/[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 455 by ph10, Sat Sep 26 19:12:32 2009 UTC revision 1100 by chpe, Tue Oct 16 15:56:26 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] && \       RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
308       (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
309      ) \      ) \
310    )    )
311    
# Line 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] && \       RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
322       (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
323      ) \      ) \
324    )    )
325    
# Line 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
 byte-mode, and more complicated ones for UTF-8 characters. BACKCHAR should  
 never be called in byte mode. To make sure it can never even appear when UTF-8  
 support is omitted, we don't even define it. */  
535    
536  #ifndef SUPPORT_UTF8  #define GET2(a,n) \
537       a[n]
538    
539    #elif defined COMPILE_PCRE32
540    
541    #define IMM2_SIZE 1
542    
543    #define PUT2(a,n,d)   \
544       a[n] = d
545    
546    #define GET2(a,n) \
547       a[n]
548    
549    #else
550    #error Unsupported compiling mode
551    #endif /* COMPILE_PCRE[8|16|32] */
552    
553    #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
554    
555    /* The maximum length of a MARK name is currently one data unit; it may be
556    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
557    
558    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
559    #define MAX_MARK ((1u << 16) - 1)
560    #else
561    #define MAX_MARK ((1u << 8) - 1)
562    #endif
563    
564    /* When UTF encoding is being used, a character is no longer just a single
565    byte. The macros for character handling generate simple sequences when used in
566    character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
567    and other macros are not used when UTF is not supported, so they are not
568    defined. To make sure they can never even appear when UTF support is omitted,
569    we don't even define them. */
570    
571    #ifndef SUPPORT_UTF
572    
573    /* #define MAX_VALUE_FOR_SINGLE_CHAR */
574    /* #define HAS_EXTRALEN(c) */
575    /* #define GET_EXTRALEN(c) */
576    /* #define NOT_FIRSTCHAR(c) */
577  #define GETCHAR(c, eptr) c = *eptr;  #define GETCHAR(c, eptr) c = *eptr;
578  #define GETCHARTEST(c, eptr) c = *eptr;  #define GETCHARTEST(c, eptr) c = *eptr;
579  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
580  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
581  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
582    #define RAWUCHAR(eptr) (*(eptr))
583    #define RAWUCHARINC(eptr) (*(eptr)++)
584    #define RAWUCHARTEST(eptr) (*(eptr))
585    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
586    /* #define GETCHARLENTEST(c, eptr, len) */
587  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
588    /* #define FORWARDCHAR(eptr) */
589    /* #define ACROSSCHAR(condition, eptr, action) */
590    
591  #else   /* SUPPORT_UTF8 */  #else   /* SUPPORT_UTF */
592    
593  /* Get the next UTF-8 character, not advancing the pointer. This is called when  /* Tests whether the code point needs extra characters to decode. */
 we know we are in UTF-8 mode. */  
594    
595  #define GETCHAR(c, eptr) \  #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
596    c = *eptr; \  
597    if (c >= 0xc0) \  /* Base macro to pick up the remaining bytes of a UTF-8 character, not
598    advancing the pointer. */
599    
600    #define GETUTF8(c, eptr) \
601      { \      { \
602      int gcii; \      if ((c & 0x20) == 0) \
603      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
604      int gcss = 6*gcaa; \      else if ((c & 0x10) == 0) \
605      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
606      for (gcii = 1; gcii <= gcaa; gcii++) \      else if ((c & 0x08) == 0) \
607          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
608          ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
609        else if ((c & 0x04) == 0) \
610          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
611              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
612              (eptr[4] & 0x3f); \
613        else \
614          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
615              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
616              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
617        }
618    
619    /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
620    the pointer. */
621    
622    #define GETUTF8INC(c, eptr) \
623        { \
624        if ((c & 0x20) == 0) \
625          c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
626        else if ((c & 0x10) == 0) \
627          { \
628          c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
629          eptr += 2; \
630          } \
631        else if ((c & 0x08) == 0) \
632        { \        { \
633        gcss -= 6; \        c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
634        c |= (eptr[gcii] & 0x3f) << gcss; \            ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
635          eptr += 3; \
636          } \
637        else if ((c & 0x04) == 0) \
638          { \
639          c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
640              ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
641              (eptr[3] & 0x3f); \
642          eptr += 4; \
643          } \
644        else \
645          { \
646          c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
647              ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
648              ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
649          eptr += 5; \
650        } \        } \
651      }      }
652    
653    #if defined COMPILE_PCRE8
654    
655    /* These macros were originally written in the form of loops that used data
656    from the tables whose names start with PRIV(utf8_table). They were rewritten by
657    a user so as not to use loops, because in some environments this gives a
658    significant performance advantage, and it seems never to do any harm. */
659    
660    /* Tells the biggest code point which can be encoded as a single character. */
661    
662    #define MAX_VALUE_FOR_SINGLE_CHAR 127
663    
664    /* Tests whether the code point needs extra characters to decode. */
665    
666    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
667    
668    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
669    Otherwise it has an undefined behaviour. */
670    
671    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
672    
673    /* Returns TRUE, if the given character is not the first character
674    of a UTF sequence. */
675    
676    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
677    
678    /* Get the next UTF-8 character, not advancing the pointer. This is called when
679    we know we are in UTF-8 mode. */
680    
681    #define GETCHAR(c, eptr) \
682      c = *eptr; \
683      if (c >= 0xc0) GETUTF8(c, eptr);
684    
685  /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the  /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
686  pointer. */  pointer. */
687    
688  #define GETCHARTEST(c, eptr) \  #define GETCHARTEST(c, eptr) \
689    c = *eptr; \    c = *eptr; \
690    if (utf8 && c >= 0xc0) \    if (utf && 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; \  
       } \  
     }  
691    
692  /* Get the next UTF-8 character, advancing the pointer. This is called when we  /* Get the next UTF-8 character, advancing the pointer. This is called when we
693  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
694    
695  #define GETCHARINC(c, eptr) \  #define GETCHARINC(c, eptr) \
696    c = *eptr++; \    c = *eptr++; \
697    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; \  
       } \  
     }  
698    
699  /* 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.
700    This is called when we don't know if we are in UTF-8 mode. */
701    
702  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
703    c = *eptr++; \    c = *eptr++; \
704    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
705    
706    /* Base macro to pick up the remaining bytes of a UTF-8 character, not
707    advancing the pointer, incrementing the length. */
708    
709    #define GETUTF8LEN(c, eptr, len) \
710      { \      { \
711      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \      if ((c & 0x20) == 0) \
712      int gcss = 6*gcaa; \        { \
713      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
714      while (gcaa-- > 0) \        len++; \
715          } \
716        else if ((c & 0x10)  == 0) \
717        { \        { \
718        gcss -= 6; \        c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
719        c |= (*eptr++ & 0x3f) << gcss; \        len += 2; \
720          } \
721        else if ((c & 0x08)  == 0) \
722          {\
723          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
724              ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
725          len += 3; \
726          } \
727        else if ((c & 0x04)  == 0) \
728          { \
729          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
730              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
731              (eptr[4] & 0x3f); \
732          len += 4; \
733          } \
734        else \
735          {\
736          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
737              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
738              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
739          len += 5; \
740        } \        } \
741      }      }
742    
# Line 478  if there are extra bytes. This is called Line 745  if there are extra bytes. This is called
745    
746  #define GETCHARLEN(c, eptr, len) \  #define GETCHARLEN(c, eptr, len) \
747    c = *eptr; \    c = *eptr; \
748    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; \  
     }  
749    
750  /* 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
751  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
752  know we are in UTF-8 mode. */  do not know if we are in UTF-8 mode. */
753    
754  #define GETCHARLENTEST(c, eptr, len) \  #define GETCHARLENTEST(c, eptr, len) \
755    c = *eptr; \    c = *eptr; \
