/[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 510 by ph10, Sat Mar 27 17:45:29 2010 UTC revision 1379 by ph10, Mon Oct 14 13:54:07 2013 UTC
# Line 7  Line 7 
7  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2010 University of Cambridge             Copyright (c) 1997-2013 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 40  POSSIBILITY OF SUCH DAMAGE. Line 40  POSSIBILITY OF SUCH DAMAGE.
40    
41  /* This header contains definitions that are shared between the different  /* This header contains definitions that are shared between the different
42  modules, but which are not relevant to the exported API. This includes some  modules, but which are not relevant to the exported API. This includes some
43  functions whose names all begin with "_pcre_". */  functions whose names all begin with "_pcre_", "_pcre16_" or "_pcre32_"
44    depending on the PRIV macro. */
45    
46  #ifndef PCRE_INTERNAL_H  #ifndef PCRE_INTERNAL_H
47  #define PCRE_INTERNAL_H  #define PCRE_INTERNAL_H
# Line 51  functions whose names all begin with "_p Line 52  functions whose names all begin with "_p
52  #define PCRE_DEBUG  #define PCRE_DEBUG
53  #endif  #endif
54    
55  /* We do not support both EBCDIC and UTF-8 at the same time. The "configure"  /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
 script prevents both being selected, but not everybody uses "configure". */  
56    
57  #if defined EBCDIC && defined SUPPORT_UTF8  #if !defined COMPILE_PCRE16 && !defined COMPILE_PCRE32
58  #error The use of both EBCDIC and SUPPORT_UTF8 is not supported.  #define COMPILE_PCRE8
59  #endif  #endif
60    
61  /* If SUPPORT_UCP is defined, SUPPORT_UTF8 must also be defined. The  /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
62  "configure" script ensures this, but not everybody uses "configure". */  "configure" script ensures this, but not everybody uses "configure". */
63    
64  #if defined SUPPORT_UCP && !defined SUPPORT_UTF8  #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
65    #define SUPPORT_UTF 1
66    #endif
67    
68    /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
69    reasons with existing code. */
70    
71    #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
72    #define SUPPORT_UTF 1
73    #endif
74    
75    /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
76    Until then we define it if SUPPORT_UTF is defined. */
77    
78    #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
79  #define SUPPORT_UTF8 1  #define SUPPORT_UTF8 1
80  #endif  #endif
81    
82    /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
83    script prevents both being selected, but not everybody uses "configure". */
84    
85    #if defined EBCDIC && defined SUPPORT_UTF
86    #error The use of both EBCDIC and SUPPORT_UTF is not supported.
87    #endif
88    
89  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
90  inline, and there are *still* stupid compilers about that don't like indented  inline, and there are *still* stupid compilers about that don't like indented
91  pre-processor statements, or at least there were when I first wrote this. After  pre-processor statements, or at least there were when I first wrote this. After
# Line 91  setjmp and stdarg are used is when NO_RE Line 112  setjmp and stdarg are used is when NO_RE
112  #include <stdlib.h>  #include <stdlib.h>
113  #include <string.h>  #include <string.h>
114    
115    /* Valgrind (memcheck) support */
116    
117    #ifdef SUPPORT_VALGRIND
118    #include <valgrind/memcheck.h>
119    #endif
120    
121  /* When compiling a DLL for Windows, the exported symbols have to be declared  /* When compiling a DLL for Windows, the exported symbols have to be declared
122  using some MS magic. I found some useful information on this web page:  using some MS magic. I found some useful information on this web page:
123  http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the  http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
# Line 158  set, we ensure here that it has no effec Line 185  set, we ensure here that it has no effec
185  #define PCRE_CALL_CONVENTION  #define PCRE_CALL_CONVENTION
186  #endif  #endif
187    
188  /* We need to have types that specify unsigned 16-bit and 32-bit integers. We  /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
189  cannot determine these outside the compilation (e.g. by running a program as  cannot determine these outside the compilation (e.g. by running a program as
190  part of "configure") because PCRE is often cross-compiled for use on other  part of "configure") because PCRE is often cross-compiled for use on other
191  systems. Instead we make use of the maximum sizes that are available at  systems. Instead we make use of the maximum sizes that are available at
192  preprocessor time in standard C environments. */  preprocessor time in standard C environments. */
193    
194    typedef unsigned char pcre_uint8;
195    
196  #if USHRT_MAX == 65535  #if USHRT_MAX == 65535
197    typedef unsigned short pcre_uint16;  typedef unsigned short pcre_uint16;
198    typedef short pcre_int16;  typedef short pcre_int16;
199    #define PCRE_UINT16_MAX USHRT_MAX
200    #define PCRE_INT16_MAX SHRT_MAX
201  #elif UINT_MAX == 65535  #elif UINT_MAX == 65535
202    typedef unsigned int pcre_uint16;  typedef unsigned int pcre_uint16;
203    typedef int pcre_int16;  typedef int pcre_int16;
204    #define PCRE_UINT16_MAX UINT_MAX
205    #define PCRE_INT16_MAX INT_MAX
206  #else  #else
207    #error Cannot determine a type for 16-bit unsigned integers  #error Cannot determine a type for 16-bit integers
208  #endif  #endif
209    
210  #if UINT_MAX == 4294967295  #if UINT_MAX == 4294967295U
211    typedef unsigned int pcre_uint32;  typedef unsigned int pcre_uint32;
212    typedef int pcre_int32;  typedef int pcre_int32;
213  #elif ULONG_MAX == 4294967295  #define PCRE_UINT32_MAX UINT_MAX
214    typedef unsigned long int pcre_uint32;  #define PCRE_INT32_MAX INT_MAX
215    typedef long int pcre_int32;  #elif ULONG_MAX == 4294967295UL
216    typedef unsigned long int pcre_uint32;
217    typedef long int pcre_int32;
218    #define PCRE_UINT32_MAX ULONG_MAX
219    #define PCRE_INT32_MAX LONG_MAX
220  #else  #else
221    #error Cannot determine a type for 32-bit unsigned integers  #error Cannot determine a type for 32-bit integers
222  #endif  #endif
223    
224  /* When checking for integer overflow in pcre_compile(), we need to handle  /* When checking for integer overflow in pcre_compile(), we need to handle
# Line 192  stdint.h is available, include it; it ma Line 229  stdint.h is available, include it; it ma
229  have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set  have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
230  by "configure". */  by "configure". */
231    
232  #if HAVE_STDINT_H  #if defined HAVE_STDINT_H
233  #include <stdint.h>  #include <stdint.h>
234  #elif HAVE_INTTYPES_H  #elif defined HAVE_INTTYPES_H
235  #include <inttypes.h>  #include <inttypes.h>
236  #endif  #endif
237    
# Line 206  by "configure". */ Line 243  by "configure". */
243    
244  /* All character handling must be done as unsigned characters. Otherwise there  /* All character handling must be done as unsigned characters. Otherwise there
245  are problems with top-bit-set characters and functions such as isspace().  are problems with top-bit-set characters and functions such as isspace().
246  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 *,
247  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
248  to save lots of typing. I tried "uchar", but it causes problems on Digital  called pcre_uchar.
249  Unix, where it is defined in sys/types, so use "uschar" instead. */  
250    The IN_UCHARS macro multiply its argument with the byte size of the current
251    pcre_uchar type. Useful for memcpy and such operations, whose require the
252    byte size of their input/output buffers.
253    
254    The MAX_255 macro checks whether its pcre_uchar input is less than 256.
255    
256    The TABLE_GET macro is designed for accessing elements of tables whose contain
257    exactly 256 items. When the character is able to contain more than 256
258    items, some check is needed before accessing these tables.
259    */
260    
261  typedef unsigned char uschar;  #if defined COMPILE_PCRE8
262    
263    typedef unsigned char pcre_uchar;
264    #define IN_UCHARS(x) (x)
265    #define MAX_255(c) 1
266    #define TABLE_GET(c, table, default) ((table)[c])
267    
268    #elif defined COMPILE_PCRE16
269    
270    #if USHRT_MAX != 65535
271    /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
272    pcre.h(.in) and disable (comment out) this message. */
273    #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
274    #endif
275    
276    typedef pcre_uint16 pcre_uchar;
277    #define UCHAR_SHIFT (1)
278    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
279    #define MAX_255(c) ((c) <= 255u)
280    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
281    
282    #elif defined COMPILE_PCRE32
283    
284    typedef pcre_uint32 pcre_uchar;
285    #define UCHAR_SHIFT (2)
286    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
287    #define MAX_255(c) ((c) <= 255u)
288    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
289    
290    #else
291    #error Unsupported compiling mode
292    #endif /* COMPILE_PCRE[8|16|32] */
293    
294  /* 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
295  characters only go up to 0x7fffffff (though Unicode doesn't go beyond  characters only go up to 0x7fffffff (though Unicode doesn't go beyond
# Line 234  start/end of string field names are. */ Line 312  start/end of string field names are. */
312  #define IS_NEWLINE(p) \  #define IS_NEWLINE(p) \
313    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
314      ((p) < NLBLOCK->PSEND && \      ((p) < NLBLOCK->PSEND && \
315       _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\       PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
316         utf8)) \         &(NLBLOCK->nllen), utf)) \
317      : \      : \
318      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
319       (p)[0] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
320       (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
321      ) \      ) \
322    )    )
323    
# Line 248  start/end of string field names are. */ Line 326  start/end of string field names are. */
326  #define WAS_NEWLINE(p) \  #define WAS_NEWLINE(p) \
327    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
328      ((p) > NLBLOCK->PSSTART && \      ((p) > NLBLOCK->PSSTART && \
329       _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \       PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
330         &(NLBLOCK->nllen), utf8)) \         &(NLBLOCK->nllen), utf)) \
331      : \      : \
332      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
333       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
334       (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
335      ) \      ) \
336    )    )
337    
# Line 267  used for the external interface and appe Line 345  used for the external interface and appe
345  must begin with PCRE_. */  must begin with PCRE_. */
346    
347  #ifdef CUSTOM_SUBJECT_PTR  #ifdef CUSTOM_SUBJECT_PTR
348  #define PCRE_SPTR CUSTOM_SUBJECT_PTR  #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
 #define USPTR CUSTOM_SUBJECT_PTR  
349  #else  #else
350  #define PCRE_SPTR const char *  #define PCRE_PUCHAR const pcre_uchar *
 #define USPTR const unsigned char *  
351  #endif  #endif
352    
   
   
353  /* Include the public PCRE header and the definitions of UCP character property  /* Include the public PCRE header and the definitions of UCP character property
354  values. */  values. */
355    
356  #include "pcre.h"  #include "pcre.h"
357  #include "ucp.h"  #include "ucp.h"
358    
359    #ifdef COMPILE_PCRE32
360    /* Assert that the public PCRE_UCHAR32 is a 32-bit type */
361    typedef int __assert_pcre_uchar32_size[sizeof(PCRE_UCHAR32) == 4 ? 1 : -1];
362    #endif
363    
364  /* When compiling for use with the Virtual Pascal compiler, these functions  /* When compiling for use with the Virtual Pascal compiler, these functions
365  need to have their names changed. PCRE must be compiled with the -DVPCOMPAT  need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
366  option on the command line. */  option on the command line. */
# Line 343  The macros are controlled by the value o Line 422  The macros are controlled by the value o
422  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
423  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
424    
425    #if defined COMPILE_PCRE8
426    
427  #if LINK_SIZE == 2  #if LINK_SIZE == 2
428    
429  #define PUT(a,n,d)   \  #define PUT(a,n,d)   \
# Line 379  is automated on Unix systems via the "co Line 460  is automated on Unix systems via the "co
460  #define GET(a,n) \  #define GET(a,n) \
461    (((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])
462    
463  #define MAX_PATTERN_SIZE (1 << 30)   /* Keep it positive */  /* Keep it positive */
464    #define MAX_PATTERN_SIZE (1 << 30)
465    
466    #else
467    #error LINK_SIZE must be either 2, 3, or 4
468    #endif
469    
470    #elif defined COMPILE_PCRE16
471    
472    #if LINK_SIZE == 2
473    
474    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
475    #undef LINK_SIZE
476    #define LINK_SIZE 1
477    
478    #define PUT(a,n,d)   \
479      (a[n] = (d))
480    
481    #define GET(a,n) \
482      (a[n])
483    
484    #define MAX_PATTERN_SIZE (1 << 16)
485    
486    #elif LINK_SIZE == 3 || LINK_SIZE == 4
487    
488    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
489    #undef LINK_SIZE
490    #define LINK_SIZE 2
491    
492    #define PUT(a,n,d)   \
493      (a[n] = (d) >> 16), \
494      (a[(n)+1] = (d) & 65535)
495    
496    #define GET(a,n) \
497      (((a)[n] << 16) | (a)[(n)+1])
498    
499    /* Keep it positive */
500    #define MAX_PATTERN_SIZE (1 << 30)
501    
502  #else  #else
503  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
504  #endif  #endif
505    
506    #elif defined COMPILE_PCRE32
507    
508    /* Only supported LINK_SIZE is 4 */
509    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
510    #undef LINK_SIZE
511    #define LINK_SIZE 1
512    
513    #define PUT(a,n,d)   \
514      (a[n] = (d))
515    
516    #define GET(a,n) \
517      (a[n])
518    
519    /* Keep it positive */
520    #define MAX_PATTERN_SIZE (1 << 30)
521    
522    #else
523    #error Unsupported compiling mode
524    #endif /* COMPILE_PCRE[8|16|32] */
525    
526  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
527    
# Line 396  is automated on Unix systems via the "co Line 532  is automated on Unix systems via the "co
532  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
533  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
534    
535    #if defined COMPILE_PCRE8
536    
537    #define IMM2_SIZE 2
538    
539  #define PUT2(a,n,d)   \  #define PUT2(a,n,d)   \
540    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
541    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
542    
543    /* For reasons that I do not understand, the expression in this GET2 macro is
544    treated by gcc as a signed expression, even when a is declared as unsigned. It
545    seems that any kind of arithmetic results in a signed value. */
546    
547  #define GET2(a,n) \  #define GET2(a,n) \
548    (((a)[n] << 8) | (a)[(n)+1])    (unsigned int)(((a)[n] << 8) | (a)[(n)+1])
549    
550  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += 2  #elif defined COMPILE_PCRE16
551    
552    #define IMM2_SIZE 1
553    
554  /* When UTF-8 encoding is being used, a character is no longer just a single  #define PUT2(a,n,d)   \
555  byte. The macros for character handling generate simple sequences when used in     a[n] = d
556  byte-mode, and more complicated ones for UTF-8 characters. BACKCHAR should  
557  never be called in byte mode. To make sure it can never even appear when UTF-8  #define GET2(a,n) \
558  support is omitted, we don't even define it. */     a[n]
559    
560    #elif defined COMPILE_PCRE32
561    
562    #define IMM2_SIZE 1
563    
564  #ifndef SUPPORT_UTF8  #define PUT2(a,n,d)   \
565       a[n] = d
566    
567    #define GET2(a,n) \
568       a[n]
569    
570    #else
571    #error Unsupported compiling mode
572    #endif /* COMPILE_PCRE[8|16|32] */
573    
574    #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
575    
