1 /*************************************************
2 * Perl-Compatible Regular Expressions *
3 *************************************************/
4
5
6 /* PCRE is a library of functions to support regular expressions whose syntax
7 and semantics are as close as possible to those of the Perl 5 language.
8
9 Written by Philip Hazel
10 Copyright (c) 1997-2012 University of Cambridge
11
12 -----------------------------------------------------------------------------
13 Redistribution and use in source and binary forms, with or without
14 modification, are permitted provided that the following conditions are met:
15
16 * Redistributions of source code must retain the above copyright notice,
17 this list of conditions and the following disclaimer.
18
19 * Redistributions in binary form must reproduce the above copyright
20 notice, this list of conditions and the following disclaimer in the
21 documentation and/or other materials provided with the distribution.
22
23 * Neither the name of the University of Cambridge nor the names of its
24 contributors may be used to endorse or promote products derived from
25 this software without specific prior written permission.
26
27 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
28 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
29 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
30 ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
31 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
34 INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
35 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
36 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
37 POSSIBILITY OF SUCH DAMAGE.
38 -----------------------------------------------------------------------------
39 */
40
41 /* This header contains definitions that are shared between the different
42 modules, but which are not relevant to the exported API. This includes some
43 functions whose names all begin with "_pcre_", "_pcre16_" or "_pcre32_"
44 depending on the PRIV macro. */
45
46 #ifndef PCRE_INTERNAL_H
47 #define PCRE_INTERNAL_H
48
49 /* Define PCRE_DEBUG to get debugging output on stdout. */
50
51 #if 0
52 #define PCRE_DEBUG
53 #endif
54
55 /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
56
57 #if !defined COMPILE_PCRE16 && !defined COMPILE_PCRE32
58 #define COMPILE_PCRE8
59 #endif
60
61 /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
62 "configure" script ensures this, but not everybody uses "configure". */
63
64 #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
65 #define SUPPORT_UTF 1
66 #endif
67
68 /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
69 reasons with existing code. */
70
71 #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
72 #define SUPPORT_UTF 1
73 #endif
74
75 /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
76 Until then we define it if SUPPORT_UTF is defined. */
77
78 #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
79 #define SUPPORT_UTF8 1
80 #endif
81
82 /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
83 script prevents both being selected, but not everybody uses "configure". */
84
85 #if defined EBCDIC && defined SUPPORT_UTF
86 #error The use of both EBCDIC and SUPPORT_UTF is not supported.
87 #endif
88
89 /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
90 inline, and there are *still* stupid compilers about that don't like indented
91 pre-processor statements, or at least there were when I first wrote this. After
92 all, it had only been about 10 years then...
93
94 It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so
95 be absolutely sure we get our version. */
96
97 #undef DPRINTF
98 #ifdef PCRE_DEBUG
99 #define DPRINTF(p) printf p
100 #else
101 #define DPRINTF(p) /* Nothing */
102 #endif
103
104
105 /* Standard C headers plus the external interface definition. The only time
106 setjmp and stdarg are used is when NO_RECURSE is set. */
107
108 #include <ctype.h>
109 #include <limits.h>
110 #include <stddef.h>
111 #include <stdio.h>
112 #include <stdlib.h>
113 #include <string.h>
114
115 /* Valgrind (memcheck) support */
116
117 #ifdef SUPPORT_VALGRIND
118 #include <valgrind/memcheck.h>
119 #endif
120
121 /* When compiling a DLL for Windows, the exported symbols have to be declared
122 using some MS magic. I found some useful information on this web page:
123 http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
124 information there, using __declspec(dllexport) without "extern" we have a
125 definition; with "extern" we have a declaration. The settings here override the
126 setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL,
127 which is all that is needed for applications (they just import the symbols). We
128 use:
129
130 PCRE_EXP_DECL for declarations
131 PCRE_EXP_DEFN for definitions of exported functions
132 PCRE_EXP_DATA_DEFN for definitions of exported variables
133
134 The reason for the two DEFN macros is that in non-Windows environments, one
135 does not want to have "extern" before variable definitions because it leads to
136 compiler warnings. So we distinguish between functions and variables. In
137 Windows, the two should always be the same.
138
139 The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest,
140 which is an application, but needs to import this file in order to "peek" at
141 internals, can #include pcre.h first to get an application's-eye view.
142
143 In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon,
144 special-purpose environments) might want to stick other stuff in front of
145 exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and
146 PCRE_EXP_DATA_DEFN only if they are not already set. */
147
148 #ifndef PCRE_EXP_DECL
149 # ifdef _WIN32
150 # ifndef PCRE_STATIC
151 # define PCRE_EXP_DECL extern __declspec(dllexport)
152 # define PCRE_EXP_DEFN __declspec(dllexport)
153 # define PCRE_EXP_DATA_DEFN __declspec(dllexport)
154 # else
155 # define PCRE_EXP_DECL extern
156 # define PCRE_EXP_DEFN
157 # define PCRE_EXP_DATA_DEFN
158 # endif
159 # else
160 # ifdef __cplusplus
161 # define PCRE_EXP_DECL extern "C"
162 # else
163 # define PCRE_EXP_DECL extern
164 # endif
165 # ifndef PCRE_EXP_DEFN
166 # define PCRE_EXP_DEFN PCRE_EXP_DECL
167 # endif
168 # ifndef PCRE_EXP_DATA_DEFN
169 # define PCRE_EXP_DATA_DEFN
170 # endif
171 # endif
172 #endif
173
174 /* When compiling with the MSVC compiler, it is sometimes necessary to include
175 a "calling convention" before exported function names. (This is secondhand
176 information; I know nothing about MSVC myself). For example, something like
177
178 void __cdecl function(....)
179
180 might be needed. In order so make this easy, all the exported functions have
181 PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
182 set, we ensure here that it has no effect. */
183
184 #ifndef PCRE_CALL_CONVENTION
185 #define PCRE_CALL_CONVENTION
186 #endif
187
188 /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
189 cannot determine these outside the compilation (e.g. by running a program as
190 part of "configure") because PCRE is often cross-compiled for use on other
191 systems. Instead we make use of the maximum sizes that are available at
192 preprocessor time in standard C environments. */
193
194 typedef unsigned char pcre_uint8;
195
196 #if USHRT_MAX == 65535
197 typedef unsigned short pcre_uint16;
198 typedef short pcre_int16;
199 #elif UINT_MAX == 65535
200 typedef unsigned int pcre_uint16;
201 typedef int pcre_int16;
202 #else
203 # error Cannot determine a type for 16-bit unsigned integers
204 #endif
205
206 #if UINT_MAX == 4294967295
207 typedef unsigned int pcre_uint32;
208 typedef int pcre_int32;
209 #elif ULONG_MAX == 4294967295
210 typedef unsigned long int pcre_uint32;
211 typedef long int pcre_int32;
212 #else
213 # error Cannot determine a type for 32-bit unsigned integers
214 #endif
215
216 /* When checking for integer overflow in pcre_compile(), we need to handle
217 large integers. If a 64-bit integer type is available, we can use that.
218 Otherwise we have to cast to double, which of course requires floating point
219 arithmetic. Handle this by defining a macro for the appropriate type. If
220 stdint.h is available, include it; it may define INT64_MAX. Systems that do not
221 have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
222 by "configure". */
223
224 #ifdef PHP_WIN32
225 #include "win32/php_stdint.h"
226 #elif HAVE_STDINT_H
227 #include <stdint.h>
228 #elif HAVE_INTTYPES_H
229 #include <inttypes.h>
230 #endif
231
232 #if defined INT64_MAX || defined int64_t
233 #define INT64_OR_DOUBLE int64_t
234 #else
235 #define INT64_OR_DOUBLE double
236 #endif
237
238 /* All character handling must be done as unsigned characters. Otherwise there
239 are problems with top-bit-set characters and functions such as isspace().
240 However, we leave the interface to the outside world as char * or short *,
241 because that should make things easier for callers. This character type is
242 called pcre_uchar.
243
244 The IN_UCHARS macro multiply its argument with the byte size of the current
245 pcre_uchar type. Useful for memcpy and such operations, whose require the
246 byte size of their input/output buffers.
247
248 The MAX_255 macro checks whether its pcre_uchar input is less than 256.
249
250 The TABLE_GET macro is designed for accessing elements of tables whose contain
251 exactly 256 items. When the character is able to contain more than 256
252 items, some check is needed before accessing these tables.
253 */
254
255 #if defined COMPILE_PCRE8
256
257 typedef unsigned char pcre_uchar;
258 #define IN_UCHARS(x) (x)
259 #define MAX_255(c) 1
260 #define TABLE_GET(c, table, default) ((table)[c])
261
262 #elif defined COMPILE_PCRE16
263
264 #if USHRT_MAX != 65535
265 /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
266 pcre.h(.in) and disable (comment out) this message. */
267 #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
268 #endif
269
270 typedef pcre_uint16 pcre_uchar;
271 #define UCHAR_SHIFT (1)
272 #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
273 #define MAX_255(c) ((c) <= 255u)
274 #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
275
276 #elif defined COMPILE_PCRE32
277
278 typedef pcre_uint32 pcre_uchar;
279 #define UCHAR_SHIFT (2)
280 #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
281 #define MAX_255(c) ((c) <= 255u)
282 #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
283
284 #else
285 #error Unsupported compiling mode
286 #endif /* COMPILE_PCRE[8|16|32] */
287
288 /* This is an unsigned int value that no character can ever have. UTF-8
289 characters only go up to 0x7fffffff (though Unicode doesn't go beyond
290 0x0010ffff). */
291
292 #define NOTACHAR 0xffffffff
293
294 /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
295 "any" and "anycrlf" at present). The following macros are used to package up
296 testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
297 modules to indicate in which datablock the parameters exist, and what the
298 start/end of string field names are. */
299
300 #define NLTYPE_FIXED 0 /* Newline is a fixed length string */
301 #define NLTYPE_ANY 1 /* Newline is any Unicode line ending */
302 #define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */
303
304 /* This macro checks for a newline at the given position */
305
306 #define IS_NEWLINE(p) \
307 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
308 ((p) < NLBLOCK->PSEND && \
309 PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
310 &(NLBLOCK->nllen), utf)) \
311 : \
312 ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
313 RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
314 (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1]) \
315 ) \
316 )
317
318 /* This macro checks for a newline immediately preceding the given position */
319
320 #define WAS_NEWLINE(p) \
321 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
322 ((p) > NLBLOCK->PSSTART && \
323 PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
324 &(NLBLOCK->nllen), utf)) \
325 : \
326 ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
327 RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] && \
328 (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
329 ) \
330 )
331
332 /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
333 with a custom type. This makes it possible, for example, to allow pcre_exec()
334 to process subject strings that are discontinuous by using a smart pointer
335 class. It must always be possible to inspect all of the subject string in
336 pcre_exec() because of the way it backtracks. Two macros are required in the
337 normal case, for sign-unspecified and unsigned char pointers. The former is
338 used for the external interface and appears in pcre.h, which is why its name
339 must begin with PCRE_. */
340
341 #ifdef CUSTOM_SUBJECT_PTR
342 #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
343 #else
344 #define PCRE_PUCHAR const pcre_uchar *
345 #endif
346
347 /* Include the public PCRE header and the definitions of UCP character property
348 values. */
349
350 #include "pcre.h"
351 #include "ucp.h"
352
353 #ifdef COMPILE_PCRE32
354 /* Assert that the public PCRE_UCHAR32 is a 32-bit type */
355 typedef int __assert_pcre_uchar32_size[sizeof(PCRE_UCHAR32) == 4 ? 1 : -1];
356 #endif
357
358 /* When compiling for use with the Virtual Pascal compiler, these functions
359 need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
360 option on the command line. */
361
362 #ifdef VPCOMPAT
363 #define strlen(s) _strlen(s)
364 #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
365 #define memcmp(s,c,n) _memcmp(s,c,n)
366 #define memcpy(d,s,n) _memcpy(d,s,n)
367 #define memmove(d,s,n) _memmove(d,s,n)
368 #define memset(s,c,n) _memset(s,c,n)
369 #else /* VPCOMPAT */
370
371 /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
372 define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
373 is set. Otherwise, include an emulating function for those systems that have
374 neither (there some non-Unix environments where this is the case). */
375
376 #ifndef HAVE_MEMMOVE
377 #undef memmove /* some systems may have a macro */
378 #ifdef HAVE_BCOPY
379 #define memmove(a, b, c) bcopy(b, a, c)
380 #else /* HAVE_BCOPY */
381 static void *
pcre_memmove(void * d,const void * s,size_t n)382 pcre_memmove(void *d, const void *s, size_t n)
383 {
384 size_t i;
385 unsigned char *dest = (unsigned char *)d;
386 const unsigned char *src = (const unsigned char *)s;
387 if (dest > src)
388 {
389 dest += n;
390 src += n;
391 for (i = 0; i < n; ++i) *(--dest) = *(--src);
392 return (void *)dest;
393 }
394 else
395 {
396 for (i = 0; i < n; ++i) *dest++ = *src++;
397 return (void *)(dest - n);
398 }
399 }
400 #define memmove(a, b, c) pcre_memmove(a, b, c)
401 #endif /* not HAVE_BCOPY */
402 #endif /* not HAVE_MEMMOVE */
403 #endif /* not VPCOMPAT */
404
405
406 /* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
407 in big-endian order) by default. These are used, for example, to link from the
408 start of a subpattern to its alternatives and its end. The use of 2 bytes per
409 offset limits the size of the compiled regex to around 64K, which is big enough
410 for almost everybody. However, I received a request for an even bigger limit.
