/[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 1141 by ph10, Fri Oct 19 15:44:19 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. */
 script prevents both being selected, but not everybody uses "configure". */  
56    
57  #if defined EBCDIC && defined SUPPORT_UTF8  #if !defined COMPILE_PCRE16 && !defined COMPILE_PCRE32
58  #error The use of both EBCDIC and SUPPORT_UTF8 is not supported.  #define COMPILE_PCRE8
59  #endif  #endif
60    
61  /* If SUPPORT_UCP is defined, SUPPORT_UTF8 must also be defined. The  /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
62  "configure" script ensures this, but not everybody uses "configure". */  "configure" script ensures this, but not everybody uses "configure". */
63    
64  #if defined SUPPORT_UCP && !defined SUPPORT_UTF8  #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
65    #define SUPPORT_UTF 1
66    #endif
67    
68    /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
69    reasons with existing code. */
70    
71    #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
72    #define SUPPORT_UTF 1
73    #endif
74    
75    /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
76    Until then we define it if SUPPORT_UTF is defined. */
77    
78    #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
79  #define SUPPORT_UTF8 1  #define SUPPORT_UTF8 1
80  #endif  #endif
81    
82    /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
83    script prevents both being selected, but not everybody uses "configure". */
84    
85    #if defined EBCDIC && defined SUPPORT_UTF
86    #error The use of both EBCDIC and SUPPORT_UTF is not supported.
87    #endif
88    
89  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
90  inline, and there are *still* stupid compilers about that don't like indented  inline, and there are *still* stupid compilers about that don't like indented
91  pre-processor statements, or at least there were when I first wrote this. After  pre-processor statements, or at least there were when I first wrote this. After
# Line 74  It turns out that the Mac Debugging.h he Line 95  It turns out that the Mac Debugging.h he
95  be absolutely sure we get our version. */  be absolutely sure we get our version. */
96    
97  #undef DPRINTF  #undef DPRINTF
98  #ifdef DEBUG  #ifdef PCRE_DEBUG
99  #define DPRINTF(p) printf p  #define DPRINTF(p) printf p
100  #else  #else
101  #define DPRINTF(p) /* Nothing */  #define DPRINTF(p) /* Nothing */
# Line 86  setjmp and stdarg are used is when NO_RE Line 107  setjmp and stdarg are used is when NO_RE
107    
108  #include <ctype.h>  #include <ctype.h>
109  #include <limits.h>  #include <limits.h>
 #include <setjmp.h>  
 #include <stdarg.h>  
110  #include <stddef.h>  #include <stddef.h>
111  #include <stdio.h>  #include <stdio.h>
112  #include <stdlib.h>  #include <stdlib.h>
# Line 160  set, we ensure here that it has no effec Line 179  set, we ensure here that it has no effec
179  #define PCRE_CALL_CONVENTION  #define PCRE_CALL_CONVENTION
180  #endif  #endif
181    
182  /* 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
183  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
184  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
185  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
186  preprocessor time in standard C environments. */  preprocessor time in standard C environments. */
187    
188    typedef unsigned char pcre_uint8;
189    
190  #if USHRT_MAX == 65535  #if USHRT_MAX == 65535
191    typedef unsigned short pcre_uint16;    typedef unsigned short pcre_uint16;
192    typedef short pcre_int16;    typedef short pcre_int16;
# Line 186  preprocessor time in standard C environm Line 207  preprocessor time in standard C environm
207    #error Cannot determine a type for 32-bit unsigned integers    #error Cannot determine a type for 32-bit unsigned integers
208  #endif  #endif
209    
210    /* When checking for integer overflow in pcre_compile(), we need to handle
211    large integers. If a 64-bit integer type is available, we can use that.
212    Otherwise we have to cast to double, which of course requires floating point
213    arithmetic. Handle this by defining a macro for the appropriate type. If
214    stdint.h is available, include it; it may define INT64_MAX. Systems that do not
215    have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
216    by "configure". */
217    
218    #if HAVE_STDINT_H
219    #include <stdint.h>
220    #elif HAVE_INTTYPES_H
221    #include <inttypes.h>
222    #endif
223    
224    #if defined INT64_MAX || defined int64_t
225    #define INT64_OR_DOUBLE int64_t
226    #else
227    #define INT64_OR_DOUBLE double
228    #endif
229    
230  /* All character handling must be done as unsigned characters. Otherwise there  /* All character handling must be done as unsigned characters. Otherwise there
231  are problems with top-bit-set characters and functions such as isspace().  are problems with top-bit-set characters and functions such as isspace().
232  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 *,
233  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
234  to save lots of typing. I tried "uchar", but it causes problems on Digital  called pcre_uchar.
235  Unix, where it is defined in sys/types, so use "uschar" instead. */  
236    The IN_UCHARS macro multiply its argument with the byte size of the current
237    pcre_uchar type. Useful for memcpy and such operations, whose require the
238    byte size of their input/output buffers.
239    
240    The MAX_255 macro checks whether its pcre_uchar input is less than 256.
241    
242    The TABLE_GET macro is designed for accessing elements of tables whose contain
243    exactly 256 items. When the character is able to contain more than 256
244    items, some check is needed before accessing these tables.
245    */
246    
247    #if defined COMPILE_PCRE8
248    
249    typedef unsigned char pcre_uchar;
250    #define IN_UCHARS(x) (x)
251    #define MAX_255(c) 1
252    #define TABLE_GET(c, table, default) ((table)[c])
253    
254  typedef unsigned char uschar;  #elif defined COMPILE_PCRE16
255    
256    #if USHRT_MAX != 65535
257    /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
258    pcre.h(.in) and disable (comment out) this message. */
259    #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
260    #endif
261    
262    typedef pcre_uint16 pcre_uchar;
263    #define UCHAR_SHIFT (1)
264    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
265    #define MAX_255(c) ((c) <= 255u)
266    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
267    
268    #elif defined COMPILE_PCRE32
269    
270    typedef pcre_uint32 pcre_uchar;
271    #define UCHAR_SHIFT (2)
272    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
273    #define MAX_255(c) ((c) <= 255u)
274    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
275    
276    /* Assert that pcre_uchar32 is a 32-bit type */
277    typedef int __assert_pcre_uchar32_size[sizeof(pcre_uchar) == 4 ? 1 : -1];
278    
279    #else
280    #error Unsupported compiling mode
281    #endif /* COMPILE_PCRE[8|16|32] */
282    
283  /* 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
284  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 301  start/end of string field names are. */
301  #define IS_NEWLINE(p) \  #define IS_NEWLINE(p) \
302    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
303      ((p) < NLBLOCK->PSEND && \      ((p) < NLBLOCK->PSEND && \
304       _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\       PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
305         utf8)) \         &(NLBLOCK->nllen), utf)) \
306      : \      : \
307      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
308       (p)[0] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
309       (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
310      ) \      ) \
311    )    )
312    
# Line 230  start/end of string field names are. */ Line 315  start/end of string field names are. */
315  #define WAS_NEWLINE(p) \  #define WAS_NEWLINE(p) \
316    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
317      ((p) > NLBLOCK->PSSTART && \      ((p) > NLBLOCK->PSSTART && \
318       _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \       PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
319         &(NLBLOCK->nllen), utf8)) \         &(NLBLOCK->nllen), utf)) \
320      : \      : \
321      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
322       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
323       (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
324      ) \      ) \
325    )    )
326    
# Line 249  used for the external interface and appe Line 334  used for the external interface and appe
334  must begin with PCRE_. */  must begin with PCRE_. */
335    
336  #ifdef CUSTOM_SUBJECT_PTR  #ifdef CUSTOM_SUBJECT_PTR
337  #define PCRE_SPTR CUSTOM_SUBJECT_PTR  #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
 #define USPTR CUSTOM_SUBJECT_PTR  
338  #else  #else
339  #define PCRE_SPTR const char *  #define PCRE_PUCHAR const pcre_uchar *
 #define USPTR const unsigned char *  
340  #endif  #endif
341    
   
   
342  /* Include the public PCRE header and the definitions of UCP character property  /* Include the public PCRE header and the definitions of UCP character property
343  values. */  values. */
344    
# Line 325  The macros are controlled by the value o Line 406  The macros are controlled by the value o
406  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
407  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
408    
409    #if defined COMPILE_PCRE8
410    
411  #if LINK_SIZE == 2  #if LINK_SIZE == 2
412    
413  #define PUT(a,n,d)   \  #define PUT(a,n,d)   \
# Line 361  is automated on Unix systems via the "co Line 444  is automated on Unix systems via the "co
444  #define GET(a,n) \  #define GET(a,n) \
445    (((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])
446    
447  #define MAX_PATTERN_SIZE (1 << 30)   /* Keep it positive */  /* Keep it positive */
448    #define MAX_PATTERN_SIZE (1 << 30)
449    
450    #else
451    #error LINK_SIZE must be either 2, 3, or 4
452    #endif
453    
454    #elif defined COMPILE_PCRE16
455    
456    #if LINK_SIZE == 2
457    
458    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
459    #undef LINK_SIZE
460    #define LINK_SIZE 1
461    
462    #define PUT(a,n,d)   \
463      (a[n] = (d))
464    
465    #define GET(a,n) \
466      (a[n])
467    
468    #define MAX_PATTERN_SIZE (1 << 16)
469    
470    #elif LINK_SIZE == 3 || LINK_SIZE == 4
471    
472    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
473    #undef LINK_SIZE
474    #define LINK_SIZE 2
475    
476    #define PUT(a,n,d)   \
477      (a[n] = (d) >> 16), \
478      (a[(n)+1] = (d) & 65535)
479    
480    #define GET(a,n) \
481      (((a)[n] << 16) | (a)[(n)+1])
482    
483    /* Keep it positive */
484    #define MAX_PATTERN_SIZE (1 << 30)
485    
486  #else  #else
487  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
488  #endif  #endif
489    
490    #elif defined COMPILE_PCRE32
491    
492    /* Only supported LINK_SIZE is 4 */
493    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
494    #undef LINK_SIZE
495    #define LINK_SIZE 1
496    
497    #define PUT(a,n,d)   \
498      (a[n] = (d))
499    
500    #define GET(a,n) \
501      (a[n])
502    
503    /* Keep it positive */
504    #define MAX_PATTERN_SIZE (1 << 30)
505    
506    #else
507    #error Unsupported compiling mode
508    #endif /* COMPILE_PCRE[8|16|32] */
509    
510  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
511    
# Line 378  is automated on Unix systems via the "co Line 516  is automated on Unix systems via the "co
516  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
517  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
518    
519    #if defined COMPILE_PCRE8
520    
521    #define IMM2_SIZE 2
522    
523  #define PUT2(a,n,d)   \  #define PUT2(a,n,d)   \
524    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
525    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
# Line 385  capturing parenthesis numbers in back re Line 527  capturing parenthesis numbers in back re
527  #define GET2(a,n) \  #define GET2(a,n) \
528    (((a)[n] << 8) | (a)[(n)+1])    (((a)[n] << 8) | (a)[(n)+1])
529    
530  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += 2  #elif defined COMPILE_PCRE16
531    
532    #define IMM2_SIZE 1
533    
534  /* When UTF-8 encoding is being used, a character is no longer just a single  #define PUT2(a,n,d)   \
535  byte. The macros for character handling generate simple sequences when used in     a[n] = d
536  byte-mode, and more complicated ones for UTF-8 characters. BACKCHAR should  
537  never be called in byte mode. To make sure it can never even appear when UTF-8  #define GET2(a,n) \
538  support is omitted, we don't even define it. */     a[n]
539    
540    #elif defined COMPILE_PCRE32
541    
542    #define IMM2_SIZE 1
543    
544    #define PUT2(a,n,d)   \
545       a[n] = d
546    
547    #define GET2(a,n) \
548       a[n]
549    
550    #else
551    #error Unsupported compiling mode
552    #endif /* COMPILE_PCRE[8|16|32] */
553    
554    #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
555    
556    /* The maximum length of a MARK name is currently one data unit; it may be
557    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
558    
559  #ifndef SUPPORT_UTF8  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
560    #define MAX_MARK ((1u << 16) - 1)
561    #else
562    #define MAX_MARK ((1u << 8) - 1)
563    #endif
564    
565    /* When UTF encoding is being used, a character is no longer just a single
566    byte. The macros for character handling generate simple sequences when used in
567    character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
568    and other macros are not used when UTF is not supported, so they are not
569    defined. To make sure they can never even appear when UTF support is omitted,
570    we don't even define them. */
571    
572    #ifndef SUPPORT_UTF
573    
574    /* #define MAX_VALUE_FOR_SINGLE_CHAR */
575    /* #define HAS_EXTRALEN(c) */
576    /* #define GET_EXTRALEN(c) */
577    /* #define NOT_FIRSTCHAR(c) */
578  #define GETCHAR(c, eptr) c = *eptr;  #define GETCHAR(c, eptr) c = *eptr;
579  #define GETCHARTEST(c, eptr) c = *eptr;  #define GETCHARTEST(c, eptr) c = *eptr;
580  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
581  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
582  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
583    #define RAWUCHAR(eptr) (*(eptr))
584    #define RAWUCHARINC(eptr) (*(eptr)++)
585    #define RAWUCHARTEST(eptr) (*(eptr))
586    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
587    /* #define GETCHARLENTEST(c, eptr, len) */
588  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
589    /* #define FORWARDCHAR(eptr) */