576    /* The maximum length of a MARK name is currently one data unit; it may be
577    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
578    
579    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
580    #define MAX_MARK ((1u << 16) - 1)
581    #else
582    #define MAX_MARK ((1u << 8) - 1)
583    #endif
584    
585    /* When UTF encoding is being used, a character is no longer just a single
586    byte. The macros for character handling generate simple sequences when used in
587    character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
588    and other macros are not used when UTF is not supported, so they are not
589    defined. To make sure they can never even appear when UTF support is omitted,
590    we don't even define them. */
591    
592    #ifndef SUPPORT_UTF
593    
594    /* #define MAX_VALUE_FOR_SINGLE_CHAR */
595    /* #define HAS_EXTRALEN(c) */
596    /* #define GET_EXTRALEN(c) */
597    /* #define NOT_FIRSTCHAR(c) */
598  #define GETCHAR(c, eptr) c = *eptr;  #define GETCHAR(c, eptr) c = *eptr;
599  #define GETCHARTEST(c, eptr) c = *eptr;  #define GETCHARTEST(c, eptr) c = *eptr;
600  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
601  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
602  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
603    #define RAWUCHAR(eptr) (*(eptr))
604    #define RAWUCHARINC(eptr) (*(eptr)++)
605    #define RAWUCHARTEST(eptr) (*(eptr))
606    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
607    /* #define GETCHARLENTEST(c, eptr, len) */
608  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
609    /* #define FORWARDCHAR(eptr) */
610    /* #define ACROSSCHAR(condition, eptr, action) */
611    
612  #else   /* SUPPORT_UTF8 */  #else   /* SUPPORT_UTF */
613    
614  /* 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. */  
615    
616  #define GETCHAR(c, eptr) \  #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
617    c = *eptr; \  
618    if (c >= 0xc0) \  /* Base macro to pick up the remaining bytes of a UTF-8 character, not
619    advancing the pointer. */
620    
621    #define GETUTF8(c, eptr) \
622        { \
623        if ((c & 0x20) == 0) \
624          c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
625        else if ((c & 0x10) == 0) \
626          c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
627        else if ((c & 0x08) == 0) \
628          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
629          ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
630        else if ((c & 0x04) == 0) \
631          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
632              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
633              (eptr[4] & 0x3f); \
634        else \
635          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
636              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
637              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
638        }
639    
640    /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
641    the pointer. */
642    
643    #define GETUTF8INC(c, eptr) \
644      { \      { \
645      int gcii; \      if ((c & 0x20) == 0) \
646      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \        c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
647      int gcss = 6*gcaa; \      else if ((c & 0x10) == 0) \
648      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        { \
649      for (gcii = 1; gcii <= gcaa; gcii++) \        c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
650          eptr += 2; \
651          } \
652        else if ((c & 0x08) == 0) \
653          { \
654          c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
655              ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
656          eptr += 3; \
657          } \
658        else if ((c & 0x04) == 0) \
659        { \        { \
660        gcss -= 6; \        c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
661        c |= (eptr[gcii] & 0x3f) << gcss; \            ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
662              (eptr[3] & 0x3f); \
663          eptr += 4; \
664          } \
665        else \
666          { \
667          c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
668              ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
669              ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
670          eptr += 5; \
671        } \        } \
672      }      }
673    
674    #if defined COMPILE_PCRE8
675    
676    /* These macros were originally written in the form of loops that used data
677    from the tables whose names start with PRIV(utf8_table). They were rewritten by
678    a user so as not to use loops, because in some environments this gives a
679    significant performance advantage, and it seems never to do any harm. */
680    
681    /* Tells the biggest code point which can be encoded as a single character. */
682    
683    #define MAX_VALUE_FOR_SINGLE_CHAR 127
684    
685    /* Tests whether the code point needs extra characters to decode. */
686    
687    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
688    
689    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
690    Otherwise it has an undefined behaviour. */
691    
692    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
693    
694    /* Returns TRUE, if the given character is not the first character
695    of a UTF sequence. */
696    
697    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
698    
699    /* Get the next UTF-8 character, not advancing the pointer. This is called when
700    we know we are in UTF-8 mode. */
701    
702    #define GETCHAR(c, eptr) \
703      c = *eptr; \
704      if (c >= 0xc0) GETUTF8(c, eptr);
705    
706  /* 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
707  pointer. */  pointer. */
708    
709  #define GETCHARTEST(c, eptr) \  #define GETCHARTEST(c, eptr) \
710    c = *eptr; \    c = *eptr; \
711    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; \  
       } \  
     }  
712    
713  /* 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
714  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
715    
716  #define GETCHARINC(c, eptr) \  #define GETCHARINC(c, eptr) \
717    c = *eptr++; \    c = *eptr++; \
718    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; \  
       } \  
     }  
719    
720  /* 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.
721    This is called when we don't know if we are in UTF-8 mode. */
722    
723  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
724    c = *eptr++; \    c = *eptr++; \
725    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
726    
727    /* Base macro to pick up the remaining bytes of a UTF-8 character, not
728    advancing the pointer, incrementing the length. */
729    
730    #define GETUTF8LEN(c, eptr, len) \
731      { \      { \
732      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \      if ((c & 0x20) == 0) \
733      int gcss = 6*gcaa; \        { \
734      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
735      while (gcaa-- > 0) \        len++; \
736          } \
737        else if ((c & 0x10)  == 0) \
738        { \        { \
739        gcss -= 6; \        c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
740        c |= (*eptr++ & 0x3f) << gcss; \        len += 2; \
741          } \
742        else if ((c & 0x08)  == 0) \
743          {\
744          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
745              ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
746          len += 3; \
747          } \
748        else if ((c & 0x04)  == 0) \
749          { \
750          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
751              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
752              (eptr[4] & 0x3f); \
753          len += 4; \
754          } \
755        else \
756          {\
757          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
758              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
759              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
760          len += 5; \
761        } \        } \
762      }      }
763    
# Line 496  if there are extra bytes. This is called Line 766  if there are extra bytes. This is called
766    
767  #define GETCHARLEN(c, eptr, len) \  #define GETCHARLEN(c, eptr, len) \
768    c = *eptr; \    c = *eptr; \
769    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; \  
     }  
770    
771  /* 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
772  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
773  know we are in UTF-8 mode. */  do not know if we are in UTF-8 mode. */
774    
775  #define GETCHARLENTEST(c, eptr, len) \  #define GETCHARLENTEST(c, eptr, len) \
776    c = *eptr; \    c = *eptr; \
777    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
778      { \  
779      int gcii; \  /* Returns the next uchar, not advancing the pointer. This is called when
780      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  we know we are in UTF mode. */
781      int gcss = 6*gcaa; \  
782      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  #define RAWUCHAR(eptr) \
783      for (gcii = 1; gcii <= gcaa; gcii++) \    (*(eptr))
784        { \  
785        gcss -= 6; \  /* Returns the next uchar, advancing the pointer. This is called when
786        c |= (eptr[gcii] & 0x3f) << gcss; \  we know we are in UTF mode. */
787        } \  
788      len += gcaa; \  #define RAWUCHARINC(eptr) \
789      }    (*((eptr)++))
790    
791    /* Returns the next uchar, testing for UTF mode, and not advancing the
792    pointer. */
793    
794    #define RAWUCHARTEST(eptr) \
795      (*(eptr))
796    
797    /* Returns the next uchar, testing for UTF mode, advancing the
798    pointer. */
799    
800    #define RAWUCHARINCTEST(eptr) \
801      (*((eptr)++))
802    
803  /* 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
804  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 536  because almost all calls are already wit Line 806  because almost all calls are already wit
806    
807  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
808    
809  #endif  /* Same as above, just in the other direction. */
810    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
811    
812    /* Same as above, but it allows a fully customizable form. */
813    #define ACROSSCHAR(condition, eptr, action) \
814      while((condition) && ((eptr) & 0xc0) == 0x80) action
815    
816    #elif defined COMPILE_PCRE16
817    
818    /* Tells the biggest code point which can be encoded as a single character. */
819    
820    #define MAX_VALUE_FOR_SINGLE_CHAR 65535
821    
822    /* Tests whether the code point needs extra characters to decode. */
823    
824    #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
825    
826    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
827    Otherwise it has an undefined behaviour. */
828    
829    #define GET_EXTRALEN(c) 1
830    
831    /* Returns TRUE, if the given character is not the first character
832    of a UTF sequence. */
833    
834    #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
835    
836    /* Base macro to pick up the low surrogate of a UTF-16 character, not
837    advancing the pointer. */
838    
839    #define GETUTF16(c, eptr) \
840       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
841    
842    /* Get the next UTF-16 character, not advancing the pointer. This is called when
843    we know we are in UTF-16 mode. */
844    
845    #define GETCHAR(c, eptr) \
846      c = *eptr; \
847      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
848    
849    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
850    pointer. */
851    
852    #define GETCHARTEST(c, eptr) \
853      c = *eptr; \
854      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
855    
856    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
857    the pointer. */
858    
859    #define GETUTF16INC(c, eptr) \
860       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
861    
862  /* In case there is no definition of offsetof() provided - though any proper  /* Get the next UTF-16 character, advancing the pointer. This is called when we
863  Standard C system should have one. */  know we are in UTF-16 mode. */
864    
865  #ifndef offsetof  #define GETCHARINC(c, eptr) \
866  #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))    c = *eptr++; \
867      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
868    
869    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
870    This is called when we don't know if we are in UTF-16 mode. */
871    
872    #define GETCHARINCTEST(c, eptr) \
873      c = *eptr++; \
874      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
875    
876    /* Base macro to pick up the low surrogate of a UTF-16 character, not
877    advancing the pointer, incrementing the length. */
878    
879    #define GETUTF16LEN(c, eptr, len) \
880       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
881    
882    /* Get the next UTF-16 character, not advancing the pointer, incrementing
883    length if there is a low surrogate. This is called when we know we are in
884    UTF-16 mode. */
885    
886    #define GETCHARLEN(c, eptr, len) \
887      c = *eptr; \
888      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
889    
890    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
891    pointer, incrementing length if there is a low surrogate. This is called when
892    we do not know if we are in UTF-16 mode. */
893    
894    #define GETCHARLENTEST(c, eptr, len) \
895      c = *eptr; \
896      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
897    
898    /* Returns the next uchar, not advancing the pointer. This is called when
899    we know we are in UTF mode. */
900    
901    #define RAWUCHAR(eptr) \
902      (*(eptr))
903    
904    /* Returns the next uchar, advancing the pointer. This is called when
905    we know we are in UTF mode. */
906    
907    #define RAWUCHARINC(eptr) \
908      (*((eptr)++))
909    
910    /* Returns the next uchar, testing for UTF mode, and not advancing the
911    pointer. */
912    
913    #define RAWUCHARTEST(eptr) \
914      (*(eptr))
915    
916    /* Returns the next uchar, testing for UTF mode, advancing the
917    pointer. */
918    
919    #define RAWUCHARINCTEST(eptr) \
920      (*((eptr)++))
921    
922    /* If the pointer is not at the start of a character, move it back until
923    it is. This is called only in UTF-16 mode - we don't put a test within the
924    macro because almost all calls are already within a block of UTF-16 only
925    code. */
926    
927    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
928    
929    /* Same as above, just in the other direction. */
930    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
931    
932    /* Same as above, but it allows a fully customizable form. */
933    #define ACROSSCHAR(condition, eptr, action) \
934      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
935    
936    #elif defined COMPILE_PCRE32
937    
938    /* These are trivial for the 32-bit library, since all UTF-32 characters fit
939    into one pcre_uchar unit. */
940    #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
941    #define HAS_EXTRALEN(c) (0)
942    #define GET_EXTRALEN(c) (0)
943    #define NOT_FIRSTCHAR(c) (0)
944    
945    /* Get the next UTF-32 character, not advancing the pointer. This is called when
946    we know we are in UTF-32 mode. */
947    
948    #define GETCHAR(c, eptr) \
949      c = *(eptr);
950    
951    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
952    pointer. */
953    
954    #define GETCHARTEST(c, eptr) \
955      c = *(eptr);
956    
957    /* Get the next UTF-32 character, advancing the pointer. This is called when we
958    know we are in UTF-32 mode. */
959    
960    #define GETCHARINC(c, eptr) \
961      c = *((eptr)++);
962    
963    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
964    This is called when we don't know if we are in UTF-32 mode. */
965    
966    #define GETCHARINCTEST(c, eptr) \
967      c = *((eptr)++);
968    
969    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
970    length (since all UTF-32 is of length 1). This is called when we know we are in
971    UTF-32 mode. */
972    
973    #define GETCHARLEN(c, eptr, len) \
974      GETCHAR(c, eptr)
975    
976    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
977    pointer, not incrementing the length (since all UTF-32 is of length 1).
