/[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 412 by ph10, Sat Apr 11 10:34:37 2009 UTC revision 1078 by chpe, Tue Oct 16 15:55:00 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) 1997-2009 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
48    
49  /* Define DEBUG to get debugging output on stdout. */  /* Define PCRE_DEBUG to get debugging output on stdout. */
50    
51  #if 0  #if 0
52  #define 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 74  It turns out that the Mac Debugging.h he Line 94  It turns out that the Mac Debugging.h he
94  be absolutely sure we get our version. */  be absolutely sure we get our version. */
95    
96  #undef DPRINTF  #undef DPRINTF
97  #ifdef DEBUG  #ifdef PCRE_DEBUG
98  #define DPRINTF(p) printf p  #define DPRINTF(p) printf p
99  #else  #else
100  #define DPRINTF(p) /* Nothing */  #define DPRINTF(p) /* Nothing */
# Line 86  setjmp and stdarg are used is when NO_RE Line 106  setjmp and stdarg are used is when NO_RE
106    
107  #include <ctype.h>  #include <ctype.h>
108  #include <limits.h>  #include <limits.h>
 #include <setjmp.h>  
 #include <stdarg.h>  
109  #include <stddef.h>  #include <stddef.h>
110  #include <stdio.h>  #include <stdio.h>
111  #include <stdlib.h>  #include <stdlib.h>
# Line 160  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 186  preprocessor time in standard C environm Line 206  preprocessor time in standard C environm
206    #error Cannot determine a type for 32-bit unsigned integers    #error Cannot determine a type for 32-bit unsigned integers
207  #endif  #endif
208    
209    /* When checking for integer overflow in pcre_compile(), we need to handle
210    large integers. If a 64-bit integer type is available, we can use that.
211    Otherwise we have to cast to double, which of course requires floating point
212    arithmetic. Handle this by defining a macro for the appropriate type. If
213    stdint.h is available, include it; it may define INT64_MAX. Systems that do not
214    have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
215    by "configure". */
216    
217    #if HAVE_STDINT_H
218    #include <stdint.h>
219    #elif HAVE_INTTYPES_H
220    #include <inttypes.h>
221    #endif
222    
223    #if defined INT64_MAX || defined int64_t
224    #define INT64_OR_DOUBLE int64_t
225    #else
226    #define INT64_OR_DOUBLE double
227    #endif
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 uschar;  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 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 216  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] && \       (p)[0] == NLBLOCK->nl[0] && \
# Line 230  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] && \       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \
# Line 249  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 325  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 361  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 378  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 385  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. BACKCHAR should  
536  never be called in byte mode. To make sure it can never even appear when UTF-8  #define GET2(a,n) \
537  support is omitted, we don't even define it. */     a[n]
538    
539    #elif defined COMPILE_PCRE32
540    
541    #define IMM2_SIZE 1
542    
543  #ifndef SUPPORT_UTF8  #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 GETCHARLENTEST(c, eptr, len) */
583  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
584    /* #define FORWARDCHAR(eptr) */
585    /* #define ACROSSCHAR(condition, eptr, action) */
586    
587    #else   /* SUPPORT_UTF */
588    
589    #if defined COMPILE_PCRE8
590    
591    /* These macros were originally written in the form of loops that used data
592    from the tables whose names start with PRIV(utf8_table). They were rewritten by
593    a user so as not to use loops, because in some environments this gives a
594    significant performance advantage, and it seems never to do any harm. */
595    
596    /* Tells the biggest code point which can be encoded as a single character. */
597    
598    #define MAX_VALUE_FOR_SINGLE_CHAR 127
599    
600    /* Tests whether the code point needs extra characters to decode. */
601    
602    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
603    
604  #else   /* SUPPORT_UTF8 */  /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
605    Otherwise it has an undefined behaviour. */
606    
607    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
608    
609    /* Returns TRUE, if the given character is not the first character
610    of a UTF sequence. */
611    
612    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
613    
614    /* Base macro to pick up the remaining bytes of a UTF-8 character, not
615    advancing the pointer. */
616    
617    #define GETUTF8(c, eptr) \
618        { \
619        if ((c & 0x20) == 0) \
620          c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
621        else if ((c & 0x10) == 0) \
622          c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
623        else if ((c & 0x08) == 0) \
624          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
625          ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
626        else if ((c & 0x04) == 0) \
627          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
628              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
629              (eptr[4] & 0x3f); \
630        else \
631          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
632              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
633              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
634        }
635    
636  /* Get the next UTF-8 character, not advancing the pointer. This is called when  /* Get the next UTF-8 character, not advancing the pointer. This is called when
637  we know we are in UTF-8 mode. */  we know we are in UTF-8 mode. */
638    
639  #define GETCHAR(c, eptr) \  #define GETCHAR(c, eptr) \
640    c = *eptr; \    c = *eptr; \
641    if (c >= 0xc0) \    if (c >= 0xc0) GETUTF8(c, eptr);
     { \  
     int gcii; \  
     int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     for (gcii = 1; gcii <= gcaa; gcii++) \  
       { \  
       gcss -= 6; \  
       c |= (eptr[gcii] & 0x3f) << gcss; \  
       } \  
     }  
642    
643  /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the  /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
644  pointer. */  pointer. */
645    
646  #define GETCHARTEST(c, eptr) \  #define GETCHARTEST(c, eptr) \
647    c = *eptr; \    c = *eptr; \
648    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8(c, eptr);
649    
650    /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
651    the pointer. */
652    
653    #define GETUTF8INC(c, eptr) \
654      { \      { \
655      int gcii; \      if ((c & 0x20) == 0) \
656      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \        c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
657      int gcss = 6*gcaa; \      else if ((c & 0x10) == 0) \
658      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        { \
659      for (gcii = 1; gcii <= gcaa; gcii++) \        c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
660          eptr += 2; \
661          } \
662        else if ((c & 0x08) == 0) \
663          { \
664          c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
665              ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
666          eptr += 3; \
667          } \
668        else if ((c & 0x04) == 0) \
669          { \
670          c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
671              ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
672              (eptr[3] & 0x3f); \
673          eptr += 4; \
674          } \
675        else \
676        { \        { \
677        gcss -= 6; \        c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
678        c |= (eptr[gcii] & 0x3f) << gcss; \            ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
679              ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
680          eptr += 5; \
681        } \        } \
682      }      }
683    
# Line 445  know we are in UTF-8 mode. */ Line 686  know we are in UTF-8 mode. */
686    
687  #define GETCHARINC(c, eptr) \  #define GETCHARINC(c, eptr) \
688    c = *eptr++; \    c = *eptr++; \
689    if (c >= 0xc0) \    if (c >= 0xc0) GETUTF8INC(c, eptr);
     { \  
     int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     while (gcaa-- > 0) \  
       { \  
       gcss -= 6; \  
       c |= (*eptr++ & 0x3f) << gcss; \  
       } \  
     }  
690    
691  /* Get the next character, testing for UTF-8 mode, and advancing the pointer */  /* Get the next character, testing for UTF-8 mode, and advancing the pointer.