756    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
757      { \  
758      int gcii; \  /* Returns the next uchar, not advancing the pointer. This is called when
759      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  we know we are in UTF mode. */
760      int gcss = 6*gcaa; \  
761      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  #define RAWUCHAR(eptr) \
762      for (gcii = 1; gcii <= gcaa; gcii++) \    (*(eptr))
763        { \  
764        gcss -= 6; \  /* Returns the next uchar, advancing the pointer. This is called when
765        c |= (eptr[gcii] & 0x3f) << gcss; \  we know we are in UTF mode. */
766        } \  
767      len += gcaa; \  #define RAWUCHARINC(eptr) \
768      }    (*(eptr)++)
769    
770    /* Returns the next uchar, testing for UTF mode, and not advancing the
771    pointer. */
772    
773    #define RAWUCHARTEST(eptr) \
774      (*(eptr))
775    
776    /* Returns the next uchar, testing for UTF mode, advancing the
777    pointer. */
778    
779    #define RAWUCHARINCTEST(eptr) \
780      (*(eptr)++)
781    
782  /* If the pointer is not at the start of a character, move it back until  /* If the pointer is not at the start of a character, move it back until
783  it is. This is called only in UTF-8 mode - we don't put a test within the macro  it is. This is called only in UTF-8 mode - we don't put a test within the macro
# Line 518  because almost all calls are already wit Line 785  because almost all calls are already wit
785    
786  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
787    
788  #endif  /* Same as above, just in the other direction. */
789    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
790    
791    /* Same as above, but it allows a fully customizable form. */
792    #define ACROSSCHAR(condition, eptr, action) \
793      while((condition) && ((eptr) & 0xc0) == 0x80) action
794    
795  /* In case there is no definition of offsetof() provided - though any proper  #elif defined COMPILE_PCRE16
 Standard C system should have one. */  
796    
797  #ifndef offsetof  /* Tells the biggest code point which can be encoded as a single character. */
798  #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
799  #endif  #define MAX_VALUE_FOR_SINGLE_CHAR 65535
800    
801    /* Tests whether the code point needs extra characters to decode. */
802    
803    #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
804    
805    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
806    Otherwise it has an undefined behaviour. */
807    
808    #define GET_EXTRALEN(c) 1
809    
810    /* Returns TRUE, if the given character is not the first character
811    of a UTF sequence. */
812    
813    #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
814    
815    /* Base macro to pick up the low surrogate of a UTF-16 character, not
816    advancing the pointer. */
817    
818    #define GETUTF16(c, eptr) \
819       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
820    
821    /* Get the next UTF-16 character, not advancing the pointer. This is called when
822    we know we are in UTF-16 mode. */
823    
824    #define GETCHAR(c, eptr) \
825      c = *eptr; \
826      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
827    
828    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
829    pointer. */
830    
831    #define GETCHARTEST(c, eptr) \
832      c = *eptr; \
833      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
834    
835    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
836    the pointer. */
837    
838    #define GETUTF16INC(c, eptr) \
839       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
840    
841    /* Get the next UTF-16 character, advancing the pointer. This is called when we
842    know we are in UTF-16 mode. */
843    
844    #define GETCHARINC(c, eptr) \
845      c = *eptr++; \
846      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
847    
848    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
849    This is called when we don't know if we are in UTF-16 mode. */
850    
851    #define GETCHARINCTEST(c, eptr) \
852      c = *eptr++; \
853      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
854    
855    /* Base macro to pick up the low surrogate of a UTF-16 character, not
856    advancing the pointer, incrementing the length. */
857    
858    #define GETUTF16LEN(c, eptr, len) \
859       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
860    
861    /* Get the next UTF-16 character, not advancing the pointer, incrementing
862    length if there is a low surrogate. This is called when we know we are in
863    UTF-16 mode. */
864    
865    #define GETCHARLEN(c, eptr, len) \
866      c = *eptr; \
867      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
868    
869    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
870    pointer, incrementing length if there is a low surrogate. This is called when
871    we do not know if we are in UTF-16 mode. */
872    
873    #define GETCHARLENTEST(c, eptr, len) \
874      c = *eptr; \
875      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
876    
877    /* Returns the next uchar, not advancing the pointer. This is called when
878    we know we are in UTF mode. */
879    
880    #define RAWUCHAR(eptr) \
881      (*(eptr))
882    
883    /* Returns the next uchar, advancing the pointer. This is called when
884    we know we are in UTF mode. */
885    
886    #define RAWUCHARINC(eptr) \
887      (*(eptr)++)
888    
889    /* Returns the next uchar, testing for UTF mode, and not advancing the
890    pointer. */
891    
892    #define RAWUCHARTEST(eptr) \
893      (*(eptr))
894    
895    /* Returns the next uchar, testing for UTF mode, advancing the
896    pointer. */
897    
898    #define RAWUCHARINCTEST(eptr) \
899      (*(eptr)++)
900    
901    /* If the pointer is not at the start of a character, move it back until
902    it is. This is called only in UTF-16 mode - we don't put a test within the
903    macro because almost all calls are already within a block of UTF-16 only
904    code. */
905    
906    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
907    
908    /* Same as above, just in the other direction. */
909    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
910    
911    /* Same as above, but it allows a fully customizable form. */
912    #define ACROSSCHAR(condition, eptr, action) \
913      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
914    
915    #elif defined COMPILE_PCRE32
916    
917    /* These are unnecessary for the 32-bit library */
918    #undef MAX_VALUE_FOR_SINGLE_CHAR
919    #undef HAS_EXTRALEN
920    #undef GET_EXTRALEN
921    #undef NOT_FIRSTCHAR
922    
923    #define UTF32_MASK (0x1ffffful)
924    
925    /* Get the next UTF-32 character, not advancing the pointer. This is called when
926    we know we are in UTF-32 mode. */
927    
928    #define GETCHAR(c, eptr) \
929      c = (*eptr) & UTF32_MASK;
930    
931    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
932    pointer. */
933    
934    #define GETCHARTEST(c, eptr) \
935      c = *eptr; \
936      if (utf) c &= UTF32_MASK;
937    
938    /* Get the next UTF-32 character, advancing the pointer. This is called when we
939    know we are in UTF-32 mode. */
940    
941    #define GETCHARINC(c, eptr) \
942      c = (*eptr++) & UTF32_MASK;
943    
944    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
945    This is called when we don't know if we are in UTF-32 mode. */
946    
947    #define GETCHARINCTEST(c, eptr) \
948      c = *eptr++; \
949      if (utf) c &= UTF32_MASK;
950    
951    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
952    length (since all UTF-32 is of length 1). This is called when we know we are in
953    UTF-32 mode. */
954    
955    #define GETCHARLEN(c, eptr, len) \
956      GETCHAR(c, eptr)
957    
958    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
959    pointer, not incrementing the length (since all UTF-32 is of length 1).
960    This is called when we do not know if we are in UTF-32 mode. */
961    
962    #define GETCHARLENTEST(c, eptr, len) \
963      GETCHARTEST(c, eptr)
964    
965    /* Returns the next uchar, not advancing the pointer. This is called when
966    we know we are in UTF mode. */
967    
968    #define RAWUCHAR(eptr) \
969      (*(eptr) & UTF32_MASK)
970    
971    /* Returns the next uchar, advancing the pointer. This is called when
972    we know we are in UTF mode. */
973    
974    #define RAWUCHARINC(eptr) \
975      (*(eptr)++ & UTF32_MASK)
976    
977    /* Returns the next uchar, testing for UTF mode, and not advancing the
978    pointer. */
979    
980    #define RAWUCHARTEST(eptr) \
981      (utf ? (*(eptr) & UTF32_MASK) : *(eptr))
982    
983    /* Returns the next uchar, testing for UTF mode, advancing the
984    pointer. */
985    
986    #define RAWUCHARINCTEST(eptr) \
987      (utf ? (*(eptr)++ & UTF32_MASK) : *(eptr)++)
988    
989    /* If the pointer is not at the start of a character, move it back until
990    it is. This is called only in UTF-32 mode - we don't put a test within the
991    macro because almost all calls are already within a block of UTF-32 only
992    code. */
993    
994    #define BACKCHAR(eptr) do { } while (0)
995    
996    /* Same as above, just in the other direction. */
997    #define FORWARDCHAR(eptr) do { } while (0)
998    
999    /* Same as above, but it allows a fully customizable form. */
1000    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1001    
1002    #else
1003    #error Unsupported compiling mode
1004    #endif /* COMPILE_PCRE[8|16|32] */
1005    
1006    #endif  /* SUPPORT_UTF */
1007    
1008    /* Tests for Unicode horizontal and vertical whitespace characters must check a
1009    number of different values. Using a switch statement for this generates the
1010    fastest code (no loop, no memory access), and there are several places in the
1011    interpreter code where this happens. In order to ensure that all the case lists
1012    remain in step, we use macros so that there is only one place where the lists
1013    are defined.
1014    
1015    These values are also required as lists in pcre_compile.c when processing \h,
1016    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1017    macros that define the values are here so that all the definitions are