978    This is called when we do not know if we are in UTF-32 mode. */
979    
980    #define GETCHARLENTEST(c, eptr, len) \
981      GETCHARTEST(c, eptr)
982    
983    /* Returns the next uchar, not advancing the pointer. This is called when
984    we know we are in UTF mode. */
985    
986    #define RAWUCHAR(eptr) \
987      (*(eptr))
988    
989    /* Returns the next uchar, advancing the pointer. This is called when
990    we know we are in UTF mode. */
991    
992    #define RAWUCHARINC(eptr) \
993      (*((eptr)++))
994    
995    /* Returns the next uchar, testing for UTF mode, and not advancing the
996    pointer. */
997    
998    #define RAWUCHARTEST(eptr) \
999      (*(eptr))
1000    
1001    /* Returns the next uchar, testing for UTF mode, advancing the
1002    pointer. */
1003    
1004    #define RAWUCHARINCTEST(eptr) \
1005      (*((eptr)++))
1006    
1007    /* If the pointer is not at the start of a character, move it back until
1008    it is. This is called only in UTF-32 mode - we don't put a test within the
1009    macro because almost all calls are already within a block of UTF-32 only
1010    code.
1011    These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
1012    
1013    #define BACKCHAR(eptr) do { } while (0)
1014    
1015    /* Same as above, just in the other direction. */
1016    #define FORWARDCHAR(eptr) do { } while (0)
1017    
1018    /* Same as above, but it allows a fully customizable form. */
1019    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1020    
1021    #else
1022    #error Unsupported compiling mode
1023    #endif /* COMPILE_PCRE[8|16|32] */
1024    
1025    #endif  /* SUPPORT_UTF */
1026    
1027    /* Tests for Unicode horizontal and vertical whitespace characters must check a
1028    number of different values. Using a switch statement for this generates the
1029    fastest code (no loop, no memory access), and there are several places in the
1030    interpreter code where this happens. In order to ensure that all the case lists
1031    remain in step, we use macros so that there is only one place where the lists
1032    are defined.
1033    
1034    These values are also required as lists in pcre_compile.c when processing \h,
1035    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1036    macros that define the values are here so that all the definitions are
1037    together. The lists must be in ascending character order, terminated by
1038    NOTACHAR (which is 0xffffffff).
1039    
1040    Any changes should ensure that the various macros are kept in step with each
1041    other. NOTE: The values also appear in pcre_jit_compile.c. */
1042    
1043    /* ------ ASCII/Unicode environments ------ */
1044    
1045    #ifndef EBCDIC
1046    
1047    #define HSPACE_LIST \
1048      CHAR_HT, CHAR_SPACE, 0xa0, \
1049      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1050      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1051      NOTACHAR
1052    
1053    #define HSPACE_MULTIBYTE_CASES \
1054      case 0x1680:  /* OGHAM SPACE MARK */ \
1055      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1056      case 0x2000:  /* EN QUAD */ \
1057      case 0x2001:  /* EM QUAD */ \
1058      case 0x2002:  /* EN SPACE */ \
1059      case 0x2003:  /* EM SPACE */ \
1060      case 0x2004:  /* THREE-PER-EM SPACE */ \
1061      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1062      case 0x2006:  /* SIX-PER-EM SPACE */ \
1063      case 0x2007:  /* FIGURE SPACE */ \
1064      case 0x2008:  /* PUNCTUATION SPACE */ \
1065      case 0x2009:  /* THIN SPACE */ \
1066      case 0x200A:  /* HAIR SPACE */ \
1067      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1068      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1069      case 0x3000   /* IDEOGRAPHIC SPACE */
1070    
1071    #define HSPACE_BYTE_CASES \
1072      case CHAR_HT: \
1073      case CHAR_SPACE: \
1074      case 0xa0     /* NBSP */
1075    
1076    #define HSPACE_CASES \
1077      HSPACE_BYTE_CASES: \
1078      HSPACE_MULTIBYTE_CASES
1079    
1080    #define VSPACE_LIST \
1081      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1082    
1083    #define VSPACE_MULTIBYTE_CASES \
1084      case 0x2028:    /* LINE SEPARATOR */ \
1085      case 0x2029     /* PARAGRAPH SEPARATOR */
1086    
1087    #define VSPACE_BYTE_CASES \
1088      case CHAR_LF: \
1089      case CHAR_VT: \
1090      case CHAR_FF: \
1091      case CHAR_CR: \
1092      case CHAR_NEL
1093    
1094    #define VSPACE_CASES \
1095      VSPACE_BYTE_CASES: \
1096      VSPACE_MULTIBYTE_CASES
1097    
1098    /* ------ EBCDIC environments ------ */
1099    
1100    #else
1101    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1102    
1103    #define HSPACE_BYTE_CASES \
1104      case CHAR_HT: \
1105      case CHAR_SPACE
1106    
1107    #define HSPACE_CASES HSPACE_BYTE_CASES
1108    
1109    #ifdef EBCDIC_NL25
1110    #define VSPACE_LIST \
1111      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1112    #else
1113    #define VSPACE_LIST \
1114      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1115  #endif  #endif
1116    
1117    #define VSPACE_BYTE_CASES \
1118      case CHAR_LF: \
1119      case CHAR_VT: \
1120      case CHAR_FF: \
1121      case CHAR_CR: \
1122      case CHAR_NEL
1123    
1124    #define VSPACE_CASES VSPACE_BYTE_CASES
1125    #endif  /* EBCDIC */
1126    
1127    /* ------ End of whitespace macros ------ */
1128    
 /* These are the public options that can change during matching. */  
1129    
 #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)  
1130    
1131  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
1132  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 they were
1133  are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as  moved to a 16-bit flags word - which got almost full, so now they are in a
1134  the restrictions on partial matching have been lifted. It remains for backwards  32-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as the
1135    restrictions on partial matching have been lifted. It remains for backwards
1136  compatibility. */  compatibility. */
1137    
1138  #define PCRE_NOPARTIAL     0x0001  /* can't use partial with this regex */  #define PCRE_MODE8         0x00000001  /* compiled in 8 bit mode */
1139  #define PCRE_FIRSTSET      0x0002  /* first_byte is set */  #define PCRE_MODE16        0x00000002  /* compiled in 16 bit mode */
1140  #define PCRE_REQCHSET      0x0004  /* req_byte is set */  #define PCRE_MODE32        0x00000004  /* compiled in 32 bit mode */
1141  #define PCRE_STARTLINE     0x0008  /* start after \n for multiline */  #define PCRE_FIRSTSET      0x00000010  /* first_char is set */
1142  #define PCRE_JCHANGED      0x0010  /* j option used in regex */  #define PCRE_FCH_CASELESS  0x00000020  /* caseless first char */
1143  #define PCRE_HASCRORLF     0x0020  /* explicit \r or \n in pattern */  #define PCRE_REQCHSET      0x00000040  /* req_byte is set */
1144    #define PCRE_RCH_CASELESS  0x00000080  /* caseless requested char */
1145    #define PCRE_STARTLINE     0x00000100  /* start after \n for multiline */
1146    #define PCRE_NOPARTIAL     0x00000200  /* can't use partial with this regex */
1147    #define PCRE_JCHANGED      0x00000400  /* j option used in regex */
1148    #define PCRE_HASCRORLF     0x00000800  /* explicit \r or \n in pattern */
1149    #define PCRE_HASTHEN       0x00001000  /* pattern contains (*THEN) */
1150    #define PCRE_MLSET         0x00002000  /* match limit set by regex */
1151    #define PCRE_RLSET         0x00004000  /* recursion limit set by regex */
1152    #define PCRE_MATCH_EMPTY   0x00008000  /* pattern can match empty string */
1153    
1154    #if defined COMPILE_PCRE8
1155    #define PCRE_MODE          PCRE_MODE8
1156    #elif defined COMPILE_PCRE16
1157    #define PCRE_MODE          PCRE_MODE16
1158    #elif defined COMPILE_PCRE32
1159    #define PCRE_MODE          PCRE_MODE32
1160    #endif
1161    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1162    
1163  /* Options for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1164    
1165  #define PCRE_STUDY_MAPPED   0x01     /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
1166  #define PCRE_STUDY_MINLEN   0x02     /* a minimum length field exists */  #define PCRE_STUDY_MINLEN  0x0002  /* a minimum length field exists */
1167    
1168  /* Masks for identifying the public options that are permitted at compile  /* Masks for identifying the public options that are permitted at compile
1169  time, run time, or study time, respectively. */  time, run time, or study time, respectively. */
# Line 578  time, run time, or study time, respectiv Line 1174  time, run time, or study time, respectiv
1174  #define PUBLIC_COMPILE_OPTIONS \  #define PUBLIC_COMPILE_OPTIONS \
1175    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
1176     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1177     PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \     PCRE_NO_AUTO_CAPTURE|PCRE_NO_AUTO_POSSESSIFY| \
1178       PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
1179     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1180     PCRE_JAVASCRIPT_COMPAT)     PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE|PCRE_NEVER_UTF)
1181    
1182  #define PUBLIC_EXEC_OPTIONS \  #define PUBLIC_EXEC_OPTIONS \
1183    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
# Line 593  time, run time, or study time, respectiv Line 1190  time, run time, or study time, respectiv
1190     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \     PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1191     PCRE_NO_START_OPTIMIZE)     PCRE_NO_START_OPTIMIZE)
1192    
1193  #define PUBLIC_STUDY_OPTIONS 0   /* None defined */  #define PUBLIC_STUDY_OPTIONS \
1194       (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1195        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1196    
1197    #define PUBLIC_JIT_EXEC_OPTIONS \
1198       (PCRE_NO_UTF8_CHECK|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|\
1199        PCRE_NOTEMPTY_ATSTART|PCRE_PARTIAL_SOFT|PCRE_PARTIAL_HARD)
1200    
1201  /* 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. */  
1202    
1203  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1204    
1205  /* Negative values for the firstchar and reqchar variables */  /* This variable is used to detect a loaded regular expression
1206    in different endianness. */
1207    
1208  #define REQ_UNSET (-2)  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
 #define REQ_NONE  (-1)  
1209    
1210  /* 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
1211  req_byte match. */  req_byte match. */
1212    
1213  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1214    
 /* 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 */  
   
1215  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1216  environments where these macros are defined elsewhere. Unfortunately, there  environments where these macros are defined elsewhere. Unfortunately, there
1217  is no way to do the same for the typedef. */  is no way to do the same for the typedef. */
# Line 644  for) in a minority area (EBCDIC platform Line 1240  for) in a minority area (EBCDIC platform
1240  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
1241  macros to give the functions distinct names. */  macros to give the functions distinct names. */
1242    
1243  #ifndef SUPPORT_UTF8  #ifndef SUPPORT_UTF
1244    
1245  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1246  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
1247    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1248    characters, a common practice has been to use its NL (0x15) character as the
1249    line terminator in C-like processing environments. However, sometimes the LF
1250    (0x25) character is used instead, according to this Unicode document:
1251    
1252    http://unicode.org/standard/reports/tr13/tr13-5.html
1253    
1254    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1255    instead. Whichever is *not* chosen is defined as NEL.
