/[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 336 by ph10, Sat Apr 12 15:59:03 2008 UTC revision 341 by ph10, Sat Apr 19 16:41:04 2008 UTC
# Line 605  contain UTF-8 characters with values gre Line 605  contain UTF-8 characters with values gre
605  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
606  their negation. Also, they must appear in the same order as in the opcode  their negation. Also, they must appear in the same order as in the opcode
607  definitions below, up to ESC_z. There's a dummy for OP_ANY because it  definitions below, up to ESC_z. There's a dummy for OP_ANY because it
608  corresponds to "." rather than an escape sequence. The final one must be  corresponds to "." rather than an escape sequence, and another for OP_ALLANY
609  ESC_REF as subsequent values are used for backreferences (\1, \2, \3, etc).  (which is used for [^] in JavaScript compatibility mode).
610  There are two tests in the code for an escape greater than ESC_b and less than  
611  ESC_Z to detect the types that may be repeated. These are the types that  The final escape must be ESC_REF as subsequent values are used for
612  consume characters. If any new escapes are put in between that don't consume a  backreferences (\1, \2, \3, etc). There are two tests in the code for an escape
613  character, that code will have to change. */  greater than ESC_b and less than ESC_Z to detect the types that may be
614    repeated. These are the types that consume characters. If any new escapes are
615    put in between that don't consume a character, that code will have to change.
616    */
617    
618  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
619         ESC_W, ESC_w, ESC_dum1, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H, ESC_h,         ESC_W, ESC_w, ESC_dum1, ESC_dum2, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
620         ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_g, ESC_k, ESC_REF };         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_g, ESC_k,
621           ESC_REF };
622    
623    
624  /* 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
# Line 640  enum { Line 644  enum {
644    OP_WHITESPACE,         /*  9 \s */    OP_WHITESPACE,         /*  9 \s */
645    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
646    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
647    OP_ANY,            /* 12 Match any character */    OP_ANY,            /* 12 Match any character (subject to DOTALL) */
648    OP_ANYBYTE,        /* 13 Match any byte (\C); different to OP_ANY for UTF-8 */    OP_ALLANY,         /* 13 Match any character (not subject to DOTALL) */
649    OP_NOTPROP,        /* 14 \P (not Unicode property) */    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
650    OP_PROP,           /* 15 \p (Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
651    OP_ANYNL,          /* 16 \R (any newline sequence) */    OP_PROP,           /* 16 \p (Unicode property) */
652    OP_NOT_HSPACE,     /* 17 \H (not horizontal whitespace) */    OP_ANYNL,          /* 17 \R (any newline sequence) */
653    OP_HSPACE,         /* 18 \h (horizontal whitespace) */    OP_NOT_HSPACE,     /* 18 \H (not horizontal whitespace) */
654    OP_NOT_VSPACE,     /* 19 \V (not vertical whitespace) */    OP_HSPACE,         /* 19 \h (horizontal whitespace) */
655    OP_VSPACE,         /* 20 \v (vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
656    OP_EXTUNI,         /* 21 \X (extended Unicode sequence */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
657    OP_EODN,           /* 22 End of data or \n at end of data: \Z. */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
658    OP_EOD,            /* 23 End of data: \z */    OP_EODN,           /* 23 End of data or \n at end of data: \Z. */
659      OP_EOD,            /* 24 End of data: \z */
660    OP_OPT,            /* 24 Set runtime options */  
661    OP_CIRC,           /* 25 Start of line - varies with multiline switch */    OP_OPT,            /* 25 Set runtime options */
662    OP_DOLL,           /* 26 End of line - varies with multiline switch */    OP_CIRC,           /* 26 Start of line - varies with multiline switch */
