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

Diff of /code/trunk/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 850 by zherczeg, Wed Jan 4 17:29:11 2012 UTC revision 1309 by ph10, Fri Apr 5 15:35:59 2013 UTC
# Line 7  Line 7 
7  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2012 University of Cambridge             Copyright (c) 1997-2013 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 40  POSSIBILITY OF SUCH DAMAGE. Line 40  POSSIBILITY OF SUCH DAMAGE.
40    
41  /* This header contains definitions that are shared between the different  /* This header contains definitions that are shared between the different
42  modules, but which are not relevant to the exported API. This includes some  modules, but which are not relevant to the exported API. This includes some
43  functions whose names all begin with "_pcre_" or "_pcre16_" depending on  functions whose names all begin with "_pcre_", "_pcre16_" or "_pcre32_"
44  the PRIV macro. */  depending on the PRIV macro. */
45    
46  #ifndef PCRE_INTERNAL_H  #ifndef PCRE_INTERNAL_H
47  #define PCRE_INTERNAL_H  #define PCRE_INTERNAL_H
# Line 53  the PRIV macro. */ Line 53  the PRIV macro. */
53  #endif  #endif
54    
55  /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */  /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
56  #ifndef COMPILE_PCRE16  
57    #if !defined COMPILE_PCRE16 && !defined COMPILE_PCRE32
58  #define COMPILE_PCRE8  #define COMPILE_PCRE8
59  #endif  #endif
60    
# Line 78  Until then we define it if SUPPORT_UTF i Line 79  Until then we define it if SUPPORT_UTF i
79  #define SUPPORT_UTF8 1  #define SUPPORT_UTF8 1
80  #endif  #endif
81    
82  /* We do not support both EBCDIC and UTF-8/16 at the same time. The "configure"  /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
83  script prevents both being selected, but not everybody uses "configure". */  script prevents both being selected, but not everybody uses "configure". */
84    
85  #if defined EBCDIC && defined SUPPORT_UTF  #if defined EBCDIC && defined SUPPORT_UTF
86  #error The use of both EBCDIC and SUPPORT_UTF8/16 is not supported.  #error The use of both EBCDIC and SUPPORT_UTF is not supported.
87  #endif  #endif
88    
89  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
# Line 111  setjmp and stdarg are used is when NO_RE Line 112  setjmp and stdarg are used is when NO_RE
112  #include <stdlib.h>  #include <stdlib.h>
113  #include <string.h>  #include <string.h>
114    
115    /* Valgrind (memcheck) support */
116    
117    #ifdef SUPPORT_VALGRIND
118    #include <valgrind/memcheck.h>
119    #endif
120    
121  /* When compiling a DLL for Windows, the exported symbols have to be declared  /* When compiling a DLL for Windows, the exported symbols have to be declared
122  using some MS magic. I found some useful information on this web page:  using some MS magic. I found some useful information on this web page:
123  http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the  http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
# Line 193  typedef unsigned char pcre_uint8; Line 200  typedef unsigned char pcre_uint8;
200    typedef unsigned int pcre_uint16;    typedef unsigned int pcre_uint16;
201    typedef int pcre_int16;    typedef int pcre_int16;
202  #else  #else
203    #error Cannot determine a type for 16-bit unsigned integers  # error Cannot determine a type for 16-bit unsigned integers
204  #endif  #endif
205    
206  #if UINT_MAX == 4294967295  #if UINT_MAX == 4294967295
# Line 203  typedef unsigned char pcre_uint8; Line 210  typedef unsigned char pcre_uint8;
210    typedef unsigned long int pcre_uint32;    typedef unsigned long int pcre_uint32;
211    typedef long int pcre_int32;    typedef long int pcre_int32;
212  #else  #else
213    #error Cannot determine a type for 32-bit unsigned integers  # error Cannot determine a type for 32-bit unsigned integers
214  #endif  #endif
215    
216  /* When checking for integer overflow in pcre_compile(), we need to handle  /* When checking for integer overflow in pcre_compile(), we need to handle
# Line 214  stdint.h is available, include it; it ma Line 221  stdint.h is available, include it; it ma
221  have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set  have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
222  by "configure". */  by "configure". */
223    
224  #if HAVE_STDINT_H  #if defined HAVE_STDINT_H
225  #include <stdint.h>  #include <stdint.h>
226  #elif HAVE_INTTYPES_H  #elif defined HAVE_INTTYPES_H
227  #include <inttypes.h>  #include <inttypes.h>
228  #endif  #endif
229    
# Line 243  exactly 256 items. When the character is Line 250  exactly 256 items. When the character is
250  items, some check is needed before accessing these tables.  items, some check is needed before accessing these tables.
251  */  */
252    
253  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
254    
255  typedef unsigned char pcre_uchar;  typedef unsigned char pcre_uchar;
256  #define IN_UCHARS(x) (x)  #define IN_UCHARS(x) (x)
257  #define MAX_255(c) 1  #define MAX_255(c) 1
258  #define TABLE_GET(c, table, default) ((table)[c])  #define TABLE_GET(c, table, default) ((table)[c])
259    
260  #else  #elif defined COMPILE_PCRE16
261    
 #ifdef COMPILE_PCRE16  
262  #if USHRT_MAX != 65535  #if USHRT_MAX != 65535
263  /* This is a warning message. Change PCRE_SCHAR16 to a 16 bit data type in  /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
264  pcre.h(.in) and disable (comment out) this message. */  pcre.h(.in) and disable (comment out) this message. */
265  #error Warning: PCRE_SCHAR16 is not a 16 bit data type.  #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
266  #endif  #endif
267    
268  typedef pcre_uint16 pcre_uchar;  typedef pcre_uint16 pcre_uchar;
269  #define IN_UCHARS(x) ((x) << 1)  #define UCHAR_SHIFT (1)
270    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
271    #define MAX_255(c) ((c) <= 255u)
272    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
273    
274    #elif defined COMPILE_PCRE32
275    
276    typedef pcre_uint32 pcre_uchar;
277    #define UCHAR_SHIFT (2)
278    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
279  #define MAX_255(c) ((c) <= 255u)  #define MAX_255(c) ((c) <= 255u)
280  #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))  #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
281    
282  #else  #else
283  #error Unsupported compiling mode  #error Unsupported compiling mode
284  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
285    
286  /* This is an unsigned int value that no character can ever have. UTF-8  /* This is an unsigned int value that no character can ever have. UTF-8
287  characters only go up to 0x7fffffff (though Unicode doesn't go beyond  characters only go up to 0x7fffffff (though Unicode doesn't go beyond
# Line 295  start/end of string field names are. */ Line 308  start/end of string field names are. */
308         &(NLBLOCK->nllen), utf)) \         &(NLBLOCK->nllen), utf)) \
309      : \      : \
310      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
311       (p)[0] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
312       (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
313      ) \      ) \
314    )    )
315    
# Line 309  start/end of string field names are. */ Line 322  start/end of string field names are. */
322         &(NLBLOCK->nllen), utf)) \         &(NLBLOCK->nllen), utf)) \
323      : \      : \
324      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
325       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
326       (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
327      ) \      ) \
328    )    )
329    
# Line 335  values. */ Line 348  values. */
348  #include "pcre.h"  #include "pcre.h"
349  #include "ucp.h"  #include "ucp.h"
350    
351    #ifdef COMPILE_PCRE32
352    /* Assert that the public PCRE_UCHAR32 is a 32-bit type */
353    typedef int __assert_pcre_uchar32_size[sizeof(PCRE_UCHAR32) == 4 ? 1 : -1];
354    #endif
355    
356  /* When compiling for use with the Virtual Pascal compiler, these functions  /* When compiling for use with the Virtual Pascal compiler, these functions
357  need to have their names changed. PCRE must be compiled with the -DVPCOMPAT  need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
358  option on the command line. */  option on the command line. */
# Line 396  The macros are controlled by the value o Line 414  The macros are controlled by the value o
414  the config.h file, but can be overridden by using -D on the command line. This  the config.h file, but can be overridden by using -D on the command line. This
415  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
416    
417  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
418    
419  #if LINK_SIZE == 2  #if LINK_SIZE == 2
420    
# Line 441  is automated on Unix systems via the "co Line 459  is automated on Unix systems via the "co
459  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
460  #endif  #endif
461    
462  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
   