1256    
1257    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1258    same code point. */
1259    
1260    #ifdef EBCDIC
1261    
1262    #ifndef EBCDIC_NL25
1263    #define CHAR_NL                     '\x15'
1264    #define CHAR_NEL                    '\x25'
1265    #define STR_NL                      "\x15"
1266    #define STR_NEL                     "\x25"
1267    #else
1268    #define CHAR_NL                     '\x25'
1269    #define CHAR_NEL                    '\x15'
1270    #define STR_NL                      "\x25"
1271    #define STR_NEL                     "\x15"
1272    #endif
1273    
1274    #define CHAR_LF                     CHAR_NL
1275    #define STR_LF                      STR_NL
1276    
1277    #define CHAR_ESC                    '\047'
1278    #define CHAR_DEL                    '\007'
1279    #define STR_ESC                     "\047"
1280    #define STR_DEL                     "\007"
1281    
1282    #else  /* Not EBCDIC */
1283    
1284    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1285    compatibility. NEL is the Unicode newline character; make sure it is
1286    a positive value. */
1287    
1288    #define CHAR_LF                     '\n'
1289    #define CHAR_NL                     CHAR_LF
1290    #define CHAR_NEL                    ((unsigned char)'\x85')
1291    #define CHAR_ESC                    '\033'
1292    #define CHAR_DEL                    '\177'
1293    
1294    #define STR_LF                      "\n"
1295    #define STR_NL                      STR_LF
1296    #define STR_NEL                     "\x85"
1297    #define STR_ESC                     "\033"
1298    #define STR_DEL                     "\177"
1299    
1300    #endif  /* EBCDIC */
1301    
1302    /* The remaining definitions work in both environments. */
1303    
1304    #define CHAR_NULL                   '\0'
1305  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1306  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1307  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1308  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1309  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1310  #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  
1311    
1312  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1313  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 764  so that PCRE works on both ASCII and EBC Line 1409  so that PCRE works on both ASCII and EBC
1409  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1410  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1411  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1412  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1413  #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  
1414    
1415  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1416  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 897  so that PCRE works on both ASCII and EBC Line 1534  so that PCRE works on both ASCII and EBC
1534    
1535  #define STRING_DEFINE               "DEFINE"  #define STRING_DEFINE               "DEFINE"
1536    
1537  #define STRING_CR_RIGHTPAR          "CR)"  #define STRING_CR_RIGHTPAR             "CR)"
1538  #define STRING_LF_RIGHTPAR          "LF)"  #define STRING_LF_RIGHTPAR             "LF)"
1539  #define STRING_CRLF_RIGHTPAR        "CRLF)"  #define STRING_CRLF_RIGHTPAR           "CRLF)"
1540  #define STRING_ANY_RIGHTPAR         "ANY)"  #define STRING_ANY_RIGHTPAR            "ANY)"
1541  #define STRING_ANYCRLF_RIGHTPAR     "ANYCRLF)"  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1542  #define STRING_BSR_ANYCRLF_RIGHTPAR "BSR_ANYCRLF)"  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1543  #define STRING_BSR_UNICODE_RIGHTPAR "BSR_UNICODE)"  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1544  #define STRING_UTF8_RIGHTPAR        "UTF8)"  #define STRING_UTF8_RIGHTPAR           "UTF8)"
1545    #define STRING_UTF16_RIGHTPAR          "UTF16)"
1546    #define STRING_UTF32_RIGHTPAR          "UTF32)"
1547    #define STRING_UTF_RIGHTPAR            "UTF)"
1548    #define STRING_UCP_RIGHTPAR            "UCP)"
1549    #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1550    #define STRING_LIMIT_MATCH_EQ          "LIMIT_MATCH="
1551    #define STRING_LIMIT_RECURSION_EQ      "LIMIT_RECURSION="
1552    
1553  #else  /* SUPPORT_UTF8 */  #else  /* SUPPORT_UTF */
1554    
1555  /* 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
1556  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 916  only. */ Line 1560  only. */
1560  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1561  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1562  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1563  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1564    #define CHAR_NL                     CHAR_LF
1565    #define CHAR_NEL                    ((unsigned char)'\x85')
1566  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1567  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1568  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
1569  #define CHAR_DEL                    '\177'  #define CHAR_DEL                    '\177'
1570    
1571    #define CHAR_NULL                   '\0'
1572  #define CHAR_SPACE                  '\040'  #define CHAR_SPACE                  '\040'
1573  #define CHAR_EXCLAMATION_MARK       '\041'  #define CHAR_EXCLAMATION_MARK       '\041'
1574  #define CHAR_QUOTATION_MARK         '\042'  #define CHAR_QUOTATION_MARK         '\042'
# Line 1150  only. */ Line 1797  only. */
1797    
1798  #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
1799    
1800  #define STRING_CR_RIGHTPAR          STR_C STR_R STR_RIGHT_PARENTHESIS  #define STRING_CR_RIGHTPAR             STR_C STR_R STR_RIGHT_PARENTHESIS
1801  #define STRING_LF_RIGHTPAR          STR_L STR_F STR_RIGHT_PARENTHESIS  #define STRING_LF_RIGHTPAR             STR_L STR_F STR_RIGHT_PARENTHESIS
1802  #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
1803  #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
1804  #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
1805  #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
1806  #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
1807  #define STRING_UTF8_RIGHTPAR        STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS  #define STRING_UTF8_RIGHTPAR           STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1808    #define STRING_UTF16_RIGHTPAR          STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1809    #define STRING_UTF32_RIGHTPAR          STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1810    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_RIGHT_PARENTHESIS
1811    #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1812    #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
1813    #define STRING_LIMIT_MATCH_EQ          STR_L STR_I STR_M STR_I STR_T STR_UNDERSCORE STR_M STR_A STR_T STR_C STR_H STR_EQUALS_SIGN
1814    #define STRING_LIMIT_RECURSION_EQ      STR_L STR_I STR_M STR_I STR_T STR_UNDERSCORE STR_R STR_E STR_C STR_U STR_R STR_S STR_I STR_O STR_N STR_EQUALS_SIGN
1815    
1816  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
1817    
1818  /* Escape items that are just an encoding of a particular data value. */  /* Escape items that are just an encoding of a particular data value. */
1819    
# Line 1172  only. */ Line 1826  only. */
1826  #endif  #endif
1827    
1828  #ifndef ESC_n  #ifndef ESC_n
1829  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1830  #endif  #endif
1831    
1832  #ifndef ESC_r  #ifndef ESC_r
# Line 1190  only. */ Line 1844  only. */
1844    
1845  #define PT_ANY        0    /* Any property - matches all chars */  #define PT_ANY        0    /* Any property - matches all chars */
1846  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */
1847  #define PT_GC         2    /* General characteristic (e.g. L) */  #define PT_GC         2    /* Specified general characteristic (e.g. L) */
1848  #define PT_PC         3    /* Particular characteristic (e.g. Lu) */  #define PT_PC         3    /* Specified particular characteristic (e.g. Lu) */
1849  #define PT_SC         4    /* Script (e.g. Han) */  #define PT_SC         4    /* Script (e.g. Han) */
1850    #define PT_ALNUM      5    /* Alphanumeric - the union of L and N */
1851    #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1852    #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1853    #define PT_WORD       8    /* Word - L plus N plus underscore */
1854    #define PT_CLIST      9    /* Pseudo-property: match character list */
1855    #define PT_UCNC      10    /* Universal Character nameable character */
1856    #define PT_TABSIZE   11    /* Size of square table for autopossessify tests */
1857    
1858  /* 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
1859  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1860    
1861  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1862  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */
# Line 1208  contain UTF-8 characters with values gre Line 1869  contain UTF-8 characters with values gre
1869    
1870  /* 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
1871  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
1872  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
1873  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
1874  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
1875  (which is used for [^] in JavaScript compatibility mode).  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1876    non-DOTALL mode, "." behaves like \N.
1877    
1878    The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1879    when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1880    They must be contiguous, and remain in order so that the replacements can be
1881    looked up from a table.
1882    
1883  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
1884  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
1885  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
1886  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
1887  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.
1888  */  */
1889    
1890  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,
1891         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,
1892         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,
1893         ESC_REF };         ESC_E, ESC_Q, ESC_g, ESC_k,
1894           ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1895    
1896    /* Opcode table */
1897    
1898    
1899    /****** NOTE NOTE NOTE ******
1900    
1901    Starting from 1 (i.e. after OP_END), the values up to OP_EOD must correspond in
1902    order to the list of escapes immediately above. Furthermore, values up to
1903    OP_DOLLM must not be changed without adjusting the table called autoposstab in
1904    pcre_compile.c
1905    
1906  /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to  Whenever this list is updated, the two macro definitions that follow must also
1907  OP_EOD must correspond in order to the list of escapes immediately above.  be updated to match. There are also tables called "coptable" and "poptable" in
1908    pcre_dfa_exec.c that must be updated.
