/[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 932 by ph10, Fri Feb 24 18:54:43 2012 UTC revision 1100 by chpe, Tue Oct 16 15:56:26 2012 UTC
# 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  #if !defined COMPILE_PCRE16 && ! defined COMPILE_PCRE32
57  #define COMPILE_PCRE8  #define COMPILE_PCRE8
58  #endif  #endif
59    
# Line 78  Until then we define it if SUPPORT_UTF i Line 78  Until then we define it if SUPPORT_UTF i
78  #define SUPPORT_UTF8 1  #define SUPPORT_UTF8 1
79  #endif  #endif
80    
81  /* 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"
82  script prevents both being selected, but not everybody uses "configure". */  script prevents both being selected, but not everybody uses "configure". */
83    
84  #if defined EBCDIC && defined SUPPORT_UTF  #if defined EBCDIC && defined SUPPORT_UTF
85  #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.
86  #endif  #endif
87    
88  /* 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 243  exactly 256 items. When the character is Line 243  exactly 256 items. When the character is
243  items, some check is needed before accessing these tables.  items, some check is needed before accessing these tables.
244  */  */
245    
246  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
247    
248  typedef unsigned char pcre_uchar;  typedef unsigned char pcre_uchar;
249  #define IN_UCHARS(x) (x)  #define IN_UCHARS(x) (x)
250  #define MAX_255(c) 1  #define MAX_255(c) 1
251  #define TABLE_GET(c, table, default) ((table)[c])  #define TABLE_GET(c, table, default) ((table)[c])
252    
253  #else  #elif defined COMPILE_PCRE16
254    
 #ifdef COMPILE_PCRE16  
255  #if USHRT_MAX != 65535  #if USHRT_MAX != 65535
256  /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in  /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
257  pcre.h(.in) and disable (comment out) this message. */  pcre.h(.in) and disable (comment out) this message. */
# Line 260  pcre.h(.in) and disable (comment out) th Line 259  pcre.h(.in) and disable (comment out) th
259  #endif  #endif
260    
261  typedef pcre_uint16 pcre_uchar;  typedef pcre_uint16 pcre_uchar;
262  #define IN_UCHARS(x) ((x) << 1)  #define UCHAR_SHIFT (1)
263    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
264  #define MAX_255(c) ((c) <= 255u)  #define MAX_255(c) ((c) <= 255u)
265  #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))  #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
266    
267    #elif defined COMPILE_PCRE32
268    
269    typedef pcre_uint32 pcre_uchar;
270    #define UCHAR_SHIFT (2)
271    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
272    #define MAX_255(c) ((c) <= 255u)
273    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
274    
275    /* Assert that pcre_uchar32 is a 32-bit type */
276    typedef int __assert_pcre_uchar32_size[sizeof(pcre_uchar) == 4 ? 1 : -1];
277    
278  #else  #else
279  #error Unsupported compiling mode  #error Unsupported compiling mode
280  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
281    
282  /* 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
283  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 304  start/end of string field names are. */
304         &(NLBLOCK->nllen), utf)) \         &(NLBLOCK->nllen), utf)) \
305      : \      : \
306      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
307       (p)[0] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
308       (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
309      ) \      ) \
310    )    )
311    
# Line 309  start/end of string field names are. */ Line 318  start/end of string field names are. */
318         &(NLBLOCK->nllen), utf)) \         &(NLBLOCK->nllen), utf)) \
319      : \      : \
320      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
321       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
322       (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
323      ) \      ) \
324    )    )
325    
# Line 396  The macros are controlled by the value o Line 405  The macros are controlled by the value o
405  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
406  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
407    
408  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
409    
410  #if LINK_SIZE == 2  #if LINK_SIZE == 2
411    
# Line 441  is automated on Unix systems via the "co Line 450  is automated on Unix systems via the "co
450  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
451  #endif  #endif
452    
453  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
   
 #ifdef COMPILE_PCRE16  
454    
455  #if LINK_SIZE == 2  #if LINK_SIZE == 2
456    
457    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
458  #undef LINK_SIZE  #undef LINK_SIZE
459  #define LINK_SIZE 1  #define LINK_SIZE 1
460    
# Line 460  is automated on Unix systems via the "co Line 468  is automated on Unix systems via the "co
468    
469  #elif LINK_SIZE == 3 || LINK_SIZE == 4  #elif LINK_SIZE == 3 || LINK_SIZE == 4
470    
471    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
472  #undef LINK_SIZE  #undef LINK_SIZE
473  #define LINK_SIZE 2  #define LINK_SIZE 2
474    
# Line 477  is automated on Unix systems via the "co Line 486  is automated on Unix systems via the "co
486  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
487  #endif  #endif
488    
489    #elif defined COMPILE_PCRE32
490    
491    /* Only supported LINK_SIZE is 4 */
492    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
493    #undef LINK_SIZE
494    #define LINK_SIZE 1
495    
496    #define PUT(a,n,d)   \
497      (a[n] = (d))
498    
499    #define GET(a,n) \
500      (a[n])
501    
502    /* Keep it positive */
503    #define MAX_PATTERN_SIZE (1 << 30)
504    
505  #else  #else
506  #error Unsupported compiling mode  #error Unsupported compiling mode
507  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
508    
509  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
510    
# Line 492  is automated on Unix systems via the "co Line 515  is automated on Unix systems via the "co
515  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
516  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
517    
518  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
519    
520  #define IMM2_SIZE 2  #define IMM2_SIZE 2
521    
# Line 503  capturing parenthesis numbers in back re Line 526  capturing parenthesis numbers in back re
526  #define GET2(a,n) \  #define GET2(a,n) \
