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

Contents of /code/trunk/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log


Revision 494 - (hide annotations) (download)
Tue Mar 2 11:53:12 2010 UTC (4 years, 4 months ago) by ph10
File MIME type: text/plain
File size: 74605 byte(s)
If stdint.h is not available, include inttypes.h if available.

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 475 Copyright (c) 1997-2010 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 208 byte-mode, and more complicated ones for UTF-8 characters. BACKCHAR should
412     never be called in byte mode. To make sure it can never even appear when UTF-8
413 ph10 207 support is omitted, we don't even define it. */
414 nigel 77
415     #ifndef SUPPORT_UTF8
416     #define GETCHAR(c, eptr) c = *eptr;
417     #define GETCHARTEST(c, eptr) c = *eptr;
418     #define GETCHARINC(c, eptr) c = *eptr++;
419     #define GETCHARINCTEST(c, eptr) c = *eptr++;
420     #define GETCHARLEN(c, eptr, len) c = *eptr;
421 ph10 207 /* #define BACKCHAR(eptr) */
422 nigel 77
423     #else /* SUPPORT_UTF8 */
424    
425     /* Get the next UTF-8 character, not advancing the pointer. This is called when
426     we know we are in UTF-8 mode. */
427    
428     #define GETCHAR(c, eptr) \
429     c = *eptr; \
430 nigel 93 if (c >= 0xc0) \
431 nigel 77 { \
432     int gcii; \
433     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
434     int gcss = 6*gcaa; \
435     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
436     for (gcii = 1; gcii <= gcaa; gcii++) \
437     { \
438     gcss -= 6; \
439     c |= (eptr[gcii] & 0x3f) << gcss; \
440     } \
441     }
442    
443     /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
444     pointer. */
445    
446     #define GETCHARTEST(c, eptr) \
447     c = *eptr; \
448 nigel 93 if (utf8 && c >= 0xc0) \
449 nigel 77 { \
450     int gcii; \
451     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
452     int gcss = 6*gcaa; \
453     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
454     for (gcii = 1; gcii <= gcaa; gcii++) \
455     { \
456     gcss -= 6; \
457     c |= (eptr[gcii] & 0x3f) << gcss; \
458     } \
459     }
460    
461     /* Get the next UTF-8 character, advancing the pointer. This is called when we
462     know we are in UTF-8 mode. */
463    
464     #define GETCHARINC(c, eptr) \
465     c = *eptr++; \
466 nigel 93 if (c >= 0xc0) \
467 nigel 77 { \
468     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
469     int gcss = 6*gcaa; \
470     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
471     while (gcaa-- > 0) \
472     { \
473     gcss -= 6; \
474     c |= (*eptr++ & 0x3f) << gcss; \
475     } \
476     }
477    
478     /* Get the next character, testing for UTF-8 mode, and advancing the pointer */
479    
480     #define GETCHARINCTEST(c, eptr) \
481     c = *eptr++; \
482 nigel 93 if (utf8 && c >= 0xc0) \
483 nigel 77 { \
484     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
485     int gcss = 6*gcaa; \
486     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
487     while (gcaa-- > 0) \
488     { \
489     gcss -= 6; \
490     c |= (*eptr++ & 0x3f) << gcss; \
491     } \
492     }
493    
494     /* Get the next UTF-8 character, not advancing the pointer, incrementing length
495     if there are extra bytes. This is called when we know we are in UTF-8 mode. */
496    
497     #define GETCHARLEN(c, eptr, len) \
498     c = *eptr; \
499 nigel 93 if (c >= 0xc0) \
500 nigel 77 { \
501     int gcii; \
502     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
503     int gcss = 6*gcaa; \
504     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
505     for (gcii = 1; gcii <= gcaa; gcii++) \
506     { \
507     gcss -= 6; \
508     c |= (eptr[gcii] & 0x3f) << gcss; \
509     } \
510     len += gcaa; \
511     }
512    
513 ph10 384 /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
514     pointer, incrementing length if there are extra bytes. This is called when we
515     know we are in UTF-8 mode. */
516    
517     #define GETCHARLENTEST(c, eptr, len) \
518     c = *eptr; \
519     if (utf8 && c >= 0xc0) \
520     { \
521     int gcii; \
522     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
523     int gcss = 6*gcaa; \
524     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
525     for (gcii = 1; gcii <= gcaa; gcii++) \
526     { \
527     gcss -= 6; \
528     c |= (eptr[gcii] & 0x3f) << gcss; \
529     } \
530     len += gcaa; \
531     }
532    
533 nigel 77 /* If the pointer is not at the start of a character, move it back until
534 ph10 208 it is. This is called only in UTF-8 mode - we don't put a test within the macro
535 ph10 207 because almost all calls are already within a block of UTF-8 only code. */
536 nigel 77
537 ph10 207 #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
538 nigel 77
539     #endif
540    
541    
542     /* In case there is no definition of offsetof() provided - though any proper
543     Standard C system should have one. */
544    
545     #ifndef offsetof
546     #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))
547     #endif
548    
549    
550     /* These are the public options that can change during matching. */
551    
552     #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)
553    
554 ph10 231 /* Private flags containing information about the compiled regex. They used to
555     live at the top end of the options word, but that got almost full, so now they
556 ph10 426 are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
557 ph10 427 the restrictions on partial matching have been lifted. It remains for backwards
558 ph10 426 compatibility. */
559 nigel 77
560 ph10 230 #define PCRE_NOPARTIAL 0x0001 /* can't use partial with this regex */
561     #define PCRE_FIRSTSET 0x0002 /* first_byte is set */
562     #define PCRE_REQCHSET 0x0004 /* req_byte is set */
563     #define PCRE_STARTLINE 0x0008 /* start after \n for multiline */
564     #define PCRE_JCHANGED 0x0010 /* j option used in regex */
565     #define PCRE_HASCRORLF 0x0020 /* explicit \r or \n in pattern */
566 nigel 77
567     /* Options for the "extra" block produced by pcre_study(). */
568    
569     #define PCRE_STUDY_MAPPED 0x01 /* a map of starting chars exists */
570 ph10 455 #define PCRE_STUDY_MINLEN 0x02 /* a minimum length field exists */
571 nigel 77
572     /* Masks for identifying the public options that are permitted at compile
573     time, run time, or study time, respectively. */
574    
575 ph10 149 #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
576     PCRE_NEWLINE_ANYCRLF)
577 nigel 93
578 ph10 389 #define PUBLIC_COMPILE_OPTIONS \
579 nigel 77 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
580     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
581 nigel 91 PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
582 ph10 336 PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
583     PCRE_JAVASCRIPT_COMPAT)
584 nigel 77
585     #define PUBLIC_EXEC_OPTIONS \
586 ph10 442 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
587     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
588     PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
589 nigel 77
590     #define PUBLIC_DFA_EXEC_OPTIONS \
591 ph10 442 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
592     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
593     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
594     PCRE_NO_START_OPTIMIZE)
595 nigel 77
596     #define PUBLIC_STUDY_OPTIONS 0 /* None defined */
597    
598     /* Magic number to provide a small check against being handed junk. Also used
599     to detect whether a pattern was compiled on a host of different endianness. */
600    
601     #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
602    
603     /* Negative values for the firstchar and reqchar variables */
604    
605     #define REQ_UNSET (-2)
606     #define REQ_NONE (-1)
607    
608     /* The maximum remaining length of subject we are prepared to search for a
609     req_byte match. */
610    
611     #define REQ_BYTE_MAX 1000
612    
613     /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a
614     variable-length repeat, or a anything other than literal characters. */
615    
616     #define REQ_CASELESS 0x0100 /* indicates caselessness */
617     #define REQ_VARY 0x0200 /* reqbyte followed non-literal item */
618    
619 ph10 369 /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
620 ph10 376 environments where these macros are defined elsewhere. Unfortunately, there
621     is no way to do the same for the typedef. */
622 nigel 77
623     typedef int BOOL;
624    
625 ph10 376 #ifndef FALSE
626 nigel 77 #define FALSE 0
627     #define TRUE 1
628 ph10 369 #endif
629 nigel 77
630 ph10 391 /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
631     character constants like '*' because the compiler would emit their EBCDIC code,
632 ph10 392 which is different from their ASCII/UTF-8 code. Instead we define macros for
633     the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
634     is enabled. When UTF-8 support is not enabled, the definitions use character
635     literals. Both character and string versions of each character are needed, and
636 ph10 391 there are some longer strings as well.
