/[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 1046 by ph10, Tue Sep 25 16:27:58 2012 UTC revision 1248 by ph10, Wed Feb 13 17:36:38 2013 UTC
# Line 40  POSSIBILITY OF SUCH DAMAGE. Line 40  POSSIBILITY OF SUCH DAMAGE.
40    
41  /* This header contains definitions that are shared between the different  /* This header contains definitions that are shared between the different
42  modules, but which are not relevant to the exported API. This includes some  modules, but which are not relevant to the exported API. This includes some
43  functions whose names all begin with "_pcre_" or "_pcre16_" depending on  functions whose names all begin with "_pcre_", "_pcre16_" or "_pcre32_"
44  the PRIV macro. */  depending on the PRIV macro. */
45    
46  #ifndef PCRE_INTERNAL_H  #ifndef PCRE_INTERNAL_H
47  #define PCRE_INTERNAL_H  #define PCRE_INTERNAL_H
# Line 53  the PRIV macro. */ Line 53  the PRIV macro. */
53  #endif  #endif
54    
55  /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */  /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
56  #ifndef COMPILE_PCRE16  
57    #if !defined COMPILE_PCRE16 && !defined COMPILE_PCRE32
58  #define COMPILE_PCRE8  #define COMPILE_PCRE8
59  #endif  #endif
60    
# Line 78  Until then we define it if SUPPORT_UTF i Line 79  Until then we define it if SUPPORT_UTF i
79  #define SUPPORT_UTF8 1  #define SUPPORT_UTF8 1
80  #endif  #endif
81    
82  /* We do not support both EBCDIC and UTF-8/16 at the same time. The "configure"  /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
83  script prevents both being selected, but not everybody uses "configure". */  script prevents both being selected, but not everybody uses "configure". */
84    
85  #if defined EBCDIC && defined SUPPORT_UTF  #if defined EBCDIC && defined SUPPORT_UTF
86  #error The use of both EBCDIC and SUPPORT_UTF8/16 is not supported.  #error The use of both EBCDIC and SUPPORT_UTF is not supported.
87  #endif  #endif
88    
89  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
# Line 111  setjmp and stdarg are used is when NO_RE Line 112  setjmp and stdarg are used is when NO_RE
112  #include <stdlib.h>  #include <stdlib.h>
113  #include <string.h>  #include <string.h>
114    
115    /* Valgrind (memcheck) support */
116    
117    #ifdef SUPPORT_VALGRIND
118    #include <valgrind/memcheck.h>
119    #endif
120    
121  /* When compiling a DLL for Windows, the exported symbols have to be declared  /* When compiling a DLL for Windows, the exported symbols have to be declared
122  using some MS magic. I found some useful information on this web page:  using some MS magic. I found some useful information on this web page:
123  http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the  http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
# Line 193  typedef unsigned char pcre_uint8; Line 200  typedef unsigned char pcre_uint8;
200    typedef unsigned int pcre_uint16;    typedef unsigned int pcre_uint16;
201    typedef int pcre_int16;    typedef int pcre_int16;
202  #else  #else
203    #error Cannot determine a type for 16-bit unsigned integers  # error Cannot determine a type for 16-bit unsigned integers
204  #endif  #endif
205    
206  #if UINT_MAX == 4294967295  #if UINT_MAX == 4294967295
# Line 203  typedef unsigned char pcre_uint8; Line 210  typedef unsigned char pcre_uint8;
210    typedef unsigned long int pcre_uint32;    typedef unsigned long int pcre_uint32;
211    typedef long int pcre_int32;    typedef long int pcre_int32;
212  #else  #else
213    #error Cannot determine a type for 32-bit unsigned integers  # error Cannot determine a type for 32-bit unsigned integers
214  #endif  #endif
215    
216  /* When checking for integer overflow in pcre_compile(), we need to handle  /* When checking for integer overflow in pcre_compile(), we need to handle
# Line 214  stdint.h is available, include it; it ma Line 221  stdint.h is available, include it; it ma
221  have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set  have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
222  by "configure". */  by "configure". */
223    
224  #if HAVE_STDINT_H  #if defined HAVE_STDINT_H
225  #include <stdint.h>  #include <stdint.h>
226  #elif HAVE_INTTYPES_H  #elif defined HAVE_INTTYPES_H
227  #include <inttypes.h>  #include <inttypes.h>
228  #endif  #endif
229    
# Line 243  exactly 256 items. When the character is Line 250  exactly 256 items. When the character is
250  items, some check is needed before accessing these tables.  items, some check is needed before accessing these tables.
251  */  */
252    
253  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
254    
255  typedef unsigned char pcre_uchar;  typedef unsigned char pcre_uchar;
256  #define IN_UCHARS(x) (x)  #define IN_UCHARS(x) (x)
257  #define MAX_255(c) 1  #define MAX_255(c) 1
258  #define TABLE_GET(c, table, default) ((table)[c])  #define TABLE_GET(c, table, default) ((table)[c])
259    
260  #else  #elif defined COMPILE_PCRE16
261    
 #ifdef COMPILE_PCRE16  
262  #if USHRT_MAX != 65535  #if USHRT_MAX != 65535
263  /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in  /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
264  pcre.h(.in) and disable (comment out) this message. */  pcre.h(.in) and disable (comment out) this message. */
# Line 260  pcre.h(.in) and disable (comment out) th Line 266  pcre.h(.in) and disable (comment out) th
266  #endif  #endif
267    
268  typedef pcre_uint16 pcre_uchar;  typedef pcre_uint16 pcre_uchar;
269  #define IN_UCHARS(x) ((x) << 1)  #define UCHAR_SHIFT (1)
270    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
271    #define MAX_255(c) ((c) <= 255u)
272    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
273    
274    #elif defined COMPILE_PCRE32
275    
276    typedef pcre_uint32 pcre_uchar;
277    #define UCHAR_SHIFT (2)
278    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
279  #define MAX_255(c) ((c) <= 255u)  #define MAX_255(c) ((c) <= 255u)
280  #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))  #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
281    
282  #else  #else
283  #error Unsupported compiling mode  #error Unsupported compiling mode
284  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
285    
286  /* This is an unsigned int value that no character can ever have. UTF-8  /* This is an unsigned int value that no character can ever have. UTF-8
287  characters only go up to 0x7fffffff (though Unicode doesn't go beyond  characters only go up to 0x7fffffff (though Unicode doesn't go beyond
# Line 295  start/end of string field names are. */ Line 308  start/end of string field names are. */
308         &(NLBLOCK->nllen), utf)) \         &(NLBLOCK->nllen), utf)) \
309      : \      : \
310      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
311       (p)[0] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
312       (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
313      ) \      ) \
314    )    )
315    
# Line 309  start/end of string field names are. */ Line 322  start/end of string field names are. */
322         &(NLBLOCK->nllen), utf)) \         &(NLBLOCK->nllen), utf)) \
323      : \      : \
324      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
325       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
326       (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
327      ) \      ) \
328    )    )
329    
# Line 335  values. */ Line 348  values. */
348  #include "pcre.h"  #include "pcre.h"
349  #include "ucp.h"  #include "ucp.h"
350    
351    #ifdef COMPILE_PCRE32
352    /* Assert that the public PCRE_UCHAR32 is a 32-bit type */
353    typedef int __assert_pcre_uchar32_size[sizeof(PCRE_UCHAR32) == 4 ? 1 : -1];
354    #endif
355    
356  /* When compiling for use with the Virtual Pascal compiler, these functions  /* When compiling for use with the Virtual Pascal compiler, these functions
357  need to have their names changed. PCRE must be compiled with the -DVPCOMPAT  need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
358  option on the command line. */  option on the command line. */
# Line 396  The macros are controlled by the value o Line 414  The macros are controlled by the value o
414  the config.h file, but can be overridden by using -D on the command line. This  the config.h file, but can be overridden by using -D on the command line. This
415  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
416    
417  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
418    
419  #if LINK_SIZE == 2  #if LINK_SIZE == 2
420    
# Line 441  is automated on Unix systems via the "co Line 459  is automated on Unix systems via the "co
459  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
460  #endif  #endif
461    
462  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
   
