/[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 1022 by ph10, Tue Aug 28 12:28:15 2012 UTC revision 1128 by chpe, Thu Oct 18 18:35:08 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  /* 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. */  changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
557    
558  #define MAX_MARK ((1 << (sizeof(pcre_uchar)*8)) - 1)  #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 546  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 553  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 600  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 648  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 714  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 727  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 811  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 825  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 trivial for the 32-bit library, since all UTF-32 characters fit
918    into one pcre_uchar unit. */
919    #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
920    #define HAS_EXTRALEN(c) (0)
921    #define GET_EXTRALEN(c) (0)
922    #define NOT_FIRSTCHAR(c) (0)
923    
924    #define UTF32_MASK (0x1fffffu)
925    
926    /* Get the next UTF-32 character, not advancing the pointer. This is called when
927    we know we are in UTF-32 mode. */
928    
929    #define GETCHAR(c, eptr) \
930      c = (*eptr) & UTF32_MASK;
931    
932    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
933    pointer. */
934    
935    #define GETCHARTEST(c, eptr) \
936      c = *eptr; \
937      if (utf) c &= UTF32_MASK;
938    
939    /* Get the next UTF-32 character, advancing the pointer. This is called when we
940    know we are in UTF-32 mode. */
941    
942    #define GETCHARINC(c, eptr) \
943      c = (*eptr++) & UTF32_MASK;
944    
945    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
946    This is called when we don't know if we are in UTF-32 mode. */
947    
948    #define GETCHARINCTEST(c, eptr) \
949      c = *eptr++; \
950      if (utf) c &= UTF32_MASK;
951    
952    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
953    length (since all UTF-32 is of length 1). This is called when we know we are in
954    UTF-32 mode. */
955    
956    #define GETCHARLEN(c, eptr, len) \
957      GETCHAR(c, eptr)
958    
959    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
960    pointer, not incrementing the length (since all UTF-32 is of length 1).
961    This is called when we do not know if we are in UTF-32 mode. */
962    
963    #define GETCHARLENTEST(c, eptr, len) \
964      GETCHARTEST(c, eptr)
965    
966    /* Returns the next uchar, not advancing the pointer. This is called when
967    we know we are in UTF mode. */
968    
969    #define RAWUCHAR(eptr) \
970      (*(eptr) & UTF32_MASK)
971    
972    /* Returns the next uchar, advancing the pointer. This is called when
973    we know we are in UTF mode. */
974    
975    #define RAWUCHARINC(eptr) \
976      (*(eptr)++ & UTF32_MASK)
977    
978    /* Returns the next uchar, testing for UTF mode, and not advancing the
979    pointer. */
980    
981    #define RAWUCHARTEST(eptr) \
982      (utf ? (*(eptr) & UTF32_MASK) : *(eptr))
983    
984    /* Returns the next uchar, testing for UTF mode, advancing the
985    pointer. */
986    
987    #define RAWUCHARINCTEST(eptr) \
988      (utf ? (*(eptr)++ & UTF32_MASK) : *(eptr)++)
989    
990    /* If the pointer is not at the start of a character, move it back until
991    it is. This is called only in UTF-32 mode - we don't put a test within the
992    macro because almost all calls are already within a block of UTF-32 only
993    code.
994    These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
995    
996    #define BACKCHAR(eptr) do { } while (0)
997    
998    /* Same as above, just in the other direction. */
999    #define FORWARDCHAR(eptr) do { } while (0)
1000    
1001    /* Same as above, but it allows a fully customizable form. */
1002    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1003    
1004    #else
1005    #error Unsupported compiling mode
1006    #endif /* COMPILE_PCRE[8|16|32] */
1007    
1008  #endif  /* SUPPORT_UTF */  #endif  /* SUPPORT_UTF */
1009    
1010    /* Tests for Unicode horizontal and vertical whitespace characters must check a
1011    number of different values. Using a switch statement for this generates the
1012    fastest code (no loop, no memory access), and there are several places in the
1013    interpreter code where this happens. In order to ensure that all the case lists
1014    remain in step, we use macros so that there is only one place where the lists
1015    are defined.
