/[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 231 by ph10, Tue Sep 11 11:15:33 2007 UTC revision 364 by ph10, Fri Jul 11 14:53:41 2008 UTC
# Line 7  Line 7 
7  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2007 University of Cambridge             Copyright (c) 1997-2008 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 132  PCRE_EXP_DATA_DEFN only if they are not Line 132  PCRE_EXP_DATA_DEFN only if they are not
132  #  endif  #  endif
133  #endif  #endif
134    
135    /* When compiling with the MSVC compiler, it is sometimes necessary to include
136    a "calling convention" before exported function names. (This is secondhand
137    information; I know nothing about MSVC myself). For example, something like
138    
139      void __cdecl function(....)
140    
141    might be needed. In order so make this easy, all the exported functions have
142    PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
143    set, we ensure here that it has no effect. */
144    
145    #ifndef PCRE_CALL_CONVENTION
146    #define PCRE_CALL_CONVENTION
147    #endif
148    
149  /* We need to have types that specify unsigned 16-bit and 32-bit integers. We  /* We need to have types that specify unsigned 16-bit and 32-bit integers. We
150  cannot determine these outside the compilation (e.g. by running a program as  cannot determine these outside the compilation (e.g. by running a program as
151  part of "configure") because PCRE is often cross-compiled for use on other  part of "configure") because PCRE is often cross-compiled for use on other
# Line 140  preprocessor time in standard C environm Line 154  preprocessor time in standard C environm
154    
155  #if USHRT_MAX == 65535  #if USHRT_MAX == 65535
156    typedef unsigned short pcre_uint16;    typedef unsigned short pcre_uint16;
157      typedef short pcre_int16;
158  #elif UINT_MAX == 65535  #elif UINT_MAX == 65535
159    typedef unsigned int pcre_uint16;    typedef unsigned int pcre_uint16;
160      typedef int pcre_int16;
161  #else  #else
162    #error Cannot determine a type for 16-bit unsigned integers    #error Cannot determine a type for 16-bit unsigned integers
163  #endif  #endif
164    
165  #if UINT_MAX == 4294967295  #if UINT_MAX == 4294967295
166    typedef unsigned int pcre_uint32;    typedef unsigned int pcre_uint32;
167      typedef int pcre_int32;
168  #elif ULONG_MAX == 4294967295  #elif ULONG_MAX == 4294967295
169    typedef unsigned long int pcre_uint32;    typedef unsigned long int pcre_uint32;
170      typedef long int pcre_int32;
171  #else  #else
172    #error Cannot determine a type for 32-bit unsigned integers    #error Cannot determine a type for 32-bit unsigned integers
173  #endif  #endif
# Line 506  time, run time, or study time, respectiv Line 524  time, run time, or study time, respectiv
524    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
525     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
526     PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \     PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
527     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
528       PCRE_JAVASCRIPT_COMPAT)
529    
530  #define PUBLIC_EXEC_OPTIONS \  #define PUBLIC_EXEC_OPTIONS \
531    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \
# Line 596  contain UTF-8 characters with values gre Line 615  contain UTF-8 characters with values gre
615  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
616  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
617  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
618  corresponds to "." rather than an escape sequence. The final one must be  corresponds to "." rather than an escape sequence, and another for OP_ALLANY
619  ESC_REF as subsequent values are used for backreferences (\1, \2, \3, etc).  (which is used for [^] in JavaScript compatibility mode).
620  There are two tests in the code for an escape greater than ESC_b and less than  
621  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
622  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
623  character, that code will have to change. */  greater than ESC_b and less than ESC_Z to detect the types that may be
624    repeated. These are the types that consume characters. If any new escapes are
625    put in between that don't consume a character, that code will have to change.
