/[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 758 by ph10, Mon Nov 21 12:05:36 2011 UTC revision 1160 by chpe, Mon Oct 22 22:05:03 2012 UTC
# Line 7  Line 7 
7  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2011 University of Cambridge             Copyright (c) 1997-2012 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 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_". */  functions whose names all begin with "_pcre_", "_pcre16_" or "_pcre32_"
44    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 51  functions whose names all begin with "_p Line 52  functions whose names all begin with "_p
52  #define PCRE_DEBUG  #define PCRE_DEBUG
53  #endif  #endif
54    
55  /* We do not support both EBCDIC and UTF-8 at the same time. The "configure"  /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
 script prevents both being selected, but not everybody uses "configure". */  
56    
57  #if defined EBCDIC && defined SUPPORT_UTF8  #if !defined COMPILE_PCRE16 && !defined COMPILE_PCRE32
58  #error The use of both EBCDIC and SUPPORT_UTF8 is not supported.  #define COMPILE_PCRE8
59  #endif  #endif
60    
61  /* If SUPPORT_UCP is defined, SUPPORT_UTF8 must also be defined. The  /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
62  "configure" script ensures this, but not everybody uses "configure". */  "configure" script ensures this, but not everybody uses "configure". */
63    
64  #if defined SUPPORT_UCP && !defined SUPPORT_UTF8  #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
65    #define SUPPORT_UTF 1
66    #endif
67    
68    /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
69    reasons with existing code. */
70    
71    #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
72    #define SUPPORT_UTF 1
73    #endif
74    
75    /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
76    Until then we define it if SUPPORT_UTF is defined. */
77    
78    #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
79  #define SUPPORT_UTF8 1  #define SUPPORT_UTF8 1
80  #endif  #endif
81    
82    /* 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". */
84    
85    #if defined EBCDIC && defined SUPPORT_UTF
86    #error The use of both EBCDIC and SUPPORT_UTF is not supported.
87    #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
90  inline, and there are *still* stupid compilers about that don't like indented  inline, and there are *still* stupid compilers about that don't like indented
91  pre-processor statements, or at least there were when I first wrote this. After  pre-processor statements, or at least there were when I first wrote this. After
# Line 91  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 158  set, we ensure here that it has no effec Line 185  set, we ensure here that it has no effec
185  #define PCRE_CALL_CONVENTION  #define PCRE_CALL_CONVENTION
186  #endif  #endif
187    
188  /* We need to have types that specify unsigned 16-bit and 32-bit integers. We  /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
189  cannot determine these outside the compilation (e.g. by running a program as  cannot determine these outside the compilation (e.g. by running a program as
190  part of "configure") because PCRE is often cross-compiled for use on other  part of "configure") because PCRE is often cross-compiled for use on other
191  systems. Instead we make use of the maximum sizes that are available at  systems. Instead we make use of the maximum sizes that are available at
192  preprocessor time in standard C environments. */  preprocessor time in standard C environments. */
193    
194    typedef unsigned char pcre_uint8;
195    
196  #if USHRT_MAX == 65535  #if USHRT_MAX == 65535
197    typedef unsigned short pcre_uint16;    typedef unsigned short pcre_uint16;
198    typedef short pcre_int16;    typedef short pcre_int16;
# Line 192  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 206  by "configure". */ Line 235  by "configure". */
235    
236  /* All character handling must be done as unsigned characters. Otherwise there  /* All character handling must be done as unsigned characters. Otherwise there
237  are problems with top-bit-set characters and functions such as isspace().  are problems with top-bit-set characters and functions such as isspace().
238  However, we leave the interface to the outside world as char *, because that  However, we leave the interface to the outside world as char * or short *,
239  should make things easier for callers. We define a short type for unsigned char  because that should make things easier for callers. This character type is
240  to save lots of typing. I tried "uchar", but it causes problems on Digital  called pcre_uchar.
241  Unix, where it is defined in sys/types, so use "uschar" instead. */  
242    The IN_UCHARS macro multiply its argument with the byte size of the current
243    pcre_uchar type. Useful for memcpy and such operations, whose require the
244    byte size of their input/output buffers.
245    
246    The MAX_255 macro checks whether its pcre_uchar input is less than 256.
247    
248    The TABLE_GET macro is designed for accessing elements of tables whose contain
249    exactly 256 items. When the character is able to contain more than 256
250    items, some check is needed before accessing these tables.
251    */
252    
253    #if defined COMPILE_PCRE8
254    
255    typedef unsigned char pcre_uchar;
256    #define IN_UCHARS(x) (x)
257    #define MAX_255(c) 1
258    #define TABLE_GET(c, table, default) ((table)[c])
259    
260    #elif defined COMPILE_PCRE16
261    
262    #if USHRT_MAX != 65535
263    /* 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. */
265    #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
266    #endif
267    
268    typedef pcre_uint16 pcre_uchar;
269    #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)
280    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
281    
282    /* Assert that pcre_uchar32 is a 32-bit type */
283    typedef int __assert_pcre_uchar32_size[sizeof(pcre_uchar) == 4 ? 1 : -1];
284    
285  typedef unsigned char uschar;  #else
286    #error Unsupported compiling mode
287    #endif /* COMPILE_PCRE[8|16|32] */
288    
289  /* 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
290  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 234  start/end of string field names are. */ Line 307  start/end of string field names are. */
307  #define IS_NEWLINE(p) \  #define IS_NEWLINE(p) \
308    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
309      ((p) < NLBLOCK->PSEND && \      ((p) < NLBLOCK->PSEND && \
310       _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\       PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
311         utf8)) \         &(NLBLOCK->nllen), utf)) \
312      : \      : \
313      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
314       (p)[0] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
315       (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
316      ) \      ) \
317    )    )
318    
# Line 248  start/end of string field names are. */ Line 321  start/end of string field names are. */
321  #define WAS_NEWLINE(p) \  #define WAS_NEWLINE(p) \
322    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
323      ((p) > NLBLOCK->PSSTART && \      ((p) > NLBLOCK->PSSTART && \
324       _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \       PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
325         &(NLBLOCK->nllen), utf8)) \         &(NLBLOCK->nllen), utf)) \
326      : \      : \
327      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
328       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
329       (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
330      ) \      ) \
331    )    )
332    
# Line 267  used for the external interface and appe Line 340  used for the external interface and appe
340  must begin with PCRE_. */  must begin with PCRE_. */
341    
342  #ifdef CUSTOM_SUBJECT_PTR  #ifdef CUSTOM_SUBJECT_PTR
343  #define PCRE_SPTR CUSTOM_SUBJECT_PTR  #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
 #define USPTR CUSTOM_SUBJECT_PTR  
344  #else  #else
345  #define PCRE_SPTR const char *  #define PCRE_PUCHAR const pcre_uchar *
 #define USPTR const unsigned char *  
346  #endif  #endif
347    
   
   
348  /* Include the public PCRE header and the definitions of UCP character property  /* Include the public PCRE header and the definitions of UCP character property
349  values. */  values. */
350    
# Line 343  The macros are controlled by the value o Line 412  The macros are controlled by the value o
412  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
413  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
414    
415    #if defined COMPILE_PCRE8
416    
417  #if LINK_SIZE == 2  #if LINK_SIZE == 2
418    
419  #define PUT(a,n,d)   \  #define PUT(a,n,d)   \
# Line 379  is automated on Unix systems via the "co Line 450  is automated on Unix systems via the "co
450  #define GET(a,n) \  #define GET(a,n) \
451    (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])    (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
452    
