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

Contents of /code/trunk/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log


Revision 454 - (hide annotations) (download)
Tue Sep 22 09:42:11 2009 UTC (4 years, 10 months ago) by ph10
File MIME type: text/plain
File size: 72837 byte(s)
Allow fixed-length subroutine calls in lookbehinds.

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 376 Copyright (c) 1997-2009 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     /* Define DEBUG to get debugging output on stdout. */
49    
50 nigel 85 #if 0
51 nigel 77 #define 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 nigel 77 #ifdef DEBUG
78     #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 <setjmp.h>
90     #include <stdarg.h>
91     #include <stddef.h>
92     #include <stdio.h>
93     #include <stdlib.h>
94     #include <string.h>
95    
96 ph10 145 /* When compiling a DLL for Windows, the exported symbols have to be declared
97     using some MS magic. I found some useful information on this web page:
98     http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
99 ph10 154 information there, using __declspec(dllexport) without "extern" we have a
100 ph10 145 definition; with "extern" we have a declaration. The settings here override the
101     setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL,
102 ph10 154 which is all that is needed for applications (they just import the symbols). We
103     use:
104 ph10 145
105     PCRE_EXP_DECL for declarations
106     PCRE_EXP_DEFN for definitions of exported functions
107     PCRE_EXP_DATA_DEFN for definitions of exported variables
108    
109     The reason for the two DEFN macros is that in non-Windows environments, one
110     does not want to have "extern" before variable definitions because it leads to
111     compiler warnings. So we distinguish between functions and variables. In
112     Windows, the two should always be the same.
113    
114     The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest,
115     which is an application, but needs to import this file in order to "peek" at
116 ph10 154 internals, can #include pcre.h first to get an application's-eye view.
117 ph10 145
118     In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon,
119     special-purpose environments) might want to stick other stuff in front of
120     exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and
121     PCRE_EXP_DATA_DEFN only if they are not already set. */
122    
123     #ifndef PCRE_EXP_DECL
124     # ifdef _WIN32
125 ph10 199 # ifndef PCRE_STATIC
126 ph10 145 # define PCRE_EXP_DECL extern __declspec(dllexport)
127     # define PCRE_EXP_DEFN __declspec(dllexport)
128     # define PCRE_EXP_DATA_DEFN __declspec(dllexport)
129     # else
130     # define PCRE_EXP_DECL extern
131     # define PCRE_EXP_DEFN
132     # define PCRE_EXP_DATA_DEFN
133     # endif
134     # else
135     # ifdef __cplusplus
136     # define PCRE_EXP_DECL extern "C"
137     # else
138     # define PCRE_EXP_DECL extern
139     # endif
140     # ifndef PCRE_EXP_DEFN
141     # define PCRE_EXP_DEFN PCRE_EXP_DECL
142     # endif
143     # ifndef PCRE_EXP_DATA_DEFN
144     # define PCRE_EXP_DATA_DEFN
145     # endif
146     # endif
147 nigel 77 #endif
148    
149 ph10 359 /* When compiling with the MSVC compiler, it is sometimes necessary to include
150 ph10 371 a "calling convention" before exported function names. (This is secondhand
151 ph10 359 information; I know nothing about MSVC myself). For example, something like
152    
153     void __cdecl function(....)
154 ph10 371
155 ph10 359 might be needed. In order so make this easy, all the exported functions have
156     PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
157 ph10 371 set, we ensure here that it has no effect. */
158 ph10 359
159     #ifndef PCRE_CALL_CONVENTION
160     #define PCRE_CALL_CONVENTION
161     #endif
162    
163 nigel 77 /* We need to have types that specify unsigned 16-bit and 32-bit integers. We
164     cannot determine these outside the compilation (e.g. by running a program as
165     part of "configure") because PCRE is often cross-compiled for use on other
166     systems. Instead we make use of the maximum sizes that are available at
167     preprocessor time in standard C environments. */
168    
169     #if USHRT_MAX == 65535
170     typedef unsigned short pcre_uint16;
171 ph10 349 typedef short pcre_int16;
172 nigel 77 #elif UINT_MAX == 65535
173     typedef unsigned int pcre_uint16;
174 ph10 349 typedef int pcre_int16;
175 nigel 77 #else
176     #error Cannot determine a type for 16-bit unsigned integers
177     #endif
178    
179     #if UINT_MAX == 4294967295
180     typedef unsigned int pcre_uint32;
181 ph10 349 typedef int pcre_int32;
182 nigel 77 #elif ULONG_MAX == 4294967295
183     typedef unsigned long int pcre_uint32;
184 ph10 349 typedef long int pcre_int32;
185 nigel 77 #else
186     #error Cannot determine a type for 32-bit unsigned integers
187     #endif
188    
189     /* All character handling must be done as unsigned characters. Otherwise there
190     are problems with top-bit-set characters and functions such as isspace().
191     However, we leave the interface to the outside world as char *, because that
192     should make things easier for callers. We define a short type for unsigned char
193     to save lots of typing. I tried "uchar", but it causes problems on Digital
194     Unix, where it is defined in sys/types, so use "uschar" instead. */
195    
196     typedef unsigned char uschar;
197    
198 nigel 93 /* This is an unsigned int value that no character can ever have. UTF-8
199     characters only go up to 0x7fffffff (though Unicode doesn't go beyond
200     0x0010ffff). */
201 nigel 91
202 nigel 93 #define NOTACHAR 0xffffffff
203    
204     /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
205 ph10 149 "any" and "anycrlf" at present). The following macros are used to package up
206     testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
207     modules to indicate in which datablock the parameters exist, and what the
208     start/end of string field names are. */
209 nigel 93
210 ph10 149 #define NLTYPE_FIXED 0 /* Newline is a fixed length string */
211     #define NLTYPE_ANY 1 /* Newline is any Unicode line ending */
212     #define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */
213 nigel 93
214     /* This macro checks for a newline at the given position */
215    
216 nigel 91 #define IS_NEWLINE(p) \
217 nigel 93 ((NLBLOCK->nltype != NLTYPE_FIXED)? \
218     ((p) < NLBLOCK->PSEND && \
219 ph10 149 _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\
220     utf8)) \
221 nigel 93 : \
222     ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
223     (p)[0] == NLBLOCK->nl[0] && \
224     (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \
225     ) \
226     )
227 nigel 91
228 nigel 93 /* This macro checks for a newline immediately preceding the given position */
229    
230     #define WAS_NEWLINE(p) \
231     ((NLBLOCK->nltype != NLTYPE_FIXED)? \
232     ((p) > NLBLOCK->PSSTART && \
233 ph10 149 _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
234     &(NLBLOCK->nllen), utf8)) \
235 nigel 93 : \
236     ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
237     (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \
238     (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \
239     ) \
240     )
241    
242 nigel 87 /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
243     with a custom type. This makes it possible, for example, to allow pcre_exec()
244     to process subject strings that are discontinuous by using a smart pointer
245     class. It must always be possible to inspect all of the subject string in
246     pcre_exec() because of the way it backtracks. Two macros are required in the
247     normal case, for sign-unspecified and unsigned char pointers. The former is
248     used for the external interface and appears in pcre.h, which is why its name
249     must begin with PCRE_. */
250    
251     #ifdef CUSTOM_SUBJECT_PTR
252     #define PCRE_SPTR CUSTOM_SUBJECT_PTR
253     #define USPTR CUSTOM_SUBJECT_PTR
254     #else
255     #define PCRE_SPTR const char *
256     #define USPTR const unsigned char *
257     #endif
258    
259 ph10 145
260    
261 nigel 85 /* Include the public PCRE header and the definitions of UCP character property
262     values. */
263 nigel 77
264 ph10 189 #include "pcre.h"
265 nigel 77 #include "ucp.h"
266    
267     /* When compiling for use with the Virtual Pascal compiler, these functions
268     need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
269     option on the command line. */
270    
271     #ifdef VPCOMPAT
272 ph10 101 #define strlen(s) _strlen(s)
273 nigel 77 #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
274 ph10 101 #define memcmp(s,c,n) _memcmp(s,c,n)
275 nigel 77 #define memcpy(d,s,n) _memcpy(d,s,n)
276     #define memmove(d,s,n) _memmove(d,s,n)
277     #define memset(s,c,n) _memset(s,c,n)
278     #else /* VPCOMPAT */
279    
280     /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
281     define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
282     is set. Otherwise, include an emulating function for those systems that have
283 ph10 118 neither (there some non-Unix environments where this is the case). */
284 nigel 77
285 ph10 137 #ifndef HAVE_MEMMOVE
286 nigel 77 #undef memmove /* some systems may have a macro */
287 ph10 137 #ifdef HAVE_BCOPY
288 nigel 77 #define memmove(a, b, c) bcopy(b, a, c)
289     #else /* HAVE_BCOPY */
290 nigel 91 static void *
291 ph10 108 pcre_memmove(void *d, const void *s, size_t n)
292 nigel 77 {
293 nigel 87 size_t i;
294 ph10 108 unsigned char *dest = (unsigned char *)d;
295     const unsigned char *src = (const unsigned char *)s;
296 ph10 118 if (dest > src)
297     {
298     dest += n;
299     src += n;
300     for (i = 0; i < n; ++i) *(--dest) = *(--src);
301     return (void *)dest;
302     }
303     else
304     {
305     for (i = 0; i < n; ++i) *dest++ = *src++;
306 ph10 123 return (void *)(dest - n);
307     }
308 nigel 77 }
309     #define memmove(a, b, c) pcre_memmove(a, b, c)
310     #endif /* not HAVE_BCOPY */
311     #endif /* not HAVE_MEMMOVE */
312     #endif /* not VPCOMPAT */
313    
314    
315     /* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
316     in big-endian order) by default. These are used, for example, to link from the
317     start of a subpattern to its alternatives and its end. The use of 2 bytes per
318     offset limits the size of the compiled regex to around 64K, which is big enough
319     for almost everybody. However, I received a request for an even bigger limit.