626    */
627    
628  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,
629         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,
630         ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_k, ESC_REF };         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_g, ESC_k,
631           ESC_REF };
632    
633    
634  /* 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 631  enum { Line 654  enum {
654    OP_WHITESPACE,         /*  9 \s */    OP_WHITESPACE,         /*  9 \s */
655    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
656    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
657    OP_ANY,            /* 12 Match any character */    OP_ANY,            /* 12 Match any character (subject to DOTALL) */
658    OP_ANYBYTE,        /* 13 Match any byte (\C); different to OP_ANY for UTF-8 */    OP_ALLANY,         /* 13 Match any character (not subject to DOTALL) */
659    OP_NOTPROP,        /* 14 \P (not Unicode property) */    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
660    OP_PROP,           /* 15 \p (Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
661    OP_ANYNL,          /* 16 \R (any newline sequence) */    OP_PROP,           /* 16 \p (Unicode property) */
662    OP_NOT_HSPACE,     /* 17 \H (not horizontal whitespace) */    OP_ANYNL,          /* 17 \R (any newline sequence) */
663    OP_HSPACE,         /* 18 \h (horizontal whitespace) */    OP_NOT_HSPACE,     /* 18 \H (not horizontal whitespace) */
664    OP_NOT_VSPACE,     /* 19 \V (not vertical whitespace) */    OP_HSPACE,         /* 19 \h (horizontal whitespace) */
665    OP_VSPACE,         /* 20 \v (vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
666    OP_EXTUNI,         /* 21 \X (extended Unicode sequence */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
667    OP_EODN,           /* 22 End of data or \n at end of data: \Z. */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
668    OP_EOD,            /* 23 End of data: \z */    OP_EODN,           /* 23 End of data or \n at end of data: \Z. */
669      OP_EOD,            /* 24 End of data: \z */
670    OP_OPT,            /* 24 Set runtime options */  
671    OP_CIRC,           /* 25 Start of line - varies with multiline switch */    OP_OPT,            /* 25 Set runtime options */
672    OP_DOLL,           /* 26 End of line - varies with multiline switch */    OP_CIRC,           /* 26 Start of line - varies with multiline switch */
673    OP_CHAR,           /* 27 Match one character, casefully */    OP_DOLL,           /* 27 End of line - varies with multiline switch */
674    OP_CHARNC,         /* 28 Match one character, caselessly */    OP_CHAR,           /* 28 Match one character, casefully */
675    OP_NOT,            /* 29 Match one character, not the following one */    OP_CHARNC,         /* 29 Match one character, caselessly */
676      OP_NOT,            /* 30 Match one character, not the following one */
677    OP_STAR,           /* 30 The maximizing and minimizing versions of */  
678    OP_MINSTAR,        /* 31 these six opcodes must come in pairs, with */    OP_STAR,           /* 31 The maximizing and minimizing versions of */
679    OP_PLUS,           /* 32 the minimizing one second. */    OP_MINSTAR,        /* 32 these six opcodes must come in pairs, with */
680    OP_MINPLUS,        /* 33 This first set applies to single characters.*/    OP_PLUS,           /* 33 the minimizing one second. */
681    OP_QUERY,          /* 34 */    OP_MINPLUS,        /* 34 This first set applies to single characters.*/
682    OP_MINQUERY,       /* 35 */    OP_QUERY,          /* 35 */
683      OP_MINQUERY,       /* 36 */
684    OP_UPTO,           /* 36 From 0 to n matches */  
685    OP_MINUPTO,        /* 37 */    OP_UPTO,           /* 37 From 0 to n matches */
686    OP_EXACT,          /* 38 Exactly n matches */    OP_MINUPTO,        /* 38 */
687      OP_EXACT,          /* 39 Exactly n matches */
688    OP_POSSTAR,        /* 39 Possessified star */  
689    OP_POSPLUS,        /* 40 Possessified plus */    OP_POSSTAR,        /* 40 Possessified star */
690    OP_POSQUERY,       /* 41 Posesssified query */    OP_POSPLUS,        /* 41 Possessified plus */
691    OP_POSUPTO,        /* 42 Possessified upto */    OP_POSQUERY,       /* 42 Posesssified query */
692      OP_POSUPTO,        /* 43 Possessified upto */
693    OP_NOTSTAR,        /* 43 The maximizing and minimizing versions of */  
694    OP_NOTMINSTAR,     /* 44 these six opcodes must come in pairs, with */    OP_NOTSTAR,        /* 44 The maximizing and minimizing versions of */