590    /* #define ACROSSCHAR(condition, eptr, action) */
591    
592  #else   /* SUPPORT_UTF8 */  #else   /* SUPPORT_UTF */
593    
594  /* 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. */  
595    
596  #define GETCHAR(c, eptr) \  #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
597    c = *eptr; \  
598    if (c >= 0xc0) \  /* Base macro to pick up the remaining bytes of a UTF-8 character, not
599    advancing the pointer. */
600    
601    #define GETUTF8(c, eptr) \
602      { \      { \
603      int gcii; \      if ((c & 0x20) == 0) \
604      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
605      int gcss = 6*gcaa; \      else if ((c & 0x10) == 0) \
606      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
607      for (gcii = 1; gcii <= gcaa; gcii++) \      else if ((c & 0x08) == 0) \
608          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
609          ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
610        else if ((c & 0x04) == 0) \
611          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
612              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
613              (eptr[4] & 0x3f); \
614        else \
615          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
616              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
617              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
618        }
619    
620    /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
621    the pointer. */
622    
623    #define GETUTF8INC(c, eptr) \
624        { \
625        if ((c & 0x20) == 0) \
626          c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
627        else if ((c & 0x10) == 0) \
628          { \
629          c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
630          eptr += 2; \
631          } \
632        else if ((c & 0x08) == 0) \
633        { \        { \
634        gcss -= 6; \        c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
635        c |= (eptr[gcii] & 0x3f) << gcss; \            ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
636          eptr += 3; \
637          } \
638        else if ((c & 0x04) == 0) \
639          { \
640          c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
641              ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
642              (eptr[3] & 0x3f); \
643          eptr += 4; \
644          } \
645        else \
646          { \
647          c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
648              ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
649              ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
650          eptr += 5; \
651        } \        } \
652      }      }
653    
654    #if defined COMPILE_PCRE8
655    
656    /* These macros were originally written in the form of loops that used data
657    from the tables whose names start with PRIV(utf8_table). They were rewritten by
658    a user so as not to use loops, because in some environments this gives a
659    significant performance advantage, and it seems never to do any harm. */
660    
661    /* Tells the biggest code point which can be encoded as a single character. */
662    
663    #define MAX_VALUE_FOR_SINGLE_CHAR 127
664    
665    /* Tests whether the code point needs extra characters to decode. */
666    
667    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
668    
669    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
670    Otherwise it has an undefined behaviour. */
671    
672    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
673    
674    /* Returns TRUE, if the given character is not the first character
675    of a UTF sequence. */
676    
677    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
678    
679    /* Get the next UTF-8 character, not advancing the pointer. This is called when
680    we know we are in UTF-8 mode. */
681    
682    #define GETCHAR(c, eptr) \
683      c = *eptr; \
684      if (c >= 0xc0) GETUTF8(c, eptr);
685    
686  /* 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
687  pointer. */  pointer. */
688    
689  #define GETCHARTEST(c, eptr) \  #define GETCHARTEST(c, eptr) \
690    c = *eptr; \    c = *eptr; \
691    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; \  
       } \  
     }  
692    
693  /* 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
694  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
695    
696  #define GETCHARINC(c, eptr) \  #define GETCHARINC(c, eptr) \
697    c = *eptr++; \    c = *eptr++; \
698    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; \  
       } \  
     }  
699    
700  /* 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.
701    This is called when we don't know if we are in UTF-8 mode. */
702    
703  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
704    c = *eptr++; \    c = *eptr++; \
705    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
706    
707    /* Base macro to pick up the remaining bytes of a UTF-8 character, not
708    advancing the pointer, incrementing the length. */
709    
710    #define GETUTF8LEN(c, eptr, len) \
711      { \      { \
712      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \      if ((c & 0x20) == 0) \
713      int gcss = 6*gcaa; \        { \
714      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
715      while (gcaa-- > 0) \        len++; \
716          } \
717        else if ((c & 0x10)  == 0) \
718        { \        { \
719        gcss -= 6; \        c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
720        c |= (*eptr++ & 0x3f) << gcss; \        len += 2; \
721          } \
722        else if ((c & 0x08)  == 0) \
723          {\
724          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
725              ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
726          len += 3; \
727          } \
728        else if ((c & 0x04)  == 0) \
729          { \
730          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
731              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
732              (eptr[4] & 0x3f); \
733          len += 4; \
734          } \
735        else \
736          {\
737          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
738              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
739              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
740          len += 5; \
741        } \        } \
742      }      }
743    
# Line 478  if there are extra bytes. This is called Line 746  if there are extra bytes. This is called
746    
747  #define GETCHARLEN(c, eptr, len) \  #define GETCHARLEN(c, eptr, len) \
748    c = *eptr; \    c = *eptr; \
749    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; \  
     }  
750    
751  /* 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
752  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
753  know we are in UTF-8 mode. */  do not know if we are in UTF-8 mode. */
754    
755  #define GETCHARLENTEST(c, eptr, len) \  #define GETCHARLENTEST(c, eptr, len) \
756    c = *eptr; \    c = *eptr; \
757    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
758      { \  
759      int gcii; \  /* Returns the next uchar, not advancing the pointer. This is called when
760      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  we know we are in UTF mode. */
761      int gcss = 6*gcaa; \  
762      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  #define RAWUCHAR(eptr) \
763      for (gcii = 1; gcii <= gcaa; gcii++) \    (*(eptr))
764        { \  
765        gcss -= 6; \  /* Returns the next uchar, advancing the pointer. This is called when
766        c |= (eptr[gcii] & 0x3f) << gcss; \  we know we are in UTF mode. */
767        } \  
768      len += gcaa; \  #define RAWUCHARINC(eptr) \
769      }    (*(eptr)++)
770    
771    /* Returns the next uchar, testing for UTF mode, and not advancing the
772    pointer. */
773    
774    #define RAWUCHARTEST(eptr) \
775      (*(eptr))
776    
777    /* Returns the next uchar, testing for UTF mode, advancing the
778    pointer. */
779    
780    #define RAWUCHARINCTEST(eptr) \
781      (*(eptr)++)
782    
783  /* 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
784  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 786  because almost all calls are already wit
786    
787  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
788    
789  #endif  /* Same as above, just in the other direction. */
790    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
791    
792    /* Same as above, but it allows a fully customizable form. */
793    #define ACROSSCHAR(condition, eptr, action) \
794      while((condition) && ((eptr) & 0xc0) == 0x80) action
795    
796  /* In case there is no definition of offsetof() provided - though any proper  #elif defined COMPILE_PCRE16
 Standard C system should have one. */  
797    
798  #ifndef offsetof  /* Tells the biggest code point which can be encoded as a single character. */
799  #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
800  #endif  #define MAX_VALUE_FOR_SINGLE_CHAR 65535
801    
802    /* Tests whether the code point needs extra characters to decode. */
803    
804    #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
805    
806    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
807    Otherwise it has an undefined behaviour. */
808    
809    #define GET_EXTRALEN(c) 1
810    
811    /* Returns TRUE, if the given character is not the first character
812    of a UTF sequence. */
813    
814    #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
815    
816    /* Base macro to pick up the low surrogate of a UTF-16 character, not
817    advancing the pointer. */
818    
819    #define GETUTF16(c, eptr) \
820       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
821    
822    /* Get the next UTF-16 character, not advancing the pointer. This is called when
823    we know we are in UTF-16 mode. */
824    
825    #define GETCHAR(c, eptr) \
826      c = *eptr; \
827      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
828    
829    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
830    pointer. */
831    
832    #define GETCHARTEST(c, eptr) \
833      c = *eptr; \
834      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
835    
836    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
837    the pointer. */
838    
839    #define GETUTF16INC(c, eptr) \
840       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
841    
842    /* Get the next UTF-16 character, advancing the pointer. This is called when we
843    know we are in UTF-16 mode. */
844    
845    #define GETCHARINC(c, eptr) \
846      c = *eptr++; \
847      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
848    
849    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
850    This is called when we don't know if we are in UTF-16 mode. */
851    
852    #define GETCHARINCTEST(c, eptr) \
853      c = *eptr++; \
854      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
855    
856    /* Base macro to pick up the low surrogate of a UTF-16 character, not
857    advancing the pointer, incrementing the length. */
858    
859    #define GETUTF16LEN(c, eptr, len) \
860       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
861    
862    /* Get the next UTF-16 character, not advancing the pointer, incrementing
863    length if there is a low surrogate. This is called when we know we are in
864    UTF-16 mode. */
865    
866    #define GETCHARLEN(c, eptr, len) \
867      c = *eptr; \
868      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
869    
870    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
871    pointer, incrementing length if there is a low surrogate. This is called when
872    we do not know if we are in UTF-16 mode. */
873    
874    #define GETCHARLENTEST(c, eptr, len) \
875      c = *eptr; \
876      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
877    
878    /* Returns the next uchar, not advancing the pointer. This is called when
879    we know we are in UTF mode. */
880    
881    #define RAWUCHAR(eptr) \
882      (*(eptr))
883    
884    /* Returns the next uchar, advancing the pointer. This is called when
885    we know we are in UTF mode. */
886    
887    #define RAWUCHARINC(eptr) \
888      (*(eptr)++)
889    
890    /* Returns the next uchar, testing for UTF mode, and not advancing the
891    pointer. */
892    
893    #define RAWUCHARTEST(eptr) \
894      (*(eptr))
895    
896    /* Returns the next uchar, testing for UTF mode, advancing the
897    pointer. */
898    
899    #define RAWUCHARINCTEST(eptr) \
900      (*(eptr)++)
901    
902    /* If the pointer is not at the start of a character, move it back until
903    it is. This is called only in UTF-16 mode - we don't put a test within the
904    macro because almost all calls are already within a block of UTF-16 only
905    code. */
906    
907    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
908    
909    /* Same as above, just in the other direction. */
910    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
911    
912    /* Same as above, but it allows a fully customizable form. */
913    #define ACROSSCHAR(condition, eptr, action) \
914      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
915    
916    #elif defined COMPILE_PCRE32
917    
918    /* These are trivial for the 32-bit library, since all UTF-32 characters fit
919    into one pcre_uchar unit. */
920    #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
921    #define HAS_EXTRALEN(c) (0)
922    #define GET_EXTRALEN(c) (0)
923    #define NOT_FIRSTCHAR(c) (0)
924    
925    #define UTF32_MASK (0x1fffffu)
926    
927    /* Get the next UTF-32 character, not advancing the pointer. This is called when
928    we know we are in UTF-32 mode. */
929    
930    #define GETCHAR(c, eptr) \
931      c = (*eptr) & UTF32_MASK;
932    
933    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
934    pointer. */
935    
936    #define GETCHARTEST(c, eptr) \
937      c = *eptr; \
938      if (utf) c &= UTF32_MASK;
939    
940    /* Get the next UTF-32 character, advancing the pointer. This is called when we
941    know we are in UTF-32 mode. */
942    
943    #define GETCHARINC(c, eptr) \
944      c = (*eptr++) & UTF32_MASK;
945    
946    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
947    This is called when we don't know if we are in UTF-32 mode. */
948    
949    #define GETCHARINCTEST(c, eptr) \
950      c = *eptr++; \
951      if (utf) c &= UTF32_MASK;
952    
953    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
954    length (since all UTF-32 is of length 1). This is called when we know we are in
955    UTF-32 mode. */
956    
957    #define GETCHARLEN(c, eptr, len) \
958      GETCHAR(c, eptr)
959    
960    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
961    pointer, not incrementing the length (since all UTF-32 is of length 1).