320     For this reason, and also to make the code easier to maintain, the storing and
321     loading of offsets from the byte string is now handled by the macros that are
322     defined here.
323    
324     The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
325     the config.h file, but can be overridden by using -D on the command line. This
326     is automated on Unix systems via the "configure" command. */
327    
328     #if LINK_SIZE == 2
329    
330     #define PUT(a,n,d) \
331     (a[n] = (d) >> 8), \
332     (a[(n)+1] = (d) & 255)
333    
334     #define GET(a,n) \
335     (((a)[n] << 8) | (a)[(n)+1])
336    
337     #define MAX_PATTERN_SIZE (1 << 16)
338    
339    
340     #elif LINK_SIZE == 3
341    
342     #define PUT(a,n,d) \
343     (a[n] = (d) >> 16), \
344     (a[(n)+1] = (d) >> 8), \
345     (a[(n)+2] = (d) & 255)
346    
347     #define GET(a,n) \
348     (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
349    
350     #define MAX_PATTERN_SIZE (1 << 24)
351    
352    
353     #elif LINK_SIZE == 4
354    
355     #define PUT(a,n,d) \
356     (a[n] = (d) >> 24), \
357     (a[(n)+1] = (d) >> 16), \
358     (a[(n)+2] = (d) >> 8), \
359     (a[(n)+3] = (d) & 255)
360    
361     #define GET(a,n) \
362     (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
363    
364     #define MAX_PATTERN_SIZE (1 << 30) /* Keep it positive */
365    
366    
367     #else
368     #error LINK_SIZE must be either 2, 3, or 4
369     #endif
370    
371    
372     /* Convenience macro defined in terms of the others */
373    
374     #define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
375    
376    
377     /* PCRE uses some other 2-byte quantities that do not change when the size of
378     offsets changes. There are used for repeat counts and for other things such as
379     capturing parenthesis numbers in back references. */
380    
381     #define PUT2(a,n,d) \
382     a[n] = (d) >> 8; \
383     a[(n)+1] = (d) & 255
384    
385     #define GET2(a,n) \
386     (((a)[n] << 8) | (a)[(n)+1])
387    
388     #define PUT2INC(a,n,d) PUT2(a,n,d), a += 2
389    
390    
391     /* When UTF-8 encoding is being used, a character is no longer just a single
392     byte. The macros for character handling generate simple sequences when used in
393 ph10 208 byte-mode, and more complicated ones for UTF-8 characters. BACKCHAR should
394     never be called in byte mode. To make sure it can never even appear when UTF-8
395 ph10 207 support is omitted, we don't even define it. */
396 nigel 77
397     #ifndef SUPPORT_UTF8
398     #define GETCHAR(c, eptr) c = *eptr;
399     #define GETCHARTEST(c, eptr) c = *eptr;
400     #define GETCHARINC(c, eptr) c = *eptr++;
401     #define GETCHARINCTEST(c, eptr) c = *eptr++;
402     #define GETCHARLEN(c, eptr, len) c = *eptr;
403 ph10 207 /* #define BACKCHAR(eptr) */
404 nigel 77
405     #else /* SUPPORT_UTF8 */
406    
407     /* Get the next UTF-8 character, not advancing the pointer. This is called when
408     we know we are in UTF-8 mode. */
409    
410     #define GETCHAR(c, eptr) \
411     c = *eptr; \
412 nigel 93 if (c >= 0xc0) \
413 nigel 77 { \
414     int gcii; \
415     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
416     int gcss = 6*gcaa; \
417     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
418     for (gcii = 1; gcii <= gcaa; gcii++) \
419     { \
420     gcss -= 6; \
421     c |= (eptr[gcii] & 0x3f) << gcss; \
422     } \
423     }
424    
425     /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
426     pointer. */
427    
428     #define GETCHARTEST(c, eptr) \
429     c = *eptr; \
430 nigel 93 if (utf8 && 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, advancing the pointer. This is called when we
444     know we are in UTF-8 mode. */
445    
446     #define GETCHARINC(c, eptr) \
447     c = *eptr++; \
448 nigel 93 if (c >= 0xc0) \
449 nigel 77 { \
450     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
451     int gcss = 6*gcaa; \
452     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
453     while (gcaa-- > 0) \
454     { \
455     gcss -= 6; \
456     c |= (*eptr++ & 0x3f) << gcss; \
457     } \
458     }
459    
460     /* Get the next character, testing for UTF-8 mode, and advancing the pointer */
461    
462     #define GETCHARINCTEST(c, eptr) \
463     c = *eptr++; \
464 nigel 93 if (utf8 && c >= 0xc0) \
465 nigel 77 { \
466     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
467     int gcss = 6*gcaa; \
468     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
469     while (gcaa-- > 0) \
470     { \
471     gcss -= 6; \
472     c |= (*eptr++ & 0x3f) << gcss; \
473     } \
474     }
475    
476     /* Get the next UTF-8 character, not advancing the pointer, incrementing length
477     if there are extra bytes. This is called when we know we are in UTF-8 mode. */
478    
479     #define GETCHARLEN(c, eptr, len) \
480     c = *eptr; \
481 nigel 93 if (c >= 0xc0) \
482 nigel 77 { \
483     int gcii; \
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     for (gcii = 1; gcii <= gcaa; gcii++) \
488     { \
489     gcss -= 6; \
490     c |= (eptr[gcii] & 0x3f) << gcss; \
491     } \
492     len += gcaa; \
493     }
494    
495 ph10 384 /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
496     pointer, incrementing length if there are extra bytes. This is called when we
497     know we are in UTF-8 mode. */
498    
499     #define GETCHARLENTEST(c, eptr, len) \
500     c = *eptr; \
501     if (utf8 && c >= 0xc0) \
502     { \
503     int gcii; \
504     int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
505     int gcss = 6*gcaa; \
506     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \
507     for (gcii = 1; gcii <= gcaa; gcii++) \
508     { \
509     gcss -= 6; \
510     c |= (eptr[gcii] & 0x3f) << gcss; \
511     } \
512     len += gcaa; \
513     }
514    
515 nigel 77 /* If the pointer is not at the start of a character, move it back until
516 ph10 208 it is. This is called only in UTF-8 mode - we don't put a test within the macro
517 ph10 207 because almost all calls are already within a block of UTF-8 only code. */
518 nigel 77
519 ph10 207 #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
520 nigel 77
521     #endif
522    
523    
524     /* In case there is no definition of offsetof() provided - though any proper
525     Standard C system should have one. */
526    
527     #ifndef offsetof
528     #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))
529     #endif
530    
531    
532     /* These are the public options that can change during matching. */
533    
534     #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)
535    
536 ph10 231 /* Private flags containing information about the compiled regex. They used to
537     live at the top end of the options word, but that got almost full, so now they
538 ph10 426 are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
539 ph10 427 the restrictions on partial matching have been lifted. It remains for backwards
540 ph10 426 compatibility. */
541 nigel 77
542 ph10 230 #define PCRE_NOPARTIAL 0x0001 /* can't use partial with this regex */
543     #define PCRE_FIRSTSET 0x0002 /* first_byte is set */
544     #define PCRE_REQCHSET 0x0004 /* req_byte is set */
545     #define PCRE_STARTLINE 0x0008 /* start after \n for multiline */
546     #define PCRE_JCHANGED 0x0010 /* j option used in regex */
547     #define PCRE_HASCRORLF 0x0020 /* explicit \r or \n in pattern */
548 nigel 77
549     /* Options for the "extra" block produced by pcre_study(). */
550    
551     #define PCRE_STUDY_MAPPED 0x01 /* a map of starting chars exists */
552    
553     /* Masks for identifying the public options that are permitted at compile
554     time, run time, or study time, respectively. */
555    
556 ph10 149 #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
557     PCRE_NEWLINE_ANYCRLF)
558 nigel 93
559 ph10 389 #define PUBLIC_COMPILE_OPTIONS \
560 nigel 77 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
561     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
562 nigel 91 PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
563 ph10 336 PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
564     PCRE_JAVASCRIPT_COMPAT)
565 nigel 77
566     #define PUBLIC_EXEC_OPTIONS \
567 ph10 442 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
568     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
569     PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
570 nigel 77
571     #define PUBLIC_DFA_EXEC_OPTIONS \
572 ph10 442 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
573     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
574     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
575     PCRE_NO_START_OPTIMIZE)
576 nigel 77
577     #define PUBLIC_STUDY_OPTIONS 0 /* None defined */
578    
579     /* Magic number to provide a small check against being handed junk. Also used
580     to detect whether a pattern was compiled on a host of different endianness. */
581    
582     #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
583    
584     /* Negative values for the firstchar and reqchar variables */
585    
586     #define REQ_UNSET (-2)
587     #define REQ_NONE (-1)
588    
589     /* The maximum remaining length of subject we are prepared to search for a
590     req_byte match. */
591    
592     #define REQ_BYTE_MAX 1000
593    
594     /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a
595     variable-length repeat, or a anything other than literal characters. */
596    
597     #define REQ_CASELESS 0x0100 /* indicates caselessness */
598     #define REQ_VARY 0x0200 /* reqbyte followed non-literal item */
599    
600 ph10 369 /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
601 ph10 376 environments where these macros are defined elsewhere. Unfortunately, there
602     is no way to do the same for the typedef. */
603 nigel 77
604     typedef int BOOL;
605    
606 ph10 376 #ifndef FALSE
607 nigel 77 #define FALSE 0
608     #define TRUE 1
609 ph10 369 #endif
610 nigel 77
611 ph10 391 /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
612     character constants like '*' because the compiler would emit their EBCDIC code,
613 ph10 392 which is different from their ASCII/UTF-8 code. Instead we define macros for
614     the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
615     is enabled. When UTF-8 support is not enabled, the definitions use character
616     literals. Both character and string versions of each character are needed, and
617 ph10 391 there are some longer strings as well.