527    (((a)[n] << 8) | (a)[(n)+1])    (((a)[n] << 8) | (a)[(n)+1])
528    
529  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
530    
531  #ifdef COMPILE_PCRE16  #define IMM2_SIZE 1
532    
533    #define PUT2(a,n,d)   \
534       a[n] = d
535    
536    #define GET2(a,n) \
537       a[n]
538    
539    #elif defined COMPILE_PCRE32
540    
541  #define IMM2_SIZE 1  #define IMM2_SIZE 1
542    
# Line 517  capturing parenthesis numbers in back re Line 548  capturing parenthesis numbers in back re
548    
549  #else  #else
550  #error Unsupported compiling mode  #error Unsupported compiling mode
551  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
552    
553  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
554    
555    /* The maximum length of a MARK name is currently one data unit; it may be
556    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
557    
558    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
559    #define MAX_MARK ((1u << 16) - 1)
560    #else
561    #define MAX_MARK ((1u << 8) - 1)
562    #endif
563    
564  /* 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
565  character. The macros for character handling generate simple sequences when  byte. The macros for character handling generate simple sequences when used in
566  used in character-mode, and more complicated ones for UTF characters.  character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
567  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
568  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,
569  UTF support is omitted, we don't even define them. */  we don't even define them. */
570    
571  #ifndef SUPPORT_UTF  #ifndef SUPPORT_UTF
572    
# Line 541  UTF support is omitted, we don't even de Line 579  UTF support is omitted, we don't even de
579  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
580  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
581  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
582    #define RAWUCHAR(eptr) (*(eptr))
583    #define RAWUCHARINC(eptr) (*(eptr)++)
584    #define RAWUCHARTEST(eptr) (*(eptr))
585    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
586  /* #define GETCHARLENTEST(c, eptr, len) */  /* #define GETCHARLENTEST(c, eptr, len) */
587  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
588  /* #define FORWARDCHAR(eptr) */  /* #define FORWARDCHAR(eptr) */
# Line 548  UTF support is omitted, we don't even de Line 590  UTF support is omitted, we don't even de
590    
591  #else   /* SUPPORT_UTF */  #else   /* SUPPORT_UTF */
592    
 #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  
   
593  /* Tests whether the code point needs extra characters to decode. */  /* Tests whether the code point needs extra characters to decode. */
594    
595  #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)  
596    
597  /* 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
598  advancing the pointer. */  advancing the pointer. */
# Line 595  advancing the pointer. */ Line 616  advancing the pointer. */
616            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
617      }      }
618    
 /* 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);  
   
619  /* 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
620  the pointer. */  the pointer. */
621    
# Line 643  the pointer. */ Line 650  the pointer. */
650        } \        } \
651      }      }
652    
653    #if defined COMPILE_PCRE8
654    
655    /* These macros were originally written in the form of loops that used data
656    from the tables whose names start with PRIV(utf8_table). They were rewritten by
657    a user so as not to use loops, because in some environments this gives a
658    significant performance advantage, and it seems never to do any harm. */
659    
660    /* Tells the biggest code point which can be encoded as a single character. */
661    
662    #define MAX_VALUE_FOR_SINGLE_CHAR 127
663    
664    /* Tests whether the code point needs extra characters to decode. */
665    
666    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
667    
668    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
669    Otherwise it has an undefined behaviour. */
670    
671    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
672    
673    /* Returns TRUE, if the given character is not the first character
674    of a UTF sequence. */
675    
676    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
677    
678    /* Get the next UTF-8 character, not advancing the pointer. This is called when
679    we know we are in UTF-8 mode. */
680    
681    #define GETCHAR(c, eptr) \
682      c = *eptr; \
683      if (c >= 0xc0) GETUTF8(c, eptr);
684    
685    /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
686    pointer. */
687    
688    #define GETCHARTEST(c, eptr) \
689      c = *eptr; \
690      if (utf && c >= 0xc0) GETUTF8(c, eptr);
691    
692  /* 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
693  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
694    
# Line 709  do not know if we are in UTF-8 mode. */ Line 755  do not know if we are in UTF-8 mode. */
755    c = *eptr; \    c = *eptr; \
756    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
757    
758    /* Returns the next uchar, not advancing the pointer. This is called when
759    we know we are in UTF mode. */
760    
761    #define RAWUCHAR(eptr) \
762      (*(eptr))
763    
764    /* Returns the next uchar, advancing the pointer. This is called when
765    we know we are in UTF mode. */
766    
767    #define RAWUCHARINC(eptr) \
768      (*(eptr)++)
769    
770    /* Returns the next uchar, testing for UTF mode, and not advancing the
771    pointer. */
772    
773    #define RAWUCHARTEST(eptr) \
774      (*(eptr))
775    
776    /* Returns the next uchar, testing for UTF mode, advancing the
777    pointer. */
778    
779    #define RAWUCHARINCTEST(eptr) \
780      (*(eptr)++)
781    
782  /* 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
783  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
784  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 792  because almost all calls are already wit
792  #define ACROSSCHAR(condition, eptr, action) \  #define ACROSSCHAR(condition, eptr, action) \
793    while((condition) && ((eptr) & 0xc0) == 0x80) action    while((condition) && ((eptr) & 0xc0) == 0x80) action
794    
795  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
   
