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

Contents of /code/trunk/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log


Revision 483 - (hide annotations) (download)
Mon Jan 4 16:19:17 2010 UTC (4 years, 7 months ago) by ph10
File MIME type: text/plain
File size: 74298 byte(s)
Fix some discrepancies between "USPTR" and "unsigned char *".

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 475 /* 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     stdint.h is available, include it; it may define INT64_MAX. The macro int64_t
192     may be set by "configure". */
193    
194     #if HAVE_STDINT_H
195     #include <stdint.h>
196     #endif
197    
198     #if defined INT64_MAX || defined int64_t
199     #define INT64_OR_DOUBLE int64_t
200     #else
201     #define INT64_OR_DOUBLE double
202     #endif
203    
204 nigel 77 /* All character handling must be done as unsigned characters. Otherwise there
205     are problems with top-bit-set characters and functions such as isspace().
206     However, we leave the interface to the outside world as char *, because that
207     should make things easier for callers. We define a short type for unsigned char
208     to save lots of typing. I tried "uchar", but it causes problems on Digital
209     Unix, where it is defined in sys/types, so use "uschar" instead. */
210    
211     typedef unsigned char uschar;
212    
213 nigel 93 /* This is an unsigned int value that no character can ever have. UTF-8
214     characters only go up to 0x7fffffff (though Unicode doesn't go beyond
215     0x0010ffff). */
216 nigel 91
217 nigel 93 #define NOTACHAR 0xffffffff
218    
219     /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
220 ph10 149 "any" and "anycrlf" at present). The following macros are used to package up
221     testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
222     modules to indicate in which datablock the parameters exist, and what the
223     start/end of string field names are. */
224 nigel 93
225 ph10 149 #define NLTYPE_FIXED 0 /* Newline is a fixed length string */
226     #define NLTYPE_ANY 1 /* Newline is any Unicode line ending */
227     #define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */
228 nigel 93
229     /* This macro checks for a newline at the given position */
230    
231 nigel 91 #define IS_NEWLINE(p) \
232 nigel 93 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
233     ((p) < NLBLOCK->PSEND && \
234 ph10 149 _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\
235     utf8)) \
236 nigel 93 : \
237     ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
238     (p)[0] == NLBLOCK->nl[0] && \
239     (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \
240     ) \
241     )
242 nigel 91
243 nigel 93 /* This macro checks for a newline immediately preceding the given position */
244    
245     #define WAS_NEWLINE(p) \
246     ((NLBLOCK->nltype != NLTYPE_FIXED)? \
247     ((p) > NLBLOCK->PSSTART && \
248 ph10 149 _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
249     &(NLBLOCK->nllen), utf8)) \
250 nigel 93 : \
251     ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
252     (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \
253     (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \
254     ) \
255     )
256    
257 nigel 87 /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
258     with a custom type. This makes it possible, for example, to allow pcre_exec()
259     to process subject strings that are discontinuous by using a smart pointer
260     class. It must always be possible to inspect all of the subject string in
261     pcre_exec() because of the way it backtracks. Two macros are required in the
262     normal case, for sign-unspecified and unsigned char pointers. The former is
263     used for the external interface and appears in pcre.h, which is why its name
264     must begin with PCRE_. */
265    
266     #ifdef CUSTOM_SUBJECT_PTR
267     #define PCRE_SPTR CUSTOM_SUBJECT_PTR
268     #define USPTR CUSTOM_SUBJECT_PTR
269     #else
270     #define PCRE_SPTR const char *
271     #define USPTR const unsigned char *
272     #endif
273    
274 ph10 145
275    
276 nigel 85 /* Include the public PCRE header and the definitions of UCP character property
277     values. */
278 nigel 77
279 ph10 189 #include "pcre.h"
280 nigel 77 #include "ucp.h"
281    
282     /* When compiling for use with the Virtual Pascal compiler, these functions
283     need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
284     option on the command line. */
285    
286     #ifdef VPCOMPAT
287 ph10 101 #define strlen(s) _strlen(s)
288 nigel 77 #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
289 ph10 101 #define memcmp(s,c,n) _memcmp(s,c,n)
290 nigel 77 #define memcpy(d,s,n) _memcpy(d,s,n)
291     #define memmove(d,s,n) _memmove(d,s,n)
292     #define memset(s,c,n) _memset(s,c,n)
293     #else /* VPCOMPAT */
294    
295     /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
296     define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
297     is set. Otherwise, include an emulating function for those systems that have
298 ph10 118 neither (there some non-Unix environments where this is the case). */
299 nigel 77
300 ph10 137 #ifndef HAVE_MEMMOVE
301 nigel 77 #undef memmove /* some systems may have a macro */
302 ph10 137 #ifdef HAVE_BCOPY
303 nigel 77 #define memmove(a, b, c) bcopy(b, a, c)
304     #else /* HAVE_BCOPY */
305 nigel 91 static void *
306 ph10 108 pcre_memmove(void *d, const void *s, size_t n)
307 nigel 77 {
308 nigel 87 size_t i;
309 ph10 108 unsigned char *dest = (unsigned char *)d;
310     const unsigned char *src = (const unsigned char *)s;
311 ph10 118 if (dest > src)
312     {
313     dest += n;
314     src += n;
315     for (i = 0; i < n; ++i) *(--dest) = *(--src);
316     return (void *)dest;
317     }
318     else
319     {
320     for (i = 0; i < n; ++i) *dest++ = *src++;
321 ph10 123 return (void *)(dest - n);
322     }
323 nigel 77 }
324     #define memmove(a, b, c) pcre_memmove(a, b, c)
325     #endif /* not HAVE_BCOPY */
326     #endif /* not HAVE_MEMMOVE */
327     #endif /* not VPCOMPAT */
328    
329    
330     /* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
331     in big-endian order) by default. These are used, for example, to link from the
332     start of a subpattern to its alternatives and its end. The use of 2 bytes per
333     offset limits the size of the compiled regex to around 64K, which is big enough
334     for almost everybody. However, I received a request for an even bigger limit.
335     For this reason, and also to make the code easier to maintain, the storing and
336     loading of offsets from the byte string is now handled by the macros that are
337     defined here.
