/[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 576 by ph10, Sun Nov 21 18:45:10 2010 UTC revision 1128 by chpe, Thu Oct 18 18:35:08 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) 1997-2010 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. */
56  script prevents both being selected, but not everybody uses "configure". */  #if !defined COMPILE_PCRE16 && !defined COMPILE_PCRE32
57    #define COMPILE_PCRE8
 #if defined EBCDIC && defined SUPPORT_UTF8  
 #error The use of both EBCDIC and SUPPORT_UTF8 is not supported.  
58  #endif  #endif
59    
60  /* If SUPPORT_UCP is defined, SUPPORT_UTF8 must also be defined. The  /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
61  "configure" script ensures this, but not everybody uses "configure". */  "configure" script ensures this, but not everybody uses "configure". */
62    
63  #if defined SUPPORT_UCP && !defined SUPPORT_UTF8  #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
64    #define SUPPORT_UTF 1
65    #endif
66    
67    /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
68    reasons with existing code. */
69    
70    #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
71    #define SUPPORT_UTF 1
72    #endif
73    
74    /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
75    Until then we define it if SUPPORT_UTF is defined. */
76    
77    #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
78  #define SUPPORT_UTF8 1  #define SUPPORT_UTF8 1
79  #endif  #endif
80    
81    /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
82    script prevents both being selected, but not everybody uses "configure". */
83    
84    #if defined EBCDIC && defined SUPPORT_UTF
85    #error The use of both EBCDIC and SUPPORT_UTF is not supported.
86    #endif
87    
88  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
89  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
90  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 158  set, we ensure here that it has no effec Line 178  set, we ensure here that it has no effec
178  #define PCRE_CALL_CONVENTION  #define PCRE_CALL_CONVENTION
179  #endif  #endif
180    
181  /* 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
182  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
183  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
184  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
185  preprocessor time in standard C environments. */  preprocessor time in standard C environments. */
186    
187    typedef unsigned char pcre_uint8;
188    
189  #if USHRT_MAX == 65535  #if USHRT_MAX == 65535
190    typedef unsigned short pcre_uint16;    typedef unsigned short pcre_uint16;
191    typedef short pcre_int16;    typedef short pcre_int16;
# Line 206  by "configure". */ Line 228  by "configure". */
228    
229  /* All character handling must be done as unsigned characters. Otherwise there  /* All character handling must be done as unsigned characters. Otherwise there
230  are problems with top-bit-set characters and functions such as isspace().  are problems with top-bit-set characters and functions such as isspace().
231  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 *,
232  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
233  to save lots of typing. I tried "uchar", but it causes problems on Digital  called pcre_uchar.
234  Unix, where it is defined in sys/types, so use "uschar" instead. */  
235    The IN_UCHARS macro multiply its argument with the byte size of the current
236    pcre_uchar type. Useful for memcpy and such operations, whose require the
237    byte size of their input/output buffers.
238    
239    The MAX_255 macro checks whether its pcre_uchar input is less than 256.
240    
241    The TABLE_GET macro is designed for accessing elements of tables whose contain
242    exactly 256 items. When the character is able to contain more than 256
243    items, some check is needed before accessing these tables.
244    */
245    
246    #if defined COMPILE_PCRE8
247    
248    typedef unsigned char pcre_uchar;
249    #define IN_UCHARS(x) (x)
250    #define MAX_255(c) 1
251    #define TABLE_GET(c, table, default) ((table)[c])
252    
253    #elif defined COMPILE_PCRE16
254    
255    #if USHRT_MAX != 65535
256    /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
257    pcre.h(.in) and disable (comment out) this message. */
258    #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
259    #endif
260    
261    typedef pcre_uint16 pcre_uchar;
262    #define UCHAR_SHIFT (1)
263    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
264    #define MAX_255(c) ((c) <= 255u)
265    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
266    
267    #elif defined COMPILE_PCRE32
268    
269  typedef unsigned char uschar;  typedef pcre_uint32 pcre_uchar;
270    #define UCHAR_SHIFT (2)
271    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
272    #define MAX_255(c) ((c) <= 255u)
273    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
274    
275    /* Assert that pcre_uchar32 is a 32-bit type */
276    typedef int __assert_pcre_uchar32_size[sizeof(pcre_uchar) == 4 ? 1 : -1];
277    
278    #else
279    #error Unsupported compiling mode
280    #endif /* COMPILE_PCRE[8|16|32] */
281    
282  /* This is an unsigned int value that no character can ever have. UTF-8  /* This is an unsigned int value that no character can ever have. UTF-8
283  characters only go up to 0x7fffffff (though Unicode doesn't go beyond  characters only go up to 0x7fffffff (though Unicode doesn't go beyond
# Line 234  start/end of string field names are. */ Line 300  start/end of string field names are. */
300  #define IS_NEWLINE(p) \  #define IS_NEWLINE(p) \
301    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
302      ((p) < NLBLOCK->PSEND && \      ((p) < NLBLOCK->PSEND && \
303       _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\       PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
304         utf8)) \         &(NLBLOCK->nllen), utf)) \
305      : \      : \
306      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
307       (p)[0] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
308       (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
309      ) \      ) \
310    )    )
311    
# Line 248  start/end of string field names are. */ Line 314  start/end of string field names are. */
314  #define WAS_NEWLINE(p) \  #define WAS_NEWLINE(p) \
315    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
316      ((p) > NLBLOCK->PSSTART && \      ((p) > NLBLOCK->PSSTART && \
317       _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \       PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
318         &(NLBLOCK->nllen), utf8)) \         &(NLBLOCK->nllen), utf)) \
319      : \      : \
320      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
321       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
322       (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
323      ) \      ) \
324    )    )
325    
# Line 267  used for the external interface and appe Line 333  used for the external interface and appe
333  must begin with PCRE_. */  must begin with PCRE_. */
334    
335  #ifdef CUSTOM_SUBJECT_PTR  #ifdef CUSTOM_SUBJECT_PTR
336  #define PCRE_SPTR CUSTOM_SUBJECT_PTR  #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
 #define USPTR CUSTOM_SUBJECT_PTR  
337  #else  #else
338  #define PCRE_SPTR const char *  #define PCRE_PUCHAR const pcre_uchar *
 #define USPTR const unsigned char *  
339  #endif  #endif
340    
   
   
341  /* Include the public PCRE header and the definitions of UCP character property  /* Include the public PCRE header and the definitions of UCP character property
342  values. */  values. */
343    
# Line 343  The macros are controlled by the value o Line 405  The macros are controlled by the value o
405  the config.h file, but can be overridden by using -D on the command line. This  the config.h file, but can be overridden by using -D on the command line. This
406  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
407    
408    #if defined COMPILE_PCRE8
409    
410  #if LINK_SIZE == 2  #if LINK_SIZE == 2
411    
412  #define PUT(a,n,d)   \  #define PUT(a,n,d)   \
# Line 379  is automated on Unix systems via the "co Line 443  is automated on Unix systems via the "co
443  #define GET(a,n) \  #define GET(a,n) \
444    (((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])
445    
446  #define MAX_PATTERN_SIZE (1 << 30)   /* Keep it positive */  /* Keep it positive */
447    #define MAX_PATTERN_SIZE (1 << 30)
448    
449    #else
450    #error LINK_SIZE must be either 2, 3, or 4
451    #endif
452    
453    #elif defined COMPILE_PCRE16
454    
455    #if LINK_SIZE == 2
456    
457    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
458    #undef LINK_SIZE
459    #define LINK_SIZE 1
460    
461    #define PUT(a,n,d)   \
462      (a[n] = (d))
463    
464    #define GET(a,n) \
465      (a[n])
466    
467    #define MAX_PATTERN_SIZE (1 << 16)
468    
469    #elif LINK_SIZE == 3 || LINK_SIZE == 4
470    
471    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
472    #undef LINK_SIZE
473    #define LINK_SIZE 2
474    
475    #define PUT(a,n,d)   \
476      (a[n] = (d) >> 16), \
477      (a[(n)+1] = (d) & 65535)
478    
479    #define GET(a,n) \
480      (((a)[n] << 16) | (a)[(n)+1])
481    
482    /* Keep it positive */
483    #define MAX_PATTERN_SIZE (1 << 30)
484    
485  #else  #else
486  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
487  #endif  #endif
488    
489    #elif defined COMPILE_PCRE32
490    
491    /* Only supported LINK_SIZE is 4 */
492    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
493    #undef LINK_SIZE
494    #define LINK_SIZE 1
495    
496    #define PUT(a,n,d)   \
497      (a[n] = (d))
498    
499    #define GET(a,n) \
500      (a[n])
501    
502    /* Keep it positive */
503    #define MAX_PATTERN_SIZE (1 << 30)
504    
505    #else
506    #error Unsupported compiling mode
507    #endif /* COMPILE_PCRE[8|16|32] */
508    
509  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
510    
# Line 396  is automated on Unix systems via the "co Line 515  is automated on Unix systems via the "co
515  offsets changes. There are used for repeat counts and for other things such as  offsets changes. There are used for repeat counts and for other things such as
516  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
517    
518    #if defined COMPILE_PCRE8
519    
520    #define IMM2_SIZE 2
521    
522  #define PUT2(a,n,d)   \  #define PUT2(a,n,d)   \
523    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
524    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
# Line 403  capturing parenthesis numbers in back re Line 526  capturing parenthesis numbers in back re
526  #define GET2(a,n) \  #define GET2(a,n) \
527    (((a)[n] << 8) | (a)[(n)+1])    (((a)[n] << 8) | (a)[(n)+1])
528    
529  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += 2  #elif defined COMPILE_PCRE16
530    
531    #define IMM2_SIZE 1
532    
533  /* When UTF-8 encoding is being used, a character is no longer just a single  #define PUT2(a,n,d)   \
534  byte. The macros for character handling generate simple sequences when used in     a[n] = d
535  byte-mode, and more complicated ones for UTF-8 characters. GETCHARLENTEST is  
536  not used when UTF-8 is not supported, so it is not defined, and BACKCHAR should  #define GET2(a,n) \
537  never be called in byte mode. To make sure they can never even appear when     a[n]
 UTF-8 support is omitted, we don't even define them. */  
538    
539  #ifndef SUPPORT_UTF8  #elif defined COMPILE_PCRE32
540    
541    #define IMM2_SIZE 1
542    
543    #define PUT2(a,n,d)   \
544       a[n] = d
545    
546    #define GET2(a,n) \
547       a[n]
548    
549    #else
550    #error Unsupported compiling mode
551    #endif /* COMPILE_PCRE[8|16|32] */
552    
553    #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
554    
555    /* The maximum length of a MARK name is currently one data unit; it may be
556    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
557    
558    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
559    #define MAX_MARK ((1u << 16) - 1)
560    #else
561    #define MAX_MARK ((1u << 8) - 1)
562    #endif
563    