1909    
1910  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions  ****** NOTE NOTE NOTE ******/
1911  that follow must also be updated to match. There are also tables called  
1912  "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */  
1913    /* The values between FIRST_AUTOTAB_OP and LAST_AUTOTAB_RIGHT_OP, inclusive,
1914    are used in a table for deciding whether a repeated character type can be
1915    auto-possessified. */
1916    
1917    #define FIRST_AUTOTAB_OP       OP_NOT_DIGIT
1918    #define LAST_AUTOTAB_LEFT_OP   OP_EXTUNI
1919    #define LAST_AUTOTAB_RIGHT_OP  OP_DOLLM
1920    
1921  enum {  enum {
1922    OP_END,            /* 0 End of pattern */    OP_END,            /* 0 End of pattern */
# Line 1249  enum { Line 1934  enum {
1934    OP_WHITESPACE,         /*  9 \s */    OP_WHITESPACE,         /*  9 \s */
1935    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1936    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1937    OP_ANY,            /* 12 Match any character (subject to DOTALL) */  
1938    OP_ALLANY,         /* 13 Match any character (not subject to DOTALL) */    OP_ANY,            /* 12 Match any character except newline (\N) */
1939      OP_ALLANY,         /* 13 Match any character */
1940    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 */
1941    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
1942    OP_PROP,           /* 16 \p (Unicode property) */    OP_PROP,           /* 16 \p (Unicode property) */
# Line 1260  enum { Line 1946  enum {
1946    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1947    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1948    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1949    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) */
1950    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1951    
1952    OP_OPT,            /* 25 Set runtime options */    /* Line end assertions */
1953    OP_CIRC,           /* 26 Start of line - varies with multiline switch */  
1954    OP_DOLL,           /* 27 End of line - varies with multiline switch */    OP_DOLL,           /* 25 End of line - not multiline */
1955    OP_CHAR,           /* 28 Match one character, casefully */    OP_DOLLM,          /* 26 End of line - multiline */
1956    OP_CHARNC,         /* 29 Match one character, caselessly */    OP_CIRC,           /* 27 Start of line - not multiline */
1957    OP_NOT,            /* 30 Match one character, not the following one */    OP_CIRCM,          /* 28 Start of line - multiline */
1958    
1959    OP_STAR,           /* 31 The maximizing and minimizing versions of */    /* Single characters; caseful must precede the caseless ones */
1960    OP_MINSTAR,        /* 32 these six opcodes must come in pairs, with */  
1961    OP_PLUS,           /* 33 the minimizing one second. */    OP_CHAR,           /* 29 Match one character, casefully */
1962    OP_MINPLUS,        /* 34 This first set applies to single characters.*/    OP_CHARI,          /* 30 Match one character, caselessly */
1963    OP_QUERY,          /* 35 */    OP_NOT,            /* 31 Match one character, not the given one, casefully */
1964    OP_MINQUERY,       /* 36 */    OP_NOTI,           /* 32 Match one character, not the given one, caselessly */
1965    
1966    OP_UPTO,           /* 37 From 0 to n matches */    /* The following sets of 13 opcodes must always be kept in step because
1967    OP_MINUPTO,        /* 38 */    the offset from the first one is used to generate the others. */
1968    OP_EXACT,          /* 39 Exactly n matches */  
1969      /* Repeated characters; caseful must precede the caseless ones */
1970    OP_POSSTAR,        /* 40 Possessified star */  
1971    OP_POSPLUS,        /* 41 Possessified plus */    OP_STAR,           /* 33 The maximizing and minimizing versions of */
1972    OP_POSQUERY,       /* 42 Posesssified query */    OP_MINSTAR,        /* 34 these six opcodes must come in pairs, with */
1973    OP_POSUPTO,        /* 43 Possessified upto */    OP_PLUS,           /* 35 the minimizing one second. */
1974      OP_MINPLUS,        /* 36 */
1975    OP_NOTSTAR,        /* 44 The maximizing and minimizing versions of */    OP_QUERY,          /* 37 */
1976    OP_NOTMINSTAR,     /* 45 these six opcodes must come in pairs, with */    OP_MINQUERY,       /* 38 */
1977    OP_NOTPLUS,        /* 46 the minimizing one second. They must be in */  
1978    OP_NOTMINPLUS,     /* 47 exactly the same order as those above. */    OP_UPTO,           /* 39 From 0 to n matches of one character, caseful*/
1979    OP_NOTQUERY,       /* 48 This set applies to "not" single characters. */    OP_MINUPTO,        /* 40 */
1980    OP_NOTMINQUERY,    /* 49 */    OP_EXACT,          /* 41 Exactly n matches */
1981    
1982    OP_NOTUPTO,        /* 50 From 0 to n matches */    OP_POSSTAR,        /* 42 Possessified star, caseful */
1983    OP_NOTMINUPTO,     /* 51 */    OP_POSPLUS,        /* 43 Possessified plus, caseful */
1984    OP_NOTEXACT,       /* 52 Exactly n matches */    OP_POSQUERY,       /* 44 Posesssified query, caseful */
1985      OP_POSUPTO,        /* 45 Possessified upto, caseful */
1986    OP_NOTPOSSTAR,     /* 53 Possessified versions */  
1987    OP_NOTPOSPLUS,     /* 54 */    /* Repeated characters; caseless must follow the caseful ones */
1988    OP_NOTPOSQUERY,    /* 55 */  
1989    OP_NOTPOSUPTO,     /* 56 */    OP_STARI,          /* 46 */
1990      OP_MINSTARI,       /* 47 */
1991    OP_TYPESTAR,       /* 57 The maximizing and minimizing versions of */    OP_PLUSI,          /* 48 */
1992    OP_TYPEMINSTAR,    /* 58 these six opcodes must come in pairs, with */    OP_MINPLUSI,       /* 49 */
1993    OP_TYPEPLUS,       /* 59 the minimizing one second. These codes must */    OP_QUERYI,         /* 50 */
1994    OP_TYPEMINPLUS,    /* 60 be in exactly the same order as those above. */    OP_MINQUERYI,      /* 51 */
1995    OP_TYPEQUERY,      /* 61 This set applies to character types such as \d */  
1996    OP_TYPEMINQUERY,   /* 62 */    OP_UPTOI,          /* 52 From 0 to n matches of one character, caseless */
1997      OP_MINUPTOI,       /* 53 */
1998    OP_TYPEUPTO,       /* 63 From 0 to n matches */    OP_EXACTI,         /* 54 */
1999    OP_TYPEMINUPTO,    /* 64 */  
2000    OP_TYPEEXACT,      /* 65 Exactly n matches */    OP_POSSTARI,       /* 55 Possessified star, caseless */
2001      OP_POSPLUSI,       /* 56 Possessified plus, caseless */
2002    OP_TYPEPOSSTAR,    /* 66 Possessified versions */    OP_POSQUERYI,      /* 57 Posesssified query, caseless */
2003    OP_TYPEPOSPLUS,    /* 67 */    OP_POSUPTOI,       /* 58 Possessified upto, caseless */
2004    OP_TYPEPOSQUERY,   /* 68 */  
2005    OP_TYPEPOSUPTO,    /* 69 */    /* The negated ones must follow the non-negated ones, and match them */
2006      /* Negated repeated character, caseful; must precede the caseless ones */
2007    OP_CRSTAR,         /* 70 The maximizing and minimizing versions of */  
2008    OP_CRMINSTAR,      /* 71 all these opcodes must come in pairs, with */    OP_NOTSTAR,        /* 59 The maximizing and minimizing versions of */
2009    OP_CRPLUS,         /* 72 the minimizing one second. These codes must */    OP_NOTMINSTAR,     /* 60 these six opcodes must come in pairs, with */
2010    OP_CRMINPLUS,      /* 73 be in exactly the same order as those above. */    OP_NOTPLUS,        /* 61 the minimizing one second. They must be in */
2011    OP_CRQUERY,        /* 74 These are for character classes and back refs */    OP_NOTMINPLUS,     /* 62 exactly the same order as those above. */
2012    OP_CRMINQUERY,     /* 75 */    OP_NOTQUERY,       /* 63 */
2013    OP_CRRANGE,        /* 76 These are different to the three sets above. */    OP_NOTMINQUERY,    /* 64 */
2014    OP_CRMINRANGE,     /* 77 */  
2015      OP_NOTUPTO,        /* 65 From 0 to n matches, caseful */
2016    OP_CLASS,          /* 78 Match a character class, chars < 256 only */    OP_NOTMINUPTO,     /* 66 */
2017    OP_NCLASS,         /* 79 Same, but the bitmap was created from a negative    OP_NOTEXACT,       /* 67 Exactly n matches */
2018                             class - the difference is relevant only when a UTF-8  
2019                             character > 255 is encountered. */    OP_NOTPOSSTAR,     /* 68 Possessified versions, caseful */
2020      OP_NOTPOSPLUS,     /* 69 */
2021    OP_XCLASS,         /* 80 Extended class for handling UTF-8 chars within the    OP_NOTPOSQUERY,    /* 70 */
2022                             class. This does both positive and negative. */    OP_NOTPOSUPTO,     /* 71 */
2023    
2024    OP_REF,            /* 81 Match a back reference */    /* Negated repeated character, caseless; must follow the caseful ones */
2025    OP_RECURSE,        /* 82 Match a numbered subpattern (possibly recursive) */  
2026    OP_CALLOUT,        /* 83 Call out to external function if provided */    OP_NOTSTARI,       /* 72 */
2027      OP_NOTMINSTARI,    /* 73 */
2028    OP_ALT,            /* 84 Start of alternation */    OP_NOTPLUSI,       /* 74 */
2029    OP_KET,            /* 85 End of group that doesn't have an unbounded repeat */    OP_NOTMINPLUSI,    /* 75 */
2030    OP_KETRMAX,        /* 86 These two must remain together and in this */    OP_NOTQUERYI,      /* 76 */
2031    OP_KETRMIN,        /* 87 order. They are for groups the repeat for ever. */    OP_NOTMINQUERYI,   /* 77 */
2032    
2033    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/    OP_NOTUPTOI,       /* 78 From 0 to n matches, caseless */
2034      OP_NOTMINUPTOI,    /* 79 */
2035    OP_ASSERT,         /* 88 Positive lookahead */    OP_NOTEXACTI,      /* 80 Exactly n matches */
2036    OP_ASSERT_NOT,     /* 89 Negative lookahead */  
2037    OP_ASSERTBACK,     /* 90 Positive lookbehind */    OP_NOTPOSSTARI,    /* 81 Possessified versions, caseless */
2038    OP_ASSERTBACK_NOT, /* 91 Negative lookbehind */    OP_NOTPOSPLUSI,    /* 82 */
2039    OP_REVERSE,        /* 92 Move pointer back - used in lookbehind assertions */    OP_NOTPOSQUERYI,   /* 83 */
2040      OP_NOTPOSUPTOI,    /* 84 */
2041    /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,  
2042    as there's a test for >= ONCE for a subpattern that isn't an assertion. */    /* Character types */
2043    
2044    OP_ONCE,           /* 93 Atomic group */    OP_TYPESTAR,       /* 85 The maximizing and minimizing versions of */
2045    OP_BRA,            /* 94 Start of non-capturing bracket */    OP_TYPEMINSTAR,    /* 86 these six opcodes must come in pairs, with */
2046    OP_CBRA,           /* 95 Start of capturing bracket */    OP_TYPEPLUS,       /* 87 the minimizing one second. These codes must */
2047    OP_COND,           /* 96 Conditional group */    OP_TYPEMINPLUS,    /* 88 be in exactly the same order as those above. */
2048      OP_TYPEQUERY,      /* 89 */
2049      OP_TYPEMINQUERY,   /* 90 */
2050    
2051      OP_TYPEUPTO,       /* 91 From 0 to n matches */
2052      OP_TYPEMINUPTO,    /* 92 */
2053      OP_TYPEEXACT,      /* 93 Exactly n matches */
2054    
2055      OP_TYPEPOSSTAR,    /* 94 Possessified versions */
2056      OP_TYPEPOSPLUS,    /* 95 */
2057      OP_TYPEPOSQUERY,   /* 96 */
2058      OP_TYPEPOSUPTO,    /* 97 */
2059    
2060      /* These are used for character classes and back references; only the
2061      first six are the same as the sets above. */
2062    
2063      OP_CRSTAR,         /* 98 The maximizing and minimizing versions of */
2064      OP_CRMINSTAR,      /* 99 all these opcodes must come in pairs, with */
2065      OP_CRPLUS,         /* 100 the minimizing one second. These codes must */
2066      OP_CRMINPLUS,      /* 101 be in exactly the same order as those above. */
2067      OP_CRQUERY,        /* 102 */
2068      OP_CRMINQUERY,     /* 103 */
2069    
2070      OP_CRRANGE,        /* 104 These are different to the three sets above. */
2071      OP_CRMINRANGE,     /* 105 */
2072    
2073      OP_CRPOSSTAR,      /* 106 Possessified versions */
2074      OP_CRPOSPLUS,      /* 107 */
2075      OP_CRPOSQUERY,     /* 108 */
2076      OP_CRPOSRANGE,     /* 109 */
2077    
2078      /* End of quantifier opcodes */
2079    
2080      OP_CLASS,          /* 110 Match a character class, chars < 256 only */
2081      OP_NCLASS,         /* 111 Same, but the bitmap was created from a negative
2082                                  class - the difference is relevant only when a
2083                                  character > 255 is encountered. */
2084      OP_XCLASS,         /* 112 Extended class for handling > 255 chars within the
2085                                  class. This does both positive and negative. */
2086      OP_REF,            /* 113 Match a back reference, casefully */
2087      OP_REFI,           /* 114 Match a back reference, caselessly */
2088      OP_DNREF,          /* 115 Match a duplicate name backref, casefully */
2089      OP_DNREFI,         /* 116 Match a duplicate name backref, caselessly */
2090      OP_RECURSE,        /* 117 Match a numbered subpattern (possibly recursive) */
2091      OP_CALLOUT,        /* 118 Call out to external function if provided */
2092    
2093      OP_ALT,            /* 119 Start of alternation */
2094      OP_KET,            /* 120 End of group that doesn't have an unbounded repeat */
2095      OP_KETRMAX,        /* 121 These two must remain together and in this */
2096      OP_KETRMIN,        /* 122 order. They are for groups the repeat for ever. */
2097      OP_KETRPOS,        /* 123 Possessive unlimited repeat. */
2098    
2099      /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
2100      asserts must remain in order. */
2101    
2102      OP_REVERSE,        /* 124 Move pointer back - used in lookbehind assertions */
2103      OP_ASSERT,         /* 125 Positive lookahead */
2104      OP_ASSERT_NOT,     /* 126 Negative lookahead */
2105      OP_ASSERTBACK,     /* 127 Positive lookbehind */
2106      OP_ASSERTBACK_NOT, /* 128 Negative lookbehind */
2107    
2108      /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
2109      after the assertions, with ONCE first, as there's a test for >= ONCE for a
2110      subpattern that isn't an assertion. The POS versions must immediately follow
2111      the non-POS versions in each case. */
2112    
2113      OP_ONCE,           /* 129 Atomic group, contains captures */
2114      OP_ONCE_NC,        /* 130 Atomic group containing no captures */
2115      OP_BRA,            /* 131 Start of non-capturing bracket */
2116      OP_BRAPOS,         /* 132 Ditto, with unlimited, possessive repeat */
2117      OP_CBRA,           /* 133 Start of capturing bracket */
2118      OP_CBRAPOS,        /* 134 Ditto, with unlimited, possessive repeat */
2119      OP_COND,           /* 135 Conditional group */
2120    
2121    /* 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
2122    check for >= SBRA to distinguish the two sets. */    check for >= SBRA to distinguish the two sets. */
2123    
2124    OP_SBRA,           /* 97 Start of non-capturing bracket, check empty  */    OP_SBRA,           /* 136 Start of non-capturing bracket, check empty  */
2125    OP_SCBRA,          /* 98 Start of capturing bracket, check empty */    OP_SBRAPOS,        /* 137 Ditto, with unlimited, possessive repeat */
2126    OP_SCOND,          /* 99 Conditional group, check empty */    OP_SCBRA,          /* 138 Start of capturing bracket, check empty */
2127      OP_SCBRAPOS,       /* 139 Ditto, with unlimited, possessive repeat */
2128      OP_SCOND,          /* 140 Conditional group, check empty */
2129    
2130    /* The next two pairs must (respectively) be kept together. */    /* The next two pairs must (respectively) be kept together. */
2131    
2132    OP_CREF,           /* 100 Used to hold a capture number as condition */    OP_CREF,           /* 141 Used to hold a capture number as condition */
2133    OP_NCREF,          /* 101 Same, but generaged by a name reference*/    OP_DNCREF,         /* 142 Used to point to duplicate names as a condition */
