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

Contents of /code/trunk/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log


Revision 642 - (hide annotations) (download)
Thu Jul 28 18:59:40 2011 UTC (3 years, 4 months ago) by ph10
File MIME type: text/plain
File size: 83409 byte(s)
Avoid false positive for infinite recursion by not checking conditionals at 
compile time, but add tests at runtime that also catch infinite mutual 
recursion.

1 nigel 77 /*************************************************
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 ph10 642 Copyright (c) 1997-2011 University of Cambridge
11 nigel 77
12     -----------------------------------------------------------------------------
13     Redistribution and use in source and binary forms, with or without
14     modification, are permitted provided that the following conditions are met:
15    
16     * Redistributions of source code must retain the above copyright notice,
17     this list of conditions and the following disclaimer.
18    
19     * Redistributions in binary form must reproduce the above copyright
20     notice, this list of conditions and the following disclaimer in the
21     documentation and/or other materials provided with the distribution.
22    
23     * Neither the name of the University of Cambridge nor the names of its
24     contributors may be used to endorse or promote products derived from
25     this software without specific prior written permission.
26    
27     THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
28     AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
29     IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
30     ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
31     LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32     CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33     SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
34     INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
35     CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
36     ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
37     POSSIBILITY OF SUCH DAMAGE.
38     -----------------------------------------------------------------------------
39     */
40    
41     /* This header contains definitions that are shared between the different
42     modules, but which are not relevant to the exported API. This includes some
43     functions whose names all begin with "_pcre_". */
44    
45 nigel 85 #ifndef PCRE_INTERNAL_H
46     #define PCRE_INTERNAL_H
47 nigel 77
48 ph10 475 /* Define PCRE_DEBUG to get debugging output on stdout. */
49 nigel 77
50 nigel 85 #if 0
51 ph10 475 #define PCRE_DEBUG
52 nigel 85 #endif
53 nigel 77
54 ph10 392 /* We do not support both EBCDIC and UTF-8 at the same time. The "configure"
55 ph10 391 script prevents both being selected, but not everybody uses "configure". */
56    
57     #if defined EBCDIC && defined SUPPORT_UTF8
58     #error The use of both EBCDIC and SUPPORT_UTF8 is not supported.
59     #endif
60    
61 ph10 406 /* If SUPPORT_UCP is defined, SUPPORT_UTF8 must also be defined. The
62 ph10 393 "configure" script ensures this, but not everybody uses "configure". */
63    
64     #if defined SUPPORT_UCP && !defined SUPPORT_UTF8
65     #define SUPPORT_UTF8 1
66     #endif
67    
68 nigel 77 /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
69     inline, and there are *still* stupid compilers about that don't like indented
70     pre-processor statements, or at least there were when I first wrote this. After
71 nigel 93 all, it had only been about 10 years then...
72 nigel 77
73 nigel 93 It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so
74     be absolutely sure we get our version. */
75    
76     #undef DPRINTF
77 ph10 475 #ifdef PCRE_DEBUG
78 nigel 77 #define DPRINTF(p) printf p
79     #else
80 nigel 93 #define DPRINTF(p) /* Nothing */
81 nigel 77 #endif
82    
83    
84     /* Standard C headers plus the external interface definition. The only time
85     setjmp and stdarg are used is when NO_RECURSE is set. */
86    
87     #include <ctype.h>
88     #include <limits.h>
89     #include <stddef.h>
90     #include <stdio.h>
91     #include <stdlib.h>
92     #include <string.h>
93    
94 ph10 145 /* When compiling a DLL for Windows, the exported symbols have to be declared
95     using some MS magic. I found some useful information on this web page:
96     http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
97 ph10 154 information there, using __declspec(dllexport) without "extern" we have a
98 ph10 145 definition; with "extern" we have a declaration. The settings here override the
99     setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL,
100 ph10 154 which is all that is needed for applications (they just import the symbols). We
101     use:
102 ph10 145
103     PCRE_EXP_DECL for declarations
104     PCRE_EXP_DEFN for definitions of exported functions
105     PCRE_EXP_DATA_DEFN for definitions of exported variables
106    
107     The reason for the two DEFN macros is that in non-Windows environments, one
108     does not want to have "extern" before variable definitions because it leads to
109     compiler warnings. So we distinguish between functions and variables. In
110     Windows, the two should always be the same.
111    
112     The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest,
113     which is an application, but needs to import this file in order to "peek" at
114 ph10 154 internals, can #include pcre.h first to get an application's-eye view.
115 ph10 145
116     In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon,
117     special-purpose environments) might want to stick other stuff in front of
118     exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and
119     PCRE_EXP_DATA_DEFN only if they are not already set. */
120    
121     #ifndef PCRE_EXP_DECL
122     # ifdef _WIN32
123 ph10 199 # ifndef PCRE_STATIC
124 ph10 145 # define PCRE_EXP_DECL extern __declspec(dllexport)
125     # define PCRE_EXP_DEFN __declspec(dllexport)
126     # define PCRE_EXP_DATA_DEFN __declspec(dllexport)
127     # else
128     # define PCRE_EXP_DECL extern
129     # define PCRE_EXP_DEFN
130     # define PCRE_EXP_DATA_DEFN
131     # endif
132     # else
133     # ifdef __cplusplus
134     # define PCRE_EXP_DECL extern "C"
135     # else
136     # define PCRE_EXP_DECL extern
137     # endif
138     # ifndef PCRE_EXP_DEFN
139     # define PCRE_EXP_DEFN PCRE_EXP_DECL
140     # endif
141     # ifndef PCRE_EXP_DATA_DEFN
142     # define PCRE_EXP_DATA_DEFN
143     # endif
144     # endif
145 nigel 77 #endif
146    
147 ph10 359 /* When compiling with the MSVC compiler, it is sometimes necessary to include
148 ph10 371 a "calling convention" before exported function names. (This is secondhand
149 ph10 359 information; I know nothing about MSVC myself). For example, something like
150    
151     void __cdecl function(....)
152 ph10 371
153 ph10 359 might be needed. In order so make this easy, all the exported functions have
154     PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
155 ph10 371 set, we ensure here that it has no effect. */
156 ph10 359
157     #ifndef PCRE_CALL_CONVENTION
158     #define PCRE_CALL_CONVENTION
159     #endif
160    
161 nigel 77 /* We need to have types that specify unsigned 16-bit and 32-bit integers. We
162     cannot determine these outside the compilation (e.g. by running a program as
163     part of "configure") because PCRE is often cross-compiled for use on other
164     systems. Instead we make use of the maximum sizes that are available at
165     preprocessor time in standard C environments. */
166    
167     #if USHRT_MAX == 65535
168     typedef unsigned short pcre_uint16;
169 ph10 349 typedef short pcre_int16;
170 nigel 77 #elif UINT_MAX == 65535
171     typedef unsigned int pcre_uint16;
172 ph10 349 typedef int pcre_int16;
173 nigel 77 #else
174     #error Cannot determine a type for 16-bit unsigned integers
175     #endif
176    
177     #if UINT_MAX == 4294967295
178     typedef unsigned int pcre_uint32;
179 ph10 349 typedef int pcre_int32;
180 nigel 77 #elif ULONG_MAX == 4294967295
181     typedef unsigned long int pcre_uint32;
182 ph10 349 typedef long int pcre_int32;
183 nigel 77 #else
184     #error Cannot determine a type for 32-bit unsigned integers
185     #endif
186    
187 ph10 487 /* When checking for integer overflow in pcre_compile(), we need to handle
188     large integers. If a 64-bit integer type is available, we can use that.
189     Otherwise we have to cast to double, which of course requires floating point
190     arithmetic. Handle this by defining a macro for the appropriate type. If
191 ph10 494 stdint.h is available, include it; it may define INT64_MAX. Systems that do not
192     have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
193     by "configure". */
194 ph10 475
195     #if HAVE_STDINT_H
196     #include <stdint.h>
197 ph10 494 #elif HAVE_INTTYPES_H
198     #include <inttypes.h>
199 ph10 475 #endif
200    
201     #if defined INT64_MAX || defined int64_t
202     #define INT64_OR_DOUBLE int64_t
203     #else
204     #define INT64_OR_DOUBLE double
205     #endif
206    
207 nigel 77 /* All character handling must be done as unsigned characters. Otherwise there
208     are problems with top-bit-set characters and functions such as isspace().