 #ifdef COMPILE_PCRE16  
796    
797  /* 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. */
798    
# Line 806  we do not know if we are in UTF-16 mode. Line 874  we do not know if we are in UTF-16 mode.
874    c = *eptr; \    c = *eptr; \
875    if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);    if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
876    
877    /* Returns the next uchar, not advancing the pointer. This is called when
878    we know we are in UTF mode. */
879    
880    #define RAWUCHAR(eptr) \
881      (*(eptr))
882    
883    /* Returns the next uchar, advancing the pointer. This is called when
884    we know we are in UTF mode. */
885    
886    #define RAWUCHARINC(eptr) \
887      (*(eptr)++)
888    
889    /* Returns the next uchar, testing for UTF mode, and not advancing the
890    pointer. */
891    
892    #define RAWUCHARTEST(eptr) \
893      (*(eptr))
894    
895    /* Returns the next uchar, testing for UTF mode, advancing the
896    pointer. */
897    
898    #define RAWUCHARINCTEST(eptr) \
899      (*(eptr)++)
900    
901  /* 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
902  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
903  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 912  code. */
912  #define ACROSSCHAR(condition, eptr, action) \  #define ACROSSCHAR(condition, eptr, action) \
913    if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action    if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
914    
915  #endif  #elif defined COMPILE_PCRE32
916    
917  #endif /* COMPILE_PCRE8 */  /* These are unnecessary for the 32-bit library */
918    #undef MAX_VALUE_FOR_SINGLE_CHAR
919    #undef HAS_EXTRALEN
920    #undef GET_EXTRALEN
921    #undef NOT_FIRSTCHAR
922    
923    #define UTF32_MASK (0x1ffffful)
924    
925    /* Get the next UTF-32 character, not advancing the pointer. This is called when
926    we know we are in UTF-32 mode. */
927    
928    #define GETCHAR(c, eptr) \
929      c = (*eptr) & UTF32_MASK;
930    
931    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
932    pointer. */
933    
934    #define GETCHARTEST(c, eptr) \
935      c = *eptr; \
936      if (utf) c &= UTF32_MASK;
937    
938    /* Get the next UTF-32 character, advancing the pointer. This is called when we
939    know we are in UTF-32 mode. */
940    
941    #define GETCHARINC(c, eptr) \
942      c = (*eptr++) & UTF32_MASK;
943    
944    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
945    This is called when we don't know if we are in UTF-32 mode. */
946    
947    #define GETCHARINCTEST(c, eptr) \
948      c = *eptr++; \
949      if (utf) c &= UTF32_MASK;
950    
951    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
952    length (since all UTF-32 is of length 1). This is called when we know we are in
953    UTF-32 mode. */
954    
955    #define GETCHARLEN(c, eptr, len) \
956      GETCHAR(c, eptr)
957    
958    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
959    pointer, not incrementing the length (since all UTF-32 is of length 1).
