/[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 836 by ph10, Wed Dec 28 17:16:11 2011 UTC revision 1145 by ph10, Fri Oct 19 16:25:32 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  
57    #if !defined COMPILE_PCRE16 && !defined COMPILE_PCRE32
58  #define COMPILE_PCRE8  #define COMPILE_PCRE8
59  #endif  #endif
60    
# Line 78  Until then we define it if SUPPORT_UTF i Line 79  Until then we define it if SUPPORT_UTF i
79  #define SUPPORT_UTF8 1  #define SUPPORT_UTF8 1
80  #endif  #endif
81    
82  /* We do not support both EBCDIC and UTF-8/16 at the same time. The "configure"  /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
83  script prevents both being selected, but not everybody uses "configure". */  script prevents both being selected, but not everybody uses "configure". */
84    
85  #if defined EBCDIC && defined SUPPORT_UTF  #if defined EBCDIC && defined SUPPORT_UTF
86  #error The use of both EBCDIC and SUPPORT_UTF8/16 is not supported.  #error The use of both EBCDIC and SUPPORT_UTF is not supported.
87  #endif  #endif
88    
89  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
# Line 243  exactly 256 items. When the character is Line 244  exactly 256 items. When the character is
244  items, some check is needed before accessing these tables.  items, some check is needed before accessing these tables.
245  */  */
246    
247  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
248    
249  typedef unsigned char pcre_uchar;  typedef unsigned char pcre_uchar;
250  #define IN_UCHARS(x) (x)  #define IN_UCHARS(x) (x)
251  #define MAX_255(c) 1  #define MAX_255(c) 1
252  #define TABLE_GET(c, table, default) ((table)[c])  #define TABLE_GET(c, table, default) ((table)[c])
253    
254  #else  #elif defined COMPILE_PCRE16
255    
 #ifdef COMPILE_PCRE16  
256  #if USHRT_MAX != 65535  #if USHRT_MAX != 65535
257  /* This is a warning message. Change PCRE_SCHAR16 to a 16 bit data type in  /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
258  pcre.h(.in) and disable (comment out) this message. */  pcre.h(.in) and disable (comment out) this message. */
259  #error Warning: PCRE_SCHAR16 is not a 16 bit data type.  #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
260  #endif  #endif
261    
262  typedef pcre_uint16 pcre_uchar;  typedef pcre_uint16 pcre_uchar;
263  #define IN_UCHARS(x) ((x) << 1)  #define UCHAR_SHIFT (1)
264    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
265  #define MAX_255(c) ((c) <= 255u)  #define MAX_255(c) ((c) <= 255u)
266  #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))  #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
267    
268    #elif defined COMPILE_PCRE32
269    
270    typedef pcre_uint32 pcre_uchar;
271    #define UCHAR_SHIFT (2)
272    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
273    #define MAX_255(c) ((c) <= 255u)
274    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
275    
276    /* Assert that pcre_uchar32 is a 32-bit type */
277    typedef int __assert_pcre_uchar32_size[sizeof(pcre_uchar) == 4 ? 1 : -1];
278    
279  #else  #else
280  #error Unsupported compiling mode  #error Unsupported compiling mode
281  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
282    
283  /* 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
284  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 305  start/end of string field names are. */
305         &(NLBLOCK->nllen), utf)) \         &(NLBLOCK->nllen), utf)) \
306      : \      : \
307      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
308       (p)[0] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
309       (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
310      ) \      ) \
311    )    )
312    
# Line 309  start/end of string field names are. */ Line 319  start/end of string field names are. */
319         &(NLBLOCK->nllen), utf)) \         &(NLBLOCK->nllen), utf)) \
320      : \      : \
321      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
322       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
323       (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
324      ) \      ) \
325    )    )
326    
# Line 396  The macros are controlled by the value o Line 406  The macros are controlled by the value o
406  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
407  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
408    
409  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
410    
411  #if LINK_SIZE == 2  #if LINK_SIZE == 2
412    
# Line 441  is automated on Unix systems via the "co Line 451  is automated on Unix systems via the "co
451  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
452  #endif  #endif
453    
454  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
   
 #ifdef COMPILE_PCRE16  
455    
456  #if LINK_SIZE == 2  #if LINK_SIZE == 2
457    
458    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
459  #undef LINK_SIZE  #undef LINK_SIZE
460  #define LINK_SIZE 1  #define LINK_SIZE 1
461    
# Line 460  is automated on Unix systems via the "co Line 469  is automated on Unix systems via the "co
469    
470  #elif LINK_SIZE == 3 || LINK_SIZE == 4  #elif LINK_SIZE == 3 || LINK_SIZE == 4
471    
472    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
473  #undef LINK_SIZE  #undef LINK_SIZE
474  #define LINK_SIZE 2  #define LINK_SIZE 2
475    
# Line 477  is automated on Unix systems via the "co Line 487  is automated on Unix systems via the "co
487  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
488  #endif  #endif
489    
490    #elif defined COMPILE_PCRE32
491    
492    /* Only supported LINK_SIZE is 4 */
493    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
494    #undef LINK_SIZE
495    #define LINK_SIZE 1
496    
497    #define PUT(a,n,d)   \
498      (a[n] = (d))
499    
500    #define GET(a,n) \
501      (a[n])
502    
503    /* Keep it positive */
504    #define MAX_PATTERN_SIZE (1 << 30)
505    
506  #else  #else
507  #error Unsupported compiling mode  #error Unsupported compiling mode
508  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
509    
510  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
511    
# Line 492  is automated on Unix systems via the "co Line 516  is automated on Unix systems via the "co
516  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
517  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
518    
519  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
520    
521  #define IMM2_SIZE 2  #define IMM2_SIZE 2
522    
523  #define PUT2(a,n,d)   \  #define PUT2(a,n,d)   \
524    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
525    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
526    
