_data; /* To pass back to callouts */
2441 dfa_recursion_info *recursive; /* Linked list of recursion data */
2442 } dfa_match_data;
2443
2444 /* Bit definitions for entries in the pcre_ctypes table. */
2445
2446 #define ctype_space 0x01
2447 #define ctype_letter 0x02
2448 #define ctype_digit 0x04
2449 #define ctype_xdigit 0x08
2450 #define ctype_word 0x10 /* alphanumeric or '_' */
2451 #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
2452
2453 /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
2454 of bits for a class map. Some classes are built by combining these tables. */
2455
2456 #define cbit_space 0 /* [:space:] or \s */
2457 #define cbit_xdigit 32 /* [:xdigit:] */
2458 #define cbit_digit 64 /* [:digit:] or \d */
2459 #define cbit_upper 96 /* [:upper:] */
2460 #define cbit_lower 128 /* [:lower:] */
2461 #define cbit_word 160 /* [:word:] or \w */
2462 #define cbit_graph 192 /* [:graph:] */
2463 #define cbit_print 224 /* [:print:] */
2464 #define cbit_punct 256 /* [:punct:] */
2465 #define cbit_cntrl 288 /* [:cntrl:] */
2466 #define cbit_length 320 /* Length of the cbits table */
2467
2468 /* Offsets of the various tables from the base tables pointer, and
2469 total length. */
2470
2471 #define lcc_offset 0
2472 #define fcc_offset 256
2473 #define cbits_offset 512
2474 #define ctypes_offset (cbits_offset + cbit_length)
2475 #define tables_length (ctypes_offset + 256)
2476
2477 /* Internal function and data prefixes. */
2478
2479 #if defined COMPILE_PCRE8
2480 #ifndef PUBL
2481 #define PUBL(name) pcre_##name
2482 #endif
2483 #ifndef PRIV
2484 #define PRIV(name) _pcre_##name
2485 #endif
2486 #elif defined COMPILE_PCRE16
2487 #ifndef PUBL
2488 #define PUBL(name) pcre16_##name
2489 #endif
2490 #ifndef PRIV
2491 #define PRIV(name) _pcre16_##name
2492 #endif
2493 #elif defined COMPILE_PCRE32
2494 #ifndef PUBL
2495 #define PUBL(name) pcre32_##name
2496 #endif
2497 #ifndef PRIV
2498 #define PRIV(name) _pcre32_##name
2499 #endif
2500 #else
2501 #error Unsupported compiling mode
2502 #endif /* COMPILE_PCRE[8|16|32] */
2503
2504 /* Layout of the UCP type table that translates property names into types and
2505 codes. Each entry used to point directly to a name, but to reduce the number of
2506 relocations in shared libraries, it now has an offset into a single string
2507 instead. */
2508
2509 typedef struct {
2510 pcre_uint16 name_offset;
2511 pcre_uint16 type;
2512 pcre_uint16 value;
2513 } ucp_type_table;
2514
2515
2516 /* Internal shared data tables. These are tables that are used by more than one
2517 of the exported public functions. They have to be "external" in the C sense,
2518 but are not part of the PCRE public API. The data for these tables is in the
2519 pcre_tables.c module. */
2520
2521 #ifdef COMPILE_PCRE8
2522 extern const int PRIV(utf8_table1)[];
2523 extern const int PRIV(utf8_table1_size);
2524 extern const int PRIV(utf8_table2)[];
2525 extern const int PRIV(utf8_table3)[];
2526 extern const pcre_uint8 PRIV(utf8_table4)[];
2527 #endif /* COMPILE_PCRE8 */
2528
2529 extern const char PRIV(utt_names)[];
2530 extern const ucp_type_table PRIV(utt)[];
2531 extern const int PRIV(utt_size);
2532
2533 extern const pcre_uint8 PRIV(OP_lengths)[];
2534 extern const pcre_uint8 PRIV(default_tables)[];
2535
2536 extern const pcre_uint32 PRIV(hspace_list)[];
2537 extern const pcre_uint32 PRIV(vspace_list)[];
2538
2539
2540 /* Internal shared functions. These are functions that are used by more than
2541 one of the exported public functions. They have to be "external" in the C
2542 sense, but are not part of the PCRE public API. */
2543
2544 /* String comparison functions. */
2545 #if defined COMPILE_PCRE8
2546
2547 #define STRCMP_UC_UC(str1, str2) \