564    /* When UTF encoding is being used, a character is no longer just a single
565    byte. The macros for character handling generate simple sequences when used in
566    character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
567    and other macros are not used when UTF is not supported, so they are not
568    defined. To make sure they can never even appear when UTF support is omitted,
569    we don't even define them. */
570    
571    #ifndef SUPPORT_UTF
572    
573    /* #define MAX_VALUE_FOR_SINGLE_CHAR */
574    /* #define HAS_EXTRALEN(c) */
575    /* #define GET_EXTRALEN(c) */
576    /* #define NOT_FIRSTCHAR(c) */
577  #define GETCHAR(c, eptr) c = *eptr;  #define GETCHAR(c, eptr) c = *eptr;
578  #define GETCHARTEST(c, eptr) c = *eptr;  #define GETCHARTEST(c, eptr) c = *eptr;
579  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
580  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
581  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
582    #define RAWUCHAR(eptr) (*(eptr))
583    #define RAWUCHARINC(eptr) (*(eptr)++)
584    #define RAWUCHARTEST(eptr) (*(eptr))
585    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
586  /* #define GETCHARLENTEST(c, eptr, len) */  /* #define GETCHARLENTEST(c, eptr, len) */
587  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
588    /* #define FORWARDCHAR(eptr) */
589    /* #define ACROSSCHAR(condition, eptr, action) */
590    
591  #else   /* SUPPORT_UTF8 */  #else   /* SUPPORT_UTF */
592    
593  /* These macros were originally written in the form of loops that used data  /* Tests whether the code point needs extra characters to decode. */
594  from the tables whose names start with _pcre_utf8_table. They were rewritten by  
595  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. */  
596    
597  /* 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
598  advancing the pointer. */  advancing the pointer. */
# Line 451  advancing the pointer. */ Line 616  advancing the pointer. */
616            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \            ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
617      }      }
618    
 /* 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);  
   
619  /* 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
620  the pointer. */  the pointer. */
621    
# Line 499  the pointer. */ Line 650  the pointer. */
650        } \        } \
651      }      }
652    
653    #if defined COMPILE_PCRE8
654    
655    /* These macros were originally written in the form of loops that used data
656    from the tables whose names start with PRIV(utf8_table). They were rewritten by
657    a user so as not to use loops, because in some environments this gives a
658    significant performance advantage, and it seems never to do any harm. */
659    
660    /* Tells the biggest code point which can be encoded as a single character. */
661    
662    #define MAX_VALUE_FOR_SINGLE_CHAR 127
663    
664    /* Tests whether the code point needs extra characters to decode. */
665    
666    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
667    
668    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
669    Otherwise it has an undefined behaviour. */
670    
671    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
672    
673    /* Returns TRUE, if the given character is not the first character
674    of a UTF sequence. */
675    
676    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
677    
678    /* Get the next UTF-8 character, not advancing the pointer. This is called when
679    we know we are in UTF-8 mode. */
680    
681    #define GETCHAR(c, eptr) \
682      c = *eptr; \
683      if (c >= 0xc0) GETUTF8(c, eptr);
684    
685    /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
686    pointer. */
687    
688    #define GETCHARTEST(c, eptr) \
689      c = *eptr; \
690      if (utf && c >= 0xc0) GETUTF8(c, eptr);
691    
692  /* 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
693  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
694    
# Line 511  This is called when we don't know if we Line 701  This is called when we don't know if we
701    
702  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
703    c = *eptr++; \    c = *eptr++; \
704    if (utf8 && c >= 0xc0) GETUTF8INC(c, eptr);    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
705    
706  /* 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
707  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 753  do not know if we are in UTF-8 mode. */
753    
754  #define GETCHARLENTEST(c, eptr, len) \  #define GETCHARLENTEST(c, eptr, len) \
755    c = *eptr; \    c = *eptr; \
756    if (utf8 && c >= 0xc0) GETUTF8LEN(c, eptr, len);    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
757    
758    /* Returns the next uchar, not advancing the pointer. This is called when
759    we know we are in UTF mode. */
760    
761    #define RAWUCHAR(eptr) \
762      (*(eptr))
763    
764    /* Returns the next uchar, advancing the pointer. This is called when
765    we know we are in UTF mode. */
766    
767    #define RAWUCHARINC(eptr) \
768      (*(eptr)++)
769    
770    /* Returns the next uchar, testing for UTF mode, and not advancing the
771    pointer. */
772    
773    #define RAWUCHARTEST(eptr) \
774      (*(eptr))
775    
776    /* Returns the next uchar, testing for UTF mode, advancing the
777    pointer. */
778    
779    #define RAWUCHARINCTEST(eptr) \
780      (*(eptr)++)
781    
782  /* 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
783  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 785  because almost all calls are already wit
785    
786  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
787    
788  #endif  /* SUPPORT_UTF8 */  /* Same as above, just in the other direction. */
789    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
790    
791    /* Same as above, but it allows a fully customizable form. */
792    #define ACROSSCHAR(condition, eptr, action) \
793      while((condition) && ((eptr) & 0xc0) == 0x80) action
794    
795  /* In case there is no definition of offsetof() provided - though any proper  #elif defined COMPILE_PCRE16
 Standard C system should have one. */  
796    
797  #ifndef offsetof  /* Tells the biggest code point which can be encoded as a single character. */
798  #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
799  #endif  #define MAX_VALUE_FOR_SINGLE_CHAR 65535
800    
801    /* Tests whether the code point needs extra characters to decode. */
802    
803    #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
804    
805    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
806    Otherwise it has an undefined behaviour. */
807    
808    #define GET_EXTRALEN(c) 1
809    
810    /* Returns TRUE, if the given character is not the first character
811    of a UTF sequence. */
812    
813    #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
814    
815    /* Base macro to pick up the low surrogate of a UTF-16 character, not
816    advancing the pointer. */
817    
818    #define GETUTF16(c, eptr) \
819       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
820    
821    /* Get the next UTF-16 character, not advancing the pointer. This is called when
822    we know we are in UTF-16 mode. */
823    
824    #define GETCHAR(c, eptr) \
825      c = *eptr; \
826      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
827    
828    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
829    pointer. */
830    
831    #define GETCHARTEST(c, eptr) \
832      c = *eptr; \
833      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
834    
835    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
836    the pointer. */
837    
838    #define GETUTF16INC(c, eptr) \
839       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
840    
841    /* Get the next UTF-16 character, advancing the pointer. This is called when we
842    know we are in UTF-16 mode. */
843    
844    #define GETCHARINC(c, eptr) \
845      c = *eptr++; \
846      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
847    
848    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
849    This is called when we don't know if we are in UTF-16 mode. */
850    
851    #define GETCHARINCTEST(c, eptr) \
852      c = *eptr++; \
853      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
854    
855    /* Base macro to pick up the low surrogate of a UTF-16 character, not
856    advancing the pointer, incrementing the length. */
857    
858    #define GETUTF16LEN(c, eptr, len) \
859       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
860    
861    /* Get the next UTF-16 character, not advancing the pointer, incrementing
862    length if there is a low surrogate. This is called when we know we are in
863    UTF-16 mode. */
864    
865    #define GETCHARLEN(c, eptr, len) \
866      c = *eptr; \
867      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
868    
869    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
870    pointer, incrementing length if there is a low surrogate. This is called when
871    we do not know if we are in UTF-16 mode. */
872    
873    #define GETCHARLENTEST(c, eptr, len) \
874      c = *eptr; \
875      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
876    
877    /* Returns the next uchar, not advancing the pointer. This is called when
878    we know we are in UTF mode. */
879    
880    #define RAWUCHAR(eptr) \
881      (*(eptr))
882    
883    /* Returns the next uchar, advancing the pointer. This is called when
884    we know we are in UTF mode. */
885    
886    #define RAWUCHARINC(eptr) \
887      (*(eptr)++)
888    
889    /* Returns the next uchar, testing for UTF mode, and not advancing the
890    pointer. */
891    
892    #define RAWUCHARTEST(eptr) \
893      (*(eptr))
894    
895    /* Returns the next uchar, testing for UTF mode, advancing the
896    pointer. */
897    
898    #define RAWUCHARINCTEST(eptr) \
899      (*(eptr)++)
900    
901    /* If the pointer is not at the start of a character, move it back until
902    it is. This is called only in UTF-16 mode - we don't put a test within the
903    macro because almost all calls are already within a block of UTF-16 only
904    code. */
905    
906    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
907    
908    /* Same as above, just in the other direction. */
909    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
910    
911    /* Same as above, but it allows a fully customizable form. */
912    #define ACROSSCHAR(condition, eptr, action) \
913      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
914    
915    #elif defined COMPILE_PCRE32
916    
917    /* These are trivial for the 32-bit library, since all UTF-32 characters fit
918    into one pcre_uchar unit. */
919    #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
920    #define HAS_EXTRALEN(c) (0)
921    #define GET_EXTRALEN(c) (0)
922    #define NOT_FIRSTCHAR(c) (0)
923    
924    #define UTF32_MASK (0x1fffffu)
925    
926    /* Get the next UTF-32 character, not advancing the pointer. This is called when
927    we know we are in UTF-32 mode. */
928    
929    #define GETCHAR(c, eptr) \
930      c = (*eptr) & UTF32_MASK;
931    
932    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
933    pointer. */
934    
935    #define GETCHARTEST(c, eptr) \
936      c = *eptr; \
937      if (utf) c &= UTF32_MASK;
938    
939    /* Get the next UTF-32 character, advancing the pointer. This is called when we
940    know we are in UTF-32 mode. */
941    
942    #define GETCHARINC(c, eptr) \
943      c = (*eptr++) & UTF32_MASK;
944    
945    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
946    This is called when we don't know if we are in UTF-32 mode. */
947    
948    #define GETCHARINCTEST(c, eptr) \
949      c = *eptr++; \
950      if (utf) c &= UTF32_MASK;
951    
952    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
953    length (since all UTF-32 is of length 1). This is called when we know we are in
954    UTF-32 mode. */
955    
956    #define GETCHARLEN(c, eptr, len) \
957      GETCHAR(c, eptr)
958    
959    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
960    pointer, not incrementing the length (since all UTF-32 is of length 1).
