/[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 461 by ph10, Mon Oct 5 10:59:35 2009 UTC revision 499 by ph10, Wed Mar 3 20:09:39 2010 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) 1997-2009 University of Cambridge             Copyright (c) 1997-2010 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 45  functions whose names all begin with "_p Line 45  functions whose names all begin with "_p
45  #ifndef PCRE_INTERNAL_H  #ifndef PCRE_INTERNAL_H
46  #define PCRE_INTERNAL_H  #define PCRE_INTERNAL_H
47    
48  /* Define DEBUG to get debugging output on stdout. */  /* Define PCRE_DEBUG to get debugging output on stdout. */
49    
50  #if 0  #if 0
51  #define DEBUG  #define PCRE_DEBUG
52  #endif  #endif
53    
54  /* We do not support both EBCDIC and UTF-8 at the same time. The "configure"  /* We do not support both EBCDIC and UTF-8 at the same time. The "configure"
# Line 74  It turns out that the Mac Debugging.h he Line 74  It turns out that the Mac Debugging.h he
74  be absolutely sure we get our version. */  be absolutely sure we get our version. */
75    
76  #undef DPRINTF  #undef DPRINTF
77  #ifdef DEBUG  #ifdef PCRE_DEBUG
78  #define DPRINTF(p) printf p  #define DPRINTF(p) printf p
79  #else  #else
80  #define DPRINTF(p) /* Nothing */  #define DPRINTF(p) /* Nothing */
# Line 86  setjmp and stdarg are used is when NO_RE Line 86  setjmp and stdarg are used is when NO_RE
86    
87  #include <ctype.h>  #include <ctype.h>
88  #include <limits.h>  #include <limits.h>
 #include <setjmp.h>  
 #include <stdarg.h>  
89  #include <stddef.h>  #include <stddef.h>
90  #include <stdio.h>  #include <stdio.h>
91  #include <stdlib.h>  #include <stdlib.h>
# Line 186  preprocessor time in standard C environm Line 184  preprocessor time in standard C environm
184    #error Cannot determine a type for 32-bit unsigned integers    #error Cannot determine a type for 32-bit unsigned integers
185  #endif  #endif
186    
187    /* When checking for integer overflow in pcre_compile(), we need to handle
188    large integers. If a 64-bit integer type is available, we can use that.
189    Otherwise we have to cast to double, which of course requires floating point
190    arithmetic. Handle this by defining a macro for the appropriate type. If
191    stdint.h is available, include it; it may define INT64_MAX. Systems that do not
192    have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
193    by "configure". */
194    
195    #if HAVE_STDINT_H
196    #include <stdint.h>
197    #elif HAVE_INTTYPES_H
198    #include <inttypes.h>
199    #endif
200    
201    #if defined INT64_MAX || defined int64_t
202    #define INT64_OR_DOUBLE int64_t
203    #else
204    #define INT64_OR_DOUBLE double
205    #endif
206    
207  /* All character handling must be done as unsigned characters. Otherwise there  /* All character handling must be done as unsigned characters. Otherwise there
208  are problems with top-bit-set characters and functions such as isspace().  are problems with top-bit-set characters and functions such as isspace().
209  However, we leave the interface to the outside world as char *, because that  However, we leave the interface to the outside world as char *, because that
# Line 1210  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E Line 1228  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E
1228  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.
