/[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 87 by nigel, Sat Feb 24 21:41:21 2007 UTC revision 1090 by chpe, Tue Oct 16 15:55:48 2012 UTC
# Line 7  Line 7 
7  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2006 University of Cambridge             Copyright (c) 1997-2012 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 40  POSSIBILITY OF SUCH DAMAGE. Line 40  POSSIBILITY OF SUCH DAMAGE.
40    
41  /* This header contains definitions that are shared between the different  /* This header contains definitions that are shared between the different
42  modules, but which are not relevant to the exported API. This includes some  modules, but which are not relevant to the exported API. This includes some
43  functions whose names all begin with "_pcre_". */  functions whose names all begin with "_pcre_", "_pcre16_" or "_pcre32_"
44    depending on the PRIV macro. */
45    
46  #ifndef PCRE_INTERNAL_H  #ifndef PCRE_INTERNAL_H
47  #define PCRE_INTERNAL_H  #define PCRE_INTERNAL_H
48    
49  /* Define DEBUG to get debugging output on stdout. */  /* Define PCRE_DEBUG to get debugging output on stdout. */
50    
51  #if 0  #if 0
52  #define DEBUG  #define PCRE_DEBUG
53    #endif
54    
55    /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
56    #if !defined COMPILE_PCRE16 && ! defined COMPILE_PCRE32
57    #define COMPILE_PCRE8
58    #endif
59    
60    /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
61    "configure" script ensures this, but not everybody uses "configure". */
62    
63    #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
64    #define SUPPORT_UTF 1
65    #endif
66    
67    /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
68    reasons with existing code. */
69    
70    #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
71    #define SUPPORT_UTF 1
72    #endif
73    
74    /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
75    Until then we define it if SUPPORT_UTF is defined. */
76    
77    #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
78    #define SUPPORT_UTF8 1
79    #endif
80    
81    /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
82    script prevents both being selected, but not everybody uses "configure". */
83    
84    #if defined EBCDIC && defined SUPPORT_UTF
85    #error The use of both EBCDIC and SUPPORT_UTF is not supported.
86  #endif  #endif
87    
88  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
89  inline, and there are *still* stupid compilers about that don't like indented  inline, and there are *still* stupid compilers about that don't like indented
90  pre-processor statements, or at least there were when I first wrote this. After  pre-processor statements, or at least there were when I first wrote this. After
91  all, it had only been about 10 years then... */  all, it had only been about 10 years then...
92    
93  #ifdef DEBUG  It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so
94    be absolutely sure we get our version. */
95    
96    #undef DPRINTF
97    #ifdef PCRE_DEBUG
98  #define DPRINTF(p) printf p  #define DPRINTF(p) printf p
99  #else  #else
100  #define DPRINTF(p) /*nothing*/  #define DPRINTF(p) /* Nothing */
101  #endif  #endif
102    
103    
 /* Get the definitions provided by running "configure" */  
   
 #include "config.h"  
   
104  /* Standard C headers plus the external interface definition. The only time  /* Standard C headers plus the external interface definition. The only time
105  setjmp and stdarg are used is when NO_RECURSE is set. */  setjmp and stdarg are used is when NO_RECURSE is set. */
106    
107  #include <ctype.h>  #include <ctype.h>
108  #include <limits.h>  #include <limits.h>
 #include <setjmp.h>  
 #include <stdarg.h>  
109  #include <stddef.h>  #include <stddef.h>
110  #include <stdio.h>  #include <stdio.h>
111  #include <stdlib.h>  #include <stdlib.h>
112  #include <string.h>  #include <string.h>
113    
114  #ifndef PCRE_SPY  /* When compiling a DLL for Windows, the exported symbols have to be declared
115  #define PCRE_DEFINITION       /* Win32 __declspec(export) trigger for .dll */  using some MS magic. I found some useful information on this web page:
116    http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
117    information there, using __declspec(dllexport) without "extern" we have a
118    definition; with "extern" we have a declaration. The settings here override the
119    setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL,
120    which is all that is needed for applications (they just import the symbols). We
121    use:
122    
123      PCRE_EXP_DECL       for declarations
124      PCRE_EXP_DEFN       for definitions of exported functions
125      PCRE_EXP_DATA_DEFN  for definitions of exported variables
126    
127    The reason for the two DEFN macros is that in non-Windows environments, one
128    does not want to have "extern" before variable definitions because it leads to
129    compiler warnings. So we distinguish between functions and variables. In
130    Windows, the two should always be the same.
131    
132    The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest,
133    which is an application, but needs to import this file in order to "peek" at
134    internals, can #include pcre.h first to get an application's-eye view.
135    
136    In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon,
137    special-purpose environments) might want to stick other stuff in front of
138    exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and
139    PCRE_EXP_DATA_DEFN only if they are not already set. */
140    
141    #ifndef PCRE_EXP_DECL
142    #  ifdef _WIN32
143    #    ifndef PCRE_STATIC
144    #      define PCRE_EXP_DECL       extern __declspec(dllexport)
145    #      define PCRE_EXP_DEFN       __declspec(dllexport)
146    #      define PCRE_EXP_DATA_DEFN  __declspec(dllexport)
147    #    else
148    #      define PCRE_EXP_DECL       extern
149    #      define PCRE_EXP_DEFN
150    #      define PCRE_EXP_DATA_DEFN
151    #    endif
152    #  else
153    #    ifdef __cplusplus
154    #      define PCRE_EXP_DECL       extern "C"
155    #    else
156    #      define PCRE_EXP_DECL       extern
157    #    endif
158    #    ifndef PCRE_EXP_DEFN
159    #      define PCRE_EXP_DEFN       PCRE_EXP_DECL
160    #    endif
161    #    ifndef PCRE_EXP_DATA_DEFN
162    #      define PCRE_EXP_DATA_DEFN
163    #    endif
164    #  endif
165    #endif
166    
167    /* When compiling with the MSVC compiler, it is sometimes necessary to include
168    a "calling convention" before exported function names. (This is secondhand
169    information; I know nothing about MSVC myself). For example, something like
170    
171      void __cdecl function(....)
172    
173    might be needed. In order so make this easy, all the exported functions have
174    PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
175    set, we ensure here that it has no effect. */
176    
177    #ifndef PCRE_CALL_CONVENTION
178    #define PCRE_CALL_CONVENTION
179  #endif  #endif
180    
181  /* We need to have types that specify unsigned 16-bit and 32-bit integers. We  /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
182  cannot determine these outside the compilation (e.g. by running a program as  cannot determine these outside the compilation (e.g. by running a program as
183  part of "configure") because PCRE is often cross-compiled for use on other  part of "configure") because PCRE is often cross-compiled for use on other
184  systems. Instead we make use of the maximum sizes that are available at  systems. Instead we make use of the maximum sizes that are available at
185  preprocessor time in standard C environments. */  preprocessor time in standard C environments. */
186    
187    typedef unsigned char pcre_uint8;
188    
189  #if USHRT_MAX == 65535  #if USHRT_MAX == 65535
190    typedef unsigned short pcre_uint16;    typedef unsigned short pcre_uint16;
191      typedef short pcre_int16;
192  #elif UINT_MAX == 65535  #elif UINT_MAX == 65535
193    typedef unsigned int pcre_uint16;    typedef unsigned int pcre_uint16;
194      typedef int pcre_int16;
195  #else  #else
196    #error Cannot determine a type for 16-bit unsigned integers    #error Cannot determine a type for 16-bit unsigned integers
197  #endif  #endif
198    
199  #if UINT_MAX == 4294967295  #if UINT_MAX == 4294967295
200    typedef unsigned int pcre_uint32;    typedef unsigned int pcre_uint32;
201      typedef int pcre_int32;
202  #elif ULONG_MAX == 4294967295  #elif ULONG_MAX == 4294967295
203    typedef unsigned long int pcre_uint32;    typedef unsigned long int pcre_uint32;
204      typedef long int pcre_int32;
205  #else  #else
206    #error Cannot determine a type for 32-bit unsigned integers    #error Cannot determine a type for 32-bit unsigned integers
207  #endif  #endif
208    
209    /* When checking for integer overflow in pcre_compile(), we need to handle
210    large integers. If a 64-bit integer type is available, we can use that.
211    Otherwise we have to cast to double, which of course requires floating point
212    arithmetic. Handle this by defining a macro for the appropriate type. If
213    stdint.h is available, include it; it may define INT64_MAX. Systems that do not
214    have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
215    by "configure". */
216    
217    #if HAVE_STDINT_H
218    #include <stdint.h>
219    #elif HAVE_INTTYPES_H
220    #include <inttypes.h>
221    #endif
222    
223    #if defined INT64_MAX || defined int64_t
224    #define INT64_OR_DOUBLE int64_t
225    #else
226    #define INT64_OR_DOUBLE double
227    #endif
228    
229  /* All character handling must be done as unsigned characters. Otherwise there  /* All character handling must be done as unsigned characters. Otherwise there
230  are problems with top-bit-set characters and functions such as isspace().  are problems with top-bit-set characters and functions such as isspace().
231  However, we leave the interface to the outside world as char *, because that  However, we leave the interface to the outside world as char * or short *,
232  should make things easier for callers. We define a short type for unsigned char  because that should make things easier for callers. This character type is
233  to save lots of typing. I tried "uchar", but it causes problems on Digital  called pcre_uchar.
234  Unix, where it is defined in sys/types, so use "uschar" instead. */  
235    The IN_UCHARS macro multiply its argument with the byte size of the current
236    pcre_uchar type. Useful for memcpy and such operations, whose require the
237    byte size of their input/output buffers.
238    
239    The MAX_255 macro checks whether its pcre_uchar input is less than 256.
240    
241    The TABLE_GET macro is designed for accessing elements of tables whose contain
242    exactly 256 items. When the character is able to contain more than 256
243    items, some check is needed before accessing these tables.
244    */
245    
246  typedef unsigned char uschar;  #if defined COMPILE_PCRE8
247    
248    typedef unsigned char pcre_uchar;
249    #define IN_UCHARS(x) (x)
250    #define MAX_255(c) 1
251    #define TABLE_GET(c, table, default) ((table)[c])
252    
253    #elif defined COMPILE_PCRE16
254    
255    #if USHRT_MAX != 65535
256    /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
257    pcre.h(.in) and disable (comment out) this message. */
258    #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
259    #endif
260    
261    typedef pcre_uint16 pcre_uchar;
262    #define UCHAR_SHIFT (1)
263    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
264    #define MAX_255(c) ((c) <= 255u)
265    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
266    
267    #elif defined COMPILE_PCRE32
268    
269    typedef pcre_uint32 pcre_uchar;
270    #define UCHAR_SHIFT (2)
271    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
272    #define MAX_255(c) ((c) <= 255u)
273    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
274    
275    /* Assert that pcre_uchar32 is a 32-bit type */
276    typedef int __assert_pcre_uchar32_size[sizeof(pcre_uchar) == 4 ? 1 : -1];
277    
278    #else
279    #error Unsupported compiling mode
280    #endif /* COMPILE_PCRE[8|16|32] */
281    
282    /* This is an unsigned int value that no character can ever have. UTF-8
283    characters only go up to 0x7fffffff (though Unicode doesn't go beyond
284    0x0010ffff). */
285    
286    #define NOTACHAR 0xffffffff
287    
288    /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
289    "any" and "anycrlf" at present). The following macros are used to package up
290    testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
291    modules to indicate in which datablock the parameters exist, and what the
292    start/end of string field names are. */
293    
294    #define NLTYPE_FIXED    0     /* Newline is a fixed length string */
295    #define NLTYPE_ANY      1     /* Newline is any Unicode line ending */
296    #define NLTYPE_ANYCRLF  2     /* Newline is CR, LF, or CRLF */
297    
298    /* This macro checks for a newline at the given position */
299    
300    #define IS_NEWLINE(p) \
301      ((NLBLOCK->nltype != NLTYPE_FIXED)? \
302        ((p) < NLBLOCK->PSEND && \
303         PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
304           &(NLBLOCK->nllen), utf)) \
305        : \
306        ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
307         (p)[0] == NLBLOCK->nl[0] && \
308         (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \
309        ) \
310      )
311    
312    /* This macro checks for a newline immediately preceding the given position */
313    
314    #define WAS_NEWLINE(p) \
315      ((NLBLOCK->nltype != NLTYPE_FIXED)? \
316        ((p) > NLBLOCK->PSSTART && \
317         PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
318           &(NLBLOCK->nllen), utf)) \
319        : \
320        ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
321         (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \
322         (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \
323        ) \
324      )
325    
326  /* When PCRE is compiled as a C++ library, the subject pointer can be replaced  /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
327  with a custom type. This makes it possible, for example, to allow pcre_exec()  with a custom type. This makes it possible, for example, to allow pcre_exec()
# Line 124  used for the external interface and appe Line 333  used for the external interface and appe
333  must begin with PCRE_. */  must begin with PCRE_. */
334    
335  #ifdef CUSTOM_SUBJECT_PTR  #ifdef CUSTOM_SUBJECT_PTR
336  #define PCRE_SPTR CUSTOM_SUBJECT_PTR  #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
 #define USPTR CUSTOM_SUBJECT_PTR  
337  #else  #else
338  #define PCRE_SPTR const char *  #define PCRE_PUCHAR const pcre_uchar *
 #define USPTR const unsigned char *  
339  #endif  #endif
340    
341  /* Include the public PCRE header and the definitions of UCP character property  /* Include the public PCRE header and the definitions of UCP character property
# Line 142  need to have their names changed. PCRE m Line 349  need to have their names changed. PCRE m
349  option on the command line. */  option on the command line. */
350    
351  #ifdef VPCOMPAT  #ifdef VPCOMPAT
352    #define strlen(s)        _strlen(s)
353  #define strncmp(s1,s2,m) _strncmp(s1,s2,m)  #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
354    #define memcmp(s,c,n)    _memcmp(s,c,n)
355  #define memcpy(d,s,n)    _memcpy(d,s,n)  #define memcpy(d,s,n)    _memcpy(d,s,n)
356  #define memmove(d,s,n)   _memmove(d,s,n)  #define memmove(d,s,n)   _memmove(d,s,n)
357  #define memset(s,c,n)    _memset(s,c,n)  #define memset(s,c,n)    _memset(s,c,n)
# Line 151  option on the command line. */ Line 360  option on the command line. */
360  /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),  /* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
361  define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY  define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
362  is set. Otherwise, include an emulating function for those systems that have  is set. Otherwise, include an emulating function for those systems that have
363  neither (there some non-Unix environments where this is the case). This assumes  neither (there some non-Unix environments where this is the case). */
 that all calls to memmove are moving strings upwards in store, which is the  
 case in PCRE. */  
364    
365  #if ! HAVE_MEMMOVE  #ifndef HAVE_MEMMOVE
366  #undef  memmove        /* some systems may have a macro */  #undef  memmove        /* some systems may have a macro */
367  #if HAVE_BCOPY  #ifdef HAVE_BCOPY
368  #define memmove(a, b, c) bcopy(b, a, c)  #define memmove(a, b, c) bcopy(b, a, c)
369  #else  /* HAVE_BCOPY */  #else  /* HAVE_BCOPY */
370  void *  static void *
371  pcre_memmove(unsigned char *dest, const unsigned char *src, size_t n)  pcre_memmove(void *d, const void *s, size_t n)
372  {  {
373  size_t i;  size_t i;
374  dest += n;  unsigned char *dest = (unsigned char *)d;
375  src += n;  const unsigned char *src = (const unsigned char *)s;
376  for (i = 0; i < n; ++i) *(--dest) =  *(--src);  if (dest > src)
377  return dest;    {
378      dest += n;
379      src += n;
380      for (i = 0; i < n; ++i) *(--dest) = *(--src);
381      return (void *)dest;
382      }
383    else
384      {
385      for (i = 0; i < n; ++i) *dest++ = *src++;
386      return (void *)(dest - n);
387      }
388  }  }
389  #define memmove(a, b, c) pcre_memmove(a, b, c)  #define memmove(a, b, c) pcre_memmove(a, b, c)
390  #endif   /* not HAVE_BCOPY */  #endif   /* not HAVE_BCOPY */
# Line 188  The macros are controlled by the value o Line 405  The macros are controlled by the value o
405  the config.h file, but can be overridden by using -D on the command line. This  the config.h file, but can be overridden by using -D on the command line. This
406  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
407    
408    #if defined COMPILE_PCRE8
409    
410  #if LINK_SIZE == 2  #if LINK_SIZE == 2
411    
412  #define PUT(a,n,d)   \  #define PUT(a,n,d)   \
# Line 224  is automated on Unix systems via the "co Line 443  is automated on Unix systems via the "co
443  #define GET(a,n) \  #define GET(a,n) \
444    (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])    (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
445    
446  #define MAX_PATTERN_SIZE (1 << 30)   /* Keep it positive */  /* Keep it positive */
447    #define MAX_PATTERN_SIZE (1 << 30)
448    
449    #else
450    #error LINK_SIZE must be either 2, 3, or 4
451    #endif
452    
453    #elif defined COMPILE_PCRE16
454    
455    #if LINK_SIZE == 2
456    
457    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
458    #undef LINK_SIZE
459    #define LINK_SIZE 1
460    
461    #define PUT(a,n,d)   \
462      (a[n] = (d))
463    