961    This is called when we do not know if we are in UTF-32 mode. */
962    
963    #define GETCHARLENTEST(c, eptr, len) \
964      GETCHARTEST(c, eptr)
965    
966    /* Returns the next uchar, not advancing the pointer. This is called when
967    we know we are in UTF mode. */
968    
969    #define RAWUCHAR(eptr) \
970      (*(eptr) & UTF32_MASK)
971    
972    /* Returns the next uchar, advancing the pointer. This is called when
973    we know we are in UTF mode. */
974    
975    #define RAWUCHARINC(eptr) \
976      (*(eptr)++ & UTF32_MASK)
977    
978    /* Returns the next uchar, testing for UTF mode, and not advancing the
979    pointer. */
980    
981    #define RAWUCHARTEST(eptr) \
982      (utf ? (*(eptr) & UTF32_MASK) : *(eptr))
983    
984    /* Returns the next uchar, testing for UTF mode, advancing the
985    pointer. */
986    
987    #define RAWUCHARINCTEST(eptr) \
988      (utf ? (*(eptr)++ & UTF32_MASK) : *(eptr)++)
989    
990    /* If the pointer is not at the start of a character, move it back until
991    it is. This is called only in UTF-32 mode - we don't put a test within the
992    macro because almost all calls are already within a block of UTF-32 only
993    code.
994    These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
995    
996    #define BACKCHAR(eptr) do { } while (0)
997    
998    /* Same as above, just in the other direction. */
999    #define FORWARDCHAR(eptr) do { } while (0)
1000    
1001    /* Same as above, but it allows a fully customizable form. */
1002    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1003    
1004    #else
1005    #error Unsupported compiling mode
1006    #endif /* COMPILE_PCRE[8|16|32] */
1007    
1008    #endif  /* SUPPORT_UTF */
1009    
1010    /* Tests for Unicode horizontal and vertical whitespace characters must check a
1011    number of different values. Using a switch statement for this generates the
1012    fastest code (no loop, no memory access), and there are several places in the
1013    interpreter code where this happens. In order to ensure that all the case lists
1014    remain in step, we use macros so that there is only one place where the lists
1015    are defined.
1016    
1017    These values are also required as lists in pcre_compile.c when processing \h,
1018    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1019    macros that define the values are here so that all the definitions are
1020    together. The lists must be in ascending character order, terminated by
1021    NOTACHAR (which is 0xffffffff).
1022    
1023    Any changes should ensure that the various macros are kept in step with each
1024    other. NOTE: The values also appear in pcre_jit_compile.c. */
1025    
1026    /* ------ ASCII/Unicode environments ------ */
1027    
1028    #ifndef EBCDIC
1029    
1030    #define HSPACE_LIST \
1031      CHAR_HT, CHAR_SPACE, 0xa0, \
1032      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1033      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1034      NOTACHAR
1035    
1036    #define HSPACE_MULTIBYTE_CASES \
1037      case 0x1680:  /* OGHAM SPACE MARK */ \
1038      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1039      case 0x2000:  /* EN QUAD */ \
1040      case 0x2001:  /* EM QUAD */ \
1041      case 0x2002:  /* EN SPACE */ \
1042      case 0x2003:  /* EM SPACE */ \
1043      case 0x2004:  /* THREE-PER-EM SPACE */ \
1044      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1045      case 0x2006:  /* SIX-PER-EM SPACE */ \
1046      case 0x2007:  /* FIGURE SPACE */ \
1047      case 0x2008:  /* PUNCTUATION SPACE */ \
1048      case 0x2009:  /* THIN SPACE */ \
1049      case 0x200A:  /* HAIR SPACE */ \
1050      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1051      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1052      case 0x3000   /* IDEOGRAPHIC SPACE */
1053    
1054    #define HSPACE_BYTE_CASES \
1055      case CHAR_HT: \
1056      case CHAR_SPACE: \
1057      case 0xa0     /* NBSP */
1058    
1059    #define HSPACE_CASES \
1060      HSPACE_BYTE_CASES: \
1061      HSPACE_MULTIBYTE_CASES
1062    
1063    #define VSPACE_LIST \
1064      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1065    
1066    #define VSPACE_MULTIBYTE_CASES \
1067      case 0x2028:    /* LINE SEPARATOR */ \
1068      case 0x2029     /* PARAGRAPH SEPARATOR */
1069    
1070    #define VSPACE_BYTE_CASES \
1071      case CHAR_LF: \
1072      case CHAR_VT: \
1073      case CHAR_FF: \
1074      case CHAR_CR: \
1075      case CHAR_NEL
1076    
1077    #define VSPACE_CASES \
1078      VSPACE_BYTE_CASES: \
1079      VSPACE_MULTIBYTE_CASES
1080    
1081    /* ------ EBCDIC environments ------ */
1082    
1083    #else
1084    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1085    
1086    #define HSPACE_BYTE_CASES \
1087      case CHAR_HT: \
1088      case CHAR_SPACE
1089    
1090    #define HSPACE_CASES HSPACE_BYTE_CASES
1091    
1092    #ifdef EBCDIC_NL25
1093    #define VSPACE_LIST \
1094      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1095    #else
1096    #define VSPACE_LIST \
1097      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1098    #endif
1099    
1100    #define VSPACE_BYTE_CASES \
1101      case CHAR_LF: \
1102      case CHAR_VT: \
1103      case CHAR_FF: \
1104      case CHAR_CR: \
1105      case CHAR_NEL
1106    
1107    #define VSPACE_CASES VSPACE_BYTE_CASES
1108    #endif  /* EBCDIC */
1109    
1110    /* ------ End of whitespace macros ------ */
1111    
 /* These are the public options that can change during matching. */  
1112    
 #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)  
1113    
1114  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
1115  live at the top end of the options word, but that got almost full, so now they  live at the top end of the options word, but that got almost full, so now they
# Line 592  are in a 16-bit flags word. From release Line 1117  are in a 16-bit flags word. From release
1117  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1118  compatibility. */  compatibility. */
1119    
1120  #define PCRE_NOPARTIAL     0x0001  /* can't use partial with this regex */  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1121  #define PCRE_FIRSTSET      0x0002  /* first_byte is set */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1122  #define PCRE_REQCHSET      0x0004  /* req_byte is set */  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
1123  #define PCRE_STARTLINE     0x0008  /* start after \n for multiline */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1124  #define PCRE_JCHANGED      0x0010  /* j option used in regex */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1125  #define PCRE_HASCRORLF     0x0020  /* explicit \r or \n in pattern */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
1126    #define PCRE_RCH_CASELESS  0x0080  /* caseless requested char */
1127    #define PCRE_STARTLINE     0x0100  /* start after \n for multiline */
1128    #define PCRE_NOPARTIAL     0x0200  /* can't use partial with this regex */
1129    #define PCRE_JCHANGED      0x0400  /* j option used in regex */
1130    #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1131    #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1132    
1133    #if defined COMPILE_PCRE8
1134    #define PCRE_MODE          PCRE_MODE8
1135    #elif defined COMPILE_PCRE16
1136    #define PCRE_MODE          PCRE_MODE16
1137    #elif defined COMPILE_PCRE32
1138    #define PCRE_MODE          PCRE_MODE32
1139    #endif
1140    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1141    
1142  /* Options for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1143    
1144  #define PCRE_STUDY_MAPPED   0x01     /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
1145  #define PCRE_STUDY_MINLEN   0x02     /* a minimum length field exists */  #define PCRE_STUDY_MINLEN  0x0002  /* a minimum length field exists */
1146    
1147  /* Masks for identifying the public options that are permitted at compile  /* Masks for identifying the public options that are permitted at compile
1148  time, run time, or study time, respectively. */  time, run time, or study time, respectively. */
# Line 628  time, run time, or study time, respectiv Line 1168  time, run time, or study time, respectiv
1168     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1169     PCRE_NO_START_OPTIMIZE)     PCRE_NO_START_OPTIMIZE)
1170    
1171  #define PUBLIC_STUDY_OPTIONS 0   /* None defined */  #define PUBLIC_STUDY_OPTIONS \
1172       (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1173        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1174    
1175  /* 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. */  
1176    
1177  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1178    
1179  /* Negative values for the firstchar and reqchar variables */  /* This variable is used to detect a loaded regular expression
1180    in different endianness. */
1181    
1182  #define REQ_UNSET (-2)  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
 #define REQ_NONE  (-1)  
1183    
1184  /* 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
1185  req_byte match. */  req_byte match. */
1186    
1187  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1188    
 /* 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 */  
   
1189  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1190  environments where these macros are defined elsewhere. Unfortunately, there  environments where these macros are defined elsewhere. Unfortunately, there
1191  is no way to do the same for the typedef. */  is no way to do the same for the typedef. */
# Line 679  for) in a minority area (EBCDIC platform Line 1214  for) in a minority area (EBCDIC platform
1214  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
1215  macros to give the functions distinct names. */  macros to give the functions distinct names. */
1216    
1217  #ifndef SUPPORT_UTF8  #ifndef SUPPORT_UTF
1218    
1219  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1220  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
1221    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1222    characters, a common practice has been to use its NL (0x15) character as the
1223    line terminator in C-like processing environments. However, sometimes the LF
1224    (0x25) character is used instead, according to this Unicode document:
1225    
1226    http://unicode.org/standard/reports/tr13/tr13-5.html
1227    
1228    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1229    instead. Whichever is *not* chosen is defined as NEL.
