Annotation of embedaddon/pcre/pcre_internal.h, revision 1.1.1.2
1.1 misho 1: /*************************************************
2: * Perl-Compatible Regular Expressions *
3: *************************************************/
4:
5:
6: /* PCRE is a library of functions to support regular expressions whose syntax
7: and semantics are as close as possible to those of the Perl 5 language.
8:
9: Written by Philip Hazel
1.1.1.2 ! misho 10: Copyright (c) 1997-2012 University of Cambridge
1.1 misho 11:
12: -----------------------------------------------------------------------------
13: Redistribution and use in source and binary forms, with or without
14: modification, are permitted provided that the following conditions are met:
15:
16: * Redistributions of source code must retain the above copyright notice,
17: this list of conditions and the following disclaimer.
18:
19: * Redistributions in binary form must reproduce the above copyright
20: notice, this list of conditions and the following disclaimer in the
21: documentation and/or other materials provided with the distribution.
22:
23: * Neither the name of the University of Cambridge nor the names of its
24: contributors may be used to endorse or promote products derived from
25: this software without specific prior written permission.
26:
27: THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
28: AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
29: IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
30: ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
31: LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32: CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33: SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
34: INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
35: CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
36: ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
37: POSSIBILITY OF SUCH DAMAGE.
38: -----------------------------------------------------------------------------
39: */
40:
41: /* This header contains definitions that are shared between the different
42: modules, but which are not relevant to the exported API. This includes some
1.1.1.2 ! misho 43: functions whose names all begin with "_pcre_" or "_pcre16_" depending on
! 44: the PRIV macro. */
1.1 misho 45:
46: #ifndef PCRE_INTERNAL_H
47: #define PCRE_INTERNAL_H
48:
49: /* Define PCRE_DEBUG to get debugging output on stdout. */
50:
51: #if 0
52: #define PCRE_DEBUG
53: #endif
54:
1.1.1.2 ! misho 55: /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
! 56: #ifndef COMPILE_PCRE16
! 57: #define COMPILE_PCRE8
1.1 misho 58: #endif
59:
1.1.1.2 ! misho 60: /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
1.1 misho 61: "configure" script ensures this, but not everybody uses "configure". */
62:
1.1.1.2 ! misho 63: #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
! 64: #define SUPPORT_UTF 1
! 65: #endif
! 66:
! 67: /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
! 68: reasons with existing code. */
! 69:
! 70: #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
! 71: #define SUPPORT_UTF 1
! 72: #endif
! 73:
! 74: /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
! 75: Until then we define it if SUPPORT_UTF is defined. */
! 76:
! 77: #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
1.1 misho 78: #define SUPPORT_UTF8 1
79: #endif
80:
1.1.1.2 ! misho 81: /* We do not support both EBCDIC and UTF-8/16 at the same time. The "configure"
! 82: script prevents both being selected, but not everybody uses "configure". */
! 83:
! 84: #if defined EBCDIC && defined SUPPORT_UTF
! 85: #error The use of both EBCDIC and SUPPORT_UTF8/16 is not supported.
! 86: #endif
! 87:
1.1 misho 88: /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
89: inline, and there are *still* stupid compilers about that don't like indented
90: pre-processor statements, or at least there were when I first wrote this. After
91: all, it had only been about 10 years then...
92:
93: It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so
94: be absolutely sure we get our version. */
95:
96: #undef DPRINTF
97: #ifdef PCRE_DEBUG
98: #define DPRINTF(p) printf p
99: #else
100: #define DPRINTF(p) /* Nothing */
101: #endif
102:
103:
104: /* Standard C headers plus the external interface definition. The only time
105: setjmp and stdarg are used is when NO_RECURSE is set. */
106:
107: #include <ctype.h>
108: #include <limits.h>
109: #include <stddef.h>
110: #include <stdio.h>
111: #include <stdlib.h>
112: #include <string.h>
113:
114: /* When compiling a DLL for Windows, the exported symbols have to be declared
115: using some MS magic. I found some useful information on this web page:
116: http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
117: information there, using __declspec(dllexport) without "extern" we have a
118: definition; with "extern" we have a declaration. The settings here override the
119: setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL,
120: which is all that is needed for applications (they just import the symbols). We
121: use:
122:
123: PCRE_EXP_DECL for declarations
124: PCRE_EXP_DEFN for definitions of exported functions
125: PCRE_EXP_DATA_DEFN for definitions of exported variables
126:
127: The reason for the two DEFN macros is that in non-Windows environments, one
128: does not want to have "extern" before variable definitions because it leads to
129: compiler warnings. So we distinguish between functions and variables. In
130: Windows, the two should always be the same.
131:
132: The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest,
133: which is an application, but needs to import this file in order to "peek" at
134: internals, can #include pcre.h first to get an application's-eye view.
135:
136: In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon,
137: special-purpose environments) might want to stick other stuff in front of
138: exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and
139: PCRE_EXP_DATA_DEFN only if they are not already set. */
140:
141: #ifndef PCRE_EXP_DECL
142: # ifdef _WIN32
143: # ifndef PCRE_STATIC
144: # define PCRE_EXP_DECL extern __declspec(dllexport)
145: # define PCRE_EXP_DEFN __declspec(dllexport)
146: # define PCRE_EXP_DATA_DEFN __declspec(dllexport)
147: # else
148: # define PCRE_EXP_DECL extern
149: # define PCRE_EXP_DEFN
150: # define PCRE_EXP_DATA_DEFN
151: # endif
152: # else
153: # ifdef __cplusplus
154: # define PCRE_EXP_DECL extern "C"
155: # else
156: # define PCRE_EXP_DECL extern
157: # endif
158: # ifndef PCRE_EXP_DEFN
159: # define PCRE_EXP_DEFN PCRE_EXP_DECL
160: # endif
161: # ifndef PCRE_EXP_DATA_DEFN
162: # define PCRE_EXP_DATA_DEFN
163: # endif
164: # endif
165: #endif
166:
167: /* When compiling with the MSVC compiler, it is sometimes necessary to include
168: a "calling convention" before exported function names. (This is secondhand
169: information; I know nothing about MSVC myself). For example, something like
170:
171: void __cdecl function(....)
172:
173: might be needed. In order so make this easy, all the exported functions have
174: PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
175: set, we ensure here that it has no effect. */
176:
177: #ifndef PCRE_CALL_CONVENTION
178: #define PCRE_CALL_CONVENTION
179: #endif
180:
1.1.1.2 ! misho 181: /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
1.1 misho 182: cannot determine these outside the compilation (e.g. by running a program as
183: part of "configure") because PCRE is often cross-compiled for use on other
184: systems. Instead we make use of the maximum sizes that are available at
185: preprocessor time in standard C environments. */
186:
1.1.1.2 ! misho 187: typedef unsigned char pcre_uint8;
! 188:
1.1 misho 189: #if USHRT_MAX == 65535
190: typedef unsigned short pcre_uint16;
191: typedef short pcre_int16;
192: #elif UINT_MAX == 65535
193: typedef unsigned int pcre_uint16;
194: typedef int pcre_int16;
195: #else
196: #error Cannot determine a type for 16-bit unsigned integers
197: #endif
198:
199: #if UINT_MAX == 4294967295
200: typedef unsigned int pcre_uint32;
201: typedef int pcre_int32;
202: #elif ULONG_MAX == 4294967295
203: typedef unsigned long int pcre_uint32;
204: typedef long int pcre_int32;
205: #else
206: #error Cannot determine a type for 32-bit unsigned integers
207: #endif
208:
209: /* When checking for integer overflow in pcre_compile(), we need to handle
210: large integers. If a 64-bit integer type is available, we can use that.
211: Otherwise we have to cast to double, which of course requires floating point
212: arithmetic. Handle this by defining a macro for the appropriate type. If
213: stdint.h is available, include it; it may define INT64_MAX. Systems that do not
214: have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
215: by "configure". */
216:
217: #if HAVE_STDINT_H
218: #include <stdint.h>
219: #elif HAVE_INTTYPES_H
220: #include <inttypes.h>
221: #endif
222:
223: #if defined INT64_MAX || defined int64_t
224: #define INT64_OR_DOUBLE int64_t
225: #else
226: #define INT64_OR_DOUBLE double
227: #endif
228:
229: /* All character handling must be done as unsigned characters. Otherwise there
230: are problems with top-bit-set characters and functions such as isspace().
1.1.1.2 ! misho 231: However, we leave the interface to the outside world as char * or short *,
! 232: because that should make things easier for callers. This character type is
! 233: called pcre_uchar.
! 234:
! 235: The IN_UCHARS macro multiply its argument with the byte size of the current
! 236: pcre_uchar type. Useful for memcpy and such operations, whose require the
! 237: byte size of their input/output buffers.
! 238:
! 239: The MAX_255 macro checks whether its pcre_uchar input is less than 256.
! 240:
! 241: The TABLE_GET macro is designed for accessing elements of tables whose contain
! 242: exactly 256 items. When the character is able to contain more than 256
! 243: items, some check is needed before accessing these tables.
! 244: */
! 245:
! 246: #ifdef COMPILE_PCRE8
1.1 misho 247:
1.1.1.2 ! misho 248: typedef unsigned char pcre_uchar;
! 249: #define IN_UCHARS(x) (x)
! 250: #define MAX_255(c) 1
! 251: #define TABLE_GET(c, table, default) ((table)[c])
! 252:
! 253: #else
! 254:
! 255: #ifdef COMPILE_PCRE16
! 256: #if USHRT_MAX != 65535
! 257: /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
! 258: pcre.h(.in) and disable (comment out) this message. */
! 259: #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
! 260: #endif
! 261:
! 262: typedef pcre_uint16 pcre_uchar;
! 263: #define IN_UCHARS(x) ((x) << 1)
! 264: #define MAX_255(c) ((c) <= 255u)
! 265: #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
! 266:
! 267: #else
! 268: #error Unsupported compiling mode
! 269: #endif /* COMPILE_PCRE16 */
! 270:
! 271: #endif /* COMPILE_PCRE8 */
1.1 misho 272:
273: /* This is an unsigned int value that no character can ever have. UTF-8
274: characters only go up to 0x7fffffff (though Unicode doesn't go beyond
275: 0x0010ffff). */
276:
277: #define NOTACHAR 0xffffffff
278:
279: /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
280: "any" and "anycrlf" at present). The following macros are used to package up
281: testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
282: modules to indicate in which datablock the parameters exist, and what the
283: start/end of string field names are. */
284:
285: #define NLTYPE_FIXED 0 /* Newline is a fixed length string */
286: #define NLTYPE_ANY 1 /* Newline is any Unicode line ending */
287: #define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */
288:
289: /* This macro checks for a newline at the given position */
290:
291: #define IS_NEWLINE(p) \
292: ((NLBLOCK->nltype != NLTYPE_FIXED)? \
293: ((p) < NLBLOCK->PSEND && \
1.1.1.2 ! misho 294: PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
! 295: &(NLBLOCK->nllen), utf)) \
1.1 misho 296: : \
297: ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
298: (p)[0] == NLBLOCK->nl[0] && \
299: (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \
300: ) \
301: )
302:
303: /* This macro checks for a newline immediately preceding the given position */
304:
305: #define WAS_NEWLINE(p) \
306: ((NLBLOCK->nltype != NLTYPE_FIXED)? \
307: ((p) > NLBLOCK->PSSTART && \
1.1.1.2 ! misho 308: PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
! 309: &(NLBLOCK->nllen), utf)) \
1.1 misho 310: : \
311: ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
312: (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \
313: (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \
314: ) \
315: )
316:
317: /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
318: with a custom type. This makes it possible, for example, to allow pcre_exec()
319: to process subject strings that are discontinuous by using a smart pointer
320: class. It must always be possible to inspect all of the subject string in
321: pcre_exec() because of the way it backtracks. Two macros are required in the
322: normal case, for sign-unspecified and unsigned char pointers. The former is
323: used for the external interface and appears in pcre.h, which is why its name
324: must begin with PCRE_. */
325:
326: #ifdef CUSTOM_SUBJECT_PTR
1.1.1.2 ! misho 327: #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
1.1 misho 328: #else
1.1.1.2 ! misho 329: #define PCRE_PUCHAR const pcre_uchar *
1.1 misho 330: #endif
331:
332: /* Include the public PCRE header and the definitions of UCP character property
333: values. */
334:
335: #include "pcre.h"
336: #include "ucp.h"
337:
338: /* When compiling for use with the Virtual Pascal compiler, these functions
339: need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
340: option on the command line. */
341:
342: #ifdef VPCOMPAT
343: #define strlen(s) _strlen(s)
344: #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
345: #define memcmp(s,c,n) _memcmp(s,c,n)
346: #define memcpy(d,s,n) _memcpy(d,s,n)
347: #define memmove(d,s,n) _memmove(d,s,n)
348: #define memset(s,c,n) _memset(s,c,n)
349: #else /* VPCOMPAT */
350:
351: /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
352: define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
353: is set. Otherwise, include an emulating function for those systems that have
354: neither (there some non-Unix environments where this is the case). */
355:
356: #ifndef HAVE_MEMMOVE
357: #undef memmove /* some systems may have a macro */
358: #ifdef HAVE_BCOPY
359: #define memmove(a, b, c) bcopy(b, a, c)
360: #else /* HAVE_BCOPY */
361: static void *
362: pcre_memmove(void *d, const void *s, size_t n)
363: {
364: size_t i;
365: unsigned char *dest = (unsigned char *)d;
366: const unsigned char *src = (const unsigned char *)s;
367: if (dest > src)
368: {
369: dest += n;
370: src += n;
371: for (i = 0; i < n; ++i) *(--dest) = *(--src);
372: return (void *)dest;
373: }
374: else
375: {
376: for (i = 0; i < n; ++i) *dest++ = *src++;
377: return (void *)(dest - n);
378: }
379: }
380: #define memmove(a, b, c) pcre_memmove(a, b, c)
381: #endif /* not HAVE_BCOPY */
382: #endif /* not HAVE_MEMMOVE */
383: #endif /* not VPCOMPAT */
384:
385:
386: /* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
387: in big-endian order) by default. These are used, for example, to link from the
388: start of a subpattern to its alternatives and its end. The use of 2 bytes per
389: offset limits the size of the compiled regex to around 64K, which is big enough
390: for almost everybody. However, I received a request for an even bigger limit.