695    OP_NOTPLUS,        /* 45 the minimizing one second. They must be in */    OP_NOTMINSTAR,     /* 45 these six opcodes must come in pairs, with */
696    OP_NOTMINPLUS,     /* 46 exactly the same order as those above. */    OP_NOTPLUS,        /* 46 the minimizing one second. They must be in */
697    OP_NOTQUERY,       /* 47 This set applies to "not" single characters. */    OP_NOTMINPLUS,     /* 47 exactly the same order as those above. */
698    OP_NOTMINQUERY,    /* 48 */    OP_NOTQUERY,       /* 48 This set applies to "not" single characters. */
699      OP_NOTMINQUERY,    /* 49 */
700    OP_NOTUPTO,        /* 49 From 0 to n matches */  
701    OP_NOTMINUPTO,     /* 50 */    OP_NOTUPTO,        /* 50 From 0 to n matches */
702    OP_NOTEXACT,       /* 51 Exactly n matches */    OP_NOTMINUPTO,     /* 51 */
703      OP_NOTEXACT,       /* 52 Exactly n matches */
704    OP_NOTPOSSTAR,     /* 52 Possessified versions */  
705    OP_NOTPOSPLUS,     /* 53 */    OP_NOTPOSSTAR,     /* 53 Possessified versions */
706    OP_NOTPOSQUERY,    /* 54 */    OP_NOTPOSPLUS,     /* 54 */
707    OP_NOTPOSUPTO,     /* 55 */    OP_NOTPOSQUERY,    /* 55 */
708      OP_NOTPOSUPTO,     /* 56 */
709    OP_TYPESTAR,       /* 56 The maximizing and minimizing versions of */  
710    OP_TYPEMINSTAR,    /* 57 these six opcodes must come in pairs, with */    OP_TYPESTAR,       /* 57 The maximizing and minimizing versions of */
711    OP_TYPEPLUS,       /* 58 the minimizing one second. These codes must */    OP_TYPEMINSTAR,    /* 58 these six opcodes must come in pairs, with */
712    OP_TYPEMINPLUS,    /* 59 be in exactly the same order as those above. */    OP_TYPEPLUS,       /* 59 the minimizing one second. These codes must */
713    OP_TYPEQUERY,      /* 60 This set applies to character types such as \d */    OP_TYPEMINPLUS,    /* 60 be in exactly the same order as those above. */
714    OP_TYPEMINQUERY,   /* 61 */    OP_TYPEQUERY,      /* 61 This set applies to character types such as \d */
715      OP_TYPEMINQUERY,   /* 62 */
716    OP_TYPEUPTO,       /* 62 From 0 to n matches */  
717    OP_TYPEMINUPTO,    /* 63 */    OP_TYPEUPTO,       /* 63 From 0 to n matches */
718    OP_TYPEEXACT,      /* 64 Exactly n matches */    OP_TYPEMINUPTO,    /* 64 */
719      OP_TYPEEXACT,      /* 65 Exactly n matches */
720    OP_TYPEPOSSTAR,    /* 65 Possessified versions */  
721    OP_TYPEPOSPLUS,    /* 66 */    OP_TYPEPOSSTAR,    /* 66 Possessified versions */
722    OP_TYPEPOSQUERY,   /* 67 */    OP_TYPEPOSPLUS,    /* 67 */
723    OP_TYPEPOSUPTO,    /* 68 */    OP_TYPEPOSQUERY,   /* 68 */
724      OP_TYPEPOSUPTO,    /* 69 */
725    OP_CRSTAR,         /* 69 The maximizing and minimizing versions of */  
726    OP_CRMINSTAR,      /* 70 all these opcodes must come in pairs, with */    OP_CRSTAR,         /* 70 The maximizing and minimizing versions of */
727    OP_CRPLUS,         /* 71 the minimizing one second. These codes must */    OP_CRMINSTAR,      /* 71 all these opcodes must come in pairs, with */
728    OP_CRMINPLUS,      /* 72 be in exactly the same order as those above. */    OP_CRPLUS,         /* 72 the minimizing one second. These codes must */
729    OP_CRQUERY,        /* 73 These are for character classes and back refs */    OP_CRMINPLUS,      /* 73 be in exactly the same order as those above. */
730    OP_CRMINQUERY,     /* 74 */    OP_CRQUERY,        /* 74 These are for character classes and back refs */
731    OP_CRRANGE,        /* 75 These are different to the three sets above. */    OP_CRMINQUERY,     /* 75 */
732    OP_CRMINRANGE,     /* 76 */    OP_CRRANGE,        /* 76 These are different to the three sets above. */
733      OP_CRMINRANGE,     /* 77 */
734    
735    OP_CLASS,          /* 77 Match a character class, chars < 256 only */    OP_CLASS,          /* 78 Match a character class, chars < 256 only */
736    OP_NCLASS,         /* 78 Same, but the bitmap was created from a negative    OP_NCLASS,         /* 79 Same, but the bitmap was created from a negative
737                             class - the difference is relevant only when a UTF-8                             class - the difference is relevant only when a UTF-8
738                             character > 255 is encountered. */                             character > 255 is encountered. */
739    
740    OP_XCLASS,         /* 79 Extended class for handling UTF-8 chars within the    OP_XCLASS,         /* 80 Extended class for handling UTF-8 chars within the
741                             class. This does both positive and negative. */                             class. This does both positive and negative. */
742    