527    /* For reasons that I do not understand, the expression in this GET2 macro is
528    treated by gcc as a signed expression, even when a is declared as unsigned. It
529    seems that any kind of arithmetic results in a signed value. */
530    
531  #define GET2(a,n) \  #define GET2(a,n) \
532    (((a)[n] << 8) | (a)[(n)+1])    (unsigned int)(((a)[n] << 8) | (a)[(n)+1])
533    
534  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
535    
536  #ifdef COMPILE_PCRE16  #define IMM2_SIZE 1
537    
538    #define PUT2(a,n,d)   \
539       a[n] = d
540    
541    #define GET2(a,n) \
542       a[n]
543    
544    #elif defined COMPILE_PCRE32
545    
546  #define IMM2_SIZE 1  #define IMM2_SIZE 1
547    
# Line 517  capturing parenthesis numbers in back re Line 553  capturing parenthesis numbers in back re
553    
554  #else  #else
555  #error Unsupported compiling mode  #error Unsupported compiling mode
556  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
557    
558  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
559    
560    /* The maximum length of a MARK name is currently one data unit; it may be
561    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
562    
563    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
564    #define MAX_MARK ((1u << 16) - 1)
565    #else
566    #define MAX_MARK ((1u << 8) - 1)
567    #endif
568    
569  /* 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
570  character. The macros for character handling generate simple sequences when  byte. The macros for character handling generate simple sequences when used in
571  used in character-mode, and more complicated ones for UTF characters.  character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
572  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
573  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,
574  UTF support is omitted, we don't even define them. */  we don't even define them. */
575    
576  #ifndef SUPPORT_UTF  #ifndef SUPPORT_UTF
577    
# Line 541  UTF support is omitted, we don't even de Line 584  UTF support is omitted, we don't even de
584  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
585  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
586  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
587    #define RAWUCHAR(eptr) (*(eptr))
588    #define RAWUCHARINC(eptr) (*(eptr)++)
589    #define RAWUCHARTEST(eptr) (*(eptr))
590    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
591  /* #define GETCHARLENTEST(c, eptr, len) */  /* #define GETCHARLENTEST(c, eptr, len) */
592  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
593  /* #define FORWARDCHAR(eptr) */  /* #define FORWARDCHAR(eptr) */
# Line 548  UTF support is omitted, we don't even de Line 595  UTF support is omitted, we don't even de
595    
596  #else   /* SUPPORT_UTF */  #else   /* SUPPORT_UTF */
597    
 #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  
   
598  /* Tests whether the code point needs extra characters to decode. */  /* Tests whether the code point needs extra characters to decode. */
599    
600  #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)  
601    
602  /* 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
603  advancing the pointer. */  advancing the pointer. */
# Line 595  advancing the pointer. */ Line 621  advancing the pointer. */
621            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
622      }      }
623    
 /* 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);  
   
624  /* 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
625  the pointer. */  the pointer. */
626    
# Line 643  the pointer. */ Line 655  the pointer. */
655        } \        } \
656      }      }
657    
658    #if defined COMPILE_PCRE8
659    
660    /* These macros were originally written in the form of loops that used data
661    from the tables whose names start with PRIV(utf8_table). They were rewritten by
662    a user so as not to use loops, because in some environments this gives a
663    significant performance advantage, and it seems never to do any harm. */
664    
665    /* Tells the biggest code point which can be encoded as a single character. */
666    
667    #define MAX_VALUE_FOR_SINGLE_CHAR 127
668    
669    /* Tests whether the code point needs extra characters to decode. */
670    
671    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
672    
673    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
674    Otherwise it has an undefined behaviour. */
675    
676    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
677    
678    /* Returns TRUE, if the given character is not the first character
679    of a UTF sequence. */
680    
681    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
682    
683    /* Get the next UTF-8 character, not advancing the pointer. This is called when
684    we know we are in UTF-8 mode. */
685    
686    #define GETCHAR(c, eptr) \
687      c = *eptr; \
688      if (c >= 0xc0) GETUTF8(c, eptr);
689    
690    /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
691    pointer. */
692    
693    #define GETCHARTEST(c, eptr) \
694      c = *eptr; \
695      if (utf && c >= 0xc0) GETUTF8(c, eptr);
696    
697  /* 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
698  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
699    
# Line 709  do not know if we are in UTF-8 mode. */ Line 760  do not know if we are in UTF-8 mode. */
760    c = *eptr; \    c = *eptr; \
761    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
762    
763    /* Returns the next uchar, not advancing the pointer. This is called when
764    we know we are in UTF mode. */
765    
766    #define RAWUCHAR(eptr) \
767      (*(eptr))
768    
769    /* Returns the next uchar, advancing the pointer. This is called when
770    we know we are in UTF mode. */
771    
772    #define RAWUCHARINC(eptr) \
773      (*(eptr)++)
774    
775    /* Returns the next uchar, testing for UTF mode, and not advancing the
776    pointer. */
777    
778    #define RAWUCHARTEST(eptr) \
779      (*(eptr))
780    
781    /* Returns the next uchar, testing for UTF mode, advancing the
782    pointer. */
783    
784    #define RAWUCHARINCTEST(eptr) \
785      (*(eptr)++)
786    
787  /* 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
788  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
789  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 797  because almost all calls are already wit
797  #define ACROSSCHAR(condition, eptr, action) \  #define ACROSSCHAR(condition, eptr, action) \
798    while((condition) && ((eptr) & 0xc0) == 0x80) action    while((condition) && ((eptr) & 0xc0) == 0x80) action
799    
800  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
   
