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) 1997-2016 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 # define PCRE_EXP_DECL extern
150 # define PCRE_EXP_DEFN
151 # define PCRE_EXP_DATA_DEFN
152 #endif
153
154 /* When compiling with the MSVC compiler, it is sometimes necessary to include
155 a "calling convention" before exported function names. (This is secondhand
156 information; I know nothing about MSVC myself). For example, something like
157
158 void __cdecl function(....)
159
160 might be needed. In order so make this easy, all the exported functions have
161 PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
162 set, we ensure here that it has no effect. */
163
164 #ifndef PCRE_CALL_CONVENTION
165 #define PCRE_CALL_CONVENTION
166 #endif
167
168 /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
169 cannot determine these outside the compilation (e.g. by running a program as
170 part of "configure") because PCRE is often cross-compiled for use on other
171 systems. Instead we make use of the maximum sizes that are available at
172 preprocessor time in standard C environments. */
173
174 typedef unsigned char pcre_uint8;
175
176 #if USHRT_MAX == 65535
177 typedef unsigned short pcre_uint16;
178 typedef short pcre_int16;
179 #define PCRE_UINT16_MAX USHRT_MAX
180 #define PCRE_INT16_MAX SHRT_MAX
181 #elif UINT_MAX == 65535
182 typedef unsigned int pcre_uint16;
183 typedef int pcre_int16;
184 #define PCRE_UINT16_MAX UINT_MAX
185 #define PCRE_INT16_MAX INT_MAX
186 #else
187 #error Cannot determine a type for 16-bit integers
188 #endif
189
190 #if UINT_MAX == 4294967295U
191 typedef unsigned int pcre_uint32;
192 typedef int pcre_int32;
193 #define PCRE_UINT32_MAX UINT_MAX
194 #define PCRE_INT32_MAX INT_MAX
195 #elif ULONG_MAX == 4294967295UL
196 typedef unsigned long int pcre_uint32;
197 typedef long int pcre_int32;
198 #define PCRE_UINT32_MAX ULONG_MAX
199 #define PCRE_INT32_MAX LONG_MAX
200 #else
201 #error Cannot determine a type for 32-bit integers
202 #endif
203
204 /* When checking for integer overflow in pcre_compile(), we need to handle
205 large integers. If a 64-bit integer type is available, we can use that.
206 Otherwise we have to cast to double, which of course requires floating point
207 arithmetic. Handle this by defining a macro for the appropriate type. If
208 stdint.h is available, include it; it may define INT64_MAX. Systems that do not
209 have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
210 by "configure". */
211
212 #if defined HAVE_STDINT_H
213 #include <stdint.h>
214 #elif defined HAVE_INTTYPES_H
215 #include <inttypes.h>
216 #endif
217
218 #if defined INT64_MAX || defined int64_t
219 #define INT64_OR_DOUBLE int64_t
220 #else
221 #define INT64_OR_DOUBLE double
222 #endif
223
224 /* All character handling must be done as unsigned characters. Otherwise there
225 are problems with top-bit-set characters and functions such as isspace().
226 However, we leave the interface to the outside world as char * or short *,
227 because that should make things easier for callers. This character type is
228 called pcre_uchar.
229
230 The IN_UCHARS macro multiply its argument with the byte size of the current
231 pcre_uchar type. Useful for memcpy and such operations, whose require the
232 byte size of their input/output buffers.
233
234 The MAX_255 macro checks whether its pcre_uchar input is less than 256.
235
236 The TABLE_GET macro is designed for accessing elements of tables whose contain
237 exactly 256 items. When the character is able to contain more than 256
238 items, some check is needed before accessing these tables.
239 */
240
241 #if defined COMPILE_PCRE8
242
243 typedef unsigned char pcre_uchar;
244 #define IN_UCHARS(x) (x)
245 #define MAX_255(c) 1
246 #define TABLE_GET(c, table, default) ((table)[c])
247
248 #elif defined COMPILE_PCRE16
249
250 #if USHRT_MAX != 65535
251 /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
252 pcre.h(.in) and disable (comment out) this message. */
253 #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
254 #endif
255
256 typedef pcre_uint16 pcre_uchar;
257 #define UCHAR_SHIFT (1)
258 #define IN_UCHARS(x) ((x) * 2)
259 #define MAX_255(c) ((c) <= 255u)
260 #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
261
262 #elif defined COMPILE_PCRE32
263
264 typedef pcre_uint32 pcre_uchar;
265 #define UCHAR_SHIFT (2)
266 #define IN_UCHARS(x) ((x) * 4)
267 #define MAX_255(c) ((c) <= 255u)
268 #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
269
270 #else
271 #error Unsupported compiling mode
272 #endif /* COMPILE_PCRE[8|16|32] */
273
274 /* This is an unsigned int value that no character can ever have. UTF-8
275 characters only go up to 0x7fffffff (though Unicode doesn't go beyond
276 0x0010ffff). */
277
278 #define NOTACHAR 0xffffffff
279
280 /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
281 "any" and "anycrlf" at present). The following macros are used to package up
282 testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
283 modules to indicate in which datablock the parameters exist, and what the
284 start/end of string field names are. */
285
286 #define NLTYPE_FIXED 0 /* Newline is a fixed length string */
287 #define NLTYPE_ANY 1 /* Newline is any Unicode line ending */
288 #define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */
289
290 /* This macro checks for a newline at the given position */
291
292 #define IS_NEWLINE(p) \
293 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
294 ((p) < NLBLOCK->PSEND && \
295 PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
296 &(NLBLOCK->nllen), utf)) \
297 : \
298 ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
299 UCHAR21TEST(p) == NLBLOCK->nl[0] && \
300 (NLBLOCK->nllen == 1 || UCHAR21TEST(p+1) == NLBLOCK->nl[1]) \
301 ) \
302 )
303
304 /* This macro checks for a newline immediately preceding the given position */
305
306 #define WAS_NEWLINE(p) \
307 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
308 ((p) > NLBLOCK->PSSTART && \
309 PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
310 &(NLBLOCK->nllen), utf)) \
311 : \
312 ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
313 UCHAR21TEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] && \
314 (NLBLOCK->nllen == 1 || UCHAR21TEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
315 ) \
316 )
317
318 /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
319 with a custom type. This makes it possible, for example, to allow pcre_exec()
320 to process subject strings that are discontinuous by using a smart pointer
321 class. It must always be possible to inspect all of the subject string in
322 pcre_exec() because of the way it backtracks. Two macros are required in the
323 normal case, for sign-unspecified and unsigned char pointers. The former is
324 used for the external interface and appears in pcre.h, which is why its name
325 must begin with PCRE_. */
326
327 #ifdef CUSTOM_SUBJECT_PTR
328 #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
329 #else
330 #define PCRE_PUCHAR const pcre_uchar *
331 #endif
332
333 /* Include the public PCRE header and the definitions of UCP character property
334 values. */
335
336 #include "pcre.h"
337 #include "ucp.h"
338
339 #ifdef COMPILE_PCRE32
340 /* Assert that the public PCRE_UCHAR32 is a 32-bit type */
341 typedef int __assert_pcre_uchar32_size[sizeof(PCRE_UCHAR32) == 4 ? 1 : -1];
342 #endif
343
344 /* When compiling for use with the Virtual Pascal compiler, these functions
345 need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
346 option on the command line. */
347
348 #ifdef VPCOMPAT
349 #define strlen(s) _strlen(s)
350 #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
351 #define memcmp(s,c,n) _memcmp(s,c,n)
352 #define memcpy(d,s,n) _memcpy(d,s,n)
353 #define memmove(d,s,n) _memmove(d,s,n)
354 #define memset(s,c,n) _memset(s,c,n)
355 #else /* VPCOMPAT */
356
357 /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
358 define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
359 is set. Otherwise, include an emulating function for those systems that have
360 neither (there some non-Unix environments where this is the case). */
361
362 #ifndef HAVE_MEMMOVE
363 #undef memmove /* some systems may have a macro */
364 #ifdef HAVE_BCOPY
365 #define memmove(a, b, c) bcopy(b, a, c)
366 #else /* HAVE_BCOPY */
367 static void *
pcre_memmove(void * d,const void * s,size_t n)368 pcre_memmove(void *d, const void *s, size_t n)
369 {
370 size_t i;
371 unsigned char *dest = (unsigned char *)d;
372 const unsigned char *src = (const unsigned char *)s;
373 if (dest > src)
374 {
375 dest += n;
376 src += n;
377 for (i = 0; i < n; ++i) *(--dest) = *(--src);
378 return (void *)dest;
379 }
380 else
381 {
382 for (i = 0; i < n; ++i) *dest++ = *src++;
383 return (void *)(dest - n);
384 }
385 }
386 #define memmove(a, b, c) pcre_memmove(a, b, c)
387 #endif /* not HAVE_BCOPY */
388 #endif /* not HAVE_MEMMOVE */
389 #endif /* not VPCOMPAT */
390
391
392 /* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
393 in big-endian order) by default. These are used, for example, to link from the
394 start of a subpattern to its alternatives and its end. The use of 2 bytes per
395 offset limits the size of the compiled regex to around 64K, which is big enough
396 for almost everybody. However, I received a request for an even bigger limit.
397 For this reason, and also to make the code easier to maintain, the storing and
398 loading of offsets from the byte string is now handled by the macros that are
399 defined here.
400
401 The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
402 the config.h file, but can be overridden by using -D on the command line. This
403 is automated on Unix systems via the "configure" command. */
404
405 #if defined COMPILE_PCRE8
406
407 #if LINK_SIZE == 2
408
409 #define PUT(a,n,d) \
410 (a[n] = (d) >> 8), \
411 (a[(n)+1] = (d) & 255)
412
413 #define GET(a,n) \
414 (((a)[n] << 8) | (a)[(n)+1])
415
416 #define MAX_PATTERN_SIZE (1 << 16)
417
418
419 #elif LINK_SIZE == 3
420
421 #define PUT(a,n,d) \
422 (a[n] = (d) >> 16), \
423 (a[(n)+1] = (d) >> 8), \
424 (a[(n)+2] = (d) & 255)
425
426 #define GET(a,n) \
427 (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
428
429 #define MAX_PATTERN_SIZE (1 << 24)
430
431
432 #elif LINK_SIZE == 4
433
434 #define PUT(a,n,d) \
435 (a[n] = (d) >> 24), \
436 (a[(n)+1] = (d) >> 16), \
437 (a[(n)+2] = (d) >> 8), \
438 (a[(n)+3] = (d) & 255)
439
440 #define GET(a,n) \
441 (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
442
443 /* Keep it positive */
444 #define MAX_PATTERN_SIZE (1 << 30)
445
446 #else
447 #error LINK_SIZE must be either 2, 3, or 4
448 #endif
449
450 #elif defined COMPILE_PCRE16
451
452 #if LINK_SIZE == 2
453
454 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
455 #undef LINK_SIZE
456 #define LINK_SIZE 1
457
458 #define PUT(a,n,d) \
459 (a[n] = (d))
460
461 #define GET(a,n) \
462 (a[n])
463
464 #define MAX_PATTERN_SIZE (1 << 16)
465
466 #elif LINK_SIZE == 3 || LINK_SIZE == 4
467
468 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
469 #undef LINK_SIZE
470 #define LINK_SIZE 2
471
472 #define PUT(a,n,d) \
473 (a[n] = (d) >> 16), \
474 (a[(n)+1] = (d) & 65535)
475
476 #define GET(a,n) \
477 (((a)[n] << 16) | (a)[(n)+1])
478
479 /* Keep it positive */
480 #define MAX_PATTERN_SIZE (1 << 30)
481
482 #else
483 #error LINK_SIZE must be either 2, 3, or 4
484 #endif
485
486 #elif defined COMPILE_PCRE32
487
488 /* Only supported LINK_SIZE is 4 */
489 /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
490 #undef LINK_SIZE
491 #define LINK_SIZE 1
492
493 #define PUT(a,n,d) \
494 (a[n] = (d))
495
496 #define GET(a,n) \
497 (a[n])
498
499 /* Keep it positive */
500 #define MAX_PATTERN_SIZE (1 << 30)
501
502 #else
503 #error Unsupported compiling mode
504 #endif /* COMPILE_PCRE[8|16|32] */
505
506 /* Convenience macro defined in terms of the others */
507
508 #define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
509
510
511 /* PCRE uses some other 2-byte quantities that do not change when the size of
512 offsets changes. There are used for repeat counts and for other things such as
513 capturing parenthesis numbers in back references. */
514
515 #if defined COMPILE_PCRE8
516
517 #define IMM2_SIZE 2
518
519 #define PUT2(a,n,d) \
520 a[n] = (d) >> 8; \
521 a[(n)+1] = (d) & 255
522
523 /* For reasons that I do not understand, the expression in this GET2 macro is
524 treated by gcc as a signed expression, even when a is declared as unsigned. It
525 seems that any kind of arithmetic results in a signed value. */
526
527 #define GET2(a,n) \
528 (unsigned int)(((a)[n] << 8) | (a)[(n)+1])
529
530 #elif defined COMPILE_PCRE16
531
532 #define IMM2_SIZE 1
533
534 #define PUT2(a,n,d) \
535 a[n] = d
536
537 #define GET2(a,n) \
538 a[n]
539
540 #elif defined COMPILE_PCRE32
541
542 #define IMM2_SIZE 1
543
544 #define PUT2(a,n,d) \
545 a[n] = d
546
547 #define GET2(a,n) \
548 a[n]
549
550 #else
551 #error Unsupported compiling mode
552 #endif /* COMPILE_PCRE[8|16|32] */
553
554 #define PUT2INC(a,n,d) PUT2(a,n,d), a += IMM2_SIZE
555
556 /* The maximum length of a MARK name is currently one data unit; it may be
557 changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
558
559 #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
560 #define MAX_MARK ((1u << 16) - 1)
561 #else
562 #define MAX_MARK ((1u << 8) - 1)
563 #endif
564
565 /* There is a proposed future special "UTF-21" mode, in which only the lowest
566 21 bits of a 32-bit character are interpreted as UTF, with the remaining 11
567 high-order bits available to the application for other uses. In preparation for
568 the future implementation of this mode, there are macros that load a data item