960    This is called when we do not know if we are in UTF-32 mode. */
961    
962    #define GETCHARLENTEST(c, eptr, len) \
963      GETCHARTEST(c, eptr)
964    
965    /* Returns the next uchar, not advancing the pointer. This is called when
966    we know we are in UTF mode. */
967    
968    #define RAWUCHAR(eptr) \
969      (*(eptr) & UTF32_MASK)
970    
971    /* Returns the next uchar, advancing the pointer. This is called when
972    we know we are in UTF mode. */
973    
974    #define RAWUCHARINC(eptr) \
975      (*(eptr)++ & UTF32_MASK)
976    
977    /* Returns the next uchar, testing for UTF mode, and not advancing the
978    pointer. */
979    
980    #define RAWUCHARTEST(eptr) \
981      (utf ? (*(eptr) & UTF32_MASK) : *(eptr))
982    
983    /* Returns the next uchar, testing for UTF mode, advancing the
984    pointer. */
985    
986    #define RAWUCHARINCTEST(eptr) \
987      (utf ? (*(eptr)++ & UTF32_MASK) : *(eptr)++)
988    
989    /* If the pointer is not at the start of a character, move it back until
990    it is. This is called only in UTF-32 mode - we don't put a test within the
991    macro because almost all calls are already within a block of UTF-32 only
992    code. */
993    
994    #define BACKCHAR(eptr) do { } while (0)
995    
996    /* Same as above, just in the other direction. */
997    #define FORWARDCHAR(eptr) do { } while (0)
998    
999    /* Same as above, but it allows a fully customizable form. */
1000    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1001    
1002    #else
1003    #error Unsupported compiling mode
1004    #endif /* COMPILE_PCRE[8|16|32] */
1005    
1006  #endif  /* SUPPORT_UTF */  #endif  /* SUPPORT_UTF */
1007    
1008    /* Tests for Unicode horizontal and vertical whitespace characters must check a
1009    number of different values. Using a switch statement for this generates the
1010    fastest code (no loop, no memory access), and there are several places in the
1011    interpreter code where this happens. In order to ensure that all the case lists
1012    remain in step, we use macros so that there is only one place where the lists
1013    are defined.
1014    
1015    These values are also required as lists in pcre_compile.c when processing \h,
1016    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1017    macros that define the values are here so that all the definitions are
1018    together. The lists must be in ascending character order, terminated by
1019    NOTACHAR (which is 0xffffffff).
1020    
1021    Any changes should ensure that the various macros are kept in step with each
1022    other. NOTE: The values also appear in pcre_jit_compile.c. */
1023    
1024    /* ------ ASCII/Unicode environments ------ */
1025    
1026    #ifndef EBCDIC
1027    
1028    #define HSPACE_LIST \
1029      CHAR_HT, CHAR_SPACE, 0xa0, \
1030      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1031      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1032      NOTACHAR
1033    
1034    #define HSPACE_MULTIBYTE_CASES \
1035      case 0x1680:  /* OGHAM SPACE MARK */ \
1036      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1037      case 0x2000:  /* EN QUAD */ \
1038      case 0x2001:  /* EM QUAD */ \
1039      case 0x2002:  /* EN SPACE */ \
1040      case 0x2003:  /* EM SPACE */ \
1041      case 0x2004:  /* THREE-PER-EM SPACE */ \
1042      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1043      case 0x2006:  /* SIX-PER-EM SPACE */ \
1044      case 0x2007:  /* FIGURE SPACE */ \
1045      case 0x2008:  /* PUNCTUATION SPACE */ \
1046      case 0x2009:  /* THIN SPACE */ \
1047      case 0x200A:  /* HAIR SPACE */ \
1048      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1049      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1050      case 0x3000   /* IDEOGRAPHIC SPACE */
1051    
1052    #define HSPACE_BYTE_CASES \
1053      case CHAR_HT: \
1054      case CHAR_SPACE: \
1055      case 0xa0     /* NBSP */
1056    
1057    #define HSPACE_CASES \
1058      HSPACE_BYTE_CASES: \
1059      HSPACE_MULTIBYTE_CASES
1060    
1061    #define VSPACE_LIST \
1062      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1063    
1064    #define VSPACE_MULTIBYTE_CASES \
1065      case 0x2028:    /* LINE SEPARATOR */ \
1066      case 0x2029     /* PARAGRAPH SEPARATOR */
1067    
1068    #define VSPACE_BYTE_CASES \
1069      case CHAR_LF: \
1070      case CHAR_VT: \
1071      case CHAR_FF: \
1072      case CHAR_CR: \
1073      case CHAR_NEL
1074    
1075    #define VSPACE_CASES \
1076      VSPACE_BYTE_CASES: \
1077      VSPACE_MULTIBYTE_CASES
1078    
1079  /* In case there is no definition of offsetof() provided - though any proper  /* ------ EBCDIC environments ------ */
1080  Standard C system should have one. */  
1081    #else
1082    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1083    
1084    #define HSPACE_BYTE_CASES \
1085      case CHAR_HT: \
1086      case CHAR_SPACE
1087    
1088    #define HSPACE_CASES HSPACE_BYTE_CASES
1089    
1090    #ifdef EBCDIC_NL25
1091    #define VSPACE_LIST \
1092      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1093    #else
1094    #define VSPACE_LIST \
1095      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1096    #endif
1097    
1098    #define VSPACE_BYTE_CASES \
1099      case CHAR_LF: \
1100      case CHAR_VT: \
1101      case CHAR_FF: \
1102      case CHAR_CR: \
1103      case CHAR_NEL
1104    
1105    #define VSPACE_CASES VSPACE_BYTE_CASES
1106    #endif  /* EBCDIC */
1107    
1108    /* ------ End of whitespace macros ------ */
1109    
 #ifndef offsetof  
 #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
 #endif  
1110    
1111    
1112  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
# Line 841  are in a 16-bit flags word. From release Line 1115  are in a 16-bit flags word. From release
1115  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1116  compatibility. */  compatibility. */
1117    