 #ifdef COMPILE_PCRE16  
463    
464  #if LINK_SIZE == 2  #if LINK_SIZE == 2
465    
466    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
467  #undef LINK_SIZE  #undef LINK_SIZE
468  #define LINK_SIZE 1  #define LINK_SIZE 1
469    
# Line 460  is automated on Unix systems via the "co Line 477  is automated on Unix systems via the "co
477    
478  #elif LINK_SIZE == 3 || LINK_SIZE == 4  #elif LINK_SIZE == 3 || LINK_SIZE == 4
479    
480    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
481  #undef LINK_SIZE  #undef LINK_SIZE
482  #define LINK_SIZE 2  #define LINK_SIZE 2
483    
# Line 477  is automated on Unix systems via the "co Line 495  is automated on Unix systems via the "co
495  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
496  #endif  #endif
497    
498    #elif defined COMPILE_PCRE32
499    
500    /* Only supported LINK_SIZE is 4 */
501    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
502    #undef LINK_SIZE
503    #define LINK_SIZE 1
504    
505    #define PUT(a,n,d)   \
506      (a[n] = (d))
507    
508    #define GET(a,n) \
509      (a[n])
510    
511    /* Keep it positive */
512    #define MAX_PATTERN_SIZE (1 << 30)
513    
514  #else  #else
515  #error Unsupported compiling mode  #error Unsupported compiling mode
516  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
517    
518  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
519    
# Line 492  is automated on Unix systems via the "co Line 524  is automated on Unix systems via the "co
524  offsets changes. There are used for repeat counts and for other things such as  offsets changes. There are used for repeat counts and for other things such as
525  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
526    
527  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
528    
529  #define IMM2_SIZE 2  #define IMM2_SIZE 2
530    
# Line 500  capturing parenthesis numbers in back re Line 532  capturing parenthesis numbers in back re
532    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
533    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
534    
535    /* For reasons that I do not understand, the expression in this GET2 macro is
536    treated by gcc as a signed expression, even when a is declared as unsigned. It
537    seems that any kind of arithmetic results in a signed value. */
538    
539  #define GET2(a,n) \  #define GET2(a,n) \
540    (((a)[n] << 8) | (a)[(n)+1])    (unsigned int)(((a)[n] << 8) | (a)[(n)+1])
541    
542  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
543    
544  #ifdef COMPILE_PCRE16  #define IMM2_SIZE 1
545    
546    #define PUT2(a,n,d)   \
547       a[n] = d
548    
549    #define GET2(a,n) \
550       a[n]
551    
552    #elif defined COMPILE_PCRE32
553    
554  #define IMM2_SIZE 1  #define IMM2_SIZE 1
555    
# Line 517  capturing parenthesis numbers in back re Line 561  capturing parenthesis numbers in back re
561    
562  #else  #else
563  #error Unsupported compiling mode  #error Unsupported compiling mode
564  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
565    
566  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
567    
568  /* The maximum length of a MARK name is currently one data unit; it may be  /* The maximum length of a MARK name is currently one data unit; it may be
569  changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */  changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
570    
571  #define MAX_MARK ((1 << (sizeof(pcre_uchar)*8)) - 1)  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
572    #define MAX_MARK ((1u << 16) - 1)
573    #else
574    #define MAX_MARK ((1u << 8) - 1)
575    #endif
576    
577  /* When UTF encoding is being used, a character is no longer just a single  /* When UTF encoding is being used, a character is no longer just a single
578  byte. The macros for character handling generate simple sequences when used in  byte. The macros for character handling generate simple sequences when used in
# Line 546  we don't even define them. */ Line 592  we don't even define them. */
592  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
593  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
594  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
595    #define RAWUCHAR(eptr) (*(eptr))
596    #define RAWUCHARINC(eptr) (*(eptr)++)
597    #define RAWUCHARTEST(eptr) (*(eptr))
598    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
599  /* #define GETCHARLENTEST(c, eptr, len) */  /* #define GETCHARLENTEST(c, eptr, len) */
600  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
601  /* #define FORWARDCHAR(eptr) */  /* #define FORWARDCHAR(eptr) */
# Line 553  we don't even define them. */ Line 603  we don't even define them. */
603    
604  #else   /* SUPPORT_UTF */  #else   /* SUPPORT_UTF */
605    
 #ifdef COMPILE_PCRE8  
   