 #ifdef COMPILE_PCRE16  
463    
464  #if LINK_SIZE == 2  #if LINK_SIZE == 2
465    
466    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
467  #undef LINK_SIZE  #undef LINK_SIZE
468  #define LINK_SIZE 1  #define LINK_SIZE 1
469    
# Line 460  is automated on Unix systems via the "co Line 477  is automated on Unix systems via the "co
477    
478  #elif LINK_SIZE == 3 || LINK_SIZE == 4  #elif LINK_SIZE == 3 || LINK_SIZE == 4
479    
480    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
481  #undef LINK_SIZE  #undef LINK_SIZE
482  #define LINK_SIZE 2  #define LINK_SIZE 2
483    
# Line 477  is automated on Unix systems via the "co Line 495  is automated on Unix systems via the "co
495  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
496  #endif  #endif
497    
498    #elif defined COMPILE_PCRE32
499    
500    /* Only supported LINK_SIZE is 4 */
501    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
502    #undef LINK_SIZE
503    #define LINK_SIZE 1
504    
505    #define PUT(a,n,d)   \
506      (a[n] = (d))
507    
508    #define GET(a,n) \
509      (a[n])
510    
511    /* Keep it positive */
512    #define MAX_PATTERN_SIZE (1 << 30)
513    
514  #else  #else
515  #error Unsupported compiling mode  #error Unsupported compiling mode
516  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
517    
518  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
519    
# Line 492  is automated on Unix systems via the "co Line 524  is automated on Unix systems via the "co
524  offsets changes. There are used for repeat counts and for other things such as  offsets changes. There are used for repeat counts and for other things such as
525  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
526    
527  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
528    
529  #define IMM2_SIZE 2  #define IMM2_SIZE 2
530    
# Line 500  capturing parenthesis numbers in back re Line 532  capturing parenthesis numbers in back re
532    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
533    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
534    
535    /* For reasons that I do not understand, the expression in this GET2 macro is
536    treated by gcc as a signed expression, even when a is declared as unsigned. It
537    seems that any kind of arithmetic results in a signed value. */
538    
539  #define GET2(a,n) \  #define GET2(a,n) \
540    (((a)[n] << 8) | (a)[(n)+1])    (unsigned int)(((a)[n] << 8) | (a)[(n)+1])
541    
542    #elif defined COMPILE_PCRE16
543    
544    #define IMM2_SIZE 1
545    
546  #else /* COMPILE_PCRE8 */  #define PUT2(a,n,d)   \
547       a[n] = d
548    
549  #ifdef COMPILE_PCRE16  #define GET2(a,n) \
550       a[n]
551    
552    #elif defined COMPILE_PCRE32
553    
554  #define IMM2_SIZE 1  #define IMM2_SIZE 1
555    
# Line 517  capturing parenthesis numbers in back re Line 561  capturing parenthesis numbers in back re
561    
562  #else  #else
563  #error Unsupported compiling mode  #error Unsupported compiling mode
564  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
565    
566  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
567    
568    /* The maximum length of a MARK name is currently one data unit; it may be
569    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
570    
571    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
572    #define MAX_MARK ((1u << 16) - 1)
573    #else
574    #define MAX_MARK ((1u << 8) - 1)
575    #endif
576    
577  /* When UTF encoding is being used, a character is no longer just a single  /* When UTF encoding is being used, a character is no longer just a single
578  character. The macros for character handling generate simple sequences when  byte. The macros for character handling generate simple sequences when used in
579  used in character-mode, and more complicated ones for UTF characters.  character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
580  GETCHARLENTEST and other macros are not used when UTF is not supported,  and other macros are not used when UTF is not supported, so they are not
581  so they are not defined. To make sure they can never even appear when  defined. To make sure they can never even appear when UTF support is omitted,
582  UTF support is omitted, we don't even define them. */  we don't even define them. */
583    
584  #ifndef SUPPORT_UTF  #ifndef SUPPORT_UTF
585    
# Line 541  UTF support is omitted, we don't even de Line 592  UTF support is omitted, we don't even de
592  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
593  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
594  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
595    #define RAWUCHAR(eptr) (*(eptr))
596    #define RAWUCHARINC(eptr) (*(eptr)++)
597    #define RAWUCHARTEST(eptr) (*(eptr))
598    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
599  /* #define GETCHARLENTEST(c, eptr, len) */  /* #define GETCHARLENTEST(c, eptr, len) */
600  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
601  /* #define FORWARDCHAR(eptr) */  /* #define FORWARDCHAR(eptr) */
# Line 548  UTF support is omitted, we don't even de Line 603  UTF support is omitted, we don't even de
603    
604  #else   /* SUPPORT_UTF */  #else   /* SUPPORT_UTF */
605    
 #ifdef COMPILE_PCRE8  
   
 /* These macros were originally written in the form of loops that used data  
 from the tables whose names start with PRIV(utf8_table). They were rewritten by  
 a user so as not to use loops, because in some environments this gives a  
 significant performance advantage, and it seems never to do any harm. */  
   
 /* Tells the biggest code point which can be encoded as a single character. */  
   
 #define MAX_VALUE_FOR_SINGLE_CHAR 127  
   
606  /* Tests whether the code point needs extra characters to decode. */  /* Tests whether the code point needs extra characters to decode. */
607    
608  #define HAS_EXTRALEN(c) ((c) >= 0xc0)  #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
   
 /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.  
 Otherwise it has an undefined behaviour. */  
   
 #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])  
   
 /* Returns TRUE, if the given character is not the first character  
 of a UTF sequence. */  
   
 #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)  
609    
610  /* Base macro to pick up the remaining bytes of a UTF-8 character, not  /* Base macro to pick up the remaining bytes of a UTF-8 character, not
611  advancing the pointer. */  advancing the pointer. */
# Line 595  advancing the pointer. */ Line 629  advancing the pointer. */
629            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
630      }      }
631    
 /* Get the next UTF-8 character, not advancing the pointer. This is called when  
 we know we are in UTF-8 mode. */  
   