411 For this reason, and also to make the code easier to maintain, the storing and
412 loading of offsets from the byte string is now handled by the macros that are
413 defined here.
414
415 The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
416 the config.h file, but can be overridden by using -D on the command line. This
417 is automated on Unix systems via the "configure" command. */
418
419 #if defined COMPILE_PCRE8
420
421 #if LINK_SIZE == 2
422
423 #define PUT(a,n,d) \
424 (a[n] = (d) >> 8), \
425 (a[(n)+1] = (d) & 255)
426
427 #define GET(a,n) \
428 (((a)[n] << 8) | (a)[(n)+1])
429
430 #define MAX_PATTERN_SIZE (1 << 16)
431
432
433 #elif LINK_SIZE == 3
434
435 #define PUT(a,n,d) \
436 (a[n] = (d) >> 16), \
437 (a[(n)+1] = (d) >> 8), \
438 (a[(n)+2] = (d) & 255)
439
440 #define GET(a,n) \
441 (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
442
443 #define MAX_PATTERN_SIZE (1 << 24)
444
445
446 #elif LINK_SIZE == 4
447
448 #define PUT(a,n,d) \
449 (a[n] = (d) >> 24), \
450 (a[(n)+1] = (d) >> 16), \
451 (a[(n)+2] = (d) >> 8), \
452 (a[(n)+3] = (d) & 255)
453
454 #define GET(a,n) \
455 (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
456
457 /* Keep it positive */
458 #define MAX_PATTERN_SIZE (1 << 30)
459
460 #else
461 #error LINK_SIZE must be either 2, 3, or 4
462 #endif
463
464 #elif defined COMPILE_PCRE16
465
466 #if LINK_SIZE == 2
467
468 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
469 #undef LINK_SIZE
470 #define LINK_SIZE 1
471
472 #define PUT(a,n,d) \
473 (a[n] = (d))
474
475 #define GET(a,n) \
476 (a[n])
477
478 #define MAX_PATTERN_SIZE (1 << 16)
479
480 #elif LINK_SIZE == 3 || LINK_SIZE == 4
481
482 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
483 #undef LINK_SIZE
484 #define LINK_SIZE 2
485
486 #define PUT(a,n,d) \
487 (a[n] = (d) >> 16), \
488 (a[(n)+1] = (d) & 65535)
489
490 #define GET(a,n) \
491 (((a)[n] << 16) | (a)[(n)+1])
492
493 /* Keep it positive */
494 #define MAX_PATTERN_SIZE (1 << 30)
495
496 #else
497 #error LINK_SIZE must be either 2, 3, or 4
498 #endif
499
500 #elif defined COMPILE_PCRE32
501
502 /* Only supported LINK_SIZE is 4 */
503 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
504 #undef LINK_SIZE
505 #define LINK_SIZE 1
506
507 #define PUT(a,n,d) \
508 (a[n] = (d))
509
510 #define GET(a,n) \
511 (a[n])
512
513 /* Keep it positive */
514 #define MAX_PATTERN_SIZE (1 << 30)
515
516 #else
517 #error Unsupported compiling mode
518 #endif /* COMPILE_PCRE[8|16|32] */
519
520 /* Convenience macro defined in terms of the others */
521
522 #define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
523
524
525 /* PCRE uses some other 2-byte quantities that do not change when the size of
526 offsets changes. There are used for repeat counts and for other things such as
527 capturing parenthesis numbers in back references. */
528
529 #if defined COMPILE_PCRE8
530
531 #define IMM2_SIZE 2
532
533 #define PUT2(a,n,d) \
534 a[n] = (d) >> 8; \
535 a[(n)+1] = (d) & 255
536
537 /* For reasons that I do not understand, the expression in this GET2 macro is
538 treated by gcc as a signed expression, even when a is declared as unsigned. It
539 seems that any kind of arithmetic results in a signed value. */
540
541 #define GET2(a,n) \
542 (unsigned int)(((a)[n] << 8) | (a)[(n)+1])
543
544 #elif defined COMPILE_PCRE16
545
546 #define IMM2_SIZE 1
547
548 #define PUT2(a,n,d) \
549 a[n] = d
550
551 #define GET2(a,n) \
552 a[n]
553
554 #elif defined COMPILE_PCRE32
555
556 #define IMM2_SIZE 1
557
558 #define PUT2(a,n,d) \
559 a[n] = d
560
561 #define GET2(a,n) \
562 a[n]
563
564 #else
565 #error Unsupported compiling mode
566 #endif /* COMPILE_PCRE[8|16|32] */
567
568 #define PUT2INC(a,n,d) PUT2(a,n,d), a += IMM2_SIZE
569
570 /* The maximum length of a MARK name is currently one data unit; it may be
571 changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
572
573 #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
574 #define MAX_MARK ((1u << 16) - 1)
575 #else
576 #define MAX_MARK ((1u << 8) - 1)
577 #endif
578
579 /* When UTF encoding is being used, a character is no longer just a single
580 byte. The macros for character handling generate simple sequences when used in
581 character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
582 and other macros are not used when UTF is not supported, so they are not
583 defined. To make sure they can never even appear when UTF support is omitted,
584 we don't even define them. */
585
586 #ifndef SUPPORT_UTF
587
588 /* #define MAX_VALUE_FOR_SINGLE_CHAR */
589 /* #define HAS_EXTRALEN(c) */
590 /* #define GET_EXTRALEN(c) */
591 /* #define NOT_FIRSTCHAR(c) */
592 #define GETCHAR(c, eptr) c = *eptr;
593 #define GETCHARTEST(c, eptr) c = *eptr;
594 #define GETCHARINC(c, eptr) c = *eptr++;
595 #define GETCHARINCTEST(c, eptr) c = *eptr++;
596 #define GETCHARLEN(c, eptr, len) c = *eptr;
597 #define RAWUCHAR(eptr) (*(eptr))
598 #define RAWUCHARINC(eptr) (*(eptr)++)
599 #define RAWUCHARTEST(eptr) (*(eptr))
600 #define RAWUCHARINCTEST(eptr) (*(eptr)++)
601 /* #define GETCHARLENTEST(c, eptr, len) */
602 /* #define BACKCHAR(eptr) */
603 /* #define FORWARDCHAR(eptr) */
604 /* #define ACROSSCHAR(condition, eptr, action) */
605
606 #else /* SUPPORT_UTF */
607
608 /* Tests whether the code point needs extra characters to decode. */
609
610 #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
611
612 /* Base macro to pick up the remaining bytes of a UTF-8 character, not
613 advancing the pointer. */
614
615 #define GETUTF8(c, eptr) \
616 { \
617 if ((c & 0x20) == 0) \
618 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
619 else if ((c & 0x10) == 0) \
620 c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
621 else if ((c & 0x08) == 0) \
622 c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
623 ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
624 else if ((c & 0x04) == 0) \
625 c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
626 ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
627 (eptr[4] & 0x3f); \
628 else \
629 c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
630 ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
631 ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
632 }
633
634 /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
635 the pointer. */
636
637 #define GETUTF8INC(c, eptr) \
638 { \
639 if ((c & 0x20) == 0) \
640 c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
641 else if ((c & 0x10) == 0) \
642 { \
643 c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
644 eptr += 2; \
645 } \
646 else if ((c & 0x08) == 0) \
647 { \
648 c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
649 ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
650 eptr += 3; \
651 } \
652 else if ((c & 0x04) == 0) \
653 { \
654 c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
655 ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
656 (eptr[3] & 0x3f); \
657 eptr += 4; \
658 } \
659 else \
660 { \
661 c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
662 ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
663 ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
664 eptr += 5; \
665 } \
666 }
667
668 #if defined COMPILE_PCRE8
669
670 /* These macros were originally written in the form of loops that used data
671 from the tables whose names start with PRIV(utf8_table). They were rewritten by
672 a user so as not to use loops, because in some environments this gives a
673 significant performance advantage, and it seems never to do any harm. */
674
675 /* Tells the biggest code point which can be encoded as a single character. */
676
677 #define MAX_VALUE_FOR_SINGLE_CHAR 127
678
679 /* Tests whether the code point needs extra characters to decode. */
680
681 #define HAS_EXTRALEN(c) ((c) >= 0xc0)
682
683 /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
684 Otherwise it has an undefined behaviour. */
685
686 #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
687
688 /* Returns TRUE, if the given character is not the first character
689 of a UTF sequence. */
690
691 #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
692
693 /* Get the next UTF-8 character, not advancing the pointer. This is called when
694 we know we are in UTF-8 mode. */
695
696 #define GETCHAR(c, eptr) \
697 c = *eptr; \
698 if (c >= 0xc0) GETUTF8(c, eptr);
699
700 /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
701 pointer. */
702
703 #define GETCHARTEST(c, eptr) \
704 c = *eptr; \
705 if (utf && c >= 0xc0) GETUTF8(c, eptr);
706
707 /* Get the next UTF-8 character, advancing the pointer. This is called when we
708 know we are in UTF-8 mode. */
709
710 #define GETCHARINC(c, eptr) \
711 c = *eptr++; \
712 if (c >= 0xc0) GETUTF8INC(c, eptr);
713
714 /* Get the next character, testing for UTF-8 mode, and advancing the pointer.
715 This is called when we don't know if we are in UTF-8 mode. */
716
717 #define GETCHARINCTEST(c, eptr) \
718 c = *eptr++; \
719 if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
720
721 /* Base macro to pick up the remaining bytes of a UTF-8 character, not
722 advancing the pointer, incrementing the length. */
723
724 #define GETUTF8LEN(c, eptr, len) \
725 { \
726 if ((c & 0x20) == 0) \
727 { \
728 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
729 len++; \
730 } \
731 else if ((c & 0x10) == 0) \
732 { \
733 c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
734 len += 2; \
735 } \
736 else if ((c & 0x08) == 0) \
737 {\
738 c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
739 ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
740 len += 3; \
741 } \
742 else if ((c & 0x04) == 0) \
743 { \
744 c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
745 ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
746 (eptr[4] & 0x3f); \
747 len += 4; \
748 } \
749 else \
750 {\
751 c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
752 ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
753 ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
754 len += 5; \
755 } \
756 }
757
758 /* Get the next UTF-8 character, not advancing the pointer, incrementing length
759 if there are extra bytes. This is called when we know we are in UTF-8 mode. */
760
761 #define GETCHARLEN(c, eptr, len) \
762 c = *eptr; \
763 if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
764
765 /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
766 pointer, incrementing length if there are extra bytes. This is called when we
767 do not know if we are in UTF-8 mode. */
768
769 #define GETCHARLENTEST(c, eptr, len) \
770 c = *eptr; \
771 if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
772
773 /* Returns the next uchar, not advancing the pointer. This is called when
774 we know we are in UTF mode. */
775
776 #define RAWUCHAR(eptr) \
777 (*(eptr))
778
779 /* Returns the next uchar, advancing the pointer. This is called when
780 we know we are in UTF mode. */
781
782 #define RAWUCHARINC(eptr) \
783 (*((eptr)++))
784
785 /* Returns the next uchar, testing for UTF mode, and not advancing the
786 pointer. */
787
788 #define RAWUCHARTEST(eptr) \
789 (*(eptr))
790
791 /* Returns the next uchar, testing for UTF mode, advancing the
792 pointer. */
793
794 #define RAWUCHARINCTEST(eptr) \
795 (*((eptr)++))
796
797 /* If the pointer is not at the start of a character, move it back until
798 it is. This is called only in UTF-8 mode - we don't put a test within the macro
799 because almost all calls are already within a block of UTF-8 only code. */
800
801 #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
802
803 /* Same as above, just in the other direction. */
804 #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
805
806 /* Same as above, but it allows a fully customizable form. */
807 #define ACROSSCHAR(condition, eptr, action) \
808 while((condition) && ((eptr) & 0xc0) == 0x80) action
809
810 #elif defined COMPILE_PCRE16
811
812 /* Tells the biggest code point which can be encoded as a single character. */
813
814 #define MAX_VALUE_FOR_SINGLE_CHAR 65535
815
816 /* Tests whether the code point needs extra characters to decode. */
817
818 #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
819
820 /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
821 Otherwise it has an undefined behaviour. */
822
823 #define GET_EXTRALEN(c) 1
824
825 /* Returns TRUE, if the given character is not the first character
826 of a UTF sequence. */
827
828 #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
829
830 /* Base macro to pick up the low surrogate of a UTF-16 character, not
831 advancing the pointer. */
832
833 #define GETUTF16(c, eptr) \
834 { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
835
836 /* Get the next UTF-16 character, not advancing the pointer. This is called when
837 we know we are in UTF-16 mode. */
838
839 #define GETCHAR(c, eptr) \
840 c = *eptr; \
841 if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
842
843 /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
844 pointer. */
845
846 #define GETCHARTEST(c, eptr) \
847 c = *eptr; \
848 if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
849
850 /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
851 the pointer. */
852
853 #define GETUTF16INC(c, eptr) \
854 { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
855
856 /* Get the next UTF-16 character, advancing the pointer. This is called when we
857 know we are in UTF-16 mode. */
858
859 #define GETCHARINC(c, eptr) \
860 c = *eptr++; \
861 if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
862
863 /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
864 This is called when we don't know if we are in UTF-16 mode. */
865
866 #define GETCHARINCTEST(c, eptr) \
867 c = *eptr++; \
868 if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
869
870 /* Base macro to pick up the low surrogate of a UTF-16 character, not
871 advancing the pointer, incrementing the length. */
872
873 #define GETUTF16LEN(c, eptr, len) \
874 { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
875
876 /* Get the next UTF-16 character, not advancing the pointer, incrementing
877 length if there is a low surrogate. This is called when we know we are in
878 UTF-16 mode. */
879
880 #define GETCHARLEN(c, eptr, len) \
881 c = *eptr; \
882 if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
883
884 /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
885 pointer, incrementing length if there is a low surrogate. This is called when
886 we do not know if we are in UTF-16 mode. */
887
888 #define GETCHARLENTEST(c, eptr, len) \
889 c = *eptr; \
890 if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
891
892 /* Returns the next uchar, not advancing the pointer. This is called when
893 we know we are in UTF mode. */
894
895 #define RAWUCHAR(eptr) \
896 (*(eptr))
897
898 /* Returns the next uchar, advancing the pointer. This is called when
899 we know we are in UTF mode. */
900
901 #define RAWUCHARINC(eptr) \
902 (*((eptr)++))
903
904 /* Returns the next uchar, testing for UTF mode, and not advancing the
905 pointer. */
906
907 #define RAWUCHARTEST(eptr) \
908 (*(eptr))
909
910 /* Returns the next uchar, testing for UTF mode, advancing the
911 pointer. */
912
913 #define RAWUCHARINCTEST(eptr) \
914 (*((eptr)++))
915
916 /* If the pointer is not at the start of a character, move it back until
917 it is. This is called only in UTF-16 mode - we don't put a test within the
918 macro because almost all calls are already within a block of UTF-16 only
919 code. */
920
921 #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
922
923 /* Same as above, just in the other direction. */
924 #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
925
926 /* Same as above, but it allows a fully customizable form. */
927 #define ACROSSCHAR(condition, eptr, action) \
928 if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
929
930 #elif defined COMPILE_PCRE32
931
932 /* These are trivial for the 32-bit library, since all UTF-32 characters fit
933 into one pcre_uchar unit. */
934 #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
935 #define HAS_EXTRALEN(c) (0)
936 #define GET_EXTRALEN(c) (0)
937 #define NOT_FIRSTCHAR(c) (0)
938
939 /* Get the next UTF-32 character, not advancing the pointer. This is called when
940 we know we are in UTF-32 mode. */
941
942 #define GETCHAR(c, eptr) \
943 c = *(eptr);
944
945 /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
946 pointer. */
947
948 #define GETCHARTEST(c, eptr) \
949 c = *(eptr);
950
951 /* Get the next UTF-32 character, advancing the pointer. This is called when we
952 know we are in UTF-32 mode. */
953
954 #define GETCHARINC(c, eptr) \
955 c = *((eptr)++);
956
957 /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
958 This is called when we don't know if we are in UTF-32 mode. */
959
960 #define GETCHARINCTEST(c, eptr) \
961 c = *((eptr)++);
962
963 /* Get the next UTF-32 character, not advancing the pointer, not incrementing
964 length (since all UTF-32 is of length 1). This is called when we know we are in
965 UTF-32 mode. */
966
967 #define GETCHARLEN(c, eptr, len) \
968 GETCHAR(c, eptr)
969
970 /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
971 pointer, not incrementing the length (since all UTF-32 is of length 1).