453  #define MAX_PATTERN_SIZE (1 << 30)   /* Keep it positive */  /* Keep it positive */
454    #define MAX_PATTERN_SIZE (1 << 30)
455    
456    #else
457    #error LINK_SIZE must be either 2, 3, or 4
458    #endif
459    
460    #elif defined COMPILE_PCRE16
461    
462    #if LINK_SIZE == 2
463    
464    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
465    #undef LINK_SIZE
466    #define LINK_SIZE 1
467    
468    #define PUT(a,n,d)   \
469      (a[n] = (d))
470    
471    #define GET(a,n) \
472      (a[n])
473    
474    #define MAX_PATTERN_SIZE (1 << 16)
475    
476    #elif LINK_SIZE == 3 || LINK_SIZE == 4
477    
478    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
479    #undef LINK_SIZE
480    #define LINK_SIZE 2
481    
482    #define PUT(a,n,d)   \
483      (a[n] = (d) >> 16), \
484      (a[(n)+1] = (d) & 65535)
485    
486    #define GET(a,n) \
487      (((a)[n] << 16) | (a)[(n)+1])
488    
489    /* Keep it positive */
490    #define MAX_PATTERN_SIZE (1 << 30)
491    
492  #else  #else
493  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
494  #endif  #endif
495    
496    #elif defined COMPILE_PCRE32
497    
498    /* Only supported LINK_SIZE is 4 */
499    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
500    #undef LINK_SIZE
501    #define LINK_SIZE 1
502    
503    #define PUT(a,n,d)   \
504      (a[n] = (d))
505    
506    #define GET(a,n) \
507      (a[n])
508    
509    /* Keep it positive */
510    #define MAX_PATTERN_SIZE (1 << 30)
511    
512    #else
513    #error Unsupported compiling mode
514    #endif /* COMPILE_PCRE[8|16|32] */
515    
516  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
517    
# Line 396  is automated on Unix systems via the "co Line 522  is automated on Unix systems via the "co
522  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
523  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
524    
525    #if defined COMPILE_PCRE8
526    
527    #define IMM2_SIZE 2
528    
529  #define PUT2(a,n,d)   \  #define PUT2(a,n,d)   \
530    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
531    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
532    
533    /* For reasons that I do not understand, the expression in this GET2 macro is
534    treated by gcc as a signed expression, even when a is declared as unsigned. It
535    seems that any kind of arithmetic results in a signed value. */
536    
537  #define GET2(a,n) \  #define GET2(a,n) \
538    (((a)[n] << 8) | (a)[(n)+1])    (unsigned int)(((a)[n] << 8) | (a)[(n)+1])
539    
540  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += 2  #elif defined COMPILE_PCRE16
541    
542    #define IMM2_SIZE 1
543    
544  /* When UTF-8 encoding is being used, a character is no longer just a single  #define PUT2(a,n,d)   \
545  byte. The macros for character handling generate simple sequences when used in     a[n] = d
546  byte-mode, and more complicated ones for UTF-8 characters. GETCHARLENTEST is  
547  not used when UTF-8 is not supported, so it is not defined, and BACKCHAR should  #define GET2(a,n) \
548  never be called in byte mode. To make sure they can never even appear when     a[n]
549  UTF-8 support is omitted, we don't even define them. */  
550    #elif defined COMPILE_PCRE32
551    
552    #define IMM2_SIZE 1
553    
554    #define PUT2(a,n,d)   \
555       a[n] = d
556    
557    #define GET2(a,n) \
558       a[n]
559    
560  #ifndef SUPPORT_UTF8  #else
561    #error Unsupported compiling mode
562    #endif /* COMPILE_PCRE[8|16|32] */
563    
564    #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
565    
566    /* The maximum length of a MARK name is currently one data unit; it may be
567    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
568    
569    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
570    #define MAX_MARK ((1u << 16) - 1)
571    #else
572    #define MAX_MARK ((1u << 8) - 1)
573    #endif
574    
575    /* When UTF encoding is being used, a character is no longer just a single
576    byte. The macros for character handling generate simple sequences when used in
577    character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
578    and other macros are not used when UTF is not supported, so they are not
579    defined. To make sure they can never even appear when UTF support is omitted,
580    we don't even define them. */
581    
582    #ifndef SUPPORT_UTF
583    
584    /* #define MAX_VALUE_FOR_SINGLE_CHAR */
585    /* #define HAS_EXTRALEN(c) */
586    /* #define GET_EXTRALEN(c) */
587    /* #define NOT_FIRSTCHAR(c) */
588  #define GETCHAR(c, eptr) c = *eptr;  #define GETCHAR(c, eptr) c = *eptr;
589  #define GETCHARTEST(c, eptr) c = *eptr;  #define GETCHARTEST(c, eptr) c = *eptr;
590  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
591  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
592  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
593    #define RAWUCHAR(eptr) (*(eptr))
594    #define RAWUCHARINC(eptr) (*(eptr)++)
595    #define RAWUCHARTEST(eptr) (*(eptr))
596    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
597  /* #define GETCHARLENTEST(c, eptr, len) */  /* #define GETCHARLENTEST(c, eptr, len) */
598  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
599    /* #define FORWARDCHAR(eptr) */
600    /* #define ACROSSCHAR(condition, eptr, action) */
601    
602  #else   /* SUPPORT_UTF8 */  #else   /* SUPPORT_UTF */
603    
604  /* These macros were originally written in the form of loops that used data  /* Tests whether the code point needs extra characters to decode. */
605  from the tables whose names start with _pcre_utf8_table. They were rewritten by  
606  a user so as not to use loops, because in some environments this gives a  #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
 significant performance advantage, and it seems never to do any harm. */  
607    
608  /* 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
609  advancing the pointer. */  advancing the pointer. */
# Line 451  advancing the pointer. */ Line 627  advancing the pointer. */
627            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
628      }      }
629    
 /* 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 (utf8 && c >= 0xc0) GETUTF8(c, eptr);  
   
630  /* 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
631  the pointer. */  the pointer. */
632    
# Line 499  the pointer. */ Line 661  the pointer. */
661        } \        } \
662      }      }
663    
664    #if defined COMPILE_PCRE8
665    
666    /* These macros were originally written in the form of loops that used data
667    from the tables whose names start with PRIV(utf8_table). They were rewritten by
668    a user so as not to use loops, because in some environments this gives a
669    significant performance advantage, and it seems never to do any harm. */
670    
671    /* Tells the biggest code point which can be encoded as a single character. */
672    
673    #define MAX_VALUE_FOR_SINGLE_CHAR 127
674    
675    /* Tests whether the code point needs extra characters to decode. */
676    
677    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
678    
679    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
680    Otherwise it has an undefined behaviour. */
681    
682    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
683    
684    /* Returns TRUE, if the given character is not the first character
685    of a UTF sequence. */
686    
687    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
688    
689    /* Get the next UTF-8 character, not advancing the pointer. This is called when
690    we know we are in UTF-8 mode. */
691    
692    #define GETCHAR(c, eptr) \
693      c = *eptr; \
694      if (c >= 0xc0) GETUTF8(c, eptr);
695    
696    /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
697    pointer. */
698    
699    #define GETCHARTEST(c, eptr) \
700      c = *eptr; \
701      if (utf && c >= 0xc0) GETUTF8(c, eptr);
702    
703  /* 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
704  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
705    
# Line 511  This is called when we don't know if we Line 712  This is called when we don't know if we
712    
713  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
714    c = *eptr++; \    c = *eptr++; \
715    if (utf8 && c >= 0xc0) GETUTF8INC(c, eptr);    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
716    
717  /* 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
718  advancing the pointer, incrementing the length. */  advancing the pointer, incrementing the length. */
# Line 563  do not know if we are in UTF-8 mode. */ Line 764  do not know if we are in UTF-8 mode. */
764    
765  #define GETCHARLENTEST(c, eptr, len) \  #define GETCHARLENTEST(c, eptr, len) \
766    c = *eptr; \    c = *eptr; \
767    if (utf8 && c >= 0xc0) GETUTF8LEN(c, eptr, len);    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
768    
769    /* Returns the next uchar, not advancing the pointer. This is called when
770    we know we are in UTF mode. */
771    
772    #define RAWUCHAR(eptr) \
773      (*(eptr))
774    
775    /* Returns the next uchar, advancing the pointer. This is called when
