/[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 1015 by ph10, Sun Aug 26 16:07:14 2012 UTC revision 1083 by chpe, Tue Oct 16 15:55:24 2012 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  #if !defined COMPILE_PCRE16 && ! defined COMPILE_PCRE32
57  #define COMPILE_PCRE8  #define COMPILE_PCRE8
58  #endif  #endif
59    
# Line 78  Until then we define it if SUPPORT_UTF i Line 78  Until then we define it if SUPPORT_UTF i
78  #define SUPPORT_UTF8 1  #define SUPPORT_UTF8 1
79  #endif  #endif
80    
81  /* 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"
82  script prevents both being selected, but not everybody uses "configure". */  script prevents both being selected, but not everybody uses "configure". */
83    
84  #if defined EBCDIC && defined SUPPORT_UTF  #if defined EBCDIC && defined SUPPORT_UTF
85  #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.
86  #endif  #endif
87    
88  /* 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 243  exactly 256 items. When the character is Line 243  exactly 256 items. When the character is
243  items, some check is needed before accessing these tables.  items, some check is needed before accessing these tables.
244  */  */
245    
246  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
247    
248  typedef unsigned char pcre_uchar;  typedef unsigned char pcre_uchar;
249  #define IN_UCHARS(x) (x)  #define IN_UCHARS(x) (x)
250  #define MAX_255(c) 1  #define MAX_255(c) 1
251  #define TABLE_GET(c, table, default) ((table)[c])  #define TABLE_GET(c, table, default) ((table)[c])
252    
253  #else  #elif defined COMPILE_PCRE16
254    
 #ifdef COMPILE_PCRE16  
255  #if USHRT_MAX != 65535  #if USHRT_MAX != 65535
256  /* 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
257  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 259  pcre.h(.in) and disable (comment out) th
259  #endif  #endif
260    
261  typedef pcre_uint16 pcre_uchar;  typedef pcre_uint16 pcre_uchar;
262  #define IN_UCHARS(x) ((x) << 1)  #define UCHAR_SHIFT (1)
263    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
264  #define MAX_255(c) ((c) <= 255u)  #define MAX_255(c) ((c) <= 255u)
265  #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))  #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
266    
267    #elif defined COMPILE_PCRE32
268    
269    typedef pcre_uint32 pcre_uchar;
270    #define UCHAR_SHIFT (2)
271    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
272    #define MAX_255(c) ((c) <= 255u)
273    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
274    
275    /* Assert that pcre_uchar32 is a 32-bit type */
276    typedef int __assert_pcre_uchar32_size[sizeof(pcre_uchar) == 4 ? 1 : -1];
277    
278  #else  #else
279  #error Unsupported compiling mode  #error Unsupported compiling mode
280  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
281    
282  /* 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
283  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 396  The macros are controlled by the value o Line 405  The macros are controlled by the value o
405  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
406  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
407    
408  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
409    
410  #if LINK_SIZE == 2  #if LINK_SIZE == 2
411    
# Line 441  is automated on Unix systems via the "co Line 450  is automated on Unix systems via the "co
450  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
451  #endif  #endif
452    
453  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
   
 #ifdef COMPILE_PCRE16  
454    
455  #if LINK_SIZE == 2  #if LINK_SIZE == 2
456    
457    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
458  #undef LINK_SIZE  #undef LINK_SIZE
459  #define LINK_SIZE 1  #define LINK_SIZE 1
460    
# Line 460  is automated on Unix systems via the "co Line 468  is automated on Unix systems via the "co
468    
469  #elif LINK_SIZE == 3 || LINK_SIZE == 4  #elif LINK_SIZE == 3 || LINK_SIZE == 4
470    
471    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
472  #undef LINK_SIZE  #undef LINK_SIZE
473  #define LINK_SIZE 2  #define LINK_SIZE 2
474    
# Line 477  is automated on Unix systems via the "co Line 486  is automated on Unix systems via the "co
486  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
487  #endif  #endif
488    
489    #elif defined COMPILE_PCRE32
490    
491    /* Only supported LINK_SIZE is 4 */
492    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
493    #undef LINK_SIZE
494    #define LINK_SIZE 1
495    
496    #define PUT(a,n,d)   \
497      (a[n] = (d))
498    
499    #define GET(a,n) \
500      (a[n])
501    
502    /* Keep it positive */
503    #define MAX_PATTERN_SIZE (1 << 30)
504    
505  #else  #else
506  #error Unsupported compiling mode  #error Unsupported compiling mode
507  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
508    
509  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
510    
# Line 492  is automated on Unix systems via the "co Line 515  is automated on Unix systems via the "co
515  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