1016    
1017    These values are also required as lists in pcre_compile.c when processing \h,
1018    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1019    macros that define the values are here so that all the definitions are
1020    together. The lists must be in ascending character order, terminated by
1021    NOTACHAR (which is 0xffffffff).
1022    
1023    Any changes should ensure that the various macros are kept in step with each
1024    other. NOTE: The values also appear in pcre_jit_compile.c. */
1025    
1026    /* ------ ASCII/Unicode environments ------ */
1027    
1028    #ifndef EBCDIC
1029    
1030    #define HSPACE_LIST \
1031      CHAR_HT, CHAR_SPACE, 0xa0, \
1032      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1033      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1034      NOTACHAR
1035    
1036    #define HSPACE_MULTIBYTE_CASES \
1037      case 0x1680:  /* OGHAM SPACE MARK */ \
1038      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1039      case 0x2000:  /* EN QUAD */ \
1040      case 0x2001:  /* EM QUAD */ \
1041      case 0x2002:  /* EN SPACE */ \
1042      case 0x2003:  /* EM SPACE */ \
1043      case 0x2004:  /* THREE-PER-EM SPACE */ \
1044      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1045      case 0x2006:  /* SIX-PER-EM SPACE */ \
1046      case 0x2007:  /* FIGURE SPACE */ \
1047      case 0x2008:  /* PUNCTUATION SPACE */ \
1048      case 0x2009:  /* THIN SPACE */ \
1049      case 0x200A:  /* HAIR SPACE */ \
1050      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1051      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1052      case 0x3000   /* IDEOGRAPHIC SPACE */
1053    
1054    #define HSPACE_BYTE_CASES \
1055      case CHAR_HT: \
1056      case CHAR_SPACE: \
1057      case 0xa0     /* NBSP */
1058    
1059    #define HSPACE_CASES \
1060      HSPACE_BYTE_CASES: \
1061      HSPACE_MULTIBYTE_CASES
1062    
1063    #define VSPACE_LIST \
1064      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1065    
1066    #define VSPACE_MULTIBYTE_CASES \
1067      case 0x2028:    /* LINE SEPARATOR */ \
1068      case 0x2029     /* PARAGRAPH SEPARATOR */
1069    
1070    #define VSPACE_BYTE_CASES \
1071      case CHAR_LF: \
1072      case CHAR_VT: \
1073      case CHAR_FF: \
1074      case CHAR_CR: \
1075      case CHAR_NEL
1076    
1077    #define VSPACE_CASES \
1078      VSPACE_BYTE_CASES: \
1079      VSPACE_MULTIBYTE_CASES
1080    
1081  /* In case there is no definition of offsetof() provided - though any proper  /* ------ EBCDIC environments ------ */
1082  Standard C system should have one. */  
1083    #else
1084    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1085    
1086    #define HSPACE_BYTE_CASES \
1087      case CHAR_HT: \
1088      case CHAR_SPACE
1089    
1090    #define HSPACE_CASES HSPACE_BYTE_CASES
1091    
1092    #ifdef EBCDIC_NL25
1093    #define VSPACE_LIST \
1094      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1095    #else
1096    #define VSPACE_LIST \
1097      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1098    #endif
1099    
1100    #define VSPACE_BYTE_CASES \
1101      case CHAR_LF: \
1102      case CHAR_VT: \
1103      case CHAR_FF: \
1104      case CHAR_CR: \
1105      case CHAR_NEL
1106    
1107    #define VSPACE_CASES VSPACE_BYTE_CASES
1108    #endif  /* EBCDIC */
1109    
1110    /* ------ End of whitespace macros ------ */
1111    
 #ifndef offsetof  
 #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
 #endif  
1112    
1113    
1114  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
# Line 846  are in a 16-bit flags word. From release Line 1117  are in a 16-bit flags word. From release
1117  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1118  compatibility. */  compatibility. */
1119    
1120  #ifdef COMPILE_PCRE8  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1121  #define PCRE_MODE          0x0001  /* compiled in 8 bit mode */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1122  #endif  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
 #ifdef COMPILE_PCRE16  
 #define PCRE_MODE          0x0002  /* compiled in 16 bit mode */  
 #endif  
1123  #define PCRE_FIRSTSET      0x0010  /* first_char is set */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1124  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1125  #define PCRE_REQCHSET      0x0040  /* req_byte is set */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
# Line 862  compatibility. */ Line 1130  compatibility. */
1130  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1131  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1132    
1133    #if defined COMPILE_PCRE8
1134    #define PCRE_MODE          PCRE_MODE8
1135    #elif defined COMPILE_PCRE16
1136    #define PCRE_MODE          PCRE_MODE16
1137    #elif defined COMPILE_PCRE32
1138    #define PCRE_MODE          PCRE_MODE32
1139    #endif
1140    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1141    
1142  /* Flags for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1143    
1144  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
# Line 904  in different endianness. */ Line 1181  in different endianness. */
1181    
1182  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
1183    
 /* Negative values for the firstchar and reqchar variables */  
   
 #define REQ_UNSET (-2)  
 #define REQ_NONE  (-1)  
   
1184  /* 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
1185  req_byte match. */  req_byte match. */
1186    
# Line 945  macros to give the functions distinct na Line 1217  macros to give the functions distinct na
1217  #ifndef SUPPORT_UTF  #ifndef SUPPORT_UTF
1218    
1219  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1220  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
1221    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1222    characters, a common practice has been to use its NL (0x15) character as the
1223    line terminator in C-like processing environments. However, sometimes the LF
1224    (0x25) character is used instead, according to this Unicode document:
1225    
1226    http://unicode.org/standard/reports/tr13/tr13-5.html
1227    
1228    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1229    instead. Whichever is *not* chosen is defined as NEL.