692    This is called when we don't know if we are in UTF-8 mode. */
693    
694  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
695    c = *eptr++; \    c = *eptr++; \
696    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
697    
698    /* Base macro to pick up the remaining bytes of a UTF-8 character, not
699    advancing the pointer, incrementing the length. */
700    
701    #define GETUTF8LEN(c, eptr, len) \
702      { \      { \
703      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \      if ((c & 0x20) == 0) \
704      int gcss = 6*gcaa; \        { \
705      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
706      while (gcaa-- > 0) \        len++; \
707          } \
708        else if ((c & 0x10)  == 0) \
709          { \
710          c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
711          len += 2; \
712          } \
713        else if ((c & 0x08)  == 0) \
714          {\
715          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
716              ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
717          len += 3; \
718          } \
719        else if ((c & 0x04)  == 0) \
720        { \        { \
721        gcss -= 6; \        c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
722        c |= (*eptr++ & 0x3f) << gcss; \            ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
723              (eptr[4] & 0x3f); \
724          len += 4; \
725          } \
726        else \
727          {\
728          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
729              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
730              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
731          len += 5; \
732        } \        } \
733      }      }
734    
# Line 478  if there are extra bytes. This is called Line 737  if there are extra bytes. This is called
737    
738  #define GETCHARLEN(c, eptr, len) \  #define GETCHARLEN(c, eptr, len) \
739    c = *eptr; \    c = *eptr; \
740    if (c >= 0xc0) \    if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
     { \  
     int gcii; \  
     int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     for (gcii = 1; gcii <= gcaa; gcii++) \  
       { \  
       gcss -= 6; \  
       c |= (eptr[gcii] & 0x3f) << gcss; \  
       } \  
     len += gcaa; \  
     }  
741    
742  /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the  /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
743  pointer, incrementing length if there are extra bytes. This is called when we  pointer, incrementing length if there are extra bytes. This is called when we
744  know we are in UTF-8 mode. */  do not know if we are in UTF-8 mode. */
745    
746  #define GETCHARLENTEST(c, eptr, len) \  #define GETCHARLENTEST(c, eptr, len) \
747    c = *eptr; \    c = *eptr; \
748    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
     { \  
     int gcii; \  
     int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     for (gcii = 1; gcii <= gcaa; gcii++) \  
       { \  
       gcss -= 6; \  
       c |= (eptr[gcii] & 0x3f) << gcss; \  
       } \  
     len += gcaa; \  
     }  
749    
750  /* 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
751  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 518  because almost all calls are already wit Line 753  because almost all calls are already wit
753    
754  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
755    
756  #endif  /* Same as above, just in the other direction. */
757    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
758    
759    /* Same as above, but it allows a fully customizable form. */
760    #define ACROSSCHAR(condition, eptr, action) \
761      while((condition) && ((eptr) & 0xc0) == 0x80) action
762    
763  /* In case there is no definition of offsetof() provided - though any proper  #elif defined COMPILE_PCRE16
 Standard C system should have one. */  
764    
765  #ifndef offsetof  /* Tells the biggest code point which can be encoded as a single character. */
766  #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
767  #endif  #define MAX_VALUE_FOR_SINGLE_CHAR 65535
768    
769    /* Tests whether the code point needs extra characters to decode. */
770    
771    #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
772    
773    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
774    Otherwise it has an undefined behaviour. */
775    
776    #define GET_EXTRALEN(c) 1
777    
778    /* Returns TRUE, if the given character is not the first character
779    of a UTF sequence. */
780    
781    #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
782    
783    /* Base macro to pick up the low surrogate of a UTF-16 character, not
784    advancing the pointer. */
785    
786    #define GETUTF16(c, eptr) \
787       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
788    
789    /* Get the next UTF-16 character, not advancing the pointer. This is called when
790    we know we are in UTF-16 mode. */
791    
792    #define GETCHAR(c, eptr) \
793      c = *eptr; \
794      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
795    
796    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
797    pointer. */
798    
799    #define GETCHARTEST(c, eptr) \
800      c = *eptr; \
801      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
802    
803    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
804    the pointer. */
805    
806    #define GETUTF16INC(c, eptr) \
807       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
808    
809    /* Get the next UTF-16 character, advancing the pointer. This is called when we
810    know we are in UTF-16 mode. */
811    
812    #define GETCHARINC(c, eptr) \
813      c = *eptr++; \
814      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
815    
816    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
817    This is called when we don't know if we are in UTF-16 mode. */
818    
819    #define GETCHARINCTEST(c, eptr) \
820      c = *eptr++; \
821      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
822    
823    /* Base macro to pick up the low surrogate of a UTF-16 character, not
824    advancing the pointer, incrementing the length. */
825    
826    #define GETUTF16LEN(c, eptr, len) \
827       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
828    
829    /* Get the next UTF-16 character, not advancing the pointer, incrementing
830    length if there is a low surrogate. This is called when we know we are in
831    UTF-16 mode. */
832    
833    #define GETCHARLEN(c, eptr, len) \
834      c = *eptr; \
835      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
836    
837    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
838    pointer, incrementing length if there is a low surrogate. This is called when
839    we do not know if we are in UTF-16 mode. */
840    
841    #define GETCHARLENTEST(c, eptr, len) \
842      c = *eptr; \
843      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
844    
845    /* If the pointer is not at the start of a character, move it back until
846    it is. This is called only in UTF-16 mode - we don't put a test within the
847    macro because almost all calls are already within a block of UTF-16 only
848    code. */
849    
850    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
851    
852    /* Same as above, just in the other direction. */
853    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
854    
855    /* Same as above, but it allows a fully customizable form. */
856    #define ACROSSCHAR(condition, eptr, action) \
857      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
858    
859    #elif defined COMPILE_PCRE32
860    
861    /* These are unnecessary for the 32-bit library */
862    #undef MAX_VALUE_FOR_SINGLE_CHAR
863    #undef HAS_EXTRALEN
864    #undef GET_EXTRALEN
865    #undef NOT_FIRSTCHAR
866    
867    /* Get the next UTF-32 character, not advancing the pointer. This is called when
868    we know we are in UTF-32 mode. */
869    
870    #define GETCHAR(c, eptr) \
871      c = *eptr;
872    
873    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
874    pointer. */
875    
876    #define GETCHARTEST(c, eptr) \
877      c = *eptr;
878    
879    /* Get the next UTF-32 character, advancing the pointer. This is called when we
880    know we are in UTF-32 mode. */
881    
882    #define GETCHARINC(c, eptr) \
883      c = *eptr++;
884    
885    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
886    This is called when we don't know if we are in UTF-32 mode. */
887    
888    #define GETCHARINCTEST(c, eptr) \
889      c = *eptr++;
890    
891    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
892    length (since all UTF-32 is of length 1). This is called when we know we are in
893    UTF-32 mode. */
894    
895    #define GETCHARLEN(c, eptr, len) \
896      c = *eptr;
897    
898    /* Get the next UTF-832character, testing for UTF-32 mode, not advancing the
899    pointer, not incrementing the length (since all UTF-32 is of length 1).
900    This is called when we do not know if we are in UTF-32 mode. */
901    
902    #define GETCHARLENTEST(c, eptr, len) \
903      c = *eptr;
904    
905    /* If the pointer is not at the start of a character, move it back until
906    it is. This is called only in UTF-32 mode - we don't put a test within the
907    macro because almost all calls are already within a block of UTF-32 only
908    code. */
909    
910    #define BACKCHAR(eptr) do { } while (0)
911    
912    /* Same as above, just in the other direction. */
913    #define FORWARDCHAR(eptr) do { } while (0)
914    
915    /* Same as above, but it allows a fully customizable form. */
916    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
917    
918    #else
919    #error Unsupported compiling mode
920    #endif /* COMPILE_PCRE[8|16|32] */
921    
922    #endif  /* SUPPORT_UTF */
923    
924    /* Tests for Unicode horizontal and vertical whitespace characters must check a
925    number of different values. Using a switch statement for this generates the
926    fastest code (no loop, no memory access), and there are several places in the
927    interpreter code where this happens. In order to ensure that all the case lists
928    remain in step, we use macros so that there is only one place where the lists
929    are defined.
930    
931    These values are also required as lists in pcre_compile.c when processing \h,
932    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
933    macros that define the values are here so that all the definitions are
934    together. The lists must be in ascending character order, terminated by
935    NOTACHAR (which is 0xffffffff).
936    
937    Any changes should ensure that the various macros are kept in step with each
938    other. NOTE: The values also appear in pcre_jit_compile.c. */
939    
940    /* ------ ASCII/Unicode environments ------ */
941    
942    #ifndef EBCDIC
943    
944    #define HSPACE_LIST \
945      CHAR_HT, CHAR_SPACE, 0xa0, \
946      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
947      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
948      NOTACHAR
949    
950    #define HSPACE_MULTIBYTE_CASES \
951      case 0x1680:  /* OGHAM SPACE MARK */ \
952      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
953      case 0x2000:  /* EN QUAD */ \
954      case 0x2001:  /* EM QUAD */ \
955      case 0x2002:  /* EN SPACE */ \
956      case 0x2003:  /* EM SPACE */ \
957      case 0x2004:  /* THREE-PER-EM SPACE */ \
958      case 0x2005:  /* FOUR-PER-EM SPACE */ \
959      case 0x2006:  /* SIX-PER-EM SPACE */ \
960      case 0x2007:  /* FIGURE SPACE */ \
961      case 0x2008:  /* PUNCTUATION SPACE */ \
962      case 0x2009:  /* THIN SPACE */ \
963      case 0x200A:  /* HAIR SPACE */ \
964      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
965      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
966      case 0x3000   /* IDEOGRAPHIC SPACE */
967    
968    #define HSPACE_BYTE_CASES \
969      case CHAR_HT: \
970      case CHAR_SPACE: \
971      case 0xa0     /* NBSP */
972    
973    #define HSPACE_CASES \
974      HSPACE_BYTE_CASES: \
975      HSPACE_MULTIBYTE_CASES
976    
977    #define VSPACE_LIST \
978      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
979    
980    #define VSPACE_MULTIBYTE_CASES \
981      case 0x2028:    /* LINE SEPARATOR */ \
982      case 0x2029     /* PARAGRAPH SEPARATOR */
983    
984    #define VSPACE_BYTE_CASES \
985      case CHAR_LF: \
986      case CHAR_VT: \
987      case CHAR_FF: \
988      case CHAR_CR: \
989      case CHAR_NEL
990    
991    #define VSPACE_CASES \
992      VSPACE_BYTE_CASES: \
993      VSPACE_MULTIBYTE_CASES
994    
995    /* ------ EBCDIC environments ------ */
996    
997    #else
998    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
999    
1000    #define HSPACE_BYTE_CASES \
1001      case CHAR_HT: \
1002      case CHAR_SPACE
1003    
1004    #define HSPACE_CASES HSPACE_BYTE_CASES
1005    
1006    #ifdef EBCDIC_NL25
1007    #define VSPACE_LIST \
1008      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1009    #else
1010    #define VSPACE_LIST \
1011      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1012    #endif
1013    
1014    #define VSPACE_BYTE_CASES \
1015      case CHAR_LF: \
1016      case CHAR_VT: \
1017      case CHAR_FF: \
1018      case CHAR_CR: \
1019      case CHAR_NEL
1020    
1021    #define VSPACE_CASES VSPACE_BYTE_CASES
1022    #endif  /* EBCDIC */
1023    
1024    /* ------ End of whitespace macros ------ */
1025    
 /* These are the public options that can change during matching. */  
1026    
 #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)  
1027    
1028  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
1029  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
1030  are in a 16-bit flags word. */  are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
1031    the restrictions on partial matching have been lifted. It remains for backwards
1032  #define PCRE_NOPARTIAL     0x0001  /* can't use partial with this regex */  compatibility. */
1033  #define PCRE_FIRSTSET      0x0002  /* first_byte is set */  
1034  #define PCRE_REQCHSET      0x0004  /* req_byte is set */  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1035  #define PCRE_STARTLINE     0x0008  /* start after \n for multiline */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1036  #define PCRE_JCHANGED      0x0010  /* j option used in regex */  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
1037  #define PCRE_HASCRORLF     0x0020  /* explicit \r or \n in pattern */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1038    #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1039    #define PCRE_REQCHSET      0x0040  /* req_byte is set */
1040    #define PCRE_RCH_CASELESS  0x0080  /* caseless requested char */
1041    #define PCRE_STARTLINE     0x0100  /* start after \n for multiline */