618    
619 ph10 392 This means that, on EBCDIC platforms, the PCRE library can handle either
620 ph10 391 EBCDIC, or UTF-8, but not both. To support both in the same compiled library
621     would need different lookups depending on whether PCRE_UTF8 was set or not.
622     This would make it impossible to use characters in switch/case statements,
623 ph10 392 which would reduce performance. For a theoretical use (which nobody has asked
624     for) in a minority area (EBCDIC platforms), this is not sensible. Any
625     application that did need both could compile two versions of the library, using
626 ph10 391 macros to give the functions distinct names. */
627    
628     #ifndef SUPPORT_UTF8
629    
630     /* UTF-8 support is not enabled; use the platform-dependent character literals
631     so that PCRE works on both ASCII and EBCDIC platforms, in non-UTF-mode only. */
632    
633     #define CHAR_HT '\t'
634     #define CHAR_VT '\v'
635     #define CHAR_FF '\f'
636     #define CHAR_CR '\r'
637     #define CHAR_NL '\n'
638     #define CHAR_BS '\b'
639     #define CHAR_BEL '\a'
640     #ifdef EBCDIC
641     #define CHAR_ESC '\047'
642     #define CHAR_DEL '\007'
643     #else
644     #define CHAR_ESC '\033'
645     #define CHAR_DEL '\177'
646     #endif
647    
648     #define CHAR_SPACE ' '
649     #define CHAR_EXCLAMATION_MARK '!'
650     #define CHAR_QUOTATION_MARK '"'
651     #define CHAR_NUMBER_SIGN '#'
652     #define CHAR_DOLLAR_SIGN '$'
653     #define CHAR_PERCENT_SIGN '%'
654     #define CHAR_AMPERSAND '&'
655     #define CHAR_APOSTROPHE '\''
656     #define CHAR_LEFT_PARENTHESIS '('
657     #define CHAR_RIGHT_PARENTHESIS ')'
658     #define CHAR_ASTERISK '*'
659     #define CHAR_PLUS '+'
660     #define CHAR_COMMA ','
661     #define CHAR_MINUS '-'
662     #define CHAR_DOT '.'
663     #define CHAR_SLASH '/'
664     #define CHAR_0 '0'
665     #define CHAR_1 '1'
666     #define CHAR_2 '2'
667     #define CHAR_3 '3'
668     #define CHAR_4 '4'
669     #define CHAR_5 '5'
670     #define CHAR_6 '6'
671     #define CHAR_7 '7'
672     #define CHAR_8 '8'
673     #define CHAR_9 '9'
674     #define CHAR_COLON ':'
675     #define CHAR_SEMICOLON ';'
676     #define CHAR_LESS_THAN_SIGN '<'
677     #define CHAR_EQUALS_SIGN '='
678     #define CHAR_GREATER_THAN_SIGN '>'
679     #define CHAR_QUESTION_MARK '?'
680     #define CHAR_COMMERCIAL_AT '@'
681     #define CHAR_A 'A'
682     #define CHAR_B 'B'
683     #define CHAR_C 'C'
684     #define CHAR_D 'D'
685     #define CHAR_E 'E'
686     #define CHAR_F 'F'
687     #define CHAR_G 'G'
688     #define CHAR_H 'H'
689     #define CHAR_I 'I'
690     #define CHAR_J 'J'
691     #define CHAR_K 'K'
692     #define CHAR_L 'L'
693     #define CHAR_M 'M'
694     #define CHAR_N 'N'
695     #define CHAR_O 'O'
696     #define CHAR_P 'P'
697     #define CHAR_Q 'Q'
698     #define CHAR_R 'R'
699     #define CHAR_S 'S'
700     #define CHAR_T 'T'
701     #define CHAR_U 'U'
702     #define CHAR_V 'V'
703     #define CHAR_W 'W'
704     #define CHAR_X 'X'
705     #define CHAR_Y 'Y'
706     #define CHAR_Z 'Z'
707     #define CHAR_LEFT_SQUARE_BRACKET '['
708     #define CHAR_BACKSLASH '\\'
709     #define CHAR_RIGHT_SQUARE_BRACKET ']'
710     #define CHAR_CIRCUMFLEX_ACCENT '^'
711     #define CHAR_UNDERSCORE '_'
712     #define CHAR_GRAVE_ACCENT '`'
713     #define CHAR_a 'a'
714     #define CHAR_b 'b'
715     #define CHAR_c 'c'
716     #define CHAR_d 'd'
717     #define CHAR_e 'e'
718     #define CHAR_f 'f'
719     #define CHAR_g 'g'
720     #define CHAR_h 'h'
721     #define CHAR_i 'i'
722     #define CHAR_j 'j'
723     #define CHAR_k 'k'
724     #define CHAR_l 'l'
725     #define CHAR_m 'm'
726     #define CHAR_n 'n'
727     #define CHAR_o 'o'
728     #define CHAR_p 'p'
729     #define CHAR_q 'q'
730     #define CHAR_r 'r'
731     #define CHAR_s 's'
732     #define CHAR_t 't'
733     #define CHAR_u 'u'
734     #define CHAR_v 'v'
735     #define CHAR_w 'w'
736     #define CHAR_x 'x'
737     #define CHAR_y 'y'
738     #define CHAR_z 'z'
739     #define CHAR_LEFT_CURLY_BRACKET '{'
740     #define CHAR_VERTICAL_LINE '|'
741     #define CHAR_RIGHT_CURLY_BRACKET '}'
742     #define CHAR_TILDE '~'
743    
744     #define STR_HT "\t"
745     #define STR_VT "\v"
746     #define STR_FF "\f"
747     #define STR_CR "\r"
748     #define STR_NL "\n"
749     #define STR_BS "\b"
750     #define STR_BEL "\a"
751     #ifdef EBCDIC
752     #define STR_ESC "\047"
753     #define STR_DEL "\007"
754     #else
755     #define STR_ESC "\033"
756     #define STR_DEL "\177"
757     #endif
758    
759     #define STR_SPACE " "
760     #define STR_EXCLAMATION_MARK "!"
761     #define STR_QUOTATION_MARK "\""
762     #define STR_NUMBER_SIGN "#"
763     #define STR_DOLLAR_SIGN "$"
764     #define STR_PERCENT_SIGN "%"
765     #define STR_AMPERSAND "&"
766     #define STR_APOSTROPHE "'"
767     #define STR_LEFT_PARENTHESIS "("
768     #define STR_RIGHT_PARENTHESIS ")"
769     #define STR_ASTERISK "*"
770     #define STR_PLUS "+"
771     #define STR_COMMA ","
772     #define STR_MINUS "-"
773     #define STR_DOT "."
774     #define STR_SLASH "/"
775     #define STR_0 "0"
776     #define STR_1 "1"
777     #define STR_2 "2"
778     #define STR_3 "3"
779     #define STR_4 "4"
780     #define STR_5 "5"
781     #define STR_6 "6"
782     #define STR_7 "7"
783     #define STR_8 "8"
784     #define STR_9 "9"
785     #define STR_COLON ":"
786     #define STR_SEMICOLON ";"
787     #define STR_LESS_THAN_SIGN "<"
788     #define STR_EQUALS_SIGN "="
789     #define STR_GREATER_THAN_SIGN ">"
790     #define STR_QUESTION_MARK "?"