464    #define GET(a,n) \
465      (a[n])
466    
467    #define MAX_PATTERN_SIZE (1 << 16)
468    
469    #elif LINK_SIZE == 3 || LINK_SIZE == 4
470    
471    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
472    #undef LINK_SIZE
473    #define LINK_SIZE 2
474    
475    #define PUT(a,n,d)   \
476      (a[n] = (d) >> 16), \
477      (a[(n)+1] = (d) & 65535)
478    
479    #define GET(a,n) \
480      (((a)[n] << 16) | (a)[(n)+1])
481    
482    /* Keep it positive */
483    #define MAX_PATTERN_SIZE (1 << 30)
484    
485  #else  #else
486  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
487  #endif  #endif
488    
489    #elif defined COMPILE_PCRE32
490    
491    /* Only supported LINK_SIZE is 4 */
492    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
493    #undef LINK_SIZE
494    #define LINK_SIZE 1
495    
496    #define PUT(a,n,d)   \
497      (a[n] = (d))
498    
499    #define GET(a,n) \
500      (a[n])
501    
502    /* Keep it positive */
503    #define MAX_PATTERN_SIZE (1 << 30)
504    
505    #else
506    #error Unsupported compiling mode
507    #endif /* COMPILE_PCRE[8|16|32] */
508    
509  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
510    
# Line 241  is automated on Unix systems via the "co Line 515  is automated on Unix systems via the "co
515  offsets changes. There are used for repeat counts and for other things such as  offsets changes. There are used for repeat counts and for other things such as
516  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
517    
518    #if defined COMPILE_PCRE8
519    
520    #define IMM2_SIZE 2
521    
522  #define PUT2(a,n,d)   \  #define PUT2(a,n,d)   \
523    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
524    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
# Line 248  capturing parenthesis numbers in back re Line 526  capturing parenthesis numbers in back re
526  #define GET2(a,n) \  #define GET2(a,n) \
527    (((a)[n] << 8) | (a)[(n)+1])    (((a)[n] << 8) | (a)[(n)+1])
528    
529  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += 2  #elif defined COMPILE_PCRE16
530    
531    #define IMM2_SIZE 1
532    
533  /* When UTF-8 encoding is being used, a character is no longer just a single  #define PUT2(a,n,d)   \
534  byte. The macros for character handling generate simple sequences when used in     a[n] = d
535  byte-mode, and more complicated ones for UTF-8 characters. */  
536    #define GET2(a,n) \
537       a[n]
538    
539    #elif defined COMPILE_PCRE32
540    
541    #define IMM2_SIZE 1
542    
543    #define PUT2(a,n,d)   \
544       a[n] = d
545    
546    #define GET2(a,n) \
547       a[n]
548    
549    #else
550    #error Unsupported compiling mode
551    #endif /* COMPILE_PCRE[8|16|32] */
552    
553    #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
554    
555    /* The maximum length of a MARK name is currently one data unit; it may be
556    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
557    
558    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
559    #define MAX_MARK ((1u << 16) - 1)
560    #else
561    #define MAX_MARK ((1u << 8) - 1)
562    #endif
563    
564  #ifndef SUPPORT_UTF8  /* When UTF encoding is being used, a character is no longer just a single
565    byte. The macros for character handling generate simple sequences when used in
566    character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
567    and other macros are not used when UTF is not supported, so they are not
568    defined. To make sure they can never even appear when UTF support is omitted,
569    we don't even define them. */
570    
571    #ifndef SUPPORT_UTF
572    
573    /* #define MAX_VALUE_FOR_SINGLE_CHAR */
574    /* #define HAS_EXTRALEN(c) */
575    /* #define GET_EXTRALEN(c) */
576    /* #define NOT_FIRSTCHAR(c) */
577  #define GETCHAR(c, eptr) c = *eptr;  #define GETCHAR(c, eptr) c = *eptr;
578  #define GETCHARTEST(c, eptr) c = *eptr;  #define GETCHARTEST(c, eptr) c = *eptr;
579  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
580  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
581  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
582  #define BACKCHAR(eptr)  /* #define GETCHARLENTEST(c, eptr, len) */
583    /* #define BACKCHAR(eptr) */
584    /* #define FORWARDCHAR(eptr) */
585    /* #define ACROSSCHAR(condition, eptr, action) */
586    
587  #else   /* SUPPORT_UTF8 */  #else   /* SUPPORT_UTF */
588    
589  /* 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. */  
590    
591  #define GETCHAR(c, eptr) \  #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
592    c = *eptr; \  
593    if ((c & 0xc0) == 0xc0) \  /* Base macro to pick up the remaining bytes of a UTF-8 character, not
594    advancing the pointer. */
595    
596    #define GETUTF8(c, eptr) \
597        { \
598        if ((c & 0x20) == 0) \
599          c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
600        else if ((c & 0x10) == 0) \
601          c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
602        else if ((c & 0x08) == 0) \
603          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
604          ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
605        else if ((c & 0x04) == 0) \
606          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
607              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
608              (eptr[4] & 0x3f); \
609        else \
610          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
611              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
612              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
613        }
614    
615    /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
616    the pointer. */
617    
618    #define GETUTF8INC(c, eptr) \
619      { \      { \
620      int gcii; \      if ((c & 0x20) == 0) \
621      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \        c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
622      int gcss = 6*gcaa; \      else if ((c & 0x10) == 0) \
623      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        { \
624      for (gcii = 1; gcii <= gcaa; gcii++) \        c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
625          eptr += 2; \
626          } \
627        else if ((c & 0x08) == 0) \
628          { \
629          c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
630              ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
631          eptr += 3; \
632          } \
633        else if ((c & 0x04) == 0) \
634          { \
635          c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
636              ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
637              (eptr[3] & 0x3f); \
638          eptr += 4; \
639          } \
640        else \
641        { \        { \
642        gcss -= 6; \        c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
643        c |= (eptr[gcii] & 0x3f) << gcss; \            ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
644              ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
645          eptr += 5; \
646        } \        } \
647      }      }
648    
649    #if defined COMPILE_PCRE8
650    
651    /* These macros were originally written in the form of loops that used data
652    from the tables whose names start with PRIV(utf8_table). They were rewritten by
653    a user so as not to use loops, because in some environments this gives a
654    significant performance advantage, and it seems never to do any harm. */
655    
656    /* Tells the biggest code point which can be encoded as a single character. */
657    
658    #define MAX_VALUE_FOR_SINGLE_CHAR 127
659    
660    /* Tests whether the code point needs extra characters to decode. */
661    
662    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
663    
664    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
665    Otherwise it has an undefined behaviour. */
666    
667    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
668    
669    /* Returns TRUE, if the given character is not the first character
670    of a UTF sequence. */
671    
672    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
673    
674    /* Get the next UTF-8 character, not advancing the pointer. This is called when
675    we know we are in UTF-8 mode. */
676    
677    #define GETCHAR(c, eptr) \
678      c = *eptr; \
679      if (c >= 0xc0) GETUTF8(c, eptr);
680    
681  /* 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
682  pointer. */  pointer. */
683    
684  #define GETCHARTEST(c, eptr) \  #define GETCHARTEST(c, eptr) \
685    c = *eptr; \    c = *eptr; \
686    if (utf8 && (c & 0xc0) == 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; \  
       } \  
     }  
687    
688  /* 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
689  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
690    
691  #define GETCHARINC(c, eptr) \  #define GETCHARINC(c, eptr) \
692    c = *eptr++; \    c = *eptr++; \
693    if ((c & 0xc0) == 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; \  
       } \  
     }  
694    
695  /* 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.
696    This is called when we don't know if we are in UTF-8 mode. */
697    
698  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
699    c = *eptr++; \    c = *eptr++; \
700    if (utf8 && (c & 0xc0) == 0xc0) \    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
701    
702    /* Base macro to pick up the remaining bytes of a UTF-8 character, not
703    advancing the pointer, incrementing the length. */
704    
705    #define GETUTF8LEN(c, eptr, len) \
706      { \      { \
707      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \      if ((c & 0x20) == 0) \
708      int gcss = 6*gcaa; \        { \
709      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
710      while (gcaa-- > 0) \        len++; \
711          } \
712        else if ((c & 0x10)  == 0) \
713        { \        { \
714        gcss -= 6; \        c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
715        c |= (*eptr++ & 0x3f) << gcss; \        len += 2; \
716          } \
717        else if ((c & 0x08)  == 0) \
718          {\
719          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
720              ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
721          len += 3; \
722          } \
723        else if ((c & 0x04)  == 0) \
724          { \
725          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
726              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
727              (eptr[4] & 0x3f); \
728          len += 4; \
729          } \
730        else \
731          {\
732          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
733              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
734              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
735          len += 5; \
736        } \        } \
737      }      }
738    
# Line 339  if there are extra bytes. This is called Line 741  if there are extra bytes. This is called
741    
742  #define GETCHARLEN(c, eptr, len) \  #define GETCHARLEN(c, eptr, len) \
743    c = *eptr; \    c = *eptr; \
744    if ((c & 0xc0) == 0xc0) \    if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
745      { \  
746      int gcii; \  /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
747      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  pointer, incrementing length if there are extra bytes. This is called when we
748      int gcss = 6*gcaa; \  do not know if we are in UTF-8 mode. */
749      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
750      for (gcii = 1; gcii <= gcaa; gcii++) \  #define GETCHARLENTEST(c, eptr, len) \
751        { \    c = *eptr; \
752        gcss -= 6; \    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
       c |= (eptr[gcii] & 0x3f) << gcss; \  
       } \  
     len += gcaa; \  
     }  
753    
754  /* 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
755  it is. Called only in UTF-8 mode. */  it is. This is called only in UTF-8 mode - we don't put a test within the macro
756    because almost all calls are already within a block of UTF-8 only code. */
757    
758  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--;  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
759    
760  #endif  /* Same as above, just in the other direction. */
761    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
762    
763    /* Same as above, but it allows a fully customizable form. */
764    #define ACROSSCHAR(condition, eptr, action) \
765      while((condition) && ((eptr) & 0xc0) == 0x80) action
766    
767  /* In case there is no definition of offsetof() provided - though any proper  #elif defined COMPILE_PCRE16
 Standard C system should have one. */  
768    
769  #ifndef offsetof  /* Tells the biggest code point which can be encoded as a single character. */
 #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
 #endif  
770    
771    #define MAX_VALUE_FOR_SINGLE_CHAR 65535
772    
773  /* These are the public options that can change during matching. */  /* Tests whether the code point needs extra characters to decode. */
774    
775  #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)  #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
776    
777  /* Private options flags start at the most significant end of the four bytes,  /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
778  but skip the top bit so we can use ints for convenience without getting tangled  Otherwise it has an undefined behaviour. */
 with negative values. The public options defined in pcre.h start at the least  
 significant end. Make sure they don't overlap! */  
779    
780  #define PCRE_FIRSTSET      0x40000000  /* first_byte is set */  #define GET_EXTRALEN(c) 1
 #define PCRE_REQCHSET      0x20000000  /* req_byte is set */  
 #define PCRE_STARTLINE     0x10000000  /* start after \n for multiline */  
 #define PCRE_ICHANGED      0x08000000  /* i option changes within regex */  
 #define PCRE_NOPARTIAL     0x04000000  /* can't use partial with this regex */  
781    
782  /* Options for the "extra" block produced by pcre_study(). */  /* Returns TRUE, if the given character is not the first character
783    of a UTF sequence. */
784    
785  #define PCRE_STUDY_MAPPED   0x01     /* a map of starting chars exists */  #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
786    
787    /* Base macro to pick up the low surrogate of a UTF-16 character, not
788    advancing the pointer. */
789    
790    #define GETUTF16(c, eptr) \
791       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
792    
793    /* Get the next UTF-16 character, not advancing the pointer. This is called when
794    we know we are in UTF-16 mode. */
795    
796    #define GETCHAR(c, eptr) \
797      c = *eptr; \
798      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
799    
800    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
801    pointer. */
802    
803    #define GETCHARTEST(c, eptr) \
804      c = *eptr; \
805      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
806    
807    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
808    the pointer. */
809    
810    #define GETUTF16INC(c, eptr) \
811       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
812    
813    /* Get the next UTF-16 character, advancing the pointer. This is called when we
814    know we are in UTF-16 mode. */
815    
816    #define GETCHARINC(c, eptr) \
817      c = *eptr++; \
818      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
819    
820    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
821    This is called when we don't know if we are in UTF-16 mode. */
822    
823    #define GETCHARINCTEST(c, eptr) \
824      c = *eptr++; \
825      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
826    
827    /* Base macro to pick up the low surrogate of a UTF-16 character, not
828    advancing the pointer, incrementing the length. */
829    
830    #define GETUTF16LEN(c, eptr, len) \
831       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
832    
833    /* Get the next UTF-16 character, not advancing the pointer, incrementing
834    length if there is a low surrogate. This is called when we know we are in
835    UTF-16 mode. */
836    
837    #define GETCHARLEN(c, eptr, len) \
838      c = *eptr; \
839      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
840    
841    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
842    pointer, incrementing length if there is a low surrogate. This is called when
843    we do not know if we are in UTF-16 mode. */
844    
845    #define GETCHARLENTEST(c, eptr, len) \
846      c = *eptr; \
847      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
848    
849    /* If the pointer is not at the start of a character, move it back until
850    it is. This is called only in UTF-16 mode - we don't put a test within the
851    macro because almost all calls are already within a block of UTF-16 only
852    code. */
853    
854    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
855    
856    /* Same as above, just in the other direction. */
857    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
858    
859    /* Same as above, but it allows a fully customizable form. */
860    #define ACROSSCHAR(condition, eptr, action) \
861      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
862    
863    #elif defined COMPILE_PCRE32
864    
865    /* These are unnecessary for the 32-bit library */
866    #undef MAX_VALUE_FOR_SINGLE_CHAR
867    #undef HAS_EXTRALEN
868    #undef GET_EXTRALEN
869    #undef NOT_FIRSTCHAR
870    
871    #define UTF32_MASK (0x1ffffful)
872    
873    /* Get the next UTF-32 character, not advancing the pointer. This is called when
874    we know we are in UTF-32 mode. */
875    
876    #define GETCHAR(c, eptr) \
877      c = (*eptr) & UTF32_MASK;
878    
879    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
880    pointer. */
881    
882    #define GETCHARTEST(c, eptr) \
883      c = *eptr; \
884      if (utf) c &= UTF32_MASK;
885    
886    /* Get the next UTF-32 character, advancing the pointer. This is called when we
887    know we are in UTF-32 mode. */
888    
889    #define GETCHARINC(c, eptr) \
890      c = (*eptr++) & UTF32_MASK;
891    
892    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
893    This is called when we don't know if we are in UTF-32 mode. */
894    
895    #define GETCHARINCTEST(c, eptr) \
896      c = *eptr++; \
897      if (utf) c &= UTF32_MASK;
898    
899    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
900    length (since all UTF-32 is of length 1). This is called when we know we are in
901    UTF-32 mode. */
902    
903    #define GETCHARLEN(c, eptr, len) \
904      GETCHAR(c, eptr)
905    
906    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
907    pointer, not incrementing the length (since all UTF-32 is of length 1).