663    OP_CHAR,           /* 27 Match one character, casefully */    OP_DOLL,           /* 27 End of line - varies with multiline switch */
664    OP_CHARNC,         /* 28 Match one character, caselessly */    OP_CHAR,           /* 28 Match one character, casefully */
665    OP_NOT,            /* 29 Match one character, not the following one */    OP_CHARNC,         /* 29 Match one character, caselessly */
666      OP_NOT,            /* 30 Match one character, not the following one */
667    OP_STAR,           /* 30 The maximizing and minimizing versions of */  
668    OP_MINSTAR,        /* 31 these six opcodes must come in pairs, with */    OP_STAR,           /* 31 The maximizing and minimizing versions of */
669    OP_PLUS,           /* 32 the minimizing one second. */    OP_MINSTAR,        /* 32 these six opcodes must come in pairs, with */
670    OP_MINPLUS,        /* 33 This first set applies to single characters.*/    OP_PLUS,           /* 33 the minimizing one second. */
671    OP_QUERY,          /* 34 */    OP_MINPLUS,        /* 34 This first set applies to single characters.*/
672    OP_MINQUERY,       /* 35 */    OP_QUERY,          /* 35 */
673      OP_MINQUERY,       /* 36 */
674    OP_UPTO,           /* 36 From 0 to n matches */  
675    OP_MINUPTO,        /* 37 */    OP_UPTO,           /* 37 From 0 to n matches */
676    OP_EXACT,          /* 38 Exactly n matches */    OP_MINUPTO,        /* 38 */
677      OP_EXACT,          /* 39 Exactly n matches */
678    OP_POSSTAR,        /* 39 Possessified star */  
679    OP_POSPLUS,        /* 40 Possessified plus */    OP_POSSTAR,        /* 40 Possessified star */
680    OP_POSQUERY,       /* 41 Posesssified query */    OP_POSPLUS,        /* 41 Possessified plus */
681    OP_POSUPTO,        /* 42 Possessified upto */    OP_POSQUERY,       /* 42 Posesssified query */
682      OP_POSUPTO,        /* 43 Possessified upto */
683    OP_NOTSTAR,        /* 43 The maximizing and minimizing versions of */  
684    OP_NOTMINSTAR,     /* 44 these six opcodes must come in pairs, with */    OP_NOTSTAR,        /* 44 The maximizing and minimizing versions of */
685    OP_NOTPLUS,        /* 45 the minimizing one second. They must be in */    OP_NOTMINSTAR,     /* 45 these six opcodes must come in pairs, with */
686    OP_NOTMINPLUS,     /* 46 exactly the same order as those above. */    OP_NOTPLUS,        /* 46 the minimizing one second. They must be in */
687    OP_NOTQUERY,       /* 47 This set applies to "not" single characters. */    OP_NOTMINPLUS,     /* 47 exactly the same order as those above. */
688    OP_NOTMINQUERY,    /* 48 */    OP_NOTQUERY,       /* 48 This set applies to "not" single characters. */
689      OP_NOTMINQUERY,    /* 49 */
690    OP_NOTUPTO,        /* 49 From 0 to n matches */  
691    OP_NOTMINUPTO,     /* 50 */    OP_NOTUPTO,        /* 50 From 0 to n matches */
692    OP_NOTEXACT,       /* 51 Exactly n matches */    OP_NOTMINUPTO,     /* 51 */
693      OP_NOTEXACT,       /* 52 Exactly n matches */
694    OP_NOTPOSSTAR,     /* 52 Possessified versions */  
695    OP_NOTPOSPLUS,     /* 53 */    OP_NOTPOSSTAR,     /* 53 Possessified versions */
696    OP_NOTPOSQUERY,    /* 54 */    OP_NOTPOSPLUS,     /* 54 */
697    OP_NOTPOSUPTO,     /* 55 */    OP_NOTPOSQUERY,    /* 55 */
698      OP_NOTPOSUPTO,     /* 56 */
699    OP_TYPESTAR,       /* 56 The maximizing and minimizing versions of */  
700    OP_TYPEMINSTAR,    /* 57 these six opcodes must come in pairs, with */    OP_TYPESTAR,       /* 57 The maximizing and minimizing versions of */
701    OP_TYPEPLUS,       /* 58 the minimizing one second. These codes must */    OP_TYPEMINSTAR,    /* 58 these six opcodes must come in pairs, with */
702    OP_TYPEMINPLUS,    /* 59 be in exactly the same order as those above. */    OP_TYPEPLUS,       /* 59 the minimizing one second. These codes must */
703    OP_TYPEQUERY,      /* 60 This set applies to character types such as \d */    OP_TYPEMINPLUS,    /* 60 be in exactly the same order as those above. */
704    OP_TYPEMINQUERY,   /* 61 */    OP_TYPEQUERY,      /* 61 This set applies to character types such as \d */
705      OP_TYPEMINQUERY,   /* 62 */
706    OP_TYPEUPTO,       /* 62 From 0 to n matches */  
707    OP_TYPEMINUPTO,    /* 63 */    OP_TYPEUPTO,       /* 63 From 0 to n matches */
