/[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 77 by nigel, Sat Feb 24 21:40:45 2007 UTC revision 1083 by chpe, Tue Oct 16 15:55:24 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) 1997-2005 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
47    #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
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
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    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  #ifdef DEBUG  #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  #endif
166    
167  /* We need to have types that specify unsigned 16-bit and 32-bit integers. We  /* 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
180    
181    /* 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  /* Include the public PCRE header */  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  #include "pcre.h"  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
327    with a custom type. This makes it possible, for example, to allow pcre_exec()
328    to process subject strings that are discontinuous by using a smart pointer
329    class. It must always be possible to inspect all of the subject string in
330    pcre_exec() because of the way it backtracks. Two macros are required in the
331    normal case, for sign-unspecified and unsigned char pointers. The former is
332    used for the external interface and appears in pcre.h, which is why its name
333    must begin with PCRE_. */
334    
335    #ifdef CUSTOM_SUBJECT_PTR
336    #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
337    #else
338    #define PCRE_PUCHAR const pcre_uchar *
339    #endif
340    
341  /* Include the (copy of) the public ucp header, changing the external name into  /* Include the public PCRE header and the definitions of UCP character property
342  a private one. This does no harm, even if we aren't compiling UCP support. */  values. */
343    
344  #define ucp_findchar _pcre_ucp_findchar  #include "pcre.h"
345  #include "ucp.h"  #include "ucp.h"
346    
347  /* When compiling for use with the Virtual Pascal compiler, these functions  /* When compiling for use with the Virtual Pascal compiler, these functions
# Line 127  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 136  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  int 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      {
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 172  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 208  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 225  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 232  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  #ifndef SUPPORT_UTF8  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
554    
555    /* The maximum length of a MARK name is currently one data unit; it may be
556    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
557    
558    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
559    #define MAX_MARK ((1u << 16) - 1)
560    #else
561    #define MAX_MARK ((1u << 8) - 1)
562    #endif
563    
564    /* When UTF encoding is being used, a character is no longer just a single
565    byte. The macros for character handling generate simple sequences when used in
566    character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
567    and other macros are not used when UTF is not supported, so they are not
568    defined. To make sure they can never even appear when UTF support is omitted,
569    we don't even define them. */
570    
571    #ifndef SUPPORT_UTF
572    
573    /* #define MAX_VALUE_FOR_SINGLE_CHAR */
574    /* #define HAS_EXTRALEN(c) */
575    /* #define GET_EXTRALEN(c) */
576    /* #define NOT_FIRSTCHAR(c) */
577  #define GETCHAR(c, eptr) c = *eptr;  #define GETCHAR(c, eptr) c = *eptr;
578  #define GETCHARTEST(c, eptr) c = *eptr;  #define GETCHARTEST(c, eptr) c = *eptr;
579  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
580  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
581  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
582  #define 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_UTF */
588    
589    #if defined COMPILE_PCRE8
590    
591    /* These macros were originally written in the form of loops that used data
592    from the tables whose names start with PRIV(utf8_table). They were rewritten by
593    a user so as not to use loops, because in some environments this gives a
594    significant performance advantage, and it seems never to do any harm. */
595    
596  #else   /* SUPPORT_UTF8 */  /* Tells the biggest code point which can be encoded as a single character. */
597    
598    #define MAX_VALUE_FOR_SINGLE_CHAR 127
599    
600    /* Tests whether the code point needs extra characters to decode. */
601    
602    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
603    
604    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
605    Otherwise it has an undefined behaviour. */
606    
607    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
608    
609    /* Returns TRUE, if the given character is not the first character
610    of a UTF sequence. */
611    
612    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
613    
614    /* Base macro to pick up the remaining bytes of a UTF-8 character, not
615    advancing the pointer. */
616    
617    #define GETUTF8(c, eptr) \
618        { \
619        if ((c & 0x20) == 0) \
620          c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
621        else if ((c & 0x10) == 0) \
622          c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
623        else if ((c & 0x08) == 0) \
624          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
625          ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
626        else if ((c & 0x04) == 0) \
627          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
628              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
629              (eptr[4] & 0x3f); \
630        else \
631          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
632              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
633              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
634        }
635    
636  /* Get the next UTF-8 character, not advancing the pointer. This is called when  /* Get the next UTF-8 character, not advancing the pointer. This is called when
637  we know we are in UTF-8 mode. */  we know we are in UTF-8 mode. */
638    
639  #define GETCHAR(c, eptr) \  #define GETCHAR(c, eptr) \
640    c = *eptr; \    c = *eptr; \
641    if ((c & 0xc0) == 0xc0) \    if (c >= 0xc0) GETUTF8(c, eptr);
     { \  
     int gcii; \  
     int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     for (gcii = 1; gcii <= gcaa; gcii++) \  
       { \  
       gcss -= 6; \  
       c |= (eptr[gcii] & 0x3f) << gcss; \  
       } \  
     }  
642    
643  /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the  /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
644  pointer. */  pointer. */
645    
646  #define GETCHARTEST(c, eptr) \  #define GETCHARTEST(c, eptr) \
647    c = *eptr; \    c = *eptr; \
648    if (utf8 && (c & 0xc0) == 0xc0) \    if (utf && c >= 0xc0) GETUTF8(c, eptr);
649    
650    /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
651    the pointer. */
652    
653    #define GETUTF8INC(c, eptr) \
654      { \      { \
655      int gcii; \      if ((c & 0x20) == 0) \
656      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \        c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
657      int gcss = 6*gcaa; \      else if ((c & 0x10) == 0) \
658      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        { \
659      for (gcii = 1; gcii <= gcaa; gcii++) \        c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
660          eptr += 2; \
661          } \
662        else if ((c & 0x08) == 0) \
663          { \
664          c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
665              ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
666          eptr += 3; \
667          } \
668        else if ((c & 0x04) == 0) \
669          { \
670          c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
671              ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
672              (eptr[3] & 0x3f); \
673          eptr += 4; \
674          } \
675        else \
676        { \        { \
677        gcss -= 6; \        c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
678        c |= (eptr[gcii] & 0x3f) << gcss; \            ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
679              ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
680          eptr += 5; \
681        } \        } \
682      }      }
683    
# Line 290  know we are in UTF-8 mode. */ Line 686  know we are in UTF-8 mode. */
686    
687  #define GETCHARINC(c, eptr) \  #define GETCHARINC(c, eptr) \
688    c = *eptr++; \    c = *eptr++; \
689    if ((c & 0xc0) == 0xc0) \    if (c >= 0xc0) GETUTF8INC(c, eptr);
     { \  
     int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     while (gcaa-- > 0) \  
       { \  
       gcss -= 6; \  
       c |= (*eptr++ & 0x3f) << gcss; \  
       } \  
     }  
690    
691  /* Get the next character, testing for UTF-8 mode, and advancing the pointer */  /* Get the next character, testing for UTF-8 mode, and advancing the pointer.
692    This is called when we don't know if we are in UTF-8 mode. */
693    
694  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
695    c = *eptr++; \    c = *eptr++; \
696    if (utf8 && (c & 0xc0) == 0xc0) \    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
697    
698    /* Base macro to pick up the remaining bytes of a UTF-8 character, not
699    advancing the pointer, incrementing the length. */
700    
701    #define GETUTF8LEN(c, eptr, len) \
702      { \      { \
703      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \      if ((c & 0x20) == 0) \
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     while (gcaa-- > 0) \  
704        { \        { \
705        gcss -= 6; \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
706        c |= (*eptr++ & 0x3f) << gcss; \        len++; \
707          } \
708        else if ((c & 0x10)  == 0) \
709          { \
710          c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
711          len += 2; \
712          } \
713        else if ((c & 0x08)  == 0) \
714          {\
715          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
716              ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
717          len += 3; \
718          } \
719        else if ((c & 0x04)  == 0) \
720          { \
721          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
722              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
723              (eptr[4] & 0x3f); \
724          len += 4; \
725          } \
726        else \
727          {\
728          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
729              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
730              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
731          len += 5; \
732        } \        } \
733      }      }
734    
# Line 323  if there are extra bytes. This is called Line 737  if there are extra bytes. This is called
737    
738  #define GETCHARLEN(c, eptr, len) \  #define GETCHARLEN(c, eptr, len) \
739    c = *eptr; \    c = *eptr; \
740    if ((c & 0xc0) == 0xc0) \    if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
741      { \  
742      int gcii; \  /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
743      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  pointer, incrementing length if there are extra bytes. This is called when we
744      int gcss = 6*gcaa; \  do not know if we are in UTF-8 mode. */
745      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
746      for (gcii = 1; gcii <= gcaa; gcii++) \  #define GETCHARLENTEST(c, eptr, len) \
747        { \    c = *eptr; \
748        gcss -= 6; \    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
       c |= (eptr[gcii] & 0x3f) << gcss; \  
       } \  
     len += gcaa; \  
     }  
749    
750  /* If the pointer is not at the start of a character, move it back until  /* If the pointer is not at the start of a character, move it back until
751  it is. 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
752    because almost all calls are already within a block of UTF-8 only code. */
753    
754  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--;  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
755    
756  #endif  /* Same as above, just in the other direction. */
757    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
758    
759    /* Same as above, but it allows a fully customizable form. */
760    #define ACROSSCHAR(condition, eptr, action) \
761      while((condition) && ((eptr) & 0xc0) == 0x80) action
762    
763  /* In case there is no definition of offsetof() provided - though any proper  #elif defined COMPILE_PCRE16
 Standard C system should have one. */  
764    
765  #ifndef offsetof  /* Tells the biggest code point which can be encoded as a single character. */
766  #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
767  #endif  #define MAX_VALUE_FOR_SINGLE_CHAR 65535
768    
769    /* Tests whether the code point needs extra characters to decode. */
770    
771    #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
772    
773    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
774    Otherwise it has an undefined behaviour. */
775    
776    #define GET_EXTRALEN(c) 1
777    
778    /* Returns TRUE, if the given character is not the first character
779    of a UTF sequence. */
780    
781    #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
782    
783    /* Base macro to pick up the low surrogate of a UTF-16 character, not
784    advancing the pointer. */
785    
786    #define GETUTF16(c, eptr) \
787       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
788    
789    /* Get the next UTF-16 character, not advancing the pointer. This is called when
790    we know we are in UTF-16 mode. */
791    
792    #define GETCHAR(c, eptr) \
793      c = *eptr; \
794      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
795    
796    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
797    pointer. */
798    
799    #define GETCHARTEST(c, eptr) \
800      c = *eptr; \
801      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
802    
803    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
804    the pointer. */
805    
806    #define GETUTF16INC(c, eptr) \
807       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
808    
809    /* Get the next UTF-16 character, advancing the pointer. This is called when we
810    know we are in UTF-16 mode. */
811    
812    #define GETCHARINC(c, eptr) \
813      c = *eptr++; \
814      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
815    
816    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
817    This is called when we don't know if we are in UTF-16 mode. */
818    
819    #define GETCHARINCTEST(c, eptr) \
820      c = *eptr++; \
821      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
822    
823    /* Base macro to pick up the low surrogate of a UTF-16 character, not
824    advancing the pointer, incrementing the length. */
825    
826    #define GETUTF16LEN(c, eptr, len) \
827       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
828    
829    /* Get the next UTF-16 character, not advancing the pointer, incrementing
830    length if there is a low surrogate. This is called when we know we are in
831    UTF-16 mode. */
832    
833    #define GETCHARLEN(c, eptr, len) \
834      c = *eptr; \
835      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
836    
837    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
838    pointer, incrementing length if there is a low surrogate. This is called when
839    we do not know if we are in UTF-16 mode. */
840    
841    #define GETCHARLENTEST(c, eptr, len) \
842      c = *eptr; \
843      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
844    
845    /* If the pointer is not at the start of a character, move it back until
846    it is. This is called only in UTF-16 mode - we don't put a test within the
847    macro because almost all calls are already within a block of UTF-16 only
848    code. */
849    
850    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
851    
852    /* Same as above, just in the other direction. */
853    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
854    
855    /* Same as above, but it allows a fully customizable form. */
856    #define ACROSSCHAR(condition, eptr, action) \
857      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
858    
859    #elif defined COMPILE_PCRE32
860    
861    /* These are unnecessary for the 32-bit library */
862    #undef MAX_VALUE_FOR_SINGLE_CHAR
863    #undef HAS_EXTRALEN
864    #undef GET_EXTRALEN
865    #undef NOT_FIRSTCHAR
866    
867    #define UTF32_MASK (0x1ffffful)
868    
869    /* Get the next UTF-32 character, not advancing the pointer. This is called when
870    we know we are in UTF-32 mode. */
871    
872    #define GETCHAR(c, eptr) \
873      c = (*eptr) & UTF32_MASK;
874    
875    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
876    pointer. */
877    
878    #define GETCHARTEST(c, eptr) \
879      c = *eptr; \
880      if (utf) c &= UTF32_MASK;
881    
882    /* Get the next UTF-32 character, advancing the pointer. This is called when we
883    know we are in UTF-32 mode. */
884    
885    #define GETCHARINC(c, eptr) \
886      c = (*eptr++) & UTF32_MASK;
887    
888    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
889    This is called when we don't know if we are in UTF-32 mode. */
890    
891    #define GETCHARINCTEST(c, eptr) \
892      c = *eptr++; \
893      if (utf) c &= UTF32_MASK;
894    
895    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
896    length (since all UTF-32 is of length 1). This is called when we know we are in
897    UTF-32 mode. */
898    
899    #define GETCHARLEN(c, eptr, len) \
900      GETCHAR(c, eptr)
901    
902    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
903    pointer, not incrementing the length (since all UTF-32 is of length 1).