908    This is called when we do not know if we are in UTF-32 mode. */
909    
910    #define GETCHARLENTEST(c, eptr, len) \
911      GETCHARTEST(c, eptr)
912    
913    /* If the pointer is not at the start of a character, move it back until
914    it is. This is called only in UTF-32 mode - we don't put a test within the
915    macro because almost all calls are already within a block of UTF-32 only
916    code. */
917    
918    #define BACKCHAR(eptr) do { } while (0)
919    
920    /* Same as above, just in the other direction. */
921    #define FORWARDCHAR(eptr) do { } while (0)
922    
923    /* Same as above, but it allows a fully customizable form. */
924    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
925    
926    #else
927    #error Unsupported compiling mode
928    #endif /* COMPILE_PCRE[8|16|32] */
929    
930    #endif  /* SUPPORT_UTF */
931    
932    /* Tests for Unicode horizontal and vertical whitespace characters must check a
933    number of different values. Using a switch statement for this generates the
934    fastest code (no loop, no memory access), and there are several places in the
935    interpreter code where this happens. In order to ensure that all the case lists
936    remain in step, we use macros so that there is only one place where the lists
937    are defined.
938    
939    These values are also required as lists in pcre_compile.c when processing \h,
940    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
941    macros that define the values are here so that all the definitions are
942    together. The lists must be in ascending character order, terminated by
943    NOTACHAR (which is 0xffffffff).
944    
945    Any changes should ensure that the various macros are kept in step with each
946    other. NOTE: The values also appear in pcre_jit_compile.c. */
947    
948    /* ------ ASCII/Unicode environments ------ */
949    
950    #ifndef EBCDIC
951    
952    #define HSPACE_LIST \
953      CHAR_HT, CHAR_SPACE, 0xa0, \
954      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
955      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
956      NOTACHAR
957    
958    #define HSPACE_MULTIBYTE_CASES \
959      case 0x1680:  /* OGHAM SPACE MARK */ \
960      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
961      case 0x2000:  /* EN QUAD */ \
962      case 0x2001:  /* EM QUAD */ \
963      case 0x2002:  /* EN SPACE */ \
964      case 0x2003:  /* EM SPACE */ \
965      case 0x2004:  /* THREE-PER-EM SPACE */ \
966      case 0x2005:  /* FOUR-PER-EM SPACE */ \
967      case 0x2006:  /* SIX-PER-EM SPACE */ \
968      case 0x2007:  /* FIGURE SPACE */ \
969      case 0x2008:  /* PUNCTUATION SPACE */ \
970      case 0x2009:  /* THIN SPACE */ \
971      case 0x200A:  /* HAIR SPACE */ \
972      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
973      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
974      case 0x3000   /* IDEOGRAPHIC SPACE */
975    
976    #define HSPACE_BYTE_CASES \
977      case CHAR_HT: \
978      case CHAR_SPACE: \
979      case 0xa0     /* NBSP */
980    
981    #define HSPACE_CASES \
982      HSPACE_BYTE_CASES: \
983      HSPACE_MULTIBYTE_CASES
984    
985    #define VSPACE_LIST \
986      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
987    
988    #define VSPACE_MULTIBYTE_CASES \
989      case 0x2028:    /* LINE SEPARATOR */ \
990      case 0x2029     /* PARAGRAPH SEPARATOR */
991    
992    #define VSPACE_BYTE_CASES \
993      case CHAR_LF: \
994      case CHAR_VT: \
995      case CHAR_FF: \
996      case CHAR_CR: \
997      case CHAR_NEL
998    
999    #define VSPACE_CASES \
1000      VSPACE_BYTE_CASES: \
1001      VSPACE_MULTIBYTE_CASES
1002    
1003    /* ------ EBCDIC environments ------ */
1004    
1005    #else
1006    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1007    
1008    #define HSPACE_BYTE_CASES \
1009      case CHAR_HT: \
1010      case CHAR_SPACE
1011    
1012    #define HSPACE_CASES HSPACE_BYTE_CASES
1013    
1014    #ifdef EBCDIC_NL25
1015    #define VSPACE_LIST \
1016      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1017    #else
1018    #define VSPACE_LIST \
1019      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1020    #endif
1021    
1022    #define VSPACE_BYTE_CASES \
1023      case CHAR_LF: \
1024      case CHAR_VT: \
1025      case CHAR_FF: \
1026      case CHAR_CR: \
1027      case CHAR_NEL
1028    
1029    #define VSPACE_CASES VSPACE_BYTE_CASES
1030    #endif  /* EBCDIC */
1031    
1032    /* ------ End of whitespace macros ------ */
1033    
1034    
1035    
1036    /* Private flags containing information about the compiled regex. They used to
1037    live at the top end of the options word, but that got almost full, so now they
1038    are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
1039    the restrictions on partial matching have been lifted. It remains for backwards
1040    compatibility. */
1041    
1042    #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1043    #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1044    #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
1045    #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1046    #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1047    #define PCRE_REQCHSET      0x0040  /* req_byte is set */
1048    #define PCRE_RCH_CASELESS  0x0080  /* caseless requested char */
1049    #define PCRE_STARTLINE     0x0100  /* start after \n for multiline */
1050    #define PCRE_NOPARTIAL     0x0200  /* can't use partial with this regex */
1051    #define PCRE_JCHANGED      0x0400  /* j option used in regex */
1052    #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1053    #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1054    
1055    #if defined COMPILE_PCRE8
1056    #define PCRE_MODE          PCRE_MODE8
1057    #elif defined COMPILE_PCRE16
1058    #define PCRE_MODE          PCRE_MODE16
1059    #elif defined COMPILE_PCRE32
1060    #define PCRE_MODE          PCRE_MODE32
1061    #endif
1062    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1063    
1064    /* Flags for the "extra" block produced by pcre_study(). */
1065    
1066    #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
1067    #define PCRE_STUDY_MINLEN  0x0002  /* a minimum length field exists */
1068    
1069  /* Masks for identifying the public options that are permitted at compile  /* Masks for identifying the public options that are permitted at compile
1070  time, run time, or study time, respectively. */  time, run time, or study time, respectively. */
1071    
1072  #define PUBLIC_OPTIONS \  #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
1073                               PCRE_NEWLINE_ANYCRLF)
1074    
1075    #define PUBLIC_COMPILE_OPTIONS \
1076    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
1077     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1078     PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE)     PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT|PCRE_FIRSTLINE| \
1079       PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1080       PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
1081    
1082  #define PUBLIC_EXEC_OPTIONS \  #define PUBLIC_EXEC_OPTIONS \
1083    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1084     PCRE_PARTIAL)     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
1085       PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
1086    
1087  #define PUBLIC_DFA_EXEC_OPTIONS \  #define PUBLIC_DFA_EXEC_OPTIONS \
1088    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1089     PCRE_PARTIAL|PCRE_DFA_SHORTEST|PCRE_DFA_RESTART)     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
1090       PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1091       PCRE_NO_START_OPTIMIZE)
1092    
1093    #define PUBLIC_STUDY_OPTIONS \
1094       (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1095        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1096    
1097  #define PUBLIC_STUDY_OPTIONS 0   /* None defined */  /* Magic number to provide a small check against being handed junk. */
   
 /* Magic number to provide a small check against being handed junk. Also used  
 to detect whether a pattern was compiled on a host of different endianness. */  
1098    
1099  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1100    
1101  /* Negative values for the firstchar and reqchar variables */  /* This variable is used to detect a loaded regular expression
1102    in different endianness. */
1103    
1104  #define REQ_UNSET (-2)  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
 #define REQ_NONE  (-1)  
1105    
1106  /* 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
1107  req_byte match. */  req_byte match. */
1108    
1109  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1110    
1111  /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1112  variable-length repeat, or a anything other than literal characters. */  environments where these macros are defined elsewhere. Unfortunately, there
1113    is no way to do the same for the typedef. */
 #define REQ_CASELESS 0x0100    /* indicates caselessness */  
 #define REQ_VARY     0x0200    /* reqbyte followed non-literal item */  
   
 /* Miscellaneous definitions */  
1114    
1115  typedef int BOOL;  typedef int BOOL;
1116    
1117    #ifndef FALSE
1118  #define FALSE   0  #define FALSE   0
1119  #define TRUE    1  #define TRUE    1
1120    #endif
1121    
1122    /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
1123    character constants like '*' because the compiler would emit their EBCDIC code,
1124    which is different from their ASCII/UTF-8 code. Instead we define macros for
1125    the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
1126    is enabled. When UTF-8 support is not enabled, the definitions use character
1127    literals. Both character and string versions of each character are needed, and
1128    there are some longer strings as well.
1129    
1130    This means that, on EBCDIC platforms, the PCRE library can handle either
1131    EBCDIC, or UTF-8, but not both. To support both in the same compiled library
1132    would need different lookups depending on whether PCRE_UTF8 was set or not.
1133    This would make it impossible to use characters in switch/case statements,
1134    which would reduce performance. For a theoretical use (which nobody has asked
1135    for) in a minority area (EBCDIC platforms), this is not sensible. Any
1136    application that did need both could compile two versions of the library, using
1137    macros to give the functions distinct names. */
1138    
1139    #ifndef SUPPORT_UTF
1140    
1141    /* UTF-8 support is not enabled; use the platform-dependent character literals
1142    so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1143    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1144    characters, a common practice has been to use its NL (0x15) character as the
1145    line terminator in C-like processing environments. However, sometimes the LF
1146    (0x25) character is used instead, according to this Unicode document:
1147    
1148    http://unicode.org/standard/reports/tr13/tr13-5.html
1149    
1150    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1151    instead. Whichever is *not* chosen is defined as NEL.
1152    
1153    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1154    same code point. */
1155    
1156    #ifdef EBCDIC
1157    
1158    #ifndef EBCDIC_NL25
1159    #define CHAR_NL                     '\x15'
1160    #define CHAR_NEL                    '\x25'
1161    #define STR_NL                      "\x15"
1162    #define STR_NEL                     "\x25"
1163    #else
1164    #define CHAR_NL                     '\x25'
1165    #define CHAR_NEL                    '\x15'
1166    #define STR_NL                      "\x25"
1167    #define STR_NEL                     "\x15"
1168    #endif
1169    
1170    #define CHAR_LF                     CHAR_NL
1171    #define STR_LF                      STR_NL
1172    
1173    #define CHAR_ESC                    '\047'
1174    #define CHAR_DEL                    '\007'
1175    #define STR_ESC                     "\047"
1176    #define STR_DEL                     "\007"
1177    
1178    #else  /* Not EBCDIC */
1179    
1180    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1181    compatibility. NEL is the Unicode newline character; make sure it is
1182    a positive value. */
1183    
1184    #define CHAR_LF                     '\n'
1185    #define CHAR_NL                     CHAR_LF
1186    #define CHAR_NEL                    ((unsigned char)'\x85')
1187    #define CHAR_ESC                    '\033'
1188    #define CHAR_DEL                    '\177'
1189    
1190    #define STR_LF                      "\n"
1191    #define STR_NL                      STR_LF
1192    #define STR_NEL                     "\x85"
1193    #define STR_ESC                     "\033"
1194    #define STR_DEL                     "\177"
1195    
1196    #endif  /* EBCDIC */
1197    
1198    /* The remaining definitions work in both environments. */
1199    
1200    #define CHAR_HT                     '\t'
1201    #define CHAR_VT                     '\v'
1202    #define CHAR_FF                     '\f'
1203    #define CHAR_CR                     '\r'
1204    #define CHAR_BS                     '\b'
1205    #define CHAR_BEL                    '\a'
1206    
1207    #define CHAR_SPACE                  ' '
1208    #define CHAR_EXCLAMATION_MARK       '!'