 #ifdef COMPILE_PCRE16  
801    
802  /* 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. */
803    
# Line 806  we do not know if we are in UTF-16 mode. Line 879  we do not know if we are in UTF-16 mode.
879    c = *eptr; \    c = *eptr; \
880    if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);    if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
881    
882    /* Returns the next uchar, not advancing the pointer. This is called when
883    we know we are in UTF mode. */
884    
885    #define RAWUCHAR(eptr) \
886      (*(eptr))
887    
888    /* Returns the next uchar, advancing the pointer. This is called when
889    we know we are in UTF mode. */
890    
891    #define RAWUCHARINC(eptr) \
892      (*(eptr)++)
893    
894    /* Returns the next uchar, testing for UTF mode, and not advancing the
895    pointer. */
896    
897    #define RAWUCHARTEST(eptr) \
898      (*(eptr))
899    
900    /* Returns the next uchar, testing for UTF mode, advancing the
901    pointer. */
902    
903    #define RAWUCHARINCTEST(eptr) \
904      (*(eptr)++)
905    
906  /* 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
907  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
908  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 917  code. */
917  #define ACROSSCHAR(condition, eptr, action) \  #define ACROSSCHAR(condition, eptr, action) \
918    if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action    if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
919    
920  #endif  #elif defined COMPILE_PCRE32
921    
922  #endif /* COMPILE_PCRE8 */  /* These are trivial for the 32-bit library, since all UTF-32 characters fit
923    into one pcre_uchar unit. */
924    #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
925    #define HAS_EXTRALEN(c) (0)
926    #define GET_EXTRALEN(c) (0)
927    #define NOT_FIRSTCHAR(c) (0)
928    
929    #define UTF32_MASK (0x1fffffu)
930    
931    /* Get the next UTF-32 character, not advancing the pointer. This is called when
932    we know we are in UTF-32 mode. */
933    
934    #define GETCHAR(c, eptr) \
935      c = (*eptr) & UTF32_MASK;
936    
937    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
938    pointer. */
939    
940    #define GETCHARTEST(c, eptr) \
941      c = *eptr; \
942      if (utf) c &= UTF32_MASK;
943    
944    /* Get the next UTF-32 character, advancing the pointer. This is called when we
945    know we are in UTF-32 mode. */
946    
947    #define GETCHARINC(c, eptr) \
948      c = (*eptr++) & UTF32_MASK;
949    
950    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
951    This is called when we don't know if we are in UTF-32 mode. */
952    
953    #define GETCHARINCTEST(c, eptr) \
954      c = *eptr++; \
955      if (utf) c &= UTF32_MASK;
956    
957    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
958    length (since all UTF-32 is of length 1). This is called when we know we are in
959    UTF-32 mode. */
960    
961    #define GETCHARLEN(c, eptr, len) \
962      GETCHAR(c, eptr)
963    
964    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
965    pointer, not incrementing the length (since all UTF-32 is of length 1).
966    This is called when we do not know if we are in UTF-32 mode. */
967    
968    #define GETCHARLENTEST(c, eptr, len) \
969      GETCHARTEST(c, eptr)
970    
971    /* Returns the next uchar, not advancing the pointer. This is called when
972    we know we are in UTF mode. */
973    
974    #define RAWUCHAR(eptr) \
975      (*(eptr) & UTF32_MASK)
976    
977    /* Returns the next uchar, advancing the pointer. This is called when
978    we know we are in UTF mode. */
979    
980    #define RAWUCHARINC(eptr) \
981      (*(eptr)++ & UTF32_MASK)
982    
983    /* Returns the next uchar, testing for UTF mode, and not advancing the
984    pointer. */
985    
986    #define RAWUCHARTEST(eptr) \
987      (utf ? (*(eptr) & UTF32_MASK) : *(eptr))
988    
989    /* Returns the next uchar, testing for UTF mode, advancing the
990    pointer. */
991    
992    #define RAWUCHARINCTEST(eptr) \
993      (utf ? (*(eptr)++ & UTF32_MASK) : *(eptr)++)
994    
995    /* If the pointer is not at the start of a character, move it back until
996    it is. This is called only in UTF-32 mode - we don't put a test within the
997    macro because almost all calls are already within a block of UTF-32 only
998    code.
999    These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
1000    
1001    #define BACKCHAR(eptr) do { } while (0)
1002    
1003    /* Same as above, just in the other direction. */
1004    #define FORWARDCHAR(eptr) do { } while (0)
1005    
1006    /* Same as above, but it allows a fully customizable form. */
1007    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1008    
1009    #else
1010    #error Unsupported compiling mode
1011    #endif /* COMPILE_PCRE[8|16|32] */
1012    
1013  #endif  /* SUPPORT_UTF */  #endif  /* SUPPORT_UTF */
1014    
1015    /* Tests for Unicode horizontal and vertical whitespace characters must check a
1016    number of different values. Using a switch statement for this generates the
1017    fastest code (no loop, no memory access), and there are several places in the
1018    interpreter code where this happens. In order to ensure that all the case lists
1019    remain in step, we use macros so that there is only one place where the lists
1020    are defined.