791     #define STR_COMMERCIAL_AT "@"
792     #define STR_A "A"
793     #define STR_B "B"
794     #define STR_C "C"
795     #define STR_D "D"
796     #define STR_E "E"
797     #define STR_F "F"
798     #define STR_G "G"
799     #define STR_H "H"
800     #define STR_I "I"
801     #define STR_J "J"
802     #define STR_K "K"
803     #define STR_L "L"
804     #define STR_M "M"
805     #define STR_N "N"
806     #define STR_O "O"
807     #define STR_P "P"
808     #define STR_Q "Q"
809     #define STR_R "R"
810     #define STR_S "S"
811     #define STR_T "T"
812     #define STR_U "U"
813     #define STR_V "V"
814     #define STR_W "W"
815     #define STR_X "X"
816     #define STR_Y "Y"
817     #define STR_Z "Z"
818     #define STR_LEFT_SQUARE_BRACKET "["
819     #define STR_BACKSLASH "\\"
820     #define STR_RIGHT_SQUARE_BRACKET "]"
821     #define STR_CIRCUMFLEX_ACCENT "^"
822     #define STR_UNDERSCORE "_"
823     #define STR_GRAVE_ACCENT "`"
824     #define STR_a "a"
825     #define STR_b "b"
826     #define STR_c "c"
827     #define STR_d "d"
828     #define STR_e "e"
829     #define STR_f "f"
830     #define STR_g "g"
831     #define STR_h "h"
832     #define STR_i "i"
833     #define STR_j "j"
834     #define STR_k "k"
835     #define STR_l "l"
836     #define STR_m "m"
837     #define STR_n "n"
838     #define STR_o "o"
839     #define STR_p "p"
840     #define STR_q "q"
841     #define STR_r "r"
842     #define STR_s "s"
843     #define STR_t "t"
844     #define STR_u "u"
845     #define STR_v "v"
846     #define STR_w "w"
847     #define STR_x "x"
848     #define STR_y "y"
849     #define STR_z "z"
850     #define STR_LEFT_CURLY_BRACKET "{"
851     #define STR_VERTICAL_LINE "|"
852     #define STR_RIGHT_CURLY_BRACKET "}"
853     #define STR_TILDE "~"
854    
855     #define STRING_ACCEPT0 "ACCEPT\0"
856     #define STRING_COMMIT0 "COMMIT\0"
857     #define STRING_F0 "F\0"
858     #define STRING_FAIL0 "FAIL\0"
859     #define STRING_PRUNE0 "PRUNE\0"
860     #define STRING_SKIP0 "SKIP\0"
861     #define STRING_THEN "THEN"
862    
863     #define STRING_alpha0 "alpha\0"
864     #define STRING_lower0 "lower\0"
865     #define STRING_upper0 "upper\0"
866     #define STRING_alnum0 "alnum\0"
867     #define STRING_ascii0 "ascii\0"
868     #define STRING_blank0 "blank\0"
869     #define STRING_cntrl0 "cntrl\0"
870     #define STRING_digit0 "digit\0"
871     #define STRING_graph0 "graph\0"
872     #define STRING_print0 "print\0"
873     #define STRING_punct0 "punct\0"
874     #define STRING_space0 "space\0"
875     #define STRING_word0 "word\0"
876     #define STRING_xdigit "xdigit"
877    
878     #define STRING_DEFINE "DEFINE"
879    
880     #define STRING_CR_RIGHTPAR "CR)"
881     #define STRING_LF_RIGHTPAR "LF)"
882     #define STRING_CRLF_RIGHTPAR "CRLF)"
883     #define STRING_ANY_RIGHTPAR "ANY)"
884     #define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)"
885     #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"
886     #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"
887 ph10 412 #define STRING_UTF8_RIGHTPAR "UTF8)"
888 ph10 391
889     #else /* SUPPORT_UTF8 */
890    
891     /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
892     works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
893     only. */
894    
895     #define CHAR_HT '\011'
896     #define CHAR_VT '\013'
897     #define CHAR_FF '\014'
898     #define CHAR_CR '\015'
899     #define CHAR_NL '\012'
900     #define CHAR_BS '\010'
901     #define CHAR_BEL '\007'
902     #define CHAR_ESC '\033'
903     #define CHAR_DEL '\177'
904    
905     #define CHAR_SPACE '\040'
906     #define CHAR_EXCLAMATION_MARK '\041'
907     #define CHAR_QUOTATION_MARK '\042'
908     #define CHAR_NUMBER_SIGN '\043'
909     #define CHAR_DOLLAR_SIGN '\044'
910     #define CHAR_PERCENT_SIGN '\045'
911     #define CHAR_AMPERSAND '\046'
912     #define CHAR_APOSTROPHE '\047'
913     #define CHAR_LEFT_PARENTHESIS '\050'
914     #define CHAR_RIGHT_PARENTHESIS '\051'
915     #define CHAR_ASTERISK '\052'
916     #define CHAR_PLUS '\053'
917     #define CHAR_COMMA '\054'
918     #define CHAR_MINUS '\055'
919     #define CHAR_DOT '\056'
920     #define CHAR_SLASH '\057'
921     #define CHAR_0 '\060'
922     #define CHAR_1 '\061'
923     #define CHAR_2 '\062'
924     #define CHAR_3 '\063'
925     #define CHAR_4 '\064'
926     #define CHAR_5 '\065'
927     #define CHAR_6 '\066'
928     #define CHAR_7 '\067'
929     #define CHAR_8 '\070'
930     #define CHAR_9 '\071'
931     #define CHAR_COLON '\072'
932     #define CHAR_SEMICOLON '\073'
933     #define CHAR_LESS_THAN_SIGN '\074'
934     #define CHAR_EQUALS_SIGN '\075'
935     #define CHAR_GREATER_THAN_SIGN '\076'
936     #define CHAR_QUESTION_MARK '\077'
937     #define CHAR_COMMERCIAL_AT '\100'
938     #define CHAR_A '\101'
939     #define CHAR_B '\102'
940     #define CHAR_C '\103'
941     #define CHAR_D '\104'
942     #define CHAR_E '\105'
943     #define CHAR_F '\106'
944     #define CHAR_G '\107'
945     #define CHAR_H '\110'
946     #define CHAR_I '\111'
947     #define CHAR_J '\112'
948     #define CHAR_K '\113'
949     #define CHAR_L '\114'
950     #define CHAR_M '\115'
951     #define CHAR_N '\116'
952     #define CHAR_O '\117'
953     #define CHAR_P '\120'
954     #define CHAR_Q '\121'
955     #define CHAR_R '\122'
956     #define CHAR_S '\123'
957     #define CHAR_T '\124'
958     #define CHAR_U '\125'
959     #define CHAR_V '\126'
960     #define CHAR_W '\127'
961     #define CHAR_X '\130'
962     #define CHAR_Y '\131'
963     #define CHAR_Z '\132'
964     #define CHAR_LEFT_SQUARE_BRACKET '\133'
965     #define CHAR_BACKSLASH '\134'
966     #define CHAR_RIGHT_SQUARE_BRACKET '\135'
967     #define CHAR_CIRCUMFLEX_ACCENT '\136'
968     #define CHAR_UNDERSCORE '\137'
969     #define CHAR_GRAVE_ACCENT '\140'
970     #define CHAR_a '\141'
971     #define CHAR_b '\142'
972     #define CHAR_c '\143'
973     #define CHAR_d '\144'
974     #define CHAR_e '\145'
975     #define CHAR_f '\146'
976     #define CHAR_g '\147'
977     #define CHAR_h '\150'
978     #define CHAR_i '\151'
979     #define CHAR_j '\152'
980     #define CHAR_k '\153'
981     #define CHAR_l '\154'
982     #define CHAR_m '\155'
983     #define CHAR_n '\156'
984     #define CHAR_o '\157'
985     #define CHAR_p '\160'
986     #define CHAR_q '\161'
987     #define CHAR_r '\162'
988     #define CHAR_s '\163'
989     #define CHAR_t '\164'
990     #define CHAR_u '\165'
991     #define CHAR_v '\166'
992     #define CHAR_w '\167'
993     #define CHAR_x '\170'
994     #define CHAR_y '\171'
995     #define CHAR_z '\172'
996     #define CHAR_LEFT_CURLY_BRACKET '\173'
997     #define CHAR_VERTICAL_LINE '\174'
998     #define CHAR_RIGHT_CURLY_BRACKET '\175'
999     #define CHAR_TILDE '\176'
1000    
1001     #define STR_HT "\011"
1002     #define STR_VT "\013"
1003     #define STR_FF "\014"
1004     #define STR_CR "\015"
1005     #define STR_NL "\012"
1006     #define STR_BS "\010"
1007     #define STR_BEL "\007"
1008     #define STR_ESC "\033"
1009     #define STR_DEL "\177"
1010    
1011     #define STR_SPACE "\040"
1012     #define STR_EXCLAMATION_MARK "\041"
1013     #define STR_QUOTATION_MARK "\042"
1014     #define STR_NUMBER_SIGN "\043"
1015     #define STR_DOLLAR_SIGN "\044"
1016     #define STR_PERCENT_SIGN "\045"
1017     #define STR_AMPERSAND "\046"
1018     #define STR_APOSTROPHE "\047"
1019     #define STR_LEFT_PARENTHESIS "\050"
1020     #define STR_RIGHT_PARENTHESIS "\051"
1021     #define STR_ASTERISK "\052"
1022     #define STR_PLUS "\053"
1023     #define STR_COMMA "\054"
1024     #define STR_MINUS "\055"
1025     #define STR_DOT "\056"
1026     #define STR_SLASH "\057"
1027     #define STR_0 "\060"
1028     #define STR_1 "\061"
1029     #define STR_2 "\062"