1230    
1231    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1232    same code point. */
1233    
1234    #ifdef EBCDIC
1235    
1236    #ifndef EBCDIC_NL25
1237    #define CHAR_NL                     '\x15'
1238    #define CHAR_NEL                    '\x25'
1239    #define STR_NL                      "\x15"
1240    #define STR_NEL                     "\x25"
1241    #else
1242    #define CHAR_NL                     '\x25'
1243    #define CHAR_NEL                    '\x15'
1244    #define STR_NL                      "\x25"
1245    #define STR_NEL                     "\x15"
1246    #endif
1247    
1248    #define CHAR_LF                     CHAR_NL
1249    #define STR_LF                      STR_NL
1250    
1251    #define CHAR_ESC                    '\047'
1252    #define CHAR_DEL                    '\007'
1253    #define STR_ESC                     "\047"
1254    #define STR_DEL                     "\007"
1255    
1256    #else  /* Not EBCDIC */
1257    
1258    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1259    compatibility. NEL is the Unicode newline character; make sure it is
1260    a positive value. */
1261    
1262    #define CHAR_LF                     '\n'
1263    #define CHAR_NL                     CHAR_LF
1264    #define CHAR_NEL                    ((unsigned char)'\x85')
1265    #define CHAR_ESC                    '\033'
1266    #define CHAR_DEL                    '\177'
1267    
1268    #define STR_LF                      "\n"
1269    #define STR_NL                      STR_LF
1270    #define STR_NEL                     "\x85"
1271    #define STR_ESC                     "\033"
1272    #define STR_DEL                     "\177"
1273    
1274    #endif  /* EBCDIC */
1275    
1276    /* The remaining definitions work in both environments. */
1277    
1278  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1279  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1280  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1281  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1282  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1283  #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  
1284    
1285  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1286  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 1062  so that PCRE works on both ASCII and EBC Line 1382  so that PCRE works on both ASCII and EBC
1382  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1383  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1384  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1385  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1386  #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  
1387    
1388  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1389  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 1208  so that PCRE works on both ASCII and EBC Line 1520  so that PCRE works on both ASCII and EBC
1520  #ifdef COMPILE_PCRE16  #ifdef COMPILE_PCRE16
1521  #define STRING_UTF_RIGHTPAR            "UTF16)"  #define STRING_UTF_RIGHTPAR            "UTF16)"
1522  #endif  #endif
1523    #ifdef COMPILE_PCRE32
1524    #define STRING_UTF_RIGHTPAR            "UTF32)"
1525    #endif
1526  #define STRING_UCP_RIGHTPAR            "UCP)"  #define STRING_UCP_RIGHTPAR            "UCP)"
1527  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1528    
# Line 1221  only. */ Line 1536  only. */
1536  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1537  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1538  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1539  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1540    #define CHAR_NL                     CHAR_LF
1541    #define CHAR_NEL                    ((unsigned char)'\x85')
1542  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1543  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1544  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
# Line 1468  only. */ Line 1785  only. */
1785  #ifdef COMPILE_PCRE16  #ifdef COMPILE_PCRE16
1786  #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
1787  #endif  #endif
1788    #ifdef COMPILE_PCRE32
1789    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1790    #endif
1791  #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
1792  #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
1793    
# Line 1484  only. */ Line 1804  only. */
1804  #endif  #endif
1805    
1806  #ifndef ESC_n  #ifndef ESC_n
1807  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1808  #endif  #endif
1809    
1810  #ifndef ESC_r  #ifndef ESC_r
# Line 1509  only. */ Line 1829  only. */
1829  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1830  #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 */
1831  #define PT_WORD       8    /* Word - L plus N plus underscore */  #define PT_WORD       8    /* Word - L plus N plus underscore */
1832    #define PT_CLIST      9    /* Pseudo-property: match character list */
1833    
1834  /* 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
1835  contain characters with values greater than 255. */  contain characters with values greater than 255. */
# Line 1524  contain characters with values greater t Line 1845  contain characters with values greater t
1845    
1846  /* 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
1847  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
1848  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
1849  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
1850  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
1851  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
1852  non-DOTALL mode, "." behaves like \N.  non-DOTALL mode, "." behaves like \N.