904    This is called when we do not know if we are in UTF-32 mode. */
905    
906    #define GETCHARLENTEST(c, eptr, len) \
907      GETCHARTEST(c, eptr)
908    
909  /* These are the public options that can change during matching. */  /* If the pointer is not at the start of a character, move it back until
910    it is. This is called only in UTF-32 mode - we don't put a test within the
911    macro because almost all calls are already within a block of UTF-32 only
912    code. */
913    
914    #define BACKCHAR(eptr) do { } while (0)
915    
916    /* Same as above, just in the other direction. */
917    #define FORWARDCHAR(eptr) do { } while (0)
918    
919    /* Same as above, but it allows a fully customizable form. */
920    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
921    
922    #else
923    #error Unsupported compiling mode
924    #endif /* COMPILE_PCRE[8|16|32] */
925    
926    #endif  /* SUPPORT_UTF */
927    
928    /* Tests for Unicode horizontal and vertical whitespace characters must check a
929    number of different values. Using a switch statement for this generates the
930    fastest code (no loop, no memory access), and there are several places in the
931    interpreter code where this happens. In order to ensure that all the case lists
932    remain in step, we use macros so that there is only one place where the lists
933    are defined.
934    
935    These values are also required as lists in pcre_compile.c when processing \h,
936    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
937    macros that define the values are here so that all the definitions are
938    together. The lists must be in ascending character order, terminated by
939    NOTACHAR (which is 0xffffffff).
940    
941    Any changes should ensure that the various macros are kept in step with each
942    other. NOTE: The values also appear in pcre_jit_compile.c. */
943    
944    /* ------ ASCII/Unicode environments ------ */
945    
946    #ifndef EBCDIC
947    
948    #define HSPACE_LIST \
949      CHAR_HT, CHAR_SPACE, 0xa0, \
950      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
951      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
952      NOTACHAR
953    
954    #define HSPACE_MULTIBYTE_CASES \
955      case 0x1680:  /* OGHAM SPACE MARK */ \
956      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
957      case 0x2000:  /* EN QUAD */ \
958      case 0x2001:  /* EM QUAD */ \
959      case 0x2002:  /* EN SPACE */ \
960      case 0x2003:  /* EM SPACE */ \
961      case 0x2004:  /* THREE-PER-EM SPACE */ \
962      case 0x2005:  /* FOUR-PER-EM SPACE */ \
963      case 0x2006:  /* SIX-PER-EM SPACE */ \
964      case 0x2007:  /* FIGURE SPACE */ \
965      case 0x2008:  /* PUNCTUATION SPACE */ \
966      case 0x2009:  /* THIN SPACE */ \
967      case 0x200A:  /* HAIR SPACE */ \
968      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
969      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
970      case 0x3000   /* IDEOGRAPHIC SPACE */
971    
972    #define HSPACE_BYTE_CASES \
973      case CHAR_HT: \
974      case CHAR_SPACE: \
975      case 0xa0     /* NBSP */
976    
977    #define HSPACE_CASES \
978      HSPACE_BYTE_CASES: \
979      HSPACE_MULTIBYTE_CASES
980    
981    #define VSPACE_LIST \
982      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
983    
984    #define VSPACE_MULTIBYTE_CASES \
985      case 0x2028:    /* LINE SEPARATOR */ \
986      case 0x2029     /* PARAGRAPH SEPARATOR */
987    
988    #define VSPACE_BYTE_CASES \
989      case CHAR_LF: \
990      case CHAR_VT: \
991      case CHAR_FF: \
992      case CHAR_CR: \
993      case CHAR_NEL
994    
995    #define VSPACE_CASES \
996      VSPACE_BYTE_CASES: \
997      VSPACE_MULTIBYTE_CASES
998    
999    /* ------ EBCDIC environments ------ */
1000    
1001  #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)  #else
1002    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1003    
1004    #define HSPACE_BYTE_CASES \
1005      case CHAR_HT: \
1006      case CHAR_SPACE
1007    
1008  /* Private options flags start at the most significant end of the four bytes,  #define HSPACE_CASES HSPACE_BYTE_CASES
 but skip the top bit so we can use ints for convenience without getting tangled  
 with negative values. The public options defined in pcre.h start at the least  
 significant end. Make sure they don't overlap! */  
1009    
1010  #define PCRE_FIRSTSET      0x40000000  /* first_byte is set */  #ifdef EBCDIC_NL25
1011  #define PCRE_REQCHSET      0x20000000  /* req_byte is set */  #define VSPACE_LIST \
1012  #define PCRE_STARTLINE     0x10000000  /* start after \n for multiline */    CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1013  #define PCRE_ICHANGED      0x08000000  /* i option changes within regex */  #else
1014  #define PCRE_NOPARTIAL     0x04000000  /* can't use partial with this regex */  #define VSPACE_LIST \
1015      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1016    #endif
1017    
1018    #define VSPACE_BYTE_CASES \
1019      case CHAR_LF: \
1020      case CHAR_VT: \
1021      case CHAR_FF: \
1022      case CHAR_CR: \
1023      case CHAR_NEL
1024    
1025    #define VSPACE_CASES VSPACE_BYTE_CASES
1026    #endif  /* EBCDIC */
1027    
1028    /* ------ End of whitespace macros ------ */
1029    
 /* Options for the "extra" block produced by pcre_study(). */  
1030    
1031  #define PCRE_STUDY_MAPPED   0x01     /* a map of starting chars exists */  
1032    /* Private flags containing information about the compiled regex. They used to
1033    live at the top end of the options word, but that got almost full, so now they
1034    are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
1035    the restrictions on partial matching have been lifted. It remains for backwards
1036    compatibility. */
1037    
1038    #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1039    #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1040    #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
1041    #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1042    #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1043    #define PCRE_REQCHSET      0x0040  /* req_byte is set */
1044    #define PCRE_RCH_CASELESS  0x0080  /* caseless requested char */
1045    #define PCRE_STARTLINE     0x0100  /* start after \n for multiline */
1046    #define PCRE_NOPARTIAL     0x0200  /* can't use partial with this regex */
1047    #define PCRE_JCHANGED      0x0400  /* j option used in regex */
1048    #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1049    #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1050    
1051    #if defined COMPILE_PCRE8
1052    #define PCRE_MODE          PCRE_MODE8
1053    #elif defined COMPILE_PCRE16
1054    #define PCRE_MODE          PCRE_MODE16
1055    #elif defined COMPILE_PCRE32
1056    #define PCRE_MODE          PCRE_MODE32
1057    #endif
1058    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1059    
1060    /* Flags for the "extra" block produced by pcre_study(). */
1061    
1062    #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
1063    #define PCRE_STUDY_MINLEN  0x0002  /* a minimum length field exists */
1064    
1065  /* Masks for identifying the public options that are permitted at compile  /* Masks for identifying the public options that are permitted at compile
1066  time, run time, or study time, respectively. */  time, run time, or study time, respectively. */
1067    
1068  #define PUBLIC_OPTIONS \  #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
1069                               PCRE_NEWLINE_ANYCRLF)
1070    
1071    #define PUBLIC_COMPILE_OPTIONS \
1072    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
1073     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1074     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| \
1075       PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1076       PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
1077    
1078  #define PUBLIC_EXEC_OPTIONS \  #define PUBLIC_EXEC_OPTIONS \
1079    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1080     PCRE_PARTIAL)     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
1081       PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
1082    
1083  #define PUBLIC_DFA_EXEC_OPTIONS \  #define PUBLIC_DFA_EXEC_OPTIONS \
1084    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1085     PCRE_PARTIAL|PCRE_DFA_SHORTEST|PCRE_DFA_RESTART)     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
1086       PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1087  #define PUBLIC_STUDY_OPTIONS 0   /* None defined */     PCRE_NO_START_OPTIMIZE)
1088    
1089    #define PUBLIC_STUDY_OPTIONS \
1090       (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1091        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1092    
1093  /* Magic number to provide a small check against being handed junk. Also used  /* Magic number to provide a small check against being handed junk. */
 to detect whether a pattern was compiled on a host of different endianness. */  
1094    
1095  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1096    
1097  /* Negative values for the firstchar and reqchar variables */  /* This variable is used to detect a loaded regular expression
1098    in different endianness. */
1099    
1100  #define REQ_UNSET (-2)  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
 #define REQ_NONE  (-1)  
1101    
1102  /* 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
1103  req_byte match. */  req_byte match. */
1104    
1105  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1106    
1107  /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1108  variable-length repeat, or a anything other than literal characters. */  environments where these macros are defined elsewhere. Unfortunately, there
1109    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 */  
1110    
1111  typedef int BOOL;  typedef int BOOL;
1112    
1113    #ifndef FALSE
1114  #define FALSE   0  #define FALSE   0
1115  #define TRUE    1  #define TRUE    1
1116    #endif
1117    
1118    /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
1119    character constants like '*' because the compiler would emit their EBCDIC code,
1120    which is different from their ASCII/UTF-8 code. Instead we define macros for
1121    the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
1122    is enabled. When UTF-8 support is not enabled, the definitions use character
1123    literals. Both character and string versions of each character are needed, and
1124    there are some longer strings as well.
1125    
1126    This means that, on EBCDIC platforms, the PCRE library can handle either
1127    EBCDIC, or UTF-8, but not both. To support both in the same compiled library
1128    would need different lookups depending on whether PCRE_UTF8 was set or not.
1129    This would make it impossible to use characters in switch/case statements,
1130    which would reduce performance. For a theoretical use (which nobody has asked
1131    for) in a minority area (EBCDIC platforms), this is not sensible. Any
1132    application that did need both could compile two versions of the library, using
1133    macros to give the functions distinct names. */
1134    
1135    #ifndef SUPPORT_UTF
1136    
1137    /* UTF-8 support is not enabled; use the platform-dependent character literals
1138    so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1139    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1140    characters, a common practice has been to use its NL (0x15) character as the
1141    line terminator in C-like processing environments. However, sometimes the LF
1142    (0x25) character is used instead, according to this Unicode document:
1143    
1144    http://unicode.org/standard/reports/tr13/tr13-5.html
1145    
1146    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1147    instead. Whichever is *not* chosen is defined as NEL.
1148    
1149    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1150    same code point. */
1151    
1152    #ifdef EBCDIC
1153    
1154    #ifndef EBCDIC_NL25
1155    #define CHAR_NL                     '\x15'
1156    #define CHAR_NEL                    '\x25'
1157    #define STR_NL                      "\x15"
1158    #define STR_NEL                     "\x25"
1159    #else
1160    #define CHAR_NL                     '\x25'
1161    #define CHAR_NEL                    '\x15'
1162    #define STR_NL                      "\x25"
1163    #define STR_NEL                     "\x15"
1164    #endif
1165    
1166    #define CHAR_LF                     CHAR_NL
1167    #define STR_LF                      STR_NL
1168    
1169    #define CHAR_ESC                    '\047'
1170    #define CHAR_DEL                    '\007'
1171    #define STR_ESC                     "\047"
1172    #define STR_DEL                     "\007"
1173    
1174    #else  /* Not EBCDIC */
1175    
1176    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1177    compatibility. NEL is the Unicode newline character; make sure it is
1178    a positive value. */
1179    
1180    #define CHAR_LF                     '\n'
1181    #define CHAR_NL                     CHAR_LF
1182    #define CHAR_NEL                    ((unsigned char)'\x85')
1183    #define CHAR_ESC                    '\033'
1184    #define CHAR_DEL                    '\177'
1185    
1186    #define STR_LF                      "\n"
1187    #define STR_NL                      STR_LF
1188    #define STR_NEL                     "\x85"
1189    #define STR_ESC                     "\033"
1190    #define STR_DEL                     "\177"
1191    
1192    #endif  /* EBCDIC */
1193    
1194    /* The remaining definitions work in both environments. */
1195    
1196    #define CHAR_HT                     '\t'
1197    #define CHAR_VT                     '\v'
1198    #define CHAR_FF                     '\f'
1199    #define CHAR_CR                     '\r'
1200    #define CHAR_BS                     '\b'
1201    #define CHAR_BEL                    '\a'
1202    
1203    #define CHAR_SPACE                  ' '
1204    #define CHAR_EXCLAMATION_MARK       '!'