962    This is called when we do not know if we are in UTF-32 mode. */
963    
964    #define GETCHARLENTEST(c, eptr, len) \
965      GETCHARTEST(c, eptr)
966    
967    /* Returns the next uchar, not advancing the pointer. This is called when
968    we know we are in UTF mode. */
969    
970    #define RAWUCHAR(eptr) \
971      (*(eptr) & UTF32_MASK)
972    
973    /* Returns the next uchar, advancing the pointer. This is called when
974    we know we are in UTF mode. */
975    
976    #define RAWUCHARINC(eptr) \
977      (*(eptr)++ & UTF32_MASK)
978    
979    /* Returns the next uchar, testing for UTF mode, and not advancing the
980    pointer. */
981    
982    #define RAWUCHARTEST(eptr) \
983      (utf ? (*(eptr) & UTF32_MASK) : *(eptr))
984    
985    /* Returns the next uchar, testing for UTF mode, advancing the
986    pointer. */
987    
988    #define RAWUCHARINCTEST(eptr) \
989      (utf ? (*(eptr)++ & UTF32_MASK) : *(eptr)++)
990    
991    /* If the pointer is not at the start of a character, move it back until
992    it is. This is called only in UTF-32 mode - we don't put a test within the
993    macro because almost all calls are already within a block of UTF-32 only
994    code.
995    These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
996    
997    #define BACKCHAR(eptr) do { } while (0)
998    
999    /* Same as above, just in the other direction. */
1000    #define FORWARDCHAR(eptr) do { } while (0)
1001    
1002    /* Same as above, but it allows a fully customizable form. */
1003    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1004    
1005    #else
1006    #error Unsupported compiling mode
1007    #endif /* COMPILE_PCRE[8|16|32] */
1008    
1009    #endif  /* SUPPORT_UTF */
1010    
1011    /* Tests for Unicode horizontal and vertical whitespace characters must check a
1012    number of different values. Using a switch statement for this generates the
1013    fastest code (no loop, no memory access), and there are several places in the
1014    interpreter code where this happens. In order to ensure that all the case lists
1015    remain in step, we use macros so that there is only one place where the lists
1016    are defined.
1017    
1018    These values are also required as lists in pcre_compile.c when processing \h,
1019    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1020    macros that define the values are here so that all the definitions are
1021    together. The lists must be in ascending character order, terminated by
1022    NOTACHAR (which is 0xffffffff).
1023    
1024    Any changes should ensure that the various macros are kept in step with each
1025    other. NOTE: The values also appear in pcre_jit_compile.c. */
1026    
1027    /* ------ ASCII/Unicode environments ------ */
1028    
1029    #ifndef EBCDIC
1030    
1031    #define HSPACE_LIST \
1032      CHAR_HT, CHAR_SPACE, 0xa0, \
1033      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1034      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1035      NOTACHAR
1036    
1037    #define HSPACE_MULTIBYTE_CASES \
1038      case 0x1680:  /* OGHAM SPACE MARK */ \
1039      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1040      case 0x2000:  /* EN QUAD */ \
1041      case 0x2001:  /* EM QUAD */ \
1042      case 0x2002:  /* EN SPACE */ \
1043      case 0x2003:  /* EM SPACE */ \
1044      case 0x2004:  /* THREE-PER-EM SPACE */ \
1045      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1046      case 0x2006:  /* SIX-PER-EM SPACE */ \
1047      case 0x2007:  /* FIGURE SPACE */ \
1048      case 0x2008:  /* PUNCTUATION SPACE */ \
1049      case 0x2009:  /* THIN SPACE */ \
1050      case 0x200A:  /* HAIR SPACE */ \
1051      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1052      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1053      case 0x3000   /* IDEOGRAPHIC SPACE */
1054    
1055    #define HSPACE_BYTE_CASES \
1056      case CHAR_HT: \
1057      case CHAR_SPACE: \
1058      case 0xa0     /* NBSP */
1059    
1060    #define HSPACE_CASES \
1061      HSPACE_BYTE_CASES: \
1062      HSPACE_MULTIBYTE_CASES
1063    
1064    #define VSPACE_LIST \
1065      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1066    
1067    #define VSPACE_MULTIBYTE_CASES \
1068      case 0x2028:    /* LINE SEPARATOR */ \
1069      case 0x2029     /* PARAGRAPH SEPARATOR */
1070    
1071    #define VSPACE_BYTE_CASES \
1072      case CHAR_LF: \
1073      case CHAR_VT: \
1074      case CHAR_FF: \
1075      case CHAR_CR: \
1076      case CHAR_NEL
1077    
1078    #define VSPACE_CASES \
1079      VSPACE_BYTE_CASES: \
1080      VSPACE_MULTIBYTE_CASES
1081    
1082    /* ------ EBCDIC environments ------ */
1083    
1084    #else
1085    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1086    
1087    #define HSPACE_BYTE_CASES \
1088      case CHAR_HT: \
1089      case CHAR_SPACE
1090    
1091    #define HSPACE_CASES HSPACE_BYTE_CASES
1092    
1093    #ifdef EBCDIC_NL25
1094    #define VSPACE_LIST \
1095      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1096    #else
1097    #define VSPACE_LIST \
1098      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1099    #endif
1100    
1101    #define VSPACE_BYTE_CASES \
1102      case CHAR_LF: \
1103      case CHAR_VT: \
1104      case CHAR_FF: \
1105      case CHAR_CR: \
1106      case CHAR_NEL
1107    
1108    #define VSPACE_CASES VSPACE_BYTE_CASES
1109    #endif  /* EBCDIC */
1110    
1111    /* ------ End of whitespace macros ------ */
1112    
 /* These are the public options that can change during matching. */  
1113    
 #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)  
1114    
1115  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
1116  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 1118  are in a 16-bit flags word. From release
1118  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1119  compatibility. */  compatibility. */
1120    
1121  #define PCRE_NOPARTIAL     0x0001  /* can't use partial with this regex */  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1122  #define PCRE_FIRSTSET      0x0002  /* first_byte is set */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1123  #define PCRE_REQCHSET      0x0004  /* req_byte is set */  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
1124  #define PCRE_STARTLINE     0x0008  /* start after \n for multiline */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1125  #define PCRE_JCHANGED      0x0010  /* j option used in regex */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1126  #define PCRE_HASCRORLF     0x0020  /* explicit \r or \n in pattern */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
1127    #define PCRE_RCH_CASELESS  0x0080  /* caseless requested char */
1128    #define PCRE_STARTLINE     0x0100  /* start after \n for multiline */
1129    #define PCRE_NOPARTIAL     0x0200  /* can't use partial with this regex */
1130    #define PCRE_JCHANGED      0x0400  /* j option used in regex */
1131    #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1132    #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1133    
1134    #if defined COMPILE_PCRE8
1135    #define PCRE_MODE          PCRE_MODE8
1136    #elif defined COMPILE_PCRE16
1137    #define PCRE_MODE          PCRE_MODE16
1138    #elif defined COMPILE_PCRE32
1139    #define PCRE_MODE          PCRE_MODE32
1140    #endif
1141    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1142    
1143  /* Options for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1144    
1145  #define PCRE_STUDY_MAPPED   0x01     /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
1146  #define PCRE_STUDY_MINLEN   0x02     /* a minimum length field exists */  #define PCRE_STUDY_MINLEN  0x0002  /* a minimum length field exists */
1147    
1148  /* Masks for identifying the public options that are permitted at compile  /* Masks for identifying the public options that are permitted at compile
1149  time, run time, or study time, respectively. */  time, run time, or study time, respectively. */
# Line 562  time, run time, or study time, respectiv Line 1156  time, run time, or study time, respectiv
1156     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1157     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| \
1158     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1159     PCRE_JAVASCRIPT_COMPAT)     PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
1160    
1161  #define PUBLIC_EXEC_OPTIONS \  #define PUBLIC_EXEC_OPTIONS \
1162    (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 1169  time, run time, or study time, respectiv
1169     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1170     PCRE_NO_START_OPTIMIZE)     PCRE_NO_START_OPTIMIZE)
1171    
1172  #define PUBLIC_STUDY_OPTIONS 0   /* None defined */  #define PUBLIC_STUDY_OPTIONS \
1173       (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1174        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1175    
1176  /* 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. */  
1177    
1178  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1179    
1180  /* Negative values for the firstchar and reqchar variables */  /* This variable is used to detect a loaded regular expression
1181    in different endianness. */
1182    
1183  #define REQ_UNSET (-2)  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
 #define REQ_NONE  (-1)  
1184    
1185  /* 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
1186  req_byte match. */  req_byte match. */
1187    
1188  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1189    
 /* 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 */  
   
1190  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1191  environments where these macros are defined elsewhere. Unfortunately, there  environments where these macros are defined elsewhere. Unfortunately, there
1192  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 1215  for) in a minority area (EBCDIC platform
1215  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
1216  macros to give the functions distinct names. */  macros to give the functions distinct names. */
1217    
1218  #ifndef SUPPORT_UTF8  #ifndef SUPPORT_UTF
1219    
1220  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1221  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
1222    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1223    characters, a common practice has been to use its NL (0x15) character as the
1224    line terminator in C-like processing environments. However, sometimes the LF
1225    (0x25) character is used instead, according to this Unicode document:
1226    
1227    http://unicode.org/standard/reports/tr13/tr13-5.html
1228    
1229    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1230    instead. Whichever is *not* chosen is defined as NEL.