743    OP_REF,            /* 80 Match a back reference */    OP_REF,            /* 81 Match a back reference */
744    OP_RECURSE,        /* 81 Match a numbered subpattern (possibly recursive) */    OP_RECURSE,        /* 82 Match a numbered subpattern (possibly recursive) */
745    OP_CALLOUT,        /* 82 Call out to external function if provided */    OP_CALLOUT,        /* 83 Call out to external function if provided */
746    
747    OP_ALT,            /* 83 Start of alternation */    OP_ALT,            /* 84 Start of alternation */
748    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 */
749    OP_KETRMAX,        /* 85 These two must remain together and in this */    OP_KETRMAX,        /* 86 These two must remain together and in this */
750    OP_KETRMIN,        /* 86 order. They are for groups the repeat for ever. */    OP_KETRMIN,        /* 87 order. They are for groups the repeat for ever. */
751    
752    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/
753    
754    OP_ASSERT,         /* 87 Positive lookahead */    OP_ASSERT,         /* 88 Positive lookahead */
755    OP_ASSERT_NOT,     /* 88 Negative lookahead */    OP_ASSERT_NOT,     /* 89 Negative lookahead */
756    OP_ASSERTBACK,     /* 89 Positive lookbehind */    OP_ASSERTBACK,     /* 90 Positive lookbehind */
757    OP_ASSERTBACK_NOT, /* 90 Negative lookbehind */    OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */
758    OP_REVERSE,        /* 91 Move pointer back - used in lookbehind assertions */    OP_REVERSE,        /* 92 Move pointer back - used in lookbehind assertions */
759    
760    /* 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,
761    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. */
762    
763    OP_ONCE,           /* 92 Atomic group */    OP_ONCE,           /* 93 Atomic group */
764    OP_BRA,            /* 93 Start of non-capturing bracket */    OP_BRA,            /* 94 Start of non-capturing bracket */
765    OP_CBRA,           /* 94 Start of capturing bracket */    OP_CBRA,           /* 95 Start of capturing bracket */
766    OP_COND,           /* 95 Conditional group */    OP_COND,           /* 96 Conditional group */
767    
768    /* 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
769    check for >= SBRA to distinguish the two sets. */    check for >= SBRA to distinguish the two sets. */
770    
771    OP_SBRA,           /* 96 Start of non-capturing bracket, check empty  */    OP_SBRA,           /* 97 Start of non-capturing bracket, check empty  */
772    OP_SCBRA,          /* 97 Start of capturing bracket, check empty */    OP_SCBRA,          /* 98 Start of capturing bracket, check empty */
773    OP_SCOND,          /* 98 Conditional group, check empty */    OP_SCOND,          /* 99 Conditional group, check empty */
774    
775    OP_CREF,           /* 99 Used to hold a capture number as condition */    OP_CREF,           /* 100 Used to hold a capture number as condition */
776    OP_RREF,           /* 100 Used to hold a recursion number as condition */    OP_RREF,           /* 101 Used to hold a recursion number as condition */
777    OP_DEF,            /* 101 The DEFINE condition */    OP_DEF,            /* 102 The DEFINE condition */
778    
779    OP_BRAZERO,        /* 102 These two must remain together and in this */    OP_BRAZERO,        /* 103 These two must remain together and in this */
780    OP_BRAMINZERO,     /* 103 order. */    OP_BRAMINZERO,     /* 104 order. */
781    
782    /* These are backtracking control verbs */    /* These are backtracking control verbs */
783    
784    OP_PRUNE,          /* 104 */    OP_PRUNE,          /* 105 */
785    OP_SKIP,           /* 105 */    OP_SKIP,           /* 106 */
786    OP_THEN,           /* 106 */    OP_THEN,           /* 107 */
787    OP_COMMIT,         /* 107 */    OP_COMMIT,         /* 108 */
788    
789    /* These are forced failure and success verbs */    /* These are forced failure and success verbs */
790    
791    OP_FAIL,           /* 108 */    OP_FAIL,           /* 109 */
792    OP_ACCEPT          /* 109 */    OP_ACCEPT,         /* 110 */
793    
794      /* This is used to skip a subpattern with a {0} quantifier */
795    
796      OP_SKIPZERO        /* 111 */
797  };  };
798    
799    
# Line 774  for debugging. The macro is referenced o Line 802  for debugging. The macro is referenced o
802    
803  #define OP_NAME_LIST \  #define OP_NAME_LIST \
804    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
805    "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte",                   \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