1030     #define STR_3 "\063"
1031     #define STR_4 "\064"
1032     #define STR_5 "\065"
1033     #define STR_6 "\066"
1034     #define STR_7 "\067"
1035     #define STR_8 "\070"
1036     #define STR_9 "\071"
1037     #define STR_COLON "\072"
1038     #define STR_SEMICOLON "\073"
1039     #define STR_LESS_THAN_SIGN "\074"
1040     #define STR_EQUALS_SIGN "\075"
1041     #define STR_GREATER_THAN_SIGN "\076"
1042     #define STR_QUESTION_MARK "\077"
1043     #define STR_COMMERCIAL_AT "\100"
1044     #define STR_A "\101"
1045     #define STR_B "\102"
1046     #define STR_C "\103"
1047     #define STR_D "\104"
1048     #define STR_E "\105"
1049     #define STR_F "\106"
1050     #define STR_G "\107"
1051     #define STR_H "\110"
1052     #define STR_I "\111"
1053     #define STR_J "\112"
1054     #define STR_K "\113"
1055     #define STR_L "\114"
1056     #define STR_M "\115"
1057     #define STR_N "\116"
1058     #define STR_O "\117"
1059     #define STR_P "\120"
1060     #define STR_Q "\121"
1061     #define STR_R "\122"
1062     #define STR_S "\123"
1063     #define STR_T "\124"
1064     #define STR_U "\125"
1065     #define STR_V "\126"
1066     #define STR_W "\127"
1067     #define STR_X "\130"
1068     #define STR_Y "\131"
1069     #define STR_Z "\132"
1070     #define STR_LEFT_SQUARE_BRACKET "\133"
1071     #define STR_BACKSLASH "\134"
1072     #define STR_RIGHT_SQUARE_BRACKET "\135"
1073     #define STR_CIRCUMFLEX_ACCENT "\136"
1074     #define STR_UNDERSCORE "\137"
1075     #define STR_GRAVE_ACCENT "\140"
1076     #define STR_a "\141"
1077     #define STR_b "\142"
1078     #define STR_c "\143"
1079     #define STR_d "\144"
1080     #define STR_e "\145"
1081     #define STR_f "\146"
1082     #define STR_g "\147"
1083     #define STR_h "\150"
1084     #define STR_i "\151"
1085     #define STR_j "\152"
1086     #define STR_k "\153"
1087     #define STR_l "\154"
1088     #define STR_m "\155"
1089     #define STR_n "\156"
1090     #define STR_o "\157"
1091     #define STR_p "\160"
1092     #define STR_q "\161"
1093     #define STR_r "\162"
1094     #define STR_s "\163"
1095     #define STR_t "\164"
1096     #define STR_u "\165"
1097     #define STR_v "\166"
1098     #define STR_w "\167"
1099     #define STR_x "\170"
1100     #define STR_y "\171"
1101     #define STR_z "\172"
1102     #define STR_LEFT_CURLY_BRACKET "\173"
1103     #define STR_VERTICAL_LINE "\174"
1104     #define STR_RIGHT_CURLY_BRACKET "\175"
1105     #define STR_TILDE "\176"
1106    
1107     #define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1108     #define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1109     #define STRING_F0 STR_F "\0"
1110     #define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0"
1111     #define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0"
1112     #define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0"
1113     #define STRING_THEN STR_T STR_H STR_E STR_N
1114    
1115     #define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0"
1116     #define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0"
1117     #define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0"
1118     #define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0"
1119     #define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0"
1120     #define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0"
1121     #define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0"
1122     #define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0"
1123     #define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0"
1124     #define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0"
1125     #define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0"
1126     #define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0"
1127     #define STRING_word0 STR_w STR_o STR_r STR_d "\0"
1128     #define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t
1129    
1130     #define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E
1131    
1132     #define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS
1133     #define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS
1134     #define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1135     #define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1136     #define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1137     #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
1138     #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
1139 ph10 412 #define STRING_UTF8_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1140 ph10 391
1141     #endif /* SUPPORT_UTF8 */
1142    
1143 nigel 93 /* Escape items that are just an encoding of a particular data value. */
1144 nigel 77
1145     #ifndef ESC_e
1146 ph10 391 #define ESC_e CHAR_ESC
1147 nigel 77 #endif
1148    
1149     #ifndef ESC_f
1150 ph10 391 #define ESC_f CHAR_FF
1151 nigel 77 #endif
1152    
1153     #ifndef ESC_n
1154 ph10 391 #define ESC_n CHAR_NL
1155 nigel 77 #endif
1156    
1157     #ifndef ESC_r
1158 ph10 391 #define ESC_r CHAR_CR
1159 nigel 77 #endif
1160    
1161     /* We can't officially use ESC_t because it is a POSIX reserved identifier
1162     (presumably because of all the others like size_t). */
1163    
1164     #ifndef ESC_tee
1165 ph10 391 #define ESC_tee CHAR_HT
1166 nigel 77 #endif
1167    
1168 nigel 87 /* Codes for different types of Unicode property */
1169    
1170     #define PT_ANY 0 /* Any property - matches all chars */
1171     #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
1172     #define PT_GC 2 /* General characteristic (e.g. L) */
1173     #define PT_PC 3 /* Particular characteristic (e.g. Lu) */
1174     #define PT_SC 4 /* Script (e.g. Han) */
1175    
1176     /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1177     contain UTF-8 characters with values greater than 255. */
1178    
1179     #define XCL_NOT 0x01 /* Flag: this is a negative class */
1180     #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
1181    
1182     #define XCL_END 0 /* Marks end of individual items */
1183     #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
1184     #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
1185     #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
1186     #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
1187    
1188 nigel 77 /* These are escaped items that aren't just an encoding of a particular data
1189     value such as \n. They must have non-zero values, as check_escape() returns
1190     their negation. Also, they must appear in the same order as in the opcode
1191     definitions below, up to ESC_z. There's a dummy for OP_ANY because it
1192 ph10 345 corresponds to "." rather than an escape sequence, and another for OP_ALLANY
1193 ph10 341 (which is used for [^] in JavaScript compatibility mode).
1194 nigel 77
1195 ph10 341 The final escape must be ESC_REF as subsequent values are used for
1196     backreferences (\1, \2, \3, etc). There are two tests in the code for an escape
1197     greater than ESC_b and less than ESC_Z to detect the types that may be
1198     repeated. These are the types that consume characters. If any new escapes are
1199     put in between that don't consume a character, that code will have to change.
1200     */
1201    
1202 ph10 172 enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1203 ph10 345 ESC_W, ESC_w, ESC_dum1, ESC_dum2, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1204     ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_g, ESC_k,
1205 ph10 341 ESC_REF };
1206 nigel 77
1207 nigel 93
1208 ph10 210 /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1209 nigel 77 OP_EOD must correspond in order to the list of escapes immediately above.