637    
638 ph10 392 This means that, on EBCDIC platforms, the PCRE library can handle either
639 ph10 391 EBCDIC, or UTF-8, but not both. To support both in the same compiled library
640     would need different lookups depending on whether PCRE_UTF8 was set or not.
641     This would make it impossible to use characters in switch/case statements,
642 ph10 392 which would reduce performance. For a theoretical use (which nobody has asked
643     for) in a minority area (EBCDIC platforms), this is not sensible. Any
644     application that did need both could compile two versions of the library, using
645 ph10 391 macros to give the functions distinct names. */
646    
647     #ifndef SUPPORT_UTF8
648    
649     /* UTF-8 support is not enabled; use the platform-dependent character literals
650     so that PCRE works on both ASCII and EBCDIC platforms, in non-UTF-mode only. */
651    
652     #define CHAR_HT '\t'
653     #define CHAR_VT '\v'
654     #define CHAR_FF '\f'
655     #define CHAR_CR '\r'
656     #define CHAR_NL '\n'
657     #define CHAR_BS '\b'
658     #define CHAR_BEL '\a'
659     #ifdef EBCDIC
660     #define CHAR_ESC '\047'
661     #define CHAR_DEL '\007'
662     #else
663     #define CHAR_ESC '\033'
664     #define CHAR_DEL '\177'
665     #endif
666    
667     #define CHAR_SPACE ' '
668     #define CHAR_EXCLAMATION_MARK '!'
669     #define CHAR_QUOTATION_MARK '"'
670     #define CHAR_NUMBER_SIGN '#'
671     #define CHAR_DOLLAR_SIGN '$'
672     #define CHAR_PERCENT_SIGN '%'
673     #define CHAR_AMPERSAND '&'
674     #define CHAR_APOSTROPHE '\''
675     #define CHAR_LEFT_PARENTHESIS '('
676     #define CHAR_RIGHT_PARENTHESIS ')'
677     #define CHAR_ASTERISK '*'
678     #define CHAR_PLUS '+'
679     #define CHAR_COMMA ','
680     #define CHAR_MINUS '-'
681     #define CHAR_DOT '.'
682     #define CHAR_SLASH '/'
683     #define CHAR_0 '0'
684     #define CHAR_1 '1'
685     #define CHAR_2 '2'
686     #define CHAR_3 '3'
687     #define CHAR_4 '4'
688     #define CHAR_5 '5'
689     #define CHAR_6 '6'
690     #define CHAR_7 '7'
691     #define CHAR_8 '8'
692     #define CHAR_9 '9'
693     #define CHAR_COLON ':'
694     #define CHAR_SEMICOLON ';'
695     #define CHAR_LESS_THAN_SIGN '<'
696     #define CHAR_EQUALS_SIGN '='
697     #define CHAR_GREATER_THAN_SIGN '>'
698     #define CHAR_QUESTION_MARK '?'
699     #define CHAR_COMMERCIAL_AT '@'
700     #define CHAR_A 'A'
701     #define CHAR_B 'B'
702     #define CHAR_C 'C'
703     #define CHAR_D 'D'
704     #define CHAR_E 'E'
705     #define CHAR_F 'F'
706     #define CHAR_G 'G'
707     #define CHAR_H 'H'
708     #define CHAR_I 'I'
709     #define CHAR_J 'J'
710     #define CHAR_K 'K'
711     #define CHAR_L 'L'
712     #define CHAR_M 'M'
713     #define CHAR_N 'N'
714     #define CHAR_O 'O'
715     #define CHAR_P 'P'
716     #define CHAR_Q 'Q'
717     #define CHAR_R 'R'
718     #define CHAR_S 'S'
719     #define CHAR_T 'T'
720     #define CHAR_U 'U'
721     #define CHAR_V 'V'
722     #define CHAR_W 'W'
723     #define CHAR_X 'X'
724     #define CHAR_Y 'Y'
725     #define CHAR_Z 'Z'
726     #define CHAR_LEFT_SQUARE_BRACKET '['
727     #define CHAR_BACKSLASH '\\'
728     #define CHAR_RIGHT_SQUARE_BRACKET ']'
729     #define CHAR_CIRCUMFLEX_ACCENT '^'
730     #define CHAR_UNDERSCORE '_'
731     #define CHAR_GRAVE_ACCENT '`'
732     #define CHAR_a 'a'
733     #define CHAR_b 'b'
734     #define CHAR_c 'c'
735     #define CHAR_d 'd'
736     #define CHAR_e 'e'
737     #define CHAR_f 'f'
738     #define CHAR_g 'g'
739     #define CHAR_h 'h'
740     #define CHAR_i 'i'
741     #define CHAR_j 'j'
742     #define CHAR_k 'k'
743     #define CHAR_l 'l'
744     #define CHAR_m 'm'
745     #define CHAR_n 'n'
746     #define CHAR_o 'o'
747     #define CHAR_p 'p'
748     #define CHAR_q 'q'
749     #define CHAR_r 'r'
750     #define CHAR_s 's'
751     #define CHAR_t 't'
752     #define CHAR_u 'u'
753     #define CHAR_v 'v'
754     #define CHAR_w 'w'
755     #define CHAR_x 'x'
756     #define CHAR_y 'y'
757     #define CHAR_z 'z'
758     #define CHAR_LEFT_CURLY_BRACKET '{'
759     #define CHAR_VERTICAL_LINE '|'
760     #define CHAR_RIGHT_CURLY_BRACKET '}'
761     #define CHAR_TILDE '~'
762    
763     #define STR_HT "\t"
764     #define STR_VT "\v"
765     #define STR_FF "\f"
766     #define STR_CR "\r"
767     #define STR_NL "\n"
768     #define STR_BS "\b"
769     #define STR_BEL "\a"
770     #ifdef EBCDIC
771     #define STR_ESC "\047"
772     #define STR_DEL "\007"
773     #else
774     #define STR_ESC "\033"
775     #define STR_DEL "\177"
776     #endif
777    
778     #define STR_SPACE " "
779     #define STR_EXCLAMATION_MARK "!"
780     #define STR_QUOTATION_MARK "\""
781     #define STR_NUMBER_SIGN "#"
782     #define STR_DOLLAR_SIGN "$"
783     #define STR_PERCENT_SIGN "%"
784     #define STR_AMPERSAND "&"
785     #define STR_APOSTROPHE "'"
786     #define STR_LEFT_PARENTHESIS "("
787     #define STR_RIGHT_PARENTHESIS ")"
788     #define STR_ASTERISK "*"
789     #define STR_PLUS "+"
790     #define STR_COMMA ","
791     #define STR_MINUS "-"
792     #define STR_DOT "."
793     #define STR_SLASH "/"
794     #define STR_0 "0"
795     #define STR_1 "1"
796     #define STR_2 "2"
797     #define STR_3 "3"
798     #define STR_4 "4"
799     #define STR_5 "5"
800     #define STR_6 "6"
801     #define STR_7 "7"
802     #define STR_8 "8"
803     #define STR_9 "9"
804     #define STR_COLON ":"
805     #define STR_SEMICOLON ";"
806     #define STR_LESS_THAN_SIGN "<"
807     #define STR_EQUALS_SIGN "="
808     #define STR_GREATER_THAN_SIGN ">"
809     #define STR_QUESTION_MARK "?"
