-/* $Cambridge: exim/src/src/pcre/pcre_internal.h,v 1.6 2007/06/29 08:49:17 ph10 Exp $ */
+/* $Cambridge: exim/src/src/pcre/pcre_internal.h,v 1.7 2007/11/12 13:02:19 nm4 Exp $ */
/*************************************************
* Perl-Compatible Regular Expressions *
#endif
-/* Get the definitions provided by running "configure" */
-
-#include "config.h"
-
/* Standard C headers plus the external interface definition. The only time
setjmp and stdarg are used is when NO_RECURSE is set. */
#ifndef PCRE_EXP_DECL
# ifdef _WIN32
-# ifdef DLL_EXPORT
+# ifndef PCRE_STATIC
# define PCRE_EXP_DECL extern __declspec(dllexport)
# define PCRE_EXP_DEFN __declspec(dllexport)
# define PCRE_EXP_DATA_DEFN __declspec(dllexport)
# define PCRE_EXP_DEFN
# define PCRE_EXP_DATA_DEFN
# endif
-#
# else
# ifdef __cplusplus
# define PCRE_EXP_DECL extern "C"
/* When UTF-8 encoding is being used, a character is no longer just a single
byte. The macros for character handling generate simple sequences when used in
-byte-mode, and more complicated ones for UTF-8 characters. */
+byte-mode, and more complicated ones for UTF-8 characters. BACKCHAR should
+never be called in byte mode. To make sure it can never even appear when UTF-8
+support is omitted, we don't even define it. */
#ifndef SUPPORT_UTF8
#define GETCHAR(c, eptr) c = *eptr;
#define GETCHARINC(c, eptr) c = *eptr++;
#define GETCHARINCTEST(c, eptr) c = *eptr++;
#define GETCHARLEN(c, eptr, len) c = *eptr;
-#define BACKCHAR(eptr)
+/* #define BACKCHAR(eptr) */
#else /* SUPPORT_UTF8 */
}
/* If the pointer is not at the start of a character, move it back until
-it is. Called only in UTF-8 mode. */
+it is. This is called only in UTF-8 mode - we don't put a test within the macro
+because almost all calls are already within a block of UTF-8 only code. */
-#define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--;
+#define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
#endif
#define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)
-/* Private options flags start at the most significant end of the four bytes.
-The public options defined in pcre.h start at the least significant end. Make
-sure they don't overlap! The bits are getting a bit scarce now -- when we run
-out, there is a dummy word in the structure that could be used for the private
-bits. */
+/* Private flags containing information about the compiled regex. They used to
+live at the top end of the options word, but that got almost full, so now they
+are in a 16-bit flags word. */
-#define PCRE_NOPARTIAL 0x80000000 /* can't use partial with this regex */
-#define PCRE_FIRSTSET 0x40000000 /* first_byte is set */
-#define PCRE_REQCHSET 0x20000000 /* req_byte is set */
-#define PCRE_STARTLINE 0x10000000 /* start after \n for multiline */
-#define PCRE_JCHANGED 0x08000000 /* j option changes within regex */
+#define PCRE_NOPARTIAL 0x0001 /* can't use partial with this regex */
+#define PCRE_FIRSTSET 0x0002 /* first_byte is set */
+#define PCRE_REQCHSET 0x0004 /* req_byte is set */
+#define PCRE_STARTLINE 0x0008 /* start after \n for multiline */
+#define PCRE_JCHANGED 0x0010 /* j option used in regex */
+#define PCRE_HASCRORLF 0x0020 /* explicit \r or \n in pattern */
/* Options for the "extra" block produced by pcre_study(). */
(PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
- PCRE_DUPNAMES|PCRE_NEWLINE_BITS)
+ PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)
#define PUBLIC_EXEC_OPTIONS \
(PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \
- PCRE_PARTIAL|PCRE_NEWLINE_BITS)
+ PCRE_PARTIAL|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)
#define PUBLIC_DFA_EXEC_OPTIONS \
(PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \
- PCRE_PARTIAL|PCRE_DFA_SHORTEST|PCRE_DFA_RESTART|PCRE_NEWLINE_BITS)
+ PCRE_PARTIAL|PCRE_DFA_SHORTEST|PCRE_DFA_RESTART|PCRE_NEWLINE_BITS| \
+ PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)
#define PUBLIC_STUDY_OPTIONS 0 /* None defined */
ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_k, ESC_REF };
-/* Opcode table: OP_BRA must be last, as all values >= it are used for brackets
-that extract substrings. 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
OP_EOD must correspond in order to the list of escapes immediately above.