776    we know we are in UTF mode. */
777    
778    #define RAWUCHARINC(eptr) \
779      (*(eptr)++)
780    
781    /* Returns the next uchar, testing for UTF mode, and not advancing the
782    pointer. */
783    
784    #define RAWUCHARTEST(eptr) \
785      (*(eptr))
786    
787    /* Returns the next uchar, testing for UTF mode, advancing the
788    pointer. */
789    
790    #define RAWUCHARINCTEST(eptr) \
791      (*(eptr)++)
792    
793  /* 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
794  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
# Line 571  because almost all calls are already wit Line 796  because almost all calls are already wit
796    
797  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
798    
799  #endif  /* SUPPORT_UTF8 */  /* Same as above, just in the other direction. */
800    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
801    
802    /* Same as above, but it allows a fully customizable form. */
803    #define ACROSSCHAR(condition, eptr, action) \
804      while((condition) && ((eptr) & 0xc0) == 0x80) action
805    
806  /* In case there is no definition of offsetof() provided - though any proper  #elif defined COMPILE_PCRE16
807  Standard C system should have one. */  
808    /* Tells the biggest code point which can be encoded as a single character. */
809    
810    #define MAX_VALUE_FOR_SINGLE_CHAR 65535
811    
812    /* Tests whether the code point needs extra characters to decode. */
813    
814    #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
815    
816    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
817    Otherwise it has an undefined behaviour. */
818    
819    #define GET_EXTRALEN(c) 1
820    
821    /* Returns TRUE, if the given character is not the first character
822    of a UTF sequence. */
823    
824    #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
825    
826    /* Base macro to pick up the low surrogate of a UTF-16 character, not
827    advancing the pointer. */
828    
829    #define GETUTF16(c, eptr) \
830       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
831    
832    /* Get the next UTF-16 character, not advancing the pointer. This is called when
833    we know we are in UTF-16 mode. */
834    
835    #define GETCHAR(c, eptr) \
836      c = *eptr; \
837      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
838    
839    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
840    pointer. */
841    
842    #define GETCHARTEST(c, eptr) \
843      c = *eptr; \
844      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
845    
846    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
847    the pointer. */
848    
849    #define GETUTF16INC(c, eptr) \
850       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
851    
852    /* Get the next UTF-16 character, advancing the pointer. This is called when we
853    know we are in UTF-16 mode. */
854    
855    #define GETCHARINC(c, eptr) \
856      c = *eptr++; \
857      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
858    
859    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
860    This is called when we don't know if we are in UTF-16 mode. */
861    
862    #define GETCHARINCTEST(c, eptr) \
863      c = *eptr++; \
864      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
865    
866    /* Base macro to pick up the low surrogate of a UTF-16 character, not
867    advancing the pointer, incrementing the length. */
868    
869    #define GETUTF16LEN(c, eptr, len) \
870       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
871    
872    /* Get the next UTF-16 character, not advancing the pointer, incrementing
873    length if there is a low surrogate. This is called when we know we are in
874    UTF-16 mode. */
875    
876    #define GETCHARLEN(c, eptr, len) \
877      c = *eptr; \
878      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
879    
880    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
881    pointer, incrementing length if there is a low surrogate. This is called when
882    we do not know if we are in UTF-16 mode. */
883    
884    #define GETCHARLENTEST(c, eptr, len) \
885      c = *eptr; \
886      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
887    
888    /* Returns the next uchar, not advancing the pointer. This is called when
889    we know we are in UTF mode. */
890    
891    #define RAWUCHAR(eptr) \
892      (*(eptr))
893    
894    /* Returns the next uchar, advancing the pointer. This is called when
895    we know we are in UTF mode. */
896    
897    #define RAWUCHARINC(eptr) \
898      (*(eptr)++)
899    
900    /* Returns the next uchar, testing for UTF mode, and not advancing the
901    pointer. */
902    
903    #define RAWUCHARTEST(eptr) \
904      (*(eptr))
905    
906    /* Returns the next uchar, testing for UTF mode, advancing the
907    pointer. */
908    
909    #define RAWUCHARINCTEST(eptr) \
910      (*(eptr)++)
911    
912    /* If the pointer is not at the start of a character, move it back until
913    it is. This is called only in UTF-16 mode - we don't put a test within the
914    macro because almost all calls are already within a block of UTF-16 only
915    code. */
916    
917    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
918    
919    /* Same as above, just in the other direction. */
920    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
921    
922    /* Same as above, but it allows a fully customizable form. */
923    #define ACROSSCHAR(condition, eptr, action) \
924      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
925    
926    #elif defined COMPILE_PCRE32
927    
928    /* These are trivial for the 32-bit library, since all UTF-32 characters fit
929    into one pcre_uchar unit. */
930    #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
931    #define HAS_EXTRALEN(c) (0)
932    #define GET_EXTRALEN(c) (0)
933    #define NOT_FIRSTCHAR(c) (0)
934    
935    #define UTF32_MASK (0x1fffffu)
936    
937    /* 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) & UTF32_MASK;
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      if (utf) c &= UTF32_MASK;
949    
950    /* Get the next UTF-32 character, advancing the pointer. This is called when we
951    know we are in UTF-32 mode. */
952    
953    #define GETCHARINC(c, eptr) \
954      c = (*eptr++) & UTF32_MASK;
955    
956    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
957    This is called when we don't know if we are in UTF-32 mode. */
958    
959    #define GETCHARINCTEST(c, eptr) \
960      c = *eptr++; \
961      if (utf) c &= UTF32_MASK;
962    
963    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
964    length (since all UTF-32 is of length 1). This is called when we know we are in
965    UTF-32 mode. */
966    
967    #define GETCHARLEN(c, eptr, len) \
968      GETCHAR(c, eptr)
969    
970    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
971    pointer, not incrementing the length (since all UTF-32 is of length 1).
972    This is called when we do not know if we are in UTF-32 mode. */
973    
974    #define GETCHARLENTEST(c, eptr, len) \
975      GETCHARTEST(c, eptr)
976    
977    /* Returns the next uchar, not advancing the pointer. This is called when
978    we know we are in UTF mode. */
979    
980    #define RAWUCHAR(eptr) \
981      (*(eptr) & UTF32_MASK)
982    
983    /* Returns the next uchar, advancing the pointer. This is called when
984    we know we are in UTF mode. */
985    
986    #define RAWUCHARINC(eptr) \
987      (*(eptr)++ & UTF32_MASK)
988    
989    /* Returns the next uchar, testing for UTF mode, and not advancing the
990    pointer. */
991    
992    #define RAWUCHARTEST(eptr) \
993      (utf ? (*(eptr) & UTF32_MASK) : *(eptr))
994    
995    /* Returns the next uchar, testing for UTF mode, advancing the
996    pointer. */
997    
998    #define RAWUCHARINCTEST(eptr) \
999      (utf ? (*(eptr)++ & UTF32_MASK) : *(eptr)++)
1000    
1001    /* If the pointer is not at the start of a character, move it back until
1002    it is. This is called only in UTF-32 mode - we don't put a test within the
1003    macro because almost all calls are already within a block of UTF-32 only
1004    code.
1005    These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
1006    
1007    #define BACKCHAR(eptr) do { } while (0)
1008    
1009    /* Same as above, just in the other direction. */
1010    #define FORWARDCHAR(eptr) do { } while (0)
1011    
1012    /* Same as above, but it allows a fully customizable form. */
1013    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1014    
1015    #else
1016    #error Unsupported compiling mode
1017    #endif /* COMPILE_PCRE[8|16|32] */
1018    
1019    #endif  /* SUPPORT_UTF */
1020    
1021    /* Tests for Unicode horizontal and vertical whitespace characters must check a
1022    number of different values. Using a switch statement for this generates the
1023    fastest code (no loop, no memory access), and there are several places in the
1024    interpreter code where this happens. In order to ensure that all the case lists
1025    remain in step, we use macros so that there is only one place where the lists
1026    are defined.