516  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
517    
518  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
519    
520  #define IMM2_SIZE 2  #define IMM2_SIZE 2
521    
# Line 503  capturing parenthesis numbers in back re Line 526  capturing parenthesis numbers in back re
526  #define GET2(a,n) \  #define GET2(a,n) \
527    (((a)[n] << 8) | (a)[(n)+1])    (((a)[n] << 8) | (a)[(n)+1])
528    
529  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
530    
531  #ifdef COMPILE_PCRE16  #define IMM2_SIZE 1
532    
533    #define PUT2(a,n,d)   \
534       a[n] = d
535    
536    #define GET2(a,n) \
537       a[n]
538    
539    #elif defined COMPILE_PCRE32
540    
541  #define IMM2_SIZE 1  #define IMM2_SIZE 1
542    
# Line 517  capturing parenthesis numbers in back re Line 548  capturing parenthesis numbers in back re
548    
549  #else  #else
550  #error Unsupported compiling mode  #error Unsupported compiling mode
551  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
   
 #endif /* COMPILE_PCRE8 */  
552    
553  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
554    
555  /* 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
556  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. */
557    
558  #define MAX_MARK ((1 << (sizeof(pcre_uchar)*8)) - 1)  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
559    #define MAX_MARK ((1u << 16) - 1)
560    #else
561    #define MAX_MARK ((1u << 8) - 1)
562    #endif
563    
564  /* 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
565  character. The macros for character handling generate simple sequences when  byte. The macros for character handling generate simple sequences when used in
566  used in character-mode, and more complicated ones for UTF characters.  character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
567  GETCHARLENTEST and other macros are not used when UTF is not supported,  and other macros are not used when UTF is not supported, so they are not
568  so they are not defined. To make sure they can never even appear when  defined. To make sure they can never even appear when UTF support is omitted,
569  UTF support is omitted, we don't even define them. */  we don't even define them. */
570    
571  #ifndef SUPPORT_UTF  #ifndef SUPPORT_UTF
572    
# Line 553  UTF support is omitted, we don't even de Line 586  UTF support is omitted, we don't even de
586    
587  #else   /* SUPPORT_UTF */  #else   /* SUPPORT_UTF */
588    
589  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
590    
591  /* These macros were originally written in the form of loops that used data  /* These macros were originally written in the form of loops that used data
592  from the tables whose names start with PRIV(utf8_table). They were rewritten by  from the tables whose names start with PRIV(utf8_table). They were rewritten by
# Line 727  because almost all calls are already wit Line 760  because almost all calls are already wit
760  #define ACROSSCHAR(condition, eptr, action) \  #define ACROSSCHAR(condition, eptr, action) \
761    while((condition) && ((eptr) & 0xc0) == 0x80) action    while((condition) && ((eptr) & 0xc0) == 0x80) action
762    
763  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
   
 #ifdef COMPILE_PCRE16  
764    
765  /* 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. */
766    
# Line 825  code. */ Line 856  code. */
856  #define ACROSSCHAR(condition, eptr, action) \  #define ACROSSCHAR(condition, eptr, action) \
857    if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action    if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
858    
859  #endif  #elif defined COMPILE_PCRE32
860    
861  #endif /* COMPILE_PCRE8 */  /* These are unnecessary for the 32-bit library */
862    #undef MAX_VALUE_FOR_SINGLE_CHAR
863    #undef HAS_EXTRALEN
864    #undef GET_EXTRALEN
865    #undef NOT_FIRSTCHAR
866    
867    #define UTF32_MASK (0x1ffffful)
868    
869    /* Get the next UTF-32 character, not advancing the pointer. This is called when
870    we know we are in UTF-32 mode. */
871    
872    #define GETCHAR(c, eptr) \
873      c = (*eptr) & UTF32_MASK;
874    
875    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
876    pointer. */
877    
878    #define GETCHARTEST(c, eptr) \
879      c = *eptr; \
880      if (utf) c &= UTF32_MASK;
881    
882    /* Get the next UTF-32 character, advancing the pointer. This is called when we
883    know we are in UTF-32 mode. */
884    
885    #define GETCHARINC(c, eptr) \
886      c = (*eptr++) & UTF32_MASK;
887    
888    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
889    This is called when we don't know if we are in UTF-32 mode. */
890    
891    #define GETCHARINCTEST(c, eptr) \
892      c = *eptr++; \
893      if (utf) c &= UTF32_MASK;
894    
895    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
896    length (since all UTF-32 is of length 1). This is called when we know we are in
897    UTF-32 mode. */
898    
899    #define GETCHARLEN(c, eptr, len) \
900      GETCHAR(c, eptr)
901    
902    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
903    pointer, not incrementing the length (since all UTF-32 is of length 1).