972 This is called when we do not know if we are in UTF-32 mode. */
973
974 #define GETCHARLENTEST(c, eptr, len) \
975 GETCHARTEST(c, eptr)
976
977 /* Returns the next uchar, not advancing the pointer. This is called when
978 we know we are in UTF mode. */
979
980 #define RAWUCHAR(eptr) \
981 (*(eptr))
982
983 /* Returns the next uchar, advancing the pointer. This is called when
984 we know we are in UTF mode. */
985
986 #define RAWUCHARINC(eptr) \
987 (*((eptr)++))
988
989 /* Returns the next uchar, testing for UTF mode, and not advancing the
990 pointer. */
991
992 #define RAWUCHARTEST(eptr) \
993 (*(eptr))
994
995 /* Returns the next uchar, testing for UTF mode, advancing the
996 pointer. */
997
998 #define RAWUCHARINCTEST(eptr) \
999 (*((eptr)++))
1000
1001 /* If the pointer is not at the start of a character, move it back until
1002 it is. This is called only in UTF-32 mode - we don't put a test within the
1003 macro because almost all calls are already within a block of UTF-32 only
1004 code.
1005 These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
1006
1007 #define BACKCHAR(eptr) do { } while (0)
1008
1009 /* Same as above, just in the other direction. */
1010 #define FORWARDCHAR(eptr) do { } while (0)
1011
1012 /* Same as above, but it allows a fully customizable form. */
1013 #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1014
1015 #else
1016 #error Unsupported compiling mode
1017 #endif /* COMPILE_PCRE[8|16|32] */
1018
1019 #endif /* SUPPORT_UTF */
1020
1021 /* Tests for Unicode horizontal and vertical whitespace characters must check a
1022 number of different values. Using a switch statement for this generates the
1023 fastest code (no loop, no memory access), and there are several places in the
1024 interpreter code where this happens. In order to ensure that all the case lists
1025 remain in step, we use macros so that there is only one place where the lists
1026 are defined.
1027
1028 These values are also required as lists in pcre_compile.c when processing \h,
1029 \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1030 macros that define the values are here so that all the definitions are
1031 together. The lists must be in ascending character order, terminated by
1032 NOTACHAR (which is 0xffffffff).
1033
1034 Any changes should ensure that the various macros are kept in step with each
1035 other. NOTE: The values also appear in pcre_jit_compile.c. */
1036
1037 /* ------ ASCII/Unicode environments ------ */
1038
1039 #ifndef EBCDIC
1040
1041 #define HSPACE_LIST \
1042 CHAR_HT, CHAR_SPACE, 0xa0, \
1043 0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1044 0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1045 NOTACHAR
1046
1047 #define HSPACE_MULTIBYTE_CASES \
1048 case 0x1680: /* OGHAM SPACE MARK */ \
1049 case 0x180e: /* MONGOLIAN VOWEL SEPARATOR */ \
1050 case 0x2000: /* EN QUAD */ \
1051 case 0x2001: /* EM QUAD */ \
1052 case 0x2002: /* EN SPACE */ \
1053 case 0x2003: /* EM SPACE */ \
1054 case 0x2004: /* THREE-PER-EM SPACE */ \
1055 case 0x2005: /* FOUR-PER-EM SPACE */ \
1056 case 0x2006: /* SIX-PER-EM SPACE */ \
1057 case 0x2007: /* FIGURE SPACE */ \
1058 case 0x2008: /* PUNCTUATION SPACE */ \
1059 case 0x2009: /* THIN SPACE */ \
1060 case 0x200A: /* HAIR SPACE */ \
1061 case 0x202f: /* NARROW NO-BREAK SPACE */ \
1062 case 0x205f: /* MEDIUM MATHEMATICAL SPACE */ \
1063 case 0x3000 /* IDEOGRAPHIC SPACE */
1064
1065 #define HSPACE_BYTE_CASES \
1066 case CHAR_HT: \
1067 case CHAR_SPACE: \
1068 case 0xa0 /* NBSP */
1069
1070 #define HSPACE_CASES \
1071 HSPACE_BYTE_CASES: \
1072 HSPACE_MULTIBYTE_CASES
1073
1074 #define VSPACE_LIST \
1075 CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1076
1077 #define VSPACE_MULTIBYTE_CASES \
1078 case 0x2028: /* LINE SEPARATOR */ \
1079 case 0x2029 /* PARAGRAPH SEPARATOR */
1080
1081 #define VSPACE_BYTE_CASES \
1082 case CHAR_LF: \
1083 case CHAR_VT: \
1084 case CHAR_FF: \
1085 case CHAR_CR: \
1086 case CHAR_NEL
1087
1088 #define VSPACE_CASES \
1089 VSPACE_BYTE_CASES: \
1090 VSPACE_MULTIBYTE_CASES
1091
1092 /* ------ EBCDIC environments ------ */
1093
1094 #else
1095 #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1096
1097 #define HSPACE_BYTE_CASES \
1098 case CHAR_HT: \
1099 case CHAR_SPACE
1100
1101 #define HSPACE_CASES HSPACE_BYTE_CASES
1102
1103 #ifdef EBCDIC_NL25
1104 #define VSPACE_LIST \
1105 CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1106 #else
1107 #define VSPACE_LIST \
1108 CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1109 #endif
1110
1111 #define VSPACE_BYTE_CASES \
1112 case CHAR_LF: \
1113 case CHAR_VT: \
1114 case CHAR_FF: \
1115 case CHAR_CR: \
1116 case CHAR_NEL
1117
1118 #define VSPACE_CASES VSPACE_BYTE_CASES
1119 #endif /* EBCDIC */
1120
1121 /* ------ End of whitespace macros ------ */
1122
1123
1124
1125 /* Private flags containing information about the compiled regex. They used to
1126 live at the top end of the options word, but that got almost full, so now they
1127 are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
1128 the restrictions on partial matching have been lifted. It remains for backwards
1129 compatibility. */
1130
1131 #define PCRE_MODE8 0x0001 /* compiled in 8 bit mode */
1132 #define PCRE_MODE16 0x0002 /* compiled in 16 bit mode */
1133 #define PCRE_MODE32 0x0004 /* compiled in 32 bit mode */
1134 #define PCRE_FIRSTSET 0x0010 /* first_char is set */
1135 #define PCRE_FCH_CASELESS 0x0020 /* caseless first char */
1136 #define PCRE_REQCHSET 0x0040 /* req_byte is set */
1137 #define PCRE_RCH_CASELESS 0x0080 /* caseless requested char */
1138 #define PCRE_STARTLINE 0x0100 /* start after \n for multiline */
1139 #define PCRE_NOPARTIAL 0x0200 /* can't use partial with this regex */
1140 #define PCRE_JCHANGED 0x0400 /* j option used in regex */
1141 #define PCRE_HASCRORLF 0x0800 /* explicit \r or \n in pattern */
1142 #define PCRE_HASTHEN 0x1000 /* pattern contains (*THEN) */
1143
1144 #if defined COMPILE_PCRE8
1145 #define PCRE_MODE PCRE_MODE8
1146 #elif defined COMPILE_PCRE16
1147 #define PCRE_MODE PCRE_MODE16
1148 #elif defined COMPILE_PCRE32
1149 #define PCRE_MODE PCRE_MODE32
1150 #endif
1151 #define PCRE_MODE_MASK (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1152
1153 /* Flags for the "extra" block produced by pcre_study(). */
1154
1155 #define PCRE_STUDY_MAPPED 0x0001 /* a map of starting chars exists */
1156 #define PCRE_STUDY_MINLEN 0x0002 /* a minimum length field exists */
1157
1158 /* Masks for identifying the public options that are permitted at compile
1159 time, run time, or study time, respectively. */
1160
1161 #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
1162 PCRE_NEWLINE_ANYCRLF)
1163
1164 #define PUBLIC_COMPILE_OPTIONS \
1165 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
1166 PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1167 PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
1168 PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1169 PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
1170
1171 #define PUBLIC_EXEC_OPTIONS \
1172 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1173 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
1174 PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
1175
1176 #define PUBLIC_DFA_EXEC_OPTIONS \
1177 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1178 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
1179 PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1180 PCRE_NO_START_OPTIMIZE)
1181
1182 #define PUBLIC_STUDY_OPTIONS \
1183 (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1184 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1185
1186 #define PUBLIC_JIT_EXEC_OPTIONS \
1187 (PCRE_NO_UTF8_CHECK|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|\
1188 PCRE_NOTEMPTY_ATSTART|PCRE_PARTIAL_SOFT|PCRE_PARTIAL_HARD)
1189
1190 /* Magic number to provide a small check against being handed junk. */
1191
1192 #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
1193
1194 /* This variable is used to detect a loaded regular expression
1195 in different endianness. */
1196
1197 #define REVERSED_MAGIC_NUMBER 0x45524350UL /* 'ERCP' */
1198
1199 /* The maximum remaining length of subject we are prepared to search for a
1200 req_byte match. */
1201
1202 #define REQ_BYTE_MAX 1000
1203
1204 /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1205 environments where these macros are defined elsewhere. Unfortunately, there
1206 is no way to do the same for the typedef. */
1207
1208 typedef int BOOL;
1209
1210 #ifndef FALSE
1211 #define FALSE 0
1212 #define TRUE 1
1213 #endif
1214
1215 /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
1216 character constants like '*' because the compiler would emit their EBCDIC code,
1217 which is different from their ASCII/UTF-8 code. Instead we define macros for
1218 the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
1219 is enabled. When UTF-8 support is not enabled, the definitions use character
1220 literals. Both character and string versions of each character are needed, and
1221 there are some longer strings as well.
1222
1223 This means that, on EBCDIC platforms, the PCRE library can handle either
1224 EBCDIC, or UTF-8, but not both. To support both in the same compiled library
1225 would need different lookups depending on whether PCRE_UTF8 was set or not.