209     However, we leave the interface to the outside world as char *, because that
210     should make things easier for callers. We define a short type for unsigned char
211     to save lots of typing. I tried "uchar", but it causes problems on Digital
212     Unix, where it is defined in sys/types, so use "uschar" instead. */
213    
214     typedef unsigned char uschar;
215    
216 nigel 93 /* This is an unsigned int value that no character can ever have. UTF-8
217     characters only go up to 0x7fffffff (though Unicode doesn't go beyond
218     0x0010ffff). */
219 nigel 91
220 nigel 93 #define NOTACHAR 0xffffffff
221    
222     /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
223 ph10 149 "any" and "anycrlf" at present). The following macros are used to package up
224     testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
225     modules to indicate in which datablock the parameters exist, and what the
226     start/end of string field names are. */
227 nigel 93
228 ph10 149 #define NLTYPE_FIXED 0 /* Newline is a fixed length string */
229     #define NLTYPE_ANY 1 /* Newline is any Unicode line ending */
230     #define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */
231 nigel 93
232     /* This macro checks for a newline at the given position */
233    
234 nigel 91 #define IS_NEWLINE(p) \
235 nigel 93 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
236     ((p) < NLBLOCK->PSEND && \
237 ph10 149 _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\
238     utf8)) \
239 nigel 93 : \
240     ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
241     (p)[0] == NLBLOCK->nl[0] && \
242     (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \
243     ) \
244     )
245 nigel 91
246 nigel 93 /* This macro checks for a newline immediately preceding the given position */
247    
248     #define WAS_NEWLINE(p) \
249     ((NLBLOCK->nltype != NLTYPE_FIXED)? \
250     ((p) > NLBLOCK->PSSTART && \
251 ph10 149 _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
252     &(NLBLOCK->nllen), utf8)) \
253 nigel 93 : \
254     ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
255     (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \
256     (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \
257     ) \
258     )
259    
260 nigel 87 /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
261     with a custom type. This makes it possible, for example, to allow pcre_exec()
262     to process subject strings that are discontinuous by using a smart pointer
263     class. It must always be possible to inspect all of the subject string in
264     pcre_exec() because of the way it backtracks. Two macros are required in the
265     normal case, for sign-unspecified and unsigned char pointers. The former is
266     used for the external interface and appears in pcre.h, which is why its name
267     must begin with PCRE_. */
268    
269     #ifdef CUSTOM_SUBJECT_PTR
270     #define PCRE_SPTR CUSTOM_SUBJECT_PTR
271     #define USPTR CUSTOM_SUBJECT_PTR
272     #else
273     #define PCRE_SPTR const char *
274     #define USPTR const unsigned char *
275     #endif
276    
277 ph10 145
278    
279 nigel 85 /* Include the public PCRE header and the definitions of UCP character property
280     values. */
281 nigel 77
282 ph10 189 #include "pcre.h"
283 nigel 77 #include "ucp.h"
284    
285     /* When compiling for use with the Virtual Pascal compiler, these functions
286     need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
287     option on the command line. */
288    
289     #ifdef VPCOMPAT
290 ph10 101 #define strlen(s) _strlen(s)
291 nigel 77 #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
292 ph10 101 #define memcmp(s,c,n) _memcmp(s,c,n)
293 nigel 77 #define memcpy(d,s,n) _memcpy(d,s,n)
294     #define memmove(d,s,n) _memmove(d,s,n)
295     #define memset(s,c,n) _memset(s,c,n)
296     #else /* VPCOMPAT */
297    
298     /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
299     define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
300     is set. Otherwise, include an emulating function for those systems that have
301 ph10 118 neither (there some non-Unix environments where this is the case). */
302 nigel 77
303 ph10 137 #ifndef HAVE_MEMMOVE
304 nigel 77 #undef memmove /* some systems may have a macro */
305 ph10 137 #ifdef HAVE_BCOPY
306 nigel 77 #define memmove(a, b, c) bcopy(b, a, c)
307     #else /* HAVE_BCOPY */
308 nigel 91 static void *
309 ph10 108 pcre_memmove(void *d, const void *s, size_t n)
310 nigel 77 {
311 nigel 87 size_t i;
312 ph10 108 unsigned char *dest = (unsigned char *)d;
313     const unsigned char *src = (const unsigned char *)s;
314 ph10 118 if (dest > src)
315     {
316     dest += n;
317     src += n;
318     for (i = 0; i < n; ++i) *(--dest) = *(--src);
319     return (void *)dest;
320     }
321     else
322     {
323     for (i = 0; i < n; ++i) *dest++ = *src++;
324 ph10 123 return (void *)(dest - n);
325     }
326 nigel 77 }
327     #define memmove(a, b, c) pcre_memmove(a, b, c)
328     #endif /* not HAVE_BCOPY */
329     #endif /* not HAVE_MEMMOVE */
330     #endif /* not VPCOMPAT */
331    
332    
333     /* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
334     in big-endian order) by default. These are used, for example, to link from the
335     start of a subpattern to its alternatives and its end. The use of 2 bytes per
336     offset limits the size of the compiled regex to around 64K, which is big enough
337     for almost everybody. However, I received a request for an even bigger limit.
338     For this reason, and also to make the code easier to maintain, the storing and
339     loading of offsets from the byte string is now handled by the macros that are
340     defined here.
341    
342     The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
343     the config.h file, but can be overridden by using -D on the command line. This
344     is automated on Unix systems via the "configure" command. */
345    
346     #if LINK_SIZE == 2
347    
348     #define PUT(a,n,d) \
349     (a[n] = (d) >> 8), \
350     (a[(n)+1] = (d) & 255)
351    
352     #define GET(a,n) \
353     (((a)[n] << 8) | (a)[(n)+1])
354    
355     #define MAX_PATTERN_SIZE (1 << 16)
356    
357    
358     #elif LINK_SIZE == 3
359    
360     #define PUT(a,n,d) \
361     (a[n] = (d) >> 16), \
362     (a[(n)+1] = (d) >> 8), \
363     (a[(n)+2] = (d) & 255)
364    
365     #define GET(a,n) \
366     (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
367    
368     #define MAX_PATTERN_SIZE (1 << 24)
369    
370    
371     #elif LINK_SIZE == 4
372    
373     #define PUT(a,n,d) \
374     (a[n] = (d) >> 24), \
375     (a[(n)+1] = (d) >> 16), \
376     (a[(n)+2] = (d) >> 8), \
377     (a[(n)+3] = (d) & 255)
378    
379     #define GET(a,n) \
380     (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
381    
382     #define MAX_PATTERN_SIZE (1 << 30) /* Keep it positive */
383    
384    
385     #else
386     #error LINK_SIZE must be either 2, 3, or 4
387     #endif
388    
389    
390     /* Convenience macro defined in terms of the others */
391    
392     #define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
393    
394    
395     /* PCRE uses some other 2-byte quantities that do not change when the size of
396     offsets changes. There are used for repeat counts and for other things such as
397     capturing parenthesis numbers in back references. */
398    
399     #define PUT2(a,n,d) \
400     a[n] = (d) >> 8; \
401     a[(n)+1] = (d) & 255
402    
403     #define GET2(a,n) \
404     (((a)[n] << 8) | (a)[(n)+1])
405    
406     #define PUT2INC(a,n,d) PUT2(a,n,d), a += 2
407    
408    
409     /* When UTF-8 encoding is being used, a character is no longer just a single
410     byte. The macros for character handling generate simple sequences when used in
411 ph10 573 byte-mode, and more complicated ones for UTF-8 characters. GETCHARLENTEST is
412     not used when UTF-8 is not supported, so it is not defined, and BACKCHAR should
413     never be called in byte mode. To make sure they can never even appear when
414     UTF-8 support is omitted, we don't even define them. */
415 nigel 77
416     #ifndef SUPPORT_UTF8
417     #define GETCHAR(c, eptr) c = *eptr;
418     #define GETCHARTEST(c, eptr) c = *eptr;
419     #define GETCHARINC(c, eptr) c = *eptr++;
420     #define GETCHARINCTEST(c, eptr) c = *eptr++;
421     #define GETCHARLEN(c, eptr, len) c = *eptr;
422 ph10 573 /* #define GETCHARLENTEST(c, eptr, len) */
423 ph10 207 /* #define BACKCHAR(eptr) */
424 nigel 77
425     #else /* SUPPORT_UTF8 */
426    
427 ph10 573 /* These macros were originally written in the form of loops that used data
428     from the tables whose names start with _pcre_utf8_table. They were rewritten by
429     a user so as not to use loops, because in some environments this gives a
430     significant performance advantage, and it seems never to do any harm. */
431    
432     /* Base macro to pick up the remaining bytes of a UTF-8 character, not
433     advancing the pointer. */
434    
435     #define GETUTF8(c, eptr) \
436     { \
437     if ((c & 0x20) == 0) \
438     c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
439     else if ((c & 0x10) == 0) \
440     c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
441     else if ((c & 0x08) == 0) \
442     c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
443     ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
444     else if ((c & 0x04) == 0) \
445     c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
446     ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
447     (eptr[4] & 0x3f); \
448     else \
449     c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
450     ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
451     ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
452     }
453    
454 nigel 77 /* Get the next UTF-8 character, not advancing the pointer. This is called when
455     we know we are in UTF-8 mode. */
456    
457     #define GETCHAR(c, eptr) \
458     c = *eptr; \
459 ph10 573 if (c >= 0xc0) GETUTF8(c, eptr);
460 nigel 77
461     /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
462     pointer. */
463    
464     #define GETCHARTEST(c, eptr) \
465     c = *eptr; \
466 ph10 573 if (utf8 && c >= 0xc0) GETUTF8(c, eptr);
467    
468     /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
469     the pointer. */
470    
471     #define GETUTF8INC(c, eptr) \
472 nigel 77 { \
473 ph10 573 if ((c & 0x20) == 0) \
474     c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
475     else if ((c & 0x10) == 0) \
476 nigel 77 { \
477 ph10 573 c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
478     eptr += 2; \
479 nigel 77 } \
480 ph10 573 else if ((c & 0x08) == 0) \
481     { \
482     c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
483     ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
484     eptr += 3; \
485     } \
486     else if ((c & 0x04) == 0) \
487     { \
488     c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
489     ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
490     (eptr[3] & 0x3f); \
491     eptr += 4; \
492     } \
493     else \
494     { \
495     c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
496     ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
497     ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
498     eptr += 5; \
499     } \
500 nigel 77 }
501    
502     /* Get the next UTF-8 character, advancing the pointer. This is called when we
503     know we are in UTF-8 mode. */
504    
505     #define GETCHARINC(c, eptr) \
506     c = *eptr++; \
507 ph10 573 if (c >= 0xc0) GETUTF8INC(c, eptr);
508 nigel 77
509 ph10 529 /* Get the next character, testing for UTF-8 mode, and advancing the pointer.