1027    
1028    These values are also required as lists in pcre_compile.c when processing \h,
1029    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1030    macros that define the values are here so that all the definitions are
1031    together. The lists must be in ascending character order, terminated by
1032    NOTACHAR (which is 0xffffffff).
1033    
1034    Any changes should ensure that the various macros are kept in step with each
1035    other. NOTE: The values also appear in pcre_jit_compile.c. */
1036    
1037    /* ------ ASCII/Unicode environments ------ */
1038    
1039    #ifndef EBCDIC
1040    
1041    #define HSPACE_LIST \
1042      CHAR_HT, CHAR_SPACE, 0xa0, \
1043      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1044      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1045      NOTACHAR
1046    
1047    #define HSPACE_MULTIBYTE_CASES \
1048      case 0x1680:  /* OGHAM SPACE MARK */ \
1049      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1050      case 0x2000:  /* EN QUAD */ \
1051      case 0x2001:  /* EM QUAD */ \
1052      case 0x2002:  /* EN SPACE */ \
1053      case 0x2003:  /* EM SPACE */ \
1054      case 0x2004:  /* THREE-PER-EM SPACE */ \
1055      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1056      case 0x2006:  /* SIX-PER-EM SPACE */ \
1057      case 0x2007:  /* FIGURE SPACE */ \
1058      case 0x2008:  /* PUNCTUATION SPACE */ \
1059      case 0x2009:  /* THIN SPACE */ \
1060      case 0x200A:  /* HAIR SPACE */ \
1061      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1062      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1063      case 0x3000   /* IDEOGRAPHIC SPACE */
1064    
1065    #define HSPACE_BYTE_CASES \
1066      case CHAR_HT: \
1067      case CHAR_SPACE: \
1068      case 0xa0     /* NBSP */
1069    
1070    #define HSPACE_CASES \
1071      HSPACE_BYTE_CASES: \
1072      HSPACE_MULTIBYTE_CASES
1073    
1074    #define VSPACE_LIST \
1075      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1076    
1077    #define VSPACE_MULTIBYTE_CASES \
1078      case 0x2028:    /* LINE SEPARATOR */ \
1079      case 0x2029     /* PARAGRAPH SEPARATOR */
1080    
1081    #define VSPACE_BYTE_CASES \
1082      case CHAR_LF: \
1083      case CHAR_VT: \
1084      case CHAR_FF: \
1085      case CHAR_CR: \
1086      case CHAR_NEL
1087    
1088    #define VSPACE_CASES \
1089      VSPACE_BYTE_CASES: \
1090      VSPACE_MULTIBYTE_CASES
1091    
1092    /* ------ EBCDIC environments ------ */
1093    
1094    #else
1095    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1096    
1097    #define HSPACE_BYTE_CASES \
1098      case CHAR_HT: \
1099      case CHAR_SPACE
1100    
1101    #define HSPACE_CASES HSPACE_BYTE_CASES
1102    
1103    #ifdef EBCDIC_NL25
1104    #define VSPACE_LIST \
1105      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1106    #else
1107    #define VSPACE_LIST \
1108      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1109    #endif
1110    
1111    #define VSPACE_BYTE_CASES \
1112      case CHAR_LF: \
1113      case CHAR_VT: \
1114      case CHAR_FF: \
1115      case CHAR_CR: \
1116      case CHAR_NEL
1117    
1118    #define VSPACE_CASES VSPACE_BYTE_CASES
1119    #endif  /* EBCDIC */
1120    
1121    /* ------ End of whitespace macros ------ */
1122    
 #ifndef offsetof  
 #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
 #endif  
1123    
1124    
1125  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
# Line 588  are in a 16-bit flags word. From release Line 1128  are in a 16-bit flags word. From release
1128  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1129  compatibility. */  compatibility. */
1130    
1131  #define PCRE_NOPARTIAL     0x0001  /* can't use partial with this regex */  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1132  #define PCRE_FIRSTSET      0x0002  /* first_byte is set */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1133  #define PCRE_REQCHSET      0x0004  /* req_byte is set */  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
1134  #define PCRE_STARTLINE     0x0008  /* start after \n for multiline */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1135  #define PCRE_JCHANGED      0x0010  /* j option used in regex */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1136  #define PCRE_HASCRORLF     0x0020  /* explicit \r or \n in pattern */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
1137  #define PCRE_HASTHEN       0x0040  /* pattern contains (*THEN) */  #define PCRE_RCH_CASELESS  0x0080  /* caseless requested char */
1138    #define PCRE_STARTLINE     0x0100  /* start after \n for multiline */
1139    #define PCRE_NOPARTIAL     0x0200  /* can't use partial with this regex */
1140    #define PCRE_JCHANGED      0x0400  /* j option used in regex */
1141    #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1142    #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1143    
1144    #if defined COMPILE_PCRE8
1145    #define PCRE_MODE          PCRE_MODE8
1146    #elif defined COMPILE_PCRE16
1147    #define PCRE_MODE          PCRE_MODE16
1148    #elif defined COMPILE_PCRE32
1149    #define PCRE_MODE          PCRE_MODE32
1150    #endif
1151    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1152    
1153  /* Flags for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1154    
# Line 626  time, run time, or study time, respectiv Line 1180  time, run time, or study time, respectiv
1180     PCRE_NO_START_OPTIMIZE)     PCRE_NO_START_OPTIMIZE)
1181    
1182  #define PUBLIC_STUDY_OPTIONS \  #define PUBLIC_STUDY_OPTIONS \
1183     PCRE_STUDY_JIT_COMPILE     (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1184        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1185    
1186  /* Magic number to provide a small check against being handed junk. Also used  /* Magic number to provide a small check against being handed junk. */
 to detect whether a pattern was compiled on a host of different endianness. */  
1187    
1188  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1189    
1190  /* Negative values for the firstchar and reqchar variables */  /* This variable is used to detect a loaded regular expression
1191    in different endianness. */
1192    
1193  #define REQ_UNSET (-2)  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
 #define REQ_NONE  (-1)  
1194    
1195  /* 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
1196  req_byte match. */  req_byte match. */
1197    
1198  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1199    
 /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a  
 variable-length repeat, or a anything other than literal characters. */  
   
 #define REQ_CASELESS 0x0100    /* indicates caselessness */  
 #define REQ_VARY     0x0200    /* reqbyte followed non-literal item */  
   
1200  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1201  environments where these macros are defined elsewhere. Unfortunately, there  environments where these macros are defined elsewhere. Unfortunately, there
1202  is no way to do the same for the typedef. */  is no way to do the same for the typedef. */
# Line 677  for) in a minority area (EBCDIC platform Line 1225  for) in a minority area (EBCDIC platform
1225  application that did need both could compile two versions of the library, using  application that did need both could compile two versions of the library, using
1226  macros to give the functions distinct names. */  macros to give the functions distinct names. */
1227    
1228  #ifndef SUPPORT_UTF8  #ifndef SUPPORT_UTF
1229    
1230  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1231  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
1232    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1233    characters, a common practice has been to use its NL (0x15) character as the
1234    line terminator in C-like processing environments. However, sometimes the LF
1235    (0x25) character is used instead, according to this Unicode document:
1236    
1237    http://unicode.org/standard/reports/tr13/tr13-5.html
1238    
1239    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1240    instead. Whichever is *not* chosen is defined as NEL.