391: For this reason, and also to make the code easier to maintain, the storing and
392: loading of offsets from the byte string is now handled by the macros that are
393: defined here.
394:
395: The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
396: the config.h file, but can be overridden by using -D on the command line. This
397: is automated on Unix systems via the "configure" command. */
398:
1.1.1.2 ! misho 399: #ifdef COMPILE_PCRE8
! 400:
1.1 misho 401: #if LINK_SIZE == 2
402:
403: #define PUT(a,n,d) \
404: (a[n] = (d) >> 8), \
405: (a[(n)+1] = (d) & 255)
406:
407: #define GET(a,n) \
408: (((a)[n] << 8) | (a)[(n)+1])
409:
410: #define MAX_PATTERN_SIZE (1 << 16)
411:
412:
413: #elif LINK_SIZE == 3
414:
415: #define PUT(a,n,d) \
416: (a[n] = (d) >> 16), \
417: (a[(n)+1] = (d) >> 8), \
418: (a[(n)+2] = (d) & 255)
419:
420: #define GET(a,n) \
421: (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
422:
423: #define MAX_PATTERN_SIZE (1 << 24)
424:
425:
426: #elif LINK_SIZE == 4
427:
428: #define PUT(a,n,d) \
429: (a[n] = (d) >> 24), \
430: (a[(n)+1] = (d) >> 16), \
431: (a[(n)+2] = (d) >> 8), \
432: (a[(n)+3] = (d) & 255)
433:
434: #define GET(a,n) \
435: (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
436:
1.1.1.2 ! misho 437: /* Keep it positive */
! 438: #define MAX_PATTERN_SIZE (1 << 30)
! 439:
! 440: #else
! 441: #error LINK_SIZE must be either 2, 3, or 4
! 442: #endif
! 443:
! 444: #else /* COMPILE_PCRE8 */
! 445:
! 446: #ifdef COMPILE_PCRE16
! 447:
! 448: #if LINK_SIZE == 2
! 449:
! 450: #undef LINK_SIZE
! 451: #define LINK_SIZE 1
! 452:
! 453: #define PUT(a,n,d) \
! 454: (a[n] = (d))
! 455:
! 456: #define GET(a,n) \
! 457: (a[n])
! 458:
! 459: #define MAX_PATTERN_SIZE (1 << 16)
! 460:
! 461: #elif LINK_SIZE == 3 || LINK_SIZE == 4
! 462:
! 463: #undef LINK_SIZE
! 464: #define LINK_SIZE 2
! 465:
! 466: #define PUT(a,n,d) \
! 467: (a[n] = (d) >> 16), \
! 468: (a[(n)+1] = (d) & 65535)
! 469:
! 470: #define GET(a,n) \
! 471: (((a)[n] << 16) | (a)[(n)+1])
1.1 misho 472:
1.1.1.2 ! misho 473: /* Keep it positive */
! 474: #define MAX_PATTERN_SIZE (1 << 30)
1.1 misho 475:
476: #else
477: #error LINK_SIZE must be either 2, 3, or 4
478: #endif
479:
1.1.1.2 ! misho 480: #else
! 481: #error Unsupported compiling mode
! 482: #endif /* COMPILE_PCRE16 */
! 483:
! 484: #endif /* COMPILE_PCRE8 */
1.1 misho 485:
486: /* Convenience macro defined in terms of the others */
487:
488: #define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
489:
490:
491: /* PCRE uses some other 2-byte quantities that do not change when the size of
492: offsets changes. There are used for repeat counts and for other things such as
493: capturing parenthesis numbers in back references. */
494:
1.1.1.2 ! misho 495: #ifdef COMPILE_PCRE8
! 496:
! 497: #define IMM2_SIZE 2
! 498:
1.1 misho 499: #define PUT2(a,n,d) \
500: a[n] = (d) >> 8; \
501: a[(n)+1] = (d) & 255
502:
503: #define GET2(a,n) \
504: (((a)[n] << 8) | (a)[(n)+1])
505:
1.1.1.2 ! misho 506: #else /* COMPILE_PCRE8 */
1.1 misho 507:
1.1.1.2 ! misho 508: #ifdef COMPILE_PCRE16
1.1 misho 509:
1.1.1.2 ! misho 510: #define IMM2_SIZE 1
1.1 misho 511:
1.1.1.2 ! misho 512: #define PUT2(a,n,d) \
! 513: a[n] = d
! 514:
! 515: #define GET2(a,n) \
! 516: a[n]
! 517:
! 518: #else
! 519: #error Unsupported compiling mode
! 520: #endif /* COMPILE_PCRE16 */
! 521:
! 522: #endif /* COMPILE_PCRE8 */
! 523:
! 524: #define PUT2INC(a,n,d) PUT2(a,n,d), a += IMM2_SIZE
! 525:
! 526: /* When UTF encoding is being used, a character is no longer just a single
! 527: character. The macros for character handling generate simple sequences when
! 528: used in character-mode, and more complicated ones for UTF characters.
! 529: GETCHARLENTEST and other macros are not used when UTF is not supported,
! 530: so they are not defined. To make sure they can never even appear when
! 531: UTF support is omitted, we don't even define them. */
! 532:
! 533: #ifndef SUPPORT_UTF
! 534:
! 535: /* #define MAX_VALUE_FOR_SINGLE_CHAR */
! 536: /* #define HAS_EXTRALEN(c) */
! 537: /* #define GET_EXTRALEN(c) */
! 538: /* #define NOT_FIRSTCHAR(c) */
1.1 misho 539: #define GETCHAR(c, eptr) c = *eptr;
540: #define GETCHARTEST(c, eptr) c = *eptr;
541: #define GETCHARINC(c, eptr) c = *eptr++;
542: #define GETCHARINCTEST(c, eptr) c = *eptr++;
543: #define GETCHARLEN(c, eptr, len) c = *eptr;
544: /* #define GETCHARLENTEST(c, eptr, len) */
545: /* #define BACKCHAR(eptr) */
1.1.1.2 ! misho 546: /* #define FORWARDCHAR(eptr) */
! 547: /* #define ACROSSCHAR(condition, eptr, action) */
1.1 misho 548:
1.1.1.2 ! misho 549: #else /* SUPPORT_UTF */
! 550:
! 551: #ifdef COMPILE_PCRE8
1.1 misho 552:
553: /* These macros were originally written in the form of loops that used data
1.1.1.2 ! misho 554: from the tables whose names start with PRIV(utf8_table). They were rewritten by
1.1 misho 555: a user so as not to use loops, because in some environments this gives a
556: significant performance advantage, and it seems never to do any harm. */
557:
1.1.1.2 ! misho 558: /* Tells the biggest code point which can be encoded as a single character. */
! 559:
! 560: #define MAX_VALUE_FOR_SINGLE_CHAR 127
! 561:
! 562: /* Tests whether the code point needs extra characters to decode. */
! 563:
! 564: #define HAS_EXTRALEN(c) ((c) >= 0xc0)
! 565:
! 566: /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
! 567: Otherwise it has an undefined behaviour. */
! 568:
! 569: #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
! 570:
! 571: /* Returns TRUE, if the given character is not the first character
! 572: of a UTF sequence. */
! 573:
! 574: #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
! 575:
1.1 misho 576: /* Base macro to pick up the remaining bytes of a UTF-8 character, not
577: advancing the pointer. */
578:
579: #define GETUTF8(c, eptr) \
580: { \
581: if ((c & 0x20) == 0) \
582: c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
583: else if ((c & 0x10) == 0) \
584: c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
585: else if ((c & 0x08) == 0) \
586: c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
587: ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
588: else if ((c & 0x04) == 0) \
589: c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
590: ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
591: (eptr[4] & 0x3f); \
592: else \
593: c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
594: ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
595: ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
596: }
597:
598: /* Get the next UTF-8 character, not advancing the pointer. This is called when
599: we know we are in UTF-8 mode. */
600:
601: #define GETCHAR(c, eptr) \
602: c = *eptr; \
603: if (c >= 0xc0) GETUTF8(c, eptr);
604:
605: /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
606: pointer. */
607:
608: #define GETCHARTEST(c, eptr) \
609: c = *eptr; \
1.1.1.2 ! misho 610: if (utf && c >= 0xc0) GETUTF8(c, eptr);
1.1 misho 611:
612: /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
613: the pointer. */
614:
615: #define GETUTF8INC(c, eptr) \
616: { \
617: if ((c & 0x20) == 0) \
618: c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
619: else if ((c & 0x10) == 0) \
620: { \
621: c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
622: eptr += 2; \
623: } \
624: else if ((c & 0x08) == 0) \
625: { \
626: c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
627: ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
628: eptr += 3; \
629: } \
630: else if ((c & 0x04) == 0) \
631: { \
632: c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
633: ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
634: (eptr[3] & 0x3f); \
635: eptr += 4; \
636: } \
637: else \
638: { \
639: c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
640: ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
641: ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
642: eptr += 5; \
643: } \
644: }
645:
646: /* Get the next UTF-8 character, advancing the pointer. This is called when we
647: know we are in UTF-8 mode. */
648:
649: #define GETCHARINC(c, eptr) \
650: c = *eptr++; \
651: if (c >= 0xc0) GETUTF8INC(c, eptr);
652:
653: /* Get the next character, testing for UTF-8 mode, and advancing the pointer.