569 and, if in this special mode, mask it to 21 bits. These macros all have names
570 starting with UCHAR21. In all other modes, including the normal 32-bit
571 library, the macros all have the same simple definitions. When the new mode is
572 implemented, it is expected that these definitions will be varied appropriately
573 using #ifdef when compiling the library that supports the special mode. */
574
575 #define UCHAR21(eptr) (*(eptr))
576 #define UCHAR21TEST(eptr) (*(eptr))
577 #define UCHAR21INC(eptr) (*(eptr)++)
578 #define UCHAR21INCTEST(eptr) (*(eptr)++)
579
580 /* When UTF encoding is being used, a character is no longer just a single
581 byte in 8-bit mode or a single short in 16-bit mode. The macros for character
582 handling generate simple sequences when used in the basic mode, and more
583 complicated ones for UTF characters. GETCHARLENTEST and other macros are not
584 used when UTF is not supported. To make sure they can never even appear when
585 UTF support is omitted, we don't even define them. */
586
587 #ifndef SUPPORT_UTF
588
589 /* #define MAX_VALUE_FOR_SINGLE_CHAR */
590 /* #define HAS_EXTRALEN(c) */
591 /* #define GET_EXTRALEN(c) */
592 /* #define NOT_FIRSTCHAR(c) */
593 #define GETCHAR(c, eptr) c = *eptr;
594 #define GETCHARTEST(c, eptr) c = *eptr;
595 #define GETCHARINC(c, eptr) c = *eptr++;
596 #define GETCHARINCTEST(c, eptr) c = *eptr++;
597 #define GETCHARLEN(c, eptr, len) c = *eptr;
598 /* #define GETCHARLENTEST(c, eptr, len) */
599 /* #define BACKCHAR(eptr) */
600 /* #define FORWARDCHAR(eptr) */
601 /* #define ACROSSCHAR(condition, eptr, action) */
602
603 #else /* SUPPORT_UTF */
604
605 /* Tests whether the code point needs extra characters to decode. */
606
607 #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
608
609 /* Base macro to pick up the remaining bytes of a UTF-8 character, not
610 advancing the pointer. */
611
612 #define GETUTF8(c, eptr) \
613 { \
614 if ((c & 0x20) == 0) \
615 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
616 else if ((c & 0x10) == 0) \
617 c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
618 else if ((c & 0x08) == 0) \
619 c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
620 ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
621 else if ((c & 0x04) == 0) \
622 c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
623 ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
624 (eptr[4] & 0x3f); \
625 else \
626 c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
627 ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
628 ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
629 }
630
631 /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
632 the pointer. */
633
634 #define GETUTF8INC(c, eptr) \
635 { \
636 if ((c & 0x20) == 0) \
637 c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
638 else if ((c & 0x10) == 0) \
639 { \
640 c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
641 eptr += 2; \
642 } \
643 else if ((c & 0x08) == 0) \
644 { \
645 c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
646 ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
647 eptr += 3; \
648 } \
649 else if ((c & 0x04) == 0) \
650 { \
651 c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
652 ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
653 (eptr[3] & 0x3f); \
654 eptr += 4; \
655 } \
656 else \
657 { \
658 c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
659 ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
660 ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
661 eptr += 5; \
662 } \
663 }
664
665 #if defined COMPILE_PCRE8
666
667 /* These macros were originally written in the form of loops that used data
668 from the tables whose names start with PRIV(utf8_table). They were rewritten by
669 a user so as not to use loops, because in some environments this gives a
670 significant performance advantage, and it seems never to do any harm. */
671
672 /* Tells the biggest code point which can be encoded as a single character. */
673
674 #define MAX_VALUE_FOR_SINGLE_CHAR 127
675
676 /* Tests whether the code point needs extra characters to decode. */
677
678 #define HAS_EXTRALEN(c) ((c) >= 0xc0)
679
680 /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
681 Otherwise it has an undefined behaviour. */
682
683 #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
684
685 /* Returns TRUE, if the given character is not the first character
686 of a UTF sequence. */
687
688 #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
689
690 /* Get the next UTF-8 character, not advancing the pointer. This is called when
691 we know we are in UTF-8 mode. */
692
693 #define GETCHAR(c, eptr) \
694 c = *eptr; \
695 if (c >= 0xc0) GETUTF8(c, eptr);
696
697 /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
698 pointer. */
699
700 #define GETCHARTEST(c, eptr) \
701 c = *eptr; \
702 if (utf && c >= 0xc0) GETUTF8(c, eptr);
703
704 /* Get the next UTF-8 character, advancing the pointer. This is called when we
705 know we are in UTF-8 mode. */
706
707 #define GETCHARINC(c, eptr) \
708 c = *eptr++; \
709 if (c >= 0xc0) GETUTF8INC(c, eptr);
710
711 /* Get the next character, testing for UTF-8 mode, and advancing the pointer.
712 This is called when we don't know if we are in UTF-8 mode. */
713
714 #define GETCHARINCTEST(c, eptr) \
715 c = *eptr++; \
716 if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
717
718 /* Base macro to pick up the remaining bytes of a UTF-8 character, not
719 advancing the pointer, incrementing the length. */
720
721 #define GETUTF8LEN(c, eptr, len) \
722 { \
723 if ((c & 0x20) == 0) \
724 { \
725 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
726 len++; \
727 } \
728 else if ((c & 0x10) == 0) \
729 { \
730 c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
731 len += 2; \
732 } \
733 else if ((c & 0x08) == 0) \
734 {\
735 c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
736 ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
737 len += 3; \
738 } \
739 else if ((c & 0x04) == 0) \
740 { \
741 c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
742 ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
743 (eptr[4] & 0x3f); \
744 len += 4; \
745 } \
746 else \
747 {\
748 c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
749 ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
750 ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
751 len += 5; \
752 } \
753 }
754
755 /* Get the next UTF-8 character, not advancing the pointer, incrementing length
756 if there are extra bytes. This is called when we know we are in UTF-8 mode. */
757
758 #define GETCHARLEN(c, eptr, len) \
759 c = *eptr; \
760 if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
761
762 /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
763 pointer, incrementing length if there are extra bytes. This is called when we
764 do not know if we are in UTF-8 mode. */
765
766 #define GETCHARLENTEST(c, eptr, len) \
767 c = *eptr; \
768 if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
769
770 /* If the pointer is not at the start of a character, move it back until
771 it is. This is called only in UTF-8 mode - we don't put a test within the macro
772 because almost all calls are already within a block of UTF-8 only code. */
773
774 #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
775
776 /* Same as above, just in the other direction. */
777 #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
778
779 /* Same as above, but it allows a fully customizable form. */
780 #define ACROSSCHAR(condition, eptr, action) \
781 while((condition) && ((eptr) & 0xc0) == 0x80) action
782
783 #elif defined COMPILE_PCRE16
784
785 /* Tells the biggest code point which can be encoded as a single character. */
786
787 #define MAX_VALUE_FOR_SINGLE_CHAR 65535
788
789 /* Tests whether the code point needs extra characters to decode. */
790
791 #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
792
793 /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
794 Otherwise it has an undefined behaviour. */
795
796 #define GET_EXTRALEN(c) 1
797
798 /* Returns TRUE, if the given character is not the first character
799 of a UTF sequence. */
800
801 #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
802
803 /* Base macro to pick up the low surrogate of a UTF-16 character, not
804 advancing the pointer. */
805
806 #define GETUTF16(c, eptr) \
807 { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
808
809 /* Get the next UTF-16 character, not advancing the pointer. This is called when
810 we know we are in UTF-16 mode. */
811
812 #define GETCHAR(c, eptr) \
813 c = *eptr; \
814 if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
815
816 /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
817 pointer. */
818
819 #define GETCHARTEST(c, eptr) \
820 c = *eptr; \
821 if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
822
823 /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
824 the pointer. */
825
826 #define GETUTF16INC(c, eptr) \
827 { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
828
829 /* Get the next UTF-16 character, advancing the pointer. This is called when we
830 know we are in UTF-16 mode. */
831
832 #define GETCHARINC(c, eptr) \
833 c = *eptr++; \
834 if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
835
836 /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
837 This is called when we don't know if we are in UTF-16 mode. */
838
839 #define GETCHARINCTEST(c, eptr) \
840 c = *eptr++; \
841 if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
842
843 /* Base macro to pick up the low surrogate of a UTF-16 character, not
844 advancing the pointer, incrementing the length. */
845
846 #define GETUTF16LEN(c, eptr, len) \
847 { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
848
849 /* Get the next UTF-16 character, not advancing the pointer, incrementing
850 length if there is a low surrogate. This is called when we know we are in
851 UTF-16 mode. */
852
853 #define GETCHARLEN(c, eptr, len) \
854 c = *eptr; \
855 if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
856
857 /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
858 pointer, incrementing length if there is a low surrogate. This is called when
859 we do not know if we are in UTF-16 mode. */
860
861 #define GETCHARLENTEST(c, eptr, len) \
862 c = *eptr; \
863 if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
864
865 /* If the pointer is not at the start of a character, move it back until
866 it is. This is called only in UTF-16 mode - we don't put a test within the
867 macro because almost all calls are already within a block of UTF-16 only
868 code. */
869
870 #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
871
872 /* Same as above, just in the other direction. */
873 #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
874
875 /* Same as above, but it allows a fully customizable form. */
876 #define ACROSSCHAR(condition, eptr, action) \
877 if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
878
879 #elif defined COMPILE_PCRE32
880
881 /* These are trivial for the 32-bit library, since all UTF-32 characters fit
882 into one pcre_uchar unit. */
883 #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
884 #define HAS_EXTRALEN(c) (0)
885 #define GET_EXTRALEN(c) (0)
886 #define NOT_FIRSTCHAR(c) (0)
887
888 /* Get the next UTF-32 character, not advancing the pointer. This is called when
889 we know we are in UTF-32 mode. */
890
891 #define GETCHAR(c, eptr) \
892 c = *(eptr);
893
894 /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
895 pointer. */
896
897 #define GETCHARTEST(c, eptr) \
898 c = *(eptr);
899
900 /* Get the next UTF-32 character, advancing the pointer. This is called when we
901 know we are in UTF-32 mode. */
902
903 #define GETCHARINC(c, eptr) \
904 c = *((eptr)++);
905
906 /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
907 This is called when we don't know if we are in UTF-32 mode. */
908
909 #define GETCHARINCTEST(c, eptr) \
910 c = *((eptr)++);
911
912 /* Get the next UTF-32 character, not advancing the pointer, not incrementing
913 length (since all UTF-32 is of length 1). This is called when we know we are in
914 UTF-32 mode. */
915
916 #define GETCHARLEN(c, eptr, len) \
917 GETCHAR(c, eptr)
918
919 /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
920 pointer, not incrementing the length (since all UTF-32 is of length 1).
921 This is called when we do not know if we are in UTF-32 mode. */
922
923 #define GETCHARLENTEST(c, eptr, len) \
924 GETCHARTEST(c, eptr)
925
926 /* If the pointer is not at the start of a character, move it back until
927 it is. This is called only in UTF-32 mode - we don't put a test within the
928 macro because almost all calls are already within a block of UTF-32 only
929 code.
930 These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
931
932 #define BACKCHAR(eptr) do { } while (0)
933
934 /* Same as above, just in the other direction. */
935 #define FORWARDCHAR(eptr) do { } while (0)
936
937 /* Same as above, but it allows a fully customizable form. */
938 #define ACROSSCHAR(condition, eptr, action) do { } while (0)
939
940 #else
941 #error Unsupported compiling mode
942 #endif /* COMPILE_PCRE[8|16|32] */
943
944 #endif /* SUPPORT_UTF */
945
946 /* Tests for Unicode horizontal and vertical whitespace characters must check a
947 number of different values. Using a switch statement for this generates the
948 fastest code (no loop, no memory access), and there are several places in the
949 interpreter code where this happens. In order to ensure that all the case lists
950 remain in step, we use macros so that there is only one place where the lists
951 are defined.