 #define GETCHAR(c, eptr) \  
   c = *eptr; \  
   if (c >= 0xc0) GETUTF8(c, eptr);  
   
 /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the  
 pointer. */  
   
 #define GETCHARTEST(c, eptr) \  
   c = *eptr; \  
   if (utf && c >= 0xc0) GETUTF8(c, eptr);  
   
632  /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing  /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
633  the pointer. */  the pointer. */
634    
# Line 643  the pointer. */ Line 663  the pointer. */
663        } \        } \
664      }      }
665    
666    #if defined COMPILE_PCRE8
667    
668    /* These macros were originally written in the form of loops that used data
669    from the tables whose names start with PRIV(utf8_table). They were rewritten by
670    a user so as not to use loops, because in some environments this gives a
671    significant performance advantage, and it seems never to do any harm. */
672    
673    /* Tells the biggest code point which can be encoded as a single character. */
674    
675    #define MAX_VALUE_FOR_SINGLE_CHAR 127
676    
677    /* Tests whether the code point needs extra characters to decode. */
678    
679    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
680    
681    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
682    Otherwise it has an undefined behaviour. */
683    
684    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
685    
686    /* Returns TRUE, if the given character is not the first character
687    of a UTF sequence. */
688    
689    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
690    
691    /* Get the next UTF-8 character, not advancing the pointer. This is called when
692    we know we are in UTF-8 mode. */
693    
694    #define GETCHAR(c, eptr) \
695      c = *eptr; \
696      if (c >= 0xc0) GETUTF8(c, eptr);
697    
698    /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
699    pointer. */
700    
701    #define GETCHARTEST(c, eptr) \
702      c = *eptr; \
703      if (utf && c >= 0xc0) GETUTF8(c, eptr);
704    
705  /* Get the next UTF-8 character, advancing the pointer. This is called when we  /* Get the next UTF-8 character, advancing the pointer. This is called when we
706  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
707    
# Line 709  do not know if we are in UTF-8 mode. */ Line 768  do not know if we are in UTF-8 mode. */
768    c = *eptr; \    c = *eptr; \
769    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
770    
771    /* Returns the next uchar, not advancing the pointer. This is called when
772    we know we are in UTF mode. */
773    
774    #define RAWUCHAR(eptr) \
775      (*(eptr))
776    
777    /* Returns the next uchar, advancing the pointer. This is called when
778    we know we are in UTF mode. */
779    
780    #define RAWUCHARINC(eptr) \
781      (*((eptr)++))
782    
783    /* Returns the next uchar, testing for UTF mode, and not advancing the
784    pointer. */
785    
786    #define RAWUCHARTEST(eptr) \
787      (*(eptr))
788    
789    /* Returns the next uchar, testing for UTF mode, advancing the
790    pointer. */
791    
792    #define RAWUCHARINCTEST(eptr) \
793      (*((eptr)++))
794    
795  /* If the pointer is not at the start of a character, move it back until  /* If the pointer is not at the start of a character, move it back until
796  it is. This is called only in UTF-8 mode - we don't put a test within the macro  it is. This is called only in UTF-8 mode - we don't put a test within the macro
797  because almost all calls are already within a block of UTF-8 only code. */  because almost all calls are already within a block of UTF-8 only code. */
# Line 722  because almost all calls are already wit Line 805  because almost all calls are already wit
805  #define ACROSSCHAR(condition, eptr, action) \  #define ACROSSCHAR(condition, eptr, action) \
806    while((condition) && ((eptr) & 0xc0) == 0x80) action    while((condition) && ((eptr) & 0xc0) == 0x80) action
807    
808  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
   