810     #define STR_COMMERCIAL_AT "@"
811     #define STR_A "A"
812     #define STR_B "B"
813     #define STR_C "C"
814     #define STR_D "D"
815     #define STR_E "E"
816     #define STR_F "F"
817     #define STR_G "G"
818     #define STR_H "H"
819     #define STR_I "I"
820     #define STR_J "J"
821     #define STR_K "K"
822     #define STR_L "L"
823     #define STR_M "M"
824     #define STR_N "N"
825     #define STR_O "O"
826     #define STR_P "P"
827     #define STR_Q "Q"
828     #define STR_R "R"
829     #define STR_S "S"
830     #define STR_T "T"
831     #define STR_U "U"
832     #define STR_V "V"
833     #define STR_W "W"
834     #define STR_X "X"
835     #define STR_Y "Y"
836     #define STR_Z "Z"
837     #define STR_LEFT_SQUARE_BRACKET "["
838     #define STR_BACKSLASH "\\"
839     #define STR_RIGHT_SQUARE_BRACKET "]"
840     #define STR_CIRCUMFLEX_ACCENT "^"
841     #define STR_UNDERSCORE "_"
842     #define STR_GRAVE_ACCENT "`"
843     #define STR_a "a"
844     #define STR_b "b"
845     #define STR_c "c"
846     #define STR_d "d"
847     #define STR_e "e"
848     #define STR_f "f"
849     #define STR_g "g"
850     #define STR_h "h"
851     #define STR_i "i"
852     #define STR_j "j"
853     #define STR_k "k"
854     #define STR_l "l"
855     #define STR_m "m"
856     #define STR_n "n"
857     #define STR_o "o"
858     #define STR_p "p"
859     #define STR_q "q"
860     #define STR_r "r"
861     #define STR_s "s"
862     #define STR_t "t"
863     #define STR_u "u"
864     #define STR_v "v"
865     #define STR_w "w"
866     #define STR_x "x"
867     #define STR_y "y"
868     #define STR_z "z"
869     #define STR_LEFT_CURLY_BRACKET "{"
870     #define STR_VERTICAL_LINE "|"
871     #define STR_RIGHT_CURLY_BRACKET "}"
872     #define STR_TILDE "~"
873    
874     #define STRING_ACCEPT0 "ACCEPT\0"
875     #define STRING_COMMIT0 "COMMIT\0"
876     #define STRING_F0 "F\0"
877     #define STRING_FAIL0 "FAIL\0"
878     #define STRING_PRUNE0 "PRUNE\0"
879     #define STRING_SKIP0 "SKIP\0"
880     #define STRING_THEN "THEN"
881    
882     #define STRING_alpha0 "alpha\0"
883     #define STRING_lower0 "lower\0"
884     #define STRING_upper0 "upper\0"
885     #define STRING_alnum0 "alnum\0"
886     #define STRING_ascii0 "ascii\0"
887     #define STRING_blank0 "blank\0"
888     #define STRING_cntrl0 "cntrl\0"
889     #define STRING_digit0 "digit\0"
890     #define STRING_graph0 "graph\0"
891     #define STRING_print0 "print\0"
892     #define STRING_punct0 "punct\0"
893     #define STRING_space0 "space\0"
894     #define STRING_word0 "word\0"
895     #define STRING_xdigit "xdigit"
896    
897     #define STRING_DEFINE "DEFINE"
898    
899     #define STRING_CR_RIGHTPAR "CR)"
900     #define STRING_LF_RIGHTPAR "LF)"
901     #define STRING_CRLF_RIGHTPAR "CRLF)"
902     #define STRING_ANY_RIGHTPAR "ANY)"
903     #define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)"
904     #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"
905     #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"
906 ph10 412 #define STRING_UTF8_RIGHTPAR "UTF8)"
907 ph10 391
908     #else /* SUPPORT_UTF8 */
909    
910     /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
911     works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
912     only. */
913    
914     #define CHAR_HT '\011'
915     #define CHAR_VT '\013'
916     #define CHAR_FF '\014'
917     #define CHAR_CR '\015'
918     #define CHAR_NL '\012'
919     #define CHAR_BS '\010'
920     #define CHAR_BEL '\007'
921     #define CHAR_ESC '\033'
922     #define CHAR_DEL '\177'
923    
924     #define CHAR_SPACE '\040'
925     #define CHAR_EXCLAMATION_MARK '\041'
926     #define CHAR_QUOTATION_MARK '\042'
927     #define CHAR_NUMBER_SIGN '\043'
928     #define CHAR_DOLLAR_SIGN '\044'
929     #define CHAR_PERCENT_SIGN '\045'
930     #define CHAR_AMPERSAND '\046'
931     #define CHAR_APOSTROPHE '\047'
932     #define CHAR_LEFT_PARENTHESIS '\050'
933     #define CHAR_RIGHT_PARENTHESIS '\051'
934     #define CHAR_ASTERISK '\052'
935     #define CHAR_PLUS '\053'
936     #define CHAR_COMMA '\054'
937     #define CHAR_MINUS '\055'
938     #define CHAR_DOT '\056'
939     #define CHAR_SLASH '\057'
940     #define CHAR_0 '\060'
941     #define CHAR_1 '\061'
942     #define CHAR_2 '\062'
943     #define CHAR_3 '\063'
944     #define CHAR_4 '\064'
945     #define CHAR_5 '\065'
946     #define CHAR_6 '\066'
947     #define CHAR_7 '\067'
948     #define CHAR_8 '\070'
949     #define CHAR_9 '\071'
950     #define CHAR_COLON '\072'
951     #define CHAR_SEMICOLON '\073'
952     #define CHAR_LESS_THAN_SIGN '\074'
953     #define CHAR_EQUALS_SIGN '\075'
954     #define CHAR_GREATER_THAN_SIGN '\076'
955     #define CHAR_QUESTION_MARK '\077'
956     #define CHAR_COMMERCIAL_AT '\100'
957     #define CHAR_A '\101'
958     #define CHAR_B '\102'
959     #define CHAR_C '\103'
960     #define CHAR_D '\104'
961     #define CHAR_E '\105'
962     #define CHAR_F '\106'
963     #define CHAR_G '\107'
964     #define CHAR_H '\110'
965     #define CHAR_I '\111'
966     #define CHAR_J '\112'
967     #define CHAR_K '\113'
968     #define CHAR_L '\114'
969     #define CHAR_M '\115'
970     #define CHAR_N '\116'
971     #define CHAR_O '\117'
972     #define CHAR_P '\120'
973     #define CHAR_Q '\121'
974     #define CHAR_R '\122'
975     #define CHAR_S '\123'
976     #define CHAR_T '\124'
977     #define CHAR_U '\125'
978     #define CHAR_V '\126'
979     #define CHAR_W '\127'
980     #define CHAR_X '\130'
981     #define CHAR_Y '\131'
982     #define CHAR_Z '\132'
983     #define CHAR_LEFT_SQUARE_BRACKET '\133'
984     #define CHAR_BACKSLASH '\134'
985     #define CHAR_RIGHT_SQUARE_BRACKET '\135'
986     #define CHAR_CIRCUMFLEX_ACCENT '\136'
987     #define CHAR_UNDERSCORE '\137'
988     #define CHAR_GRAVE_ACCENT '\140'
989     #define CHAR_a '\141'
990     #define CHAR_b '\142'
991     #define CHAR_c '\143'
992     #define CHAR_d '\144'
993     #define CHAR_e '\145'
994     #define CHAR_f '\146'
995     #define CHAR_g '\147'
996     #define CHAR_h '\150'
997     #define CHAR_i '\151'
998     #define CHAR_j '\152'
999     #define CHAR_k '\153'
1000     #define CHAR_l '\154'
1001     #define CHAR_m '\155'
1002     #define CHAR_n '\156'
1003     #define CHAR_o '\157'
1004     #define CHAR_p '\160'
1005     #define CHAR_q '\161'
1006     #define CHAR_r '\162'
1007     #define CHAR_s '\163'
1008     #define CHAR_t '\164'
1009     #define CHAR_u '\165'
1010     #define CHAR_v '\166'
1011     #define CHAR_w '\167'
1012     #define CHAR_x '\170'
1013     #define CHAR_y '\171'
1014     #define CHAR_z '\172'
1015     #define CHAR_LEFT_CURLY_BRACKET '\173'
1016     #define CHAR_VERTICAL_LINE '\174'
1017     #define CHAR_RIGHT_CURLY_BRACKET '\175'
1018     #define CHAR_TILDE '\176'
1019    
1020     #define STR_HT "\011"
1021     #define STR_VT "\013"
1022     #define STR_FF "\014"
1023     #define STR_CR "\015"
1024     #define STR_NL "\012"
1025     #define STR_BS "\010"
1026     #define STR_BEL "\007"
1027     #define STR_ESC "\033"
1028     #define STR_DEL "\177"
1029    
1030     #define STR_SPACE "\040"
1031     #define STR_EXCLAMATION_MARK "\041"
1032     #define STR_QUOTATION_MARK "\042"
1033     #define STR_NUMBER_SIGN "\043"
1034     #define STR_DOLLAR_SIGN "\044"
1035     #define STR_PERCENT_SIGN "\045"