1230    
1231    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1232    same code point. */
1233    
1234    #ifdef EBCDIC
1235    
1236    #ifndef EBCDIC_NL25
1237    #define CHAR_NL                     '\x15'
1238    #define CHAR_NEL                    '\x25'
1239    #define STR_NL                      "\x15"
1240    #define STR_NEL                     "\x25"
1241    #else
1242    #define CHAR_NL                     '\x25'
1243    #define CHAR_NEL                    '\x15'
1244    #define STR_NL                      "\x25"
1245    #define STR_NEL                     "\x15"
1246    #endif
1247    
1248    #define CHAR_LF                     CHAR_NL
1249    #define STR_LF                      STR_NL
1250    
1251    #define CHAR_ESC                    '\047'
1252    #define CHAR_DEL                    '\007'
1253    #define STR_ESC                     "\047"
1254    #define STR_DEL                     "\007"
1255    
1256    #else  /* Not EBCDIC */
1257    
1258    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1259    compatibility. NEL is the Unicode newline character; make sure it is
1260    a positive value. */
1261    
1262    #define CHAR_LF                     '\n'
1263    #define CHAR_NL                     CHAR_LF
1264    #define CHAR_NEL                    ((unsigned char)'\x85')
1265    #define CHAR_ESC                    '\033'
1266    #define CHAR_DEL                    '\177'
1267    
1268    #define STR_LF                      "\n"
1269    #define STR_NL                      STR_LF
1270    #define STR_NEL                     "\x85"
1271    #define STR_ESC                     "\033"
1272    #define STR_DEL                     "\177"
1273    
1274    #endif  /* EBCDIC */
1275    
1276    /* The remaining definitions work in both environments. */
1277    
1278  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1279  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1280  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1281  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1282  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1283  #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  
1284    
1285  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1286  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 799  so that PCRE works on both ASCII and EBC Line 1382  so that PCRE works on both ASCII and EBC
1382  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1383  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1384  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1385  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1386  #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  
1387    
1388  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1389  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 939  so that PCRE works on both ASCII and EBC Line 1514  so that PCRE works on both ASCII and EBC
1514  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1515  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1516  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1517  #define STRING_UTF8_RIGHTPAR           "UTF8)"  #ifdef COMPILE_PCRE8
1518    #define STRING_UTF_RIGHTPAR            "UTF8)"
1519    #endif
1520    #ifdef COMPILE_PCRE16
1521    #define STRING_UTF_RIGHTPAR            "UTF16)"
1522    #endif
1523    #ifdef COMPILE_PCRE32
1524    #define STRING_UTF_RIGHTPAR            "UTF32)"
1525    #endif
1526  #define STRING_UCP_RIGHTPAR            "UCP)"  #define STRING_UCP_RIGHTPAR            "UCP)"
1527  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1528    
1529  #else  /* SUPPORT_UTF8 */  #else  /* SUPPORT_UTF */
1530    
1531  /* 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
1532  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 953  only. */ Line 1536  only. */
1536  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1537  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1538  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1539  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1540    #define CHAR_NL                     CHAR_LF
1541    #define CHAR_NEL                    ((unsigned char)'\x85')
1542  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1543  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1544  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
# Line 1194  only. */ Line 1779  only. */
1779  #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
1780  #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
1781  #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
1782  #define STRING_UTF8_RIGHTPAR           STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS  #ifdef COMPILE_PCRE8
1783    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1784    #endif
1785    #ifdef COMPILE_PCRE16
1786    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1787    #endif
1788    #ifdef COMPILE_PCRE32
1789    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1790    #endif
1791  #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
1792  #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
1793    
1794  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
1795    
1796  /* Escape items that are just an encoding of a particular data value. */  /* Escape items that are just an encoding of a particular data value. */
1797    
# Line 1211  only. */ Line 1804  only. */
1804  #endif  #endif
1805    
1806  #ifndef ESC_n  #ifndef ESC_n
1807  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1808  #endif  #endif
1809    
1810  #ifndef ESC_r  #ifndef ESC_r
# Line 1236  only. */ Line 1829  only. */
1829  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1830  #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 */
1831  #define PT_WORD       8    /* Word - L plus N plus underscore */  #define PT_WORD       8    /* Word - L plus N plus underscore */
1832    #define PT_CLIST      9    /* Pseudo-property: match character list */
1833    
1834  /* 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
1835  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1836    
1837  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1838  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */
# Line 1251  contain UTF-8 characters with values gre Line 1845  contain UTF-8 characters with values gre
1845    
1846  /* 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
1847  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
1848  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
1849  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
1850  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
1851  used for [^] in JavaScript compatibility mode. In non-DOTALL mode, "." behaves  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1852  like \N.  non-DOTALL mode, "." behaves like \N.
1853    
1854  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.
1855  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.
1856  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
1857  looked up from a table.  looked up from a table.
1858    
1859  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
1860  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
1861  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
1862  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
1863  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 1273  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E Line 1867  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, E
1867         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,
1868         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1869         ESC_E, ESC_Q, ESC_g, ESC_k,         ESC_E, ESC_Q, ESC_g, ESC_k,
1870         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 };  
1871    
1872  /* 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
1873  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 1299  enum { Line 1892  enum {
1892    OP_WHITESPACE,         /*  9 \s */    OP_WHITESPACE,         /*  9 \s */
1893    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1894    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1895    OP_ANY,            /* 12 Match any character except newline */  
1896      OP_ANY,            /* 12 Match any character except newline (\N) */
1897    OP_ALLANY,         /* 13 Match any character */    OP_ALLANY,         /* 13 Match any character */
1898    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 */
1899    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
# Line 1310  enum { Line 1904  enum {
1904    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1905    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1906    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1907    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) */
1908    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1909    
1910    OP_OPT,            /* 25 Set runtime options */    OP_CIRC,           /* 25 Start of line - not multiline */
1911    OP_CIRC,           /* 26 Start of line - varies with multiline switch */    OP_CIRCM,          /* 26 Start of line - multiline */
1912    OP_DOLL,           /* 27 End of line - varies with multiline switch */    OP_DOLL,           /* 27 End of line - not multiline */
1913    OP_CHAR,           /* 28 Match one character, casefully */    OP_DOLLM,          /* 28 End of line - multiline */
1914    OP_CHARNC,         /* 29 Match one character, caselessly */    OP_CHAR,           /* 29 Match one character, casefully */
1915    OP_NOT,            /* 30 Match one character, not the following one */    OP_CHARI,          /* 30 Match one character, caselessly */
1916      OP_NOT,            /* 31 Match one character, not the given one, casefully */
1917    OP_STAR,           /* 31 The maximizing and minimizing versions of */    OP_NOTI,           /* 32 Match one character, not the given one, caselessly */
1918    OP_MINSTAR,        /* 32 these six opcodes must come in pairs, with */  
1919    OP_PLUS,           /* 33 the minimizing one second. */    /* The following sets of 13 opcodes must always be kept in step because
1920    OP_MINPLUS,        /* 34 This first set applies to single characters.*/    the offset from the first one is used to generate the others. */
1921    OP_QUERY,          /* 35 */  
1922    OP_MINQUERY,       /* 36 */    /**** Single characters, caseful, must precede the caseless ones ****/
1923    
1924    OP_UPTO,           /* 37 From 0 to n matches */    OP_STAR,           /* 33 The maximizing and minimizing versions of */
1925    OP_MINUPTO,        /* 38 */    OP_MINSTAR,        /* 34 these six opcodes must come in pairs, with */
1926    OP_EXACT,          /* 39 Exactly n matches */    OP_PLUS,           /* 35 the minimizing one second. */
1927      OP_MINPLUS,        /* 36 */
1928    OP_POSSTAR,        /* 40 Possessified star */    OP_QUERY,          /* 37 */
1929    OP_POSPLUS,        /* 41 Possessified plus */    OP_MINQUERY,       /* 38 */
1930    OP_POSQUERY,       /* 42 Posesssified query */  
1931    OP_POSUPTO,        /* 43 Possessified upto */    OP_UPTO,           /* 39 From 0 to n matches of one character, caseful*/
1932      OP_MINUPTO,        /* 40 */
1933    OP_NOTSTAR,        /* 44 The maximizing and minimizing versions of */    OP_EXACT,          /* 41 Exactly n matches */
1934    OP_NOTMINSTAR,     /* 45 these six opcodes must come in pairs, with */  
1935    OP_NOTPLUS,        /* 46 the minimizing one second. They must be in */    OP_POSSTAR,        /* 42 Possessified star, caseful */
1936    OP_NOTMINPLUS,     /* 47 exactly the same order as those above. */    OP_POSPLUS,        /* 43 Possessified plus, caseful */
1937    OP_NOTQUERY,       /* 48 This set applies to "not" single characters. */    OP_POSQUERY,       /* 44 Posesssified query, caseful */
1938    OP_NOTMINQUERY,    /* 49 */    OP_POSUPTO,        /* 45 Possessified upto, caseful */
1939    
1940    OP_NOTUPTO,        /* 50 From 0 to n matches */    /**** Single characters, caseless, must follow the caseful ones */
1941    OP_NOTMINUPTO,     /* 51 */  
1942    OP_NOTEXACT,       /* 52 Exactly n matches */    OP_STARI,          /* 46 */
1943      OP_MINSTARI,       /* 47 */
1944    OP_NOTPOSSTAR,     /* 53 Possessified versions */    OP_PLUSI,          /* 48 */
1945    OP_NOTPOSPLUS,     /* 54 */    OP_MINPLUSI,       /* 49 */
1946    OP_NOTPOSQUERY,    /* 55 */    OP_QUERYI,         /* 50 */
1947    OP_NOTPOSUPTO,     /* 56 */    OP_MINQUERYI,      /* 51 */
1948    
1949    OP_TYPESTAR,       /* 57 The maximizing and minimizing versions of */    OP_UPTOI,          /* 52 From 0 to n matches of one character, caseless */
1950    OP_TYPEMINSTAR,    /* 58 these six opcodes must come in pairs, with */    OP_MINUPTOI,       /* 53 */
1951    OP_TYPEPLUS,       /* 59 the minimizing one second. These codes must */    OP_EXACTI,         /* 54 */