1209    #define CHAR_QUOTATION_MARK         '"'
1210    #define CHAR_NUMBER_SIGN            '#'
1211    #define CHAR_DOLLAR_SIGN            '$'
1212    #define CHAR_PERCENT_SIGN           '%'
1213    #define CHAR_AMPERSAND              '&'
1214    #define CHAR_APOSTROPHE             '\''
1215    #define CHAR_LEFT_PARENTHESIS       '('
1216    #define CHAR_RIGHT_PARENTHESIS      ')'
1217    #define CHAR_ASTERISK               '*'
1218    #define CHAR_PLUS                   '+'
1219    #define CHAR_COMMA                  ','
1220    #define CHAR_MINUS                  '-'
1221    #define CHAR_DOT                    '.'
1222    #define CHAR_SLASH                  '/'
1223    #define CHAR_0                      '0'
1224    #define CHAR_1                      '1'
1225    #define CHAR_2                      '2'
1226    #define CHAR_3                      '3'
1227    #define CHAR_4                      '4'
1228    #define CHAR_5                      '5'
1229    #define CHAR_6                      '6'
1230    #define CHAR_7                      '7'
1231    #define CHAR_8                      '8'
1232    #define CHAR_9                      '9'
1233    #define CHAR_COLON                  ':'
1234    #define CHAR_SEMICOLON              ';'
1235    #define CHAR_LESS_THAN_SIGN         '<'
1236    #define CHAR_EQUALS_SIGN            '='
1237    #define CHAR_GREATER_THAN_SIGN      '>'
1238    #define CHAR_QUESTION_MARK          '?'
1239    #define CHAR_COMMERCIAL_AT          '@'
1240    #define CHAR_A                      'A'
1241    #define CHAR_B                      'B'
1242    #define CHAR_C                      'C'
1243    #define CHAR_D                      'D'
1244    #define CHAR_E                      'E'
1245    #define CHAR_F                      'F'
1246    #define CHAR_G                      'G'
1247    #define CHAR_H                      'H'
1248    #define CHAR_I                      'I'
1249    #define CHAR_J                      'J'
1250    #define CHAR_K                      'K'
1251    #define CHAR_L                      'L'
1252    #define CHAR_M                      'M'
1253    #define CHAR_N                      'N'
1254    #define CHAR_O                      'O'
1255    #define CHAR_P                      'P'
1256    #define CHAR_Q                      'Q'
1257    #define CHAR_R                      'R'
1258    #define CHAR_S                      'S'
1259    #define CHAR_T                      'T'
1260    #define CHAR_U                      'U'
1261    #define CHAR_V                      'V'
1262    #define CHAR_W                      'W'
1263    #define CHAR_X                      'X'
1264    #define CHAR_Y                      'Y'
1265    #define CHAR_Z                      'Z'
1266    #define CHAR_LEFT_SQUARE_BRACKET    '['
1267    #define CHAR_BACKSLASH              '\\'
1268    #define CHAR_RIGHT_SQUARE_BRACKET   ']'
1269    #define CHAR_CIRCUMFLEX_ACCENT      '^'
1270    #define CHAR_UNDERSCORE             '_'
1271    #define CHAR_GRAVE_ACCENT           '`'
1272    #define CHAR_a                      'a'
1273    #define CHAR_b                      'b'
1274    #define CHAR_c                      'c'
1275    #define CHAR_d                      'd'
1276    #define CHAR_e                      'e'
1277    #define CHAR_f                      'f'
1278    #define CHAR_g                      'g'
1279    #define CHAR_h                      'h'
1280    #define CHAR_i                      'i'
1281    #define CHAR_j                      'j'
1282    #define CHAR_k                      'k'
1283    #define CHAR_l                      'l'
1284    #define CHAR_m                      'm'
1285    #define CHAR_n                      'n'
1286    #define CHAR_o                      'o'
1287    #define CHAR_p                      'p'
1288    #define CHAR_q                      'q'
1289    #define CHAR_r                      'r'
1290    #define CHAR_s                      's'
1291    #define CHAR_t                      't'
1292    #define CHAR_u                      'u'
1293    #define CHAR_v                      'v'
1294    #define CHAR_w                      'w'
1295    #define CHAR_x                      'x'
1296    #define CHAR_y                      'y'
1297    #define CHAR_z                      'z'
1298    #define CHAR_LEFT_CURLY_BRACKET     '{'
1299    #define CHAR_VERTICAL_LINE          '|'
1300    #define CHAR_RIGHT_CURLY_BRACKET    '}'
1301    #define CHAR_TILDE                  '~'
1302    
1303    #define STR_HT                      "\t"
1304    #define STR_VT                      "\v"
1305    #define STR_FF                      "\f"
1306    #define STR_CR                      "\r"
1307    #define STR_BS                      "\b"
1308    #define STR_BEL                     "\a"
1309    
1310    #define STR_SPACE                   " "
1311    #define STR_EXCLAMATION_MARK        "!"
1312    #define STR_QUOTATION_MARK          "\""
1313    #define STR_NUMBER_SIGN             "#"
1314    #define STR_DOLLAR_SIGN             "$"
1315    #define STR_PERCENT_SIGN            "%"
1316    #define STR_AMPERSAND               "&"
1317    #define STR_APOSTROPHE              "'"
1318    #define STR_LEFT_PARENTHESIS        "("
1319    #define STR_RIGHT_PARENTHESIS       ")"
1320    #define STR_ASTERISK                "*"
1321    #define STR_PLUS                    "+"
1322    #define STR_COMMA                   ","
1323    #define STR_MINUS                   "-"
1324    #define STR_DOT                     "."
1325    #define STR_SLASH                   "/"
1326    #define STR_0                       "0"
1327    #define STR_1                       "1"
1328    #define STR_2                       "2"
1329    #define STR_3                       "3"
1330    #define STR_4                       "4"
1331    #define STR_5                       "5"
1332    #define STR_6                       "6"
1333    #define STR_7                       "7"
1334    #define STR_8                       "8"
1335    #define STR_9                       "9"
1336    #define STR_COLON                   ":"
1337    #define STR_SEMICOLON               ";"
1338    #define STR_LESS_THAN_SIGN          "<"
1339    #define STR_EQUALS_SIGN             "="
1340    #define STR_GREATER_THAN_SIGN       ">"
1341    #define STR_QUESTION_MARK           "?"
1342    #define STR_COMMERCIAL_AT           "@"
1343    #define STR_A                       "A"
1344    #define STR_B                       "B"
1345    #define STR_C                       "C"
1346    #define STR_D                       "D"
1347    #define STR_E                       "E"
1348    #define STR_F                       "F"
1349    #define STR_G                       "G"
1350    #define STR_H                       "H"
1351    #define STR_I                       "I"
1352    #define STR_J                       "J"
1353    #define STR_K                       "K"
1354    #define STR_L                       "L"
1355    #define STR_M                       "M"
1356    #define STR_N                       "N"
1357    #define STR_O                       "O"
1358    #define STR_P                       "P"
1359    #define STR_Q                       "Q"
1360    #define STR_R                       "R"
1361    #define STR_S                       "S"
1362    #define STR_T                       "T"
1363    #define STR_U                       "U"
1364    #define STR_V                       "V"
1365    #define STR_W                       "W"
1366    #define STR_X                       "X"
1367    #define STR_Y                       "Y"
1368    #define STR_Z                       "Z"
1369    #define STR_LEFT_SQUARE_BRACKET     "["
1370    #define STR_BACKSLASH               "\\"
1371    #define STR_RIGHT_SQUARE_BRACKET    "]"
1372    #define STR_CIRCUMFLEX_ACCENT       "^"
1373    #define STR_UNDERSCORE              "_"
1374    #define STR_GRAVE_ACCENT            "`"
1375    #define STR_a                       "a"
1376    #define STR_b                       "b"
1377    #define STR_c                       "c"
1378    #define STR_d                       "d"
1379    #define STR_e                       "e"
1380    #define STR_f                       "f"
1381    #define STR_g                       "g"
1382    #define STR_h                       "h"
1383    #define STR_i                       "i"
1384    #define STR_j                       "j"
1385    #define STR_k                       "k"
1386    #define STR_l                       "l"
1387    #define STR_m                       "m"
1388    #define STR_n                       "n"
1389    #define STR_o                       "o"
1390    #define STR_p                       "p"
1391    #define STR_q                       "q"
1392    #define STR_r                       "r"
1393    #define STR_s                       "s"
1394    #define STR_t                       "t"
1395    #define STR_u                       "u"
1396    #define STR_v                       "v"
1397    #define STR_w                       "w"
1398    #define STR_x                       "x"
1399    #define STR_y                       "y"
1400    #define STR_z                       "z"
1401    #define STR_LEFT_CURLY_BRACKET      "{"
1402    #define STR_VERTICAL_LINE           "|"
1403    #define STR_RIGHT_CURLY_BRACKET     "}"
1404    #define STR_TILDE                   "~"
1405    
1406    #define STRING_ACCEPT0              "ACCEPT\0"
1407    #define STRING_COMMIT0              "COMMIT\0"
1408    #define STRING_F0                   "F\0"
1409    #define STRING_FAIL0                "FAIL\0"
1410    #define STRING_MARK0                "MARK\0"
1411    #define STRING_PRUNE0               "PRUNE\0"
1412    #define STRING_SKIP0                "SKIP\0"
1413    #define STRING_THEN                 "THEN"
1414    
1415    #define STRING_alpha0               "alpha\0"
1416    #define STRING_lower0               "lower\0"
1417    #define STRING_upper0               "upper\0"
1418    #define STRING_alnum0               "alnum\0"
1419    #define STRING_ascii0               "ascii\0"
1420    #define STRING_blank0               "blank\0"
1421    #define STRING_cntrl0               "cntrl\0"
1422    #define STRING_digit0               "digit\0"
1423    #define STRING_graph0               "graph\0"
1424    #define STRING_print0               "print\0"
1425    #define STRING_punct0               "punct\0"
1426    #define STRING_space0               "space\0"
1427    #define STRING_word0                "word\0"
1428    #define STRING_xdigit               "xdigit"
1429    
1430    #define STRING_DEFINE               "DEFINE"
1431    
1432    #define STRING_CR_RIGHTPAR             "CR)"
1433    #define STRING_LF_RIGHTPAR             "LF)"
1434    #define STRING_CRLF_RIGHTPAR           "CRLF)"
1435    #define STRING_ANY_RIGHTPAR            "ANY)"
1436    #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1437    #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1438    #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1439    #ifdef COMPILE_PCRE8
1440    #define STRING_UTF_RIGHTPAR            "UTF8)"
1441    #endif
1442    #ifdef COMPILE_PCRE16
1443    #define STRING_UTF_RIGHTPAR            "UTF16)"
1444    #endif
1445    #ifdef COMPILE_PCRE32
1446    #define STRING_UTF_RIGHTPAR            "UTF32)"
1447    #endif
1448    #define STRING_UCP_RIGHTPAR            "UCP)"
1449    #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1450    
1451    #else  /* SUPPORT_UTF */
1452    
1453  /* Escape items that are just an encoding of a particular data value. Note that  /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1454  ESC_n is defined as yet another macro, which is set in config.h to either \n  works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
1455  (the default) or \r (which some people want). */  only. */
1456    
1457    #define CHAR_HT                     '\011'
1458    #define CHAR_VT                     '\013'
1459    #define CHAR_FF                     '\014'
1460    #define CHAR_CR                     '\015'
1461    #define CHAR_LF                     '\012'
1462    #define CHAR_NL                     CHAR_LF
1463    #define CHAR_NEL                    ((unsigned char)'\x85')
1464    #define CHAR_BS                     '\010'
1465    #define CHAR_BEL                    '\007'
1466    #define CHAR_ESC                    '\033'
1467    #define CHAR_DEL                    '\177'
1468    
1469    #define CHAR_SPACE                  '\040'
1470    #define CHAR_EXCLAMATION_MARK       '\041'
1471    #define CHAR_QUOTATION_MARK         '\042'
1472    #define CHAR_NUMBER_SIGN            '\043'
1473    #define CHAR_DOLLAR_SIGN            '\044'
1474    #define CHAR_PERCENT_SIGN           '\045'
1475    #define CHAR_AMPERSAND              '\046'
1476    #define CHAR_APOSTROPHE             '\047'
1477    #define CHAR_LEFT_PARENTHESIS       '\050'
1478    #define CHAR_RIGHT_PARENTHESIS      '\051'
1479    #define CHAR_ASTERISK               '\052'
1480    #define CHAR_PLUS                   '\053'
1481    #define CHAR_COMMA                  '\054'
1482    #define CHAR_MINUS                  '\055'
1483    #define CHAR_DOT                    '\056'
1484    #define CHAR_SLASH                  '\057'
1485    #define CHAR_0                      '\060'
1486    #define CHAR_1                      '\061'
1487    #define CHAR_2                      '\062'
1488    #define CHAR_3                      '\063'
1489    #define CHAR_4                      '\064'
1490    #define CHAR_5                      '\065'
1491    #define CHAR_6                      '\066'
1492    #define CHAR_7                      '\067'
1493    #define CHAR_8                      '\070'
1494    #define CHAR_9                      '\071'
1495    #define CHAR_COLON                  '\072'
1496    #define CHAR_SEMICOLON              '\073'
1497    #define CHAR_LESS_THAN_SIGN         '\074'
1498    #define CHAR_EQUALS_SIGN            '\075'
1499    #define CHAR_GREATER_THAN_SIGN      '\076'
1500    #define CHAR_QUESTION_MARK          '\077'
1501    #define CHAR_COMMERCIAL_AT          '\100'
1502    #define CHAR_A                      '\101'
1503    #define CHAR_B                      '\102'
1504    #define CHAR_C                      '\103'
1505    #define CHAR_D                      '\104'
1506    #define CHAR_E                      '\105'
1507    #define CHAR_F                      '\106'
1508    #define CHAR_G                      '\107'
1509    #define CHAR_H                      '\110'
1510    #define CHAR_I                      '\111'
1511    #define CHAR_J                      '\112'
1512    #define CHAR_K                      '\113'
1513    #define CHAR_L                      '\114'
1514    #define CHAR_M                      '\115'
1515    #define CHAR_N                      '\116'
1516    #define CHAR_O                      '\117'
1517    #define CHAR_P                      '\120'
1518    #define CHAR_Q                      '\121'
1519    #define CHAR_R                      '\122'
1520    #define CHAR_S                      '\123'
1521    #define CHAR_T                      '\124'
1522    #define CHAR_U                      '\125'