1036     #define STR_AMPERSAND "\046"
1037     #define STR_APOSTROPHE "\047"
1038     #define STR_LEFT_PARENTHESIS "\050"
1039     #define STR_RIGHT_PARENTHESIS "\051"
1040     #define STR_ASTERISK "\052"
1041     #define STR_PLUS "\053"
1042     #define STR_COMMA "\054"
1043     #define STR_MINUS "\055"
1044     #define STR_DOT "\056"
1045     #define STR_SLASH "\057"
1046     #define STR_0 "\060"
1047     #define STR_1 "\061"
1048     #define STR_2 "\062"
1049     #define STR_3 "\063"
1050     #define STR_4 "\064"
1051     #define STR_5 "\065"
1052     #define STR_6 "\066"
1053     #define STR_7 "\067"
1054     #define STR_8 "\070"
1055     #define STR_9 "\071"
1056     #define STR_COLON "\072"
1057     #define STR_SEMICOLON "\073"
1058     #define STR_LESS_THAN_SIGN "\074"
1059     #define STR_EQUALS_SIGN "\075"
1060     #define STR_GREATER_THAN_SIGN "\076"
1061     #define STR_QUESTION_MARK "\077"
1062     #define STR_COMMERCIAL_AT "\100"
1063     #define STR_A "\101"
1064     #define STR_B "\102"
1065     #define STR_C "\103"
1066     #define STR_D "\104"
1067     #define STR_E "\105"
1068     #define STR_F "\106"
1069     #define STR_G "\107"
1070     #define STR_H "\110"
1071     #define STR_I "\111"
1072     #define STR_J "\112"
1073     #define STR_K "\113"
1074     #define STR_L "\114"
1075     #define STR_M "\115"
1076     #define STR_N "\116"
1077     #define STR_O "\117"
1078     #define STR_P "\120"
1079     #define STR_Q "\121"
1080     #define STR_R "\122"
1081     #define STR_S "\123"
1082     #define STR_T "\124"
1083     #define STR_U "\125"
1084     #define STR_V "\126"
1085     #define STR_W "\127"
1086     #define STR_X "\130"
1087     #define STR_Y "\131"
1088     #define STR_Z "\132"
1089     #define STR_LEFT_SQUARE_BRACKET "\133"
1090     #define STR_BACKSLASH "\134"
1091     #define STR_RIGHT_SQUARE_BRACKET "\135"
1092     #define STR_CIRCUMFLEX_ACCENT "\136"
1093     #define STR_UNDERSCORE "\137"
1094     #define STR_GRAVE_ACCENT "\140"
1095     #define STR_a "\141"
1096     #define STR_b "\142"
1097     #define STR_c "\143"
1098     #define STR_d "\144"
1099     #define STR_e "\145"
1100     #define STR_f "\146"
1101     #define STR_g "\147"
1102     #define STR_h "\150"
1103     #define STR_i "\151"
1104     #define STR_j "\152"
1105     #define STR_k "\153"
1106     #define STR_l "\154"
1107     #define STR_m "\155"
1108     #define STR_n "\156"
1109     #define STR_o "\157"
1110     #define STR_p "\160"
1111     #define STR_q "\161"
1112     #define STR_r "\162"
1113     #define STR_s "\163"
1114     #define STR_t "\164"
1115     #define STR_u "\165"
1116     #define STR_v "\166"
1117     #define STR_w "\167"
1118     #define STR_x "\170"
1119     #define STR_y "\171"
1120     #define STR_z "\172"
1121     #define STR_LEFT_CURLY_BRACKET "\173"
1122     #define STR_VERTICAL_LINE "\174"
1123     #define STR_RIGHT_CURLY_BRACKET "\175"
1124     #define STR_TILDE "\176"
1125    
1126     #define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1127     #define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1128     #define STRING_F0 STR_F "\0"
1129     #define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0"
1130     #define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0"
1131     #define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0"
1132     #define STRING_THEN STR_T STR_H STR_E STR_N
1133    
1134     #define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0"
1135     #define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0"
1136     #define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0"
1137     #define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0"
1138     #define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0"
1139     #define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0"
1140     #define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0"
1141     #define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0"
1142     #define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0"
1143     #define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0"
1144     #define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0"
1145     #define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0"
1146     #define STRING_word0 STR_w STR_o STR_r STR_d "\0"
1147     #define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t
1148    
1149     #define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E
1150    
1151     #define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS
1152     #define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS
1153     #define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1154     #define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1155     #define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1156     #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
1157     #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
1158 ph10 412 #define STRING_UTF8_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1159 ph10 391
1160     #endif /* SUPPORT_UTF8 */
1161    
1162 nigel 93 /* Escape items that are just an encoding of a particular data value. */
1163 nigel 77
1164     #ifndef ESC_e
1165 ph10 391 #define ESC_e CHAR_ESC
1166 nigel 77 #endif
1167    
1168     #ifndef ESC_f
1169 ph10 391 #define ESC_f CHAR_FF
1170 nigel 77 #endif
1171    
1172     #ifndef ESC_n
1173 ph10 391 #define ESC_n CHAR_NL
1174 nigel 77 #endif
1175    
1176     #ifndef ESC_r
1177 ph10 391 #define ESC_r CHAR_CR
1178 nigel 77 #endif
1179    
1180     /* We can't officially use ESC_t because it is a POSIX reserved identifier
1181     (presumably because of all the others like size_t). */
1182    
1183     #ifndef ESC_tee
1184 ph10 391 #define ESC_tee CHAR_HT
1185 nigel 77 #endif
1186    
1187 nigel 87 /* Codes for different types of Unicode property */
1188    
1189     #define PT_ANY 0 /* Any property - matches all chars */
1190     #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
1191     #define PT_GC 2 /* General characteristic (e.g. L) */
1192     #define PT_PC 3 /* Particular characteristic (e.g. Lu) */
1193     #define PT_SC 4 /* Script (e.g. Han) */
1194    
1195     /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1196     contain UTF-8 characters with values greater than 255. */
1197    
1198     #define XCL_NOT 0x01 /* Flag: this is a negative class */
1199     #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
1200    
1201     #define XCL_END 0 /* Marks end of individual items */
1202     #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
1203     #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
1204     #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
1205     #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
1206    
1207 nigel 77 /* These are escaped items that aren't just an encoding of a particular data
1208     value such as \n. They must have non-zero values, as check_escape() returns
1209     their negation. Also, they must appear in the same order as in the opcode
1210     definitions below, up to ESC_z. There's a dummy for OP_ANY because it
1211 ph10 345 corresponds to "." rather than an escape sequence, and another for OP_ALLANY
1212 ph10 341 (which is used for [^] in JavaScript compatibility mode).