952
953 These values are also required as lists in pcre_compile.c when processing \h,
954 \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
955 macros that define the values are here so that all the definitions are
956 together. The lists must be in ascending character order, terminated by
957 NOTACHAR (which is 0xffffffff).
958
959 Any changes should ensure that the various macros are kept in step with each
960 other. NOTE: The values also appear in pcre_jit_compile.c. */
961
962 /* ------ ASCII/Unicode environments ------ */
963
964 #ifndef EBCDIC
965
966 #define HSPACE_LIST \
967 CHAR_HT, CHAR_SPACE, CHAR_NBSP, \
968 0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
969 0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
970 NOTACHAR
971
972 #define HSPACE_MULTIBYTE_CASES \
973 case 0x1680: /* OGHAM SPACE MARK */ \
974 case 0x180e: /* MONGOLIAN VOWEL SEPARATOR */ \
975 case 0x2000: /* EN QUAD */ \
976 case 0x2001: /* EM QUAD */ \
977 case 0x2002: /* EN SPACE */ \
978 case 0x2003: /* EM SPACE */ \
979 case 0x2004: /* THREE-PER-EM SPACE */ \
980 case 0x2005: /* FOUR-PER-EM SPACE */ \
981 case 0x2006: /* SIX-PER-EM SPACE */ \
982 case 0x2007: /* FIGURE SPACE */ \
983 case 0x2008: /* PUNCTUATION SPACE */ \
984 case 0x2009: /* THIN SPACE */ \
985 case 0x200A: /* HAIR SPACE */ \
986 case 0x202f: /* NARROW NO-BREAK SPACE */ \
987 case 0x205f: /* MEDIUM MATHEMATICAL SPACE */ \
988 case 0x3000 /* IDEOGRAPHIC SPACE */
989
990 #define HSPACE_BYTE_CASES \
991 case CHAR_HT: \
992 case CHAR_SPACE: \
993 case CHAR_NBSP
994
995 #define HSPACE_CASES \
996 HSPACE_BYTE_CASES: \
997 HSPACE_MULTIBYTE_CASES
998
999 #define VSPACE_LIST \
1000 CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1001
1002 #define VSPACE_MULTIBYTE_CASES \
1003 case 0x2028: /* LINE SEPARATOR */ \
1004 case 0x2029 /* PARAGRAPH SEPARATOR */
1005
1006 #define VSPACE_BYTE_CASES \
1007 case CHAR_LF: \
1008 case CHAR_VT: \
1009 case CHAR_FF: \
1010 case CHAR_CR: \
1011 case CHAR_NEL
1012
1013 #define VSPACE_CASES \
1014 VSPACE_BYTE_CASES: \
1015 VSPACE_MULTIBYTE_CASES
1016
1017 /* ------ EBCDIC environments ------ */
1018
1019 #else
1020 #define HSPACE_LIST CHAR_HT, CHAR_SPACE, CHAR_NBSP, NOTACHAR
1021
1022 #define HSPACE_BYTE_CASES \
1023 case CHAR_HT: \
1024 case CHAR_SPACE: \
1025 case CHAR_NBSP
1026
1027 #define HSPACE_CASES HSPACE_BYTE_CASES
1028
1029 #ifdef EBCDIC_NL25
1030 #define VSPACE_LIST \
1031 CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1032 #else
1033 #define VSPACE_LIST \
1034 CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1035 #endif
1036
1037 #define VSPACE_BYTE_CASES \
1038 case CHAR_LF: \
1039 case CHAR_VT: \
1040 case CHAR_FF: \
1041 case CHAR_CR: \
1042 case CHAR_NEL
1043
1044 #define VSPACE_CASES VSPACE_BYTE_CASES
1045 #endif /* EBCDIC */
1046
1047 /* ------ End of whitespace macros ------ */
1048
1049
1050
1051 /* Private flags containing information about the compiled regex. They used to
1052 live at the top end of the options word, but that got almost full, so they were
1053 moved to a 16-bit flags word - which got almost full, so now they are in a
1054 32-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as the
1055 restrictions on partial matching have been lifted. It remains for backwards
1056 compatibility. */
1057
1058 #define PCRE_MODE8 0x00000001 /* compiled in 8 bit mode */
1059 #define PCRE_MODE16 0x00000002 /* compiled in 16 bit mode */
1060 #define PCRE_MODE32 0x00000004 /* compiled in 32 bit mode */
1061 #define PCRE_FIRSTSET 0x00000010 /* first_char is set */
1062 #define PCRE_FCH_CASELESS 0x00000020 /* caseless first char */
1063 #define PCRE_REQCHSET 0x00000040 /* req_byte is set */
1064 #define PCRE_RCH_CASELESS 0x00000080 /* caseless requested char */
1065 #define PCRE_STARTLINE 0x00000100 /* start after \n for multiline */
1066 #define PCRE_NOPARTIAL 0x00000200 /* can't use partial with this regex */
1067 #define PCRE_JCHANGED 0x00000400 /* j option used in regex */
1068 #define PCRE_HASCRORLF 0x00000800 /* explicit \r or \n in pattern */
1069 #define PCRE_HASTHEN 0x00001000 /* pattern contains (*THEN) */
1070 #define PCRE_MLSET 0x00002000 /* match limit set by regex */
1071 #define PCRE_RLSET 0x00004000 /* recursion limit set by regex */
1072 #define PCRE_MATCH_EMPTY 0x00008000 /* pattern can match empty string */
1073
1074 #if defined COMPILE_PCRE8
1075 #define PCRE_MODE PCRE_MODE8
1076 #elif defined COMPILE_PCRE16
1077 #define PCRE_MODE PCRE_MODE16
1078 #elif defined COMPILE_PCRE32
1079 #define PCRE_MODE PCRE_MODE32
1080 #endif
1081 #define PCRE_MODE_MASK (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1082
1083 /* Flags for the "extra" block produced by pcre_study(). */
1084
1085 #define PCRE_STUDY_MAPPED 0x0001 /* a map of starting chars exists */
1086 #define PCRE_STUDY_MINLEN 0x0002 /* a minimum length field exists */
1087
1088 /* Masks for identifying the public options that are permitted at compile
1089 time, run time, or study time, respectively. */
1090
1091 #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
1092 PCRE_NEWLINE_ANYCRLF)
1093
1094 #define PUBLIC_COMPILE_OPTIONS \
1095 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
1096 PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1097 PCRE_NO_AUTO_CAPTURE|PCRE_NO_AUTO_POSSESS| \
1098 PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
1099 PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1100 PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE|PCRE_NEVER_UTF)
1101
1102 #define PUBLIC_EXEC_OPTIONS \
1103 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1104 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
1105 PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
1106
1107 #define PUBLIC_DFA_EXEC_OPTIONS \
1108 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1109 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
1110 PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1111 PCRE_NO_START_OPTIMIZE)
1112
1113 #define PUBLIC_STUDY_OPTIONS \
1114 (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1115 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1116
1117 #define PUBLIC_JIT_EXEC_OPTIONS \
1118 (PCRE_NO_UTF8_CHECK|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|\
1119 PCRE_NOTEMPTY_ATSTART|PCRE_PARTIAL_SOFT|PCRE_PARTIAL_HARD)
1120
1121 /* Magic number to provide a small check against being handed junk. */
1122
1123 #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
1124
1125 /* This variable is used to detect a loaded regular expression
1126 in different endianness. */
1127
1128 #define REVERSED_MAGIC_NUMBER 0x45524350UL /* 'ERCP' */
1129
1130 /* The maximum remaining length of subject we are prepared to search for a
1131 req_byte match. */
1132
1133 #define REQ_BYTE_MAX 1000
1134
1135 /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1136 environments where these macros are defined elsewhere. Unfortunately, there
1137 is no way to do the same for the typedef. */
1138
1139 typedef int BOOL;
1140
1141 #ifndef FALSE
1142 #define FALSE 0
1143 #define TRUE 1
1144 #endif
1145
1146 /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
1147 character constants like '*' because the compiler would emit their EBCDIC code,
1148 which is different from their ASCII/UTF-8 code. Instead we define macros for
1149 the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
1150 is enabled. When UTF-8 support is not enabled, the definitions use character
1151 literals. Both character and string versions of each character are needed, and
1152 there are some longer strings as well.
1153
1154 This means that, on EBCDIC platforms, the PCRE library can handle either
1155 EBCDIC, or UTF-8, but not both. To support both in the same compiled library
1156 would need different lookups depending on whether PCRE_UTF8 was set or not.
1157 This would make it impossible to use characters in switch/case statements,
1158 which would reduce performance. For a theoretical use (which nobody has asked
1159 for) in a minority area (EBCDIC platforms), this is not sensible. Any
1160 application that did need both could compile two versions of the library, using
1161 macros to give the functions distinct names. */
1162
1163 #ifndef SUPPORT_UTF
1164
1165 /* UTF-8 support is not enabled; use the platform-dependent character literals
1166 so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1167 mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1168 characters, a common practice has been to use its NL (0x15) character as the
1169 line terminator in C-like processing environments. However, sometimes the LF
1170 (0x25) character is used instead, according to this Unicode document:
1171
1172 http://unicode.org/standard/reports/tr13/tr13-5.html
1173
1174 PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1175 instead. Whichever is *not* chosen is defined as NEL.
1176
1177 In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1178 same code point. */
1179
1180 #ifdef EBCDIC
1181
1182 #ifndef EBCDIC_NL25
1183 #define CHAR_NL '\x15'
1184 #define CHAR_NEL '\x25'
1185 #define STR_NL "\x15"
1186 #define STR_NEL "\x25"
1187 #else
1188 #define CHAR_NL '\x25'
1189 #define CHAR_NEL '\x15'
1190 #define STR_NL "\x25"
1191 #define STR_NEL "\x15"
1192 #endif
1193
1194 #define CHAR_LF CHAR_NL
1195 #define STR_LF STR_NL
1196
1197 #define CHAR_ESC '\047'
1198 #define CHAR_DEL '\007'
1199 #define CHAR_NBSP '\x41'
1200 #define STR_ESC "\047"
1201 #define STR_DEL "\007"
1202
1203 #else /* Not EBCDIC */
1204
1205 /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1206 compatibility. NEL is the Unicode newline character; make sure it is
1207 a positive value. */
1208
1209 #define CHAR_LF '\n'
1210 #define CHAR_NL CHAR_LF
1211 #define CHAR_NEL ((unsigned char)'\x85')
1212 #define CHAR_ESC '\033'
1213 #define CHAR_DEL '\177'
1214 #define CHAR_NBSP ((unsigned char)'\xa0')
1215
1216 #define STR_LF "\n"
1217 #define STR_NL STR_LF
1218 #define STR_NEL "\x85"
1219 #define STR_ESC "\033"
1220 #define STR_DEL "\177"
1221
1222 #endif /* EBCDIC */
1223
1224 /* The remaining definitions work in both environments. */
1225
1226 #define CHAR_NULL '\0'
1227 #define CHAR_HT '\t'
1228 #define CHAR_VT '\v'
1229 #define CHAR_FF '\f'
1230 #define CHAR_CR '\r'
1231 #define CHAR_BS '\b'
1232 #define CHAR_BEL '\a'
1233
1234 #define CHAR_SPACE ' '
1235 #define CHAR_EXCLAMATION_MARK '!'
1236 #define CHAR_QUOTATION_MARK '"'
1237 #define CHAR_NUMBER_SIGN '#'
1238 #define CHAR_DOLLAR_SIGN '$'
1239 #define CHAR_PERCENT_SIGN '%'
1240 #define CHAR_AMPERSAND '&'
1241 #define CHAR_APOSTROPHE '\''
1242 #define CHAR_LEFT_PARENTHESIS '('
1243 #define CHAR_RIGHT_PARENTHESIS ')'
1244 #define CHAR_ASTERISK '*'
1245 #define CHAR_PLUS '+'
1246 #define CHAR_COMMA ','
1247 #define CHAR_MINUS '-'
1248 #define CHAR_DOT '.'
1249 #define CHAR_SLASH '/'
1250 #define CHAR_0 '0'
1251 #define CHAR_1 '1'
1252 #define CHAR_2 '2'
1253 #define CHAR_3 '3'
1254 #define CHAR_4 '4'
1255 #define CHAR_5 '5'
1256 #define CHAR_6 '6'
1257 #define CHAR_7 '7'
1258 #define CHAR_8 '8'
1259 #define CHAR_9 '9'
1260 #define CHAR_COLON ':'
1261 #define CHAR_SEMICOLON ';'
1262 #define CHAR_LESS_THAN_SIGN '<'
1263 #define CHAR_EQUALS_SIGN '='
1264 #define CHAR_GREATER_THAN_SIGN '>'
1265 #define CHAR_QUESTION_MARK '?'
1266 #define CHAR_COMMERCIAL_AT '@'
1267 #define CHAR_A 'A'
1268 #define CHAR_B 'B'
1269 #define CHAR_C 'C'
1270 #define CHAR_D 'D'
1271 #define CHAR_E 'E'
1272 #define CHAR_F 'F'
1273 #define CHAR_G 'G'
1274 #define CHAR_H 'H'
1275 #define CHAR_I 'I'
1276 #define CHAR_J 'J'
1277 #define CHAR_K 'K'
1278 #define CHAR_L 'L'
1279 #define CHAR_M 'M'
1280 #define CHAR_N 'N'
1281 #define CHAR_O 'O'
1282 #define CHAR_P 'P'
1283 #define CHAR_Q 'Q'
1284 #define CHAR_R 'R'
1285 #define CHAR_S 'S'
1286 #define CHAR_T 'T'
1287 #define CHAR_U 'U'
1288 #define CHAR_V 'V'
1289 #define CHAR_W 'W'
1290 #define CHAR_X 'X'
1291 #define CHAR_Y 'Y'
1292 #define CHAR_Z 'Z'
1293 #define CHAR_LEFT_SQUARE_BRACKET '['
1294 #define CHAR_BACKSLASH '\\'
1295 #define CHAR_RIGHT_SQUARE_BRACKET ']'
1296 #define CHAR_CIRCUMFLEX_ACCENT '^'
1297 #define CHAR_UNDERSCORE '_'
1298 #define CHAR_GRAVE_ACCENT '`'
1299 #define CHAR_a 'a'
1300 #define CHAR_b 'b'
1301 #define CHAR_c 'c'
1302 #define CHAR_d 'd'
1303 #define CHAR_e 'e'
1304 #define CHAR_f 'f'
1305 #define CHAR_g 'g'
1306 #define CHAR_h 'h'
1307 #define CHAR_i 'i'
1308 #define CHAR_j 'j'
1309 #define CHAR_k 'k'
1310 #define CHAR_l 'l'
1311 #define CHAR_m 'm'
1312 #define CHAR_n 'n'
1313 #define CHAR_o 'o'
1314 #define CHAR_p 'p'
1315 #define CHAR_q 'q'
1316 #define CHAR_r 'r'
1317 #define CHAR_s 's'
1318 #define CHAR_t 't'
1319 #define CHAR_u 'u'
1320 #define CHAR_v 'v'
1321 #define CHAR_w 'w'
1322 #define CHAR_x 'x'
1323 #define CHAR_y 'y'
1324 #define CHAR_z 'z'
1325 #define CHAR_LEFT_CURLY_BRACKET '{'
1326 #define CHAR_VERTICAL_LINE '|'
1327 #define CHAR_RIGHT_CURLY_BRACKET '}'
1328 #define CHAR_TILDE '~'
1329
1330 #define STR_HT "\t"
1331 #define STR_VT "\v"
1332 #define STR_FF "\f"
1333 #define STR_CR "\r"
1334 #define STR_BS "\b"
1335 #define STR_BEL "\a"
1336
1337 #define STR_SPACE " "
1338 #define STR_EXCLAMATION_MARK "!"
