1/************************************************* 2* Perl-Compatible Regular Expressions * 3*************************************************/ 4 5 6/* PCRE is a library of functions to support regular expressions whose syntax 7and semantics are as close as possible to those of the Perl 5 language. 8 9 Written by Philip Hazel 10 Copyright (c) 1997-2010 University of Cambridge 11 12----------------------------------------------------------------------------- 13Redistribution and use in source and binary forms, with or without 14modification, 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 27THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" 28AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 29IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 30ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE 31LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 32CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 33SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 34INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 35CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 36ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 37POSSIBILITY OF SUCH DAMAGE. 38----------------------------------------------------------------------------- 39*/ 40 41/* This header contains definitions that are shared between the different 42modules, but which are not relevant to the exported API. This includes some 43functions whose names all begin with "_pcre_". */ 44 45#ifndef PCRE_INTERNAL_H 46#define PCRE_INTERNAL_H 47 48/* Define PCRE_DEBUG to get debugging output on stdout. */ 49 50#if 0 51#define PCRE_DEBUG 52#endif 53 54/* We do not support both EBCDIC and UTF-8 at the same time. The "configure" 55script prevents both being selected, but not everybody uses "configure". */ 56 57#if defined EBCDIC && defined SUPPORT_UTF8 58#error The use of both EBCDIC and SUPPORT_UTF8 is not supported. 59#endif 60 61/* If SUPPORT_UCP is defined, SUPPORT_UTF8 must also be defined. The 62"configure" script ensures this, but not everybody uses "configure". */ 63 64#if defined SUPPORT_UCP && !defined SUPPORT_UTF8 65#define SUPPORT_UTF8 1 66#endif 67 68/* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef 69inline, and there are *still* stupid compilers about that don't like indented 70pre-processor statements, or at least there were when I first wrote this. After 71all, it had only been about 10 years then... 72 73It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so 74be absolutely sure we get our version. */ 75 76#undef DPRINTF 77#ifdef PCRE_DEBUG 78#define DPRINTF(p) printf p 79#else 80#define DPRINTF(p) /* Nothing */ 81#endif 82 83 84/* Standard C headers plus the external interface definition. The only time 85setjmp and stdarg are used is when NO_RECURSE is set. */ 86 87#include <ctype.h> 88#include <limits.h> 89#include <stddef.h> 90#include <stdio.h> 91#include <stdlib.h> 92#include <string.h> 93 94/* When compiling a DLL for Windows, the exported symbols have to be declared 95using some MS magic. I found some useful information on this web page: 96http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the 97information there, using __declspec(dllexport) without "extern" we have a 98definition; with "extern" we have a declaration. The settings here override the 99setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL, 100which is all that is needed for applications (they just import the symbols). We 101use: 102 103 PCRE_EXP_DECL for declarations 104 PCRE_EXP_DEFN for definitions of exported functions 105 PCRE_EXP_DATA_DEFN for definitions of exported variables 106 107The reason for the two DEFN macros is that in non-Windows environments, one 108does not want to have "extern" before variable definitions because it leads to 109compiler warnings. So we distinguish between functions and variables. In 110Windows, the two should always be the same. 111 112The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest, 113which is an application, but needs to import this file in order to "peek" at 114internals, can #include pcre.h first to get an application's-eye view. 115 116In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon, 117special-purpose environments) might want to stick other stuff in front of 118exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and 119PCRE_EXP_DATA_DEFN only if they are not already set. */ 120 121#ifndef PCRE_EXP_DECL 122# ifdef _WIN32 123# ifndef PCRE_STATIC 124# define PCRE_EXP_DECL extern __declspec(dllexport) 125# define PCRE_EXP_DEFN __declspec(dllexport) 126# define PCRE_EXP_DATA_DEFN __declspec(dllexport) 127# else 128# define PCRE_EXP_DECL extern 129# define PCRE_EXP_DEFN 130# define PCRE_EXP_DATA_DEFN 131# endif 132# else 133# ifdef __cplusplus 134# define PCRE_EXP_DECL extern "C" 135# else 136# define PCRE_EXP_DECL extern 137# endif 138# ifndef PCRE_EXP_DEFN 139# define PCRE_EXP_DEFN PCRE_EXP_DECL 140# endif 141# ifndef PCRE_EXP_DATA_DEFN 142# define PCRE_EXP_DATA_DEFN 143# endif 144# endif 145#endif 146 147/* When compiling with the MSVC compiler, it is sometimes necessary to include 148a "calling convention" before exported function names. (This is secondhand 149information; I know nothing about MSVC myself). For example, something like 150 151 void __cdecl function(....) 152 153might be needed. In order so make this easy, all the exported functions have 154PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not 155set, we ensure here that it has no effect. */ 156 157#ifndef PCRE_CALL_CONVENTION 158#define PCRE_CALL_CONVENTION 159#endif 160 161/* We need to have types that specify unsigned 16-bit and 32-bit integers. We 162cannot determine these outside the compilation (e.g. by running a program as 163part of "configure") because PCRE is often cross-compiled for use on other 164systems. Instead we make use of the maximum sizes that are available at 165preprocessor time in standard C environments. */ 166 167#if USHRT_MAX == 65535 168 typedef unsigned short pcre_uint16; 169 typedef short pcre_int16; 170#elif UINT_MAX == 65535 171 typedef unsigned int pcre_uint16; 172 typedef int pcre_int16; 173#else 174 #error Cannot determine a type for 16-bit unsigned integers 175#endif 176 177#if UINT_MAX == 4294967295 178 typedef unsigned int pcre_uint32; 179 typedef int pcre_int32; 180#elif ULONG_MAX == 4294967295 181 typedef unsigned long int pcre_uint32; 182 typedef long int pcre_int32; 183#else 184 #error Cannot determine a type for 32-bit unsigned integers 185#endif 186 187/* When checking for integer overflow in pcre_compile(), we need to handle 188large integers. If a 64-bit integer type is available, we can use that. 189Otherwise we have to cast to double, which of course requires floating point 190arithmetic. Handle this by defining a macro for the appropriate type. If 191stdint.h is available, include it; it may define INT64_MAX. Systems that do not 192have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set 193by "configure". */ 194 195#if HAVE_STDINT_H 196#include <stdint.h> 197#elif HAVE_INTTYPES_H 198#include <inttypes.h> 199#endif 200 201#if defined INT64_MAX || defined int64_t 202#define INT64_OR_DOUBLE int64_t 203#else 204#define INT64_OR_DOUBLE double 205#endif 206 207/* All character handling must be done as unsigned characters. Otherwise there 208are problems with top-bit-set characters and functions such as isspace(). 