1226 This would make it impossible to use characters in switch/case statements,
1227 which would reduce performance. For a theoretical use (which nobody has asked
1228 for) in a minority area (EBCDIC platforms), this is not sensible. Any
1229 application that did need both could compile two versions of the library, using
1230 macros to give the functions distinct names. */
1231
1232 #ifndef SUPPORT_UTF
1233
1234 /* UTF-8 support is not enabled; use the platform-dependent character literals
1235 so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1236 mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1237 characters, a common practice has been to use its NL (0x15) character as the
1238 line terminator in C-like processing environments. However, sometimes the LF
1239 (0x25) character is used instead, according to this Unicode document:
1240
1241 http://unicode.org/standard/reports/tr13/tr13-5.html
1242
1243 PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1244 instead. Whichever is *not* chosen is defined as NEL.
1245
1246 In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1247 same code point. */
1248
1249 #ifdef EBCDIC
1250
1251 #ifndef EBCDIC_NL25
1252 #define CHAR_NL '\x15'
1253 #define CHAR_NEL '\x25'
1254 #define STR_NL "\x15"
1255 #define STR_NEL "\x25"
1256 #else
1257 #define CHAR_NL '\x25'
1258 #define CHAR_NEL '\x15'
1259 #define STR_NL "\x25"
1260 #define STR_NEL "\x15"
1261 #endif
1262
1263 #define CHAR_LF CHAR_NL
1264 #define STR_LF STR_NL
1265
1266 #define CHAR_ESC '\047'
1267 #define CHAR_DEL '\007'
1268 #define STR_ESC "\047"
1269 #define STR_DEL "\007"
1270
1271 #else /* Not EBCDIC */
1272
1273 /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1274 compatibility. NEL is the Unicode newline character; make sure it is
1275 a positive value. */
1276
1277 #define CHAR_LF '\n'
1278 #define CHAR_NL CHAR_LF
1279 #define CHAR_NEL ((unsigned char)'\x85')
1280 #define CHAR_ESC '\033'
1281 #define CHAR_DEL '\177'
1282
1283 #define STR_LF "\n"
1284 #define STR_NL STR_LF
1285 #define STR_NEL "\x85"
1286 #define STR_ESC "\033"
1287 #define STR_DEL "\177"
1288
1289 #endif /* EBCDIC */
1290
1291 /* The remaining definitions work in both environments. */
1292
1293 #define CHAR_NULL '\0'
1294 #define CHAR_HT '\t'
1295 #define CHAR_VT '\v'
1296 #define CHAR_FF '\f'
1297 #define CHAR_CR '\r'
1298 #define CHAR_BS '\b'
1299 #define CHAR_BEL '\a'
1300
1301 #define CHAR_SPACE ' '
1302 #define CHAR_EXCLAMATION_MARK '!'
1303 #define CHAR_QUOTATION_MARK '"'
1304 #define CHAR_NUMBER_SIGN '#'
1305 #define CHAR_DOLLAR_SIGN '$'
1306 #define CHAR_PERCENT_SIGN '%'
1307 #define CHAR_AMPERSAND '&'
1308 #define CHAR_APOSTROPHE '\''
1309 #define CHAR_LEFT_PARENTHESIS '('
1310 #define CHAR_RIGHT_PARENTHESIS ')'
1311 #define CHAR_ASTERISK '*'
1312 #define CHAR_PLUS '+'
1313 #define CHAR_COMMA ','
1314 #define CHAR_MINUS '-'
1315 #define CHAR_DOT '.'
1316 #define CHAR_SLASH '/'
1317 #define CHAR_0 '0'
1318 #define CHAR_1 '1'
1319 #define CHAR_2 '2'
1320 #define CHAR_3 '3'
1321 #define CHAR_4 '4'
1322 #define CHAR_5 '5'
1323 #define CHAR_6 '6'
1324 #define CHAR_7 '7'
1325 #define CHAR_8 '8'
1326 #define CHAR_9 '9'
1327 #define CHAR_COLON ':'
1328 #define CHAR_SEMICOLON ';'
1329 #define CHAR_LESS_THAN_SIGN '<'
1330 #define CHAR_EQUALS_SIGN '='
1331 #define CHAR_GREATER_THAN_SIGN '>'
1332 #define CHAR_QUESTION_MARK '?'
1333 #define CHAR_COMMERCIAL_AT '@'
1334 #define CHAR_A 'A'
1335 #define CHAR_B 'B'
1336 #define CHAR_C 'C'
1337 #define CHAR_D 'D'
1338 #define CHAR_E 'E'
1339 #define CHAR_F 'F'
1340 #define CHAR_G 'G'
1341 #define CHAR_H 'H'
1342 #define CHAR_I 'I'
1343 #define CHAR_J 'J'
1344 #define CHAR_K 'K'
1345 #define CHAR_L 'L'
1346 #define CHAR_M 'M'
1347 #define CHAR_N 'N'
1348 #define CHAR_O 'O'
1349 #define CHAR_P 'P'
1350 #define CHAR_Q 'Q'
1351 #define CHAR_R 'R'
1352 #define CHAR_S 'S'
1353 #define CHAR_T 'T'
1354 #define CHAR_U 'U'
1355 #define CHAR_V 'V'
1356 #define CHAR_W 'W'
1357 #define CHAR_X 'X'
1358 #define CHAR_Y 'Y'
1359 #define CHAR_Z 'Z'
1360 #define CHAR_LEFT_SQUARE_BRACKET '['
1361 #define CHAR_BACKSLASH '\\'
1362 #define CHAR_RIGHT_SQUARE_BRACKET ']'
1363 #define CHAR_CIRCUMFLEX_ACCENT '^'
1364 #define CHAR_UNDERSCORE '_'
1365 #define CHAR_GRAVE_ACCENT '`'
1366 #define CHAR_a 'a'
1367 #define CHAR_b 'b'
1368 #define CHAR_c 'c'
1369 #define CHAR_d 'd'
1370 #define CHAR_e 'e'
1371 #define CHAR_f 'f'
1372 #define CHAR_g 'g'
1373 #define CHAR_h 'h'
1374 #define CHAR_i 'i'
1375 #define CHAR_j 'j'
1376 #define CHAR_k 'k'
1377 #define CHAR_l 'l'
1378 #define CHAR_m 'm'
1379 #define CHAR_n 'n'
1380 #define CHAR_o 'o'
1381 #define CHAR_p 'p'
1382 #define CHAR_q 'q'
1383 #define CHAR_r 'r'
1384 #define CHAR_s 's'
1385 #define CHAR_t 't'
1386 #define CHAR_u 'u'
1387 #define CHAR_v 'v'
1388 #define CHAR_w 'w'
1389 #define CHAR_x 'x'
1390 #define CHAR_y 'y'
1391 #define CHAR_z 'z'
1392 #define CHAR_LEFT_CURLY_BRACKET '{'
1393 #define CHAR_VERTICAL_LINE '|'
1394 #define CHAR_RIGHT_CURLY_BRACKET '}'
1395 #define CHAR_TILDE '~'
1396
1397 #define STR_HT "\t"
1398 #define STR_VT "\v"
1399 #define STR_FF "\f"
1400 #define STR_CR "\r"
1401 #define STR_BS "\b"
1402 #define STR_BEL "\a"
1403
1404 #define STR_SPACE " "
1405 #define STR_EXCLAMATION_MARK "!"
1406 #define STR_QUOTATION_MARK "\""
1407 #define STR_NUMBER_SIGN "#"
1408 #define STR_DOLLAR_SIGN "$"
1409 #define STR_PERCENT_SIGN "%"
1410 #define STR_AMPERSAND "&"
1411 #define STR_APOSTROPHE "'"
1412 #define STR_LEFT_PARENTHESIS "("
1413 #define STR_RIGHT_PARENTHESIS ")"
1414 #define STR_ASTERISK "*"
1415 #define STR_PLUS "+"
1416 #define STR_COMMA ","
1417 #define STR_MINUS "-"
1418 #define STR_DOT "."
1419 #define STR_SLASH "/"
1420 #define STR_0 "0"
1421 #define STR_1 "1"
1422 #define STR_2 "2"
1423 #define STR_3 "3"
1424 #define STR_4 "4"
1425 #define STR_5 "5"
1426 #define STR_6 "6"
1427 #define STR_7 "7"
1428 #define STR_8 "8"
1429 #define STR_9 "9"
1430 #define STR_COLON ":"
1431 #define STR_SEMICOLON ";"
1432 #define STR_LESS_THAN_SIGN "<"
1433 #define STR_EQUALS_SIGN "="
1434 #define STR_GREATER_THAN_SIGN ">"
1435 #define STR_QUESTION_MARK "?"
1436 #define STR_COMMERCIAL_AT "@"
1437 #define STR_A "A"
1438 #define STR_B "B"
1439 #define STR_C "C"
1440 #define STR_D "D"
1441 #define STR_E "E"
1442 #define STR_F "F"
1443 #define STR_G "G"
1444 #define STR_H "H"
1445 #define STR_I "I"
1446 #define STR_J "J"
1447 #define STR_K "K"
1448 #define STR_L "L"
1449 #define STR_M "M"
1450 #define STR_N "N"
1451 #define STR_O "O"
1452 #define STR_P "P"
1453 #define STR_Q "Q"
1454 #define STR_R "R"
1455 #define STR_S "S"
1456 #define STR_T "T"
1457 #define STR_U "U"
1458 #define STR_V "V"
1459 #define STR_W "W"
1460 #define STR_X "X"
1461 #define STR_Y "Y"
1462 #define STR_Z "Z"
1463 #define STR_LEFT_SQUARE_BRACKET "["
1464 #define STR_BACKSLASH "\\"
1465 #define STR_RIGHT_SQUARE_BRACKET "]"
1466 #define STR_CIRCUMFLEX_ACCENT "^"
1467 #define STR_UNDERSCORE "_"
1468 #define STR_GRAVE_ACCENT "`"
1469 #define STR_a "a"
1470 #define STR_b "b"
1471 #define STR_c "c"
1472 #define STR_d "d"
1473 #define STR_e "e"
1474 #define STR_f "f"
1475 #define STR_g "g"
1476 #define STR_h "h"
1477 #define STR_i "i"
1478 #define STR_j "j"
1479 #define STR_k "k"
1480 #define STR_l "l"
1481 #define STR_m "m"
1482 #define STR_n "n"
1483 #define STR_o "o"
1484 #define STR_p "p"
1485 #define STR_q "q"
1486 #define STR_r "r"
1487 #define STR_s "s"
1488 #define STR_t "t"
1489 #define STR_u "u"
1490 #define STR_v "v"
1491 #define STR_w "w"
1492 #define STR_x "x"
1493 #define STR_y "y"
1494 #define STR_z "z"
1495 #define STR_LEFT_CURLY_BRACKET "{"
1496 #define STR_VERTICAL_LINE "|"
1497 #define STR_RIGHT_CURLY_BRACKET "}"
1498 #define STR_TILDE "~"
1499
1500 #define STRING_ACCEPT0 "ACCEPT\0"
1501 #define STRING_COMMIT0 "COMMIT\0"
1502 #define STRING_F0 "F\0"
1503 #define STRING_FAIL0 "FAIL\0"
1504 #define STRING_MARK0 "MARK\0"
1505 #define STRING_PRUNE0 "PRUNE\0"
1506 #define STRING_SKIP0 "SKIP\0"
1507 #define STRING_THEN "THEN"
1508
1509 #define STRING_alpha0 "alpha\0"
1510 #define STRING_lower0 "lower\0"
1511 #define STRING_upper0 "upper\0"
1512 #define STRING_alnum0 "alnum\0"
1513 #define STRING_ascii0 "ascii\0"
1514 #define STRING_blank0 "blank\0"
1515 #define STRING_cntrl0 "cntrl\0"
1516 #define STRING_digit0 "digit\0"
1517 #define STRING_graph0 "graph\0"
1518 #define STRING_print0 "print\0"
1519 #define STRING_punct0 "punct\0"
1520 #define STRING_space0 "space\0"
1521 #define STRING_word0 "word\0"
1522 #define STRING_xdigit "xdigit"
1523
1524 #define STRING_DEFINE "DEFINE"
1525
1526 #define STRING_CR_RIGHTPAR "CR)"
1527 #define STRING_LF_RIGHTPAR "LF)"
1528 #define STRING_CRLF_RIGHTPAR "CRLF)"
1529 #define STRING_ANY_RIGHTPAR "ANY)"
1530 #define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)"
1531 #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"
1532 #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"
1533 #define STRING_UTF8_RIGHTPAR "UTF8)"
1534 #define STRING_UTF16_RIGHTPAR "UTF16)"
1535 #define STRING_UTF32_RIGHTPAR "UTF32)"
1536 #define STRING_UTF_RIGHTPAR "UTF)"
1537 #define STRING_UCP_RIGHTPAR "UCP)"
1538 #define STRING_NO_START_OPT_RIGHTPAR "NO_START_OPT)"
1539
1540 #else /* SUPPORT_UTF */
1541
1542 /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1543 works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
1544 only. */
1545
1546 #define CHAR_HT '\011'
1547 #define CHAR_VT '\013'
1548 #define CHAR_FF '\014'
1549 #define CHAR_CR '\015'
1550 #define CHAR_LF '\012'
1551 #define CHAR_NL CHAR_LF
1552 #define CHAR_NEL ((unsigned char)'\x85')
1553 #define CHAR_BS '\010'
1554 #define CHAR_BEL '\007'
1555 #define CHAR_ESC '\033'
1556 #define CHAR_DEL '\177'
1557
1558 #define CHAR_NULL '\0'
1559 #define CHAR_SPACE '\040'
1560 #define CHAR_EXCLAMATION_MARK '\041'
1561 #define CHAR_QUOTATION_MARK '\042'
1562 #define CHAR_NUMBER_SIGN '\043'
1563 #define CHAR_DOLLAR_SIGN '\044'
1564 #define CHAR_PERCENT_SIGN '\045'
1565 #define CHAR_AMPERSAND '\046'
1566 #define CHAR_APOSTROPHE '\047'
1567 #define CHAR_LEFT_PARENTHESIS '\050'
1568 #define CHAR_RIGHT_PARENTHESIS '\051'
1569 #define CHAR_ASTERISK '\052'
1570 #define CHAR_PLUS '\053'
1571 #define CHAR_COMMA '\054'
1572 #define CHAR_MINUS '\055'
1573 #define CHAR_DOT '\056'
1574 #define CHAR_SLASH '\057'
1575 #define CHAR_0 '\060'
1576 #define CHAR_1 '\061'
1577 #define CHAR_2 '\062'
1578 #define CHAR_3 '\063'
1579 #define CHAR_4 '\064'
1580 #define CHAR_5 '\065'
1581 #define CHAR_6 '\066'
1582 #define CHAR_7 '\067'
1583 #define CHAR_8 '\070'
1584 #define CHAR_9 '\071'
1585 #define CHAR_COLON '\072'
1586 #define CHAR_SEMICOLON '\073'
1587 #define CHAR_LESS_THAN_SIGN '\074'
1588 #define CHAR_EQUALS_SIGN '\075'
1589 #define CHAR_GREATER_THAN_SIGN '\076'