1952    OP_TYPEMINPLUS,    /* 60 be in exactly the same order as those above. */  
1953    OP_TYPEQUERY,      /* 61 This set applies to character types such as \d */    OP_POSSTARI,       /* 55 Possessified star, caseless */
1954    OP_TYPEMINQUERY,   /* 62 */    OP_POSPLUSI,       /* 56 Possessified plus, caseless */
1955      OP_POSQUERYI,      /* 57 Posesssified query, caseless */
1956    OP_TYPEUPTO,       /* 63 From 0 to n matches */    OP_POSUPTOI,       /* 58 Possessified upto, caseless */
1957    OP_TYPEMINUPTO,    /* 64 */  
1958    OP_TYPEEXACT,      /* 65 Exactly n matches */    /**** The negated ones must follow the non-negated ones, and match them ****/
1959      /**** Negated single character, caseful; must precede the caseless ones ****/
1960    OP_TYPEPOSSTAR,    /* 66 Possessified versions */  
1961    OP_TYPEPOSPLUS,    /* 67 */    OP_NOTSTAR,        /* 59 The maximizing and minimizing versions of */
1962    OP_TYPEPOSQUERY,   /* 68 */    OP_NOTMINSTAR,     /* 60 these six opcodes must come in pairs, with */
1963    OP_TYPEPOSUPTO,    /* 69 */    OP_NOTPLUS,        /* 61 the minimizing one second. They must be in */
1964      OP_NOTMINPLUS,     /* 62 exactly the same order as those above. */
1965    OP_CRSTAR,         /* 70 The maximizing and minimizing versions of */    OP_NOTQUERY,       /* 63 */
1966    OP_CRMINSTAR,      /* 71 all these opcodes must come in pairs, with */    OP_NOTMINQUERY,    /* 64 */
1967    OP_CRPLUS,         /* 72 the minimizing one second. These codes must */  
1968    OP_CRMINPLUS,      /* 73 be in exactly the same order as those above. */    OP_NOTUPTO,        /* 65 From 0 to n matches, caseful */
1969    OP_CRQUERY,        /* 74 These are for character classes and back refs */    OP_NOTMINUPTO,     /* 66 */
1970    OP_CRMINQUERY,     /* 75 */    OP_NOTEXACT,       /* 67 Exactly n matches */
1971    OP_CRRANGE,        /* 76 These are different to the three sets above. */  
1972    OP_CRMINRANGE,     /* 77 */    OP_NOTPOSSTAR,     /* 68 Possessified versions, caseful */
1973      OP_NOTPOSPLUS,     /* 69 */
1974    OP_CLASS,          /* 78 Match a character class, chars < 256 only */    OP_NOTPOSQUERY,    /* 70 */
1975    OP_NCLASS,         /* 79 Same, but the bitmap was created from a negative    OP_NOTPOSUPTO,     /* 71 */
1976                             class - the difference is relevant only when a UTF-8  
1977                             character > 255 is encountered. */    /**** Negated single character, caseless; must follow the caseful ones ****/
1978    
1979    OP_XCLASS,         /* 80 Extended class for handling UTF-8 chars within the    OP_NOTSTARI,       /* 72 */
1980                             class. This does both positive and negative. */    OP_NOTMINSTARI,    /* 73 */
1981      OP_NOTPLUSI,       /* 74 */
1982    OP_REF,            /* 81 Match a back reference */    OP_NOTMINPLUSI,    /* 75 */
1983    OP_RECURSE,        /* 82 Match a numbered subpattern (possibly recursive) */    OP_NOTQUERYI,      /* 76 */
1984    OP_CALLOUT,        /* 83 Call out to external function if provided */    OP_NOTMINQUERYI,   /* 77 */
1985    
1986    OP_ALT,            /* 84 Start of alternation */    OP_NOTUPTOI,       /* 78 From 0 to n matches, caseless */
1987    OP_KET,            /* 85 End of group that doesn't have an unbounded repeat */    OP_NOTMINUPTOI,    /* 79 */
1988    OP_KETRMAX,        /* 86 These two must remain together and in this */    OP_NOTEXACTI,      /* 80 Exactly n matches */
1989    OP_KETRMIN,        /* 87 order. They are for groups the repeat for ever. */  
1990      OP_NOTPOSSTARI,    /* 81 Possessified versions, caseless */
1991    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/    OP_NOTPOSPLUSI,    /* 82 */
1992      OP_NOTPOSQUERYI,   /* 83 */
1993    OP_ASSERT,         /* 88 Positive lookahead */    OP_NOTPOSUPTOI,    /* 84 */
1994    OP_ASSERT_NOT,     /* 89 Negative lookahead */  
1995    OP_ASSERTBACK,     /* 90 Positive lookbehind */    /**** Character types ****/
1996    OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */  
1997    OP_REVERSE,        /* 92 Move pointer back - used in lookbehind assertions */    OP_TYPESTAR,       /* 85 The maximizing and minimizing versions of */
1998      OP_TYPEMINSTAR,    /* 86 these six opcodes must come in pairs, with */
1999    /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,    OP_TYPEPLUS,       /* 87 the minimizing one second. These codes must */
2000    as there's a test for >= ONCE for a subpattern that isn't an assertion. */    OP_TYPEMINPLUS,    /* 88 be in exactly the same order as those above. */
2001      OP_TYPEQUERY,      /* 89 */
2002    OP_ONCE,           /* 93 Atomic group */    OP_TYPEMINQUERY,   /* 90 */
2003    OP_BRA,            /* 94 Start of non-capturing bracket */  
2004    OP_CBRA,           /* 95 Start of capturing bracket */    OP_TYPEUPTO,       /* 91 From 0 to n matches */
2005    OP_COND,           /* 96 Conditional group */    OP_TYPEMINUPTO,    /* 92 */
2006      OP_TYPEEXACT,      /* 93 Exactly n matches */
2007    
2008      OP_TYPEPOSSTAR,    /* 94 Possessified versions */
2009      OP_TYPEPOSPLUS,    /* 95 */
2010      OP_TYPEPOSQUERY,   /* 96 */
2011      OP_TYPEPOSUPTO,    /* 97 */
2012    
2013      /* These are used for character classes and back references; only the
2014      first six are the same as the sets above. */
2015    
2016      OP_CRSTAR,         /* 98 The maximizing and minimizing versions of */
2017      OP_CRMINSTAR,      /* 99 all these opcodes must come in pairs, with */
2018      OP_CRPLUS,         /* 100 the minimizing one second. These codes must */
2019      OP_CRMINPLUS,      /* 101 be in exactly the same order as those above. */
2020      OP_CRQUERY,        /* 102 */
2021      OP_CRMINQUERY,     /* 103 */
2022    
2023      OP_CRRANGE,        /* 104 These are different to the three sets above. */
2024      OP_CRMINRANGE,     /* 105 */
2025    
2026      /* End of quantifier opcodes */
2027    
2028      OP_CLASS,          /* 106 Match a character class, chars < 256 only */
2029      OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
2030                                  class - the difference is relevant only when a
2031                                  character > 255 is encountered. */
2032      OP_XCLASS,         /* 108 Extended class for handling > 255 chars within the
2033                                  class. This does both positive and negative. */
2034      OP_REF,            /* 109 Match a back reference, casefully */
2035      OP_REFI,           /* 110 Match a back reference, caselessly */
2036      OP_RECURSE,        /* 111 Match a numbered subpattern (possibly recursive) */
2037      OP_CALLOUT,        /* 112 Call out to external function if provided */
2038    
2039      OP_ALT,            /* 113 Start of alternation */
2040      OP_KET,            /* 114 End of group that doesn't have an unbounded repeat */
2041      OP_KETRMAX,        /* 115 These two must remain together and in this */
2042      OP_KETRMIN,        /* 116 order. They are for groups the repeat for ever. */
2043      OP_KETRPOS,        /* 117 Possessive unlimited repeat. */
2044    
2045      /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
2046      asserts must remain in order. */
2047    
2048      OP_REVERSE,        /* 118 Move pointer back - used in lookbehind assertions */
2049      OP_ASSERT,         /* 119 Positive lookahead */
2050      OP_ASSERT_NOT,     /* 120 Negative lookahead */
2051      OP_ASSERTBACK,     /* 121 Positive lookbehind */
2052      OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
2053    
2054      /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
2055      after the assertions, with ONCE first, as there's a test for >= ONCE for a
2056      subpattern that isn't an assertion. The POS versions must immediately follow
2057      the non-POS versions in each case. */
2058    
2059      OP_ONCE,           /* 123 Atomic group, contains captures */
2060      OP_ONCE_NC,        /* 124 Atomic group containing no captures */
2061      OP_BRA,            /* 125 Start of non-capturing bracket */
2062      OP_BRAPOS,         /* 126 Ditto, with unlimited, possessive repeat */
2063      OP_CBRA,           /* 127 Start of capturing bracket */
2064      OP_CBRAPOS,        /* 128 Ditto, with unlimited, possessive repeat */
2065      OP_COND,           /* 129 Conditional group */
2066    
2067    /* These three must follow the previous three, in the same order. There's a    /* These five must follow the previous five, in the same order. There's a
2068    check for >= SBRA to distinguish the two sets. */    check for >= SBRA to distinguish the two sets. */
2069    
2070    OP_SBRA,           /* 97 Start of non-capturing bracket, check empty  */    OP_SBRA,           /* 130 Start of non-capturing bracket, check empty  */
2071    OP_SCBRA,          /* 98 Start of capturing bracket, check empty */    OP_SBRAPOS,        /* 131 Ditto, with unlimited, possessive repeat */
2072    OP_SCOND,          /* 99 Conditional group, check empty */    OP_SCBRA,          /* 132 Start of capturing bracket, check empty */
2073      OP_SCBRAPOS,       /* 133 Ditto, with unlimited, possessive repeat */
2074      OP_SCOND,          /* 134 Conditional group, check empty */
2075    
2076    /* The next two pairs must (respectively) be kept together. */    /* The next two pairs must (respectively) be kept together. */
2077    
2078    OP_CREF,           /* 100 Used to hold a capture number as condition */    OP_CREF,           /* 135 Used to hold a capture number as condition */
2079    OP_NCREF,          /* 101 Same, but generaged by a name reference*/    OP_NCREF,          /* 136 Same, but generated by a name reference*/
2080    OP_RREF,           /* 102 Used to hold a recursion number as condition */    OP_RREF,           /* 137 Used to hold a recursion number as condition */
2081    OP_NRREF,          /* 103 Same, but generaged by a name reference*/    OP_NRREF,          /* 138 Same, but generated by a name reference*/
2082    OP_DEF,            /* 104 The DEFINE condition */    OP_DEF,            /* 139 The DEFINE condition */
2083    
2084    OP_BRAZERO,        /* 105 These two must remain together and in this */    OP_BRAZERO,        /* 140 These two must remain together and in this */
2085    OP_BRAMINZERO,     /* 106 order. */    OP_BRAMINZERO,     /* 141 order. */
2086      OP_BRAPOSZERO,     /* 142 */
2087    
2088    /* These are backtracking control verbs */    /* These are backtracking control verbs */
2089    
2090    OP_MARK,           /* 107 always has an argument */    OP_MARK,           /* 143 always has an argument */
2091    OP_PRUNE,          /* 108 */    OP_PRUNE,          /* 144 */
2092    OP_PRUNE_ARG,      /* 109 same, but with argument */    OP_PRUNE_ARG,      /* 145 same, but with argument */
2093    OP_SKIP,           /* 110 */    OP_SKIP,           /* 146 */
2094    OP_SKIP_ARG,       /* 111 same, but with argument */    OP_SKIP_ARG,       /* 147 same, but with argument */
2095    OP_THEN,           /* 112 */    OP_THEN,           /* 148 */
2096    OP_THEN_ARG,       /* 113 same, but with argument */    OP_THEN_ARG,       /* 149 same, but with argument */
2097    OP_COMMIT,         /* 114 */    OP_COMMIT,         /* 150 */
2098    
2099    /* These are forced failure and success verbs */    /* These are forced failure and success verbs */
2100    
2101    OP_FAIL,           /* 115 */    OP_FAIL,           /* 151 */
2102    OP_ACCEPT,         /* 116 */    OP_ACCEPT,         /* 152 */
2103    OP_CLOSE,          /* 117 Used before OP_ACCEPT to close open captures */    OP_ASSERT_ACCEPT,  /* 153 Used inside assertions */
2104      OP_CLOSE,          /* 154 Used before OP_ACCEPT to close open captures */
2105    
2106    /* This is used to skip a subpattern with a {0} quantifier */    /* This is used to skip a subpattern with a {0} quantifier */
2107    
2108    OP_SKIPZERO,       /* 118 */    OP_SKIPZERO,       /* 155 */
2109    
2110    /* This is not an opcode, but is used to check that tables indexed by opcode    /* This is not an opcode, but is used to check that tables indexed by opcode
2111    are the correct length, in order to catch updating errors - there have been    are the correct length, in order to catch updating errors - there have been
# Line 1462  called "coptable" and "poptable" in pcre Line 2120  called "coptable" and "poptable" in pcre
2120    
2121    
2122  /* This macro defines textual names for all the opcodes. These are used only  /* This macro defines textual names for all the opcodes. These are used only
2123  for debugging. The macro is referenced only in pcre_printint.c. */  for debugging, and some of them are only partial names. The macro is referenced
2124    only in pcre_printint.c, which fills out the full names in many cases (and in
2125    some cases doesn't actually use these names at all). */
2126    
2127  #define OP_NAME_LIST \  #define OP_NAME_LIST \