1231    
1232    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1233    same code point. */
1234    
1235    #ifdef EBCDIC
1236    
1237    #ifndef EBCDIC_NL25
1238    #define CHAR_NL                     '\x15'
1239    #define CHAR_NEL                    '\x25'
1240    #define STR_NL                      "\x15"
1241    #define STR_NEL                     "\x25"
1242    #else
1243    #define CHAR_NL                     '\x25'
1244    #define CHAR_NEL                    '\x15'
1245    #define STR_NL                      "\x25"
1246    #define STR_NEL                     "\x15"
1247    #endif
1248    
1249    #define CHAR_LF                     CHAR_NL
1250    #define STR_LF                      STR_NL
1251    
1252    #define CHAR_ESC                    '\047'
1253    #define CHAR_DEL                    '\007'
1254    #define STR_ESC                     "\047"
1255    #define STR_DEL                     "\007"
1256    
1257    #else  /* Not EBCDIC */
1258    
1259    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1260    compatibility. NEL is the Unicode newline character; make sure it is
1261    a positive value. */
1262    
1263    #define CHAR_LF                     '\n'
1264    #define CHAR_NL                     CHAR_LF
1265    #define CHAR_NEL                    ((unsigned char)'\x85')
1266    #define CHAR_ESC                    '\033'
1267    #define CHAR_DEL                    '\177'
1268    
1269    #define STR_LF                      "\n"
1270    #define STR_NL                      STR_LF
1271    #define STR_NEL                     "\x85"
1272    #define STR_ESC                     "\033"
1273    #define STR_DEL                     "\177"
1274    
1275    #endif  /* EBCDIC */
1276    
1277    /* The remaining definitions work in both environments. */
1278    
1279  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1280  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1281  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1282  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1283  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1284  #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  
1285    
1286  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1287  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 746  so that PCRE works on both ASCII and EBC Line 1383  so that PCRE works on both ASCII and EBC
1383  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1384  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1385  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1386  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1387  #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  
1388    
1389  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1390  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 857  so that PCRE works on both ASCII and EBC Line 1486  so that PCRE works on both ASCII and EBC
1486  #define STRING_COMMIT0              "COMMIT\0"  #define STRING_COMMIT0              "COMMIT\0"
1487  #define STRING_F0                   "F\0"  #define STRING_F0                   "F\0"
1488  #define STRING_FAIL0                "FAIL\0"  #define STRING_FAIL0                "FAIL\0"
1489    #define STRING_MARK0                "MARK\0"
1490  #define STRING_PRUNE0               "PRUNE\0"  #define STRING_PRUNE0               "PRUNE\0"
1491  #define STRING_SKIP0                "SKIP\0"  #define STRING_SKIP0                "SKIP\0"
1492  #define STRING_THEN                 "THEN"  #define STRING_THEN                 "THEN"
# Line 878  so that PCRE works on both ASCII and EBC Line 1508  so that PCRE works on both ASCII and EBC
1508    
1509  #define STRING_DEFINE               "DEFINE"  #define STRING_DEFINE               "DEFINE"
1510    
1511  #define STRING_CR_RIGHTPAR          "CR)"  #define STRING_CR_RIGHTPAR             "CR)"
1512  #define STRING_LF_RIGHTPAR          "LF)"  #define STRING_LF_RIGHTPAR             "LF)"
1513  #define STRING_CRLF_RIGHTPAR        "CRLF)"  #define STRING_CRLF_RIGHTPAR           "CRLF)"
1514  #define STRING_ANY_RIGHTPAR         "ANY)"  #define STRING_ANY_RIGHTPAR            "ANY)"
1515  #define STRING_ANYCRLF_RIGHTPAR     "ANYCRLF)"  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1516  #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1517  #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1518  #define STRING_UTF8_RIGHTPAR        "UTF8)"  #ifdef COMPILE_PCRE8
1519    #define STRING_UTF_RIGHTPAR            "UTF8)"
1520    #endif
1521    #ifdef COMPILE_PCRE16
1522    #define STRING_UTF_RIGHTPAR            "UTF16)"
1523    #endif
1524    #ifdef COMPILE_PCRE32
1525    #define STRING_UTF_RIGHTPAR            "UTF32)"
1526    #endif
1527    #define STRING_UCP_RIGHTPAR            "UCP)"
1528    #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1529    
1530  #else  /* SUPPORT_UTF8 */  #else  /* SUPPORT_UTF */
1531    
1532  /* 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
1533  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 1537  only. */
1537  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1538  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1539  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1540  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1541    #define CHAR_NL                     CHAR_LF
1542    #define CHAR_NEL                    ((unsigned char)'\x85')
1543  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1544  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1545  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
# Line 1109  only. */ Line 1751  only. */
1751  #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"
1752  #define STRING_F0                   STR_F "\0"  #define STRING_F0                   STR_F "\0"
1753  #define STRING_FAIL0                STR_F STR_A STR_I STR_L "\0"  #define STRING_FAIL0                STR_F STR_A STR_I STR_L "\0"
1754    #define STRING_MARK0                STR_M STR_A STR_R STR_K "\0"
1755  #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"
1756  #define STRING_SKIP0                STR_S STR_K STR_I STR_P "\0"  #define STRING_SKIP0                STR_S STR_K STR_I STR_P "\0"
1757  #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 1773  only. */
1773    
1774  #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
1775    
1776  #define STRING_CR_RIGHTPAR          STR_C STR_R STR_RIGHT_PARENTHESIS  #define STRING_CR_RIGHTPAR             STR_C STR_R STR_RIGHT_PARENTHESIS
1777  #define STRING_LF_RIGHTPAR          STR_L STR_F STR_RIGHT_PARENTHESIS  #define STRING_LF_RIGHTPAR             STR_L STR_F STR_RIGHT_PARENTHESIS
1778  #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
1779  #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
1780  #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
1781  #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
1782  #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
1783  #define STRING_UTF8_RIGHTPAR        STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS  #ifdef COMPILE_PCRE8
1784    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1785    #endif
1786    #ifdef COMPILE_PCRE16
1787    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1788    #endif
1789    #ifdef COMPILE_PCRE32
1790    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1791    #endif
1792    #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1793    #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
1794    
1795  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
1796    
1797  /* Escape items that are just an encoding of a particular data value. */  /* Escape items that are just an encoding of a particular data value. */
1798    
# Line 1152  only. */ Line 1805  only. */
1805  #endif  #endif
1806    
1807  #ifndef ESC_n  #ifndef ESC_n
1808  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1809  #endif  #endif
1810    
1811  #ifndef ESC_r  #ifndef ESC_r
# Line 1170  only. */ Line 1823  only. */
1823    
1824  #define PT_ANY        0    /* Any property - matches all chars */  #define PT_ANY        0    /* Any property - matches all chars */
1825  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */
1826  #define PT_GC         2    /* General characteristic (e.g. L) */  #define PT_GC         2    /* Specified general characteristic (e.g. L) */
1827  #define PT_PC         3    /* Particular characteristic (e.g. Lu) */  #define PT_PC         3    /* Specified particular characteristic (e.g. Lu) */
1828  #define PT_SC         4    /* Script (e.g. Han) */  #define PT_SC         4    /* Script (e.g. Han) */
1829    #define PT_ALNUM      5    /* Alphanumeric - the union of L and N */
1830    #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1831    #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1832    #define PT_WORD       8    /* Word - L plus N plus underscore */
1833    #define PT_CLIST      9    /* Pseudo-property: match character list */
1834    
1835  /* 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
1836  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1837    
1838  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1839  #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 1846  contain UTF-8 characters with values gre
1846    
1847  /* 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
1848  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
1849  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
1850  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
1851  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
1852  (which is used for [^] in JavaScript compatibility mode).  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1853    non-DOTALL mode, "." behaves like \N.
1854    
1855    The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1856    when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1857    They must be contiguous, and remain in order so that the replacements can be
1858    looked up from a table.
1859    
1860  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
1861  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
1862  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
1863  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
1864  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.
1865  */  */
1866    
1867  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,
1868         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,
1869         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,
1870         ESC_REF };         ESC_E, ESC_Q, ESC_g, ESC_k,
1871           ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1872    
1873  /* 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
1874  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.