1213 nigel 77
1214 ph10 341 The final escape must be ESC_REF as subsequent values are used for
1215     backreferences (\1, \2, \3, etc). There are two tests in the code for an escape
1216     greater than ESC_b and less than ESC_Z to detect the types that may be
1217     repeated. These are the types that consume characters. If any new escapes are
1218     put in between that don't consume a character, that code will have to change.
1219     */
1220    
1221 ph10 172 enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1222 ph10 345 ESC_W, ESC_w, ESC_dum1, ESC_dum2, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1223     ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_g, ESC_k,
1224 ph10 341 ESC_REF };
1225 nigel 77
1226 nigel 93
1227 ph10 210 /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1228 nigel 77 OP_EOD must correspond in order to the list of escapes immediately above.
1229    
1230 nigel 93 *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1231 ph10 462 that follow must also be updated to match. There are also tables called
1232     "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1233 nigel 93
1234 nigel 77 enum {
1235     OP_END, /* 0 End of pattern */
1236    
1237     /* Values corresponding to backslashed metacharacters */
1238    
1239     OP_SOD, /* 1 Start of data: \A */
1240     OP_SOM, /* 2 Start of match (subject + offset): \G */
1241 ph10 172 OP_SET_SOM, /* 3 Set start of match (\K) */
1242 ph10 168 OP_NOT_WORD_BOUNDARY, /* 4 \B */
1243     OP_WORD_BOUNDARY, /* 5 \b */
1244     OP_NOT_DIGIT, /* 6 \D */
1245     OP_DIGIT, /* 7 \d */
1246     OP_NOT_WHITESPACE, /* 8 \S */
1247     OP_WHITESPACE, /* 9 \s */
1248     OP_NOT_WORDCHAR, /* 10 \W */
1249     OP_WORDCHAR, /* 11 \w */
1250 ph10 341 OP_ANY, /* 12 Match any character (subject to DOTALL) */
1251     OP_ALLANY, /* 13 Match any character (not subject to DOTALL) */
1252     OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1253     OP_NOTPROP, /* 15 \P (not Unicode property) */
1254     OP_PROP, /* 16 \p (Unicode property) */
1255     OP_ANYNL, /* 17 \R (any newline sequence) */
1256     OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */
1257     OP_HSPACE, /* 19 \h (horizontal whitespace) */
1258     OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */
1259     OP_VSPACE, /* 21 \v (vertical whitespace) */
1260     OP_EXTUNI, /* 22 \X (extended Unicode sequence */
1261     OP_EODN, /* 23 End of data or \n at end of data: \Z. */
1262     OP_EOD, /* 24 End of data: \z */
1263 nigel 77
1264 ph10 341 OP_OPT, /* 25 Set runtime options */
1265     OP_CIRC, /* 26 Start of line - varies with multiline switch */
1266     OP_DOLL, /* 27 End of line - varies with multiline switch */
1267     OP_CHAR, /* 28 Match one character, casefully */
1268     OP_CHARNC, /* 29 Match one character, caselessly */
1269     OP_NOT, /* 30 Match one character, not the following one */
1270 nigel 77
1271 ph10 341 OP_STAR, /* 31 The maximizing and minimizing versions of */
1272     OP_MINSTAR, /* 32 these six opcodes must come in pairs, with */
1273     OP_PLUS, /* 33 the minimizing one second. */
1274     OP_MINPLUS, /* 34 This first set applies to single characters.*/
1275     OP_QUERY, /* 35 */
1276     OP_MINQUERY, /* 36 */
1277 nigel 77
1278 ph10 341 OP_UPTO, /* 37 From 0 to n matches */
1279     OP_MINUPTO, /* 38 */
1280     OP_EXACT, /* 39 Exactly n matches */
1281 nigel 77
1282 ph10 341 OP_POSSTAR, /* 40 Possessified star */
1283     OP_POSPLUS, /* 41 Possessified plus */
1284     OP_POSQUERY, /* 42 Posesssified query */
1285     OP_POSUPTO, /* 43 Possessified upto */
1286 nigel 77
1287 ph10 341 OP_NOTSTAR, /* 44 The maximizing and minimizing versions of */
1288     OP_NOTMINSTAR, /* 45 these six opcodes must come in pairs, with */
1289     OP_NOTPLUS, /* 46 the minimizing one second. They must be in */
1290     OP_NOTMINPLUS, /* 47 exactly the same order as those above. */
1291     OP_NOTQUERY, /* 48 This set applies to "not" single characters. */
1292     OP_NOTMINQUERY, /* 49 */
1293 nigel 77
1294 ph10 341 OP_NOTUPTO, /* 50 From 0 to n matches */
1295     OP_NOTMINUPTO, /* 51 */
1296     OP_NOTEXACT, /* 52 Exactly n matches */
1297 nigel 93
1298 ph10 341 OP_NOTPOSSTAR, /* 53 Possessified versions */
1299     OP_NOTPOSPLUS, /* 54 */
1300     OP_NOTPOSQUERY, /* 55 */
1301     OP_NOTPOSUPTO, /* 56 */
1302 nigel 93
1303 ph10 341 OP_TYPESTAR, /* 57 The maximizing and minimizing versions of */
1304     OP_TYPEMINSTAR, /* 58 these six opcodes must come in pairs, with */
1305     OP_TYPEPLUS, /* 59 the minimizing one second. These codes must */
1306     OP_TYPEMINPLUS, /* 60 be in exactly the same order as those above. */
1307     OP_TYPEQUERY, /* 61 This set applies to character types such as \d */
1308     OP_TYPEMINQUERY, /* 62 */
1309 nigel 93
1310 ph10 341 OP_TYPEUPTO, /* 63 From 0 to n matches */
1311     OP_TYPEMINUPTO, /* 64 */
1312     OP_TYPEEXACT, /* 65 Exactly n matches */
1313 nigel 93
1314 ph10 341 OP_TYPEPOSSTAR, /* 66 Possessified versions */
1315     OP_TYPEPOSPLUS, /* 67 */
1316     OP_TYPEPOSQUERY, /* 68 */
1317     OP_TYPEPOSUPTO, /* 69 */
1318 nigel 93
1319 ph10 341 OP_CRSTAR, /* 70 The maximizing and minimizing versions of */
1320     OP_CRMINSTAR, /* 71 all these opcodes must come in pairs, with */
1321     OP_CRPLUS, /* 72 the minimizing one second. These codes must */
1322     OP_CRMINPLUS, /* 73 be in exactly the same order as those above. */
1323     OP_CRQUERY, /* 74 These are for character classes and back refs */
1324     OP_CRMINQUERY, /* 75 */
1325     OP_CRRANGE, /* 76 These are different to the three sets above. */
1326     OP_CRMINRANGE, /* 77 */
1327 nigel 93
1328 ph10 341 OP_CLASS, /* 78 Match a character class, chars < 256 only */
1329     OP_NCLASS, /* 79 Same, but the bitmap was created from a negative
1330 nigel 77 class - the difference is relevant only when a UTF-8
1331     character > 255 is encountered. */
1332    
1333 ph10 341 OP_XCLASS, /* 80 Extended class for handling UTF-8 chars within the
1334 nigel 77 class. This does both positive and negative. */
1335    
1336 ph10 341 OP_REF, /* 81 Match a back reference */
1337     OP_RECURSE, /* 82 Match a numbered subpattern (possibly recursive) */
1338     OP_CALLOUT, /* 83 Call out to external function if provided */
1339 nigel 77
1340 ph10 341 OP_ALT, /* 84 Start of alternation */
1341     OP_KET, /* 85 End of group that doesn't have an unbounded repeat */
1342     OP_KETRMAX, /* 86 These two must remain together and in this */
1343     OP_KETRMIN, /* 87 order. They are for groups the repeat for ever. */
1344 nigel 77
1345 nigel 93 /* The assertions must come before BRA, CBRA, ONCE, and COND.*/
1346 nigel 77
1347 ph10 341 OP_ASSERT, /* 88 Positive lookahead */
1348     OP_ASSERT_NOT, /* 89 Negative lookahead */
1349     OP_ASSERTBACK, /* 90 Positive lookbehind */
1350     OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */
1351     OP_REVERSE, /* 92 Move pointer back - used in lookbehind assertions */
1352 nigel 77
1353 nigel 93 /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,
1354     as there's a test for >= ONCE for a subpattern that isn't an assertion. */
1355 nigel 77
1356 ph10 341 OP_ONCE, /* 93 Atomic group */
1357     OP_BRA, /* 94 Start of non-capturing bracket */
1358     OP_CBRA, /* 95 Start of capturing bracket */
1359     OP_COND, /* 96 Conditional group */
1360 nigel 77
1361 nigel 93 /* These three must follow the previous three, in the same order. There's a