1018    together. The lists must be in ascending character order, terminated by
1019    NOTACHAR (which is 0xffffffff).
1020    
1021    Any changes should ensure that the various macros are kept in step with each
1022    other. NOTE: The values also appear in pcre_jit_compile.c. */
1023    
1024    /* ------ ASCII/Unicode environments ------ */
1025    
1026    #ifndef EBCDIC
1027    
1028    #define HSPACE_LIST \
1029      CHAR_HT, CHAR_SPACE, 0xa0, \
1030      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1031      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1032      NOTACHAR
1033    
1034    #define HSPACE_MULTIBYTE_CASES \
1035      case 0x1680:  /* OGHAM SPACE MARK */ \
1036      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1037      case 0x2000:  /* EN QUAD */ \
1038      case 0x2001:  /* EM QUAD */ \
1039      case 0x2002:  /* EN SPACE */ \
1040      case 0x2003:  /* EM SPACE */ \
1041      case 0x2004:  /* THREE-PER-EM SPACE */ \
1042      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1043      case 0x2006:  /* SIX-PER-EM SPACE */ \
1044      case 0x2007:  /* FIGURE SPACE */ \
1045      case 0x2008:  /* PUNCTUATION SPACE */ \
1046      case 0x2009:  /* THIN SPACE */ \
1047      case 0x200A:  /* HAIR SPACE */ \
1048      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1049      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1050      case 0x3000   /* IDEOGRAPHIC SPACE */
1051    
1052    #define HSPACE_BYTE_CASES \
1053      case CHAR_HT: \
1054      case CHAR_SPACE: \
1055      case 0xa0     /* NBSP */
1056    
1057    #define HSPACE_CASES \
1058      HSPACE_BYTE_CASES: \
1059      HSPACE_MULTIBYTE_CASES
1060    
1061    #define VSPACE_LIST \
1062      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1063    
1064    #define VSPACE_MULTIBYTE_CASES \
1065      case 0x2028:    /* LINE SEPARATOR */ \
1066      case 0x2029     /* PARAGRAPH SEPARATOR */
1067    
1068    #define VSPACE_BYTE_CASES \
1069      case CHAR_LF: \
1070      case CHAR_VT: \
1071      case CHAR_FF: \
1072      case CHAR_CR: \
1073      case CHAR_NEL
1074    
1075    #define VSPACE_CASES \
1076      VSPACE_BYTE_CASES: \
1077      VSPACE_MULTIBYTE_CASES
1078    
1079    /* ------ EBCDIC environments ------ */
1080    
1081    #else
1082    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1083    
1084    #define HSPACE_BYTE_CASES \
1085      case CHAR_HT: \
1086      case CHAR_SPACE
1087    
1088    #define HSPACE_CASES HSPACE_BYTE_CASES
1089    
1090    #ifdef EBCDIC_NL25
1091    #define VSPACE_LIST \
1092      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1093    #else
1094    #define VSPACE_LIST \
1095      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1096    #endif
1097    
1098    #define VSPACE_BYTE_CASES \
1099      case CHAR_LF: \
1100      case CHAR_VT: \
1101      case CHAR_FF: \
1102      case CHAR_CR: \
1103      case CHAR_NEL
1104    
1105    #define VSPACE_CASES VSPACE_BYTE_CASES
1106    #endif  /* EBCDIC */
1107    
1108    /* ------ End of whitespace macros ------ */
1109    
 /* These are the public options that can change during matching. */  
1110    
 #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)  
1111    
1112  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
1113  live at the top end of the options word, but that got almost full, so now they  live at the top end of the options word, but that got almost full, so now they
# Line 539  are in a 16-bit flags word. From release Line 1115  are in a 16-bit flags word. From release
1115  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1116  compatibility. */  compatibility. */
1117    
1118  #define PCRE_NOPARTIAL     0x0001  /* can't use partial with this regex */  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1119  #define PCRE_FIRSTSET      0x0002  /* first_byte is set */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1120  #define PCRE_REQCHSET      0x0004  /* req_byte is set */  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
1121  #define PCRE_STARTLINE     0x0008  /* start after \n for multiline */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1122  #define PCRE_JCHANGED      0x0010  /* j option used in regex */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1123  #define PCRE_HASCRORLF     0x0020  /* explicit \r or \n in pattern */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
1124    #define PCRE_RCH_CASELESS  0x0080  /* caseless requested char */
1125    #define PCRE_STARTLINE     0x0100  /* start after \n for multiline */
1126    #define PCRE_NOPARTIAL     0x0200  /* can't use partial with this regex */
1127    #define PCRE_JCHANGED      0x0400  /* j option used in regex */
1128    #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1129    #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1130    
1131    #if defined COMPILE_PCRE8
1132    #define PCRE_MODE          PCRE_MODE8
1133    #elif defined COMPILE_PCRE16
1134    #define PCRE_MODE          PCRE_MODE16
1135    #elif defined COMPILE_PCRE32
1136    #define PCRE_MODE          PCRE_MODE32
1137    #endif
1138    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1139    
1140  /* Options for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1141    
1142  #define PCRE_STUDY_MAPPED   0x01     /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
1143  #define PCRE_STUDY_MINLEN   0x02     /* a minimum length field exists */  #define PCRE_STUDY_MINLEN  0x0002  /* a minimum length field exists */
1144    
1145  /* Masks for identifying the public options that are permitted at compile  /* Masks for identifying the public options that are permitted at compile
1146  time, run time, or study time, respectively. */  time, run time, or study time, respectively. */
# Line 562  time, run time, or study time, respectiv Line 1153  time, run time, or study time, respectiv
1153     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1154     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| \
1155     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1156     PCRE_JAVASCRIPT_COMPAT)     PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
1157    
1158  #define PUBLIC_EXEC_OPTIONS \  #define PUBLIC_EXEC_OPTIONS \
1159    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
# Line 575  time, run time, or study time, respectiv Line 1166  time, run time, or study time, respectiv
1166     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1167     PCRE_NO_START_OPTIMIZE)     PCRE_NO_START_OPTIMIZE)
1168    
1169  #define PUBLIC_STUDY_OPTIONS 0   /* None defined */  #define PUBLIC_STUDY_OPTIONS \
1170       (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1171        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1172    
1173  /* 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. */  
1174    
1175  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1176    
1177  /* Negative values for the firstchar and reqchar variables */  /* This variable is used to detect a loaded regular expression
1178    in different endianness. */
1179    
1180  #define REQ_UNSET (-2)  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
 #define REQ_NONE  (-1)  
1181    
1182  /* 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
1183  req_byte match. */  req_byte match. */
1184    
1185  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1186    
 /* 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 */  
   
1187  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1188  environments where these macros are defined elsewhere. Unfortunately, there  environments where these macros are defined elsewhere. Unfortunately, there
1189  is no way to do the same for the typedef. */  is no way to do the same for the typedef. */
# Line 626  for) in a minority area (EBCDIC platform Line 1212  for) in a minority area (EBCDIC platform
1212  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
1213  macros to give the functions distinct names. */  macros to give the functions distinct names. */
1214    
1215  #ifndef SUPPORT_UTF8  #ifndef SUPPORT_UTF
1216    
1217  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1218  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
1219    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1220    characters, a common practice has been to use its NL (0x15) character as the
1221    line terminator in C-like processing environments. However, sometimes the LF
1222    (0x25) character is used instead, according to this Unicode document:
1223    
1224    http://unicode.org/standard/reports/tr13/tr13-5.html
1225    
1226    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1227    instead. Whichever is *not* chosen is defined as NEL.
1228    
1229    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1230    same code point. */
1231    
1232    #ifdef EBCDIC
1233    
1234    #ifndef EBCDIC_NL25
1235    #define CHAR_NL                     '\x15'
1236    #define CHAR_NEL                    '\x25'
1237    #define STR_NL                      "\x15"
1238    #define STR_NEL                     "\x25"
1239    #else
1240    #define CHAR_NL                     '\x25'
1241    #define CHAR_NEL                    '\x15'
1242    #define STR_NL                      "\x25"
1243    #define STR_NEL                     "\x15"
1244    #endif
1245    
1246    #define CHAR_LF                     CHAR_NL
1247    #define STR_LF                      STR_NL
1248    
1249    #define CHAR_ESC                    '\047'
1250    #define CHAR_DEL                    '\007'
1251    #define STR_ESC                     "\047"
1252    #define STR_DEL                     "\007"
1253    
1254    #else  /* Not EBCDIC */
1255    
1256    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1257    compatibility. NEL is the Unicode newline character; make sure it is
1258    a positive value. */
1259    
1260    #define CHAR_LF                     '\n'
1261    #define CHAR_NL                     CHAR_LF
1262    #define CHAR_NEL                    ((unsigned char)'\x85')
1263    #define CHAR_ESC                    '\033'
1264    #define CHAR_DEL                    '\177'
1265    
1266    #define STR_LF                      "\n"
1267    #define STR_NL                      STR_LF
1268    #define STR_NEL                     "\x85"
1269    #define STR_ESC                     "\033"
1270    #define STR_DEL                     "\177"
1271    
1272    #endif  /* EBCDIC */
1273    
1274    /* The remaining definitions work in both environments. */
1275    
1276  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1277  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1278  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1279  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1280  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1281  #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  
1282    
1283  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1284  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 746  so that PCRE works on both ASCII and EBC Line 1380  so that PCRE works on both ASCII and EBC
1380  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1381  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1382  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1383  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1384  #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  
1385    
1386  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1387  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 857  so that PCRE works on both ASCII and EBC Line 1483  so that PCRE works on both ASCII and EBC
1483  #define STRING_COMMIT0              "COMMIT\0"  #define STRING_COMMIT0              "COMMIT\0"
1484  #define STRING_F0                   "F\0"  #define STRING_F0                   "F\0"
1485  #define STRING_FAIL0                "FAIL\0"  #define STRING_FAIL0                "FAIL\0"
1486    #define STRING_MARK0                "MARK\0"
1487  #define STRING_PRUNE0               "PRUNE\0"  #define STRING_PRUNE0               "PRUNE\0"
1488  #define STRING_SKIP0                "SKIP\0"  #define STRING_SKIP0                "SKIP\0"
1489  #define STRING_THEN                 "THEN"  #define STRING_THEN                 "THEN"
# Line 878  so that PCRE works on both ASCII and EBC Line 1505  so that PCRE works on both ASCII and EBC
1505    
1506  #define STRING_DEFINE               "DEFINE"  #define STRING_DEFINE               "DEFINE"
1507    
1508  #define STRING_CR_RIGHTPAR          "CR)"  #define STRING_CR_RIGHTPAR             "CR)"
1509  #define STRING_LF_RIGHTPAR          "LF)"  #define STRING_LF_RIGHTPAR             "LF)"
1510  #define STRING_CRLF_RIGHTPAR        "CRLF)"  #define STRING_CRLF_RIGHTPAR           "CRLF)"