1590 #define CHAR_QUESTION_MARK '\077'
1591 #define CHAR_COMMERCIAL_AT '\100'
1592 #define CHAR_A '\101'
1593 #define CHAR_B '\102'
1594 #define CHAR_C '\103'
1595 #define CHAR_D '\104'
1596 #define CHAR_E '\105'
1597 #define CHAR_F '\106'
1598 #define CHAR_G '\107'
1599 #define CHAR_H '\110'
1600 #define CHAR_I '\111'
1601 #define CHAR_J '\112'
1602 #define CHAR_K '\113'
1603 #define CHAR_L '\114'
1604 #define CHAR_M '\115'
1605 #define CHAR_N '\116'
1606 #define CHAR_O '\117'
1607 #define CHAR_P '\120'
1608 #define CHAR_Q '\121'
1609 #define CHAR_R '\122'
1610 #define CHAR_S '\123'
1611 #define CHAR_T '\124'
1612 #define CHAR_U '\125'
1613 #define CHAR_V '\126'
1614 #define CHAR_W '\127'
1615 #define CHAR_X '\130'
1616 #define CHAR_Y '\131'
1617 #define CHAR_Z '\132'
1618 #define CHAR_LEFT_SQUARE_BRACKET '\133'
1619 #define CHAR_BACKSLASH '\134'
1620 #define CHAR_RIGHT_SQUARE_BRACKET '\135'
1621 #define CHAR_CIRCUMFLEX_ACCENT '\136'
1622 #define CHAR_UNDERSCORE '\137'
1623 #define CHAR_GRAVE_ACCENT '\140'
1624 #define CHAR_a '\141'
1625 #define CHAR_b '\142'
1626 #define CHAR_c '\143'
1627 #define CHAR_d '\144'
1628 #define CHAR_e '\145'
1629 #define CHAR_f '\146'
1630 #define CHAR_g '\147'
1631 #define CHAR_h '\150'
1632 #define CHAR_i '\151'
1633 #define CHAR_j '\152'
1634 #define CHAR_k '\153'
1635 #define CHAR_l '\154'
1636 #define CHAR_m '\155'
1637 #define CHAR_n '\156'
1638 #define CHAR_o '\157'
1639 #define CHAR_p '\160'
1640 #define CHAR_q '\161'
1641 #define CHAR_r '\162'
1642 #define CHAR_s '\163'
1643 #define CHAR_t '\164'
1644 #define CHAR_u '\165'
1645 #define CHAR_v '\166'
1646 #define CHAR_w '\167'
1647 #define CHAR_x '\170'
1648 #define CHAR_y '\171'
1649 #define CHAR_z '\172'
1650 #define CHAR_LEFT_CURLY_BRACKET '\173'
1651 #define CHAR_VERTICAL_LINE '\174'
1652 #define CHAR_RIGHT_CURLY_BRACKET '\175'
1653 #define CHAR_TILDE '\176'
1654
1655 #define STR_HT "\011"
1656 #define STR_VT "\013"
1657 #define STR_FF "\014"
1658 #define STR_CR "\015"
1659 #define STR_NL "\012"
1660 #define STR_BS "\010"
1661 #define STR_BEL "\007"
1662 #define STR_ESC "\033"
1663 #define STR_DEL "\177"
1664
1665 #define STR_SPACE "\040"
1666 #define STR_EXCLAMATION_MARK "\041"
1667 #define STR_QUOTATION_MARK "\042"
1668 #define STR_NUMBER_SIGN "\043"
1669 #define STR_DOLLAR_SIGN "\044"
1670 #define STR_PERCENT_SIGN "\045"
1671 #define STR_AMPERSAND "\046"
1672 #define STR_APOSTROPHE "\047"
1673 #define STR_LEFT_PARENTHESIS "\050"
1674 #define STR_RIGHT_PARENTHESIS "\051"
1675 #define STR_ASTERISK "\052"
1676 #define STR_PLUS "\053"
1677 #define STR_COMMA "\054"
1678 #define STR_MINUS "\055"
1679 #define STR_DOT "\056"
1680 #define STR_SLASH "\057"
1681 #define STR_0 "\060"
1682 #define STR_1 "\061"
1683 #define STR_2 "\062"
1684 #define STR_3 "\063"
1685 #define STR_4 "\064"
1686 #define STR_5 "\065"
1687 #define STR_6 "\066"
1688 #define STR_7 "\067"
1689 #define STR_8 "\070"
1690 #define STR_9 "\071"
1691 #define STR_COLON "\072"
1692 #define STR_SEMICOLON "\073"
1693 #define STR_LESS_THAN_SIGN "\074"
1694 #define STR_EQUALS_SIGN "\075"
1695 #define STR_GREATER_THAN_SIGN "\076"
1696 #define STR_QUESTION_MARK "\077"
1697 #define STR_COMMERCIAL_AT "\100"
1698 #define STR_A "\101"
1699 #define STR_B "\102"
1700 #define STR_C "\103"
1701 #define STR_D "\104"
1702 #define STR_E "\105"
1703 #define STR_F "\106"
1704 #define STR_G "\107"
1705 #define STR_H "\110"
1706 #define STR_I "\111"
1707 #define STR_J "\112"
1708 #define STR_K "\113"
1709 #define STR_L "\114"
1710 #define STR_M "\115"
1711 #define STR_N "\116"
1712 #define STR_O "\117"
1713 #define STR_P "\120"
1714 #define STR_Q "\121"
1715 #define STR_R "\122"
1716 #define STR_S "\123"
1717 #define STR_T "\124"
1718 #define STR_U "\125"
1719 #define STR_V "\126"
1720 #define STR_W "\127"
1721 #define STR_X "\130"
1722 #define STR_Y "\131"
1723 #define STR_Z "\132"
1724 #define STR_LEFT_SQUARE_BRACKET "\133"
1725 #define STR_BACKSLASH "\134"
1726 #define STR_RIGHT_SQUARE_BRACKET "\135"
1727 #define STR_CIRCUMFLEX_ACCENT "\136"
1728 #define STR_UNDERSCORE "\137"
1729 #define STR_GRAVE_ACCENT "\140"
1730 #define STR_a "\141"
1731 #define STR_b "\142"
1732 #define STR_c "\143"
1733 #define STR_d "\144"
1734 #define STR_e "\145"
1735 #define STR_f "\146"
1736 #define STR_g "\147"
1737 #define STR_h "\150"
1738 #define STR_i "\151"
1739 #define STR_j "\152"
1740 #define STR_k "\153"
1741 #define STR_l "\154"
1742 #define STR_m "\155"
1743 #define STR_n "\156"
1744 #define STR_o "\157"
1745 #define STR_p "\160"
1746 #define STR_q "\161"
1747 #define STR_r "\162"
1748 #define STR_s "\163"
1749 #define STR_t "\164"
1750 #define STR_u "\165"
1751 #define STR_v "\166"
1752 #define STR_w "\167"
1753 #define STR_x "\170"
1754 #define STR_y "\171"
1755 #define STR_z "\172"
1756 #define STR_LEFT_CURLY_BRACKET "\173"
1757 #define STR_VERTICAL_LINE "\174"
1758 #define STR_RIGHT_CURLY_BRACKET "\175"
1759 #define STR_TILDE "\176"
1760
1761 #define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1762 #define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1763 #define STRING_F0 STR_F "\0"
1764 #define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0"
1765 #define STRING_MARK0 STR_M STR_A STR_R STR_K "\0"
1766 #define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0"
1767 #define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0"
1768 #define STRING_THEN STR_T STR_H STR_E STR_N
1769
1770 #define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0"
1771 #define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0"
1772 #define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0"
1773 #define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0"
1774 #define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0"
1775 #define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0"
1776 #define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0"
1777 #define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0"
1778 #define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0"
1779 #define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0"
1780 #define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0"
1781 #define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0"
1782 #define STRING_word0 STR_w STR_o STR_r STR_d "\0"
1783 #define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t
1784
1785 #define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E
1786
1787 #define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS
1788 #define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS
1789 #define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1790 #define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1791 #define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1792 #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
1793 #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
1794 #define STRING_UTF8_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1795 #define STRING_UTF16_RIGHTPAR STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1796 #define STRING_UTF32_RIGHTPAR STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1797 #define STRING_UTF_RIGHTPAR STR_U STR_T STR_F STR_RIGHT_PARENTHESIS
1798 #define STRING_UCP_RIGHTPAR STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1799 #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
1800
1801 #endif /* SUPPORT_UTF */
1802
1803 /* Escape items that are just an encoding of a particular data value. */
1804
1805 #ifndef ESC_e
1806 #define ESC_e CHAR_ESC
1807 #endif
1808
1809 #ifndef ESC_f
1810 #define ESC_f CHAR_FF
1811 #endif
1812
1813 #ifndef ESC_n
1814 #define ESC_n CHAR_LF
1815 #endif
1816
1817 #ifndef ESC_r
1818 #define ESC_r CHAR_CR
1819 #endif
1820
1821 /* We can't officially use ESC_t because it is a POSIX reserved identifier
1822 (presumably because of all the others like size_t). */
1823
1824 #ifndef ESC_tee
1825 #define ESC_tee CHAR_HT
1826 #endif
1827
1828 /* Codes for different types of Unicode property */
1829
1830 #define PT_ANY 0 /* Any property - matches all chars */
1831 #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
1832 #define PT_GC 2 /* Specified general characteristic (e.g. L) */
1833 #define PT_PC 3 /* Specified particular characteristic (e.g. Lu) */
1834 #define PT_SC 4 /* Script (e.g. Han) */
1835 #define PT_ALNUM 5 /* Alphanumeric - the union of L and N */
1836 #define PT_SPACE 6 /* Perl space - Z plus 9,10,12,13 */
1837 #define PT_PXSPACE 7 /* POSIX space - Z plus 9,10,11,12,13 */
1838 #define PT_WORD 8 /* Word - L plus N plus underscore */
1839 #define PT_CLIST 9 /* Pseudo-property: match character list */
1840
1841 /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1842 contain characters with values greater than 255. */
1843
1844 #define XCL_NOT 0x01 /* Flag: this is a negative class */
1845 #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
1846
1847 #define XCL_END 0 /* Marks end of individual items */
1848 #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
1849 #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
1850 #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
1851 #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
1852
1853 /* These are escaped items that aren't just an encoding of a particular data
1854 value such as \n. They must have non-zero values, as check_escape() returns
1855 0 for a data character. Also, they must appear in the same order as in the opcode
1856 definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1857 corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1858 used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1859 non-DOTALL mode, "." behaves like \N.
1860
1861 The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1862 when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1863 They must be contiguous, and remain in order so that the replacements can be
1864 looked up from a table.
1865
1866 Negative numbers are used to encode a backreference (\1, \2, \3, etc.) in
1867 check_escape(). There are two tests in the code for an escape
1868 greater than ESC_b and less than ESC_Z to detect the types that may be
1869 repeated. These are the types that consume characters. If any new escapes are
1870 put in between that don't consume a character, that code will have to change.
1871 */
1872
1873 enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1874 ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1875 ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1876 ESC_E, ESC_Q, ESC_g, ESC_k,
1877 ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1878
1879 /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1880 OP_EOD must correspond in order to the list of escapes immediately above.