708    OP_TYPEEXACT,      /* 64 Exactly n matches */    OP_TYPEMINUPTO,    /* 64 */
709      OP_TYPEEXACT,      /* 65 Exactly n matches */
710    OP_TYPEPOSSTAR,    /* 65 Possessified versions */  
711    OP_TYPEPOSPLUS,    /* 66 */    OP_TYPEPOSSTAR,    /* 66 Possessified versions */
712    OP_TYPEPOSQUERY,   /* 67 */    OP_TYPEPOSPLUS,    /* 67 */
713    OP_TYPEPOSUPTO,    /* 68 */    OP_TYPEPOSQUERY,   /* 68 */
714      OP_TYPEPOSUPTO,    /* 69 */
715    OP_CRSTAR,         /* 69 The maximizing and minimizing versions of */  
716    OP_CRMINSTAR,      /* 70 all these opcodes must come in pairs, with */    OP_CRSTAR,         /* 70 The maximizing and minimizing versions of */
717    OP_CRPLUS,         /* 71 the minimizing one second. These codes must */    OP_CRMINSTAR,      /* 71 all these opcodes must come in pairs, with */
718    OP_CRMINPLUS,      /* 72 be in exactly the same order as those above. */    OP_CRPLUS,         /* 72 the minimizing one second. These codes must */
719    OP_CRQUERY,        /* 73 These are for character classes and back refs */    OP_CRMINPLUS,      /* 73 be in exactly the same order as those above. */
720    OP_CRMINQUERY,     /* 74 */    OP_CRQUERY,        /* 74 These are for character classes and back refs */
721    OP_CRRANGE,        /* 75 These are different to the three sets above. */    OP_CRMINQUERY,     /* 75 */
722    OP_CRMINRANGE,     /* 76 */    OP_CRRANGE,        /* 76 These are different to the three sets above. */
723      OP_CRMINRANGE,     /* 77 */
724    
725    OP_CLASS,          /* 77 Match a character class, chars < 256 only */    OP_CLASS,          /* 78 Match a character class, chars < 256 only */
726    OP_NCLASS,         /* 78 Same, but the bitmap was created from a negative    OP_NCLASS,         /* 79 Same, but the bitmap was created from a negative
727                             class - the difference is relevant only when a UTF-8                             class - the difference is relevant only when a UTF-8
728                             character > 255 is encountered. */                             character > 255 is encountered. */
729    
730    OP_XCLASS,         /* 79 Extended class for handling UTF-8 chars within the    OP_XCLASS,         /* 80 Extended class for handling UTF-8 chars within the
731                             class. This does both positive and negative. */                             class. This does both positive and negative. */
732    
733    OP_REF,            /* 80 Match a back reference */    OP_REF,            /* 81 Match a back reference */
734    OP_RECURSE,        /* 81 Match a numbered subpattern (possibly recursive) */    OP_RECURSE,        /* 82 Match a numbered subpattern (possibly recursive) */
735    OP_CALLOUT,        /* 82 Call out to external function if provided */    OP_CALLOUT,        /* 83 Call out to external function if provided */
736    
737    OP_ALT,            /* 83 Start of alternation */    OP_ALT,            /* 84 Start of alternation */
738    OP_KET,            /* 84 End of group that doesn't have an unbounded repeat */    OP_KET,            /* 85 End of group that doesn't have an unbounded repeat */
739    OP_KETRMAX,        /* 85 These two must remain together and in this */    OP_KETRMAX,        /* 86 These two must remain together and in this */
740    OP_KETRMIN,        /* 86 order. They are for groups the repeat for ever. */    OP_KETRMIN,        /* 87 order. They are for groups the repeat for ever. */
741    
742    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/
743    
744    OP_ASSERT,         /* 87 Positive lookahead */    OP_ASSERT,         /* 88 Positive lookahead */
745    OP_ASSERT_NOT,     /* 88 Negative lookahead */    OP_ASSERT_NOT,     /* 89 Negative lookahead */
746    OP_ASSERTBACK,     /* 89 Positive lookbehind */    OP_ASSERTBACK,     /* 90 Positive lookbehind */
747    OP_ASSERTBACK_NOT, /* 90 Negative lookbehind */    OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */
748    OP_REVERSE,        /* 91 Move pointer back - used in lookbehind assertions */    OP_REVERSE,        /* 92 Move pointer back - used in lookbehind assertions */
749    
750    /* 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,