1511  #define STRING_ANY_RIGHTPAR         "ANY)"  #define STRING_ANY_RIGHTPAR            "ANY)"
1512  #define STRING_ANYCRLF_RIGHTPAR     "ANYCRLF)"  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1513  #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1514  #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1515  #define STRING_UTF8_RIGHTPAR        "UTF8)"  #ifdef COMPILE_PCRE8
1516    #define STRING_UTF_RIGHTPAR            "UTF8)"
1517    #endif
1518    #ifdef COMPILE_PCRE16
1519    #define STRING_UTF_RIGHTPAR            "UTF16)"
1520    #endif
1521    #ifdef COMPILE_PCRE32
1522    #define STRING_UTF_RIGHTPAR            "UTF32)"
1523    #endif
1524    #define STRING_UCP_RIGHTPAR            "UCP)"
1525    #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1526    
1527  #else  /* SUPPORT_UTF8 */  #else  /* SUPPORT_UTF */
1528    
1529  /* 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
1530  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 897  only. */ Line 1534  only. */
1534  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1535  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1536  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1537  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1538    #define CHAR_NL                     CHAR_LF
1539    #define CHAR_NEL                    ((unsigned char)'\x85')
1540  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1541  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1542  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
# Line 1109  only. */ Line 1748  only. */
1748  #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"
1749  #define STRING_F0                   STR_F "\0"  #define STRING_F0                   STR_F "\0"
1750  #define STRING_FAIL0                STR_F STR_A STR_I STR_L "\0"  #define STRING_FAIL0                STR_F STR_A STR_I STR_L "\0"
1751    #define STRING_MARK0                STR_M STR_A STR_R STR_K "\0"
1752  #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"
1753  #define STRING_SKIP0                STR_S STR_K STR_I STR_P "\0"  #define STRING_SKIP0                STR_S STR_K STR_I STR_P "\0"
1754  #define STRING_THEN                 STR_T STR_H STR_E STR_N  #define STRING_THEN                 STR_T STR_H STR_E STR_N
# Line 1130  only. */ Line 1770  only. */
1770    
1771  #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
1772    
1773  #define STRING_CR_RIGHTPAR          STR_C STR_R STR_RIGHT_PARENTHESIS  #define STRING_CR_RIGHTPAR             STR_C STR_R STR_RIGHT_PARENTHESIS
1774  #define STRING_LF_RIGHTPAR          STR_L STR_F STR_RIGHT_PARENTHESIS  #define STRING_LF_RIGHTPAR             STR_L STR_F STR_RIGHT_PARENTHESIS
1775  #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
1776  #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
1777  #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
1778  #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
1779  #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
1780  #define STRING_UTF8_RIGHTPAR        STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS  #ifdef COMPILE_PCRE8
1781    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1782    #endif
1783    #ifdef COMPILE_PCRE16
1784    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1785    #endif
1786    #ifdef COMPILE_PCRE32
1787    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1788    #endif
1789    #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1790    #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
1791    
1792  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
1793    
1794  /* Escape items that are just an encoding of a particular data value. */  /* Escape items that are just an encoding of a particular data value. */
1795    
# Line 1152  only. */ Line 1802  only. */
1802  #endif  #endif
1803    
1804  #ifndef ESC_n  #ifndef ESC_n
1805  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1806  #endif  #endif
1807    
1808  #ifndef ESC_r  #ifndef ESC_r
# Line 1170  only. */ Line 1820  only. */
1820    
1821  #define PT_ANY        0    /* Any property - matches all chars */  #define PT_ANY        0    /* Any property - matches all chars */
1822  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */
1823  #define PT_GC         2    /* General characteristic (e.g. L) */  #define PT_GC         2    /* Specified general characteristic (e.g. L) */
1824  #define PT_PC         3    /* Particular characteristic (e.g. Lu) */  #define PT_PC         3    /* Specified particular characteristic (e.g. Lu) */
1825  #define PT_SC         4    /* Script (e.g. Han) */  #define PT_SC         4    /* Script (e.g. Han) */
1826    #define PT_ALNUM      5    /* Alphanumeric - the union of L and N */
1827    #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1828    #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1829    #define PT_WORD       8    /* Word - L plus N plus underscore */
1830    #define PT_CLIST      9    /* Pseudo-property: match character list */
1831    
1832  /* 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
1833  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1834    
1835  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1836  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */
# Line 1188  contain UTF-8 characters with values gre Line 1843  contain UTF-8 characters with values gre
1843    
1844  /* 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
1845  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
1846  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
1847  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
1848  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
1849  (which is used for [^] in JavaScript compatibility mode).  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1850    non-DOTALL mode, "." behaves like \N.
1851    
1852    The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1853    when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1854    They must be contiguous, and remain in order so that the replacements can be
1855    looked up from a table.
1856    
1857  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
1858  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
1859  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
1860  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
1861  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.
1862  */  */
1863    
1864  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,
1865         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,
1866         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,
1867         ESC_REF };         ESC_E, ESC_Q, ESC_g, ESC_k,
1868           ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1869    
1870  /* 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
1871  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.
1872    
1873  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1874  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
1875  "coptable" in pcre_dfa_exec.c that must be updated. */  "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1876    
1877  enum {  enum {
1878    OP_END,            /* 0 End of pattern */    OP_END,            /* 0 End of pattern */
# Line 1229  enum { Line 1890  enum {
1890    OP_WHITESPACE,         /*  9 \s */    OP_WHITESPACE,         /*  9 \s */
1891    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1892    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1893    OP_ANY,            /* 12 Match any character (subject to DOTALL) */  
1894    OP_ALLANY,         /* 13 Match any character (not subject to DOTALL) */    OP_ANY,            /* 12 Match any character except newline (\N) */
1895      OP_ALLANY,         /* 13 Match any character */
1896    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 */
1897    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
1898    OP_PROP,           /* 16 \p (Unicode property) */    OP_PROP,           /* 16 \p (Unicode property) */
# Line 1240  enum { Line 1902  enum {
1902    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1903    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1904    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1905    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) */
1906    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1907    
1908    OP_OPT,            /* 25 Set runtime options */    OP_CIRC,           /* 25 Start of line - not multiline */
1909    OP_CIRC,           /* 26 Start of line - varies with multiline switch */    OP_CIRCM,          /* 26 Start of line - multiline */
1910    OP_DOLL,           /* 27 End of line - varies with multiline switch */    OP_DOLL,           /* 27 End of line - not multiline */
1911    OP_CHAR,           /* 28 Match one character, casefully */    OP_DOLLM,          /* 28 End of line - multiline */
1912    OP_CHARNC,         /* 29 Match one character, caselessly */    OP_CHAR,           /* 29 Match one character, casefully */
1913    OP_NOT,            /* 30 Match one character, not the following one */    OP_CHARI,          /* 30 Match one character, caselessly */
1914      OP_NOT,            /* 31 Match one character, not the given one, casefully */
1915    OP_STAR,           /* 31 The maximizing and minimizing versions of */    OP_NOTI,           /* 32 Match one character, not the given one, caselessly */
1916    OP_MINSTAR,        /* 32 these six opcodes must come in pairs, with */  
1917    OP_PLUS,           /* 33 the minimizing one second. */    /* The following sets of 13 opcodes must always be kept in step because
1918    OP_MINPLUS,        /* 34 This first set applies to single characters.*/    the offset from the first one is used to generate the others. */
1919    OP_QUERY,          /* 35 */  
1920    OP_MINQUERY,       /* 36 */    /**** Single characters, caseful, must precede the caseless ones ****/
1921    
1922    OP_UPTO,           /* 37 From 0 to n matches */    OP_STAR,           /* 33 The maximizing and minimizing versions of */
1923    OP_MINUPTO,        /* 38 */    OP_MINSTAR,        /* 34 these six opcodes must come in pairs, with */
1924    OP_EXACT,          /* 39 Exactly n matches */    OP_PLUS,           /* 35 the minimizing one second. */
1925      OP_MINPLUS,        /* 36 */
1926    OP_POSSTAR,        /* 40 Possessified star */    OP_QUERY,          /* 37 */
1927    OP_POSPLUS,        /* 41 Possessified plus */    OP_MINQUERY,       /* 38 */
1928    OP_POSQUERY,       /* 42 Posesssified query */  
1929    OP_POSUPTO,        /* 43 Possessified upto */    OP_UPTO,           /* 39 From 0 to n matches of one character, caseful*/
1930      OP_MINUPTO,        /* 40 */
1931    OP_NOTSTAR,        /* 44 The maximizing and minimizing versions of */    OP_EXACT,          /* 41 Exactly n matches */
1932    OP_NOTMINSTAR,     /* 45 these six opcodes must come in pairs, with */  
1933    OP_NOTPLUS,        /* 46 the minimizing one second. They must be in */    OP_POSSTAR,        /* 42 Possessified star, caseful */
1934    OP_NOTMINPLUS,     /* 47 exactly the same order as those above. */    OP_POSPLUS,        /* 43 Possessified plus, caseful */
1935    OP_NOTQUERY,       /* 48 This set applies to "not" single characters. */    OP_POSQUERY,       /* 44 Posesssified query, caseful */
1936    OP_NOTMINQUERY,    /* 49 */    OP_POSUPTO,        /* 45 Possessified upto, caseful */
1937    
1938    OP_NOTUPTO,        /* 50 From 0 to n matches */    /**** Single characters, caseless, must follow the caseful ones */
1939    OP_NOTMINUPTO,     /* 51 */  
1940    OP_NOTEXACT,       /* 52 Exactly n matches */    OP_STARI,          /* 46 */
1941      OP_MINSTARI,       /* 47 */
1942    OP_NOTPOSSTAR,     /* 53 Possessified versions */    OP_PLUSI,          /* 48 */
1943    OP_NOTPOSPLUS,     /* 54 */    OP_MINPLUSI,       /* 49 */
1944    OP_NOTPOSQUERY,    /* 55 */    OP_QUERYI,         /* 50 */
1945    OP_NOTPOSUPTO,     /* 56 */    OP_MINQUERYI,      /* 51 */
1946    
1947    OP_TYPESTAR,       /* 57 The maximizing and minimizing versions of */    OP_UPTOI,          /* 52 From 0 to n matches of one character, caseless */
1948    OP_TYPEMINSTAR,    /* 58 these six opcodes must come in pairs, with */    OP_MINUPTOI,       /* 53 */
1949    OP_TYPEPLUS,       /* 59 the minimizing one second. These codes must */    OP_EXACTI,         /* 54 */
1950    OP_TYPEMINPLUS,    /* 60 be in exactly the same order as those above. */  
1951    OP_TYPEQUERY,      /* 61 This set applies to character types such as \d */    OP_POSSTARI,       /* 55 Possessified star, caseless */
1952    OP_TYPEMINQUERY,   /* 62 */    OP_POSPLUSI,       /* 56 Possessified plus, caseless */
1953      OP_POSQUERYI,      /* 57 Posesssified query, caseless */