654: This is called when we don't know if we are in UTF-8 mode. */
655:
656: #define GETCHARINCTEST(c, eptr) \
657: c = *eptr++; \
1.1.1.2 ! misho 658: if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
1.1 misho 659:
660: /* Base macro to pick up the remaining bytes of a UTF-8 character, not
661: advancing the pointer, incrementing the length. */
662:
663: #define GETUTF8LEN(c, eptr, len) \
664: { \
665: if ((c & 0x20) == 0) \
666: { \
667: c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
668: len++; \
669: } \
670: else if ((c & 0x10) == 0) \
671: { \
672: c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
673: len += 2; \
674: } \
675: else if ((c & 0x08) == 0) \
676: {\
677: c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
678: ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
679: len += 3; \
680: } \
681: else if ((c & 0x04) == 0) \
682: { \
683: c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
684: ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
685: (eptr[4] & 0x3f); \
686: len += 4; \
687: } \
688: else \
689: {\
690: c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
691: ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
692: ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
693: len += 5; \
694: } \
695: }
696:
697: /* Get the next UTF-8 character, not advancing the pointer, incrementing length
698: if there are extra bytes. This is called when we know we are in UTF-8 mode. */
699:
700: #define GETCHARLEN(c, eptr, len) \
701: c = *eptr; \
702: if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
703:
704: /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
705: pointer, incrementing length if there are extra bytes. This is called when we
706: do not know if we are in UTF-8 mode. */
707:
708: #define GETCHARLENTEST(c, eptr, len) \
709: c = *eptr; \
1.1.1.2 ! misho 710: if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
1.1 misho 711:
712: /* If the pointer is not at the start of a character, move it back until
713: it is. This is called only in UTF-8 mode - we don't put a test within the macro
714: because almost all calls are already within a block of UTF-8 only code. */
715:
716: #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
717:
1.1.1.2 ! misho 718: /* Same as above, just in the other direction. */
! 719: #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
! 720:
! 721: /* Same as above, but it allows a fully customizable form. */
! 722: #define ACROSSCHAR(condition, eptr, action) \
! 723: while((condition) && ((eptr) & 0xc0) == 0x80) action
! 724:
! 725: #else /* COMPILE_PCRE8 */
! 726:
! 727: #ifdef COMPILE_PCRE16
! 728:
! 729: /* Tells the biggest code point which can be encoded as a single character. */
! 730:
! 731: #define MAX_VALUE_FOR_SINGLE_CHAR 65535
! 732:
! 733: /* Tests whether the code point needs extra characters to decode. */
! 734:
! 735: #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
! 736:
! 737: /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
! 738: Otherwise it has an undefined behaviour. */
! 739:
! 740: #define GET_EXTRALEN(c) 1
! 741:
! 742: /* Returns TRUE, if the given character is not the first character
! 743: of a UTF sequence. */
! 744:
! 745: #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
! 746:
! 747: /* Base macro to pick up the low surrogate of a UTF-16 character, not
! 748: advancing the pointer. */
! 749:
! 750: #define GETUTF16(c, eptr) \
! 751: { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
! 752:
! 753: /* Get the next UTF-16 character, not advancing the pointer. This is called when
! 754: we know we are in UTF-16 mode. */
! 755:
! 756: #define GETCHAR(c, eptr) \
! 757: c = *eptr; \
! 758: if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
! 759:
! 760: /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
! 761: pointer. */
! 762:
! 763: #define GETCHARTEST(c, eptr) \
! 764: c = *eptr; \
! 765: if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
! 766:
! 767: /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
! 768: the pointer. */
! 769:
! 770: #define GETUTF16INC(c, eptr) \
! 771: { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
! 772:
! 773: /* Get the next UTF-16 character, advancing the pointer. This is called when we
! 774: know we are in UTF-16 mode. */
! 775:
! 776: #define GETCHARINC(c, eptr) \
! 777: c = *eptr++; \
! 778: if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
! 779:
! 780: /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
! 781: This is called when we don't know if we are in UTF-16 mode. */
! 782:
! 783: #define GETCHARINCTEST(c, eptr) \
! 784: c = *eptr++; \
! 785: if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
! 786:
! 787: /* Base macro to pick up the low surrogate of a UTF-16 character, not
! 788: advancing the pointer, incrementing the length. */
! 789:
! 790: #define GETUTF16LEN(c, eptr, len) \
! 791: { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
! 792:
! 793: /* Get the next UTF-16 character, not advancing the pointer, incrementing
! 794: length if there is a low surrogate. This is called when we know we are in
! 795: UTF-16 mode. */
! 796:
! 797: #define GETCHARLEN(c, eptr, len) \
! 798: c = *eptr; \
! 799: if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
! 800:
! 801: /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
! 802: pointer, incrementing length if there is a low surrogate. This is called when
! 803: we do not know if we are in UTF-16 mode. */
! 804:
! 805: #define GETCHARLENTEST(c, eptr, len) \
! 806: c = *eptr; \
! 807: if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
! 808:
! 809: /* If the pointer is not at the start of a character, move it back until
! 810: it is. This is called only in UTF-16 mode - we don't put a test within the
! 811: macro because almost all calls are already within a block of UTF-16 only
! 812: code. */
! 813:
! 814: #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
! 815:
! 816: /* Same as above, just in the other direction. */
! 817: #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
! 818:
! 819: /* Same as above, but it allows a fully customizable form. */
! 820: #define ACROSSCHAR(condition, eptr, action) \
! 821: if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
! 822:
! 823: #endif
! 824:
! 825: #endif /* COMPILE_PCRE8 */
! 826:
! 827: #endif /* SUPPORT_UTF */
1.1 misho 828:
829:
830: /* In case there is no definition of offsetof() provided - though any proper
831: Standard C system should have one. */
832:
833: #ifndef offsetof
834: #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))
835: #endif
836:
837:
838: /* Private flags containing information about the compiled regex. They used to
839: live at the top end of the options word, but that got almost full, so now they
840: are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
841: the restrictions on partial matching have been lifted. It remains for backwards
842: compatibility. */
843:
1.1.1.2 ! misho 844: #ifdef COMPILE_PCRE8
! 845: #define PCRE_MODE 0x0001 /* compiled in 8 bit mode */
! 846: #endif
! 847: #ifdef COMPILE_PCRE16
! 848: #define PCRE_MODE 0x0002 /* compiled in 16 bit mode */
! 849: #endif
! 850: #define PCRE_FIRSTSET 0x0010 /* first_char is set */
! 851: #define PCRE_FCH_CASELESS 0x0020 /* caseless first char */
! 852: #define PCRE_REQCHSET 0x0040 /* req_byte is set */
! 853: #define PCRE_RCH_CASELESS 0x0080 /* caseless requested char */
! 854: #define PCRE_STARTLINE 0x0100 /* start after \n for multiline */
! 855: #define PCRE_NOPARTIAL 0x0200 /* can't use partial with this regex */
! 856: #define PCRE_JCHANGED 0x0400 /* j option used in regex */
! 857: #define PCRE_HASCRORLF 0x0800 /* explicit \r or \n in pattern */
! 858: #define PCRE_HASTHEN 0x1000 /* pattern contains (*THEN) */
1.1 misho 859:
860: /* Flags for the "extra" block produced by pcre_study(). */
861:
862: #define PCRE_STUDY_MAPPED 0x0001 /* a map of starting chars exists */
863: #define PCRE_STUDY_MINLEN 0x0002 /* a minimum length field exists */
864:
865: /* Masks for identifying the public options that are permitted at compile
866: time, run time, or study time, respectively. */
867:
868: #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
869: PCRE_NEWLINE_ANYCRLF)
870:
871: #define PUBLIC_COMPILE_OPTIONS \
872: (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
873: PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
874: PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
875: PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
876: PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
877:
878: #define PUBLIC_EXEC_OPTIONS \
879: (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
880: PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
881: PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
882:
883: #define PUBLIC_DFA_EXEC_OPTIONS \
884: (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
885: PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
886: PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
887: PCRE_NO_START_OPTIMIZE)
888:
889: #define PUBLIC_STUDY_OPTIONS \
890: PCRE_STUDY_JIT_COMPILE
891:
1.1.1.2 ! misho 892: /* Magic number to provide a small check against being handed junk. */
1.1 misho 893:
894: #define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
895:
1.1.1.2 ! misho 896: /* This variable is used to detect a loaded regular expression
! 897: in different endianness. */
! 898:
! 899: #define REVERSED_MAGIC_NUMBER 0x45524350UL /* 'ERCP' */
! 900:
1.1 misho 901: /* Negative values for the firstchar and reqchar variables */
902:
903: #define REQ_UNSET (-2)
904: #define REQ_NONE (-1)
905:
906: /* The maximum remaining length of subject we are prepared to search for a
907: req_byte match. */
908:
909: #define REQ_BYTE_MAX 1000
910:
911: /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
912: environments where these macros are defined elsewhere. Unfortunately, there
913: is no way to do the same for the typedef. */
914:
915: typedef int BOOL;
916:
917: #ifndef FALSE
918: #define FALSE 0
919: #define TRUE 1
920: #endif
921:
922: /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
923: character constants like '*' because the compiler would emit their EBCDIC code,
924: which is different from their ASCII/UTF-8 code. Instead we define macros for
925: the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
926: is enabled. When UTF-8 support is not enabled, the definitions use character
927: literals. Both character and string versions of each character are needed, and
928: there are some longer strings as well.
929:
930: This means that, on EBCDIC platforms, the PCRE library can handle either
931: EBCDIC, or UTF-8, but not both. To support both in the same compiled library
932: would need different lookups depending on whether PCRE_UTF8 was set or not.
933: This would make it impossible to use characters in switch/case statements,
934: which would reduce performance. For a theoretical use (which nobody has asked
935: for) in a minority area (EBCDIC platforms), this is not sensible. Any
936: application that did need both could compile two versions of the library, using
937: macros to give the functions distinct names. */
938:
1.1.1.2 ! misho 939: #ifndef SUPPORT_UTF
1.1 misho 940:
941: /* UTF-8 support is not enabled; use the platform-dependent character literals
942: so that PCRE works on both ASCII and EBCDIC platforms, in non-UTF-mode only. */
943:
944: #define CHAR_HT '\t'
945: #define CHAR_VT '\v'
946: #define CHAR_FF '\f'
947: #define CHAR_CR '\r'
948: #define CHAR_NL '\n'
949: #define CHAR_BS '\b'
950: #define CHAR_BEL '\a'
951: #ifdef EBCDIC
952: #define CHAR_ESC '\047'
953: #define CHAR_DEL '\007'
954: #else
955: #define CHAR_ESC '\033'
956: #define CHAR_DEL '\177'
957: #endif
958:
959: #define CHAR_SPACE ' '
960: #define CHAR_EXCLAMATION_MARK '!'
961: #define CHAR_QUOTATION_MARK '"'
962: #define CHAR_NUMBER_SIGN '#'
963: #define CHAR_DOLLAR_SIGN '$'
964: #define CHAR_PERCENT_SIGN '%'
965: #define CHAR_AMPERSAND '&'
966: #define CHAR_APOSTROPHE '\''
967: #define CHAR_LEFT_PARENTHESIS '('
968: #define CHAR_RIGHT_PARENTHESIS ')'
969: #define CHAR_ASTERISK '*'
970: #define CHAR_PLUS '+'
971: #define CHAR_COMMA ','
972: #define CHAR_MINUS '-'
973: #define CHAR_DOT '.'
974: #define CHAR_SLASH '/'
975: #define CHAR_0 '0'
976: #define CHAR_1 '1'
977: #define CHAR_2 '2'
978: #define CHAR_3 '3'
979: #define CHAR_4 '4'
980: #define CHAR_5 '5'
981: #define CHAR_6 '6'
982: #define CHAR_7 '7'
983: #define CHAR_8 '8'
984: #define CHAR_9 '9'
985: #define CHAR_COLON ':'
986: #define CHAR_SEMICOLON ';'
987: #define CHAR_LESS_THAN_SIGN '<'
988: #define CHAR_EQUALS_SIGN '='
989: #define CHAR_GREATER_THAN_SIGN '>'
990: #define CHAR_QUESTION_MARK '?'
991: #define CHAR_COMMERCIAL_AT '@'
992: #define CHAR_A 'A'
993: #define CHAR_B 'B'
994: #define CHAR_C 'C'
995: #define CHAR_D 'D'
996: #define CHAR_E 'E'
997: #define CHAR_F 'F'
998: #define CHAR_G 'G'
999: #define CHAR_H 'H'
1000: #define CHAR_I 'I'
1001: #define CHAR_J 'J'
1002: #define CHAR_K 'K'
1003: #define CHAR_L 'L'
1004: #define CHAR_M 'M'
1005: #define CHAR_N 'N'
1006: #define CHAR_O 'O'
1007: #define CHAR_P 'P'
1008: #define CHAR_Q 'Q'
1009: #define CHAR_R 'R'
1010: #define CHAR_S 'S'
1011: #define CHAR_T 'T'
1012: #define CHAR_U 'U'
1013: #define CHAR_V 'V'
1014: #define CHAR_W 'W'
1015: #define CHAR_X 'X'
1016: #define CHAR_Y 'Y'
1017: #define CHAR_Z 'Z'
1018: #define CHAR_LEFT_SQUARE_BRACKET '['
1019: #define CHAR_BACKSLASH '\\'
1020: #define CHAR_RIGHT_SQUARE_BRACKET ']'
1021: #define CHAR_CIRCUMFLEX_ACCENT '^'
1022: #define CHAR_UNDERSCORE '_'
1023: #define CHAR_GRAVE_ACCENT '`'
1024: #define CHAR_a 'a'
1025: #define CHAR_b 'b'
1026: #define CHAR_c 'c'
1027: #define CHAR_d 'd'
1028: #define CHAR_e 'e'
1029: #define CHAR_f 'f'
1030: #define CHAR_g 'g'
1031: #define CHAR_h 'h'
1032: #define CHAR_i 'i'
1033: #define CHAR_j 'j'
1034: #define CHAR_k 'k'
1035: #define CHAR_l 'l'
1036: #define CHAR_m 'm'
1037: #define CHAR_n 'n'
1038: #define CHAR_o 'o'
1039: #define CHAR_p 'p'
1040: #define CHAR_q 'q'
1041: #define CHAR_r 'r'
1042: #define CHAR_s 's'
1043: #define CHAR_t 't'
1044: #define CHAR_u 'u'
1045: #define CHAR_v 'v'
1046: #define CHAR_w 'w'
1047: #define CHAR_x 'x'
1048: #define CHAR_y 'y'
1049: #define CHAR_z 'z'
1050: #define CHAR_LEFT_CURLY_BRACKET '{'
1051: #define CHAR_VERTICAL_LINE '|'
1052: #define CHAR_RIGHT_CURLY_BRACKET '}'
1053: #define CHAR_TILDE '~'
1054:
1055: #define STR_HT "\t"
1056: #define STR_VT "\v"
1057: #define STR_FF "\f"
1058: #define STR_CR "\r"
1059: #define STR_NL "\n"
1060: #define STR_BS "\b"
1061: #define STR_BEL "\a"
1062: #ifdef EBCDIC
1063: #define STR_ESC "\047"
1064: #define STR_DEL "\007"
1065: #else
1066: #define STR_ESC "\033"
1067: #define STR_DEL "\177"
1068: #endif
1069:
1070: #define STR_SPACE " "
1071: #define STR_EXCLAMATION_MARK "!"