1205    #define CHAR_QUOTATION_MARK         '"'
1206    #define CHAR_NUMBER_SIGN            '#'
1207    #define CHAR_DOLLAR_SIGN            '$'
1208    #define CHAR_PERCENT_SIGN           '%'
1209    #define CHAR_AMPERSAND              '&'
1210    #define CHAR_APOSTROPHE             '\''
1211    #define CHAR_LEFT_PARENTHESIS       '('
1212    #define CHAR_RIGHT_PARENTHESIS      ')'
1213    #define CHAR_ASTERISK               '*'
1214    #define CHAR_PLUS                   '+'
1215    #define CHAR_COMMA                  ','
1216    #define CHAR_MINUS                  '-'
1217    #define CHAR_DOT                    '.'
1218    #define CHAR_SLASH                  '/'
1219    #define CHAR_0                      '0'
1220    #define CHAR_1                      '1'
1221    #define CHAR_2                      '2'
1222    #define CHAR_3                      '3'
1223    #define CHAR_4                      '4'
1224    #define CHAR_5                      '5'
1225    #define CHAR_6                      '6'
1226    #define CHAR_7                      '7'
1227    #define CHAR_8                      '8'
1228    #define CHAR_9                      '9'
1229    #define CHAR_COLON                  ':'
1230    #define CHAR_SEMICOLON              ';'
1231    #define CHAR_LESS_THAN_SIGN         '<'
1232    #define CHAR_EQUALS_SIGN            '='
1233    #define CHAR_GREATER_THAN_SIGN      '>'
1234    #define CHAR_QUESTION_MARK          '?'
1235    #define CHAR_COMMERCIAL_AT          '@'
1236    #define CHAR_A                      'A'
1237    #define CHAR_B                      'B'
1238    #define CHAR_C                      'C'
1239    #define CHAR_D                      'D'
1240    #define CHAR_E                      'E'
1241    #define CHAR_F                      'F'
1242    #define CHAR_G                      'G'
1243    #define CHAR_H                      'H'
1244    #define CHAR_I                      'I'
1245    #define CHAR_J                      'J'
1246    #define CHAR_K                      'K'
1247    #define CHAR_L                      'L'
1248    #define CHAR_M                      'M'
1249    #define CHAR_N                      'N'
1250    #define CHAR_O                      'O'
1251    #define CHAR_P                      'P'
1252    #define CHAR_Q                      'Q'
1253    #define CHAR_R                      'R'
1254    #define CHAR_S                      'S'
1255    #define CHAR_T                      'T'
1256    #define CHAR_U                      'U'
1257    #define CHAR_V                      'V'
1258    #define CHAR_W                      'W'
1259    #define CHAR_X                      'X'
1260    #define CHAR_Y                      'Y'
1261    #define CHAR_Z                      'Z'
1262    #define CHAR_LEFT_SQUARE_BRACKET    '['
1263    #define CHAR_BACKSLASH              '\\'
1264    #define CHAR_RIGHT_SQUARE_BRACKET   ']'
1265    #define CHAR_CIRCUMFLEX_ACCENT      '^'
1266    #define CHAR_UNDERSCORE             '_'
1267    #define CHAR_GRAVE_ACCENT           '`'
1268    #define CHAR_a                      'a'
1269    #define CHAR_b                      'b'
1270    #define CHAR_c                      'c'
1271    #define CHAR_d                      'd'
1272    #define CHAR_e                      'e'
1273    #define CHAR_f                      'f'
1274    #define CHAR_g                      'g'
1275    #define CHAR_h                      'h'
1276    #define CHAR_i                      'i'
1277    #define CHAR_j                      'j'
1278    #define CHAR_k                      'k'
1279    #define CHAR_l                      'l'
1280    #define CHAR_m                      'm'
1281    #define CHAR_n                      'n'
1282    #define CHAR_o                      'o'
1283    #define CHAR_p                      'p'
1284    #define CHAR_q                      'q'
1285    #define CHAR_r                      'r'
1286    #define CHAR_s                      's'
1287    #define CHAR_t                      't'
1288    #define CHAR_u                      'u'
1289    #define CHAR_v                      'v'
1290    #define CHAR_w                      'w'
1291    #define CHAR_x                      'x'
1292    #define CHAR_y                      'y'
1293    #define CHAR_z                      'z'
1294    #define CHAR_LEFT_CURLY_BRACKET     '{'
1295    #define CHAR_VERTICAL_LINE          '|'
1296    #define CHAR_RIGHT_CURLY_BRACKET    '}'
1297    #define CHAR_TILDE                  '~'
1298    
1299    #define STR_HT                      "\t"
1300    #define STR_VT                      "\v"
1301    #define STR_FF                      "\f"
1302    #define STR_CR                      "\r"
1303    #define STR_BS                      "\b"
1304    #define STR_BEL                     "\a"
1305    
1306    #define STR_SPACE                   " "
1307    #define STR_EXCLAMATION_MARK        "!"
1308    #define STR_QUOTATION_MARK          "\""
1309    #define STR_NUMBER_SIGN             "#"
1310    #define STR_DOLLAR_SIGN             "$"
1311    #define STR_PERCENT_SIGN            "%"
1312    #define STR_AMPERSAND               "&"
1313    #define STR_APOSTROPHE              "'"
1314    #define STR_LEFT_PARENTHESIS        "("
1315    #define STR_RIGHT_PARENTHESIS       ")"
1316    #define STR_ASTERISK                "*"
1317    #define STR_PLUS                    "+"
1318    #define STR_COMMA                   ","
1319    #define STR_MINUS                   "-"
1320    #define STR_DOT                     "."
1321    #define STR_SLASH                   "/"
1322    #define STR_0                       "0"
1323    #define STR_1                       "1"
1324    #define STR_2                       "2"
1325    #define STR_3                       "3"
1326    #define STR_4                       "4"
1327    #define STR_5                       "5"
1328    #define STR_6                       "6"
1329    #define STR_7                       "7"
1330    #define STR_8                       "8"
1331    #define STR_9                       "9"
1332    #define STR_COLON                   ":"
1333    #define STR_SEMICOLON               ";"
1334    #define STR_LESS_THAN_SIGN          "<"
1335    #define STR_EQUALS_SIGN             "="
1336    #define STR_GREATER_THAN_SIGN       ">"
1337    #define STR_QUESTION_MARK           "?"
1338    #define STR_COMMERCIAL_AT           "@"
1339    #define STR_A                       "A"
1340    #define STR_B                       "B"
1341    #define STR_C                       "C"
1342    #define STR_D                       "D"
1343    #define STR_E                       "E"
1344    #define STR_F                       "F"
1345    #define STR_G                       "G"
1346    #define STR_H                       "H"
1347    #define STR_I                       "I"
1348    #define STR_J                       "J"
1349    #define STR_K                       "K"
1350    #define STR_L                       "L"
1351    #define STR_M                       "M"
1352    #define STR_N                       "N"
1353    #define STR_O                       "O"
1354    #define STR_P                       "P"
1355    #define STR_Q                       "Q"
1356    #define STR_R                       "R"
1357    #define STR_S                       "S"
1358    #define STR_T                       "T"
1359    #define STR_U                       "U"
1360    #define STR_V                       "V"
1361    #define STR_W                       "W"
1362    #define STR_X                       "X"
1363    #define STR_Y                       "Y"
1364    #define STR_Z                       "Z"
1365    #define STR_LEFT_SQUARE_BRACKET     "["
1366    #define STR_BACKSLASH               "\\"
1367    #define STR_RIGHT_SQUARE_BRACKET    "]"
1368    #define STR_CIRCUMFLEX_ACCENT       "^"
1369    #define STR_UNDERSCORE              "_"
1370    #define STR_GRAVE_ACCENT            "`"
1371    #define STR_a                       "a"
1372    #define STR_b                       "b"
1373    #define STR_c                       "c"
1374    #define STR_d                       "d"
1375    #define STR_e                       "e"
1376    #define STR_f                       "f"
1377    #define STR_g                       "g"
1378    #define STR_h                       "h"
1379    #define STR_i                       "i"
1380    #define STR_j                       "j"
1381    #define STR_k                       "k"
1382    #define STR_l                       "l"
1383    #define STR_m                       "m"
1384    #define STR_n                       "n"
1385    #define STR_o                       "o"
1386    #define STR_p                       "p"
1387    #define STR_q                       "q"
1388    #define STR_r                       "r"
1389    #define STR_s                       "s"
1390    #define STR_t                       "t"
1391    #define STR_u                       "u"
1392    #define STR_v                       "v"
1393    #define STR_w                       "w"
1394    #define STR_x                       "x"
1395    #define STR_y                       "y"
1396    #define STR_z                       "z"
1397    #define STR_LEFT_CURLY_BRACKET      "{"
1398    #define STR_VERTICAL_LINE           "|"
1399    #define STR_RIGHT_CURLY_BRACKET     "}"
1400    #define STR_TILDE                   "~"
1401    
1402    #define STRING_ACCEPT0              "ACCEPT\0"
1403    #define STRING_COMMIT0              "COMMIT\0"
1404    #define STRING_F0                   "F\0"
1405    #define STRING_FAIL0                "FAIL\0"
1406    #define STRING_MARK0                "MARK\0"
1407    #define STRING_PRUNE0               "PRUNE\0"
1408    #define STRING_SKIP0                "SKIP\0"
1409    #define STRING_THEN                 "THEN"
1410    
1411    #define STRING_alpha0               "alpha\0"
1412    #define STRING_lower0               "lower\0"
1413    #define STRING_upper0               "upper\0"
1414    #define STRING_alnum0               "alnum\0"
1415    #define STRING_ascii0               "ascii\0"
1416    #define STRING_blank0               "blank\0"
1417    #define STRING_cntrl0               "cntrl\0"
1418    #define STRING_digit0               "digit\0"
1419    #define STRING_graph0               "graph\0"
1420    #define STRING_print0               "print\0"
1421    #define STRING_punct0               "punct\0"
1422    #define STRING_space0               "space\0"
1423    #define STRING_word0                "word\0"
1424    #define STRING_xdigit               "xdigit"
1425    
1426    #define STRING_DEFINE               "DEFINE"
1427    
1428    #define STRING_CR_RIGHTPAR             "CR)"
1429    #define STRING_LF_RIGHTPAR             "LF)"
1430    #define STRING_CRLF_RIGHTPAR           "CRLF)"
1431    #define STRING_ANY_RIGHTPAR            "ANY)"
1432    #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1433    #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1434    #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1435    #ifdef COMPILE_PCRE8
1436    #define STRING_UTF_RIGHTPAR            "UTF8)"
1437    #endif
1438    #ifdef COMPILE_PCRE16
1439    #define STRING_UTF_RIGHTPAR            "UTF16)"
1440    #endif
1441    #ifdef COMPILE_PCRE32
1442    #define STRING_UTF_RIGHTPAR            "UTF32)"
1443    #endif
1444    #define STRING_UCP_RIGHTPAR            "UCP)"
1445    #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1446    
1447  /* Escape items that are just an encoding of a particular data value. Note that  #else  /* SUPPORT_UTF */
1448  ESC_n is defined as yet another macro, which is set in config.h to either \n  
1449  (the default) or \r (which some people want). */  /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1450    works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
1451    only. */
1452    
1453    #define CHAR_HT                     '\011'
1454    #define CHAR_VT                     '\013'
1455    #define CHAR_FF                     '\014'
1456    #define CHAR_CR                     '\015'
1457    #define CHAR_LF                     '\012'
1458    #define CHAR_NL                     CHAR_LF
1459    #define CHAR_NEL                    ((unsigned char)'\x85')
1460    #define CHAR_BS                     '\010'
1461    #define CHAR_BEL                    '\007'
1462    #define CHAR_ESC                    '\033'
1463    #define CHAR_DEL                    '\177'
1464    
1465    #define CHAR_SPACE                  '\040'
1466    #define CHAR_EXCLAMATION_MARK       '\041'
1467    #define CHAR_QUOTATION_MARK         '\042'
1468    #define CHAR_NUMBER_SIGN            '\043'
1469    #define CHAR_DOLLAR_SIGN            '\044'
1470    #define CHAR_PERCENT_SIGN           '\045'
1471    #define CHAR_AMPERSAND              '\046'
1472    #define CHAR_APOSTROPHE             '\047'
1473    #define CHAR_LEFT_PARENTHESIS       '\050'
1474    #define CHAR_RIGHT_PARENTHESIS      '\051'
1475    #define CHAR_ASTERISK               '\052'
1476    #define CHAR_PLUS                   '\053'
1477    #define CHAR_COMMA                  '\054'
1478    #define CHAR_MINUS                  '\055'
1479    #define CHAR_DOT                    '\056'
1480    #define CHAR_SLASH                  '\057'
1481    #define CHAR_0                      '\060'
1482    #define CHAR_1                      '\061'
1483    #define CHAR_2                      '\062'
1484    #define CHAR_3                      '\063'
1485    #define CHAR_4                      '\064'
1486    #define CHAR_5                      '\065'
1487    #define CHAR_6                      '\066'
1488    #define CHAR_7                      '\067'
1489    #define CHAR_8                      '\070'
1490    #define CHAR_9                      '\071'
1491    #define CHAR_COLON                  '\072'
1492    #define CHAR_SEMICOLON              '\073'
1493    #define CHAR_LESS_THAN_SIGN         '\074'
1494    #define CHAR_EQUALS_SIGN            '\075'
1495    #define CHAR_GREATER_THAN_SIGN      '\076'
1496    #define CHAR_QUESTION_MARK          '\077'
1497    #define CHAR_COMMERCIAL_AT          '\100'
1498    #define CHAR_A                      '\101'
1499    #define CHAR_B                      '\102'
1500    #define CHAR_C                      '\103'
1501    #define CHAR_D                      '\104'
1502    #define CHAR_E                      '\105'
1503    #define CHAR_F                      '\106'
1504    #define CHAR_G                      '\107'
1505    #define CHAR_H                      '\110'
1506    #define CHAR_I                      '\111'
1507    #define CHAR_J                      '\112'
1508    #define CHAR_K                      '\113'
1509    #define CHAR_L                      '\114'
1510    #define CHAR_M                      '\115'
1511    #define CHAR_N                      '\116'