806    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
807    "extuni",  "\\Z", "\\z",                                        \    "extuni",  "\\Z", "\\z",                                        \
808    "Opt", "^", "$", "char", "charnc", "not",                       \    "Opt", "^", "$", "char", "charnc", "not",                       \
# Line 790  for debugging. The macro is referenced o Line 818  for debugging. The macro is referenced o
818    "AssertB", "AssertB not", "Reverse",                            \    "AssertB", "AssertB not", "Reverse",                            \
819    "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond",        \    "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond",        \
820    "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero",    \    "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero",    \
821    "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT"    "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT",      \
822      "Skip zero"
823    
824    
825  /* This macro defines the length of fixed length operations in the compiled  /* This macro defines the length of fixed length operations in the compiled
# Line 806  in UTF-8 mode. The code that uses this t Line 835  in UTF-8 mode. The code that uses this t
835    1,                             /* End                                    */ \    1,                             /* End                                    */ \
836    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \
837    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
838    1, 1,                          /* Any, Anybyte                           */ \    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \
839    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \
840    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
841    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \
# Line 855  in UTF-8 mode. The code that uses this t Line 884  in UTF-8 mode. The code that uses this t
884    1,                             /* DEF                                    */ \    1,                             /* DEF                                    */ \
885    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \
886    1, 1, 1, 1,                    /* PRUNE, SKIP, THEN, COMMIT,             */ \    1, 1, 1, 1,                    /* PRUNE, SKIP, THEN, COMMIT,             */ \
887    1, 1                           /* FAIL, ACCEPT                           */    1, 1, 1                        /* FAIL, ACCEPT, SKIPZERO                 */
888    
889    
890  /* A magic value for OP_RREF to indicate the "any recursion" condition. */  /* A magic value for OP_RREF to indicate the "any recursion" condition. */
# Line 871  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 900  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
900         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
901         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
902         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
903         ERR60, ERR61 };         ERR60, ERR61, ERR62, ERR63, ERR64 };
904    
905  /* 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
906  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 934  typedef struct compile_data { Line 963  typedef struct compile_data {
963    uschar *name_table;           /* The name/number table */    uschar *name_table;           /* The name/number table */
964    int  names_found;             /* Number of entries so far */    int  names_found;             /* Number of entries so far */
965    int  name_entry_size;         /* Size of each entry */    int  name_entry_size;         /* Size of each entry */
966    int  bracount;                /* Count of capturing parens */    int  bracount;                /* Count of capturing parens as we compile */
967      int  final_bracount;          /* Saved value after first pass */
968    int  top_backref;             /* Maximum back reference */    int  top_backref;             /* Maximum back reference */
969    unsigned int backref_map;     /* Bitmap of low back refs */    unsigned int backref_map;     /* Bitmap of low back refs */
970    int  external_options;        /* External (initial) options */    int  external_options;        /* External (initial) options */
# Line 995  typedef struct match_data { Line 1025  typedef struct match_data {
1025    BOOL   notbol;                /* NOTBOL flag */    BOOL   notbol;                /* NOTBOL flag */
1026    BOOL   noteol;                /* NOTEOL flag */    BOOL   noteol;                /* NOTEOL flag */