1853    
1854  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.
1855  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.
1856  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
1857  looked up from a table.  looked up from a table.
1858    
1859  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
1860  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
1861  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
1862  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
1863  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 1546  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E Line 1867  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E
1867         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,
1868         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1869         ESC_E, ESC_Q, ESC_g, ESC_k,         ESC_E, ESC_Q, ESC_g, ESC_k,
1870         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 };  
1871    
1872  /* 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
1873  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 1573  enum { Line 1893  enum {
1893    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1894    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1895    
1896    OP_ANY,            /* 12 Match any character except newline */    OP_ANY,            /* 12 Match any character except newline (\N) */
1897    OP_ALLANY,         /* 13 Match any character */    OP_ALLANY,         /* 13 Match any character */
1898    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 */
1899    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
# Line 1584  enum { Line 1904  enum {
1904    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1905    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1906    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1907    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) */
1908    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1909    
1910    OP_CIRC,           /* 25 Start of line - not multiline */    OP_CIRC,           /* 25 Start of line - not multiline */
1911    OP_CIRCM,          /* 26 Start of line - multiline */    OP_CIRCM,          /* 26 Start of line - multiline */
# Line 1945  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2265  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2265         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2266         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2267         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2268         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERRCOUNT };         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2269    
2270  /* JIT compiling modes. The function list is indexed by them. */  /* JIT compiling modes. The function list is indexed by them. */
2271  enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,  enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
# Line 1968  fields are present. Currently PCRE alway Line 2288  fields are present. Currently PCRE alway
2288  NOTE NOTE NOTE  NOTE NOTE NOTE
2289  */  */
2290    
2291  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2292  #define REAL_PCRE real_pcre  #define REAL_PCRE real_pcre
2293  #else  #elif defined COMPILE_PCRE16
2294  #define REAL_PCRE real_pcre16  #define REAL_PCRE real_pcre16
2295    #elif defined COMPILE_PCRE32
2296    #define REAL_PCRE real_pcre32
2297  #endif  #endif
2298    
2299  typedef struct REAL_PCRE {  /* It is necessary to fork the struct for 32 bit, since it needs to use
2300     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2301     * typedef since pcretest needs access to  the struct of the 8-, 16-
2302     * and 32-bit variants. */
2303    
2304    typedef struct real_pcre8_or_16 {
2305    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2306    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2307    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
# Line 1990  typedef struct REAL_PCRE { Line 2317  typedef struct REAL_PCRE {
2317    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2318    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2319    const pcre_uint8 *nullpad;      /* NULL padding */    const pcre_uint8 *nullpad;      /* NULL padding */
2320  } REAL_PCRE;  } real_pcre8_or_16;
2321    
2322    typedef struct real_pcre8_or_16 real_pcre;
2323    typedef struct real_pcre8_or_16 real_pcre16;