1512    #define CHAR_O                      '\117'
1513    #define CHAR_P                      '\120'
1514    #define CHAR_Q                      '\121'
1515    #define CHAR_R                      '\122'
1516    #define CHAR_S                      '\123'
1517    #define CHAR_T                      '\124'
1518    #define CHAR_U                      '\125'
1519    #define CHAR_V                      '\126'
1520    #define CHAR_W                      '\127'
1521    #define CHAR_X                      '\130'
1522    #define CHAR_Y                      '\131'
1523    #define CHAR_Z                      '\132'
1524    #define CHAR_LEFT_SQUARE_BRACKET    '\133'
1525    #define CHAR_BACKSLASH              '\134'
1526    #define CHAR_RIGHT_SQUARE_BRACKET   '\135'
1527    #define CHAR_CIRCUMFLEX_ACCENT      '\136'
1528    #define CHAR_UNDERSCORE             '\137'
1529    #define CHAR_GRAVE_ACCENT           '\140'
1530    #define CHAR_a                      '\141'
1531    #define CHAR_b                      '\142'
1532    #define CHAR_c                      '\143'
1533    #define CHAR_d                      '\144'
1534    #define CHAR_e                      '\145'
1535    #define CHAR_f                      '\146'
1536    #define CHAR_g                      '\147'
1537    #define CHAR_h                      '\150'
1538    #define CHAR_i                      '\151'
1539    #define CHAR_j                      '\152'
1540    #define CHAR_k                      '\153'
1541    #define CHAR_l                      '\154'
1542    #define CHAR_m                      '\155'
1543    #define CHAR_n                      '\156'
1544    #define CHAR_o                      '\157'
1545    #define CHAR_p                      '\160'
1546    #define CHAR_q                      '\161'
1547    #define CHAR_r                      '\162'
1548    #define CHAR_s                      '\163'
1549    #define CHAR_t                      '\164'
1550    #define CHAR_u                      '\165'
1551    #define CHAR_v                      '\166'
1552    #define CHAR_w                      '\167'
1553    #define CHAR_x                      '\170'
1554    #define CHAR_y                      '\171'
1555    #define CHAR_z                      '\172'
1556    #define CHAR_LEFT_CURLY_BRACKET     '\173'
1557    #define CHAR_VERTICAL_LINE          '\174'
1558    #define CHAR_RIGHT_CURLY_BRACKET    '\175'
1559    #define CHAR_TILDE                  '\176'
1560    
1561    #define STR_HT                      "\011"
1562    #define STR_VT                      "\013"
1563    #define STR_FF                      "\014"
1564    #define STR_CR                      "\015"
1565    #define STR_NL                      "\012"
1566    #define STR_BS                      "\010"
1567    #define STR_BEL                     "\007"
1568    #define STR_ESC                     "\033"
1569    #define STR_DEL                     "\177"
1570    
1571    #define STR_SPACE                   "\040"
1572    #define STR_EXCLAMATION_MARK        "\041"
1573    #define STR_QUOTATION_MARK          "\042"
1574    #define STR_NUMBER_SIGN             "\043"
1575    #define STR_DOLLAR_SIGN             "\044"
1576    #define STR_PERCENT_SIGN            "\045"
1577    #define STR_AMPERSAND               "\046"
1578    #define STR_APOSTROPHE              "\047"
1579    #define STR_LEFT_PARENTHESIS        "\050"
1580    #define STR_RIGHT_PARENTHESIS       "\051"
1581    #define STR_ASTERISK                "\052"
1582    #define STR_PLUS                    "\053"
1583    #define STR_COMMA                   "\054"
1584    #define STR_MINUS                   "\055"
1585    #define STR_DOT                     "\056"
1586    #define STR_SLASH                   "\057"
1587    #define STR_0                       "\060"
1588    #define STR_1                       "\061"
1589    #define STR_2                       "\062"
1590    #define STR_3                       "\063"
1591    #define STR_4                       "\064"
1592    #define STR_5                       "\065"
1593    #define STR_6                       "\066"
1594    #define STR_7                       "\067"
1595    #define STR_8                       "\070"
1596    #define STR_9                       "\071"
1597    #define STR_COLON                   "\072"
1598    #define STR_SEMICOLON               "\073"
1599    #define STR_LESS_THAN_SIGN          "\074"
1600    #define STR_EQUALS_SIGN             "\075"
1601    #define STR_GREATER_THAN_SIGN       "\076"
1602    #define STR_QUESTION_MARK           "\077"
1603    #define STR_COMMERCIAL_AT           "\100"
1604    #define STR_A                       "\101"
1605    #define STR_B                       "\102"
1606    #define STR_C                       "\103"
1607    #define STR_D                       "\104"
1608    #define STR_E                       "\105"
1609    #define STR_F                       "\106"
1610    #define STR_G                       "\107"
1611    #define STR_H                       "\110"
1612    #define STR_I                       "\111"
1613    #define STR_J                       "\112"
1614    #define STR_K                       "\113"
1615    #define STR_L                       "\114"
1616    #define STR_M                       "\115"
1617    #define STR_N                       "\116"
1618    #define STR_O                       "\117"
1619    #define STR_P                       "\120"
1620    #define STR_Q                       "\121"
1621    #define STR_R                       "\122"
1622    #define STR_S                       "\123"
1623    #define STR_T                       "\124"
1624    #define STR_U                       "\125"
1625    #define STR_V                       "\126"
1626    #define STR_W                       "\127"
1627    #define STR_X                       "\130"
1628    #define STR_Y                       "\131"
1629    #define STR_Z                       "\132"
1630    #define STR_LEFT_SQUARE_BRACKET     "\133"
1631    #define STR_BACKSLASH               "\134"
1632    #define STR_RIGHT_SQUARE_BRACKET    "\135"
1633    #define STR_CIRCUMFLEX_ACCENT       "\136"
1634    #define STR_UNDERSCORE              "\137"
1635    #define STR_GRAVE_ACCENT            "\140"
1636    #define STR_a                       "\141"
1637    #define STR_b                       "\142"
1638    #define STR_c                       "\143"
1639    #define STR_d                       "\144"
1640    #define STR_e                       "\145"
1641    #define STR_f                       "\146"
1642    #define STR_g                       "\147"
1643    #define STR_h                       "\150"
1644    #define STR_i                       "\151"
1645    #define STR_j                       "\152"
1646    #define STR_k                       "\153"
1647    #define STR_l                       "\154"
1648    #define STR_m                       "\155"
1649    #define STR_n                       "\156"
1650    #define STR_o                       "\157"
1651    #define STR_p                       "\160"
1652    #define STR_q                       "\161"
1653    #define STR_r                       "\162"
1654    #define STR_s                       "\163"
1655    #define STR_t                       "\164"
1656    #define STR_u                       "\165"
1657    #define STR_v                       "\166"
1658    #define STR_w                       "\167"
1659    #define STR_x                       "\170"
1660    #define STR_y                       "\171"
1661    #define STR_z                       "\172"
1662    #define STR_LEFT_CURLY_BRACKET      "\173"
1663    #define STR_VERTICAL_LINE           "\174"
1664    #define STR_RIGHT_CURLY_BRACKET     "\175"
1665    #define STR_TILDE                   "\176"
1666    
1667    #define STRING_ACCEPT0              STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1668    #define STRING_COMMIT0              STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1669    #define STRING_F0                   STR_F "\0"
1670    #define STRING_FAIL0                STR_F STR_A STR_I STR_L "\0"
1671    #define STRING_MARK0                STR_M STR_A STR_R STR_K "\0"
1672    #define STRING_PRUNE0               STR_P STR_R STR_U STR_N STR_E "\0"
1673    #define STRING_SKIP0                STR_S STR_K STR_I STR_P "\0"
1674    #define STRING_THEN                 STR_T STR_H STR_E STR_N
1675    
1676    #define STRING_alpha0               STR_a STR_l STR_p STR_h STR_a "\0"
1677    #define STRING_lower0               STR_l STR_o STR_w STR_e STR_r "\0"
1678    #define STRING_upper0               STR_u STR_p STR_p STR_e STR_r "\0"
1679    #define STRING_alnum0               STR_a STR_l STR_n STR_u STR_m "\0"
1680    #define STRING_ascii0               STR_a STR_s STR_c STR_i STR_i "\0"
1681    #define STRING_blank0               STR_b STR_l STR_a STR_n STR_k "\0"
1682    #define STRING_cntrl0               STR_c STR_n STR_t STR_r STR_l "\0"
1683    #define STRING_digit0               STR_d STR_i STR_g STR_i STR_t "\0"
1684    #define STRING_graph0               STR_g STR_r STR_a STR_p STR_h "\0"
1685    #define STRING_print0               STR_p STR_r STR_i STR_n STR_t "\0"
1686    #define STRING_punct0               STR_p STR_u STR_n STR_c STR_t "\0"
1687    #define STRING_space0               STR_s STR_p STR_a STR_c STR_e "\0"
1688    #define STRING_word0                STR_w STR_o STR_r STR_d       "\0"
1689    #define STRING_xdigit               STR_x STR_d STR_i STR_g STR_i STR_t
1690    
1691    #define STRING_DEFINE               STR_D STR_E STR_F STR_I STR_N STR_E
1692    
1693    #define STRING_CR_RIGHTPAR             STR_C STR_R STR_RIGHT_PARENTHESIS
1694    #define STRING_LF_RIGHTPAR             STR_L STR_F STR_RIGHT_PARENTHESIS
1695    #define STRING_CRLF_RIGHTPAR           STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1696    #define STRING_ANY_RIGHTPAR            STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1697    #define STRING_ANYCRLF_RIGHTPAR        STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1698    #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
1699    #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
1700    #ifdef COMPILE_PCRE8
1701    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1702    #endif
1703    #ifdef COMPILE_PCRE16
1704    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1705    #endif
1706    #ifdef COMPILE_PCRE32
1707    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1708    #endif
1709    #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1710    #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
1711    
1712    #endif  /* SUPPORT_UTF */
1713    
1714    /* Escape items that are just an encoding of a particular data value. */
1715    
1716  #ifndef ESC_e  #ifndef ESC_e
1717  #define ESC_e 27  #define ESC_e CHAR_ESC
1718  #endif  #endif
1719    
1720  #ifndef ESC_f  #ifndef ESC_f
1721  #define ESC_f '\f'  #define ESC_f CHAR_FF
1722  #endif  #endif
1723    
1724  #ifndef ESC_n  #ifndef ESC_n
1725  #define ESC_n NEWLINE  #define ESC_n CHAR_LF
1726  #endif  #endif
1727    
1728  #ifndef ESC_r  #ifndef ESC_r
1729  #define ESC_r '\r'  #define ESC_r CHAR_CR
1730  #endif  #endif
1731    
1732  /* 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
1733  (presumably because of all the others like size_t). */  (presumably because of all the others like size_t). */
1734    
1735  #ifndef ESC_tee  #ifndef ESC_tee
1736  #define ESC_tee '\t'  #define ESC_tee CHAR_HT
1737  #endif  #endif
1738    
1739  /* These are escaped items that aren't just an encoding of a particular data  /* Codes for different types of Unicode property */
1740  value such as \n. They must have non-zero values, as check_escape() returns  
1741  their negation. Also, they must appear in the same order as in the opcode  #define PT_ANY        0    /* Any property - matches all chars */
1742  definitions below, up to ESC_z. There's a dummy for OP_ANY because it  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */
1743  corresponds to "." rather than an escape sequence. The final one must be  #define PT_GC         2    /* Specified general characteristic (e.g. L) */
1744  ESC_REF as subsequent values are used for \1, \2, \3, etc. There is are two  #define PT_PC         3    /* Specified particular characteristic (e.g. Lu) */
1745  tests in the code for an escape greater than ESC_b and less than ESC_Z to  #define PT_SC         4    /* Script (e.g. Han) */
1746  detect the types that may be repeated. These are the types that consume  #define PT_ALNUM      5    /* Alphanumeric - the union of L and N */
1747  characters. If any new escapes are put in between that don't consume a  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1748  character, that code will have to change. */  #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1749    #define PT_WORD       8    /* Word - L plus N plus underscore */
1750  enum { ESC_A = 1, ESC_G, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, ESC_W,  #define PT_CLIST      9    /* Pseudo-property: match character list */
        ESC_w, ESC_dum1, ESC_C, ESC_P, ESC_p, ESC_X, ESC_Z, ESC_z, ESC_E,  
        ESC_Q, ESC_REF };  
1751    
1752  /* 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
1753  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1754    
1755  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1756  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */
# Line 469  contain UTF-8 characters with values gre Line 1758  contain UTF-8 characters with values gre
1758  #define XCL_END       0    /* Marks end of individual items */  #define XCL_END       0    /* Marks end of individual items */
1759  #define XCL_SINGLE    1    /* Single item (one multibyte char) follows */  #define XCL_SINGLE    1    /* Single item (one multibyte char) follows */
1760  #define XCL_RANGE     2    /* A range (two multibyte chars) follows */  #define XCL_RANGE     2    /* A range (two multibyte chars) follows */
1761  #define XCL_PROP      3    /* Unicode property (one property code) follows */  #define XCL_PROP      3    /* Unicode property (2-byte property code follows) */
1762  #define XCL_NOTPROP   4    /* Unicode inverted property (ditto) */  #define XCL_NOTPROP   4    /* Unicode inverted property (ditto) */
1763    
1764    /* These are escaped items that aren't just an encoding of a particular data
1765    value such as \n. They must have non-zero values, as check_escape() returns
1766    0 for a data character.  Also, they must appear in the same order as in the opcode
1767    definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1768    corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1769    used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1770    non-DOTALL mode, "." behaves like \N.