1118  #ifdef COMPILE_PCRE8  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1119  #define PCRE_MODE          0x0001  /* compiled in 8 bit mode */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1120  #endif  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
 #ifdef COMPILE_PCRE16  
 #define PCRE_MODE          0x0002  /* compiled in 16 bit mode */  
 #endif  
1121  #define PCRE_FIRSTSET      0x0010  /* first_char is set */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1122  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1123  #define PCRE_REQCHSET      0x0040  /* req_byte is set */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
# Line 857  compatibility. */ Line 1128  compatibility. */
1128  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1129  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1130    
1131    #if defined COMPILE_PCRE8
1132    #define PCRE_MODE          PCRE_MODE8
1133    #elif defined COMPILE_PCRE16
1134    #define PCRE_MODE          PCRE_MODE16
1135    #elif defined COMPILE_PCRE32
1136    #define PCRE_MODE          PCRE_MODE32
1137    #endif
1138    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1139    
1140  /* Flags for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1141    
1142  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
# Line 888  time, run time, or study time, respectiv Line 1168  time, run time, or study time, respectiv
1168    
1169  #define PUBLIC_STUDY_OPTIONS \  #define PUBLIC_STUDY_OPTIONS \
1170     (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \     (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1171      PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE)      PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1172    
1173  /* Magic number to provide a small check against being handed junk. */  /* Magic number to provide a small check against being handed junk. */
1174    
# Line 899  in different endianness. */ Line 1179  in different endianness. */
1179    
1180  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
1181    
 /* Negative values for the firstchar and reqchar variables */  
   
 #define REQ_UNSET (-2)  
 #define REQ_NONE  (-1)  
   
1182  /* 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
1183  req_byte match. */  req_byte match. */
1184    
# Line 940  macros to give the functions distinct na Line 1215  macros to give the functions distinct na
1215  #ifndef SUPPORT_UTF  #ifndef SUPPORT_UTF
1216    
1217  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1218  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
1219    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1220    characters, a common practice has been to use its NL (0x15) character as the
1221    line terminator in C-like processing environments. However, sometimes the LF
1222    (0x25) character is used instead, according to this Unicode document:
1223    
1224    http://unicode.org/standard/reports/tr13/tr13-5.html
1225    
1226    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1227    instead. Whichever is *not* chosen is defined as NEL.
1228    
1229    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1230    same code point. */
1231    
1232    #ifdef EBCDIC
1233    
1234    #ifndef EBCDIC_NL25
1235    #define CHAR_NL                     '\x15'
1236    #define CHAR_NEL                    '\x25'
1237    #define STR_NL                      "\x15"
1238    #define STR_NEL                     "\x25"
1239    #else
1240    #define CHAR_NL                     '\x25'
1241    #define CHAR_NEL                    '\x15'
1242    #define STR_NL                      "\x25"
1243    #define STR_NEL                     "\x15"
1244    #endif
1245    
1246    #define CHAR_LF                     CHAR_NL
1247    #define STR_LF                      STR_NL
1248    
1249    #define CHAR_ESC                    '\047'
1250    #define CHAR_DEL                    '\007'
1251    #define STR_ESC                     "\047"
1252    #define STR_DEL                     "\007"
1253    
1254    #else  /* Not EBCDIC */
1255    
1256    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1257    compatibility. NEL is the Unicode newline character; make sure it is
1258    a positive value. */
1259    
1260    #define CHAR_LF                     '\n'
1261    #define CHAR_NL                     CHAR_LF
1262    #define CHAR_NEL                    ((unsigned char)'\x85')
1263    #define CHAR_ESC                    '\033'
1264    #define CHAR_DEL                    '\177'
1265    
1266    #define STR_LF                      "\n"
1267    #define STR_NL                      STR_LF
1268    #define STR_NEL                     "\x85"
1269    #define STR_ESC                     "\033"
1270    #define STR_DEL                     "\177"
1271    
1272    #endif  /* EBCDIC */
1273    
1274    /* The remaining definitions work in both environments. */
1275    
1276  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1277  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1278  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1279  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1280  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1281  #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  
1282    
1283  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1284  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 1057  so that PCRE works on both ASCII and EBC Line 1380  so that PCRE works on both ASCII and EBC
1380  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1381  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1382  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1383  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1384  #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  
1385    
1386  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1387  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 1203  so that PCRE works on both ASCII and EBC Line 1518  so that PCRE works on both ASCII and EBC
1518  #ifdef COMPILE_PCRE16  #ifdef COMPILE_PCRE16
1519  #define STRING_UTF_RIGHTPAR            "UTF16)"  #define STRING_UTF_RIGHTPAR            "UTF16)"
