/[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 164 by ph10, Fri May 4 15:11:14 2007 UTC revision 189 by ph10, Thu Jul 5 11:50:49 2007 UTC
# Line 234  must begin with PCRE_. */ Line 234  must begin with PCRE_. */
234  /* Include the public PCRE header and the definitions of UCP character property  /* Include the public PCRE header and the definitions of UCP character property
235  values. */  values. */
236    
237  #include <pcre.h>  #include "pcre.h"
238  #include "ucp.h"  #include "ucp.h"
239    
240  /* When compiling for use with the Virtual Pascal compiler, these functions  /* When compiling for use with the Virtual Pascal compiler, these functions
# Line 605  ESC_Z to detect the types that may be re Line 605  ESC_Z to detect the types that may be re
605  consume characters. If any new escapes are put in between that don't consume a  consume characters. If any new escapes are put in between that don't consume a
606  character, that code will have to change. */  character, that code will have to change. */
607    
608  enum { ESC_A = 1, ESC_G, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, ESC_W,  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
609         ESC_w, ESC_dum1, ESC_C, ESC_P, ESC_p, ESC_R, ESC_X, ESC_Z, ESC_z,         ESC_W, ESC_w, ESC_dum1, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H, ESC_h,
610         ESC_E, ESC_Q, ESC_k, ESC_REF };         ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_k, ESC_REF };
611    
612    
613  /* Opcode table: OP_BRA must be last, as all values >= it are used for brackets  /* Opcode table: OP_BRA must be last, as all values >= it are used for brackets
# Line 629  enum { Line 629  enum {
629    
630    OP_SOD,            /* 1 Start of data: \A */    OP_SOD,            /* 1 Start of data: \A */
631    OP_SOM,            /* 2 Start of match (subject + offset): \G */    OP_SOM,            /* 2 Start of match (subject + offset): \G */
632    OP_NOT_WORD_BOUNDARY,  /*  3 \B */    OP_SET_SOM,        /* 3 Set start of match (\K) */
633    OP_WORD_BOUNDARY,      /*  4 \b */    OP_NOT_WORD_BOUNDARY,  /*  4 \B */
634    OP_NOT_DIGIT,          /*  5 \D */    OP_WORD_BOUNDARY,      /*  5 \b */
635    OP_DIGIT,              /*  6 \d */    OP_NOT_DIGIT,          /*  6 \D */
636    OP_NOT_WHITESPACE,     /*  7 \S */    OP_DIGIT,              /*  7 \d */
637    OP_WHITESPACE,         /*  8 \s */    OP_NOT_WHITESPACE,     /*  8 \S */
638    OP_NOT_WORDCHAR,       /*  9 \W */    OP_WHITESPACE,         /*  9 \s */
639    OP_WORDCHAR,           /* 10 \w */    OP_NOT_WORDCHAR,       /* 10 \W */
640    OP_ANY,            /* 11 Match any character */    OP_WORDCHAR,           /* 11 \w */
641    OP_ANYBYTE,        /* 12 Match any byte (\C); different to OP_ANY for UTF-8 */    OP_ANY,            /* 12 Match any character */
642    OP_NOTPROP,        /* 13 \P (not Unicode property) */    OP_ANYBYTE,        /* 13 Match any byte (\C); different to OP_ANY for UTF-8 */
643    OP_PROP,           /* 14 \p (Unicode property) */    OP_NOTPROP,        /* 14 \P (not Unicode property) */
644    OP_ANYNL,          /* 15 \R (any newline sequence) */    OP_PROP,           /* 15 \p (Unicode property) */
645    OP_EXTUNI,         /* 16 \X (extended Unicode sequence */    OP_ANYNL,          /* 16 \R (any newline sequence) */
646    OP_EODN,           /* 17 End of data or \n at end of data: \Z. */    OP_NOT_HSPACE,     /* 17 \H (not horizontal whitespace) */
647    OP_EOD,            /* 18 End of data: \z */    OP_HSPACE,         /* 18 \h (horizontal whitespace) */
648      OP_NOT_VSPACE,     /* 19 \V (not vertical whitespace) */
649    OP_OPT,            /* 19 Set runtime options */    OP_VSPACE,         /* 20 \v (vertical whitespace) */
650    OP_CIRC,           /* 20 Start of line - varies with multiline switch */    OP_EXTUNI,         /* 21 \X (extended Unicode sequence */