1027    BOOL   utf8;                  /* UTF8 flag */    BOOL   utf8;                  /* UTF8 flag */
1028      BOOL   jscript_compat;        /* JAVASCRIPT_COMPAT flag */
1029    BOOL   endonly;               /* Dollar not before final \n */    BOOL   endonly;               /* Dollar not before final \n */
1030    BOOL   notempty;              /* Empty string match not wanted */    BOOL   notempty;              /* Empty string match not wanted */
1031    BOOL   partial;               /* PARTIAL flag */    BOOL   partial;               /* PARTIAL flag */
# Line 1036  typedef struct dfa_match_data { Line 1067  typedef struct dfa_match_data {
1067  #define ctype_letter  0x02  #define ctype_letter  0x02
1068  #define ctype_digit   0x04  #define ctype_digit   0x04
1069  #define ctype_xdigit  0x08  #define ctype_xdigit  0x08
1070  #define ctype_word    0x10   /* alphameric or '_' */  #define ctype_word    0x10   /* alphanumeric or '_' */
1071  #define ctype_meta    0x80   /* regexp meta char or zero (end pattern) */  #define ctype_meta    0x80   /* regexp meta char or zero (end pattern) */
1072    
1073  /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set  /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
# Line 1064  total length. */ Line 1095  total length. */
1095  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
1096    
1097  /* 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
1098  codes. */  codes. Each entry used to point directly to a name, but to reduce the number of
1099    relocations in shared libraries, it now has an offset into a single string
1100    instead. */
1101    
1102  typedef struct {  typedef struct {
1103    const char *name;    pcre_uint16 name_offset;
1104    pcre_uint16 type;    pcre_uint16 type;
1105    pcre_uint16 value;    pcre_uint16 value;
1106  } ucp_type_table;  } ucp_type_table;
# Line 1085  extern const uschar _pcre_utf8_table4[]; Line 1118  extern const uschar _pcre_utf8_table4[];
1118    
1119  extern const int    _pcre_utf8_table1_size;  extern const int    _pcre_utf8_table1_size;
1120    
1121    extern const char   _pcre_utt_names[];
1122  extern const ucp_type_table _pcre_utt[];  extern const ucp_type_table _pcre_utt[];
1123  extern const int _pcre_utt_size;  extern const int _pcre_utt_size;
1124    
# Line 1102  extern BOOL         _pcre_is_newline(con Line 1136  extern BOOL         _pcre_is_newline(con
1136  extern int          _pcre_ord2utf8(int, uschar *);  extern int          _pcre_ord2utf8(int, uschar *);
1137  extern real_pcre   *_pcre_try_flipped(const real_pcre *, real_pcre *,  extern real_pcre   *_pcre_try_flipped(const real_pcre *, real_pcre *,
1138                        const pcre_study_data *, pcre_study_data *);                        const pcre_study_data *, pcre_study_data *);
 extern int          _pcre_ucp_findprop(const unsigned int, int *, int *);  
 extern unsigned int _pcre_ucp_othercase(const unsigned int);  
1139  extern int          _pcre_valid_utf8(const uschar *, int);  extern int          _pcre_valid_utf8(const uschar *, int);
1140  extern BOOL         _pcre_was_newline(const uschar *, int, const uschar *,  extern BOOL         _pcre_was_newline(const uschar *, int, const uschar *,
1141                        int *, BOOL);                        int *, BOOL);
1142  extern BOOL         _pcre_xclass(int, const uschar *);  extern BOOL         _pcre_xclass(int, const uschar *);
1143    
1144    
1145    /* Unicode character database (UCD) */
1146    
1147    typedef struct {
1148      uschar script;
1149      uschar chartype;
1150      pcre_int32 other_case;
1151    } ucd_record;
1152    
1153    extern const ucd_record  _pcre_ucd_records[];
1154    extern const uschar      _pcre_ucd_stage1[];
1155    extern const pcre_uint16 _pcre_ucd_stage2[];
1156    extern const int         _pcre_ucp_gentype[];
1157    
1158    
1159    /* UCD access macros */
1160    
1161    #define UCD_BLOCK_SIZE 128
1162    #define GET_UCD(ch) (_pcre_ucd_records + \
1163            _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \
1164            UCD_BLOCK_SIZE + ch % UCD_BLOCK_SIZE])
1165    
1166    #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype
1167    #define UCD_SCRIPT(ch)    GET_UCD(ch)->script
1168    #define UCD_CATEGORY(ch)  _pcre_ucp_gentype[UCD_CHARTYPE(ch)]
1169    #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)
1170    
1171  #endif  #endif
1172    
1173  /* End of pcre_internal.h */  /* End of pcre_internal.h */

Legend:
Removed from v.231  
changed lines
  Added in v.364

  ViewVC Help
Powered by ViewVC 1.1.5