1520  #endif  #endif
1521    #ifdef COMPILE_PCRE32
1522    #define STRING_UTF_RIGHTPAR            "UTF32)"
1523    #endif
1524  #define STRING_UCP_RIGHTPAR            "UCP)"  #define STRING_UCP_RIGHTPAR            "UCP)"
1525  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1526    
# Line 1216  only. */ Line 1534  only. */
1534  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1535  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1536  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1537  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1538    #define CHAR_NL                     CHAR_LF
1539    #define CHAR_NEL                    ((unsigned char)'\x85')
1540  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1541  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1542  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
# Line 1463  only. */ Line 1783  only. */
1783  #ifdef COMPILE_PCRE16  #ifdef COMPILE_PCRE16
1784  #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS  #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1785  #endif  #endif
1786    #ifdef COMPILE_PCRE32
1787    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1788    #endif
1789  #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
1790  #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
1791    
# Line 1479  only. */ Line 1802  only. */
1802  #endif  #endif
1803    
1804  #ifndef ESC_n  #ifndef ESC_n
1805  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1806  #endif  #endif
1807    
1808  #ifndef ESC_r  #ifndef ESC_r
# Line 1504  only. */ Line 1827  only. */
1827  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1828  #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 */
1829  #define PT_WORD       8    /* Word - L plus N plus underscore */  #define PT_WORD       8    /* Word - L plus N plus underscore */
1830    #define PT_CLIST      9    /* Pseudo-property: match character list */
1831    
1832  /* 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
1833  contain characters with values greater than 255. */  contain characters with values greater than 255. */
# Line 1519  contain characters with values greater t Line 1843  contain characters with values greater t
1843    
1844  /* 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
1845  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
1846  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
1847  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
1848  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
1849  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
1850  non-DOTALL mode, "." behaves like \N.  non-DOTALL mode, "." behaves like \N.
1851    
1852  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.
1853  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.
1854  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
1855  looked up from a table.  looked up from a table.
1856    
1857  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
1858  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
1859  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
1860  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
1861  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 1541  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E Line 1865  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E
1865         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,
1866         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1867         ESC_E, ESC_Q, ESC_g, ESC_k,         ESC_E, ESC_Q, ESC_g, ESC_k,
1868         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 };  
1869    
1870  /* 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
1871  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 1568  enum { Line 1891  enum {
1891    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1892    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1893    
1894    OP_ANY,            /* 12 Match any character except newline */    OP_ANY,            /* 12 Match any character except newline (\N) */
1895    OP_ALLANY,         /* 13 Match any character */    OP_ALLANY,         /* 13 Match any character */
1896    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 */
1897    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
# Line 1579  enum { Line 1902  enum {
1902    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1903    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1904    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1905    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) */
1906    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1907    
1908    OP_CIRC,           /* 25 Start of line - not multiline */    OP_CIRC,           /* 25 Start of line - not multiline */
1909    OP_CIRCM,          /* 26 Start of line - multiline */    OP_CIRCM,          /* 26 Start of line - multiline */
# Line 1940  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2263  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2263         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2264         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2265         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2266         ERR70, ERR71, ERR72, ERR73, ERR74, ERRCOUNT };         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2267    
2268  /* JIT compiling modes. The function list is indexed by them. */  /* JIT compiling modes. The function list is indexed by them. */
2269  enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,  enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
# Line 1963  fields are present. Currently PCRE alway Line 2286  fields are present. Currently PCRE alway
2286  NOTE NOTE NOTE  NOTE NOTE NOTE
2287  */  */
2288    
2289  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2290  #define REAL_PCRE real_pcre  #define REAL_PCRE real_pcre
2291  #else  #elif defined COMPILE_PCRE16
2292  #define REAL_PCRE real_pcre16  #define REAL_PCRE real_pcre16
2293    #elif defined COMPILE_PCRE32