 #ifdef COMPILE_PCRE16  
809    
810  /* Tells the biggest code point which can be encoded as a single character. */  /* Tells the biggest code point which can be encoded as a single character. */
811    
# Line 806  we do not know if we are in UTF-16 mode. Line 887  we do not know if we are in UTF-16 mode.
887    c = *eptr; \    c = *eptr; \
888    if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);    if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
889    
890    /* Returns the next uchar, not advancing the pointer. This is called when
891    we know we are in UTF mode. */
892    
893    #define RAWUCHAR(eptr) \
894      (*(eptr))
895    
896    /* Returns the next uchar, advancing the pointer. This is called when
897    we know we are in UTF mode. */
898    
899    #define RAWUCHARINC(eptr) \
900      (*((eptr)++))
901    
902    /* Returns the next uchar, testing for UTF mode, and not advancing the
903    pointer. */
904    
905    #define RAWUCHARTEST(eptr) \
906      (*(eptr))
907    
908    /* Returns the next uchar, testing for UTF mode, advancing the
909    pointer. */
910    
911    #define RAWUCHARINCTEST(eptr) \
912      (*((eptr)++))
913    
914  /* If the pointer is not at the start of a character, move it back until  /* If the pointer is not at the start of a character, move it back until
915  it is. This is called only in UTF-16 mode - we don't put a test within the  it is. This is called only in UTF-16 mode - we don't put a test within the
916  macro because almost all calls are already within a block of UTF-16 only  macro because almost all calls are already within a block of UTF-16 only
# Line 820  code. */ Line 925  code. */
925  #define ACROSSCHAR(condition, eptr, action) \  #define ACROSSCHAR(condition, eptr, action) \
926    if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action    if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
927    
928  #endif  #elif defined COMPILE_PCRE32
929    
930  #endif /* COMPILE_PCRE8 */  /* These are trivial for the 32-bit library, since all UTF-32 characters fit
931    into one pcre_uchar unit. */
932    #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
933    #define HAS_EXTRALEN(c) (0)
934    #define GET_EXTRALEN(c) (0)
935    #define NOT_FIRSTCHAR(c) (0)
936    
937    /* Get the next UTF-32 character, not advancing the pointer. This is called when
938    we know we are in UTF-32 mode. */
939    
940    #define GETCHAR(c, eptr) \
941      c = *(eptr);
942    
943    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
944    pointer. */
945    
946    #define GETCHARTEST(c, eptr) \
947      c = *(eptr);
948    
949    /* Get the next UTF-32 character, advancing the pointer. This is called when we
950    know we are in UTF-32 mode. */
951    
952    #define GETCHARINC(c, eptr) \
953      c = *((eptr)++);
954    
955    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
956    This is called when we don't know if we are in UTF-32 mode. */
957    
958    #define GETCHARINCTEST(c, eptr) \
959      c = *((eptr)++);
960    
961    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
962    length (since all UTF-32 is of length 1). This is called when we know we are in
963    UTF-32 mode. */
964    
965    #define GETCHARLEN(c, eptr, len) \
966      GETCHAR(c, eptr)
967    
968    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
969    pointer, not incrementing the length (since all UTF-32 is of length 1).
970    This is called when we do not know if we are in UTF-32 mode. */
971    
972    #define GETCHARLENTEST(c, eptr, len) \
973      GETCHARTEST(c, eptr)
974    
975    /* Returns the next uchar, not advancing the pointer. This is called when
976    we know we are in UTF mode. */
977    
978    #define RAWUCHAR(eptr) \
979      (*(eptr))
980    
981    /* Returns the next uchar, advancing the pointer. This is called when
982    we know we are in UTF mode. */
983    
984    #define RAWUCHARINC(eptr) \
985      (*((eptr)++))
986    
987    /* Returns the next uchar, testing for UTF mode, and not advancing the
988    pointer. */
989    
990    #define RAWUCHARTEST(eptr) \
991      (*(eptr))
992    
993    /* Returns the next uchar, testing for UTF mode, advancing the
994    pointer. */
995    
996    #define RAWUCHARINCTEST(eptr) \
997      (*((eptr)++))
998    
999    /* If the pointer is not at the start of a character, move it back until
1000    it is. This is called only in UTF-32 mode - we don't put a test within the
1001    macro because almost all calls are already within a block of UTF-32 only
1002    code.
1003    These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
1004    
1005    #define BACKCHAR(eptr) do { } while (0)
1006    
1007    /* Same as above, just in the other direction. */
1008    #define FORWARDCHAR(eptr) do { } while (0)
1009    
1010    /* Same as above, but it allows a fully customizable form. */
1011    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1012    
1013    #else
1014    #error Unsupported compiling mode
1015    #endif /* COMPILE_PCRE[8|16|32] */
1016    
1017  #endif  /* SUPPORT_UTF */  #endif  /* SUPPORT_UTF */
1018    
1019    /* Tests for Unicode horizontal and vertical whitespace characters must check a
1020    number of different values. Using a switch statement for this generates the
1021    fastest code (no loop, no memory access), and there are several places in the
1022    interpreter code where this happens. In order to ensure that all the case lists
1023    remain in step, we use macros so that there is only one place where the lists
1024    are defined.
1025    
1026    These values are also required as lists in pcre_compile.c when processing \h,
1027    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1028    macros that define the values are here so that all the definitions are
1029    together. The lists must be in ascending character order, terminated by
1030    NOTACHAR (which is 0xffffffff).
1031    
1032    Any changes should ensure that the various macros are kept in step with each
1033    other. NOTE: The values also appear in pcre_jit_compile.c. */
1034    
1035    /* ------ ASCII/Unicode environments ------ */
1036    
1037    #ifndef EBCDIC
1038    