209However, we leave the interface to the outside world as char *, because that 210should make things easier for callers. We define a short type for unsigned char 211to save lots of typing. I tried "uchar", but it causes problems on Digital 212Unix, where it is defined in sys/types, so use "uschar" instead. */ 213 214typedef unsigned char uschar; 215 216/* This is an unsigned int value that no character can ever have. UTF-8 217characters only go up to 0x7fffffff (though Unicode doesn't go beyond 2180x0010ffff). */ 219 220#define NOTACHAR 0xffffffff 221 222/* PCRE is able to support several different kinds of newline (CR, LF, CRLF, 223"any" and "anycrlf" at present). The following macros are used to package up 224testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various 225modules to indicate in which datablock the parameters exist, and what the 226start/end of string field names are. */ 227 228#define NLTYPE_FIXED 0 /* Newline is a fixed length string */ 229#define NLTYPE_ANY 1 /* Newline is any Unicode line ending */ 230#define NLTYPE_ANYCRLF 2 /* Newline is CR, LF, or CRLF */ 231 232/* This macro checks for a newline at the given position */ 233 234#define IS_NEWLINE(p) \ 235 ((NLBLOCK->nltype != NLTYPE_FIXED)? \ 236 ((p) < NLBLOCK->PSEND && \ 237 _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\ 238 utf8)) \ 239 : \ 240 ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \ 241 (p)[0] == NLBLOCK->nl[0] && \ 242 (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \ 243 ) \ 244 ) 245 246/* This macro checks for a newline immediately preceding the given position */ 247 248#define WAS_NEWLINE(p) \ 249 ((NLBLOCK->nltype != NLTYPE_FIXED)? \ 250 ((p) > NLBLOCK->PSSTART && \ 251 _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \ 252 &(NLBLOCK->nllen), utf8)) \ 253 : \ 254 ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \ 255 (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \ 256 (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \ 257 ) \ 258 ) 259 260/* When PCRE is compiled as a C++ library, the subject pointer can be replaced 261with a custom type. This makes it possible, for example, to allow pcre_exec() 262to process subject strings that are discontinuous by using a smart pointer 263class. It must always be possible to inspect all of the subject string in 264pcre_exec() because of the way it backtracks. Two macros are required in the 265normal case, for sign-unspecified and unsigned char pointers. The former is 266used for the external interface and appears in pcre.h, which is why its name 267must begin with PCRE_. */ 268 269#ifdef CUSTOM_SUBJECT_PTR 270#define PCRE_SPTR CUSTOM_SUBJECT_PTR 271#define USPTR CUSTOM_SUBJECT_PTR 272#else 273#define PCRE_SPTR const char * 274#define USPTR const unsigned char * 275#endif 276 277 278 279/* Include the public PCRE header and the definitions of UCP character property 280values. */ 281 282#include "pcre.h" 283#include "ucp.h" 284 285/* When compiling for use with the Virtual Pascal compiler, these functions 286need to have their names changed. PCRE must be compiled with the -DVPCOMPAT 287option on the command line. */ 288 289#ifdef VPCOMPAT 290#define strlen(s) _strlen(s) 291#define strncmp(s1,s2,m) _strncmp(s1,s2,m) 292#define memcmp(s,c,n) _memcmp(s,c,n) 293#define memcpy(d,s,n) _memcpy(d,s,n) 294#define memmove(d,s,n) _memmove(d,s,n) 295#define memset(s,c,n) _memset(s,c,n) 296#else /* VPCOMPAT */ 297 298/* To cope with SunOS4 and other systems that lack memmove() but have bcopy(), 299define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY 300is set. Otherwise, include an emulating function for those systems that have 301neither (there some non-Unix environments where this is the case). */ 302 303#ifndef HAVE_MEMMOVE 304#undef memmove /* some systems may have a macro */ 305#ifdef HAVE_BCOPY 306#define memmove(a, b, c) bcopy(b, a, c) 307#else /* HAVE_BCOPY */ 308static void * 309pcre_memmove(void *d, const void *s, size_t n) 310{ 311size_t i; 312unsigned char *dest = (unsigned char *)d; 313const unsigned char *src = (const unsigned char *)s; 314if (dest > src) 315 { 316 dest += n; 317 src += n; 318 for (i = 0; i < n; ++i) *(--dest) = *(--src); 319 return (void *)dest; 320 } 321else 322 { 323 for (i = 0; i < n; ++i) *dest++ = *src++; 324 return (void *)(dest - n); 325 } 326} 327#define memmove(a, b, c) pcre_memmove(a, b, c) 328#endif /* not HAVE_BCOPY */ 329#endif /* not HAVE_MEMMOVE */ 330#endif /* not VPCOMPAT */ 331 332 333/* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored 334in big-endian order) by default. These are used, for example, to link from the 335start of a subpattern to its alternatives and its end. The use of 2 bytes per 336offset limits the size of the compiled regex to around 64K, which is big enough 337for almost everybody. However, I received a request for an even bigger limit. 338For this reason, and also to make the code easier to maintain, the storing and 339loading of offsets from the byte string is now handled by the macros that are 340defined here. 341 342The macros are controlled by the value of LINK_SIZE. This defaults to 2 in 343the config.h file, but can be overridden by using -D on the command line. This 344is automated on Unix systems via the "configure" command. */ 345 346#if LINK_SIZE == 2 347 348#define PUT(a,n,d) \ 349 (a[n] = (d) >> 8), \ 350 (a[(n)+1] = (d) & 255) 351 352#define GET(a,n) \ 353 (((a)[n] << 8) | (a)[(n)+1]) 354 355#define MAX_PATTERN_SIZE (1 << 16) 356 357 358#elif LINK_SIZE == 3 359 360#define PUT(a,n,d) \ 361 (a[n] = (d) >> 16), \ 362 (a[(n)+1] = (d) >> 8), \ 363 (a[(n)+2] = (d) & 255) 364 365#define GET(a,n) \ 366 (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2]) 367 368#define MAX_PATTERN_SIZE (1 << 24) 369 370 371#elif LINK_SIZE == 4 372 373#define PUT(a,n,d) \ 374 (a[n] = (d) >> 24), \ 375 (a[(n)+1] = (d) >> 16), \ 376 (a[(n)+2] = (d) >> 8), \ 377 (a[(n)+3] = (d) & 255) 378 379#define GET(a,n) \ 380 (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3]) 381 382#define MAX_PATTERN_SIZE (1 << 30) /* Keep it positive */ 383 384 385#else 386#error LINK_SIZE must be either 2, 3, or 4 387#endif 388 389 390/* Convenience macro defined in terms of the others */ 391 392#define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE 393 394 395/* PCRE uses some other 2-byte quantities that do not change when the size of 396offsets changes. There are used for repeat counts and for other things such as 397capturing parenthesis numbers in back references. */ 398 399#define PUT2(a,n,d) \ 400 a[n] = (d) >> 8; \ 401 a[(n)+1] = (d) & 255 402 403#define GET2(a,n) \ 404 (((a)[n] << 8) | (a)[(n)+1]) 405 406#define PUT2INC(a,n,d) PUT2(a,n,d), a += 2 407 408 409/* When UTF-8 encoding is being used, a character is no longer just a single 410byte. The macros for character handling generate simple sequences when used in 411byte-mode, and more complicated ones for UTF-8 characters. BACKCHAR should 412never be called in byte mode. To make sure it can never even appear when UTF-8 413support is omitted, we don't even define it. */ 414 415#ifndef SUPPORT_UTF8 416#define GETCHAR(c, eptr) c = *eptr; 417#define GETCHARTEST(c, eptr) c = *eptr; 418#define GETCHARINC(c, eptr) c = *eptr++; 419#define GETCHARINCTEST(c, eptr) c = *eptr++; 420#define GETCHARLEN(c, eptr, len) c = *eptr; 421/* #define BACKCHAR(eptr) */ 422 423#else /* SUPPORT_UTF8 */ 424 425/* Get the next UTF-8 character, not advancing the pointer. This is called when 426we know we are in UTF-8 mode. */ 427 428#define GETCHAR(c, eptr) \ 429 c = *eptr; \ 430 if (c >= 0xc0) \ 431 { \ 432 int gcii; \ 433 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ 434 int gcss = 6*gcaa; \ 435 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \ 436 for (gcii = 1; gcii <= gcaa; gcii++) \ 437 { \ 438 gcss -= 6; \ 439 c |= (eptr[gcii] & 0x3f) << gcss; \ 440 } \ 441 } 442 443/* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the 444pointer. */ 445 446#define GETCHARTEST(c, eptr) \ 447 c = *eptr; \ 448 if (utf8 && c >= 0xc0) \ 449 { \ 450 int gcii; \ 451 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ 452 int gcss = 6*gcaa; \ 453 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \ 454 for (gcii = 1; gcii <= gcaa; gcii++) \ 455 { \ 456 gcss -= 6; \ 457 c |= (eptr[gcii] & 0x3f) << gcss; \ 458 } \ 459 } 460 461/* Get the next UTF-8 character, advancing the pointer. This is called when we 462know we are in UTF-8 mode. */ 463 464#define GETCHARINC(c, eptr) \ 465 c = *eptr++; \ 466 if (c >= 0xc0) \ 467 { \ 468 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ 469 int gcss = 6*gcaa; \ 470 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \ 471 while (gcaa-- > 0) \ 472 { \ 473 gcss -= 6; \ 474 c |= (*eptr++ & 0x3f) << gcss; \ 475 } \ 476 } 477 478/* Get the next character, testing for UTF-8 mode, and advancing the pointer */ 479 