338    
339     The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
340     the config.h file, but can be overridden by using -D on the command line. This
341     is automated on Unix systems via the "configure" command. */
342    
343     #if LINK_SIZE == 2
344    
345     #define PUT(a,n,d) \
346     (a[n] = (d) >> 8), \
347     (a[(n)+1] = (d) & 255)
348    
349     #define GET(a,n) \
350     (((a)[n] << 8) | (a)[(n)+1])
351    
352     #define MAX_PATTERN_SIZE (1 << 16)
353    
354    
355     #elif LINK_SIZE == 3
356    
357     #define PUT(a,n,d) \
358     (a[n] = (d) >> 16), \
359     (a[(n)+1] = (d) >> 8), \
360     (a[(n)+2] = (d) & 255)
361    
362     #define GET(a,n) \
363     (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
364    
365     #define MAX_PATTERN_SIZE (1 << 24)
366    
367    
368     #elif LINK_SIZE == 4
369    
370     #define PUT(a,n,d) \
371     (a[n] = (d) >> 24), \
372     (a[(n)+1] = (d) >> 16), \
373     (a[(n)+2] = (d) >> 8), \
374     (a[(n)+3] = (d) & 255)
375    
376     #define GET(a,n) \
377     (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
378    
379     #define MAX_PATTERN_SIZE (1 << 30) /* Keep it positive */
380    
381    
382     #else
383     #error LINK_SIZE must be either 2, 3, or 4
384     #endif
385    
386    
387     /* Convenience macro defined in terms of the others */
388    
389     #define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
390    
391    
392     /* PCRE uses some other 2-byte quantities that do not change when the size of
393     offsets changes. There are used for repeat counts and for other things such as
394     capturing parenthesis numbers in back references. */
395    
396     #define PUT2(a,n,d) \
397     a[n] = (d) >> 8; \
398     a[(n)+1] = (d) & 255
399    
400     #define GET2(a,n) \
401     (((a)[n] << 8) | (a)[(n)+1])
402    
403     #define PUT2INC(a,n,d) PUT2(a,n,d), a += 2
404    
405    
406     /* When UTF-8 encoding is being used, a character is no longer just a single
407     byte. The macros for character handling generate simple sequences when used in
408 ph10 208 byte-mode, and more complicated ones for UTF-8 characters. BACKCHAR should
409     never be called in byte mode. To make sure it can never even appear when UTF-8
410 ph10 207 support is omitted, we don't even define it. */
411 nigel 77
412     #ifndef SUPPORT_UTF8
413     #define GETCHAR(c, eptr) c = *eptr;
414     #define GETCHARTEST(c, eptr) c = *eptr;
415     #define GETCHARINC(c, eptr) c = *eptr++;
416     #define GETCHARINCTEST(c, eptr) c = *eptr++;
417     #define GETCHARLEN(c, eptr, len) c = *eptr;
418 ph10 207 /* #define BACKCHAR(eptr) */
419 nigel 77
420     #else /* SUPPORT_UTF8 */
421    
422     /* Get the next UTF-8 character, not advancing the pointer. This is called when
423     we know we are in UTF-8 mode. */
424    
425     #define GETCHAR(c, eptr) \
426     c = *eptr; \
427 nigel 93 if (c >= 0xc0) \
428 nigel 77 { \
429     int gcii; \
430     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
431     int gcss = 6*gcaa; \
432     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
433     for (gcii = 1; gcii <= gcaa; gcii++) \
434     { \
435     gcss -= 6; \
436     c |= (eptr[gcii] & 0x3f) << gcss; \
437     } \
438     }
439    
440     /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
441     pointer. */
442    
443     #define GETCHARTEST(c, eptr) \
444     c = *eptr; \
445 nigel 93 if (utf8 && c >= 0xc0) \
446 nigel 77 { \
447     int gcii; \
448     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
449     int gcss = 6*gcaa; \
450     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
451     for (gcii = 1; gcii <= gcaa; gcii++) \
452     { \
453     gcss -= 6; \
454     c |= (eptr[gcii] & 0x3f) << gcss; \
455     } \
456     }
457    
458     /* Get the next UTF-8 character, advancing the pointer. This is called when we
459     know we are in UTF-8 mode. */
460    
461     #define GETCHARINC(c, eptr) \
462     c = *eptr++; \
463 nigel 93 if (c >= 0xc0) \
464 nigel 77 { \
465     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
466     int gcss = 6*gcaa; \
467     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
468     while (gcaa-- > 0) \
469     { \
470     gcss -= 6; \
471     c |= (*eptr++ & 0x3f) << gcss; \
472     } \
473     }
474    
475     /* Get the next character, testing for UTF-8 mode, and advancing the pointer */
476    
477     #define GETCHARINCTEST(c, eptr) \
478     c = *eptr++; \
479 nigel 93 if (utf8 && c >= 0xc0) \
480 nigel 77 { \
481     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
482     int gcss = 6*gcaa; \
483     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
484     while (gcaa-- > 0) \
485     { \
486     gcss -= 6; \
487     c |= (*eptr++ & 0x3f) << gcss; \
488     } \
489     }
490    
491     /* Get the next UTF-8 character, not advancing the pointer, incrementing length
492     if there are extra bytes. This is called when we know we are in UTF-8 mode. */
493    
494     #define GETCHARLEN(c, eptr, len) \
495     c = *eptr; \
496 nigel 93 if (c >= 0xc0) \
497 nigel 77 { \
498     int gcii; \
499     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
500     int gcss = 6*gcaa; \
501     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
502     for (gcii = 1; gcii <= gcaa; gcii++) \
503     { \
504     gcss -= 6; \
505     c |= (eptr[gcii] & 0x3f) << gcss; \
506     } \
507     len += gcaa; \
508     }
509    
510 ph10 384 /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
511     pointer, incrementing length if there are extra bytes. This is called when we
512     know we are in UTF-8 mode. */
513    
514     #define GETCHARLENTEST(c, eptr, len) \
515     c = *eptr; \
516     if (utf8 && c >= 0xc0) \
517     { \
518     int gcii; \
519     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
520     int gcss = 6*gcaa; \
521     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
522     for (gcii = 1; gcii <= gcaa; gcii++) \
523     { \
524     gcss -= 6; \
525     c |= (eptr[gcii] & 0x3f) << gcss; \
526     } \
527     len += gcaa; \
528     }
529    
530 nigel 77 /* If the pointer is not at the start of a character, move it back until
531 ph10 208 it is. This is called only in UTF-8 mode - we don't put a test within the macro
532 ph10 207 because almost all calls are already within a block of UTF-8 only code. */
533 nigel 77
534 ph10 207 #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
535 nigel 77
536     #endif
537    
538    
539     /* In case there is no definition of offsetof() provided - though any proper
540     Standard C system should have one. */
541    
542     #ifndef offsetof
543     #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))
544     #endif
545    
546    
547     /* These are the public options that can change during matching. */
548    
549     #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)
550    
551 ph10 231 /* Private flags containing information about the compiled regex. They used to
552     live at the top end of the options word, but that got almost full, so now they
553 ph10 426 are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
554 ph10 427 the restrictions on partial matching have been lifted. It remains for backwards
555 ph10 426 compatibility. */
556 nigel 77
557 ph10 230 #define PCRE_NOPARTIAL 0x0001 /* can't use partial with this regex */
558     #define PCRE_FIRSTSET 0x0002 /* first_byte is set */
559     #define PCRE_REQCHSET 0x0004 /* req_byte is set */
560     #define PCRE_STARTLINE 0x0008 /* start after \n for multiline */
561     #define PCRE_JCHANGED 0x0010 /* j option used in regex */
562     #define PCRE_HASCRORLF 0x0020 /* explicit \r or \n in pattern */
563 nigel 77
564     /* Options for the "extra" block produced by pcre_study(). */
565    
566     #define PCRE_STUDY_MAPPED 0x01 /* a map of starting chars exists */
567 ph10 455 #define PCRE_STUDY_MINLEN 0x02 /* a minimum length field exists */
568 nigel 77
569     /* Masks for identifying the public options that are permitted at compile
570     time, run time, or study time, respectively. */
571    
572 ph10 149 #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
573     PCRE_NEWLINE_ANYCRLF)
574 nigel 93
575 ph10 389 #define PUBLIC_COMPILE_OPTIONS \
576 nigel 77 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
577     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
578 nigel 91 PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
579 ph10 336 PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
580     PCRE_JAVASCRIPT_COMPAT)
581 nigel 77
582     #define PUBLIC_EXEC_OPTIONS \
583 ph10 442 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
584     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
585     PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
586 nigel 77
587     #define PUBLIC_DFA_EXEC_OPTIONS \
588 ph10 442 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
589     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
590     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
591     PCRE_NO_START_OPTIMIZE)
592 nigel 77
593     #define PUBLIC_STUDY_OPTIONS 0 /* None defined */
594    
595     /* Magic number to provide a small check against being handed junk. Also used
596     to detect whether a pattern was compiled on a host of different endianness. */
597    
598     #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
599    
600     /* Negative values for the firstchar and reqchar variables */
601    
602     #define REQ_UNSET (-2)
603     #define REQ_NONE (-1)
604    
605     /* The maximum remaining length of subject we are prepared to search for a
606     req_byte match. */
607    
608     #define REQ_BYTE_MAX 1000
609    
610     /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a
611     variable-length repeat, or a anything other than literal characters. */
612    
613     #define REQ_CASELESS 0x0100 /* indicates caselessness */
614     #define REQ_VARY 0x0200 /* reqbyte followed non-literal item */
615    
616 ph10 369 /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
617 ph10 376 environments where these macros are defined elsewhere. Unfortunately, there
618     is no way to do the same for the typedef. */
619 nigel 77
620     typedef int BOOL;
621    
622 ph10 376 #ifndef FALSE
623 nigel 77 #define FALSE 0
624     #define TRUE 1
625 ph10 369 #endif
626 nigel 77
627 ph10 391 /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
628     character constants like '*' because the compiler would emit their EBCDIC code,
629 ph10 392 which is different from their ASCII/UTF-8 code. Instead we define macros for
630     the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
631     is enabled. When UTF-8 support is not enabled, the definitions use character
632     literals. Both character and string versions of each character are needed, and
633 ph10 391 there are some longer strings as well.
634    
635 ph10 392 This means that, on EBCDIC platforms, the PCRE library can handle either
636 ph10 391 EBCDIC, or UTF-8, but not both. To support both in the same compiled library
637     would need different lookups depending on whether PCRE_UTF8 was set or not.
638     This would make it impossible to use characters in switch/case statements,
639 ph10 392 which would reduce performance. For a theoretical use (which nobody has asked
640     for) in a minority area (EBCDIC platforms), this is not sensible. Any
641     application that did need both could compile two versions of the library, using
642 ph10 391 macros to give the functions distinct names. */
643    
644     #ifndef SUPPORT_UTF8
645    
646     /* UTF-8 support is not enabled; use the platform-dependent character literals
647     so that PCRE works on both ASCII and EBCDIC platforms, in non-UTF-mode only. */
648    
649     #define CHAR_HT '\t'
650     #define CHAR_VT '\v'
651     #define CHAR_FF '\f'
652     #define CHAR_CR '\r'
653     #define CHAR_NL '\n'
654     #define CHAR_BS '\b'
655     #define CHAR_BEL '\a'
656     #ifdef EBCDIC
657     #define CHAR_ESC '\047'
658     #define CHAR_DEL '\007'
659     #else
660     #define CHAR_ESC '\033'
661     #define CHAR_DEL '\177'
662     #endif
663    
664     #define CHAR_SPACE ' '
665     #define CHAR_EXCLAMATION_MARK '!'