1072: #define STR_QUOTATION_MARK "\""
1073: #define STR_NUMBER_SIGN "#"
1074: #define STR_DOLLAR_SIGN "$"
1075: #define STR_PERCENT_SIGN "%"
1076: #define STR_AMPERSAND "&"
1077: #define STR_APOSTROPHE "'"
1078: #define STR_LEFT_PARENTHESIS "("
1079: #define STR_RIGHT_PARENTHESIS ")"
1080: #define STR_ASTERISK "*"
1081: #define STR_PLUS "+"
1082: #define STR_COMMA ","
1083: #define STR_MINUS "-"
1084: #define STR_DOT "."
1085: #define STR_SLASH "/"
1086: #define STR_0 "0"
1087: #define STR_1 "1"
1088: #define STR_2 "2"
1089: #define STR_3 "3"
1090: #define STR_4 "4"
1091: #define STR_5 "5"
1092: #define STR_6 "6"
1093: #define STR_7 "7"
1094: #define STR_8 "8"
1095: #define STR_9 "9"
1096: #define STR_COLON ":"
1097: #define STR_SEMICOLON ";"
1098: #define STR_LESS_THAN_SIGN "<"
1099: #define STR_EQUALS_SIGN "="
1100: #define STR_GREATER_THAN_SIGN ">"
1101: #define STR_QUESTION_MARK "?"
1102: #define STR_COMMERCIAL_AT "@"
1103: #define STR_A "A"
1104: #define STR_B "B"
1105: #define STR_C "C"
1106: #define STR_D "D"
1107: #define STR_E "E"
1108: #define STR_F "F"
1109: #define STR_G "G"
1110: #define STR_H "H"
1111: #define STR_I "I"
1112: #define STR_J "J"
1113: #define STR_K "K"
1114: #define STR_L "L"
1115: #define STR_M "M"
1116: #define STR_N "N"
1117: #define STR_O "O"
1118: #define STR_P "P"
1119: #define STR_Q "Q"
1120: #define STR_R "R"
1121: #define STR_S "S"
1122: #define STR_T "T"
1123: #define STR_U "U"
1124: #define STR_V "V"
1125: #define STR_W "W"
1126: #define STR_X "X"
1127: #define STR_Y "Y"
1128: #define STR_Z "Z"
1129: #define STR_LEFT_SQUARE_BRACKET "["
1130: #define STR_BACKSLASH "\\"
1131: #define STR_RIGHT_SQUARE_BRACKET "]"
1132: #define STR_CIRCUMFLEX_ACCENT "^"
1133: #define STR_UNDERSCORE "_"
1134: #define STR_GRAVE_ACCENT "`"
1135: #define STR_a "a"
1136: #define STR_b "b"
1137: #define STR_c "c"
1138: #define STR_d "d"
1139: #define STR_e "e"
1140: #define STR_f "f"
1141: #define STR_g "g"
1142: #define STR_h "h"
1143: #define STR_i "i"
1144: #define STR_j "j"
1145: #define STR_k "k"
1146: #define STR_l "l"
1147: #define STR_m "m"
1148: #define STR_n "n"
1149: #define STR_o "o"
1150: #define STR_p "p"
1151: #define STR_q "q"
1152: #define STR_r "r"
1153: #define STR_s "s"
1154: #define STR_t "t"
1155: #define STR_u "u"
1156: #define STR_v "v"
1157: #define STR_w "w"
1158: #define STR_x "x"
1159: #define STR_y "y"
1160: #define STR_z "z"
1161: #define STR_LEFT_CURLY_BRACKET "{"
1162: #define STR_VERTICAL_LINE "|"
1163: #define STR_RIGHT_CURLY_BRACKET "}"
1164: #define STR_TILDE "~"
1165:
1166: #define STRING_ACCEPT0 "ACCEPT\0"
1167: #define STRING_COMMIT0 "COMMIT\0"
1168: #define STRING_F0 "F\0"
1169: #define STRING_FAIL0 "FAIL\0"
1170: #define STRING_MARK0 "MARK\0"
1171: #define STRING_PRUNE0 "PRUNE\0"
1172: #define STRING_SKIP0 "SKIP\0"
1173: #define STRING_THEN "THEN"
1174:
1175: #define STRING_alpha0 "alpha\0"
1176: #define STRING_lower0 "lower\0"
1177: #define STRING_upper0 "upper\0"
1178: #define STRING_alnum0 "alnum\0"
1179: #define STRING_ascii0 "ascii\0"
1180: #define STRING_blank0 "blank\0"
1181: #define STRING_cntrl0 "cntrl\0"
1182: #define STRING_digit0 "digit\0"
1183: #define STRING_graph0 "graph\0"
1184: #define STRING_print0 "print\0"
1185: #define STRING_punct0 "punct\0"
1186: #define STRING_space0 "space\0"
1187: #define STRING_word0 "word\0"
1188: #define STRING_xdigit "xdigit"
1189:
1190: #define STRING_DEFINE "DEFINE"
1191:
1192: #define STRING_CR_RIGHTPAR "CR)"
1193: #define STRING_LF_RIGHTPAR "LF)"
1194: #define STRING_CRLF_RIGHTPAR "CRLF)"
1195: #define STRING_ANY_RIGHTPAR "ANY)"
1196: #define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)"
1197: #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"
1198: #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"
1.1.1.2 ! misho 1199: #ifdef COMPILE_PCRE8
! 1200: #define STRING_UTF_RIGHTPAR "UTF8)"
! 1201: #endif
! 1202: #ifdef COMPILE_PCRE16
! 1203: #define STRING_UTF_RIGHTPAR "UTF16)"
! 1204: #endif
1.1 misho 1205: #define STRING_UCP_RIGHTPAR "UCP)"
1206: #define STRING_NO_START_OPT_RIGHTPAR "NO_START_OPT)"
1207:
1.1.1.2 ! misho 1208: #else /* SUPPORT_UTF */
1.1 misho 1209:
1210: /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1211: works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
1212: only. */
1213:
1214: #define CHAR_HT '\011'
1215: #define CHAR_VT '\013'
1216: #define CHAR_FF '\014'
1217: #define CHAR_CR '\015'
1218: #define CHAR_NL '\012'
1219: #define CHAR_BS '\010'
1220: #define CHAR_BEL '\007'
1221: #define CHAR_ESC '\033'
1222: #define CHAR_DEL '\177'
1223:
1224: #define CHAR_SPACE '\040'
1225: #define CHAR_EXCLAMATION_MARK '\041'
1226: #define CHAR_QUOTATION_MARK '\042'
1227: #define CHAR_NUMBER_SIGN '\043'
1228: #define CHAR_DOLLAR_SIGN '\044'
1229: #define CHAR_PERCENT_SIGN '\045'
1230: #define CHAR_AMPERSAND '\046'
1231: #define CHAR_APOSTROPHE '\047'
1232: #define CHAR_LEFT_PARENTHESIS '\050'
1233: #define CHAR_RIGHT_PARENTHESIS '\051'
1234: #define CHAR_ASTERISK '\052'
1235: #define CHAR_PLUS '\053'
1236: #define CHAR_COMMA '\054'
1237: #define CHAR_MINUS '\055'
1238: #define CHAR_DOT '\056'
1239: #define CHAR_SLASH '\057'
1240: #define CHAR_0 '\060'
1241: #define CHAR_1 '\061'
1242: #define CHAR_2 '\062'
1243: #define CHAR_3 '\063'
1244: #define CHAR_4 '\064'
1245: #define CHAR_5 '\065'
1246: #define CHAR_6 '\066'
1247: #define CHAR_7 '\067'
1248: #define CHAR_8 '\070'
1249: #define CHAR_9 '\071'
1250: #define CHAR_COLON '\072'
1251: #define CHAR_SEMICOLON '\073'
1252: #define CHAR_LESS_THAN_SIGN '\074'
1253: #define CHAR_EQUALS_SIGN '\075'
1254: #define CHAR_GREATER_THAN_SIGN '\076'
1255: #define CHAR_QUESTION_MARK '\077'
1256: #define CHAR_COMMERCIAL_AT '\100'
1257: #define CHAR_A '\101'
1258: #define CHAR_B '\102'
1259: #define CHAR_C '\103'
1260: #define CHAR_D '\104'
1261: #define CHAR_E '\105'
1262: #define CHAR_F '\106'
1263: #define CHAR_G '\107'
1264: #define CHAR_H '\110'
1265: #define CHAR_I '\111'
1266: #define CHAR_J '\112'
1267: #define CHAR_K '\113'
1268: #define CHAR_L '\114'
1269: #define CHAR_M '\115'
1270: #define CHAR_N '\116'
1271: #define CHAR_O '\117'
1272: #define CHAR_P '\120'
1273: #define CHAR_Q '\121'
1274: #define CHAR_R '\122'
1275: #define CHAR_S '\123'
1276: #define CHAR_T '\124'
1277: #define CHAR_U '\125'
1278: #define CHAR_V '\126'
1279: #define CHAR_W '\127'
1280: #define CHAR_X '\130'
1281: #define CHAR_Y '\131'
1282: #define CHAR_Z '\132'
1283: #define CHAR_LEFT_SQUARE_BRACKET '\133'
1284: #define CHAR_BACKSLASH '\134'
1285: #define CHAR_RIGHT_SQUARE_BRACKET '\135'
1286: #define CHAR_CIRCUMFLEX_ACCENT '\136'
1287: #define CHAR_UNDERSCORE '\137'
1288: #define CHAR_GRAVE_ACCENT '\140'
1289: #define CHAR_a '\141'
1290: #define CHAR_b '\142'
1291: #define CHAR_c '\143'
1292: #define CHAR_d '\144'
1293: #define CHAR_e '\145'
1294: #define CHAR_f '\146'
1295: #define CHAR_g '\147'
1296: #define CHAR_h '\150'
1297: #define CHAR_i '\151'
1298: #define CHAR_j '\152'
1299: #define CHAR_k '\153'
1300: #define CHAR_l '\154'
1301: #define CHAR_m '\155'
1302: #define CHAR_n '\156'
1303: #define CHAR_o '\157'
1304: #define CHAR_p '\160'
1305: #define CHAR_q '\161'
1306: #define CHAR_r '\162'
1307: #define CHAR_s '\163'
1308: #define CHAR_t '\164'
1309: #define CHAR_u '\165'
1310: #define CHAR_v '\166'
1311: #define CHAR_w '\167'
1312: #define CHAR_x '\170'
1313: #define CHAR_y '\171'
1314: #define CHAR_z '\172'
1315: #define CHAR_LEFT_CURLY_BRACKET '\173'
1316: #define CHAR_VERTICAL_LINE '\174'
1317: #define CHAR_RIGHT_CURLY_BRACKET '\175'
1318: #define CHAR_TILDE '\176'
1319:
1320: #define STR_HT "\011"
1321: #define STR_VT "\013"
1322: #define STR_FF "\014"
1323: #define STR_CR "\015"
1324: #define STR_NL "\012"
1325: #define STR_BS "\010"
1326: #define STR_BEL "\007"
1327: #define STR_ESC "\033"
1328: #define STR_DEL "\177"
1329:
1330: #define STR_SPACE "\040"
1331: #define STR_EXCLAMATION_MARK "\041"
1332: #define STR_QUOTATION_MARK "\042"
1333: #define STR_NUMBER_SIGN "\043"
1334: #define STR_DOLLAR_SIGN "\044"
1335: #define STR_PERCENT_SIGN "\045"
1336: #define STR_AMPERSAND "\046"
1337: #define STR_APOSTROPHE "\047"
1338: #define STR_LEFT_PARENTHESIS "\050"
1339: #define STR_RIGHT_PARENTHESIS "\051"
1340: #define STR_ASTERISK "\052"
1341: #define STR_PLUS "\053"
1342: #define STR_COMMA "\054"
1343: #define STR_MINUS "\055"
1344: #define STR_DOT "\056"
1345: #define STR_SLASH "\057"
1346: #define STR_0 "\060"
1347: #define STR_1 "\061"
1348: #define STR_2 "\062"
1349: #define STR_3 "\063"
1350: #define STR_4 "\064"
1351: #define STR_5 "\065"
1352: #define STR_6 "\066"
1353: #define STR_7 "\067"
1354: #define STR_8 "\070"
1355: #define STR_9 "\071"
1356: #define STR_COLON "\072"
1357: #define STR_SEMICOLON "\073"
1358: #define STR_LESS_THAN_SIGN "\074"
1359: #define STR_EQUALS_SIGN "\075"
1360: #define STR_GREATER_THAN_SIGN "\076"
1361: #define STR_QUESTION_MARK "\077"
1362: #define STR_COMMERCIAL_AT "\100"
1363: #define STR_A "\101"
1364: #define STR_B "\102"
1365: #define STR_C "\103"
1366: #define STR_D "\104"
1367: #define STR_E "\105"
1368: #define STR_F "\106"
1369: #define STR_G "\107"
1370: #define STR_H "\110"
1371: #define STR_I "\111"
1372: #define STR_J "\112"
1373: #define STR_K "\113"
1374: #define STR_L "\114"
1375: #define STR_M "\115"
1376: #define STR_N "\116"
1377: #define STR_O "\117"
1378: #define STR_P "\120"
1379: #define STR_Q "\121"
1380: #define STR_R "\122"
1381: #define STR_S "\123"
1382: #define STR_T "\124"
1383: #define STR_U "\125"
1384: #define STR_V "\126"
1385: #define STR_W "\127"
1386: #define STR_X "\130"
1387: #define STR_Y "\131"
1388: #define STR_Z "\132"
1389: #define STR_LEFT_SQUARE_BRACKET "\133"
1390: #define STR_BACKSLASH "\134"
1391: #define STR_RIGHT_SQUARE_BRACKET "\135"