480#define GETCHARINCTEST(c, eptr) \ 481 c = *eptr++; \ 482 if (utf8 && c >= 0xc0) \ 483 { \ 484 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ 485 int gcss = 6*gcaa; \ 486 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \ 487 while (gcaa-- > 0) \ 488 { \ 489 gcss -= 6; \ 490 c |= (*eptr++ & 0x3f) << gcss; \ 491 } \ 492 } 493 494/* Get the next UTF-8 character, not advancing the pointer, incrementing length 495if there are extra bytes. This is called when we know we are in UTF-8 mode. */ 496 497#define GETCHARLEN(c, eptr, len) \ 498 c = *eptr; \ 499 if (c >= 0xc0) \ 500 { \ 501 int gcii; \ 502 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ 503 int gcss = 6*gcaa; \ 504 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \ 505 for (gcii = 1; gcii <= gcaa; gcii++) \ 506 { \ 507 gcss -= 6; \ 508 c |= (eptr[gcii] & 0x3f) << gcss; \ 509 } \ 510 len += gcaa; \ 511 } 512 513/* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the 514pointer, incrementing length if there are extra bytes. This is called when we 515know we are in UTF-8 mode. */ 516 517#define GETCHARLENTEST(c, eptr, len) \ 518 c = *eptr; \ 519 if (utf8 && c >= 0xc0) \ 520 { \ 521 int gcii; \ 522 int gcaa = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ \ 523 int gcss = 6*gcaa; \ 524 c = (c & _pcre_utf8_table3[gcaa]) << gcss; \ 525 for (gcii = 1; gcii <= gcaa; gcii++) \ 526 { \ 527 gcss -= 6; \ 528 c |= (eptr[gcii] & 0x3f) << gcss; \ 529 } \ 530 len += gcaa; \ 531 } 532 533/* If the pointer is not at the start of a character, move it back until 534it is. This is called only in UTF-8 mode - we don't put a test within the macro 535because almost all calls are already within a block of UTF-8 only code. */ 536 537#define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr-- 538 539#endif 540 541 542/* In case there is no definition of offsetof() provided - though any proper 543Standard C system should have one. */ 544 545#ifndef offsetof 546#define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field)) 547#endif 548 549 550/* These are the public options that can change during matching. */ 551 552#define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL) 553 554/* Private flags containing information about the compiled regex. They used to 555live at the top end of the options word, but that got almost full, so now they 556are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as 557the restrictions on partial matching have been lifted. It remains for backwards 558compatibility. */ 559 560#define PCRE_NOPARTIAL 0x0001 /* can't use partial with this regex */ 561#define PCRE_FIRSTSET 0x0002 /* first_byte is set */ 562#define PCRE_REQCHSET 0x0004 /* req_byte is set */ 563#define PCRE_STARTLINE 0x0008 /* start after \n for multiline */ 564#define PCRE_JCHANGED 0x0010 /* j option used in regex */ 565#define PCRE_HASCRORLF 0x0020 /* explicit \r or \n in pattern */ 566 567/* Options for the "extra" block produced by pcre_study(). */ 568 569#define PCRE_STUDY_MAPPED 0x01 /* a map of starting chars exists */ 570#define PCRE_STUDY_MINLEN 0x02 /* a minimum length field exists */ 571 572/* Masks for identifying the public options that are permitted at compile 573time, run time, or study time, respectively. */ 574 575#define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \ 576 PCRE_NEWLINE_ANYCRLF) 577 578#define PUBLIC_COMPILE_OPTIONS \ 579 (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \ 580 PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \ 581 PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \ 582 PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \ 583 PCRE_JAVASCRIPT_COMPAT) 584 585#define PUBLIC_EXEC_OPTIONS \ 586 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \ 587 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \ 588 PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE) 589 590#define PUBLIC_DFA_EXEC_OPTIONS \ 591 (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \ 592 PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \ 593 PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \ 594 PCRE_NO_START_OPTIMIZE) 595 596#define PUBLIC_STUDY_OPTIONS 0 /* None defined */ 597 598/* Magic number to provide a small check against being handed junk. Also used 599to detect whether a pattern was compiled on a host of different endianness. */ 600 601#define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */ 602 603/* Negative values for the firstchar and reqchar variables */ 604 605#define REQ_UNSET (-2) 606#define REQ_NONE (-1) 607 608/* The maximum remaining length of subject we are prepared to search for a 609req_byte match. */ 610 611#define REQ_BYTE_MAX 1000 612 613/* Flags added to firstbyte or reqbyte; a "non-literal" item is either a 614variable-length repeat, or a anything other than literal characters. */ 615 616#define REQ_CASELESS 0x0100 /* indicates caselessness */ 617#define REQ_VARY 0x0200 /* reqbyte followed non-literal item */ 618 619/* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in 620environments where these macros are defined elsewhere. Unfortunately, there 621is no way to do the same for the typedef. */ 622 623typedef int BOOL; 624 625#ifndef FALSE 626#define FALSE 0 627#define TRUE 1 628#endif 629 630/* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal 631character constants like '*' because the compiler would emit their EBCDIC code, 632which is different from their ASCII/UTF-8 code. Instead we define macros for 633the characters so that they always use the ASCII/UTF-8 code when UTF-8 support 634is enabled. When UTF-8 support is not enabled, the definitions use character 635literals. Both character and string versions of each character are needed, and 636there are some longer strings as well. 637 638This means that, on EBCDIC platforms, the PCRE library can handle either 639EBCDIC, or UTF-8, but not both. To support both in the same compiled library 640would need different lookups depending on whether PCRE_UTF8 was set or not. 641This would make it impossible to use characters in switch/case statements, 642which would reduce performance. For a theoretical use (which nobody has asked 643for) in a minority area (EBCDIC platforms), this is not sensible. Any 644application that did need both could compile two versions of the library, using 645macros to give the functions distinct names. */ 646 647#ifndef SUPPORT_UTF8 648 649/* UTF-8 support is not enabled; use the platform-dependent character literals 650so that PCRE works on both ASCII and EBCDIC platforms, in non-UTF-mode only. */ 651 652#define CHAR_HT '\t' 653#define CHAR_VT '\v' 654#define CHAR_FF '\f' 655#define CHAR_CR '\r' 656#define CHAR_NL '\n' 657#define CHAR_BS '\b' 658#define CHAR_BEL '\a' 659#ifdef EBCDIC 660#define CHAR_ESC '\047' 661#define CHAR_DEL '\007' 662#else 663#define CHAR_ESC '\033' 664#define CHAR_DEL '\177' 665#endif 666 667#define CHAR_SPACE ' ' 668#define CHAR_EXCLAMATION_MARK '!' 669#define CHAR_QUOTATION_MARK '"' 670#define CHAR_NUMBER_SIGN '#' 671#define CHAR_DOLLAR_SIGN '$' 672#define CHAR_PERCENT_SIGN '%' 673#define CHAR_AMPERSAND '&' 674#define CHAR_APOSTROPHE '\'' 675#define CHAR_LEFT_PARENTHESIS '(' 676#define CHAR_RIGHT_PARENTHESIS ')' 677#define CHAR_ASTERISK '*' 678#define CHAR_PLUS '+' 679#define CHAR_COMMA ',' 680#define CHAR_MINUS '-' 681#define CHAR_DOT '.' 682#define CHAR_SLASH '/' 683#define CHAR_0 '0' 684#define CHAR_1 '1' 685#define CHAR_2 '2' 686#define CHAR_3 '3' 687#define CHAR_4 '4' 688#define CHAR_5 '5' 689#define CHAR_6 '6' 690#define CHAR_7 '7' 691#define CHAR_8 '8' 692#define CHAR_9 '9' 693#define CHAR_COLON ':' 694#define CHAR_SEMICOLON ';' 695#define CHAR_LESS_THAN_SIGN '<' 696#define CHAR_EQUALS_SIGN '=' 697#define CHAR_GREATER_THAN_SIGN '>' 698#define CHAR_QUESTION_MARK '?' 