1954    OP_TYPEUPTO,       /* 63 From 0 to n matches */    OP_POSUPTOI,       /* 58 Possessified upto, caseless */
1955    OP_TYPEMINUPTO,    /* 64 */  
1956    OP_TYPEEXACT,      /* 65 Exactly n matches */    /**** The negated ones must follow the non-negated ones, and match them ****/
1957      /**** Negated single character, caseful; must precede the caseless ones ****/
1958    OP_TYPEPOSSTAR,    /* 66 Possessified versions */  
1959    OP_TYPEPOSPLUS,    /* 67 */    OP_NOTSTAR,        /* 59 The maximizing and minimizing versions of */
1960    OP_TYPEPOSQUERY,   /* 68 */    OP_NOTMINSTAR,     /* 60 these six opcodes must come in pairs, with */
1961    OP_TYPEPOSUPTO,    /* 69 */    OP_NOTPLUS,        /* 61 the minimizing one second. They must be in */
1962      OP_NOTMINPLUS,     /* 62 exactly the same order as those above. */
1963    OP_CRSTAR,         /* 70 The maximizing and minimizing versions of */    OP_NOTQUERY,       /* 63 */
1964    OP_CRMINSTAR,      /* 71 all these opcodes must come in pairs, with */    OP_NOTMINQUERY,    /* 64 */
1965    OP_CRPLUS,         /* 72 the minimizing one second. These codes must */  
1966    OP_CRMINPLUS,      /* 73 be in exactly the same order as those above. */    OP_NOTUPTO,        /* 65 From 0 to n matches, caseful */
1967    OP_CRQUERY,        /* 74 These are for character classes and back refs */    OP_NOTMINUPTO,     /* 66 */
1968    OP_CRMINQUERY,     /* 75 */    OP_NOTEXACT,       /* 67 Exactly n matches */
1969    OP_CRRANGE,        /* 76 These are different to the three sets above. */  
1970    OP_CRMINRANGE,     /* 77 */    OP_NOTPOSSTAR,     /* 68 Possessified versions, caseful */
1971      OP_NOTPOSPLUS,     /* 69 */
1972    OP_CLASS,          /* 78 Match a character class, chars < 256 only */    OP_NOTPOSQUERY,    /* 70 */
1973    OP_NCLASS,         /* 79 Same, but the bitmap was created from a negative    OP_NOTPOSUPTO,     /* 71 */
1974                             class - the difference is relevant only when a UTF-8  
1975                             character > 255 is encountered. */    /**** Negated single character, caseless; must follow the caseful ones ****/
1976    
1977    OP_XCLASS,         /* 80 Extended class for handling UTF-8 chars within the    OP_NOTSTARI,       /* 72 */
1978                             class. This does both positive and negative. */    OP_NOTMINSTARI,    /* 73 */
1979      OP_NOTPLUSI,       /* 74 */
1980    OP_REF,            /* 81 Match a back reference */    OP_NOTMINPLUSI,    /* 75 */
1981    OP_RECURSE,        /* 82 Match a numbered subpattern (possibly recursive) */    OP_NOTQUERYI,      /* 76 */
1982    OP_CALLOUT,        /* 83 Call out to external function if provided */    OP_NOTMINQUERYI,   /* 77 */
1983    
1984    OP_ALT,            /* 84 Start of alternation */    OP_NOTUPTOI,       /* 78 From 0 to n matches, caseless */
1985    OP_KET,            /* 85 End of group that doesn't have an unbounded repeat */    OP_NOTMINUPTOI,    /* 79 */
1986    OP_KETRMAX,        /* 86 These two must remain together and in this */    OP_NOTEXACTI,      /* 80 Exactly n matches */
1987    OP_KETRMIN,        /* 87 order. They are for groups the repeat for ever. */  
1988      OP_NOTPOSSTARI,    /* 81 Possessified versions, caseless */
1989    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/    OP_NOTPOSPLUSI,    /* 82 */
1990      OP_NOTPOSQUERYI,   /* 83 */
1991    OP_ASSERT,         /* 88 Positive lookahead */    OP_NOTPOSUPTOI,    /* 84 */
1992    OP_ASSERT_NOT,     /* 89 Negative lookahead */  
1993    OP_ASSERTBACK,     /* 90 Positive lookbehind */    /**** Character types ****/
1994    OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */  
1995    OP_REVERSE,        /* 92 Move pointer back - used in lookbehind assertions */    OP_TYPESTAR,       /* 85 The maximizing and minimizing versions of */
1996      OP_TYPEMINSTAR,    /* 86 these six opcodes must come in pairs, with */
1997    /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,    OP_TYPEPLUS,       /* 87 the minimizing one second. These codes must */
1998    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. */
1999      OP_TYPEQUERY,      /* 89 */
2000    OP_ONCE,           /* 93 Atomic group */    OP_TYPEMINQUERY,   /* 90 */
2001    OP_BRA,            /* 94 Start of non-capturing bracket */  
2002    OP_CBRA,           /* 95 Start of capturing bracket */    OP_TYPEUPTO,       /* 91 From 0 to n matches */
2003    OP_COND,           /* 96 Conditional group */    OP_TYPEMINUPTO,    /* 92 */
2004      OP_TYPEEXACT,      /* 93 Exactly n matches */
2005    
2006      OP_TYPEPOSSTAR,    /* 94 Possessified versions */
2007      OP_TYPEPOSPLUS,    /* 95 */
2008      OP_TYPEPOSQUERY,   /* 96 */
2009      OP_TYPEPOSUPTO,    /* 97 */
2010    
2011      /* These are used for character classes and back references; only the
2012      first six are the same as the sets above. */
2013    
2014      OP_CRSTAR,         /* 98 The maximizing and minimizing versions of */
2015      OP_CRMINSTAR,      /* 99 all these opcodes must come in pairs, with */
2016      OP_CRPLUS,         /* 100 the minimizing one second. These codes must */
2017      OP_CRMINPLUS,      /* 101 be in exactly the same order as those above. */
2018      OP_CRQUERY,        /* 102 */
2019      OP_CRMINQUERY,     /* 103 */
2020    
2021      OP_CRRANGE,        /* 104 These are different to the three sets above. */
2022      OP_CRMINRANGE,     /* 105 */
2023    
2024      /* End of quantifier opcodes */
2025    
2026      OP_CLASS,          /* 106 Match a character class, chars < 256 only */
2027      OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
2028                                  class - the difference is relevant only when a
2029                                  character > 255 is encountered. */
2030      OP_XCLASS,         /* 108 Extended class for handling > 255 chars within the
2031                                  class. This does both positive and negative. */
2032      OP_REF,            /* 109 Match a back reference, casefully */
2033      OP_REFI,           /* 110 Match a back reference, caselessly */
2034      OP_RECURSE,        /* 111 Match a numbered subpattern (possibly recursive) */
2035      OP_CALLOUT,        /* 112 Call out to external function if provided */
2036    
2037      OP_ALT,            /* 113 Start of alternation */
2038      OP_KET,            /* 114 End of group that doesn't have an unbounded repeat */
2039      OP_KETRMAX,        /* 115 These two must remain together and in this */
2040      OP_KETRMIN,        /* 116 order. They are for groups the repeat for ever. */
2041      OP_KETRPOS,        /* 117 Possessive unlimited repeat. */
2042    
2043      /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
2044      asserts must remain in order. */
2045    
2046      OP_REVERSE,        /* 118 Move pointer back - used in lookbehind assertions */
2047      OP_ASSERT,         /* 119 Positive lookahead */
2048      OP_ASSERT_NOT,     /* 120 Negative lookahead */
2049      OP_ASSERTBACK,     /* 121 Positive lookbehind */
2050      OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
2051    
2052      /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
2053      after the assertions, with ONCE first, as there's a test for >= ONCE for a
2054      subpattern that isn't an assertion. The POS versions must immediately follow
2055      the non-POS versions in each case. */
2056    
2057      OP_ONCE,           /* 123 Atomic group, contains captures */
2058      OP_ONCE_NC,        /* 124 Atomic group containing no captures */
2059      OP_BRA,            /* 125 Start of non-capturing bracket */
2060      OP_BRAPOS,         /* 126 Ditto, with unlimited, possessive repeat */
2061      OP_CBRA,           /* 127 Start of capturing bracket */
2062      OP_CBRAPOS,        /* 128 Ditto, with unlimited, possessive repeat */
2063      OP_COND,           /* 129 Conditional group */
2064    
2065    /* 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
2066    check for >= SBRA to distinguish the two sets. */    check for >= SBRA to distinguish the two sets. */
2067    
2068    OP_SBRA,           /* 97 Start of non-capturing bracket, check empty  */    OP_SBRA,           /* 130 Start of non-capturing bracket, check empty  */
2069    OP_SCBRA,          /* 98 Start of capturing bracket, check empty */    OP_SBRAPOS,        /* 131 Ditto, with unlimited, possessive repeat */
2070    OP_SCOND,          /* 99 Conditional group, check empty */    OP_SCBRA,          /* 132 Start of capturing bracket, check empty */
2071      OP_SCBRAPOS,       /* 133 Ditto, with unlimited, possessive repeat */
2072    OP_CREF,           /* 100 Used to hold a capture number as condition */    OP_SCOND,          /* 134 Conditional group, check empty */
2073    OP_RREF,           /* 101 Used to hold a recursion number as condition */  
2074    OP_DEF,            /* 102 The DEFINE condition */    /* The next two pairs must (respectively) be kept together. */
2075    
2076    OP_BRAZERO,        /* 103 These two must remain together and in this */    OP_CREF,           /* 135 Used to hold a capture number as condition */
2077    OP_BRAMINZERO,     /* 104 order. */    OP_NCREF,          /* 136 Same, but generated by a name reference*/
2078      OP_RREF,           /* 137 Used to hold a recursion number as condition */
2079      OP_NRREF,          /* 138 Same, but generated by a name reference*/
2080      OP_DEF,            /* 139 The DEFINE condition */
2081    
2082      OP_BRAZERO,        /* 140 These two must remain together and in this */
2083      OP_BRAMINZERO,     /* 141 order. */
2084      OP_BRAPOSZERO,     /* 142 */
2085    
2086    /* These are backtracking control verbs */    /* These are backtracking control verbs */
2087    
2088    OP_PRUNE,          /* 105 */    OP_MARK,           /* 143 always has an argument */
2089    OP_SKIP,           /* 106 */    OP_PRUNE,          /* 144 */
2090    OP_THEN,           /* 107 */    OP_PRUNE_ARG,      /* 145 same, but with argument */
2091    OP_COMMIT,         /* 108 */    OP_SKIP,           /* 146 */
2092      OP_SKIP_ARG,       /* 147 same, but with argument */
2093      OP_THEN,           /* 148 */
2094      OP_THEN_ARG,       /* 149 same, but with argument */
2095      OP_COMMIT,         /* 150 */
2096    
2097    /* These are forced failure and success verbs */    /* These are forced failure and success verbs */
2098    
2099    OP_FAIL,           /* 109 */    OP_FAIL,           /* 151 */
2100    OP_ACCEPT,         /* 110 */    OP_ACCEPT,         /* 152 */
2101    OP_CLOSE,          /* 111 Used before OP_ACCEPT to close open captures */    OP_ASSERT_ACCEPT,  /* 153 Used inside assertions */
2102      OP_CLOSE,          /* 154 Used before OP_ACCEPT to close open captures */
2103    
2104    /* This is used to skip a subpattern with a {0} quantifier */    /* This is used to skip a subpattern with a {0} quantifier */
2105    
2106    OP_SKIPZERO        /* 112 */    OP_SKIPZERO,       /* 155 */
2107    
2108      /* This is not an opcode, but is used to check that tables indexed by opcode
2109      are the correct length, in order to catch updating errors - there have been
2110      some in the past. */
2111    
2112      OP_TABLE_LENGTH
2113  };  };
2114    
2115    /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2116    definitions that follow must also be updated to match. There are also tables
2117    called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
2118    
2119    
2120  /* 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
2121  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
2122    only in pcre_printint.c, which fills out the full names in many cases (and in
2123    some cases doesn't actually use these names at all). */
2124    
2125  #define OP_NAME_LIST \  #define OP_NAME_LIST \
2126    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
2127    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
2128    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
2129    "extuni",  "\\Z", "\\z",                                        \    "extuni",  "\\Z", "\\z",                                        \
2130    "Opt", "^", "$", "char", "charnc", "not",                       \    "^", "^", "$", "$", "char", "chari", "not", "noti",             \
2131    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2132      "{", "{", "{",                                                  \
2133    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2134    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2135      "{", "{", "{",                                                  \
2136      "*+","++", "?+", "{",                                           \