751    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. */
752    
753    OP_ONCE,           /* 92 Atomic group */    OP_ONCE,           /* 93 Atomic group */
754    OP_BRA,            /* 93 Start of non-capturing bracket */    OP_BRA,            /* 94 Start of non-capturing bracket */
755    OP_CBRA,           /* 94 Start of capturing bracket */    OP_CBRA,           /* 95 Start of capturing bracket */
756    OP_COND,           /* 95 Conditional group */    OP_COND,           /* 96 Conditional group */
757    
758    /* 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
759    check for >= SBRA to distinguish the two sets. */    check for >= SBRA to distinguish the two sets. */
760    
761    OP_SBRA,           /* 96 Start of non-capturing bracket, check empty  */    OP_SBRA,           /* 97 Start of non-capturing bracket, check empty  */
762    OP_SCBRA,          /* 97 Start of capturing bracket, check empty */    OP_SCBRA,          /* 98 Start of capturing bracket, check empty */
763    OP_SCOND,          /* 98 Conditional group, check empty */    OP_SCOND,          /* 99 Conditional group, check empty */
764    
765    OP_CREF,           /* 99 Used to hold a capture number as condition */    OP_CREF,           /* 100 Used to hold a capture number as condition */
766    OP_RREF,           /* 100 Used to hold a recursion number as condition */    OP_RREF,           /* 101 Used to hold a recursion number as condition */
767    OP_DEF,            /* 101 The DEFINE condition */    OP_DEF,            /* 102 The DEFINE condition */
768    
769    OP_BRAZERO,        /* 102 These two must remain together and in this */    OP_BRAZERO,        /* 103 These two must remain together and in this */
770    OP_BRAMINZERO,     /* 103 order. */    OP_BRAMINZERO,     /* 104 order. */
771    
772    /* These are backtracking control verbs */    /* These are backtracking control verbs */
773    
774    OP_PRUNE,          /* 104 */    OP_PRUNE,          /* 105 */
775    OP_SKIP,           /* 105 */    OP_SKIP,           /* 106 */
776    OP_THEN,           /* 106 */    OP_THEN,           /* 107 */
777    OP_COMMIT,         /* 107 */    OP_COMMIT,         /* 108 */
778    
779    /* These are forced failure and success verbs */    /* These are forced failure and success verbs */
780    
781    OP_FAIL,           /* 108 */    OP_FAIL,           /* 109 */
782    OP_ACCEPT,         /* 109 */    OP_ACCEPT,         /* 110 */
783    
784    /* This is used to skip a subpattern with a {0} quantifier */    /* This is used to skip a subpattern with a {0} quantifier */
785    
786    OP_SKIPZERO        /* 110 */    OP_SKIPZERO        /* 111 */
787  };  };
788    
789    
# Line 787  for debugging. The macro is referenced o Line 792  for debugging. The macro is referenced o
792    
793  #define OP_NAME_LIST \  #define OP_NAME_LIST \
794    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
795    "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte",                   \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
796    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
797    "extuni",  "\\Z", "\\z",                                        \    "extuni",  "\\Z", "\\z",                                        \
798    "Opt", "^", "$", "char", "charnc", "not",                       \    "Opt", "^", "$", "char", "charnc", "not",                       \
# Line 820  in UTF-8 mode. The code that uses this t Line 825  in UTF-8 mode. The code that uses this t
825    1,                             /* End                                    */ \    1,                             /* End                                    */ \
826    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \
827    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
828    1, 1,                          /* Any, Anybyte                           */ \    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \
829    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \
830    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
831    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \
Legend:
Removed from v.336  
changed lines
  Added in v.341
  ViewVC Help
Powered by ViewVC 1.1.5