1042    #define PCRE_NOPARTIAL     0x0200  /* can't use partial with this regex */
1043    #define PCRE_JCHANGED      0x0400  /* j option used in regex */
1044    #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1045    #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1046    
1047    #if defined COMPILE_PCRE8
1048    #define PCRE_MODE          PCRE_MODE8
1049    #elif defined COMPILE_PCRE16
1050    #define PCRE_MODE          PCRE_MODE16
1051    #elif defined COMPILE_PCRE32
1052    #define PCRE_MODE          PCRE_MODE32
1053    #endif
1054    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1055    
1056  /* Options for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1057    
1058  #define PCRE_STUDY_MAPPED   0x01     /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
1059    #define PCRE_STUDY_MINLEN  0x0002  /* a minimum length field exists */
1060    
1061  /* Masks for identifying the public options that are permitted at compile  /* Masks for identifying the public options that are permitted at compile
1062  time, run time, or study time, respectively. */  time, run time, or study time, respectively. */
# Line 559  time, run time, or study time, respectiv Line 1069  time, run time, or study time, respectiv
1069     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1070     PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \     PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
1071     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1072     PCRE_JAVASCRIPT_COMPAT)     PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
1073    
1074  #define PUBLIC_EXEC_OPTIONS \  #define PUBLIC_EXEC_OPTIONS \
1075    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1076     PCRE_PARTIAL|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
1077     PCRE_NO_START_OPTIMIZE)     PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
1078    
1079  #define PUBLIC_DFA_EXEC_OPTIONS \  #define PUBLIC_DFA_EXEC_OPTIONS \
1080    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1081     PCRE_PARTIAL|PCRE_DFA_SHORTEST|PCRE_DFA_RESTART|PCRE_NEWLINE_BITS| \     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
1082     PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1083       PCRE_NO_START_OPTIMIZE)
1084    
1085  #define PUBLIC_STUDY_OPTIONS 0   /* None defined */  #define PUBLIC_STUDY_OPTIONS \
1086       (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1087        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1088    
1089  /* 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. */  
1090    
1091  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1092    
1093  /* Negative values for the firstchar and reqchar variables */  /* This variable is used to detect a loaded regular expression
1094    in different endianness. */
1095    
1096  #define REQ_UNSET (-2)  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
 #define REQ_NONE  (-1)  
1097    
1098  /* 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
1099  req_byte match. */  req_byte match. */
1100    
1101  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1102    
 /* 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 */  
   
1103  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1104  environments where these macros are defined elsewhere. Unfortunately, there  environments where these macros are defined elsewhere. Unfortunately, there
1105  is no way to do the same for the typedef. */  is no way to do the same for the typedef. */
# Line 622  for) in a minority area (EBCDIC platform Line 1128  for) in a minority area (EBCDIC platform
1128  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
1129  macros to give the functions distinct names. */  macros to give the functions distinct names. */
1130    
1131  #ifndef SUPPORT_UTF8  #ifndef SUPPORT_UTF
1132    
1133  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1134  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
1135    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1136    characters, a common practice has been to use its NL (0x15) character as the
1137    line terminator in C-like processing environments. However, sometimes the LF
1138    (0x25) character is used instead, according to this Unicode document:
1139    
1140    http://unicode.org/standard/reports/tr13/tr13-5.html
1141    
1142    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1143    instead. Whichever is *not* chosen is defined as NEL.
1144    
1145    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1146    same code point. */
1147    
1148    #ifdef EBCDIC
1149    
1150    #ifndef EBCDIC_NL25
1151    #define CHAR_NL                     '\x15'
1152    #define CHAR_NEL                    '\x25'
1153    #define STR_NL                      "\x15"
1154    #define STR_NEL                     "\x25"
1155    #else
1156    #define CHAR_NL                     '\x25'
1157    #define CHAR_NEL                    '\x15'
1158    #define STR_NL                      "\x25"
1159    #define STR_NEL                     "\x15"
1160    #endif
1161    
1162    #define CHAR_LF                     CHAR_NL
1163    #define STR_LF                      STR_NL
1164    
1165    #define CHAR_ESC                    '\047'
1166    #define CHAR_DEL                    '\007'
1167    #define STR_ESC                     "\047"
1168    #define STR_DEL                     "\007"
1169    
1170    #else  /* Not EBCDIC */
1171    
1172    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1173    compatibility. NEL is the Unicode newline character; make sure it is
1174    a positive value. */
1175    
1176    #define CHAR_LF                     '\n'
1177    #define CHAR_NL                     CHAR_LF
1178    #define CHAR_NEL                    ((unsigned char)'\x85')
1179    #define CHAR_ESC                    '\033'
1180    #define CHAR_DEL                    '\177'
1181    
1182    #define STR_LF                      "\n"
1183    #define STR_NL                      STR_LF
1184    #define STR_NEL                     "\x85"
1185    #define STR_ESC                     "\033"
1186    #define STR_DEL                     "\177"
1187    
1188    #endif  /* EBCDIC */
1189    
1190    /* The remaining definitions work in both environments. */
1191    
1192  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1193  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1194  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1195  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1196  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1197  #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  
1198    
1199  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1200  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 742  so that PCRE works on both ASCII and EBC Line 1296  so that PCRE works on both ASCII and EBC
1296  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1297  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1298  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1299  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1300  #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  
1301    
1302  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1303  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 853  so that PCRE works on both ASCII and EBC Line 1399  so that PCRE works on both ASCII and EBC
1399  #define STRING_COMMIT0              "COMMIT\0"  #define STRING_COMMIT0              "COMMIT\0"
1400  #define STRING_F0                   "F\0"  #define STRING_F0                   "F\0"
1401  #define STRING_FAIL0                "FAIL\0"  #define STRING_FAIL0                "FAIL\0"
1402    #define STRING_MARK0                "MARK\0"
1403  #define STRING_PRUNE0               "PRUNE\0"  #define STRING_PRUNE0               "PRUNE\0"
1404  #define STRING_SKIP0                "SKIP\0"  #define STRING_SKIP0                "SKIP\0"
1405  #define STRING_THEN                 "THEN"  #define STRING_THEN                 "THEN"
# Line 874  so that PCRE works on both ASCII and EBC Line 1421  so that PCRE works on both ASCII and EBC
1421    
1422  #define STRING_DEFINE               "DEFINE"  #define STRING_DEFINE               "DEFINE"
1423    
1424  #define STRING_CR_RIGHTPAR          "CR)"  #define STRING_CR_RIGHTPAR             "CR)"
1425  #define STRING_LF_RIGHTPAR          "LF)"  #define STRING_LF_RIGHTPAR             "LF)"
1426  #define STRING_CRLF_RIGHTPAR        "CRLF)"  #define STRING_CRLF_RIGHTPAR           "CRLF)"
1427  #define STRING_ANY_RIGHTPAR         "ANY)"  #define STRING_ANY_RIGHTPAR            "ANY)"
1428  #define STRING_ANYCRLF_RIGHTPAR     "ANYCRLF)"  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1429  #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1430  #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1431  #define STRING_UTF8_RIGHTPAR        "UTF8)"  #ifdef COMPILE_PCRE8
1432    #define STRING_UTF_RIGHTPAR            "UTF8)"
1433    #endif
1434    #ifdef COMPILE_PCRE16
1435    #define STRING_UTF_RIGHTPAR            "UTF16)"
1436    #endif
1437    #ifdef COMPILE_PCRE32
1438    #define STRING_UTF_RIGHTPAR            "UTF32)"
1439    #endif
1440    #define STRING_UCP_RIGHTPAR            "UCP)"
1441    #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1442    
1443  #else  /* SUPPORT_UTF8 */  #else  /* SUPPORT_UTF */
1444    
1445  /* 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
1446  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 893  only. */ Line 1450  only. */
1450  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1451  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1452  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1453  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1454    #define CHAR_NL                     CHAR_LF
1455    #define CHAR_NEL                    ((unsigned char)'\x85')
1456  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1457  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1458  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
# Line 1105  only. */ Line 1664  only. */
1664  #define STRING_COMMIT0              STR_C STR_O STR_M STR_M STR_I STR_T "\0"  #define STRING_COMMIT0              STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1665  #define STRING_F0                   STR_F "\0"  #define STRING_F0                   STR_F "\0"
1666  #define STRING_FAIL0                STR_F STR_A STR_I STR_L "\0"  #define STRING_FAIL0                STR_F STR_A STR_I STR_L "\0"
1667    #define STRING_MARK0                STR_M STR_A STR_R STR_K "\0"
1668  #define STRING_PRUNE0               STR_P STR_R STR_U STR_N STR_E "\0"  #define STRING_PRUNE0               STR_P STR_R STR_U STR_N STR_E "\0"
1669  #define STRING_SKIP0                STR_S STR_K STR_I STR_P "\0"  #define STRING_SKIP0                STR_S STR_K STR_I STR_P "\0"
1670  #define STRING_THEN                 STR_T STR_H STR_E STR_N  #define STRING_THEN                 STR_T STR_H STR_E STR_N
# Line 1126  only. */ Line 1686  only. */
1686    
1687  #define STRING_DEFINE               STR_D STR_E STR_F STR_I STR_N STR_E  #define STRING_DEFINE               STR_D STR_E STR_F STR_I STR_N STR_E
1688    
1689  #define STRING_CR_RIGHTPAR          STR_C STR_R STR_RIGHT_PARENTHESIS  #define STRING_CR_RIGHTPAR             STR_C STR_R STR_RIGHT_PARENTHESIS
1690  #define STRING_LF_RIGHTPAR          STR_L STR_F STR_RIGHT_PARENTHESIS  #define STRING_LF_RIGHTPAR             STR_L STR_F STR_RIGHT_PARENTHESIS
1691  #define STRING_CRLF_RIGHTPAR        STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS  #define STRING_CRLF_RIGHTPAR           STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1692  #define STRING_ANY_RIGHTPAR         STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS  #define STRING_ANY_RIGHTPAR            STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1693  #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
1694  #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
1695  #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
1696  #define STRING_UTF8_RIGHTPAR        STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS  #ifdef COMPILE_PCRE8
1697    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1698    #endif
1699    #ifdef COMPILE_PCRE16
1700    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1701    #endif
1702    #ifdef COMPILE_PCRE32
1703    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1704    #endif
1705    #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1706    #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
1707    
1708  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
1709    
1710  /* Escape items that are just an encoding of a particular data value. */  /* Escape items that are just an encoding of a particular data value. */
1711    
# Line 1148  only. */ Line 1718  only. */
1718  #endif  #endif
1719    
1720  #ifndef ESC_n  #ifndef ESC_n
1721  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1722  #endif  #endif
1723    
1724  #ifndef ESC_r  #ifndef ESC_r
# Line 1166  only. */ Line 1736  only. */
1736    
1737  #define PT_ANY        0    /* Any property - matches all chars */  #define PT_ANY        0    /* Any property - matches all chars */
1738  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */
1739  #define PT_GC         2    /* General characteristic (e.g. L) */  #define PT_GC         2    /* Specified general characteristic (e.g. L) */
1740  #define PT_PC         3    /* Particular characteristic (e.g. Lu) */  #define PT_PC         3    /* Specified particular characteristic (e.g. Lu) */
1741  #define PT_SC         4    /* Script (e.g. Han) */  #define PT_SC         4    /* Script (e.g. Han) */
1742    #define PT_ALNUM      5    /* Alphanumeric - the union of L and N */
1743    #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1744    #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1745    #define PT_WORD       8    /* Word - L plus N plus underscore */
1746    #define PT_CLIST      9    /* Pseudo-property: match character list */
1747    
1748  /* 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
1749  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1750    
1751  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1752  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */
# Line 1184  contain UTF-8 characters with values gre Line 1759  contain UTF-8 characters with values gre
1759    
1760  /* 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
1761  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
1762  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
1763  definitions below, up to ESC_z. There's a dummy for OP_ANY because it  definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1764  corresponds to "." rather than an escape sequence, and another for OP_ALLANY  corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1765  (which is used for [^] in JavaScript compatibility mode).  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1766    non-DOTALL mode, "." behaves like \N.