2137      "*", "*?", "+", "+?", "?", "??",                                \
2138      "{", "{", "{",                                                  \
2139      "*+","++", "?+", "{",                                           \
2140      "*", "*?", "+", "+?", "?", "??",                                \
2141      "{", "{", "{",                                                  \
2142    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2143    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
2144    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2145    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \
2146    "class", "nclass", "xclass", "Ref", "Recurse", "Callout",       \    "class", "nclass", "xclass", "Ref", "Refi",                     \
2147    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",     \    "Recurse", "Callout",                                           \
2148    "AssertB", "AssertB not", "Reverse",                            \    "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos",                  \
2149    "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond",        \    "Reverse", "Assert", "Assert not", "AssertB", "AssertB not",    \
2150    "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero",    \    "Once", "Once_NC",                                              \
2151    "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT",      \    "Bra", "BraPos", "CBra", "CBraPos",                             \
2152      "Cond",                                                         \
2153      "SBra", "SBraPos", "SCBra", "SCBraPos",                         \
2154      "SCond",                                                        \
2155      "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \
2156      "Brazero", "Braminzero", "Braposzero",                          \
2157      "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \
2158      "*THEN", "*THEN", "*COMMIT", "*FAIL",                           \
2159      "*ACCEPT", "*ASSERT_ACCEPT",                                    \
2160    "Close", "Skip zero"    "Close", "Skip zero"
2161    
2162    
# Line 1412  in UTF-8 mode. The code that uses this t Line 2174  in UTF-8 mode. The code that uses this t
2174    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \
2175    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
2176    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \
2177    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \    3, 3,                          /* \P, \p                                 */ \
2178    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
2179    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    1,                             /* \X                                     */ \
2180      1, 1, 1, 1, 1, 1,              /* \Z, \z, ^, ^M, $, $M                   */ \
2181    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
2182    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Chari  - the minimum length            */ \
2183    2,                             /* not                                    */ \    2,                             /* not                                    */ \
2184    /* Positive single-char repeats                            ** These are  */ \    2,                             /* noti                                   */ \
2185    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??      ** minima in  */ \    /* Positive single-char repeats                             ** These are */ \
2186    4, 4, 4,                       /* upto, minupto, exact     ** UTF-8 mode */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2187    2, 2, 2, 4,                    /* *+, ++, ?+, upto+                      */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2188      2+IMM2_SIZE,                   /* exact                                  */ \
2189      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2190      2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2191      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2192      2+IMM2_SIZE,                   /* exact I                                */ \
2193      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2194    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2195    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2196    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2197    2, 2, 2, 4,                    /* Possessive *, +, ?, upto               */ \    2+IMM2_SIZE,                   /* NOT exact                              */ \
2198      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2199      2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2200      2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2201      2+IMM2_SIZE,                   /* NOT exact I                            */ \
2202      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2203    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2204    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2205    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2206    2, 2, 2, 4,                    /* Possessive *+, ++, ?+, upto+           */ \    2+IMM2_SIZE,                   /* Type exact                             */ \
2207      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2208    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2209    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2210    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2211   33,                             /* CLASS                                  */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2212   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2213    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2214    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2215      1+IMM2_SIZE,                   /* REFI                                   */ \
2216    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2217    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2218    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
2219    1+LINK_SIZE,                   /* Ket                                    */ \    1+LINK_SIZE,                   /* Ket                                    */ \
2220    1+LINK_SIZE,                   /* KetRmax                                */ \    1+LINK_SIZE,                   /* KetRmax                                */ \
2221    1+LINK_SIZE,                   /* KetRmin                                */ \    1+LINK_SIZE,                   /* KetRmin                                */ \
2222      1+LINK_SIZE,                   /* KetRpos                                */ \
2223      1+LINK_SIZE,                   /* Reverse                                */ \
2224    1+LINK_SIZE,                   /* Assert                                 */ \    1+LINK_SIZE,                   /* Assert                                 */ \
2225    1+LINK_SIZE,                   /* Assert not                             */ \    1+LINK_SIZE,                   /* Assert not                             */ \
2226    1+LINK_SIZE,                   /* Assert behind                          */ \    1+LINK_SIZE,                   /* Assert behind                          */ \
2227    1+LINK_SIZE,                   /* Assert behind not                      */ \    1+LINK_SIZE,                   /* Assert behind not                      */ \
   1+LINK_SIZE,                   /* Reverse                                */ \  
2228    1+LINK_SIZE,                   /* ONCE                                   */ \    1+LINK_SIZE,                   /* ONCE                                   */ \
2229      1+LINK_SIZE,                   /* ONCE_NC                                */ \
2230    1+LINK_SIZE,                   /* BRA                                    */ \    1+LINK_SIZE,                   /* BRA                                    */ \
2231    3+LINK_SIZE,                   /* CBRA                                   */ \    1+LINK_SIZE,                   /* BRAPOS                                 */ \
2232      1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2233      1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2234    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2235    1+LINK_SIZE,                   /* SBRA                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2236    3+LINK_SIZE,                   /* SCBRA                                  */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2237      1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2238      1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2239    1+LINK_SIZE,                   /* SCOND                                  */ \    1+LINK_SIZE,                   /* SCOND                                  */ \
2240    3,                             /* CREF                                   */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* CREF, NCREF                            */ \
2241    3,                             /* RREF                                   */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* RREF, NRREF                            */ \
2242    1,                             /* DEF                                    */ \    1,                             /* DEF                                    */ \
2243    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2244    1, 1, 1, 1,                    /* PRUNE, SKIP, THEN, COMMIT,             */ \    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2245    1, 1, 3, 1                     /* FAIL, ACCEPT, CLOSE, SKIPZERO          */    1, 3,                          /* SKIP, SKIP_ARG                         */ \
2246      1, 3,                          /* THEN, THEN_ARG                         */ \
2247      1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2248      1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2249    
2250  /* 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"
2251    condition. */
2252    
2253  #define RREF_ANY  0xffff  #define RREF_ANY  0xffff
2254    
2255  /* 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
2256  tracked. */  easily be tracked. When a new number is added, the table called eint in
2257    pcreposix.c must be updated. */
2258    
2259  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,
2260         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
# Line 1476  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2262  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2262         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2263         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2264         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2265         ERR60, ERR61, ERR62, ERR63, ERR64 };         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2266           ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2267    
2268    /* JIT compiling modes. The function list is indexed by them. */
2269    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2270           JIT_NUMBER_OF_COMPILE_MODES };
2271    
2272  /* 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
2273  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 1495  fields are present. Currently PCRE alway Line 2286  fields are present. Currently PCRE alway
2286  NOTE NOTE NOTE  NOTE NOTE NOTE
2287  */  */
2288    
2289  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2290    #define REAL_PCRE real_pcre
2291    #elif defined COMPILE_PCRE16
2292    #define REAL_PCRE real_pcre16
2293    #elif defined COMPILE_PCRE32
2294    #define REAL_PCRE real_pcre32
2295    #endif
2296    
2297    /* It is necessary to fork the struct for 32 bit, since it needs to use
2298     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2299     * typedef since pcretest needs access to  the struct of the 8-, 16-
2300     * and 32-bit variants. */
2301    
2302    typedef struct real_pcre8_or_16 {
2303    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2304    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2305    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
2306    pcre_uint16 flags;              /* Private flags */    pcre_uint16 flags;              /* Private flags */
2307    pcre_uint16 dummy1;             /* For future use */    pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2308    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2309    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2310    pcre_uint16 first_byte;    pcre_uint16 first_char;         /* Starting character */
2311    pcre_uint16 req_byte;    pcre_uint16 req_char;           /* This character must be seen */