2134    OP_RREF,           /* 102 Used to hold a recursion number as condition */    OP_RREF,           /* 143 Used to hold a recursion number as condition */
2135    OP_NRREF,          /* 103 Same, but generaged by a name reference*/    OP_DNRREF,         /* 144 Used to point to duplicate names as a condition */
2136    OP_DEF,            /* 104 The DEFINE condition */    OP_DEF,            /* 145 The DEFINE condition */
2137    
2138    OP_BRAZERO,        /* 105 These two must remain together and in this */    OP_BRAZERO,        /* 146 These two must remain together and in this */
2139    OP_BRAMINZERO,     /* 106 order. */    OP_BRAMINZERO,     /* 147 order. */
2140      OP_BRAPOSZERO,     /* 148 */
2141    
2142    /* These are backtracking control verbs */    /* These are backtracking control verbs */
2143    
2144    OP_MARK,           /* 107 always has an argument */    OP_MARK,           /* 149 always has an argument */
2145    OP_PRUNE,          /* 108 */    OP_PRUNE,          /* 150 */
2146    OP_PRUNE_ARG,      /* 109 same, but with argument */    OP_PRUNE_ARG,      /* 151 same, but with argument */
2147    OP_SKIP,           /* 110 */    OP_SKIP,           /* 152 */
2148    OP_SKIP_ARG,       /* 111 same, but with argument */    OP_SKIP_ARG,       /* 153 same, but with argument */
2149    OP_THEN,           /* 112 */    OP_THEN,           /* 154 */
2150    OP_THEN_ARG,       /* 113 same, but with argument */    OP_THEN_ARG,       /* 155 same, but with argument */
2151    OP_COMMIT,         /* 114 */    OP_COMMIT,         /* 156 */
2152    
2153    /* These are forced failure and success verbs */    /* These are forced failure and success verbs */
2154    
2155    OP_FAIL,           /* 115 */    OP_FAIL,           /* 157 */
2156    OP_ACCEPT,         /* 116 */    OP_ACCEPT,         /* 158 */
2157    OP_CLOSE,          /* 117 Used before OP_ACCEPT to close open captures */    OP_ASSERT_ACCEPT,  /* 159 Used inside assertions */
2158      OP_CLOSE,          /* 160 Used before OP_ACCEPT to close open captures */
2159    
2160    /* This is used to skip a subpattern with a {0} quantifier */    /* This is used to skip a subpattern with a {0} quantifier */
2161    
2162    OP_SKIPZERO,       /* 118 */    OP_SKIPZERO,       /* 161 */
2163    
2164    /* This is not an opcode, but is used to check that tables indexed by opcode    /* This is not an opcode, but is used to check that tables indexed by opcode
2165    are the correct length, in order to catch updating errors - there have been    are the correct length, in order to catch updating errors - there have been
# Line 1412  called "coptable" and "poptable" in pcre Line 2174  called "coptable" and "poptable" in pcre
2174    
2175    
2176  /* 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
2177  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
2178    only in pcre_printint.c, which fills out the full names in many cases (and in
2179    some cases doesn't actually use these names at all). */
2180    
2181  #define OP_NAME_LIST \  #define OP_NAME_LIST \
2182    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
2183    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
2184    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
2185    "extuni",  "\\Z", "\\z",                                        \    "extuni",  "\\Z", "\\z",                                        \
2186    "Opt", "^", "$", "char", "charnc", "not",                       \    "$", "$", "^", "^", "char", "chari", "not", "noti",             \
2187    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2188      "{", "{", "{",                                                  \
2189    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2190    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2191      "{", "{", "{",                                                  \
2192      "*+","++", "?+", "{",                                           \
2193      "*", "*?", "+", "+?", "?", "??",                                \
2194      "{", "{", "{",                                                  \
2195      "*+","++", "?+", "{",                                           \
2196      "*", "*?", "+", "+?", "?", "??",                                \
2197      "{", "{", "{",                                                  \
2198    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2199    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
2200    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2201    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \
2202    "class", "nclass", "xclass", "Ref", "Recurse", "Callout",       \    "*+","++", "?+", "{",                                           \
2203    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",     \    "class", "nclass", "xclass", "Ref", "Refi", "DnRef", "DnRefi",  \
2204    "AssertB", "AssertB not", "Reverse",                            \    "Recurse", "Callout",                                           \
2205    "Once", "Bra", "CBra", "Cond", "SBra", "SCBra", "SCond",        \    "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos",                  \
2206    "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \    "Reverse", "Assert", "Assert not", "AssertB", "AssertB not",    \
2207    "Brazero", "Braminzero",                                        \    "Once", "Once_NC",                                              \
2208      "Bra", "BraPos", "CBra", "CBraPos",                             \
2209      "Cond",                                                         \
2210      "SBra", "SBraPos", "SCBra", "SCBraPos",                         \
2211      "SCond",                                                        \
2212      "Cond ref", "Cond dnref", "Cond rec", "Cond dnrec", "Cond def", \
2213      "Brazero", "Braminzero", "Braposzero",                          \
2214    "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \    "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \
2215    "*THEN", "*THEN", "*COMMIT", "*FAIL", "*ACCEPT",                \    "*THEN", "*THEN", "*COMMIT", "*FAIL",                           \
2216      "*ACCEPT", "*ASSERT_ACCEPT",                                    \
2217    "Close", "Skip zero"    "Close", "Skip zero"
2218    
2219    
# Line 1455  in UTF-8 mode. The code that uses this t Line 2234  in UTF-8 mode. The code that uses this t
2234    3, 3,                          /* \P, \p                                 */ \    3, 3,                          /* \P, \p                                 */ \
2235    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \    1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
2236    1,                             /* \X                                     */ \    1,                             /* \X                                     */ \
2237    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    1, 1, 1, 1, 1, 1,              /* \Z, \z, $, $M ^, ^M                    */ \
2238    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
2239    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Chari  - the minimum length            */ \
2240    2,                             /* not                                    */ \    2,                             /* not                                    */ \
2241    /* Positive single-char repeats                            ** These are  */ \    2,                             /* noti                                   */ \
2242    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??      ** minima in  */ \    /* Positive single-char repeats                             ** These are */ \
2243    4, 4, 4,                       /* upto, minupto, exact     ** UTF-8 mode */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2244    2, 2, 2, 4,                    /* *+, ++, ?+, upto+                      */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2245      2+IMM2_SIZE,                   /* exact                                  */ \
2246      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2247      2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2248      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2249      2+IMM2_SIZE,                   /* exact I                                */ \
2250      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2251    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2252    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2253    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2254    2, 2, 2, 4,                    /* Possessive *, +, ?, upto               */ \    2+IMM2_SIZE,                   /* NOT exact                              */ \
2255      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2256      2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2257      2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2258      2+IMM2_SIZE,                   /* NOT exact I                            */ \
2259      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2260    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2261    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2262    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2263    2, 2, 2, 4,                    /* Possessive *+, ++, ?+, upto+           */ \    2+IMM2_SIZE,                   /* Type exact                             */ \
2264      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2265    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2266    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2267    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2268   33,                             /* CLASS                                  */ \    1, 1, 1, 1+2*IMM2_SIZE,        /* Possessive *+, ++, ?+, CRPOSRANGE      */ \
2269   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2270      1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2271    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2272    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2273      1+IMM2_SIZE,                   /* REFI                                   */ \
2274      1+2*IMM2_SIZE,                 /* DNREF                                  */ \
2275      1+2*IMM2_SIZE,                 /* DNREFI                                 */ \
2276    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2277    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2278    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
2279    1+LINK_SIZE,                   /* Ket                                    */ \    1+LINK_SIZE,                   /* Ket                                    */ \
2280    1+LINK_SIZE,                   /* KetRmax                                */ \    1+LINK_SIZE,                   /* KetRmax                                */ \
2281    1+LINK_SIZE,                   /* KetRmin                                */ \    1+LINK_SIZE,                   /* KetRmin                                */ \
2282      1+LINK_SIZE,                   /* KetRpos                                */ \
2283      1+LINK_SIZE,                   /* Reverse                                */ \
2284    1+LINK_SIZE,                   /* Assert                                 */ \    1+LINK_SIZE,                   /* Assert                                 */ \
2285    1+LINK_SIZE,                   /* Assert not                             */ \    1+LINK_SIZE,                   /* Assert not                             */ \
2286    1+LINK_SIZE,                   /* Assert behind                          */ \    1+LINK_SIZE,                   /* Assert behind                          */ \
2287    1+LINK_SIZE,                   /* Assert behind not                      */ \    1+LINK_SIZE,                   /* Assert behind not                      */ \
   1+LINK_SIZE,                   /* Reverse                                */ \  
2288    1+LINK_SIZE,                   /* ONCE                                   */ \    1+LINK_SIZE,                   /* ONCE                                   */ \
2289      1+LINK_SIZE,                   /* ONCE_NC                                */ \
2290    1+LINK_SIZE,                   /* BRA                                    */ \    1+LINK_SIZE,                   /* BRA                                    */ \
2291    3+LINK_SIZE,                   /* CBRA                                   */ \    1+LINK_SIZE,                   /* BRAPOS                                 */ \
2292      1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2293      1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2294    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2295    1+LINK_SIZE,                   /* SBRA                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2296    3+LINK_SIZE,                   /* SCBRA                                  */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2297      1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2298      1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2299    1+LINK_SIZE,                   /* SCOND                                  */ \    1+LINK_SIZE,                   /* SCOND                                  */ \
2300    3, 3,                          /* CREF, NCREF                            */ \    1+IMM2_SIZE, 1+2*IMM2_SIZE,    /* CREF, DNCREF                           */ \
2301    3, 3,                          /* RREF, NRREF                            */ \    1+IMM2_SIZE, 1+2*IMM2_SIZE,    /* RREF, DNRREF                           */ \
2302    1,                             /* DEF                                    */ \    1,                             /* DEF                                    */ \
2303    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2304    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG,                */ \    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2305    1, 3, 1, 3,                    /* SKIP, SKIP_ARG, THEN, THEN_ARG,        */ \    1, 3,                          /* SKIP, SKIP_ARG                         */ \
2306    1, 1, 1, 3, 1                  /* COMMIT, FAIL, ACCEPT, CLOSE, SKIPZERO  */    1, 3,                          /* THEN, THEN_ARG                         */ \
2307      1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2308      1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2309    
2310    /* 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"  
 condition. */  
2311    
2312  #define RREF_ANY  0xffff  #define RREF_ANY  0xffff
2313    
# Line 1520  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2321  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2321         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2322         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2323         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2324         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERRCOUNT };         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2325           ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERR78, ERR79,
2326           ERR80, ERR81, ERRCOUNT };
2327    
2328    /* JIT compiling modes. The function list is indexed by them. */
2329    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2330           JIT_NUMBER_OF_COMPILE_MODES };
2331    
2332  /* 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
2333  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
2334  offset to the name table so that if a regex is compiled on one host, saved, and  offset to the name table so that if a regex is compiled on one host, saved, and
2335  then run on another where the size of pointers is different, all might still  then run on another where the size of pointers is different, all might still
2336  be well. For the case of compiled-on-4 and run-on-8, we include an extra  be well.