904    This is called when we do not know if we are in UTF-32 mode. */
905    
906    #define GETCHARLENTEST(c, eptr, len) \
907      GETCHARTEST(c, eptr)
908    
909    /* If the pointer is not at the start of a character, move it back until
910    it is. This is called only in UTF-32 mode - we don't put a test within the
911    macro because almost all calls are already within a block of UTF-32 only
912    code. */
913    
914    #define BACKCHAR(eptr) do { } while (0)
915    
916    /* Same as above, just in the other direction. */
917    #define FORWARDCHAR(eptr) do { } while (0)
918    
919    /* Same as above, but it allows a fully customizable form. */
920    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
921    
922    #else
923    #error Unsupported compiling mode
924    #endif /* COMPILE_PCRE[8|16|32] */
925    
926  #endif  /* SUPPORT_UTF */  #endif  /* SUPPORT_UTF */
927    
928    /* Tests for Unicode horizontal and vertical whitespace characters must check a
929    number of different values. Using a switch statement for this generates the
930    fastest code (no loop, no memory access), and there are several places in the
931    interpreter code where this happens. In order to ensure that all the case lists
932    remain in step, we use macros so that there is only one place where the lists
933    are defined.
934    
935    These values are also required as lists in pcre_compile.c when processing \h,
936    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
937    macros that define the values are here so that all the definitions are
938    together. The lists must be in ascending character order, terminated by
939    NOTACHAR (which is 0xffffffff).
940    
941    Any changes should ensure that the various macros are kept in step with each
942    other. NOTE: The values also appear in pcre_jit_compile.c. */
943    
944    /* ------ ASCII/Unicode environments ------ */
945    
946    #ifndef EBCDIC
947    
948    #define HSPACE_LIST \
949      CHAR_HT, CHAR_SPACE, 0xa0, \
950      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
951      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
952      NOTACHAR
953    
954    #define HSPACE_MULTIBYTE_CASES \
955      case 0x1680:  /* OGHAM SPACE MARK */ \
956      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
957      case 0x2000:  /* EN QUAD */ \
958      case 0x2001:  /* EM QUAD */ \
959      case 0x2002:  /* EN SPACE */ \
960      case 0x2003:  /* EM SPACE */ \
961      case 0x2004:  /* THREE-PER-EM SPACE */ \
962      case 0x2005:  /* FOUR-PER-EM SPACE */ \
963      case 0x2006:  /* SIX-PER-EM SPACE */ \
964      case 0x2007:  /* FIGURE SPACE */ \
965      case 0x2008:  /* PUNCTUATION SPACE */ \
966      case 0x2009:  /* THIN SPACE */ \
967      case 0x200A:  /* HAIR SPACE */ \
968      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
969      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
970      case 0x3000   /* IDEOGRAPHIC SPACE */
971    
972    #define HSPACE_BYTE_CASES \
973      case CHAR_HT: \
974      case CHAR_SPACE: \
975      case 0xa0     /* NBSP */
976    
977    #define HSPACE_CASES \
978      HSPACE_BYTE_CASES: \
979      HSPACE_MULTIBYTE_CASES
980    
981    #define VSPACE_LIST \
982      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
983    
984    #define VSPACE_MULTIBYTE_CASES \
985      case 0x2028:    /* LINE SEPARATOR */ \
986      case 0x2029     /* PARAGRAPH SEPARATOR */
987    
988    #define VSPACE_BYTE_CASES \
989      case CHAR_LF: \
990      case CHAR_VT: \
991      case CHAR_FF: \
992      case CHAR_CR: \
993      case CHAR_NEL
994    
995    #define VSPACE_CASES \
996      VSPACE_BYTE_CASES: \
997      VSPACE_MULTIBYTE_CASES
998    
999  /* In case there is no definition of offsetof() provided - though any proper  /* ------ EBCDIC environments ------ */
1000  Standard C system should have one. */  
1001    #else
1002    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1003    
1004    #define HSPACE_BYTE_CASES \
1005      case CHAR_HT: \
1006      case CHAR_SPACE
1007    
1008    #define HSPACE_CASES HSPACE_BYTE_CASES
1009    
1010    #ifdef EBCDIC_NL25
1011    #define VSPACE_LIST \