651    OP_DOLL,           /* 21 End of line - varies with multiline switch */    OP_EODN,           /* 22 End of data or \n at end of data: \Z. */
652    OP_CHAR,           /* 22 Match one character, casefully */    OP_EOD,            /* 23 End of data: \z */
653    OP_CHARNC,         /* 23 Match one character, caselessly */  
654    OP_NOT,            /* 24 Match one character, not the following one */    OP_OPT,            /* 24 Set runtime options */
655      OP_CIRC,           /* 25 Start of line - varies with multiline switch */
656    OP_STAR,           /* 25 The maximizing and minimizing versions of */    OP_DOLL,           /* 26 End of line - varies with multiline switch */
657    OP_MINSTAR,        /* 26 these six opcodes must come in pairs, with */    OP_CHAR,           /* 27 Match one character, casefully */
658    OP_PLUS,           /* 27 the minimizing one second. */    OP_CHARNC,         /* 28 Match one character, caselessly */
659    OP_MINPLUS,        /* 28 This first set applies to single characters.*/    OP_NOT,            /* 29 Match one character, not the following one */
660    OP_QUERY,          /* 29 */  
661    OP_MINQUERY,       /* 30 */    OP_STAR,           /* 30 The maximizing and minimizing versions of */
662      OP_MINSTAR,        /* 31 these six opcodes must come in pairs, with */
663    OP_UPTO,           /* 31 From 0 to n matches */    OP_PLUS,           /* 32 the minimizing one second. */
664    OP_MINUPTO,        /* 32 */    OP_MINPLUS,        /* 33 This first set applies to single characters.*/
665    OP_EXACT,          /* 33 Exactly n matches */    OP_QUERY,          /* 34 */
666      OP_MINQUERY,       /* 35 */
667    OP_POSSTAR,        /* 34 Possessified star */  
668    OP_POSPLUS,        /* 35 Possessified plus */    OP_UPTO,           /* 36 From 0 to n matches */
669    OP_POSQUERY,       /* 36 Posesssified query */    OP_MINUPTO,        /* 37 */
670    OP_POSUPTO,        /* 37 Possessified upto */    OP_EXACT,          /* 38 Exactly n matches */
671    
672    OP_NOTSTAR,        /* 38 The maximizing and minimizing versions of */    OP_POSSTAR,        /* 39 Possessified star */
673    OP_NOTMINSTAR,     /* 39 these six opcodes must come in pairs, with */    OP_POSPLUS,        /* 40 Possessified plus */
674    OP_NOTPLUS,        /* 40 the minimizing one second. They must be in */    OP_POSQUERY,       /* 41 Posesssified query */
675    OP_NOTMINPLUS,     /* 41 exactly the same order as those above. */    OP_POSUPTO,        /* 42 Possessified upto */
676    OP_NOTQUERY,       /* 42 This set applies to "not" single characters. */  
677    OP_NOTMINQUERY,    /* 43 */    OP_NOTSTAR,        /* 43 The maximizing and minimizing versions of */
678      OP_NOTMINSTAR,     /* 44 these six opcodes must come in pairs, with */
679    OP_NOTUPTO,        /* 44 From 0 to n matches */    OP_NOTPLUS,        /* 45 the minimizing one second. They must be in */
680    OP_NOTMINUPTO,     /* 45 */    OP_NOTMINPLUS,     /* 46 exactly the same order as those above. */
681    OP_NOTEXACT,       /* 46 Exactly n matches */    OP_NOTQUERY,       /* 47 This set applies to "not" single characters. */
682      OP_NOTMINQUERY,    /* 48 */
683    OP_NOTPOSSTAR,     /* 47 Possessified versions */  
684    OP_NOTPOSPLUS,     /* 48 */    OP_NOTUPTO,        /* 49 From 0 to n matches */
685    OP_NOTPOSQUERY,    /* 49 */    OP_NOTMINUPTO,     /* 50 */
686    OP_NOTPOSUPTO,     /* 50 */    OP_NOTEXACT,       /* 51 Exactly n matches */
687    
688    OP_TYPESTAR,       /* 51 The maximizing and minimizing versions of */    OP_NOTPOSSTAR,     /* 52 Possessified versions */
689    OP_TYPEMINSTAR,    /* 52 these six opcodes must come in pairs, with */    OP_NOTPOSPLUS,     /* 53 */
690    OP_TYPEPLUS,       /* 53 the minimizing one second. These codes must */    OP_NOTPOSQUERY,    /* 54 */
691    OP_TYPEMINPLUS,    /* 54 be in exactly the same order as those above. */    OP_NOTPOSUPTO,     /* 55 */
692    OP_TYPEQUERY,      /* 55 This set applies to character types such as \d */  