2128    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
2129    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
2130    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
2131    "extuni",  "\\Z", "\\z",                                        \    "extuni",  "\\Z", "\\z",                                        \
2132    "Opt", "^", "$", "char", "charnc", "not",                       \    "^", "^", "$", "$", "char", "chari", "not", "noti",             \
2133    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2134      "{", "{", "{",                                                  \
2135    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2136    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2137      "{", "{", "{",                                                  \
2138      "*+","++", "?+", "{",                                           \
2139      "*", "*?", "+", "+?", "?", "??",                                \
2140      "{", "{", "{",                                                  \
2141      "*+","++", "?+", "{",                                           \
2142      "*", "*?", "+", "+?", "?", "??",                                \
2143      "{", "{", "{",                                                  \
2144    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2145    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
2146    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2147    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \
2148    "class", "nclass", "xclass", "Ref", "Recurse", "Callout",       \    "class", "nclass", "xclass", "Ref", "Refi",                     \
2149    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",     \    "Recurse", "Callout",                                           \
2150    "AssertB", "AssertB not", "Reverse",                            \    "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos",                  \
2151    "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond",        \    "Reverse", "Assert", "Assert not", "AssertB", "AssertB not",    \
2152      "Once", "Once_NC",                                              \
2153      "Bra", "BraPos", "CBra", "CBraPos",                             \
2154      "Cond",                                                         \
2155      "SBra", "SBraPos", "SCBra", "SCBraPos",                         \
2156      "SCond",                                                        \
2157    "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \    "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \
2158    "Brazero", "Braminzero",                                        \    "Brazero", "Braminzero", "Braposzero",                          \
2159    "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \    "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \
2160    "*THEN", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT",                \    "*THEN", "*THEN", "*COMMIT", "*FAIL",                           \
2161      "*ACCEPT", "*ASSERT_ACCEPT",                                    \
2162    "Close", "Skip zero"    "Close", "Skip zero"
2163    
2164    
# Line 1505  in UTF-8 mode. The code that uses this t Line 2179  in UTF-8 mode. The code that uses this t
2179    3, 3,                          /* \P, \p                                 */ \    3, 3,                          /* \P, \p                                 */ \
2180    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
2181    1,                             /* \X                                     */ \    1,                             /* \X                                     */ \
2182    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    1, 1, 1, 1, 1, 1,              /* \Z, \z, ^, ^M, $, $M                   */ \
2183    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
2184    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Chari  - the minimum length            */ \
2185    2,                             /* not                                    */ \    2,                             /* not                                    */ \
2186    /* Positive single-char repeats                            ** These are  */ \    2,                             /* noti                                   */ \
2187    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??      ** minima in  */ \    /* Positive single-char repeats                             ** These are */ \
2188    4, 4, 4,                       /* upto, minupto, exact     ** UTF-8 mode */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2189    2, 2, 2, 4,                    /* *+, ++, ?+, upto+                      */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2190      2+IMM2_SIZE,                   /* exact                                  */ \
2191      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2192      2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2193      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2194      2+IMM2_SIZE,                   /* exact I                                */ \
2195      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2196    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2197    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2198    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2199    2, 2, 2, 4,                    /* Possessive *, +, ?, upto               */ \    2+IMM2_SIZE,                   /* NOT exact                              */ \
2200      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2201      2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2202      2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2203      2+IMM2_SIZE,                   /* NOT exact I                            */ \
2204      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2205    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2206    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2207    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2208    2, 2, 2, 4,                    /* Possessive *+, ++, ?+, upto+           */ \    2+IMM2_SIZE,                   /* Type exact                             */ \
2209      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2210    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2211    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2212    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2213   33,                             /* CLASS                                  */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2214   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2215    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2216    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2217      1+IMM2_SIZE,                   /* REFI                                   */ \
2218    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2219    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2220    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
2221    1+LINK_SIZE,                   /* Ket                                    */ \    1+LINK_SIZE,                   /* Ket                                    */ \
2222    1+LINK_SIZE,                   /* KetRmax                                */ \    1+LINK_SIZE,                   /* KetRmax                                */ \
2223    1+LINK_SIZE,                   /* KetRmin                                */ \    1+LINK_SIZE,                   /* KetRmin                                */ \
2224      1+LINK_SIZE,                   /* KetRpos                                */ \
2225      1+LINK_SIZE,                   /* Reverse                                */ \
2226    1+LINK_SIZE,                   /* Assert                                 */ \    1+LINK_SIZE,                   /* Assert                                 */ \
2227    1+LINK_SIZE,                   /* Assert not                             */ \    1+LINK_SIZE,                   /* Assert not                             */ \
2228    1+LINK_SIZE,                   /* Assert behind                          */ \    1+LINK_SIZE,                   /* Assert behind                          */ \
2229    1+LINK_SIZE,                   /* Assert behind not                      */ \    1+LINK_SIZE,                   /* Assert behind not                      */ \
   1+LINK_SIZE,                   /* Reverse                                */ \  
2230    1+LINK_SIZE,                   /* ONCE                                   */ \    1+LINK_SIZE,                   /* ONCE                                   */ \
2231      1+LINK_SIZE,                   /* ONCE_NC                                */ \
2232    1+LINK_SIZE,                   /* BRA                                    */ \    1+LINK_SIZE,                   /* BRA                                    */ \
2233    3+LINK_SIZE,                   /* CBRA                                   */ \    1+LINK_SIZE,                   /* BRAPOS                                 */ \
2234      1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2235      1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2236    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2237    1+LINK_SIZE,                   /* SBRA                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2238    3+LINK_SIZE,                   /* SCBRA                                  */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2239      1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2240      1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2241    1+LINK_SIZE,                   /* SCOND                                  */ \    1+LINK_SIZE,                   /* SCOND                                  */ \
2242    3, 3,                          /* CREF, NCREF                            */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* CREF, NCREF                            */ \
2243    3, 3,                          /* RREF, NRREF                            */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* RREF, NRREF                            */ \
2244    1,                             /* DEF                                    */ \    1,                             /* DEF                                    */ \
2245    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2246    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2247    1, 3,                          /* SKIP, SKIP_ARG                         */ \    1, 3,                          /* SKIP, SKIP_ARG                         */ \
2248    1+LINK_SIZE, 3+LINK_SIZE,      /* THEN, THEN_ARG                         */ \    1, 3,                          /* THEN, THEN_ARG                         */ \
2249    1, 1, 1, 3, 1                  /* COMMIT, FAIL, ACCEPT, CLOSE, SKIPZERO  */    1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2250      1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2251    
2252  /* 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"
2253  condition. */  condition. */
# Line 1571  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2264  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2264         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2265         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2266         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2267         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68,         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2268         ERRCOUNT };         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2269    
2270    /* JIT compiling modes. The function list is indexed by them. */
2271    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2272           JIT_NUMBER_OF_COMPILE_MODES };
2273    
2274  /* 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
2275  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 1591  fields are present. Currently PCRE alway Line 2288  fields are present. Currently PCRE alway
2288  NOTE NOTE NOTE  NOTE NOTE NOTE
2289  */  */
2290    
2291  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2292    #define REAL_PCRE real_pcre
2293    #elif defined COMPILE_PCRE16
2294    #define REAL_PCRE real_pcre16
2295    #elif defined COMPILE_PCRE32
2296    #define REAL_PCRE real_pcre32
2297    #endif
2298    
2299    /* It is necessary to fork the struct for 32 bit, since it needs to use
2300     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2301     * typedef since pcretest needs access to  the struct of the 8-, 16-
2302     * and 32-bit variants. */
2303    
2304    typedef struct real_pcre8_or_16 {
2305    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2306    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2307    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
2308    pcre_uint16 flags;              /* Private flags */    pcre_uint16 flags;              /* Private flags */