1875    
1876  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1877  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
1878  "coptable" in pcre_dfa_exec.c that must be updated. */  "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1879    
1880  enum {  enum {
1881    OP_END,            /* 0 End of pattern */    OP_END,            /* 0 End of pattern */
# Line 1229  enum { Line 1893  enum {
1893    OP_WHITESPACE,         /*  9 \s */    OP_WHITESPACE,         /*  9 \s */
1894    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1895    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1896    OP_ANY,            /* 12 Match any character (subject to DOTALL) */  
1897    OP_ALLANY,         /* 13 Match any character (not subject to DOTALL) */    OP_ANY,            /* 12 Match any character except newline (\N) */
1898      OP_ALLANY,         /* 13 Match any character */
1899    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 */
1900    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
1901    OP_PROP,           /* 16 \p (Unicode property) */    OP_PROP,           /* 16 \p (Unicode property) */
# Line 1240  enum { Line 1905  enum {
1905    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1906    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1907    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1908    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) */
1909    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1910    
1911    OP_OPT,            /* 25 Set runtime options */    OP_CIRC,           /* 25 Start of line - not multiline */
1912    OP_CIRC,           /* 26 Start of line - varies with multiline switch */    OP_CIRCM,          /* 26 Start of line - multiline */
1913    OP_DOLL,           /* 27 End of line - varies with multiline switch */    OP_DOLL,           /* 27 End of line - not multiline */
1914    OP_CHAR,           /* 28 Match one character, casefully */    OP_DOLLM,          /* 28 End of line - multiline */
1915    OP_CHARNC,         /* 29 Match one character, caselessly */    OP_CHAR,           /* 29 Match one character, casefully */
1916    OP_NOT,            /* 30 Match one character, not the following one */    OP_CHARI,          /* 30 Match one character, caselessly */
1917      OP_NOT,            /* 31 Match one character, not the given one, casefully */
1918    OP_STAR,           /* 31 The maximizing and minimizing versions of */    OP_NOTI,           /* 32 Match one character, not the given one, caselessly */
1919    OP_MINSTAR,        /* 32 these six opcodes must come in pairs, with */  
1920    OP_PLUS,           /* 33 the minimizing one second. */    /* The following sets of 13 opcodes must always be kept in step because
1921    OP_MINPLUS,        /* 34 This first set applies to single characters.*/    the offset from the first one is used to generate the others. */
1922    OP_QUERY,          /* 35 */  
1923    OP_MINQUERY,       /* 36 */    /**** Single characters, caseful, must precede the caseless ones ****/
1924    
1925    OP_UPTO,           /* 37 From 0 to n matches */    OP_STAR,           /* 33 The maximizing and minimizing versions of */
1926    OP_MINUPTO,        /* 38 */    OP_MINSTAR,        /* 34 these six opcodes must come in pairs, with */
1927    OP_EXACT,          /* 39 Exactly n matches */    OP_PLUS,           /* 35 the minimizing one second. */
1928      OP_MINPLUS,        /* 36 */
1929    OP_POSSTAR,        /* 40 Possessified star */    OP_QUERY,          /* 37 */
1930    OP_POSPLUS,        /* 41 Possessified plus */    OP_MINQUERY,       /* 38 */
1931    OP_POSQUERY,       /* 42 Posesssified query */  
1932    OP_POSUPTO,        /* 43 Possessified upto */    OP_UPTO,           /* 39 From 0 to n matches of one character, caseful*/
1933      OP_MINUPTO,        /* 40 */
1934    OP_NOTSTAR,        /* 44 The maximizing and minimizing versions of */    OP_EXACT,          /* 41 Exactly n matches */
1935    OP_NOTMINSTAR,     /* 45 these six opcodes must come in pairs, with */  
1936    OP_NOTPLUS,        /* 46 the minimizing one second. They must be in */    OP_POSSTAR,        /* 42 Possessified star, caseful */
1937    OP_NOTMINPLUS,     /* 47 exactly the same order as those above. */    OP_POSPLUS,        /* 43 Possessified plus, caseful */
1938    OP_NOTQUERY,       /* 48 This set applies to "not" single characters. */    OP_POSQUERY,       /* 44 Posesssified query, caseful */
1939    OP_NOTMINQUERY,    /* 49 */    OP_POSUPTO,        /* 45 Possessified upto, caseful */
1940    
1941    OP_NOTUPTO,        /* 50 From 0 to n matches */    /**** Single characters, caseless, must follow the caseful ones */
1942    OP_NOTMINUPTO,     /* 51 */  
1943    OP_NOTEXACT,       /* 52 Exactly n matches */    OP_STARI,          /* 46 */
1944      OP_MINSTARI,       /* 47 */
1945    OP_NOTPOSSTAR,     /* 53 Possessified versions */    OP_PLUSI,          /* 48 */
1946    OP_NOTPOSPLUS,     /* 54 */    OP_MINPLUSI,       /* 49 */
1947    OP_NOTPOSQUERY,    /* 55 */    OP_QUERYI,         /* 50 */
1948    OP_NOTPOSUPTO,     /* 56 */    OP_MINQUERYI,      /* 51 */
1949    
1950    OP_TYPESTAR,       /* 57 The maximizing and minimizing versions of */    OP_UPTOI,          /* 52 From 0 to n matches of one character, caseless */
1951    OP_TYPEMINSTAR,    /* 58 these six opcodes must come in pairs, with */    OP_MINUPTOI,       /* 53 */
1952    OP_TYPEPLUS,       /* 59 the minimizing one second. These codes must */    OP_EXACTI,         /* 54 */
1953    OP_TYPEMINPLUS,    /* 60 be in exactly the same order as those above. */  
1954    OP_TYPEQUERY,      /* 61 This set applies to character types such as \d */    OP_POSSTARI,       /* 55 Possessified star, caseless */
1955    OP_TYPEMINQUERY,   /* 62 */    OP_POSPLUSI,       /* 56 Possessified plus, caseless */
1956      OP_POSQUERYI,      /* 57 Posesssified query, caseless */
1957    OP_TYPEUPTO,       /* 63 From 0 to n matches */    OP_POSUPTOI,       /* 58 Possessified upto, caseless */
1958    OP_TYPEMINUPTO,    /* 64 */  
1959    OP_TYPEEXACT,      /* 65 Exactly n matches */    /**** The negated ones must follow the non-negated ones, and match them ****/
1960      /**** Negated single character, caseful; must precede the caseless ones ****/
1961    OP_TYPEPOSSTAR,    /* 66 Possessified versions */  
1962    OP_TYPEPOSPLUS,    /* 67 */    OP_NOTSTAR,        /* 59 The maximizing and minimizing versions of */
1963    OP_TYPEPOSQUERY,   /* 68 */    OP_NOTMINSTAR,     /* 60 these six opcodes must come in pairs, with */
1964    OP_TYPEPOSUPTO,    /* 69 */    OP_NOTPLUS,        /* 61 the minimizing one second. They must be in */
1965      OP_NOTMINPLUS,     /* 62 exactly the same order as those above. */
1966    OP_CRSTAR,         /* 70 The maximizing and minimizing versions of */    OP_NOTQUERY,       /* 63 */
1967    OP_CRMINSTAR,      /* 71 all these opcodes must come in pairs, with */    OP_NOTMINQUERY,    /* 64 */
1968    OP_CRPLUS,         /* 72 the minimizing one second. These codes must */  
1969    OP_CRMINPLUS,      /* 73 be in exactly the same order as those above. */    OP_NOTUPTO,        /* 65 From 0 to n matches, caseful */
1970    OP_CRQUERY,        /* 74 These are for character classes and back refs */    OP_NOTMINUPTO,     /* 66 */
1971    OP_CRMINQUERY,     /* 75 */    OP_NOTEXACT,       /* 67 Exactly n matches */
1972    OP_CRRANGE,        /* 76 These are different to the three sets above. */  
1973    OP_CRMINRANGE,     /* 77 */    OP_NOTPOSSTAR,     /* 68 Possessified versions, caseful */
1974      OP_NOTPOSPLUS,     /* 69 */
1975    OP_CLASS,          /* 78 Match a character class, chars < 256 only */    OP_NOTPOSQUERY,    /* 70 */
1976    OP_NCLASS,         /* 79 Same, but the bitmap was created from a negative    OP_NOTPOSUPTO,     /* 71 */
1977                             class - the difference is relevant only when a UTF-8  
1978                             character > 255 is encountered. */    /**** Negated single character, caseless; must follow the caseful ones ****/
1979    
1980    OP_XCLASS,         /* 80 Extended class for handling UTF-8 chars within the    OP_NOTSTARI,       /* 72 */
1981                             class. This does both positive and negative. */    OP_NOTMINSTARI,    /* 73 */
1982      OP_NOTPLUSI,       /* 74 */
1983    OP_REF,            /* 81 Match a back reference */    OP_NOTMINPLUSI,    /* 75 */
1984    OP_RECURSE,        /* 82 Match a numbered subpattern (possibly recursive) */    OP_NOTQUERYI,      /* 76 */
1985    OP_CALLOUT,        /* 83 Call out to external function if provided */    OP_NOTMINQUERYI,   /* 77 */
1986    
1987    OP_ALT,            /* 84 Start of alternation */    OP_NOTUPTOI,       /* 78 From 0 to n matches, caseless */
1988    OP_KET,            /* 85 End of group that doesn't have an unbounded repeat */    OP_NOTMINUPTOI,    /* 79 */
1989    OP_KETRMAX,        /* 86 These two must remain together and in this */    OP_NOTEXACTI,      /* 80 Exactly n matches */
1990    OP_KETRMIN,        /* 87 order. They are for groups the repeat for ever. */  
1991      OP_NOTPOSSTARI,    /* 81 Possessified versions, caseless */
1992    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/    OP_NOTPOSPLUSI,    /* 82 */
1993      OP_NOTPOSQUERYI,   /* 83 */
1994    OP_ASSERT,         /* 88 Positive lookahead */    OP_NOTPOSUPTOI,    /* 84 */
1995    OP_ASSERT_NOT,     /* 89 Negative lookahead */  
1996    OP_ASSERTBACK,     /* 90 Positive lookbehind */    /**** Character types ****/
1997    OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */  
1998    OP_REVERSE,        /* 92 Move pointer back - used in lookbehind assertions */    OP_TYPESTAR,       /* 85 The maximizing and minimizing versions of */
1999      OP_TYPEMINSTAR,    /* 86 these six opcodes must come in pairs, with */
2000    /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,    OP_TYPEPLUS,       /* 87 the minimizing one second. These codes must */
2001    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. */
2002      OP_TYPEQUERY,      /* 89 */
2003    OP_ONCE,           /* 93 Atomic group */    OP_TYPEMINQUERY,   /* 90 */
2004    OP_BRA,            /* 94 Start of non-capturing bracket */  
2005    OP_CBRA,           /* 95 Start of capturing bracket */    OP_TYPEUPTO,       /* 91 From 0 to n matches */
2006    OP_COND,           /* 96 Conditional group */    OP_TYPEMINUPTO,    /* 92 */
2007      OP_TYPEEXACT,      /* 93 Exactly n matches */
2008    
2009      OP_TYPEPOSSTAR,    /* 94 Possessified versions */
2010      OP_TYPEPOSPLUS,    /* 95 */
2011      OP_TYPEPOSQUERY,   /* 96 */
2012      OP_TYPEPOSUPTO,    /* 97 */
2013    
2014      /* These are used for character classes and back references; only the
2015      first six are the same as the sets above. */
2016    
2017      OP_CRSTAR,         /* 98 The maximizing and minimizing versions of */
2018      OP_CRMINSTAR,      /* 99 all these opcodes must come in pairs, with */
2019      OP_CRPLUS,         /* 100 the minimizing one second. These codes must */
2020      OP_CRMINPLUS,      /* 101 be in exactly the same order as those above. */
2021      OP_CRQUERY,        /* 102 */
2022      OP_CRMINQUERY,     /* 103 */
2023    
2024      OP_CRRANGE,        /* 104 These are different to the three sets above. */
2025      OP_CRMINRANGE,     /* 105 */
2026    
2027      /* End of quantifier opcodes */
2028    
2029      OP_CLASS,          /* 106 Match a character class, chars < 256 only */
2030      OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
2031                                  class - the difference is relevant only when a
2032                                  character > 255 is encountered. */
2033      OP_XCLASS,         /* 108 Extended class for handling > 255 chars within the
2034                                  class. This does both positive and negative. */
2035      OP_REF,            /* 109 Match a back reference, casefully */
2036      OP_REFI,           /* 110 Match a back reference, caselessly */
2037      OP_RECURSE,        /* 111 Match a numbered subpattern (possibly recursive) */
2038      OP_CALLOUT,        /* 112 Call out to external function if provided */
2039    
2040      OP_ALT,            /* 113 Start of alternation */
2041      OP_KET,            /* 114 End of group that doesn't have an unbounded repeat */
2042      OP_KETRMAX,        /* 115 These two must remain together and in this */
2043      OP_KETRMIN,        /* 116 order. They are for groups the repeat for ever. */
2044      OP_KETRPOS,        /* 117 Possessive unlimited repeat. */
2045    
2046      /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
2047      asserts must remain in order. */
2048    
2049      OP_REVERSE,        /* 118 Move pointer back - used in lookbehind assertions */
2050      OP_ASSERT,         /* 119 Positive lookahead */
2051      OP_ASSERT_NOT,     /* 120 Negative lookahead */
2052      OP_ASSERTBACK,     /* 121 Positive lookbehind */
2053      OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
2054    
2055      /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
2056      after the assertions, with ONCE first, as there's a test for >= ONCE for a
2057      subpattern that isn't an assertion. The POS versions must immediately follow
2058      the non-POS versions in each case. */
2059    
2060      OP_ONCE,           /* 123 Atomic group, contains captures */
2061      OP_ONCE_NC,        /* 124 Atomic group containing no captures */
2062      OP_BRA,            /* 125 Start of non-capturing bracket */
2063      OP_BRAPOS,         /* 126 Ditto, with unlimited, possessive repeat */
2064      OP_CBRA,           /* 127 Start of capturing bracket */
2065      OP_CBRAPOS,        /* 128 Ditto, with unlimited, possessive repeat */
2066      OP_COND,           /* 129 Conditional group */
2067    
2068    /* 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
2069    check for >= SBRA to distinguish the two sets. */    check for >= SBRA to distinguish the two sets. */
2070    
2071    OP_SBRA,           /* 97 Start of non-capturing bracket, check empty  */    OP_SBRA,           /* 130 Start of non-capturing bracket, check empty  */