1771    
1772    The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1773    when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1774    They must be contiguous, and remain in order so that the replacements can be
1775    looked up from a table.
1776    
1777    Negative numbers are used to encode a backreference (\1, \2, \3, etc.) in
1778    check_escape(). There are two tests in the code for an escape
1779    greater than ESC_b and less than ESC_Z to detect the types that may be
1780    repeated. These are the types that consume characters. If any new escapes are
1781    put in between that don't consume a character, that code will have to change.
1782    */
1783    
1784  /* Opcode table: OP_BRA must be last, as all values >= it are used for brackets  enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1785  that extract substrings. Starting from 1 (i.e. after OP_END), the values up to         ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1786           ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1787           ESC_E, ESC_Q, ESC_g, ESC_k,
1788           ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1789    
1790    /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
1791  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.
1792  Note that whenever this list is updated, the two macro definitions that follow  
1793  must also be updated to match. */  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1794    that follow must also be updated to match. There are also tables called
1795    "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1796    
1797  enum {  enum {
1798    OP_END,            /* 0 End of pattern */    OP_END,            /* 0 End of pattern */
# Line 486  enum { Line 1801  enum {
1801    
1802    OP_SOD,            /* 1 Start of data: \A */    OP_SOD,            /* 1 Start of data: \A */
1803    OP_SOM,            /* 2 Start of match (subject + offset): \G */    OP_SOM,            /* 2 Start of match (subject + offset): \G */
1804    OP_NOT_WORD_BOUNDARY,  /*  3 \B */    OP_SET_SOM,        /* 3 Set start of match (\K) */
1805    OP_WORD_BOUNDARY,      /*  4 \b */    OP_NOT_WORD_BOUNDARY,  /*  4 \B */
1806    OP_NOT_DIGIT,          /*  5 \D */    OP_WORD_BOUNDARY,      /*  5 \b */
1807    OP_DIGIT,              /*  6 \d */    OP_NOT_DIGIT,          /*  6 \D */
1808    OP_NOT_WHITESPACE,     /*  7 \S */    OP_DIGIT,              /*  7 \d */
1809    OP_WHITESPACE,         /*  8 \s */    OP_NOT_WHITESPACE,     /*  8 \S */
1810    OP_NOT_WORDCHAR,       /*  9 \W */    OP_WHITESPACE,         /*  9 \s */
1811    OP_WORDCHAR,           /* 10 \w */    OP_NOT_WORDCHAR,       /* 10 \W */
1812    OP_ANY,            /* 11 Match any character */    OP_WORDCHAR,           /* 11 \w */
1813    OP_ANYBYTE,        /* 12 Match any byte (\C); different to OP_ANY for UTF-8 */  
1814    OP_NOTPROP,        /* 13 \P (not Unicode property) */    OP_ANY,            /* 12 Match any character except newline (\N) */
1815    OP_PROP,           /* 14 \p (Unicode property) */    OP_ALLANY,         /* 13 Match any character */
1816    OP_EXTUNI,         /* 15 \X (extended Unicode sequence */    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1817    OP_EODN,           /* 16 End of data or \n at end of data: \Z. */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
1818    OP_EOD,            /* 17 End of data: \z */    OP_PROP,           /* 16 \p (Unicode property) */
1819      OP_ANYNL,          /* 17 \R (any newline sequence) */
1820    OP_OPT,            /* 18 Set runtime options */    OP_NOT_HSPACE,     /* 18 \H (not horizontal whitespace) */
1821    OP_CIRC,           /* 19 Start of line - varies with multiline switch */    OP_HSPACE,         /* 19 \h (horizontal whitespace) */
1822    OP_DOLL,           /* 20 End of line - varies with multiline switch */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1823    OP_CHAR,           /* 21 Match one character, casefully */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1824    OP_CHARNC,         /* 22 Match one character, caselessly */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1825    OP_NOT,            /* 23 Match anything but the following char */    OP_EODN,           /* 23 End of data or \n at end of data (\Z) */
1826      OP_EOD,            /* 24 End of data (\z) */
1827    OP_STAR,           /* 24 The maximizing and minimizing versions of */  
1828    OP_MINSTAR,        /* 25 all these opcodes must come in pairs, with */    OP_CIRC,           /* 25 Start of line - not multiline */
1829    OP_PLUS,           /* 26 the minimizing one second. */    OP_CIRCM,          /* 26 Start of line - multiline */
1830    OP_MINPLUS,        /* 27 This first set applies to single characters */    OP_DOLL,           /* 27 End of line - not multiline */
1831    OP_QUERY,          /* 28 */    OP_DOLLM,          /* 28 End of line - multiline */
1832    OP_MINQUERY,       /* 29 */    OP_CHAR,           /* 29 Match one character, casefully */
1833    OP_UPTO,           /* 30 From 0 to n matches */    OP_CHARI,          /* 30 Match one character, caselessly */
1834    OP_MINUPTO,        /* 31 */    OP_NOT,            /* 31 Match one character, not the given one, casefully */
1835    OP_EXACT,          /* 32 Exactly n matches */    OP_NOTI,           /* 32 Match one character, not the given one, caselessly */
1836    
1837    OP_NOTSTAR,        /* 33 The maximizing and minimizing versions of */    /* The following sets of 13 opcodes must always be kept in step because
1838    OP_NOTMINSTAR,     /* 34 all these opcodes must come in pairs, with */    the offset from the first one is used to generate the others. */
1839    OP_NOTPLUS,        /* 35 the minimizing one second. */  
1840    OP_NOTMINPLUS,     /* 36 This set applies to "not" single characters */    /**** Single characters, caseful, must precede the caseless ones ****/
1841    OP_NOTQUERY,       /* 37 */  
1842    OP_NOTMINQUERY,    /* 38 */    OP_STAR,           /* 33 The maximizing and minimizing versions of */
1843    OP_NOTUPTO,        /* 39 From 0 to n matches */    OP_MINSTAR,        /* 34 these six opcodes must come in pairs, with */
1844    OP_NOTMINUPTO,     /* 40 */    OP_PLUS,           /* 35 the minimizing one second. */
1845    OP_NOTEXACT,       /* 41 Exactly n matches */    OP_MINPLUS,        /* 36 */
1846      OP_QUERY,          /* 37 */
1847    OP_TYPESTAR,       /* 42 The maximizing and minimizing versions of */    OP_MINQUERY,       /* 38 */
1848    OP_TYPEMINSTAR,    /* 43 all these opcodes must come in pairs, with */  
1849    OP_TYPEPLUS,       /* 44 the minimizing one second. These codes must */    OP_UPTO,           /* 39 From 0 to n matches of one character, caseful*/
1850    OP_TYPEMINPLUS,    /* 45 be in exactly the same order as those above. */    OP_MINUPTO,        /* 40 */
1851    OP_TYPEQUERY,      /* 46 This set applies to character types such as \d */    OP_EXACT,          /* 41 Exactly n matches */
1852    OP_TYPEMINQUERY,   /* 47 */  
1853    OP_TYPEUPTO,       /* 48 From 0 to n matches */    OP_POSSTAR,        /* 42 Possessified star, caseful */
1854    OP_TYPEMINUPTO,    /* 49 */    OP_POSPLUS,        /* 43 Possessified plus, caseful */
1855    OP_TYPEEXACT,      /* 50 Exactly n matches */    OP_POSQUERY,       /* 44 Posesssified query, caseful */
1856      OP_POSUPTO,        /* 45 Possessified upto, caseful */
1857    OP_CRSTAR,         /* 51 The maximizing and minimizing versions of */  
1858    OP_CRMINSTAR,      /* 52 all these opcodes must come in pairs, with */    /**** Single characters, caseless, must follow the caseful ones */
1859    OP_CRPLUS,         /* 53 the minimizing one second. These codes must */  
1860    OP_CRMINPLUS,      /* 54 be in exactly the same order as those above. */    OP_STARI,          /* 46 */
1861    OP_CRQUERY,        /* 55 These are for character classes and back refs */    OP_MINSTARI,       /* 47 */
1862    OP_CRMINQUERY,     /* 56 */    OP_PLUSI,          /* 48 */
1863    OP_CRRANGE,        /* 57 These are different to the three sets above. */    OP_MINPLUSI,       /* 49 */
1864    OP_CRMINRANGE,     /* 58 */    OP_QUERYI,         /* 50 */
1865      OP_MINQUERYI,      /* 51 */
1866    OP_CLASS,          /* 59 Match a character class, chars < 256 only */  
1867    OP_NCLASS,         /* 60 Same, but the bitmap was created from a negative    OP_UPTOI,          /* 52 From 0 to n matches of one character, caseless */
1868                             class - the difference is relevant only when a UTF-8    OP_MINUPTOI,       /* 53 */
1869                             character > 255 is encountered. */    OP_EXACTI,         /* 54 */
1870    
1871    OP_XCLASS,         /* 61 Extended class for handling UTF-8 chars within the    OP_POSSTARI,       /* 55 Possessified star, caseless */
1872                             class. This does both positive and negative. */    OP_POSPLUSI,       /* 56 Possessified plus, caseless */
1873      OP_POSQUERYI,      /* 57 Posesssified query, caseless */
1874    OP_REF,            /* 62 Match a back reference */    OP_POSUPTOI,       /* 58 Possessified upto, caseless */
1875    OP_RECURSE,        /* 63 Match a numbered subpattern (possibly recursive) */  
1876    OP_CALLOUT,        /* 64 Call out to external function if provided */    /**** The negated ones must follow the non-negated ones, and match them ****/
1877      /**** Negated single character, caseful; must precede the caseless ones ****/
1878    OP_ALT,            /* 65 Start of alternation */  
1879    OP_KET,            /* 66 End of group that doesn't have an unbounded repeat */    OP_NOTSTAR,        /* 59 The maximizing and minimizing versions of */
1880    OP_KETRMAX,        /* 67 These two must remain together and in this */    OP_NOTMINSTAR,     /* 60 these six opcodes must come in pairs, with */
1881    OP_KETRMIN,        /* 68 order. They are for groups the repeat for ever. */    OP_NOTPLUS,        /* 61 the minimizing one second. They must be in */
1882      OP_NOTMINPLUS,     /* 62 exactly the same order as those above. */
1883    /* The assertions must come before ONCE and COND */    OP_NOTQUERY,       /* 63 */
1884      OP_NOTMINQUERY,    /* 64 */
1885    OP_ASSERT,         /* 69 Positive lookahead */  
1886    OP_ASSERT_NOT,     /* 70 Negative lookahead */    OP_NOTUPTO,        /* 65 From 0 to n matches, caseful */
1887    OP_ASSERTBACK,     /* 71 Positive lookbehind */    OP_NOTMINUPTO,     /* 66 */
1888    OP_ASSERTBACK_NOT, /* 72 Negative lookbehind */    OP_NOTEXACT,       /* 67 Exactly n matches */
1889    OP_REVERSE,        /* 73 Move pointer back - used in lookbehind assertions */  
1890      OP_NOTPOSSTAR,     /* 68 Possessified versions, caseful */
1891    /* ONCE and COND must come after the assertions, with ONCE first, as there's    OP_NOTPOSPLUS,     /* 69 */
1892    a test for >= ONCE for a subpattern that isn't an assertion. */    OP_NOTPOSQUERY,    /* 70 */
1893      OP_NOTPOSUPTO,     /* 71 */
1894    OP_ONCE,           /* 74 Once matched, don't back up into the subpattern */  
1895    OP_COND,           /* 75 Conditional group */    /**** Negated single character, caseless; must follow the caseful ones ****/
1896    OP_CREF,           /* 76 Used to hold an extraction string number (cond ref) */  
1897      OP_NOTSTARI,       /* 72 */
1898    OP_BRAZERO,        /* 77 These two must remain together and in this */    OP_NOTMINSTARI,    /* 73 */
1899    OP_BRAMINZERO,     /* 78 order. */    OP_NOTPLUSI,       /* 74 */
1900      OP_NOTMINPLUSI,    /* 75 */
1901    OP_BRANUMBER,      /* 79 Used for extracting brackets whose number is greater    OP_NOTQUERYI,      /* 76 */
1902                             than can fit into an opcode. */    OP_NOTMINQUERYI,   /* 77 */
1903    
1904    OP_BRA             /* 80 This and greater values are used for brackets that    OP_NOTUPTOI,       /* 78 From 0 to n matches, caseless */
1905                             extract substrings up to EXTRACT_BASIC_MAX. After    OP_NOTMINUPTOI,    /* 79 */
1906                             that, use is made of OP_BRANUMBER. */    OP_NOTEXACTI,      /* 80 Exactly n matches */
1907  };  
1908      OP_NOTPOSSTARI,    /* 81 Possessified versions, caseless */
1909      OP_NOTPOSPLUSI,    /* 82 */
1910      OP_NOTPOSQUERYI,   /* 83 */
1911      OP_NOTPOSUPTOI,    /* 84 */
1912    
1913      /**** Character types ****/
1914    
1915      OP_TYPESTAR,       /* 85 The maximizing and minimizing versions of */
1916      OP_TYPEMINSTAR,    /* 86 these six opcodes must come in pairs, with */
1917      OP_TYPEPLUS,       /* 87 the minimizing one second. These codes must */
1918      OP_TYPEMINPLUS,    /* 88 be in exactly the same order as those above. */
1919      OP_TYPEQUERY,      /* 89 */
1920      OP_TYPEMINQUERY,   /* 90 */
1921    
1922      OP_TYPEUPTO,       /* 91 From 0 to n matches */
1923      OP_TYPEMINUPTO,    /* 92 */
1924      OP_TYPEEXACT,      /* 93 Exactly n matches */
1925    
1926      OP_TYPEPOSSTAR,    /* 94 Possessified versions */
1927      OP_TYPEPOSPLUS,    /* 95 */
1928      OP_TYPEPOSQUERY,   /* 96 */
1929      OP_TYPEPOSUPTO,    /* 97 */
1930    
1931      /* These are used for character classes and back references; only the
1932      first six are the same as the sets above. */
1933    
1934      OP_CRSTAR,         /* 98 The maximizing and minimizing versions of */
1935      OP_CRMINSTAR,      /* 99 all these opcodes must come in pairs, with */
1936      OP_CRPLUS,         /* 100 the minimizing one second. These codes must */
1937      OP_CRMINPLUS,      /* 101 be in exactly the same order as those above. */
1938      OP_CRQUERY,        /* 102 */
1939      OP_CRMINQUERY,     /* 103 */
1940    
1941      OP_CRRANGE,        /* 104 These are different to the three sets above. */
1942      OP_CRMINRANGE,     /* 105 */
1943    
1944      /* End of quantifier opcodes */
1945    
1946      OP_CLASS,          /* 106 Match a character class, chars < 256 only */
1947      OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
1948                                  class - the difference is relevant only when a
1949                                  character > 255 is encountered. */
1950      OP_XCLASS,         /* 108 Extended class for handling > 255 chars within the