666     #define CHAR_QUOTATION_MARK '"'
667     #define CHAR_NUMBER_SIGN '#'
668     #define CHAR_DOLLAR_SIGN '$'
669     #define CHAR_PERCENT_SIGN '%'
670     #define CHAR_AMPERSAND '&'
671     #define CHAR_APOSTROPHE '\''
672     #define CHAR_LEFT_PARENTHESIS '('
673     #define CHAR_RIGHT_PARENTHESIS ')'
674     #define CHAR_ASTERISK '*'
675     #define CHAR_PLUS '+'
676     #define CHAR_COMMA ','
677     #define CHAR_MINUS '-'
678     #define CHAR_DOT '.'
679     #define CHAR_SLASH '/'
680     #define CHAR_0 '0'
681     #define CHAR_1 '1'
682     #define CHAR_2 '2'
683     #define CHAR_3 '3'
684     #define CHAR_4 '4'
685     #define CHAR_5 '5'
686     #define CHAR_6 '6'
687     #define CHAR_7 '7'
688     #define CHAR_8 '8'
689     #define CHAR_9 '9'
690     #define CHAR_COLON ':'
691     #define CHAR_SEMICOLON ';'
692     #define CHAR_LESS_THAN_SIGN '<'
693     #define CHAR_EQUALS_SIGN '='
694     #define CHAR_GREATER_THAN_SIGN '>'
695     #define CHAR_QUESTION_MARK '?'
696     #define CHAR_COMMERCIAL_AT '@'
697     #define CHAR_A 'A'
698     #define CHAR_B 'B'
699     #define CHAR_C 'C'
700     #define CHAR_D 'D'
701     #define CHAR_E 'E'
702     #define CHAR_F 'F'
703     #define CHAR_G 'G'
704     #define CHAR_H 'H'
705     #define CHAR_I 'I'
706     #define CHAR_J 'J'
707     #define CHAR_K 'K'
708     #define CHAR_L 'L'
709     #define CHAR_M 'M'
710     #define CHAR_N 'N'
711     #define CHAR_O 'O'
712     #define CHAR_P 'P'
713     #define CHAR_Q 'Q'
714     #define CHAR_R 'R'
715     #define CHAR_S 'S'
716     #define CHAR_T 'T'
717     #define CHAR_U 'U'
718     #define CHAR_V 'V'
719     #define CHAR_W 'W'
720     #define CHAR_X 'X'
721     #define CHAR_Y 'Y'
722     #define CHAR_Z 'Z'
723     #define CHAR_LEFT_SQUARE_BRACKET '['
724     #define CHAR_BACKSLASH '\\'
725     #define CHAR_RIGHT_SQUARE_BRACKET ']'
726     #define CHAR_CIRCUMFLEX_ACCENT '^'
727     #define CHAR_UNDERSCORE '_'
728     #define CHAR_GRAVE_ACCENT '`'
729     #define CHAR_a 'a'
730     #define CHAR_b 'b'
731     #define CHAR_c 'c'
732     #define CHAR_d 'd'
733     #define CHAR_e 'e'
734     #define CHAR_f 'f'
735     #define CHAR_g 'g'
736     #define CHAR_h 'h'
737     #define CHAR_i 'i'
738     #define CHAR_j 'j'
739     #define CHAR_k 'k'
740     #define CHAR_l 'l'
741     #define CHAR_m 'm'
742     #define CHAR_n 'n'
743     #define CHAR_o 'o'
744     #define CHAR_p 'p'
745     #define CHAR_q 'q'
746     #define CHAR_r 'r'
747     #define CHAR_s 's'
748     #define CHAR_t 't'
749     #define CHAR_u 'u'
750     #define CHAR_v 'v'
751     #define CHAR_w 'w'
752     #define CHAR_x 'x'
753     #define CHAR_y 'y'
754     #define CHAR_z 'z'
755     #define CHAR_LEFT_CURLY_BRACKET '{'
756     #define CHAR_VERTICAL_LINE '|'
757     #define CHAR_RIGHT_CURLY_BRACKET '}'
758     #define CHAR_TILDE '~'
759    
760     #define STR_HT "\t"
761     #define STR_VT "\v"
762     #define STR_FF "\f"
763     #define STR_CR "\r"
764     #define STR_NL "\n"
765     #define STR_BS "\b"
766     #define STR_BEL "\a"
767     #ifdef EBCDIC
768     #define STR_ESC "\047"
769     #define STR_DEL "\007"
770     #else
771     #define STR_ESC "\033"
772     #define STR_DEL "\177"
773     #endif
774    
775     #define STR_SPACE " "
776     #define STR_EXCLAMATION_MARK "!"
777     #define STR_QUOTATION_MARK "\""
778     #define STR_NUMBER_SIGN "#"
779     #define STR_DOLLAR_SIGN "$"
780     #define STR_PERCENT_SIGN "%"
781     #define STR_AMPERSAND "&"
782     #define STR_APOSTROPHE "'"
783     #define STR_LEFT_PARENTHESIS "("
784     #define STR_RIGHT_PARENTHESIS ")"
785     #define STR_ASTERISK "*"
786     #define STR_PLUS "+"
787     #define STR_COMMA ","
788     #define STR_MINUS "-"
789     #define STR_DOT "."
790     #define STR_SLASH "/"
791     #define STR_0 "0"
792     #define STR_1 "1"
793     #define STR_2 "2"
794     #define STR_3 "3"
795     #define STR_4 "4"
796     #define STR_5 "5"
797     #define STR_6 "6"
798     #define STR_7 "7"
799     #define STR_8 "8"
800     #define STR_9 "9"
801     #define STR_COLON ":"
802     #define STR_SEMICOLON ";"
803     #define STR_LESS_THAN_SIGN "<"
804     #define STR_EQUALS_SIGN "="
805     #define STR_GREATER_THAN_SIGN ">"
806     #define STR_QUESTION_MARK "?"