2337  pointer that is always NULL. For future-proofing, a few dummy fields were  
2338  originally included - even though you can never get this planning right - but  The size of the structure must be a multiple of 8 bytes. For the case of
2339  there is only one left now.  compiled-on-4 and run-on-8, we include an extra pointer that is always NULL so
2340    that there are an even number of pointers which therefore are a multiple of 8
2341  NOTE NOTE NOTE:  bytes.
2342  Because people can now save and re-use compiled patterns, any additions to this  
2343  structure should be made at the end, and something earlier (e.g. a new  It is necessary to fork the struct for the 32 bit library, since it needs to
2344  flag in the options or one of the dummy fields) should indicate that the new  use pcre_uint32 for first_char and req_char. We can't put an ifdef inside the
2345  fields are present. Currently PCRE always sets the dummy fields to zero.  typedef because pcretest needs access to the struct of the 8-, 16- and 32-bit
2346  NOTE NOTE NOTE  variants.
2347    
2348    *** WARNING ***
2349    When new fields are added to these structures, remember to adjust the code in
2350    pcre_byte_order.c that is concerned with swapping the byte order of the fields
2351    when a compiled regex is reloaded on a host with different endianness.
2352    *** WARNING ***
2353    There is also similar byte-flipping code in pcretest.c, which is used for
2354    testing the byte-flipping features. It must also be kept in step.
2355    *** WARNING ***
2356  */  */
2357    
2358  typedef struct real_pcre {  typedef struct real_pcre8_or_16 {
2359    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2360    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2361    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
2362    pcre_uint16 flags;              /* Private flags */    pcre_uint32 flags;              /* Private flags */
2363    pcre_uint16 dummy1;             /* For future use */    pcre_uint32 limit_match;        /* Limit set from regex */
2364    pcre_uint16 top_bracket;    pcre_uint32 limit_recursion;    /* Limit set from regex */
2365    pcre_uint16 top_backref;    pcre_uint16 first_char;         /* Starting character */
2366    pcre_uint16 first_byte;    pcre_uint16 req_char;           /* This character must be seen */
2367    pcre_uint16 req_byte;    pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2368      pcre_uint16 top_bracket;        /* Highest numbered group */
2369      pcre_uint16 top_backref;        /* Highest numbered back reference */
2370    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2371    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2372    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2373    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2374      pcre_uint16 dummy1;             /* To ensure size is a multiple of 8 */
2375      pcre_uint16 dummy2;             /* To ensure size is a multiple of 8 */
2376      pcre_uint16 dummy3;             /* To ensure size is a multiple of 8 */
2377      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2378      void             *nullpad;      /* NULL padding */
2379    } real_pcre8_or_16;
2380    
2381    typedef struct real_pcre8_or_16 real_pcre;
2382    typedef struct real_pcre8_or_16 real_pcre16;
2383    
2384    const unsigned char *tables;    /* Pointer to tables or NULL for std */  typedef struct real_pcre32 {
2385    const unsigned char *nullpad;   /* NULL padding */    pcre_uint32 magic_number;
2386  } real_pcre;    pcre_uint32 size;               /* Total that was malloced */
2387      pcre_uint32 options;            /* Public options */
2388      pcre_uint32 flags;              /* Private flags */
2389      pcre_uint32 limit_match;        /* Limit set from regex */
2390      pcre_uint32 limit_recursion;    /* Limit set from regex */
2391      pcre_uint32 first_char;         /* Starting character */
2392      pcre_uint32 req_char;           /* This character must be seen */
2393      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2394      pcre_uint16 top_bracket;        /* Highest numbered group */
2395      pcre_uint16 top_backref;        /* Highest numbered back reference */
2396      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2397      pcre_uint16 name_entry_size;    /* Size of any name items */
2398      pcre_uint16 name_count;         /* Number of name items */
2399      pcre_uint16 ref_count;          /* Reference count */
2400      pcre_uint16 dummy;              /* To ensure size is a multiple of 8 */
2401      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2402      void             *nullpad;      /* NULL padding */
2403    } real_pcre32;
2404    
2405    #if defined COMPILE_PCRE8
2406    #define REAL_PCRE real_pcre
2407    #elif defined COMPILE_PCRE16
2408    #define REAL_PCRE real_pcre16
2409    #elif defined COMPILE_PCRE32
2410    #define REAL_PCRE real_pcre32
2411    #endif
2412    
2413    /* Assert that the size of REAL_PCRE is divisible by 8 */
2414    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2415    
2416    /* Needed in pcretest to access some fields in the real_pcre* structures
2417     * directly. They're unified for 8/16/32 bits since the structs only differ
2418     * after these fields; if that ever changes, need to fork those defines into
2419     * 8/16 and 32 bit versions. */
2420    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2421    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2422    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2423    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2424    
2425  /* 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
2426  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 1564  remark (see NOTE above) about extending Line 2428  remark (see NOTE above) about extending
2428  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2429    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2430    pcre_uint32 flags;              /* Private flags */    pcre_uint32 flags;              /* Private flags */
2431    uschar start_bits[32];          /* Starting char bits */    pcre_uint8 start_bits[32];      /* Starting char bits */
2432    pcre_uint32 minlength;          /* Minimum subject length */    pcre_uint32 minlength;          /* Minimum subject length */
2433  } pcre_study_data;  } pcre_study_data;
2434    
# Line 1579  typedef struct open_capitem { Line 2443  typedef struct open_capitem {
2443    pcre_uint16 flag;             /* Set TRUE if recursive back ref */    pcre_uint16 flag;             /* Set TRUE if recursive back ref */
2444  } open_capitem;  } open_capitem;
2445    
2446    /* Structure for building a list of named groups during the first pass of
2447    compiling. */
2448    
2449    typedef struct named_group {
2450      const pcre_uchar  *name;          /* Points to the name in the pattern */
2451      int                length;        /* Length of the name */
2452      pcre_uint32        number;        /* Group number */
2453    } named_group;
2454    
2455  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2456  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2457    
2458  typedef struct compile_data {  typedef struct compile_data {
2459    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2460    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2461    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2462    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2463    const uschar *start_workspace;/* The start of working space */    const pcre_uchar *start_workspace;/* The start of working space */
2464    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_code;     /* The start of the compiled code */
2465    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2466    const uschar *end_pattern;    /* The end of the pattern */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2467    open_capitem *open_caps;      /* Chain of open capture items */    pcre_uchar *hwm;                  /* High watermark of workspace */
2468    uschar *hwm;                  /* High watermark of workspace */    open_capitem *open_caps;          /* Chain of open capture items */
2469    uschar *name_table;           /* The name/number table */    named_group *named_groups;        /* Points to vector in pre-compile */
2470    int  names_found;             /* Number of entries so far */    pcre_uchar *name_table;           /* The name/number table */
2471    int  name_entry_size;         /* Size of each entry */    int  names_found;                 /* Number of entries so far */
2472    int  bracount;                /* Count of capturing parens as we compile */    int  name_entry_size;             /* Size of each entry */
2473    int  final_bracount;          /* Saved value after first pass */    int  named_group_list_size;       /* Number of entries in the list */
2474    int  top_backref;             /* Maximum back reference */    int  workspace_size;              /* Size of workspace */
2475    unsigned int backref_map;     /* Bitmap of low back refs */    unsigned int bracount;            /* Count of capturing parens as we compile */
2476    int  external_options;        /* External (initial) options */    int  final_bracount;              /* Saved value after first pass */
2477    int  external_flags;          /* External flag bits to be set */    int  max_lookbehind;              /* Maximum lookbehind (characters) */
2478    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  top_backref;                 /* Maximum back reference */
2479    BOOL had_accept;              /* (*ACCEPT) encountered */    unsigned int backref_map;         /* Bitmap of low back refs */
2480    BOOL check_lookbehind;        /* Lookbehinds need later checking */    unsigned int namedrefcount;       /* Number of backreferences by name */
2481    int  nltype;                  /* Newline type */    int  assert_depth;                /* Depth of nested assertions */
2482    int  nllen;                   /* Newline string length */    pcre_uint32 external_options;     /* External (initial) options */
2483    uschar nl[4];                 /* Newline string when fixed length */    pcre_uint32 external_flags;       /* External flag bits to be set */
2484      int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2485      BOOL had_accept;                  /* (*ACCEPT) encountered */
2486      BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2487      BOOL check_lookbehind;            /* Lookbehinds need later checking */
2488      BOOL dupnames;                    /* Duplicate names exist */
2489      int  nltype;                      /* Newline type */
2490      int  nllen;                       /* Newline string length */
2491      pcre_uchar nl[4];                 /* Newline string when fixed length */
2492  } compile_data;  } compile_data;
2493    
2494  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
2495  branches, for testing for left recursion. */  branches, for testing for left recursion while compiling. */
2496    
2497  typedef struct branch_chain {  typedef struct branch_chain {
2498    struct branch_chain *outer;    struct branch_chain *outer;
2499    uschar *current_branch;    pcre_uchar *current_branch;
2500  } branch_chain;  } branch_chain;
2501    
2502  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
2503  call within the pattern. */  call within the pattern; used by pcre_exec(). */
2504    
2505  typedef struct recursion_info {  typedef struct recursion_info {
2506    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2507    int group_num;                /* Number of group that was called */    unsigned int group_num;         /* Number of group that was called */
2508    const uschar *after_call;     /* "Return value": points after the call in the expr */    int *offset_save;               /* Pointer to start of saved offsets */
2509    int *offset_save;             /* Pointer to start of saved offsets */    int saved_max;                  /* Number of saved offsets */
2510    int saved_max;                /* Number of saved offsets */    int saved_capture_last;         /* Last capture number */
2511    int save_offset_top;          /* Current value of offset_top */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
2512  } recursion_info;  } recursion_info;
2513    
2514    /* A similar structure for pcre_dfa_exec(). */
2515    
2516    typedef struct dfa_recursion_info {
2517      struct dfa_recursion_info *prevrec;
2518      int group_num;
2519      PCRE_PUCHAR subject_position;
2520    } dfa_recursion_info;
2521    
2522  /* 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
2523  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
2524  has been matched by a subpattern - to break infinite loops. */  has been matched by a subpattern - to break infinite loops; used by
2525    pcre_exec(). */
2526    
2527  typedef struct eptrblock {  typedef struct eptrblock {
2528    struct eptrblock *epb_prev;    struct eptrblock *epb_prev;
2529    USPTR epb_saved_eptr;    PCRE_PUCHAR epb_saved_eptr;
2530  } eptrblock;  } eptrblock;
2531    
2532    
# Line 1647  typedef struct match_data { Line 2537  typedef struct match_data {
2537    unsigned long int match_call_count;      /* As it says */    unsigned long int match_call_count;      /* As it says */
2538    unsigned long int match_limit;           /* As it says */    unsigned long int match_limit;           /* As it says */
2539    unsigned long int match_limit_recursion; /* As it says */    unsigned long int match_limit_recursion; /* As it says */
2540    int   *offset_vector;         /* Offset vector */    int   *offset_vector;           /* Offset vector */
2541    int    offset_end;            /* One past the end */    int    offset_end;              /* One past the end */
2542    int    offset_max;            /* The maximum usable for return data */    int    offset_max;              /* The maximum usable for return data */
2543    int    nltype;                /* Newline type */    int    nltype;                  /* Newline type */
2544    int    nllen;                 /* Newline string length */    int    nllen;                   /* Newline string length */
2545    int    name_count;            /* Number of names in name table */    int    name_count;              /* Number of names in name table */
2546    int    name_entry_size;       /* Size of entry in names table */    int    name_entry_size;         /* Size of entry in names table */
2547    uschar *name_table;           /* Table of names */    unsigned int skip_arg_count;    /* For counting SKIP_ARGs */