2309    pcre_uint16 dummy1;             /* For future use */    pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2310    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2311    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2312    pcre_uint16 first_byte;    pcre_uint16 first_char;         /* Starting character */
2313    pcre_uint16 req_byte;    pcre_uint16 req_char;           /* This character must be seen */
2314    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2315    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2316    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2317    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2318      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2319      const pcre_uint8 *nullpad;      /* NULL padding */
2320    } real_pcre8_or_16;
2321    
2322    const unsigned char *tables;    /* Pointer to tables or NULL for std */  typedef struct real_pcre8_or_16 real_pcre;
2323    const unsigned char *nullpad;   /* NULL padding */  typedef struct real_pcre8_or_16 real_pcre16;
2324  } real_pcre;  
2325    typedef struct real_pcre32 {
2326      pcre_uint32 magic_number;
2327      pcre_uint32 size;               /* Total that was malloced */
2328      pcre_uint32 options;            /* Public options */
2329      pcre_uint16 flags;              /* Private flags */
2330      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2331      pcre_uint16 top_bracket;        /* Highest numbered group */
2332      pcre_uint16 top_backref;        /* Highest numbered back reference */
2333      pcre_uint32 first_char;         /* Starting character */
2334      pcre_uint32 req_char;           /* This character must be seen */
2335      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2336      pcre_uint16 name_entry_size;    /* Size of any name items */
2337      pcre_uint16 name_count;         /* Number of name items */
2338      pcre_uint16 ref_count;          /* Reference count */
2339      pcre_uint16 dummy1;             /* for later expansion */
2340      pcre_uint16 dummy2;             /* for later expansion */
2341      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2342      void *nullpad;                  /* for later expansion */
2343    } real_pcre32;
2344    
2345    /* Assert that the size of REAL_PCRE is divisible by 8 */
2346    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2347    
2348    /* Needed in pcretest to access some fields in the real_pcre* structures
2349     * directly. They're unified for 8/16/32 bits since the structs only differ
2350     * after these fields; if that ever changes, need to fork those defines into
2351     * 8/16 and 32 bit versions. */
2352    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2353    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2354    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2355    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2356    
2357  /* 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
2358  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 1616  remark (see NOTE above) about extending Line 2360  remark (see NOTE above) about extending
2360  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2361    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2362    pcre_uint32 flags;              /* Private flags */    pcre_uint32 flags;              /* Private flags */
2363    uschar start_bits[32];          /* Starting char bits */    pcre_uint8 start_bits[32];      /* Starting char bits */
2364    pcre_uint32 minlength;          /* Minimum subject length */    pcre_uint32 minlength;          /* Minimum subject length */
2365  } pcre_study_data;  } pcre_study_data;
2366    
# Line 1635  typedef struct open_capitem { Line 2379  typedef struct open_capitem {
2379  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2380    
2381  typedef struct compile_data {  typedef struct compile_data {
2382    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2383    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2384    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2385    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2386    const uschar *start_workspace;/* The start of working space */    const pcre_uchar *start_workspace;/* The start of working space */
2387    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_code;     /* The start of the compiled code */
2388    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2389    const uschar *end_pattern;    /* The end of the pattern */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2390    open_capitem *open_caps;      /* Chain of open capture items */    open_capitem *open_caps;          /* Chain of open capture items */
2391    uschar *hwm;                  /* High watermark of workspace */    pcre_uchar *hwm;                  /* High watermark of workspace */
2392    uschar *name_table;           /* The name/number table */    pcre_uchar *name_table;           /* The name/number table */
2393    int  names_found;             /* Number of entries so far */    int  names_found;                 /* Number of entries so far */
2394    int  name_entry_size;         /* Size of each entry */    int  name_entry_size;             /* Size of each entry */
2395    int  bracount;                /* Count of capturing parens as we compile */    int  workspace_size;              /* Size of workspace */
2396    int  final_bracount;          /* Saved value after first pass */    int  bracount;                    /* Count of capturing parens as we compile */
2397    int  top_backref;             /* Maximum back reference */    int  final_bracount;              /* Saved value after first pass */
2398    unsigned int backref_map;     /* Bitmap of low back refs */    int  max_lookbehind;              /* Maximum lookbehind (characters) */
2399    int  external_options;        /* External (initial) options */    int  top_backref;                 /* Maximum back reference */
2400    int  external_flags;          /* External flag bits to be set */    unsigned int backref_map;         /* Bitmap of low back refs */
2401    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  assert_depth;                /* Depth of nested assertions */
2402    BOOL had_accept;              /* (*ACCEPT) encountered */    int  external_options;            /* External (initial) options */
2403    BOOL check_lookbehind;        /* Lookbehinds need later checking */    int  external_flags;              /* External flag bits to be set */
2404    int  nltype;                  /* Newline type */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2405    int  nllen;                   /* Newline string length */    BOOL had_accept;                  /* (*ACCEPT) encountered */
2406    uschar nl[4];                 /* Newline string when fixed length */    BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2407      BOOL check_lookbehind;            /* Lookbehinds need later checking */
2408      int  nltype;                      /* Newline type */
2409      int  nllen;                       /* Newline string length */
2410      pcre_uchar nl[4];                 /* Newline string when fixed length */
2411  } compile_data;  } compile_data;
2412    
2413  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
2414  branches, for testing for left recursion. */  branches, for testing for left recursion while compiling. */
2415    
2416  typedef struct branch_chain {  typedef struct branch_chain {
2417    struct branch_chain *outer;    struct branch_chain *outer;
2418    uschar *current_branch;    pcre_uchar *current_branch;
2419  } branch_chain;  } branch_chain;
2420    
2421  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
2422  call within the pattern. */  call within the pattern; used by pcre_exec(). */
2423    
2424  typedef struct recursion_info {  typedef struct recursion_info {
2425    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2426    int group_num;                /* Number of group that was called */    int group_num;                  /* Number of group that was called */
2427    const uschar *after_call;     /* "Return value": points after the call in the expr */    int *offset_save;               /* Pointer to start of saved offsets */
2428    int *offset_save;             /* Pointer to start of saved offsets */    int saved_max;                  /* Number of saved offsets */
2429    int saved_max;                /* Number of saved offsets */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
   int save_offset_top;          /* Current value of offset_top */  
2430  } recursion_info;  } recursion_info;
2431    
2432    /* A similar structure for pcre_dfa_exec(). */
2433    
2434    typedef struct dfa_recursion_info {
2435      struct dfa_recursion_info *prevrec;
2436      int group_num;
2437      PCRE_PUCHAR subject_position;
2438    } dfa_recursion_info;
2439    
2440  /* 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
2441  pointer at the start of each subpattern, so as to detect when an empty string  pointer at the start of each subpattern, so as to detect when an empty string
2442  has been matched by a subpattern - to break infinite loops. */  has been matched by a subpattern - to break infinite loops; used by
2443    pcre_exec(). */
2444    
2445  typedef struct eptrblock {  typedef struct eptrblock {
2446    struct eptrblock *epb_prev;    struct eptrblock *epb_prev;
2447    USPTR epb_saved_eptr;    PCRE_PUCHAR epb_saved_eptr;
2448  } eptrblock;  } eptrblock;
2449    
2450    
# Line 1699  typedef struct match_data { Line 2455  typedef struct match_data {
2455    unsigned long int match_call_count;      /* As it says */    unsigned long int match_call_count;      /* As it says */
2456    unsigned long int match_limit;           /* As it says */    unsigned long int match_limit;           /* As it says */
2457    unsigned long int match_limit_recursion; /* As it says */    unsigned long int match_limit_recursion; /* As it says */
2458    int   *offset_vector;         /* Offset vector */    int   *offset_vector;           /* Offset vector */
2459    int    offset_end;            /* One past the end */    int    offset_end;              /* One past the end */
2460    int    offset_max;            /* The maximum usable for return data */    int    offset_max;              /* The maximum usable for return data */
2461    int    nltype;                /* Newline type */    int    nltype;                  /* Newline type */
2462    int    nllen;                 /* Newline string length */    int    nllen;                   /* Newline string length */
2463    int    name_count;            /* Number of names in name table */    int    name_count;              /* Number of names in name table */
2464    int    name_entry_size;       /* Size of entry in names table */    int    name_entry_size;         /* Size of entry in names table */
2465    uschar *name_table;           /* Table of names */    pcre_uchar *name_table;         /* Table of names */
2466    uschar nl[4];                 /* Newline string when fixed */    pcre_uchar nl[4];               /* Newline string when fixed */
2467    const uschar *lcc;            /* Points to lower casing table */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2468    const uschar *ctypes;         /* Points to table of type maps */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2469    BOOL   offset_overflow;       /* Set if too many extractions */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2470    BOOL   notbol;                /* NOTBOL flag */    BOOL   offset_overflow;         /* Set if too many extractions */
2471    BOOL   noteol;                /* NOTEOL flag */    BOOL   notbol;                  /* NOTBOL flag */
2472    BOOL   utf8;                  /* UTF8 flag */    BOOL   noteol;                  /* NOTEOL flag */
2473    BOOL   jscript_compat;        /* JAVASCRIPT_COMPAT flag */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2474    BOOL   use_ucp;               /* PCRE_UCP flag */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2475    BOOL   endonly;               /* Dollar not before final \n */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2476    BOOL   notempty;              /* Empty string match not wanted */    BOOL   endonly;                 /* Dollar not before final \n */
2477    BOOL   notempty_atstart;      /* Empty string match at start not wanted */    BOOL   notempty;                /* Empty string match not wanted */
2478    BOOL   hitend;                /* Hit the end of the subject at some point */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2479    BOOL   bsr_anycrlf;           /* \R is just any CRLF, not full Unicode */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2480    const uschar *start_code;     /* For use when recursing */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2481    USPTR  start_subject;         /* Start of the subject string */    BOOL   hasthen;                 /* Pattern contains (*THEN) */