1767    
1768    The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1769    when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1770    They must be contiguous, and remain in order so that the replacements can be
1771    looked up from a table.
1772    
1773  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
1774  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
1775  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
1776  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
1777  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.
1778  */  */
1779    
1780  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1781         ESC_W, ESC_w, ESC_dum1, ESC_dum2, 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,
1782         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z, ESC_E, ESC_Q, ESC_g, ESC_k,         ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1783         ESC_REF };         ESC_E, ESC_Q, ESC_g, ESC_k,
1784           ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1785    
1786  /* 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
1787  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.
1788    
1789  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1790  that follow must also be updated to match. There is also a table called  that follow must also be updated to match. There are also tables called
1791  "coptable" in pcre_dfa_exec.c that must be updated. */  "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1792    
1793  enum {  enum {
1794    OP_END,            /* 0 End of pattern */    OP_END,            /* 0 End of pattern */
# Line 1225  enum { Line 1806  enum {
1806    OP_WHITESPACE,         /*  9 \s */    OP_WHITESPACE,         /*  9 \s */
1807    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1808    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1809    OP_ANY,            /* 12 Match any character (subject to DOTALL) */  
1810    OP_ALLANY,         /* 13 Match any character (not subject to DOTALL) */    OP_ANY,            /* 12 Match any character except newline (\N) */
1811      OP_ALLANY,         /* 13 Match any character */
1812    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 */
1813    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
1814    OP_PROP,           /* 16 \p (Unicode property) */    OP_PROP,           /* 16 \p (Unicode property) */
# Line 1236  enum { Line 1818  enum {
1818    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1819    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1820    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1821    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) */
1822    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1823    
1824    OP_OPT,            /* 25 Set runtime options */    OP_CIRC,           /* 25 Start of line - not multiline */
1825    OP_CIRC,           /* 26 Start of line - varies with multiline switch */    OP_CIRCM,          /* 26 Start of line - multiline */
1826    OP_DOLL,           /* 27 End of line - varies with multiline switch */    OP_DOLL,           /* 27 End of line - not multiline */
1827    OP_CHAR,           /* 28 Match one character, casefully */    OP_DOLLM,          /* 28 End of line - multiline */
1828    OP_CHARNC,         /* 29 Match one character, caselessly */    OP_CHAR,           /* 29 Match one character, casefully */
1829    OP_NOT,            /* 30 Match one character, not the following one */    OP_CHARI,          /* 30 Match one character, caselessly */
1830      OP_NOT,            /* 31 Match one character, not the given one, casefully */
1831    OP_STAR,           /* 31 The maximizing and minimizing versions of */    OP_NOTI,           /* 32 Match one character, not the given one, caselessly */
1832    OP_MINSTAR,        /* 32 these six opcodes must come in pairs, with */  
1833    OP_PLUS,           /* 33 the minimizing one second. */    /* The following sets of 13 opcodes must always be kept in step because
1834    OP_MINPLUS,        /* 34 This first set applies to single characters.*/    the offset from the first one is used to generate the others. */
1835    OP_QUERY,          /* 35 */  
1836    OP_MINQUERY,       /* 36 */    /**** Single characters, caseful, must precede the caseless ones ****/
1837    
1838    OP_UPTO,           /* 37 From 0 to n matches */    OP_STAR,           /* 33 The maximizing and minimizing versions of */
1839    OP_MINUPTO,        /* 38 */    OP_MINSTAR,        /* 34 these six opcodes must come in pairs, with */
1840    OP_EXACT,          /* 39 Exactly n matches */    OP_PLUS,           /* 35 the minimizing one second. */
1841      OP_MINPLUS,        /* 36 */
1842    OP_POSSTAR,        /* 40 Possessified star */    OP_QUERY,          /* 37 */
1843    OP_POSPLUS,        /* 41 Possessified plus */    OP_MINQUERY,       /* 38 */
1844    OP_POSQUERY,       /* 42 Posesssified query */  
1845    OP_POSUPTO,        /* 43 Possessified upto */    OP_UPTO,           /* 39 From 0 to n matches of one character, caseful*/
1846      OP_MINUPTO,        /* 40 */
1847    OP_NOTSTAR,        /* 44 The maximizing and minimizing versions of */    OP_EXACT,          /* 41 Exactly n matches */
1848    OP_NOTMINSTAR,     /* 45 these six opcodes must come in pairs, with */  
1849    OP_NOTPLUS,        /* 46 the minimizing one second. They must be in */    OP_POSSTAR,        /* 42 Possessified star, caseful */
1850    OP_NOTMINPLUS,     /* 47 exactly the same order as those above. */    OP_POSPLUS,        /* 43 Possessified plus, caseful */
1851    OP_NOTQUERY,       /* 48 This set applies to "not" single characters. */    OP_POSQUERY,       /* 44 Posesssified query, caseful */
1852    OP_NOTMINQUERY,    /* 49 */    OP_POSUPTO,        /* 45 Possessified upto, caseful */
1853    
1854    OP_NOTUPTO,        /* 50 From 0 to n matches */    /**** Single characters, caseless, must follow the caseful ones */
1855    OP_NOTMINUPTO,     /* 51 */  
1856    OP_NOTEXACT,       /* 52 Exactly n matches */    OP_STARI,          /* 46 */
1857      OP_MINSTARI,       /* 47 */
1858    OP_NOTPOSSTAR,     /* 53 Possessified versions */    OP_PLUSI,          /* 48 */
1859    OP_NOTPOSPLUS,     /* 54 */    OP_MINPLUSI,       /* 49 */
1860    OP_NOTPOSQUERY,    /* 55 */    OP_QUERYI,         /* 50 */
1861    OP_NOTPOSUPTO,     /* 56 */    OP_MINQUERYI,      /* 51 */
1862    
1863    OP_TYPESTAR,       /* 57 The maximizing and minimizing versions of */    OP_UPTOI,          /* 52 From 0 to n matches of one character, caseless */
1864    OP_TYPEMINSTAR,    /* 58 these six opcodes must come in pairs, with */    OP_MINUPTOI,       /* 53 */
1865    OP_TYPEPLUS,       /* 59 the minimizing one second. These codes must */    OP_EXACTI,         /* 54 */
1866    OP_TYPEMINPLUS,    /* 60 be in exactly the same order as those above. */  
1867    OP_TYPEQUERY,      /* 61 This set applies to character types such as \d */    OP_POSSTARI,       /* 55 Possessified star, caseless */
1868    OP_TYPEMINQUERY,   /* 62 */    OP_POSPLUSI,       /* 56 Possessified plus, caseless */
1869      OP_POSQUERYI,      /* 57 Posesssified query, caseless */
1870    OP_TYPEUPTO,       /* 63 From 0 to n matches */    OP_POSUPTOI,       /* 58 Possessified upto, caseless */
1871    OP_TYPEMINUPTO,    /* 64 */  
1872    OP_TYPEEXACT,      /* 65 Exactly n matches */    /**** The negated ones must follow the non-negated ones, and match them ****/
1873      /**** Negated single character, caseful; must precede the caseless ones ****/
1874    OP_TYPEPOSSTAR,    /* 66 Possessified versions */  
1875    OP_TYPEPOSPLUS,    /* 67 */    OP_NOTSTAR,        /* 59 The maximizing and minimizing versions of */
1876    OP_TYPEPOSQUERY,   /* 68 */    OP_NOTMINSTAR,     /* 60 these six opcodes must come in pairs, with */
1877    OP_TYPEPOSUPTO,    /* 69 */    OP_NOTPLUS,        /* 61 the minimizing one second. They must be in */
1878      OP_NOTMINPLUS,     /* 62 exactly the same order as those above. */
1879    OP_CRSTAR,         /* 70 The maximizing and minimizing versions of */    OP_NOTQUERY,       /* 63 */
1880    OP_CRMINSTAR,      /* 71 all these opcodes must come in pairs, with */    OP_NOTMINQUERY,    /* 64 */
1881    OP_CRPLUS,         /* 72 the minimizing one second. These codes must */  
1882    OP_CRMINPLUS,      /* 73 be in exactly the same order as those above. */    OP_NOTUPTO,        /* 65 From 0 to n matches, caseful */
1883    OP_CRQUERY,        /* 74 These are for character classes and back refs */    OP_NOTMINUPTO,     /* 66 */
1884    OP_CRMINQUERY,     /* 75 */    OP_NOTEXACT,       /* 67 Exactly n matches */
1885    OP_CRRANGE,        /* 76 These are different to the three sets above. */  
1886    OP_CRMINRANGE,     /* 77 */    OP_NOTPOSSTAR,     /* 68 Possessified versions, caseful */
1887      OP_NOTPOSPLUS,     /* 69 */
1888    OP_CLASS,          /* 78 Match a character class, chars < 256 only */    OP_NOTPOSQUERY,    /* 70 */
1889    OP_NCLASS,         /* 79 Same, but the bitmap was created from a negative    OP_NOTPOSUPTO,     /* 71 */
1890                             class - the difference is relevant only when a UTF-8  
1891                             character > 255 is encountered. */    /**** Negated single character, caseless; must follow the caseful ones ****/
1892    
1893    OP_XCLASS,         /* 80 Extended class for handling UTF-8 chars within the    OP_NOTSTARI,       /* 72 */
1894                             class. This does both positive and negative. */    OP_NOTMINSTARI,    /* 73 */
1895      OP_NOTPLUSI,       /* 74 */
1896    OP_REF,            /* 81 Match a back reference */    OP_NOTMINPLUSI,    /* 75 */
1897    OP_RECURSE,        /* 82 Match a numbered subpattern (possibly recursive) */    OP_NOTQUERYI,      /* 76 */
1898    OP_CALLOUT,        /* 83 Call out to external function if provided */    OP_NOTMINQUERYI,   /* 77 */
1899    
1900    OP_ALT,            /* 84 Start of alternation */    OP_NOTUPTOI,       /* 78 From 0 to n matches, caseless */
1901    OP_KET,            /* 85 End of group that doesn't have an unbounded repeat */    OP_NOTMINUPTOI,    /* 79 */
1902    OP_KETRMAX,        /* 86 These two must remain together and in this */    OP_NOTEXACTI,      /* 80 Exactly n matches */
1903    OP_KETRMIN,        /* 87 order. They are for groups the repeat for ever. */  
1904      OP_NOTPOSSTARI,    /* 81 Possessified versions, caseless */
1905    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/    OP_NOTPOSPLUSI,    /* 82 */
1906      OP_NOTPOSQUERYI,   /* 83 */
1907    OP_ASSERT,         /* 88 Positive lookahead */    OP_NOTPOSUPTOI,    /* 84 */
1908    OP_ASSERT_NOT,     /* 89 Negative lookahead */  