 /* These macros were originally written in the form of loops that used data  
 from the tables whose names start with PRIV(utf8_table). They were rewritten by  
 a user so as not to use loops, because in some environments this gives a  
 significant performance advantage, and it seems never to do any harm. */  
   
 /* Tells the biggest code point which can be encoded as a single character. */  
   
 #define MAX_VALUE_FOR_SINGLE_CHAR 127  
   
606  /* Tests whether the code point needs extra characters to decode. */  /* Tests whether the code point needs extra characters to decode. */
607    
608  #define HAS_EXTRALEN(c) ((c) >= 0xc0)  #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
   
 /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.  
 Otherwise it has an undefined behaviour. */  
   
 #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])  
   
 /* Returns TRUE, if the given character is not the first character  
 of a UTF sequence. */  
   
 #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)  
609    
610  /* Base macro to pick up the remaining bytes of a UTF-8 character, not  /* Base macro to pick up the remaining bytes of a UTF-8 character, not
611  advancing the pointer. */  advancing the pointer. */
# Line 600  advancing the pointer. */ Line 629  advancing the pointer. */
629            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
630      }      }
631    
 /* Get the next UTF-8 character, not advancing the pointer. This is called when  
 we know we are in UTF-8 mode. */  
   
 #define GETCHAR(c, eptr) \  
   c = *eptr; \  
   if (c >= 0xc0) GETUTF8(c, eptr);  
   
 /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the  
 pointer. */  
   
 #define GETCHARTEST(c, eptr) \  
   c = *eptr; \  
   if (utf && c >= 0xc0) GETUTF8(c, eptr);  
   
632  /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing  /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
633  the pointer. */  the pointer. */
634    
# Line 648  the pointer. */ Line 663  the pointer. */
663        } \        } \
664      }      }
665    
666    #if defined COMPILE_PCRE8
667    
668    /* These macros were originally written in the form of loops that used data
669    from the tables whose names start with PRIV(utf8_table). They were rewritten by
670    a user so as not to use loops, because in some environments this gives a
671    significant performance advantage, and it seems never to do any harm. */
672    
673    /* Tells the biggest code point which can be encoded as a single character. */
674    
675    #define MAX_VALUE_FOR_SINGLE_CHAR 127
676    
677    /* Tests whether the code point needs extra characters to decode. */
678    
679    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
680    
681    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
682    Otherwise it has an undefined behaviour. */
683    
684    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
685    
686    /* Returns TRUE, if the given character is not the first character
687    of a UTF sequence. */
688    
689    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
690    
691    /* Get the next UTF-8 character, not advancing the pointer. This is called when
692    we know we are in UTF-8 mode. */
693    
694    #define GETCHAR(c, eptr) \
695      c = *eptr; \
696      if (c >= 0xc0) GETUTF8(c, eptr);
697    
698    /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
699    pointer. */
700    
701    #define GETCHARTEST(c, eptr) \
702      c = *eptr; \
703      if (utf && c >= 0xc0) GETUTF8(c, eptr);
704    
705  /* Get the next UTF-8 character, advancing the pointer. This is called when we  /* Get the next UTF-8 character, advancing the pointer. This is called when we
706  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
707    
# Line 714  do not know if we are in UTF-8 mode. */ Line 768  do not know if we are in UTF-8 mode. */
768    c = *eptr; \    c = *eptr; \
769    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
770    
771    /* Returns the next uchar, not advancing the pointer. This is called when
772    we know we are in UTF mode. */
773    
774    #define RAWUCHAR(eptr) \
775      (*(eptr))
776    
777    /* Returns the next uchar, advancing the pointer. This is called when
778    we know we are in UTF mode. */
779    
780    #define RAWUCHARINC(eptr) \
781      (*((eptr)++))
782    
783    /* Returns the next uchar, testing for UTF mode, and not advancing the
784    pointer. */
785    
786    #define RAWUCHARTEST(eptr) \
787      (*(eptr))
788    
789    /* Returns the next uchar, testing for UTF mode, advancing the
790    pointer. */
791    
792    #define RAWUCHARINCTEST(eptr) \
793      (*((eptr)++))
794    
795  /* If the pointer is not at the start of a character, move it back until  /* If the pointer is not at the start of a character, move it back until
796  it is. This is called only in UTF-8 mode - we don't put a test within the macro  it is. This is called only in UTF-8 mode - we don't put a test within the macro
797  because almost all calls are already within a block of UTF-8 only code. */  because almost all calls are already within a block of UTF-8 only code. */
# Line 727  because almost all calls are already wit Line 805  because almost all calls are already wit
805  #define ACROSSCHAR(condition, eptr, action) \  #define ACROSSCHAR(condition, eptr, action) \
806    while((condition) && ((eptr) & 0xc0) == 0x80) action    while((condition) && ((eptr) & 0xc0) == 0x80) action
807    
808  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
   