2294    #define REAL_PCRE real_pcre32
2295  #endif  #endif
2296    
2297  typedef struct REAL_PCRE {  /* It is necessary to fork the struct for 32 bit, since it needs to use
2298     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2299     * typedef since pcretest needs access to  the struct of the 8-, 16-
2300     * and 32-bit variants. */
2301    
2302    typedef struct real_pcre8_or_16 {
2303    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2304    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2305    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
# Line 1985  typedef struct REAL_PCRE { Line 2315  typedef struct REAL_PCRE {
2315    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2316    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2317    const pcre_uint8 *nullpad;      /* NULL padding */    const pcre_uint8 *nullpad;      /* NULL padding */
2318  } REAL_PCRE;  } real_pcre8_or_16;
2319    
2320    typedef struct real_pcre8_or_16 real_pcre;
2321    typedef struct real_pcre8_or_16 real_pcre16;
2322    
2323    typedef struct real_pcre32 {
2324      pcre_uint32 magic_number;
2325      pcre_uint32 size;               /* Total that was malloced */
2326      pcre_uint32 options;            /* Public options */
2327      pcre_uint16 flags;              /* Private flags */
2328      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2329      pcre_uint16 top_bracket;        /* Highest numbered group */
2330      pcre_uint16 top_backref;        /* Highest numbered back reference */
2331      pcre_uint32 first_char;         /* Starting character */
2332      pcre_uint32 req_char;           /* This character must be seen */
2333      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2334      pcre_uint16 name_entry_size;    /* Size of any name items */
2335      pcre_uint16 name_count;         /* Number of name items */
2336      pcre_uint16 ref_count;          /* Reference count */
2337      pcre_uint16 dummy1;             /* for later expansion */
2338      pcre_uint16 dummy2;             /* for later expansion */
2339      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2340      void *nullpad;                  /* for later expansion */
2341    } real_pcre32;
2342    
2343    /* Assert that the size of REAL_PCRE is divisible by 8 */
2344    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2345    
2346    /* Needed in pcretest to access some fields in the real_pcre* structures
2347     * directly. They're unified for 8/16/32 bits since the structs only differ
2348     * after these fields; if that ever changes, need to fork those defines into
2349     * 8/16 and 32 bit versions. */
2350    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2351    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2352    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2353    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2354    
2355  /* 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
2356  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 2036  typedef struct compile_data { Line 2401  typedef struct compile_data {
2401    int  external_flags;              /* External flag bits to be set */    int  external_flags;              /* External flag bits to be set */
2402    int  req_varyopt;                 /* "After variable item" flag for reqbyte */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2403    BOOL had_accept;                  /* (*ACCEPT) encountered */    BOOL had_accept;                  /* (*ACCEPT) encountered */
2404      BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2405    BOOL check_lookbehind;            /* Lookbehinds need later checking */    BOOL check_lookbehind;            /* Lookbehinds need later checking */
2406    int  nltype;                      /* Newline type */    int  nltype;                      /* Newline type */
2407    int  nllen;                       /* Newline string length */    int  nllen;                       /* Newline string length */
# Line 2130  typedef struct match_data { Line 2496  typedef struct match_data {
2496    const  pcre_uchar *mark;        /* Mark pointer to pass back on success */    const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2497    const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */    const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2498    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 */
2499    #ifdef NO_RECURSE
2500      void  *match_frames_base;       /* For remembering malloc'd frames */
2501    #endif
2502  } match_data;  } match_data;
2503    
2504  /* 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 2186  total length. */ Line 2555  total length. */
2555    
2556  /* Internal function and data prefixes. */  /* Internal function and data prefixes. */
2557    
2558  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2559  #ifndef PUBL  #ifndef PUBL
2560  #define PUBL(name) pcre_##name  #define PUBL(name) pcre_##name
2561  #endif  #endif
2562  #ifndef PRIV  #ifndef PRIV
2563  #define PRIV(name) _pcre_##name  #define PRIV(name) _pcre_##name
2564  #endif  #endif
2565  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
 #ifdef COMPILE_PCRE16  
2566  #ifndef PUBL  #ifndef PUBL
2567  #define PUBL(name) pcre16_##name  #define PUBL(name) pcre16_##name
2568  #endif  #endif
2569  #ifndef PRIV  #ifndef PRIV
2570  #define PRIV(name) _pcre16_##name  #define PRIV(name) _pcre16_##name
2571  #endif  #endif
2572    #elif defined COMPILE_PCRE32
2573    #ifndef PUBL
2574    #define PUBL(name) pcre32_##name
2575    #endif
2576    #ifndef PRIV
2577    #define PRIV(name) _pcre32_##name
2578    #endif
2579  #else  #else
2580  #error Unsupported compiling mode  #error Unsupported compiling mode
2581  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
 #endif /* COMPILE_PCRE8 */  
2582    
2583  /* 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
2584  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 2224  but are not part of the PCRE public API. Line 2598  but are not part of the PCRE public API.