807     #define STR_COMMERCIAL_AT "@"
808     #define STR_A "A"
809     #define STR_B "B"
810     #define STR_C "C"
811     #define STR_D "D"
812     #define STR_E "E"
813     #define STR_F "F"
814     #define STR_G "G"
815     #define STR_H "H"
816     #define STR_I "I"
817     #define STR_J "J"
818     #define STR_K "K"
819     #define STR_L "L"
820     #define STR_M "M"
821     #define STR_N "N"
822     #define STR_O "O"
823     #define STR_P "P"
824     #define STR_Q "Q"
825     #define STR_R "R"
826     #define STR_S "S"
827     #define STR_T "T"
828     #define STR_U "U"
829     #define STR_V "V"
830     #define STR_W "W"
831     #define STR_X "X"
832     #define STR_Y "Y"
833     #define STR_Z "Z"
834     #define STR_LEFT_SQUARE_BRACKET "["
835     #define STR_BACKSLASH "\\"
836     #define STR_RIGHT_SQUARE_BRACKET "]"
837     #define STR_CIRCUMFLEX_ACCENT "^"
838     #define STR_UNDERSCORE "_"
839     #define STR_GRAVE_ACCENT "`"
840     #define STR_a "a"
841     #define STR_b "b"
842     #define STR_c "c"
843     #define STR_d "d"
844     #define STR_e "e"
845     #define STR_f "f"
846     #define STR_g "g"
847     #define STR_h "h"
848     #define STR_i "i"
849     #define STR_j "j"
850     #define STR_k "k"
851     #define STR_l "l"
852     #define STR_m "m"
853     #define STR_n "n"
854     #define STR_o "o"
855     #define STR_p "p"
856     #define STR_q "q"
857     #define STR_r "r"
858     #define STR_s "s"
859     #define STR_t "t"
860     #define STR_u "u"
861     #define STR_v "v"
862     #define STR_w "w"
863     #define STR_x "x"
864     #define STR_y "y"
865     #define STR_z "z"
866     #define STR_LEFT_CURLY_BRACKET "{"
867     #define STR_VERTICAL_LINE "|"
868     #define STR_RIGHT_CURLY_BRACKET "}"
869     #define STR_TILDE "~"
870    
871     #define STRING_ACCEPT0 "ACCEPT\0"
872     #define STRING_COMMIT0 "COMMIT\0"
873     #define STRING_F0 "F\0"
874     #define STRING_FAIL0 "FAIL\0"
875     #define STRING_PRUNE0 "PRUNE\0"
876     #define STRING_SKIP0 "SKIP\0"
877     #define STRING_THEN "THEN"
878    
879     #define STRING_alpha0 "alpha\0"
880     #define STRING_lower0 "lower\0"
881     #define STRING_upper0 "upper\0"
882     #define STRING_alnum0 "alnum\0"
883     #define STRING_ascii0 "ascii\0"
884     #define STRING_blank0 "blank\0"
885     #define STRING_cntrl0 "cntrl\0"
886     #define STRING_digit0 "digit\0"
887     #define STRING_graph0 "graph\0"
888     #define STRING_print0 "print\0"
889     #define STRING_punct0 "punct\0"
890     #define STRING_space0 "space\0"
891     #define STRING_word0 "word\0"
892     #define STRING_xdigit "xdigit"
893    
894     #define STRING_DEFINE "DEFINE"
895    
896     #define STRING_CR_RIGHTPAR "CR)"
897     #define STRING_LF_RIGHTPAR "LF)"
898     #define STRING_CRLF_RIGHTPAR "CRLF)"
899     #define STRING_ANY_RIGHTPAR "ANY)"
900     #define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)"
901     #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"
902     #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"
903 ph10 412 #define STRING_UTF8_RIGHTPAR "UTF8)"
904 ph10 391
905     #else /* SUPPORT_UTF8 */
906    
907     /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
908     works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
909     only. */
910    
911     #define CHAR_HT '\011'
912     #define CHAR_VT '\013'
913     #define CHAR_FF '\014'
914     #define CHAR_CR '\015'
915     #define CHAR_NL '\012'
916     #define CHAR_BS '\010'
917     #define CHAR_BEL '\007'
918     #define CHAR_ESC '\033'
919     #define CHAR_DEL '\177'
920    
921     #define CHAR_SPACE '\040'
922     #define CHAR_EXCLAMATION_MARK '\041'
923     #define CHAR_QUOTATION_MARK '\042'
924     #define CHAR_NUMBER_SIGN '\043'
925     #define CHAR_DOLLAR_SIGN '\044'
926     #define CHAR_PERCENT_SIGN '\045'
927     #define CHAR_AMPERSAND '\046'
928     #define CHAR_APOSTROPHE '\047'
929     #define CHAR_LEFT_PARENTHESIS '\050'
930     #define CHAR_RIGHT_PARENTHESIS '\051'
931     #define CHAR_ASTERISK '\052'
932     #define CHAR_PLUS '\053'
933     #define CHAR_COMMA '\054'
934     #define CHAR_MINUS '\055'
935     #define CHAR_DOT '\056'
936     #define CHAR_SLASH '\057'
937     #define CHAR_0 '\060'
938     #define CHAR_1 '\061'
939     #define CHAR_2 '\062'
940     #define CHAR_3 '\063'
941     #define CHAR_4 '\064'
942     #define CHAR_5 '\065'
943     #define CHAR_6 '\066'
944     #define CHAR_7 '\067'
945     #define CHAR_8 '\070'
946     #define CHAR_9 '\071'
947     #define CHAR_COLON '\072'
948     #define CHAR_SEMICOLON '\073'
949     #define CHAR_LESS_THAN_SIGN '\074'
950     #define CHAR_EQUALS_SIGN '\075'
951     #define CHAR_GREATER_THAN_SIGN '\076'
952     #define CHAR_QUESTION_MARK '\077'
953     #define CHAR_COMMERCIAL_AT '\100'
954     #define CHAR_A '\101'
955     #define CHAR_B '\102'
956     #define CHAR_C '\103'
957     #define CHAR_D '\104'
958     #define CHAR_E '\105'
959     #define CHAR_F '\106'
960     #define CHAR_G '\107'
961     #define CHAR_H '\110'
962     #define CHAR_I '\111'
963     #define CHAR_J '\112'
964     #define CHAR_K '\113'
965     #define CHAR_L '\114'
966     #define CHAR_M '\115'
967     #define CHAR_N '\116'
968     #define CHAR_O '\117'
969     #define CHAR_P '\120'
970     #define CHAR_Q '\121'
971     #define CHAR_R '\122'
972     #define CHAR_S '\123'
973     #define CHAR_T '\124'
974     #define CHAR_U '\125'
975     #define CHAR_V '\126'
976     #define CHAR_W '\127'
977     #define CHAR_X '\130'
978     #define CHAR_Y '\131'
979     #define CHAR_Z '\132'
980     #define CHAR_LEFT_SQUARE_BRACKET '\133'
981     #define CHAR_BACKSLASH '\134'
982     #define CHAR_RIGHT_SQUARE_BRACKET '\135'
983     #define CHAR_CIRCUMFLEX_ACCENT '\136'
984     #define CHAR_UNDERSCORE '\137'
985     #define CHAR_GRAVE_ACCENT '\140'
986     #define CHAR_a '\141'
987     #define CHAR_b '\142'
988     #define CHAR_c '\143'
989     #define CHAR_d '\144'
990     #define CHAR_e '\145'
991     #define CHAR_f '\146'
992     #define CHAR_g '\147'
993     #define CHAR_h '\150'
994     #define CHAR_i '\151'
995     #define CHAR_j '\152'
996     #define CHAR_k '\153'
997     #define CHAR_l '\154'
998     #define CHAR_m '\155'
999     #define CHAR_n '\156'
1000     #define CHAR_o '\157'
1001     #define CHAR_p '\160'
1002     #define CHAR_q '\161'
1003     #define CHAR_r '\162'
1004     #define CHAR_s '\163'
1005     #define CHAR_t '\164'
1006     #define CHAR_u '\165'
1007     #define CHAR_v '\166'
1008     #define CHAR_w '\167'
1009     #define CHAR_x '\170'
1010     #define CHAR_y '\171'
1011     #define CHAR_z '\172'
1012     #define CHAR_LEFT_CURLY_BRACKET '\173'
1013     #define CHAR_VERTICAL_LINE '\174'
1014     #define CHAR_RIGHT_CURLY_BRACKET '\175'
1015     #define CHAR_TILDE '\176'
1016    
1017     #define STR_HT "\011"
1018     #define STR_VT "\013"
1019     #define STR_FF "\014"
1020     #define STR_CR "\015"
1021     #define STR_NL "\012"
1022     #define STR_BS "\010"
1023     #define STR_BEL "\007"
1024     #define STR_ESC "\033"
1025     #define STR_DEL "\177"
1026    
1027     #define STR_SPACE "\040"
1028     #define STR_EXCLAMATION_MARK "\041"
1029     #define STR_QUOTATION_MARK "\042"
1030     #define STR_NUMBER_SIGN "\043"
1031     #define STR_DOLLAR_SIGN "\044"
1032     #define STR_PERCENT_SIGN "\045"
1033     #define STR_AMPERSAND "\046"
1034     #define STR_APOSTROPHE "\047"
1035     #define STR_LEFT_PARENTHESIS "\050"
1036     #define STR_RIGHT_PARENTHESIS "\051"
1037     #define STR_ASTERISK "\052"
1038     #define STR_PLUS "\053"
1039     #define STR_COMMA "\054"
1040     #define STR_MINUS "\055"
1041     #define STR_DOT "\056"
1042     #define STR_SLASH "\057"
1043     #define STR_0 "\060"
1044     #define STR_1 "\061"
1045     #define STR_2 "\062"
1046     #define STR_3 "\063"
1047     #define STR_4 "\064"
1048     #define STR_5 "\065"
1049     #define STR_6 "\066"
1050     #define STR_7 "\067"
1051     #define STR_8 "\070"
1052     #define STR_9 "\071"
1053     #define STR_COLON "\072"
1054     #define STR_SEMICOLON "\073"
1055     #define STR_LESS_THAN_SIGN "\074"
1056     #define STR_EQUALS_SIGN "\075"
1057     #define STR_GREATER_THAN_SIGN "\076"
1058     #define STR_QUESTION_MARK "\077"
1059     #define STR_COMMERCIAL_AT "\100"
1060     #define STR_A "\101"
1061     #define STR_B "\102"
1062     #define STR_C "\103"
1063     #define STR_D "\104"
1064     #define STR_E "\105"
1065     #define STR_F "\106"
1066     #define STR_G "\107"