510     This is called when we don't know if we are in UTF-8 mode. */
511 nigel 77
512     #define GETCHARINCTEST(c, eptr) \
513     c = *eptr++; \
514 ph10 573 if (utf8 && c >= 0xc0) GETUTF8INC(c, eptr);
515    
516     /* Base macro to pick up the remaining bytes of a UTF-8 character, not
517     advancing the pointer, incrementing the length. */
518    
519     #define GETUTF8LEN(c, eptr, len) \
520 nigel 77 { \
521 ph10 573 if ((c & 0x20) == 0) \
522 nigel 77 { \
523 ph10 573 c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
524     len++; \
525 nigel 77 } \
526 ph10 573 else if ((c & 0x10) == 0) \
527     { \
528     c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
529     len += 2; \
530     } \
531     else if ((c & 0x08) == 0) \
532     {\
533     c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
534     ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
535     len += 3; \
536     } \
537     else if ((c & 0x04) == 0) \
538     { \
539     c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
540     ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
541     (eptr[4] & 0x3f); \
542     len += 4; \
543     } \
544     else \
545     {\
546     c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
547     ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
548     ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
549     len += 5; \
550     } \
551 nigel 77 }
552    
553     /* Get the next UTF-8 character, not advancing the pointer, incrementing length
554     if there are extra bytes. This is called when we know we are in UTF-8 mode. */
555    
556     #define GETCHARLEN(c, eptr, len) \
557     c = *eptr; \
558 ph10 573 if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
559 nigel 77
560 ph10 384 /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
561     pointer, incrementing length if there are extra bytes. This is called when we
562 ph10 529 do not know if we are in UTF-8 mode. */
563 ph10 384
564     #define GETCHARLENTEST(c, eptr, len) \
565     c = *eptr; \
566 ph10 573 if (utf8 && c >= 0xc0) GETUTF8LEN(c, eptr, len);
567 ph10 384
568 nigel 77 /* If the pointer is not at the start of a character, move it back until
569 ph10 208 it is. This is called only in UTF-8 mode - we don't put a test within the macro
570 ph10 207 because almost all calls are already within a block of UTF-8 only code. */
571 nigel 77
572 ph10 207 #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
573 nigel 77
574 ph10 573 #endif /* SUPPORT_UTF8 */
575 nigel 77
576    
577     /* In case there is no definition of offsetof() provided - though any proper
578     Standard C system should have one. */
579    
580     #ifndef offsetof
581     #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))
582     #endif
583    
584    
585 ph10 231 /* Private flags containing information about the compiled regex. They used to
586     live at the top end of the options word, but that got almost full, so now they
587 ph10 426 are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
588 ph10 427 the restrictions on partial matching have been lifted. It remains for backwards
589 ph10 426 compatibility. */
590 nigel 77
591 ph10 230 #define PCRE_NOPARTIAL 0x0001 /* can't use partial with this regex */
592     #define PCRE_FIRSTSET 0x0002 /* first_byte is set */
593     #define PCRE_REQCHSET 0x0004 /* req_byte is set */
594     #define PCRE_STARTLINE 0x0008 /* start after \n for multiline */
595     #define PCRE_JCHANGED 0x0010 /* j option used in regex */
596     #define PCRE_HASCRORLF 0x0020 /* explicit \r or \n in pattern */
597 nigel 77
598 ph10 612 /* Flags for the "extra" block produced by pcre_study(). */
599 nigel 77
600 ph10 612 #define PCRE_STUDY_MAPPED 0x0001 /* a map of starting chars exists */
601     #define PCRE_STUDY_MINLEN 0x0002 /* a minimum length field exists */
602 nigel 77
603     /* Masks for identifying the public options that are permitted at compile
604     time, run time, or study time, respectively. */
605    
606 ph10 149 #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
607     PCRE_NEWLINE_ANYCRLF)
608 nigel 93
609 ph10 389 #define PUBLIC_COMPILE_OPTIONS \
610 nigel 77 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
611     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
612 nigel 91 PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
613 ph10 336 PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
614 ph10 576 PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
615 nigel 77
616     #define PUBLIC_EXEC_OPTIONS \
617 ph10 442 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
618     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
619     PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
620 nigel 77
621     #define PUBLIC_DFA_EXEC_OPTIONS \
622 ph10 442 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
623     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
624     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
625     PCRE_NO_START_OPTIMIZE)
626 nigel 77
627     #define PUBLIC_STUDY_OPTIONS 0 /* None defined */
628    
629     /* Magic number to provide a small check against being handed junk. Also used
630     to detect whether a pattern was compiled on a host of different endianness. */
631    
632     #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
633    
634     /* Negative values for the firstchar and reqchar variables */
635    
636     #define REQ_UNSET (-2)
637     #define REQ_NONE (-1)
638    
639     /* The maximum remaining length of subject we are prepared to search for a
640     req_byte match. */
641    
642     #define REQ_BYTE_MAX 1000
643    
644     /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a
645     variable-length repeat, or a anything other than literal characters. */
646    
647     #define REQ_CASELESS 0x0100 /* indicates caselessness */
648     #define REQ_VARY 0x0200 /* reqbyte followed non-literal item */
649    
650 ph10 369 /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
651 ph10 376 environments where these macros are defined elsewhere. Unfortunately, there
652     is no way to do the same for the typedef. */
653 nigel 77
654     typedef int BOOL;
655    
656 ph10 376 #ifndef FALSE
657 nigel 77 #define FALSE 0
658     #define TRUE 1
659 ph10 369 #endif
660 nigel 77
661 ph10 391 /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
662     character constants like '*' because the compiler would emit their EBCDIC code,
663 ph10 392 which is different from their ASCII/UTF-8 code. Instead we define macros for
664     the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
665     is enabled. When UTF-8 support is not enabled, the definitions use character
666     literals. Both character and string versions of each character are needed, and
667 ph10 391 there are some longer strings as well.
668    
669 ph10 392 This means that, on EBCDIC platforms, the PCRE library can handle either
670 ph10 391 EBCDIC, or UTF-8, but not both. To support both in the same compiled library
671     would need different lookups depending on whether PCRE_UTF8 was set or not.
672     This would make it impossible to use characters in switch/case statements,
673 ph10 392 which would reduce performance. For a theoretical use (which nobody has asked
674     for) in a minority area (EBCDIC platforms), this is not sensible. Any
675     application that did need both could compile two versions of the library, using
676 ph10 391 macros to give the functions distinct names. */
677    
678     #ifndef SUPPORT_UTF8
679    
680     /* UTF-8 support is not enabled; use the platform-dependent character literals
681     so that PCRE works on both ASCII and EBCDIC platforms, in non-UTF-mode only. */
682    
683     #define CHAR_HT '\t'
684     #define CHAR_VT '\v'
685     #define CHAR_FF '\f'
686     #define CHAR_CR '\r'
687     #define CHAR_NL '\n'
688     #define CHAR_BS '\b'
689     #define CHAR_BEL '\a'
690     #ifdef EBCDIC
691     #define CHAR_ESC '\047'
692     #define CHAR_DEL '\007'
693     #else
694     #define CHAR_ESC '\033'
695     #define CHAR_DEL '\177'
696     #endif
697    
698     #define CHAR_SPACE ' '
699     #define CHAR_EXCLAMATION_MARK '!'
700     #define CHAR_QUOTATION_MARK '"'
701     #define CHAR_NUMBER_SIGN '#'
702     #define CHAR_DOLLAR_SIGN '$'
703     #define CHAR_PERCENT_SIGN '%'
704     #define CHAR_AMPERSAND '&'
705     #define CHAR_APOSTROPHE '\''
706     #define CHAR_LEFT_PARENTHESIS '('
707     #define CHAR_RIGHT_PARENTHESIS ')'
708     #define CHAR_ASTERISK '*'
709     #define CHAR_PLUS '+'
710     #define CHAR_COMMA ','
711     #define CHAR_MINUS '-'
712     #define CHAR_DOT '.'
713     #define CHAR_SLASH '/'
714     #define CHAR_0 '0'
715     #define CHAR_1 '1'
716     #define CHAR_2 '2'
717     #define CHAR_3 '3'
718     #define CHAR_4 '4'
719     #define CHAR_5 '5'
720     #define CHAR_6 '6'
721     #define CHAR_7 '7'
722     #define CHAR_8 '8'
723     #define CHAR_9 '9'
724     #define CHAR_COLON ':'
725     #define CHAR_SEMICOLON ';'
726     #define CHAR_LESS_THAN_SIGN '<'
727     #define CHAR_EQUALS_SIGN '='
728     #define CHAR_GREATER_THAN_SIGN '>'
729     #define CHAR_QUESTION_MARK '?'
730     #define CHAR_COMMERCIAL_AT '@'
731     #define CHAR_A 'A'
732     #define CHAR_B 'B'
733     #define CHAR_C 'C'
734     #define CHAR_D 'D'
735     #define CHAR_E 'E'
736     #define CHAR_F 'F'
737     #define CHAR_G 'G'
738     #define CHAR_H 'H'
739     #define CHAR_I 'I'
740     #define CHAR_J 'J'
741     #define CHAR_K 'K'
742     #define CHAR_L 'L'
743     #define CHAR_M 'M'
744     #define CHAR_N 'N'
745     #define CHAR_O 'O'
746     #define CHAR_P 'P'
747     #define CHAR_Q 'Q'
748     #define CHAR_R 'R'
749     #define CHAR_S 'S'
750     #define CHAR_T 'T'
751     #define CHAR_U 'U'
752     #define CHAR_V 'V'
753     #define CHAR_W 'W'
754     #define CHAR_X 'X'
755     #define CHAR_Y 'Y'
756     #define CHAR_Z 'Z'
757     #define CHAR_LEFT_SQUARE_BRACKET '['
758     #define CHAR_BACKSLASH '\\'
759     #define CHAR_RIGHT_SQUARE_BRACKET ']'
760     #define CHAR_CIRCUMFLEX_ACCENT '^'
761     #define CHAR_UNDERSCORE '_'
762     #define CHAR_GRAVE_ACCENT '`'
763     #define CHAR_a 'a'
764     #define CHAR_b 'b'
765     #define CHAR_c 'c'
766     #define CHAR_d 'd'
767     #define CHAR_e 'e'
768     #define CHAR_f 'f'
769     #define CHAR_g 'g'
770     #define CHAR_h 'h'
771     #define CHAR_i 'i'
772     #define CHAR_j 'j'
773     #define CHAR_k 'k'
774     #define CHAR_l 'l'
775     #define CHAR_m 'm'
776     #define CHAR_n 'n'
777     #define CHAR_o 'o'
778     #define CHAR_p 'p'
779     #define CHAR_q 'q'
780     #define CHAR_r 'r'
781     #define CHAR_s 's'
782     #define CHAR_t 't'
783     #define CHAR_u 'u'
784     #define CHAR_v 'v'
785     #define CHAR_w 'w'
786     #define CHAR_x 'x'
787     #define CHAR_y 'y'
788     #define CHAR_z 'z'
789     #define CHAR_LEFT_CURLY_BRACKET '{'
790     #define CHAR_VERTICAL_LINE '|'
791     #define CHAR_RIGHT_CURLY_BRACKET '}'
792     #define CHAR_TILDE '~'
793    
794     #define STR_HT "\t"
795     #define STR_VT "\v"
796     #define STR_FF "\f"
797     #define STR_CR "\r"
798     #define STR_NL "\n"
799     #define STR_BS "\b"
800     #define STR_BEL "\a"
801     #ifdef EBCDIC
802     #define STR_ESC "\047"
803     #define STR_DEL "\007"
804     #else
805     #define STR_ESC "\033"
806     #define STR_DEL "\177"
807     #endif
808    
809     #define STR_SPACE " "
810     #define STR_EXCLAMATION_MARK "!"
811     #define STR_QUOTATION_MARK "\""
812     #define STR_NUMBER_SIGN "#"
813     #define STR_DOLLAR_SIGN "$"
814     #define STR_PERCENT_SIGN "%"
815     #define STR_AMPERSAND "&"
816     #define STR_APOSTROPHE "'"
817     #define STR_LEFT_PARENTHESIS "("
818     #define STR_RIGHT_PARENTHESIS ")"
819     #define STR_ASTERISK "*"
820     #define STR_PLUS "+"
821     #define STR_COMMA ","
822     #define STR_MINUS "-"
823     #define STR_DOT "."
824     #define STR_SLASH "/"
825     #define STR_0 "0"
826     #define STR_1 "1"
827     #define STR_2 "2"
828     #define STR_3 "3"
829     #define STR_4 "4"
830     #define STR_5 "5"
831     #define STR_6 "6"
832     #define STR_7 "7"
833     #define STR_8 "8"
834     #define STR_9 "9"
835     #define STR_COLON ":"
836     #define STR_SEMICOLON ";"
837     #define STR_LESS_THAN_SIGN "<"
838     #define STR_EQUALS_SIGN "="
839     #define STR_GREATER_THAN_SIGN ">"
840     #define STR_QUESTION_MARK "?"