1951                                  class. This does both positive and negative. */
1952      OP_REF,            /* 109 Match a back reference, casefully */
1953      OP_REFI,           /* 110 Match a back reference, caselessly */
1954      OP_RECURSE,        /* 111 Match a numbered subpattern (possibly recursive) */
1955      OP_CALLOUT,        /* 112 Call out to external function if provided */
1956    
1957      OP_ALT,            /* 113 Start of alternation */
1958      OP_KET,            /* 114 End of group that doesn't have an unbounded repeat */
1959      OP_KETRMAX,        /* 115 These two must remain together and in this */
1960      OP_KETRMIN,        /* 116 order. They are for groups the repeat for ever. */
1961      OP_KETRPOS,        /* 117 Possessive unlimited repeat. */
1962    
1963      /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
1964      asserts must remain in order. */
1965    
1966      OP_REVERSE,        /* 118 Move pointer back - used in lookbehind assertions */
1967      OP_ASSERT,         /* 119 Positive lookahead */
1968      OP_ASSERT_NOT,     /* 120 Negative lookahead */
1969      OP_ASSERTBACK,     /* 121 Positive lookbehind */
1970      OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
1971    
1972      /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
1973      after the assertions, with ONCE first, as there's a test for >= ONCE for a
1974      subpattern that isn't an assertion. The POS versions must immediately follow
1975      the non-POS versions in each case. */
1976    
1977      OP_ONCE,           /* 123 Atomic group, contains captures */
1978      OP_ONCE_NC,        /* 124 Atomic group containing no captures */
1979      OP_BRA,            /* 125 Start of non-capturing bracket */
1980      OP_BRAPOS,         /* 126 Ditto, with unlimited, possessive repeat */
1981      OP_CBRA,           /* 127 Start of capturing bracket */
1982      OP_CBRAPOS,        /* 128 Ditto, with unlimited, possessive repeat */
1983      OP_COND,           /* 129 Conditional group */
1984    
1985      /* These five must follow the previous five, in the same order. There's a
1986      check for >= SBRA to distinguish the two sets. */
1987    
1988      OP_SBRA,           /* 130 Start of non-capturing bracket, check empty  */
1989      OP_SBRAPOS,        /* 131 Ditto, with unlimited, possessive repeat */
1990      OP_SCBRA,          /* 132 Start of capturing bracket, check empty */
1991      OP_SCBRAPOS,       /* 133 Ditto, with unlimited, possessive repeat */
1992      OP_SCOND,          /* 134 Conditional group, check empty */
1993    
1994      /* The next two pairs must (respectively) be kept together. */
1995    
1996      OP_CREF,           /* 135 Used to hold a capture number as condition */
1997      OP_NCREF,          /* 136 Same, but generated by a name reference*/
1998      OP_RREF,           /* 137 Used to hold a recursion number as condition */
1999      OP_NRREF,          /* 138 Same, but generated by a name reference*/
2000      OP_DEF,            /* 139 The DEFINE condition */
2001    
2002      OP_BRAZERO,        /* 140 These two must remain together and in this */
2003      OP_BRAMINZERO,     /* 141 order. */
2004      OP_BRAPOSZERO,     /* 142 */
2005    
2006      /* These are backtracking control verbs */
2007    
2008      OP_MARK,           /* 143 always has an argument */
2009      OP_PRUNE,          /* 144 */
2010      OP_PRUNE_ARG,      /* 145 same, but with argument */
2011      OP_SKIP,           /* 146 */
2012      OP_SKIP_ARG,       /* 147 same, but with argument */
2013      OP_THEN,           /* 148 */
2014      OP_THEN_ARG,       /* 149 same, but with argument */
2015      OP_COMMIT,         /* 150 */
2016    
2017      /* These are forced failure and success verbs */
2018    
2019      OP_FAIL,           /* 151 */
2020      OP_ACCEPT,         /* 152 */
2021      OP_ASSERT_ACCEPT,  /* 153 Used inside assertions */
2022      OP_CLOSE,          /* 154 Used before OP_ACCEPT to close open captures */
2023    
2024      /* This is used to skip a subpattern with a {0} quantifier */
2025    
2026      OP_SKIPZERO,       /* 155 */
2027    
2028      /* This is not an opcode, but is used to check that tables indexed by opcode
2029      are the correct length, in order to catch updating errors - there have been
2030      some in the past. */
2031    
2032  /* WARNING WARNING WARNING: There is an implicit assumption in pcre.c and    OP_TABLE_LENGTH
2033  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. */  
2034    
2035  #define EXTRACT_BASIC_MAX  100  /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2036    definitions that follow must also be updated to match. There are also tables
2037    called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
2038    
2039    
2040  /* 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
2041  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
2042    only in pcre_printint.c, which fills out the full names in many cases (and in
2043    some cases doesn't actually use these names at all). */
2044    
2045  #define OP_NAME_LIST \  #define OP_NAME_LIST \
2046    "End", "\\A", "\\G", "\\B", "\\b", "\\D", "\\d",                \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
2047    "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte",                   \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
2048    "notprop", "prop", "extuni",                                    \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
2049    "\\Z", "\\z",                                                   \    "extuni",  "\\Z", "\\z",                                        \
2050    "Opt", "^", "$", "char", "charnc", "not",                       \    "^", "^", "$", "$", "char", "chari", "not", "noti",             \
2051    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2052    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "{", "{", "{",                                                  \
2053      "*+","++", "?+", "{",                                           \
2054      "*", "*?", "+", "+?", "?", "??",                                \
2055      "{", "{", "{",                                                  \
2056      "*+","++", "?+", "{",                                           \
2057      "*", "*?", "+", "+?", "?", "??",                                \
2058      "{", "{", "{",                                                  \
2059      "*+","++", "?+", "{",                                           \
2060      "*", "*?", "+", "+?", "?", "??",                                \
2061      "{", "{", "{",                                                  \
2062      "*+","++", "?+", "{",                                           \
2063    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
2064      "*+","++", "?+", "{",                                           \
2065    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \
2066    "class", "nclass", "xclass", "Ref", "Recurse", "Callout",       \    "class", "nclass", "xclass", "Ref", "Refi",                     \
2067    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",     \    "Recurse", "Callout",                                           \
2068    "AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cond ref",\    "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos",                  \
2069    "Brazero", "Braminzero", "Branumber", "Bra"    "Reverse", "Assert", "Assert not", "AssertB", "AssertB not",    \
2070      "Once", "Once_NC",                                              \
2071      "Bra", "BraPos", "CBra", "CBraPos",                             \
2072      "Cond",                                                         \
2073      "SBra", "SBraPos", "SCBra", "SCBraPos",                         \
2074      "SCond",                                                        \
2075      "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \
2076      "Brazero", "Braminzero", "Braposzero",                          \
2077      "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \
2078      "*THEN", "*THEN", "*COMMIT", "*FAIL",                           \
2079      "*ACCEPT", "*ASSERT_ACCEPT",                                    \
2080      "Close", "Skip zero"
2081    
2082    
2083  /* 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 634  in UTF-8 mode. The code that uses this t Line 2091  in UTF-8 mode. The code that uses this t
2091    
2092  #define OP_LENGTHS \  #define OP_LENGTHS \
2093    1,                             /* End                                    */ \    1,                             /* End                                    */ \
2094    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                     */ \
2095    1, 1,                          /* Any, Anybyte                           */ \    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
2096    2, 2, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \
2097    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    3, 3,                          /* \P, \p                                 */ \
2098      1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
2099      1,                             /* \X                                     */ \
2100      1, 1, 1, 1, 1, 1,              /* \Z, \z, ^, ^M, $, $M                   */ \
2101    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
2102    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Chari  - the minimum length            */ \
2103    2,                             /* not                                    */ \    2,                             /* not                                    */ \
2104    /* Positive single-char repeats                            ** These are  */ \    2,                             /* noti                                   */ \
2105    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??      ** minima in  */ \    /* Positive single-char repeats                             ** These are */ \
2106    4, 4, 4,                       /* upto, minupto, exact     ** UTF-8 mode */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2107      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2108      2+IMM2_SIZE,                   /* exact                                  */ \
2109      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2110      2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2111      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2112      2+IMM2_SIZE,                   /* exact I                                */ \
2113      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2114    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2115    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2116    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2117      2+IMM2_SIZE,                   /* NOT exact                              */ \
2118      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2119      2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2120      2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2121      2+IMM2_SIZE,                   /* NOT exact I                            */ \
2122      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2123    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2124    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2125    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2126      2+IMM2_SIZE,                   /* Type exact                             */ \
2127      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2128    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2129    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2130    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2131   33,                             /* CLASS                                  */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2132   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2133    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2134    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2135      1+IMM2_SIZE,                   /* REFI                                   */ \
2136    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2137    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2138    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
2139    1+LINK_SIZE,                   /* Ket                                    */ \    1+LINK_SIZE,                   /* Ket                                    */ \
2140    1+LINK_SIZE,                   /* KetRmax                                */ \    1+LINK_SIZE,                   /* KetRmax                                */ \
2141    1+LINK_SIZE,                   /* KetRmin                                */ \    1+LINK_SIZE,                   /* KetRmin                                */ \
2142      1+LINK_SIZE,                   /* KetRpos                                */ \
2143      1+LINK_SIZE,                   /* Reverse                                */ \
2144    1+LINK_SIZE,                   /* Assert                                 */ \    1+LINK_SIZE,                   /* Assert                                 */ \
2145    1+LINK_SIZE,                   /* Assert not                             */ \    1+LINK_SIZE,                   /* Assert not                             */ \
2146    1+LINK_SIZE,                   /* Assert behind                          */ \    1+LINK_SIZE,                   /* Assert behind                          */ \
2147    1+LINK_SIZE,                   /* Assert behind not                      */ \    1+LINK_SIZE,                   /* Assert behind not                      */ \
2148    1+LINK_SIZE,                   /* Reverse                                */ \    1+LINK_SIZE,                   /* ONCE                                   */ \
2149    1+LINK_SIZE,                   /* Once                                   */ \    1+LINK_SIZE,                   /* ONCE_NC                                */ \
2150      1+LINK_SIZE,                   /* BRA                                    */ \
2151      1+LINK_SIZE,                   /* BRAPOS                                 */ \