699#define CHAR_COMMERCIAL_AT '@' 700#define CHAR_A 'A' 701#define CHAR_B 'B' 702#define CHAR_C 'C' 703#define CHAR_D 'D' 704#define CHAR_E 'E' 705#define CHAR_F 'F' 706#define CHAR_G 'G' 707#define CHAR_H 'H' 708#define CHAR_I 'I' 709#define CHAR_J 'J' 710#define CHAR_K 'K' 711#define CHAR_L 'L' 712#define CHAR_M 'M' 713#define CHAR_N 'N' 714#define CHAR_O 'O' 715#define CHAR_P 'P' 716#define CHAR_Q 'Q' 717#define CHAR_R 'R' 718#define CHAR_S 'S' 719#define CHAR_T 'T' 720#define CHAR_U 'U' 721#define CHAR_V 'V' 722#define CHAR_W 'W' 723#define CHAR_X 'X' 724#define CHAR_Y 'Y' 725#define CHAR_Z 'Z' 726#define CHAR_LEFT_SQUARE_BRACKET '[' 727#define CHAR_BACKSLASH '\\' 728#define CHAR_RIGHT_SQUARE_BRACKET ']' 729#define CHAR_CIRCUMFLEX_ACCENT '^' 730#define CHAR_UNDERSCORE '_' 731#define CHAR_GRAVE_ACCENT '`' 732#define CHAR_a 'a' 733#define CHAR_b 'b' 734#define CHAR_c 'c' 735#define CHAR_d 'd' 736#define CHAR_e 'e' 737#define CHAR_f 'f' 738#define CHAR_g 'g' 739#define CHAR_h 'h' 740#define CHAR_i 'i' 741#define CHAR_j 'j' 742#define CHAR_k 'k' 743#define CHAR_l 'l' 744#define CHAR_m 'm' 745#define CHAR_n 'n' 746#define CHAR_o 'o' 747#define CHAR_p 'p' 748#define CHAR_q 'q' 749#define CHAR_r 'r' 750#define CHAR_s 's' 751#define CHAR_t 't' 752#define CHAR_u 'u' 753#define CHAR_v 'v' 754#define CHAR_w 'w' 755#define CHAR_x 'x' 756#define CHAR_y 'y' 757#define CHAR_z 'z' 758#define CHAR_LEFT_CURLY_BRACKET '{' 759#define CHAR_VERTICAL_LINE '|' 760#define CHAR_RIGHT_CURLY_BRACKET '}' 761#define CHAR_TILDE '~' 762 763#define STR_HT "\t" 764#define STR_VT "\v" 765#define STR_FF "\f" 766#define STR_CR "\r" 767#define STR_NL "\n" 768#define STR_BS "\b" 769#define STR_BEL "\a" 770#ifdef EBCDIC 771#define STR_ESC "\047" 772#define STR_DEL "\007" 773#else 774#define STR_ESC "\033" 775#define STR_DEL "\177" 776#endif 777 778#define STR_SPACE " " 779#define STR_EXCLAMATION_MARK "!" 780#define STR_QUOTATION_MARK "\"" 781#define STR_NUMBER_SIGN "#" 782#define STR_DOLLAR_SIGN "$" 783#define STR_PERCENT_SIGN "%" 784#define STR_AMPERSAND "&" 785#define STR_APOSTROPHE "'" 786#define STR_LEFT_PARENTHESIS "(" 787#define STR_RIGHT_PARENTHESIS ")" 788#define STR_ASTERISK "*" 789#define STR_PLUS "+" 790#define STR_COMMA "," 791#define STR_MINUS "-" 792#define STR_DOT "." 793#define STR_SLASH "/" 794#define STR_0 "0" 795#define STR_1 "1" 796#define STR_2 "2" 797#define STR_3 "3" 798#define STR_4 "4" 799#define STR_5 "5" 800#define STR_6 "6" 801#define STR_7 "7" 802#define STR_8 "8" 803#define STR_9 "9" 804#define STR_COLON ":" 805#define STR_SEMICOLON ";" 806#define STR_LESS_THAN_SIGN "<" 807#define STR_EQUALS_SIGN "=" 808#define STR_GREATER_THAN_SIGN ">" 809#define STR_QUESTION_MARK "?" 810#define STR_COMMERCIAL_AT "@" 811#define STR_A "A" 812#define STR_B "B" 813#define STR_C "C" 814#define STR_D "D" 815#define STR_E "E" 816#define STR_F "F" 817#define STR_G "G" 818#define STR_H "H" 819#define STR_I "I" 820#define STR_J "J" 821#define STR_K "K" 822#define STR_L "L" 823#define STR_M "M" 824#define STR_N "N" 825#define STR_O "O" 826#define STR_P "P" 827#define STR_Q "Q" 828#define STR_R "R" 829#define STR_S "S" 830#define STR_T "T" 831#define STR_U "U" 832#define STR_V "V" 833#define STR_W "W" 834#define STR_X "X" 835#define STR_Y "Y" 836#define STR_Z "Z" 837#define STR_LEFT_SQUARE_BRACKET "[" 838#define STR_BACKSLASH "\\" 839#define STR_RIGHT_SQUARE_BRACKET "]" 840#define STR_CIRCUMFLEX_ACCENT "^" 841#define STR_UNDERSCORE "_" 842#define STR_GRAVE_ACCENT "`" 843#define STR_a "a" 844#define STR_b "b" 845#define STR_c "c" 846#define STR_d "d" 847#define STR_e "e" 848#define STR_f "f" 849#define STR_g "g" 850#define STR_h "h" 851#define STR_i "i" 852#define STR_j "j" 853#define STR_k "k" 854#define STR_l "l" 855#define STR_m "m" 856#define STR_n "n" 857#define STR_o "o" 858#define STR_p "p" 859#define STR_q "q" 860#define STR_r "r" 861#define STR_s "s" 862#define STR_t "t" 863#define STR_u "u" 864#define STR_v "v" 865#define STR_w "w" 866#define STR_x "x" 867#define STR_y "y" 868#define STR_z "z" 869#define STR_LEFT_CURLY_BRACKET "{" 870#define STR_VERTICAL_LINE "|" 871#define STR_RIGHT_CURLY_BRACKET "}" 872#define STR_TILDE "~" 873 874#define STRING_ACCEPT0 "ACCEPT\0" 875#define STRING_COMMIT0 "COMMIT\0" 876#define STRING_F0 "F\0" 877#define STRING_FAIL0 "FAIL\0" 878#define STRING_PRUNE0 "PRUNE\0" 879#define STRING_SKIP0 "SKIP\0" 880#define STRING_THEN "THEN" 881 882#define STRING_alpha0 "alpha\0" 883#define STRING_lower0 "lower\0" 884#define STRING_upper0 "upper\0" 885#define STRING_alnum0 "alnum\0" 886#define STRING_ascii0 "ascii\0" 887#define STRING_blank0 "blank\0" 888#define STRING_cntrl0 "cntrl\0" 889#define STRING_digit0 "digit\0" 890#define STRING_graph0 "graph\0" 891#define STRING_print0 "print\0" 892#define STRING_punct0 "punct\0" 893#define STRING_space0 "space\0" 894#define STRING_word0 "word\0" 895#define STRING_xdigit "xdigit" 896 897#define STRING_DEFINE "DEFINE" 898 899#define STRING_CR_RIGHTPAR "CR)" 900#define STRING_LF_RIGHTPAR "LF)" 901#define STRING_CRLF_RIGHTPAR "CRLF)" 902#define STRING_ANY_RIGHTPAR "ANY)" 903#define STRING_ANYCRLF_RIGHTPAR "ANYCRLF)" 904#define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)" 905#define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)" 906#define STRING_UTF8_RIGHTPAR "UTF8)" 907 908#else /* SUPPORT_UTF8 */ 909 910/* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This 911works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode 912only. */ 913 914#define CHAR_HT '\011' 915#define CHAR_VT '\013' 916#define CHAR_FF '\014' 917#define CHAR_CR '\015' 918#define CHAR_NL '\012' 919#define CHAR_BS '\010' 920#define CHAR_BEL '\007' 921#define CHAR_ESC '\033' 922#define CHAR_DEL '\177' 923 924#define CHAR_SPACE '\040' 925#define CHAR_EXCLAMATION_MARK '\041' 926#define CHAR_QUOTATION_MARK '\042' 927#define CHAR_NUMBER_SIGN '\043' 928#define CHAR_DOLLAR_SIGN '\044' 929#define CHAR_PERCENT_SIGN '\045' 930#define CHAR_AMPERSAND '\046' 931#define CHAR_APOSTROPHE '\047' 932#define CHAR_LEFT_PARENTHESIS '\050' 933#define CHAR_RIGHT_PARENTHESIS '\051' 934#define CHAR_ASTERISK '\052' 935#define CHAR_PLUS '\053' 936#define CHAR_COMMA '\054' 937#define CHAR_MINUS '\055' 938#define CHAR_DOT '\056' 939#define CHAR_SLASH '\057' 940#define CHAR_0 '\060' 941#define CHAR_1 '\061' 942#define CHAR_2 '\062' 943#define CHAR_3 '\063' 944#define CHAR_4 '\064' 945#define CHAR_5 '\065' 946#define CHAR_6 '\066' 947#define CHAR_7 '\067' 948#define CHAR_8 '\070' 949#define CHAR_9 '\071' 950#define CHAR_COLON '\072' 951#define CHAR_SEMICOLON '\073' 952#define CHAR_LESS_THAN_SIGN '\074' 953#define CHAR_EQUALS_SIGN '\075' 954#define CHAR_GREATER_THAN_SIGN '\076' 955#define CHAR_QUESTION_MARK '\077' 956#define CHAR_COMMERCIAL_AT '\100' 957#define CHAR_A '\101' 958#define CHAR_B '\102' 959#define CHAR_C '\103' 960#define CHAR_D '\104' 961#define CHAR_E '\105' 962#define CHAR_F '\106' 963#define CHAR_G '\107' 964#define CHAR_H '\110' 965#define CHAR_I '\111' 966#define CHAR_J '\112' 967#define CHAR_K '\113' 968#define CHAR_L '\114' 969#define CHAR_M '\115' 970#define CHAR_N '\116' 971#define CHAR_O '\117' 972#define CHAR_P '\120' 973#define CHAR_Q '\121' 974#define CHAR_R '\122' 975#define CHAR_S '\123' 976#define CHAR_T '\124' 977#define CHAR_U '\125' 978#define CHAR_V '\126' 979#define CHAR_W '\127' 980#define CHAR_X '\130' 981#define CHAR_Y '\131' 982#define CHAR_Z '\132' 983#define CHAR_LEFT_SQUARE_BRACKET '\133' 984#define CHAR_BACKSLASH '\134' 985#define CHAR_RIGHT_SQUARE_BRACKET '\135' 986#define CHAR_CIRCUMFLEX_ACCENT '\136' 987#define CHAR_UNDERSCORE '\137' 988#define CHAR_GRAVE_ACCENT '\140' 989#define CHAR_a '\141' 990#define CHAR_b '\142' 991#define CHAR_c '\143' 992#define CHAR_d '\144' 993#define CHAR_e '\145' 994#define CHAR_f '\146' 995#define CHAR_g '\147' 996#define CHAR_h '\150' 997#define CHAR_i '\151' 998#define CHAR_j '\152' 999#define CHAR_k '\153' 1000#define CHAR_l '\154' 1001#define CHAR_m '\155' 1002#define CHAR_n '\156' 1003#define CHAR_o '\157' 1004#define CHAR_p '\160' 1005#define CHAR_q '\161' 1006#define CHAR_r '\162' 1007#define CHAR_s '\163' 1008#define CHAR_t '\164' 1009#define CHAR_u '\165' 1010#define CHAR_v '\166' 1011#define CHAR_w '\167' 1012#define CHAR_x '\170' 1013#define CHAR_y '\171' 1014#define CHAR_z '\172' 1015#define CHAR_LEFT_CURLY_BRACKET '\173' 1016#define CHAR_VERTICAL_LINE '\174' 1017#define CHAR_RIGHT_CURLY_BRACKET '\175' 1018#define CHAR_TILDE '\176' 1019 1020#define STR_HT "\011" 1021#define STR_VT "\013" 1022#define STR_FF "\014" 1023#define STR_CR "\015" 1024#define STR_NL "\012" 1025#define STR_BS "\010" 1026#define STR_BEL "\007" 1027#define STR_ESC "\033" 1028#define STR_DEL "\177" 1029 1030#define STR_SPACE "\040" 1031#define STR_EXCLAMATION_MARK "\041" 1032#define STR_QUOTATION_MARK "\042" 1033#define STR_NUMBER_SIGN "\043" 1034#define STR_DOLLAR_SIGN "\044" 1035#define STR_PERCENT_SIGN "\045" 1036#define STR_AMPERSAND "\046" 1037#define STR_APOSTROPHE "\047" 1038#define STR_LEFT_PARENTHESIS "\050" 1039#define STR_RIGHT_PARENTHESIS "\051" 1040#define STR_ASTERISK "\052" 1041#define STR_PLUS "\053" 1042#define STR_COMMA "\054" 1043#define STR_MINUS "\055" 1044#define STR_DOT "\056" 1045#define STR_SLASH "\057" 1046#define STR_0 "\060" 1047#define STR_1 "\061" 1048#define STR_2 "\062" 1049#define STR_3 "\063" 1050#define STR_4 "\064" 1051#define STR_5 "\065" 1052#define STR_6 "\066" 1053#define STR_7 "\067" 1054#define STR_8 "\070" 1055#define STR_9 "\071" 1056#define STR_COLON "\072" 1057#define STR_SEMICOLON "\073" 1058#define STR_LESS_THAN_SIGN "\074" 1059#define STR_EQUALS_SIGN "\075" 1060#define STR_GREATER_THAN_SIGN "\076" 1061#define STR_QUESTION_MARK "\077" 1062#define STR_COMMERCIAL_AT "\100" 1063#define STR_A "\101" 1064#define STR_B "\102" 1065#define STR_C "\103" 1066#define STR_D "\104" 1067#define STR_E "\105" 1068#define STR_F "\106" 1069#define STR_G "\107" 1070#define STR_H "\110" 1071#define STR_I "\111" 1072#define STR_J "\112" 1073#define STR_K "\113" 1074#define STR_L "\114" 1075#define STR_M "\115" 1076#define STR_N "\116" 1077#define STR_O "\117" 1078#define STR_P "\120" 1079#define STR_Q "\121" 1080#define STR_R "\122" 1081#define STR_S "\123" 1082#define STR_T "\124" 1083#define STR_U "\125" 1084#define STR_V "\126" 1085#define STR_W "\127" 1086#define STR_X "\130" 1087#define STR_Y "\131" 1088#define STR_Z "\132" 1089#define STR_LEFT_SQUARE_BRACKET "\133" 1090#define STR_BACKSLASH "\134" 1091#define STR_RIGHT_SQUARE_BRACKET "\135" 1092#define STR_CIRCUMFLEX_ACCENT "\136" 1093#define STR_UNDERSCORE "\137" 1094#define STR_GRAVE_ACCENT "\140" 1095#define STR_a "\141" 1096#define STR_b "\142" 1097#define STR_c "\143" 1098#define STR_d "\144" 1099#define STR_e "\145" 1100#define STR_f "\146" 1101#define STR_g "\147" 1102#define STR_h "\150" 1103#define STR_i "\151" 1104#define STR_j "\152" 1105#define STR_k "\153" 1106#define STR_l "\154" 1107#define STR_m "\155" 1108#define STR_n "\156" 1109#define STR_o "\157" 1110#define STR_p "\160" 1111#define STR_q "\161" 1112#define STR_r "\162" 1113#define STR_s "\163" 1114#define STR_t "\164" 1115#define STR_u "\165" 1116#define STR_v "\166" 1117#define STR_w "\167" 1118#define STR_x "\170" 1119#define STR_y "\171" 1120#define STR_z "\172" 1121#define STR_LEFT_CURLY_BRACKET "\173" 1122#define STR_VERTICAL_LINE "\174" 1123#define STR_RIGHT_CURLY_BRACKET "\175" 1124#define STR_TILDE "\176" 1125 1126#define STRING_ACCEPT0 STR_A STR_C STR_C STR_E STR_P STR_T "\0" 1127#define STRING_COMMIT0 STR_C STR_O STR_M STR_M STR_I STR_T "\0" 1128#define STRING_F0 STR_F "\0" 1129#define STRING_FAIL0 STR_F STR_A STR_I STR_L "\0" 1130#define STRING_PRUNE0 STR_P STR_R STR_U STR_N STR_E "\0" 1131#define STRING_SKIP0 STR_S STR_K STR_I STR_P "\0" 1132#define STRING_THEN STR_T STR_H STR_E STR_N 1133 1134#define STRING_alpha0 STR_a STR_l STR_p STR_h STR_a "\0" 1135#define STRING_lower0 STR_l STR_o STR_w STR_e STR_r "\0" 1136#define STRING_upper0 STR_u STR_p STR_p STR_e STR_r "\0" 1137#define STRING_alnum0 STR_a STR_l STR_n STR_u STR_m "\0" 1138#define STRING_ascii0 STR_a STR_s STR_c STR_i STR_i "\0" 1139#define STRING_blank0 STR_b STR_l STR_a STR_n STR_k "\0" 1140#define STRING_cntrl0 STR_c STR_n STR_t STR_r STR_l "\0" 1141#define STRING_digit0 STR_d STR_i STR_g STR_i STR_t "\0" 1142#define STRING_graph0 STR_g STR_r STR_a STR_p STR_h "\0" 1143#define STRING_print0 STR_p STR_r STR_i STR_n STR_t "\0" 1144#define STRING_punct0 STR_p STR_u STR_n STR_c STR_t "\0" 1145#define STRING_space0 STR_s STR_p STR_a STR_c STR_e "\0" 1146#define STRING_word0 STR_w STR_o STR_r STR_d "\0" 1147#define STRING_xdigit STR_x STR_d STR_i STR_g STR_i STR_t 1148 1149#define STRING_DEFINE STR_D STR_E STR_F STR_I STR_N STR_E 1150 1151#define STRING_CR_RIGHTPAR STR_C STR_R STR_RIGHT_PARENTHESIS 1152#define STRING_LF_RIGHTPAR STR_L STR_F STR_RIGHT_PARENTHESIS 1153#define STRING_CRLF_RIGHTPAR STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS 1154#define STRING_ANY_RIGHTPAR STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS 1155#define STRING_ANYCRLF_RIGHTPAR STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS 1156#define STRING_BSR_ANYCRLF_RIGHTPAR STR_B STR_S STR_R STR_UNDERSCORE STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS 1157#define STRING_BSR_UNICODE_RIGHTPAR STR_B STR_S STR_R STR_UNDERSCORE STR_U STR_N STR_I STR_C STR_O STR_D STR_E STR_RIGHT_PARENTHESIS 1158#define STRING_UTF8_RIGHTPAR STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS 1159 1160#endif /* SUPPORT_UTF8 */ 1161 1162/* Escape items that are just an encoding of a particular data value. */ 1163 1164#ifndef ESC_e 1165#define ESC_e CHAR_ESC 1166#endif 1167 1168#ifndef ESC_f 1169#define ESC_f CHAR_FF 1170#endif 1171 1172#ifndef ESC_n 1173#define ESC_n CHAR_NL 1174#endif 1175 1176#ifndef ESC_r 1177#define ESC_r CHAR_CR 1178#endif 1179 1180/* We can't officially use ESC_t because it is a POSIX reserved identifier 1181(presumably because of all the others like size_t). */ 1182 1183#ifndef ESC_tee 1184#define ESC_tee CHAR_HT 1185#endif 1186 1187/* Codes for different types of Unicode property */ 1188 1189#define PT_ANY 0 /* Any property - matches all chars */ 1190#define PT_LAMP 1 /* L& - the union of Lu, Ll, Lt */ 1191#define PT_GC 2 /* General characteristic (e.g. L) */ 1192#define PT_PC 3 /* Particular characteristic (e.g. Lu) */ 1193#define PT_SC 4 /* Script (e.g. Han) */ 1194 1195/* Flag bits and data types for the extended class (OP_XCLASS) for classes that 1196contain UTF-8 characters with values greater than 255. */ 1197 1198#define XCL_NOT 0x01 /* Flag: this is a negative class */ 1199#define XCL_MAP 0x02 /* Flag: a 32-byte map is present */ 1200 1201#define XCL_END 0 /* Marks end of individual items */ 1202#define XCL_SINGLE 1 /* Single item (one multibyte char) follows */ 1203#define XCL_RANGE 2 /* A range (two multibyte chars) follows */ 1204#define XCL_PROP 3 /* Unicode property (2-byte property code follows) */ 1205#define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */ 1206 1207/* These are escaped items that aren't just an encoding of a particular data 1208value such as \n. They must have non-zero values, as check_escape() returns 1209their negation. Also, they must appear in the same order as in the opcode 1210definitions below, up to ESC_z. There's a dummy for OP_ANY because it 1211corresponds to "." rather than an escape sequence, and another for OP_ALLANY 1212(which is used for [^] in JavaScript compatibility mode). 1213 1214The final escape must be ESC_REF as subsequent values are used for 1215backreferences (\1, \2, \3, etc). There are two tests in the code for an escape 1216greater than ESC_b and less than ESC_Z to detect the types that may be 1217repeated. These are the types that consume characters. If any new escapes are 1218put in between that don't consume a character, that code will have to change. 1219*/ 1220 1221enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, 1222 ESC_W, ESC_w, ESC_dum1, ESC_dum2, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H, 1223 ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_g, ESC_k, 1224 ESC_REF }; 1225 1226 1227/* Opcode table: Starting from 1 (i.e. after OP_END), the values up to 1228OP_EOD must correspond in order to the list of escapes immediately above. 