841     #define STR_COMMERCIAL_AT "@"
842     #define STR_A "A"
843     #define STR_B "B"
844     #define STR_C "C"
845     #define STR_D "D"
846     #define STR_E "E"
847     #define STR_F "F"
848     #define STR_G "G"
849     #define STR_H "H"
850     #define STR_I "I"
851     #define STR_J "J"
852     #define STR_K "K"
853     #define STR_L "L"
854     #define STR_M "M"
855     #define STR_N "N"
856     #define STR_O "O"
857     #define STR_P "P"
858     #define STR_Q "Q"
859     #define STR_R "R"
860     #define STR_S "S"
861     #define STR_T "T"
862     #define STR_U "U"
863     #define STR_V "V"
864     #define STR_W "W"
865     #define STR_X "X"
866     #define STR_Y "Y"
867     #define STR_Z "Z"
868     #define STR_LEFT_SQUARE_BRACKET "["
869     #define STR_BACKSLASH "\\"
870     #define STR_RIGHT_SQUARE_BRACKET "]"
871     #define STR_CIRCUMFLEX_ACCENT "^"
872     #define STR_UNDERSCORE "_"
873     #define STR_GRAVE_ACCENT "`"
874     #define STR_a "a"
875     #define STR_b "b"
876     #define STR_c "c"
877     #define STR_d "d"
878     #define STR_e "e"
879     #define STR_f "f"
880     #define STR_g "g"
881     #define STR_h "h"
882     #define STR_i "i"
883     #define STR_j "j"
884     #define STR_k "k"
885     #define STR_l "l"
886     #define STR_m "m"
887     #define STR_n "n"
888     #define STR_o "o"
889     #define STR_p "p"
890     #define STR_q "q"
891     #define STR_r "r"
892     #define STR_s "s"
893     #define STR_t "t"
894     #define STR_u "u"
895     #define STR_v "v"
896     #define STR_w "w"
897     #define STR_x "x"
898     #define STR_y "y"
899     #define STR_z "z"
900     #define STR_LEFT_CURLY_BRACKET "{"
901     #define STR_VERTICAL_LINE "|"
902     #define STR_RIGHT_CURLY_BRACKET "}"
903     #define STR_TILDE "~"
904    
905     #define STRING_ACCEPT0 "ACCEPT\0"
906     #define STRING_COMMIT0 "COMMIT\0"
907     #define STRING_F0 "F\0"
908     #define STRING_FAIL0 "FAIL\0"
909 ph10 510 #define STRING_MARK0 "MARK\0"
910 ph10 391 #define STRING_PRUNE0 "PRUNE\0"
911     #define STRING_SKIP0 "SKIP\0"
912     #define STRING_THEN "THEN"
913    
914     #define STRING_alpha0 "alpha\0"
915     #define STRING_lower0 "lower\0"
916     #define STRING_upper0 "upper\0"
917     #define STRING_alnum0 "alnum\0"
918     #define STRING_ascii0 "ascii\0"
919     #define STRING_blank0 "blank\0"
920     #define STRING_cntrl0 "cntrl\0"
921     #define STRING_digit0 "digit\0"
922     #define STRING_graph0 "graph\0"
923     #define STRING_print0 "print\0"
924     #define STRING_punct0 "punct\0"
925     #define STRING_space0 "space\0"
926     #define STRING_word0 "word\0"
927     #define STRING_xdigit "xdigit"
928    
929     #define STRING_DEFINE "DEFINE"
930    
931 ph10 576 #define STRING_CR_RIGHTPAR "CR)"
932     #define STRING_LF_RIGHTPAR "LF)"
933     #define STRING_CRLF_RIGHTPAR "CRLF)"
934     #define STRING_ANY_RIGHTPAR "ANY)"
935     #define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)"
936     #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"
937     #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"
938     #define STRING_UTF8_RIGHTPAR "UTF8)"
939     #define STRING_UCP_RIGHTPAR "UCP)"
940     #define STRING_NO_START_OPT_RIGHTPAR "NO_START_OPT)"
941 ph10 391
942     #else /* SUPPORT_UTF8 */
943    
944     /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
945     works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
946     only. */
947    
948     #define CHAR_HT '\011'
949     #define CHAR_VT '\013'
950     #define CHAR_FF '\014'
951     #define CHAR_CR '\015'
952     #define CHAR_NL '\012'
953     #define CHAR_BS '\010'
954     #define CHAR_BEL '\007'
955     #define CHAR_ESC '\033'
956     #define CHAR_DEL '\177'
957    
958     #define CHAR_SPACE '\040'
959     #define CHAR_EXCLAMATION_MARK '\041'
960     #define CHAR_QUOTATION_MARK '\042'
961     #define CHAR_NUMBER_SIGN '\043'
962     #define CHAR_DOLLAR_SIGN '\044'
963     #define CHAR_PERCENT_SIGN '\045'
964     #define CHAR_AMPERSAND '\046'
965     #define CHAR_APOSTROPHE '\047'
966     #define CHAR_LEFT_PARENTHESIS '\050'
967     #define CHAR_RIGHT_PARENTHESIS '\051'
968     #define CHAR_ASTERISK '\052'
969     #define CHAR_PLUS '\053'
970     #define CHAR_COMMA '\054'
971     #define CHAR_MINUS '\055'
972     #define CHAR_DOT '\056'
973     #define CHAR_SLASH '\057'
974     #define CHAR_0 '\060'
975     #define CHAR_1 '\061'
976     #define CHAR_2 '\062'
977     #define CHAR_3 '\063'
978     #define CHAR_4 '\064'
979     #define CHAR_5 '\065'
980     #define CHAR_6 '\066'
981     #define CHAR_7 '\067'
982     #define CHAR_8 '\070'
983     #define CHAR_9 '\071'
984     #define CHAR_COLON '\072'
985     #define CHAR_SEMICOLON '\073'
986     #define CHAR_LESS_THAN_SIGN '\074'
987     #define CHAR_EQUALS_SIGN '\075'
988     #define CHAR_GREATER_THAN_SIGN '\076'
989     #define CHAR_QUESTION_MARK '\077'
990     #define CHAR_COMMERCIAL_AT '\100'
991     #define CHAR_A '\101'
992     #define CHAR_B '\102'
993     #define CHAR_C '\103'
994     #define CHAR_D '\104'
995     #define CHAR_E '\105'
996     #define CHAR_F '\106'
997     #define CHAR_G '\107'
998     #define CHAR_H '\110'
999     #define CHAR_I '\111'
1000     #define CHAR_J '\112'
1001     #define CHAR_K '\113'
1002     #define CHAR_L '\114'
1003     #define CHAR_M '\115'
1004     #define CHAR_N '\116'
1005     #define CHAR_O '\117'
1006     #define CHAR_P '\120'
1007     #define CHAR_Q '\121'
1008     #define CHAR_R '\122'
1009     #define CHAR_S '\123'
1010     #define CHAR_T '\124'
1011     #define CHAR_U '\125'
1012     #define CHAR_V '\126'
1013     #define CHAR_W '\127'
1014     #define CHAR_X '\130'
1015     #define CHAR_Y '\131'
1016     #define CHAR_Z '\132'
1017     #define CHAR_LEFT_SQUARE_BRACKET '\133'
1018     #define CHAR_BACKSLASH '\134'
1019     #define CHAR_RIGHT_SQUARE_BRACKET '\135'
1020     #define CHAR_CIRCUMFLEX_ACCENT '\136'
1021     #define CHAR_UNDERSCORE '\137'
1022     #define CHAR_GRAVE_ACCENT '\140'
1023     #define CHAR_a '\141'
1024     #define CHAR_b '\142'
1025     #define CHAR_c '\143'
1026     #define CHAR_d '\144'
1027     #define CHAR_e '\145'
1028     #define CHAR_f '\146'
1029     #define CHAR_g '\147'
1030     #define CHAR_h '\150'
1031     #define CHAR_i '\151'
1032     #define CHAR_j '\152'
1033     #define CHAR_k '\153'
1034     #define CHAR_l '\154'
1035     #define CHAR_m '\155'
1036     #define CHAR_n '\156'
1037     #define CHAR_o '\157'
1038     #define CHAR_p '\160'
1039     #define CHAR_q '\161'
1040     #define CHAR_r '\162'
1041     #define CHAR_s '\163'
1042     #define CHAR_t '\164'
1043     #define CHAR_u '\165'
1044     #define CHAR_v '\166'
1045     #define CHAR_w '\167'
1046     #define CHAR_x '\170'
1047     #define CHAR_y '\171'
1048     #define CHAR_z '\172'
1049     #define CHAR_LEFT_CURLY_BRACKET '\173'
1050     #define CHAR_VERTICAL_LINE '\174'
1051     #define CHAR_RIGHT_CURLY_BRACKET '\175'
1052     #define CHAR_TILDE '\176'
1053    
1054     #define STR_HT "\011"
1055     #define STR_VT "\013"
1056     #define STR_FF "\014"
1057     #define STR_CR "\015"
1058     #define STR_NL "\012"
1059     #define STR_BS "\010"
1060     #define STR_BEL "\007"
1061     #define STR_ESC "\033"
1062     #define STR_DEL "\177"
1063    
1064     #define STR_SPACE "\040"
1065     #define STR_EXCLAMATION_MARK "\041"
1066     #define STR_QUOTATION_MARK "\042"
1067     #define STR_NUMBER_SIGN "\043"
1068     #define STR_DOLLAR_SIGN "\044"
1069     #define STR_PERCENT_SIGN "\045"
1070     #define STR_AMPERSAND "\046"
1071     #define STR_APOSTROPHE "\047"
1072     #define STR_LEFT_PARENTHESIS "\050"
1073     #define STR_RIGHT_PARENTHESIS "\051"
1074     #define STR_ASTERISK "\052"
1075     #define STR_PLUS "\053"
1076     #define STR_COMMA "\054"
1077     #define STR_MINUS "\055"
1078     #define STR_DOT "\056"
1079     #define STR_SLASH "\057"
1080     #define STR_0 "\060"
1081     #define STR_1 "\061"
1082     #define STR_2 "\062"
1083     #define STR_3 "\063"
1084     #define STR_4 "\064"
1085     #define STR_5 "\065"
1086     #define STR_6 "\066"
1087     #define STR_7 "\067"
1088     #define STR_8 "\070"
1089     #define STR_9 "\071"
1090     #define STR_COLON "\072"
1091     #define STR_SEMICOLON "\073"
1092     #define STR_LESS_THAN_SIGN "\074"
1093     #define STR_EQUALS_SIGN "\075"
1094     #define STR_GREATER_THAN_SIGN "\076"
1095     #define STR_QUESTION_MARK "\077"
1096     #define STR_COMMERCIAL_AT "\100"
1097     #define STR_A "\101"
1098     #define STR_B "\102"
1099     #define STR_C "\103"
1100     #define STR_D "\104"
1101     #define STR_E "\105"
1102     #define STR_F "\106"
1103     #define STR_G "\107"
1104     #define STR_H "\110"
1105     #define STR_I "\111"
1106     #define STR_J "\112"
1107     #define STR_K "\113"
1108     #define STR_L "\114"
1109     #define STR_M "\115"
1110     #define STR_N "\116"
1111     #define STR_O "\117"
1112     #define STR_P "\120"
1113     #define STR_Q "\121"
1114     #define STR_R "\122"
1115     #define STR_S "\123"
1116     #define STR_T "\124"
1117     #define STR_U "\125"
1118     #define STR_V "\126"
1119     #define STR_W "\127"
1120     #define STR_X "\130"
1121     #define STR_Y "\131"
1122     #define STR_Z "\132"
1123     #define STR_LEFT_SQUARE_BRACKET "\133"
1124     #define STR_BACKSLASH "\134"
1125     #define STR_RIGHT_SQUARE_BRACKET "\135"
1126     #define STR_CIRCUMFLEX_ACCENT "\136"
1127     #define STR_UNDERSCORE "\137"
1128     #define STR_GRAVE_ACCENT "\140"
1129     #define STR_a "\141"
1130     #define STR_b "\142"
1131     #define STR_c "\143"
1132     #define STR_d "\144"
1133     #define STR_e "\145"
1134     #define STR_f "\146"
1135     #define STR_g "\147"
1136     #define STR_h "\150"
1137     #define STR_i "\151"
1138     #define STR_j "\152"
1139     #define STR_k "\153"
1140     #define STR_l "\154"
1141     #define STR_m "\155"
1142     #define STR_n "\156"
1143     #define STR_o "\157"
1144     #define STR_p "\160"
1145     #define STR_q "\161"
1146     #define STR_r "\162"
1147     #define STR_s "\163"
1148     #define STR_t "\164"
1149     #define STR_u "\165"
1150     #define STR_v "\166"