1241    
1242    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1243    same code point. */
1244    
1245    #ifdef EBCDIC
1246    
1247    #ifndef EBCDIC_NL25
1248    #define CHAR_NL                     '\x15'
1249    #define CHAR_NEL                    '\x25'
1250    #define STR_NL                      "\x15"
1251    #define STR_NEL                     "\x25"
1252    #else
1253    #define CHAR_NL                     '\x25'
1254    #define CHAR_NEL                    '\x15'
1255    #define STR_NL                      "\x25"
1256    #define STR_NEL                     "\x15"
1257    #endif
1258    
1259    #define CHAR_LF                     CHAR_NL
1260    #define STR_LF                      STR_NL
1261    
1262    #define CHAR_ESC                    '\047'
1263    #define CHAR_DEL                    '\007'
1264    #define STR_ESC                     "\047"
1265    #define STR_DEL                     "\007"
1266    
1267    #else  /* Not EBCDIC */
1268    
1269    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1270    compatibility. NEL is the Unicode newline character; make sure it is
1271    a positive value. */
1272    
1273    #define CHAR_LF                     '\n'
1274    #define CHAR_NL                     CHAR_LF
1275    #define CHAR_NEL                    ((unsigned char)'\x85')
1276    #define CHAR_ESC                    '\033'
1277    #define CHAR_DEL                    '\177'
1278    
1279    #define STR_LF                      "\n"
1280    #define STR_NL                      STR_LF
1281    #define STR_NEL                     "\x85"
1282    #define STR_ESC                     "\033"
1283    #define STR_DEL                     "\177"
1284    
1285    #endif  /* EBCDIC */
1286    
1287    /* The remaining definitions work in both environments. */
1288    
1289  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1290  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1291  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1292  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1293  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1294  #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  
1295    
1296  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1297  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 797  so that PCRE works on both ASCII and EBC Line 1393  so that PCRE works on both ASCII and EBC
1393  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1394  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1395  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1396  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1397  #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  
1398    
1399  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1400  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 937  so that PCRE works on both ASCII and EBC Line 1525  so that PCRE works on both ASCII and EBC
1525  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1526  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1527  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1528  #define STRING_UTF8_RIGHTPAR           "UTF8)"  #ifdef COMPILE_PCRE8
1529    #define STRING_UTF_RIGHTPAR            "UTF8)"
1530    #endif
1531    #ifdef COMPILE_PCRE16
1532    #define STRING_UTF_RIGHTPAR            "UTF16)"
1533    #endif
1534    #ifdef COMPILE_PCRE32
1535    #define STRING_UTF_RIGHTPAR            "UTF32)"
1536    #endif
1537  #define STRING_UCP_RIGHTPAR            "UCP)"  #define STRING_UCP_RIGHTPAR            "UCP)"
1538  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1539    
1540  #else  /* SUPPORT_UTF8 */  #else  /* SUPPORT_UTF */
1541    
1542  /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This  /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1543  works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode  works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
# Line 951  only. */ Line 1547  only. */
1547  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1548  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1549  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1550  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1551    #define CHAR_NL                     CHAR_LF
1552    #define CHAR_NEL                    ((unsigned char)'\x85')
1553  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1554  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1555  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
# Line 1192  only. */ Line 1790  only. */
1790  #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
1791  #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
1792  #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
1793  #define STRING_UTF8_RIGHTPAR           STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS  #ifdef COMPILE_PCRE8
1794    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1795    #endif
1796    #ifdef COMPILE_PCRE16
1797    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1798    #endif
1799    #ifdef COMPILE_PCRE32
1800    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1801    #endif
1802  #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
1803  #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
1804    
1805  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
1806    
1807  /* Escape items that are just an encoding of a particular data value. */  /* Escape items that are just an encoding of a particular data value. */
1808    
# Line 1209  only. */ Line 1815  only. */
1815  #endif  #endif
1816    
1817  #ifndef ESC_n  #ifndef ESC_n
1818  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1819  #endif  #endif
1820    
1821  #ifndef ESC_r  #ifndef ESC_r
# Line 1234  only. */ Line 1840  only. */
1840  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1841  #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 */
1842  #define PT_WORD       8    /* Word - L plus N plus underscore */  #define PT_WORD       8    /* Word - L plus N plus underscore */
1843    #define PT_CLIST      9    /* Pseudo-property: match character list */
1844    
1845  /* 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
1846  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1847    
1848  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1849  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */
# Line 1249  contain UTF-8 characters with values gre Line 1856  contain UTF-8 characters with values gre
1856    
1857  /* 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
1858  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
1859  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
1860  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
1861  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
1862  used for [^] in JavaScript compatibility mode, and for \C in non-utf8 mode. In  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1863  non-DOTALL mode, "." behaves like \N.  non-DOTALL mode, "." behaves like \N.
1864    
1865  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.
1866  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.
1867  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
1868  looked up from a table.  looked up from a table.
1869    
1870  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
1871  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
1872  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
1873  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
1874  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 1271  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E Line 1878  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E
1878         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,
1879         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1880         ESC_E, ESC_Q, ESC_g, ESC_k,         ESC_E, ESC_Q, ESC_g, ESC_k,
1881         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 };  
1882    
1883  /* 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
1884  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 1298  enum { Line 1904  enum {
1904    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1905    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1906    
1907    OP_ANY,            /* 12 Match any character except newline */    OP_ANY,            /* 12 Match any character except newline (\N) */
1908    OP_ALLANY,         /* 13 Match any character */    OP_ALLANY,         /* 13 Match any character */
1909    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 */
1910    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
# Line 1309  enum { Line 1915  enum {
1915    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1916    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1917    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1918    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) */
1919    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1920    
1921    OP_CIRC,           /* 25 Start of line - not multiline */    OP_CIRC,           /* 25 Start of line - not multiline */
1922    OP_CIRCM,          /* 26 Start of line - multiline */    OP_CIRCM,          /* 26 Start of line - multiline */
# Line 1433  enum { Line 2039  enum {
2039    OP_CLASS,          /* 106 Match a character class, chars < 256 only */    OP_CLASS,          /* 106 Match a character class, chars < 256 only */
2040    OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative    OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
2041                                class - the difference is relevant only when a                                class - the difference is relevant only when a
2042                                UTF-8 character > 255 is encountered. */                                character > 255 is encountered. */
2043    OP_XCLASS,         /* 108 Extended class for handling UTF-8 chars within the    OP_XCLASS,         /* 108 Extended class for handling > 255 chars within the
2044                                class. This does both positive and negative. */                                class. This does both positive and negative. */
2045    OP_REF,            /* 109 Match a back reference, casefully */    OP_REF,            /* 109 Match a back reference, casefully */
2046    OP_REFI,           /* 110 Match a back reference, caselessly */    OP_REFI,           /* 110 Match a back reference, caselessly */
# Line 1591  in UTF-8 mode. The code that uses this t Line 2197  in UTF-8 mode. The code that uses this t
2197    2,                             /* noti                                   */ \    2,                             /* noti                                   */ \
2198    /* Positive single-char repeats                             ** These are */ \    /* Positive single-char repeats                             ** These are */ \
2199    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2200    4, 4, 4,                       /* upto, minupto, exact      ** mode      */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2201    2, 2, 2, 4,                    /* *+, ++, ?+, upto+                      */ \    2+IMM2_SIZE,                   /* exact                                  */ \
2202      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2203    2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \    2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2204    4, 4, 4,                       /* upto I, minupto I, exact I             */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2205    2, 2, 2, 4,                    /* *+I, ++I, ?+I, upto+I                  */ \    2+IMM2_SIZE,                   /* exact I                                */ \