1012      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1013    #else
1014    #define VSPACE_LIST \
1015      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1016    #endif
1017    
1018    #define VSPACE_BYTE_CASES \
1019      case CHAR_LF: \
1020      case CHAR_VT: \
1021      case CHAR_FF: \
1022      case CHAR_CR: \
1023      case CHAR_NEL
1024    
1025    #define VSPACE_CASES VSPACE_BYTE_CASES
1026    #endif  /* EBCDIC */
1027    
1028    /* ------ End of whitespace macros ------ */
1029    
 #ifndef offsetof  
 #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
 #endif  
1030    
1031    
1032  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
# Line 846  are in a 16-bit flags word. From release Line 1035  are in a 16-bit flags word. From release
1035  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1036  compatibility. */  compatibility. */
1037    
1038  #ifdef COMPILE_PCRE8  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1039  #define PCRE_MODE          0x0001  /* compiled in 8 bit mode */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1040  #endif  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
 #ifdef COMPILE_PCRE16  
 #define PCRE_MODE          0x0002  /* compiled in 16 bit mode */  
 #endif  
1041  #define PCRE_FIRSTSET      0x0010  /* first_char is set */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1042  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1043  #define PCRE_REQCHSET      0x0040  /* req_byte is set */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
# Line 862  compatibility. */ Line 1048  compatibility. */
1048  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */  #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1049  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */  #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1050    
1051    #if defined COMPILE_PCRE8
1052    #define PCRE_MODE          PCRE_MODE8
1053    #elif defined COMPILE_PCRE16
1054    #define PCRE_MODE          PCRE_MODE16
1055    #elif defined COMPILE_PCRE32
1056    #define PCRE_MODE          PCRE_MODE32
1057    #endif
1058    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1059    
1060  /* Flags for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1061    
1062  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
# Line 893  time, run time, or study time, respectiv Line 1088  time, run time, or study time, respectiv
1088    
1089  #define PUBLIC_STUDY_OPTIONS \  #define PUBLIC_STUDY_OPTIONS \
1090     (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \     (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1091      PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE)      PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1092    
1093  /* Magic number to provide a small check against being handed junk. */  /* Magic number to provide a small check against being handed junk. */
1094    
# Line 904  in different endianness. */ Line 1099  in different endianness. */
1099    
1100  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
1101    
 /* Negative values for the firstchar and reqchar variables */  
   
 #define REQ_UNSET (-2)  
 #define REQ_NONE  (-1)  
   
1102  /* 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
1103  req_byte match. */  req_byte match. */
1104    
# Line 945  macros to give the functions distinct na Line 1135  macros to give the functions distinct na
1135  #ifndef SUPPORT_UTF  #ifndef SUPPORT_UTF
1136    
1137  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1138  so that PCRE works on both ASCII and EBCDIC platforms, in non-UTF-mode only. */  so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1139    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1140    characters, a common practice has been to use its NL (0x15) character as the
1141    line terminator in C-like processing environments. However, sometimes the LF
1142    (0x25) character is used instead, according to this Unicode document:
1143    
1144    http://unicode.org/standard/reports/tr13/tr13-5.html
1145    
1146    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1147    instead. Whichever is *not* chosen is defined as NEL.