1039    #define HSPACE_LIST \
1040      CHAR_HT, CHAR_SPACE, 0xa0, \
1041      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1042      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1043      NOTACHAR
1044    
1045    #define HSPACE_MULTIBYTE_CASES \
1046      case 0x1680:  /* OGHAM SPACE MARK */ \
1047      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1048      case 0x2000:  /* EN QUAD */ \
1049      case 0x2001:  /* EM QUAD */ \
1050      case 0x2002:  /* EN SPACE */ \
1051      case 0x2003:  /* EM SPACE */ \
1052      case 0x2004:  /* THREE-PER-EM SPACE */ \
1053      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1054      case 0x2006:  /* SIX-PER-EM SPACE */ \
1055      case 0x2007:  /* FIGURE SPACE */ \
1056      case 0x2008:  /* PUNCTUATION SPACE */ \
1057      case 0x2009:  /* THIN SPACE */ \
1058      case 0x200A:  /* HAIR SPACE */ \
1059      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1060      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1061      case 0x3000   /* IDEOGRAPHIC SPACE */
1062    
1063    #define HSPACE_BYTE_CASES \
1064      case CHAR_HT: \
1065      case CHAR_SPACE: \
1066      case 0xa0     /* NBSP */
1067    
1068    #define HSPACE_CASES \
1069      HSPACE_BYTE_CASES: \
1070      HSPACE_MULTIBYTE_CASES
1071    
1072    #define VSPACE_LIST \
1073      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1074    
1075    #define VSPACE_MULTIBYTE_CASES \
1076      case 0x2028:    /* LINE SEPARATOR */ \
1077      case 0x2029     /* PARAGRAPH SEPARATOR */
1078    
1079    #define VSPACE_BYTE_CASES \
1080      case CHAR_LF: \
1081      case CHAR_VT: \
1082      case CHAR_FF: \
1083      case CHAR_CR: \
1084      case CHAR_NEL
1085    
1086    #define VSPACE_CASES \
1087      VSPACE_BYTE_CASES: \
1088      VSPACE_MULTIBYTE_CASES
1089    
1090  /* In case there is no definition of offsetof() provided - though any proper  /* ------ EBCDIC environments ------ */
 Standard C system should have one. */  
1091    
1092  #ifndef offsetof  #else
1093  #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1094    
1095    #define HSPACE_BYTE_CASES \
1096      case CHAR_HT: \
1097      case CHAR_SPACE
1098    
1099    #define HSPACE_CASES HSPACE_BYTE_CASES
1100    
1101    #ifdef EBCDIC_NL25
1102    #define VSPACE_LIST \
1103      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1104    #else
1105    #define VSPACE_LIST \
1106      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1107  #endif  #endif
1108    
1109    #define VSPACE_BYTE_CASES \
1110      case CHAR_LF: \
1111      case CHAR_VT: \
1112      case CHAR_FF: \
1113      case CHAR_CR: \
1114      case CHAR_NEL
1115    
1116    #define VSPACE_CASES VSPACE_BYTE_CASES
1117    #endif  /* EBCDIC */
1118    
1119    /* ------ End of whitespace macros ------ */
1120    
1121    
1122    
1123  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
1124  live at the top end of the options word, but that got almost full, so now they  live at the top end of the options word, but that got almost full, so now they
# Line 841  are in a 16-bit flags word. From release Line 1126  are in a 16-bit flags word. From release
1126  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1127  compatibility. */  compatibility. */
1128    
1129  #ifdef COMPILE_PCRE8  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1130  #define PCRE_MODE          0x0001  /* compiled in 8 bit mode */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1131  #endif  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
 #ifdef COMPILE_PCRE16  
 #define PCRE_MODE          0x0002  /* compiled in 16 bit mode */  
 #endif  
1132  #define PCRE_FIRSTSET      0x0010  /* first_char is set */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1133  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1134  #define PCRE_REQCHSET      0x0040  /* req_byte is set */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
# Line 857  compatibility. */ Line 1139  compatibility. */
1139  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1140  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1141    
1142    #if defined COMPILE_PCRE8
1143    #define PCRE_MODE          PCRE_MODE8
1144    #elif defined COMPILE_PCRE16
1145    #define PCRE_MODE          PCRE_MODE16
1146    #elif defined COMPILE_PCRE32
1147    #define PCRE_MODE          PCRE_MODE32
1148    #endif
1149    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1150    
1151  /* Flags for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1152    
1153  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
# Line 873  time, run time, or study time, respectiv Line 1164  time, run time, or study time, respectiv
1164     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1165     PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \     PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
1166     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1167     PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)     PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE|PCRE_NEVER_UTF)
1168    
1169  #define PUBLIC_EXEC_OPTIONS \  #define PUBLIC_EXEC_OPTIONS \
1170    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
# Line 887  time, run time, or study time, respectiv Line 1178  time, run time, or study time, respectiv
1178     PCRE_NO_START_OPTIMIZE)     PCRE_NO_START_OPTIMIZE)
1179    
1180  #define PUBLIC_STUDY_OPTIONS \  #define PUBLIC_STUDY_OPTIONS \
1181     PCRE_STUDY_JIT_COMPILE     (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1182        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1183    
1184    #define PUBLIC_JIT_EXEC_OPTIONS \
1185       (PCRE_NO_UTF8_CHECK|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|\
1186        PCRE_NOTEMPTY_ATSTART|PCRE_PARTIAL_SOFT|PCRE_PARTIAL_HARD)
1187    
1188  /* Magic number to provide a small check against being handed junk. */  /* Magic number to provide a small check against being handed junk. */
1189    
# Line 898  in different endianness. */ Line 1194  in different endianness. */
1194    
1195  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
1196    
 /* Negative values for the firstchar and reqchar variables */  
   