2152      1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2153      1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2154    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2155    3,                             /* CREF                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2156    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2157    3,                             /* BRANUMBER                              */ \    1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2158    1+LINK_SIZE                    /* BRA                                    */ \    1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2159      1+LINK_SIZE,                   /* SCOND                                  */ \
2160      1+IMM2_SIZE, 1+IMM2_SIZE,      /* CREF, NCREF                            */ \
2161  /* A magic value for OP_CREF to indicate the "in recursion" condition. */    1+IMM2_SIZE, 1+IMM2_SIZE,      /* RREF, NRREF                            */ \
2162      1,                             /* DEF                                    */ \
2163  #define CREF_RECURSE  0xffff    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2164      3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2165  /* Error code numbers. They are given names so that they can more easily be    1, 3,                          /* SKIP, SKIP_ARG                         */ \
2166  tracked. */    1, 3,                          /* THEN, THEN_ARG                         */ \
2167      1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2168      1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2169    
2170    /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
2171    condition. */
2172    
2173    #define RREF_ANY  0xffff
2174    
2175    /* Compile time error code numbers. They are given names so that they can more
2176    easily be tracked. When a new number is added, the table called eint in
2177    pcreposix.c must be updated. */
2178    
2179  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,
2180         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
2181         ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,         ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
2182         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2183         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47 };         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2184           ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2185           ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2186           ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2187    
2188    /* JIT compiling modes. The function list is indexed by them. */
2189    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2190           JIT_NUMBER_OF_COMPILE_MODES };
2191    
2192  /* 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
2193  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 703  Because people can now save and re-use c Line 2203  Because people can now save and re-use c
2203  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
2204  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
2205  fields are present. Currently PCRE always sets the dummy fields to zero.  fields are present. Currently PCRE always sets the dummy fields to zero.
2206  NOTE NOTE NOTE:  NOTE NOTE NOTE
2207  */  */
2208    
2209  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2210    #define REAL_PCRE real_pcre
2211    #elif defined COMPILE_PCRE16
2212    #define REAL_PCRE real_pcre16
2213    #elif defined COMPILE_PCRE32
2214    #define REAL_PCRE real_pcre32
2215    #endif
2216    
2217    /* It is necessary to fork the struct for 32 bit, since it needs to use
2218     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2219     * typedef since pcretest needs access to  the struct of the 8-, 16-
2220     * and 32-bit variants. */
2221    
2222    typedef struct real_pcre8_or_16 {
2223    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2224    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2225    pcre_uint32 options;    pcre_uint32 options;            /* Public options */
2226    pcre_uint32 dummy1;             /* For future use, maybe */    pcre_uint16 flags;              /* Private flags */
2227      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2228    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2229    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2230    pcre_uint16 first_byte;    pcre_uint16 first_char;         /* Starting character */
2231    pcre_uint16 req_byte;    pcre_uint16 req_char;           /* This character must be seen */
2232    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2233    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2234    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2235    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2236      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2237      const pcre_uint8 *nullpad;      /* NULL padding */
2238    } real_pcre8_or_16;
2239    
2240    typedef struct real_pcre8_or_16 real_pcre;
2241    typedef struct real_pcre8_or_16 real_pcre16;
2242    
2243    const unsigned char *tables;    /* Pointer to tables or NULL for std */  typedef struct real_pcre32 {
2244    const unsigned char *nullpad;   /* NULL padding */    pcre_uint32 magic_number;
2245  } real_pcre;    pcre_uint32 size;               /* Total that was malloced */
2246      pcre_uint32 options;            /* Public options */
2247      pcre_uint16 flags;              /* Private flags */
2248      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2249      pcre_uint16 top_bracket;        /* Highest numbered group */
2250      pcre_uint16 top_backref;        /* Highest numbered back reference */
2251      pcre_uint32 first_char;         /* Starting character */
2252      pcre_uint32 req_char;           /* This character must be seen */
2253      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2254      pcre_uint16 name_entry_size;    /* Size of any name items */
2255      pcre_uint16 name_count;         /* Number of name items */
2256      pcre_uint16 ref_count;          /* Reference count */
2257      pcre_uint16 dummy1;             /* for later expansion */
2258      pcre_uint16 dummy2;             /* for later expansion */
2259      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2260      void *nullpad;                  /* for later expansion */
2261    } real_pcre32;
2262    
2263    /* Assert that the size of REAL_PCRE is divisible by 8 */
2264    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2265    
2266    /* Needed in pcretest to access some fields in the real_pcre* structures
2267     * directly. They're unified for 8/16/32 bits since the structs only differ
2268     * after these fields; if that ever changes, need to fork those defines into
2269     * 8/16 and 32 bit versions. */
2270    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2271    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2272    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2273    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2274    
2275  /* 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
2276  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
2277    
2278  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2279    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2280    pcre_uint32 options;    pcre_uint32 flags;              /* Private flags */
2281    uschar start_bits[32];    pcre_uint8 start_bits[32];      /* Starting char bits */
2282      pcre_uint32 minlength;          /* Minimum subject length */
2283  } pcre_study_data;  } pcre_study_data;
2284    
2285    /* Structure for building a chain of open capturing subpatterns during
2286    compiling, so that instructions to close them can be compiled when (*ACCEPT) is
2287    encountered. This is also used to identify subpatterns that contain recursive
2288    back references to themselves, so that they can be made atomic. */
2289    
2290    typedef struct open_capitem {
2291      struct open_capitem *next;    /* Chain link */
2292      pcre_uint16 number;           /* Capture number */
2293      pcre_uint16 flag;             /* Set TRUE if recursive back ref */
2294    } open_capitem;
2295    
2296  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2297  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2298    
2299  typedef struct compile_data {  typedef struct compile_data {
2300    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2301    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2302    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2303    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2304    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_workspace;/* The start of working space */
2305    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_code;     /* The start of the compiled code */
2306    uschar *name_table;           /* The name/number table */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2307    int  names_found;             /* Number of entries so far */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2308    int  name_entry_size;         /* Size of each entry */    open_capitem *open_caps;          /* Chain of open capture items */
2309    int  top_backref;             /* Maximum back reference */    pcre_uchar *hwm;                  /* High watermark of workspace */
2310    unsigned int backref_map;     /* Bitmap of low back refs */    pcre_uchar *name_table;           /* The name/number table */
2311    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  names_found;                 /* Number of entries so far */
2312    BOOL nopartial;               /* Set TRUE if partial won't work */    int  name_entry_size;             /* Size of each entry */
2313      int  workspace_size;              /* Size of workspace */
2314      int  bracount;                    /* Count of capturing parens as we compile */
2315      int  final_bracount;              /* Saved value after first pass */
2316      int  max_lookbehind;              /* Maximum lookbehind (characters) */
2317      int  top_backref;                 /* Maximum back reference */
2318      unsigned int backref_map;         /* Bitmap of low back refs */
2319      int  assert_depth;                /* Depth of nested assertions */
2320      int  external_options;            /* External (initial) options */
2321      int  external_flags;              /* External flag bits to be set */
2322      int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2323      BOOL had_accept;                  /* (*ACCEPT) encountered */
2324      BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2325      BOOL check_lookbehind;            /* Lookbehinds need later checking */
2326      int  nltype;                      /* Newline type */
2327      int  nllen;                       /* Newline string length */
2328      pcre_uchar nl[4];                 /* Newline string when fixed length */
2329  } compile_data;  } compile_data;
2330    
2331  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
2332  branches, for testing for left recursion. */  branches, for testing for left recursion while compiling. */
2333    
2334  typedef struct branch_chain {  typedef struct branch_chain {
2335    struct branch_chain *outer;    struct branch_chain *outer;
2336    uschar *current;    pcre_uchar *current_branch;
2337  } branch_chain;  } branch_chain;
2338    
2339  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
2340  call within the pattern. */  call within the pattern; used by pcre_exec(). */
2341    
2342  typedef struct recursion_info {  typedef struct recursion_info {
2343    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2344    int group_num;                /* Number of group that was called */    int group_num;                  /* Number of group that was called */
2345    const uschar *after_call;     /* "Return value": points after the call in the expr */    int *offset_save;               /* Pointer to start of saved offsets */
2346    const uschar *save_start;     /* Old value of md->start_match */    int saved_max;                  /* Number of saved offsets */
2347    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 */  
2348  } recursion_info;  } recursion_info;
2349    
2350  /* When compiling in a mode that doesn't use recursive calls to match(),  /* A similar structure for pcre_dfa_exec(). */
2351  a structure is used to remember local variables on the heap. It is defined in  
2352  pcre.c, close to the match() function, so that it is easy to keep it in step  typedef struct dfa_recursion_info {
2353  with any changes of local variable. However, the pointer to the current frame    struct dfa_recursion_info *prevrec;
2354  must be saved in some "static" place over a longjmp(). We declare the    int group_num;
2355  structure here so that we can put a pointer in the match_data structure.    PCRE_PUCHAR subject_position;
2356  NOTE: This isn't used for a "normal" compilation of pcre. */  } dfa_recursion_info;
2357    
2358    /* Structure for building a chain of data for holding the values of the subject
2359    pointer at the start of each subpattern, so as to detect when an empty string
2360    has been matched by a subpattern - to break infinite loops; used by
2361    pcre_exec(). */
2362    
2363    typedef struct eptrblock {
2364      struct eptrblock *epb_prev;
2365      PCRE_PUCHAR epb_saved_eptr;
2366    } eptrblock;
2367    
 struct heapframe;  
2368    
2369  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2370  doing traditional NFA matching, so that they are thread-safe. */  doing traditional NFA matching, so that they are thread-safe. */
2371    
2372  typedef struct match_data {  typedef struct match_data {
2373    unsigned long int match_call_count; /* As it says */    unsigned long int match_call_count;      /* As it says */
2374    unsigned long int match_limit;/* As it says */    unsigned long int match_limit;           /* As it says */
2375    int   *offset_vector;         /* Offset vector */    unsigned long int match_limit_recursion; /* As it says */
2376    int    offset_end;            /* One past the end */    int   *offset_vector;           /* Offset vector */