1148    
1149    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1150    same code point. */
1151    
1152    #ifdef EBCDIC
1153    
1154    #ifndef EBCDIC_NL25
1155    #define CHAR_NL                     '\x15'
1156    #define CHAR_NEL                    '\x25'
1157    #define STR_NL                      "\x15"
1158    #define STR_NEL                     "\x25"
1159    #else
1160    #define CHAR_NL                     '\x25'
1161    #define CHAR_NEL                    '\x15'
1162    #define STR_NL                      "\x25"
1163    #define STR_NEL                     "\x15"
1164    #endif
1165    
1166    #define CHAR_LF                     CHAR_NL
1167    #define STR_LF                      STR_NL
1168    
1169    #define CHAR_ESC                    '\047'
1170    #define CHAR_DEL                    '\007'
1171    #define STR_ESC                     "\047"
1172    #define STR_DEL                     "\007"
1173    
1174    #else  /* Not EBCDIC */
1175    
1176    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1177    compatibility. NEL is the Unicode newline character; make sure it is
1178    a positive value. */
1179    
1180    #define CHAR_LF                     '\n'
1181    #define CHAR_NL                     CHAR_LF
1182    #define CHAR_NEL                    ((unsigned char)'\x85')
1183    #define CHAR_ESC                    '\033'
1184    #define CHAR_DEL                    '\177'
1185    
1186    #define STR_LF                      "\n"
1187    #define STR_NL                      STR_LF
1188    #define STR_NEL                     "\x85"
1189    #define STR_ESC                     "\033"
1190    #define STR_DEL                     "\177"
1191    
1192    #endif  /* EBCDIC */
1193    
1194    /* The remaining definitions work in both environments. */
1195    
1196  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1197  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1198  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1199  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1200  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1201  #define CHAR_BEL                    '\a'  #define CHAR_BEL                    '\a'
 #ifdef EBCDIC  
 #define CHAR_ESC                    '\047'  
 #define CHAR_DEL                    '\007'  
 #else  
 #define CHAR_ESC                    '\033'  
 #define CHAR_DEL                    '\177'  
 #endif  
1202    
1203  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1204  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 1062  so that PCRE works on both ASCII and EBC Line 1300  so that PCRE works on both ASCII and EBC
1300  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1301  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1302  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1303  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1304  #define STR_BEL                     "\a"  #define STR_BEL                     "\a"
 #ifdef EBCDIC  
 #define STR_ESC                     "\047"  
 #define STR_DEL                     "\007"  
 #else  
 #define STR_ESC                     "\033"  
 #define STR_DEL                     "\177"  
 #endif  
1305    
1306  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1307  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 1208  so that PCRE works on both ASCII and EBC Line 1438  so that PCRE works on both ASCII and EBC
1438  #ifdef COMPILE_PCRE16  #ifdef COMPILE_PCRE16
1439  #define STRING_UTF_RIGHTPAR            "UTF16)"  #define STRING_UTF_RIGHTPAR            "UTF16)"
1440  #endif  #endif
1441    #ifdef COMPILE_PCRE32
1442    #define STRING_UTF_RIGHTPAR            "UTF32)"
1443    #endif
1444  #define STRING_UCP_RIGHTPAR            "UCP)"  #define STRING_UCP_RIGHTPAR            "UCP)"
1445  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1446    
# Line 1221  only. */ Line 1454  only. */
1454  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1455  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1456  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1457  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1458    #define CHAR_NL                     CHAR_LF
1459    #define CHAR_NEL                    ((unsigned char)'\x85')
1460  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1461  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1462  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
# Line 1468  only. */ Line 1703  only. */
1703  #ifdef COMPILE_PCRE16  #ifdef COMPILE_PCRE16
1704  #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS  #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1705  #endif  #endif
1706    #ifdef COMPILE_PCRE32
1707    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1708    #endif
1709  #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
1710  #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
1711    
# Line 1484  only. */ Line 1722  only. */
1722  #endif  #endif
1723    
1724  #ifndef ESC_n  #ifndef ESC_n
1725  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1726  #endif  #endif
1727    
1728  #ifndef ESC_r  #ifndef ESC_r
# Line 1509  only. */ Line 1747  only. */
1747  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1748  #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */  #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1749  #define PT_WORD       8    /* Word - L plus N plus underscore */  #define PT_WORD       8    /* Word - L plus N plus underscore */
1750    #define PT_CLIST      9    /* Pseudo-property: match character list */
1751    
1752  /* Flag bits and data types for the extended class (OP_XCLASS) for classes that  /* Flag bits and data types for the extended class (OP_XCLASS) for classes that
1753  contain characters with values greater than 255. */  contain characters with values greater than 255. */
# Line 1524  contain characters with values greater t Line 1763  contain characters with values greater t
1763    
1764  /* 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
1765  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
1766  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
1767  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
1768  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
1769  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
1770  non-DOTALL mode, "." behaves like \N.  non-DOTALL mode, "." behaves like \N.