1021    
1022    These values are also required as lists in pcre_compile.c when processing \h,
1023    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1024    macros that define the values are here so that all the definitions are
1025    together. The lists must be in ascending character order, terminated by
1026    NOTACHAR (which is 0xffffffff).
1027    
1028    Any changes should ensure that the various macros are kept in step with each
1029    other. NOTE: The values also appear in pcre_jit_compile.c. */
1030    
1031    /* ------ ASCII/Unicode environments ------ */
1032    
1033    #ifndef EBCDIC
1034    
1035    #define HSPACE_LIST \
1036      CHAR_HT, CHAR_SPACE, 0xa0, \
1037      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1038      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1039      NOTACHAR
1040    
1041    #define HSPACE_MULTIBYTE_CASES \
1042      case 0x1680:  /* OGHAM SPACE MARK */ \
1043      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1044      case 0x2000:  /* EN QUAD */ \
1045      case 0x2001:  /* EM QUAD */ \
1046      case 0x2002:  /* EN SPACE */ \
1047      case 0x2003:  /* EM SPACE */ \
1048      case 0x2004:  /* THREE-PER-EM SPACE */ \
1049      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1050      case 0x2006:  /* SIX-PER-EM SPACE */ \
1051      case 0x2007:  /* FIGURE SPACE */ \
1052      case 0x2008:  /* PUNCTUATION SPACE */ \
1053      case 0x2009:  /* THIN SPACE */ \
1054      case 0x200A:  /* HAIR SPACE */ \
1055      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1056      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1057      case 0x3000   /* IDEOGRAPHIC SPACE */
1058    
1059    #define HSPACE_BYTE_CASES \
1060      case CHAR_HT: \
1061      case CHAR_SPACE: \
1062      case 0xa0     /* NBSP */
1063    
1064    #define HSPACE_CASES \
1065      HSPACE_BYTE_CASES: \
1066      HSPACE_MULTIBYTE_CASES
1067    
1068    #define VSPACE_LIST \
1069      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1070    
1071    #define VSPACE_MULTIBYTE_CASES \
1072      case 0x2028:    /* LINE SEPARATOR */ \
1073      case 0x2029     /* PARAGRAPH SEPARATOR */
1074    
1075    #define VSPACE_BYTE_CASES \
1076      case CHAR_LF: \
1077      case CHAR_VT: \
1078      case CHAR_FF: \
1079      case CHAR_CR: \
1080      case CHAR_NEL
1081    
1082    #define VSPACE_CASES \
1083      VSPACE_BYTE_CASES: \
1084      VSPACE_MULTIBYTE_CASES
1085    
1086  /* In case there is no definition of offsetof() provided - though any proper  /* ------ EBCDIC environments ------ */
1087  Standard C system should have one. */  
1088    #else
1089    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1090    
1091    #define HSPACE_BYTE_CASES \
1092      case CHAR_HT: \
1093      case CHAR_SPACE
1094    
1095    #define HSPACE_CASES HSPACE_BYTE_CASES
1096    
1097    #ifdef EBCDIC_NL25
1098    #define VSPACE_LIST \
1099      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1100    #else
1101    #define VSPACE_LIST \
1102      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1103    #endif
1104    
1105    #define VSPACE_BYTE_CASES \
1106      case CHAR_LF: \
1107      case CHAR_VT: \
1108      case CHAR_FF: \
1109      case CHAR_CR: \
1110      case CHAR_NEL
1111    
1112    #define VSPACE_CASES VSPACE_BYTE_CASES
1113    #endif  /* EBCDIC */
1114    
1115    /* ------ End of whitespace macros ------ */
1116    
 #ifndef offsetof  
 #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
 #endif  
1117    
1118    
1119  /* 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 1122  are in a 16-bit flags word. From release
1122  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1123  compatibility. */  compatibility. */
1124    
1125  #ifdef COMPILE_PCRE8  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1126  #define PCRE_MODE          0x0001  /* compiled in 8 bit mode */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1127  #endif  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
 #ifdef COMPILE_PCRE16  
 #define PCRE_MODE          0x0002  /* compiled in 16 bit mode */  
 #endif  
1128  #define PCRE_FIRSTSET      0x0010  /* first_char is set */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1129  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1130  #define PCRE_REQCHSET      0x0040  /* req_byte is set */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
# Line 857  compatibility. */ Line 1135  compatibility. */
1135  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1136  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1137    
1138    #if defined COMPILE_PCRE8
1139    #define PCRE_MODE          PCRE_MODE8
1140    #elif defined COMPILE_PCRE16
1141    #define PCRE_MODE          PCRE_MODE16
1142    #elif defined COMPILE_PCRE32
1143    #define PCRE_MODE          PCRE_MODE32
1144    #endif
1145    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1146    
1147  /* Flags for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1148    
1149  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
# Line 887  time, run time, or study time, respectiv Line 1174  time, run time, or study time, respectiv
1174     PCRE_NO_START_OPTIMIZE)     PCRE_NO_START_OPTIMIZE)
1175    
1176  #define PUBLIC_STUDY_OPTIONS \  #define PUBLIC_STUDY_OPTIONS \
1177     PCRE_STUDY_JIT_COMPILE     (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1178        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1179    
1180  /* Magic number to provide a small check against being handed junk. */  /* Magic number to provide a small check against being handed junk. */
1181    
# Line 898  in different endianness. */ Line 1186  in different endianness. */
1186    
1187  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
1188    
 /* Negative values for the firstchar and reqchar variables */  
   
 #define REQ_UNSET (-2)  
 #define REQ_NONE  (-1)  
   
1189  /* 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
1190  req_byte match. */  req_byte match. */
1191    
# Line 939  macros to give the functions distinct na Line 1222  macros to give the functions distinct na
1222  #ifndef SUPPORT_UTF  #ifndef SUPPORT_UTF
1223    
1224  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1225  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
1226    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1227    characters, a common practice has been to use its NL (0x15) character as the
1228    line terminator in C-like processing environments. However, sometimes the LF
1229    (0x25) character is used instead, according to this Unicode document:
1230    
1231    http://unicode.org/standard/reports/tr13/tr13-5.html
1232    
1233    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1234    instead. Whichever is *not* chosen is defined as NEL.