1881
1882 *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1883 that follow must also be updated to match. There are also tables called
1884 "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1885
1886 enum {
1887 OP_END, /* 0 End of pattern */
1888
1889 /* Values corresponding to backslashed metacharacters */
1890
1891 OP_SOD, /* 1 Start of data: \A */
1892 OP_SOM, /* 2 Start of match (subject + offset): \G */
1893 OP_SET_SOM, /* 3 Set start of match (\K) */
1894 OP_NOT_WORD_BOUNDARY, /* 4 \B */
1895 OP_WORD_BOUNDARY, /* 5 \b */
1896 OP_NOT_DIGIT, /* 6 \D */
1897 OP_DIGIT, /* 7 \d */
1898 OP_NOT_WHITESPACE, /* 8 \S */
1899 OP_WHITESPACE, /* 9 \s */
1900 OP_NOT_WORDCHAR, /* 10 \W */
1901 OP_WORDCHAR, /* 11 \w */
1902
1903 OP_ANY, /* 12 Match any character except newline (\N) */
1904 OP_ALLANY, /* 13 Match any character */
1905 OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1906 OP_NOTPROP, /* 15 \P (not Unicode property) */
1907 OP_PROP, /* 16 \p (Unicode property) */
1908 OP_ANYNL, /* 17 \R (any newline sequence) */
1909 OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */
1910 OP_HSPACE, /* 19 \h (horizontal whitespace) */
1911 OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */
1912 OP_VSPACE, /* 21 \v (vertical whitespace) */
1913 OP_EXTUNI, /* 22 \X (extended Unicode sequence */
1914 OP_EODN, /* 23 End of data or \n at end of data (\Z) */
1915 OP_EOD, /* 24 End of data (\z) */
1916
1917 OP_CIRC, /* 25 Start of line - not multiline */
1918 OP_CIRCM, /* 26 Start of line - multiline */
1919 OP_DOLL, /* 27 End of line - not multiline */
1920 OP_DOLLM, /* 28 End of line - multiline */
1921 OP_CHAR, /* 29 Match one character, casefully */
1922 OP_CHARI, /* 30 Match one character, caselessly */
1923 OP_NOT, /* 31 Match one character, not the given one, casefully */
1924 OP_NOTI, /* 32 Match one character, not the given one, caselessly */
1925
1926 /* The following sets of 13 opcodes must always be kept in step because
1927 the offset from the first one is used to generate the others. */
1928
1929 /**** Single characters, caseful, must precede the caseless ones ****/
1930
1931 OP_STAR, /* 33 The maximizing and minimizing versions of */
1932 OP_MINSTAR, /* 34 these six opcodes must come in pairs, with */
1933 OP_PLUS, /* 35 the minimizing one second. */
1934 OP_MINPLUS, /* 36 */
1935 OP_QUERY, /* 37 */
1936 OP_MINQUERY, /* 38 */
1937
1938 OP_UPTO, /* 39 From 0 to n matches of one character, caseful*/
1939 OP_MINUPTO, /* 40 */
1940 OP_EXACT, /* 41 Exactly n matches */
1941
1942 OP_POSSTAR, /* 42 Possessified star, caseful */
1943 OP_POSPLUS, /* 43 Possessified plus, caseful */
1944 OP_POSQUERY, /* 44 Posesssified query, caseful */
1945 OP_POSUPTO, /* 45 Possessified upto, caseful */
1946
1947 /**** Single characters, caseless, must follow the caseful ones */
1948
1949 OP_STARI, /* 46 */
1950 OP_MINSTARI, /* 47 */
1951 OP_PLUSI, /* 48 */
1952 OP_MINPLUSI, /* 49 */
1953 OP_QUERYI, /* 50 */
1954 OP_MINQUERYI, /* 51 */
1955
1956 OP_UPTOI, /* 52 From 0 to n matches of one character, caseless */
1957 OP_MINUPTOI, /* 53 */
1958 OP_EXACTI, /* 54 */
1959
1960 OP_POSSTARI, /* 55 Possessified star, caseless */
1961 OP_POSPLUSI, /* 56 Possessified plus, caseless */
1962 OP_POSQUERYI, /* 57 Posesssified query, caseless */
1963 OP_POSUPTOI, /* 58 Possessified upto, caseless */
1964
1965 /**** The negated ones must follow the non-negated ones, and match them ****/
1966 /**** Negated single character, caseful; must precede the caseless ones ****/
1967
1968 OP_NOTSTAR, /* 59 The maximizing and minimizing versions of */
1969 OP_NOTMINSTAR, /* 60 these six opcodes must come in pairs, with */
1970 OP_NOTPLUS, /* 61 the minimizing one second. They must be in */
1971 OP_NOTMINPLUS, /* 62 exactly the same order as those above. */
1972 OP_NOTQUERY, /* 63 */
1973 OP_NOTMINQUERY, /* 64 */
1974
1975 OP_NOTUPTO, /* 65 From 0 to n matches, caseful */
1976 OP_NOTMINUPTO, /* 66 */
1977 OP_NOTEXACT, /* 67 Exactly n matches */
1978
1979 OP_NOTPOSSTAR, /* 68 Possessified versions, caseful */
1980 OP_NOTPOSPLUS, /* 69 */
1981 OP_NOTPOSQUERY, /* 70 */
1982 OP_NOTPOSUPTO, /* 71 */
1983
1984 /**** Negated single character, caseless; must follow the caseful ones ****/
1985
1986 OP_NOTSTARI, /* 72 */
1987 OP_NOTMINSTARI, /* 73 */
1988 OP_NOTPLUSI, /* 74 */
1989 OP_NOTMINPLUSI, /* 75 */
1990 OP_NOTQUERYI, /* 76 */
1991 OP_NOTMINQUERYI, /* 77 */
1992
1993 OP_NOTUPTOI, /* 78 From 0 to n matches, caseless */
1994 OP_NOTMINUPTOI, /* 79 */
1995 OP_NOTEXACTI, /* 80 Exactly n matches */
1996
1997 OP_NOTPOSSTARI, /* 81 Possessified versions, caseless */
1998 OP_NOTPOSPLUSI, /* 82 */
1999 OP_NOTPOSQUERYI, /* 83 */
2000 OP_NOTPOSUPTOI, /* 84 */
2001
2002 /**** Character types ****/
2003
2004 OP_TYPESTAR, /* 85 The maximizing and minimizing versions of */
2005 OP_TYPEMINSTAR, /* 86 these six opcodes must come in pairs, with */
2006 OP_TYPEPLUS, /* 87 the minimizing one second. These codes must */
2007 OP_TYPEMINPLUS, /* 88 be in exactly the same order as those above. */
2008 OP_TYPEQUERY, /* 89 */
2009 OP_TYPEMINQUERY, /* 90 */
2010
2011 OP_TYPEUPTO, /* 91 From 0 to n matches */
2012 OP_TYPEMINUPTO, /* 92 */
2013 OP_TYPEEXACT, /* 93 Exactly n matches */
2014
2015 OP_TYPEPOSSTAR, /* 94 Possessified versions */
2016 OP_TYPEPOSPLUS, /* 95 */
2017 OP_TYPEPOSQUERY, /* 96 */
2018 OP_TYPEPOSUPTO, /* 97 */
2019
2020 /* These are used for character classes and back references; only the
2021 first six are the same as the sets above. */
2022
2023 OP_CRSTAR, /* 98 The maximizing and minimizing versions of */
2024 OP_CRMINSTAR, /* 99 all these opcodes must come in pairs, with */
2025 OP_CRPLUS, /* 100 the minimizing one second. These codes must */
2026 OP_CRMINPLUS, /* 101 be in exactly the same order as those above. */
2027 OP_CRQUERY, /* 102 */
2028 OP_CRMINQUERY, /* 103 */
2029
2030 OP_CRRANGE, /* 104 These are different to the three sets above. */
2031 OP_CRMINRANGE, /* 105 */
2032
2033 /* End of quantifier opcodes */
2034
2035 OP_CLASS, /* 106 Match a character class, chars < 256 only */
2036 OP_NCLASS, /* 107 Same, but the bitmap was created from a negative
2037 class - the difference is relevant only when a
2038 character > 255 is encountered. */
2039 OP_XCLASS, /* 108 Extended class for handling > 255 chars within the
2040 class. This does both positive and negative. */
2041 OP_REF, /* 109 Match a back reference, casefully */
2042 OP_REFI, /* 110 Match a back reference, caselessly */
2043 OP_RECURSE, /* 111 Match a numbered subpattern (possibly recursive) */
2044 OP_CALLOUT, /* 112 Call out to external function if provided */
2045
2046 OP_ALT, /* 113 Start of alternation */
2047 OP_KET, /* 114 End of group that doesn't have an unbounded repeat */
2048 OP_KETRMAX, /* 115 These two must remain together and in this */
2049 OP_KETRMIN, /* 116 order. They are for groups the repeat for ever. */
2050 OP_KETRPOS, /* 117 Possessive unlimited repeat. */
2051
2052 /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
2053 asserts must remain in order. */
2054
2055 OP_REVERSE, /* 118 Move pointer back - used in lookbehind assertions */
2056 OP_ASSERT, /* 119 Positive lookahead */
2057 OP_ASSERT_NOT, /* 120 Negative lookahead */
2058 OP_ASSERTBACK, /* 121 Positive lookbehind */
2059 OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
2060
2061 /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
2062 after the assertions, with ONCE first, as there's a test for >= ONCE for a
2063 subpattern that isn't an assertion. The POS versions must immediately follow
2064 the non-POS versions in each case. */
2065
2066 OP_ONCE, /* 123 Atomic group, contains captures */
2067 OP_ONCE_NC, /* 124 Atomic group containing no captures */
2068 OP_BRA, /* 125 Start of non-capturing bracket */
2069 OP_BRAPOS, /* 126 Ditto, with unlimited, possessive repeat */
2070 OP_CBRA, /* 127 Start of capturing bracket */
2071 OP_CBRAPOS, /* 128 Ditto, with unlimited, possessive repeat */
2072 OP_COND, /* 129 Conditional group */
2073
2074 /* These five must follow the previous five, in the same order. There's a
2075 check for >= SBRA to distinguish the two sets. */
2076
2077 OP_SBRA, /* 130 Start of non-capturing bracket, check empty */
2078 OP_SBRAPOS, /* 131 Ditto, with unlimited, possessive repeat */
2079 OP_SCBRA, /* 132 Start of capturing bracket, check empty */
2080 OP_SCBRAPOS, /* 133 Ditto, with unlimited, possessive repeat */
2081 OP_SCOND, /* 134 Conditional group, check empty */
2082
2083 /* The next two pairs must (respectively) be kept together. */
2084
2085 OP_CREF, /* 135 Used to hold a capture number as condition */
2086 OP_NCREF, /* 136 Same, but generated by a name reference*/
2087 OP_RREF, /* 137 Used to hold a recursion number as condition */
2088 OP_NRREF, /* 138 Same, but generated by a name reference*/
2089 OP_DEF, /* 139 The DEFINE condition */
2090
2091 OP_BRAZERO, /* 140 These two must remain together and in this */
2092 OP_BRAMINZERO, /* 141 order. */
2093 OP_BRAPOSZERO, /* 142 */
2094
2095 /* These are backtracking control verbs */
2096
2097 OP_MARK, /* 143 always has an argument */
2098 OP_PRUNE, /* 144 */
2099 OP_PRUNE_ARG, /* 145 same, but with argument */
2100 OP_SKIP, /* 146 */
2101 OP_SKIP_ARG, /* 147 same, but with argument */
2102 OP_THEN, /* 148 */
2103 OP_THEN_ARG, /* 149 same, but with argument */
2104 OP_COMMIT, /* 150 */
2105
2106 /* These are forced failure and success verbs */
2107
2108 OP_FAIL, /* 151 */
2109 OP_ACCEPT, /* 152 */
2110 OP_ASSERT_ACCEPT, /* 153 Used inside assertions */
2111 OP_CLOSE, /* 154 Used before OP_ACCEPT to close open captures */
2112
2113 /* This is used to skip a subpattern with a {0} quantifier */
2114
2115 OP_SKIPZERO, /* 155 */
2116
2117 /* This is not an opcode, but is used to check that tables indexed by opcode
2118 are the correct length, in order to catch updating errors - there have been
2119 some in the past. */
2120
2121 OP_TABLE_LENGTH
2122 };
2123
2124 /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2125 definitions that follow must also be updated to match. There are also tables
2126 called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
2127
2128
2129 /* This macro defines textual names for all the opcodes. These are used only
2130 for debugging, and some of them are only partial names. The macro is referenced
2131 only in pcre_printint.c, which fills out the full names in many cases (and in
2132 some cases doesn't actually use these names at all). */
2133
2134 #define OP_NAME_LIST \
2135 "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \
2136 "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \
2137 "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \
2138 "extuni", "\\Z", "\\z", \
2139 "^", "^", "$", "$", "char", "chari", "not", "noti", \
2140 "*", "*?", "+", "+?", "?", "??", \
2141 "{", "{", "{", \
2142 "*+","++", "?+", "{", \
2143 "*", "*?", "+", "+?", "?", "??", \
2144 "{", "{", "{", \
2145 "*+","++", "?+", "{", \
2146 "*", "*?", "+", "+?", "?", "??", \
2147 "{", "{", "{", \
2148 "*+","++", "?+", "{", \
2149 "*", "*?", "+", "+?", "?", "??", \
2150 "{", "{", "{", \
2151 "*+","++", "?+", "{", \
2152 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
2153 "*+","++", "?+", "{", \
2154 "*", "*?", "+", "+?", "?", "??", "{", "{", \
2155 "class", "nclass", "xclass", "Ref", "Refi", \
2156 "Recurse", "Callout", \
2157 "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos", \
2158 "Reverse", "Assert", "Assert not", "AssertB", "AssertB not", \
2159 "Once", "Once_NC", \
2160 "Bra", "BraPos", "CBra", "CBraPos", \
2161 "Cond", \
2162 "SBra", "SBraPos", "SCBra", "SCBraPos", \
2163 "SCond", \
2164 "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def", \
2165 "Brazero", "Braminzero", "Braposzero", \
2166 "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP", \
2167 "*THEN", "*THEN", "*COMMIT", "*FAIL", \
2168 "*ACCEPT", "*ASSERT_ACCEPT", \
2169 "Close", "Skip zero"
2170
2171
2172 /* This macro defines the length of fixed length operations in the compiled
2173 regex. The lengths are used when searching for specific things, and also in the
2174 debugging printing of a compiled regex. We use a macro so that it can be
2175 defined close to the definitions of the opcodes themselves.