1362     check for >= SBRA to distinguish the two sets. */
1363 nigel 77
1364 ph10 341 OP_SBRA, /* 97 Start of non-capturing bracket, check empty */
1365     OP_SCBRA, /* 98 Start of capturing bracket, check empty */
1366     OP_SCOND, /* 99 Conditional group, check empty */
1367 nigel 77
1368 ph10 459 /* The next two pairs must (respectively) be kept together. */
1369 ph10 461
1370 ph10 341 OP_CREF, /* 100 Used to hold a capture number as condition */
1371 ph10 459 OP_NCREF, /* 101 Same, but generaged by a name reference*/
1372     OP_RREF, /* 102 Used to hold a recursion number as condition */
1373     OP_NRREF, /* 103 Same, but generaged by a name reference*/
1374     OP_DEF, /* 104 The DEFINE condition */
1375 nigel 93
1376 ph10 459 OP_BRAZERO, /* 105 These two must remain together and in this */
1377     OP_BRAMINZERO, /* 106 order. */
1378 ph10 210
1379     /* These are backtracking control verbs */
1380    
1381 ph10 459 OP_PRUNE, /* 107 */
1382     OP_SKIP, /* 108 */
1383     OP_THEN, /* 109 */
1384     OP_COMMIT, /* 110 */
1385 ph10 210
1386     /* These are forced failure and success verbs */
1387    
1388 ph10 459 OP_FAIL, /* 111 */
1389     OP_ACCEPT, /* 112 */
1390     OP_CLOSE, /* 113 Used before OP_ACCEPT to close open captures */
1391 ph10 345
1392 ph10 335 /* This is used to skip a subpattern with a {0} quantifier */
1393 ph10 345
1394 ph10 459 OP_SKIPZERO /* 114 */
1395 nigel 77 };
1396    
1397 ph10 462 /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
1398     definitions that follow must also be updated to match. There are also tables
1399     called "coptable" cna "poptable" in pcre_dfa_exec.c that must be updated. */
1400 nigel 77
1401 ph10 462
1402 nigel 77 /* This macro defines textual names for all the opcodes. These are used only
1403     for debugging. The macro is referenced only in pcre_printint.c. */
1404    
1405     #define OP_NAME_LIST \
1406 ph10 168 "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \
1407 ph10 341 "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \
1408 ph10 178 "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \
1409     "extuni", "\\Z", "\\z", \
1410 nigel 77 "Opt", "^", "$", "char", "charnc", "not", \
1411     "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1412 nigel 93 "*+","++", "?+", "{", \
1413 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1414 nigel 93 "*+","++", "?+", "{", \
1415 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1416 nigel 93 "*+","++", "?+", "{", \
1417 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", \
1418     "class", "nclass", "xclass", "Ref", "Recurse", "Callout", \
1419     "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", \
1420 nigel 93 "AssertB", "AssertB not", "Reverse", \
1421 ph10 192 "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond", \
1422 ph10 459 "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def", \
1423     "Brazero", "Braminzero", \
1424 ph10 335 "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT", \
1425 ph10 447 "Close", "Skip zero"
1426 nigel 77
1427    
1428     /* This macro defines the length of fixed length operations in the compiled
1429     regex. The lengths are used when searching for specific things, and also in the
1430     debugging printing of a compiled regex. We use a macro so that it can be
1431     defined close to the definitions of the opcodes themselves.
1432    
1433     As things have been extended, some of these are no longer fixed lenths, but are
1434     minima instead. For example, the length of a single-character repeat may vary
1435     in UTF-8 mode. The code that uses this table must know about such things. */
1436    
1437     #define OP_LENGTHS \
1438     1, /* End */ \
1439 ph10 168 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \
1440     1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \
1441 ph10 341 1, 1, 1, /* Any, AllAny, Anybyte */ \
1442 ph10 178 3, 3, 1, /* NOTPROP, PROP, EXTUNI */ \
1443     1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \
1444 nigel 77 1, 1, 2, 1, 1, /* \Z, \z, Opt, ^, $ */ \
1445     2, /* Char - the minimum length */ \
1446     2, /* Charnc - the minimum length */ \
1447     2, /* not */ \
1448     /* Positive single-char repeats ** These are */ \
1449     2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
1450     4, 4, 4, /* upto, minupto, exact ** UTF-8 mode */ \
1451 nigel 93 2, 2, 2, 4, /* *+, ++, ?+, upto+ */ \
1452 nigel 77 /* Negative single-char repeats - only for chars < 256 */ \
1453     2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
1454     4, 4, 4, /* NOT upto, minupto, exact */ \
1455 nigel 93 2, 2, 2, 4, /* Possessive *, +, ?, upto */ \
1456 nigel 77 /* Positive type repeats */ \
1457     2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
1458     4, 4, 4, /* Type upto, minupto, exact */ \
1459 nigel 93 2, 2, 2, 4, /* Possessive *+, ++, ?+, upto+ */ \
1460 nigel 77 /* Character class & ref repeats */ \
1461     1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
1462     5, 5, /* CRRANGE, CRMINRANGE */ \
1463     33, /* CLASS */ \
1464     33, /* NCLASS */ \
1465     0, /* XCLASS - variable length */ \
1466     3, /* REF */ \
1467     1+LINK_SIZE, /* RECURSE */ \
1468     2+2*LINK_SIZE, /* CALLOUT */ \
1469     1+LINK_SIZE, /* Alt */ \
1470     1+LINK_SIZE, /* Ket */ \
1471     1+LINK_SIZE, /* KetRmax */ \
1472     1+LINK_SIZE, /* KetRmin */ \
1473     1+LINK_SIZE, /* Assert */ \
1474     1+LINK_SIZE, /* Assert not */ \
1475     1+LINK_SIZE, /* Assert behind */ \
1476     1+LINK_SIZE, /* Assert behind not */ \
1477     1+LINK_SIZE, /* Reverse */ \
1478 nigel 93 1+LINK_SIZE, /* ONCE */ \
1479     1+LINK_SIZE, /* BRA */ \
1480     3+LINK_SIZE, /* CBRA */ \
1481 nigel 77 1+LINK_SIZE, /* COND */ \
1482 nigel 93 1+LINK_SIZE, /* SBRA */ \
1483     3+LINK_SIZE, /* SCBRA */ \
1484     1+LINK_SIZE, /* SCOND */ \
1485 ph10 459 3, 3, /* CREF, NCREF */ \
1486     3, 3, /* RREF, NRREF */ \
1487 nigel 93 1, /* DEF */ \
1488 nigel 77 1, 1, /* BRAZERO, BRAMINZERO */ \
1489 ph10 210 1, 1, 1, 1, /* PRUNE, SKIP, THEN, COMMIT, */ \
1490 ph10 447 1, 1, 3, 1 /* FAIL, ACCEPT, CLOSE, SKIPZERO */
1491 nigel 77
1492    
1493 ph10 459 /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
1494     condition. */
1495 nigel 77
1496 nigel 93 #define RREF_ANY 0xffff
1497 nigel 77
1498     /* Error code numbers. They are given names so that they can more easily be
1499     tracked. */
1500    
1501     enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
1502     ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
1503     ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
1504     ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
1505 nigel 91 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
1506 ph10 210 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
1507 ph10 457 ERR60, ERR61, ERR62, ERR63, ERR64, ERR65 };
1508 nigel 77
1509     /* The real format of the start of the pcre block; the index of names and the
1510     code vector run on as long as necessary after the end. We store an explicit
1511     offset to the name table so that if a regex is compiled on one host, saved, and
1512     then run on another where the size of pointers is different, all might still
1513     be well. For the case of compiled-on-4 and run-on-8, we include an extra
1514     pointer that is always NULL. For future-proofing, a few dummy fields were
1515     originally included - even though you can never get this planning right - but
1516     there is only one left now.