1392: #define STR_CIRCUMFLEX_ACCENT "\136"
1393: #define STR_UNDERSCORE "\137"
1394: #define STR_GRAVE_ACCENT "\140"
1395: #define STR_a "\141"
1396: #define STR_b "\142"
1397: #define STR_c "\143"
1398: #define STR_d "\144"
1399: #define STR_e "\145"
1400: #define STR_f "\146"
1401: #define STR_g "\147"
1402: #define STR_h "\150"
1403: #define STR_i "\151"
1404: #define STR_j "\152"
1405: #define STR_k "\153"
1406: #define STR_l "\154"
1407: #define STR_m "\155"
1408: #define STR_n "\156"
1409: #define STR_o "\157"
1410: #define STR_p "\160"
1411: #define STR_q "\161"
1412: #define STR_r "\162"
1413: #define STR_s "\163"
1414: #define STR_t "\164"
1415: #define STR_u "\165"
1416: #define STR_v "\166"
1417: #define STR_w "\167"
1418: #define STR_x "\170"
1419: #define STR_y "\171"
1420: #define STR_z "\172"
1421: #define STR_LEFT_CURLY_BRACKET "\173"
1422: #define STR_VERTICAL_LINE "\174"
1423: #define STR_RIGHT_CURLY_BRACKET "\175"
1424: #define STR_TILDE "\176"
1425:
1426: #define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1427: #define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1428: #define STRING_F0 STR_F "\0"
1429: #define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0"
1430: #define STRING_MARK0 STR_M STR_A STR_R STR_K "\0"
1431: #define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0"
1432: #define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0"
1433: #define STRING_THEN STR_T STR_H STR_E STR_N
1434:
1435: #define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0"
1436: #define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0"
1437: #define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0"
1438: #define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0"
1439: #define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0"
1440: #define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0"
1441: #define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0"
1442: #define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0"
1443: #define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0"
1444: #define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0"
1445: #define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0"
1446: #define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0"
1447: #define STRING_word0 STR_w STR_o STR_r STR_d "\0"
1448: #define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t
1449:
1450: #define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E
1451:
1452: #define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS
1453: #define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS
1454: #define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1455: #define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1456: #define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1457: #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
1458: #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
1.1.1.2 ! misho 1459: #ifdef COMPILE_PCRE8
! 1460: #define STRING_UTF_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
! 1461: #endif
! 1462: #ifdef COMPILE_PCRE16
! 1463: #define STRING_UTF_RIGHTPAR STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
! 1464: #endif
1.1 misho 1465: #define STRING_UCP_RIGHTPAR STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1466: #define STRING_NO_START_OPT_RIGHTPAR STR_N STR_O STR_UNDERSCORE STR_S STR_T STR_A STR_R STR_T STR_UNDERSCORE STR_O STR_P STR_T STR_RIGHT_PARENTHESIS
1467:
1.1.1.2 ! misho 1468: #endif /* SUPPORT_UTF */
1.1 misho 1469:
1470: /* Escape items that are just an encoding of a particular data value. */
1471:
1472: #ifndef ESC_e
1473: #define ESC_e CHAR_ESC
1474: #endif
1475:
1476: #ifndef ESC_f
1477: #define ESC_f CHAR_FF
1478: #endif
1479:
1480: #ifndef ESC_n
1481: #define ESC_n CHAR_NL
1482: #endif
1483:
1484: #ifndef ESC_r
1485: #define ESC_r CHAR_CR
1486: #endif
1487:
1488: /* We can't officially use ESC_t because it is a POSIX reserved identifier
1489: (presumably because of all the others like size_t). */
1490:
1491: #ifndef ESC_tee
1492: #define ESC_tee CHAR_HT
1493: #endif
1494:
1495: /* Codes for different types of Unicode property */
1496:
1497: #define PT_ANY 0 /* Any property - matches all chars */
1498: #define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */
1499: #define PT_GC 2 /* Specified general characteristic (e.g. L) */
1500: #define PT_PC 3 /* Specified particular characteristic (e.g. Lu) */
1501: #define PT_SC 4 /* Script (e.g. Han) */
1502: #define PT_ALNUM 5 /* Alphanumeric - the union of L and N */
1503: #define PT_SPACE 6 /* Perl space - Z plus 9,10,12,13 */
1504: #define PT_PXSPACE 7 /* POSIX space - Z plus 9,10,11,12,13 */
1505: #define PT_WORD 8 /* Word - L plus N plus underscore */
1506:
1507: /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1.1.1.2 ! misho 1508: contain characters with values greater than 255. */
1.1 misho 1509:
1510: #define XCL_NOT 0x01 /* Flag: this is a negative class */
1511: #define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
1512:
1513: #define XCL_END 0 /* Marks end of individual items */
1514: #define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
1515: #define XCL_RANGE 2 /* A range (two multibyte chars) follows */
1516: #define XCL_PROP 3 /* Unicode property (2-byte property code follows) */
1517: #define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
1518:
1519: /* These are escaped items that aren't just an encoding of a particular data
1520: value such as \n. They must have non-zero values, as check_escape() returns
1521: their negation. Also, they must appear in the same order as in the opcode
1522: definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1523: corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1.1.1.2 ! misho 1524: used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1.1 misho 1525: non-DOTALL mode, "." behaves like \N.
1526:
1527: The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1528: when PCRE_UCP is set, when replacement of \d etc by \p sequences is required.
1529: They must be contiguous, and remain in order so that the replacements can be
1530: looked up from a table.
1531:
1532: The final escape must be ESC_REF as subsequent values are used for
1533: backreferences (\1, \2, \3, etc). There are two tests in the code for an escape
1534: greater than ESC_b and less than ESC_Z to detect the types that may be
1535: repeated. These are the types that consume characters. If any new escapes are
1536: put in between that don't consume a character, that code will have to change.
1537: */
1538:
1539: enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1540: ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1541: ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1542: ESC_E, ESC_Q, ESC_g, ESC_k,
1543: ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu,
1544: ESC_REF };
1545:
1546: /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1547: OP_EOD must correspond in order to the list of escapes immediately above.
1548:
1549: *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1550: that follow must also be updated to match. There are also tables called
1551: "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1552:
1553: enum {
1554: OP_END, /* 0 End of pattern */
1555:
1556: /* Values corresponding to backslashed metacharacters */
1557:
1558: OP_SOD, /* 1 Start of data: \A */
1559: OP_SOM, /* 2 Start of match (subject + offset): \G */
1560: OP_SET_SOM, /* 3 Set start of match (\K) */
1561: OP_NOT_WORD_BOUNDARY, /* 4 \B */
1562: OP_WORD_BOUNDARY, /* 5 \b */
1563: OP_NOT_DIGIT, /* 6 \D */
1564: OP_DIGIT, /* 7 \d */
1565: OP_NOT_WHITESPACE, /* 8 \S */
1566: OP_WHITESPACE, /* 9 \s */
1567: OP_NOT_WORDCHAR, /* 10 \W */
1568: OP_WORDCHAR, /* 11 \w */
1569:
1570: OP_ANY, /* 12 Match any character except newline */
1571: OP_ALLANY, /* 13 Match any character */
1572: OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1573: OP_NOTPROP, /* 15 \P (not Unicode property) */
1574: OP_PROP, /* 16 \p (Unicode property) */
1575: OP_ANYNL, /* 17 \R (any newline sequence) */
1576: OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */
1577: OP_HSPACE, /* 19 \h (horizontal whitespace) */
1578: OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */
1579: OP_VSPACE, /* 21 \v (vertical whitespace) */
1580: OP_EXTUNI, /* 22 \X (extended Unicode sequence */
1581: OP_EODN, /* 23 End of data or \n at end of data: \Z. */
1582: OP_EOD, /* 24 End of data: \z */
1583:
1584: OP_CIRC, /* 25 Start of line - not multiline */
1585: OP_CIRCM, /* 26 Start of line - multiline */
1586: OP_DOLL, /* 27 End of line - not multiline */
1587: OP_DOLLM, /* 28 End of line - multiline */
1588: OP_CHAR, /* 29 Match one character, casefully */
1589: OP_CHARI, /* 30 Match one character, caselessly */
1590: OP_NOT, /* 31 Match one character, not the given one, casefully */
1591: OP_NOTI, /* 32 Match one character, not the given one, caselessly */
1592:
1593: /* The following sets of 13 opcodes must always be kept in step because
1594: the offset from the first one is used to generate the others. */
1595:
1596: /**** Single characters, caseful, must precede the caseless ones ****/
1597:
1598: OP_STAR, /* 33 The maximizing and minimizing versions of */
1599: OP_MINSTAR, /* 34 these six opcodes must come in pairs, with */
1600: OP_PLUS, /* 35 the minimizing one second. */
1601: OP_MINPLUS, /* 36 */
1602: OP_QUERY, /* 37 */
1603: OP_MINQUERY, /* 38 */
1604:
1605: OP_UPTO, /* 39 From 0 to n matches of one character, caseful*/
1606: OP_MINUPTO, /* 40 */
1607: OP_EXACT, /* 41 Exactly n matches */
1608:
1609: OP_POSSTAR, /* 42 Possessified star, caseful */
1610: OP_POSPLUS, /* 43 Possessified plus, caseful */
1611: OP_POSQUERY, /* 44 Posesssified query, caseful */