1210    
1211 nigel 93 *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1212     that follow must also be updated to match. There is also a table called
1213     "coptable" in pcre_dfa_exec.c that must be updated. */
1214    
1215 nigel 77 enum {
1216     OP_END, /* 0 End of pattern */
1217    
1218     /* Values corresponding to backslashed metacharacters */
1219    
1220     OP_SOD, /* 1 Start of data: \A */
1221     OP_SOM, /* 2 Start of match (subject + offset): \G */
1222 ph10 172 OP_SET_SOM, /* 3 Set start of match (\K) */
1223 ph10 168 OP_NOT_WORD_BOUNDARY, /* 4 \B */
1224     OP_WORD_BOUNDARY, /* 5 \b */
1225     OP_NOT_DIGIT, /* 6 \D */
1226     OP_DIGIT, /* 7 \d */
1227     OP_NOT_WHITESPACE, /* 8 \S */
1228     OP_WHITESPACE, /* 9 \s */
1229     OP_NOT_WORDCHAR, /* 10 \W */
1230     OP_WORDCHAR, /* 11 \w */
1231 ph10 341 OP_ANY, /* 12 Match any character (subject to DOTALL) */
1232     OP_ALLANY, /* 13 Match any character (not subject to DOTALL) */
1233     OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1234     OP_NOTPROP, /* 15 \P (not Unicode property) */
1235     OP_PROP, /* 16 \p (Unicode property) */
1236     OP_ANYNL, /* 17 \R (any newline sequence) */
1237     OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */
1238     OP_HSPACE, /* 19 \h (horizontal whitespace) */
1239     OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */
1240     OP_VSPACE, /* 21 \v (vertical whitespace) */
1241     OP_EXTUNI, /* 22 \X (extended Unicode sequence */
1242     OP_EODN, /* 23 End of data or \n at end of data: \Z. */
1243     OP_EOD, /* 24 End of data: \z */
1244 nigel 77
1245 ph10 341 OP_OPT, /* 25 Set runtime options */
1246     OP_CIRC, /* 26 Start of line - varies with multiline switch */
1247     OP_DOLL, /* 27 End of line - varies with multiline switch */
1248     OP_CHAR, /* 28 Match one character, casefully */
1249     OP_CHARNC, /* 29 Match one character, caselessly */
1250     OP_NOT, /* 30 Match one character, not the following one */
1251 nigel 77
1252 ph10 341 OP_STAR, /* 31 The maximizing and minimizing versions of */
1253     OP_MINSTAR, /* 32 these six opcodes must come in pairs, with */
1254     OP_PLUS, /* 33 the minimizing one second. */
1255     OP_MINPLUS, /* 34 This first set applies to single characters.*/
1256     OP_QUERY, /* 35 */
1257     OP_MINQUERY, /* 36 */
1258 nigel 77
1259 ph10 341 OP_UPTO, /* 37 From 0 to n matches */
1260     OP_MINUPTO, /* 38 */
1261     OP_EXACT, /* 39 Exactly n matches */
1262 nigel 77
1263 ph10 341 OP_POSSTAR, /* 40 Possessified star */
1264     OP_POSPLUS, /* 41 Possessified plus */
1265     OP_POSQUERY, /* 42 Posesssified query */
1266     OP_POSUPTO, /* 43 Possessified upto */
1267 nigel 77
1268 ph10 341 OP_NOTSTAR, /* 44 The maximizing and minimizing versions of */
1269     OP_NOTMINSTAR, /* 45 these six opcodes must come in pairs, with */
1270     OP_NOTPLUS, /* 46 the minimizing one second. They must be in */
1271     OP_NOTMINPLUS, /* 47 exactly the same order as those above. */
1272     OP_NOTQUERY, /* 48 This set applies to "not" single characters. */
1273     OP_NOTMINQUERY, /* 49 */
1274 nigel 77
1275 ph10 341 OP_NOTUPTO, /* 50 From 0 to n matches */
1276     OP_NOTMINUPTO, /* 51 */
1277     OP_NOTEXACT, /* 52 Exactly n matches */
1278 nigel 93
1279 ph10 341 OP_NOTPOSSTAR, /* 53 Possessified versions */
1280     OP_NOTPOSPLUS, /* 54 */
1281     OP_NOTPOSQUERY, /* 55 */
1282     OP_NOTPOSUPTO, /* 56 */
1283 nigel 93
1284 ph10 341 OP_TYPESTAR, /* 57 The maximizing and minimizing versions of */
1285     OP_TYPEMINSTAR, /* 58 these six opcodes must come in pairs, with */
1286     OP_TYPEPLUS, /* 59 the minimizing one second. These codes must */
1287     OP_TYPEMINPLUS, /* 60 be in exactly the same order as those above. */
1288     OP_TYPEQUERY, /* 61 This set applies to character types such as \d */
1289     OP_TYPEMINQUERY, /* 62 */
1290 nigel 93
1291 ph10 341 OP_TYPEUPTO, /* 63 From 0 to n matches */
1292     OP_TYPEMINUPTO, /* 64 */
1293     OP_TYPEEXACT, /* 65 Exactly n matches */
1294 nigel 93
1295 ph10 341 OP_TYPEPOSSTAR, /* 66 Possessified versions */
1296     OP_TYPEPOSPLUS, /* 67 */
1297     OP_TYPEPOSQUERY, /* 68 */
1298     OP_TYPEPOSUPTO, /* 69 */
1299 nigel 93
1300 ph10 341 OP_CRSTAR, /* 70 The maximizing and minimizing versions of */
1301     OP_CRMINSTAR, /* 71 all these opcodes must come in pairs, with */
1302     OP_CRPLUS, /* 72 the minimizing one second. These codes must */
1303     OP_CRMINPLUS, /* 73 be in exactly the same order as those above. */
1304     OP_CRQUERY, /* 74 These are for character classes and back refs */
1305     OP_CRMINQUERY, /* 75 */
1306     OP_CRRANGE, /* 76 These are different to the three sets above. */
1307     OP_CRMINRANGE, /* 77 */
1308 nigel 93
1309 ph10 341 OP_CLASS, /* 78 Match a character class, chars < 256 only */
1310     OP_NCLASS, /* 79 Same, but the bitmap was created from a negative
1311 nigel 77 class - the difference is relevant only when a UTF-8
1312     character > 255 is encountered. */
1313    
1314 ph10 341 OP_XCLASS, /* 80 Extended class for handling UTF-8 chars within the
1315 nigel 77 class. This does both positive and negative. */
1316    
1317 ph10 341 OP_REF, /* 81 Match a back reference */
1318     OP_RECURSE, /* 82 Match a numbered subpattern (possibly recursive) */
1319     OP_CALLOUT, /* 83 Call out to external function if provided */
1320 nigel 77
1321 ph10 341 OP_ALT, /* 84 Start of alternation */
1322     OP_KET, /* 85 End of group that doesn't have an unbounded repeat */
1323     OP_KETRMAX, /* 86 These two must remain together and in this */
1324     OP_KETRMIN, /* 87 order. They are for groups the repeat for ever. */
1325 nigel 77
1326 nigel 93 /* The assertions must come before BRA, CBRA, ONCE, and COND.*/
1327 nigel 77
1328 ph10 341 OP_ASSERT, /* 88 Positive lookahead */
1329     OP_ASSERT_NOT, /* 89 Negative lookahead */
1330     OP_ASSERTBACK, /* 90 Positive lookbehind */
1331     OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */
1332     OP_REVERSE, /* 92 Move pointer back - used in lookbehind assertions */
1333 nigel 77
1334 nigel 93 /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,
1335     as there's a test for >= ONCE for a subpattern that isn't an assertion. */
1336 nigel 77
1337 ph10 341 OP_ONCE, /* 93 Atomic group */
1338     OP_BRA, /* 94 Start of non-capturing bracket */
1339     OP_CBRA, /* 95 Start of capturing bracket */
1340     OP_COND, /* 96 Conditional group */
1341 nigel 77
1342 nigel 93 /* These three must follow the previous three, in the same order. There's a
1343     check for >= SBRA to distinguish the two sets. */
1344 nigel 77
1345 ph10 341 OP_SBRA, /* 97 Start of non-capturing bracket, check empty */
1346     OP_SCBRA, /* 98 Start of capturing bracket, check empty */
1347     OP_SCOND, /* 99 Conditional group, check empty */
1348 nigel 77
1349 ph10 341 OP_CREF, /* 100 Used to hold a capture number as condition */
1350     OP_RREF, /* 101 Used to hold a recursion number as condition */
1351     OP_DEF, /* 102 The DEFINE condition */
1352 nigel 93
1353 ph10 341 OP_BRAZERO, /* 103 These two must remain together and in this */
1354     OP_BRAMINZERO, /* 104 order. */
1355 ph10 210
1356     /* These are backtracking control verbs */
1357    
1358 ph10 341 OP_PRUNE, /* 105 */
1359     OP_SKIP, /* 106 */
1360     OP_THEN, /* 107 */
1361     OP_COMMIT, /* 108 */
1362 ph10 210
1363     /* These are forced failure and success verbs */
1364    
1365 ph10 341 OP_FAIL, /* 109 */
1366     OP_ACCEPT, /* 110 */
1367 ph10 447 OP_CLOSE, /* 111 Used before OP_ACCEPT to close open captures */
1368 ph10 345
1369 ph10 335 /* This is used to skip a subpattern with a {0} quantifier */
1370 ph10 345
1371 ph10 447 OP_SKIPZERO /* 112 */
1372 nigel 77 };
1373    
1374    
1375     /* This macro defines textual names for all the opcodes. These are used only
1376     for debugging. The macro is referenced only in pcre_printint.c. */
1377    
1378     #define OP_NAME_LIST \
1379 ph10 168 "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \
1380 ph10 341 "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \
1381 ph10 178 "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \
1382     "extuni", "\\Z", "\\z", \
1383 nigel 77 "Opt", "^", "$", "char", "charnc", "not", \
1384     "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1385 nigel 93 "*+","++", "?+", "{", \
1386 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1387 nigel 93 "*+","++", "?+", "{", \
1388 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1389 nigel 93 "*+","++", "?+", "{", \
1390 nigel 77 "*", "*?", "+", "+?", "?", "??", "{", "{", \
1391     "class", "nclass", "xclass", "Ref", "Recurse", "Callout", \
1392     "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", \
1393 nigel 93 "AssertB", "AssertB not", "Reverse", \
1394 ph10 192 "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond", \
1395 ph10 210 "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero", \
1396 ph10 335 "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT", \
1397 ph10 447 "Close", "Skip zero"
1398 nigel 77
1399    
1400     /* This macro defines the length of fixed length operations in the compiled
1401     regex. The lengths are used when searching for specific things, and also in the
1402     debugging printing of a compiled regex. We use a macro so that it can be
1403     defined close to the definitions of the opcodes themselves.