-To keep stored, compiled patterns compatible, new opcodes should be added
-immediately before OP_BRA, where (since release 7.0) a gap is left for this
-purpose.
-
*** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
that follow must also be updated to match. There is also a table called
"coptable" in pcre_dfa_exec.c that must be updated. */
as there's a test for >= ONCE for a subpattern that isn't an assertion. */
OP_ONCE, /* 92 Atomic group */
- OP_BRA, /* 83 Start of non-capturing bracket */
+ OP_BRA, /* 93 Start of non-capturing bracket */
OP_CBRA, /* 94 Start of capturing bracket */
OP_COND, /* 95 Conditional group */
OP_DEF, /* 101 The DEFINE condition */
OP_BRAZERO, /* 102 These two must remain together and in this */
- OP_BRAMINZERO /* 103 order. */
+ OP_BRAMINZERO, /* 103 order. */
+
+ /* These are backtracking control verbs */
+
+ OP_PRUNE, /* 104 */
+ OP_SKIP, /* 105 */
+ OP_THEN, /* 106 */
+ OP_COMMIT, /* 107 */
+
+ /* These are forced failure and success verbs */
+
+ OP_FAIL, /* 108 */
+ OP_ACCEPT /* 109 */
};
"class", "nclass", "xclass", "Ref", "Recurse", "Callout", \
"Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", \
"AssertB", "AssertB not", "Reverse", \
- "Once", "Bra 0", "Bra", "Cond", "SBra 0", "SBra", "SCond", \
- "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero"
+ "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond", \
+ "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero", \
+ "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT"
/* This macro defines the length of fixed length operations in the compiled
3, /* RREF */ \
1, /* DEF */ \
1, 1, /* BRAZERO, BRAMINZERO */ \
+ 1, 1, 1, 1, /* PRUNE, SKIP, THEN, COMMIT, */ \
+ 1, 1 /* FAIL, ACCEPT */
/* A magic value for OP_RREF to indicate the "any recursion" condition. */
ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
- ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58 };
+ ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
+ ERR60, ERR61 };
/* The real format of the start of the pcre block; the index of names and the
code vector run on as long as necessary after the end. We store an explicit
typedef struct real_pcre {
pcre_uint32 magic_number;
pcre_uint32 size; /* Total that was malloced */
- pcre_uint32 options;
- pcre_uint32 dummy1; /* For future use, maybe */
-
+ pcre_uint32 options; /* Public options */
+ pcre_uint16 flags; /* Private flags */
+ pcre_uint16 dummy1; /* For future use */
pcre_uint16 top_bracket;
pcre_uint16 top_backref;
pcre_uint16 first_byte;
int top_backref; /* Maximum back reference */
unsigned int backref_map; /* Bitmap of low back refs */
int external_options; /* External (initial) options */
+ int external_flags; /* External flag bits to be set */
int req_varyopt; /* "After variable item" flag for reqbyte */
- BOOL nopartial; /* Set TRUE if partial won't work */
+ BOOL had_accept; /* (*ACCEPT) encountered */
int nltype; /* Newline type */
int nllen; /* Newline string length */
uschar nl[4]; /* Newline string when fixed length */
BOOL notempty; /* Empty string match not wanted */
BOOL partial; /* PARTIAL flag */
BOOL hitend; /* Hit the end of the subject at some point */
+ BOOL bsr_anycrlf; /* \R is just any CRLF, not full Unicode */
const uschar *start_code; /* For use when recursing */
USPTR start_subject; /* Start of the subject string */
USPTR end_subject; /* End of the subject string */
#define tables_length (ctypes_offset + 256)
/* Layout of the UCP type table that translates property names into types and
-codes. */
+codes. Each entry used to point directly to a name, but to reduce the number of
+relocations in shared libraries, it now has an offset into a single string
+instead. */
typedef struct {
- const char *name;
+ pcre_uint16 name_offset;
pcre_uint16 type;
pcre_uint16 value;
} ucp_type_table;
extern const int _pcre_utf8_table1_size;
+extern const char _pcre_utt_names[];
extern const ucp_type_table _pcre_utt[];
extern const int _pcre_utt_size;
projects / exim.git / blobdiff