1771    
1772  The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.  The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1773  when PCRE_UCP is set, when replacement of \d etc by \p sequences is required.  when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1774  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
1775  looked up from a table.  looked up from a table.
1776    
1777  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
1778  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
1779  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
1780  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
1781  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 1546  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E Line 1785  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E
1785         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,
1786         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1787         ESC_E, ESC_Q, ESC_g, ESC_k,         ESC_E, ESC_Q, ESC_g, ESC_k,
1788         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 };  
1789    
1790  /* 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
1791  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 1573  enum { Line 1811  enum {
1811    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1812    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1813    
1814    OP_ANY,            /* 12 Match any character except newline */    OP_ANY,            /* 12 Match any character except newline (\N) */
1815    OP_ALLANY,         /* 13 Match any character */    OP_ALLANY,         /* 13 Match any character */
1816    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1817    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
# Line 1584  enum { Line 1822  enum {
1822    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1823    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1824    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1825    OP_EODN,           /* 23 End of data or \n at end of data: \Z. */    OP_EODN,           /* 23 End of data or \n at end of data (\Z) */
1826    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1827    
1828    OP_CIRC,           /* 25 Start of line - not multiline */    OP_CIRC,           /* 25 Start of line - not multiline */
1829    OP_CIRCM,          /* 26 Start of line - multiline */    OP_CIRCM,          /* 26 Start of line - multiline */
# Line 1945  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2183  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2183         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2184         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2185         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2186         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERRCOUNT };         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2187    
2188  /* JIT compiling modes. The function list is indexed by them. */  /* JIT compiling modes. The function list is indexed by them. */
2189  enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,  enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
# Line 1968  fields are present. Currently PCRE alway Line 2206  fields are present. Currently PCRE alway
2206  NOTE NOTE NOTE  NOTE NOTE NOTE
2207  */  */
2208    
2209  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2210  #define REAL_PCRE real_pcre  #define REAL_PCRE real_pcre
2211  #else  #elif defined COMPILE_PCRE16
2212  #define REAL_PCRE real_pcre16  #define REAL_PCRE real_pcre16
2213    #elif defined COMPILE_PCRE32
2214    #define REAL_PCRE real_pcre32
2215  #endif  #endif
2216    
2217  typedef struct REAL_PCRE {  /* It is necessary to fork the struct for 32 bit, since it needs to use
2218     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2219     * typedef since pcretest needs access to  the struct of the 8-, 16-
2220     * and 32-bit variants. */
2221    
2222    typedef struct real_pcre8_or_16 {
2223    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2224    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2225    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
# Line 1990  typedef struct REAL_PCRE { Line 2235  typedef struct REAL_PCRE {
2235    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2236    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */    const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2237    const pcre_uint8 *nullpad;      /* NULL padding */    const pcre_uint8 *nullpad;      /* NULL padding */
2238  } REAL_PCRE;  } real_pcre8_or_16;