1151     #define STR_w "\167"
1152     #define STR_x "\170"
1153     #define STR_y "\171"
1154     #define STR_z "\172"
1155     #define STR_LEFT_CURLY_BRACKET "\173"
1156     #define STR_VERTICAL_LINE "\174"
1157     #define STR_RIGHT_CURLY_BRACKET "\175"
1158     #define STR_TILDE "\176"
1159    
1160     #define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1161     #define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1162     #define STRING_F0 STR_F "\0"
1163     #define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0"
1164 ph10 510 #define STRING_MARK0 STR_M STR_A STR_R STR_K "\0"
1165 ph10 391 #define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0"
1166     #define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0"
1167     #define STRING_THEN STR_T STR_H STR_E STR_N
1168    
1169     #define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0"
1170     #define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0"
1171     #define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0"
1172     #define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0"
1173     #define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0"
1174     #define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0"
1175     #define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0"
1176     #define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0"
1177     #define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0"
1178     #define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0"
1179     #define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0"
1180     #define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0"
1181     #define STRING_word0 STR_w STR_o STR_r STR_d "\0"
1182     #define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t
1183    
1184     #define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E
1185    
1186 ph10 576 #define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS
1187     #define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS
1188     #define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1189     #define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1190     #define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1191     #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
1192     #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
1193     #define STRING_UTF8_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1194     #define STRING_UCP_RIGHTPAR STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1195     #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
1196 ph10 391
1197     #endif /* SUPPORT_UTF8 */
1198    
1199 nigel 93 /* Escape items that are just an encoding of a particular data value. */
1200 nigel 77
1201     #ifndef ESC_e
1202 ph10 391 #define ESC_e CHAR_ESC
1203 nigel 77 #endif
1204    
1205     #ifndef ESC_f
1206 ph10 391 #define ESC_f CHAR_FF
1207 nigel 77 #endif
1208    
1209     #ifndef ESC_n
1210 ph10 391 #define ESC_n CHAR_NL
1211 nigel 77 #endif
1212    
1213     #ifndef ESC_r
1214 ph10 391 #define ESC_r CHAR_CR
1215 nigel 77 #endif
1216    
1217     /* We can't officially use ESC_t because it is a POSIX reserved identifier
1218     (presumably because of all the others like size_t). */
1219    
1220     #ifndef ESC_tee
1221 ph10 391 #define ESC_tee CHAR_HT
1222 nigel 77 #endif
1223    
1224 nigel 87 /* Codes for different types of Unicode property */
1225    
1226     #define PT_ANY 0 /* Any property - matches all chars */
1227     #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
1228 ph10 517 #define PT_GC 2 /* Specified general characteristic (e.g. L) */
1229     #define PT_PC 3 /* Specified particular characteristic (e.g. Lu) */
1230 nigel 87 #define PT_SC 4 /* Script (e.g. Han) */
1231 ph10 517 #define PT_ALNUM 5 /* Alphanumeric - the union of L and N */
1232     #define PT_SPACE 6 /* Perl space - Z plus 9,10,12,13 */
1233     #define PT_PXSPACE 7 /* POSIX space - Z plus 9,10,11,12,13 */
1234     #define PT_WORD 8 /* Word - L plus N plus underscore */
1235 nigel 87
1236     /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1237     contain UTF-8 characters with values greater than 255. */
1238    
1239     #define XCL_NOT 0x01 /* Flag: this is a negative class */
1240     #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
1241    
1242     #define XCL_END 0 /* Marks end of individual items */
1243     #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
1244     #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
1245     #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
1246     #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
1247    
1248 nigel 77 /* These are escaped items that aren't just an encoding of a particular data
1249     value such as \n. They must have non-zero values, as check_escape() returns
1250     their negation. Also, they must appear in the same order as in the opcode
1251 ph10 514 definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1252     corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1253 ph10 535 used for [^] in JavaScript compatibility mode. In non-DOTALL mode, "." behaves
1254 ph10 514 like \N.
1255 nigel 77
1256 ph10 518 The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1257     when PCRE_UCP is set, when replacement of \d etc by \p sequences is required.
1258     They must be contiguous, and remain in order so that the replacements can be
1259     looked up from a table.
1260    
1261 ph10 341 The final escape must be ESC_REF as subsequent values are used for
1262     backreferences (\1, \2, \3, etc). There are two tests in the code for an escape
1263     greater than ESC_b and less than ESC_Z to detect the types that may be
1264     repeated. These are the types that consume characters. If any new escapes are
1265     put in between that don't consume a character, that code will have to change.
1266     */
1267    
1268 ph10 172 enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1269 ph10 514 ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1270 ph10 535 ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1271 ph10 518 ESC_E, ESC_Q, ESC_g, ESC_k,
1272 ph10 535 ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu,
1273 ph10 341 ESC_REF };
1274 nigel 77
1275 ph10 210 /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1276 nigel 77 OP_EOD must correspond in order to the list of escapes immediately above.
1277    
1278 nigel 93 *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1279 ph10 462 that follow must also be updated to match. There are also tables called
1280     "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1281 nigel 93
1282 nigel 77 enum {
1283     OP_END, /* 0 End of pattern */
1284    
1285     /* Values corresponding to backslashed metacharacters */
1286    
1287     OP_SOD, /* 1 Start of data: \A */
1288     OP_SOM, /* 2 Start of match (subject + offset): \G */
1289 ph10 172 OP_SET_SOM, /* 3 Set start of match (\K) */
1290 ph10 168 OP_NOT_WORD_BOUNDARY, /* 4 \B */
1291     OP_WORD_BOUNDARY, /* 5 \b */
1292     OP_NOT_DIGIT, /* 6 \D */
1293     OP_DIGIT, /* 7 \d */
1294     OP_NOT_WHITESPACE, /* 8 \S */
1295     OP_WHITESPACE, /* 9 \s */
1296     OP_NOT_WORDCHAR, /* 10 \W */
1297     OP_WORDCHAR, /* 11 \w */
1298 ph10 602
1299 ph10 514 OP_ANY, /* 12 Match any character except newline */
1300     OP_ALLANY, /* 13 Match any character */
1301 ph10 341 OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1302     OP_NOTPROP, /* 15 \P (not Unicode property) */
1303     OP_PROP, /* 16 \p (Unicode property) */
1304     OP_ANYNL, /* 17 \R (any newline sequence) */
1305     OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */
1306     OP_HSPACE, /* 19 \h (horizontal whitespace) */
1307     OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */
1308     OP_VSPACE, /* 21 \v (vertical whitespace) */
1309     OP_EXTUNI, /* 22 \X (extended Unicode sequence */
1310     OP_EODN, /* 23 End of data or \n at end of data: \Z. */
1311     OP_EOD, /* 24 End of data: \z */
1312 nigel 77
1313 ph10 602 OP_CIRC, /* 25 Start of line - not multiline */
1314     OP_CIRCM, /* 26 Start of line - multiline */
1315     OP_DOLL, /* 27 End of line - not multiline */
1316     OP_DOLLM, /* 28 End of line - multiline */
1317     OP_CHAR, /* 29 Match one character, casefully */
1318     OP_CHARI, /* 30 Match one character, caselessly */
1319     OP_NOT, /* 31 Match one character, not the given one, casefully */
1320     OP_NOTI, /* 32 Match one character, not the given one, caselessly */
1321 nigel 77
1322 ph10 602 /* The following sets of 13 opcodes must always be kept in step because
1323     the offset from the first one is used to generate the others. */
1324 ph10 604
1325 ph10 602 /**** Single characters, caseful, must precede the caseless ones ****/
1326 nigel 77
1327 ph10 602 OP_STAR, /* 33 The maximizing and minimizing versions of */
1328     OP_MINSTAR, /* 34 these six opcodes must come in pairs, with */
1329     OP_PLUS, /* 35 the minimizing one second. */
1330     OP_MINPLUS, /* 36 */
1331     OP_QUERY, /* 37 */
1332     OP_MINQUERY, /* 38 */
1333 nigel 77
1334 ph10 602 OP_UPTO, /* 39 From 0 to n matches of one character, caseful*/
1335     OP_MINUPTO, /* 40 */
1336     OP_EXACT, /* 41 Exactly n matches */
1337 nigel 77
1338 ph10 602 OP_POSSTAR, /* 42 Possessified star, caseful */
1339     OP_POSPLUS, /* 43 Possessified plus, caseful */
1340     OP_POSQUERY, /* 44 Posesssified query, caseful */
1341     OP_POSUPTO, /* 45 Possessified upto, caseful */
1342 nigel 77
1343 ph10 602 /**** Single characters, caseless, must follow the caseful ones */
1344 ph10 604
1345 ph10 602 OP_STARI, /* 46 */
1346     OP_MINSTARI, /* 47 */
1347     OP_PLUSI, /* 48 */
1348     OP_MINPLUSI, /* 49 */
1349     OP_QUERYI, /* 50 */
1350     OP_MINQUERYI, /* 51 */
1351 nigel 93
1352 ph10 602 OP_UPTOI, /* 52 From 0 to n matches of one character, caseless */
1353     OP_MINUPTOI, /* 53 */
1354     OP_EXACTI, /* 54 */
1355 nigel 93
1356 ph10 602 OP_POSSTARI, /* 55 Possessified star, caseless */
1357     OP_POSPLUSI, /* 56 Possessified plus, caseless */