1404    
1405     As things have been extended, some of these are no longer fixed lenths, but are
1406     minima instead. For example, the length of a single-character repeat may vary
1407     in UTF-8 mode. The code that uses this table must know about such things. */
1408    
1409     #define OP_LENGTHS \
1410     1, /* End */ \
1411 ph10 168 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \
1412     1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \
1413 ph10 341 1, 1, 1, /* Any, AllAny, Anybyte */ \
1414 ph10 178 3, 3, 1, /* NOTPROP, PROP, EXTUNI */ \
1415     1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \
1416 nigel 77 1, 1, 2, 1, 1, /* \Z, \z, Opt, ^, $ */ \
1417     2, /* Char - the minimum length */ \
1418     2, /* Charnc - the minimum length */ \
1419     2, /* not */ \
1420     /* Positive single-char repeats ** These are */ \
1421     2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
1422     4, 4, 4, /* upto, minupto, exact ** UTF-8 mode */ \
1423 nigel 93 2, 2, 2, 4, /* *+, ++, ?+, upto+ */ \
1424 nigel 77 /* Negative single-char repeats - only for chars < 256 */ \
1425     2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
1426     4, 4, 4, /* NOT upto, minupto, exact */ \
1427 nigel 93 2, 2, 2, 4, /* Possessive *, +, ?, upto */ \
1428 nigel 77 /* Positive type repeats */ \
1429     2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
1430     4, 4, 4, /* Type upto, minupto, exact */ \
1431 nigel 93 2, 2, 2, 4, /* Possessive *+, ++, ?+, upto+ */ \
1432 nigel 77 /* Character class & ref repeats */ \
1433     1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
1434     5, 5, /* CRRANGE, CRMINRANGE */ \
1435     33, /* CLASS */ \
1436     33, /* NCLASS */ \
1437     0, /* XCLASS - variable length */ \
1438     3, /* REF */ \
1439     1+LINK_SIZE, /* RECURSE */ \
1440     2+2*LINK_SIZE, /* CALLOUT */ \
1441     1+LINK_SIZE, /* Alt */ \
1442     1+LINK_SIZE, /* Ket */ \
1443     1+LINK_SIZE, /* KetRmax */ \
1444     1+LINK_SIZE, /* KetRmin */ \
1445     1+LINK_SIZE, /* Assert */ \
1446     1+LINK_SIZE, /* Assert not */ \
1447     1+LINK_SIZE, /* Assert behind */ \
1448     1+LINK_SIZE, /* Assert behind not */ \
1449     1+LINK_SIZE, /* Reverse */ \
1450 nigel 93 1+LINK_SIZE, /* ONCE */ \
1451     1+LINK_SIZE, /* BRA */ \
1452     3+LINK_SIZE, /* CBRA */ \
1453 nigel 77 1+LINK_SIZE, /* COND */ \
1454 nigel 93 1+LINK_SIZE, /* SBRA */ \
1455     3+LINK_SIZE, /* SCBRA */ \
1456     1+LINK_SIZE, /* SCOND */ \
1457 nigel 77 3, /* CREF */ \
1458 nigel 93 3, /* RREF */ \
1459     1, /* DEF */ \
1460 nigel 77 1, 1, /* BRAZERO, BRAMINZERO */ \
1461 ph10 210 1, 1, 1, 1, /* PRUNE, SKIP, THEN, COMMIT, */ \
1462 ph10 447 1, 1, 3, 1 /* FAIL, ACCEPT, CLOSE, SKIPZERO */
1463 nigel 77
1464    
1465 nigel 93 /* A magic value for OP_RREF to indicate the "any recursion" condition. */
1466 nigel 77
1467 nigel 93 #define RREF_ANY 0xffff
1468 nigel 77
1469     /* Error code numbers. They are given names so that they can more easily be
1470     tracked. */
1471    
1472     enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
1473     ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
1474     ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
1475     ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
1476 nigel 91 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
1477 ph10 210 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
1478 ph10 336 ERR60, ERR61, ERR62, ERR63, ERR64 };
1479 nigel 77
1480     /* The real format of the start of the pcre block; the index of names and the
1481     code vector run on as long as necessary after the end. We store an explicit
1482     offset to the name table so that if a regex is compiled on one host, saved, and
1483     then run on another where the size of pointers is different, all might still
1484     be well. For the case of compiled-on-4 and run-on-8, we include an extra
1485     pointer that is always NULL. For future-proofing, a few dummy fields were
1486     originally included - even though you can never get this planning right - but
1487     there is only one left now.
1488    
1489     NOTE NOTE NOTE:
1490     Because people can now save and re-use compiled patterns, any additions to this
1491     structure should be made at the end, and something earlier (e.g. a new
1492     flag in the options or one of the dummy fields) should indicate that the new
1493     fields are present. Currently PCRE always sets the dummy fields to zero.
1494     NOTE NOTE NOTE:
1495     */
1496    
1497     typedef struct real_pcre {
1498     pcre_uint32 magic_number;
1499     pcre_uint32 size; /* Total that was malloced */
1500 ph10 230 pcre_uint32 options; /* Public options */
1501     pcre_uint16 flags; /* Private flags */
1502     pcre_uint16 dummy1; /* For future use */
1503 nigel 77 pcre_uint16 top_bracket;
1504     pcre_uint16 top_backref;
1505     pcre_uint16 first_byte;
1506     pcre_uint16 req_byte;
1507     pcre_uint16 name_table_offset; /* Offset to name table that follows */
1508     pcre_uint16 name_entry_size; /* Size of any name items */
1509     pcre_uint16 name_count; /* Number of name items */
1510     pcre_uint16 ref_count; /* Reference count */
1511    
1512     const unsigned char *tables; /* Pointer to tables or NULL for std */
1513     const unsigned char *nullpad; /* NULL padding */
1514     } real_pcre;
1515    
1516     /* The format of the block used to store data from pcre_study(). The same
1517     remark (see NOTE above) about extending this structure applies. */
1518    
1519     typedef struct pcre_study_data {
1520     pcre_uint32 size; /* Total that was malloced */
1521     pcre_uint32 options;
1522     uschar start_bits[32];
1523     } pcre_study_data;
1524    
1525 ph10 447 /* Structure for building a chain of open capturing subpatterns during
1526     compiling, so that instructions to close them can be compiled when (*ACCEPT) is
1527     encountered. */
1528    
1529     typedef struct open_capitem {
1530     struct open_capitem *next; /* Chain link */
1531     pcre_uint16 number; /* Capture number */
1532     } open_capitem;
1533    
1534 nigel 77 /* Structure for passing "static" information around between the functions
1535     doing the compiling, so that they are thread-safe. */
1536    
1537     typedef struct compile_data {
1538     const uschar *lcc; /* Points to lower casing table */
1539     const uschar *fcc; /* Points to case-flipping table */
1540     const uschar *cbits; /* Points to character type table */
1541     const uschar *ctypes; /* Points to table of type maps */
1542 nigel 93 const uschar *start_workspace;/* The start of working space */
1543 nigel 77 const uschar *start_code; /* The start of the compiled code */
1544     const uschar *start_pattern; /* The start of the pattern */
1545 nigel 93 const uschar *end_pattern; /* The end of the pattern */
1546 ph10 447 open_capitem *open_caps; /* Chain of open capture items */
1547 nigel 93 uschar *hwm; /* High watermark of workspace */
1548 nigel 77 uschar *name_table; /* The name/number table */
1549     int names_found; /* Number of entries so far */
1550     int name_entry_size; /* Size of each entry */
1551 ph10 270 int bracount; /* Count of capturing parens as we compile */
1552 ph10 286 int final_bracount; /* Saved value after first pass */
1553 nigel 77 int top_backref; /* Maximum back reference */
1554     unsigned int backref_map; /* Bitmap of low back refs */
1555 nigel 93 int external_options; /* External (initial) options */
1556 ph10 231 int external_flags; /* External flag bits to be set */
1557 nigel 77 int req_varyopt; /* "After variable item" flag for reqbyte */
1558 ph10 211 BOOL had_accept; /* (*ACCEPT) encountered */
1559 ph10 454 BOOL check_lookbehind; /* Lookbehinds need later checking */
1560 nigel 93 int nltype; /* Newline type */
1561     int nllen; /* Newline string length */
1562     uschar nl[4]; /* Newline string when fixed length */
1563 nigel 77 } compile_data;
1564    
1565     /* Structure for maintaining a chain of pointers to the currently incomplete
1566     branches, for testing for left recursion. */
1567    
1568     typedef struct branch_chain {
1569     struct branch_chain *outer;
1570     uschar *current;
1571     } branch_chain;
1572    
1573     /* Structure for items in a linked list that represents an explicit recursive
1574     call within the pattern. */
1575    
1576     typedef struct recursion_info {
1577     struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
1578     int group_num; /* Number of group that was called */