 #define REQ_UNSET (-2)  
 #define REQ_NONE  (-1)  
   
1197  /* The maximum remaining length of subject we are prepared to search for a  /* The maximum remaining length of subject we are prepared to search for a
1198  req_byte match. */  req_byte match. */
1199    
# Line 939  macros to give the functions distinct na Line 1230  macros to give the functions distinct na
1230  #ifndef SUPPORT_UTF  #ifndef SUPPORT_UTF
1231    
1232  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1233  so that PCRE works on both ASCII and EBCDIC platforms, in non-UTF-mode only. */  so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1234    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1235    characters, a common practice has been to use its NL (0x15) character as the
1236    line terminator in C-like processing environments. However, sometimes the LF
1237    (0x25) character is used instead, according to this Unicode document:
1238    
1239    http://unicode.org/standard/reports/tr13/tr13-5.html
1240    
1241    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1242    instead. Whichever is *not* chosen is defined as NEL.
1243    
1244    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1245    same code point. */
1246    
1247    #ifdef EBCDIC
1248    
1249    #ifndef EBCDIC_NL25
1250    #define CHAR_NL                     '\x15'
1251    #define CHAR_NEL                    '\x25'
1252    #define STR_NL                      "\x15"
1253    #define STR_NEL                     "\x25"
1254    #else
1255    #define CHAR_NL                     '\x25'
1256    #define CHAR_NEL                    '\x15'
1257    #define STR_NL                      "\x25"
1258    #define STR_NEL                     "\x15"
1259    #endif
1260    
1261    #define CHAR_LF                     CHAR_NL
1262    #define STR_LF                      STR_NL
1263    
1264    #define CHAR_ESC                    '\047'
1265    #define CHAR_DEL                    '\007'
1266    #define STR_ESC                     "\047"
1267    #define STR_DEL                     "\007"
1268    
1269    #else  /* Not EBCDIC */
1270    
1271    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1272    compatibility. NEL is the Unicode newline character; make sure it is
1273    a positive value. */
1274    
1275    #define CHAR_LF                     '\n'
1276    #define CHAR_NL                     CHAR_LF
1277    #define CHAR_NEL                    ((unsigned char)'\x85')
1278    #define CHAR_ESC                    '\033'
1279    #define CHAR_DEL                    '\177'
1280    
1281    #define STR_LF                      "\n"
1282    #define STR_NL                      STR_LF
1283    #define STR_NEL                     "\x85"
1284    #define STR_ESC                     "\033"
1285    #define STR_DEL                     "\177"
1286    
1287    #endif  /* EBCDIC */
1288    
1289    /* The remaining definitions work in both environments. */
1290    
1291    #define CHAR_NULL                   '\0'
1292  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1293  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1294  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1295  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1296  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1297  #define CHAR_BEL                    '\a'  #define CHAR_BEL                    '\a'
 #ifdef EBCDIC  
 #define CHAR_ESC                    '\047'  
 #define CHAR_DEL                    '\007'  
 #else  
 #define CHAR_ESC                    '\033'  
 #define CHAR_DEL                    '\177'  
 #endif  
1298    
1299  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1300  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 1056  so that PCRE works on both ASCII and EBC Line 1396  so that PCRE works on both ASCII and EBC
1396  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1397  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1398  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1399  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1400  #define STR_BEL                     "\a"  #define STR_BEL                     "\a"
 #ifdef EBCDIC  
 #define STR_ESC                     "\047"  
 #define STR_DEL                     "\007"  
 #else  
 #define STR_ESC                     "\033"  
 #define STR_DEL                     "\177"  
 #endif  
1401    
1402  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1403  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 1196  so that PCRE works on both ASCII and EBC Line 1528  so that PCRE works on both ASCII and EBC
1528  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1529  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1530  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1531  #ifdef COMPILE_PCRE8  #define STRING_UTF8_RIGHTPAR           "UTF8)"
1532  #define STRING_UTF_RIGHTPAR            "UTF8)"  #define STRING_UTF16_RIGHTPAR          "UTF16)"
1533  #endif  #define STRING_UTF32_RIGHTPAR          "UTF32)"
1534  #ifdef COMPILE_PCRE16  #define STRING_UTF_RIGHTPAR            "UTF)"
 #define STRING_UTF_RIGHTPAR            "UTF16)"  
 #endif  
1535  #define STRING_UCP_RIGHTPAR            "UCP)"  #define STRING_UCP_RIGHTPAR            "UCP)"
1536  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1537    
# Line 1215  only. */ Line 1545  only. */
1545  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1546  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1547  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1548  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1549    #define CHAR_NL                     CHAR_LF
1550    #define CHAR_NEL                    ((unsigned char)'\x85')
1551  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1552  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1553  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
1554  #define CHAR_DEL                    '\177'  #define CHAR_DEL                    '\177'
1555    
1556    #define CHAR_NULL                   '\0'
1557  #define CHAR_SPACE                  '\040'  #define CHAR_SPACE                  '\040'
1558  #define CHAR_EXCLAMATION_MARK       '\041'  #define CHAR_EXCLAMATION_MARK       '\041'
1559  #define CHAR_QUOTATION_MARK         '\042'  #define CHAR_QUOTATION_MARK         '\042'
# Line 1456  only. */ Line 1789  only. */
1789  #define STRING_ANYCRLF_RIGHTPAR        STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS  #define STRING_ANYCRLF_RIGHTPAR        STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1790  #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  #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
1791  #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  #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
1792  #ifdef COMPILE_PCRE8  #define STRING_UTF8_RIGHTPAR           STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1793  #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS  #define STRING_UTF16_RIGHTPAR          STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1794  #endif  #define STRING_UTF32_RIGHTPAR          STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1795  #ifdef COMPILE_PCRE16  #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_RIGHT_PARENTHESIS
 #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS  
 #endif  
1796  #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS  #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1797  #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  #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
1798    
# Line 1478  only. */ Line 1809  only. */
1809  #endif  #endif
1810    
1811  #ifndef ESC_n  #ifndef ESC_n
1812  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1813  #endif  #endif
1814    
1815  #ifndef ESC_r  #ifndef ESC_r
# Line 1503  only. */ Line 1834  only. */
1834  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1835  #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */  #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1836  #define PT_WORD       8    /* Word - L plus N plus underscore */  #define PT_WORD       8    /* Word - L plus N plus underscore */
1837    #define PT_CLIST      9    /* Pseudo-property: match character list */
1838    #define PT_UCNC      10    /* Universal Character nameable character */
1839    
1840  /* Flag bits and data types for the extended class (OP_XCLASS) for classes that  /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1841  contain characters with values greater than 255. */  contain characters with values greater than 255. */
# Line 1518  contain characters with values greater t Line 1851  contain characters with values greater t
1851    
1852  /* These are escaped items that aren't just an encoding of a particular data  /* These are escaped items that aren't just an encoding of a particular data
1853  value such as \n. They must have non-zero values, as check_escape() returns  value such as \n. They must have non-zero values, as check_escape() returns
1854  their negation. Also, they must appear in the same order as in the opcode  0 for a data character.  Also, they must appear in the same order as in the opcode
1855  definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it  definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1856  corresponds to "." in DOTALL mode rather than an escape sequence. It is also  corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1857  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1858  non-DOTALL mode, "." behaves like \N.  non-DOTALL mode, "." behaves like \N.
1859    
1860  The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.  The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1861  when PCRE_UCP is set, when replacement of \d etc by \p sequences is required.  when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1862  They must be contiguous, and remain in order so that the replacements can be  They must be contiguous, and remain in order so that the replacements can be
1863  looked up from a table.  looked up from a table.
1864    
1865  The final escape must be ESC_REF as subsequent values are used for  Negative numbers are used to encode a backreference (\1, \2, \3, etc.) in
1866  backreferences (\1, \2, \3, etc). There are two tests in the code for an escape  check_escape(). There are two tests in the code for an escape
1867  greater than ESC_b and less than ESC_Z to detect the types that may be  greater than ESC_b and less than ESC_Z to detect the types that may be
1868  repeated. These are the types that consume characters. If any new escapes are  repeated. These are the types that consume characters. If any new escapes are
1869  put in between that don't consume a character, that code will have to change.  put in between that don't consume a character, that code will have to change.
# Line 1540  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E Line 1873  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E
1873         ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,         ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1874         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1875         ESC_E, ESC_Q, ESC_g, ESC_k,         ESC_E, ESC_Q, ESC_g, ESC_k,
1876         ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu,         ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
        ESC_REF };  
1877    
1878  /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to  /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1879  OP_EOD must correspond in order to the list of escapes immediately above.  OP_EOD must correspond in order to the list of escapes immediately above.
# Line 1567  enum { Line 1899  enum {
1899    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1900    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1901    
1902    OP_ANY,            /* 12 Match any character except newline */    OP_ANY,            /* 12 Match any character except newline (\N) */
1903    OP_ALLANY,         /* 13 Match any character */    OP_ALLANY,         /* 13 Match any character */
1904    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1905    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
# Line 1578  enum { Line 1910  enum {
1910    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1911    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1912    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1913    OP_EODN,           /* 23 End of data or \n at end of data: \Z. */    OP_EODN,           /* 23 End of data or \n at end of data (\Z) */
1914    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1915    
1916    OP_CIRC,           /* 25 Start of line - not multiline */    OP_CIRC,           /* 25 Start of line - not multiline */
1917    OP_CIRCM,          /* 26 Start of line - multiline */    OP_CIRCM,          /* 26 Start of line - multiline */
# Line 1939  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2271  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2271         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2272         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2273         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2274         ERR70, ERR71, ERR72, ERR73, ERR74, ERRCOUNT };         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERR78, ERRCOUNT };
2275    
2276    /* JIT compiling modes. The function list is indexed by them. */
2277    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2278           JIT_NUMBER_OF_COMPILE_MODES };
2279    
2280  /* The real format of the start of the pcre block; the index of names and the  /* The real format of the start of the pcre block; the index of names and the
2281  code vector run on as long as necessary after the end. We store an explicit  code vector run on as long as necessary after the end. We store an explicit
# Line 1958  fields are present. Currently PCRE alway Line 2294  fields are present. Currently PCRE alway
2294  NOTE NOTE NOTE  NOTE NOTE NOTE
2295  */  */
2296    
2297  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2298    #define REAL_PCRE real_pcre
2299    #elif defined COMPILE_PCRE16
2300    #define REAL_PCRE real_pcre16
2301    #elif defined COMPILE_PCRE32
2302    #define REAL_PCRE real_pcre32
2303    #endif
2304    
2305    /* It is necessary to fork the struct for 32 bit, since it needs to use
2306     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2307     * typedef since pcretest needs access to  the struct of the 8-, 16-
2308     * and 32-bit variants. */
2309    
2310    typedef struct real_pcre8_or_16 {
2311    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2312    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2313    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
2314    pcre_uint16 flags;              /* Private flags */    pcre_uint16 flags;              /* Private flags */
2315    pcre_uint16 dummy1;             /* For future use */    pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2316    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2317    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2318    pcre_uint16 first_char;         /* Starting character */    pcre_uint16 first_char;         /* Starting character */
2319    pcre_uint16 req_char;           /* This character must be seen */    pcre_uint16 req_char;           /* This character must be seen */
2320    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2321    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2322    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2323    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
   