1229 1230*** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions 1231that follow must also be updated to match. There are also tables called 1232"coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */ 1233 1234enum { 1235 OP_END, /* 0 End of pattern */ 1236 1237 /* Values corresponding to backslashed metacharacters */ 1238 1239 OP_SOD, /* 1 Start of data: \A */ 1240 OP_SOM, /* 2 Start of match (subject + offset): \G */ 1241 OP_SET_SOM, /* 3 Set start of match (\K) */ 1242 OP_NOT_WORD_BOUNDARY, /* 4 \B */ 1243 OP_WORD_BOUNDARY, /* 5 \b */ 1244 OP_NOT_DIGIT, /* 6 \D */ 1245 OP_DIGIT, /* 7 \d */ 1246 OP_NOT_WHITESPACE, /* 8 \S */ 1247 OP_WHITESPACE, /* 9 \s */ 1248 OP_NOT_WORDCHAR, /* 10 \W */ 1249 OP_WORDCHAR, /* 11 \w */ 1250 OP_ANY, /* 12 Match any character (subject to DOTALL) */ 1251 OP_ALLANY, /* 13 Match any character (not subject to DOTALL) */ 1252 OP_ANYBYTE, /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */ 1253 OP_NOTPROP, /* 15 \P (not Unicode property) */ 1254 OP_PROP, /* 16 \p (Unicode property) */ 1255 OP_ANYNL, /* 17 \R (any newline sequence) */ 1256 OP_NOT_HSPACE, /* 18 \H (not horizontal whitespace) */ 1257 OP_HSPACE, /* 19 \h (horizontal whitespace) */ 1258 OP_NOT_VSPACE, /* 20 \V (not vertical whitespace) */ 1259 OP_VSPACE, /* 21 \v (vertical whitespace) */ 1260 OP_EXTUNI, /* 22 \X (extended Unicode sequence */ 1261 OP_EODN, /* 23 End of data or \n at end of data: \Z. */ 1262 OP_EOD, /* 24 End of data: \z */ 1263 1264 OP_OPT, /* 25 Set runtime options */ 1265 OP_CIRC, /* 26 Start of line - varies with multiline switch */ 1266 OP_DOLL, /* 27 End of line - varies with multiline switch */ 1267 OP_CHAR, /* 28 Match one character, casefully */ 1268 OP_CHARNC, /* 29 Match one character, caselessly */ 1269 OP_NOT, /* 30 Match one character, not the following one */ 1270 1271 OP_STAR, /* 31 The maximizing and minimizing versions of */ 1272 OP_MINSTAR, /* 32 these six opcodes must come in pairs, with */ 1273 OP_PLUS, /* 33 the minimizing one second. */ 1274 OP_MINPLUS, /* 34 This first set applies to single characters.*/ 1275 OP_QUERY, /* 35 */ 1276 OP_MINQUERY, /* 36 */ 1277 1278 OP_UPTO, /* 37 From 0 to n matches */ 1279 OP_MINUPTO, /* 38 */ 1280 OP_EXACT, /* 39 Exactly n matches */ 1281 1282 OP_POSSTAR, /* 40 Possessified star */ 1283 OP_POSPLUS, /* 41 Possessified plus */ 1284 OP_POSQUERY, /* 42 Posesssified query */ 1285 OP_POSUPTO, /* 43 Possessified upto */ 1286 1287 OP_NOTSTAR, /* 44 The maximizing and minimizing versions of */ 1288 OP_NOTMINSTAR, /* 45 these six opcodes must come in pairs, with */ 1289 OP_NOTPLUS, /* 46 the minimizing one second. They must be in */ 1290 OP_NOTMINPLUS, /* 47 exactly the same order as those above. */ 1291 OP_NOTQUERY, /* 48 This set applies to "not" single characters. */ 1292 OP_NOTMINQUERY, /* 49 */ 1293 1294 OP_NOTUPTO, /* 50 From 0 to n matches */ 1295 OP_NOTMINUPTO, /* 51 */ 1296 OP_NOTEXACT, /* 52 Exactly n matches */ 1297 1298 OP_NOTPOSSTAR, /* 53 Possessified versions */ 1299 OP_NOTPOSPLUS, /* 54 */ 1300 OP_NOTPOSQUERY, /* 55 */ 1301 OP_NOTPOSUPTO, /* 56 */ 1302 1303 OP_TYPESTAR, /* 57 The maximizing and minimizing versions of */ 1304 OP_TYPEMINSTAR, /* 58 these six opcodes must come in pairs, with */ 1305 OP_TYPEPLUS, /* 59 the minimizing one second. These codes must */ 1306 OP_TYPEMINPLUS, /* 60 be in exactly the same order as those above. */ 1307 OP_TYPEQUERY, /* 61 This set applies to character types such as \d */ 1308 OP_TYPEMINQUERY, /* 62 */ 1309 1310 OP_TYPEUPTO, /* 63 From 0 to n matches */ 1311 OP_TYPEMINUPTO, /* 64 */ 1312 OP_TYPEEXACT, /* 65 Exactly n matches */ 1313 1314 OP_TYPEPOSSTAR, /* 66 Possessified versions */ 1315 OP_TYPEPOSPLUS, /* 67 */ 1316 OP_TYPEPOSQUERY, /* 68 */ 1317 OP_TYPEPOSUPTO, /* 69 */ 1318 1319 OP_CRSTAR, /* 70 The maximizing and minimizing versions of */ 1320 OP_CRMINSTAR, /* 71 all these opcodes must come in pairs, with */ 1321 OP_CRPLUS, /* 72 the minimizing one second. These codes must */ 1322 OP_CRMINPLUS, /* 73 be in exactly the same order as those above. */ 1323 OP_CRQUERY, /* 74 These are for character classes and back refs */ 1324 OP_CRMINQUERY, /* 75 */ 1325 OP_CRRANGE, /* 76 These are different to the three sets above. */ 1326 OP_CRMINRANGE, /* 77 */ 1327 1328 OP_CLASS, /* 78 Match a character class, chars < 256 only */ 1329 OP_NCLASS, /* 79 Same, but the bitmap was created from a negative 1330 class - the difference is relevant only when a UTF-8 1331 character > 255 is encountered. */ 1332 1333 OP_XCLASS, /* 80 Extended class for handling UTF-8 chars within the 1334 class. This does both positive and negative. */ 1335 1336 OP_REF, /* 81 Match a back reference */ 1337 OP_RECURSE, /* 82 Match a numbered subpattern (possibly recursive) */ 1338 OP_CALLOUT, /* 83 Call out to external function if provided */ 1339 1340 OP_ALT, /* 84 Start of alternation */ 1341 OP_KET, /* 85 End of group that doesn't have an unbounded repeat */ 1342 OP_KETRMAX, /* 86 These two must remain together and in this */ 1343 OP_KETRMIN, /* 87 order. They are for groups the repeat for ever. */ 1344 1345 /* The assertions must come before BRA, CBRA, ONCE, and COND.*/ 1346 1347 OP_ASSERT, /* 88 Positive lookahead */ 1348 OP_ASSERT_NOT, /* 89 Negative lookahead */ 1349 OP_ASSERTBACK, /* 90 Positive lookbehind */ 1350 OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */ 1351 OP_REVERSE, /* 92 Move pointer back - used in lookbehind assertions */ 1352 1353 /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first, 1354 as there's a test for >= ONCE for a subpattern that isn't an assertion. */ 1355 1356 OP_ONCE, /* 93 Atomic group */ 1357 OP_BRA, /* 94 Start of non-capturing bracket */ 1358 OP_CBRA, /* 95 Start of capturing bracket */ 1359 OP_COND, /* 96 Conditional group */ 1360 1361 /* These three must follow the previous three, in the same order. There's a 1362 check for >= SBRA to distinguish the two sets. */ 1363 1364 OP_SBRA, /* 97 Start of non-capturing bracket, check empty */ 1365 OP_SCBRA, /* 98 Start of capturing bracket, check empty */ 1366 OP_SCOND, /* 99 Conditional group, check empty */ 1367 1368 /* The next two pairs must (respectively) be kept together. */ 1369 1370 OP_CREF, /* 100 Used to hold a capture number as condition */ 1371 OP_NCREF, /* 101 Same, but generaged by a name reference*/ 1372 OP_RREF, /* 102 Used to hold a recursion number as condition */ 1373 OP_NRREF, /* 103 Same, but generaged by a name reference*/ 1374 OP_DEF, /* 104 The DEFINE condition */ 1375 1376 OP_BRAZERO, /* 105 These two must remain together and in this */ 1377 OP_BRAMINZERO, /* 106 order. */ 1378 1379 /* These are backtracking control verbs */ 1380 1381 OP_PRUNE, /* 107 */ 1382 OP_SKIP, /* 108 */ 1383 OP_THEN, /* 109 */ 1384 OP_COMMIT, /* 110 */ 1385 1386 /* These are forced failure and success verbs */ 1387 1388 OP_FAIL, /* 111 */ 1389 OP_ACCEPT, /* 112 */ 1390 OP_CLOSE, /* 113 Used before OP_ACCEPT to close open captures */ 1391 1392 /* This is used to skip a subpattern with a {0} quantifier */ 1393 1394 OP_SKIPZERO, /* 114 */ 1395 1396 /* This is not an opcode, but is used to check that tables indexed by opcode 1397 are the correct length, in order to catch updating errors - there have been 1398 some in the past. */ 1399 1400 OP_TABLE_LENGTH 1401}; 1402 1403/* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro 1404definitions that follow must also be updated to match. There are also tables 1405called "coptable" cna "poptable" in pcre_dfa_exec.c that must be updated. */ 1406 1407 1408/* This macro defines textual names for all the opcodes. These are used only 1409for debugging. The macro is referenced only in pcre_printint.c. */ 1410 1411#define OP_NAME_LIST \ 1412 "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d", \ 1413 "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte", \ 1414 "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v", \ 1415 "extuni", "\\Z", "\\z", \ 1416 "Opt", "^", "$", "char", "charnc", "not", \ 1417 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \ 1418 "*+","++", "?+", "{", \ 1419 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \ 1420 "*+","++", "?+", "{", \ 1421 "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \ 1422 "*+","++", "?