1579     const uschar *after_call; /* "Return value": points after the call in the expr */
1580 ph10 168 USPTR save_start; /* Old value of mstart */
1581 nigel 77 int *offset_save; /* Pointer to start of saved offsets */
1582     int saved_max; /* Number of saved offsets */
1583 ph10 446 int offset_top; /* Current value of offset_top */
1584 nigel 77 } recursion_info;
1585    
1586 nigel 93 /* Structure for building a chain of data for holding the values of the subject
1587     pointer at the start of each subpattern, so as to detect when an empty string
1588     has been matched by a subpattern - to break infinite loops. */
1589    
1590     typedef struct eptrblock {
1591     struct eptrblock *epb_prev;
1592     USPTR epb_saved_eptr;
1593     } eptrblock;
1594    
1595    
1596 nigel 77 /* Structure for passing "static" information around between the functions
1597     doing traditional NFA matching, so that they are thread-safe. */
1598    
1599     typedef struct match_data {
1600 nigel 87 unsigned long int match_call_count; /* As it says */
1601     unsigned long int match_limit; /* As it says */
1602     unsigned long int match_limit_recursion; /* As it says */
1603 nigel 77 int *offset_vector; /* Offset vector */
1604     int offset_end; /* One past the end */
1605     int offset_max; /* The maximum usable for return data */
1606 nigel 93 int nltype; /* Newline type */
1607     int nllen; /* Newline string length */
1608     uschar nl[4]; /* Newline string when fixed */
1609 nigel 77 const uschar *lcc; /* Points to lower casing table */
1610     const uschar *ctypes; /* Points to table of type maps */
1611     BOOL offset_overflow; /* Set if too many extractions */
1612     BOOL notbol; /* NOTBOL flag */
1613     BOOL noteol; /* NOTEOL flag */
1614     BOOL utf8; /* UTF8 flag */
1615 ph10 345 BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */
1616 nigel 77 BOOL endonly; /* Dollar not before final \n */
1617     BOOL notempty; /* Empty string match not wanted */
1618 ph10 442 BOOL notempty_atstart; /* Empty string match at start not wanted */
1619 nigel 77 BOOL hitend; /* Hit the end of the subject at some point */
1620 ph10 231 BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
1621 nigel 77 const uschar *start_code; /* For use when recursing */
1622 nigel 87 USPTR start_subject; /* Start of the subject string */
1623     USPTR end_subject; /* End of the subject string */
1624 ph10 168 USPTR start_match_ptr; /* Start of matched string */
1625 nigel 87 USPTR end_match_ptr; /* Subject position at end match */
1626 ph10 442 USPTR start_used_ptr; /* Earliest consulted character */
1627 ph10 435 int partial; /* PARTIAL options */
1628 nigel 77 int end_offset_top; /* Highwater mark at end of match */
1629     int capture_last; /* Most recent capture number */
1630     int start_offset; /* The start offset value */
1631 nigel 93 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
1632     int eptrn; /* Next free eptrblock */
1633 nigel 77 recursion_info *recursive; /* Linked list of recursion data */
1634     void *callout_data; /* To pass back to callouts */
1635     } match_data;
1636    
1637     /* A similar structure is used for the same purpose by the DFA matching
1638     functions. */
1639    
1640     typedef struct dfa_match_data {
1641     const uschar *start_code; /* Start of the compiled pattern */
1642     const uschar *start_subject; /* Start of the subject string */
1643     const uschar *end_subject; /* End of subject string */
1644 ph10 442 const uschar *start_used_ptr; /* Earliest consulted character */
1645 nigel 77 const uschar *tables; /* Character tables */
1646 ph10 442 int start_offset; /* The start offset value */
1647 nigel 77 int moptions; /* Match options */
1648     int poptions; /* Pattern options */
1649 nigel 93 int nltype; /* Newline type */
1650     int nllen; /* Newline string length */
1651     uschar nl[4]; /* Newline string when fixed */
1652 nigel 77 void *callout_data; /* To pass back to callouts */
1653     } dfa_match_data;
1654    
1655     /* Bit definitions for entries in the pcre_ctypes table. */
1656    
1657     #define ctype_space 0x01
1658     #define ctype_letter 0x02
1659     #define ctype_digit 0x04
1660     #define ctype_xdigit 0x08
1661 ph10 274 #define ctype_word 0x10 /* alphanumeric or '_' */
1662 nigel 77 #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
1663    
1664     /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
1665     of bits for a class map. Some classes are built by combining these tables. */
1666    
1667     #define cbit_space 0 /* [:space:] or \s */
1668     #define cbit_xdigit 32 /* [:xdigit:] */
1669     #define cbit_digit 64 /* [:digit:] or \d */
1670     #define cbit_upper 96 /* [:upper:] */
1671     #define cbit_lower 128 /* [:lower:] */
1672     #define cbit_word 160 /* [:word:] or \w */
1673     #define cbit_graph 192 /* [:graph:] */
1674     #define cbit_print 224 /* [:print:] */
1675     #define cbit_punct 256 /* [:punct:] */
1676     #define cbit_cntrl 288 /* [:cntrl:] */
1677     #define cbit_length 320 /* Length of the cbits table */
1678    
1679     /* Offsets of the various tables from the base tables pointer, and
1680     total length. */
1681    
1682     #define lcc_offset 0
1683     #define fcc_offset 256
1684     #define cbits_offset 512
1685     #define ctypes_offset (cbits_offset + cbit_length)
1686     #define tables_length (ctypes_offset + 256)
1687    
1688 nigel 87 /* Layout of the UCP type table that translates property names into types and
1689 ph10 240 codes. Each entry used to point directly to a name, but to reduce the number of
1690 ph10 243 relocations in shared libraries, it now has an offset into a single string
1691 ph10 240 instead. */
1692 nigel 77
1693     typedef struct {
1694 ph10 243 pcre_uint16 name_offset;
1695 nigel 87 pcre_uint16 type;
1696     pcre_uint16 value;
1697 nigel 77 } ucp_type_table;
1698    
1699    
1700     /* Internal shared data tables. These are tables that are used by more than one
1701     of the exported public functions. They have to be "external" in the C sense,
1702     but are not part of the PCRE public API. The data for these tables is in the
1703     pcre_tables.c module. */
1704    
1705     extern const int _pcre_utf8_table1[];
1706     extern const int _pcre_utf8_table2[];
1707     extern const int _pcre_utf8_table3[];
1708     extern const uschar _pcre_utf8_table4[];
1709    
1710     extern const int _pcre_utf8_table1_size;
1711    
1712 ph10 240 extern const char _pcre_utt_names[];
1713 nigel 77 extern const ucp_type_table _pcre_utt[];
1714     extern const int _pcre_utt_size;
1715    
1716     extern const uschar _pcre_default_tables[];
1717    
1718     extern const uschar _pcre_OP_lengths[];
1719    
1720    
1721     /* Internal shared functions. These are functions that are used by more than
1722     one of the exported public functions. They have to be "external" in the C
1723     sense, but are not part of the PCRE public API. */
1724    
1725 ph10 150 extern BOOL _pcre_is_newline(const uschar *, int, const uschar *,
1726 ph10 149 int *, BOOL);
1727 nigel 93 extern int _pcre_ord2utf8(int, uschar *);
1728     extern real_pcre *_pcre_try_flipped(const real_pcre *, real_pcre *,
1729     const pcre_study_data *, pcre_study_data *);
1730     extern int _pcre_valid_utf8(const uschar *, int);
1731 ph10 150 extern BOOL _pcre_was_newline(const uschar *, int, const uschar *,
1732 ph10 149 int *, BOOL);
1733 nigel 93 extern BOOL _pcre_xclass(int, const uschar *);
1734 nigel 77
1735 ph10 349
1736     /* Unicode character database (UCD) */
1737    
1738     typedef struct {
1739     uschar script;
1740     uschar chartype;
1741 ph10 351 pcre_int32 other_case;
1742 ph10 349 } ucd_record;
1743    
1744 ph10 351 extern const ucd_record _pcre_ucd_records[];
1745     extern const uschar _pcre_ucd_stage1[];
1746     extern const pcre_uint16 _pcre_ucd_stage2[];
1747     extern const int _pcre_ucp_gentype[];
1748 ph10 349
1749 ph10 351
1750 ph10 349 /* UCD access macros */
1751    
1752     #define UCD_BLOCK_SIZE 128
1753 ph10 351 #define GET_UCD(ch) (_pcre_ucd_records + \
1754     _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \
1755 ph10 349 UCD_BLOCK_SIZE + ch % UCD_BLOCK_SIZE])
1756    
1757     #define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype
1758     #define UCD_SCRIPT(ch) GET_UCD(ch)->script
1759 ph10 351 #define UCD_CATEGORY(ch) _pcre_ucp_gentype[UCD_CHARTYPE(ch)]
1760     #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)
1761 ph10 349
1762 nigel 85 #endif
1763    
1764 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