1067     #define STR_H "\110"
1068     #define STR_I "\111"
1069     #define STR_J "\112"
1070     #define STR_K "\113"
1071     #define STR_L "\114"
1072     #define STR_M "\115"
1073     #define STR_N "\116"
1074     #define STR_O "\117"
1075     #define STR_P "\120"
1076     #define STR_Q "\121"
1077     #define STR_R "\122"
1078     #define STR_S "\123"
1079     #define STR_T "\124"
1080     #define STR_U "\125"
1081     #define STR_V "\126"
1082     #define STR_W "\127"
1083     #define STR_X "\130"
1084     #define STR_Y "\131"
1085     #define STR_Z "\132"
1086     #define STR_LEFT_SQUARE_BRACKET "\133"
1087     #define STR_BACKSLASH "\134"
1088     #define STR_RIGHT_SQUARE_BRACKET "\135"
1089     #define STR_CIRCUMFLEX_ACCENT "\136"
1090     #define STR_UNDERSCORE "\137"
1091     #define STR_GRAVE_ACCENT "\140"
1092     #define STR_a "\141"
1093     #define STR_b "\142"
1094     #define STR_c "\143"
1095     #define STR_d "\144"
1096     #define STR_e "\145"
1097     #define STR_f "\146"
1098     #define STR_g "\147"
1099     #define STR_h "\150"
1100     #define STR_i "\151"
1101     #define STR_j "\152"
1102     #define STR_k "\153"
1103     #define STR_l "\154"
1104     #define STR_m "\155"
1105     #define STR_n "\156"
1106     #define STR_o "\157"
1107     #define STR_p "\160"
1108     #define STR_q "\161"
1109     #define STR_r "\162"
1110     #define STR_s "\163"
1111     #define STR_t "\164"
1112     #define STR_u "\165"
1113     #define STR_v "\166"
1114     #define STR_w "\167"
1115     #define STR_x "\170"
1116     #define STR_y "\171"
1117     #define STR_z "\172"
1118     #define STR_LEFT_CURLY_BRACKET "\173"
1119     #define STR_VERTICAL_LINE "\174"
1120     #define STR_RIGHT_CURLY_BRACKET "\175"
1121     #define STR_TILDE "\176"
1122    
1123     #define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1124     #define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1125     #define STRING_F0 STR_F "\0"
1126     #define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0"
1127     #define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0"
1128     #define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0"
1129     #define STRING_THEN STR_T STR_H STR_E STR_N
1130    
1131     #define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0"
1132     #define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0"
1133     #define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0"
1134     #define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0"
1135     #define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0"
1136     #define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0"
1137     #define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0"
1138     #define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0"
1139     #define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0"
1140     #define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0"
1141     #define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0"
1142     #define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0"
1143     #define STRING_word0 STR_w STR_o STR_r STR_d "\0"
1144     #define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t
1145    
1146     #define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E
1147    
1148     #define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS
1149     #define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS
1150     #define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1151     #define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1152     #define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1153     #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
1154     #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
1155 ph10 412 #define STRING_UTF8_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1156 ph10 391
1157     #endif /* SUPPORT_UTF8 */
1158    
1159 nigel 93 /* Escape items that are just an encoding of a particular data value. */
1160 nigel 77
1161     #ifndef ESC_e
1162 ph10 391 #define ESC_e CHAR_ESC
1163 nigel 77 #endif
1164    
1165     #ifndef ESC_f
1166 ph10 391 #define ESC_f CHAR_FF
1167 nigel 77 #endif
1168    
1169     #ifndef ESC_n
1170 ph10 391 #define ESC_n CHAR_NL
1171 nigel 77 #endif
1172    
1173     #ifndef ESC_r
1174 ph10 391 #define ESC_r CHAR_CR
1175 nigel 77 #endif
1176    
1177     /* We can't officially use ESC_t because it is a POSIX reserved identifier
1178     (presumably because of all the others like size_t). */
1179    
1180     #ifndef ESC_tee
1181 ph10 391 #define ESC_tee CHAR_HT
1182 nigel 77 #endif
1183    
1184 nigel 87 /* Codes for different types of Unicode property */
1185    
1186     #define PT_ANY 0 /* Any property - matches all chars */
1187     #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
1188     #define PT_GC 2 /* General characteristic (e.g. L) */
1189     #define PT_PC 3 /* Particular characteristic (e.g. Lu) */
1190     #define PT_SC 4 /* Script (e.g. Han) */
1191    
1192     /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1193     contain UTF-8 characters with values greater than 255. */
1194    
1195     #define XCL_NOT 0x01 /* Flag: this is a negative class */
1196     #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
1197    
1198     #define XCL_END 0 /* Marks end of individual items */
1199     #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
1200     #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
1201     #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
1202     #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
1203    
1204 nigel 77 /* These are escaped items that aren't just an encoding of a particular data
1205     value such as \n. They must have non-zero values, as check_escape() returns
1206     their negation. Also, they must appear in the same order as in the opcode
1207     definitions below, up to ESC_z. There's a dummy for OP_ANY because it
1208 ph10 345 corresponds to "." rather than an escape sequence, and another for OP_ALLANY
1209 ph10 341 (which is used for [^] in JavaScript compatibility mode).
1210 nigel 77
1211 ph10 341 The final escape must be ESC_REF as subsequent values are used for
1212     backreferences (\1, \2, \3, etc). There are two tests in the code for an escape
1213     greater than ESC_b and less than ESC_Z to detect the types that may be
1214     repeated. These are the types that consume characters. If any new escapes are
1215     put in between that don't consume a character, that code will have to change.
1216     */
1217    
1218 ph10 172 enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1219 ph10 345 ESC_W, ESC_w, ESC_dum1, ESC_dum2, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1220     ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_g, ESC_k,
1221 ph10 341 ESC_REF };
1222 nigel 77
1223 nigel 93
1224 ph10 210 /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1225 nigel 77 OP_EOD must correspond in order to the list of escapes immediately above.
1226    
1227 nigel 93 *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1228 ph10 462 that follow must also be updated to match. There are also tables called
1229     "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1230 nigel 93
1231 nigel 77 enum {
1232     OP_END, /* 0 End of pattern */
1233    
1234     /* Values corresponding to backslashed metacharacters */
1235    
1236     OP_SOD, /* 1 Start of data: \A */
1237     OP_SOM, /* 2 Start of match (subject + offset): \G */
1238 ph10 172 OP_SET_SOM, /* 3 Set start of match (\K) */
1239 ph10 168 OP_NOT_WORD_BOUNDARY, /* 4 \B */
1240     OP_WORD_BOUNDARY, /* 5 \b */
1241     OP_NOT_DIGIT, /* 6 \D */
1242     OP_DIGIT, /* 7 \d */
1243     OP_NOT_WHITESPACE, /* 8 \S */
1244     OP_WHITESPACE, /* 9 \s */
1245     OP_NOT_WORDCHAR, /* 10 \W */
1246     OP_WORDCHAR, /* 11 \w */
1247 ph10 341 OP_ANY, /* 12 Match any character (subject to DOTALL) */
1248     OP_ALLANY, /* 13 Match any character (not subject to DOTALL) */
1249     OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1250     OP_NOTPROP, /* 15 \P (not Unicode property) */
1251     OP_PROP, /* 16 \p (Unicode property) */
1252     OP_ANYNL, /* 17 \R (any newline sequence) */
1253     OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */
1254     OP_HSPACE, /* 19 \h (horizontal whitespace) */
1255     OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */
1256     OP_VSPACE, /* 21 \v (vertical whitespace) */