2324    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2325    const pcre_uint8 *nullpad;      /* NULL padding */    const pcre_uint8 *nullpad;      /* NULL padding */
2326  } real_pcre;  } real_pcre8_or_16;
2327    
2328    typedef struct real_pcre8_or_16 real_pcre;
2329    typedef struct real_pcre8_or_16 real_pcre16;
2330    
2331    typedef struct real_pcre32 {
2332      pcre_uint32 magic_number;
2333      pcre_uint32 size;               /* Total that was malloced */
2334      pcre_uint32 options;            /* Public options */
2335      pcre_uint16 flags;              /* Private flags */
2336      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2337      pcre_uint16 top_bracket;        /* Highest numbered group */
2338      pcre_uint16 top_backref;        /* Highest numbered back reference */
2339      pcre_uint32 first_char;         /* Starting character */
2340      pcre_uint32 req_char;           /* This character must be seen */
2341      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2342      pcre_uint16 name_entry_size;    /* Size of any name items */
2343      pcre_uint16 name_count;         /* Number of name items */
2344      pcre_uint16 ref_count;          /* Reference count */
2345      pcre_uint16 dummy1;             /* for later expansion */
2346      pcre_uint16 dummy2;             /* for later expansion */
2347      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2348      void *nullpad;                  /* for later expansion */
2349    } real_pcre32;
2350    
2351    /* Assert that the size of REAL_PCRE is divisible by 8 */
2352    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2353    
2354    /* Needed in pcretest to access some fields in the real_pcre* structures
2355     * directly. They're unified for 8/16/32 bits since the structs only differ
2356     * after these fields; if that ever changes, need to fork those defines into
2357     * 8/16 and 32 bit versions. */
2358    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2359    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2360    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2361    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2362    
2363  /* The format of the block used to store data from pcre_study(). The same  /* The format of the block used to store data from pcre_study(). The same
2364  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 2016  typedef struct compile_data { Line 2399  typedef struct compile_data {
2399    int  names_found;                 /* Number of entries so far */    int  names_found;                 /* Number of entries so far */
2400    int  name_entry_size;             /* Size of each entry */    int  name_entry_size;             /* Size of each entry */
2401    int  workspace_size;              /* Size of workspace */    int  workspace_size;              /* Size of workspace */
2402    int  bracount;                    /* Count of capturing parens as we compile */    unsigned int  bracount;           /* Count of capturing parens as we compile */
2403    int  final_bracount;              /* Saved value after first pass */    int  final_bracount;              /* Saved value after first pass */
2404      int  max_lookbehind;              /* Maximum lookbehind (characters) */
2405    int  top_backref;                 /* Maximum back reference */    int  top_backref;                 /* Maximum back reference */
2406    unsigned int backref_map;         /* Bitmap of low back refs */    unsigned int backref_map;         /* Bitmap of low back refs */
2407    int  assert_depth;                /* Depth of nested assertions */    int  assert_depth;                /* Depth of nested assertions */
# Line 2025  typedef struct compile_data { Line 2409  typedef struct compile_data {
2409    int  external_flags;              /* External flag bits to be set */    int  external_flags;              /* External flag bits to be set */
2410    int  req_varyopt;                 /* "After variable item" flag for reqbyte */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2411    BOOL had_accept;                  /* (*ACCEPT) encountered */    BOOL had_accept;                  /* (*ACCEPT) encountered */
2412      BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2413    BOOL check_lookbehind;            /* Lookbehinds need later checking */    BOOL check_lookbehind;            /* Lookbehinds need later checking */
2414    int  nltype;                      /* Newline type */    int  nltype;                      /* Newline type */
2415    int  nllen;                       /* Newline string length */    int  nllen;                       /* Newline string length */
# Line 2044  call within the pattern; used by pcre_ex Line 2429  call within the pattern; used by pcre_ex
2429    
2430  typedef struct recursion_info {  typedef struct recursion_info {
2431    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2432    int group_num;                  /* Number of group that was called */    unsigned int group_num;         /* Number of group that was called */
2433    int *offset_save;               /* Pointer to start of saved offsets */    int *offset_save;               /* Pointer to start of saved offsets */
2434    int saved_max;                  /* Number of saved offsets */    int saved_max;                  /* Number of saved offsets */
2435      int saved_capture_last;         /* Last capture number */
2436    PCRE_PUCHAR subject_position;   /* Position at start of recursion */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
2437  } recursion_info;  } recursion_info;
2438    
# Line 2083  typedef struct match_data { Line 2469  typedef struct match_data {
2469    int    nllen;                   /* Newline string length */    int    nllen;                   /* Newline string length */
2470    int    name_count;              /* Number of names in name table */    int    name_count;              /* Number of names in name table */
2471    int    name_entry_size;         /* Size of entry in names table */    int    name_entry_size;         /* Size of entry in names table */
2472      unsigned int skip_arg_count;    /* For counting SKIP_ARGs */
2473      unsigned int ignore_skip_arg;   /* For re-run when SKIP arg name not found */
2474    pcre_uchar *name_table;         /* Table of names */    pcre_uchar *name_table;         /* Table of names */
2475    pcre_uchar nl[4];               /* Newline string when fixed */    pcre_uchar nl[4];               /* Newline string when fixed */
2476    const  pcre_uint8 *lcc;         /* Points to lower casing table */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2477    const  pcre_uint8 *fcc;         /* Points to case-flipping table */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2478    const  pcre_uint8 *ctypes;      /* Points to table of type maps */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
   BOOL   offset_overflow;         /* Set if too many extractions */  
2479    BOOL   notbol;                  /* NOTBOL flag */    BOOL   notbol;                  /* NOTBOL flag */
2480    BOOL   noteol;                  /* NOTEOL flag */    BOOL   noteol;                  /* NOTEOL flag */
2481    BOOL   utf;                     /* UTF-8 / UTF-16 flag */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
# Line 2100  typedef struct match_data { Line 2487  typedef struct match_data {
2487    BOOL   hitend;                  /* Hit the end of the subject at some point */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2488    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2489    BOOL   hasthen;                 /* Pattern contains (*THEN) */    BOOL   hasthen;                 /* Pattern contains (*THEN) */
   BOOL   ignore_skip_arg;         /* For re-run when SKIP name not found */  
2490    const  pcre_uchar *start_code;  /* For use when recursing */    const  pcre_uchar *start_code;  /* For use when recursing */
2491    PCRE_PUCHAR start_subject;      /* Start of the subject string */    PCRE_PUCHAR start_subject;      /* Start of the subject string */
2492    PCRE_PUCHAR end_subject;        /* End of the subject string */    PCRE_PUCHAR end_subject;        /* End of the subject string */
# Line 2109  typedef struct match_data { Line 2495  typedef struct match_data {
2495    PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */    PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2496    int    partial;                 /* PARTIAL options */    int    partial;                 /* PARTIAL options */
2497    int    end_offset_top;          /* Highwater mark at end of match */    int    end_offset_top;          /* Highwater mark at end of match */
2498    int    capture_last;            /* Most recent capture number */    pcre_int32 capture_last;        /* Most recent capture number + overflow flag */
2499    int    start_offset;            /* The start offset value */    int    start_offset;            /* The start offset value */
2500    int    match_function_type;     /* Set for certain special calls of MATCH() */    int    match_function_type;     /* Set for certain special calls of MATCH() */
2501    eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */    eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
# Line 2119  typedef struct match_data { Line 2505  typedef struct match_data {
2505    const  pcre_uchar *mark;        /* Mark pointer to pass back on success */    const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2506    const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */    const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2507    const  pcre_uchar *once_target; /* Where to back up to for atomic groups */    const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2508    #ifdef NO_RECURSE
2509      void  *match_frames_base;       /* For remembering malloc'd frames */
2510    #endif
2511  } match_data;  } match_data;
2512    
2513  /* A similar structure is used for the same purpose by the DFA matching  /* A similar structure is used for the same purpose by the DFA matching
# Line 2173  total length. */ Line 2562  total length. */
2562  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2563  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2564    
2565  /* Internal function prefix */  /* Internal function and data prefixes. */
2566    
2567  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2568    #ifndef PUBL
2569  #define PUBL(name) pcre_##name  #define PUBL(name) pcre_##name
2570    #endif
2571    #ifndef PRIV
2572  #define PRIV(name) _pcre_##name  #define PRIV(name) _pcre_##name
2573  #else  #endif
2574  #ifdef COMPILE_PCRE16  #elif defined COMPILE_PCRE16
2575    #ifndef PUBL
2576  #define PUBL(name) pcre16_##name  #define PUBL(name) pcre16_##name
2577    #endif
2578    #ifndef PRIV
2579  #define PRIV(name) _pcre16_##name  #define PRIV(name) _pcre16_##name
2580    #endif
2581    #elif defined COMPILE_PCRE32
2582    #ifndef PUBL
2583    #define PUBL(name) pcre32_##name
2584    #endif
2585    #ifndef PRIV
2586    #define PRIV(name) _pcre32_##name
2587    #endif
2588  #else  #else
2589  #error Unsupported compiling mode  #error Unsupported compiling mode
2590  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
 #endif /* COMPILE_PCRE8 */  
2591    
2592  /* Layout of the UCP type table that translates property names into types and  /* Layout of the UCP type table that translates property names into types and
2593  codes. Each entry used to point directly to a name, but to reduce the number of  codes. Each entry used to point directly to a name, but to reduce the number of
# Line 2205  but are not part of the PCRE public API. Line 2607  but are not part of the PCRE public API.
2607  pcre_tables.c module. */  pcre_tables.c module. */
2608    
2609  #ifdef COMPILE_PCRE8  #ifdef COMPILE_PCRE8
   