1235    
1236    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1237    same code point. */
1238    
1239    #ifdef EBCDIC
1240    
1241    #ifndef EBCDIC_NL25
1242    #define CHAR_NL                     '\x15'
1243    #define CHAR_NEL                    '\x25'
1244    #define STR_NL                      "\x15"
1245    #define STR_NEL                     "\x25"
1246    #else
1247    #define CHAR_NL                     '\x25'
1248    #define CHAR_NEL                    '\x15'
1249    #define STR_NL                      "\x25"
1250    #define STR_NEL                     "\x15"
1251    #endif
1252    
1253    #define CHAR_LF                     CHAR_NL
1254    #define STR_LF                      STR_NL
1255    
1256    #define CHAR_ESC                    '\047'
1257    #define CHAR_DEL                    '\007'
1258    #define STR_ESC                     "\047"
1259    #define STR_DEL                     "\007"
1260    
1261    #else  /* Not EBCDIC */
1262    
1263    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1264    compatibility. NEL is the Unicode newline character; make sure it is
1265    a positive value. */
1266    
1267    #define CHAR_LF                     '\n'
1268    #define CHAR_NL                     CHAR_LF
1269    #define CHAR_NEL                    ((unsigned char)'\x85')
1270    #define CHAR_ESC                    '\033'
1271    #define CHAR_DEL                    '\177'
1272    
1273    #define STR_LF                      "\n"
1274    #define STR_NL                      STR_LF
1275    #define STR_NEL                     "\x85"
1276    #define STR_ESC                     "\033"
1277    #define STR_DEL                     "\177"
1278    
1279    #endif  /* EBCDIC */
1280    
1281    /* The remaining definitions work in both environments. */
1282    
1283  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1284  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1285  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1286  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1287  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1288  #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  
1289    
1290  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1291  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 1056  so that PCRE works on both ASCII and EBC Line 1387  so that PCRE works on both ASCII and EBC
1387  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1388  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1389  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1390  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1391  #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  
1392    
1393  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1394  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 1202  so that PCRE works on both ASCII and EBC Line 1525  so that PCRE works on both ASCII and EBC
1525  #ifdef COMPILE_PCRE16  #ifdef COMPILE_PCRE16
1526  #define STRING_UTF_RIGHTPAR            "UTF16)"  #define STRING_UTF_RIGHTPAR            "UTF16)"
1527  #endif  #endif
1528    #ifdef COMPILE_PCRE32
1529    #define STRING_UTF_RIGHTPAR            "UTF32)"
1530    #endif
1531  #define STRING_UCP_RIGHTPAR            "UCP)"  #define STRING_UCP_RIGHTPAR            "UCP)"
1532  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1533    
# Line 1215  only. */ Line 1541  only. */
1541  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1542  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1543  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1544  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1545    #define CHAR_NL                     CHAR_LF
1546    #define CHAR_NEL                    ((unsigned char)'\x85')
1547  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1548  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1549  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
# Line 1462  only. */ Line 1790  only. */
1790  #ifdef COMPILE_PCRE16  #ifdef COMPILE_PCRE16
1791  #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
1792  #endif  #endif
1793    #ifdef COMPILE_PCRE32
1794    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1795    #endif
1796  #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS  #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1797  #define STRING_NO_START_OPT_RIGHTPAR   STR_N STR_O STR_UNDERSCORE STR_S STR_T STR_A STR_R STR_T STR_UNDERSCORE STR_O STR_P STR_T STR_RIGHT_PARENTHESIS  #define STRING_NO_START_OPT_RIGHTPAR   STR_N STR_O STR_UNDERSCORE STR_S STR_T STR_A STR_R STR_T STR_UNDERSCORE STR_O STR_P STR_T STR_RIGHT_PARENTHESIS
1798    
# Line 1478  only. */ Line 1809  only. */
1809  #endif  #endif
1810    
1811  #ifndef ESC_n  #ifndef ESC_n
1812  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1813  #endif  #endif
1814    
1815  #ifndef ESC_r  #ifndef ESC_r
# Line 1503  only. */ Line 1834  only. */
1834  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1835  #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */  #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1836  #define PT_WORD       8    /* Word - L plus N plus underscore */  #define PT_WORD       8    /* Word - L plus N plus underscore */
1837    #define PT_CLIST      9    /* Pseudo-property: match character list */
1838    
1839  /* 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
1840  contain characters with values greater than 255. */  contain characters with values greater than 255. */
# Line 1518  contain characters with values greater t Line 1850  contain characters with values greater t
1850    
1851  /* 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
1852  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
1853  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
1854  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
1855  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
1856  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
1857  non-DOTALL mode, "." behaves like \N.  non-DOTALL mode, "." behaves like \N.