1257     OP_EXTUNI, /* 22 \X (extended Unicode sequence */
1258     OP_EODN, /* 23 End of data or \n at end of data: \Z. */
1259     OP_EOD, /* 24 End of data: \z */
1260 nigel 77
1261 ph10 341 OP_OPT, /* 25 Set runtime options */
1262     OP_CIRC, /* 26 Start of line - varies with multiline switch */
1263     OP_DOLL, /* 27 End of line - varies with multiline switch */
1264     OP_CHAR, /* 28 Match one character, casefully */
1265     OP_CHARNC, /* 29 Match one character, caselessly */
1266     OP_NOT, /* 30 Match one character, not the following one */
1267 nigel 77
1268 ph10 341 OP_STAR, /* 31 The maximizing and minimizing versions of */
1269     OP_MINSTAR, /* 32 these six opcodes must come in pairs, with */
1270     OP_PLUS, /* 33 the minimizing one second. */
1271     OP_MINPLUS, /* 34 This first set applies to single characters.*/
1272     OP_QUERY, /* 35 */
1273     OP_MINQUERY, /* 36 */
1274 nigel 77
1275 ph10 341 OP_UPTO, /* 37 From 0 to n matches */
1276     OP_MINUPTO, /* 38 */
1277     OP_EXACT, /* 39 Exactly n matches */
1278 nigel 77
1279 ph10 341 OP_POSSTAR, /* 40 Possessified star */
1280     OP_POSPLUS, /* 41 Possessified plus */
1281     OP_POSQUERY, /* 42 Posesssified query */
1282     OP_POSUPTO, /* 43 Possessified upto */
1283 nigel 77
1284 ph10 341 OP_NOTSTAR, /* 44 The maximizing and minimizing versions of */
1285     OP_NOTMINSTAR, /* 45 these six opcodes must come in pairs, with */
1286     OP_NOTPLUS, /* 46 the minimizing one second. They must be in */
1287     OP_NOTMINPLUS, /* 47 exactly the same order as those above. */
1288     OP_NOTQUERY, /* 48 This set applies to "not" single characters. */
1289     OP_NOTMINQUERY, /* 49 */
1290 nigel 77
1291 ph10 341 OP_NOTUPTO, /* 50 From 0 to n matches */
1292     OP_NOTMINUPTO, /* 51 */
1293     OP_NOTEXACT, /* 52 Exactly n matches */
1294 nigel 93
1295 ph10 341 OP_NOTPOSSTAR, /* 53 Possessified versions */
1296     OP_NOTPOSPLUS, /* 54 */
1297     OP_NOTPOSQUERY, /* 55 */
1298     OP_NOTPOSUPTO, /* 56 */
1299 nigel 93
1300 ph10 341 OP_TYPESTAR, /* 57 The maximizing and minimizing versions of */
1301     OP_TYPEMINSTAR, /* 58 these six opcodes must come in pairs, with */
1302     OP_TYPEPLUS, /* 59 the minimizing one second. These codes must */
1303     OP_TYPEMINPLUS, /* 60 be in exactly the same order as those above. */
1304     OP_TYPEQUERY, /* 61 This set applies to character types such as \d */
1305     OP_TYPEMINQUERY, /* 62 */
1306 nigel 93
1307 ph10 341 OP_TYPEUPTO, /* 63 From 0 to n matches */
1308     OP_TYPEMINUPTO, /* 64 */
1309     OP_TYPEEXACT, /* 65 Exactly n matches */
1310 nigel 93
1311 ph10 341 OP_TYPEPOSSTAR, /* 66 Possessified versions */
1312     OP_TYPEPOSPLUS, /* 67 */
1313     OP_TYPEPOSQUERY, /* 68 */
1314     OP_TYPEPOSUPTO, /* 69 */
1315 nigel 93
1316 ph10 341 OP_CRSTAR, /* 70 The maximizing and minimizing versions of */
1317     OP_CRMINSTAR, /* 71 all these opcodes must come in pairs, with */
1318     OP_CRPLUS, /* 72 the minimizing one second. These codes must */
1319     OP_CRMINPLUS, /* 73 be in exactly the same order as those above. */
1320     OP_CRQUERY, /* 74 These are for character classes and back refs */
1321     OP_CRMINQUERY, /* 75 */
1322     OP_CRRANGE, /* 76 These are different to the three sets above. */
1323     OP_CRMINRANGE, /* 77 */
1324 nigel 93
1325 ph10 341 OP_CLASS, /* 78 Match a character class, chars < 256 only */
1326     OP_NCLASS, /* 79 Same, but the bitmap was created from a negative
1327 nigel 77 class - the difference is relevant only when a UTF-8
1328     character > 255 is encountered. */
1329    
1330 ph10 341 OP_XCLASS, /* 80 Extended class for handling UTF-8 chars within the
1331 nigel 77 class. This does both positive and negative. */
1332    
1333 ph10 341 OP_REF, /* 81 Match a back reference */
1334     OP_RECURSE, /* 82 Match a numbered subpattern (possibly recursive) */
1335     OP_CALLOUT, /* 83 Call out to external function if provided */
1336 nigel 77
1337 ph10 341 OP_ALT, /* 84 Start of alternation */
1338     OP_KET, /* 85 End of group that doesn't have an unbounded repeat */
1339     OP_KETRMAX, /* 86 These two must remain together and in this */
1340     OP_KETRMIN, /* 87 order. They are for groups the repeat for ever. */
1341 nigel 77
1342 nigel 93 /* The assertions must come before BRA, CBRA, ONCE, and COND.*/
1343 nigel 77
1344 ph10 341 OP_ASSERT, /* 88 Positive lookahead */
1345     OP_ASSERT_NOT, /* 89 Negative lookahead */
1346     OP_ASSERTBACK, /* 90 Positive lookbehind */
1347     OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */
1348     OP_REVERSE, /* 92 Move pointer back - used in lookbehind assertions */
1349 nigel 77
1350 nigel 93 /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,
1351     as there's a test for >= ONCE for a subpattern that isn't an assertion. */
1352 nigel 77
1353 ph10 341 OP_ONCE, /* 93 Atomic group */
1354     OP_BRA, /* 94 Start of non-capturing bracket */
1355     OP_CBRA, /* 95 Start of capturing bracket */
1356     OP_COND, /* 96 Conditional group */
1357 nigel 77
1358 nigel 93 /* These three must follow the previous three, in the same order. There's a
1359     check for >= SBRA to distinguish the two sets. */
1360 nigel 77
1361 ph10 341 OP_SBRA, /* 97 Start of non-capturing bracket, check empty */
1362     OP_SCBRA, /* 98 Start of capturing bracket, check empty */
1363     OP_SCOND, /* 99 Conditional group, check empty */
1364 nigel 77
1365 ph10 459 /* The next two pairs must (respectively) be kept together. */
1366 ph10 461
1367 ph10 341 OP_CREF, /* 100 Used to hold a capture number as condition */
1368 ph10 459 OP_NCREF, /* 101 Same, but generaged by a name reference*/
1369     OP_RREF, /* 102 Used to hold a recursion number as condition */
1370     OP_NRREF, /* 103 Same, but generaged by a name reference*/
1371     OP_DEF, /* 104 The DEFINE condition */
1372 nigel 93
1373 ph10 459 OP_BRAZERO, /* 105 These two must remain together and in this */
1374     OP_BRAMINZERO, /* 106 order. */
1375 ph10 210
1376     /* These are backtracking control verbs */
1377    
1378 ph10 459 OP_PRUNE, /* 107 */
1379     OP_SKIP, /* 108 */
1380     OP_THEN, /* 109 */
1381     OP_COMMIT, /* 110 */
1382 ph10 210
1383     /* These are forced failure and success verbs */
1384    
1385 ph10 459 OP_FAIL, /* 111 */
1386     OP_ACCEPT, /* 112 */
1387     OP_CLOSE, /* 113 Used before OP_ACCEPT to close open captures */
1388 ph10 345
1389 ph10 335 /* This is used to skip a subpattern with a {0} quantifier */
1390 ph10 345
1391 ph10 459 OP_SKIPZERO /* 114 */
1392 nigel 77 };
1393    
1394 ph10 462 /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
1395     definitions that follow must also be updated to match. There are also tables
1396     called "coptable" cna "poptable" in pcre_dfa_exec.c that must be updated. */
1397 nigel 77
1398 ph10 462
1399 nigel 77 /* This macro defines textual names for all the opcodes. These are used only
1400     for debugging. The macro is referenced only in pcre_printint.c. */
1401    
1402     #define OP_NAME_LIST \
1403 ph10 168 "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \
1404 ph10 341 "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \
1405 ph10 178 "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \
1406     "extuni", "\\Z", "\\z", \
1407 nigel 77 "Opt", "^", "$", "char", "charnc", "not", \
1408     "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1409 nigel 93 "*+","++", "?+", "{", \
1410 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1411 nigel 93 "*+","++", "?+", "{", \
1412 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1413 nigel 93 "*+","++", "?+", "{", \
1414 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", \
1415     "class", "nclass", "xclass", "Ref", "Recurse", "Callout", \
1416     "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", \
1417 nigel 93 "AssertB", "AssertB not", "Reverse", \
1418 ph10 192 "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond", \
1419 ph10 459 "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def", \
1420     "Brazero", "Braminzero", \
1421 ph10 335 "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT", \
1422 ph10 447 "Close", "Skip zero"
1423 nigel 77
1424    
1425     /* This macro defines the length of fixed length operations in the compiled
1426     regex. The lengths are used when searching for specific things, and also in the
1427     debugging printing of a compiled regex. We use a macro so that it can be
1428     defined close to the definitions of the opcodes themselves.