2239    
2240    typedef struct real_pcre8_or_16 real_pcre;
2241    typedef struct real_pcre8_or_16 real_pcre16;
2242    
2243    typedef struct real_pcre32 {
2244      pcre_uint32 magic_number;
2245      pcre_uint32 size;               /* Total that was malloced */
2246      pcre_uint32 options;            /* Public options */
2247      pcre_uint16 flags;              /* Private flags */
2248      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2249      pcre_uint16 top_bracket;        /* Highest numbered group */
2250      pcre_uint16 top_backref;        /* Highest numbered back reference */
2251      pcre_uint32 first_char;         /* Starting character */
2252      pcre_uint32 req_char;           /* This character must be seen */
2253      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2254      pcre_uint16 name_entry_size;    /* Size of any name items */
2255      pcre_uint16 name_count;         /* Number of name items */
2256      pcre_uint16 ref_count;          /* Reference count */
2257      pcre_uint16 dummy1;             /* for later expansion */
2258      pcre_uint16 dummy2;             /* for later expansion */
2259      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2260      void *nullpad;                  /* for later expansion */
2261    } real_pcre32;
2262    
2263    /* Assert that the size of REAL_PCRE is divisible by 8 */
2264    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2265    
2266    /* Needed in pcretest to access some fields in the real_pcre* structures
2267     * directly. They're unified for 8/16/32 bits since the structs only differ
2268     * after these fields; if that ever changes, need to fork those defines into
2269     * 8/16 and 32 bit versions. */
2270    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2271    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2272    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2273    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2274    
2275  /* 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
2276  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 2041  typedef struct compile_data { Line 2321  typedef struct compile_data {
2321    int  external_flags;              /* External flag bits to be set */    int  external_flags;              /* External flag bits to be set */
2322    int  req_varyopt;                 /* "After variable item" flag for reqbyte */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2323    BOOL had_accept;                  /* (*ACCEPT) encountered */    BOOL had_accept;                  /* (*ACCEPT) encountered */
2324    BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */    BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2325    BOOL check_lookbehind;            /* Lookbehinds need later checking */    BOOL check_lookbehind;            /* Lookbehinds need later checking */
2326    int  nltype;                      /* Newline type */    int  nltype;                      /* Newline type */
2327    int  nllen;                       /* Newline string length */    int  nllen;                       /* Newline string length */
# Line 2195  total length. */ Line 2475  total length. */
2475    
2476  /* Internal function and data prefixes. */  /* Internal function and data prefixes. */
2477    
2478  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2479  #ifndef PUBL  #ifndef PUBL
2480  #define PUBL(name) pcre_##name  #define PUBL(name) pcre_##name
2481  #endif  #endif
2482  #ifndef PRIV  #ifndef PRIV
2483  #define PRIV(name) _pcre_##name  #define PRIV(name) _pcre_##name
2484  #endif  #endif
2485  #else /* COMPILE_PCRE8 */  #elif defined COMPILE_PCRE16
 #ifdef COMPILE_PCRE16  
2486  #ifndef PUBL  #ifndef PUBL
2487  #define PUBL(name) pcre16_##name  #define PUBL(name) pcre16_##name
2488  #endif  #endif
2489  #ifndef PRIV  #ifndef PRIV
2490  #define PRIV(name) _pcre16_##name  #define PRIV(name) _pcre16_##name
2491  #endif  #endif
2492    #elif defined COMPILE_PCRE32
2493    #ifndef PUBL
2494    #define PUBL(name) pcre32_##name
2495    #endif
2496    #ifndef PRIV
2497    #define PRIV(name) _pcre32_##name
2498    #endif
2499  #else  #else
2500  #error Unsupported compiling mode  #error Unsupported compiling mode
2501  #endif /* COMPILE_PCRE16 */  #endif /* COMPILE_PCRE[8|16|32] */
 #endif /* COMPILE_PCRE8 */  
2502    
2503  /* 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
2504  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 2233  but are not part of the PCRE public API. Line 2518  but are not part of the PCRE public API.