 #ifdef COMPILE_PCRE16  
809    
810  /* Tells the biggest code point which can be encoded as a single character. */  /* Tells the biggest code point which can be encoded as a single character. */
811    
# Line 811  we do not know if we are in UTF-16 mode. Line 887  we do not know if we are in UTF-16 mode.
887    c = *eptr; \    c = *eptr; \
888    if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);    if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
889    
890    /* Returns the next uchar, not advancing the pointer. This is called when
891    we know we are in UTF mode. */
892    
893    #define RAWUCHAR(eptr) \
894      (*(eptr))
895    
896    /* Returns the next uchar, advancing the pointer. This is called when
897    we know we are in UTF mode. */
898    
899    #define RAWUCHARINC(eptr) \
900      (*((eptr)++))
901    
902    /* Returns the next uchar, testing for UTF mode, and not advancing the
903    pointer. */
904    
905    #define RAWUCHARTEST(eptr) \
906      (*(eptr))
907    
908    /* Returns the next uchar, testing for UTF mode, advancing the
909    pointer. */
910    
911    #define RAWUCHARINCTEST(eptr) \
912      (*((eptr)++))
913    
914  /* If the pointer is not at the start of a character, move it back until  /* If the pointer is not at the start of a character, move it back until
915  it is. This is called only in UTF-16 mode - we don't put a test within the  it is. This is called only in UTF-16 mode - we don't put a test within the
916  macro because almost all calls are already within a block of UTF-16 only  macro because almost all calls are already within a block of UTF-16 only
# Line 825  code. */ Line 925  code. */
925  #define ACROSSCHAR(condition, eptr, action) \  #define ACROSSCHAR(condition, eptr, action) \
926    if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action    if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
927    
928  #endif  #elif defined COMPILE_PCRE32
929    
930  #endif /* COMPILE_PCRE8 */  /* These are trivial for the 32-bit library, since all UTF-32 characters fit
931    into one pcre_uchar unit. */
932    #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
933    #define HAS_EXTRALEN(c) (0)
934    #define GET_EXTRALEN(c) (0)
935    #define NOT_FIRSTCHAR(c) (0)
936    
937  #endif  /* SUPPORT_UTF */  /* Get the next UTF-32 character, not advancing the pointer. This is called when
938    we know we are in UTF-32 mode. */
939    
940    #define GETCHAR(c, eptr) \
941      c = *(eptr);
942    
943    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
944    pointer. */
945    
946    #define GETCHARTEST(c, eptr) \
947      c = *(eptr);
948    
949    /* Get the next UTF-32 character, advancing the pointer. This is called when we
950    know we are in UTF-32 mode. */
951    
952    #define GETCHARINC(c, eptr) \
953      c = *((eptr)++);
954    
955    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
956    This is called when we don't know if we are in UTF-32 mode. */
957    
958    #define GETCHARINCTEST(c, eptr) \
959      c = *((eptr)++);
960    
961    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
962    length (since all UTF-32 is of length 1). This is called when we know we are in
963    UTF-32 mode. */
964    
965    #define GETCHARLEN(c, eptr, len) \
966      GETCHAR(c, eptr)
967    
968    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
969    pointer, not incrementing the length (since all UTF-32 is of length 1).
970    This is called when we do not know if we are in UTF-32 mode. */
971    
972    #define GETCHARLENTEST(c, eptr, len) \
973      GETCHARTEST(c, eptr)
974    
975    /* Returns the next uchar, not advancing the pointer. This is called when
976    we know we are in UTF mode. */
977    
978    #define RAWUCHAR(eptr) \
979      (*(eptr))
980    
981    /* Returns the next uchar, advancing the pointer. This is called when
982    we know we are in UTF mode. */
983    
984    #define RAWUCHARINC(eptr) \
985      (*((eptr)++))
986    
987    /* Returns the next uchar, testing for UTF mode, and not advancing the
988    pointer. */
989    
990    #define RAWUCHARTEST(eptr) \
991      (*(eptr))
992    
993    /* Returns the next uchar, testing for UTF mode, advancing the
994    pointer. */
995    
996    #define RAWUCHARINCTEST(eptr) \
997      (*((eptr)++))
998    
999    /* If the pointer is not at the start of a character, move it back until
1000    it is. This is called only in UTF-32 mode - we don't put a test within the
1001    macro because almost all calls are already within a block of UTF-32 only
1002    code.
1003    These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
1004    
1005    #define BACKCHAR(eptr) do { } while (0)
1006    
1007    /* Same as above, just in the other direction. */
1008    #define FORWARDCHAR(eptr) do { } while (0)
1009    
1010    /* Same as above, but it allows a fully customizable form. */
1011    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1012    
1013    #else
1014    #error Unsupported compiling mode
1015    #endif /* COMPILE_PCRE[8|16|32] */
1016    
1017    #endif  /* SUPPORT_UTF */
1018    
1019  /* Tests for Unicode horizontal and vertical whitespace characters must check a  /* Tests for Unicode horizontal and vertical whitespace characters must check a
1020  number of different values. Using a switch statement for this generates the  number of different values. Using a switch statement for this generates the
# Line 842  are defined. Line 1026  are defined.
1026  These values are also required as lists in pcre_compile.c when processing \h,  These values are also required as lists in pcre_compile.c when processing \h,
1027  \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but  \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1028  macros that define the values are here so that all the definitions are  macros that define the values are here so that all the definitions are
1029  together. The lists must be in ascending character order, terminated by  together. The lists must be in ascending character order, terminated by
1030  NOTACHAR (which is 0xffffffff).  NOTACHAR (which is 0xffffffff).
1031    
1032  Any changes should ensure that the various macros are kept in step with each  Any changes should ensure that the various macros are kept in step with each
# Line 856  other. NOTE: The values also appear in p Line 1040  other. NOTE: The values also appear in p
1040    CHAR_HT, CHAR_SPACE, 0xa0, \    CHAR_HT, CHAR_SPACE, 0xa0, \
1041    0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \    0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1042    0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \    0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1043    NOTACHAR    NOTACHAR
1044    
1045  #define HSPACE_MULTIBYTE_CASES \  #define HSPACE_MULTIBYTE_CASES \
1046    case 0x1680:  /* OGHAM SPACE MARK */ \    case 0x1680:  /* OGHAM SPACE MARK */ \
# Line 880  other. NOTE: The values also appear in p Line 1064  other. NOTE: The values also appear in p
1064    case CHAR_HT: \    case CHAR_HT: \
1065    case CHAR_SPACE: \    case CHAR_SPACE: \
1066    case 0xa0     /* NBSP */    case 0xa0     /* NBSP */
1067    
1068  #define HSPACE_CASES \  #define HSPACE_CASES \
1069    HSPACE_BYTE_CASES: \    HSPACE_BYTE_CASES: \
1070    HSPACE_MULTIBYTE_CASES    HSPACE_MULTIBYTE_CASES
1071    
1072  #define VSPACE_LIST \  #define VSPACE_LIST \
1073    CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR    CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1074    
1075  #define VSPACE_MULTIBYTE_CASES \  #define VSPACE_MULTIBYTE_CASES \
1076    case 0x2028:    /* LINE SEPARATOR */ \    case 0x2028:    /* LINE SEPARATOR */ \
# Line 916  other. NOTE: The values also appear in p Line 1100  other. NOTE: The values also appear in p
1100    
1101  #ifdef EBCDIC_NL25  #ifdef EBCDIC_NL25
1102  #define VSPACE_LIST \  #define VSPACE_LIST \
1103    CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR    CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1104  #else  #else
1105  #define VSPACE_LIST \  #define VSPACE_LIST \
1106    CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR    CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1107  #endif  #endif
1108    
1109  #define VSPACE_BYTE_CASES \  #define VSPACE_BYTE_CASES \
1110    case CHAR_LF: \    case CHAR_LF: \
# Line 935  other. NOTE: The values also appear in p Line 1119  other. NOTE: The values also appear in p
1119  /* ------ End of whitespace macros ------ */  /* ------ End of whitespace macros ------ */
1120    
1121    
 /* In case there is no definition of offsetof() provided - though any proper  
 Standard C system should have one. */  
   