1429    
1430     As things have been extended, some of these are no longer fixed lenths, but are
1431     minima instead. For example, the length of a single-character repeat may vary
1432     in UTF-8 mode. The code that uses this table must know about such things. */
1433    
1434     #define OP_LENGTHS \
1435     1, /* End */ \
1436 ph10 168 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \
1437     1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \
1438 ph10 341 1, 1, 1, /* Any, AllAny, Anybyte */ \
1439 ph10 178 3, 3, 1, /* NOTPROP, PROP, EXTUNI */ \
1440     1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \
1441 nigel 77 1, 1, 2, 1, 1, /* \Z, \z, Opt, ^, $ */ \
1442     2, /* Char - the minimum length */ \
1443     2, /* Charnc - the minimum length */ \
1444     2, /* not */ \
1445     /* Positive single-char repeats ** These are */ \
1446     2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
1447     4, 4, 4, /* upto, minupto, exact ** UTF-8 mode */ \
1448 nigel 93 2, 2, 2, 4, /* *+, ++, ?+, upto+ */ \
1449 nigel 77 /* Negative single-char repeats - only for chars < 256 */ \
1450     2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
1451     4, 4, 4, /* NOT upto, minupto, exact */ \
1452 nigel 93 2, 2, 2, 4, /* Possessive *, +, ?, upto */ \
1453 nigel 77 /* Positive type repeats */ \
1454     2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
1455     4, 4, 4, /* Type upto, minupto, exact */ \
1456 nigel 93 2, 2, 2, 4, /* Possessive *+, ++, ?+, upto+ */ \
1457 nigel 77 /* Character class & ref repeats */ \
1458     1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
1459     5, 5, /* CRRANGE, CRMINRANGE */ \
1460     33, /* CLASS */ \
1461     33, /* NCLASS */ \
1462     0, /* XCLASS - variable length */ \
1463     3, /* REF */ \
1464     1+LINK_SIZE, /* RECURSE */ \
1465     2+2*LINK_SIZE, /* CALLOUT */ \
1466     1+LINK_SIZE, /* Alt */ \
1467     1+LINK_SIZE, /* Ket */ \
1468     1+LINK_SIZE, /* KetRmax */ \
1469     1+LINK_SIZE, /* KetRmin */ \
1470     1+LINK_SIZE, /* Assert */ \
1471     1+LINK_SIZE, /* Assert not */ \
1472     1+LINK_SIZE, /* Assert behind */ \
1473     1+LINK_SIZE, /* Assert behind not */ \
1474     1+LINK_SIZE, /* Reverse */ \
1475 nigel 93 1+LINK_SIZE, /* ONCE */ \
1476     1+LINK_SIZE, /* BRA */ \
1477     3+LINK_SIZE, /* CBRA */ \
1478 nigel 77 1+LINK_SIZE, /* COND */ \
1479 nigel 93 1+LINK_SIZE, /* SBRA */ \
1480     3+LINK_SIZE, /* SCBRA */ \
1481     1+LINK_SIZE, /* SCOND */ \
1482 ph10 459 3, 3, /* CREF, NCREF */ \
1483     3, 3, /* RREF, NRREF */ \
1484 nigel 93 1, /* DEF */ \
1485 nigel 77 1, 1, /* BRAZERO, BRAMINZERO */ \
1486 ph10 210 1, 1, 1, 1, /* PRUNE, SKIP, THEN, COMMIT, */ \
1487 ph10 447 1, 1, 3, 1 /* FAIL, ACCEPT, CLOSE, SKIPZERO */
1488 nigel 77
1489    
1490 ph10 459 /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
1491     condition. */
1492 nigel 77
1493 nigel 93 #define RREF_ANY 0xffff
1494 nigel 77
1495     /* Error code numbers. They are given names so that they can more easily be
1496     tracked. */
1497    
1498     enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
1499     ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
1500     ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
1501     ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
1502 nigel 91 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
1503 ph10 210 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
1504 ph10 457 ERR60, ERR61, ERR62, ERR63, ERR64, ERR65 };
1505 nigel 77
1506     /* The real format of the start of the pcre block; the index of names and the
1507     code vector run on as long as necessary after the end. We store an explicit
1508     offset to the name table so that if a regex is compiled on one host, saved, and
1509     then run on another where the size of pointers is different, all might still
1510     be well. For the case of compiled-on-4 and run-on-8, we include an extra
1511     pointer that is always NULL. For future-proofing, a few dummy fields were
1512     originally included - even though you can never get this planning right - but
1513     there is only one left now.
1514    
1515     NOTE NOTE NOTE:
1516     Because people can now save and re-use compiled patterns, any additions to this
1517     structure should be made at the end, and something earlier (e.g. a new
1518     flag in the options or one of the dummy fields) should indicate that the new
1519     fields are present. Currently PCRE always sets the dummy fields to zero.
1520 ph10 455 NOTE NOTE NOTE
1521 nigel 77 */
1522    
1523     typedef struct real_pcre {
1524     pcre_uint32 magic_number;
1525     pcre_uint32 size; /* Total that was malloced */
1526 ph10 230 pcre_uint32 options; /* Public options */
1527     pcre_uint16 flags; /* Private flags */
1528     pcre_uint16 dummy1; /* For future use */
1529 nigel 77 pcre_uint16 top_bracket;
1530     pcre_uint16 top_backref;
1531     pcre_uint16 first_byte;
1532     pcre_uint16 req_byte;
1533     pcre_uint16 name_table_offset; /* Offset to name table that follows */
1534     pcre_uint16 name_entry_size; /* Size of any name items */
1535     pcre_uint16 name_count; /* Number of name items */
1536     pcre_uint16 ref_count; /* Reference count */
1537    
1538     const unsigned char *tables; /* Pointer to tables or NULL for std */
1539     const unsigned char *nullpad; /* NULL padding */
1540     } real_pcre;
1541    
1542     /* The format of the block used to store data from pcre_study(). The same
1543     remark (see NOTE above) about extending this structure applies. */
1544    
1545     typedef struct pcre_study_data {
1546     pcre_uint32 size; /* Total that was malloced */
1547 ph10 455 pcre_uint32 flags; /* Private flags */
1548     uschar start_bits[32]; /* Starting char bits */
1549 ph10 459 pcre_uint32 minlength; /* Minimum subject length */
1550 nigel 77 } pcre_study_data;
1551    
1552 ph10 459 /* Structure for building a chain of open capturing subpatterns during
1553     compiling, so that instructions to close them can be compiled when (*ACCEPT) is
1554 ph10 447 encountered. */
1555    
1556     typedef struct open_capitem {
1557     struct open_capitem *next; /* Chain link */
1558     pcre_uint16 number; /* Capture number */
1559 ph10 459 } open_capitem;
1560 ph10 447
1561 nigel 77 /* Structure for passing "static" information around between the functions
1562     doing the compiling, so that they are thread-safe. */
1563    
1564     typedef struct compile_data {
1565     const uschar *lcc; /* Points to lower casing table */
1566     const uschar *fcc; /* Points to case-flipping table */
1567     const uschar *cbits; /* Points to character type table */
1568     const uschar *ctypes; /* Points to table of type maps */
1569 nigel 93 const uschar *start_workspace;/* The start of working space */
1570 nigel 77 const uschar *start_code; /* The start of the compiled code */
1571     const uschar *start_pattern; /* The start of the pattern */
1572 nigel 93 const uschar *end_pattern; /* The end of the pattern */
1573 ph10 459 open_capitem *open_caps; /* Chain of open capture items */
1574 nigel 93 uschar *hwm; /* High watermark of workspace */
1575 nigel 77 uschar *name_table; /* The name/number table */
1576     int names_found; /* Number of entries so far */
1577     int name_entry_size; /* Size of each entry */
1578 ph10 270 int bracount; /* Count of capturing parens as we compile */
1579 ph10 286 int final_bracount; /* Saved value after first pass */
1580 nigel 77 int top_backref; /* Maximum back reference */
1581     unsigned int backref_map; /* Bitmap of low back refs */
1582 nigel 93 int external_options; /* External (initial) options */
1583 ph10 231 int external_flags; /* External flag bits to be set */
1584 nigel 77 int req_varyopt; /* "After variable item" flag for reqbyte */
1585 ph10 211 BOOL had_accept; /* (*ACCEPT) encountered */
1586 ph10 459 BOOL check_lookbehind; /* Lookbehinds need later checking */
1587 nigel 93 int nltype; /* Newline type */
1588     int nllen; /* Newline string length */
1589     uschar nl[4]; /* Newline string when fixed length */
1590 nigel 77 } compile_data;
1591    
1592     /* Structure for maintaining a chain of pointers to the currently incomplete
1593     branches, for testing for left recursion. */
1594    
1595     typedef struct branch_chain {
1596     struct branch_chain *outer;
1597 ph10 475 uschar *current_branch;
1598 nigel 77 } branch_chain;
1599    
1600     /* Structure for items in a linked list that represents an explicit recursive
1601     call within the pattern. */
1602    
1603     typedef struct recursion_info {
1604     struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
1605     int group_num; /* Number of group that was called */