1358     OP_POSQUERYI, /* 57 Posesssified query, caseless */
1359     OP_POSUPTOI, /* 58 Possessified upto, caseless */
1360 nigel 93
1361 ph10 602 /**** The negated ones must follow the non-negated ones, and match them ****/
1362     /**** Negated single character, caseful; must precede the caseless ones ****/
1363 nigel 93
1364 ph10 602 OP_NOTSTAR, /* 59 The maximizing and minimizing versions of */
1365     OP_NOTMINSTAR, /* 60 these six opcodes must come in pairs, with */
1366     OP_NOTPLUS, /* 61 the minimizing one second. They must be in */
1367     OP_NOTMINPLUS, /* 62 exactly the same order as those above. */
1368     OP_NOTQUERY, /* 63 */
1369     OP_NOTMINQUERY, /* 64 */
1370 nigel 93
1371 ph10 602 OP_NOTUPTO, /* 65 From 0 to n matches, caseful */
1372     OP_NOTMINUPTO, /* 66 */
1373     OP_NOTEXACT, /* 67 Exactly n matches */
1374 nigel 93
1375 ph10 602 OP_NOTPOSSTAR, /* 68 Possessified versions, caseful */
1376     OP_NOTPOSPLUS, /* 69 */
1377     OP_NOTPOSQUERY, /* 70 */
1378     OP_NOTPOSUPTO, /* 71 */
1379 nigel 77
1380 ph10 602 /**** Negated single character, caseless; must follow the caseful ones ****/
1381 ph10 604
1382 ph10 602 OP_NOTSTARI, /* 72 */
1383     OP_NOTMINSTARI, /* 73 */
1384     OP_NOTPLUSI, /* 74 */
1385     OP_NOTMINPLUSI, /* 75 */
1386     OP_NOTQUERYI, /* 76 */
1387     OP_NOTMINQUERYI, /* 77 */
1388 nigel 77
1389 ph10 602 OP_NOTUPTOI, /* 78 From 0 to n matches, caseless */
1390     OP_NOTMINUPTOI, /* 79 */
1391     OP_NOTEXACTI, /* 80 Exactly n matches */
1392 nigel 77
1393 ph10 602 OP_NOTPOSSTARI, /* 81 Possessified versions, caseless */
1394     OP_NOTPOSPLUSI, /* 82 */
1395     OP_NOTPOSQUERYI, /* 83 */
1396     OP_NOTPOSUPTOI, /* 84 */
1397 nigel 77
1398 ph10 602 /**** Character types ****/
1399 ph10 604
1400 ph10 602 OP_TYPESTAR, /* 85 The maximizing and minimizing versions of */
1401     OP_TYPEMINSTAR, /* 86 these six opcodes must come in pairs, with */
1402     OP_TYPEPLUS, /* 87 the minimizing one second. These codes must */
1403     OP_TYPEMINPLUS, /* 88 be in exactly the same order as those above. */
1404     OP_TYPEQUERY, /* 89 */
1405     OP_TYPEMINQUERY, /* 90 */
1406    
1407     OP_TYPEUPTO, /* 91 From 0 to n matches */
1408     OP_TYPEMINUPTO, /* 92 */
1409     OP_TYPEEXACT, /* 93 Exactly n matches */
1410    
1411     OP_TYPEPOSSTAR, /* 94 Possessified versions */
1412     OP_TYPEPOSPLUS, /* 95 */
1413     OP_TYPEPOSQUERY, /* 96 */
1414     OP_TYPEPOSUPTO, /* 97 */
1415    
1416     /* These are used for character classes and back references; only the
1417 ph10 604 first six are the same as the sets above. */
1418 ph10 602
1419     OP_CRSTAR, /* 98 The maximizing and minimizing versions of */
1420     OP_CRMINSTAR, /* 99 all these opcodes must come in pairs, with */
1421     OP_CRPLUS, /* 100 the minimizing one second. These codes must */
1422     OP_CRMINPLUS, /* 101 be in exactly the same order as those above. */
1423     OP_CRQUERY, /* 102 */
1424     OP_CRMINQUERY, /* 103 */
1425    
1426     OP_CRRANGE, /* 104 These are different to the three sets above. */
1427     OP_CRMINRANGE, /* 105 */
1428    
1429     /* End of quantifier opcodes */
1430    
1431     OP_CLASS, /* 106 Match a character class, chars < 256 only */
1432     OP_NCLASS, /* 107 Same, but the bitmap was created from a negative
1433     class - the difference is relevant only when a
1434     UTF-8 character > 255 is encountered. */
1435     OP_XCLASS, /* 108 Extended class for handling UTF-8 chars within the
1436     class. This does both positive and negative. */
1437     OP_REF, /* 109 Match a back reference, casefully */
1438     OP_REFI, /* 110 Match a back reference, caselessly */
1439     OP_RECURSE, /* 111 Match a numbered subpattern (possibly recursive) */
1440     OP_CALLOUT, /* 112 Call out to external function if provided */
1441    
1442     OP_ALT, /* 113 Start of alternation */
1443     OP_KET, /* 114 End of group that doesn't have an unbounded repeat */
1444     OP_KETRMAX, /* 115 These two must remain together and in this */
1445     OP_KETRMIN, /* 116 order. They are for groups the repeat for ever. */
1446 ph10 604 OP_KETRPOS, /* 117 Possessive unlimited repeat. */
1447 ph10 602
1448 ph10 613 /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
1449     asserts must remain in order. */
1450 nigel 77
1451 ph10 637 OP_REVERSE, /* 118 Move pointer back - used in lookbehind assertions */
1452     OP_ASSERT, /* 119 Positive lookahead */
1453     OP_ASSERT_NOT, /* 120 Negative lookahead */
1454     OP_ASSERTBACK, /* 121 Positive lookbehind */
1455     OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
1456 nigel 77
1457 ph10 637 /* ONCE, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately after the
1458     assertions, with ONCE first, as there's a test for >= ONCE for a subpattern
1459     that isn't an assertion. The POS versions must immediately follow the non-POS
1460     versions in each case. */
1461 nigel 77
1462 ph10 604 OP_ONCE, /* 123 Atomic group */
1463     OP_BRA, /* 124 Start of non-capturing bracket */
1464     OP_BRAPOS, /* 125 Ditto, with unlimited, possessive repeat */
1465     OP_CBRA, /* 126 Start of capturing bracket */
1466     OP_CBRAPOS, /* 127 Ditto, with unlimited, possessive repeat */
1467     OP_COND, /* 128 Conditional group */
1468 nigel 77
1469 ph10 604 /* These five must follow the previous five, in the same order. There's a
1470 nigel 93 check for >= SBRA to distinguish the two sets. */
1471 nigel 77
1472 ph10 604 OP_SBRA, /* 129 Start of non-capturing bracket, check empty */
1473     OP_SBRAPOS, /* 130 Ditto, with unlimited, possessive repeat */
1474     OP_SCBRA, /* 131 Start of capturing bracket, check empty */
1475     OP_SCBRAPOS, /* 132 Ditto, with unlimited, possessive repeat */
1476     OP_SCOND, /* 133 Conditional group, check empty */
1477 nigel 77
1478 ph10 459 /* The next two pairs must (respectively) be kept together. */
1479 ph10 461
1480 ph10 604 OP_CREF, /* 134 Used to hold a capture number as condition */
1481     OP_NCREF, /* 135 Same, but generated by a name reference*/
1482     OP_RREF, /* 136 Used to hold a recursion number as condition */
1483     OP_NRREF, /* 137 Same, but generated by a name reference*/
1484     OP_DEF, /* 138 The DEFINE condition */
1485 nigel 93
1486 ph10 604 OP_BRAZERO, /* 139 These two must remain together and in this */
1487     OP_BRAMINZERO, /* 140 order. */
1488 ph10 613 OP_BRAPOSZERO, /* 141 */
1489 ph10 210
1490     /* These are backtracking control verbs */
1491    
1492 ph10 604 OP_MARK, /* 142 always has an argument */
1493     OP_PRUNE, /* 143 */
1494     OP_PRUNE_ARG, /* 144 same, but with argument */
1495     OP_SKIP, /* 145 */
1496     OP_SKIP_ARG, /* 146 same, but with argument */
1497     OP_THEN, /* 147 */
1498     OP_THEN_ARG, /* 148 same, but with argument */
1499     OP_COMMIT, /* 149 */
1500 ph10 210
1501     /* These are forced failure and success verbs */
1502    
1503 ph10 604 OP_FAIL, /* 150 */
1504     OP_ACCEPT, /* 151 */
1505 ph10 613 OP_ASSERT_ACCEPT, /* 152 Used inside assertions */
1506     OP_CLOSE, /* 153 Used before OP_ACCEPT to close open captures */
1507 ph10 345
1508 ph10 335 /* This is used to skip a subpattern with a {0} quantifier */
1509 ph10 345
1510 ph10 613 OP_SKIPZERO, /* 154 */
1511 ph10 498
1512     /* This is not an opcode, but is used to check that tables indexed by opcode
1513     are the correct length, in order to catch updating errors - there have been
1514     some in the past. */
1515    
1516     OP_TABLE_LENGTH
1517 nigel 77 };
1518    
1519 ph10 462 /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
1520     definitions that follow must also be updated to match. There are also tables
1521 ph10 510 called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1522 nigel 77
1523 ph10 462
1524 nigel 77 /* This macro defines textual names for all the opcodes. These are used only
1525 ph10 602 for debugging, and some of them are only partial names. The macro is referenced
1526 ph10 604 only in pcre_printint.c, which fills out the full names in many cases (and in
1527 ph10 602 some cases doesn't actually use these names at all). */
1528 nigel 77
1529     #define OP_NAME_LIST \
1530 ph10 168 "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \
1531 ph10 341 "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \
1532 ph10 178 "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \
1533     "extuni", "\\Z", "\\z", \
1534 ph10 602 "^", "^", "$", "$", "char", "chari", "not", "noti", \
1535     "*", "*?", "+", "+?", "?", "??", \
1536     "{", "{", "{", \
1537 nigel 93 "*+","++", "?+", "{", \
1538 ph10 602 "*", "*?", "+", "+?", "?", "??", \
1539     "{", "{", "{", \
1540 nigel 93 "*+","++", "?+", "{", \
1541 ph10 602 "*", "*?", "+", "+?", "?", "??", \
1542     "{", "{", "{", \
1543     "*+","++", "?+", "{", \
1544     "*", "*?", "+", "+?", "?", "??", \
1545     "{", "{", "{", \
1546     "*+","++", "?+", "{", \
1547 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1548 nigel 93 "*+","++", "?+", "{", \
1549 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", \
1550 ph10 602 "class", "nclass", "xclass", "Ref", "Refi", \
1551     "Recurse", "Callout", \
1552 ph10 604 "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos", \
1553 ph10 637 "Reverse", "Assert", "Assert not", "AssertB", "AssertB not", \
1554 ph10 604 "Once", \
1555     "Bra", "BraPos", "CBra", "CBraPos", \
1556     "Cond", \
1557     "SBra", "SBraPos", "SCBra", "SCBraPos", \
1558     "SCond", \
1559 ph10 459 "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def", \
1560 ph10 604 "Brazero", "Braminzero", "Braposzero", \
1561 ph10 510 "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP", \
1562 ph10 613 "*THEN", "*THEN", "*COMMIT", "*FAIL", \
1563     "*ACCEPT", "*ASSERT_ACCEPT", \
1564 ph10 447 "Close", "Skip zero"
1565 nigel 77
1566    
1567     /* This macro defines the length of fixed length operations in the compiled
1568     regex. The lengths are used when searching for specific things, and also in the
1569     debugging printing of a compiled regex. We use a macro so that it can be
1570     defined close to the definitions of the opcodes themselves.