1858    
1859  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.
1860  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.
1861  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
1862  looked up from a table.  looked up from a table.
1863    
1864  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
1865  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
1866  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
1867  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
1868  put in between that don't consume a character, that code will have to change.  put in between that don't consume a character, that code will have to change.
# Line 1540  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E Line 1872  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E
1872         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,
1873         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1874         ESC_E, ESC_Q, ESC_g, ESC_k,         ESC_E, ESC_Q, ESC_g, ESC_k,
1875         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 };  
1876    
1877  /* 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
1878  OP_EOD must correspond in order to the list of escapes immediately above.  OP_EOD must correspond in order to the list of escapes immediately above.
# Line 1567  enum { Line 1898  enum {
1898    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1899    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1900    
1901    OP_ANY,            /* 12 Match any character except newline */    OP_ANY,            /* 12 Match any character except newline (\N) */
1902    OP_ALLANY,         /* 13 Match any character */    OP_ALLANY,         /* 13 Match any character */
1903    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 */
1904    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
# Line 1578  enum { Line 1909  enum {
1909    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1910    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1911    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1912    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) */
1913    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1914    
1915    OP_CIRC,           /* 25 Start of line - not multiline */    OP_CIRC,           /* 25 Start of line - not multiline */
1916    OP_CIRCM,          /* 26 Start of line - multiline */    OP_CIRCM,          /* 26 Start of line - multiline */
# Line 1939  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2270  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2270         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2271         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2272         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2273         ERR70, ERR71, ERR72, ERR73, ERRCOUNT };         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2274    
2275    /* JIT compiling modes. The function list is indexed by them. */
2276    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2277           JIT_NUMBER_OF_COMPILE_MODES };
2278    
2279  /* The real format of the start of the pcre block; the index of names and the  /* The real format of the start of the pcre block; the index of names and the
2280  code vector run on as long as necessary after the end. We store an explicit  code vector run on as long as necessary after the end. We store an explicit
# Line 1958  fields are present. Currently PCRE alway Line 2293  fields are present. Currently PCRE alway
2293  NOTE NOTE NOTE  NOTE NOTE NOTE
2294  */  */
2295    
2296  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2297    #define REAL_PCRE real_pcre
2298    #elif defined COMPILE_PCRE16
2299    #define REAL_PCRE real_pcre16
2300    #elif defined COMPILE_PCRE32
2301    #define REAL_PCRE real_pcre32
2302    #endif
2303    
2304    /* It is necessary to fork the struct for 32 bit, since it needs to use
2305     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2306     * typedef since pcretest needs access to  the struct of the 8-, 16-
2307     * and 32-bit variants. */
2308    
2309    typedef struct real_pcre8_or_16 {
2310    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2311    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2312    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
2313    pcre_uint16 flags;              /* Private flags */    pcre_uint16 flags;              /* Private flags */
2314    pcre_uint16 dummy1;             /* For future use */    pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2315    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2316    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2317    pcre_uint16 first_char;         /* Starting character */    pcre_uint16 first_char;         /* Starting character */
2318    pcre_uint16 req_char;           /* This character must be seen */    pcre_uint16 req_char;           /* This character must be seen */
2319    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2320    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2321    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2322    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
   
2323    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2324    const pcre_uint8 *nullpad;      /* NULL padding */    const pcre_uint8 *nullpad;      /* NULL padding */
2325  } real_pcre;  } real_pcre8_or_16;
2326    
2327    typedef struct real_pcre8_or_16 real_pcre;
2328    typedef struct real_pcre8_or_16 real_pcre16;
2329    
2330    typedef struct real_pcre32 {
2331      pcre_uint32 magic_number;
2332      pcre_uint32 size;               /* Total that was malloced */
2333      pcre_uint32 options;            /* Public options */
2334      pcre_uint16 flags;              /* Private flags */
2335      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2336      pcre_uint16 top_bracket;        /* Highest numbered group */
2337      pcre_uint16 top_backref;        /* Highest numbered back reference */
2338      pcre_uint32 first_char;         /* Starting character */
2339      pcre_uint32 req_char;           /* This character must be seen */
2340      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2341      pcre_uint16 name_entry_size;    /* Size of any name items */
2342      pcre_uint16 name_count;         /* Number of name items */
2343      pcre_uint16 ref_count;          /* Reference count */