 #ifndef offsetof  
 #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
 #endif  
   
1122    
1123  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
1124  live at the top end of the options word, but that got almost full, so now they  live at the top end of the options word, but that got almost full, so now they
# Line 949  are in a 16-bit flags word. From release Line 1126  are in a 16-bit flags word. From release
1126  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1127  compatibility. */  compatibility. */
1128    
1129  #ifdef COMPILE_PCRE8  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1130  #define PCRE_MODE          0x0001  /* compiled in 8 bit mode */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1131  #endif  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
 #ifdef COMPILE_PCRE16  
 #define PCRE_MODE          0x0002  /* compiled in 16 bit mode */  
 #endif  
1132  #define PCRE_FIRSTSET      0x0010  /* first_char is set */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1133  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1134  #define PCRE_REQCHSET      0x0040  /* req_byte is set */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
# Line 965  compatibility. */ Line 1139  compatibility. */
1139  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1140  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1141    
1142    #if defined COMPILE_PCRE8
1143    #define PCRE_MODE          PCRE_MODE8
1144    #elif defined COMPILE_PCRE16
1145    #define PCRE_MODE          PCRE_MODE16
1146    #elif defined COMPILE_PCRE32
1147    #define PCRE_MODE          PCRE_MODE32
1148    #endif
1149    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1150    
1151  /* Flags for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1152    
1153  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
# Line 998  time, run time, or study time, respectiv Line 1181  time, run time, or study time, respectiv
1181     (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \     (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1182      PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)      PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1183    
1184    #define PUBLIC_JIT_EXEC_OPTIONS \
1185       (PCRE_NO_UTF8_CHECK|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|\
1186        PCRE_NOTEMPTY_ATSTART|PCRE_PARTIAL_SOFT|PCRE_PARTIAL_HARD)
1187    
1188  /* Magic number to provide a small check against being handed junk. */  /* Magic number to provide a small check against being handed junk. */
1189    
1190  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
# Line 1007  in different endianness. */ Line 1194  in different endianness. */
1194    
1195  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
1196    
 /* Negative values for the firstchar and reqchar variables */  
   
 #define REQ_UNSET (-2)  
 #define REQ_NONE  (-1)  
   