1571    
1572     As things have been extended, some of these are no longer fixed lenths, but are
1573     minima instead. For example, the length of a single-character repeat may vary
1574     in UTF-8 mode. The code that uses this table must know about such things. */
1575    
1576     #define OP_LENGTHS \
1577     1, /* End */ \
1578 ph10 168 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \
1579     1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \
1580 ph10 341 1, 1, 1, /* Any, AllAny, Anybyte */ \
1581 ph10 498 3, 3, /* \P, \p */ \
1582 ph10 178 1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \
1583 ph10 498 1, /* \X */ \
1584 ph10 602 1, 1, 1, 1, 1, 1, /* \Z, \z, ^, ^M, $, $M */ \
1585 nigel 77 2, /* Char - the minimum length */ \
1586 ph10 602 2, /* Chari - the minimum length */ \
1587 nigel 77 2, /* not */ \
1588 ph10 602 2, /* noti */ \
1589     /* Positive single-char repeats ** These are */ \
1590     2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
1591     4, 4, 4, /* upto, minupto, exact ** mode */ \
1592 nigel 93 2, 2, 2, 4, /* *+, ++, ?+, upto+ */ \
1593 ph10 602 2, 2, 2, 2, 2, 2, /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8 */ \
1594     4, 4, 4, /* upto I, minupto I, exact I */ \
1595     2, 2, 2, 4, /* *+I, ++I, ?+I, upto+I */ \
1596 nigel 77 /* Negative single-char repeats - only for chars < 256 */ \
1597     2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
1598     4, 4, 4, /* NOT upto, minupto, exact */ \
1599 ph10 602 2, 2, 2, 4, /* Possessive NOT *, +, ?, upto */ \
1600     2, 2, 2, 2, 2, 2, /* NOT *I, *?I, +I, +?I, ?I, ??I */ \
1601     4, 4, 4, /* NOT upto I, minupto I, exact I */ \
1602     2, 2, 2, 4, /* Possessive NOT *I, +I, ?I, upto I */ \
1603 nigel 77 /* Positive type repeats */ \
1604     2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
1605     4, 4, 4, /* Type upto, minupto, exact */ \
1606 nigel 93 2, 2, 2, 4, /* Possessive *+, ++, ?+, upto+ */ \
1607 nigel 77 /* Character class & ref repeats */ \
1608     1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
1609     5, 5, /* CRRANGE, CRMINRANGE */ \
1610     33, /* CLASS */ \
1611     33, /* NCLASS */ \
1612     0, /* XCLASS - variable length */ \
1613     3, /* REF */ \
1614 ph10 602 3, /* REFI */ \
1615 nigel 77 1+LINK_SIZE, /* RECURSE */ \
1616     2+2*LINK_SIZE, /* CALLOUT */ \
1617     1+LINK_SIZE, /* Alt */ \
1618     1+LINK_SIZE, /* Ket */ \
1619     1+LINK_SIZE, /* KetRmax */ \
1620     1+LINK_SIZE, /* KetRmin */ \
1621 ph10 604 1+LINK_SIZE, /* KetRpos */ \
1622 ph10 637 1+LINK_SIZE, /* Reverse */ \
1623 nigel 77 1+LINK_SIZE, /* Assert */ \
1624     1+LINK_SIZE, /* Assert not */ \
1625     1+LINK_SIZE, /* Assert behind */ \
1626     1+LINK_SIZE, /* Assert behind not */ \
1627 nigel 93 1+LINK_SIZE, /* ONCE */ \
1628     1+LINK_SIZE, /* BRA */ \
1629 ph10 604 1+LINK_SIZE, /* BRAPOS */ \
1630 nigel 93 3+LINK_SIZE, /* CBRA */ \
1631 ph10 604 3+LINK_SIZE, /* CBRAPOS */ \
1632 nigel 77 1+LINK_SIZE, /* COND */ \
1633 nigel 93 1+LINK_SIZE, /* SBRA */ \
1634 ph10 604 1+LINK_SIZE, /* SBRAPOS */ \
1635 nigel 93 3+LINK_SIZE, /* SCBRA */ \
1636 ph10 604 3+LINK_SIZE, /* SCBRAPOS */ \
1637 nigel 93 1+LINK_SIZE, /* SCOND */ \
1638 ph10 459 3, 3, /* CREF, NCREF */ \
1639     3, 3, /* RREF, NRREF */ \
1640 nigel 93 1, /* DEF */ \
1641 ph10 604 1, 1, 1, /* BRAZERO, BRAMINZERO, BRAPOSZERO */ \
1642 ph10 550 3, 1, 3, /* MARK, PRUNE, PRUNE_ARG */ \
1643     1, 3, /* SKIP, SKIP_ARG */ \
1644     1+LINK_SIZE, 3+LINK_SIZE, /* THEN, THEN_ARG */ \
1645 ph10 613 1, 1, 1, 1, /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT */ \
1646     3, 1 /* CLOSE, SKIPZERO */
1647 nigel 77
1648 ph10 459 /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
1649     condition. */
1650 nigel 77
1651 nigel 93 #define RREF_ANY 0xffff
1652 nigel 77
1653 ph10 499 /* Compile time error code numbers. They are given names so that they can more
1654 ph10 507 easily be tracked. When a new number is added, the table called eint in
1655 ph10 499 pcreposix.c must be updated. */
1656 nigel 77
1657     enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
1658     ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
1659     ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
1660     ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
1661 nigel 91 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
1662 ph10 210 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
1663 ph10 629 ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
1664 ph10 574 ERRCOUNT };
1665 nigel 77
1666     /* The real format of the start of the pcre block; the index of names and the
1667     code vector run on as long as necessary after the end. We store an explicit
1668     offset to the name table so that if a regex is compiled on one host, saved, and
1669     then run on another where the size of pointers is different, all might still
1670     be well. For the case of compiled-on-4 and run-on-8, we include an extra
1671     pointer that is always NULL. For future-proofing, a few dummy fields were
1672     originally included - even though you can never get this planning right - but
1673     there is only one left now.
1674    
1675     NOTE NOTE NOTE:
1676     Because people can now save and re-use compiled patterns, any additions to this
1677     structure should be made at the end, and something earlier (e.g. a new
1678     flag in the options or one of the dummy fields) should indicate that the new
1679     fields are present. Currently PCRE always sets the dummy fields to zero.
1680 ph10 455 NOTE NOTE NOTE
1681 nigel 77 */
1682    
1683     typedef struct real_pcre {
1684     pcre_uint32 magic_number;
1685     pcre_uint32 size; /* Total that was malloced */
1686 ph10 230 pcre_uint32 options; /* Public options */
1687     pcre_uint16 flags; /* Private flags */
1688     pcre_uint16 dummy1; /* For future use */
1689 nigel 77 pcre_uint16 top_bracket;
1690     pcre_uint16 top_backref;
1691     pcre_uint16 first_byte;
1692     pcre_uint16 req_byte;
1693     pcre_uint16 name_table_offset; /* Offset to name table that follows */
1694     pcre_uint16 name_entry_size; /* Size of any name items */
1695     pcre_uint16 name_count; /* Number of name items */
1696     pcre_uint16 ref_count; /* Reference count */
1697    
1698     const unsigned char *tables; /* Pointer to tables or NULL for std */
1699     const unsigned char *nullpad; /* NULL padding */
1700     } real_pcre;
1701    
1702     /* The format of the block used to store data from pcre_study(). The same
1703     remark (see NOTE above) about extending this structure applies. */
1704    
1705     typedef struct pcre_study_data {
1706     pcre_uint32 size; /* Total that was malloced */
1707 ph10 455 pcre_uint32 flags; /* Private flags */
1708     uschar start_bits[32]; /* Starting char bits */
1709 ph10 459 pcre_uint32 minlength; /* Minimum subject length */
1710 nigel 77 } pcre_study_data;
1711    
1712 ph10 459 /* Structure for building a chain of open capturing subpatterns during
1713     compiling, so that instructions to close them can be compiled when (*ACCEPT) is
1714 ph10 488 encountered. This is also used to identify subpatterns that contain recursive
1715     back references to themselves, so that they can be made atomic. */
1716 ph10 447
1717     typedef struct open_capitem {
1718     struct open_capitem *next; /* Chain link */
1719     pcre_uint16 number; /* Capture number */
1720 ph10 488 pcre_uint16 flag; /* Set TRUE if recursive back ref */
1721 ph10 459 } open_capitem;
1722 ph10 447
1723 nigel 77 /* Structure for passing "static" information around between the functions
1724     doing the compiling, so that they are thread-safe. */
1725    
1726     typedef struct compile_data {
1727     const uschar *lcc; /* Points to lower casing table */
1728     const uschar *fcc; /* Points to case-flipping table */
1729     const uschar *cbits; /* Points to character type table */
1730     const uschar *ctypes; /* Points to table of type maps */
1731 nigel 93 const uschar *start_workspace;/* The start of working space */
1732 nigel 77 const uschar *start_code; /* The start of the compiled code */
1733     const uschar *start_pattern; /* The start of the pattern */
1734 nigel 93 const uschar *end_pattern; /* The end of the pattern */
1735 ph10 459 open_capitem *open_caps; /* Chain of open capture items */
1736 nigel 93 uschar *hwm; /* High watermark of workspace */
1737 nigel 77 uschar *name_table; /* The name/number table */
1738     int names_found; /* Number of entries so far */
1739     int name_entry_size; /* Size of each entry */
1740 ph10 270 int bracount; /* Count of capturing parens as we compile */
1741 ph10 286 int final_bracount; /* Saved value after first pass */
1742 nigel 77 int top_backref; /* Maximum back reference */
1743     unsigned int backref_map; /* Bitmap of low back refs */
1744 ph10 613 int assert_depth; /* Depth of nested assertions */
1745 nigel 93 int external_options; /* External (initial) options */
1746 ph10 231 int external_flags; /* External flag bits to be set */
1747 nigel 77 int req_varyopt; /* "After variable item" flag for reqbyte */
1748 ph10 211 BOOL had_accept; /* (*ACCEPT) encountered */
1749 ph10 459 BOOL check_lookbehind; /* Lookbehinds need later checking */
1750 nigel 93 int nltype; /* Newline type */
1751     int nllen; /* Newline string length */
1752     uschar nl[4]; /* Newline string when fixed length */
1753 nigel 77 } compile_data;
1754    
1755     /* Structure for maintaining a chain of pointers to the currently incomplete
1756 ph10 642 branches, for testing for left recursion while compiling. */
1757 nigel 77
1758     typedef struct branch_chain {
1759     struct branch_chain *outer;
1760 ph10 475 uschar *current_branch;
1761 nigel 77 } branch_chain;