1517    
1518     NOTE NOTE NOTE:
1519     Because people can now save and re-use compiled patterns, any additions to this
1520     structure should be made at the end, and something earlier (e.g. a new
1521     flag in the options or one of the dummy fields) should indicate that the new
1522     fields are present. Currently PCRE always sets the dummy fields to zero.
1523 ph10 455 NOTE NOTE NOTE
1524 nigel 77 */
1525    
1526     typedef struct real_pcre {
1527     pcre_uint32 magic_number;
1528     pcre_uint32 size; /* Total that was malloced */
1529 ph10 230 pcre_uint32 options; /* Public options */
1530     pcre_uint16 flags; /* Private flags */
1531     pcre_uint16 dummy1; /* For future use */
1532 nigel 77 pcre_uint16 top_bracket;
1533     pcre_uint16 top_backref;
1534     pcre_uint16 first_byte;
1535     pcre_uint16 req_byte;
1536     pcre_uint16 name_table_offset; /* Offset to name table that follows */
1537     pcre_uint16 name_entry_size; /* Size of any name items */
1538     pcre_uint16 name_count; /* Number of name items */
1539     pcre_uint16 ref_count; /* Reference count */
1540    
1541     const unsigned char *tables; /* Pointer to tables or NULL for std */
1542     const unsigned char *nullpad; /* NULL padding */
1543     } real_pcre;
1544    
1545     /* The format of the block used to store data from pcre_study(). The same
1546     remark (see NOTE above) about extending this structure applies. */
1547    
1548     typedef struct pcre_study_data {
1549     pcre_uint32 size; /* Total that was malloced */
1550 ph10 455 pcre_uint32 flags; /* Private flags */
1551     uschar start_bits[32]; /* Starting char bits */
1552 ph10 459 pcre_uint32 minlength; /* Minimum subject length */
1553 nigel 77 } pcre_study_data;
1554    
1555 ph10 459 /* Structure for building a chain of open capturing subpatterns during
1556     compiling, so that instructions to close them can be compiled when (*ACCEPT) is
1557 ph10 488 encountered. This is also used to identify subpatterns that contain recursive
1558     back references to themselves, so that they can be made atomic. */
1559 ph10 447
1560     typedef struct open_capitem {
1561     struct open_capitem *next; /* Chain link */
1562     pcre_uint16 number; /* Capture number */
1563 ph10 488 pcre_uint16 flag; /* Set TRUE if recursive back ref */
1564 ph10 459 } open_capitem;
1565 ph10 447
1566 nigel 77 /* Structure for passing "static" information around between the functions
1567     doing the compiling, so that they are thread-safe. */
1568    
1569     typedef struct compile_data {
1570     const uschar *lcc; /* Points to lower casing table */
1571     const uschar *fcc; /* Points to case-flipping table */
1572     const uschar *cbits; /* Points to character type table */
1573     const uschar *ctypes; /* Points to table of type maps */
1574 nigel 93 const uschar *start_workspace;/* The start of working space */
1575 nigel 77 const uschar *start_code; /* The start of the compiled code */
1576     const uschar *start_pattern; /* The start of the pattern */
1577 nigel 93 const uschar *end_pattern; /* The end of the pattern */
1578 ph10 459 open_capitem *open_caps; /* Chain of open capture items */
1579 nigel 93 uschar *hwm; /* High watermark of workspace */
1580 nigel 77 uschar *name_table; /* The name/number table */
1581     int names_found; /* Number of entries so far */
1582     int name_entry_size; /* Size of each entry */
1583 ph10 270 int bracount; /* Count of capturing parens as we compile */
1584 ph10 286 int final_bracount; /* Saved value after first pass */
1585 nigel 77 int top_backref; /* Maximum back reference */
1586     unsigned int backref_map; /* Bitmap of low back refs */
1587 nigel 93 int external_options; /* External (initial) options */
1588 ph10 231 int external_flags; /* External flag bits to be set */
1589 nigel 77 int req_varyopt; /* "After variable item" flag for reqbyte */
1590 ph10 211 BOOL had_accept; /* (*ACCEPT) encountered */
1591 ph10 459 BOOL check_lookbehind; /* Lookbehinds need later checking */
1592 nigel 93 int nltype; /* Newline type */
1593     int nllen; /* Newline string length */
1594     uschar nl[4]; /* Newline string when fixed length */
1595 nigel 77 } compile_data;
1596    
1597     /* Structure for maintaining a chain of pointers to the currently incomplete
1598     branches, for testing for left recursion. */
1599    
1600     typedef struct branch_chain {
1601     struct branch_chain *outer;
1602 ph10 475 uschar *current_branch;
1603 nigel 77 } branch_chain;
1604    
1605     /* Structure for items in a linked list that represents an explicit recursive
1606     call within the pattern. */
1607    
1608     typedef struct recursion_info {
1609     struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
1610     int group_num; /* Number of group that was called */
1611     const uschar *after_call; /* "Return value": points after the call in the expr */
1612 ph10 168 USPTR save_start; /* Old value of mstart */
1613 nigel 77 int *offset_save; /* Pointer to start of saved offsets */
1614     int saved_max; /* Number of saved offsets */
1615 ph10 461 int save_offset_top; /* Current value of offset_top */
1616 nigel 77 } recursion_info;
1617    
1618 nigel 93 /* Structure for building a chain of data for holding the values of the subject
1619     pointer at the start of each subpattern, so as to detect when an empty string
1620     has been matched by a subpattern - to break infinite loops. */
1621    
1622     typedef struct eptrblock {
1623     struct eptrblock *epb_prev;
1624     USPTR epb_saved_eptr;
1625     } eptrblock;
1626    
1627    
1628 nigel 77 /* Structure for passing "static" information around between the functions
1629     doing traditional NFA matching, so that they are thread-safe. */
1630    
1631     typedef struct match_data {
1632 nigel 87 unsigned long int match_call_count; /* As it says */
1633     unsigned long int match_limit; /* As it says */
1634     unsigned long int match_limit_recursion; /* As it says */
1635 nigel 77 int *offset_vector; /* Offset vector */
1636     int offset_end; /* One past the end */
1637     int offset_max; /* The maximum usable for return data */
1638 nigel 93 int nltype; /* Newline type */
1639     int nllen; /* Newline string length */
1640 ph10 459 int name_count; /* Number of names in name table */
1641     int name_entry_size; /* Size of entry in names table */
1642 ph10 461 uschar *name_table; /* Table of names */
1643 nigel 93 uschar nl[4]; /* Newline string when fixed */
1644 nigel 77 const uschar *lcc; /* Points to lower casing table */
1645     const uschar *ctypes; /* Points to table of type maps */
1646     BOOL offset_overflow; /* Set if too many extractions */
1647     BOOL notbol; /* NOTBOL flag */
1648     BOOL noteol; /* NOTEOL flag */
1649     BOOL utf8; /* UTF8 flag */