1606     const uschar *after_call; /* "Return value": points after the call in the expr */
1607 ph10 168 USPTR save_start; /* Old value of mstart */
1608 nigel 77 int *offset_save; /* Pointer to start of saved offsets */
1609     int saved_max; /* Number of saved offsets */
1610 ph10 461 int save_offset_top; /* Current value of offset_top */
1611 nigel 77 } recursion_info;
1612    
1613 nigel 93 /* Structure for building a chain of data for holding the values of the subject
1614     pointer at the start of each subpattern, so as to detect when an empty string
1615     has been matched by a subpattern - to break infinite loops. */
1616    
1617     typedef struct eptrblock {
1618     struct eptrblock *epb_prev;
1619     USPTR epb_saved_eptr;
1620     } eptrblock;
1621    
1622    
1623 nigel 77 /* Structure for passing "static" information around between the functions
1624     doing traditional NFA matching, so that they are thread-safe. */
1625    
1626     typedef struct match_data {
1627 nigel 87 unsigned long int match_call_count; /* As it says */
1628     unsigned long int match_limit; /* As it says */
1629     unsigned long int match_limit_recursion; /* As it says */
1630 nigel 77 int *offset_vector; /* Offset vector */
1631     int offset_end; /* One past the end */
1632     int offset_max; /* The maximum usable for return data */
1633 nigel 93 int nltype; /* Newline type */
1634     int nllen; /* Newline string length */
1635 ph10 459 int name_count; /* Number of names in name table */
1636     int name_entry_size; /* Size of entry in names table */
1637 ph10 461 uschar *name_table; /* Table of names */
1638 nigel 93 uschar nl[4]; /* Newline string when fixed */
1639 nigel 77 const uschar *lcc; /* Points to lower casing table */
1640     const uschar *ctypes; /* Points to table of type maps */
1641     BOOL offset_overflow; /* Set if too many extractions */
1642     BOOL notbol; /* NOTBOL flag */
1643     BOOL noteol; /* NOTEOL flag */
1644     BOOL utf8; /* UTF8 flag */
1645 ph10 345 BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */
1646 nigel 77 BOOL endonly; /* Dollar not before final \n */
1647     BOOL notempty; /* Empty string match not wanted */
1648 ph10 442 BOOL notempty_atstart; /* Empty string match at start not wanted */
1649 nigel 77 BOOL hitend; /* Hit the end of the subject at some point */
1650 ph10 231 BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
1651 nigel 77 const uschar *start_code; /* For use when recursing */
1652 nigel 87 USPTR start_subject; /* Start of the subject string */
1653     USPTR end_subject; /* End of the subject string */
1654 ph10 168 USPTR start_match_ptr; /* Start of matched string */
1655 nigel 87 USPTR end_match_ptr; /* Subject position at end match */
1656 ph10 442 USPTR start_used_ptr; /* Earliest consulted character */
1657 ph10 435 int partial; /* PARTIAL options */
1658 nigel 77 int end_offset_top; /* Highwater mark at end of match */
1659     int capture_last; /* Most recent capture number */
1660     int start_offset; /* The start offset value */
1661 nigel 93 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
1662     int eptrn; /* Next free eptrblock */
1663 nigel 77 recursion_info *recursive; /* Linked list of recursion data */
1664     void *callout_data; /* To pass back to callouts */
1665     } match_data;
1666    
1667     /* A similar structure is used for the same purpose by the DFA matching
1668     functions. */
1669    
1670     typedef struct dfa_match_data {
1671     const uschar *start_code; /* Start of the compiled pattern */
1672     const uschar *start_subject; /* Start of the subject string */
1673     const uschar *end_subject; /* End of subject string */
1674 ph10 442 const uschar *start_used_ptr; /* Earliest consulted character */
1675 nigel 77 const uschar *tables; /* Character tables */
1676 ph10 442 int start_offset; /* The start offset value */
1677 nigel 77 int moptions; /* Match options */
1678     int poptions; /* Pattern options */
1679 nigel 93 int nltype; /* Newline type */
1680     int nllen; /* Newline string length */
1681     uschar nl[4]; /* Newline string when fixed */
1682 nigel 77 void *callout_data; /* To pass back to callouts */
1683     } dfa_match_data;
1684    
1685     /* Bit definitions for entries in the pcre_ctypes table. */
1686    
1687     #define ctype_space 0x01
1688     #define ctype_letter 0x02
1689     #define ctype_digit 0x04
1690     #define ctype_xdigit 0x08
1691 ph10 274 #define ctype_word 0x10 /* alphanumeric or '_' */
1692 nigel 77 #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
1693    
1694     /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
1695     of bits for a class map. Some classes are built by combining these tables. */
1696    
1697     #define cbit_space 0 /* [:space:] or \s */
1698     #define cbit_xdigit 32 /* [:xdigit:] */
1699     #define cbit_digit 64 /* [:digit:] or \d */
1700     #define cbit_upper 96 /* [:upper:] */
1701     #define cbit_lower 128 /* [:lower:] */
1702     #define cbit_word 160 /* [:word:] or \w */
1703     #define cbit_graph 192 /* [:graph:] */
1704     #define cbit_print 224 /* [:print:] */
1705     #define cbit_punct 256 /* [:punct:] */
1706     #define cbit_cntrl 288 /* [:cntrl:] */
1707     #define cbit_length 320 /* Length of the cbits table */
1708    
1709     /* Offsets of the various tables from the base tables pointer, and
1710     total length. */
1711    
1712     #define lcc_offset 0
1713     #define fcc_offset 256
1714     #define cbits_offset 512
1715     #define ctypes_offset (cbits_offset + cbit_length)
1716     #define tables_length (ctypes_offset + 256)
1717    
1718 nigel 87 /* Layout of the UCP type table that translates property names into types and
1719 ph10 240 codes. Each entry used to point directly to a name, but to reduce the number of
1720 ph10 243 relocations in shared libraries, it now has an offset into a single string
1721 ph10 240 instead. */
1722 nigel 77
1723     typedef struct {
1724 ph10 243 pcre_uint16 name_offset;
1725 nigel 87 pcre_uint16 type;
1726     pcre_uint16 value;
1727 nigel 77 } ucp_type_table;
1728    
1729    
1730     /* Internal shared data tables. These are tables that are used by more than one
1731     of the exported public functions. They have to be "external" in the C sense,
1732     but are not part of the PCRE public API. The data for these tables is in the
1733     pcre_tables.c module. */
1734    
1735     extern const int _pcre_utf8_table1[];
1736     extern const int _pcre_utf8_table2[];
1737     extern const int _pcre_utf8_table3[];
1738     extern const uschar _pcre_utf8_table4[];
1739    
1740     extern const int _pcre_utf8_table1_size;
1741    
1742 ph10 240 extern const char _pcre_utt_names[];
1743 nigel 77 extern const ucp_type_table _pcre_utt[];
1744     extern const int _pcre_utt_size;
1745    
1746     extern const uschar _pcre_default_tables[];
1747    
1748     extern const uschar _pcre_OP_lengths[];
1749    
1750    
1751     /* Internal shared functions. These are functions that are used by more than
1752     one of the exported public functions. They have to be "external" in the C
1753     sense, but are not part of the PCRE public API. */
1754    
1755 ph10 455 extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);
1756 ph10 483 extern BOOL _pcre_is_newline(USPTR, int, USPTR, int *, BOOL);
1757 ph10 455 extern int _pcre_ord2utf8(int, uschar *);
1758     extern real_pcre *_pcre_try_flipped(const real_pcre *, real_pcre *,
1759     const pcre_study_data *, pcre_study_data *);
1760 ph10 483 extern int _pcre_valid_utf8(USPTR, int);
1761     extern BOOL _pcre_was_newline(USPTR, int, USPTR, int *, BOOL);
1762 ph10 455 extern BOOL _pcre_xclass(int, const uschar *);
1763 nigel 77
1764 ph10 349
1765     /* Unicode character database (UCD) */
1766    
1767     typedef struct {
1768     uschar script;
1769     uschar chartype;
1770 ph10 351 pcre_int32 other_case;
1771 ph10 349 } ucd_record;
1772    
1773 ph10 351 extern const ucd_record _pcre_ucd_records[];
1774     extern const uschar _pcre_ucd_stage1[];
1775     extern const pcre_uint16 _pcre_ucd_stage2[];
1776     extern const int _pcre_ucp_gentype[];
1777 ph10 349
1778 ph10 351
1779 ph10 349 /* UCD access macros */
1780    
1781     #define UCD_BLOCK_SIZE 128
1782 ph10 351 #define GET_UCD(ch) (_pcre_ucd_records + \
1783     _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \
1784 ph10 349 UCD_BLOCK_SIZE + ch % UCD_BLOCK_SIZE])
1785    
1786     #define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype
1787     #define UCD_SCRIPT(ch) GET_UCD(ch)->script
1788 ph10 351 #define UCD_CATEGORY(ch) _pcre_ucp_gentype[UCD_CHARTYPE(ch)]
1789     #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)
1790 ph10 349
1791 nigel 85 #endif
1792    
1793 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