2518  pcre_tables.c module. */  pcre_tables.c module. */
2519    
2520  #ifdef COMPILE_PCRE8  #ifdef COMPILE_PCRE8
   
2521  extern const int            PRIV(utf8_table1)[];  extern const int            PRIV(utf8_table1)[];
2522  extern const int            PRIV(utf8_table1_size);  extern const int            PRIV(utf8_table1_size);
2523  extern const int            PRIV(utf8_table2)[];  extern const int            PRIV(utf8_table2)[];
2524  extern const int            PRIV(utf8_table3)[];  extern const int            PRIV(utf8_table3)[];
2525  extern const pcre_uint8     PRIV(utf8_table4)[];  extern const pcre_uint8     PRIV(utf8_table4)[];
   
2526  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE8 */
2527    
2528  extern const char           PRIV(utt_names)[];  extern const char           PRIV(utt_names)[];
2529  extern const ucp_type_table PRIV(utt)[];  extern const ucp_type_table PRIV(utt)[];
2530  extern const int            PRIV(utt_size);  extern const int            PRIV(utt_size);
2531    
2532    extern const pcre_uint8     PRIV(OP_lengths)[];
2533  extern const pcre_uint8     PRIV(default_tables)[];  extern const pcre_uint8     PRIV(default_tables)[];
2534    
2535  extern const pcre_uint8     PRIV(OP_lengths)[];  extern const pcre_uint32    PRIV(hspace_list)[];
2536    extern const pcre_uint32    PRIV(vspace_list)[];
2537    
2538    
2539  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
# Line 2256  one of the exported public functions. Th Line 2541  one of the exported public functions. Th
2541  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2542    
2543  /* String comparison functions. */  /* String comparison functions. */
2544  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
2545    
2546  #define STRCMP_UC_UC(str1, str2) \  #define STRCMP_UC_UC(str1, str2) \
2547    strcmp((char *)(str1), (char *)(str2))    strcmp((char *)(str1), (char *)(str2))
# Line 2268  sense, but are not part of the PCRE publ Line 2553  sense, but are not part of the PCRE publ
2553    strncmp((char *)(str1), (str2), (num))    strncmp((char *)(str1), (str2), (num))
2554  #define STRLEN_UC(str) strlen((const char *)str)  #define STRLEN_UC(str) strlen((const char *)str)
2555    
2556  #else  #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2557    
2558  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,  extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2559                             const pcre_uchar *);                             const pcre_uchar *);
# Line 2290  extern unsigned int      PRIV(strlen_uc) Line 2575  extern unsigned int      PRIV(strlen_uc)
2575    PRIV(strncmp_uc_c8)((str1), (str2), (num))    PRIV(strncmp_uc_c8)((str1), (str2), (num))
2576  #define STRLEN_UC(str) PRIV(strlen_uc)(str)  #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2577    
2578  #endif /* COMPILE_PCRE8 */  #endif /* COMPILE_PCRE[8|16|32] */
2579    
2580  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);  extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2581  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
# Line 2299  extern int               PRIV(ord2utf)(p Line 2584  extern int               PRIV(ord2utf)(p
2584  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);  extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2585  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,  extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2586                             int *, BOOL);                             int *, BOOL);
2587  extern BOOL              PRIV(xclass)(int, const pcre_uchar *, BOOL);  extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2588    
2589  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2590  extern void              PRIV(jit_compile)(const REAL_PCRE *,  extern void              PRIV(jit_compile)(const REAL_PCRE *,
# Line 2317  typedef struct { Line 2602  typedef struct {
2602    pcre_uint8 script;     /* ucp_Arabic, etc. */    pcre_uint8 script;     /* ucp_Arabic, etc. */
2603    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2604    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2605      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2606    pcre_int32 other_case; /* offset to other case, or zero if none */    pcre_int32 other_case; /* offset to other case, or zero if none */
2607  } ucd_record;  } ucd_record;
2608    
2609    extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2610  extern const ucd_record  PRIV(ucd_records)[];  extern const ucd_record  PRIV(ucd_records)[];
2611  extern const pcre_uint8  PRIV(ucd_stage1)[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2612  extern const pcre_uint16 PRIV(ucd_stage2)[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
# Line 2334  extern const int         PRIV(ucp_typera Line 2621  extern const int         PRIV(ucp_typera
2621    
2622  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2623  #define GET_UCD(ch) (PRIV(ucd_records) + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2624          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2625          UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2626    
2627  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2628  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2629  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2630  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2631  #define UCD_OTHERCASE(ch)   (ch + GET_UCD(ch)->other_case)  #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2632    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2633    
2634  #endif /* SUPPORT_UCP */  #endif /* SUPPORT_UCP */
2635    

Legend:
Removed from v.1015  
changed lines
  Added in v.1083

  ViewVC Help
Powered by ViewVC 1.1.5