2072    OP_SCBRA,          /* 98 Start of capturing bracket, check empty */    OP_SBRAPOS,        /* 131 Ditto, with unlimited, possessive repeat */
2073    OP_SCOND,          /* 99 Conditional group, check empty */    OP_SCBRA,          /* 132 Start of capturing bracket, check empty */
2074      OP_SCBRAPOS,       /* 133 Ditto, with unlimited, possessive repeat */
2075    OP_CREF,           /* 100 Used to hold a capture number as condition */    OP_SCOND,          /* 134 Conditional group, check empty */
2076    OP_RREF,           /* 101 Used to hold a recursion number as condition */  
2077    OP_DEF,            /* 102 The DEFINE condition */    /* The next two pairs must (respectively) be kept together. */
2078    
2079    OP_BRAZERO,        /* 103 These two must remain together and in this */    OP_CREF,           /* 135 Used to hold a capture number as condition */
2080    OP_BRAMINZERO,     /* 104 order. */    OP_NCREF,          /* 136 Same, but generated by a name reference*/
2081      OP_RREF,           /* 137 Used to hold a recursion number as condition */
2082      OP_NRREF,          /* 138 Same, but generated by a name reference*/
2083      OP_DEF,            /* 139 The DEFINE condition */
2084    
2085      OP_BRAZERO,        /* 140 These two must remain together and in this */
2086      OP_BRAMINZERO,     /* 141 order. */
2087      OP_BRAPOSZERO,     /* 142 */
2088    
2089    /* These are backtracking control verbs */    /* These are backtracking control verbs */
2090    
2091    OP_PRUNE,          /* 105 */    OP_MARK,           /* 143 always has an argument */
2092    OP_SKIP,           /* 106 */    OP_PRUNE,          /* 144 */
2093    OP_THEN,           /* 107 */    OP_PRUNE_ARG,      /* 145 same, but with argument */
2094    OP_COMMIT,         /* 108 */    OP_SKIP,           /* 146 */
2095      OP_SKIP_ARG,       /* 147 same, but with argument */
2096      OP_THEN,           /* 148 */
2097      OP_THEN_ARG,       /* 149 same, but with argument */
2098      OP_COMMIT,         /* 150 */
2099    
2100    /* These are forced failure and success verbs */    /* These are forced failure and success verbs */
2101    
2102    OP_FAIL,           /* 109 */    OP_FAIL,           /* 151 */
2103    OP_ACCEPT,         /* 110 */    OP_ACCEPT,         /* 152 */
2104    OP_CLOSE,          /* 111 Used before OP_ACCEPT to close open captures */    OP_ASSERT_ACCEPT,  /* 153 Used inside assertions */
2105      OP_CLOSE,          /* 154 Used before OP_ACCEPT to close open captures */
2106    
2107    /* This is used to skip a subpattern with a {0} quantifier */    /* This is used to skip a subpattern with a {0} quantifier */
2108    
2109    OP_SKIPZERO        /* 112 */    OP_SKIPZERO,       /* 155 */
2110    
2111      /* This is not an opcode, but is used to check that tables indexed by opcode
2112      are the correct length, in order to catch updating errors - there have been
2113      some in the past. */
2114    
2115      OP_TABLE_LENGTH
2116  };  };
2117    
2118    /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2119    definitions that follow must also be updated to match. There are also tables
2120    called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
2121    
2122    
2123  /* 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
2124  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
2125    only in pcre_printint.c, which fills out the full names in many cases (and in
2126    some cases doesn't actually use these names at all). */
2127    
2128  #define OP_NAME_LIST \  #define OP_NAME_LIST \
2129    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
2130    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
2131    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
2132    "extuni",  "\\Z", "\\z",                                        \    "extuni",  "\\Z", "\\z",                                        \
2133    "Opt", "^", "$", "char", "charnc", "not",                       \    "^", "^", "$", "$", "char", "chari", "not", "noti",             \
2134    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2135      "{", "{", "{",                                                  \
2136    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2137    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2138      "{", "{", "{",                                                  \
2139      "*+","++", "?+", "{",                                           \
2140      "*", "*?", "+", "+?", "?", "??",                                \
2141      "{", "{", "{",                                                  \
2142      "*+","++", "?+", "{",                                           \
2143      "*", "*?", "+", "+?", "?", "??",                                \
2144      "{", "{", "{",                                                  \
2145    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2146    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
2147    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2148    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \
2149    "class", "nclass", "xclass", "Ref", "Recurse", "Callout",       \    "class", "nclass", "xclass", "Ref", "Refi",                     \
2150    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",     \    "Recurse", "Callout",                                           \
2151    "AssertB", "AssertB not", "Reverse",                            \    "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos",                  \
2152    "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond",        \    "Reverse", "Assert", "Assert not", "AssertB", "AssertB not",    \
2153    "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero",    \    "Once", "Once_NC",                                              \
2154    "*PRUNE", "*SKIP", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT",      \    "Bra", "BraPos", "CBra", "CBraPos",                             \
2155      "Cond",                                                         \
2156      "SBra", "SBraPos", "SCBra", "SCBraPos",                         \
2157      "SCond",                                                        \
2158      "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \
2159      "Brazero", "Braminzero", "Braposzero",                          \
2160      "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \
2161      "*THEN", "*THEN", "*COMMIT", "*FAIL",                           \
2162      "*ACCEPT", "*ASSERT_ACCEPT",                                    \
2163    "Close", "Skip zero"    "Close", "Skip zero"
2164    
2165    
# Line 1412  in UTF-8 mode. The code that uses this t Line 2177  in UTF-8 mode. The code that uses this t
2177    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \
2178    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
2179    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \
2180    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \    3, 3,                          /* \P, \p                                 */ \
2181    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
2182    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    1,                             /* \X                                     */ \
2183      1, 1, 1, 1, 1, 1,              /* \Z, \z, ^, ^M, $, $M                   */ \
2184    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
2185    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Chari  - the minimum length            */ \
2186    2,                             /* not                                    */ \    2,                             /* not                                    */ \
2187    /* Positive single-char repeats                            ** These are  */ \    2,                             /* noti                                   */ \
2188    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??      ** minima in  */ \    /* Positive single-char repeats                             ** These are */ \
2189    4, 4, 4,                       /* upto, minupto, exact     ** UTF-8 mode */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2190    2, 2, 2, 4,                    /* *+, ++, ?+, upto+                      */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2191      2+IMM2_SIZE,                   /* exact                                  */ \
2192      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2193      2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2194      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2195      2+IMM2_SIZE,                   /* exact I                                */ \
2196      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2197    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2198    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2199    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2200    2, 2, 2, 4,                    /* Possessive *, +, ?, upto               */ \    2+IMM2_SIZE,                   /* NOT exact                              */ \
2201      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2202      2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2203      2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2204      2+IMM2_SIZE,                   /* NOT exact I                            */ \
2205      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2206    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2207    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2208    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2209    2, 2, 2, 4,                    /* Possessive *+, ++, ?+, upto+           */ \    2+IMM2_SIZE,                   /* Type exact                             */ \
2210      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2211    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2212    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2213    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2214   33,                             /* CLASS                                  */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2215   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2216    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2217    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2218      1+IMM2_SIZE,                   /* REFI                                   */ \
2219    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2220    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2221    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
2222    1+LINK_SIZE,                   /* Ket                                    */ \    1+LINK_SIZE,                   /* Ket                                    */ \
2223    1+LINK_SIZE,                   /* KetRmax                                */ \    1+LINK_SIZE,                   /* KetRmax                                */ \
2224    1+LINK_SIZE,                   /* KetRmin                                */ \    1+LINK_SIZE,                   /* KetRmin                                */ \
2225      1+LINK_SIZE,                   /* KetRpos                                */ \
2226      1+LINK_SIZE,                   /* Reverse                                */ \
2227    1+LINK_SIZE,                   /* Assert                                 */ \    1+LINK_SIZE,                   /* Assert                                 */ \
2228    1+LINK_SIZE,                   /* Assert not                             */ \    1+LINK_SIZE,                   /* Assert not                             */ \
2229    1+LINK_SIZE,                   /* Assert behind                          */ \    1+LINK_SIZE,                   /* Assert behind                          */ \
2230    1+LINK_SIZE,                   /* Assert behind not                      */ \    1+LINK_SIZE,                   /* Assert behind not                      */ \
   1+LINK_SIZE,                   /* Reverse                                */ \  
2231    1+LINK_SIZE,                   /* ONCE                                   */ \    1+LINK_SIZE,                   /* ONCE                                   */ \
2232      1+LINK_SIZE,                   /* ONCE_NC                                */ \
2233    1+LINK_SIZE,                   /* BRA                                    */ \    1+LINK_SIZE,                   /* BRA                                    */ \
2234    3+LINK_SIZE,                   /* CBRA                                   */ \    1+LINK_SIZE,                   /* BRAPOS                                 */ \
2235      1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2236      1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2237    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2238    1+LINK_SIZE,                   /* SBRA                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2239    3+LINK_SIZE,                   /* SCBRA                                  */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2240      1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2241      1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2242    1+LINK_SIZE,                   /* SCOND                                  */ \    1+LINK_SIZE,                   /* SCOND                                  */ \
2243    3,                             /* CREF                                   */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* CREF, NCREF                            */ \
2244    3,                             /* RREF                                   */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* RREF, NRREF                            */ \
2245    1,                             /* DEF                                    */ \    1,                             /* DEF                                    */ \