1612: OP_POSUPTO, /* 45 Possessified upto, caseful */
1613:
1614: /**** Single characters, caseless, must follow the caseful ones */
1615:
1616: OP_STARI, /* 46 */
1617: OP_MINSTARI, /* 47 */
1618: OP_PLUSI, /* 48 */
1619: OP_MINPLUSI, /* 49 */
1620: OP_QUERYI, /* 50 */
1621: OP_MINQUERYI, /* 51 */
1622:
1623: OP_UPTOI, /* 52 From 0 to n matches of one character, caseless */
1624: OP_MINUPTOI, /* 53 */
1625: OP_EXACTI, /* 54 */
1626:
1627: OP_POSSTARI, /* 55 Possessified star, caseless */
1628: OP_POSPLUSI, /* 56 Possessified plus, caseless */
1629: OP_POSQUERYI, /* 57 Posesssified query, caseless */
1630: OP_POSUPTOI, /* 58 Possessified upto, caseless */
1631:
1632: /**** The negated ones must follow the non-negated ones, and match them ****/
1633: /**** Negated single character, caseful; must precede the caseless ones ****/
1634:
1635: OP_NOTSTAR, /* 59 The maximizing and minimizing versions of */
1636: OP_NOTMINSTAR, /* 60 these six opcodes must come in pairs, with */
1637: OP_NOTPLUS, /* 61 the minimizing one second. They must be in */
1638: OP_NOTMINPLUS, /* 62 exactly the same order as those above. */
1639: OP_NOTQUERY, /* 63 */
1640: OP_NOTMINQUERY, /* 64 */
1641:
1642: OP_NOTUPTO, /* 65 From 0 to n matches, caseful */
1643: OP_NOTMINUPTO, /* 66 */
1644: OP_NOTEXACT, /* 67 Exactly n matches */
1645:
1646: OP_NOTPOSSTAR, /* 68 Possessified versions, caseful */
1647: OP_NOTPOSPLUS, /* 69 */
1648: OP_NOTPOSQUERY, /* 70 */
1649: OP_NOTPOSUPTO, /* 71 */
1650:
1651: /**** Negated single character, caseless; must follow the caseful ones ****/
1652:
1653: OP_NOTSTARI, /* 72 */
1654: OP_NOTMINSTARI, /* 73 */
1655: OP_NOTPLUSI, /* 74 */
1656: OP_NOTMINPLUSI, /* 75 */
1657: OP_NOTQUERYI, /* 76 */
1658: OP_NOTMINQUERYI, /* 77 */
1659:
1660: OP_NOTUPTOI, /* 78 From 0 to n matches, caseless */
1661: OP_NOTMINUPTOI, /* 79 */
1662: OP_NOTEXACTI, /* 80 Exactly n matches */
1663:
1664: OP_NOTPOSSTARI, /* 81 Possessified versions, caseless */
1665: OP_NOTPOSPLUSI, /* 82 */
1666: OP_NOTPOSQUERYI, /* 83 */
1667: OP_NOTPOSUPTOI, /* 84 */
1668:
1669: /**** Character types ****/
1670:
1671: OP_TYPESTAR, /* 85 The maximizing and minimizing versions of */
1672: OP_TYPEMINSTAR, /* 86 these six opcodes must come in pairs, with */
1673: OP_TYPEPLUS, /* 87 the minimizing one second. These codes must */
1674: OP_TYPEMINPLUS, /* 88 be in exactly the same order as those above. */
1675: OP_TYPEQUERY, /* 89 */
1676: OP_TYPEMINQUERY, /* 90 */
1677:
1678: OP_TYPEUPTO, /* 91 From 0 to n matches */
1679: OP_TYPEMINUPTO, /* 92 */
1680: OP_TYPEEXACT, /* 93 Exactly n matches */
1681:
1682: OP_TYPEPOSSTAR, /* 94 Possessified versions */
1683: OP_TYPEPOSPLUS, /* 95 */
1684: OP_TYPEPOSQUERY, /* 96 */
1685: OP_TYPEPOSUPTO, /* 97 */
1686:
1687: /* These are used for character classes and back references; only the
1688: first six are the same as the sets above. */
1689:
1690: OP_CRSTAR, /* 98 The maximizing and minimizing versions of */
1691: OP_CRMINSTAR, /* 99 all these opcodes must come in pairs, with */
1692: OP_CRPLUS, /* 100 the minimizing one second. These codes must */
1693: OP_CRMINPLUS, /* 101 be in exactly the same order as those above. */
1694: OP_CRQUERY, /* 102 */
1695: OP_CRMINQUERY, /* 103 */
1696:
1697: OP_CRRANGE, /* 104 These are different to the three sets above. */
1698: OP_CRMINRANGE, /* 105 */
1699:
1700: /* End of quantifier opcodes */
1701:
1702: OP_CLASS, /* 106 Match a character class, chars < 256 only */
1703: OP_NCLASS, /* 107 Same, but the bitmap was created from a negative
1704: class - the difference is relevant only when a
1.1.1.2 ! misho 1705: character > 255 is encountered. */
! 1706: OP_XCLASS, /* 108 Extended class for handling > 255 chars within the
1.1 misho 1707: class. This does both positive and negative. */
1708: OP_REF, /* 109 Match a back reference, casefully */
1709: OP_REFI, /* 110 Match a back reference, caselessly */
1710: OP_RECURSE, /* 111 Match a numbered subpattern (possibly recursive) */
1711: OP_CALLOUT, /* 112 Call out to external function if provided */
1712:
1713: OP_ALT, /* 113 Start of alternation */
1714: OP_KET, /* 114 End of group that doesn't have an unbounded repeat */
1715: OP_KETRMAX, /* 115 These two must remain together and in this */
1716: OP_KETRMIN, /* 116 order. They are for groups the repeat for ever. */
1717: OP_KETRPOS, /* 117 Possessive unlimited repeat. */
1718:
1719: /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
1720: asserts must remain in order. */
1721:
1722: OP_REVERSE, /* 118 Move pointer back - used in lookbehind assertions */
1723: OP_ASSERT, /* 119 Positive lookahead */
1724: OP_ASSERT_NOT, /* 120 Negative lookahead */
1725: OP_ASSERTBACK, /* 121 Positive lookbehind */
1726: OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
1727:
1728: /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
1729: after the assertions, with ONCE first, as there's a test for >= ONCE for a
1730: subpattern that isn't an assertion. The POS versions must immediately follow
1731: the non-POS versions in each case. */
1732:
1733: OP_ONCE, /* 123 Atomic group, contains captures */
1734: OP_ONCE_NC, /* 124 Atomic group containing no captures */
1735: OP_BRA, /* 125 Start of non-capturing bracket */
1736: OP_BRAPOS, /* 126 Ditto, with unlimited, possessive repeat */
1737: OP_CBRA, /* 127 Start of capturing bracket */
1738: OP_CBRAPOS, /* 128 Ditto, with unlimited, possessive repeat */
1739: OP_COND, /* 129 Conditional group */
1740:
1741: /* These five must follow the previous five, in the same order. There's a
1742: check for >= SBRA to distinguish the two sets. */
1743:
1744: OP_SBRA, /* 130 Start of non-capturing bracket, check empty */
1745: OP_SBRAPOS, /* 131 Ditto, with unlimited, possessive repeat */
1746: OP_SCBRA, /* 132 Start of capturing bracket, check empty */
1747: OP_SCBRAPOS, /* 133 Ditto, with unlimited, possessive repeat */
1748: OP_SCOND, /* 134 Conditional group, check empty */
1749:
1750: /* The next two pairs must (respectively) be kept together. */
1751:
1752: OP_CREF, /* 135 Used to hold a capture number as condition */
1753: OP_NCREF, /* 136 Same, but generated by a name reference*/
1754: OP_RREF, /* 137 Used to hold a recursion number as condition */
1755: OP_NRREF, /* 138 Same, but generated by a name reference*/
1756: OP_DEF, /* 139 The DEFINE condition */
1757:
1758: OP_BRAZERO, /* 140 These two must remain together and in this */
1759: OP_BRAMINZERO, /* 141 order. */
1760: OP_BRAPOSZERO, /* 142 */
1761:
1762: /* These are backtracking control verbs */
1763:
1764: OP_MARK, /* 143 always has an argument */
1765: OP_PRUNE, /* 144 */
1766: OP_PRUNE_ARG, /* 145 same, but with argument */
1767: OP_SKIP, /* 146 */
1768: OP_SKIP_ARG, /* 147 same, but with argument */
1769: OP_THEN, /* 148 */
1770: OP_THEN_ARG, /* 149 same, but with argument */
1771: OP_COMMIT, /* 150 */
1772:
1773: /* These are forced failure and success verbs */
1774:
1775: OP_FAIL, /* 151 */
1776: OP_ACCEPT, /* 152 */
1777: OP_ASSERT_ACCEPT, /* 153 Used inside assertions */
1778: OP_CLOSE, /* 154 Used before OP_ACCEPT to close open captures */
1779:
1780: /* This is used to skip a subpattern with a {0} quantifier */
1781:
1782: OP_SKIPZERO, /* 155 */
1783:
1784: /* This is not an opcode, but is used to check that tables indexed by opcode
1785: are the correct length, in order to catch updating errors - there have been
1786: some in the past. */
1787:
1788: OP_TABLE_LENGTH
1789: };
1790:
1791: /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
1792: definitions that follow must also be updated to match. There are also tables
1793: called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1794:
1795:
1796: /* This macro defines textual names for all the opcodes. These are used only
1797: for debugging, and some of them are only partial names. The macro is referenced
1798: only in pcre_printint.c, which fills out the full names in many cases (and in
1799: some cases doesn't actually use these names at all). */
1800:
1801: #define OP_NAME_LIST \
1802: "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \
1803: "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \
1804: "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \
1805: "extuni", "\\Z", "\\z", \
1806: "^", "^", "$", "$", "char", "chari", "not", "noti", \
1807: "*", "*?", "+", "+?", "?", "??", \
1808: "{", "{", "{", \
1809: "*+","++", "?+", "{", \
1810: "*", "*?", "+", "+?", "?", "??", \
1811: "{", "{", "{", \
1812: "*+","++", "?+", "{", \
1813: "*", "*?", "+", "+?", "?", "??", \
1814: "{", "{", "{", \
1815: "*+","++", "?+", "{", \
1816: "*", "*?", "+", "+?", "?", "??", \
1817: "{", "{", "{", \
1818: "*+","++", "?+", "{", \
1819: "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
1820: "*+","++", "?+", "{", \
1821: "*", "*?", "+", "+?", "?", "??", "{", "{", \
1822: "class", "nclass", "xclass", "Ref", "Refi", \
1823: "Recurse", "Callout", \
1824: "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos", \
1825: "Reverse", "Assert", "Assert not", "AssertB", "AssertB not", \
1826: "Once", "Once_NC", \
1827: "Bra", "BraPos", "CBra", "CBraPos", \
1828: "Cond", \
1829: "SBra", "SBraPos", "SCBra", "SCBraPos", \
1830: "SCond", \
1831: "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def", \
1832: "Brazero", "Braminzero", "Braposzero", \
1833: "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP", \
1834: "*THEN", "*THEN", "*COMMIT", "*FAIL", \
1835: "*ACCEPT", "*ASSERT_ACCEPT", \
1836: "Close", "Skip zero"
1837:
1838:
1839: /* This macro defines the length of fixed length operations in the compiled
1840: regex. The lengths are used when searching for specific things, and also in the
1841: debugging printing of a compiled regex. We use a macro so that it can be
1842: defined close to the definitions of the opcodes themselves.