2377    int    offset_max;            /* The maximum usable for return data */    int    offset_end;              /* One past the end */
2378    const uschar *lcc;            /* Points to lower casing table */    int    offset_max;              /* The maximum usable for return data */
2379    const uschar *ctypes;         /* Points to table of type maps */    int    nltype;                  /* Newline type */
2380    BOOL   offset_overflow;       /* Set if too many extractions */    int    nllen;                   /* Newline string length */
2381    BOOL   notbol;                /* NOTBOL flag */    int    name_count;              /* Number of names in name table */
2382    BOOL   noteol;                /* NOTEOL flag */    int    name_entry_size;         /* Size of entry in names table */
2383    BOOL   utf8;                  /* UTF8 flag */    pcre_uchar *name_table;         /* Table of names */
2384    BOOL   endonly;               /* Dollar not before final \n */    pcre_uchar nl[4];               /* Newline string when fixed */
2385    BOOL   notempty;              /* Empty string match not wanted */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2386    BOOL   partial;               /* PARTIAL flag */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2387    BOOL   hitend;                /* Hit the end of the subject at some point */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2388    const uschar *start_code;     /* For use when recursing */    BOOL   offset_overflow;         /* Set if too many extractions */
2389    const uschar *start_subject;  /* Start of the subject string */    BOOL   notbol;                  /* NOTBOL flag */
2390    const uschar *end_subject;    /* End of the subject string */    BOOL   noteol;                  /* NOTEOL flag */
2391    const uschar *start_match;    /* Start of this match attempt */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2392    const uschar *end_match_ptr;  /* Subject position at end match */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2393    int    end_offset_top;        /* Highwater mark at end of match */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2394    int    capture_last;          /* Most recent capture number */    BOOL   endonly;                 /* Dollar not before final \n */
2395    int    start_offset;          /* The start offset value */    BOOL   notempty;                /* Empty string match not wanted */
2396    recursion_info *recursive;    /* Linked list of recursion data */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2397    void  *callout_data;          /* To pass back to callouts */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2398    struct heapframe *thisframe;  /* Used only when compiling for no recursion */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2399      BOOL   hasthen;                 /* Pattern contains (*THEN) */
2400      BOOL   ignore_skip_arg;         /* For re-run when SKIP name not found */
2401      const  pcre_uchar *start_code;  /* For use when recursing */
2402      PCRE_PUCHAR start_subject;      /* Start of the subject string */
2403      PCRE_PUCHAR end_subject;        /* End of the subject string */
2404      PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2405      PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2406      PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2407      int    partial;                 /* PARTIAL options */
2408      int    end_offset_top;          /* Highwater mark at end of match */
2409      int    capture_last;            /* Most recent capture number */
2410      int    start_offset;            /* The start offset value */
2411      int    match_function_type;     /* Set for certain special calls of MATCH() */
2412      eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2413      int    eptrn;                   /* Next free eptrblock */
2414      recursion_info *recursive;      /* Linked list of recursion data */
2415      void  *callout_data;            /* To pass back to callouts */
2416      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2417      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2418      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2419    #ifdef NO_RECURSE
2420      void  *match_frames_base;       /* For remembering malloc'd frames */
2421    #endif
2422  } match_data;  } match_data;
2423    
2424  /* 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
2425  functions. */  functions. */
2426    
2427  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2428    const uschar *start_code;     /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2429    const uschar *start_subject;  /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2430    const uschar *end_subject;    /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2431    const uschar *tables;         /* Character tables */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2432    int   moptions;               /* Match options */    const pcre_uint8 *tables;         /* Character tables */
2433    int   poptions;               /* Pattern options */    int   start_offset;               /* The start offset value */
2434    void  *callout_data;          /* To pass back to callouts */    int   moptions;                   /* Match options */
2435      int   poptions;                   /* Pattern options */
2436      int   nltype;                     /* Newline type */
2437      int   nllen;                      /* Newline string length */
2438      pcre_uchar nl[4];                 /* Newline string when fixed */
2439      void *callout_data;               /* To pass back to callouts */
2440      dfa_recursion_info *recursive;    /* Linked list of recursion data */
2441  } dfa_match_data;  } dfa_match_data;
2442    
2443  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 834  typedef struct dfa_match_data { Line 2446  typedef struct dfa_match_data {
2446  #define ctype_letter  0x02  #define ctype_letter  0x02
2447  #define ctype_digit   0x04  #define ctype_digit   0x04
2448  #define ctype_xdigit  0x08  #define ctype_xdigit  0x08
2449  #define ctype_word    0x10   /* alphameric or '_' */  #define ctype_word    0x10   /* alphanumeric or '_' */
2450  #define ctype_meta    0x80   /* regexp meta char or zero (end pattern) */  #define ctype_meta    0x80   /* regexp meta char or zero (end pattern) */
2451    
2452  /* 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 861  total length. */ Line 2473  total length. */
2473  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2474  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2475    
2476  /* Layout of the UCP type table that translates property names into codes for  /* Internal function and data prefixes. */
2477  ucp_findchar(). */  
2478    #if defined COMPILE_PCRE8
2479    #ifndef PUBL
2480    #define PUBL(name) pcre_##name
2481    #endif
2482    #ifndef PRIV
2483    #define PRIV(name) _pcre_##name
2484    #endif
2485    #elif defined COMPILE_PCRE16
2486    #ifndef PUBL
2487    #define PUBL(name) pcre16_##name
2488    #endif
2489    #ifndef PRIV
2490    #define PRIV(name) _pcre16_##name
2491    #endif
2492    #elif defined COMPILE_PCRE32
2493    #ifndef PUBL
2494    #define PUBL(name) pcre32_##name
2495    #endif
2496    #ifndef PRIV
2497    #define PRIV(name) _pcre32_##name
2498    #endif
2499    #else
2500    #error Unsupported compiling mode
2501    #endif /* COMPILE_PCRE[8|16|32] */
2502    
2503    /* Layout of the UCP type table that translates property names into types and
2504    codes. Each entry used to point directly to a name, but to reduce the number of
2505    relocations in shared libraries, it now has an offset into a single string
2506    instead. */
2507    
2508  typedef struct {  typedef struct {
2509    const char *name;    pcre_uint16 name_offset;
2510    int value;    pcre_uint16 type;
2511      pcre_uint16 value;
2512  } ucp_type_table;  } ucp_type_table;
2513    
2514    
# Line 875  of the exported public functions. They h Line 2517  of the exported public functions. They h
2517  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
2518  pcre_tables.c module. */  pcre_tables.c module. */
2519    
2520  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2521  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2522  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2523  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2524    extern const int            PRIV(utf8_table3)[];
2525    extern const pcre_uint8     PRIV(utf8_table4)[];
2526    #endif /* COMPILE_PCRE8 */
2527    
2528    extern const char           PRIV(utt_names)[];
2529    extern const ucp_type_table PRIV(utt)[];
2530    extern const int            PRIV(utt_size);
2531    
2532  extern const int    _pcre_utf8_table1_size;  extern const pcre_uint8     PRIV(OP_lengths)[];
2533    extern const pcre_uint8     PRIV(default_tables)[];
2534    
2535  extern const ucp_type_table _pcre_utt[];  extern const pcre_uint32    PRIV(hspace_list)[];
2536  extern const int _pcre_utt_size;  extern const pcre_uint32    PRIV(vspace_list)[];
   
 extern const uschar _pcre_default_tables[];  
   
 extern const uschar _pcre_OP_lengths[];  
2537    
2538    
2539  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2540  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
2541  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2542    
2543  extern int         _pcre_ord2utf8(int, uschar *);  /* String comparison functions. */
2544  extern void        _pcre_printint(pcre *, FILE *);  #if defined COMPILE_PCRE8
2545  extern real_pcre * _pcre_try_flipped(const real_pcre *, real_pcre *,  
2546                       const pcre_study_data *, pcre_study_data *);  #define STRCMP_UC_UC(str1, str2) \
2547  extern int         _pcre_ucp_findchar(const int, int *, int *);    strcmp((char *)(str1), (char *)(str2))
2548  extern int         _pcre_valid_utf8(const uschar *, int);  #define STRCMP_UC_C8(str1, str2) \
2549  extern BOOL        _pcre_xclass(int, const uschar *);    strcmp((char *)(str1), (str2))
2550    #define STRNCMP_UC_UC(str1, str2, num) \
2551      strncmp((char *)(str1), (char *)(str2), (num))
2552    #define STRNCMP_UC_C8(str1, str2, num) \
2553      strncmp((char *)(str1), (str2), (num))
2554    #define STRLEN_UC(str) strlen((const char *)str)
2555    
2556    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2557    
2558    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2559                               const pcre_uchar *);
2560    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2561                               const char *);
2562    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2563                               const pcre_uchar *, unsigned int num);
2564    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2565                               const char *, unsigned int num);
2566    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2567    
2568    #define STRCMP_UC_UC(str1, str2) \
2569      PRIV(strcmp_uc_uc)((str1), (str2))
2570    #define STRCMP_UC_C8(str1, str2) \
2571      PRIV(strcmp_uc_c8)((str1), (str2))
2572    #define STRNCMP_UC_UC(str1, str2, num) \
2573      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2574    #define STRNCMP_UC_C8(str1, str2, num) \
2575      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2576    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2577    
2578    #endif /* COMPILE_PCRE[8|16|32] */
2579    
2580    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2581    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2582                               int *, BOOL);
2583    extern int               PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2584    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2585    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2586                               int *, BOOL);
2587    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2588    
2589    #ifdef SUPPORT_JIT
2590    extern void              PRIV(jit_compile)(const REAL_PCRE *,
2591                               PUBL(extra) *, int);
2592    extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2593                               const pcre_uchar *, int, int, int, int *, int);
2594    extern void              PRIV(jit_free)(void *);
2595    extern int               PRIV(jit_get_size)(void *);
2596    extern const char*       PRIV(jit_get_target)(void);
2597    #endif
2598    
2599    /* Unicode character database (UCD) */
2600    
2601    typedef struct {
2602      pcre_uint8 script;     /* ucp_Arabic, etc. */
2603      pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2604      pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2605      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2606      pcre_int32 other_case; /* offset to other case, or zero if none */
2607    } ucd_record;
2608    
2609    extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2610    extern const ucd_record  PRIV(ucd_records)[];
2611    extern const pcre_uint8  PRIV(ucd_stage1)[];
2612    extern const pcre_uint16 PRIV(ucd_stage2)[];
2613    extern const int         PRIV(ucp_gentype)[];
2614    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2615    #ifdef SUPPORT_JIT
2616    extern const int         PRIV(ucp_typerange)[];
2617    #endif
2618    
2619    #ifdef SUPPORT_UCP
2620    /* UCD access macros */
2621    
2622    #define UCD_BLOCK_SIZE 128
2623    #define GET_UCD(ch) (PRIV(ucd_records) + \
2624            PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2625            UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2626    
2627    #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2628    #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2629    #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2630    #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2631    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2632    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2633    
2634    #endif /* SUPPORT_UCP */
2635    
2636    #endif
2637    
2638  /* End of pcre_internal.h */  /* End of pcre_internal.h */

Legend:
Removed from v.77  
changed lines
  Added in v.1083

  ViewVC Help
Powered by ViewVC 1.1.5