2598  pcre_tables.c module. */  pcre_tables.c module. */
2599    
2600  #ifdef COMPILE_PCRE8  #ifdef COMPILE_PCRE8
   
2601  extern const int            PRIV(utf8_table1)[];  extern const int            PRIV(utf8_table1)[];
2602  extern const int            PRIV(utf8_table1_size);  extern const int            PRIV(utf8_table1_size);
2603  extern const int            PRIV(utf8_table2)[];  extern const int            PRIV(utf8_table2)[];
2604  extern const int            PRIV(utf8_table3)[];  extern const int            PRIV(utf8_table3)[];
2605  extern const pcre_uint8     PRIV(utf8_table4)[];  extern const pcre_uint8     PRIV(utf8_table4)[];
   
2606  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE8 */
2607    
2608  extern const char           PRIV(utt_names)[];  extern const char           PRIV(utt_names)[];
2609  extern const ucp_type_table PRIV(utt)[];  extern const ucp_type_table PRIV(utt)[];
2610  extern const int            PRIV(utt_size);  extern const int            PRIV(utt_size);
2611    
2612    extern const pcre_uint8     PRIV(OP_lengths)[];
2613  extern const pcre_uint8     PRIV(default_tables)[];  extern const pcre_uint8     PRIV(default_tables)[];
2614    
2615  extern const pcre_uint8     PRIV(OP_lengths)[];  extern const pcre_uint32    PRIV(hspace_list)[];
2616    extern const pcre_uint32    PRIV(vspace_list)[];
2617    
2618    
2619  /* 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 2247  one of the exported public functions. Th Line 2621  one of the exported public functions. Th
2621  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2622    
2623  /* String comparison functions. */  /* String comparison functions. */
2624  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2625    
2626  #define STRCMP_UC_UC(str1, str2) \  #define STRCMP_UC_UC(str1, str2) \
2627    strcmp((char *)(str1), (char *)(str2))    strcmp((char *)(str1), (char *)(str2))
# Line 2259  sense, but are not part of the PCRE publ Line 2633  sense, but are not part of the PCRE publ
2633    strncmp((char *)(str1), (str2), (num))    strncmp((char *)(str1), (str2), (num))
2634  #define STRLEN_UC(str) strlen((const char *)str)  #define STRLEN_UC(str) strlen((const char *)str)
2635    
2636  #else  #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2637    
2638  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2639                             const pcre_uchar *);                             const pcre_uchar *);
# Line 2281  extern unsigned int      PRIV(strlen_uc) Line 2655  extern unsigned int      PRIV(strlen_uc)
2655    PRIV(strncmp_uc_c8)((str1), (str2), (num))    PRIV(strncmp_uc_c8)((str1), (str2), (num))
2656  #define STRLEN_UC(str) PRIV(strlen_uc)(str)  #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2657    
2658  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE[8|16|32] */
2659    
2660    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2661    
2662    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2663    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2664    
2665    #elif defined COMPILE_PCRE32
2666    
2667    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2668                               const pcre_uchar *);
2669    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2670                               const char *);
2671    
2672    #define STRCMP_UC_UC_TEST(str1, str2) \
2673      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2674    #define STRCMP_UC_C8_TEST(str1, str2) \
2675      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2676    
2677    #endif /* COMPILE_PCRE[8|16|32] */
2678    
2679  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2680  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
# Line 2290  extern int               PRIV(ord2utf)(p Line 2683  extern int               PRIV(ord2utf)(p
2683  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2684  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2685                             int *, BOOL);                             int *, BOOL);
2686  extern BOOL              PRIV(xclass)(int, const pcre_uchar *, BOOL);  extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2687    
2688  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2689  extern void              PRIV(jit_compile)(const REAL_PCRE *, PUBL(extra) *, int);  extern void              PRIV(jit_compile)(const REAL_PCRE *,
2690  extern int               PRIV(jit_exec)(const REAL_PCRE *, void *,                             PUBL(extra) *, int);
2691                             const pcre_uchar *, int, int, int, int, int *, int,  extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2692                             pcre_uchar **);                             const pcre_uchar *, int, int, int, int *, int);
2693  extern void              PRIV(jit_free)(void *);  extern void              PRIV(jit_free)(void *);
2694  extern int               PRIV(jit_get_size)(void *);  extern int               PRIV(jit_get_size)(void *);
2695  extern const char*       PRIV(jit_get_target)(void);  extern const char*       PRIV(jit_get_target)(void);
# Line 2305  extern const char*       PRIV(jit_get_ta Line 2698  extern const char*       PRIV(jit_get_ta
2698  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
2699    
2700  typedef struct {  typedef struct {
2701    pcre_uint8 script;    pcre_uint8 script;     /* ucp_Arabic, etc. */
2702    pcre_uint8 chartype;    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2703    pcre_int32 other_case;    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2704      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2705      pcre_int32 other_case; /* offset to other case, or zero if none */
2706  } ucd_record;  } ucd_record;
2707    
2708    extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2709  extern const ucd_record  PRIV(ucd_records)[];  extern const ucd_record  PRIV(ucd_records)[];
2710  extern const pcre_uint8  PRIV(ucd_stage1)[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2711  extern const pcre_uint16 PRIV(ucd_stage2)[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2712  extern const int         PRIV(ucp_gentype)[];  extern const int         PRIV(ucp_gentype)[];
2713    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2714  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2715  extern const int         PRIV(ucp_typerange)[];  extern const int         PRIV(ucp_typerange)[];
2716  #endif  #endif
# Line 2323  extern const int         PRIV(ucp_typera Line 2720  extern const int         PRIV(ucp_typera
2720    
2721  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2722  #define GET_UCD(ch) (PRIV(ucd_records) + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2723          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2724          UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2725    
2726  #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2727  #define UCD_SCRIPT(ch)    GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2728  #define UCD_CATEGORY(ch)  PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2729  #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2730    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2731    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2732    
2733  #endif /* SUPPORT_UCP */  #endif /* SUPPORT_UCP */
2734    

Legend:
Removed from v.932  
changed lines
  Added in v.1100

  ViewVC Help
Powered by ViewVC 1.1.5