2176
2177 As things have been extended, some of these are no longer fixed lenths, but are
2178 minima instead. For example, the length of a single-character repeat may vary
2179 in UTF-8 mode. The code that uses this table must know about such things. */
2180
2181 #define OP_LENGTHS \
2182 1, /* End */ \
2183 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \
2184 1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \
2185 1, 1, 1, /* Any, AllAny, Anybyte */ \
2186 3, 3, /* \P, \p */ \
2187 1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \
2188 1, /* \X */ \
2189 1, 1, 1, 1, 1, 1, /* \Z, \z, ^, ^M, $, $M */ \
2190 2, /* Char - the minimum length */ \
2191 2, /* Chari - the minimum length */ \
2192 2, /* not */ \
2193 2, /* noti */ \
2194 /* Positive single-char repeats ** These are */ \
2195 2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
2196 2+IMM2_SIZE, 2+IMM2_SIZE, /* upto, minupto ** mode */ \
2197 2+IMM2_SIZE, /* exact */ \
2198 2, 2, 2, 2+IMM2_SIZE, /* *+, ++, ?+, upto+ */ \
2199 2, 2, 2, 2, 2, 2, /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8 */ \
2200 2+IMM2_SIZE, 2+IMM2_SIZE, /* upto I, minupto I */ \
2201 2+IMM2_SIZE, /* exact I */ \
2202 2, 2, 2, 2+IMM2_SIZE, /* *+I, ++I, ?+I, upto+I */ \
2203 /* Negative single-char repeats - only for chars < 256 */ \
2204 2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
2205 2+IMM2_SIZE, 2+IMM2_SIZE, /* NOT upto, minupto */ \
2206 2+IMM2_SIZE, /* NOT exact */ \
2207 2, 2, 2, 2+IMM2_SIZE, /* Possessive NOT *, +, ?, upto */ \
2208 2, 2, 2, 2, 2, 2, /* NOT *I, *?I, +I, +?I, ?I, ??I */ \
2209 2+IMM2_SIZE, 2+IMM2_SIZE, /* NOT upto I, minupto I */ \
2210 2+IMM2_SIZE, /* NOT exact I */ \
2211 2, 2, 2, 2+IMM2_SIZE, /* Possessive NOT *I, +I, ?I, upto I */ \
2212 /* Positive type repeats */ \
2213 2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
2214 2+IMM2_SIZE, 2+IMM2_SIZE, /* Type upto, minupto */ \
2215 2+IMM2_SIZE, /* Type exact */ \
2216 2, 2, 2, 2+IMM2_SIZE, /* Possessive *+, ++, ?+, upto+ */ \
2217 /* Character class & ref repeats */ \
2218 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
2219 1+2*IMM2_SIZE, 1+2*IMM2_SIZE, /* CRRANGE, CRMINRANGE */ \
2220 1+(32/sizeof(pcre_uchar)), /* CLASS */ \
2221 1+(32/sizeof(pcre_uchar)), /* NCLASS */ \
2222 0, /* XCLASS - variable length */ \
2223 1+IMM2_SIZE, /* REF */ \
2224 1+IMM2_SIZE, /* REFI */ \
2225 1+LINK_SIZE, /* RECURSE */ \
2226 2+2*LINK_SIZE, /* CALLOUT */ \
2227 1+LINK_SIZE, /* Alt */ \
2228 1+LINK_SIZE, /* Ket */ \
2229 1+LINK_SIZE, /* KetRmax */ \
2230 1+LINK_SIZE, /* KetRmin */ \
2231 1+LINK_SIZE, /* KetRpos */ \
2232 1+LINK_SIZE, /* Reverse */ \
2233 1+LINK_SIZE, /* Assert */ \
2234 1+LINK_SIZE, /* Assert not */ \
2235 1+LINK_SIZE, /* Assert behind */ \
2236 1+LINK_SIZE, /* Assert behind not */ \
2237 1+LINK_SIZE, /* ONCE */ \
2238 1+LINK_SIZE, /* ONCE_NC */ \
2239 1+LINK_SIZE, /* BRA */ \
2240 1+LINK_SIZE, /* BRAPOS */ \
2241 1+LINK_SIZE+IMM2_SIZE, /* CBRA */ \
2242 1+LINK_SIZE+IMM2_SIZE, /* CBRAPOS */ \
2243 1+LINK_SIZE, /* COND */ \
2244 1+LINK_SIZE, /* SBRA */ \
2245 1+LINK_SIZE, /* SBRAPOS */ \
2246 1+LINK_SIZE+IMM2_SIZE, /* SCBRA */ \
2247 1+LINK_SIZE+IMM2_SIZE, /* SCBRAPOS */ \
2248 1+LINK_SIZE, /* SCOND */ \
2249 1+IMM2_SIZE, 1+IMM2_SIZE, /* CREF, NCREF */ \
2250 1+IMM2_SIZE, 1+IMM2_SIZE, /* RREF, NRREF */ \
2251 1, /* DEF */ \
2252 1, 1, 1, /* BRAZERO, BRAMINZERO, BRAPOSZERO */ \
2253 3, 1, 3, /* MARK, PRUNE, PRUNE_ARG */ \
2254 1, 3, /* SKIP, SKIP_ARG */ \
2255 1, 3, /* THEN, THEN_ARG */ \
2256 1, 1, 1, 1, /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT */ \
2257 1+IMM2_SIZE, 1 /* CLOSE, SKIPZERO */
2258
2259 /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
2260 condition. */
2261
2262 #define RREF_ANY 0xffff
2263
2264 /* Compile time error code numbers. They are given names so that they can more
2265 easily be tracked. When a new number is added, the table called eint in
2266 pcreposix.c must be updated. */
2267
2268 enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
2269 ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
2270 ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
2271 ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2272 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2273 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2274 ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2275 ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2276
2277 /* JIT compiling modes. The function list is indexed by them. */
2278 enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2279 JIT_NUMBER_OF_COMPILE_MODES };
2280
2281 /* The real format of the start of the pcre block; the index of names and the
2282 code vector run on as long as necessary after the end. We store an explicit
2283 offset to the name table so that if a regex is compiled on one host, saved, and
2284 then run on another where the size of pointers is different, all might still
2285 be well. For the case of compiled-on-4 and run-on-8, we include an extra
2286 pointer that is always NULL. For future-proofing, a few dummy fields were
2287 originally included - even though you can never get this planning right - but
2288 there is only one left now.
2289
2290 NOTE NOTE NOTE:
2291 Because people can now save and re-use compiled patterns, any additions to this
2292 structure should be made at the end, and something earlier (e.g. a new
2293 flag in the options or one of the dummy fields) should indicate that the new
2294 fields are present. Currently PCRE always sets the dummy fields to zero.
2295 NOTE NOTE NOTE
2296 */
2297
2298 #if defined COMPILE_PCRE8
2299 #define REAL_PCRE real_pcre
2300 #elif defined COMPILE_PCRE16
2301 #define REAL_PCRE real_pcre16
2302 #elif defined COMPILE_PCRE32
2303 #define REAL_PCRE real_pcre32
2304 #endif
2305
2306 /* It is necessary to fork the struct for 32 bit, since it needs to use
2307 * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2308 * typedef since pcretest needs access to the struct of the 8-, 16-
2309 * and 32-bit variants. */
2310
2311 typedef struct real_pcre8_or_16 {
2312 pcre_uint32 magic_number;
2313 pcre_uint32 size; /* Total that was malloced */
2314 pcre_uint32 options; /* Public options */
2315 pcre_uint16 flags; /* Private flags */
2316 pcre_uint16 max_lookbehind; /* Longest lookbehind (characters) */
2317 pcre_uint16 top_bracket; /* Highest numbered group */
2318 pcre_uint16 top_backref; /* Highest numbered back reference */
2319 pcre_uint16 first_char; /* Starting character */
2320 pcre_uint16 req_char; /* This character must be seen */
2321 pcre_uint16 name_table_offset; /* Offset to name table that follows */
2322 pcre_uint16 name_entry_size; /* Size of any name items */
2323 pcre_uint16 name_count; /* Number of name items */
2324 pcre_uint16 ref_count; /* Reference count */
2325 const pcre_uint8 *tables; /* Pointer to tables or NULL for std */
2326 const pcre_uint8 *nullpad; /* NULL padding */
2327 } real_pcre8_or_16;
2328
2329 typedef struct real_pcre8_or_16 real_pcre;
2330 typedef struct real_pcre8_or_16 real_pcre16;
2331
2332 typedef struct real_pcre32 {
2333 pcre_uint32 magic_number;
2334 pcre_uint32 size; /* Total that was malloced */
2335 pcre_uint32 options; /* Public options */
2336 pcre_uint16 flags; /* Private flags */
2337 pcre_uint16 max_lookbehind; /* Longest lookbehind (characters) */
2338 pcre_uint16 top_bracket; /* Highest numbered group */
2339 pcre_uint16 top_backref; /* Highest numbered back reference */
2340 pcre_uint32 first_char; /* Starting character */
2341 pcre_uint32 req_char; /* This character must be seen */
2342 pcre_uint16 name_table_offset; /* Offset to name table that follows */
2343 pcre_uint16 name_entry_size; /* Size of any name items */
2344 pcre_uint16 name_count; /* Number of name items */
2345 pcre_uint16 ref_count; /* Reference count */
2346 pcre_uint16 dummy1; /* for later expansion */
2347 pcre_uint16 dummy2; /* for later expansion */
2348 const pcre_uint8 *tables; /* Pointer to tables or NULL for std */
2349 void *nullpad; /* for later expansion */
2350 } real_pcre32;
2351
2352 /* Assert that the size of REAL_PCRE is divisible by 8 */
2353 typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2354
2355 /* Needed in pcretest to access some fields in the real_pcre* structures
2356 * directly. They're unified for 8/16/32 bits since the structs only differ
2357 * after these fields; if that ever changes, need to fork those defines into
2358 * 8/16 and 32 bit versions. */
2359 #define REAL_PCRE_MAGIC(re) (((REAL_PCRE*)re)->magic_number)
2360 #define REAL_PCRE_SIZE(re) (((REAL_PCRE*)re)->size)
2361 #define REAL_PCRE_OPTIONS(re) (((REAL_PCRE*)re)->options)
2362 #define REAL_PCRE_FLAGS(re) (((REAL_PCRE*)re)->flags)
2363
2364 /* The format of the block used to store data from pcre_study(). The same
2365 remark (see NOTE above) about extending this structure applies. */
2366
2367 typedef struct pcre_study_data {
2368 pcre_uint32 size; /* Total that was malloced */
2369 pcre_uint32 flags; /* Private flags */
2370 pcre_uint8 start_bits[32]; /* Starting char bits */
2371 pcre_uint32 minlength; /* Minimum subject length */
2372 } pcre_study_data;
2373
2374 /* Structure for building a chain of open capturing subpatterns during
2375 compiling, so that instructions to close them can be compiled when (*ACCEPT) is
2376 encountered. This is also used to identify subpatterns that contain recursive
2377 back references to themselves, so that they can be made atomic. */
2378
2379 typedef struct open_capitem {
2380 struct open_capitem *next; /* Chain link */
2381 pcre_uint16 number; /* Capture number */
2382 pcre_uint16 flag; /* Set TRUE if recursive back ref */
2383 } open_capitem;
2384
2385 /* Structure for passing "static" information around between the functions
2386 doing the compiling, so that they are thread-safe. */
2387
2388 typedef struct compile_data {
2389 const pcre_uint8 *lcc; /* Points to lower casing table */
2390 const pcre_uint8 *fcc; /* Points to case-flipping table */
2391 const pcre_uint8 *cbits; /* Points to character type table */
2392 const pcre_uint8 *ctypes; /* Points to table of type maps */
2393 const pcre_uchar *start_workspace;/* The start of working space */
2394 const pcre_uchar *start_code; /* The start of the compiled code */
2395 const pcre_uchar *start_pattern; /* The start of the pattern */
2396 const pcre_uchar *end_pattern; /* The end of the pattern */
2397 open_capitem *open_caps; /* Chain of open capture items */
2398 pcre_uchar *hwm; /* High watermark of workspace */
2399 pcre_uchar *name_table; /* The name/number table */
2400 int names_found; /* Number of entries so far */
2401 int name_entry_size; /* Size of each entry */
2402 int workspace_size; /* Size of workspace */
2403 unsigned int bracount; /* Count of capturing parens as we compile */
2404 int final_bracount; /* Saved value after first pass */
2405 int max_lookbehind; /* Maximum lookbehind (characters) */
2406 int top_backref; /* Maximum back reference */
2407 unsigned int backref_map; /* Bitmap of low back refs */
2408 int assert_depth; /* Depth of nested assertions */
2409 int external_options; /* External (initial) options */
2410 int external_flags; /* External flag bits to be set */
2411 int req_varyopt; /* "After variable item" flag for reqbyte */
2412 BOOL had_accept; /* (*ACCEPT) encountered */
2413 BOOL had_pruneorskip; /* (*PRUNE) or (*SKIP) encountered */
2414 BOOL check_lookbehind; /* Lookbehinds need later checking */
2415 int nltype; /* Newline type */
2416 int nllen; /* Newline string length */
2417 pcre_uchar nl[4]; /* Newline string when fixed length */
2418 } compile_data;
2419
2420 /* Structure for maintaining a chain of pointers to the currently incomplete
2421 branches, for testing for left recursion while compiling. */
2422
2423 typedef struct branch_chain {
2424 struct branch_chain *outer;
2425 pcre_uchar *current_branch;
2426 } branch_chain;
2427
2428 /* Structure for items in a linked list that represents an explicit recursive
2429 call within the pattern; used by pcre_exec(). */
2430
2431 typedef struct recursion_info {
2432 struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2433 unsigned int group_num; /* Number of group that was called */
2434 int *offset_save; /* Pointer to start of saved offsets */
2435 int saved_max; /* Number of saved offsets */
2436 PCRE_PUCHAR subject_position; /* Position at start of recursion */
2437 } recursion_info;