1523    #define CHAR_V                      '\126'
1524    #define CHAR_W                      '\127'
1525    #define CHAR_X                      '\130'
1526    #define CHAR_Y                      '\131'
1527    #define CHAR_Z                      '\132'
1528    #define CHAR_LEFT_SQUARE_BRACKET    '\133'
1529    #define CHAR_BACKSLASH              '\134'
1530    #define CHAR_RIGHT_SQUARE_BRACKET   '\135'
1531    #define CHAR_CIRCUMFLEX_ACCENT      '\136'
1532    #define CHAR_UNDERSCORE             '\137'
1533    #define CHAR_GRAVE_ACCENT           '\140'
1534    #define CHAR_a                      '\141'
1535    #define CHAR_b                      '\142'
1536    #define CHAR_c                      '\143'
1537    #define CHAR_d                      '\144'
1538    #define CHAR_e                      '\145'
1539    #define CHAR_f                      '\146'
1540    #define CHAR_g                      '\147'
1541    #define CHAR_h                      '\150'
1542    #define CHAR_i                      '\151'
1543    #define CHAR_j                      '\152'
1544    #define CHAR_k                      '\153'
1545    #define CHAR_l                      '\154'
1546    #define CHAR_m                      '\155'
1547    #define CHAR_n                      '\156'
1548    #define CHAR_o                      '\157'
1549    #define CHAR_p                      '\160'
1550    #define CHAR_q                      '\161'
1551    #define CHAR_r                      '\162'
1552    #define CHAR_s                      '\163'
1553    #define CHAR_t                      '\164'
1554    #define CHAR_u                      '\165'
1555    #define CHAR_v                      '\166'
1556    #define CHAR_w                      '\167'
1557    #define CHAR_x                      '\170'
1558    #define CHAR_y                      '\171'
1559    #define CHAR_z                      '\172'
1560    #define CHAR_LEFT_CURLY_BRACKET     '\173'
1561    #define CHAR_VERTICAL_LINE          '\174'
1562    #define CHAR_RIGHT_CURLY_BRACKET    '\175'
1563    #define CHAR_TILDE                  '\176'
1564    
1565    #define STR_HT                      "\011"
1566    #define STR_VT                      "\013"
1567    #define STR_FF                      "\014"
1568    #define STR_CR                      "\015"
1569    #define STR_NL                      "\012"
1570    #define STR_BS                      "\010"
1571    #define STR_BEL                     "\007"
1572    #define STR_ESC                     "\033"
1573    #define STR_DEL                     "\177"
1574    
1575    #define STR_SPACE                   "\040"
1576    #define STR_EXCLAMATION_MARK        "\041"
1577    #define STR_QUOTATION_MARK          "\042"
1578    #define STR_NUMBER_SIGN             "\043"
1579    #define STR_DOLLAR_SIGN             "\044"
1580    #define STR_PERCENT_SIGN            "\045"
1581    #define STR_AMPERSAND               "\046"
1582    #define STR_APOSTROPHE              "\047"
1583    #define STR_LEFT_PARENTHESIS        "\050"
1584    #define STR_RIGHT_PARENTHESIS       "\051"
1585    #define STR_ASTERISK                "\052"
1586    #define STR_PLUS                    "\053"
1587    #define STR_COMMA                   "\054"
1588    #define STR_MINUS                   "\055"
1589    #define STR_DOT                     "\056"
1590    #define STR_SLASH                   "\057"
1591    #define STR_0                       "\060"
1592    #define STR_1                       "\061"
1593    #define STR_2                       "\062"
1594    #define STR_3                       "\063"
1595    #define STR_4                       "\064"
1596    #define STR_5                       "\065"
1597    #define STR_6                       "\066"
1598    #define STR_7                       "\067"
1599    #define STR_8                       "\070"
1600    #define STR_9                       "\071"
1601    #define STR_COLON                   "\072"
1602    #define STR_SEMICOLON               "\073"
1603    #define STR_LESS_THAN_SIGN          "\074"
1604    #define STR_EQUALS_SIGN             "\075"
1605    #define STR_GREATER_THAN_SIGN       "\076"
1606    #define STR_QUESTION_MARK           "\077"
1607    #define STR_COMMERCIAL_AT           "\100"
1608    #define STR_A                       "\101"
1609    #define STR_B                       "\102"
1610    #define STR_C                       "\103"
1611    #define STR_D                       "\104"
1612    #define STR_E                       "\105"
1613    #define STR_F                       "\106"
1614    #define STR_G                       "\107"
1615    #define STR_H                       "\110"
1616    #define STR_I                       "\111"
1617    #define STR_J                       "\112"
1618    #define STR_K                       "\113"
1619    #define STR_L                       "\114"
1620    #define STR_M                       "\115"
1621    #define STR_N                       "\116"
1622    #define STR_O                       "\117"
1623    #define STR_P                       "\120"
1624    #define STR_Q                       "\121"
1625    #define STR_R                       "\122"
1626    #define STR_S                       "\123"
1627    #define STR_T                       "\124"
1628    #define STR_U                       "\125"
1629    #define STR_V                       "\126"
1630    #define STR_W                       "\127"
1631    #define STR_X                       "\130"
1632    #define STR_Y                       "\131"
1633    #define STR_Z                       "\132"
1634    #define STR_LEFT_SQUARE_BRACKET     "\133"
1635    #define STR_BACKSLASH               "\134"
1636    #define STR_RIGHT_SQUARE_BRACKET    "\135"
1637    #define STR_CIRCUMFLEX_ACCENT       "\136"
1638    #define STR_UNDERSCORE              "\137"
1639    #define STR_GRAVE_ACCENT            "\140"
1640    #define STR_a                       "\141"
1641    #define STR_b                       "\142"
1642    #define STR_c                       "\143"
1643    #define STR_d                       "\144"
1644    #define STR_e                       "\145"
1645    #define STR_f                       "\146"
1646    #define STR_g                       "\147"
1647    #define STR_h                       "\150"
1648    #define STR_i                       "\151"
1649    #define STR_j                       "\152"
1650    #define STR_k                       "\153"
1651    #define STR_l                       "\154"
1652    #define STR_m                       "\155"
1653    #define STR_n                       "\156"
1654    #define STR_o                       "\157"
1655    #define STR_p                       "\160"
1656    #define STR_q                       "\161"
1657    #define STR_r                       "\162"
1658    #define STR_s                       "\163"
1659    #define STR_t                       "\164"
1660    #define STR_u                       "\165"
1661    #define STR_v                       "\166"
1662    #define STR_w                       "\167"
1663    #define STR_x                       "\170"
1664    #define STR_y                       "\171"
1665    #define STR_z                       "\172"
1666    #define STR_LEFT_CURLY_BRACKET      "\173"
1667    #define STR_VERTICAL_LINE           "\174"
1668    #define STR_RIGHT_CURLY_BRACKET     "\175"
1669    #define STR_TILDE                   "\176"
1670    
1671    #define STRING_ACCEPT0              STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1672    #define STRING_COMMIT0              STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1673    #define STRING_F0                   STR_F "\0"
1674    #define STRING_FAIL0                STR_F STR_A STR_I STR_L "\0"
1675    #define STRING_MARK0                STR_M STR_A STR_R STR_K "\0"
1676    #define STRING_PRUNE0               STR_P STR_R STR_U STR_N STR_E "\0"
1677    #define STRING_SKIP0                STR_S STR_K STR_I STR_P "\0"
1678    #define STRING_THEN                 STR_T STR_H STR_E STR_N
1679    
1680    #define STRING_alpha0               STR_a STR_l STR_p STR_h STR_a "\0"
1681    #define STRING_lower0               STR_l STR_o STR_w STR_e STR_r "\0"
1682    #define STRING_upper0               STR_u STR_p STR_p STR_e STR_r "\0"
1683    #define STRING_alnum0               STR_a STR_l STR_n STR_u STR_m "\0"
1684    #define STRING_ascii0               STR_a STR_s STR_c STR_i STR_i "\0"
1685    #define STRING_blank0               STR_b STR_l STR_a STR_n STR_k "\0"
1686    #define STRING_cntrl0               STR_c STR_n STR_t STR_r STR_l "\0"
1687    #define STRING_digit0               STR_d STR_i STR_g STR_i STR_t "\0"
1688    #define STRING_graph0               STR_g STR_r STR_a STR_p STR_h "\0"
1689    #define STRING_print0               STR_p STR_r STR_i STR_n STR_t "\0"
1690    #define STRING_punct0               STR_p STR_u STR_n STR_c STR_t "\0"
1691    #define STRING_space0               STR_s STR_p STR_a STR_c STR_e "\0"
1692    #define STRING_word0                STR_w STR_o STR_r STR_d       "\0"
1693    #define STRING_xdigit               STR_x STR_d STR_i STR_g STR_i STR_t
1694    
1695    #define STRING_DEFINE               STR_D STR_E STR_F STR_I STR_N STR_E
1696    
1697    #define STRING_CR_RIGHTPAR             STR_C STR_R STR_RIGHT_PARENTHESIS
1698    #define STRING_LF_RIGHTPAR             STR_L STR_F STR_RIGHT_PARENTHESIS
1699    #define STRING_CRLF_RIGHTPAR           STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1700    #define STRING_ANY_RIGHTPAR            STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1701    #define STRING_ANYCRLF_RIGHTPAR        STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1702    #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
1703    #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
1704    #ifdef COMPILE_PCRE8
1705    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1706    #endif
1707    #ifdef COMPILE_PCRE16
1708    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1709    #endif
1710    #ifdef COMPILE_PCRE32
1711    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1712    #endif
1713    #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1714    #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
1715    
1716    #endif  /* SUPPORT_UTF */
1717    
1718    /* Escape items that are just an encoding of a particular data value. */
1719    
1720  #ifndef ESC_e  #ifndef ESC_e
1721  #define ESC_e 27  #define ESC_e CHAR_ESC
1722  #endif  #endif
1723    
1724  #ifndef ESC_f  #ifndef ESC_f
1725  #define ESC_f '\f'  #define ESC_f CHAR_FF
1726  #endif  #endif
1727    
1728  #ifndef ESC_n  #ifndef ESC_n
1729  #define ESC_n NEWLINE  #define ESC_n CHAR_LF
1730  #endif  #endif
1731    
1732  #ifndef ESC_r  #ifndef ESC_r
1733  #define ESC_r '\r'  #define ESC_r CHAR_CR
1734  #endif  #endif
1735    
1736  /* We can't officially use ESC_t because it is a POSIX reserved identifier  /* We can't officially use ESC_t because it is a POSIX reserved identifier
1737  (presumably because of all the others like size_t). */  (presumably because of all the others like size_t). */
1738    
1739  #ifndef ESC_tee  #ifndef ESC_tee
1740  #define ESC_tee '\t'  #define ESC_tee CHAR_HT
1741  #endif  #endif
1742    
1743  /* Codes for different types of Unicode property */  /* Codes for different types of Unicode property */
1744    
1745  #define PT_ANY        0    /* Any property - matches all chars */  #define PT_ANY        0    /* Any property - matches all chars */
1746  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */
1747  #define PT_GC         2    /* General characteristic (e.g. L) */  #define PT_GC         2    /* Specified general characteristic (e.g. L) */
1748  #define PT_PC         3    /* Particular characteristic (e.g. Lu) */  #define PT_PC         3    /* Specified particular characteristic (e.g. Lu) */
1749  #define PT_SC         4    /* Script (e.g. Han) */  #define PT_SC         4    /* Script (e.g. Han) */
1750    #define PT_ALNUM      5    /* Alphanumeric - the union of L and N */
1751    #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1752    #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1753    #define PT_WORD       8    /* Word - L plus N plus underscore */
1754    #define PT_CLIST      9    /* Pseudo-property: match character list */
1755    
1756  /* 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
1757  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1758    
1759  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1760  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */
# Line 483  contain UTF-8 characters with values gre Line 1767  contain UTF-8 characters with values gre
1767    
1768  /* 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
1769  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
1770  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
1771  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
1772  corresponds to "." rather than an escape sequence. The final one must be  corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1773  ESC_REF as subsequent values are used for \1, \2, \3, etc. There is are two  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1774  tests in the code for an escape greater than ESC_b and less than ESC_Z to  non-DOTALL mode, "." behaves like \N.
1775  detect the types that may be repeated. These are the types that consume  
1776  characters. If any new escapes are put in between that don't consume a  The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1777  character, that code will have to change. */  when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1778    They must be contiguous, and remain in order so that the replacements can be
1779  enum { ESC_A = 1, ESC_G, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, ESC_W,  looked up from a table.
1780         ESC_w, ESC_dum1, ESC_C, ESC_P, ESC_p, ESC_X, ESC_Z, ESC_z, ESC_E,  
1781         ESC_Q, ESC_REF };  Negative numbers are used to encode a backreference (\1, \2, \3, etc.) in
1782    check_escape(). There are two tests in the code for an escape
1783    greater than ESC_b and less than ESC_Z to detect the types that may be
1784    repeated. These are the types that consume characters. If any new escapes are
1785    put in between that don't consume a character, that code will have to change.
1786    */
1787    
1788    enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1789           ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1790           ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1791           ESC_E, ESC_Q, ESC_g, ESC_k,
1792           ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1793    
1794  /* Opcode table: OP_BRA must be last, as all values >= it are used for brackets  /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
 that extract substrings. Starting from 1 (i.e. after OP_END), the values up to  
1795  OP_EOD must correspond in order to the list of escapes immediately above.  OP_EOD must correspond in order to the list of escapes immediately above.