2312    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2313    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2314    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2315    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2316      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2317      const pcre_uint8 *nullpad;      /* NULL padding */
2318    } real_pcre8_or_16;
2319    
2320    typedef struct real_pcre8_or_16 real_pcre;
2321    typedef struct real_pcre8_or_16 real_pcre16;
2322    
2323    const unsigned char *tables;    /* Pointer to tables or NULL for std */  typedef struct real_pcre32 {
2324    const unsigned char *nullpad;   /* NULL padding */    pcre_uint32 magic_number;
2325  } real_pcre;    pcre_uint32 size;               /* Total that was malloced */
2326      pcre_uint32 options;            /* Public options */
2327      pcre_uint16 flags;              /* Private flags */
2328      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2329      pcre_uint16 top_bracket;        /* Highest numbered group */
2330      pcre_uint16 top_backref;        /* Highest numbered back reference */
2331      pcre_uint32 first_char;         /* Starting character */
2332      pcre_uint32 req_char;           /* This character must be seen */
2333      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2334      pcre_uint16 name_entry_size;    /* Size of any name items */
2335      pcre_uint16 name_count;         /* Number of name items */
2336      pcre_uint16 ref_count;          /* Reference count */
2337      pcre_uint16 dummy1;             /* for later expansion */
2338      pcre_uint16 dummy2;             /* for later expansion */
2339      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2340      void *nullpad;                  /* for later expansion */
2341    } real_pcre32;
2342    
2343    /* Assert that the size of REAL_PCRE is divisible by 8 */
2344    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2345    
2346    /* Needed in pcretest to access some fields in the real_pcre* structures
2347     * directly. They're unified for 8/16/32 bits since the structs only differ
2348     * after these fields; if that ever changes, need to fork those defines into
2349     * 8/16 and 32 bit versions. */
2350    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2351    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2352    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2353    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2354    
2355  /* 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
2356  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 1520  remark (see NOTE above) about extending Line 2358  remark (see NOTE above) about extending
2358  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2359    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2360    pcre_uint32 flags;              /* Private flags */    pcre_uint32 flags;              /* Private flags */
2361    uschar start_bits[32];          /* Starting char bits */    pcre_uint8 start_bits[32];      /* Starting char bits */
2362    pcre_uint32 minlength;          /* Minimum subject length */    pcre_uint32 minlength;          /* Minimum subject length */
2363  } pcre_study_data;  } pcre_study_data;
2364    
2365  /* Structure for building a chain of open capturing subpatterns during  /* Structure for building a chain of open capturing subpatterns during
2366  compiling, so that instructions to close them can be compiled when (*ACCEPT) is  compiling, so that instructions to close them can be compiled when (*ACCEPT) is
2367  encountered. */  encountered. This is also used to identify subpatterns that contain recursive
2368    back references to themselves, so that they can be made atomic. */
2369    
2370  typedef struct open_capitem {  typedef struct open_capitem {
2371    struct open_capitem *next;    /* Chain link */    struct open_capitem *next;    /* Chain link */
2372    pcre_uint16 number;           /* Capture number */    pcre_uint16 number;           /* Capture number */
2373  } open_capitem;    pcre_uint16 flag;             /* Set TRUE if recursive back ref */
2374    } open_capitem;
2375    
2376  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2377  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2378    
2379  typedef struct compile_data {  typedef struct compile_data {
2380    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2381    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2382    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2383    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2384    const uschar *start_workspace;/* The start of working space */    const pcre_uchar *start_workspace;/* The start of working space */
2385    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_code;     /* The start of the compiled code */
2386    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2387    const uschar *end_pattern;    /* The end of the pattern */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2388    open_capitem *open_caps;      /* Chain of open capture items */    open_capitem *open_caps;          /* Chain of open capture items */
2389    uschar *hwm;                  /* High watermark of workspace */    pcre_uchar *hwm;                  /* High watermark of workspace */
2390    uschar *name_table;           /* The name/number table */    pcre_uchar *name_table;           /* The name/number table */
2391    int  names_found;             /* Number of entries so far */    int  names_found;                 /* Number of entries so far */
2392    int  name_entry_size;         /* Size of each entry */    int  name_entry_size;             /* Size of each entry */
2393    int  bracount;                /* Count of capturing parens as we compile */    int  workspace_size;              /* Size of workspace */
2394    int  final_bracount;          /* Saved value after first pass */    int  bracount;                    /* Count of capturing parens as we compile */
2395    int  top_backref;             /* Maximum back reference */    int  final_bracount;              /* Saved value after first pass */
2396    unsigned int backref_map;     /* Bitmap of low back refs */    int  max_lookbehind;              /* Maximum lookbehind (characters) */
2397    int  external_options;        /* External (initial) options */    int  top_backref;                 /* Maximum back reference */
2398    int  external_flags;          /* External flag bits to be set */    unsigned int backref_map;         /* Bitmap of low back refs */
2399    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  assert_depth;                /* Depth of nested assertions */
2400    BOOL had_accept;              /* (*ACCEPT) encountered */    int  external_options;            /* External (initial) options */
2401    BOOL check_lookbehind;        /* Lookbehinds need later checking */    int  external_flags;              /* External flag bits to be set */
2402    int  nltype;                  /* Newline type */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2403    int  nllen;                   /* Newline string length */    BOOL had_accept;                  /* (*ACCEPT) encountered */
2404    uschar nl[4];                 /* Newline string when fixed length */    BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2405      BOOL check_lookbehind;            /* Lookbehinds need later checking */
2406      int  nltype;                      /* Newline type */
2407      int  nllen;                       /* Newline string length */
2408      pcre_uchar nl[4];                 /* Newline string when fixed length */
2409  } compile_data;  } compile_data;
2410    
2411  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
2412  branches, for testing for left recursion. */  branches, for testing for left recursion while compiling. */
2413    
2414  typedef struct branch_chain {  typedef struct branch_chain {
2415    struct branch_chain *outer;    struct branch_chain *outer;
2416    uschar *current;    pcre_uchar *current_branch;
2417  } branch_chain;  } branch_chain;
2418    
2419  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
2420  call within the pattern. */  call within the pattern; used by pcre_exec(). */
2421    
2422  typedef struct recursion_info {  typedef struct recursion_info {
2423    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2424    int group_num;                /* Number of group that was called */    int group_num;                  /* Number of group that was called */
2425    const uschar *after_call;     /* "Return value": points after the call in the expr */    int *offset_save;               /* Pointer to start of saved offsets */
2426    USPTR save_start;             /* Old value of mstart */    int saved_max;                  /* Number of saved offsets */
2427    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 */  
   int offset_top;               /* Current value of offset_top */  
2428  } recursion_info;  } recursion_info;
2429    
2430    /* A similar structure for pcre_dfa_exec(). */
2431    
2432    typedef struct dfa_recursion_info {
2433      struct dfa_recursion_info *prevrec;
2434      int group_num;
2435      PCRE_PUCHAR subject_position;
2436    } dfa_recursion_info;
2437    
2438  /* 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
2439  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
2440  has been matched by a subpattern - to break infinite loops. */  has been matched by a subpattern - to break infinite loops; used by
2441    pcre_exec(). */
2442    
2443  typedef struct eptrblock {  typedef struct eptrblock {
2444    struct eptrblock *epb_prev;    struct eptrblock *epb_prev;
2445    USPTR epb_saved_eptr;    PCRE_PUCHAR epb_saved_eptr;
2446  } eptrblock;  } eptrblock;
2447    
2448    
# Line 1602  typedef struct match_data { Line 2453  typedef struct match_data {
2453    unsigned long int match_call_count;      /* As it says */    unsigned long int match_call_count;      /* As it says */
2454    unsigned long int match_limit;           /* As it says */    unsigned long int match_limit;           /* As it says */
2455    unsigned long int match_limit_recursion; /* As it says */    unsigned long int match_limit_recursion; /* As it says */
2456    int   *offset_vector;         /* Offset vector */    int   *offset_vector;           /* Offset vector */
2457    int    offset_end;            /* One past the end */    int    offset_end;              /* One past the end */
2458    int    offset_max;            /* The maximum usable for return data */    int    offset_max;              /* The maximum usable for return data */
2459    int    nltype;                /* Newline type */    int    nltype;                  /* Newline type */
2460    int    nllen;                 /* Newline string length */    int    nllen;                   /* Newline string length */
2461    uschar nl[4];                 /* Newline string when fixed */    int    name_count;              /* Number of names in name table */
2462    const uschar *lcc;            /* Points to lower casing table */    int    name_entry_size;         /* Size of entry in names table */
2463    const uschar *ctypes;         /* Points to table of type maps */    pcre_uchar *name_table;         /* Table of names */
2464    BOOL   offset_overflow;       /* Set if too many extractions */    pcre_uchar nl[4];               /* Newline string when fixed */
2465    BOOL   notbol;                /* NOTBOL flag */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2466    BOOL   noteol;                /* NOTEOL flag */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2467    BOOL   utf8;                  /* UTF8 flag */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2468    BOOL   jscript_compat;        /* JAVASCRIPT_COMPAT flag */    BOOL   offset_overflow;         /* Set if too many extractions */
2469    BOOL   endonly;               /* Dollar not before final \n */    BOOL   notbol;                  /* NOTBOL flag */
2470    BOOL   notempty;              /* Empty string match not wanted */    BOOL   noteol;                  /* NOTEOL flag */
2471    BOOL   notempty_atstart;      /* Empty string match at start not wanted */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2472    BOOL   hitend;                /* Hit the end of the subject at some point */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2473    BOOL   bsr_anycrlf;           /* \R is just any CRLF, not full Unicode */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2474    const uschar *start_code;     /* For use when recursing */    BOOL   endonly;                 /* Dollar not before final \n */
2475    USPTR  start_subject;         /* Start of the subject string */    BOOL   notempty;                /* Empty string match not wanted */