2206      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2207    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2208    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2209    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2210    2, 2, 2, 4,                    /* Possessive NOT *, +, ?, upto           */ \    2+IMM2_SIZE,                   /* NOT exact                              */ \
2211      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2212    2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \    2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2213    4, 4, 4,                       /* NOT upto I, minupto I, exact I         */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2214    2, 2, 2, 4,                    /* Possessive NOT *I, +I, ?I, upto I      */ \    2+IMM2_SIZE,                   /* NOT exact I                            */ \
2215      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2216    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2217    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2218    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2219    2, 2, 2, 4,                    /* Possessive *+, ++, ?+, upto+           */ \    2+IMM2_SIZE,                   /* Type exact                             */ \
2220      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2221    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2222    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2223    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2224   33,                             /* CLASS                                  */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2225   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2226    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2227    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2228    3,                             /* REFI                                   */ \    1+IMM2_SIZE,                   /* REFI                                   */ \
2229    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2230    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2231    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
# Line 1631  in UTF-8 mode. The code that uses this t Line 2242  in UTF-8 mode. The code that uses this t
2242    1+LINK_SIZE,                   /* ONCE_NC                                */ \    1+LINK_SIZE,                   /* ONCE_NC                                */ \
2243    1+LINK_SIZE,                   /* BRA                                    */ \    1+LINK_SIZE,                   /* BRA                                    */ \
2244    1+LINK_SIZE,                   /* BRAPOS                                 */ \    1+LINK_SIZE,                   /* BRAPOS                                 */ \
2245    3+LINK_SIZE,                   /* CBRA                                   */ \    1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2246    3+LINK_SIZE,                   /* CBRAPOS                                */ \    1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2247    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2248    1+LINK_SIZE,                   /* SBRA                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2249    1+LINK_SIZE,                   /* SBRAPOS                                */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2250    3+LINK_SIZE,                   /* SCBRA                                  */ \    1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2251    3+LINK_SIZE,                   /* SCBRAPOS                               */ \    1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2252    1+LINK_SIZE,                   /* SCOND                                  */ \    1+LINK_SIZE,                   /* SCOND                                  */ \
2253    3, 3,                          /* CREF, NCREF                            */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* CREF, NCREF                            */ \
2254    3, 3,                          /* RREF, NRREF                            */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* RREF, NRREF                            */ \
2255    1,                             /* DEF                                    */ \    1,                             /* DEF                                    */ \
2256    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2257    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2258    1, 3,                          /* SKIP, SKIP_ARG                         */ \    1, 3,                          /* SKIP, SKIP_ARG                         */ \
2259    1, 3,                          /* THEN, THEN_ARG                         */ \    1, 3,                          /* THEN, THEN_ARG                         */ \
2260    1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \    1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2261    3, 1                           /* CLOSE, SKIPZERO  */    1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2262    
2263  /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"  /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
2264  condition. */  condition. */
# Line 1665  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2276  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2276         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2277         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2278         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2279         ERR70, ERR71, ERRCOUNT };         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2280    
2281    /* JIT compiling modes. The function list is indexed by them. */
2282    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2283           JIT_NUMBER_OF_COMPILE_MODES };
2284    
2285  /* The real format of the start of the pcre block; the index of names and the  /* The real format of the start of the pcre block; the index of names and the
2286  code vector run on as long as necessary after the end. We store an explicit  code vector run on as long as necessary after the end. We store an explicit
# Line 1684  fields are present. Currently PCRE alway Line 2299  fields are present. Currently PCRE alway
2299  NOTE NOTE NOTE  NOTE NOTE NOTE
2300  */  */
2301    
2302  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2303    #define REAL_PCRE real_pcre
2304    #elif defined COMPILE_PCRE16
2305    #define REAL_PCRE real_pcre16
2306    #elif defined COMPILE_PCRE32
2307    #define REAL_PCRE real_pcre32
2308    #endif
2309    
2310    /* It is necessary to fork the struct for 32 bit, since it needs to use
2311     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2312     * typedef since pcretest needs access to  the struct of the 8-, 16-
2313     * and 32-bit variants. */
2314    
2315    typedef struct real_pcre8_or_16 {
2316    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2317    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2318    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
2319    pcre_uint16 flags;              /* Private flags */    pcre_uint16 flags;              /* Private flags */
2320    pcre_uint16 dummy1;             /* For future use */    pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2321    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2322    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2323    pcre_uint16 first_byte;    pcre_uint16 first_char;         /* Starting character */
2324    pcre_uint16 req_byte;    pcre_uint16 req_char;           /* This character must be seen */
2325    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2326    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2327    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2328    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2329      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2330      const pcre_uint8 *nullpad;      /* NULL padding */
2331    } real_pcre8_or_16;
2332    
2333    const unsigned char *tables;    /* Pointer to tables or NULL for std */  typedef struct real_pcre8_or_16 real_pcre;
2334    const unsigned char *nullpad;   /* NULL padding */  typedef struct real_pcre8_or_16 real_pcre16;
2335  } real_pcre;  
2336    typedef struct real_pcre32 {
2337      pcre_uint32 magic_number;
2338      pcre_uint32 size;               /* Total that was malloced */
2339      pcre_uint32 options;            /* Public options */
2340      pcre_uint16 flags;              /* Private flags */
2341      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2342      pcre_uint16 top_bracket;        /* Highest numbered group */
2343      pcre_uint16 top_backref;        /* Highest numbered back reference */
2344      pcre_uint32 first_char;         /* Starting character */
2345      pcre_uint32 req_char;           /* This character must be seen */
2346      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2347      pcre_uint16 name_entry_size;    /* Size of any name items */
2348      pcre_uint16 name_count;         /* Number of name items */
2349      pcre_uint16 ref_count;          /* Reference count */
2350      pcre_uint16 dummy1;             /* for later expansion */
2351      pcre_uint16 dummy2;             /* for later expansion */
2352      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2353      void *nullpad;                  /* for later expansion */
2354    } real_pcre32;
2355    
2356    /* Assert that the size of REAL_PCRE is divisible by 8 */
2357    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2358    
2359    /* Needed in pcretest to access some fields in the real_pcre* structures
2360     * directly. They're unified for 8/16/32 bits since the structs only differ
2361     * after these fields; if that ever changes, need to fork those defines into
2362     * 8/16 and 32 bit versions. */
2363    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2364    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2365    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2366    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2367    
2368  /* 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
2369  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 1709  remark (see NOTE above) about extending Line 2371  remark (see NOTE above) about extending
2371  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2372    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2373    pcre_uint32 flags;              /* Private flags */    pcre_uint32 flags;              /* Private flags */
2374    uschar start_bits[32];          /* Starting char bits */    pcre_uint8 start_bits[32];      /* Starting char bits */
2375    pcre_uint32 minlength;          /* Minimum subject length */    pcre_uint32 minlength;          /* Minimum subject length */
2376  } pcre_study_data;  } pcre_study_data;
2377    
# Line 1728  typedef struct open_capitem { Line 2390  typedef struct open_capitem {
2390  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2391    
2392  typedef struct compile_data {  typedef struct compile_data {