2548 strcmp((char *)(str1), (char *)(str2))
2549 #define STRCMP_UC_C8(str1, str2) \
2550 strcmp((char *)(str1), (str2))
2551 #define STRNCMP_UC_UC(str1, str2, num) \
2552 strncmp((char *)(str1), (char *)(str2), (num))
2553 #define STRNCMP_UC_C8(str1, str2, num) \
2554 strncmp((char *)(str1), (str2), (num))
2555 #define STRLEN_UC(str) strlen((const char *)str)
2556
2557 #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2558
2559 extern int PRIV(strcmp_uc_uc)(const pcre_uchar *,
2560 const pcre_uchar *);
2561 extern int PRIV(strcmp_uc_c8)(const pcre_uchar *,
2562 const char *);
2563 extern int PRIV(strncmp_uc_uc)(const pcre_uchar *,
2564 const pcre_uchar *, unsigned int num);
2565 extern int PRIV(strncmp_uc_c8)(const pcre_uchar *,
2566 const char *, unsigned int num);
2567 extern unsigned int PRIV(strlen_uc)(const pcre_uchar *str);
2568
2569 #define STRCMP_UC_UC(str1, str2) \
2570 PRIV(strcmp_uc_uc)((str1), (str2))
2571 #define STRCMP_UC_C8(str1, str2) \
2572 PRIV(strcmp_uc_c8)((str1), (str2))
2573 #define STRNCMP_UC_UC(str1, str2, num) \
2574 PRIV(strncmp_uc_uc)((str1), (str2), (num))
2575 #define STRNCMP_UC_C8(str1, str2, num) \
2576 PRIV(strncmp_uc_c8)((str1), (str2), (num))
2577 #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2578
2579 #endif /* COMPILE_PCRE[8|16|32] */
2580
2581 extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2582 extern BOOL PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2583 int *, BOOL);
2584 extern int PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2585 extern int PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2586 extern BOOL PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2587 int *, BOOL);
2588 extern BOOL PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2589
2590 #ifdef SUPPORT_JIT
2591 extern void PRIV(jit_compile)(const REAL_PCRE *,
2592 PUBL(extra) *, int);
2593 extern int PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2594 const pcre_uchar *, int, int, int, int *, int);
2595 extern void PRIV(jit_free)(void *);
2596 extern int PRIV(jit_get_size)(void *);
2597 extern const char* PRIV(jit_get_target)(void);
2598 #endif
2599
2600 /* Unicode character database (UCD) */
2601
2602 typedef struct {
2603 pcre_uint8 script; /* ucp_Arabic, etc. */
2604 pcre_uint8 chartype; /* ucp_Cc, etc. (general categories) */
2605 pcre_uint8 gbprop; /* ucp_gbControl, etc. (grapheme break property) */
2606 pcre_uint8 caseset; /* offset to multichar other cases or zero */
2607 pcre_int32 other_case; /* offset to other case, or zero if none */
2608 } ucd_record;
2609
2610 extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2611 extern const ucd_record PRIV(ucd_records)[];
2612 extern const pcre_uint8 PRIV(ucd_stage1)[];
2613 extern const pcre_uint16 PRIV(ucd_stage2)[];
2614 extern const int PRIV(ucp_gentype)[];
2615 extern const pcre_uint32 PRIV(ucp_gbtable)[];
2616 #ifdef SUPPORT_JIT
2617 extern const int PRIV(ucp_typerange)[];
2618 #endif
2619
2620 #ifdef SUPPORT_UCP
2621 /* UCD access macros */
2622
2623 #define UCD_BLOCK_SIZE 128
2624 #define GET_UCD(ch) (PRIV(ucd_records) + \
2625 PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2626 UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2627
2628 #define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype
2629 #define UCD_SCRIPT(ch) GET_UCD(ch)->script
2630 #define UCD_CATEGORY(ch) PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2631 #define UCD_GRAPHBREAK(ch) GET_UCD(ch)->gbprop
2632 #define UCD_CASESET(ch) GET_UCD(ch)->caseset
2633 #define UCD_OTHERCASE(ch) ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2634
2635 #endif /* SUPPORT_UCP */
2636
2637 #endif
2638
2639 /* 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