1909    OP_ASSERTBACK,     /* 90 Positive lookbehind */    /**** Character types ****/
1910    OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */  
1911    OP_REVERSE,        /* 92 Move pointer back - used in lookbehind assertions */    OP_TYPESTAR,       /* 85 The maximizing and minimizing versions of */
1912      OP_TYPEMINSTAR,    /* 86 these six opcodes must come in pairs, with */
1913    /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,    OP_TYPEPLUS,       /* 87 the minimizing one second. These codes must */
1914    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. */
1915      OP_TYPEQUERY,      /* 89 */
1916    OP_ONCE,           /* 93 Atomic group */    OP_TYPEMINQUERY,   /* 90 */
1917    OP_BRA,            /* 94 Start of non-capturing bracket */  
1918    OP_CBRA,           /* 95 Start of capturing bracket */    OP_TYPEUPTO,       /* 91 From 0 to n matches */
1919    OP_COND,           /* 96 Conditional group */    OP_TYPEMINUPTO,    /* 92 */
1920      OP_TYPEEXACT,      /* 93 Exactly n matches */
1921    
1922      OP_TYPEPOSSTAR,    /* 94 Possessified versions */
1923      OP_TYPEPOSPLUS,    /* 95 */
1924      OP_TYPEPOSQUERY,   /* 96 */
1925      OP_TYPEPOSUPTO,    /* 97 */
1926    
1927      /* These are used for character classes and back references; only the
1928      first six are the same as the sets above. */
1929    
1930      OP_CRSTAR,         /* 98 The maximizing and minimizing versions of */
1931      OP_CRMINSTAR,      /* 99 all these opcodes must come in pairs, with */
1932      OP_CRPLUS,         /* 100 the minimizing one second. These codes must */
1933      OP_CRMINPLUS,      /* 101 be in exactly the same order as those above. */
1934      OP_CRQUERY,        /* 102 */
1935      OP_CRMINQUERY,     /* 103 */
1936    
1937      OP_CRRANGE,        /* 104 These are different to the three sets above. */
1938      OP_CRMINRANGE,     /* 105 */
1939    
1940      /* End of quantifier opcodes */
1941    
1942      OP_CLASS,          /* 106 Match a character class, chars < 256 only */
1943      OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
1944                                  class - the difference is relevant only when a
1945                                  character > 255 is encountered. */
1946      OP_XCLASS,         /* 108 Extended class for handling > 255 chars within the
1947                                  class. This does both positive and negative. */
1948      OP_REF,            /* 109 Match a back reference, casefully */
1949      OP_REFI,           /* 110 Match a back reference, caselessly */
1950      OP_RECURSE,        /* 111 Match a numbered subpattern (possibly recursive) */
1951      OP_CALLOUT,        /* 112 Call out to external function if provided */
1952    
1953      OP_ALT,            /* 113 Start of alternation */
1954      OP_KET,            /* 114 End of group that doesn't have an unbounded repeat */
1955      OP_KETRMAX,        /* 115 These two must remain together and in this */
1956      OP_KETRMIN,        /* 116 order. They are for groups the repeat for ever. */
1957      OP_KETRPOS,        /* 117 Possessive unlimited repeat. */
1958    
1959      /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
1960      asserts must remain in order. */
1961    
1962      OP_REVERSE,        /* 118 Move pointer back - used in lookbehind assertions */
1963      OP_ASSERT,         /* 119 Positive lookahead */
1964      OP_ASSERT_NOT,     /* 120 Negative lookahead */
1965      OP_ASSERTBACK,     /* 121 Positive lookbehind */
1966      OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
1967    
1968      /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
1969      after the assertions, with ONCE first, as there's a test for >= ONCE for a
1970      subpattern that isn't an assertion. The POS versions must immediately follow
1971      the non-POS versions in each case. */
1972    
1973      OP_ONCE,           /* 123 Atomic group, contains captures */
1974      OP_ONCE_NC,        /* 124 Atomic group containing no captures */
1975      OP_BRA,            /* 125 Start of non-capturing bracket */
1976      OP_BRAPOS,         /* 126 Ditto, with unlimited, possessive repeat */
1977      OP_CBRA,           /* 127 Start of capturing bracket */
1978      OP_CBRAPOS,        /* 128 Ditto, with unlimited, possessive repeat */
1979      OP_COND,           /* 129 Conditional group */
1980    
1981    /* 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
1982    check for >= SBRA to distinguish the two sets. */    check for >= SBRA to distinguish the two sets. */
1983    
1984    OP_SBRA,           /* 97 Start of non-capturing bracket, check empty  */    OP_SBRA,           /* 130 Start of non-capturing bracket, check empty  */
1985    OP_SCBRA,          /* 98 Start of capturing bracket, check empty */    OP_SBRAPOS,        /* 131 Ditto, with unlimited, possessive repeat */
1986    OP_SCOND,          /* 99 Conditional group, check empty */    OP_SCBRA,          /* 132 Start of capturing bracket, check empty */
1987      OP_SCBRAPOS,       /* 133 Ditto, with unlimited, possessive repeat */
1988    OP_CREF,           /* 100 Used to hold a capture number as condition */    OP_SCOND,          /* 134 Conditional group, check empty */
1989    OP_RREF,           /* 101 Used to hold a recursion number as condition */  
1990    OP_DEF,            /* 102 The DEFINE condition */    /* The next two pairs must (respectively) be kept together. */
1991    
1992    OP_BRAZERO,        /* 103 These two must remain together and in this */    OP_CREF,           /* 135 Used to hold a capture number as condition */
1993    OP_BRAMINZERO,     /* 104 order. */    OP_NCREF,          /* 136 Same, but generated by a name reference*/
1994      OP_RREF,           /* 137 Used to hold a recursion number as condition */
1995      OP_NRREF,          /* 138 Same, but generated by a name reference*/
1996      OP_DEF,            /* 139 The DEFINE condition */
1997    
1998      OP_BRAZERO,        /* 140 These two must remain together and in this */
1999      OP_BRAMINZERO,     /* 141 order. */
2000      OP_BRAPOSZERO,     /* 142 */
2001    
2002    /* These are backtracking control verbs */    /* These are backtracking control verbs */
2003    
2004    OP_PRUNE,          /* 105 */    OP_MARK,           /* 143 always has an argument */
2005    OP_SKIP,           /* 106 */    OP_PRUNE,          /* 144 */
2006    OP_THEN,           /* 107 */    OP_PRUNE_ARG,      /* 145 same, but with argument */
2007    OP_COMMIT,         /* 108 */    OP_SKIP,           /* 146 */
2008      OP_SKIP_ARG,       /* 147 same, but with argument */
2009      OP_THEN,           /* 148 */
2010      OP_THEN_ARG,       /* 149 same, but with argument */
2011      OP_COMMIT,         /* 150 */
2012    
2013    /* These are forced failure and success verbs */    /* These are forced failure and success verbs */
2014    
2015    OP_FAIL,           /* 109 */    OP_FAIL,           /* 151 */
2016    OP_ACCEPT,         /* 110 */    OP_ACCEPT,         /* 152 */
2017      OP_ASSERT_ACCEPT,  /* 153 Used inside assertions */
2018      OP_CLOSE,          /* 154 Used before OP_ACCEPT to close open captures */
2019    
2020    /* This is used to skip a subpattern with a {0} quantifier */    /* This is used to skip a subpattern with a {0} quantifier */
2021    
2022    OP_SKIPZERO        /* 111 */    OP_SKIPZERO,       /* 155 */
2023    
2024      /* This is not an opcode, but is used to check that tables indexed by opcode
2025      are the correct length, in order to catch updating errors - there have been
2026      some in the past. */
2027    
2028      OP_TABLE_LENGTH
2029  };  };
2030    
2031    /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2032    definitions that follow must also be updated to match. There are also tables
2033    called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
2034    
2035    
2036  /* 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
2037  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
2038    only in pcre_printint.c, which fills out the full names in many cases (and in
2039    some cases doesn't actually use these names at all). */
2040    
2041  #define OP_NAME_LIST \  #define OP_NAME_LIST \
2042    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
2043    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
2044    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
2045    "extuni",  "\\Z", "\\z",                                        \    "extuni",  "\\Z", "\\z",                                        \
2046    "Opt", "^", "$", "char", "charnc", "not",                       \    "^", "^", "$", "$", "char", "chari", "not", "noti",             \
2047    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2048      "{", "{", "{",                                                  \
2049    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2050    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2051      "{", "{", "{",                                                  \
2052      "*+","++", "?+", "{",                                           \
2053      "*", "*?", "+", "+?", "?", "??",                                \
2054      "{", "{", "{",                                                  \
2055      "*+","++", "?+", "{",                                           \
2056      "*", "*?", "+", "+?", "?", "??",                                \
2057      "{", "{", "{",                                                  \
2058    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2059    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
2060    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2061    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \
2062    "class", "nclass", "xclass", "Ref", "Recurse", "Callout",       \    "class", "nclass", "xclass", "Ref", "Refi",                     \
2063    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",     \    "Recurse", "Callout",                                           \
2064    "AssertB", "AssertB not", "Reverse",                            \    "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos",                  \
2065    "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond",        \    "Reverse", "Assert", "Assert not", "AssertB", "AssertB not",    \
2066    "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero",    \    "Once", "Once_NC",                                              \
2067    "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT",      \    "Bra", "BraPos", "CBra", "CBraPos",                             \