2393    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2394    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2395    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2396    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2397    const uschar *start_workspace;/* The start of working space */    const pcre_uchar *start_workspace;/* The start of working space */
2398    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_code;     /* The start of the compiled code */
2399    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2400    const uschar *end_pattern;    /* The end of the pattern */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2401    open_capitem *open_caps;      /* Chain of open capture items */    open_capitem *open_caps;          /* Chain of open capture items */
2402    uschar *hwm;                  /* High watermark of workspace */    pcre_uchar *hwm;                  /* High watermark of workspace */
2403    uschar *name_table;           /* The name/number table */    pcre_uchar *name_table;           /* The name/number table */
2404    int  names_found;             /* Number of entries so far */    int  names_found;                 /* Number of entries so far */
2405    int  name_entry_size;         /* Size of each entry */    int  name_entry_size;             /* Size of each entry */
2406    int  bracount;                /* Count of capturing parens as we compile */    int  workspace_size;              /* Size of workspace */
2407    int  final_bracount;          /* Saved value after first pass */    unsigned int  bracount;           /* Count of capturing parens as we compile */
2408    int  top_backref;             /* Maximum back reference */    int  final_bracount;              /* Saved value after first pass */
2409    unsigned int backref_map;     /* Bitmap of low back refs */    int  max_lookbehind;              /* Maximum lookbehind (characters) */
2410    int  assert_depth;            /* Depth of nested assertions */    int  top_backref;                 /* Maximum back reference */
2411    int  external_options;        /* External (initial) options */    unsigned int backref_map;         /* Bitmap of low back refs */
2412    int  external_flags;          /* External flag bits to be set */    int  assert_depth;                /* Depth of nested assertions */
2413    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  external_options;            /* External (initial) options */
2414    BOOL had_accept;              /* (*ACCEPT) encountered */    int  external_flags;              /* External flag bits to be set */
2415    BOOL check_lookbehind;        /* Lookbehinds need later checking */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2416    int  nltype;                  /* Newline type */    BOOL had_accept;                  /* (*ACCEPT) encountered */
2417    int  nllen;                   /* Newline string length */    BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2418    uschar nl[4];                 /* Newline string when fixed length */    BOOL check_lookbehind;            /* Lookbehinds need later checking */
2419      int  nltype;                      /* Newline type */
2420      int  nllen;                       /* Newline string length */
2421      pcre_uchar nl[4];                 /* Newline string when fixed length */
2422  } compile_data;  } compile_data;
2423    
2424  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
# Line 1761  branches, for testing for left recursion Line 2426  branches, for testing for left recursion
2426    
2427  typedef struct branch_chain {  typedef struct branch_chain {
2428    struct branch_chain *outer;    struct branch_chain *outer;
2429    uschar *current_branch;    pcre_uchar *current_branch;
2430  } branch_chain;  } branch_chain;
2431    
2432  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
# Line 1769  call within the pattern; used by pcre_ex Line 2434  call within the pattern; used by pcre_ex
2434    
2435  typedef struct recursion_info {  typedef struct recursion_info {
2436    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2437    int group_num;                  /* Number of group that was called */    unsigned int group_num;         /* Number of group that was called */
2438    int *offset_save;               /* Pointer to start of saved offsets */    int *offset_save;               /* Pointer to start of saved offsets */
2439    int saved_max;                  /* Number of saved offsets */    int saved_max;                  /* Number of saved offsets */
2440    USPTR subject_position;         /* Position at start of recursion */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
2441  } recursion_info;  } recursion_info;
2442    
2443  /* A similar structure for pcre_dfa_exec(). */  /* A similar structure for pcre_dfa_exec(). */
# Line 1780  typedef struct recursion_info { Line 2445  typedef struct recursion_info {
2445  typedef struct dfa_recursion_info {  typedef struct dfa_recursion_info {
2446    struct dfa_recursion_info *prevrec;    struct dfa_recursion_info *prevrec;
2447    int group_num;    int group_num;
2448    USPTR subject_position;    PCRE_PUCHAR subject_position;
2449  } dfa_recursion_info;  } dfa_recursion_info;
2450    
2451  /* Structure for building a chain of data for holding the values of the subject  /* Structure for building a chain of data for holding the values of the subject
# Line 1790  pcre_exec(). */ Line 2455  pcre_exec(). */
2455    
2456  typedef struct eptrblock {  typedef struct eptrblock {
2457    struct eptrblock *epb_prev;    struct eptrblock *epb_prev;
2458    USPTR epb_saved_eptr;    PCRE_PUCHAR epb_saved_eptr;
2459  } eptrblock;  } eptrblock;
2460    
2461    
# Line 1801  typedef struct match_data { Line 2466  typedef struct match_data {
2466    unsigned long int match_call_count;      /* As it says */    unsigned long int match_call_count;      /* As it says */
2467    unsigned long int match_limit;           /* As it says */    unsigned long int match_limit;           /* As it says */
2468    unsigned long int match_limit_recursion; /* As it says */    unsigned long int match_limit_recursion; /* As it says */
2469    int   *offset_vector;         /* Offset vector */    int   *offset_vector;           /* Offset vector */
2470    int    offset_end;            /* One past the end */    int    offset_end;              /* One past the end */
2471    int    offset_max;            /* The maximum usable for return data */    int    offset_max;              /* The maximum usable for return data */
2472    int    nltype;                /* Newline type */    int    nltype;                  /* Newline type */
2473    int    nllen;                 /* Newline string length */    int    nllen;                   /* Newline string length */
2474    int    name_count;            /* Number of names in name table */    int    name_count;              /* Number of names in name table */
2475    int    name_entry_size;       /* Size of entry in names table */    int    name_entry_size;         /* Size of entry in names table */
2476    uschar *name_table;           /* Table of names */    pcre_uchar *name_table;         /* Table of names */
2477    uschar nl[4];                 /* Newline string when fixed */    pcre_uchar nl[4];               /* Newline string when fixed */
2478    const  uschar *lcc;           /* Points to lower casing table */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2479    const  uschar *ctypes;        /* Points to table of type maps */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2480    BOOL   offset_overflow;       /* Set if too many extractions */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2481    BOOL   notbol;                /* NOTBOL flag */    BOOL   offset_overflow;         /* Set if too many extractions */
2482    BOOL   noteol;                /* NOTEOL flag */    BOOL   notbol;                  /* NOTBOL flag */
2483    BOOL   utf8;                  /* UTF8 flag */    BOOL   noteol;                  /* NOTEOL flag */
2484    BOOL   jscript_compat;        /* JAVASCRIPT_COMPAT flag */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2485    BOOL   use_ucp;               /* PCRE_UCP flag */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2486    BOOL   endonly;               /* Dollar not before final \n */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2487    BOOL   notempty;              /* Empty string match not wanted */    BOOL   endonly;                 /* Dollar not before final \n */
2488    BOOL   notempty_atstart;      /* Empty string match at start not wanted */    BOOL   notempty;                /* Empty string match not wanted */
2489    BOOL   hitend;                /* Hit the end of the subject at some point */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2490    BOOL   bsr_anycrlf;           /* \R is just any CRLF, not full Unicode */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2491    BOOL   hasthen;               /* Pattern contains (*THEN) */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2492    const  uschar *start_code;    /* For use when recursing */    BOOL   hasthen;                 /* Pattern contains (*THEN) */
2493    USPTR  start_subject;         /* Start of the subject string */    BOOL   ignore_skip_arg;         /* For re-run when SKIP name not found */
2494    USPTR  end_subject;           /* End of the subject string */    const  pcre_uchar *start_code;  /* For use when recursing */
2495    USPTR  start_match_ptr;       /* Start of matched string */    PCRE_PUCHAR start_subject;      /* Start of the subject string */
2496    USPTR  end_match_ptr;         /* Subject position at end match */    PCRE_PUCHAR end_subject;        /* End of the subject string */
2497    USPTR  start_used_ptr;        /* Earliest consulted character */    PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2498    int    partial;               /* PARTIAL options */    PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2499    int    end_offset_top;        /* Highwater mark at end of match */    PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2500    int    capture_last;          /* Most recent capture number */    int    partial;                 /* PARTIAL options */
2501    int    start_offset;          /* The start offset value */    int    end_offset_top;          /* Highwater mark at end of match */
2502    int    match_function_type;   /* Set for certain special calls of MATCH() */    int    capture_last;            /* Most recent capture number */
2503    eptrblock *eptrchain;         /* Chain of eptrblocks for tail recursions */    int    start_offset;            /* The start offset value */
2504    int    eptrn;                 /* Next free eptrblock */    int    match_function_type;     /* Set for certain special calls of MATCH() */
2505    recursion_info *recursive;    /* Linked list of recursion data */    eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2506    void  *callout_data;          /* To pass back to callouts */    int    eptrn;                   /* Next free eptrblock */
2507    const  uschar *mark;          /* Mark pointer to pass back */    recursion_info *recursive;      /* Linked list of recursion data */
2508    const  uschar *once_target;   /* Where to back up to for atomic groups */    void  *callout_data;            /* To pass back to callouts */
2509      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2510      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2511      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2512    #ifdef NO_RECURSE