2610  extern const int            PRIV(utf8_table1)[];  extern const int            PRIV(utf8_table1)[];
2611  extern const int            PRIV(utf8_table1_size);  extern const int            PRIV(utf8_table1_size);
2612  extern const int            PRIV(utf8_table2)[];  extern const int            PRIV(utf8_table2)[];
2613  extern const int            PRIV(utf8_table3)[];  extern const int            PRIV(utf8_table3)[];
2614  extern const pcre_uint8     PRIV(utf8_table4)[];  extern const pcre_uint8     PRIV(utf8_table4)[];
   
2615  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE8 */
2616    
2617  extern const char           PRIV(utt_names)[];  extern const char           PRIV(utt_names)[];
2618  extern const ucp_type_table PRIV(utt)[];  extern const ucp_type_table PRIV(utt)[];
2619  extern const int            PRIV(utt_size);  extern const int            PRIV(utt_size);
2620    
2621    extern const pcre_uint8     PRIV(OP_lengths)[];
2622  extern const pcre_uint8     PRIV(default_tables)[];  extern const pcre_uint8     PRIV(default_tables)[];
2623    
2624  extern const pcre_uint8     PRIV(OP_lengths)[];  extern const pcre_uint32    PRIV(hspace_list)[];
2625    extern const pcre_uint32    PRIV(vspace_list)[];
2626    
2627    
2628  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
# Line 2228  one of the exported public functions. Th Line 2630  one of the exported public functions. Th
2630  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2631    
2632  /* String comparison functions. */  /* String comparison functions. */
2633  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2634    
2635  #define STRCMP_UC_UC(str1, str2) \  #define STRCMP_UC_UC(str1, str2) \
2636    strcmp((char *)(str1), (char *)(str2))    strcmp((char *)(str1), (char *)(str2))
# Line 2240  sense, but are not part of the PCRE publ Line 2642  sense, but are not part of the PCRE publ
2642    strncmp((char *)(str1), (str2), (num))    strncmp((char *)(str1), (str2), (num))
2643  #define STRLEN_UC(str) strlen((const char *)str)  #define STRLEN_UC(str) strlen((const char *)str)
2644    
2645  #else  #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2646    
2647  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2648                             const pcre_uchar *);                             const pcre_uchar *);
# Line 2262  extern unsigned int      PRIV(strlen_uc) Line 2664  extern unsigned int      PRIV(strlen_uc)
2664    PRIV(strncmp_uc_c8)((str1), (str2), (num))    PRIV(strncmp_uc_c8)((str1), (str2), (num))
2665  #define STRLEN_UC(str) PRIV(strlen_uc)(str)  #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2666    
2667  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE[8|16|32] */
2668    
2669    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2670    
2671    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2672    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2673    
2674    #elif defined COMPILE_PCRE32
2675    
2676    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2677                               const pcre_uchar *);
2678    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2679                               const char *);
2680    
2681    #define STRCMP_UC_UC_TEST(str1, str2) \
2682      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2683    #define STRCMP_UC_C8_TEST(str1, str2) \
2684      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2685    
2686    #endif /* COMPILE_PCRE[8|16|32] */
2687    
2688  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2689  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2690                             int *, BOOL);                             int *, BOOL);
2691  extern int               PRIV(ord2utf)(pcre_uint32, pcre_uchar *);  extern unsigned int      PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2692  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2693  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2694                             int *, BOOL);                             int *, BOOL);
2695  extern BOOL              PRIV(xclass)(int, const pcre_uchar *, BOOL);  extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2696    
2697  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2698  extern void              PRIV(jit_compile)(const real_pcre *, PUBL(extra) *);  extern void              PRIV(jit_compile)(const REAL_PCRE *,
2699  extern int               PRIV(jit_exec)(const real_pcre *, void *,                             PUBL(extra) *, int);
2700                             const pcre_uchar *, int, int, int, int, int *, int);  extern int               PRIV(jit_exec)(const PUBL(extra) *,
2701                               const pcre_uchar *, int, int, int, int *, int);
2702  extern void              PRIV(jit_free)(void *);  extern void              PRIV(jit_free)(void *);
2703  extern int               PRIV(jit_get_size)(void *);  extern int               PRIV(jit_get_size)(void *);
2704    extern const char*       PRIV(jit_get_target)(void);
2705  #endif  #endif
2706    
2707  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
2708    
2709  typedef struct {  typedef struct {
2710    pcre_uint8 script;    pcre_uint8 script;     /* ucp_Arabic, etc. */
2711    pcre_uint8 chartype;    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2712    pcre_int32 other_case;    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2713      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2714      pcre_int32 other_case; /* offset to other case, or zero if none */
2715  } ucd_record;  } ucd_record;
2716    
2717    extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2718  extern const ucd_record  PRIV(ucd_records)[];  extern const ucd_record  PRIV(ucd_records)[];
2719  extern const pcre_uint8  PRIV(ucd_stage1)[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2720  extern const pcre_uint16 PRIV(ucd_stage2)[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2721  extern const int         PRIV(ucp_gentype)[];  extern const pcre_uint32 PRIV(ucp_gentype)[];
2722    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2723  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2724  extern const int         PRIV(ucp_typerange)[];  extern const int         PRIV(ucp_typerange)[];
2725  #endif  #endif
# Line 2302  extern const int         PRIV(ucp_typera Line 2729  extern const int         PRIV(ucp_typera
2729    
2730  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2731  #define GET_UCD(ch) (PRIV(ucd_records) + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2732          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2733          UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2734    
2735  #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2736  #define UCD_SCRIPT(ch)    GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2737  #define UCD_CATEGORY(ch)  PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2738  #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2739    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2740    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2741    
2742  #endif /* SUPPORT_UCP */  #endif /* SUPPORT_UCP */
2743    

Legend:
Removed from v.850  
changed lines
  Added in v.1309

  ViewVC Help
Powered by ViewVC 1.1.5