2482    USPTR  end_subject;           /* End of the subject string */    BOOL   ignore_skip_arg;         /* For re-run when SKIP name not found */
2483    USPTR  start_match_ptr;       /* Start of matched string */    const  pcre_uchar *start_code;  /* For use when recursing */
2484    USPTR  end_match_ptr;         /* Subject position at end match */    PCRE_PUCHAR start_subject;      /* Start of the subject string */
2485    USPTR  start_used_ptr;        /* Earliest consulted character */    PCRE_PUCHAR end_subject;        /* End of the subject string */
2486    int    partial;               /* PARTIAL options */    PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2487    int    end_offset_top;        /* Highwater mark at end of match */    PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2488    int    capture_last;          /* Most recent capture number */    PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2489    int    start_offset;          /* The start offset value */    int    partial;                 /* PARTIAL options */
2490    eptrblock *eptrchain;         /* Chain of eptrblocks for tail recursions */    int    end_offset_top;          /* Highwater mark at end of match */
2491    int    eptrn;                 /* Next free eptrblock */    int    capture_last;            /* Most recent capture number */
2492    recursion_info *recursive;    /* Linked list of recursion data */    int    start_offset;            /* The start offset value */
2493    void  *callout_data;          /* To pass back to callouts */    int    match_function_type;     /* Set for certain special calls of MATCH() */
2494    const uschar *mark;           /* Mark pointer to pass back */    eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2495      int    eptrn;                   /* Next free eptrblock */
2496      recursion_info *recursive;      /* Linked list of recursion data */
2497      void  *callout_data;            /* To pass back to callouts */
2498      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2499      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2500      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2501    #ifdef NO_RECURSE
2502      void  *match_frames_base;       /* For remembering malloc'd frames */
2503    #endif
2504  } match_data;  } match_data;
2505    
2506  /* 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
2507  functions. */  functions. */
2508    
2509  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2510    const uschar *start_code;     /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2511    const uschar *start_subject;  /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2512    const uschar *end_subject;    /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2513    const uschar *start_used_ptr; /* Earliest consulted character */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2514    const uschar *tables;         /* Character tables */    const pcre_uint8 *tables;         /* Character tables */
2515    int   start_offset;           /* The start offset value */    int   start_offset;               /* The start offset value */
2516    int   moptions;               /* Match options */    int   moptions;                   /* Match options */
2517    int   poptions;               /* Pattern options */    int   poptions;                   /* Pattern options */
2518    int    nltype;                /* Newline type */    int   nltype;                     /* Newline type */
2519    int    nllen;                 /* Newline string length */    int   nllen;                      /* Newline string length */
2520    uschar nl[4];                 /* Newline string when fixed */    pcre_uchar nl[4];                 /* Newline string when fixed */
2521    void  *callout_data;          /* To pass back to callouts */    void *callout_data;               /* To pass back to callouts */
2522      dfa_recursion_info *recursive;    /* Linked list of recursion data */
2523  } dfa_match_data;  } dfa_match_data;
2524    
2525  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 1789  total length. */ Line 2555  total length. */
2555  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2556  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2557    
2558    /* Internal function and data prefixes. */
2559    
2560    #if defined COMPILE_PCRE8
2561    #ifndef PUBL
2562    #define PUBL(name) pcre_##name
2563    #endif
2564    #ifndef PRIV
2565    #define PRIV(name) _pcre_##name
2566    #endif
2567    #elif defined COMPILE_PCRE16
2568    #ifndef PUBL
2569    #define PUBL(name) pcre16_##name
2570    #endif
2571    #ifndef PRIV
2572    #define PRIV(name) _pcre16_##name
2573    #endif
2574    #elif defined COMPILE_PCRE32
2575    #ifndef PUBL
2576    #define PUBL(name) pcre32_##name
2577    #endif
2578    #ifndef PRIV
2579    #define PRIV(name) _pcre32_##name
2580    #endif
2581    #else
2582    #error Unsupported compiling mode
2583    #endif /* COMPILE_PCRE[8|16|32] */
2584    
2585  /* 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
2586  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
2587  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 1806  of the exported public functions. They h Line 2599  of the exported public functions. They h
2599  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
2600  pcre_tables.c module. */  pcre_tables.c module. */
2601    
2602  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2603  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2604  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2605  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2606    extern const int            PRIV(utf8_table3)[];
2607  extern const int    _pcre_utf8_table1_size;  extern const pcre_uint8     PRIV(utf8_table4)[];
2608    #endif /* COMPILE_PCRE8 */
2609    
2610    extern const char           PRIV(utt_names)[];
2611    extern const ucp_type_table PRIV(utt)[];
2612    extern const int            PRIV(utt_size);
2613    
2614  extern const char   _pcre_utt_names[];  extern const pcre_uint8     PRIV(OP_lengths)[];
2615  extern const ucp_type_table _pcre_utt[];  extern const pcre_uint8     PRIV(default_tables)[];
 extern const int _pcre_utt_size;  
2616    
2617  extern const uschar _pcre_default_tables[];  extern const pcre_uint32    PRIV(hspace_list)[];
2618    extern const pcre_uint32    PRIV(vspace_list)[];
 extern const uschar _pcre_OP_lengths[];  
2619    
2620    
2621  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2622  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
2623  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2624    
2625  extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);  /* String comparison functions. */
2626  extern BOOL          _pcre_is_newline(USPTR, int, USPTR, int *, BOOL);  #if defined COMPILE_PCRE8
 extern int           _pcre_ord2utf8(int, uschar *);  
 extern real_pcre    *_pcre_try_flipped(const real_pcre *, real_pcre *,  
                        const pcre_study_data *, pcre_study_data *);  
 extern int           _pcre_valid_utf8(USPTR, int);  
 extern BOOL          _pcre_was_newline(USPTR, int, USPTR, int *, BOOL);  
 extern BOOL          _pcre_xclass(int, const uschar *);  
2627    
2628    #define STRCMP_UC_UC(str1, str2) \
2629      strcmp((char *)(str1), (char *)(str2))
2630    #define STRCMP_UC_C8(str1, str2) \
2631      strcmp((char *)(str1), (str2))
2632    #define STRNCMP_UC_UC(str1, str2, num) \
2633      strncmp((char *)(str1), (char *)(str2), (num))
2634    #define STRNCMP_UC_C8(str1, str2, num) \
2635      strncmp((char *)(str1), (str2), (num))
2636    #define STRLEN_UC(str) strlen((const char *)str)
2637    
2638    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2639    
2640    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2641                               const pcre_uchar *);
2642    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2643                               const char *);
2644    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2645                               const pcre_uchar *, unsigned int num);
2646    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2647                               const char *, unsigned int num);
2648    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2649    
2650    #define STRCMP_UC_UC(str1, str2) \
2651      PRIV(strcmp_uc_uc)((str1), (str2))
2652    #define STRCMP_UC_C8(str1, str2) \
2653      PRIV(strcmp_uc_c8)((str1), (str2))
2654    #define STRNCMP_UC_UC(str1, str2, num) \
2655      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2656    #define STRNCMP_UC_C8(str1, str2, num) \
2657      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2658    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2659    
2660    #endif /* COMPILE_PCRE[8|16|32] */
2661    
2662    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2663    
2664    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2665    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2666    
2667    #elif defined COMPILE_PCRE32
2668    
2669    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2670                               const pcre_uchar *);
2671    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2672                               const char *);
2673    
2674    #define STRCMP_UC_UC_TEST(str1, str2) \
2675      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2676    #define STRCMP_UC_C8_TEST(str1, str2) \
2677      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2678    
2679    #endif /* COMPILE_PCRE[8|16|32] */
2680    
2681    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2682    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2683                               int *, BOOL);
2684    extern int               PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2685    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2686    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2687                               int *, BOOL);
2688    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2689    
2690    #ifdef SUPPORT_JIT
2691    extern void              PRIV(jit_compile)(const REAL_PCRE *,
2692                               PUBL(extra) *, int);
2693    extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2694                               const pcre_uchar *, int, int, int, int *, int);
2695    extern void              PRIV(jit_free)(void *);
2696    extern int               PRIV(jit_get_size)(void *);
2697    extern const char*       PRIV(jit_get_target)(void);
2698    #endif
2699    
2700  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
2701    
2702  typedef struct {  typedef struct {
2703    uschar script;    pcre_uint8 script;     /* ucp_Arabic, etc. */
2704    uschar chartype;    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2705    pcre_int32 other_case;    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2706      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2707      pcre_int32 other_case; /* offset to other case, or zero if none */
2708  } ucd_record;  } ucd_record;
2709    
2710  extern const ucd_record  _pcre_ucd_records[];  extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2711  extern const uschar      _pcre_ucd_stage1[];  extern const ucd_record  PRIV(ucd_records)[];
2712  extern const pcre_uint16 _pcre_ucd_stage2[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2713  extern const int         _pcre_ucp_gentype[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2714    extern const pcre_uint32 PRIV(ucp_gentype)[];
2715    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2716    #ifdef SUPPORT_JIT
2717    extern const int         PRIV(ucp_typerange)[];
2718    #endif
2719    
2720    #ifdef SUPPORT_UCP
2721  /* UCD access macros */  /* UCD access macros */
2722    
2723  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2724  #define GET_UCD(ch) (_pcre_ucd_records + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2725          _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2726          UCD_BLOCK_SIZE + ch % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2727    
2728  #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2729  #define UCD_SCRIPT(ch)    GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2730  #define UCD_CATEGORY(ch)  _pcre_ucp_gentype[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2731  #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2732    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2733    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2734    
2735    #endif /* SUPPORT_UCP */
2736    
2737  #endif  #endif
2738    

Legend:
Removed from v.576  
changed lines
  Added in v.1128

  ViewVC Help
Powered by ViewVC 1.1.5