2513      void  *match_frames_base;       /* For remembering malloc'd frames */
2514    #endif
2515  } match_data;  } match_data;
2516    
2517  /* A similar structure is used for the same purpose by the DFA matching  /* A similar structure is used for the same purpose by the DFA matching
2518  functions. */  functions. */
2519    
2520  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2521    const uschar *start_code;      /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2522    const uschar *start_subject;   /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2523    const uschar *end_subject;     /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2524    const uschar *start_used_ptr;  /* Earliest consulted character */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2525    const uschar *tables;          /* Character tables */    const pcre_uint8 *tables;         /* Character tables */
2526    int   start_offset;            /* The start offset value */    int   start_offset;               /* The start offset value */
2527    int   moptions;                /* Match options */    int   moptions;                   /* Match options */
2528    int   poptions;                /* Pattern options */    int   poptions;                   /* Pattern options */
2529    int    nltype;                 /* Newline type */    int   nltype;                     /* Newline type */
2530    int    nllen;                  /* Newline string length */    int   nllen;                      /* Newline string length */
2531    uschar nl[4];                  /* Newline string when fixed */    pcre_uchar nl[4];                 /* Newline string when fixed */
2532    void  *callout_data;           /* To pass back to callouts */    void *callout_data;               /* To pass back to callouts */
2533    dfa_recursion_info *recursive; /* Linked list of recursion data */    dfa_recursion_info *recursive;    /* Linked list of recursion data */
2534  } dfa_match_data;  } dfa_match_data;
2535    
2536  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 1895  total length. */ Line 2566  total length. */
2566  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2567  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2568    
2569    /* Internal function and data prefixes. */
2570    
2571    #if defined COMPILE_PCRE8
2572    #ifndef PUBL
2573    #define PUBL(name) pcre_##name
2574    #endif
2575    #ifndef PRIV
2576    #define PRIV(name) _pcre_##name
2577    #endif
2578    #elif defined COMPILE_PCRE16
2579    #ifndef PUBL
2580    #define PUBL(name) pcre16_##name
2581    #endif
2582    #ifndef PRIV
2583    #define PRIV(name) _pcre16_##name
2584    #endif
2585    #elif defined COMPILE_PCRE32
2586    #ifndef PUBL
2587    #define PUBL(name) pcre32_##name
2588    #endif
2589    #ifndef PRIV
2590    #define PRIV(name) _pcre32_##name
2591    #endif
2592    #else
2593    #error Unsupported compiling mode
2594    #endif /* COMPILE_PCRE[8|16|32] */
2595    
2596  /* 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
2597  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
2598  relocations in shared libraries, it now has an offset into a single string  relocations in shared libraries, it now has an offset into a single string
# Line 1912  of the exported public functions. They h Line 2610  of the exported public functions. They h
2610  but are not part of the PCRE public API. The data for these tables is in the  but are not part of the PCRE public API. The data for these tables is in the
2611  pcre_tables.c module. */  pcre_tables.c module. */
2612    
2613  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2614  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2615  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2616  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2617    extern const int            PRIV(utf8_table3)[];
2618  #ifdef SUPPORT_JIT  extern const pcre_uint8     PRIV(utf8_table4)[];
2619  extern const uschar _pcre_utf8_char_sizes[];  #endif /* COMPILE_PCRE8 */
2620  #endif  
2621    extern const char           PRIV(utt_names)[];
2622    extern const ucp_type_table PRIV(utt)[];
2623    extern const int            PRIV(utt_size);
2624    
2625  extern const int    _pcre_utf8_table1_size;  extern const pcre_uint8     PRIV(OP_lengths)[];
2626    extern const pcre_uint8     PRIV(default_tables)[];
2627    
2628  extern const char   _pcre_utt_names[];  extern const pcre_uint32    PRIV(hspace_list)[];
2629  extern const ucp_type_table _pcre_utt[];  extern const pcre_uint32    PRIV(vspace_list)[];
 extern const int _pcre_utt_size;  
   
 extern const uschar _pcre_default_tables[];  
   
 extern const uschar _pcre_OP_lengths[];  
2630    
2631    
2632  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2633  one of the exported public functions. They have to be "external" in the C  one of the exported public functions. They have to be "external" in the C
2634  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2635    
2636  extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);  /* String comparison functions. */
2637  extern BOOL          _pcre_is_newline(USPTR, int, USPTR, int *, BOOL);  #if defined COMPILE_PCRE8
2638  extern int           _pcre_ord2utf8(int, uschar *);  
2639  extern real_pcre    *_pcre_try_flipped(const real_pcre *, real_pcre *,  #define STRCMP_UC_UC(str1, str2) \
2640                         const pcre_study_data *, pcre_study_data *);    strcmp((char *)(str1), (char *)(str2))
2641  extern int           _pcre_valid_utf8(USPTR, int, int *);  #define STRCMP_UC_C8(str1, str2) \
2642  extern BOOL          _pcre_was_newline(USPTR, int, USPTR, int *, BOOL);    strcmp((char *)(str1), (str2))
2643  extern BOOL          _pcre_xclass(int, const uschar *);  #define STRNCMP_UC_UC(str1, str2, num) \
2644      strncmp((char *)(str1), (char *)(str2), (num))
2645    #define STRNCMP_UC_C8(str1, str2, num) \
2646      strncmp((char *)(str1), (str2), (num))
2647    #define STRLEN_UC(str) strlen((const char *)str)
2648    
2649    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2650    
2651    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2652                               const pcre_uchar *);
2653    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2654                               const char *);
2655    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2656                               const pcre_uchar *, unsigned int num);
2657    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2658                               const char *, unsigned int num);
2659    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2660    
2661    #define STRCMP_UC_UC(str1, str2) \
2662      PRIV(strcmp_uc_uc)((str1), (str2))
2663    #define STRCMP_UC_C8(str1, str2) \
2664      PRIV(strcmp_uc_c8)((str1), (str2))
2665    #define STRNCMP_UC_UC(str1, str2, num) \
2666      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2667    #define STRNCMP_UC_C8(str1, str2, num) \
2668      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2669    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2670    
2671    #endif /* COMPILE_PCRE[8|16|32] */
2672    
2673    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2674    
2675    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2676    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2677    
2678    #elif defined COMPILE_PCRE32
2679    
2680    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2681                               const pcre_uchar *);
2682    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2683                               const char *);
2684    
2685    #define STRCMP_UC_UC_TEST(str1, str2) \
2686      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2687    #define STRCMP_UC_C8_TEST(str1, str2) \
2688      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2689    
2690    #endif /* COMPILE_PCRE[8|16|32] */
2691    
2692    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2693    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2694                               int *, BOOL);
2695    extern unsigned int      PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2696    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2697    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2698                               int *, BOOL);
2699    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2700    
2701  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2702  extern void          _pcre_jit_compile(const real_pcre *, pcre_extra *);  extern void              PRIV(jit_compile)(const REAL_PCRE *,
2703  extern int           _pcre_jit_exec(const real_pcre *, void *, PCRE_SPTR,                             PUBL(extra) *, int);
2704                          int, int, int, int, int *, int);  extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2705  extern void          _pcre_jit_free(void *);                             const pcre_uchar *, int, int, int, int *, int);
2706    extern void              PRIV(jit_free)(void *);
2707    extern int               PRIV(jit_get_size)(void *);
2708    extern const char*       PRIV(jit_get_target)(void);
2709  #endif  #endif
2710    
2711  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
2712    
2713  typedef struct {  typedef struct {
2714    uschar script;    pcre_uint8 script;     /* ucp_Arabic, etc. */
2715    uschar chartype;    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2716    pcre_int32 other_case;    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2717      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2718      pcre_int32 other_case; /* offset to other case, or zero if none */
2719  } ucd_record;  } ucd_record;
2720    
2721  extern const ucd_record  _pcre_ucd_records[];  extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2722  extern const uschar      _pcre_ucd_stage1[];  extern const ucd_record  PRIV(ucd_records)[];
2723  extern const pcre_uint16 _pcre_ucd_stage2[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2724  extern const int         _pcre_ucp_gentype[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2725    extern const pcre_uint32 PRIV(ucp_gentype)[];
2726    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2727  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2728  extern const int         _pcre_ucp_typerange[];  extern const int         PRIV(ucp_typerange)[];
2729  #endif  #endif
2730    
2731    #ifdef SUPPORT_UCP
2732  /* UCD access macros */  /* UCD access macros */
2733    
2734  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2735  #define GET_UCD(ch) (_pcre_ucd_records + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2736          _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2737          UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2738    
2739  #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2740  #define UCD_SCRIPT(ch)    GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2741  #define UCD_CATEGORY(ch)  _pcre_ucp_gentype[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2742  #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2743    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2744    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2745    
2746    #endif /* SUPPORT_UCP */
2747    
2748  #endif  #endif
2749    

Legend:
Removed from v.758  
changed lines
  Added in v.1160

  ViewVC Help
Powered by ViewVC 1.1.5