1796  Note that whenever this list is updated, the two macro definitions that follow  
1797  must also be updated to match. */  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1798    that follow must also be updated to match. There are also tables called
1799    "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1800    
1801  enum {  enum {
1802    OP_END,            /* 0 End of pattern */    OP_END,            /* 0 End of pattern */
# Line 509  enum { Line 1805  enum {
1805    
1806    OP_SOD,            /* 1 Start of data: \A */    OP_SOD,            /* 1 Start of data: \A */
1807    OP_SOM,            /* 2 Start of match (subject + offset): \G */    OP_SOM,            /* 2 Start of match (subject + offset): \G */
1808    OP_NOT_WORD_BOUNDARY,  /*  3 \B */    OP_SET_SOM,        /* 3 Set start of match (\K) */
1809    OP_WORD_BOUNDARY,      /*  4 \b */    OP_NOT_WORD_BOUNDARY,  /*  4 \B */
1810    OP_NOT_DIGIT,          /*  5 \D */    OP_WORD_BOUNDARY,      /*  5 \b */
1811    OP_DIGIT,              /*  6 \d */    OP_NOT_DIGIT,          /*  6 \D */
1812    OP_NOT_WHITESPACE,     /*  7 \S */    OP_DIGIT,              /*  7 \d */
1813    OP_WHITESPACE,         /*  8 \s */    OP_NOT_WHITESPACE,     /*  8 \S */
1814    OP_NOT_WORDCHAR,       /*  9 \W */    OP_WHITESPACE,         /*  9 \s */
1815    OP_WORDCHAR,           /* 10 \w */    OP_NOT_WORDCHAR,       /* 10 \W */
1816    OP_ANY,            /* 11 Match any character */    OP_WORDCHAR,           /* 11 \w */
1817    OP_ANYBYTE,        /* 12 Match any byte (\C); different to OP_ANY for UTF-8 */  
1818    OP_NOTPROP,        /* 13 \P (not Unicode property) */    OP_ANY,            /* 12 Match any character except newline (\N) */
1819    OP_PROP,           /* 14 \p (Unicode property) */    OP_ALLANY,         /* 13 Match any character */
1820    OP_EXTUNI,         /* 15 \X (extended Unicode sequence */    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1821    OP_EODN,           /* 16 End of data or \n at end of data: \Z. */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
1822    OP_EOD,            /* 17 End of data: \z */    OP_PROP,           /* 16 \p (Unicode property) */
1823      OP_ANYNL,          /* 17 \R (any newline sequence) */
1824    OP_OPT,            /* 18 Set runtime options */    OP_NOT_HSPACE,     /* 18 \H (not horizontal whitespace) */
1825    OP_CIRC,           /* 19 Start of line - varies with multiline switch */    OP_HSPACE,         /* 19 \h (horizontal whitespace) */
1826    OP_DOLL,           /* 20 End of line - varies with multiline switch */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1827    OP_CHAR,           /* 21 Match one character, casefully */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1828    OP_CHARNC,         /* 22 Match one character, caselessly */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1829    OP_NOT,            /* 23 Match anything but the following char */    OP_EODN,           /* 23 End of data or \n at end of data (\Z) */
1830      OP_EOD,            /* 24 End of data (\z) */
1831    OP_STAR,           /* 24 The maximizing and minimizing versions of */  
1832    OP_MINSTAR,        /* 25 all these opcodes must come in pairs, with */    OP_CIRC,           /* 25 Start of line - not multiline */
1833    OP_PLUS,           /* 26 the minimizing one second. */    OP_CIRCM,          /* 26 Start of line - multiline */
1834    OP_MINPLUS,        /* 27 This first set applies to single characters */    OP_DOLL,           /* 27 End of line - not multiline */
1835    OP_QUERY,          /* 28 */    OP_DOLLM,          /* 28 End of line - multiline */
1836    OP_MINQUERY,       /* 29 */    OP_CHAR,           /* 29 Match one character, casefully */
1837    OP_UPTO,           /* 30 From 0 to n matches */    OP_CHARI,          /* 30 Match one character, caselessly */
1838    OP_MINUPTO,        /* 31 */    OP_NOT,            /* 31 Match one character, not the given one, casefully */
1839    OP_EXACT,          /* 32 Exactly n matches */    OP_NOTI,           /* 32 Match one character, not the given one, caselessly */
1840    
1841    OP_NOTSTAR,        /* 33 The maximizing and minimizing versions of */    /* The following sets of 13 opcodes must always be kept in step because
1842    OP_NOTMINSTAR,     /* 34 all these opcodes must come in pairs, with */    the offset from the first one is used to generate the others. */
1843    OP_NOTPLUS,        /* 35 the minimizing one second. */  
1844    OP_NOTMINPLUS,     /* 36 This set applies to "not" single characters */    /**** Single characters, caseful, must precede the caseless ones ****/
1845    OP_NOTQUERY,       /* 37 */  
1846    OP_NOTMINQUERY,    /* 38 */    OP_STAR,           /* 33 The maximizing and minimizing versions of */
1847    OP_NOTUPTO,        /* 39 From 0 to n matches */    OP_MINSTAR,        /* 34 these six opcodes must come in pairs, with */
1848    OP_NOTMINUPTO,     /* 40 */    OP_PLUS,           /* 35 the minimizing one second. */
1849    OP_NOTEXACT,       /* 41 Exactly n matches */    OP_MINPLUS,        /* 36 */
1850      OP_QUERY,          /* 37 */
1851    OP_TYPESTAR,       /* 42 The maximizing and minimizing versions of */    OP_MINQUERY,       /* 38 */
1852    OP_TYPEMINSTAR,    /* 43 all these opcodes must come in pairs, with */  
1853    OP_TYPEPLUS,       /* 44 the minimizing one second. These codes must */    OP_UPTO,           /* 39 From 0 to n matches of one character, caseful*/
1854    OP_TYPEMINPLUS,    /* 45 be in exactly the same order as those above. */    OP_MINUPTO,        /* 40 */
1855    OP_TYPEQUERY,      /* 46 This set applies to character types such as \d */    OP_EXACT,          /* 41 Exactly n matches */
1856    OP_TYPEMINQUERY,   /* 47 */  
1857    OP_TYPEUPTO,       /* 48 From 0 to n matches */    OP_POSSTAR,        /* 42 Possessified star, caseful */
1858    OP_TYPEMINUPTO,    /* 49 */    OP_POSPLUS,        /* 43 Possessified plus, caseful */
1859    OP_TYPEEXACT,      /* 50 Exactly n matches */    OP_POSQUERY,       /* 44 Posesssified query, caseful */
1860      OP_POSUPTO,        /* 45 Possessified upto, caseful */
1861    OP_CRSTAR,         /* 51 The maximizing and minimizing versions of */  
1862    OP_CRMINSTAR,      /* 52 all these opcodes must come in pairs, with */    /**** Single characters, caseless, must follow the caseful ones */
1863    OP_CRPLUS,         /* 53 the minimizing one second. These codes must */  
1864    OP_CRMINPLUS,      /* 54 be in exactly the same order as those above. */    OP_STARI,          /* 46 */
1865    OP_CRQUERY,        /* 55 These are for character classes and back refs */    OP_MINSTARI,       /* 47 */
1866    OP_CRMINQUERY,     /* 56 */    OP_PLUSI,          /* 48 */
1867    OP_CRRANGE,        /* 57 These are different to the three sets above. */    OP_MINPLUSI,       /* 49 */
1868    OP_CRMINRANGE,     /* 58 */    OP_QUERYI,         /* 50 */
1869      OP_MINQUERYI,      /* 51 */
1870    OP_CLASS,          /* 59 Match a character class, chars < 256 only */  
1871    OP_NCLASS,         /* 60 Same, but the bitmap was created from a negative    OP_UPTOI,          /* 52 From 0 to n matches of one character, caseless */
1872                             class - the difference is relevant only when a UTF-8    OP_MINUPTOI,       /* 53 */
1873                             character > 255 is encountered. */    OP_EXACTI,         /* 54 */
1874    
1875    OP_XCLASS,         /* 61 Extended class for handling UTF-8 chars within the    OP_POSSTARI,       /* 55 Possessified star, caseless */
1876                             class. This does both positive and negative. */    OP_POSPLUSI,       /* 56 Possessified plus, caseless */
1877      OP_POSQUERYI,      /* 57 Posesssified query, caseless */
1878    OP_REF,            /* 62 Match a back reference */    OP_POSUPTOI,       /* 58 Possessified upto, caseless */
1879    OP_RECURSE,        /* 63 Match a numbered subpattern (possibly recursive) */  
1880    OP_CALLOUT,        /* 64 Call out to external function if provided */    /**** The negated ones must follow the non-negated ones, and match them ****/
1881      /**** Negated single character, caseful; must precede the caseless ones ****/
1882    OP_ALT,            /* 65 Start of alternation */  
1883    OP_KET,            /* 66 End of group that doesn't have an unbounded repeat */    OP_NOTSTAR,        /* 59 The maximizing and minimizing versions of */
1884    OP_KETRMAX,        /* 67 These two must remain together and in this */    OP_NOTMINSTAR,     /* 60 these six opcodes must come in pairs, with */
1885    OP_KETRMIN,        /* 68 order. They are for groups the repeat for ever. */    OP_NOTPLUS,        /* 61 the minimizing one second. They must be in */
1886      OP_NOTMINPLUS,     /* 62 exactly the same order as those above. */
1887    /* The assertions must come before ONCE and COND */    OP_NOTQUERY,       /* 63 */
1888      OP_NOTMINQUERY,    /* 64 */
1889    OP_ASSERT,         /* 69 Positive lookahead */  
1890    OP_ASSERT_NOT,     /* 70 Negative lookahead */    OP_NOTUPTO,        /* 65 From 0 to n matches, caseful */
1891    OP_ASSERTBACK,     /* 71 Positive lookbehind */    OP_NOTMINUPTO,     /* 66 */
1892    OP_ASSERTBACK_NOT, /* 72 Negative lookbehind */    OP_NOTEXACT,       /* 67 Exactly n matches */
1893    OP_REVERSE,        /* 73 Move pointer back - used in lookbehind assertions */  
1894      OP_NOTPOSSTAR,     /* 68 Possessified versions, caseful */
1895    /* ONCE and COND must come after the assertions, with ONCE first, as there's    OP_NOTPOSPLUS,     /* 69 */
1896    a test for >= ONCE for a subpattern that isn't an assertion. */    OP_NOTPOSQUERY,    /* 70 */
1897      OP_NOTPOSUPTO,     /* 71 */
1898    OP_ONCE,           /* 74 Once matched, don't back up into the subpattern */  
1899    OP_COND,           /* 75 Conditional group */    /**** Negated single character, caseless; must follow the caseful ones ****/
1900    OP_CREF,           /* 76 Used to hold an extraction string number (cond ref) */  
1901      OP_NOTSTARI,       /* 72 */
1902    OP_BRAZERO,        /* 77 These two must remain together and in this */    OP_NOTMINSTARI,    /* 73 */
1903    OP_BRAMINZERO,     /* 78 order. */    OP_NOTPLUSI,       /* 74 */
1904      OP_NOTMINPLUSI,    /* 75 */
1905    OP_BRANUMBER,      /* 79 Used for extracting brackets whose number is greater    OP_NOTQUERYI,      /* 76 */
1906                             than can fit into an opcode. */    OP_NOTMINQUERYI,   /* 77 */
1907    
1908    OP_BRA             /* 80 This and greater values are used for brackets that    OP_NOTUPTOI,       /* 78 From 0 to n matches, caseless */
1909                             extract substrings up to EXTRACT_BASIC_MAX. After    OP_NOTMINUPTOI,    /* 79 */
1910                             that, use is made of OP_BRANUMBER. */    OP_NOTEXACTI,      /* 80 Exactly n matches */
1911  };  
1912      OP_NOTPOSSTARI,    /* 81 Possessified versions, caseless */
1913      OP_NOTPOSPLUSI,    /* 82 */
1914      OP_NOTPOSQUERYI,   /* 83 */
1915      OP_NOTPOSUPTOI,    /* 84 */
1916    
1917      /**** Character types ****/
1918    
1919      OP_TYPESTAR,       /* 85 The maximizing and minimizing versions of */
1920      OP_TYPEMINSTAR,    /* 86 these six opcodes must come in pairs, with */
1921      OP_TYPEPLUS,       /* 87 the minimizing one second. These codes must */
1922      OP_TYPEMINPLUS,    /* 88 be in exactly the same order as those above. */
1923      OP_TYPEQUERY,      /* 89 */
1924      OP_TYPEMINQUERY,   /* 90 */
1925    
1926      OP_TYPEUPTO,       /* 91 From 0 to n matches */
1927      OP_TYPEMINUPTO,    /* 92 */
1928      OP_TYPEEXACT,      /* 93 Exactly n matches */
1929    
1930      OP_TYPEPOSSTAR,    /* 94 Possessified versions */
1931      OP_TYPEPOSPLUS,    /* 95 */
1932      OP_TYPEPOSQUERY,   /* 96 */
1933      OP_TYPEPOSUPTO,    /* 97 */
1934    
1935      /* These are used for character classes and back references; only the
1936      first six are the same as the sets above. */
1937    
1938      OP_CRSTAR,         /* 98 The maximizing and minimizing versions of */
1939      OP_CRMINSTAR,      /* 99 all these opcodes must come in pairs, with */
1940      OP_CRPLUS,         /* 100 the minimizing one second. These codes must */
1941      OP_CRMINPLUS,      /* 101 be in exactly the same order as those above. */
1942      OP_CRQUERY,        /* 102 */
1943      OP_CRMINQUERY,     /* 103 */
1944    
1945      OP_CRRANGE,        /* 104 These are different to the three sets above. */
1946      OP_CRMINRANGE,     /* 105 */
1947    
1948      /* End of quantifier opcodes */
1949    
1950      OP_CLASS,          /* 106 Match a character class, chars < 256 only */
1951      OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
1952                                  class - the difference is relevant only when a
1953                                  character > 255 is encountered. */
1954      OP_XCLASS,         /* 108 Extended class for handling > 255 chars within the
1955                                  class. This does both positive and negative. */
1956      OP_REF,            /* 109 Match a back reference, casefully */
1957      OP_REFI,           /* 110 Match a back reference, caselessly */
1958      OP_RECURSE,        /* 111 Match a numbered subpattern (possibly recursive) */
1959      OP_CALLOUT,        /* 112 Call out to external function if provided */
1960    
1961      OP_ALT,            /* 113 Start of alternation */
1962      OP_KET,            /* 114 End of group that doesn't have an unbounded repeat */
1963      OP_KETRMAX,        /* 115 These two must remain together and in this */
1964      OP_KETRMIN,        /* 116 order. They are for groups the repeat for ever. */
1965      OP_KETRPOS,        /* 117 Possessive unlimited repeat. */
1966    
1967      /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
1968      asserts must remain in order. */
1969    
1970      OP_REVERSE,        /* 118 Move pointer back - used in lookbehind assertions */
1971      OP_ASSERT,         /* 119 Positive lookahead */
1972      OP_ASSERT_NOT,     /* 120 Negative lookahead */
1973      OP_ASSERTBACK,     /* 121 Positive lookbehind */
1974      OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
1975    
1976      /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
1977      after the assertions, with ONCE first, as there's a test for >= ONCE for a
1978      subpattern that isn't an assertion. The POS versions must immediately follow
1979      the non-POS versions in each case. */
1980    
1981      OP_ONCE,           /* 123 Atomic group, contains captures */
1982      OP_ONCE_NC,        /* 124 Atomic group containing no captures */
1983      OP_BRA,            /* 125 Start of non-capturing bracket */
1984      OP_BRAPOS,         /* 126 Ditto, with unlimited, possessive repeat */
1985      OP_CBRA,           /* 127 Start of capturing bracket */
1986      OP_CBRAPOS,        /* 128 Ditto, with unlimited, possessive repeat */
1987      OP_COND,           /* 129 Conditional group */
1988    
1989      /* These five must follow the previous five, in the same order. There's a
1990      check for >= SBRA to distinguish the two sets. */
1991    
1992      OP_SBRA,           /* 130 Start of non-capturing bracket, check empty  */
1993      OP_SBRAPOS,        /* 131 Ditto, with unlimited, possessive repeat */
1994      OP_SCBRA,          /* 132 Start of capturing bracket, check empty */
1995      OP_SCBRAPOS,       /* 133 Ditto, with unlimited, possessive repeat */
1996      OP_SCOND,          /* 134 Conditional group, check empty */
1997    
1998      /* The next two pairs must (respectively) be kept together. */
1999    
2000      OP_CREF,           /* 135 Used to hold a capture number as condition */
2001      OP_NCREF,          /* 136 Same, but generated by a name reference*/
2002      OP_RREF,           /* 137 Used to hold a recursion number as condition */
2003      OP_NRREF,          /* 138 Same, but generated by a name reference*/
2004      OP_DEF,            /* 139 The DEFINE condition */
2005    
2006      OP_BRAZERO,        /* 140 These two must remain together and in this */
2007      OP_BRAMINZERO,     /* 141 order. */
2008      OP_BRAPOSZERO,     /* 142 */
2009    
2010      /* These are backtracking control verbs */
2011    
2012      OP_MARK,           /* 143 always has an argument */
2013      OP_PRUNE,          /* 144 */
2014      OP_PRUNE_ARG,      /* 145 same, but with argument */
2015      OP_SKIP,           /* 146 */
2016      OP_SKIP_ARG,       /* 147 same, but with argument */
2017      OP_THEN,           /* 148 */
2018      OP_THEN_ARG,       /* 149 same, but with argument */
2019      OP_COMMIT,         /* 150 */
2020    
2021      /* These are forced failure and success verbs */
2022    
2023      OP_FAIL,           /* 151 */
2024      OP_ACCEPT,         /* 152 */
2025      OP_ASSERT_ACCEPT,  /* 153 Used inside assertions */
2026      OP_CLOSE,          /* 154 Used before OP_ACCEPT to close open captures */
2027    
2028      /* This is used to skip a subpattern with a {0} quantifier */
2029    
2030      OP_SKIPZERO,       /* 155 */
2031    
2032      /* This is not an opcode, but is used to check that tables indexed by opcode
2033      are the correct length, in order to catch updating errors - there have been
2034      some in the past. */
2035    
2036  /* WARNING WARNING WARNING: There is an implicit assumption in pcre.c and    OP_TABLE_LENGTH
2037  study.c that all opcodes are less than 128 in value. This makes handling UTF-8  };
 character sequences easier. */  
   
 /* The highest extraction number before we have to start using additional  
 bytes. (Originally PCRE didn't have support for extraction counts highter than  
 this number.) The value is limited by the number of opcodes left after OP_BRA,  
 i.e. 255 - OP_BRA. We actually set it a bit lower to leave room for additional  
 opcodes. */  
2038    
2039  #define EXTRACT_BASIC_MAX  100  /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2040    definitions that follow must also be updated to match. There are also tables
2041    called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
2042    
2043    
2044  /* 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
2045  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
2046    only in pcre_printint.c, which fills out the full names in many cases (and in
2047    some cases doesn't actually use these names at all). */
2048    
2049  #define OP_NAME_LIST \  #define OP_NAME_LIST \
2050    "End", "\\A", "\\G", "\\B", "\\b", "\\D", "\\d",                \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
2051    "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte",                   \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
2052    "notprop", "prop", "extuni",                                    \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