1339 #define STR_QUOTATION_MARK "\""
1340 #define STR_NUMBER_SIGN "#"
1341 #define STR_DOLLAR_SIGN "$"
1342 #define STR_PERCENT_SIGN "%"
1343 #define STR_AMPERSAND "&"
1344 #define STR_APOSTROPHE "'"
1345 #define STR_LEFT_PARENTHESIS "("
1346 #define STR_RIGHT_PARENTHESIS ")"
1347 #define STR_ASTERISK "*"
1348 #define STR_PLUS "+"
1349 #define STR_COMMA ","
1350 #define STR_MINUS "-"
1351 #define STR_DOT "."
1352 #define STR_SLASH "/"
1353 #define STR_0 "0"
1354 #define STR_1 "1"
1355 #define STR_2 "2"
1356 #define STR_3 "3"
1357 #define STR_4 "4"
1358 #define STR_5 "5"
1359 #define STR_6 "6"
1360 #define STR_7 "7"
1361 #define STR_8 "8"
1362 #define STR_9 "9"
1363 #define STR_COLON ":"
1364 #define STR_SEMICOLON ";"
1365 #define STR_LESS_THAN_SIGN "<"
1366 #define STR_EQUALS_SIGN "="
1367 #define STR_GREATER_THAN_SIGN ">"
1368 #define STR_QUESTION_MARK "?"
1369 #define STR_COMMERCIAL_AT "@"
1370 #define STR_A "A"
1371 #define STR_B "B"
1372 #define STR_C "C"
1373 #define STR_D "D"
1374 #define STR_E "E"
1375 #define STR_F "F"
1376 #define STR_G "G"
1377 #define STR_H "H"
1378 #define STR_I "I"
1379 #define STR_J "J"
1380 #define STR_K "K"
1381 #define STR_L "L"
1382 #define STR_M "M"
1383 #define STR_N "N"
1384 #define STR_O "O"
1385 #define STR_P "P"
1386 #define STR_Q "Q"
1387 #define STR_R "R"
1388 #define STR_S "S"
1389 #define STR_T "T"
1390 #define STR_U "U"
1391 #define STR_V "V"
1392 #define STR_W "W"
1393 #define STR_X "X"
1394 #define STR_Y "Y"
1395 #define STR_Z "Z"
1396 #define STR_LEFT_SQUARE_BRACKET "["
1397 #define STR_BACKSLASH "\\"
1398 #define STR_RIGHT_SQUARE_BRACKET "]"
1399 #define STR_CIRCUMFLEX_ACCENT "^"
1400 #define STR_UNDERSCORE "_"
1401 #define STR_GRAVE_ACCENT "`"
1402 #define STR_a "a"
1403 #define STR_b "b"
1404 #define STR_c "c"
1405 #define STR_d "d"
1406 #define STR_e "e"
1407 #define STR_f "f"
1408 #define STR_g "g"
1409 #define STR_h "h"
1410 #define STR_i "i"
1411 #define STR_j "j"
1412 #define STR_k "k"
1413 #define STR_l "l"
1414 #define STR_m "m"
1415 #define STR_n "n"
1416 #define STR_o "o"
1417 #define STR_p "p"
1418 #define STR_q "q"
1419 #define STR_r "r"
1420 #define STR_s "s"
1421 #define STR_t "t"
1422 #define STR_u "u"
1423 #define STR_v "v"
1424 #define STR_w "w"
1425 #define STR_x "x"
1426 #define STR_y "y"
1427 #define STR_z "z"
1428 #define STR_LEFT_CURLY_BRACKET "{"
1429 #define STR_VERTICAL_LINE "|"
1430 #define STR_RIGHT_CURLY_BRACKET "}"
1431 #define STR_TILDE "~"
1432
1433 #define STRING_ACCEPT0 "ACCEPT\0"
1434 #define STRING_COMMIT0 "COMMIT\0"
1435 #define STRING_F0 "F\0"
1436 #define STRING_FAIL0 "FAIL\0"
1437 #define STRING_MARK0 "MARK\0"
1438 #define STRING_PRUNE0 "PRUNE\0"
1439 #define STRING_SKIP0 "SKIP\0"
1440 #define STRING_THEN "THEN"
1441
1442 #define STRING_alpha0 "alpha\0"
1443 #define STRING_lower0 "lower\0"
1444 #define STRING_upper0 "upper\0"
1445 #define STRING_alnum0 "alnum\0"
1446 #define STRING_ascii0 "ascii\0"
1447 #define STRING_blank0 "blank\0"
1448 #define STRING_cntrl0 "cntrl\0"
1449 #define STRING_digit0 "digit\0"
1450 #define STRING_graph0 "graph\0"
1451 #define STRING_print0 "print\0"
1452 #define STRING_punct0 "punct\0"
1453 #define STRING_space0 "space\0"
1454 #define STRING_word0 "word\0"
1455 #define STRING_xdigit "xdigit"
1456
1457 #define STRING_DEFINE "DEFINE"
1458 #define STRING_WEIRD_STARTWORD "[:<:]]"
1459 #define STRING_WEIRD_ENDWORD "[:>:]]"
1460
1461 #define STRING_CR_RIGHTPAR "CR)"
1462 #define STRING_LF_RIGHTPAR "LF)"
1463 #define STRING_CRLF_RIGHTPAR "CRLF)"
1464 #define STRING_ANY_RIGHTPAR "ANY)"
1465 #define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)"
1466 #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"
1467 #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"
1468 #define STRING_UTF8_RIGHTPAR "UTF8)"
1469 #define STRING_UTF16_RIGHTPAR "UTF16)"
1470 #define STRING_UTF32_RIGHTPAR "UTF32)"
1471 #define STRING_UTF_RIGHTPAR "UTF)"
1472 #define STRING_UCP_RIGHTPAR "UCP)"
1473 #define STRING_NO_AUTO_POSSESS_RIGHTPAR "NO_AUTO_POSSESS)"
1474 #define STRING_NO_START_OPT_RIGHTPAR "NO_START_OPT)"
1475 #define STRING_LIMIT_MATCH_EQ "LIMIT_MATCH="
1476 #define STRING_LIMIT_RECURSION_EQ "LIMIT_RECURSION="
1477
1478 #else /* SUPPORT_UTF */
1479
1480 /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1481 works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
1482 only. */
1483
1484 #define CHAR_HT '\011'
1485 #define CHAR_VT '\013'
1486 #define CHAR_FF '\014'
1487 #define CHAR_CR '\015'
1488 #define CHAR_LF '\012'
1489 #define CHAR_NL CHAR_LF
1490 #define CHAR_NEL ((unsigned char)'\x85')
1491 #define CHAR_BS '\010'
1492 #define CHAR_BEL '\007'
1493 #define CHAR_ESC '\033'
1494 #define CHAR_DEL '\177'
1495
1496 #define CHAR_NULL '\0'
1497 #define CHAR_SPACE '\040'
1498 #define CHAR_EXCLAMATION_MARK '\041'
1499 #define CHAR_QUOTATION_MARK '\042'
1500 #define CHAR_NUMBER_SIGN '\043'
1501 #define CHAR_DOLLAR_SIGN '\044'
1502 #define CHAR_PERCENT_SIGN '\045'
1503 #define CHAR_AMPERSAND '\046'
1504 #define CHAR_APOSTROPHE '\047'
1505 #define CHAR_LEFT_PARENTHESIS '\050'
1506 #define CHAR_RIGHT_PARENTHESIS '\051'
1507 #define CHAR_ASTERISK '\052'
1508 #define CHAR_PLUS '\053'
1509 #define CHAR_COMMA '\054'
1510 #define CHAR_MINUS '\055'
1511 #define CHAR_DOT '\056'
1512 #define CHAR_SLASH '\057'
1513 #define CHAR_0 '\060'
1514 #define CHAR_1 '\061'
1515 #define CHAR_2 '\062'
1516 #define CHAR_3 '\063'
1517 #define CHAR_4 '\064'
1518 #define CHAR_5 '\065'
1519 #define CHAR_6 '\066'
1520 #define CHAR_7 '\067'
1521 #define CHAR_8 '\070'
1522 #define CHAR_9 '\071'
1523 #define CHAR_COLON '\072'
1524 #define CHAR_SEMICOLON '\073'
1525 #define CHAR_LESS_THAN_SIGN '\074'
1526 #define CHAR_EQUALS_SIGN '\075'
1527 #define CHAR_GREATER_THAN_SIGN '\076'
1528 #define CHAR_QUESTION_MARK '\077'
1529 #define CHAR_COMMERCIAL_AT '\100'
1530 #define CHAR_A '\101'
1531 #define CHAR_B '\102'
1532 #define CHAR_C '\103'
1533 #define CHAR_D '\104'
1534 #define CHAR_E '\105'
1535 #define CHAR_F '\106'
1536 #define CHAR_G '\107'
1537 #define CHAR_H '\110'
1538 #define CHAR_I '\111'
1539 #define CHAR_J '\112'
1540 #define CHAR_K '\113'
1541 #define CHAR_L '\114'
1542 #define CHAR_M '\115'
1543 #define CHAR_N '\116'
1544 #define CHAR_O '\117'
1545 #define CHAR_P '\120'
1546 #define CHAR_Q '\121'
1547 #define CHAR_R '\122'
1548 #define CHAR_S '\123'
1549 #define CHAR_T '\124'
1550 #define CHAR_U '\125'
1551 #define CHAR_V '\126'
1552 #define CHAR_W '\127'
1553 #define CHAR_X '\130'
1554 #define CHAR_Y '\131'
1555 #define CHAR_Z '\132'
1556 #define CHAR_LEFT_SQUARE_BRACKET '\133'
1557 #define CHAR_BACKSLASH '\134'
1558 #define CHAR_RIGHT_SQUARE_BRACKET '\135'
1559 #define CHAR_CIRCUMFLEX_ACCENT '\136'
1560 #define CHAR_UNDERSCORE '\137'
1561 #define CHAR_GRAVE_ACCENT '\140'
1562 #define CHAR_a '\141'
1563 #define CHAR_b '\142'
1564 #define CHAR_c '\143'
1565 #define CHAR_d '\144'
1566 #define CHAR_e '\145'
1567 #define CHAR_f '\146'
1568 #define CHAR_g '\147'
1569 #define CHAR_h '\150'
1570 #define CHAR_i '\151'
1571 #define CHAR_j '\152'
1572 #define CHAR_k '\153'
1573 #define CHAR_l '\154'
1574 #define CHAR_m '\155'
1575 #define CHAR_n '\156'
1576 #define CHAR_o '\157'
1577 #define CHAR_p '\160'
1578 #define CHAR_q '\161'
1579 #define CHAR_r '\162'
1580 #define CHAR_s '\163'
1581 #define CHAR_t '\164'
1582 #define CHAR_u '\165'
1583 #define CHAR_v '\166'
1584 #define CHAR_w '\167'
1585 #define CHAR_x '\170'
1586 #define CHAR_y '\171'
1587 #define CHAR_z '\172'
1588 #define CHAR_LEFT_CURLY_BRACKET '\173'
1589 #define CHAR_VERTICAL_LINE '\174'
1590 #define CHAR_RIGHT_CURLY_BRACKET '\175'
1591 #define CHAR_TILDE '\176'
1592 #define CHAR_NBSP ((unsigned char)'\xa0')
1593
1594 #define STR_HT "\011"
1595 #define STR_VT "\013"
1596 #define STR_FF "\014"
1597 #define STR_CR "\015"
1598 #define STR_NL "\012"
1599 #define STR_BS "\010"
1600 #define STR_BEL "\007"
1601 #define STR_ESC "\033"
1602 #define STR_DEL "\177"
1603
1604 #define STR_SPACE "\040"
1605 #define STR_EXCLAMATION_MARK "\041"
1606 #define STR_QUOTATION_MARK "\042"
1607 #define STR_NUMBER_SIGN "\043"
1608 #define STR_DOLLAR_SIGN "\044"
1609 #define STR_PERCENT_SIGN "\045"
1610 #define STR_AMPERSAND "\046"
1611 #define STR_APOSTROPHE "\047"
1612 #define STR_LEFT_PARENTHESIS "\050"
1613 #define STR_RIGHT_PARENTHESIS "\051"
1614 #define STR_ASTERISK "\052"
1615 #define STR_PLUS "\053"
1616 #define STR_COMMA "\054"
1617 #define STR_MINUS "\055"
1618 #define STR_DOT "\056"
1619 #define STR_SLASH "\057"
1620 #define STR_0 "\060"
1621 #define STR_1 "\061"
1622 #define STR_2 "\062"
1623 #define STR_3 "\063"
1624 #define STR_4 "\064"
1625 #define STR_5 "\065"
1626 #define STR_6 "\066"
1627 #define STR_7 "\067"
1628 #define STR_8 "\070"
1629 #define STR_9 "\071"
1630 #define STR_COLON "\072"
1631 #define STR_SEMICOLON "\073"
1632 #define STR_LESS_THAN_SIGN "\074"
1633 #define STR_EQUALS_SIGN "\075"
1634 #define STR_GREATER_THAN_SIGN "\076"
1635 #define STR_QUESTION_MARK "\077"
1636 #define STR_COMMERCIAL_AT "\100"
1637 #define STR_A "\101"
1638 #define STR_B "\102"
1639 #define STR_C "\103"
1640 #define STR_D "\104"
1641 #define STR_E "\105"
1642 #define STR_F "\106"
1643 #define STR_G "\107"
1644 #define STR_H "\110"
1645 #define STR_I "\111"
1646 #define STR_J "\112"
1647 #define STR_K "\113"
1648 #define STR_L "\114"
1649 #define STR_M "\115"
1650 #define STR_N "\116"
1651 #define STR_O "\117"
1652 #define STR_P "\120"
1653 #define STR_Q "\121"
1654 #define STR_R "\122"
1655 #define STR_S "\123"
1656 #define STR_T "\124"
1657 #define STR_U "\125"
1658 #define STR_V "\126"
1659 #define STR_W "\127"
1660 #define STR_X "\130"
1661 #define STR_Y "\131"
1662 #define STR_Z "\132"
1663 #define STR_LEFT_SQUARE_BRACKET "\133"
1664 #define STR_BACKSLASH "\134"
1665 #define STR_RIGHT_SQUARE_BRACKET "\135"
1666 #define STR_CIRCUMFLEX_ACCENT "\136"
1667 #define STR_UNDERSCORE "\137"
1668 #define STR_GRAVE_ACCENT "\140"
1669 #define STR_a "\141"
1670 #define STR_b "\142"
1671 #define STR_c "\143"
1672 #define STR_d "\144"
1673 #define STR_e "\145"
1674 #define STR_f "\146"
1675 #define STR_g "\147"
1676 #define STR_h "\150"
1677 #define STR_i "\151"
1678 #define STR_j "\152"
1679 #define STR_k "\153"
1680 #define STR_l "\154"
1681 #define STR_m "\155"
1682 #define STR_n "\156"
1683 #define STR_o "\157"
1684 #define STR_p "\160"
1685 #define STR_q "\161"
1686 #define STR_r "\162"
1687 #define STR_s "\163"
1688 #define STR_t "\164"
1689 #define STR_u "\165"
1690 #define STR_v "\166"
1691 #define STR_w "\167"
1692 #define STR_x "\170"