2344      pcre_uint16 dummy1;             /* for later expansion */
2345      pcre_uint16 dummy2;             /* for later expansion */
2346      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2347      void *nullpad;                  /* for later expansion */
2348    } real_pcre32;
2349    
2350    /* Assert that the size of REAL_PCRE is divisible by 8 */
2351    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2352    
2353    /* Needed in pcretest to access some fields in the real_pcre* structures
2354     * directly. They're unified for 8/16/32 bits since the structs only differ
2355     * after these fields; if that ever changes, need to fork those defines into
2356     * 8/16 and 32 bit versions. */
2357    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2358    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2359    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2360    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2361    
2362  /* 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
2363  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 2016  typedef struct compile_data { Line 2398  typedef struct compile_data {
2398    int  names_found;                 /* Number of entries so far */    int  names_found;                 /* Number of entries so far */
2399    int  name_entry_size;             /* Size of each entry */    int  name_entry_size;             /* Size of each entry */
2400    int  workspace_size;              /* Size of workspace */    int  workspace_size;              /* Size of workspace */
2401    int  bracount;                    /* Count of capturing parens as we compile */    unsigned int  bracount;           /* Count of capturing parens as we compile */
2402    int  final_bracount;              /* Saved value after first pass */    int  final_bracount;              /* Saved value after first pass */
2403      int  max_lookbehind;              /* Maximum lookbehind (characters) */
2404    int  top_backref;                 /* Maximum back reference */    int  top_backref;                 /* Maximum back reference */
2405    unsigned int backref_map;         /* Bitmap of low back refs */    unsigned int backref_map;         /* Bitmap of low back refs */
2406    int  assert_depth;                /* Depth of nested assertions */    int  assert_depth;                /* Depth of nested assertions */
# Line 2025  typedef struct compile_data { Line 2408  typedef struct compile_data {
2408    int  external_flags;              /* External flag bits to be set */    int  external_flags;              /* External flag bits to be set */
2409    int  req_varyopt;                 /* "After variable item" flag for reqbyte */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2410    BOOL had_accept;                  /* (*ACCEPT) encountered */    BOOL had_accept;                  /* (*ACCEPT) encountered */
2411      BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2412    BOOL check_lookbehind;            /* Lookbehinds need later checking */    BOOL check_lookbehind;            /* Lookbehinds need later checking */
2413    int  nltype;                      /* Newline type */    int  nltype;                      /* Newline type */
2414    int  nllen;                       /* Newline string length */    int  nllen;                       /* Newline string length */
# Line 2044  call within the pattern; used by pcre_ex Line 2428  call within the pattern; used by pcre_ex
2428    
2429  typedef struct recursion_info {  typedef struct recursion_info {
2430    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2431    int group_num;                  /* Number of group that was called */    unsigned int group_num;         /* Number of group that was called */
2432    int *offset_save;               /* Pointer to start of saved offsets */    int *offset_save;               /* Pointer to start of saved offsets */
2433    int saved_max;                  /* Number of saved offsets */    int saved_max;                  /* Number of saved offsets */
2434    PCRE_PUCHAR subject_position;   /* Position at start of recursion */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
# Line 2119  typedef struct match_data { Line 2503  typedef struct match_data {
2503    const  pcre_uchar *mark;        /* Mark pointer to pass back on success */    const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2504    const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */    const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2505    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 */
2506    #ifdef NO_RECURSE
2507      void  *match_frames_base;       /* For remembering malloc'd frames */
2508    #endif
2509  } match_data;  } match_data;
2510    
2511  /* A similar structure is used for the same purpose by the DFA matching  /* A similar structure is used for the same purpose by the DFA matching
# Line 2173  total length. */ Line 2560  total length. */
2560  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2561  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2562    
2563  /* Internal function prefix */  /* Internal function and data prefixes. */
2564    
2565  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2566    #ifndef PUBL
2567  #define PUBL(name) pcre_##name  #define PUBL(name) pcre_##name
2568    #endif
2569    #ifndef PRIV
2570  #define PRIV(name) _pcre_##name  #define PRIV(name) _pcre_##name
2571  #else  #endif
2572  #ifdef COMPILE_PCRE16  #elif defined COMPILE_PCRE16
2573    #ifndef PUBL
2574  #define PUBL(name) pcre16_##name  #define PUBL(name) pcre16_##name
2575    #endif
2576    #ifndef PRIV
2577  #define PRIV(name) _pcre16_##name  #define PRIV(name) _pcre16_##name
2578    #endif
2579    #elif defined COMPILE_PCRE32
2580    #ifndef PUBL
2581    #define PUBL(name) pcre32_##name
2582    #endif
2583    #ifndef PRIV
2584    #define PRIV(name) _pcre32_##name
2585    #endif
2586  #else  #else
2587  #error Unsupported compiling mode  #error Unsupported compiling mode
2588  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
 #endif /* COMPILE_PCRE8 */  
2589    
2590  /* 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
2591  codes. Each entry used to point directly to a name, but to reduce the number of  codes. Each entry used to point directly to a name, but to reduce the number of