2476    USPTR  end_subject;           /* End of the subject string */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2477    USPTR  start_match_ptr;       /* Start of matched string */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2478    USPTR  end_match_ptr;         /* Subject position at end match */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2479    USPTR  start_used_ptr;        /* Earliest consulted character */    BOOL   hasthen;                 /* Pattern contains (*THEN) */
2480    int    partial;               /* PARTIAL options */    BOOL   ignore_skip_arg;         /* For re-run when SKIP name not found */
2481    int    end_offset_top;        /* Highwater mark at end of match */    const  pcre_uchar *start_code;  /* For use when recursing */
2482    int    capture_last;          /* Most recent capture number */    PCRE_PUCHAR start_subject;      /* Start of the subject string */
2483    int    start_offset;          /* The start offset value */    PCRE_PUCHAR end_subject;        /* End of the subject string */
2484    eptrblock *eptrchain;         /* Chain of eptrblocks for tail recursions */    PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2485    int    eptrn;                 /* Next free eptrblock */    PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2486    recursion_info *recursive;    /* Linked list of recursion data */    PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2487    void  *callout_data;          /* To pass back to callouts */    int    partial;                 /* PARTIAL options */
2488      int    end_offset_top;          /* Highwater mark at end of match */
2489      int    capture_last;            /* Most recent capture number */
2490      int    start_offset;            /* The start offset value */
2491      int    match_function_type;     /* Set for certain special calls of MATCH() */
2492      eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2493      int    eptrn;                   /* Next free eptrblock */
2494      recursion_info *recursive;      /* Linked list of recursion data */
2495      void  *callout_data;            /* To pass back to callouts */
2496      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2497      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2498      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2499    #ifdef NO_RECURSE
2500      void  *match_frames_base;       /* For remembering malloc'd frames */
2501    #endif
2502  } match_data;  } match_data;
2503    
2504  /* 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
2505  functions. */  functions. */
2506    
2507  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2508    const uschar *start_code;     /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2509    const uschar *start_subject;  /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2510    const uschar *end_subject;    /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2511    const uschar *start_used_ptr; /* Earliest consulted character */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2512    const uschar *tables;         /* Character tables */    const pcre_uint8 *tables;         /* Character tables */
2513    int   start_offset;           /* The start offset value */    int   start_offset;               /* The start offset value */
2514    int   moptions;               /* Match options */    int   moptions;                   /* Match options */
2515    int   poptions;               /* Pattern options */    int   poptions;                   /* Pattern options */
2516    int    nltype;                /* Newline type */    int   nltype;                     /* Newline type */
2517    int    nllen;                 /* Newline string length */    int   nllen;                      /* Newline string length */
2518    uschar nl[4];                 /* Newline string when fixed */    pcre_uchar nl[4];                 /* Newline string when fixed */
2519    void  *callout_data;          /* To pass back to callouts */    void *callout_data;               /* To pass back to callouts */
2520      dfa_recursion_info *recursive;    /* Linked list of recursion data */
2521  } dfa_match_data;  } dfa_match_data;
2522    
2523  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 1687  total length. */ Line 2553  total length. */
2553  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2554  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2555    
2556    /* Internal function and data prefixes. */
2557    
2558    #if defined COMPILE_PCRE8
2559    #ifndef PUBL
2560    #define PUBL(name) pcre_##name
2561    #endif
2562    #ifndef PRIV
2563    #define PRIV(name) _pcre_##name
2564    #endif
2565    #elif defined COMPILE_PCRE16
2566    #ifndef PUBL
2567    #define PUBL(name) pcre16_##name
2568    #endif
2569    #ifndef PRIV
2570    #define PRIV(name) _pcre16_##name
2571    #endif
2572    #elif defined COMPILE_PCRE32
2573    #ifndef PUBL
2574    #define PUBL(name) pcre32_##name
2575    #endif
2576    #ifndef PRIV
2577    #define PRIV(name) _pcre32_##name
2578    #endif
2579    #else
2580    #error Unsupported compiling mode
2581    #endif /* COMPILE_PCRE[8|16|32] */
2582    
2583  /* 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
2584  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
2585  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 1704  of the exported public functions. They h Line 2597  of the exported public functions. They h
2597  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
2598  pcre_tables.c module. */  pcre_tables.c module. */
2599    
2600  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2601  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2602  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2603  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2604    extern const int            PRIV(utf8_table3)[];
2605  extern const int    _pcre_utf8_table1_size;  extern const pcre_uint8     PRIV(utf8_table4)[];
2606    #endif /* COMPILE_PCRE8 */
2607    
2608    extern const char           PRIV(utt_names)[];
2609    extern const ucp_type_table PRIV(utt)[];
2610    extern const int            PRIV(utt_size);
2611    
2612  extern const char   _pcre_utt_names[];  extern const pcre_uint8     PRIV(OP_lengths)[];
2613  extern const ucp_type_table _pcre_utt[];  extern const pcre_uint8     PRIV(default_tables)[];
 extern const int _pcre_utt_size;  
2614    
2615  extern const uschar _pcre_default_tables[];  extern const pcre_uint32    PRIV(hspace_list)[];
2616    extern const pcre_uint32    PRIV(vspace_list)[];
 extern const uschar _pcre_OP_lengths[];  
2617    
2618    
2619  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2620  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
2621  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2622    
2623  extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);  /* String comparison functions. */
2624  extern BOOL          _pcre_is_newline(const uschar *, int, const uschar *,  #if defined COMPILE_PCRE8
                        int *, BOOL);  
 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 *);  
2625    
2626    #define STRCMP_UC_UC(str1, str2) \
2627      strcmp((char *)(str1), (char *)(str2))
2628    #define STRCMP_UC_C8(str1, str2) \
2629      strcmp((char *)(str1), (str2))
2630    #define STRNCMP_UC_UC(str1, str2, num) \
2631      strncmp((char *)(str1), (char *)(str2), (num))
2632    #define STRNCMP_UC_C8(str1, str2, num) \
2633      strncmp((char *)(str1), (str2), (num))
2634    #define STRLEN_UC(str) strlen((const char *)str)
2635    
2636    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2637    
2638    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2639                               const pcre_uchar *);
2640    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2641                               const char *);
2642    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2643                               const pcre_uchar *, unsigned int num);
2644    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2645                               const char *, unsigned int num);
2646    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2647    
2648    #define STRCMP_UC_UC(str1, str2) \
2649      PRIV(strcmp_uc_uc)((str1), (str2))
2650    #define STRCMP_UC_C8(str1, str2) \
2651      PRIV(strcmp_uc_c8)((str1), (str2))
2652    #define STRNCMP_UC_UC(str1, str2, num) \
2653      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2654    #define STRNCMP_UC_C8(str1, str2, num) \
2655      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2656    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2657    
2658    #endif /* COMPILE_PCRE[8|16|32] */
2659    
2660    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2661    
2662    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2663    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2664    
2665    #elif defined COMPILE_PCRE32
2666    
2667    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2668                               const pcre_uchar *);
2669    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2670                               const char *);
2671    
2672    #define STRCMP_UC_UC_TEST(str1, str2) \
2673      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2674    #define STRCMP_UC_C8_TEST(str1, str2) \
2675      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2676    
2677    #endif /* COMPILE_PCRE[8|16|32] */
2678    
2679    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2680    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2681                               int *, BOOL);
2682    extern int               PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2683    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2684    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2685                               int *, BOOL);
2686    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2687    
2688    #ifdef SUPPORT_JIT
2689    extern void              PRIV(jit_compile)(const REAL_PCRE *,
2690                               PUBL(extra) *, int);
2691    extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2692                               const pcre_uchar *, int, int, int, int *, int);
2693    extern void              PRIV(jit_free)(void *);
2694    extern int               PRIV(jit_get_size)(void *);
2695    extern const char*       PRIV(jit_get_target)(void);
2696    #endif
2697    
2698  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
2699    
2700  typedef struct {  typedef struct {
2701    uschar script;    pcre_uint8 script;     /* ucp_Arabic, etc. */
2702    uschar chartype;    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2703    pcre_int32 other_case;    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2704      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2705      pcre_int32 other_case; /* offset to other case, or zero if none */
2706  } ucd_record;  } ucd_record;
2707    
2708  extern const ucd_record  _pcre_ucd_records[];  extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2709  extern const uschar      _pcre_ucd_stage1[];  extern const ucd_record  PRIV(ucd_records)[];
2710  extern const pcre_uint16 _pcre_ucd_stage2[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2711  extern const int         _pcre_ucp_gentype[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2712    extern const int         PRIV(ucp_gentype)[];
2713    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2714    #ifdef SUPPORT_JIT
2715    extern const int         PRIV(ucp_typerange)[];
2716    #endif
2717    
2718    #ifdef SUPPORT_UCP
2719  /* UCD access macros */  /* UCD access macros */
2720    
2721  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2722  #define GET_UCD(ch) (_pcre_ucd_records + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2723          _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2724          UCD_BLOCK_SIZE + ch % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2725    
2726  #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2727  #define UCD_SCRIPT(ch)    GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2728  #define UCD_CATEGORY(ch)  _pcre_ucp_gentype[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2729  #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2730    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2731    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2732    
2733    #endif /* SUPPORT_UCP */
2734    
2735  #endif  #endif
2736    

Legend:
Removed from v.455  
changed lines
  Added in v.1100

  ViewVC Help
Powered by ViewVC 1.1.5