1693 #define STR_y "\171"
1694 #define STR_z "\172"
1695 #define STR_LEFT_CURLY_BRACKET "\173"
1696 #define STR_VERTICAL_LINE "\174"
1697 #define STR_RIGHT_CURLY_BRACKET "\175"
1698 #define STR_TILDE "\176"
1699
1700 #define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1701 #define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1702 #define STRING_F0 STR_F "\0"
1703 #define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0"
1704 #define STRING_MARK0 STR_M STR_A STR_R STR_K "\0"
1705 #define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0"
1706 #define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0"
1707 #define STRING_THEN STR_T STR_H STR_E STR_N
1708
1709 #define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0"
1710 #define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0"
1711 #define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0"
1712 #define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0"
1713 #define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0"
1714 #define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0"
1715 #define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0"
1716 #define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0"
1717 #define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0"
1718 #define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0"
1719 #define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0"
1720 #define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0"
1721 #define STRING_word0 STR_w STR_o STR_r STR_d "\0"
1722 #define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t
1723
1724 #define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E
1725 #define STRING_WEIRD_STARTWORD STR_LEFT_SQUARE_BRACKET STR_COLON STR_LESS_THAN_SIGN STR_COLON STR_RIGHT_SQUARE_BRACKET STR_RIGHT_SQUARE_BRACKET
1726 #define STRING_WEIRD_ENDWORD STR_LEFT_SQUARE_BRACKET STR_COLON STR_GREATER_THAN_SIGN STR_COLON STR_RIGHT_SQUARE_BRACKET STR_RIGHT_SQUARE_BRACKET
1727
1728 #define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS
1729 #define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS
1730 #define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1731 #define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1732 #define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1733 #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
1734 #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
1735 #define STRING_UTF8_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1736 #define STRING_UTF16_RIGHTPAR STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1737 #define STRING_UTF32_RIGHTPAR STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1738 #define STRING_UTF_RIGHTPAR STR_U STR_T STR_F STR_RIGHT_PARENTHESIS
1739 #define STRING_UCP_RIGHTPAR STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1740 #define STRING_NO_AUTO_POSSESS_RIGHTPAR STR_N STR_O STR_UNDERSCORE STR_A STR_U STR_T STR_O STR_UNDERSCORE STR_P STR_O STR_S STR_S STR_E STR_S STR_S STR_RIGHT_PARENTHESIS
1741 #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
1742 #define STRING_LIMIT_MATCH_EQ STR_L STR_I STR_M STR_I STR_T STR_UNDERSCORE STR_M STR_A STR_T STR_C STR_H STR_EQUALS_SIGN
1743 #define STRING_LIMIT_RECURSION_EQ STR_L STR_I STR_M STR_I STR_T STR_UNDERSCORE STR_R STR_E STR_C STR_U STR_R STR_S STR_I STR_O STR_N STR_EQUALS_SIGN
1744
1745 #endif /* SUPPORT_UTF */
1746
1747 /* Escape items that are just an encoding of a particular data value. */
1748
1749 #ifndef ESC_a
1750 #define ESC_a CHAR_BEL
1751 #endif
1752
1753 #ifndef ESC_e
1754 #define ESC_e CHAR_ESC
1755 #endif
1756
1757 #ifndef ESC_f
1758 #define ESC_f CHAR_FF
1759 #endif
1760
1761 #ifndef ESC_n
1762 #define ESC_n CHAR_LF
1763 #endif
1764
1765 #ifndef ESC_r
1766 #define ESC_r CHAR_CR
1767 #endif
1768
1769 /* We can't officially use ESC_t because it is a POSIX reserved identifier
1770 (presumably because of all the others like size_t). */
1771
1772 #ifndef ESC_tee
1773 #define ESC_tee CHAR_HT
1774 #endif
1775
1776 /* Codes for different types of Unicode property */
1777
1778 #define PT_ANY 0 /* Any property - matches all chars */
1779 #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
1780 #define PT_GC 2 /* Specified general characteristic (e.g. L) */
1781 #define PT_PC 3 /* Specified particular characteristic (e.g. Lu) */
1782 #define PT_SC 4 /* Script (e.g. Han) */
1783 #define PT_ALNUM 5 /* Alphanumeric - the union of L and N */
1784 #define PT_SPACE 6 /* Perl space - Z plus 9,10,12,13 */
1785 #define PT_PXSPACE 7 /* POSIX space - Z plus 9,10,11,12,13 */
1786 #define PT_WORD 8 /* Word - L plus N plus underscore */
1787 #define PT_CLIST 9 /* Pseudo-property: match character list */
1788 #define PT_UCNC 10 /* Universal Character nameable character */
1789 #define PT_TABSIZE 11 /* Size of square table for autopossessify tests */
1790
1791 /* The following special properties are used only in XCLASS items, when POSIX
1792 classes are specified and PCRE_UCP is set - in other words, for Unicode
1793 handling of these classes. They are not available via the \p or \P escapes like
1794 those in the above list, and so they do not take part in the autopossessifying
1795 table. */
1796
1797 #define PT_PXGRAPH 11 /* [:graph:] - characters that mark the paper */
1798 #define PT_PXPRINT 12 /* [:print:] - [:graph:] plus non-control spaces */
1799 #define PT_PXPUNCT 13 /* [:punct:] - punctuation characters */
1800
1801 /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1802 contain characters with values greater than 255. */
1803
1804 #define XCL_NOT 0x01 /* Flag: this is a negative class */
1805 #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
1806 #define XCL_HASPROP 0x04 /* Flag: property checks are present. */
1807
1808 #define XCL_END 0 /* Marks end of individual items */
1809 #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
1810 #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
1811 #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
1812 #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
1813
1814 /* These are escaped items that aren't just an encoding of a particular data
1815 value such as \n. They must have non-zero values, as check_escape() returns 0
1816 for a data character. Also, they must appear in the same order as in the
1817 opcode definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1818 corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1819 used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1820 non-DOTALL mode, "." behaves like \N.
1821
1822 The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1823 when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1824 They must be contiguous, and remain in order so that the replacements can be
1825 looked up from a table.
1826
1827 Negative numbers are used to encode a backreference (\1, \2, \3, etc.) in
1828 check_escape(). There are two tests in the code for an escape
1829 greater than ESC_b and less than ESC_Z to detect the types that may be
1830 repeated. These are the types that consume characters. If any new escapes are
1831 put in between that don't consume a character, that code will have to change.
1832 */
1833
1834 enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1835 ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1836 ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1837 ESC_E, ESC_Q, ESC_g, ESC_k,
1838 ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1839
1840
1841 /********************** Opcode definitions ******************/
1842
1843 /****** NOTE NOTE NOTE ******
1844
1845 Starting from 1 (i.e. after OP_END), the values up to OP_EOD must correspond in
1846 order to the list of escapes immediately above. Furthermore, values up to
1847 OP_DOLLM must not be changed without adjusting the table called autoposstab in
1848 pcre_compile.c
1849
1850 Whenever this list is updated, the two macro definitions that follow must be
1851 updated to match. The possessification table called "opcode_possessify" in
1852 pcre_compile.c must also be updated, and also the tables called "coptable"
1853 and "poptable" in pcre_dfa_exec.c.
1854
1855 ****** NOTE NOTE NOTE ******/
1856
1857
1858 /* The values between FIRST_AUTOTAB_OP and LAST_AUTOTAB_RIGHT_OP, inclusive,
1859 are used in a table for deciding whether a repeated character type can be
1860 auto-possessified. */
1861
1862 #define FIRST_AUTOTAB_OP OP_NOT_DIGIT
1863 #define LAST_AUTOTAB_LEFT_OP OP_EXTUNI
1864 #define LAST_AUTOTAB_RIGHT_OP OP_DOLLM
1865
1866 enum {
1867 OP_END, /* 0 End of pattern */
1868
1869 /* Values corresponding to backslashed metacharacters */
1870
1871 OP_SOD, /* 1 Start of data: \A */
1872 OP_SOM, /* 2 Start of match (subject + offset): \G */
1873 OP_SET_SOM, /* 3 Set start of match (\K) */
1874 OP_NOT_WORD_BOUNDARY, /* 4 \B */
1875 OP_WORD_BOUNDARY, /* 5 \b */
1876 OP_NOT_DIGIT, /* 6 \D */
1877 OP_DIGIT, /* 7 \d */
1878 OP_NOT_WHITESPACE, /* 8 \S */
1879 OP_WHITESPACE, /* 9 \s */
1880 OP_NOT_WORDCHAR, /* 10 \W */
1881 OP_WORDCHAR, /* 11 \w */
1882
1883 OP_ANY, /* 12 Match any character except newline (\N) */
1884 OP_ALLANY, /* 13 Match any character */
1885 OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1886 OP_NOTPROP, /* 15 \P (not Unicode property) */
1887 OP_PROP, /* 16 \p (Unicode property) */
1888 OP_ANYNL, /* 17 \R (any newline sequence) */
1889 OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */
1890 OP_HSPACE, /* 19 \h (horizontal whitespace) */
1891 OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */
1892 OP_VSPACE, /* 21 \v (vertical whitespace) */
1893 OP_EXTUNI, /* 22 \X (extended Unicode sequence */
1894 OP_EODN, /* 23 End of data or \n at end of data (\Z) */
1895 OP_EOD, /* 24 End of data (\z) */
1896
1897 /* Line end assertions */
1898
1899 OP_DOLL, /* 25 End of line - not multiline */
1900 OP_DOLLM, /* 26 End of line - multiline */
1901 OP_CIRC, /* 27 Start of line - not multiline */
1902 OP_CIRCM, /* 28 Start of line - multiline */
1903
1904 /* Single characters; caseful must precede the caseless ones */
1905
1906 OP_CHAR, /* 29 Match one character, casefully */
1907 OP_CHARI, /* 30 Match one character, caselessly */
1908 OP_NOT, /* 31 Match one character, not the given one, casefully */
1909 OP_NOTI, /* 32 Match one character, not the given one, caselessly */
1910
1911 /* The following sets of 13 opcodes must always be kept in step because
1912 the offset from the first one is used to generate the others. */
1913
1914 /* Repeated characters; caseful must precede the caseless ones */