2068    "Skip zero"    "Cond",                                                         \
2069      "SBra", "SBraPos", "SCBra", "SCBraPos",                         \
2070      "SCond",                                                        \
2071      "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \
2072      "Brazero", "Braminzero", "Braposzero",                          \
2073      "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \
2074      "*THEN", "*THEN", "*COMMIT", "*FAIL",                           \
2075      "*ACCEPT", "*ASSERT_ACCEPT",                                    \
2076      "Close", "Skip zero"
2077    
2078    
2079  /* This macro defines the length of fixed length operations in the compiled  /* This macro defines the length of fixed length operations in the compiled
# Line 1407  in UTF-8 mode. The code that uses this t Line 2090  in UTF-8 mode. The code that uses this t
2090    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \
2091    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
2092    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \
2093    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \    3, 3,                          /* \P, \p                                 */ \
2094    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
2095    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    1,                             /* \X                                     */ \
2096      1, 1, 1, 1, 1, 1,              /* \Z, \z, ^, ^M, $, $M                   */ \
2097    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
2098    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Chari  - the minimum length            */ \
2099    2,                             /* not                                    */ \    2,                             /* not                                    */ \
2100    /* Positive single-char repeats                            ** These are  */ \    2,                             /* noti                                   */ \
2101    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??      ** minima in  */ \    /* Positive single-char repeats                             ** These are */ \
2102    4, 4, 4,                       /* upto, minupto, exact     ** UTF-8 mode */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2103    2, 2, 2, 4,                    /* *+, ++, ?+, upto+                      */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2104      2+IMM2_SIZE,                   /* exact                                  */ \
2105      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2106      2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2107      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2108      2+IMM2_SIZE,                   /* exact I                                */ \
2109      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2110    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2111    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2112    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2113    2, 2, 2, 4,                    /* Possessive *, +, ?, upto               */ \    2+IMM2_SIZE,                   /* NOT exact                              */ \
2114      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2115      2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2116      2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2117      2+IMM2_SIZE,                   /* NOT exact I                            */ \
2118      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2119    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2120    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2121    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2122    2, 2, 2, 4,                    /* Possessive *+, ++, ?+, upto+           */ \    2+IMM2_SIZE,                   /* Type exact                             */ \
2123      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2124    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2125    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2126    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2127   33,                             /* CLASS                                  */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2128   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2129    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2130    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2131      1+IMM2_SIZE,                   /* REFI                                   */ \
2132    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2133    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2134    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
2135    1+LINK_SIZE,                   /* Ket                                    */ \    1+LINK_SIZE,                   /* Ket                                    */ \
2136    1+LINK_SIZE,                   /* KetRmax                                */ \    1+LINK_SIZE,                   /* KetRmax                                */ \
2137    1+LINK_SIZE,                   /* KetRmin                                */ \    1+LINK_SIZE,                   /* KetRmin                                */ \
2138      1+LINK_SIZE,                   /* KetRpos                                */ \
2139      1+LINK_SIZE,                   /* Reverse                                */ \
2140    1+LINK_SIZE,                   /* Assert                                 */ \    1+LINK_SIZE,                   /* Assert                                 */ \
2141    1+LINK_SIZE,                   /* Assert not                             */ \    1+LINK_SIZE,                   /* Assert not                             */ \
2142    1+LINK_SIZE,                   /* Assert behind                          */ \    1+LINK_SIZE,                   /* Assert behind                          */ \
2143    1+LINK_SIZE,                   /* Assert behind not                      */ \    1+LINK_SIZE,                   /* Assert behind not                      */ \
   1+LINK_SIZE,                   /* Reverse                                */ \  
2144    1+LINK_SIZE,                   /* ONCE                                   */ \    1+LINK_SIZE,                   /* ONCE                                   */ \
2145      1+LINK_SIZE,                   /* ONCE_NC                                */ \
2146    1+LINK_SIZE,                   /* BRA                                    */ \    1+LINK_SIZE,                   /* BRA                                    */ \
2147    3+LINK_SIZE,                   /* CBRA                                   */ \    1+LINK_SIZE,                   /* BRAPOS                                 */ \
2148      1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2149      1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2150    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2151    1+LINK_SIZE,                   /* SBRA                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2152    3+LINK_SIZE,                   /* SCBRA                                  */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2153      1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2154      1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2155    1+LINK_SIZE,                   /* SCOND                                  */ \    1+LINK_SIZE,                   /* SCOND                                  */ \
2156    3,                             /* CREF                                   */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* CREF, NCREF                            */ \
2157    3,                             /* RREF                                   */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* RREF, NRREF                            */ \
2158    1,                             /* DEF                                    */ \    1,                             /* DEF                                    */ \
2159    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2160    1, 1, 1, 1,                    /* PRUNE, SKIP, THEN, COMMIT,             */ \    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2161    1, 1, 1                        /* FAIL, ACCEPT, SKIPZERO                 */    1, 3,                          /* SKIP, SKIP_ARG                         */ \
2162      1, 3,                          /* THEN, THEN_ARG                         */ \
2163      1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2164      1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2165    
2166  /* A magic value for OP_RREF to indicate the "any recursion" condition. */  /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
2167    condition. */
2168    
2169  #define RREF_ANY  0xffff  #define RREF_ANY  0xffff
2170    
2171  /* Error code numbers. They are given names so that they can more easily be  /* Compile time error code numbers. They are given names so that they can more
2172  tracked. */  easily be tracked. When a new number is added, the table called eint in
2173    pcreposix.c must be updated. */
2174    
2175  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,
2176         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
# Line 1471  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2178  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2178         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2179         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2180         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2181         ERR60, ERR61, ERR62, ERR63, ERR64 };         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2182           ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2183    
2184    /* JIT compiling modes. The function list is indexed by them. */
2185    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2186           JIT_NUMBER_OF_COMPILE_MODES };
2187    
2188  /* 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
2189  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 1487  Because people can now save and re-use c Line 2199  Because people can now save and re-use c
2199  structure should be made at the end, and something earlier (e.g. a new  structure should be made at the end, and something earlier (e.g. a new
2200  flag in the options or one of the dummy fields) should indicate that the new  flag in the options or one of the dummy fields) should indicate that the new
2201  fields are present. Currently PCRE always sets the dummy fields to zero.  fields are present. Currently PCRE always sets the dummy fields to zero.