693    OP_TYPEMINQUERY,   /* 56 */    OP_TYPESTAR,       /* 56 The maximizing and minimizing versions of */
694      OP_TYPEMINSTAR,    /* 57 these six opcodes must come in pairs, with */
695    OP_TYPEUPTO,       /* 57 From 0 to n matches */    OP_TYPEPLUS,       /* 58 the minimizing one second. These codes must */
696    OP_TYPEMINUPTO,    /* 58 */    OP_TYPEMINPLUS,    /* 59 be in exactly the same order as those above. */
697    OP_TYPEEXACT,      /* 59 Exactly n matches */    OP_TYPEQUERY,      /* 60 This set applies to character types such as \d */
698      OP_TYPEMINQUERY,   /* 61 */
699    OP_TYPEPOSSTAR,    /* 60 Possessified versions */  
700    OP_TYPEPOSPLUS,    /* 61 */    OP_TYPEUPTO,       /* 62 From 0 to n matches */
701    OP_TYPEPOSQUERY,   /* 62 */    OP_TYPEMINUPTO,    /* 63 */
702    OP_TYPEPOSUPTO,    /* 63 */    OP_TYPEEXACT,      /* 64 Exactly n matches */
703    
704    OP_CRSTAR,         /* 64 The maximizing and minimizing versions of */    OP_TYPEPOSSTAR,    /* 65 Possessified versions */
705    OP_CRMINSTAR,      /* 65 all these opcodes must come in pairs, with */    OP_TYPEPOSPLUS,    /* 66 */
706    OP_CRPLUS,         /* 66 the minimizing one second. These codes must */    OP_TYPEPOSQUERY,   /* 67 */
707    OP_CRMINPLUS,      /* 67 be in exactly the same order as those above. */    OP_TYPEPOSUPTO,    /* 68 */
708    OP_CRQUERY,        /* 68 These are for character classes and back refs */  
709    OP_CRMINQUERY,     /* 69 */    OP_CRSTAR,         /* 69 The maximizing and minimizing versions of */
710    OP_CRRANGE,        /* 70 These are different to the three sets above. */    OP_CRMINSTAR,      /* 70 all these opcodes must come in pairs, with */
711    OP_CRMINRANGE,     /* 71 */    OP_CRPLUS,         /* 71 the minimizing one second. These codes must */
712      OP_CRMINPLUS,      /* 72 be in exactly the same order as those above. */
713      OP_CRQUERY,        /* 73 These are for character classes and back refs */
714      OP_CRMINQUERY,     /* 74 */
715      OP_CRRANGE,        /* 75 These are different to the three sets above. */
716      OP_CRMINRANGE,     /* 76 */
717    
718    OP_CLASS,          /* 72 Match a character class, chars < 256 only */    OP_CLASS,          /* 77 Match a character class, chars < 256 only */
719    OP_NCLASS,         /* 73 Same, but the bitmap was created from a negative    OP_NCLASS,         /* 78 Same, but the bitmap was created from a negative
720                             class - the difference is relevant only when a UTF-8                             class - the difference is relevant only when a UTF-8
721                             character > 255 is encountered. */                             character > 255 is encountered. */
722    
723    OP_XCLASS,         /* 74 Extended class for handling UTF-8 chars within the    OP_XCLASS,         /* 79 Extended class for handling UTF-8 chars within the
724                             class. This does both positive and negative. */                             class. This does both positive and negative. */
725    
726    OP_REF,            /* 75 Match a back reference */    OP_REF,            /* 80 Match a back reference */
727    OP_RECURSE,        /* 76 Match a numbered subpattern (possibly recursive) */    OP_RECURSE,        /* 81 Match a numbered subpattern (possibly recursive) */
728    OP_CALLOUT,        /* 77 Call out to external function if provided */    OP_CALLOUT,        /* 82 Call out to external function if provided */
729    
730    OP_ALT,            /* 78 Start of alternation */    OP_ALT,            /* 83 Start of alternation */
731    OP_KET,            /* 79 End of group that doesn't have an unbounded repeat */    OP_KET,            /* 84 End of group that doesn't have an unbounded repeat */
732    OP_KETRMAX,        /* 80 These two must remain together and in this */    OP_KETRMAX,        /* 85 These two must remain together and in this */