1197  /* The maximum remaining length of subject we are prepared to search for a  /* The maximum remaining length of subject we are prepared to search for a
1198  req_byte match. */  req_byte match. */
1199    
# Line 1106  a positive value. */ Line 1288  a positive value. */
1288    
1289  /* The remaining definitions work in both environments. */  /* The remaining definitions work in both environments. */
1290    
1291    #define CHAR_NULL                   '\0'
1292  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1293  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1294  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
# Line 1345  a positive value. */ Line 1528  a positive value. */
1528  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1529  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1530  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1531  #ifdef COMPILE_PCRE8  #define STRING_UTF8_RIGHTPAR           "UTF8)"
1532  #define STRING_UTF_RIGHTPAR            "UTF8)"  #define STRING_UTF16_RIGHTPAR          "UTF16)"
1533  #endif  #define STRING_UTF32_RIGHTPAR          "UTF32)"
1534  #ifdef COMPILE_PCRE16  #define STRING_UTF_RIGHTPAR            "UTF)"
 #define STRING_UTF_RIGHTPAR            "UTF16)"  
 #endif  
1535  #define STRING_UCP_RIGHTPAR            "UCP)"  #define STRING_UCP_RIGHTPAR            "UCP)"
1536  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1537    
# Line 1372  only. */ Line 1553  only. */
1553  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
1554  #define CHAR_DEL                    '\177'  #define CHAR_DEL                    '\177'
1555    
1556    #define CHAR_NULL                   '\0'
1557  #define CHAR_SPACE                  '\040'  #define CHAR_SPACE                  '\040'
1558  #define CHAR_EXCLAMATION_MARK       '\041'  #define CHAR_EXCLAMATION_MARK       '\041'
1559  #define CHAR_QUOTATION_MARK         '\042'  #define CHAR_QUOTATION_MARK         '\042'
# Line 1607  only. */ Line 1789  only. */
1789  #define STRING_ANYCRLF_RIGHTPAR        STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS  #define STRING_ANYCRLF_RIGHTPAR        STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1790  #define STRING_BSR_ANYCRLF_RIGHTPAR    STR_B STR_S STR_R STR_UNDERSCORE STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS  #define STRING_BSR_ANYCRLF_RIGHTPAR    STR_B STR_S STR_R STR_UNDERSCORE STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1791  #define STRING_BSR_UNICODE_RIGHTPAR    STR_B STR_S STR_R STR_UNDERSCORE STR_U STR_N STR_I STR_C STR_O STR_D STR_E STR_RIGHT_PARENTHESIS  #define STRING_BSR_UNICODE_RIGHTPAR    STR_B STR_S STR_R STR_UNDERSCORE STR_U STR_N STR_I STR_C STR_O STR_D STR_E STR_RIGHT_PARENTHESIS
1792  #ifdef COMPILE_PCRE8  #define STRING_UTF8_RIGHTPAR           STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1793  #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS  #define STRING_UTF16_RIGHTPAR          STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1794  #endif  #define STRING_UTF32_RIGHTPAR          STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1795  #ifdef COMPILE_PCRE16  #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_RIGHT_PARENTHESIS
 #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS  
 #endif  
1796  #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS  #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1797  #define STRING_NO_START_OPT_RIGHTPAR   STR_N STR_O STR_UNDERSCORE STR_S STR_T STR_A STR_R STR_T STR_UNDERSCORE STR_O STR_P STR_T STR_RIGHT_PARENTHESIS  #define STRING_NO_START_OPT_RIGHTPAR   STR_N STR_O STR_UNDERSCORE STR_S STR_T STR_A STR_R STR_T STR_UNDERSCORE STR_O STR_P STR_T STR_RIGHT_PARENTHESIS
1798    
# Line 1670  contain characters with values greater t Line 1850  contain characters with values greater t
1850    
1851  /* These are escaped items that aren't just an encoding of a particular data  /* These are escaped items that aren't just an encoding of a particular data
1852  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
1853  their negation. Also, they must appear in the same order as in the opcode  0 for a data character.  Also, they must appear in the same order as in the opcode
1854  definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it  definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1855  corresponds to "." in DOTALL mode rather than an escape sequence. It is also  corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1856  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
# Line 1681  when PCRE_UCP is set and replacement of Line 1861  when PCRE_UCP is set and replacement of
1861  They must be contiguous, and remain in order so that the replacements can be  They must be contiguous, and remain in order so that the replacements can be
1862  looked up from a table.  looked up from a table.
1863    
1864  The final escape must be ESC_REF as subsequent values are used for  Negative numbers are used to encode a backreference (\1, \2, \3, etc.) in
1865  backreferences (\1, \2, \3, etc). There are two tests in the code for an escape  check_escape(). There are two tests in the code for an escape
1866  greater than ESC_b and less than ESC_Z to detect the types that may be  greater than ESC_b and less than ESC_Z to detect the types that may be
1867  repeated. These are the types that consume characters. If any new escapes are  repeated. These are the types that consume characters. If any new escapes are
1868  put in between that don't consume a character, that code will have to change.  put in between that don't consume a character, that code will have to change.
# Line 1692  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E Line 1872  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E
1872         ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,         ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1873         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1874         ESC_E, ESC_Q, ESC_g, ESC_k,         ESC_E, ESC_Q, ESC_g, ESC_k,
1875         ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu,         ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
        ESC_REF };  
1876    
1877  /* 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
1878  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.
# Line 2091  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2270  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2270         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2271         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2272         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2273         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERRCOUNT };         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2274    
2275  /* JIT compiling modes. The function list is indexed by them. */  /* JIT compiling modes. The function list is indexed by them. */
2276  enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,  enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
# Line 2114  fields are present. Currently PCRE alway Line 2293  fields are present. Currently PCRE alway
2293  NOTE NOTE NOTE  NOTE NOTE NOTE