1229    
1230  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1231  that follow must also be updated to match. There is also a table called  that follow must also be updated to match. There are also tables called
1232  "coptable" in pcre_dfa_exec.c that must be updated. */  "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1233    
1234  enum {  enum {
1235    OP_END,            /* 0 End of pattern */    OP_END,            /* 0 End of pattern */
# Line 1373  enum { Line 1391  enum {
1391    
1392    /* This is used to skip a subpattern with a {0} quantifier */    /* This is used to skip a subpattern with a {0} quantifier */
1393    
1394    OP_SKIPZERO        /* 114 */    OP_SKIPZERO,       /* 114 */
1395    
1396      /* This is not an opcode, but is used to check that tables indexed by opcode
1397      are the correct length, in order to catch updating errors - there have been
1398      some in the past. */
1399    
1400      OP_TABLE_LENGTH
1401  };  };
1402    
1403    /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
1404    definitions that follow must also be updated to match. There are also tables
1405    called "coptable" cna "poptable" in pcre_dfa_exec.c that must be updated. */
1406    
1407    
1408  /* This macro defines textual names for all the opcodes. These are used only  /* This macro defines textual names for all the opcodes. These are used only
1409  for debugging. The macro is referenced only in pcre_printint.c. */  for debugging. The macro is referenced only in pcre_printint.c. */
# Line 1417  in UTF-8 mode. The code that uses this t Line 1445  in UTF-8 mode. The code that uses this t
1445    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \
1446    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
1447    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \
1448    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \    3, 3,                          /* \P, \p                                 */ \
1449    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
1450      1,                             /* \X                                     */ \
1451    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \
1452    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
1453    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Charnc  - the minimum length           */ \
# Line 1473  condition. */ Line 1502  condition. */
1502    
1503  #define RREF_ANY  0xffff  #define RREF_ANY  0xffff
1504    
1505  /* Error code numbers. They are given names so that they can more easily be  /* Compile time error code numbers. They are given names so that they can more
1506  tracked. */  easily be tracked. When a new number is added, the table called eint in
1507    pcreposix.c must be updated. */
1508    
1509  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,
1510         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
# Line 1482  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 1512  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
1512         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
1513         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
1514         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
1515         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65 };         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERRCOUNT };
1516    
1517  /* 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
1518  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 1532  typedef struct pcre_study_data { Line 1562  typedef struct pcre_study_data {
1562    
1563  /* Structure for building a chain of open capturing subpatterns during  /* Structure for building a chain of open capturing subpatterns during
1564  compiling, so that instructions to close them can be compiled when (*ACCEPT) is  compiling, so that instructions to close them can be compiled when (*ACCEPT) is
1565  encountered. */  encountered. This is also used to identify subpatterns that contain recursive
1566    back references to themselves, so that they can be made atomic. */
1567    
1568  typedef struct open_capitem {  typedef struct open_capitem {
1569    struct open_capitem *next;    /* Chain link */    struct open_capitem *next;    /* Chain link */
1570    pcre_uint16 number;           /* Capture number */    pcre_uint16 number;           /* Capture number */
1571      pcre_uint16 flag;             /* Set TRUE if recursive back ref */
1572  } open_capitem;  } open_capitem;
1573    
1574  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
# Line 1575  branches, for testing for left recursion Line 1607  branches, for testing for left recursion
1607    
1608  typedef struct branch_chain {  typedef struct branch_chain {
1609    struct branch_chain *outer;    struct branch_chain *outer;
1610    uschar *current;    uschar *current_branch;
1611  } branch_chain;  } branch_chain;
1612    
1613  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
# Line 1734  one of the exported public functions. Th Line 1766  one of the exported public functions. Th
1766  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
1767    
1768  extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);  extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);
1769  extern BOOL          _pcre_is_newline(const uschar *, int, const uschar *,  extern BOOL          _pcre_is_newline(USPTR, int, USPTR, int *, BOOL);
                        int *, BOOL);  
1770  extern int           _pcre_ord2utf8(int, uschar *);  extern int           _pcre_ord2utf8(int, uschar *);
1771  extern real_pcre    *_pcre_try_flipped(const real_pcre *, real_pcre *,  extern real_pcre    *_pcre_try_flipped(const real_pcre *, real_pcre *,
1772                         const pcre_study_data *, pcre_study_data *);                         const pcre_study_data *, pcre_study_data *);
1773  extern int           _pcre_valid_utf8(const uschar *, int);  extern int           _pcre_valid_utf8(USPTR, int);
1774  extern BOOL          _pcre_was_newline(const uschar *, int, const uschar *,  extern BOOL          _pcre_was_newline(USPTR, int, USPTR, int *, BOOL);
                        int *, BOOL);  
1775  extern BOOL          _pcre_xclass(int, const uschar *);  extern BOOL          _pcre_xclass(int, const uschar *);
1776    
1777    

Legend:
Removed from v.461  
changed lines
  Added in v.499

  ViewVC Help
Powered by ViewVC 1.1.5