2246    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2247    1, 1, 1, 1,                    /* PRUNE, SKIP, THEN, COMMIT,             */ \    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2248    1, 1, 3, 1                     /* FAIL, ACCEPT, CLOSE, SKIPZERO          */    1, 3,                          /* SKIP, SKIP_ARG                         */ \
2249      1, 3,                          /* THEN, THEN_ARG                         */ \
2250      1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2251      1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2252    
2253  /* 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"
2254    condition. */
2255    
2256  #define RREF_ANY  0xffff  #define RREF_ANY  0xffff
2257    
2258  /* 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
2259  tracked. */  easily be tracked. When a new number is added, the table called eint in
2260    pcreposix.c must be updated. */
2261    
2262  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,
2263         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 2265  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2265         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2266         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2267         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2268         ERR60, ERR61, ERR62, ERR63, ERR64 };         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2269           ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2270    
2271    /* JIT compiling modes. The function list is indexed by them. */
2272    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2273           JIT_NUMBER_OF_COMPILE_MODES };
2274    
2275  /* 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
2276  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 2289  fields are present. Currently PCRE alway
2289  NOTE NOTE NOTE  NOTE NOTE NOTE
2290  */  */
2291    
2292  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2293    #define REAL_PCRE real_pcre
2294    #elif defined COMPILE_PCRE16
2295    #define REAL_PCRE real_pcre16
2296    #elif defined COMPILE_PCRE32
2297    #define REAL_PCRE real_pcre32
2298    #endif
2299    
2300    /* It is necessary to fork the struct for 32 bit, since it needs to use
2301     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2302     * typedef since pcretest needs access to  the struct of the 8-, 16-
2303     * and 32-bit variants. */
2304    
2305    typedef struct real_pcre8_or_16 {
2306    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2307    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2308    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
2309    pcre_uint16 flags;              /* Private flags */    pcre_uint16 flags;              /* Private flags */
2310    pcre_uint16 dummy1;             /* For future use */    pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2311    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2312    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2313    pcre_uint16 first_byte;    pcre_uint16 first_char;         /* Starting character */
2314    pcre_uint16 req_byte;    pcre_uint16 req_char;           /* This character must be seen */
2315    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2316    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2317    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2318    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2319      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2320      const pcre_uint8 *nullpad;      /* NULL padding */
2321    } real_pcre8_or_16;
2322    
2323    typedef struct real_pcre8_or_16 real_pcre;
2324    typedef struct real_pcre8_or_16 real_pcre16;
2325    
2326    const unsigned char *tables;    /* Pointer to tables or NULL for std */  typedef struct real_pcre32 {
2327    const unsigned char *nullpad;   /* NULL padding */    pcre_uint32 magic_number;
2328  } real_pcre;    pcre_uint32 size;               /* Total that was malloced */
2329      pcre_uint32 options;            /* Public options */
2330      pcre_uint16 flags;              /* Private flags */
2331      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2332      pcre_uint16 top_bracket;        /* Highest numbered group */
2333      pcre_uint16 top_backref;        /* Highest numbered back reference */
2334      pcre_uint32 first_char;         /* Starting character */
2335      pcre_uint32 req_char;           /* This character must be seen */
2336      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2337      pcre_uint16 name_entry_size;    /* Size of any name items */
2338      pcre_uint16 name_count;         /* Number of name items */
2339      pcre_uint16 ref_count;          /* Reference count */
2340      pcre_uint16 dummy1;             /* for later expansion */
2341      pcre_uint16 dummy2;             /* for later expansion */
2342      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2343      void *nullpad;                  /* for later expansion */
2344    } real_pcre32;
2345    
2346    /* Assert that the size of REAL_PCRE is divisible by 8 */
2347    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2348    
2349    /* Needed in pcretest to access some fields in the real_pcre* structures
2350     * directly. They're unified for 8/16/32 bits since the structs only differ
2351     * after these fields; if that ever changes, need to fork those defines into
2352     * 8/16 and 32 bit versions. */
2353    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2354    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2355    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2356    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2357    
2358  /* 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
2359  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 2361  remark (see NOTE above) about extending
2361  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2362    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2363    pcre_uint32 flags;              /* Private flags */    pcre_uint32 flags;              /* Private flags */
2364    uschar start_bits[32];          /* Starting char bits */    pcre_uint8 start_bits[32];      /* Starting char bits */
2365    pcre_uint32 minlength;          /* Minimum subject length */    pcre_uint32 minlength;          /* Minimum subject length */
2366  } pcre_study_data;  } pcre_study_data;
2367    
2368  /* Structure for building a chain of open capturing subpatterns during  /* Structure for building a chain of open capturing subpatterns during
2369  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
2370  encountered. */  encountered. This is also used to identify subpatterns that contain recursive
2371    back references to themselves, so that they can be made atomic. */
2372    
2373  typedef struct open_capitem {  typedef struct open_capitem {
2374    struct open_capitem *next;    /* Chain link */    struct open_capitem *next;    /* Chain link */
2375    pcre_uint16 number;           /* Capture number */    pcre_uint16 number;           /* Capture number */
2376  } open_capitem;    pcre_uint16 flag;             /* Set TRUE if recursive back ref */
2377    } open_capitem;
2378    
2379  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2380  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2381    
2382  typedef struct compile_data {  typedef struct compile_data {
2383    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2384    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2385    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2386    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2387    const uschar *start_workspace;/* The start of working space */    const pcre_uchar *start_workspace;/* The start of working space */
2388    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_code;     /* The start of the compiled code */
2389    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2390    const uschar *end_pattern;    /* The end of the pattern */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2391    open_capitem *open_caps;      /* Chain of open capture items */    open_capitem *open_caps;          /* Chain of open capture items */
2392    uschar *hwm;                  /* High watermark of workspace */    pcre_uchar *hwm;                  /* High watermark of workspace */
2393    uschar *name_table;           /* The name/number table */    pcre_uchar *name_table;           /* The name/number table */
2394    int  names_found;             /* Number of entries so far */    int  names_found;                 /* Number of entries so far */
2395    int  name_entry_size;         /* Size of each entry */    int  name_entry_size;             /* Size of each entry */
2396    int  bracount;                /* Count of capturing parens as we compile */    int  workspace_size;              /* Size of workspace */
2397    int  final_bracount;          /* Saved value after first pass */    int  bracount;                    /* Count of capturing parens as we compile */
2398    int  top_backref;             /* Maximum back reference */    int  final_bracount;              /* Saved value after first pass */
2399    unsigned int backref_map;     /* Bitmap of low back refs */    int  max_lookbehind;              /* Maximum lookbehind (characters) */
2400    int  external_options;        /* External (initial) options */    int  top_backref;                 /* Maximum back reference */
2401    int  external_flags;          /* External flag bits to be set */    unsigned int backref_map;         /* Bitmap of low back refs */
2402    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  assert_depth;                /* Depth of nested assertions */
2403    BOOL had_accept;              /* (*ACCEPT) encountered */    int  external_options;            /* External (initial) options */
2404    BOOL check_lookbehind;        /* Lookbehinds need later checking */    int  external_flags;              /* External flag bits to be set */
2405    int  nltype;                  /* Newline type */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2406    int  nllen;                   /* Newline string length */    BOOL had_accept;                  /* (*ACCEPT) encountered */
2407    uschar nl[4];                 /* Newline string when fixed length */    BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2408      BOOL check_lookbehind;            /* Lookbehinds need later checking */
2409      int  nltype;                      /* Newline type */
2410      int  nllen;                       /* Newline string length */
2411      pcre_uchar nl[4];                 /* Newline string when fixed length */
2412  } compile_data;  } compile_data;
2413    
2414  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
2415  branches, for testing for left recursion. */  branches, for testing for left recursion while compiling. */
2416    
2417  typedef struct branch_chain {  typedef struct branch_chain {
2418    struct branch_chain *outer;    struct branch_chain *outer;
2419    uschar *current;    pcre_uchar *current_branch;
2420  } branch_chain;  } branch_chain;
2421    
2422  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
2423  call within the pattern. */  call within the pattern; used by pcre_exec(). */
2424    
2425  typedef struct recursion_info {  typedef struct recursion_info {
2426    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2427    int group_num;                /* Number of group that was called */    int group_num;                  /* Number of group that was called */
2428    const uschar *after_call;     /* "Return value": points after the call in the expr */    int *offset_save;               /* Pointer to start of saved offsets */
2429    USPTR save_start;             /* Old value of mstart */    int saved_max;                  /* Number of saved offsets */
2430    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 */  
2431  } recursion_info;  } recursion_info;
2432    
2433    /* A similar structure for pcre_dfa_exec(). */
2434    
2435    typedef struct dfa_recursion_info {
2436      struct dfa_recursion_info *prevrec;
2437      int group_num;
2438      PCRE_PUCHAR subject_position;
2439    } dfa_recursion_info;
2440    
2441  /* 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
2442  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
2443  has been matched by a subpattern - to break infinite loops. */  has been matched by a subpattern - to break infinite loops; used by
2444    pcre_exec(). */
2445    
2446  typedef struct eptrblock {  typedef struct eptrblock {
2447    struct eptrblock *epb_prev;    struct eptrblock *epb_prev;
2448    USPTR epb_saved_eptr;    PCRE_PUCHAR epb_saved_eptr;
2449  } eptrblock;  } eptrblock;
2450    
2451    
# Line 1602  typedef struct match_data { Line 2456  typedef struct match_data {
2456    unsigned long int match_call_count;      /* As it says */    unsigned long int match_call_count;      /* As it says */
2457    unsigned long int match_limit;           /* As it says */    unsigned long int match_limit;           /* As it says */
2458    unsigned long int match_limit_recursion; /* As it says */    unsigned long int match_limit_recursion; /* As it says */
2459    int   *offset_vector;         /* Offset vector */    int   *offset_vector;           /* Offset vector */
2460    int    offset_end;            /* One past the end */    int    offset_end;              /* One past the end */
2461    int    offset_max;            /* The maximum usable for return data */    int    offset_max;              /* The maximum usable for return data */
2462    int    nltype;                /* Newline type */    int    nltype;                  /* Newline type */
2463    int    nllen;                 /* Newline string length */    int    nllen;                   /* Newline string length */
2464    uschar nl[4];                 /* Newline string when fixed */    int    name_count;              /* Number of names in name table */
2465    const uschar *lcc;            /* Points to lower casing table */    int    name_entry_size;         /* Size of entry in names table */