# Line 2205  but are not part of the PCRE public API. Line 2605  but are not part of the PCRE public API.
2605  pcre_tables.c module. */  pcre_tables.c module. */
2606    
2607  #ifdef COMPILE_PCRE8  #ifdef COMPILE_PCRE8
   
2608  extern const int            PRIV(utf8_table1)[];  extern const int            PRIV(utf8_table1)[];
2609  extern const int            PRIV(utf8_table1_size);  extern const int            PRIV(utf8_table1_size);
2610  extern const int            PRIV(utf8_table2)[];  extern const int            PRIV(utf8_table2)[];
2611  extern const int            PRIV(utf8_table3)[];  extern const int            PRIV(utf8_table3)[];
2612  extern const pcre_uint8     PRIV(utf8_table4)[];  extern const pcre_uint8     PRIV(utf8_table4)[];
   
2613  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE8 */
2614    
2615  extern const char           PRIV(utt_names)[];  extern const char           PRIV(utt_names)[];
2616  extern const ucp_type_table PRIV(utt)[];  extern const ucp_type_table PRIV(utt)[];
2617  extern const int            PRIV(utt_size);  extern const int            PRIV(utt_size);
2618    
2619    extern const pcre_uint8     PRIV(OP_lengths)[];
2620  extern const pcre_uint8     PRIV(default_tables)[];  extern const pcre_uint8     PRIV(default_tables)[];
2621    
2622  extern const pcre_uint8     PRIV(OP_lengths)[];  extern const pcre_uint32    PRIV(hspace_list)[];
2623    extern const pcre_uint32    PRIV(vspace_list)[];
2624    
2625    
2626  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
# Line 2228  one of the exported public functions. Th Line 2628  one of the exported public functions. Th
2628  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2629    
2630  /* String comparison functions. */  /* String comparison functions. */
2631  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2632    
2633  #define STRCMP_UC_UC(str1, str2) \  #define STRCMP_UC_UC(str1, str2) \
2634    strcmp((char *)(str1), (char *)(str2))    strcmp((char *)(str1), (char *)(str2))
# Line 2240  sense, but are not part of the PCRE publ Line 2640  sense, but are not part of the PCRE publ
2640    strncmp((char *)(str1), (str2), (num))    strncmp((char *)(str1), (str2), (num))
2641  #define STRLEN_UC(str) strlen((const char *)str)  #define STRLEN_UC(str) strlen((const char *)str)
2642    
2643  #else  #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2644    
2645  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2646                             const pcre_uchar *);                             const pcre_uchar *);
# Line 2262  extern unsigned int      PRIV(strlen_uc) Line 2662  extern unsigned int      PRIV(strlen_uc)
2662    PRIV(strncmp_uc_c8)((str1), (str2), (num))    PRIV(strncmp_uc_c8)((str1), (str2), (num))
2663  #define STRLEN_UC(str) PRIV(strlen_uc)(str)  #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2664    
2665  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE[8|16|32] */
2666    
2667    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2668    
2669    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2670    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2671    
2672    #elif defined COMPILE_PCRE32
2673    
2674    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2675                               const pcre_uchar *);
2676    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2677                               const char *);
2678    
2679    #define STRCMP_UC_UC_TEST(str1, str2) \
2680      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2681    #define STRCMP_UC_C8_TEST(str1, str2) \
2682      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2683    
2684    #endif /* COMPILE_PCRE[8|16|32] */
2685    
2686  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2687  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2688                             int *, BOOL);                             int *, BOOL);
2689  extern int               PRIV(ord2utf)(pcre_uint32, pcre_uchar *);  extern unsigned int      PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2690  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2691  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2692                             int *, BOOL);                             int *, BOOL);
2693  extern BOOL              PRIV(xclass)(int, const pcre_uchar *, BOOL);  extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2694    
2695  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2696  extern void              PRIV(jit_compile)(const real_pcre *, pcre_extra *);  extern void              PRIV(jit_compile)(const REAL_PCRE *,
2697  extern int               PRIV(jit_exec)(const real_pcre *, void *,                             PUBL(extra) *, int);
2698                             const pcre_uchar *, int, int, int, int, int *, int);  extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2699                               const pcre_uchar *, int, int, int, int *, int);
2700  extern void              PRIV(jit_free)(void *);  extern void              PRIV(jit_free)(void *);
2701  extern int               PRIV(jit_get_size)(void *);  extern int               PRIV(jit_get_size)(void *);
2702    extern const char*       PRIV(jit_get_target)(void);
2703  #endif  #endif
2704    
2705  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
2706    
2707  typedef struct {  typedef struct {
2708    pcre_uint8 script;    pcre_uint8 script;     /* ucp_Arabic, etc. */
2709    pcre_uint8 chartype;    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2710    pcre_int32 other_case;    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2711      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2712      pcre_int32 other_case; /* offset to other case, or zero if none */
2713  } ucd_record;  } ucd_record;
2714    
2715    extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2716  extern const ucd_record  PRIV(ucd_records)[];  extern const ucd_record  PRIV(ucd_records)[];
2717  extern const pcre_uint8  PRIV(ucd_stage1)[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2718  extern const pcre_uint16 PRIV(ucd_stage2)[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2719  extern const int         PRIV(ucp_gentype)[];  extern const pcre_uint32 PRIV(ucp_gentype)[];
2720    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2721  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2722  extern const int         PRIV(ucp_typerange)[];  extern const int         PRIV(ucp_typerange)[];
2723  #endif  #endif
# Line 2302  extern const int         PRIV(ucp_typera Line 2727  extern const int         PRIV(ucp_typera
2727    
2728  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2729  #define GET_UCD(ch) (PRIV(ucd_records) + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2730          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2731          UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2732    
2733  #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2734  #define UCD_SCRIPT(ch)    GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2735  #define UCD_CATEGORY(ch)  PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2736  #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2737    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2738    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2739    
2740  #endif /* SUPPORT_UCP */  #endif /* SUPPORT_UCP */
2741    

Legend:
Removed from v.836  
changed lines
  Added in v.1145

  ViewVC Help
Powered by ViewVC 1.1.5