1915
1916 OP_STAR, /* 33 The maximizing and minimizing versions of */
1917 OP_MINSTAR, /* 34 these six opcodes must come in pairs, with */
1918 OP_PLUS, /* 35 the minimizing one second. */
1919 OP_MINPLUS, /* 36 */
1920 OP_QUERY, /* 37 */
1921 OP_MINQUERY, /* 38 */
1922
1923 OP_UPTO, /* 39 From 0 to n matches of one character, caseful*/
1924 OP_MINUPTO, /* 40 */
1925 OP_EXACT, /* 41 Exactly n matches */
1926
1927 OP_POSSTAR, /* 42 Possessified star, caseful */
1928 OP_POSPLUS, /* 43 Possessified plus, caseful */
1929 OP_POSQUERY, /* 44 Posesssified query, caseful */
1930 OP_POSUPTO, /* 45 Possessified upto, caseful */
1931
1932 /* Repeated characters; caseless must follow the caseful ones */
1933
1934 OP_STARI, /* 46 */
1935 OP_MINSTARI, /* 47 */
1936 OP_PLUSI, /* 48 */
1937 OP_MINPLUSI, /* 49 */
1938 OP_QUERYI, /* 50 */
1939 OP_MINQUERYI, /* 51 */
1940
1941 OP_UPTOI, /* 52 From 0 to n matches of one character, caseless */
1942 OP_MINUPTOI, /* 53 */
1943 OP_EXACTI, /* 54 */
1944
1945 OP_POSSTARI, /* 55 Possessified star, caseless */
1946 OP_POSPLUSI, /* 56 Possessified plus, caseless */
1947 OP_POSQUERYI, /* 57 Posesssified query, caseless */
1948 OP_POSUPTOI, /* 58 Possessified upto, caseless */
1949
1950 /* The negated ones must follow the non-negated ones, and match them */
1951 /* Negated repeated character, caseful; must precede the caseless ones */
1952
1953 OP_NOTSTAR, /* 59 The maximizing and minimizing versions of */
1954 OP_NOTMINSTAR, /* 60 these six opcodes must come in pairs, with */
1955 OP_NOTPLUS, /* 61 the minimizing one second. They must be in */
1956 OP_NOTMINPLUS, /* 62 exactly the same order as those above. */
1957 OP_NOTQUERY, /* 63 */
1958 OP_NOTMINQUERY, /* 64 */
1959
1960 OP_NOTUPTO, /* 65 From 0 to n matches, caseful */
1961 OP_NOTMINUPTO, /* 66 */
1962 OP_NOTEXACT, /* 67 Exactly n matches */
1963
1964 OP_NOTPOSSTAR, /* 68 Possessified versions, caseful */
1965 OP_NOTPOSPLUS, /* 69 */
1966 OP_NOTPOSQUERY, /* 70 */
1967 OP_NOTPOSUPTO, /* 71 */
1968
1969 /* Negated repeated character, caseless; must follow the caseful ones */
1970
1971 OP_NOTSTARI, /* 72 */
1972 OP_NOTMINSTARI, /* 73 */
1973 OP_NOTPLUSI, /* 74 */
1974 OP_NOTMINPLUSI, /* 75 */
1975 OP_NOTQUERYI, /* 76 */
1976 OP_NOTMINQUERYI, /* 77 */
1977
1978 OP_NOTUPTOI, /* 78 From 0 to n matches, caseless */
1979 OP_NOTMINUPTOI, /* 79 */
1980 OP_NOTEXACTI, /* 80 Exactly n matches */
1981
1982 OP_NOTPOSSTARI, /* 81 Possessified versions, caseless */
1983 OP_NOTPOSPLUSI, /* 82 */
1984 OP_NOTPOSQUERYI, /* 83 */
1985 OP_NOTPOSUPTOI, /* 84 */
1986
1987 /* Character types */
1988
1989 OP_TYPESTAR, /* 85 The maximizing and minimizing versions of */
1990 OP_TYPEMINSTAR, /* 86 these six opcodes must come in pairs, with */
1991 OP_TYPEPLUS, /* 87 the minimizing one second. These codes must */
1992 OP_TYPEMINPLUS, /* 88 be in exactly the same order as those above. */
1993 OP_TYPEQUERY, /* 89 */
1994 OP_TYPEMINQUERY, /* 90 */
1995
1996 OP_TYPEUPTO, /* 91 From 0 to n matches */
1997 OP_TYPEMINUPTO, /* 92 */
1998 OP_TYPEEXACT, /* 93 Exactly n matches */
1999
2000 OP_TYPEPOSSTAR, /* 94 Possessified versions */
2001 OP_TYPEPOSPLUS, /* 95 */
2002 OP_TYPEPOSQUERY, /* 96 */
2003 OP_TYPEPOSUPTO, /* 97 */
2004
2005 /* These are used for character classes and back references; only the
2006 first six are the same as the sets above. */
2007
2008 OP_CRSTAR, /* 98 The maximizing and minimizing versions of */
2009 OP_CRMINSTAR, /* 99 all these opcodes must come in pairs, with */
2010 OP_CRPLUS, /* 100 the minimizing one second. These codes must */
2011 OP_CRMINPLUS, /* 101 be in exactly the same order as those above. */
2012 OP_CRQUERY, /* 102 */
2013 OP_CRMINQUERY, /* 103 */
2014
2015 OP_CRRANGE, /* 104 These are different to the three sets above. */
2016 OP_CRMINRANGE, /* 105 */
2017
2018 OP_CRPOSSTAR, /* 106 Possessified versions */
2019 OP_CRPOSPLUS, /* 107 */
2020 OP_CRPOSQUERY, /* 108 */
2021 OP_CRPOSRANGE, /* 109 */
2022
2023 /* End of quantifier opcodes */
2024
2025 OP_CLASS, /* 110 Match a character class, chars < 256 only */
2026 OP_NCLASS, /* 111 Same, but the bitmap was created from a negative
2027 class - the difference is relevant only when a
2028 character > 255 is encountered. */
2029 OP_XCLASS, /* 112 Extended class for handling > 255 chars within the
2030 class. This does both positive and negative. */
2031 OP_REF, /* 113 Match a back reference, casefully */
2032 OP_REFI, /* 114 Match a back reference, caselessly */
2033 OP_DNREF, /* 115 Match a duplicate name backref, casefully */
2034 OP_DNREFI, /* 116 Match a duplicate name backref, caselessly */
2035 OP_RECURSE, /* 117 Match a numbered subpattern (possibly recursive) */
2036 OP_CALLOUT, /* 118 Call out to external function if provided */
2037
2038 OP_ALT, /* 119 Start of alternation */
2039 OP_KET, /* 120 End of group that doesn't have an unbounded repeat */
2040 OP_KETRMAX, /* 121 These two must remain together and in this */
2041 OP_KETRMIN, /* 122 order. They are for groups the repeat for ever. */
2042 OP_KETRPOS, /* 123 Possessive unlimited repeat. */
2043
2044 /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
2045 asserts must remain in order. */
2046
2047 OP_REVERSE, /* 124 Move pointer back - used in lookbehind assertions */
2048 OP_ASSERT, /* 125 Positive lookahead */
2049 OP_ASSERT_NOT, /* 126 Negative lookahead */
2050 OP_ASSERTBACK, /* 127 Positive lookbehind */
2051 OP_ASSERTBACK_NOT, /* 128 Negative lookbehind */
2052
2053 /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
2054 after the assertions, with ONCE first, as there's a test for >= ONCE for a
2055 subpattern that isn't an assertion. The POS versions must immediately follow
2056 the non-POS versions in each case. */
2057
2058 OP_ONCE, /* 129 Atomic group, contains captures */
2059 OP_ONCE_NC, /* 130 Atomic group containing no captures */
2060 OP_BRA, /* 131 Start of non-capturing bracket */
2061 OP_BRAPOS, /* 132 Ditto, with unlimited, possessive repeat */
2062 OP_CBRA, /* 133 Start of capturing bracket */
2063 OP_CBRAPOS, /* 134 Ditto, with unlimited, possessive repeat */
2064 OP_COND, /* 135 Conditional group */
2065
2066 /* These five must follow the previous five, in the same order. There's a
2067 check for >= SBRA to distinguish the two sets. */
2068
2069 OP_SBRA, /* 136 Start of non-capturing bracket, check empty */
2070 OP_SBRAPOS, /* 137 Ditto, with unlimited, possessive repeat */
2071 OP_SCBRA, /* 138 Start of capturing bracket, check empty */
2072 OP_SCBRAPOS, /* 139 Ditto, with unlimited, possessive repeat */
2073 OP_SCOND, /* 140 Conditional group, check empty */
2074
2075 /* The next two pairs must (respectively) be kept together. */
2076
2077 OP_CREF, /* 141 Used to hold a capture number as condition */
2078 OP_DNCREF, /* 142 Used to point to duplicate names as a condition */
2079 OP_RREF, /* 143 Used to hold a recursion number as condition */
2080 OP_DNRREF, /* 144 Used to point to duplicate names as a condition */
2081 OP_DEF, /* 145 The DEFINE condition */
2082
2083 OP_BRAZERO, /* 146 These two must remain together and in this */
2084 OP_BRAMINZERO, /* 147 order. */
2085 OP_BRAPOSZERO, /* 148 */
2086
2087 /* These are backtracking control verbs */
2088
2089 OP_MARK, /* 149 always has an argument */
2090 OP_PRUNE, /* 150 */
2091 OP_PRUNE_ARG, /* 151 same, but with argument */
2092 OP_SKIP, /* 152 */
2093 OP_SKIP_ARG, /* 153 same, but with argument */
2094 OP_THEN, /* 154 */
2095 OP_THEN_ARG, /* 155 same, but with argument */
2096 OP_COMMIT, /* 156 */
2097
2098 /* These are forced failure and success verbs */
2099
2100 OP_FAIL, /* 157 */
2101 OP_ACCEPT, /* 158 */
2102 OP_ASSERT_ACCEPT, /* 159 Used inside assertions */
2103 OP_CLOSE, /* 160 Used before OP_ACCEPT to close open captures */
2104
2105 /* This is used to skip a subpattern with a {0} quantifier */
2106
2107 OP_SKIPZERO, /* 161 */
2108
2109 /* This is not an opcode, but is used to check that tables indexed by opcode
2110 are the correct length, in order to catch updating errors - there have been
2111 some in the past. */
2112
2113 OP_TABLE_LENGTH
2114 };
2115
2116 /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2117 definitions that follow must also be updated to match. There are also tables
2118 called "opcode_possessify" in pcre_compile.c and "coptable" and "poptable" in
2119 pcre_dfa_exec.c that must be updated. */
2120
2121
2122 /* This macro defines textual names for all the opcodes. These are used only
2123 for debugging, and some of them are only partial names. The macro is referenced
2124 only in pcre_printint.c, which fills out the full names in many cases (and in
2125 some cases doesn't actually use these names at all). */
2126
2127 #define OP_NAME_LIST \
2128 "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \
2129 "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \
2130 "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \
2131 "extuni", "\\Z", "\\z", \
2132 "$", "$", "^", "^", "char", "chari", "not", "noti", \
2133 "*", "*?", "+", "+?", "?", "??", \
2134 "{", "{", "{", \
2135 "*+","++", "?+", "{", \
2136 "*", "*?", "+", "+?", "?", "??", \
2137 "{", "{", "{", \
2138 "*+","++", "?+", "{", \
2139 "*", "*?", "+", "+?", "?", "??", \
2140 "{", "{", "{", \
2141 "*+","++", "?+", "{", \
2142 "*", "*?", "+", "+?", "?", "??", \
2143 "{", "{", "{", \
2144 "*+","++", "?+", "{", \
2145 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
2146 "*+","++", "?+", "{", \
2147 "*", "*?", "+", "+?", "?", "??", "{", "{", \
2148 "*+","++", "?+", "{", \
2149 "class", "nclass", "xclass", "Ref", "Refi", "DnRef", "DnRefi", \
2150 "Recurse", "Callout", \
2151 "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos", \
2152 "Reverse", "Assert", "Assert not", "AssertB", "AssertB not", \
2153 "Once", "Once_NC", \
2154 "Bra", "BraPos", "CBra", "CBraPos", \
2155 "Cond", \
2156 "SBra", "SBraPos", "SCBra", "SCBraPos", \
2157 "SCond", \
2158 "Cond ref", "Cond dnref", "Cond rec", "Cond dnrec", "Cond def", \
2159 "Brazero", "Braminzero", "Braposzero", \
2160 "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP", \
2161 "*THEN", "*THEN", "*COMMIT", "*FAIL", \
2162 "*ACCEPT", "*ASSERT_ACCEPT", \
2163 "Close", "Skip zero"
2164
2165
2166 /* This macro defines the length of fixed length operations in the compiled
2167 regex. The lengths are used when searching for specific things, and also in the
2168 debugging printing of a compiled regex. We use a macro so that it can be
2169 defined close to the definitions of the opcodes themselves.