2294  */  */
2295    
2296  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2297  #define REAL_PCRE real_pcre  #define REAL_PCRE real_pcre
2298  #else  #elif defined COMPILE_PCRE16
2299  #define REAL_PCRE real_pcre16  #define REAL_PCRE real_pcre16
2300    #elif defined COMPILE_PCRE32
2301    #define REAL_PCRE real_pcre32
2302  #endif  #endif
2303    
2304  typedef struct REAL_PCRE {  /* It is necessary to fork the struct for 32 bit, since it needs to use
2305     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2306     * typedef since pcretest needs access to  the struct of the 8-, 16-
2307     * and 32-bit variants. */
2308    
2309    typedef struct real_pcre8_or_16 {
2310    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2311    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2312    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
# Line 2136  typedef struct REAL_PCRE { Line 2322  typedef struct REAL_PCRE {
2322    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2323    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2324    const pcre_uint8 *nullpad;      /* NULL padding */    const pcre_uint8 *nullpad;      /* NULL padding */
2325  } REAL_PCRE;  } real_pcre8_or_16;
2326    
2327    typedef struct real_pcre8_or_16 real_pcre;
2328    typedef struct real_pcre8_or_16 real_pcre16;
2329    
2330    typedef struct real_pcre32 {
2331      pcre_uint32 magic_number;
2332      pcre_uint32 size;               /* Total that was malloced */
2333      pcre_uint32 options;            /* Public options */
2334      pcre_uint16 flags;              /* Private flags */
2335      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2336      pcre_uint16 top_bracket;        /* Highest numbered group */
2337      pcre_uint16 top_backref;        /* Highest numbered back reference */
2338      pcre_uint32 first_char;         /* Starting character */
2339      pcre_uint32 req_char;           /* This character must be seen */
2340      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2341      pcre_uint16 name_entry_size;    /* Size of any name items */
2342      pcre_uint16 name_count;         /* Number of name items */
2343      pcre_uint16 ref_count;          /* Reference count */
2344      pcre_uint16 dummy1;             /* for later expansion */
2345      pcre_uint16 dummy2;             /* for later expansion */
2346      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2347      void *nullpad;                  /* for later expansion */
2348    } real_pcre32;
2349    
2350    /* Assert that the size of REAL_PCRE is divisible by 8 */
2351    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2352    
2353    /* Needed in pcretest to access some fields in the real_pcre* structures
2354     * directly. They're unified for 8/16/32 bits since the structs only differ
2355     * after these fields; if that ever changes, need to fork those defines into
2356     * 8/16 and 32 bit versions. */
2357    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2358    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2359    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2360    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2361    
2362  /* The format of the block used to store data from pcre_study(). The same  /* The format of the block used to store data from pcre_study(). The same
2363  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 2177  typedef struct compile_data { Line 2398  typedef struct compile_data {
2398    int  names_found;                 /* Number of entries so far */    int  names_found;                 /* Number of entries so far */
2399    int  name_entry_size;             /* Size of each entry */    int  name_entry_size;             /* Size of each entry */
2400    int  workspace_size;              /* Size of workspace */    int  workspace_size;              /* Size of workspace */
2401    int  bracount;                    /* Count of capturing parens as we compile */    unsigned int  bracount;           /* Count of capturing parens as we compile */
2402    int  final_bracount;              /* Saved value after first pass */    int  final_bracount;              /* Saved value after first pass */
2403    int  max_lookbehind;              /* Maximum lookbehind (characters) */    int  max_lookbehind;              /* Maximum lookbehind (characters) */
2404    int  top_backref;                 /* Maximum back reference */    int  top_backref;                 /* Maximum back reference */
# Line 2207  call within the pattern; used by pcre_ex Line 2428  call within the pattern; used by pcre_ex
2428    
2429  typedef struct recursion_info {  typedef struct recursion_info {
2430    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2431    int group_num;                  /* Number of group that was called */    unsigned int group_num;         /* Number of group that was called */
2432    int *offset_save;               /* Pointer to start of saved offsets */    int *offset_save;               /* Pointer to start of saved offsets */
2433    int saved_max;                  /* Number of saved offsets */    int saved_max;                  /* Number of saved offsets */
2434      int saved_capture_last;         /* Last capture number */
2435    PCRE_PUCHAR subject_position;   /* Position at start of recursion */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
2436  } recursion_info;  } recursion_info;
2437    
# Line 2251  typedef struct match_data { Line 2473  typedef struct match_data {
2473    const  pcre_uint8 *lcc;         /* Points to lower casing table */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2474    const  pcre_uint8 *fcc;         /* Points to case-flipping table */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2475    const  pcre_uint8 *ctypes;      /* Points to table of type maps */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
   BOOL   offset_overflow;         /* Set if too many extractions */  
2476    BOOL   notbol;                  /* NOTBOL flag */    BOOL   notbol;                  /* NOTBOL flag */
2477    BOOL   noteol;                  /* NOTEOL flag */    BOOL   noteol;                  /* NOTEOL flag */
2478    BOOL   utf;                     /* UTF-8 / UTF-16 flag */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
# Line 2272  typedef struct match_data { Line 2493  typedef struct match_data {
2493    PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */    PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2494    int    partial;                 /* PARTIAL options */    int    partial;                 /* PARTIAL options */
2495    int    end_offset_top;          /* Highwater mark at end of match */    int    end_offset_top;          /* Highwater mark at end of match */
2496    int    capture_last;            /* Most recent capture number */    pcre_int32 capture_last;        /* Most recent capture number + overflow flag */
2497    int    start_offset;            /* The start offset value */    int    start_offset;            /* The start offset value */