1843:
1844: As things have been extended, some of these are no longer fixed lenths, but are
1845: minima instead. For example, the length of a single-character repeat may vary
1846: in UTF-8 mode. The code that uses this table must know about such things. */
1847:
1848: #define OP_LENGTHS \
1849: 1, /* End */ \
1850: 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \
1851: 1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \
1852: 1, 1, 1, /* Any, AllAny, Anybyte */ \
1853: 3, 3, /* \P, \p */ \
1854: 1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \
1855: 1, /* \X */ \
1856: 1, 1, 1, 1, 1, 1, /* \Z, \z, ^, ^M, $, $M */ \
1857: 2, /* Char - the minimum length */ \
1858: 2, /* Chari - the minimum length */ \
1859: 2, /* not */ \
1860: 2, /* noti */ \
1861: /* Positive single-char repeats ** These are */ \
1862: 2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
1.1.1.2 ! misho 1863: 2+IMM2_SIZE, 2+IMM2_SIZE, /* upto, minupto ** mode */ \
! 1864: 2+IMM2_SIZE, /* exact */ \
! 1865: 2, 2, 2, 2+IMM2_SIZE, /* *+, ++, ?+, upto+ */ \
1.1 misho 1866: 2, 2, 2, 2, 2, 2, /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8 */ \
1.1.1.2 ! misho 1867: 2+IMM2_SIZE, 2+IMM2_SIZE, /* upto I, minupto I */ \
! 1868: 2+IMM2_SIZE, /* exact I */ \
! 1869: 2, 2, 2, 2+IMM2_SIZE, /* *+I, ++I, ?+I, upto+I */ \
1.1 misho 1870: /* Negative single-char repeats - only for chars < 256 */ \
1871: 2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
1.1.1.2 ! misho 1872: 2+IMM2_SIZE, 2+IMM2_SIZE, /* NOT upto, minupto */ \
! 1873: 2+IMM2_SIZE, /* NOT exact */ \
! 1874: 2, 2, 2, 2+IMM2_SIZE, /* Possessive NOT *, +, ?, upto */ \
1.1 misho 1875: 2, 2, 2, 2, 2, 2, /* NOT *I, *?I, +I, +?I, ?I, ??I */ \
1.1.1.2 ! misho 1876: 2+IMM2_SIZE, 2+IMM2_SIZE, /* NOT upto I, minupto I */ \
! 1877: 2+IMM2_SIZE, /* NOT exact I */ \
! 1878: 2, 2, 2, 2+IMM2_SIZE, /* Possessive NOT *I, +I, ?I, upto I */ \
1.1 misho 1879: /* Positive type repeats */ \
1880: 2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
1.1.1.2 ! misho 1881: 2+IMM2_SIZE, 2+IMM2_SIZE, /* Type upto, minupto */ \
! 1882: 2+IMM2_SIZE, /* Type exact */ \
! 1883: 2, 2, 2, 2+IMM2_SIZE, /* Possessive *+, ++, ?+, upto+ */ \
1.1 misho 1884: /* Character class & ref repeats */ \
1885: 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
1.1.1.2 ! misho 1886: 1+2*IMM2_SIZE, 1+2*IMM2_SIZE, /* CRRANGE, CRMINRANGE */ \
! 1887: 1+(32/sizeof(pcre_uchar)), /* CLASS */ \
! 1888: 1+(32/sizeof(pcre_uchar)), /* NCLASS */ \
1.1 misho 1889: 0, /* XCLASS - variable length */ \
1.1.1.2 ! misho 1890: 1+IMM2_SIZE, /* REF */ \
! 1891: 1+IMM2_SIZE, /* REFI */ \
1.1 misho 1892: 1+LINK_SIZE, /* RECURSE */ \
1893: 2+2*LINK_SIZE, /* CALLOUT */ \
1894: 1+LINK_SIZE, /* Alt */ \
1895: 1+LINK_SIZE, /* Ket */ \
1896: 1+LINK_SIZE, /* KetRmax */ \
1897: 1+LINK_SIZE, /* KetRmin */ \
1898: 1+LINK_SIZE, /* KetRpos */ \
1899: 1+LINK_SIZE, /* Reverse */ \
1900: 1+LINK_SIZE, /* Assert */ \
1901: 1+LINK_SIZE, /* Assert not */ \
1902: 1+LINK_SIZE, /* Assert behind */ \
1903: 1+LINK_SIZE, /* Assert behind not */ \
1904: 1+LINK_SIZE, /* ONCE */ \
1905: 1+LINK_SIZE, /* ONCE_NC */ \
1906: 1+LINK_SIZE, /* BRA */ \
1907: 1+LINK_SIZE, /* BRAPOS */ \
1.1.1.2 ! misho 1908: 1+LINK_SIZE+IMM2_SIZE, /* CBRA */ \
! 1909: 1+LINK_SIZE+IMM2_SIZE, /* CBRAPOS */ \
1.1 misho 1910: 1+LINK_SIZE, /* COND */ \
1911: 1+LINK_SIZE, /* SBRA */ \
1912: 1+LINK_SIZE, /* SBRAPOS */ \
1.1.1.2 ! misho 1913: 1+LINK_SIZE+IMM2_SIZE, /* SCBRA */ \
! 1914: 1+LINK_SIZE+IMM2_SIZE, /* SCBRAPOS */ \
1.1 misho 1915: 1+LINK_SIZE, /* SCOND */ \
1.1.1.2 ! misho 1916: 1+IMM2_SIZE, 1+IMM2_SIZE, /* CREF, NCREF */ \
! 1917: 1+IMM2_SIZE, 1+IMM2_SIZE, /* RREF, NRREF */ \
1.1 misho 1918: 1, /* DEF */ \
1919: 1, 1, 1, /* BRAZERO, BRAMINZERO, BRAPOSZERO */ \
1920: 3, 1, 3, /* MARK, PRUNE, PRUNE_ARG */ \
1921: 1, 3, /* SKIP, SKIP_ARG */ \
1922: 1, 3, /* THEN, THEN_ARG */ \
1923: 1, 1, 1, 1, /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT */ \
1.1.1.2 ! misho 1924: 1+IMM2_SIZE, 1 /* CLOSE, SKIPZERO */
1.1 misho 1925:
1926: /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
1927: condition. */
1928:
1929: #define RREF_ANY 0xffff
1930:
1931: /* Compile time error code numbers. They are given names so that they can more
1932: easily be tracked. When a new number is added, the table called eint in
1933: pcreposix.c must be updated. */
1934:
1935: enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
1936: ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
1937: ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
1938: ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
1939: ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
1940: ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
1941: ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
1.1.1.2 ! misho 1942: ERR70, ERR71, ERR72, ERR73, ERR74, ERRCOUNT };
1.1 misho 1943:
1944: /* The real format of the start of the pcre block; the index of names and the
1945: code vector run on as long as necessary after the end. We store an explicit
1946: offset to the name table so that if a regex is compiled on one host, saved, and
1947: then run on another where the size of pointers is different, all might still
1948: be well. For the case of compiled-on-4 and run-on-8, we include an extra
1949: pointer that is always NULL. For future-proofing, a few dummy fields were
1950: originally included - even though you can never get this planning right - but
1951: there is only one left now.
1952:
1953: NOTE NOTE NOTE:
1954: Because people can now save and re-use compiled patterns, any additions to this
1955: structure should be made at the end, and something earlier (e.g. a new
1956: flag in the options or one of the dummy fields) should indicate that the new
1957: fields are present. Currently PCRE always sets the dummy fields to zero.
1958: NOTE NOTE NOTE
1959: */
1960:
1.1.1.2 ! misho 1961: #ifdef COMPILE_PCRE8
! 1962: #define REAL_PCRE real_pcre
! 1963: #else
! 1964: #define REAL_PCRE real_pcre16
! 1965: #endif
! 1966:
! 1967: typedef struct REAL_PCRE {
1.1 misho 1968: pcre_uint32 magic_number;
1969: pcre_uint32 size; /* Total that was malloced */
1970: pcre_uint32 options; /* Public options */
1971: pcre_uint16 flags; /* Private flags */
1972: pcre_uint16 dummy1; /* For future use */
1973: pcre_uint16 top_bracket;
1974: pcre_uint16 top_backref;
1.1.1.2 ! misho 1975: pcre_uint16 first_char; /* Starting character */
! 1976: pcre_uint16 req_char; /* This character must be seen */
1.1 misho 1977: pcre_uint16 name_table_offset; /* Offset to name table that follows */
1978: pcre_uint16 name_entry_size; /* Size of any name items */
1979: pcre_uint16 name_count; /* Number of name items */
1980: pcre_uint16 ref_count; /* Reference count */
1981:
1.1.1.2 ! misho 1982: const pcre_uint8 *tables; /* Pointer to tables or NULL for std */
! 1983: const pcre_uint8 *nullpad; /* NULL padding */
! 1984: } REAL_PCRE;
1.1 misho 1985:
1986: /* The format of the block used to store data from pcre_study(). The same
1987: remark (see NOTE above) about extending this structure applies. */
1988:
1989: typedef struct pcre_study_data {
1990: pcre_uint32 size; /* Total that was malloced */
1991: pcre_uint32 flags; /* Private flags */
1.1.1.2 ! misho 1992: pcre_uint8 start_bits[32]; /* Starting char bits */
1.1 misho 1993: pcre_uint32 minlength; /* Minimum subject length */
1994: } pcre_study_data;
1995:
1996: /* Structure for building a chain of open capturing subpatterns during
1997: compiling, so that instructions to close them can be compiled when (*ACCEPT) is
1998: encountered. This is also used to identify subpatterns that contain recursive
1999: back references to themselves, so that they can be made atomic. */
2000:
2001: typedef struct open_capitem {
2002: struct open_capitem *next; /* Chain link */
2003: pcre_uint16 number; /* Capture number */
2004: pcre_uint16 flag; /* Set TRUE if recursive back ref */
2005: } open_capitem;
2006:
2007: /* Structure for passing "static" information around between the functions
2008: doing the compiling, so that they are thread-safe. */
2009:
2010: typedef struct compile_data {
1.1.1.2 ! misho 2011: const pcre_uint8 *lcc; /* Points to lower casing table */
! 2012: const pcre_uint8 *fcc; /* Points to case-flipping table */
! 2013: const pcre_uint8 *cbits; /* Points to character type table */
! 2014: const pcre_uint8 *ctypes; /* Points to table of type maps */
! 2015: const pcre_uchar *start_workspace;/* The start of working space */
! 2016: const pcre_uchar *start_code; /* The start of the compiled code */
! 2017: const pcre_uchar *start_pattern; /* The start of the pattern */
! 2018: const pcre_uchar *end_pattern; /* The end of the pattern */
! 2019: open_capitem *open_caps; /* Chain of open capture items */
! 2020: pcre_uchar *hwm; /* High watermark of workspace */
! 2021: pcre_uchar *name_table; /* The name/number table */
! 2022: int names_found; /* Number of entries so far */
! 2023: int name_entry_size; /* Size of each entry */
! 2024: int workspace_size; /* Size of workspace */
! 2025: int bracount; /* Count of capturing parens as we compile */
! 2026: int final_bracount; /* Saved value after first pass */
! 2027: int top_backref; /* Maximum back reference */
! 2028: unsigned int backref_map; /* Bitmap of low back refs */
! 2029: int assert_depth; /* Depth of nested assertions */
! 2030: int external_options; /* External (initial) options */
! 2031: int external_flags; /* External flag bits to be set */
! 2032: int req_varyopt; /* "After variable item" flag for reqbyte */
! 2033: BOOL had_accept; /* (*ACCEPT) encountered */
! 2034: BOOL check_lookbehind; /* Lookbehinds need later checking */
! 2035: int nltype; /* Newline type */
! 2036: int nllen; /* Newline string length */
! 2037: pcre_uchar nl[4]; /* Newline string when fixed length */
1.1 misho 2038: } compile_data;
2039:
2040: /* Structure for maintaining a chain of pointers to the currently incomplete
2041: branches, for testing for left recursion while compiling. */
2042:
2043: typedef struct branch_chain {
2044: struct branch_chain *outer;
1.1.1.2 ! misho 2045: pcre_uchar *current_branch;
1.1 misho 2046: } branch_chain;
2047:
2048: /* Structure for items in a linked list that represents an explicit recursive
2049: call within the pattern; used by pcre_exec(). */
2050:
2051: typedef struct recursion_info {
2052: struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2053: int group_num; /* Number of group that was called */
2054: int *offset_save; /* Pointer to start of saved offsets */
2055: int saved_max; /* Number of saved offsets */
1.1.1.2 ! misho 2056: PCRE_PUCHAR subject_position; /* Position at start of recursion */
1.1 misho 2057: } recursion_info;
2058:
2059: /* A similar structure for pcre_dfa_exec(). */
2060:
2061: typedef struct dfa_recursion_info {
2062: struct dfa_recursion_info *prevrec;
2063: int group_num;
1.1.1.2 ! misho 2064: PCRE_PUCHAR subject_position;
1.1 misho 2065: } dfa_recursion_info;
2066:
2067: /* Structure for building a chain of data for holding the values of the subject
2068: pointer at the start of each subpattern, so as to detect when an empty string
2069: has been matched by a subpattern - to break infinite loops; used by
2070: pcre_exec(). */
2071:
2072: typedef struct eptrblock {
2073: struct eptrblock *epb_prev;