2170
2171 As things have been extended, some of these are no longer fixed lenths, but are
2172 minima instead. For example, the length of a single-character repeat may vary
2173 in UTF-8 mode. The code that uses this table must know about such things. */
2174
2175 #define OP_LENGTHS \
2176 1, /* End */ \
2177 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \
2178 1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \
2179 1, 1, 1, /* Any, AllAny, Anybyte */ \
2180 3, 3, /* \P, \p */ \
2181 1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \
2182 1, /* \X */ \
2183 1, 1, 1, 1, 1, 1, /* \Z, \z, $, $M ^, ^M */ \
2184 2, /* Char - the minimum length */ \
2185 2, /* Chari - the minimum length */ \
2186 2, /* not */ \
2187 2, /* noti */ \
2188 /* Positive single-char repeats ** These are */ \
2189 2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
2190 2+IMM2_SIZE, 2+IMM2_SIZE, /* upto, minupto ** mode */ \
2191 2+IMM2_SIZE, /* exact */ \
2192 2, 2, 2, 2+IMM2_SIZE, /* *+, ++, ?+, upto+ */ \
2193 2, 2, 2, 2, 2, 2, /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8 */ \
2194 2+IMM2_SIZE, 2+IMM2_SIZE, /* upto I, minupto I */ \
2195 2+IMM2_SIZE, /* exact I */ \
2196 2, 2, 2, 2+IMM2_SIZE, /* *+I, ++I, ?+I, upto+I */ \
2197 /* Negative single-char repeats - only for chars < 256 */ \
2198 2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
2199 2+IMM2_SIZE, 2+IMM2_SIZE, /* NOT upto, minupto */ \
2200 2+IMM2_SIZE, /* NOT exact */ \
2201 2, 2, 2, 2+IMM2_SIZE, /* Possessive NOT *, +, ?, upto */ \
2202 2, 2, 2, 2, 2, 2, /* NOT *I, *?I, +I, +?I, ?I, ??I */ \
2203 2+IMM2_SIZE, 2+IMM2_SIZE, /* NOT upto I, minupto I */ \
2204 2+IMM2_SIZE, /* NOT exact I */ \
2205 2, 2, 2, 2+IMM2_SIZE, /* Possessive NOT *I, +I, ?I, upto I */ \
2206 /* Positive type repeats */ \
2207 2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
2208 2+IMM2_SIZE, 2+IMM2_SIZE, /* Type upto, minupto */ \
2209 2+IMM2_SIZE, /* Type exact */ \
2210 2, 2, 2, 2+IMM2_SIZE, /* Possessive *+, ++, ?+, upto+ */ \
2211 /* Character class & ref repeats */ \
2212 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
2213 1+2*IMM2_SIZE, 1+2*IMM2_SIZE, /* CRRANGE, CRMINRANGE */ \
2214 1, 1, 1, 1+2*IMM2_SIZE, /* Possessive *+, ++, ?+, CRPOSRANGE */ \
2215 1+(32/sizeof(pcre_uchar)), /* CLASS */ \
2216 1+(32/sizeof(pcre_uchar)), /* NCLASS */ \
2217 0, /* XCLASS - variable length */ \
2218 1+IMM2_SIZE, /* REF */ \
2219 1+IMM2_SIZE, /* REFI */ \
2220 1+2*IMM2_SIZE, /* DNREF */ \
2221 1+2*IMM2_SIZE, /* DNREFI */ \
2222 1+LINK_SIZE, /* RECURSE */ \
2223 2+2*LINK_SIZE, /* CALLOUT */ \
2224 1+LINK_SIZE, /* Alt */ \
2225 1+LINK_SIZE, /* Ket */ \
2226 1+LINK_SIZE, /* KetRmax */ \
2227 1+LINK_SIZE, /* KetRmin */ \
2228 1+LINK_SIZE, /* KetRpos */ \
2229 1+LINK_SIZE, /* Reverse */ \
2230 1+LINK_SIZE, /* Assert */ \
2231 1+LINK_SIZE, /* Assert not */ \
2232 1+LINK_SIZE, /* Assert behind */ \
2233 1+LINK_SIZE, /* Assert behind not */ \
2234 1+LINK_SIZE, /* ONCE */ \
2235 1+LINK_SIZE, /* ONCE_NC */ \
2236 1+LINK_SIZE, /* BRA */ \
2237 1+LINK_SIZE, /* BRAPOS */ \
2238 1+LINK_SIZE+IMM2_SIZE, /* CBRA */ \
2239 1+LINK_SIZE+IMM2_SIZE, /* CBRAPOS */ \
2240 1+LINK_SIZE, /* COND */ \
2241 1+LINK_SIZE, /* SBRA */ \
2242 1+LINK_SIZE, /* SBRAPOS */ \
2243 1+LINK_SIZE+IMM2_SIZE, /* SCBRA */ \
2244 1+LINK_SIZE+IMM2_SIZE, /* SCBRAPOS */ \
2245 1+LINK_SIZE, /* SCOND */ \
2246 1+IMM2_SIZE, 1+2*IMM2_SIZE, /* CREF, DNCREF */ \
2247 1+IMM2_SIZE, 1+2*IMM2_SIZE, /* RREF, DNRREF */ \
2248 1, /* DEF */ \
2249 1, 1, 1, /* BRAZERO, BRAMINZERO, BRAPOSZERO */ \
2250 3, 1, 3, /* MARK, PRUNE, PRUNE_ARG */ \
2251 1, 3, /* SKIP, SKIP_ARG */ \
2252 1, 3, /* THEN, THEN_ARG */ \
2253 1, 1, 1, 1, /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT */ \
2254 1+IMM2_SIZE, 1 /* CLOSE, SKIPZERO */
2255
2256 /* A magic value for OP_RREF to indicate the "any recursion" condition. */
2257
2258 #define RREF_ANY 0xffff
2259
2260 /* Compile time error code numbers. They are given names so that they can more
2261 easily be tracked. When a new number is added, the table called eint in
2262 pcreposix.c must be updated. */
2263
2264 enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
2265 ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
2266 ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
2267 ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2268 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2269 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2270 ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2271 ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERR78, ERR79,
2272 ERR80, ERR81, ERR82, ERR83, ERR84, ERR85, ERR86, ERR87, ERRCOUNT };
2273
2274 /* JIT compiling modes. The function list is indexed by them. */
2275
2276 enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2277 JIT_NUMBER_OF_COMPILE_MODES };
2278
2279 /* The real format of the start of the pcre block; the index of names and the
2280 code vector run on as long as necessary after the end. We store an explicit
2281 offset to the name table so that if a regex is compiled on one host, saved, and
2282 then run on another where the size of pointers is different, all might still
2283 be well.
2284
2285 The size of the structure must be a multiple of 8 bytes. For the case of
2286 compiled-on-4 and run-on-8, we include an extra pointer that is always NULL so
2287 that there are an even number of pointers which therefore are a multiple of 8
2288 bytes.
2289
2290 It is necessary to fork the struct for the 32 bit library, since it needs to
2291 use pcre_uint32 for first_char and req_char. We can't put an ifdef inside the
2292 typedef because pcretest needs access to the struct of the 8-, 16- and 32-bit
2293 variants.
2294
2295 *** WARNING ***
2296 When new fields are added to these structures, remember to adjust the code in
2297 pcre_byte_order.c that is concerned with swapping the byte order of the fields
2298 when a compiled regex is reloaded on a host with different endianness.
2299 *** WARNING ***
2300 There is also similar byte-flipping code in pcretest.c, which is used for
2301 testing the byte-flipping features. It must also be kept in step.
2302 *** WARNING ***
2303 */
2304
2305 typedef struct real_pcre8_or_16 {
2306 pcre_uint32 magic_number;
2307 pcre_uint32 size; /* Total that was malloced */
2308 pcre_uint32 options; /* Public options */
2309 pcre_uint32 flags; /* Private flags */
2310 pcre_uint32 limit_match; /* Limit set from regex */
2311 pcre_uint32 limit_recursion; /* Limit set from regex */
2312 pcre_uint16 first_char; /* Starting character */
2313 pcre_uint16 req_char; /* This character must be seen */
2314 pcre_uint16 max_lookbehind; /* Longest lookbehind (characters) */
2315 pcre_uint16 top_bracket; /* Highest numbered group */
2316 pcre_uint16 top_backref; /* Highest numbered back reference */
2317 pcre_uint16 name_table_offset; /* Offset to name table that follows */
2318 pcre_uint16 name_entry_size; /* Size of any name items */
2319 pcre_uint16 name_count; /* Number of name items */
2320 pcre_uint16 ref_count; /* Reference count */
2321 pcre_uint16 dummy1; /* To ensure size is a multiple of 8 */
2322 pcre_uint16 dummy2; /* To ensure size is a multiple of 8 */
2323 pcre_uint16 dummy3; /* To ensure size is a multiple of 8 */
2324 const pcre_uint8 *tables; /* Pointer to tables or NULL for std */
2325 void *nullpad; /* NULL padding */
2326 } real_pcre8_or_16;
2327
2328 typedef struct real_pcre8_or_16 real_pcre;
2329 typedef struct real_pcre8_or_16 real_pcre16;
2330
2331 typedef struct real_pcre32 {
2332 pcre_uint32 magic_number;
2333 pcre_uint32 size; /* Total that was malloced */
2334 pcre_uint32 options; /* Public options */
2335 pcre_uint32 flags; /* Private flags */
2336 pcre_uint32 limit_match; /* Limit set from regex */
2337 pcre_uint32 limit_recursion; /* Limit set from regex */
2338 pcre_uint32 first_char; /* Starting character */
2339 pcre_uint32 req_char; /* This character must be seen */
2340 pcre_uint16 max_lookbehind; /* Longest lookbehind (characters) */
2341 pcre_uint16 top_bracket; /* Highest numbered group */
2342 pcre_uint16 top_backref; /* Highest numbered back reference */
2343 pcre_uint16 name_table_offset; /* Offset to name table that follows */
2344 pcre_uint16 name_entry_size; /* Size of any name items */
2345 pcre_uint16 name_count; /* Number of name items */
2346 pcre_uint16 ref_count; /* Reference count */
2347 pcre_uint16 dummy; /* To ensure size is a multiple of 8 */
2348 const pcre_uint8 *tables; /* Pointer to tables or NULL for std */
2349 void *nullpad; /* NULL padding */
2350 } real_pcre32;
2351
2352 #if defined COMPILE_PCRE8
2353 #define REAL_PCRE real_pcre
2354 #elif defined COMPILE_PCRE16
2355 #define REAL_PCRE real_pcre16
2356 #elif defined COMPILE_PCRE32
2357 #define REAL_PCRE real_pcre32
2358 #endif
2359
2360 /* Assert that the size of REAL_PCRE is divisible by 8 */
2361 typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2362
2363 /* Needed in pcretest to access some fields in the real_pcre* structures
2364 * directly. They're unified for 8/16/32 bits since the structs only differ
2365 * after these fields; if that ever changes, need to fork those defines into
2366 * 8/16 and 32 bit versions. */
2367 #define REAL_PCRE_MAGIC(re) (((REAL_PCRE*)re)->magic_number)
2368 #define REAL_PCRE_SIZE(re) (((REAL_PCRE*)re)->size)
2369 #define REAL_PCRE_OPTIONS(re) (((REAL_PCRE*)re)->options)
2370 #define REAL_PCRE_FLAGS(re) (((REAL_PCRE*)re)->flags)
2371
2372 /* The format of the block used to store data from pcre_study(). The same
2373 remark (see NOTE above) about extending this structure applies. */
2374
2375 typedef struct pcre_study_data {
2376 pcre_uint32 size; /* Total that was malloced */
2377 pcre_uint32 flags; /* Private flags */
2378 pcre_uint8 start_bits[32]; /* Starting char bits */
2379 pcre_uint32 minlength; /* Minimum subject length */
2380 } pcre_study_data;
2381
2382 /* Structure for building a chain of open capturing subpatterns during
2383 compiling, so that instructions to close them can be compiled when (*ACCEPT) is
2384 encountered. This is also used to identify subpatterns that contain recursive
2385 back references to themselves, so that they can be made atomic. */
2386
2387 typedef struct open_capitem {
2388 struct open_capitem *next; /* Chain link */
2389 pcre_uint16 number; /* Capture number */
2390 pcre_uint16 flag; /* Set TRUE if recursive back ref */
2391 } open_capitem;
2392
2393 /* Structure for building a list of named groups during the first pass of
2394 compiling. */
2395
2396 typedef struct named_group {
2397 const pcre_uchar *name; /* Points to the name in the pattern */
2398 int length; /* Length of the name */
2399 pcre_uint32 number; /* Group number */
2400 } named_group;
2401
2402 /* Structure for passing "static" information around between the functions
2403 doing the compiling, so that they are thread-safe. */
2404
2405 typedef struct compile_data {
2406 const pcre_uint8 *lcc; /* Points to lower casing table */
2407 const pcre_uint8 *fcc; /* Points to case-flipping table */
2408 const pcre_uint8 *cbits; /* Points to character type table */
2409 const pcre_uint8 *ctypes; /* Points to table of type maps */
2410 const pcre_uchar *start_workspace;/* The start of working space */
2411 const pcre_uchar *start_code; /* The start of the compiled code */
2412 const pcre_uchar *start_pattern; /* The start of the pattern */
2413 const pcre_uchar *end_pattern; /* The end of the pattern */
2414 pcre_uchar *hwm; /* High watermark of workspace */
2415 open_capitem *open_caps; /* Chain of open capture items */
2416 named_group *named_groups; /* Points to vector in pre-compile */
2417 pcre_uchar *name_table; /* The name/number table */
2418 int names_found; /* Number of entries so far */
2419 int name_entry_size; /* Size of each entry */
2420 int named_group_list_size; /* Number of entries in the list */
2421 int workspace_size; /* Size of workspace */
2422 unsigned int bracount; /* Count of capturing parens as we compile */
2423 int final_bracount; /* Saved value after first pass */
2424 int max_lookbehind; /* Maximum lookbehind (characters) */
2425 int top_backref; /* Maximum back reference */
2426 unsigned int backref_map; /* Bitmap of low back refs */
2427 unsigned int namedrefcount; /* Number of backreferences by name */
2428 int parens_depth; /* Depth of nested parentheses */
2429 int assert_depth; /* Depth of nested assertions */
2430 pcre_uint32 external_options; /* External (initial) options */
2431 pcre_uint32 external_flags; /* External flag bits to be set */
2432 int req_varyopt; /* "After variable item" flag for reqbyte */
2433 BOOL had_accept; /* (*ACCEPT) encountered */
2434 BOOL had_pruneorskip; /* (*PRUNE) or (*SKIP) encountered */
2435 BOOL check_lookbehind; /* Lookbehinds need later checking */
2436 BOOL dupnames; /* Duplicate names exist */
2437 BOOL dupgroups; /* Duplicate groups exist: (?| found */
2438 BOOL iscondassert; /* Next assert is a condition */
2439 int nltype; /* Newline type */
2440 int nllen; /* Newline string length */
2441 pcre_uchar nl[4]; /* Newline string when fixed length */
2442 } compile_data;
2443
2444 /* Structure for maintaining a chain of pointers to the currently incomplete
2445 branches, for testing for left recursion while compiling. */
2446
2447 typedef struct branch_chain {
2448 struct branch_chain *outer;
2449 pcre_uchar *current_branch;
2450 } branch_chain;
2451
2452 /* Structure for mutual recursion detection. */