2548    uschar nl[4];                 /* Newline string when fixed */    unsigned int ignore_skip_arg;   /* For re-run when SKIP arg name not found */
2549    const uschar *lcc;            /* Points to lower casing table */    pcre_uchar *name_table;         /* Table of names */
2550    const uschar *ctypes;         /* Points to table of type maps */    pcre_uchar nl[4];               /* Newline string when fixed */
2551    BOOL   offset_overflow;       /* Set if too many extractions */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2552    BOOL   notbol;                /* NOTBOL flag */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2553    BOOL   noteol;                /* NOTEOL flag */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2554    BOOL   utf8;                  /* UTF8 flag */    BOOL   notbol;                  /* NOTBOL flag */
2555    BOOL   jscript_compat;        /* JAVASCRIPT_COMPAT flag */    BOOL   noteol;                  /* NOTEOL flag */
2556    BOOL   endonly;               /* Dollar not before final \n */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2557    BOOL   notempty;              /* Empty string match not wanted */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2558    BOOL   notempty_atstart;      /* Empty string match at start not wanted */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2559    BOOL   hitend;                /* Hit the end of the subject at some point */    BOOL   endonly;                 /* Dollar not before final \n */
2560    BOOL   bsr_anycrlf;           /* \R is just any CRLF, not full Unicode */    BOOL   notempty;                /* Empty string match not wanted */
2561    const uschar *start_code;     /* For use when recursing */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2562    USPTR  start_subject;         /* Start of the subject string */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2563    USPTR  end_subject;           /* End of the subject string */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2564    USPTR  start_match_ptr;       /* Start of matched string */    BOOL   hasthen;                 /* Pattern contains (*THEN) */
2565    USPTR  end_match_ptr;         /* Subject position at end match */    const  pcre_uchar *start_code;  /* For use when recursing */
2566    USPTR  start_used_ptr;        /* Earliest consulted character */    PCRE_PUCHAR start_subject;      /* Start of the subject string */
2567    int    partial;               /* PARTIAL options */    PCRE_PUCHAR end_subject;        /* End of the subject string */
2568    int    end_offset_top;        /* Highwater mark at end of match */    PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2569    int    capture_last;          /* Most recent capture number */    PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2570    int    start_offset;          /* The start offset value */    PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2571    eptrblock *eptrchain;         /* Chain of eptrblocks for tail recursions */    int    partial;                 /* PARTIAL options */
2572    int    eptrn;                 /* Next free eptrblock */    int    end_offset_top;          /* Highwater mark at end of match */
2573    recursion_info *recursive;    /* Linked list of recursion data */    pcre_int32 capture_last;        /* Most recent capture number + overflow flag */
2574    void  *callout_data;          /* To pass back to callouts */    int    start_offset;            /* The start offset value */
2575    const uschar *mark;           /* Mark pointer to pass back */    int    match_function_type;     /* Set for certain special calls of MATCH() */
2576      eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2577      int    eptrn;                   /* Next free eptrblock */
2578      recursion_info *recursive;      /* Linked list of recursion data */
2579      void  *callout_data;            /* To pass back to callouts */
2580      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2581      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2582      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2583    #ifdef NO_RECURSE
2584      void  *match_frames_base;       /* For remembering malloc'd frames */
2585    #endif
2586  } match_data;  } match_data;
2587    
2588  /* 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
2589  functions. */  functions. */
2590    
2591  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2592    const uschar *start_code;     /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2593    const uschar *start_subject;  /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2594    const uschar *end_subject;    /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2595    const uschar *start_used_ptr; /* Earliest consulted character */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2596    const uschar *tables;         /* Character tables */    const pcre_uint8 *tables;         /* Character tables */
2597    int   start_offset;           /* The start offset value */    int   start_offset;               /* The start offset value */
2598    int   moptions;               /* Match options */    int   moptions;                   /* Match options */
2599    int   poptions;               /* Pattern options */    int   poptions;                   /* Pattern options */
2600    int    nltype;                /* Newline type */    int   nltype;                     /* Newline type */
2601    int    nllen;                 /* Newline string length */    int   nllen;                      /* Newline string length */
2602    uschar nl[4];                 /* Newline string when fixed */    pcre_uchar nl[4];                 /* Newline string when fixed */
2603    void  *callout_data;          /* To pass back to callouts */    void *callout_data;               /* To pass back to callouts */
2604      dfa_recursion_info *recursive;    /* Linked list of recursion data */
2605  } dfa_match_data;  } dfa_match_data;
2606    
2607  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 1736  total length. */ Line 2637  total length. */
2637  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2638  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2639    
2640    /* Internal function and data prefixes. */
2641    
2642    #if defined COMPILE_PCRE8
2643    #ifndef PUBL
2644    #define PUBL(name) pcre_##name
2645    #endif
2646    #ifndef PRIV
2647    #define PRIV(name) _pcre_##name
2648    #endif
2649    #elif defined COMPILE_PCRE16
2650    #ifndef PUBL
2651    #define PUBL(name) pcre16_##name
2652    #endif
2653    #ifndef PRIV
2654    #define PRIV(name) _pcre16_##name
2655    #endif
2656    #elif defined COMPILE_PCRE32
2657    #ifndef PUBL
2658    #define PUBL(name) pcre32_##name
2659    #endif
2660    #ifndef PRIV
2661    #define PRIV(name) _pcre32_##name
2662    #endif
2663    #else
2664    #error Unsupported compiling mode
2665    #endif /* COMPILE_PCRE[8|16|32] */
2666    
2667  /* 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
2668  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
2669  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 1753  of the exported public functions. They h Line 2681  of the exported public functions. They h
2681  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
2682  pcre_tables.c module. */  pcre_tables.c module. */
2683    
2684  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2685  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2686  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2687  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2688    extern const int            PRIV(utf8_table3)[];
2689  extern const int    _pcre_utf8_table1_size;  extern const pcre_uint8     PRIV(utf8_table4)[];
2690    #endif /* COMPILE_PCRE8 */
2691    
2692    extern const char           PRIV(utt_names)[];
2693    extern const ucp_type_table PRIV(utt)[];
2694    extern const int            PRIV(utt_size);
2695    
2696  extern const char   _pcre_utt_names[];  extern const pcre_uint8     PRIV(OP_lengths)[];
2697  extern const ucp_type_table _pcre_utt[];  extern const pcre_uint8     PRIV(default_tables)[];
 extern const int _pcre_utt_size;  
2698    
2699  extern const uschar _pcre_default_tables[];  extern const pcre_uint32    PRIV(hspace_list)[];
2700    extern const pcre_uint32    PRIV(vspace_list)[];
 extern const uschar _pcre_OP_lengths[];  
2701    
2702    
2703  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2704  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
2705  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2706    
2707  extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);  /* String comparison functions. */
2708  extern BOOL          _pcre_is_newline(USPTR, int, USPTR, int *, BOOL);  #if defined COMPILE_PCRE8
 extern int           _pcre_ord2utf8(int, uschar *);  
 extern real_pcre    *_pcre_try_flipped(const real_pcre *, real_pcre *,  
                        const pcre_study_data *, pcre_study_data *);  
 extern int           _pcre_valid_utf8(USPTR, int);  
 extern BOOL          _pcre_was_newline(USPTR, int, USPTR, int *, BOOL);  
 extern BOOL          _pcre_xclass(int, const uschar *);  
2709    
2710    #define STRCMP_UC_UC(str1, str2) \
2711      strcmp((char *)(str1), (char *)(str2))
2712    #define STRCMP_UC_C8(str1, str2) \
2713      strcmp((char *)(str1), (str2))
2714    #define STRNCMP_UC_UC(str1, str2, num) \
2715      strncmp((char *)(str1), (char *)(str2), (num))
2716    #define STRNCMP_UC_C8(str1, str2, num) \
2717      strncmp((char *)(str1), (str2), (num))
2718    #define STRLEN_UC(str) strlen((const char *)str)
2719    
2720    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2721    
2722    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2723                               const pcre_uchar *);
2724    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2725                               const char *);
2726    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2727                               const pcre_uchar *, unsigned int num);
2728    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2729                               const char *, unsigned int num);
2730    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2731    
2732    #define STRCMP_UC_UC(str1, str2) \
2733      PRIV(strcmp_uc_uc)((str1), (str2))
2734    #define STRCMP_UC_C8(str1, str2) \
2735      PRIV(strcmp_uc_c8)((str1), (str2))
2736    #define STRNCMP_UC_UC(str1, str2, num) \
2737      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2738    #define STRNCMP_UC_C8(str1, str2, num) \
2739      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2740    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2741    
2742    #endif /* COMPILE_PCRE[8|16|32] */
2743    
2744    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2745    
2746    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2747    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2748    
2749    #elif defined COMPILE_PCRE32
2750    
2751    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2752                               const pcre_uchar *);
2753    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2754                               const char *);
2755    
2756    #define STRCMP_UC_UC_TEST(str1, str2) \
2757      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2758    #define STRCMP_UC_C8_TEST(str1, str2) \
2759      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2760    
2761    #endif /* COMPILE_PCRE[8|16|32] */
2762    
2763    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2764    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2765                               int *, BOOL);
2766    extern unsigned int      PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2767    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2768    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2769                               int *, BOOL);
2770    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2771    
2772    #ifdef SUPPORT_JIT
2773    extern void              PRIV(jit_compile)(const REAL_PCRE *,
2774                               PUBL(extra) *, int);
2775    extern int               PRIV(jit_exec)(const PUBL(extra) *,
2776                               const pcre_uchar *, int, int, int, int *, int);
2777    extern void              PRIV(jit_free)(void *);
2778    extern int               PRIV(jit_get_size)(void *);
2779    extern const char*       PRIV(jit_get_target)(void);
2780    #endif
2781    
2782  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
2783    
2784  typedef struct {  typedef struct {
2785    uschar script;    pcre_uint8 script;     /* ucp_Arabic, etc. */
2786    uschar chartype;    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2787    pcre_int32 other_case;    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2788      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2789      pcre_int32 other_case; /* offset to other case, or zero if none */
2790  } ucd_record;  } ucd_record;
2791    
2792  extern const ucd_record  _pcre_ucd_records[];  extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2793  extern const uschar      _pcre_ucd_stage1[];  extern const ucd_record  PRIV(ucd_records)[];
2794  extern const pcre_uint16 _pcre_ucd_stage2[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2795  extern const int         _pcre_ucp_gentype[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2796    extern const pcre_uint32 PRIV(ucp_gentype)[];
2797    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2798    #ifdef SUPPORT_JIT
2799    extern const int         PRIV(ucp_typerange)[];
2800    #endif
2801    
2802    #ifdef SUPPORT_UCP
2803  /* UCD access macros */  /* UCD access macros */
2804    
2805  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2806  #define GET_UCD(ch) (_pcre_ucd_records + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2807          _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2808          UCD_BLOCK_SIZE + ch % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2809    
2810  #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2811  #define UCD_SCRIPT(ch)    GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2812  #define UCD_CATEGORY(ch)  _pcre_ucp_gentype[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2813  #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2814    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2815    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2816    
2817    #endif /* SUPPORT_UCP */
2818    
2819  #endif  #endif
2820    

Legend:
Removed from v.510  
changed lines
  Added in v.1379

  ViewVC Help
Powered by ViewVC 1.1.5