2053    "\\Z", "\\z",                                                   \    "extuni",  "\\Z", "\\z",                                        \
2054    "Opt", "^", "$", "char", "charnc", "not",                       \    "^", "^", "$", "$", "char", "chari", "not", "noti",             \
2055    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2056    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "{", "{", "{",                                                  \
2057      "*+","++", "?+", "{",                                           \
2058      "*", "*?", "+", "+?", "?", "??",                                \
2059      "{", "{", "{",                                                  \
2060      "*+","++", "?+", "{",                                           \
2061      "*", "*?", "+", "+?", "?", "??",                                \
2062      "{", "{", "{",                                                  \
2063      "*+","++", "?+", "{",                                           \
2064      "*", "*?", "+", "+?", "?", "??",                                \
2065      "{", "{", "{",                                                  \
2066      "*+","++", "?+", "{",                                           \
2067    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
2068      "*+","++", "?+", "{",                                           \
2069    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \
2070    "class", "nclass", "xclass", "Ref", "Recurse", "Callout",       \    "class", "nclass", "xclass", "Ref", "Refi",                     \
2071    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",     \    "Recurse", "Callout",                                           \
2072    "AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cond ref",\    "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos",                  \
2073    "Brazero", "Braminzero", "Branumber", "Bra"    "Reverse", "Assert", "Assert not", "AssertB", "AssertB not",    \
2074      "Once", "Once_NC",                                              \
2075      "Bra", "BraPos", "CBra", "CBraPos",                             \
2076      "Cond",                                                         \
2077      "SBra", "SBraPos", "SCBra", "SCBraPos",                         \
2078      "SCond",                                                        \
2079      "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \
2080      "Brazero", "Braminzero", "Braposzero",                          \
2081      "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \
2082      "*THEN", "*THEN", "*COMMIT", "*FAIL",                           \
2083      "*ACCEPT", "*ASSERT_ACCEPT",                                    \
2084      "Close", "Skip zero"
2085    
2086    
2087  /* This macro defines the length of fixed length operations in the compiled  /* This macro defines the length of fixed length operations in the compiled
# Line 657  in UTF-8 mode. The code that uses this t Line 2095  in UTF-8 mode. The code that uses this t
2095    
2096  #define OP_LENGTHS \  #define OP_LENGTHS \
2097    1,                             /* End                                    */ \    1,                             /* End                                    */ \
2098    1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* \A, \G, \B, \B, \D, \d, \S, \s, \W, \w */ \    1, 1, 1, 1, 1,                 /* \A, \G, \K, \B, \b                     */ \
2099    1, 1,                          /* Any, Anybyte                           */ \    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
2100    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \
2101    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    3, 3,                          /* \P, \p                                 */ \
2102      1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
2103      1,                             /* \X                                     */ \
2104      1, 1, 1, 1, 1, 1,              /* \Z, \z, ^, ^M, $, $M                   */ \
2105    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
2106    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Chari  - the minimum length            */ \
2107    2,                             /* not                                    */ \    2,                             /* not                                    */ \
2108    /* Positive single-char repeats                            ** These are  */ \    2,                             /* noti                                   */ \
2109    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??      ** minima in  */ \    /* Positive single-char repeats                             ** These are */ \
2110    4, 4, 4,                       /* upto, minupto, exact     ** UTF-8 mode */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2111      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2112      2+IMM2_SIZE,                   /* exact                                  */ \
2113      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2114      2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2115      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2116      2+IMM2_SIZE,                   /* exact I                                */ \
2117      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2118    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2119    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2120    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2121      2+IMM2_SIZE,                   /* NOT exact                              */ \
2122      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2123      2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2124      2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2125      2+IMM2_SIZE,                   /* NOT exact I                            */ \
2126      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2127    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2128    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2129    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2130      2+IMM2_SIZE,                   /* Type exact                             */ \
2131      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2132    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2133    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2134    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2135   33,                             /* CLASS                                  */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2136   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2137    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2138    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2139      1+IMM2_SIZE,                   /* REFI                                   */ \
2140    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2141    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2142    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
2143    1+LINK_SIZE,                   /* Ket                                    */ \    1+LINK_SIZE,                   /* Ket                                    */ \
2144    1+LINK_SIZE,                   /* KetRmax                                */ \    1+LINK_SIZE,                   /* KetRmax                                */ \
2145    1+LINK_SIZE,                   /* KetRmin                                */ \    1+LINK_SIZE,                   /* KetRmin                                */ \
2146      1+LINK_SIZE,                   /* KetRpos                                */ \
2147      1+LINK_SIZE,                   /* Reverse                                */ \
2148    1+LINK_SIZE,                   /* Assert                                 */ \    1+LINK_SIZE,                   /* Assert                                 */ \
2149    1+LINK_SIZE,                   /* Assert not                             */ \    1+LINK_SIZE,                   /* Assert not                             */ \
2150    1+LINK_SIZE,                   /* Assert behind                          */ \    1+LINK_SIZE,                   /* Assert behind                          */ \
2151    1+LINK_SIZE,                   /* Assert behind not                      */ \    1+LINK_SIZE,                   /* Assert behind not                      */ \
2152    1+LINK_SIZE,                   /* Reverse                                */ \    1+LINK_SIZE,                   /* ONCE                                   */ \
2153    1+LINK_SIZE,                   /* Once                                   */ \    1+LINK_SIZE,                   /* ONCE_NC                                */ \
2154      1+LINK_SIZE,                   /* BRA                                    */ \
2155      1+LINK_SIZE,                   /* BRAPOS                                 */ \
2156      1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2157      1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2158    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2159    3,                             /* CREF                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2160    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2161    3,                             /* BRANUMBER                              */ \    1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2162    1+LINK_SIZE                    /* BRA                                    */ \    1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2163      1+LINK_SIZE,                   /* SCOND                                  */ \
2164      1+IMM2_SIZE, 1+IMM2_SIZE,      /* CREF, NCREF                            */ \
2165  /* A magic value for OP_CREF to indicate the "in recursion" condition. */    1+IMM2_SIZE, 1+IMM2_SIZE,      /* RREF, NRREF                            */ \
2166      1,                             /* DEF                                    */ \
2167  #define CREF_RECURSE  0xffff    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2168      3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2169  /* Error code numbers. They are given names so that they can more easily be    1, 3,                          /* SKIP, SKIP_ARG                         */ \
2170  tracked. */    1, 3,                          /* THEN, THEN_ARG                         */ \
2171      1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2172      1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2173    
2174    /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
2175    condition. */
2176    
2177    #define RREF_ANY  0xffff
2178    
2179    /* Compile time error code numbers. They are given names so that they can more
2180    easily be tracked. When a new number is added, the table called eint in
2181    pcreposix.c must be updated. */
2182    
2183  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,
2184         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
2185         ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,         ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
2186         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2187         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47 };         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2188           ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2189           ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2190           ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2191    
2192    /* JIT compiling modes. The function list is indexed by them. */
2193    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2194           JIT_NUMBER_OF_COMPILE_MODES };
2195    
2196  /* 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
2197  code vector run on as long as necessary after the end. We store an explicit  code vector run on as long as necessary after the end. We store an explicit
# Line 726  Because people can now save and re-use c Line 2207  Because people can now save and re-use c
2207  structure should be made at the end, and something earlier (e.g. a new  structure should be made at the end, and something earlier (e.g. a new
2208  flag in the options or one of the dummy fields) should indicate that the new  flag in the options or one of the dummy fields) should indicate that the new
2209  fields are present. Currently PCRE always sets the dummy fields to zero.  fields are present. Currently PCRE always sets the dummy fields to zero.
2210  NOTE NOTE NOTE:  NOTE NOTE NOTE
2211  */  */
2212    
2213  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2214    #define REAL_PCRE real_pcre
2215    #elif defined COMPILE_PCRE16
2216    #define REAL_PCRE real_pcre16
2217    #elif defined COMPILE_PCRE32
2218    #define REAL_PCRE real_pcre32
2219    #endif
2220    
2221    /* It is necessary to fork the struct for 32 bit, since it needs to use
2222     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2223     * typedef since pcretest needs access to  the struct of the 8-, 16-
2224     * and 32-bit variants. */
2225    
2226    typedef struct real_pcre8_or_16 {
2227    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2228    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2229    pcre_uint32 options;    pcre_uint32 options;            /* Public options */
2230    pcre_uint32 dummy1;             /* For future use, maybe */    pcre_uint16 flags;              /* Private flags */
2231      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2232    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2233    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2234    pcre_uint16 first_byte;    pcre_uint16 first_char;         /* Starting character */
2235    pcre_uint16 req_byte;    pcre_uint16 req_char;           /* This character must be seen */
2236    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2237    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2238    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2239    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2240      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2241      const pcre_uint8 *nullpad;      /* NULL padding */
2242    } real_pcre8_or_16;
2243    
2244    const unsigned char *tables;    /* Pointer to tables or NULL for std */  typedef struct real_pcre8_or_16 real_pcre;
2245    const unsigned char *nullpad;   /* NULL padding */  typedef struct real_pcre8_or_16 real_pcre16;
2246  } real_pcre;  
2247    typedef struct real_pcre32 {
2248      pcre_uint32 magic_number;
2249      pcre_uint32 size;               /* Total that was malloced */
2250      pcre_uint32 options;            /* Public options */
2251      pcre_uint16 flags;              /* Private flags */
2252      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2253      pcre_uint16 top_bracket;        /* Highest numbered group */
2254      pcre_uint16 top_backref;        /* Highest numbered back reference */
2255      pcre_uint32 first_char;         /* Starting character */
2256      pcre_uint32 req_char;           /* This character must be seen */
2257      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2258      pcre_uint16 name_entry_size;    /* Size of any name items */
2259      pcre_uint16 name_count;         /* Number of name items */
2260      pcre_uint16 ref_count;          /* Reference count */
2261      pcre_uint16 dummy1;             /* for later expansion */
2262      pcre_uint16 dummy2;             /* for later expansion */
2263      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2264      void *nullpad;                  /* for later expansion */
2265    } real_pcre32;
2266    
2267    /* Assert that the size of REAL_PCRE is divisible by 8 */
2268    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2269    
2270    /* Needed in pcretest to access some fields in the real_pcre* structures
2271     * directly. They're unified for 8/16/32 bits since the structs only differ
2272     * after these fields; if that ever changes, need to fork those defines into
2273     * 8/16 and 32 bit versions. */
2274    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2275    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2276    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2277    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2278    
2279  /* 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
2280  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
2281    
2282  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2283    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2284    pcre_uint32 options;    pcre_uint32 flags;              /* Private flags */
2285    uschar start_bits[32];    pcre_uint8 start_bits[32];      /* Starting char bits */
2286      pcre_uint32 minlength;          /* Minimum subject length */
2287  } pcre_study_data;  } pcre_study_data;
2288    
2289    /* Structure for building a chain of open capturing subpatterns during
2290    compiling, so that instructions to close them can be compiled when (*ACCEPT) is
2291    encountered. This is also used to identify subpatterns that contain recursive
2292    back references to themselves, so that they can be made atomic. */
2293    
2294    typedef struct open_capitem {
2295      struct open_capitem *next;    /* Chain link */
2296      pcre_uint16 number;           /* Capture number */
2297      pcre_uint16 flag;             /* Set TRUE if recursive back ref */
2298    } open_capitem;
2299    
2300  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2301  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2302    
2303  typedef struct compile_data {  typedef struct compile_data {
2304    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2305    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2306    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2307    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2308    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_workspace;/* The start of working space */
2309    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_code;     /* The start of the compiled code */
2310    uschar *name_table;           /* The name/number table */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2311    int  names_found;             /* Number of entries so far */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2312    int  name_entry_size;         /* Size of each entry */    open_capitem *open_caps;          /* Chain of open capture items */
2313    int  top_backref;             /* Maximum back reference */    pcre_uchar *hwm;                  /* High watermark of workspace */
2314    unsigned int backref_map;     /* Bitmap of low back refs */    pcre_uchar *name_table;           /* The name/number table */
2315    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  names_found;                 /* Number of entries so far */