2453
2454 typedef struct recurse_check {
2455 struct recurse_check *prev;
2456 const pcre_uchar *group;
2457 } recurse_check;
2458
2459 /* Structure for items in a linked list that represents an explicit recursive
2460 call within the pattern; used by pcre_exec(). */
2461
2462 typedef struct recursion_info {
2463 struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2464 unsigned int group_num; /* Number of group that was called */
2465 int *offset_save; /* Pointer to start of saved offsets */
2466 int saved_max; /* Number of saved offsets */
2467 int saved_capture_last; /* Last capture number */
2468 PCRE_PUCHAR subject_position; /* Position at start of recursion */
2469 } recursion_info;
2470
2471 /* A similar structure for pcre_dfa_exec(). */
2472
2473 typedef struct dfa_recursion_info {
2474 struct dfa_recursion_info *prevrec;
2475 int group_num;
2476 PCRE_PUCHAR subject_position;
2477 } dfa_recursion_info;
2478
2479 /* Structure for building a chain of data for holding the values of the subject
2480 pointer at the start of each subpattern, so as to detect when an empty string
2481 has been matched by a subpattern - to break infinite loops; used by
2482 pcre_exec(). */
2483
2484 typedef struct eptrblock {
2485 struct eptrblock *epb_prev;
2486 PCRE_PUCHAR epb_saved_eptr;
2487 } eptrblock;
2488
2489
2490 /* Structure for passing "static" information around between the functions
2491 doing traditional NFA matching, so that they are thread-safe. */
2492
2493 typedef struct match_data {
2494 unsigned long int match_call_count; /* As it says */
2495 unsigned long int match_limit; /* As it says */
2496 unsigned long int match_limit_recursion; /* As it says */
2497 int *offset_vector; /* Offset vector */
2498 int offset_end; /* One past the end */
2499 int offset_max; /* The maximum usable for return data */
2500 int nltype; /* Newline type */
2501 int nllen; /* Newline string length */
2502 int name_count; /* Number of names in name table */
2503 int name_entry_size; /* Size of entry in names table */
2504 unsigned int skip_arg_count; /* For counting SKIP_ARGs */
2505 unsigned int ignore_skip_arg; /* For re-run when SKIP arg name not found */
2506 pcre_uchar *name_table; /* Table of names */
2507 pcre_uchar nl[4]; /* Newline string when fixed */
2508 const pcre_uint8 *lcc; /* Points to lower casing table */
2509 const pcre_uint8 *fcc; /* Points to case-flipping table */
2510 const pcre_uint8 *ctypes; /* Points to table of type maps */
2511 BOOL notbol; /* NOTBOL flag */
2512 BOOL noteol; /* NOTEOL flag */
2513 BOOL utf; /* UTF-8 / UTF-16 flag */
2514 BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */
2515 BOOL use_ucp; /* PCRE_UCP flag */
2516 BOOL endonly; /* Dollar not before final \n */
2517 BOOL notempty; /* Empty string match not wanted */
2518 BOOL notempty_atstart; /* Empty string match at start not wanted */
2519 BOOL hitend; /* Hit the end of the subject at some point */
2520 BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
2521 BOOL hasthen; /* Pattern contains (*THEN) */
2522 const pcre_uchar *start_code; /* For use when recursing */
2523 PCRE_PUCHAR start_subject; /* Start of the subject string */
2524 PCRE_PUCHAR end_subject; /* End of the subject string */
2525 PCRE_PUCHAR start_match_ptr; /* Start of matched string */
2526 PCRE_PUCHAR end_match_ptr; /* Subject position at end match */
2527 PCRE_PUCHAR start_used_ptr; /* Earliest consulted character */
2528 int partial; /* PARTIAL options */
2529 int end_offset_top; /* Highwater mark at end of match */
2530 pcre_int32 capture_last; /* Most recent capture number + overflow flag */
2531 int start_offset; /* The start offset value */
2532 int match_function_type; /* Set for certain special calls of MATCH() */
2533 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
2534 int eptrn; /* Next free eptrblock */
2535 recursion_info *recursive; /* Linked list of recursion data */
2536 void *callout_data; /* To pass back to callouts */
2537 const pcre_uchar *mark; /* Mark pointer to pass back on success */
2538 const pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2539 const pcre_uchar *once_target; /* Where to back up to for atomic groups */
2540 #ifdef NO_RECURSE
2541 void *match_frames_base; /* For remembering malloc'd frames */
2542 #endif
2543 } match_data;
2544
2545 /* A similar structure is used for the same purpose by the DFA matching
2546 functions. */
2547
2548 typedef struct dfa_match_data {
2549 const pcre_uchar *start_code; /* Start of the compiled pattern */
2550 const pcre_uchar *start_subject ; /* Start of the subject string */
2551 const pcre_uchar *end_subject; /* End of subject string */
2552 const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2553 const pcre_uint8 *tables; /* Character tables */
2554 int start_offset; /* The start offset value */
2555 int moptions; /* Match options */
2556 int poptions; /* Pattern options */
2557 int nltype; /* Newline type */
2558 int nllen; /* Newline string length */
2559 pcre_uchar nl[4]; /* Newline string when fixed */
2560 void *callout_data; /* To pass back to callouts */
2561 dfa_recursion_info *recursive; /* Linked list of recursion data */
2562 } dfa_match_data;
2563
2564 /* Bit definitions for entries in the pcre_ctypes table. */
2565
2566 #define ctype_space 0x01
2567 #define ctype_letter 0x02
2568 #define ctype_digit 0x04
2569 #define ctype_xdigit 0x08
2570 #define ctype_word 0x10 /* alphanumeric or '_' */
2571 #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
2572
2573 /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
2574 of bits for a class map. Some classes are built by combining these tables. */
2575
2576 #define cbit_space 0 /* [:space:] or \s */
2577 #define cbit_xdigit 32 /* [:xdigit:] */
2578 #define cbit_digit 64 /* [:digit:] or \d */
2579 #define cbit_upper 96 /* [:upper:] */
2580 #define cbit_lower 128 /* [:lower:] */
2581 #define cbit_word 160 /* [:word:] or \w */
2582 #define cbit_graph 192 /* [:graph:] */
2583 #define cbit_print 224 /* [:print:] */
2584 #define cbit_punct 256 /* [:punct:] */
2585 #define cbit_cntrl 288 /* [:cntrl:] */
2586 #define cbit_length 320 /* Length of the cbits table */
2587
2588 /* Offsets of the various tables from the base tables pointer, and
2589 total length. */
2590
2591 #define lcc_offset 0
2592 #define fcc_offset 256
2593 #define cbits_offset 512
2594 #define ctypes_offset (cbits_offset + cbit_length)
2595 #define tables_length (ctypes_offset + 256)
2596
2597 /* Internal function and data prefixes. */
2598
2599 #if defined COMPILE_PCRE8
2600 #ifndef PUBL
2601 #define PUBL(name) pcre_##name
2602 #endif
2603 #ifndef PRIV
2604 #define PRIV(name) _pcre_##name
2605 #endif
2606 #elif defined COMPILE_PCRE16
2607 #ifndef PUBL
2608 #define PUBL(name) pcre16_##name
2609 #endif
2610 #ifndef PRIV
2611 #define PRIV(name) _pcre16_##name
2612 #endif
2613 #elif defined COMPILE_PCRE32
2614 #ifndef PUBL
2615 #define PUBL(name) pcre32_##name
2616 #endif
2617 #ifndef PRIV
2618 #define PRIV(name) _pcre32_##name
2619 #endif
2620 #else
2621 #error Unsupported compiling mode
2622 #endif /* COMPILE_PCRE[8|16|32] */
2623
2624 /* Layout of the UCP type table that translates property names into types and
2625 codes. Each entry used to point directly to a name, but to reduce the number of
2626 relocations in shared libraries, it now has an offset into a single string
2627 instead. */
2628
2629 typedef struct {
2630 pcre_uint16 name_offset;
2631 pcre_uint16 type;
2632 pcre_uint16 value;
2633 } ucp_type_table;
2634
2635
2636 /* Internal shared data tables. These are tables that are used by more than one
2637 of the exported public functions. They have to be "external" in the C sense,
2638 but are not part of the PCRE public API. The data for these tables is in the
2639 pcre_tables.c module. */
2640
2641 #ifdef COMPILE_PCRE8
2642 extern const int PRIV(utf8_table1)[];
2643 extern const int PRIV(utf8_table1_size);
2644 extern const int PRIV(utf8_table2)[];
2645 extern const int PRIV(utf8_table3)[];
2646 extern const pcre_uint8 PRIV(utf8_table4)[];
2647 #endif /* COMPILE_PCRE8 */
2648
2649 extern const char PRIV(utt_names)[];
2650 extern const ucp_type_table PRIV(utt)[];
2651 extern const int PRIV(utt_size);
2652
2653 extern const pcre_uint8 PRIV(OP_lengths)[];
2654 extern const pcre_uint8 PRIV(default_tables)[];
2655
2656 extern const pcre_uint32 PRIV(hspace_list)[];
2657 extern const pcre_uint32 PRIV(vspace_list)[];
2658
2659
2660 /* Internal shared functions. These are functions that are used by more than
2661 one of the exported public functions. They have to be "external" in the C
2662 sense, but are not part of the PCRE public API. */
2663
2664 /* String comparison functions. */
2665 #if defined COMPILE_PCRE8
2666
2667 #define STRCMP_UC_UC(str1, str2) \
2668 strcmp((char *)(str1), (char *)(str2))
2669 #define STRCMP_UC_C8(str1, str2) \
2670 strcmp((char *)(str1), (str2))
2671 #define STRNCMP_UC_UC(str1, str2, num) \
2672 strncmp((char *)(str1), (char *)(str2), (num))
2673 #define STRNCMP_UC_C8(str1, str2, num) \
2674 strncmp((char *)(str1), (str2), (num))
2675 #define STRLEN_UC(str) strlen((const char *)str)
2676
2677 #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2678
2679 extern int PRIV(strcmp_uc_uc)(const pcre_uchar *,
2680 const pcre_uchar *);
2681 extern int PRIV(strcmp_uc_c8)(const pcre_uchar *,
2682 const char *);
2683 extern int PRIV(strncmp_uc_uc)(const pcre_uchar *,
2684 const pcre_uchar *, unsigned int num);
2685 extern int PRIV(strncmp_uc_c8)(const pcre_uchar *,
2686 const char *, unsigned int num);
2687 extern unsigned int PRIV(strlen_uc)(const pcre_uchar *str);
2688
2689 #define STRCMP_UC_UC(str1, str2) \
2690 PRIV(strcmp_uc_uc)((str1), (str2))
2691 #define STRCMP_UC_C8(str1, str2) \
2692 PRIV(strcmp_uc_c8)((str1), (str2))
2693 #define STRNCMP_UC_UC(str1, str2, num) \
2694 PRIV(strncmp_uc_uc)((str1), (str2), (num))
2695 #define STRNCMP_UC_C8(str1, str2, num) \
2696 PRIV(strncmp_uc_c8)((str1), (str2), (num))
2697 #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2698
2699 #endif /* COMPILE_PCRE[8|16|32] */
2700
2701 #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2702
2703 #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2704 #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2705
2706 #elif defined COMPILE_PCRE32
2707
2708 extern int PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2709 const pcre_uchar *);
2710 extern int PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2711 const char *);
2712
2713 #define STRCMP_UC_UC_TEST(str1, str2) \
2714 (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2715 #define STRCMP_UC_C8_TEST(str1, str2) \
2716 (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2717
2718 #endif /* COMPILE_PCRE[8|16|32] */
2719
2720 extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2721 extern BOOL PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2722 int *, BOOL);
2723 extern unsigned int PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2724 extern int PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2725 extern BOOL PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2726 int *, BOOL);
2727 extern BOOL PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2728
2729 #ifdef SUPPORT_JIT
2730 extern void PRIV(jit_compile)(const REAL_PCRE *,
2731 PUBL(extra) *, int);
2732 extern int PRIV(jit_exec)(const PUBL(extra) *,
2733 const pcre_uchar *, int, int, int, int *, int);
2734 extern void PRIV(jit_free)(void *);
2735 extern int PRIV(jit_get_size)(void *);
2736 extern const char* PRIV(jit_get_target)(void);
2737 #endif
2738
2739 /* Unicode character database (UCD) */
2740
2741 typedef struct {
2742 pcre_uint8 script; /* ucp_Arabic, etc. */
2743 pcre_uint8 chartype; /* ucp_Cc, etc. (general categories) */
2744 pcre_uint8 gbprop; /* ucp_gbControl, etc. (grapheme break property) */
2745 pcre_uint8 caseset; /* offset to multichar other cases or zero */
2746 pcre_int32 other_case; /* offset to other case, or zero if none */
2747 } ucd_record;
2748
2749 extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2750 extern const ucd_record PRIV(ucd_records)[];
2751 extern const pcre_uint8 PRIV(ucd_stage1)[];
2752 extern const pcre_uint16 PRIV(ucd_stage2)[];
2753 extern const pcre_uint32 PRIV(ucp_gentype)[];
2754 extern const pcre_uint32 PRIV(ucp_gbtable)[];
2755 #ifdef COMPILE_PCRE32
2756 extern const ucd_record PRIV(dummy_ucd_record)[];
2757 #endif
2758 #ifdef SUPPORT_JIT
2759 extern const int PRIV(ucp_typerange)[];
2760 #endif
2761
2762 #ifdef SUPPORT_UCP
2763 /* UCD access macros */
2764
2765 #define UCD_BLOCK_SIZE 128
2766 #define REAL_GET_UCD(ch) (PRIV(ucd_records) + \
2767 PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2768 UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2769
2770 #ifdef COMPILE_PCRE32
2771 #define GET_UCD(ch) ((ch > 0x10ffff)? PRIV(dummy_ucd_record) : REAL_GET_UCD(ch))
2772 #else
2773 #define GET_UCD(ch) REAL_GET_UCD(ch)
2774 #endif
2775
2776 #define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype
2777 #define UCD_SCRIPT(ch) GET_UCD(ch)->script
2778 #define UCD_CATEGORY(ch) PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2779 #define UCD_GRAPHBREAK(ch) GET_UCD(ch)->gbprop
2780 #define UCD_CASESET(ch) GET_UCD(ch)->caseset
2781 #define UCD_OTHERCASE(ch) ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2782
2783 #endif /* SUPPORT_UCP */
2784
2785 #endif
2786
2787 /* End of pcre_internal.h */
2788