2498    int    match_function_type;     /* Set for certain special calls of MATCH() */    int    match_function_type;     /* Set for certain special calls of MATCH() */
2499    eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */    eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
# Line 2341  total length. */ Line 2562  total length. */
2562    
2563  /* Internal function and data prefixes. */  /* Internal function and data prefixes. */
2564    
2565  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2566  #ifndef PUBL  #ifndef PUBL
2567  #define PUBL(name) pcre_##name  #define PUBL(name) pcre_##name
2568  #endif  #endif
2569  #ifndef PRIV  #ifndef PRIV
2570  #define PRIV(name) _pcre_##name  #define PRIV(name) _pcre_##name
2571  #endif  #endif
2572  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
 #ifdef COMPILE_PCRE16  
2573  #ifndef PUBL  #ifndef PUBL
2574  #define PUBL(name) pcre16_##name  #define PUBL(name) pcre16_##name
2575  #endif  #endif
2576  #ifndef PRIV  #ifndef PRIV
2577  #define PRIV(name) _pcre16_##name  #define PRIV(name) _pcre16_##name
2578  #endif  #endif
2579    #elif defined COMPILE_PCRE32
2580    #ifndef PUBL
2581    #define PUBL(name) pcre32_##name
2582    #endif
2583    #ifndef PRIV
2584    #define PRIV(name) _pcre32_##name
2585    #endif
2586  #else  #else
2587  #error Unsupported compiling mode  #error Unsupported compiling mode
2588  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
 #endif /* COMPILE_PCRE8 */  
2589    
2590  /* 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
2591  codes. Each entry used to point directly to a name, but to reduce the number of  codes. Each entry used to point directly to a name, but to reduce the number of
# Line 2402  one of the exported public functions. Th Line 2628  one of the exported public functions. Th
2628  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2629    
2630  /* String comparison functions. */  /* String comparison functions. */
2631  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2632    
2633  #define STRCMP_UC_UC(str1, str2) \  #define STRCMP_UC_UC(str1, str2) \
2634    strcmp((char *)(str1), (char *)(str2))    strcmp((char *)(str1), (char *)(str2))
# Line 2414  sense, but are not part of the PCRE publ Line 2640  sense, but are not part of the PCRE publ
2640    strncmp((char *)(str1), (str2), (num))    strncmp((char *)(str1), (str2), (num))
2641  #define STRLEN_UC(str) strlen((const char *)str)  #define STRLEN_UC(str) strlen((const char *)str)
2642    
2643  #else  #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2644    
2645  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2646                             const pcre_uchar *);                             const pcre_uchar *);
# Line 2436  extern unsigned int      PRIV(strlen_uc) Line 2662  extern unsigned int      PRIV(strlen_uc)
2662    PRIV(strncmp_uc_c8)((str1), (str2), (num))    PRIV(strncmp_uc_c8)((str1), (str2), (num))
2663  #define STRLEN_UC(str) PRIV(strlen_uc)(str)  #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2664    
2665  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE[8|16|32] */
2666    
2667    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2668    
2669    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2670    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2671    
2672    #elif defined COMPILE_PCRE32
2673    
2674    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2675                               const pcre_uchar *);
2676    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2677                               const char *);
2678    
2679    #define STRCMP_UC_UC_TEST(str1, str2) \
2680      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2681    #define STRCMP_UC_C8_TEST(str1, str2) \
2682      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2683    
2684    #endif /* COMPILE_PCRE[8|16|32] */
2685    
2686  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2687  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2688                             int *, BOOL);                             int *, BOOL);
2689  extern int               PRIV(ord2utf)(pcre_uint32, pcre_uchar *);  extern unsigned int      PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2690  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2691  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2692                             int *, BOOL);                             int *, BOOL);
2693  extern BOOL              PRIV(xclass)(int, const pcre_uchar *, BOOL);  extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2694    
2695  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2696  extern void              PRIV(jit_compile)(const REAL_PCRE *,  extern void              PRIV(jit_compile)(const REAL_PCRE *,
2697                             PUBL(extra) *, int);                             PUBL(extra) *, int);
2698  extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,  extern int               PRIV(jit_exec)(const PUBL(extra) *,
2699                             const pcre_uchar *, int, int, int, int *, int);                             const pcre_uchar *, int, int, int, int *, int);
2700  extern void              PRIV(jit_free)(void *);  extern void              PRIV(jit_free)(void *);
2701  extern int               PRIV(jit_get_size)(void *);  extern int               PRIV(jit_get_size)(void *);
# Line 2471  extern const pcre_uint32 PRIV(ucd_casele Line 2716  extern const pcre_uint32 PRIV(ucd_casele
2716  extern const ucd_record  PRIV(ucd_records)[];  extern const ucd_record  PRIV(ucd_records)[];
2717  extern const pcre_uint8  PRIV(ucd_stage1)[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2718  extern const pcre_uint16 PRIV(ucd_stage2)[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2719  extern const int         PRIV(ucp_gentype)[];  extern const pcre_uint32 PRIV(ucp_gentype)[];
2720  extern const pcre_uint32 PRIV(ucp_gbtable)[];  extern const pcre_uint32 PRIV(ucp_gbtable)[];
2721  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2722  extern const int         PRIV(ucp_typerange)[];  extern const int         PRIV(ucp_typerange)[];
# Line 2482  extern const int         PRIV(ucp_typera Line 2727  extern const int         PRIV(ucp_typera
2727    
2728  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2729  #define GET_UCD(ch) (PRIV(ucd_records) + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2730          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2731          UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2732    
2733  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2734  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2735  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2736  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2737  #define UCD_CASESET(ch)     GET_UCD(ch)->caseset  #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2738  #define UCD_OTHERCASE(ch)   (ch + GET_UCD(ch)->other_case)  #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2739    
2740  #endif /* SUPPORT_UCP */  #endif /* SUPPORT_UCP */
2741    

Legend:
Removed from v.1046  
changed lines
  Added in v.1248

  ViewVC Help
Powered by ViewVC 1.1.5