733    OP_KETRMIN,        /* 81 order. They are for groups the repeat for ever. */    OP_KETRMIN,        /* 86 order. They are for groups the repeat for ever. */
734    
735    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/
736    
737    OP_ASSERT,         /* 82 Positive lookahead */    OP_ASSERT,         /* 87 Positive lookahead */
738    OP_ASSERT_NOT,     /* 83 Negative lookahead */    OP_ASSERT_NOT,     /* 88 Negative lookahead */
739    OP_ASSERTBACK,     /* 84 Positive lookbehind */    OP_ASSERTBACK,     /* 89 Positive lookbehind */
740    OP_ASSERTBACK_NOT, /* 85 Negative lookbehind */    OP_ASSERTBACK_NOT, /* 90 Negative lookbehind */
741    OP_REVERSE,        /* 86 Move pointer back - used in lookbehind assertions */    OP_REVERSE,        /* 91 Move pointer back - used in lookbehind assertions */
742    
743    /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,    /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,
744    as there's a test for >= ONCE for a subpattern that isn't an assertion. */    as there's a test for >= ONCE for a subpattern that isn't an assertion. */
745    
746    OP_ONCE,           /* 87 Atomic group */    OP_ONCE,           /* 92 Atomic group */
747    OP_BRA,            /* 88 Start of non-capturing bracket */    OP_BRA,            /* 83 Start of non-capturing bracket */
748    OP_CBRA,           /* 89 Start of capturing bracket */    OP_CBRA,           /* 94 Start of capturing bracket */
749    OP_COND,           /* 90 Conditional group */    OP_COND,           /* 95 Conditional group */
750    
751    /* These three must follow the previous three, in the same order. There's a    /* These three must follow the previous three, in the same order. There's a
752    check for >= SBRA to distinguish the two sets. */    check for >= SBRA to distinguish the two sets. */
753    
754    OP_SBRA,           /* 91 Start of non-capturing bracket, check empty  */    OP_SBRA,           /* 96 Start of non-capturing bracket, check empty  */
755    OP_SCBRA,          /* 92 Start of capturing bracket, check empty */    OP_SCBRA,          /* 97 Start of capturing bracket, check empty */
756    OP_SCOND,          /* 93 Conditional group, check empty */    OP_SCOND,          /* 98 Conditional group, check empty */
757    
758    OP_CREF,           /* 94 Used to hold a capture number as condition */    OP_CREF,           /* 99 Used to hold a capture number as condition */
759    OP_RREF,           /* 95 Used to hold a recursion number as condition */    OP_RREF,           /* 100 Used to hold a recursion number as condition */
760    OP_DEF,            /* 96 The DEFINE condition */    OP_DEF,            /* 101 The DEFINE condition */
761    
762    OP_BRAZERO,        /* 97 These two must remain together and in this */    OP_BRAZERO,        /* 102 These two must remain together and in this */
763    OP_BRAMINZERO      /* 98 order. */    OP_BRAMINZERO      /* 103 order. */
764  };  };
765    
766    
# Line 763  enum { Line 768  enum {
768  for debugging. The macro is referenced only in pcre_printint.c. */  for debugging. The macro is referenced only in pcre_printint.c. */
769    
770  #define OP_NAME_LIST \  #define OP_NAME_LIST \
771    "End", "\\A", "\\G", "\\B", "\\b", "\\D", "\\d",                \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
772    "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte",                   \    "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte",                   \
773    "notprop", "prop", "anynl", "extuni",                           \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
774    "\\Z", "\\z",                                                   \    "extuni",  "\\Z", "\\z",                                        \
775    "Opt", "^", "$", "char", "charnc", "not",                       \    "Opt", "^", "$", "char", "charnc", "not",                       \