2202  NOTE NOTE NOTE:  NOTE NOTE NOTE
2203  */  */
2204    
2205  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2206    #define REAL_PCRE real_pcre
2207    #elif defined COMPILE_PCRE16
2208    #define REAL_PCRE real_pcre16
2209    #elif defined COMPILE_PCRE32
2210    #define REAL_PCRE real_pcre32
2211    #endif
2212    
2213    /* It is necessary to fork the struct for 32 bit, since it needs to use
2214     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2215     * typedef since pcretest needs access to  the struct of the 8-, 16-
2216     * and 32-bit variants. */
2217    
2218    typedef struct real_pcre8_or_16 {
2219    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2220    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2221    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
2222    pcre_uint16 flags;              /* Private flags */    pcre_uint16 flags;              /* Private flags */
2223    pcre_uint16 dummy1;             /* For future use */    pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2224    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2225    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2226    pcre_uint16 first_byte;    pcre_uint16 first_char;         /* Starting character */
2227    pcre_uint16 req_byte;    pcre_uint16 req_char;           /* This character must be seen */
2228    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2229    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2230    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2231    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2232      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2233      const pcre_uint8 *nullpad;      /* NULL padding */
2234    } real_pcre8_or_16;
2235    
2236    const unsigned char *tables;    /* Pointer to tables or NULL for std */  typedef struct real_pcre8_or_16 real_pcre;
2237    const unsigned char *nullpad;   /* NULL padding */  typedef struct real_pcre8_or_16 real_pcre16;
2238  } real_pcre;  
2239    typedef struct real_pcre32 {
2240      pcre_uint32 magic_number;
2241      pcre_uint32 size;               /* Total that was malloced */
2242      pcre_uint32 options;            /* Public options */
2243      pcre_uint16 flags;              /* Private flags */
2244      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2245      pcre_uint16 top_bracket;        /* Highest numbered group */
2246      pcre_uint16 top_backref;        /* Highest numbered back reference */
2247      pcre_uint32 first_char;         /* Starting character */
2248      pcre_uint32 req_char;           /* This character must be seen */
2249      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2250      pcre_uint16 name_entry_size;    /* Size of any name items */
2251      pcre_uint16 name_count;         /* Number of name items */
2252      pcre_uint16 ref_count;          /* Reference count */
2253      pcre_uint16 dummy1;             /* for later expansion */
2254      pcre_uint16 dummy2;             /* for later expansion */
2255      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2256      void *nullpad;                  /* for later expansion */
2257    } real_pcre32;
2258    
2259    /* Assert that the size of REAL_PCRE is divisible by 8 */
2260    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2261    
2262    /* Needed in pcretest to access some fields in the real_pcre* structures
2263     * directly. They're unified for 8/16/32 bits since the structs only differ
2264     * after these fields; if that ever changes, need to fork those defines into
2265     * 8/16 and 32 bit versions. */
2266    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2267    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2268    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2269    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2270    
2271  /* 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
2272  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
2273    
2274  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2275    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2276    pcre_uint32 options;    pcre_uint32 flags;              /* Private flags */
2277    uschar start_bits[32];    pcre_uint8 start_bits[32];      /* Starting char bits */
2278      pcre_uint32 minlength;          /* Minimum subject length */
2279  } pcre_study_data;  } pcre_study_data;
2280    
2281    /* Structure for building a chain of open capturing subpatterns during
2282    compiling, so that instructions to close them can be compiled when (*ACCEPT) is
2283    encountered. This is also used to identify subpatterns that contain recursive
2284    back references to themselves, so that they can be made atomic. */
2285    
2286    typedef struct open_capitem {
2287      struct open_capitem *next;    /* Chain link */
2288      pcre_uint16 number;           /* Capture number */
2289      pcre_uint16 flag;             /* Set TRUE if recursive back ref */
2290    } open_capitem;
2291    
2292  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2293  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2294    
2295  typedef struct compile_data {  typedef struct compile_data {
2296    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2297    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2298    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2299    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2300    const uschar *start_workspace;/* The start of working space */    const pcre_uchar *start_workspace;/* The start of working space */
2301    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_code;     /* The start of the compiled code */
2302    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2303    const uschar *end_pattern;    /* The end of the pattern */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2304    uschar *hwm;                  /* High watermark of workspace */    open_capitem *open_caps;          /* Chain of open capture items */
2305    uschar *name_table;           /* The name/number table */    pcre_uchar *hwm;                  /* High watermark of workspace */
2306    int  names_found;             /* Number of entries so far */    pcre_uchar *name_table;           /* The name/number table */
2307    int  name_entry_size;         /* Size of each entry */    int  names_found;                 /* Number of entries so far */
2308    int  bracount;                /* Count of capturing parens as we compile */    int  name_entry_size;             /* Size of each entry */
2309    int  final_bracount;          /* Saved value after first pass */    int  workspace_size;              /* Size of workspace */
2310    int  top_backref;             /* Maximum back reference */    int  bracount;                    /* Count of capturing parens as we compile */
2311    unsigned int backref_map;     /* Bitmap of low back refs */    int  final_bracount;              /* Saved value after first pass */
2312    int  external_options;        /* External (initial) options */    int  max_lookbehind;              /* Maximum lookbehind (characters) */
2313    int  external_flags;          /* External flag bits to be set */    int  top_backref;                 /* Maximum back reference */
2314    int  req_varyopt;             /* "After variable item" flag for reqbyte */    unsigned int backref_map;         /* Bitmap of low back refs */
2315    BOOL had_accept;              /* (*ACCEPT) encountered */    int  assert_depth;                /* Depth of nested assertions */
2316    int  nltype;                  /* Newline type */    int  external_options;            /* External (initial) options */
2317    int  nllen;                   /* Newline string length */    int  external_flags;              /* External flag bits to be set */
2318    uschar nl[4];                 /* Newline string when fixed length */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2319      BOOL had_accept;                  /* (*ACCEPT) encountered */
2320      BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2321      BOOL check_lookbehind;            /* Lookbehinds need later checking */
2322      int  nltype;                      /* Newline type */
2323      int  nllen;                       /* Newline string length */
2324      pcre_uchar nl[4];                 /* Newline string when fixed length */
2325  } compile_data;  } compile_data;
2326    
2327  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
2328  branches, for testing for left recursion. */  branches, for testing for left recursion while compiling. */
2329    
2330  typedef struct branch_chain {  typedef struct branch_chain {
2331    struct branch_chain *outer;    struct branch_chain *outer;
2332    uschar *current;    pcre_uchar *current_branch;
2333  } branch_chain;  } branch_chain;
2334    
2335  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
2336  call within the pattern. */  call within the pattern; used by pcre_exec(). */
2337    
2338  typedef struct recursion_info {  typedef struct recursion_info {
2339    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2340    int group_num;                /* Number of group that was called */    int group_num;                  /* Number of group that was called */
2341    const uschar *after_call;     /* "Return value": points after the call in the expr */    int *offset_save;               /* Pointer to start of saved offsets */
2342    USPTR save_start;             /* Old value of mstart */    int saved_max;                  /* Number of saved offsets */
2343    int *offset_save;             /* Pointer to start of saved offsets */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
   int saved_max;                /* Number of saved offsets */  
2344  } recursion_info;  } recursion_info;
2345    
2346    /* A similar structure for pcre_dfa_exec(). */
2347    
2348    typedef struct dfa_recursion_info {
2349      struct dfa_recursion_info *prevrec;
2350      int group_num;
2351      PCRE_PUCHAR subject_position;
2352    } dfa_recursion_info;
2353    
2354  /* 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
2355  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
2356  has been matched by a subpattern - to break infinite loops. */  has been matched by a subpattern - to break infinite loops; used by
2357    pcre_exec(). */
2358    
2359  typedef struct eptrblock {  typedef struct eptrblock {
2360    struct eptrblock *epb_prev;    struct eptrblock *epb_prev;
2361    USPTR epb_saved_eptr;    PCRE_PUCHAR epb_saved_eptr;
2362  } eptrblock;  } eptrblock;
2363    
2364    
# Line 1584  typedef struct match_data { Line 2369  typedef struct match_data {
2369    unsigned long int match_call_count;      /* As it says */    unsigned long int match_call_count;      /* As it says */
2370    unsigned long int match_limit;           /* As it says */    unsigned long int match_limit;           /* As it says */
2371    unsigned long int match_limit_recursion; /* As it says */    unsigned long int match_limit_recursion; /* As it says */
2372    int   *offset_vector;         /* Offset vector */    int   *offset_vector;           /* Offset vector */
2373    int    offset_end;            /* One past the end */    int    offset_end;              /* One past the end */
2374    int    offset_max;            /* The maximum usable for return data */    int    offset_max;              /* The maximum usable for return data */
2375    int    nltype;                /* Newline type */    int    nltype;                  /* Newline type */
2376    int    nllen;                 /* Newline string length */    int    nllen;                   /* Newline string length */
2377    uschar nl[4];                 /* Newline string when fixed */    int    name_count;              /* Number of names in name table */
2378    const uschar *lcc;            /* Points to lower casing table */    int    name_entry_size;         /* Size of entry in names table */
2379    const uschar *ctypes;         /* Points to table of type maps */    pcre_uchar *name_table;         /* Table of names */
2380    BOOL   offset_overflow;       /* Set if too many extractions */    pcre_uchar nl[4];               /* Newline string when fixed */
2381    BOOL   notbol;                /* NOTBOL flag */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2382    BOOL   noteol;                /* NOTEOL flag */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2383    BOOL   utf8;                  /* UTF8 flag */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2384    BOOL   jscript_compat;        /* JAVASCRIPT_COMPAT flag */    BOOL   offset_overflow;         /* Set if too many extractions */