+", "{", \ 1423 "*", "*?", "+", "+?", "?", "??", "{", "{", \ 1424 "class", "nclass", "xclass", "Ref", "Recurse", "Callout", \ 1425 "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", \ 1426 "AssertB", "AssertB not", "Reverse", \ 1427 "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond", \ 1428 "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def", \ 1429 "Brazero", "Braminzero", \ 1430 "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT", \ 1431 "Close", "Skip zero" 1432 1433 1434/* This macro defines the length of fixed length operations in the compiled 1435regex. The lengths are used when searching for specific things, and also in the 1436debugging printing of a compiled regex. We use a macro so that it can be 1437defined close to the definitions of the opcodes themselves. 1438 1439As things have been extended, some of these are no longer fixed lenths, but are 1440minima instead. For example, the length of a single-character repeat may vary 1441in UTF-8 mode. The code that uses this table must know about such things. */ 1442 1443#define OP_LENGTHS \ 1444 1, /* End */ \ 1445 1, 1, 1, 1, 1, /* \A, \G, \K, \B, \b */ \ 1446 1, 1, 1, 1, 1, 1, /* \D, \d, \S, \s, \W, \w */ \ 1447 1, 1, 1, /* Any, AllAny, Anybyte */ \ 1448 3, 3, /* \P, \p */ \ 1449 1, 1, 1, 1, 1, /* \R, \H, \h, \V, \v */ \ 1450 1, /* \X */ \ 1451 1, 1, 2, 1, 1, /* \Z, \z, Opt, ^, $ */ \ 1452 2, /* Char - the minimum length */ \ 1453 2, /* Charnc - the minimum length */ \ 1454 2, /* not */ \ 1455 /* Positive single-char repeats ** These are */ \ 1456 2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \ 1457 4, 4, 4, /* upto, minupto, exact ** UTF-8 mode */ \ 1458 2, 2, 2, 4, /* *+, ++, ?+, upto+ */ \ 1459 /* Negative single-char repeats - only for chars < 256 */ \ 1460 2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \ 1461 4, 4, 4, /* NOT upto, minupto, exact */ \ 1462 2, 2, 2, 4, /* Possessive *, +, ?, upto */ \ 1463 /* Positive type repeats */ \ 1464 2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \ 1465 4, 4, 4, /* Type upto, minupto, exact */ \ 1466 2, 2, 2, 4, /* Possessive *+, ++, ?+, upto+ */ \ 1467 /* Character class & ref repeats */ \ 1468 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \ 1469 5, 5, /* CRRANGE, CRMINRANGE */ \ 1470 33, /* CLASS */ \ 1471 33, /* NCLASS */ \ 1472 0, /* XCLASS - variable length */ \ 1473 3, /* REF */ \ 1474 1+LINK_SIZE, /* RECURSE */ \ 1475 2+2*LINK_SIZE, /* CALLOUT */ \ 1476 1+LINK_SIZE, /* Alt */ \ 1477 1+LINK_SIZE, /* Ket */ \ 1478 1+LINK_SIZE, /* KetRmax */ \ 1479 1+LINK_SIZE, /* KetRmin */ \ 1480 1+LINK_SIZE, /* Assert */ \ 1481 1+LINK_SIZE, /* Assert not */ \ 1482 1+LINK_SIZE, /* Assert behind */ \ 1483 1+LINK_SIZE, /* Assert behind not */ \ 1484 1+LINK_SIZE, /* Reverse */ \ 1485 1+LINK_SIZE, /* ONCE */ \ 1486 1+LINK_SIZE, /* BRA */ \ 1487 3+LINK_SIZE, /* CBRA */ \ 1488 1+LINK_SIZE, /* COND */ \ 1489 1+LINK_SIZE, /* SBRA */ \ 1490 3+LINK_SIZE, /* SCBRA */ \ 1491 1+LINK_SIZE, /* SCOND */ \ 1492 3, 3, /* CREF, NCREF */ \ 1493 3, 3, /* RREF, NRREF */ \ 1494 1, /* DEF */ \ 1495 1, 1, /* BRAZERO, BRAMINZERO */ \ 1496 1, 1, 1, 1, /* PRUNE, SKIP, THEN, COMMIT, */ \ 1497 1, 1, 3, 1 /* FAIL, ACCEPT, CLOSE, SKIPZERO */ 1498 1499 1500/* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion" 1501condition. */ 1502 1503#define RREF_ANY 0xffff 1504 1505/* Compile time error code numbers. They are given names so that they can more 1506easily be tracked. When a new number is added, the table called eint in 1507pcreposix.c must be updated. */ 1508 1509enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9, 1510 ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19, 1511 ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29, 1512 ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39, 1513 ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49, 1514 ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59, 1515 ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERRCOUNT }; 1516 1517/* The real format of the start of the pcre block; the index of names and the 1518code vector run on as long as necessary after the end. We store an explicit 1519offset to the name table so that if a regex is compiled on one host, saved, and 1520then run on another where the size of pointers is different, all might still 1521be well. For the case of compiled-on-4 and run-on-8, we include an extra 1522pointer that is always NULL. For future-proofing, a few dummy fields were 1523originally included - even though you can never get this planning right - but 1524there is only one left now. 1525 1526NOTE NOTE NOTE: 1527Because people can now save and re-use compiled patterns, any additions to this 1528structure should be made at the end, and something earlier (e.g. a new 1529flag in the options or one of the dummy fields) should indicate that the new 1530fields are present. Currently PCRE always sets the dummy fields to zero. 1531NOTE NOTE NOTE 1532*/ 1533 1534typedef struct real_pcre { 1535 pcre_uint32 magic_number; 1536 pcre_uint32 size; /* Total that was malloced */ 1537 pcre_uint32 options; /* Public options */ 1538 pcre_uint16 flags; /* Private flags */ 1539 pcre_uint16 dummy1; /* For future use */ 1540 pcre_uint16 top_bracket; 1541 pcre_uint16 top_backref; 1542 pcre_uint16 first_byte; 1543 pcre_uint16 req_byte; 1544 pcre_uint16 name_table_offset; /* Offset to name table that follows */ 1545 pcre_uint16 name_entry_size; /* Size of any name items */ 1546 pcre_uint16 name_count; /* Number of name items */ 1547 pcre_uint16 ref_count; /* Reference count */ 1548 1549 const unsigned char *tables; /* Pointer to tables or NULL for std */ 1550 const unsigned char *nullpad; /* NULL padding */ 1551} real_pcre; 1552 1553/* The format of the block used to store data from pcre_study(). The same 1554remark (see NOTE above) about extending this structure applies. */ 1555 1556typedef struct pcre_study_data { 1557 pcre_uint32 size; /* Total that was malloced */ 1558 pcre_uint32 flags; /* Private flags */ 1559 uschar start_bits[32]; /* Starting char bits */ 1560 pcre_uint32 minlength; /* Minimum subject length */ 1561} pcre_study_data; 1562 1563/* Structure for building a chain of open capturing subpatterns during 1564compiling, so that instructions to close them can be compiled when (*ACCEPT) is 1565encountered. This is also used to identify subpatterns that contain recursive 1566back references to themselves, so that they can be made atomic. */ 1567 1568typedef struct open_capitem { 1569 struct open_capitem *next; /* Chain link */ 1570 pcre_uint16 number; /* Capture number */ 1571 pcre_uint16 flag; /* Set TRUE if recursive back ref */ 1572} open_capitem; 1573 1574/* Structure for passing "static" information around between the functions 1575doing the compiling, so that they are thread-safe. */ 1576 1577typedef struct compile_data { 1578 const uschar *lcc; /* Points to lower casing table */ 1579 const uschar *fcc; /* Points to case-flipping table */ 1580 const uschar *cbits; /* Points to character type table */ 1581 const uschar *ctypes; /* Points to table of type maps */ 1582 const uschar *start_workspace;/* The start of working space */ 1583 const uschar *start_code; /* The start of the compiled code */ 1584 const uschar *start_pattern; /* The start of the pattern */ 1585 const uschar *end_pattern; /* The end of the pattern */ 1586 open_capitem *open_caps; /* Chain of open capture items */ 1587 uschar *hwm; /* High watermark of workspace */ 1588 uschar *name_table; /* The name/number table */ 1589 int names_found; /* Number of entries so far */ 1590 int name_entry_size; /* Size of each entry */ 1591 int bracount; /* Count of capturing parens as we compile */ 1592 int final_bracount; /* Saved value after first pass */ 1593 int top_backref; /* Maximum back reference */ 1594 unsigned int backref_map; /* Bitmap of low back refs */ 1595 int external_options; /* External (initial) options */ 1596 int external_flags; /* External flag bits to be set */ 1597 int req_varyopt; /* "After variable item" flag for reqbyte */ 1598 BOOL had_accept; /* (*ACCEPT) encountered */ 1599 BOOL check_lookbehind; /* Lookbehinds need later checking */ 1600 int nltype; /* Newline type */ 1601 int nllen; /* Newline string length */ 1602 uschar nl[4]; /* Newline string when fixed length */ 1603} compile_data; 1604 1605/* Structure for maintaining a chain of pointers to the currently incomplete 