776    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
777    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
# Line 793  in UTF-8 mode. The code that uses this t Line 798  in UTF-8 mode. The code that uses this t
798    
799  #define OP_LENGTHS \  #define OP_LENGTHS \
800    1,                             /* End                                    */ \    1,                             /* End                                    */ \
801    1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* \A, \G, \B, \B, \D, \d, \S, \s, \W, \w */ \    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \
802      1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
803    1, 1,                          /* Any, Anybyte                           */ \    1, 1,                          /* Any, Anybyte                           */ \
804    3, 3, 1, 1,                    /* NOTPROP, PROP, EXTUNI, ANYNL           */ \    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \
805      1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
806    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \
807    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
808    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Charnc  - the minimum length           */ \
# Line 855  enum { ERR0, ERR1, ERR2, ERR3, ERR4, Line 862  enum { ERR0, ERR1, ERR2, ERR3, ERR4,
862         ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,         ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
863         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
864         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
865         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57 };         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58 };
866    
867  /* 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
868  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 944  typedef struct recursion_info { Line 951  typedef struct recursion_info {
951    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
952    int group_num;                /* Number of group that was called */    int group_num;                /* Number of group that was called */
953    const uschar *after_call;     /* "Return value": points after the call in the expr */    const uschar *after_call;     /* "Return value": points after the call in the expr */
954    USPTR save_start;             /* Old value of md->start_match */    USPTR save_start;             /* Old value of mstart */
955    int *offset_save;             /* Pointer to start of saved offsets */    int *offset_save;             /* Pointer to start of saved offsets */
956    int saved_max;                /* Number of saved offsets */    int saved_max;                /* Number of saved offsets */
957  } recursion_info;  } recursion_info;
# Line 985  typedef struct match_data { Line 992  typedef struct match_data {
992    const uschar *start_code;     /* For use when recursing */    const uschar *start_code;     /* For use when recursing */
993    USPTR  start_subject;         /* Start of the subject string */    USPTR  start_subject;         /* Start of the subject string */
994    USPTR  end_subject;           /* End of the subject string */    USPTR  end_subject;           /* End of the subject string */
995    USPTR  start_match;           /* Start of this match attempt */    USPTR  start_match_ptr;       /* Start of matched string */
996    USPTR  end_match_ptr;         /* Subject position at end match */    USPTR  end_match_ptr;         /* Subject position at end match */
997    int    end_offset_top;        /* Highwater mark at end of match */    int    end_offset_top;        /* Highwater mark at end of match */
998    int    capture_last;          /* Most recent capture number */    int    capture_last;          /* Most recent capture number */
Legend:
Removed from v.164  
changed lines
  Added in v.189
webmaster@exim.org
 ViewVC Help
Powered by ViewVC 1.1.12