2324    
2325    typedef struct real_pcre32 {
2326      pcre_uint32 magic_number;
2327      pcre_uint32 size;               /* Total that was malloced */
2328      pcre_uint32 options;            /* Public options */
2329      pcre_uint16 flags;              /* Private flags */
2330      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2331      pcre_uint16 top_bracket;        /* Highest numbered group */
2332      pcre_uint16 top_backref;        /* Highest numbered back reference */
2333      pcre_uint32 first_char;         /* Starting character */
2334      pcre_uint32 req_char;           /* This character must be seen */
2335      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2336      pcre_uint16 name_entry_size;    /* Size of any name items */
2337      pcre_uint16 name_count;         /* Number of name items */
2338      pcre_uint16 ref_count;          /* Reference count */
2339      pcre_uint16 dummy1;             /* for later expansion */
2340      pcre_uint16 dummy2;             /* for later expansion */
2341      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2342      void *nullpad;                  /* for later expansion */
2343    } real_pcre32;
2344    
2345    /* Assert that the size of REAL_PCRE is divisible by 8 */
2346    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2347    
2348    /* Needed in pcretest to access some fields in the real_pcre* structures
2349     * directly. They're unified for 8/16/32 bits since the structs only differ
2350     * after these fields; if that ever changes, need to fork those defines into
2351     * 8/16 and 32 bit versions. */
2352    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2353    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2354    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2355    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2356    
2357  /* 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
2358  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 2041  typedef struct compile_data { Line 2403  typedef struct compile_data {
2403    int  external_flags;              /* External flag bits to be set */    int  external_flags;              /* External flag bits to be set */
2404    int  req_varyopt;                 /* "After variable item" flag for reqbyte */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2405    BOOL had_accept;                  /* (*ACCEPT) encountered */    BOOL had_accept;                  /* (*ACCEPT) encountered */
2406    BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */    BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2407    BOOL check_lookbehind;            /* Lookbehinds need later checking */    BOOL check_lookbehind;            /* Lookbehinds need later checking */
2408    int  nltype;                      /* Newline type */    int  nltype;                      /* Newline type */
2409    int  nllen;                       /* Newline string length */    int  nllen;                       /* Newline string length */
# Line 2195  total length. */ Line 2557  total length. */
2557    
2558  /* Internal function and data prefixes. */  /* Internal function and data prefixes. */
2559    
2560  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2561  #ifndef PUBL  #ifndef PUBL
2562  #define PUBL(name) pcre_##name  #define PUBL(name) pcre_##name
2563  #endif  #endif
2564  #ifndef PRIV  #ifndef PRIV
2565  #define PRIV(name) _pcre_##name  #define PRIV(name) _pcre_##name
2566  #endif  #endif
2567  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
 #ifdef COMPILE_PCRE16  
2568  #ifndef PUBL  #ifndef PUBL
2569  #define PUBL(name) pcre16_##name  #define PUBL(name) pcre16_##name
2570  #endif  #endif
2571  #ifndef PRIV  #ifndef PRIV
2572  #define PRIV(name) _pcre16_##name  #define PRIV(name) _pcre16_##name
2573  #endif  #endif
2574    #elif defined COMPILE_PCRE32
2575    #ifndef PUBL
2576    #define PUBL(name) pcre32_##name
2577    #endif
2578    #ifndef PRIV
2579    #define PRIV(name) _pcre32_##name
2580    #endif
2581  #else  #else
2582  #error Unsupported compiling mode  #error Unsupported compiling mode
2583  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
 #endif /* COMPILE_PCRE8 */  
2584    
2585  /* 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
2586  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 2233  but are not part of the PCRE public API. Line 2600  but are not part of the PCRE public API.