1650 ph10 345 BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */
1651 nigel 77 BOOL endonly; /* Dollar not before final \n */
1652     BOOL notempty; /* Empty string match not wanted */
1653 ph10 442 BOOL notempty_atstart; /* Empty string match at start not wanted */
1654 nigel 77 BOOL hitend; /* Hit the end of the subject at some point */
1655 ph10 231 BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
1656 nigel 77 const uschar *start_code; /* For use when recursing */
1657 nigel 87 USPTR start_subject; /* Start of the subject string */
1658     USPTR end_subject; /* End of the subject string */
1659 ph10 168 USPTR start_match_ptr; /* Start of matched string */
1660 nigel 87 USPTR end_match_ptr; /* Subject position at end match */
1661 ph10 442 USPTR start_used_ptr; /* Earliest consulted character */
1662 ph10 435 int partial; /* PARTIAL options */
1663 nigel 77 int end_offset_top; /* Highwater mark at end of match */
1664     int capture_last; /* Most recent capture number */
1665     int start_offset; /* The start offset value */
1666 nigel 93 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
1667     int eptrn; /* Next free eptrblock */
1668 nigel 77 recursion_info *recursive; /* Linked list of recursion data */
1669     void *callout_data; /* To pass back to callouts */
1670     } match_data;
1671    
1672     /* A similar structure is used for the same purpose by the DFA matching
1673     functions. */
1674    
1675     typedef struct dfa_match_data {
1676     const uschar *start_code; /* Start of the compiled pattern */
1677     const uschar *start_subject; /* Start of the subject string */
1678     const uschar *end_subject; /* End of subject string */
1679 ph10 442 const uschar *start_used_ptr; /* Earliest consulted character */
1680 nigel 77 const uschar *tables; /* Character tables */
1681 ph10 442 int start_offset; /* The start offset value */
1682 nigel 77 int moptions; /* Match options */
1683     int poptions; /* Pattern options */
1684 nigel 93 int nltype; /* Newline type */
1685     int nllen; /* Newline string length */
1686     uschar nl[4]; /* Newline string when fixed */
1687 nigel 77 void *callout_data; /* To pass back to callouts */
1688     } dfa_match_data;
1689    
1690     /* Bit definitions for entries in the pcre_ctypes table. */
1691    
1692     #define ctype_space 0x01
1693     #define ctype_letter 0x02
1694     #define ctype_digit 0x04
1695     #define ctype_xdigit 0x08
1696 ph10 274 #define ctype_word 0x10 /* alphanumeric or '_' */
1697 nigel 77 #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
1698    
1699     /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
1700     of bits for a class map. Some classes are built by combining these tables. */
1701    
1702     #define cbit_space 0 /* [:space:] or \s */
1703     #define cbit_xdigit 32 /* [:xdigit:] */
1704     #define cbit_digit 64 /* [:digit:] or \d */
1705     #define cbit_upper 96 /* [:upper:] */
1706     #define cbit_lower 128 /* [:lower:] */
1707     #define cbit_word 160 /* [:word:] or \w */
1708     #define cbit_graph 192 /* [:graph:] */
1709     #define cbit_print 224 /* [:print:] */
1710     #define cbit_punct 256 /* [:punct:] */
1711     #define cbit_cntrl 288 /* [:cntrl:] */
1712     #define cbit_length 320 /* Length of the cbits table */
1713    
1714     /* Offsets of the various tables from the base tables pointer, and
1715     total length. */
1716    
1717     #define lcc_offset 0
1718     #define fcc_offset 256
1719     #define cbits_offset 512
1720     #define ctypes_offset (cbits_offset + cbit_length)
1721     #define tables_length (ctypes_offset + 256)
1722    
1723 nigel 87 /* Layout of the UCP type table that translates property names into types and
1724 ph10 240 codes. Each entry used to point directly to a name, but to reduce the number of
1725 ph10 243 relocations in shared libraries, it now has an offset into a single string
1726 ph10 240 instead. */
1727 nigel 77
1728     typedef struct {
1729 ph10 243 pcre_uint16 name_offset;
1730 nigel 87 pcre_uint16 type;
1731     pcre_uint16 value;
1732 nigel 77 } ucp_type_table;
1733    
1734    
1735     /* Internal shared data tables. These are tables that are used by more than one
1736     of the exported public functions. They have to be "external" in the C sense,
1737     but are not part of the PCRE public API. The data for these tables is in the
1738     pcre_tables.c module. */
1739    
1740     extern const int _pcre_utf8_table1[];
1741     extern const int _pcre_utf8_table2[];
1742     extern const int _pcre_utf8_table3[];
1743     extern const uschar _pcre_utf8_table4[];
1744    
1745     extern const int _pcre_utf8_table1_size;
1746    
1747 ph10 240 extern const char _pcre_utt_names[];
1748 nigel 77 extern const ucp_type_table _pcre_utt[];
1749     extern const int _pcre_utt_size;
1750    
1751     extern const uschar _pcre_default_tables[];
1752    
1753     extern const uschar _pcre_OP_lengths[];
1754    
1755    
1756     /* Internal shared functions. These are functions that are used by more than
1757     one of the exported public functions. They have to be "external" in the C
1758     sense, but are not part of the PCRE public API. */
1759    
1760 ph10 455 extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);
1761 ph10 483 extern BOOL _pcre_is_newline(USPTR, int, USPTR, int *, BOOL);
1762 ph10 455 extern int _pcre_ord2utf8(int, uschar *);
1763     extern real_pcre *_pcre_try_flipped(const real_pcre *, real_pcre *,
1764     const pcre_study_data *, pcre_study_data *);
1765 ph10 483 extern int _pcre_valid_utf8(USPTR, int);
1766     extern BOOL _pcre_was_newline(USPTR, int, USPTR, int *, BOOL);
1767 ph10 455 extern BOOL _pcre_xclass(int, const uschar *);
1768 nigel 77
1769 ph10 349
1770     /* Unicode character database (UCD) */
1771    
1772     typedef struct {
1773     uschar script;
1774     uschar chartype;
1775 ph10 351 pcre_int32 other_case;
1776 ph10 349 } ucd_record;
1777    
1778 ph10 351 extern const ucd_record _pcre_ucd_records[];
1779     extern const uschar _pcre_ucd_stage1[];
1780     extern const pcre_uint16 _pcre_ucd_stage2[];
1781     extern const int _pcre_ucp_gentype[];
1782 ph10 349
1783 ph10 351
1784 ph10 349 /* UCD access macros */
1785    
1786     #define UCD_BLOCK_SIZE 128
1787 ph10 351 #define GET_UCD(ch) (_pcre_ucd_records + \
1788     _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \
1789 ph10 349 UCD_BLOCK_SIZE + ch % UCD_BLOCK_SIZE])
1790    
1791     #define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype
1792     #define UCD_SCRIPT(ch) GET_UCD(ch)->script
1793 ph10 351 #define UCD_CATEGORY(ch) _pcre_ucp_gentype[UCD_CHARTYPE(ch)]
1794     #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)
1795 ph10 349
1796 nigel 85 #endif
1797    
1798 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