2438
2439 /* A similar structure for pcre_dfa_exec(). */
2440
2441 typedef struct dfa_recursion_info {
2442 struct dfa_recursion_info *prevrec;
2443 int group_num;
2444 PCRE_PUCHAR subject_position;
2445 } dfa_recursion_info;
2446
2447 /* Structure for building a chain of data for holding the values of the subject
2448 pointer at the start of each subpattern, so as to detect when an empty string
2449 has been matched by a subpattern - to break infinite loops; used by
2450 pcre_exec(). */
2451
2452 typedef struct eptrblock {
2453 struct eptrblock *epb_prev;
2454 PCRE_PUCHAR epb_saved_eptr;
2455 } eptrblock;
2456
2457
2458 /* Structure for passing "static" information around between the functions
2459 doing traditional NFA matching, so that they are thread-safe. */
2460
2461 typedef struct match_data {
2462 unsigned long int match_call_count; /* As it says */
2463 unsigned long int match_limit; /* As it says */
2464 unsigned long int match_limit_recursion; /* As it says */
2465 int *offset_vector; /* Offset vector */
2466 int offset_end; /* One past the end */
2467 int offset_max; /* The maximum usable for return data */
2468 int nltype; /* Newline type */
2469 int nllen; /* Newline string length */
2470 int name_count; /* Number of names in name table */
2471 int name_entry_size; /* Size of entry in names table */
2472 pcre_uchar *name_table; /* Table of names */
2473 pcre_uchar nl[4]; /* Newline string when fixed */
2474 const pcre_uint8 *lcc; /* Points to lower casing table */
2475 const pcre_uint8 *fcc; /* Points to case-flipping table */
2476 const pcre_uint8 *ctypes; /* Points to table of type maps */
2477 BOOL offset_overflow; /* Set if too many extractions */
2478 BOOL notbol; /* NOTBOL flag */
2479 BOOL noteol; /* NOTEOL flag */
2480 BOOL utf; /* UTF-8 / UTF-16 flag */
2481 BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */
2482 BOOL use_ucp; /* PCRE_UCP flag */
2483 BOOL endonly; /* Dollar not before final \n */
2484 BOOL notempty; /* Empty string match not wanted */
2485 BOOL notempty_atstart; /* Empty string match at start not wanted */
2486 BOOL hitend; /* Hit the end of the subject at some point */
2487 BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
2488 BOOL hasthen; /* Pattern contains (*THEN) */
2489 BOOL ignore_skip_arg; /* For re-run when SKIP name not found */
2490 const pcre_uchar *start_code; /* For use when recursing */
2491 PCRE_PUCHAR start_subject; /* Start of the subject string */
2492 PCRE_PUCHAR end_subject; /* End of the subject string */
2493 PCRE_PUCHAR start_match_ptr; /* Start of matched string */
2494 PCRE_PUCHAR end_match_ptr; /* Subject position at end match */
2495 PCRE_PUCHAR start_used_ptr; /* Earliest consulted character */
2496 int partial; /* PARTIAL options */
2497 int end_offset_top; /* Highwater mark at end of match */
2498 int capture_last; /* Most recent capture number */
2499 int start_offset; /* The start offset value */
2500 int match_function_type; /* Set for certain special calls of MATCH() */
2501 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
2502 int eptrn; /* Next free eptrblock */
2503 recursion_info *recursive; /* Linked list of recursion data */
2504 void *callout_data; /* To pass back to callouts */
2505 const pcre_uchar *mark; /* Mark pointer to pass back on success */
2506 const pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2507 const pcre_uchar *once_target; /* Where to back up to for atomic groups */
2508 #ifdef NO_RECURSE
2509 void *match_frames_base; /* For remembering malloc'd frames */
2510 #endif
2511 } match_data;
2512
2513 /* A similar structure is used for the same purpose by the DFA matching
2514 functions. */
2515
2516 typedef struct dfa_match_data {
2517 const pcre_uchar *start_code; /* Start of the compiled pattern */
2518 const pcre_uchar *start_subject ; /* Start of the subject string */
2519 const pcre_uchar *end_subject; /* End of subject string */
2520 const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2521 const pcre_uint8 *tables; /* Character tables */
2522 int start_offset; /* The start offset value */
2523 int moptions; /* Match options */
2524 int poptions; /* Pattern options */
2525 int nltype; /* Newline type */
2526 int nllen; /* Newline string length */
2527 pcre_uchar nl[4]; /* Newline string when fixed */
2528 void *callout_data; /* To pass back to callouts */
2529 dfa_recursion_info *recursive; /* Linked list of recursion data */
2530 } dfa_match_data;
2531
2532 /* Bit definitions for entries in the pcre_ctypes table. */
2533
2534 #define ctype_space 0x01
2535 #define ctype_letter 0x02
2536 #define ctype_digit 0x04
2537 #define ctype_xdigit 0x08
2538 #define ctype_word 0x10 /* alphanumeric or '_' */
2539 #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
2540
2541 /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
2542 of bits for a class map. Some classes are built by combining these tables. */
2543
2544 #define cbit_space 0 /* [:space:] or \s */
2545 #define cbit_xdigit 32 /* [:xdigit:] */
2546 #define cbit_digit 64 /* [:digit:] or \d */
2547 #define cbit_upper 96 /* [:upper:] */
2548 #define cbit_lower 128 /* [:lower:] */
2549 #define cbit_word 160 /* [:word:] or \w */
2550 #define cbit_graph 192 /* [:graph:] */
2551 #define cbit_print 224 /* [:print:] */
2552 #define cbit_punct 256 /* [:punct:] */
2553 #define cbit_cntrl 288 /* [:cntrl:] */
2554 #define cbit_length 320 /* Length of the cbits table */
2555
2556 /* Offsets of the various tables from the base tables pointer, and
2557 total length. */
2558
2559 #define lcc_offset 0
2560 #define fcc_offset 256
2561 #define cbits_offset 512
2562 #define ctypes_offset (cbits_offset + cbit_length)
2563 #define tables_length (ctypes_offset + 256)
2564
2565 /* Internal function and data prefixes. */
2566
2567 #if defined COMPILE_PCRE8
2568 #ifndef PUBL
2569 #define PUBL(name) pcre_##name
2570 #endif
2571 #ifndef PRIV
2572 #define PRIV(name) _pcre_##name
2573 #endif
2574 #elif defined COMPILE_PCRE16
2575 #ifndef PUBL
2576 #define PUBL(name) pcre16_##name
2577 #endif
2578 #ifndef PRIV
2579 #define PRIV(name) _pcre16_##name
2580 #endif
2581 #elif defined COMPILE_PCRE32
2582 #ifndef PUBL
2583 #define PUBL(name) pcre32_##name
2584 #endif
2585 #ifndef PRIV
2586 #define PRIV(name) _pcre32_##name
2587 #endif
2588 #else
2589 #error Unsupported compiling mode
2590 #endif /* COMPILE_PCRE[8|16|32] */
2591
2592 /* Layout of the UCP type table that translates property names into types and
2593 codes. Each entry used to point directly to a name, but to reduce the number of
2594 relocations in shared libraries, it now has an offset into a single string
2595 instead. */
2596
2597 typedef struct {
2598 pcre_uint16 name_offset;
2599 pcre_uint16 type;
2600 pcre_uint16 value;
2601 } ucp_type_table;
2602
2603
2604 /* Internal shared data tables. These are tables that are used by more than one
2605 of the exported public functions. They have to be "external" in the C sense,
2606 but are not part of the PCRE public API. The data for these tables is in the
2607 pcre_tables.c module. */
2608
2609 #ifdef COMPILE_PCRE8
2610 extern const int PRIV(utf8_table1)[];
2611 extern const int PRIV(utf8_table1_size);
2612 extern const int PRIV(utf8_table2)[];
2613 extern const int PRIV(utf8_table3)[];
2614 extern const pcre_uint8 PRIV(utf8_table4)[];
2615 #endif /* COMPILE_PCRE8 */
2616
2617 extern const char PRIV(utt_names)[];
2618 extern const ucp_type_table PRIV(utt)[];
2619 extern const int PRIV(utt_size);
2620
2621 extern const pcre_uint8 PRIV(OP_lengths)[];
2622 extern const pcre_uint8 PRIV(default_tables)[];
2623
2624 extern const pcre_uint32 PRIV(hspace_list)[];
2625 extern const pcre_uint32 PRIV(vspace_list)[];
2626
2627
2628 /* Internal shared functions. These are functions that are used by more than
2629 one of the exported public functions. They have to be "external" in the C
2630 sense, but are not part of the PCRE public API. */
2631
2632 /* String comparison functions. */
2633 #if defined COMPILE_PCRE8
2634
2635 #define STRCMP_UC_UC(str1, str2) \
2636 strcmp((char *)(str1), (char *)(str2))
2637 #define STRCMP_UC_C8(str1, str2) \
2638 strcmp((char *)(str1), (str2))
2639 #define STRNCMP_UC_UC(str1, str2, num) \
2640 strncmp((char *)(str1), (char *)(str2), (num))
2641 #define STRNCMP_UC_C8(str1, str2, num) \
2642 strncmp((char *)(str1), (str2), (num))
2643 #define STRLEN_UC(str) strlen((const char *)str)
2644
2645 #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2646
2647 extern int PRIV(strcmp_uc_uc)(const pcre_uchar *,
2648 const pcre_uchar *);
2649 extern int PRIV(strcmp_uc_c8)(const pcre_uchar *,
2650 const char *);
2651 extern int PRIV(strncmp_uc_uc)(const pcre_uchar *,
2652 const pcre_uchar *, unsigned int num);
2653 extern int PRIV(strncmp_uc_c8)(const pcre_uchar *,
2654 const char *, unsigned int num);
2655 extern unsigned int PRIV(strlen_uc)(const pcre_uchar *str);
2656
2657 #define STRCMP_UC_UC(str1, str2) \
2658 PRIV(strcmp_uc_uc)((str1), (str2))
2659 #define STRCMP_UC_C8(str1, str2) \
2660 PRIV(strcmp_uc_c8)((str1), (str2))
2661 #define STRNCMP_UC_UC(str1, str2, num) \
2662 PRIV(strncmp_uc_uc)((str1), (str2), (num))
2663 #define STRNCMP_UC_C8(str1, str2, num) \
2664 PRIV(strncmp_uc_c8)((str1), (str2), (num))
2665 #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2666
2667 #endif /* COMPILE_PCRE[8|16|32] */
2668
2669 #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2670
2671 #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2672 #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2673
2674 #elif defined COMPILE_PCRE32
2675
2676 extern int PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2677 const pcre_uchar *);
2678 extern int PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2679 const char *);
2680
2681 #define STRCMP_UC_UC_TEST(str1, str2) \
2682 (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2683 #define STRCMP_UC_C8_TEST(str1, str2) \
2684 (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2685
2686 #endif /* COMPILE_PCRE[8|16|32] */
2687
2688 extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2689 extern BOOL PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2690 int *, BOOL);
2691 extern unsigned int PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2692 extern int PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2693 extern BOOL PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2694 int *, BOOL);
2695 extern BOOL PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2696
2697 #ifdef SUPPORT_JIT
2698 extern void PRIV(jit_compile)(const REAL_PCRE *,
2699 PUBL(extra) *, int);
2700 extern int PRIV(jit_exec)(const PUBL(extra) *,
2701 const pcre_uchar *, int, int, int, int *, int);
2702 extern void PRIV(jit_free)(void *);
2703 extern int PRIV(jit_get_size)(void *);
2704 extern const char* PRIV(jit_get_target)(void);
2705 #endif
2706
2707 /* Unicode character database (UCD) */
2708
2709 typedef struct {
2710 pcre_uint8 script; /* ucp_Arabic, etc. */
2711 pcre_uint8 chartype; /* ucp_Cc, etc. (general categories) */
2712 pcre_uint8 gbprop; /* ucp_gbControl, etc. (grapheme break property) */
2713 pcre_uint8 caseset; /* offset to multichar other cases or zero */
2714 pcre_int32 other_case; /* offset to other case, or zero if none */
2715 } ucd_record;
2716
2717 extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2718 extern const ucd_record PRIV(ucd_records)[];
2719 extern const pcre_uint8 PRIV(ucd_stage1)[];
2720 extern const pcre_uint16 PRIV(ucd_stage2)[];
2721 extern const pcre_uint32 PRIV(ucp_gentype)[];
2722 extern const pcre_uint32 PRIV(ucp_gbtable)[];
2723 #ifdef SUPPORT_JIT
2724 extern const int PRIV(ucp_typerange)[];
2725 #endif
2726
2727 #ifdef SUPPORT_UCP
2728 /* UCD access macros */
2729
2730 #define UCD_BLOCK_SIZE 128
2731 #define GET_UCD(ch) (PRIV(ucd_records) + \
2732 PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2733 UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2734
2735 #define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype
2736 #define UCD_SCRIPT(ch) GET_UCD(ch)->script
2737 #define UCD_CATEGORY(ch) PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2738 #define UCD_GRAPHBREAK(ch) GET_UCD(ch)->gbprop
2739 #define UCD_CASESET(ch) GET_UCD(ch)->caseset
2740 #define UCD_OTHERCASE(ch) ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2741
2742 #endif /* SUPPORT_UCP */
2743
2744 #endif
2745
2746 /* End of pcre_internal.h */
2747