2385    BOOL   endonly;               /* Dollar not before final \n */    BOOL   notbol;                  /* NOTBOL flag */
2386    BOOL   notempty;              /* Empty string match not wanted */    BOOL   noteol;                  /* NOTEOL flag */
2387    BOOL   partial;               /* PARTIAL flag */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2388    BOOL   hitend;                /* Hit the end of the subject at some point */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2389    BOOL   bsr_anycrlf;           /* \R is just any CRLF, not full Unicode */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2390    const uschar *start_code;     /* For use when recursing */    BOOL   endonly;                 /* Dollar not before final \n */
2391    USPTR  start_subject;         /* Start of the subject string */    BOOL   notempty;                /* Empty string match not wanted */
2392    USPTR  end_subject;           /* End of the subject string */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2393    USPTR  start_match_ptr;       /* Start of matched string */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2394    USPTR  end_match_ptr;         /* Subject position at end match */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2395    int    end_offset_top;        /* Highwater mark at end of match */    BOOL   hasthen;                 /* Pattern contains (*THEN) */
2396    int    capture_last;          /* Most recent capture number */    BOOL   ignore_skip_arg;         /* For re-run when SKIP name not found */
2397    int    start_offset;          /* The start offset value */    const  pcre_uchar *start_code;  /* For use when recursing */
2398    eptrblock *eptrchain;         /* Chain of eptrblocks for tail recursions */    PCRE_PUCHAR start_subject;      /* Start of the subject string */
2399    int    eptrn;                 /* Next free eptrblock */    PCRE_PUCHAR end_subject;        /* End of the subject string */
2400    recursion_info *recursive;    /* Linked list of recursion data */    PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2401    void  *callout_data;          /* To pass back to callouts */    PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2402      PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2403      int    partial;                 /* PARTIAL options */
2404      int    end_offset_top;          /* Highwater mark at end of match */
2405      int    capture_last;            /* Most recent capture number */
2406      int    start_offset;            /* The start offset value */
2407      int    match_function_type;     /* Set for certain special calls of MATCH() */
2408      eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2409      int    eptrn;                   /* Next free eptrblock */
2410      recursion_info *recursive;      /* Linked list of recursion data */
2411      void  *callout_data;            /* To pass back to callouts */
2412      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2413      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2414      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2415    #ifdef NO_RECURSE
2416      void  *match_frames_base;       /* For remembering malloc'd frames */
2417    #endif
2418  } match_data;  } match_data;
2419    
2420  /* 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
2421  functions. */  functions. */
2422    
2423  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2424    const uschar *start_code;     /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2425    const uschar *start_subject;  /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2426    const uschar *end_subject;    /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2427    const uschar *tables;         /* Character tables */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2428    int   moptions;               /* Match options */    const pcre_uint8 *tables;         /* Character tables */
2429    int   poptions;               /* Pattern options */    int   start_offset;               /* The start offset value */
2430    int    nltype;                /* Newline type */    int   moptions;                   /* Match options */
2431    int    nllen;                 /* Newline string length */    int   poptions;                   /* Pattern options */
2432    uschar nl[4];                 /* Newline string when fixed */    int   nltype;                     /* Newline type */
2433    void  *callout_data;          /* To pass back to callouts */    int   nllen;                      /* Newline string length */
2434      pcre_uchar nl[4];                 /* Newline string when fixed */
2435      void *callout_data;               /* To pass back to callouts */
2436      dfa_recursion_info *recursive;    /* Linked list of recursion data */
2437  } dfa_match_data;  } dfa_match_data;
2438    
2439  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 1665  total length. */ Line 2469  total length. */
2469  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2470  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2471    
2472    /* Internal function and data prefixes. */
2473    
2474    #if defined COMPILE_PCRE8
2475    #ifndef PUBL
2476    #define PUBL(name) pcre_##name
2477    #endif
2478    #ifndef PRIV
2479    #define PRIV(name) _pcre_##name
2480    #endif
2481    #elif defined COMPILE_PCRE16
2482    #ifndef PUBL
2483    #define PUBL(name) pcre16_##name
2484    #endif
2485    #ifndef PRIV
2486    #define PRIV(name) _pcre16_##name
2487    #endif
2488    #elif defined COMPILE_PCRE32
2489    #ifndef PUBL
2490    #define PUBL(name) pcre32_##name
2491    #endif
2492    #ifndef PRIV
2493    #define PRIV(name) _pcre32_##name
2494    #endif
2495    #else
2496    #error Unsupported compiling mode
2497    #endif /* COMPILE_PCRE[8|16|32] */
2498    
2499  /* 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
2500  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
2501  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 1682  of the exported public functions. They h Line 2513  of the exported public functions. They h
2513  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
2514  pcre_tables.c module. */  pcre_tables.c module. */
2515    
2516  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2517  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2518  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2519  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2520    extern const int            PRIV(utf8_table3)[];
2521  extern const int    _pcre_utf8_table1_size;  extern const pcre_uint8     PRIV(utf8_table4)[];
2522    #endif /* COMPILE_PCRE8 */
2523  extern const char   _pcre_utt_names[];  
2524  extern const ucp_type_table _pcre_utt[];  extern const char           PRIV(utt_names)[];
2525  extern const int _pcre_utt_size;  extern const ucp_type_table PRIV(utt)[];
2526    extern const int            PRIV(utt_size);
2527    
2528  extern const uschar _pcre_default_tables[];  extern const pcre_uint8     PRIV(OP_lengths)[];
2529    extern const pcre_uint8     PRIV(default_tables)[];
2530    
2531  extern const uschar _pcre_OP_lengths[];  extern const pcre_uint32    PRIV(hspace_list)[];
2532    extern const pcre_uint32    PRIV(vspace_list)[];
2533    
2534    
2535  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2536  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
2537  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2538    
2539  extern BOOL         _pcre_is_newline(const uschar *, int, const uschar *,  /* String comparison functions. */
2540                        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(const uschar *, int);  
 extern BOOL         _pcre_was_newline(const uschar *, int, const uschar *,  
                       int *, BOOL);  
 extern BOOL         _pcre_xclass(int, const uschar *);  
2541    
2542    #define STRCMP_UC_UC(str1, str2) \
2543      strcmp((char *)(str1), (char *)(str2))
2544    #define STRCMP_UC_C8(str1, str2) \
2545      strcmp((char *)(str1), (str2))
2546    #define STRNCMP_UC_UC(str1, str2, num) \
2547      strncmp((char *)(str1), (char *)(str2), (num))
2548    #define STRNCMP_UC_C8(str1, str2, num) \
2549      strncmp((char *)(str1), (str2), (num))
2550    #define STRLEN_UC(str) strlen((const char *)str)
2551    
2552    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2553    
2554    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2555                               const pcre_uchar *);
2556    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2557                               const char *);
2558    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2559                               const pcre_uchar *, unsigned int num);
2560    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2561                               const char *, unsigned int num);
2562    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2563    
2564    #define STRCMP_UC_UC(str1, str2) \
2565      PRIV(strcmp_uc_uc)((str1), (str2))
2566    #define STRCMP_UC_C8(str1, str2) \
2567      PRIV(strcmp_uc_c8)((str1), (str2))
2568    #define STRNCMP_UC_UC(str1, str2, num) \
2569      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2570    #define STRNCMP_UC_C8(str1, str2, num) \
2571      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2572    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2573    
2574    #endif /* COMPILE_PCRE[8|16|32] */
2575    
2576    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2577    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2578                               int *, BOOL);
2579    extern int               PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2580    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2581    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2582                               int *, BOOL);
2583    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2584    
2585    #ifdef SUPPORT_JIT
2586    extern void              PRIV(jit_compile)(const REAL_PCRE *,
2587                               PUBL(extra) *, int);
2588    extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2589                               const pcre_uchar *, int, int, int, int *, int);
2590    extern void              PRIV(jit_free)(void *);
2591    extern int               PRIV(jit_get_size)(void *);
2592    extern const char*       PRIV(jit_get_target)(void);
2593    #endif
2594    
2595  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
2596    
2597  typedef struct {  typedef struct {
2598    uschar script;    pcre_uint8 script;     /* ucp_Arabic, etc. */
2599    uschar chartype;    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2600    pcre_int32 other_case;    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2601      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2602      pcre_int32 other_case; /* offset to other case, or zero if none */
2603  } ucd_record;  } ucd_record;
2604    
2605  extern const ucd_record  _pcre_ucd_records[];  extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2606  extern const uschar      _pcre_ucd_stage1[];  extern const ucd_record  PRIV(ucd_records)[];
2607  extern const pcre_uint16 _pcre_ucd_stage2[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2608  extern const int         _pcre_ucp_gentype[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2609    extern const int         PRIV(ucp_gentype)[];
2610    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2611    #ifdef SUPPORT_JIT
2612    extern const int         PRIV(ucp_typerange)[];
2613    #endif
2614    
2615    #ifdef SUPPORT_UCP
2616  /* UCD access macros */  /* UCD access macros */
2617    
2618  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2619  #define GET_UCD(ch) (_pcre_ucd_records + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2620          _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2621          UCD_BLOCK_SIZE + ch % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2622    
2623  #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2624  #define UCD_SCRIPT(ch)    GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2625  #define UCD_CATEGORY(ch)  _pcre_ucp_gentype[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2626  #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2627    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2628    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2629    
2630    #endif /* SUPPORT_UCP */
2631    
2632  #endif  #endif
2633    

Legend:
Removed from v.412  
changed lines
  Added in v.1078

  ViewVC Help
Powered by ViewVC 1.1.5