2316    BOOL nopartial;               /* Set TRUE if partial won't work */    int  name_entry_size;             /* Size of each entry */
2317      int  workspace_size;              /* Size of workspace */
2318      int  bracount;                    /* Count of capturing parens as we compile */
2319      int  final_bracount;              /* Saved value after first pass */
2320      int  max_lookbehind;              /* Maximum lookbehind (characters) */
2321      int  top_backref;                 /* Maximum back reference */
2322      unsigned int backref_map;         /* Bitmap of low back refs */
2323      int  assert_depth;                /* Depth of nested assertions */
2324      int  external_options;            /* External (initial) options */
2325      int  external_flags;              /* External flag bits to be set */
2326      int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2327      BOOL had_accept;                  /* (*ACCEPT) encountered */
2328      BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2329      BOOL check_lookbehind;            /* Lookbehinds need later checking */
2330      int  nltype;                      /* Newline type */
2331      int  nllen;                       /* Newline string length */
2332      pcre_uchar nl[4];                 /* Newline string when fixed length */
2333  } compile_data;  } compile_data;
2334    
2335  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
2336  branches, for testing for left recursion. */  branches, for testing for left recursion while compiling. */
2337    
2338  typedef struct branch_chain {  typedef struct branch_chain {
2339    struct branch_chain *outer;    struct branch_chain *outer;
2340    uschar *current;    pcre_uchar *current_branch;
2341  } branch_chain;  } branch_chain;
2342    
2343  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
2344  call within the pattern. */  call within the pattern; used by pcre_exec(). */
2345    
2346  typedef struct recursion_info {  typedef struct recursion_info {
2347    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2348    int group_num;                /* Number of group that was called */    int group_num;                  /* Number of group that was called */
2349    const uschar *after_call;     /* "Return value": points after the call in the expr */    int *offset_save;               /* Pointer to start of saved offsets */
2350    USPTR save_start;             /* Old value of md->start_match */    int saved_max;                  /* Number of saved offsets */
2351    int *offset_save;             /* Pointer to start of saved offsets */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
   int saved_max;                /* Number of saved offsets */  
2352  } recursion_info;  } recursion_info;
2353    
2354  /* When compiling in a mode that doesn't use recursive calls to match(),  /* A similar structure for pcre_dfa_exec(). */
2355  a structure is used to remember local variables on the heap. It is defined in  
2356  pcre.c, close to the match() function, so that it is easy to keep it in step  typedef struct dfa_recursion_info {
2357  with any changes of local variable. However, the pointer to the current frame    struct dfa_recursion_info *prevrec;
2358  must be saved in some "static" place over a longjmp(). We declare the    int group_num;
2359  structure here so that we can put a pointer in the match_data structure.    PCRE_PUCHAR subject_position;
2360  NOTE: This isn't used for a "normal" compilation of pcre. */  } dfa_recursion_info;
2361    
2362    /* Structure for building a chain of data for holding the values of the subject
2363    pointer at the start of each subpattern, so as to detect when an empty string
2364    has been matched by a subpattern - to break infinite loops; used by
2365    pcre_exec(). */
2366    
2367    typedef struct eptrblock {
2368      struct eptrblock *epb_prev;
2369      PCRE_PUCHAR epb_saved_eptr;
2370    } eptrblock;
2371    
 struct heapframe;  
2372    
2373  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2374  doing traditional NFA matching, so that they are thread-safe. */  doing traditional NFA matching, so that they are thread-safe. */
# Line 813  typedef struct match_data { Line 2377  typedef struct match_data {
2377    unsigned long int match_call_count;      /* As it says */    unsigned long int match_call_count;      /* As it says */
2378    unsigned long int match_limit;           /* As it says */    unsigned long int match_limit;           /* As it says */
2379    unsigned long int match_limit_recursion; /* As it says */    unsigned long int match_limit_recursion; /* As it says */
2380    int   *offset_vector;         /* Offset vector */    int   *offset_vector;           /* Offset vector */
2381    int    offset_end;            /* One past the end */    int    offset_end;              /* One past the end */
2382    int    offset_max;            /* The maximum usable for return data */    int    offset_max;              /* The maximum usable for return data */
2383    const uschar *lcc;            /* Points to lower casing table */    int    nltype;                  /* Newline type */
2384    const uschar *ctypes;         /* Points to table of type maps */    int    nllen;                   /* Newline string length */
2385    BOOL   offset_overflow;       /* Set if too many extractions */    int    name_count;              /* Number of names in name table */
2386    BOOL   notbol;                /* NOTBOL flag */    int    name_entry_size;         /* Size of entry in names table */
2387    BOOL   noteol;                /* NOTEOL flag */    pcre_uchar *name_table;         /* Table of names */
2388    BOOL   utf8;                  /* UTF8 flag */    pcre_uchar nl[4];               /* Newline string when fixed */
2389    BOOL   endonly;               /* Dollar not before final \n */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2390    BOOL   notempty;              /* Empty string match not wanted */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2391    BOOL   partial;               /* PARTIAL flag */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2392    BOOL   hitend;                /* Hit the end of the subject at some point */    BOOL   offset_overflow;         /* Set if too many extractions */
2393    const uschar *start_code;     /* For use when recursing */    BOOL   notbol;                  /* NOTBOL flag */
2394    USPTR  start_subject;         /* Start of the subject string */    BOOL   noteol;                  /* NOTEOL flag */
2395    USPTR  end_subject;           /* End of the subject string */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2396    USPTR  start_match;           /* Start of this match attempt */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2397    USPTR  end_match_ptr;         /* Subject position at end match */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2398    int    end_offset_top;        /* Highwater mark at end of match */    BOOL   endonly;                 /* Dollar not before final \n */
2399    int    capture_last;          /* Most recent capture number */    BOOL   notempty;                /* Empty string match not wanted */
2400    int    start_offset;          /* The start offset value */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2401    recursion_info *recursive;    /* Linked list of recursion data */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2402    void  *callout_data;          /* To pass back to callouts */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2403    struct heapframe *thisframe;  /* Used only when compiling for no recursion */    BOOL   hasthen;                 /* Pattern contains (*THEN) */
2404      BOOL   ignore_skip_arg;         /* For re-run when SKIP name not found */
2405      const  pcre_uchar *start_code;  /* For use when recursing */
2406      PCRE_PUCHAR start_subject;      /* Start of the subject string */
2407      PCRE_PUCHAR end_subject;        /* End of the subject string */
2408      PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2409      PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2410      PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2411      int    partial;                 /* PARTIAL options */
2412      int    end_offset_top;          /* Highwater mark at end of match */
2413      int    capture_last;            /* Most recent capture number */
2414      int    start_offset;            /* The start offset value */
2415      int    match_function_type;     /* Set for certain special calls of MATCH() */
2416      eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2417      int    eptrn;                   /* Next free eptrblock */
2418      recursion_info *recursive;      /* Linked list of recursion data */
2419      void  *callout_data;            /* To pass back to callouts */
2420      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2421      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2422      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2423    #ifdef NO_RECURSE
2424      void  *match_frames_base;       /* For remembering malloc'd frames */
2425    #endif
2426  } match_data;  } match_data;
2427    
2428  /* 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
2429  functions. */  functions. */
2430    
2431  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2432    const uschar *start_code;     /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2433    const uschar *start_subject;  /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2434    const uschar *end_subject;    /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2435    const uschar *tables;         /* Character tables */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2436    int   moptions;               /* Match options */    const pcre_uint8 *tables;         /* Character tables */
2437    int   poptions;               /* Pattern options */    int   start_offset;               /* The start offset value */
2438    void  *callout_data;          /* To pass back to callouts */    int   moptions;                   /* Match options */
2439      int   poptions;                   /* Pattern options */
2440      int   nltype;                     /* Newline type */
2441      int   nllen;                      /* Newline string length */
2442      pcre_uchar nl[4];                 /* Newline string when fixed */
2443      void *callout_data;               /* To pass back to callouts */
2444      dfa_recursion_info *recursive;    /* Linked list of recursion data */
2445  } dfa_match_data;  } dfa_match_data;
2446    
2447  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 858  typedef struct dfa_match_data { Line 2450  typedef struct dfa_match_data {
2450  #define ctype_letter  0x02  #define ctype_letter  0x02
2451  #define ctype_digit   0x04  #define ctype_digit   0x04
2452  #define ctype_xdigit  0x08  #define ctype_xdigit  0x08
2453  #define ctype_word    0x10   /* alphameric or '_' */  #define ctype_word    0x10   /* alphanumeric or '_' */
2454  #define ctype_meta    0x80   /* regexp meta char or zero (end pattern) */  #define ctype_meta    0x80   /* regexp meta char or zero (end pattern) */
2455    
2456  /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set  /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
# Line 885  total length. */ Line 2477  total length. */
2477  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2478  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2479    
2480    /* Internal function and data prefixes. */
2481    
2482    #if defined COMPILE_PCRE8
2483    #ifndef PUBL
2484    #define PUBL(name) pcre_##name
2485    #endif
2486    #ifndef PRIV
2487    #define PRIV(name) _pcre_##name
2488    #endif
2489    #elif defined COMPILE_PCRE16
2490    #ifndef PUBL
2491    #define PUBL(name) pcre16_##name
2492    #endif
2493    #ifndef PRIV
2494    #define PRIV(name) _pcre16_##name
2495    #endif
2496    #elif defined COMPILE_PCRE32
2497    #ifndef PUBL
2498    #define PUBL(name) pcre32_##name
2499    #endif
2500    #ifndef PRIV
2501    #define PRIV(name) _pcre32_##name
2502    #endif
2503    #else
2504    #error Unsupported compiling mode
2505    #endif /* COMPILE_PCRE[8|16|32] */
2506    
2507  /* 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
2508  codes. */  codes. Each entry used to point directly to a name, but to reduce the number of
2509    relocations in shared libraries, it now has an offset into a single string
2510    instead. */
2511    
2512  typedef struct {  typedef struct {
2513    const char *name;    pcre_uint16 name_offset;
2514    pcre_uint16 type;    pcre_uint16 type;
2515    pcre_uint16 value;    pcre_uint16 value;
2516  } ucp_type_table;  } ucp_type_table;
# Line 900  of the exported public functions. They h Line 2521  of the exported public functions. They h
2521  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
2522  pcre_tables.c module. */  pcre_tables.c module. */
2523    
2524  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2525  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2526  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2527  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2528    extern const int            PRIV(utf8_table3)[];
2529  extern const int    _pcre_utf8_table1_size;  extern const pcre_uint8     PRIV(utf8_table4)[];
2530    #endif /* COMPILE_PCRE8 */
2531    
2532    extern const char           PRIV(utt_names)[];
2533    extern const ucp_type_table PRIV(utt)[];
2534    extern const int            PRIV(utt_size);
2535    
2536  extern const ucp_type_table _pcre_utt[];  extern const pcre_uint8     PRIV(OP_lengths)[];
2537  extern const int _pcre_utt_size;  extern const pcre_uint8     PRIV(default_tables)[];
2538    
2539  extern const uschar _pcre_default_tables[];  extern const pcre_uint32    PRIV(hspace_list)[];
2540    extern const pcre_uint32    PRIV(vspace_list)[];
 extern const uschar _pcre_OP_lengths[];  
2541    
2542    
2543  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2544  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
2545  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2546    
2547  extern int         _pcre_ord2utf8(int, uschar *);  /* String comparison functions. */
2548  extern real_pcre * _pcre_try_flipped(const real_pcre *, real_pcre *,  #if defined COMPILE_PCRE8
2549                       const pcre_study_data *, pcre_study_data *);  
2550  extern int         _pcre_ucp_findprop(const int, int *, int *);  #define STRCMP_UC_UC(str1, str2) \
2551  extern int         _pcre_ucp_othercase(const int);    strcmp((char *)(str1), (char *)(str2))
2552  extern int         _pcre_valid_utf8(const uschar *, int);  #define STRCMP_UC_C8(str1, str2) \
2553  extern BOOL        _pcre_xclass(int, const uschar *);    strcmp((char *)(str1), (str2))
2554    #define STRNCMP_UC_UC(str1, str2, num) \
2555      strncmp((char *)(str1), (char *)(str2), (num))
2556    #define STRNCMP_UC_C8(str1, str2, num) \
2557      strncmp((char *)(str1), (str2), (num))
2558    #define STRLEN_UC(str) strlen((const char *)str)
2559    
2560    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2561    
2562    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2563                               const pcre_uchar *);
2564    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2565                               const char *);
2566    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2567                               const pcre_uchar *, unsigned int num);
2568    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2569                               const char *, unsigned int num);
2570    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2571    
2572    #define STRCMP_UC_UC(str1, str2) \
2573      PRIV(strcmp_uc_uc)((str1), (str2))
2574    #define STRCMP_UC_C8(str1, str2) \
2575      PRIV(strcmp_uc_c8)((str1), (str2))
2576    #define STRNCMP_UC_UC(str1, str2, num) \
2577      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2578    #define STRNCMP_UC_C8(str1, str2, num) \
2579      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2580    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2581    
2582    #endif /* COMPILE_PCRE[8|16|32] */
2583    
2584    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2585    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2586                               int *, BOOL);
2587    extern int               PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2588    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2589    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2590                               int *, BOOL);
2591    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2592    
2593    #ifdef SUPPORT_JIT
2594    extern void              PRIV(jit_compile)(const REAL_PCRE *,
2595                               PUBL(extra) *, int);
2596    extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2597                               const pcre_uchar *, int, int, int, int *, int);
2598    extern void              PRIV(jit_free)(void *);
2599    extern int               PRIV(jit_get_size)(void *);
2600    extern const char*       PRIV(jit_get_target)(void);
2601    #endif
2602    
2603    /* Unicode character database (UCD) */
2604    
2605    typedef struct {
2606      pcre_uint8 script;     /* ucp_Arabic, etc. */
2607      pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2608      pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2609      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2610      pcre_int32 other_case; /* offset to other case, or zero if none */
2611    } ucd_record;
2612    
2613    extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2614    extern const ucd_record  PRIV(ucd_records)[];
2615    extern const pcre_uint8  PRIV(ucd_stage1)[];
2616    extern const pcre_uint16 PRIV(ucd_stage2)[];
2617    extern const int         PRIV(ucp_gentype)[];
2618    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2619    #ifdef SUPPORT_JIT
2620    extern const int         PRIV(ucp_typerange)[];
2621    #endif
2622    
2623    #ifdef SUPPORT_UCP
2624    /* UCD access macros */
2625    
2626    #define UCD_BLOCK_SIZE 128
2627    #define GET_UCD(ch) (PRIV(ucd_records) + \
2628            PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2629            UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2630    
2631    #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2632    #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2633    #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2634    #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2635    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2636    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2637    
2638    #endif /* SUPPORT_UCP */
2639    
2640  #endif  #endif
2641    

Legend:
Removed from v.87  
changed lines
  Added in v.1090

  ViewVC Help
Powered by ViewVC 1.1.5