2600  pcre_tables.c module. */  pcre_tables.c module. */
2601    
2602  #ifdef COMPILE_PCRE8  #ifdef COMPILE_PCRE8
   
2603  extern const int            PRIV(utf8_table1)[];  extern const int            PRIV(utf8_table1)[];
2604  extern const int            PRIV(utf8_table1_size);  extern const int            PRIV(utf8_table1_size);
2605  extern const int            PRIV(utf8_table2)[];  extern const int            PRIV(utf8_table2)[];
2606  extern const int            PRIV(utf8_table3)[];  extern const int            PRIV(utf8_table3)[];
2607  extern const pcre_uint8     PRIV(utf8_table4)[];  extern const pcre_uint8     PRIV(utf8_table4)[];
   
2608  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE8 */
2609    
2610  extern const char           PRIV(utt_names)[];  extern const char           PRIV(utt_names)[];
2611  extern const ucp_type_table PRIV(utt)[];  extern const ucp_type_table PRIV(utt)[];
2612  extern const int            PRIV(utt_size);  extern const int            PRIV(utt_size);
2613    
2614    extern const pcre_uint8     PRIV(OP_lengths)[];
2615  extern const pcre_uint8     PRIV(default_tables)[];  extern const pcre_uint8     PRIV(default_tables)[];
2616    
2617  extern const pcre_uint8     PRIV(OP_lengths)[];  extern const pcre_uint32    PRIV(hspace_list)[];
2618    extern const pcre_uint32    PRIV(vspace_list)[];
2619    
2620    
2621  /* 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 2256  one of the exported public functions. Th Line 2623  one of the exported public functions. Th
2623  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2624    
2625  /* String comparison functions. */  /* String comparison functions. */
2626  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2627    
2628  #define STRCMP_UC_UC(str1, str2) \  #define STRCMP_UC_UC(str1, str2) \
2629    strcmp((char *)(str1), (char *)(str2))    strcmp((char *)(str1), (char *)(str2))
# Line 2268  sense, but are not part of the PCRE publ Line 2635  sense, but are not part of the PCRE publ
2635    strncmp((char *)(str1), (str2), (num))    strncmp((char *)(str1), (str2), (num))
2636  #define STRLEN_UC(str) strlen((const char *)str)  #define STRLEN_UC(str) strlen((const char *)str)
2637    
2638  #else  #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2639    
2640  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2641                             const pcre_uchar *);                             const pcre_uchar *);
# Line 2290  extern unsigned int      PRIV(strlen_uc) Line 2657  extern unsigned int      PRIV(strlen_uc)
2657    PRIV(strncmp_uc_c8)((str1), (str2), (num))    PRIV(strncmp_uc_c8)((str1), (str2), (num))
2658  #define STRLEN_UC(str) PRIV(strlen_uc)(str)  #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2659    
2660  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE[8|16|32] */
2661    
2662    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2663    
2664    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2665    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2666    
2667    #elif defined COMPILE_PCRE32
2668    
2669    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2670                               const pcre_uchar *);
2671    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2672                               const char *);
2673    
2674    #define STRCMP_UC_UC_TEST(str1, str2) \
2675      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2676    #define STRCMP_UC_C8_TEST(str1, str2) \
2677      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2678    
2679    #endif /* COMPILE_PCRE[8|16|32] */
2680    
2681  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2682  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
# Line 2299  extern int               PRIV(ord2utf)(p Line 2685  extern int               PRIV(ord2utf)(p
2685  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2686  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2687                             int *, BOOL);                             int *, BOOL);
2688  extern BOOL              PRIV(xclass)(int, const pcre_uchar *, BOOL);  extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2689    
2690  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2691  extern void              PRIV(jit_compile)(const REAL_PCRE *,  extern void              PRIV(jit_compile)(const REAL_PCRE *,
# Line 2317  typedef struct { Line 2703  typedef struct {
2703    pcre_uint8 script;     /* ucp_Arabic, etc. */    pcre_uint8 script;     /* ucp_Arabic, etc. */
2704    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2705    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2706      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2707    pcre_int32 other_case; /* offset to other case, or zero if none */    pcre_int32 other_case; /* offset to other case, or zero if none */
2708  } ucd_record;  } ucd_record;
2709    
2710    extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2711  extern const ucd_record  PRIV(ucd_records)[];  extern const ucd_record  PRIV(ucd_records)[];
2712  extern const pcre_uint8  PRIV(ucd_stage1)[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2713  extern const pcre_uint16 PRIV(ucd_stage2)[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2714  extern const int         PRIV(ucp_gentype)[];  extern const pcre_uint32 PRIV(ucp_gentype)[];
2715  extern const pcre_uint32 PRIV(ucp_gbtable)[];  extern const pcre_uint32 PRIV(ucp_gbtable)[];
2716  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2717  extern const int         PRIV(ucp_typerange)[];  extern const int         PRIV(ucp_typerange)[];
# Line 2334  extern const int         PRIV(ucp_typera Line 2722  extern const int         PRIV(ucp_typera
2722    
2723  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2724  #define GET_UCD(ch) (PRIV(ucd_records) + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2725          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2726          UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2727    
2728  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2729  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2730  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2731  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2732  #define UCD_OTHERCASE(ch)   (ch + GET_UCD(ch)->other_case)  #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2733    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2734    
2735  #endif /* SUPPORT_UCP */  #endif /* SUPPORT_UCP */
2736    

Legend:
Removed from v.1022  
changed lines
  Added in v.1128

  ViewVC Help
Powered by ViewVC 1.1.5