1.1.1.2 ! misho 2074: PCRE_PUCHAR epb_saved_eptr;
1.1 misho 2075: } eptrblock;
2076:
2077:
2078: /* Structure for passing "static" information around between the functions
2079: doing traditional NFA matching, so that they are thread-safe. */
2080:
2081: typedef struct match_data {
2082: unsigned long int match_call_count; /* As it says */
2083: unsigned long int match_limit; /* As it says */
2084: unsigned long int match_limit_recursion; /* As it says */
1.1.1.2 ! misho 2085: int *offset_vector; /* Offset vector */
! 2086: int offset_end; /* One past the end */
! 2087: int offset_max; /* The maximum usable for return data */
! 2088: int nltype; /* Newline type */
! 2089: int nllen; /* Newline string length */
! 2090: int name_count; /* Number of names in name table */
! 2091: int name_entry_size; /* Size of entry in names table */
! 2092: pcre_uchar *name_table; /* Table of names */
! 2093: pcre_uchar nl[4]; /* Newline string when fixed */
! 2094: const pcre_uint8 *lcc; /* Points to lower casing table */
! 2095: const pcre_uint8 *fcc; /* Points to case-flipping table */
! 2096: const pcre_uint8 *ctypes; /* Points to table of type maps */
! 2097: BOOL offset_overflow; /* Set if too many extractions */
! 2098: BOOL notbol; /* NOTBOL flag */
! 2099: BOOL noteol; /* NOTEOL flag */
! 2100: BOOL utf; /* UTF-8 / UTF-16 flag */
! 2101: BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */
! 2102: BOOL use_ucp; /* PCRE_UCP flag */
! 2103: BOOL endonly; /* Dollar not before final \n */
! 2104: BOOL notempty; /* Empty string match not wanted */
! 2105: BOOL notempty_atstart; /* Empty string match at start not wanted */
! 2106: BOOL hitend; /* Hit the end of the subject at some point */
! 2107: BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
! 2108: BOOL hasthen; /* Pattern contains (*THEN) */
! 2109: BOOL ignore_skip_arg; /* For re-run when SKIP name not found */
! 2110: const pcre_uchar *start_code; /* For use when recursing */
! 2111: PCRE_PUCHAR start_subject; /* Start of the subject string */
! 2112: PCRE_PUCHAR end_subject; /* End of the subject string */
! 2113: PCRE_PUCHAR start_match_ptr; /* Start of matched string */
! 2114: PCRE_PUCHAR end_match_ptr; /* Subject position at end match */
! 2115: PCRE_PUCHAR start_used_ptr; /* Earliest consulted character */
! 2116: int partial; /* PARTIAL options */
! 2117: int end_offset_top; /* Highwater mark at end of match */
! 2118: int capture_last; /* Most recent capture number */
! 2119: int start_offset; /* The start offset value */
! 2120: int match_function_type; /* Set for certain special calls of MATCH() */
! 2121: eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */
! 2122: int eptrn; /* Next free eptrblock */
! 2123: recursion_info *recursive; /* Linked list of recursion data */
! 2124: void *callout_data; /* To pass back to callouts */
! 2125: const pcre_uchar *mark; /* Mark pointer to pass back on success */
! 2126: const pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
! 2127: const pcre_uchar *once_target; /* Where to back up to for atomic groups */
1.1 misho 2128: } match_data;
2129:
2130: /* A similar structure is used for the same purpose by the DFA matching
2131: functions. */
2132:
2133: typedef struct dfa_match_data {
1.1.1.2 ! misho 2134: const pcre_uchar *start_code; /* Start of the compiled pattern */
! 2135: const pcre_uchar *start_subject ; /* Start of the subject string */
! 2136: const pcre_uchar *end_subject; /* End of subject string */
! 2137: const pcre_uchar *start_used_ptr; /* Earliest consulted character */
! 2138: const pcre_uint8 *tables; /* Character tables */
! 2139: int start_offset; /* The start offset value */
! 2140: int moptions; /* Match options */
! 2141: int poptions; /* Pattern options */
! 2142: int nltype; /* Newline type */
! 2143: int nllen; /* Newline string length */
! 2144: pcre_uchar nl[4]; /* Newline string when fixed */
! 2145: void *callout_data; /* To pass back to callouts */
! 2146: dfa_recursion_info *recursive; /* Linked list of recursion data */
1.1 misho 2147: } dfa_match_data;
2148:
2149: /* Bit definitions for entries in the pcre_ctypes table. */
2150:
2151: #define ctype_space 0x01
2152: #define ctype_letter 0x02
2153: #define ctype_digit 0x04
2154: #define ctype_xdigit 0x08
2155: #define ctype_word 0x10 /* alphanumeric or '_' */
2156: #define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
2157:
2158: /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
2159: of bits for a class map. Some classes are built by combining these tables. */
2160:
2161: #define cbit_space 0 /* [:space:] or \s */
2162: #define cbit_xdigit 32 /* [:xdigit:] */
2163: #define cbit_digit 64 /* [:digit:] or \d */
2164: #define cbit_upper 96 /* [:upper:] */
2165: #define cbit_lower 128 /* [:lower:] */
2166: #define cbit_word 160 /* [:word:] or \w */
2167: #define cbit_graph 192 /* [:graph:] */
2168: #define cbit_print 224 /* [:print:] */
2169: #define cbit_punct 256 /* [:punct:] */
2170: #define cbit_cntrl 288 /* [:cntrl:] */
2171: #define cbit_length 320 /* Length of the cbits table */
2172:
2173: /* Offsets of the various tables from the base tables pointer, and
2174: total length. */
2175:
2176: #define lcc_offset 0
2177: #define fcc_offset 256
2178: #define cbits_offset 512
2179: #define ctypes_offset (cbits_offset + cbit_length)
2180: #define tables_length (ctypes_offset + 256)
2181:
1.1.1.2 ! misho 2182: /* Internal function prefix */
! 2183:
! 2184: #ifdef COMPILE_PCRE8
! 2185: #ifndef PUBL
! 2186: #define PUBL(name) pcre_##name
! 2187: #endif
! 2188: #ifndef PRIV
! 2189: #define PRIV(name) _pcre_##name
! 2190: #endif
! 2191: #else /* COMPILE_PCRE8 */
! 2192: #ifdef COMPILE_PCRE16
! 2193: #ifndef PUBL
! 2194: #define PUBL(name) pcre16_##name
! 2195: #endif
! 2196: #ifndef PRIV
! 2197: #define PRIV(name) _pcre16_##name
! 2198: #endif
! 2199: #else
! 2200: #error Unsupported compiling mode
! 2201: #endif /* COMPILE_PCRE16 */
! 2202: #endif /* COMPILE_PCRE8 */
! 2203:
1.1 misho 2204: /* Layout of the UCP type table that translates property names into types and
2205: codes. Each entry used to point directly to a name, but to reduce the number of
2206: relocations in shared libraries, it now has an offset into a single string
2207: instead. */
2208:
2209: typedef struct {
2210: pcre_uint16 name_offset;
2211: pcre_uint16 type;
2212: pcre_uint16 value;
2213: } ucp_type_table;
2214:
2215:
2216: /* Internal shared data tables. These are tables that are used by more than one
2217: of the exported public functions. They have to be "external" in the C sense,
2218: but are not part of the PCRE public API. The data for these tables is in the
2219: pcre_tables.c module. */
2220:
1.1.1.2 ! misho 2221: #ifdef COMPILE_PCRE8
1.1 misho 2222:
1.1.1.2 ! misho 2223: extern const int PRIV(utf8_table1)[];
! 2224: extern const int PRIV(utf8_table1_size);
! 2225: extern const int PRIV(utf8_table2)[];
! 2226: extern const int PRIV(utf8_table3)[];
! 2227: extern const pcre_uint8 PRIV(utf8_table4)[];
1.1 misho 2228:
1.1.1.2 ! misho 2229: #endif /* COMPILE_PCRE8 */
1.1 misho 2230:
1.1.1.2 ! misho 2231: extern const char PRIV(utt_names)[];
! 2232: extern const ucp_type_table PRIV(utt)[];
! 2233: extern const int PRIV(utt_size);
1.1 misho 2234:
1.1.1.2 ! misho 2235: extern const pcre_uint8 PRIV(default_tables)[];
1.1 misho 2236:
1.1.1.2 ! misho 2237: extern const pcre_uint8 PRIV(OP_lengths)[];
1.1 misho 2238:
2239:
2240: /* Internal shared functions. These are functions that are used by more than
2241: one of the exported public functions. They have to be "external" in the C
2242: sense, but are not part of the PCRE public API. */
2243:
1.1.1.2 ! misho 2244: /* String comparison functions. */
! 2245: #ifdef COMPILE_PCRE8
! 2246:
! 2247: #define STRCMP_UC_UC(str1, str2) \
! 2248: strcmp((char *)(str1), (char *)(str2))
! 2249: #define STRCMP_UC_C8(str1, str2) \
! 2250: strcmp((char *)(str1), (str2))
! 2251: #define STRNCMP_UC_UC(str1, str2, num) \
! 2252: strncmp((char *)(str1), (char *)(str2), (num))
! 2253: #define STRNCMP_UC_C8(str1, str2, num) \
! 2254: strncmp((char *)(str1), (str2), (num))
! 2255: #define STRLEN_UC(str) strlen((const char *)str)
! 2256:
! 2257: #else
! 2258:
! 2259: extern int PRIV(strcmp_uc_uc)(const pcre_uchar *,
! 2260: const pcre_uchar *);
! 2261: extern int PRIV(strcmp_uc_c8)(const pcre_uchar *,
! 2262: const char *);
! 2263: extern int PRIV(strncmp_uc_uc)(const pcre_uchar *,
! 2264: const pcre_uchar *, unsigned int num);
! 2265: extern int PRIV(strncmp_uc_c8)(const pcre_uchar *,
! 2266: const char *, unsigned int num);
! 2267: extern unsigned int PRIV(strlen_uc)(const pcre_uchar *str);
! 2268:
! 2269: #define STRCMP_UC_UC(str1, str2) \
! 2270: PRIV(strcmp_uc_uc)((str1), (str2))
! 2271: #define STRCMP_UC_C8(str1, str2) \
! 2272: PRIV(strcmp_uc_c8)((str1), (str2))
! 2273: #define STRNCMP_UC_UC(str1, str2, num) \
! 2274: PRIV(strncmp_uc_uc)((str1), (str2), (num))
! 2275: #define STRNCMP_UC_C8(str1, str2, num) \
! 2276: PRIV(strncmp_uc_c8)((str1), (str2), (num))
! 2277: #define STRLEN_UC(str) PRIV(strlen_uc)(str)
! 2278:
! 2279: #endif /* COMPILE_PCRE8 */
! 2280:
! 2281: extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
! 2282: extern BOOL PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
! 2283: int *, BOOL);
! 2284: extern int PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
! 2285: extern int PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
! 2286: extern BOOL PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
! 2287: int *, BOOL);
! 2288: extern BOOL PRIV(xclass)(int, const pcre_uchar *, BOOL);
1.1 misho 2289:
2290: #ifdef SUPPORT_JIT
1.1.1.2 ! misho 2291: extern void PRIV(jit_compile)(const REAL_PCRE *, PUBL(extra) *);
! 2292: extern int PRIV(jit_exec)(const REAL_PCRE *, void *,
! 2293: const pcre_uchar *, int, int, int, int, int *, int);
! 2294: extern void PRIV(jit_free)(void *);
! 2295: extern int PRIV(jit_get_size)(void *);
! 2296: extern const char* PRIV(jit_get_target)(void);
1.1 misho 2297: #endif
2298:
2299: /* Unicode character database (UCD) */
2300:
2301: typedef struct {
1.1.1.2 ! misho 2302: pcre_uint8 script;
! 2303: pcre_uint8 chartype;
1.1 misho 2304: pcre_int32 other_case;
2305: } ucd_record;
2306:
1.1.1.2 ! misho 2307: extern const ucd_record PRIV(ucd_records)[];
! 2308: extern const pcre_uint8 PRIV(ucd_stage1)[];
! 2309: extern const pcre_uint16 PRIV(ucd_stage2)[];
! 2310: extern const int PRIV(ucp_gentype)[];
1.1 misho 2311: #ifdef SUPPORT_JIT
1.1.1.2 ! misho 2312: extern const int PRIV(ucp_typerange)[];
1.1 misho 2313: #endif
2314:
1.1.1.2 ! misho 2315: #ifdef SUPPORT_UCP
1.1 misho 2316: /* UCD access macros */
2317:
2318: #define UCD_BLOCK_SIZE 128
1.1.1.2 ! misho 2319: #define GET_UCD(ch) (PRIV(ucd_records) + \
! 2320: PRIV(ucd_stage2)[PRIV(ucd_stage1)[(ch) / UCD_BLOCK_SIZE] * \
1.1 misho 2321: UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])
2322:
2323: #define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype
2324: #define UCD_SCRIPT(ch) GET_UCD(ch)->script
1.1.1.2 ! misho 2325: #define UCD_CATEGORY(ch) PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
1.1 misho 2326: #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)
2327:
1.1.1.2 ! misho 2328: #endif /* SUPPORT_UCP */
! 2329:
1.1 misho 2330: #endif
2331:
2332: /* End of pcre_internal.h */
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>