1606branches, for testing for left recursion. */ 1607 1608typedef struct branch_chain { 1609 struct branch_chain *outer; 1610 uschar *current_branch; 1611} branch_chain; 1612 1613/* Structure for items in a linked list that represents an explicit recursive 1614call within the pattern. */ 1615 1616typedef struct recursion_info { 1617 struct recursion_info *prevrec; /* Previous recursion record (or NULL) */ 1618 int group_num; /* Number of group that was called */ 1619 const uschar *after_call; /* "Return value": points after the call in the expr */ 1620 int *offset_save; /* Pointer to start of saved offsets */ 1621 int saved_max; /* Number of saved offsets */ 1622 int save_offset_top; /* Current value of offset_top */ 1623} recursion_info; 1624 1625/* Structure for building a chain of data for holding the values of the subject 1626pointer at the start of each subpattern, so as to detect when an empty string 1627has been matched by a subpattern - to break infinite loops. */ 1628 1629typedef struct eptrblock { 1630 struct eptrblock *epb_prev; 1631 USPTR epb_saved_eptr; 1632} eptrblock; 1633 1634 1635/* Structure for passing "static" information around between the functions 1636doing traditional NFA matching, so that they are thread-safe. */ 1637 1638typedef struct match_data { 1639 unsigned long int match_call_count; /* As it says */ 1640 unsigned long int match_limit; /* As it says */ 1641 unsigned long int match_limit_recursion; /* As it says */ 1642 int *offset_vector; /* Offset vector */ 1643 int offset_end; /* One past the end */ 1644 int offset_max; /* The maximum usable for return data */ 1645 int nltype; /* Newline type */ 1646 int nllen; /* Newline string length */ 1647 int name_count; /* Number of names in name table */ 1648 int name_entry_size; /* Size of entry in names table */ 1649 uschar *name_table; /* Table of names */ 1650 uschar nl[4]; /* Newline string when fixed */ 1651 const uschar *lcc; /* Points to lower casing table */ 1652 const uschar *ctypes; /* Points to table of type maps */ 1653 BOOL offset_overflow; /* Set if too many extractions */ 1654 BOOL notbol; /* NOTBOL flag */ 1655 BOOL noteol; /* NOTEOL flag */ 1656 BOOL utf8; /* UTF8 flag */ 1657 BOOL jscript_compat; /* JAVASCRIPT_COMPAT flag */ 1658 BOOL endonly; /* Dollar not before final \n */ 1659 BOOL notempty; /* Empty string match not wanted */ 1660 BOOL notempty_atstart; /* Empty string match at start not wanted */ 1661 BOOL hitend; /* Hit the end of the subject at some point */ 1662 BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */ 1663 const uschar *start_code; /* For use when recursing */ 1664 USPTR start_subject; /* Start of the subject string */ 1665 USPTR end_subject; /* End of the subject string */ 1666 USPTR start_match_ptr; /* Start of matched string */ 1667 USPTR end_match_ptr; /* Subject position at end match */ 1668 USPTR start_used_ptr; /* Earliest consulted character */ 1669 int partial; /* PARTIAL options */ 1670 int end_offset_top; /* Highwater mark at end of match */ 1671 int capture_last; /* Most recent capture number */ 1672 int start_offset; /* The start offset value */ 1673 eptrblock *eptrchain; /* Chain of eptrblocks for tail recursions */ 1674 int eptrn; /* Next free eptrblock */ 1675 recursion_info *recursive; /* Linked list of recursion data */ 1676 void *callout_data; /* To pass back to callouts */ 1677} match_data; 1678 1679/* A similar structure is used for the same purpose by the DFA matching 1680functions. */ 1681 1682typedef struct dfa_match_data { 1683 const uschar *start_code; /* Start of the compiled pattern */ 1684 const uschar *start_subject; /* Start of the subject string */ 1685 const uschar *end_subject; /* End of subject string */ 1686 const uschar *start_used_ptr; /* Earliest consulted character */ 1687 const uschar *tables; /* Character tables */ 1688 int start_offset; /* The start offset value */ 1689 int moptions; /* Match options */ 1690 int poptions; /* Pattern options */ 1691 int nltype; /* Newline type */ 1692 int nllen; /* Newline string length */ 1693 uschar nl[4]; /* Newline string when fixed */ 1694 void *callout_data; /* To pass back to callouts */ 1695} dfa_match_data; 1696 1697/* Bit definitions for entries in the pcre_ctypes table. */ 1698 1699#define ctype_space 0x01 1700#define ctype_letter 0x02 1701#define ctype_digit 0x04 1702#define ctype_xdigit 0x08 1703#define ctype_word 0x10 /* alphanumeric or '_' */ 1704#define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */ 1705 1706/* Offsets for the bitmap tables in pcre_cbits. Each table contains a set 1707of bits for a class map. Some classes are built by combining these tables. */ 1708 1709#define cbit_space 0 /* [:space:] or \s */ 1710#define cbit_xdigit 32 /* [:xdigit:] */ 1711#define cbit_digit 64 /* [:digit:] or \d */ 1712#define cbit_upper 96 /* [:upper:] */ 1713#define cbit_lower 128 /* [:lower:] */ 1714#define cbit_word 160 /* [:word:] or \w */ 1715#define cbit_graph 192 /* [:graph:] */ 1716#define cbit_print 224 /* [:print:] */ 1717#define cbit_punct 256 /* [:punct:] */ 1718#define cbit_cntrl 288 /* [:cntrl:] */ 1719#define cbit_length 320 /* Length of the cbits table */ 1720 1721/* Offsets of the various tables from the base tables pointer, and 1722total length. */ 1723 1724#define lcc_offset 0 1725#define fcc_offset 256 1726#define cbits_offset 512 1727#define ctypes_offset (cbits_offset + cbit_length) 1728#define tables_length (ctypes_offset + 256) 1729 1730/* Layout of the UCP type table that translates property names into types and 1731codes. Each entry used to point directly to a name, but to reduce the number of 1732relocations in shared libraries, it now has an offset into a single string 1733instead. */ 1734 1735typedef struct { 1736 pcre_uint16 name_offset; 1737 pcre_uint16 type; 1738 pcre_uint16 value; 1739} ucp_type_table; 1740 1741 1742/* Internal shared data tables. These are tables that are used by more than one 1743of the exported public functions. They have to be "external" in the C sense, 1744but are not part of the PCRE public API. The data for these tables is in the 1745pcre_tables.c module. */ 1746 1747extern const int _pcre_utf8_table1[]; 1748extern const int _pcre_utf8_table2[]; 1749extern const int _pcre_utf8_table3[]; 1750extern const uschar _pcre_utf8_table4[]; 1751 1752extern const int _pcre_utf8_table1_size; 1753 1754extern const char _pcre_utt_names[]; 1755extern const ucp_type_table _pcre_utt[]; 1756extern const int _pcre_utt_size; 1757 1758extern const uschar _pcre_default_tables[]; 1759 1760extern const uschar _pcre_OP_lengths[]; 1761 1762 1763/* Internal shared functions. These are functions that are used by more than 1764one of the exported public functions. They have to be "external" in the C 1765sense, but are not part of the PCRE public API. */ 1766 1767extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int); 1768extern BOOL _pcre_is_newline(USPTR, int, USPTR, int *, BOOL); 1769extern int _pcre_ord2utf8(int, uschar *); 1770extern real_pcre *_pcre_try_flipped(const real_pcre *, real_pcre *, 1771 const pcre_study_data *, pcre_study_data *); 1772extern int _pcre_valid_utf8(USPTR, int); 1773extern BOOL _pcre_was_newline(USPTR, int, USPTR, int *, BOOL); 1774extern BOOL _pcre_xclass(int, const uschar *); 1775 1776 1777/* Unicode character database (UCD) */ 1778 1779typedef struct { 1780 uschar script; 1781 uschar chartype; 1782 pcre_int32 other_case; 1783} ucd_record; 1784 1785extern const ucd_record _pcre_ucd_records[]; 1786extern const uschar _pcre_ucd_stage1[]; 1787extern const pcre_uint16 _pcre_ucd_stage2[]; 1788extern const int _pcre_ucp_gentype[]; 1789 1790 1791/* UCD access macros */ 1792 1793#define UCD_BLOCK_SIZE 128 1794#define GET_UCD(ch) (_pcre_ucd_records + \ 1795 _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \ 1796 UCD_BLOCK_SIZE + ch % UCD_BLOCK_SIZE]) 1797 1798#define UCD_CHARTYPE(ch) GET_UCD(ch)->chartype 1799#define UCD_SCRIPT(ch) GET_UCD(ch)->script 1800#define UCD_CATEGORY(ch) _pcre_ucp_gentype[UCD_CHARTYPE(ch)] 1801#define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case) 1802 1803#endif 1804 1805/* End of pcre_internal.h */ 1806