1762    
1763     /* Structure for items in a linked list that represents an explicit recursive
1764 ph10 642 call within the pattern; used by pcre_exec(). */
1765 nigel 77
1766     typedef struct recursion_info {
1767     struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
1768 ph10 618 int group_num; /* Number of group that was called */
1769     int *offset_save; /* Pointer to start of saved offsets */
1770     int saved_max; /* Number of saved offsets */
1771 ph10 642 USPTR subject_position; /* Position at start of recursion */
1772 nigel 77 } recursion_info;
1773    
1774 ph10 642 /* A similar structure for pcre_dfa_exec(). */
1775    
1776     typedef struct dfa_recursion_info {
1777     struct dfa_recursion_info *prevrec;
1778     int group_num;
1779     USPTR subject_position;
1780     } dfa_recursion_info;
1781    
1782 nigel 93 /* Structure for building a chain of data for holding the values of the subject
1783     pointer at the start of each subpattern, so as to detect when an empty string
1784 ph10 642 has been matched by a subpattern - to break infinite loops; used by
1785     pcre_exec(). */
1786 nigel 93
1787     typedef struct eptrblock {
1788     struct eptrblock *epb_prev;
1789     USPTR epb_saved_eptr;
1790     } eptrblock;
1791    
1792    
1793 nigel 77 /* Structure for passing "static" information around between the functions
1794     doing traditional NFA matching, so that they are thread-safe. */
1795    
1796     typedef struct match_data {
1797 nigel 87 unsigned long int match_call_count; /* As it says */
1798     unsigned long int match_limit; /* As it says */
1799     unsigned long int match_limit_recursion; /* As it says */
1800 nigel 77 int *offset_vector; /* Offset vector */
1801     int offset_end; /* One past the end */
1802     int offset_max; /* The maximum usable for return data */
1803 nigel 93 int nltype; /* Newline type */
1804     int nllen; /* Newline string length */
1805 ph10 459 int name_count; /* Number of names in name table */
1806     int name_entry_size; /* Size of entry in names table */
1807 ph10 461 uschar *name_table; /* Table of names */
1808 nigel 93 uschar nl[4]; /* Newline string when fixed */
1809 ph10 613 const uschar *lcc; /* Points to lower casing table */
1810     const uschar *ctypes; /* Points to table of type maps */
1811 nigel 77 BOOL offset_overflow; /* Set if too many extractions */
1812     BOOL notbol; /* NOTBOL flag */
1813     BOOL noteol; /* NOTEOL flag */
1814     BOOL utf8; /* UTF8 flag */
1815 ph10 345 BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */
1816 ph10 535 BOOL use_ucp; /* PCRE_UCP flag */
1817 nigel 77 BOOL endonly; /* Dollar not before final \n */
1818     BOOL notempty; /* Empty string match not wanted */
1819 ph10 442 BOOL notempty_atstart; /* Empty string match at start not wanted */
1820 nigel 77 BOOL hitend; /* Hit the end of the subject at some point */
1821 ph10 231 BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
1822 ph10 613 const uschar *start_code; /* For use when recursing */
1823 nigel 87 USPTR start_subject; /* Start of the subject string */
1824     USPTR end_subject; /* End of the subject string */
1825 ph10 168 USPTR start_match_ptr; /* Start of matched string */
1826 nigel 87 USPTR end_match_ptr; /* Subject position at end match */
1827 ph10 442 USPTR start_used_ptr; /* Earliest consulted character */
1828 ph10 435 int partial; /* PARTIAL options */
1829 nigel 77 int end_offset_top; /* Highwater mark at end of match */
1830     int capture_last; /* Most recent capture number */
1831     int start_offset; /* The start offset value */
1832 ph10 613 int match_function_type; /* Set for certain special calls of MATCH() */
1833 nigel 93 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
1834     int eptrn; /* Next free eptrblock */
1835 nigel 77 recursion_info *recursive; /* Linked list of recursion data */
1836     void *callout_data; /* To pass back to callouts */
1837 ph10 613 const uschar *mark; /* Mark pointer to pass back */
1838 ph10 618 const uschar *once_target; /* Where to back up to for atomic groups */
1839 nigel 77 } match_data;
1840    
1841     /* A similar structure is used for the same purpose by the DFA matching
1842     functions. */
1843    
1844     typedef struct dfa_match_data {
1845 ph10 642 const uschar *start_code; /* Start of the compiled pattern */
1846     const uschar *start_subject; /* Start of the subject string */
1847     const uschar *end_subject; /* End of subject string */
1848     const uschar *start_used_ptr; /* Earliest consulted character */
1849     const uschar *tables; /* Character tables */
1850     int start_offset; /* The start offset value */
1851     int moptions; /* Match options */
1852     int poptions; /* Pattern options */
1853     int nltype; /* Newline type */
1854     int nllen; /* Newline string length */
1855     uschar nl[4]; /* Newline string when fixed */
1856     void *callout_data; /* To pass back to callouts */
1857     dfa_recursion_info *recursive; /* Linked list of recursion data */
1858 nigel 77 } dfa_match_data;
1859    
1860     /* Bit definitions for entries in the pcre_ctypes table. */
1861    
1862     #define ctype_space 0x01
1863     #define ctype_letter 0x02
1864     #define ctype_digit 0x04
1865     #define ctype_xdigit 0x08
1866 ph10 274 #define ctype_word 0x10 /* alphanumeric or '_' */
1867 nigel 77 #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
1868    
1869     /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
1870     of bits for a class map. Some classes are built by combining these tables. */
1871    
1872     #define cbit_space 0 /* [:space:] or \s */
1873     #define cbit_xdigit 32 /* [:xdigit:] */
1874     #define cbit_digit 64 /* [:digit:] or \d */
1875     #define cbit_upper 96 /* [:upper:] */
1876     #define cbit_lower 128 /* [:lower:] */
1877     #define cbit_word 160 /* [:word:] or \w */
1878     #define cbit_graph 192 /* [:graph:] */
1879     #define cbit_print 224 /* [:print:] */
1880     #define cbit_punct 256 /* [:punct:] */
1881     #define cbit_cntrl 288 /* [:cntrl:] */
1882     #define cbit_length 320 /* Length of the cbits table */
1883    
1884     /* Offsets of the various tables from the base tables pointer, and
1885     total length. */
1886    
1887     #define lcc_offset 0
1888     #define fcc_offset 256
1889     #define cbits_offset 512
1890     #define ctypes_offset (cbits_offset + cbit_length)
1891     #define tables_length (ctypes_offset + 256)
1892    
1893 nigel 87 /* Layout of the UCP type table that translates property names into types and
1894 ph10 240 codes. Each entry used to point directly to a name, but to reduce the number of
1895 ph10 243 relocations in shared libraries, it now has an offset into a single string
1896 ph10 240 instead. */
1897 nigel 77
1898     typedef struct {
1899 ph10 243 pcre_uint16 name_offset;
1900 nigel 87 pcre_uint16 type;
1901     pcre_uint16 value;
1902 nigel 77 } ucp_type_table;
1903    
1904    
1905     /* Internal shared data tables. These are tables that are used by more than one
1906     of the exported public functions. They have to be "external" in the C sense,
1907     but are not part of the PCRE public API. The data for these tables is in the
1908     pcre_tables.c module. */
1909    
1910     extern const int _pcre_utf8_table1[];
1911     extern const int _pcre_utf8_table2[];
1912     extern const int _pcre_utf8_table3[];
1913     extern const uschar _pcre_utf8_table4[];
1914    
1915     extern const int _pcre_utf8_table1_size;
1916    
1917 ph10 240 extern const char _pcre_utt_names[];
1918 nigel 77 extern const ucp_type_table _pcre_utt[];
1919     extern const int _pcre_utt_size;
1920    
1921     extern const uschar _pcre_default_tables[];
1922    
1923     extern const uschar _pcre_OP_lengths[];
1924    
1925    
1926     /* Internal shared functions. These are functions that are used by more than
1927     one of the exported public functions. They have to be "external" in the C
1928     sense, but are not part of the PCRE public API. */
1929    
1930 ph10 455 extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);
1931 ph10 483 extern BOOL _pcre_is_newline(USPTR, int, USPTR, int *, BOOL);
1932 ph10 455 extern int _pcre_ord2utf8(int, uschar *);
1933     extern real_pcre *_pcre_try_flipped(const real_pcre *, real_pcre *,
1934     const pcre_study_data *, pcre_study_data *);
1935 ph10 598 extern int _pcre_valid_utf8(USPTR, int, int *);
1936 ph10 483 extern BOOL _pcre_was_newline(USPTR, int, USPTR, int *, BOOL);
1937 ph10 455 extern BOOL _pcre_xclass(int, const uschar *);
1938 nigel 77
1939 ph10 349
1940     /* Unicode character database (UCD) */
1941    
1942     typedef struct {
1943     uschar script;
1944     uschar chartype;
1945 ph10 351 pcre_int32 other_case;
1946 ph10 349 } ucd_record;
1947    
1948 ph10 351 extern const ucd_record _pcre_ucd_records[];
1949     extern const uschar _pcre_ucd_stage1[];
1950     extern const pcre_uint16 _pcre_ucd_stage2[];
1951     extern const int _pcre_ucp_gentype[];
1952 ph10 349
1953 ph10 351
1954 ph10 349 /* UCD access macros */
1955    
1956     #define UCD_BLOCK_SIZE 128
1957 ph10 351 #define GET_UCD(ch) (_pcre_ucd_records + \
1958     _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \
1959 ph10 593 UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])
1960 ph10 349
1961     #define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype
1962     #define UCD_SCRIPT(ch) GET_UCD(ch)->script
1963 ph10 351 #define UCD_CATEGORY(ch) _pcre_ucp_gentype[UCD_CHARTYPE(ch)]
1964     #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)
1965 ph10 349
1966 nigel 85 #endif
1967    
1968 nigel 77 /* End of pcre_internal.h */

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12