2466    const uschar *ctypes;         /* Points to table of type maps */    pcre_uchar *name_table;         /* Table of names */
2467    BOOL   offset_overflow;       /* Set if too many extractions */    pcre_uchar nl[4];               /* Newline string when fixed */
2468    BOOL   notbol;                /* NOTBOL flag */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2469    BOOL   noteol;                /* NOTEOL flag */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2470    BOOL   utf8;                  /* UTF8 flag */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2471    BOOL   jscript_compat;        /* JAVASCRIPT_COMPAT flag */    BOOL   offset_overflow;         /* Set if too many extractions */
2472    BOOL   endonly;               /* Dollar not before final \n */    BOOL   notbol;                  /* NOTBOL flag */
2473    BOOL   notempty;              /* Empty string match not wanted */    BOOL   noteol;                  /* NOTEOL flag */
2474    BOOL   notempty_atstart;      /* Empty string match at start not wanted */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2475    BOOL   hitend;                /* Hit the end of the subject at some point */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2476    BOOL   bsr_anycrlf;           /* \R is just any CRLF, not full Unicode */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2477    const uschar *start_code;     /* For use when recursing */    BOOL   endonly;                 /* Dollar not before final \n */
2478    USPTR  start_subject;         /* Start of the subject string */    BOOL   notempty;                /* Empty string match not wanted */
2479    USPTR  end_subject;           /* End of the subject string */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2480    USPTR  start_match_ptr;       /* Start of matched string */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2481    USPTR  end_match_ptr;         /* Subject position at end match */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2482    USPTR  start_used_ptr;        /* Earliest consulted character */    BOOL   hasthen;                 /* Pattern contains (*THEN) */
2483    int    partial;               /* PARTIAL options */    BOOL   ignore_skip_arg;         /* For re-run when SKIP name not found */
2484    int    end_offset_top;        /* Highwater mark at end of match */    const  pcre_uchar *start_code;  /* For use when recursing */
2485    int    capture_last;          /* Most recent capture number */    PCRE_PUCHAR start_subject;      /* Start of the subject string */
2486    int    start_offset;          /* The start offset value */    PCRE_PUCHAR end_subject;        /* End of the subject string */
2487    eptrblock *eptrchain;         /* Chain of eptrblocks for tail recursions */    PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2488    int    eptrn;                 /* Next free eptrblock */    PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2489    recursion_info *recursive;    /* Linked list of recursion data */    PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2490    void  *callout_data;          /* To pass back to callouts */    int    partial;                 /* PARTIAL options */
2491      int    end_offset_top;          /* Highwater mark at end of match */
2492      int    capture_last;            /* Most recent capture number */
2493      int    start_offset;            /* The start offset value */
2494      int    match_function_type;     /* Set for certain special calls of MATCH() */
2495      eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2496      int    eptrn;                   /* Next free eptrblock */
2497      recursion_info *recursive;      /* Linked list of recursion data */
2498      void  *callout_data;            /* To pass back to callouts */
2499      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2500      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2501      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2502    #ifdef NO_RECURSE
2503      void  *match_frames_base;       /* For remembering malloc'd frames */
2504    #endif
2505  } match_data;  } match_data;
2506    
2507  /* 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
2508  functions. */  functions. */
2509    
2510  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2511    const uschar *start_code;     /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2512    const uschar *start_subject;  /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2513    const uschar *end_subject;    /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2514    const uschar *start_used_ptr; /* Earliest consulted character */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2515    const uschar *tables;         /* Character tables */    const pcre_uint8 *tables;         /* Character tables */
2516    int   start_offset;           /* The start offset value */    int   start_offset;               /* The start offset value */
2517    int   moptions;               /* Match options */    int   moptions;                   /* Match options */
2518    int   poptions;               /* Pattern options */    int   poptions;                   /* Pattern options */
2519    int    nltype;                /* Newline type */    int   nltype;                     /* Newline type */
2520    int    nllen;                 /* Newline string length */    int   nllen;                      /* Newline string length */
2521    uschar nl[4];                 /* Newline string when fixed */    pcre_uchar nl[4];                 /* Newline string when fixed */
2522    void  *callout_data;          /* To pass back to callouts */    void *callout_data;               /* To pass back to callouts */
2523      dfa_recursion_info *recursive;    /* Linked list of recursion data */
2524  } dfa_match_data;  } dfa_match_data;
2525    
2526  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 1687  total length. */ Line 2556  total length. */
2556  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2557  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2558    
2559    /* Internal function and data prefixes. */
2560    
2561    #if defined COMPILE_PCRE8
2562    #ifndef PUBL
2563    #define PUBL(name) pcre_##name
2564    #endif
2565    #ifndef PRIV
2566    #define PRIV(name) _pcre_##name
2567    #endif
2568    #elif defined COMPILE_PCRE16
2569    #ifndef PUBL
2570    #define PUBL(name) pcre16_##name
2571    #endif
2572    #ifndef PRIV
2573    #define PRIV(name) _pcre16_##name
2574    #endif
2575    #elif defined COMPILE_PCRE32
2576    #ifndef PUBL
2577    #define PUBL(name) pcre32_##name
2578    #endif
2579    #ifndef PRIV
2580    #define PRIV(name) _pcre32_##name
2581    #endif
2582    #else
2583    #error Unsupported compiling mode
2584    #endif /* COMPILE_PCRE[8|16|32] */
2585    
2586  /* 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
2587  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
2588  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 2600  of the exported public functions. They h
2600  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
2601  pcre_tables.c module. */  pcre_tables.c module. */
2602    
2603  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2604  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2605  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2606  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2607    extern const int            PRIV(utf8_table3)[];
2608  extern const int    _pcre_utf8_table1_size;  extern const pcre_uint8     PRIV(utf8_table4)[];
2609    #endif /* COMPILE_PCRE8 */
2610    
2611    extern const char           PRIV(utt_names)[];
2612    extern const ucp_type_table PRIV(utt)[];
2613    extern const int            PRIV(utt_size);
2614    
2615  extern const char   _pcre_utt_names[];  extern const pcre_uint8     PRIV(OP_lengths)[];
2616  extern const ucp_type_table _pcre_utt[];  extern const pcre_uint8     PRIV(default_tables)[];
 extern const int _pcre_utt_size;  
2617    
2618  extern const uschar _pcre_default_tables[];  extern const pcre_uint32    PRIV(hspace_list)[];
2619    extern const pcre_uint32    PRIV(vspace_list)[];
 extern const uschar _pcre_OP_lengths[];  
2620    
2621    
2622  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2623  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
2624  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2625    
2626  extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);  /* String comparison functions. */
2627  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 *);  
2628    
2629    #define STRCMP_UC_UC(str1, str2) \
2630      strcmp((char *)(str1), (char *)(str2))
2631    #define STRCMP_UC_C8(str1, str2) \
2632      strcmp((char *)(str1), (str2))
2633    #define STRNCMP_UC_UC(str1, str2, num) \
2634      strncmp((char *)(str1), (char *)(str2), (num))
2635    #define STRNCMP_UC_C8(str1, str2, num) \
2636      strncmp((char *)(str1), (str2), (num))
2637    #define STRLEN_UC(str) strlen((const char *)str)
2638    
2639    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2640    
2641    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2642                               const pcre_uchar *);
2643    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2644                               const char *);
2645    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2646                               const pcre_uchar *, unsigned int num);
2647    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2648                               const char *, unsigned int num);
2649    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2650    
2651    #define STRCMP_UC_UC(str1, str2) \
2652      PRIV(strcmp_uc_uc)((str1), (str2))
2653    #define STRCMP_UC_C8(str1, str2) \
2654      PRIV(strcmp_uc_c8)((str1), (str2))
2655    #define STRNCMP_UC_UC(str1, str2, num) \
2656      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2657    #define STRNCMP_UC_C8(str1, str2, num) \
2658      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2659    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2660    
2661    #endif /* COMPILE_PCRE[8|16|32] */
2662    
2663    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2664    
2665    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2666    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2667    
2668    #elif defined COMPILE_PCRE32
2669    
2670    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2671                               const pcre_uchar *);
2672    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2673                               const char *);
2674    
2675    #define STRCMP_UC_UC_TEST(str1, str2) \
2676      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2677    #define STRCMP_UC_C8_TEST(str1, str2) \
2678      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2679    
2680    #endif /* COMPILE_PCRE[8|16|32] */
2681    
2682    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2683    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2684                               int *, BOOL);
2685    extern unsigned int      PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2686    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2687    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2688                               int *, BOOL);
2689    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2690    
2691    #ifdef SUPPORT_JIT
2692    extern void              PRIV(jit_compile)(const REAL_PCRE *,
2693                               PUBL(extra) *, int);
2694    extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2695                               const pcre_uchar *, int, int, int, int *, int);
2696    extern void              PRIV(jit_free)(void *);
2697    extern int               PRIV(jit_get_size)(void *);
2698    extern const char*       PRIV(jit_get_target)(void);
2699    #endif
2700    
2701  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
2702    
2703  typedef struct {  typedef struct {
2704    uschar script;    pcre_uint8 script;     /* ucp_Arabic, etc. */
2705    uschar chartype;    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2706    pcre_int32 other_case;    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2707      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2708      pcre_int32 other_case; /* offset to other case, or zero if none */
2709  } ucd_record;  } ucd_record;
2710    
2711  extern const ucd_record  _pcre_ucd_records[];  extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2712  extern const uschar      _pcre_ucd_stage1[];  extern const ucd_record  PRIV(ucd_records)[];
2713  extern const pcre_uint16 _pcre_ucd_stage2[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2714  extern const int         _pcre_ucp_gentype[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2715    extern const pcre_uint32 PRIV(ucp_gentype)[];
2716    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2717    #ifdef SUPPORT_JIT
2718    extern const int         PRIV(ucp_typerange)[];
2719    #endif
2720    
2721    #ifdef SUPPORT_UCP
2722  /* UCD access macros */  /* UCD access macros */
2723    
2724  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2725  #define GET_UCD(ch) (_pcre_ucd_records + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2726          _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2727          UCD_BLOCK_SIZE + ch % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2728    
2729  #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2730  #define UCD_SCRIPT(ch)    GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2731  #define UCD_CATEGORY(ch)  _pcre_ucp_gentype[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2732  #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2733    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2734    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2735    
2736    #endif /* SUPPORT_UCP */
2737    
2738  #endif  #endif
2739    

Legend:
Removed from v.455  
changed lines
  Added in v.1141

  ViewVC Help
Powered by ViewVC 1.1.5