/[pcre]/code/trunk/pcre_internal.h
ViewVC logotype

Diff of /code/trunk/pcre_internal.h

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 87 by nigel, Sat Feb 24 21:41:21 2007 UTC revision 1320 by ph10, Wed May 1 16:39:35 2013 UTC
# Line 7  Line 7 
7  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2006 University of Cambridge             Copyright (c) 1997-2013 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 40  POSSIBILITY OF SUCH DAMAGE. Line 40  POSSIBILITY OF SUCH DAMAGE.
40    
41  /* This header contains definitions that are shared between the different  /* This header contains definitions that are shared between the different
42  modules, but which are not relevant to the exported API. This includes some  modules, but which are not relevant to the exported API. This includes some
43  functions whose names all begin with "_pcre_". */  functions whose names all begin with "_pcre_", "_pcre16_" or "_pcre32_"
44    depending on the PRIV macro. */
45    
46  #ifndef PCRE_INTERNAL_H  #ifndef PCRE_INTERNAL_H
47  #define PCRE_INTERNAL_H  #define PCRE_INTERNAL_H
48    
49  /* Define DEBUG to get debugging output on stdout. */  /* Define PCRE_DEBUG to get debugging output on stdout. */
50    
51  #if 0  #if 0
52  #define DEBUG  #define PCRE_DEBUG
53    #endif
54    
55    /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
56    
57    #if !defined COMPILE_PCRE16 && !defined COMPILE_PCRE32
58    #define COMPILE_PCRE8
59    #endif
60    
61    /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
62    "configure" script ensures this, but not everybody uses "configure". */
63    
64    #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
65    #define SUPPORT_UTF 1
66    #endif
67    
68    /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
69    reasons with existing code. */
70    
71    #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
72    #define SUPPORT_UTF 1
73    #endif
74    
75    /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
76    Until then we define it if SUPPORT_UTF is defined. */
77    
78    #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
79    #define SUPPORT_UTF8 1
80    #endif
81    
82    /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
83    script prevents both being selected, but not everybody uses "configure". */
84    
85    #if defined EBCDIC && defined SUPPORT_UTF
86    #error The use of both EBCDIC and SUPPORT_UTF is not supported.
87  #endif  #endif
88    
89  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
90  inline, and there are *still* stupid compilers about that don't like indented  inline, and there are *still* stupid compilers about that don't like indented
91  pre-processor statements, or at least there were when I first wrote this. After  pre-processor statements, or at least there were when I first wrote this. After
92  all, it had only been about 10 years then... */  all, it had only been about 10 years then...
93    
94  #ifdef DEBUG  It turns out that the Mac Debugging.h header also defines the macro DPRINTF, so
95    be absolutely sure we get our version. */
96    
97    #undef DPRINTF
98    #ifdef PCRE_DEBUG
99  #define DPRINTF(p) printf p  #define DPRINTF(p) printf p
100  #else  #else
101  #define DPRINTF(p) /*nothing*/  #define DPRINTF(p) /* Nothing */
102  #endif  #endif
103    
104    
 /* Get the definitions provided by running "configure" */  
   
 #include "config.h"  
   
105  /* Standard C headers plus the external interface definition. The only time  /* Standard C headers plus the external interface definition. The only time
106  setjmp and stdarg are used is when NO_RECURSE is set. */  setjmp and stdarg are used is when NO_RECURSE is set. */
107    
108  #include <ctype.h>  #include <ctype.h>
109  #include <limits.h>  #include <limits.h>
 #include <setjmp.h>  
 #include <stdarg.h>  
110  #include <stddef.h>  #include <stddef.h>
111  #include <stdio.h>  #include <stdio.h>
112  #include <stdlib.h>  #include <stdlib.h>
113  #include <string.h>  #include <string.h>
114    
115  #ifndef PCRE_SPY  /* Valgrind (memcheck) support */
116  #define PCRE_DEFINITION       /* Win32 __declspec(export) trigger for .dll */  
117    #ifdef SUPPORT_VALGRIND
118    #include <valgrind/memcheck.h>
119  #endif  #endif
120    
121  /* We need to have types that specify unsigned 16-bit and 32-bit integers. We  /* When compiling a DLL for Windows, the exported symbols have to be declared
122    using some MS magic. I found some useful information on this web page:
123    http://msdn2.microsoft.com/en-us/library/y4h7bcy6(VS.80).aspx. According to the
124    information there, using __declspec(dllexport) without "extern" we have a
125    definition; with "extern" we have a declaration. The settings here override the
126    setting in pcre.h (which is included below); it defines only PCRE_EXP_DECL,
127    which is all that is needed for applications (they just import the symbols). We
128    use:
129    
130      PCRE_EXP_DECL       for declarations
131      PCRE_EXP_DEFN       for definitions of exported functions
132      PCRE_EXP_DATA_DEFN  for definitions of exported variables
133    
134    The reason for the two DEFN macros is that in non-Windows environments, one
135    does not want to have "extern" before variable definitions because it leads to
136    compiler warnings. So we distinguish between functions and variables. In
137    Windows, the two should always be the same.
138    
139    The reason for wrapping this in #ifndef PCRE_EXP_DECL is so that pcretest,
140    which is an application, but needs to import this file in order to "peek" at
141    internals, can #include pcre.h first to get an application's-eye view.
142    
143    In principle, people compiling for non-Windows, non-Unix-like (i.e. uncommon,
144    special-purpose environments) might want to stick other stuff in front of
145    exported symbols. That's why, in the non-Windows case, we set PCRE_EXP_DEFN and
146    PCRE_EXP_DATA_DEFN only if they are not already set. */
147    
148    #ifndef PCRE_EXP_DECL
149    #  ifdef _WIN32
150    #    ifndef PCRE_STATIC
151    #      define PCRE_EXP_DECL       extern __declspec(dllexport)
152    #      define PCRE_EXP_DEFN       __declspec(dllexport)
153    #      define PCRE_EXP_DATA_DEFN  __declspec(dllexport)
154    #    else
155    #      define PCRE_EXP_DECL       extern
156    #      define PCRE_EXP_DEFN
157    #      define PCRE_EXP_DATA_DEFN
158    #    endif
159    #  else
160    #    ifdef __cplusplus
161    #      define PCRE_EXP_DECL       extern "C"
162    #    else
163    #      define PCRE_EXP_DECL       extern
164    #    endif
165    #    ifndef PCRE_EXP_DEFN
166    #      define PCRE_EXP_DEFN       PCRE_EXP_DECL
167    #    endif
168    #    ifndef PCRE_EXP_DATA_DEFN
169    #      define PCRE_EXP_DATA_DEFN
170    #    endif
171    #  endif
172    #endif
173    
174    /* When compiling with the MSVC compiler, it is sometimes necessary to include
175    a "calling convention" before exported function names. (This is secondhand
176    information; I know nothing about MSVC myself). For example, something like
177    
178      void __cdecl function(....)
179    
180    might be needed. In order so make this easy, all the exported functions have
181    PCRE_CALL_CONVENTION just before their names. It is rarely needed; if not
182    set, we ensure here that it has no effect. */
183    
184    #ifndef PCRE_CALL_CONVENTION
185    #define PCRE_CALL_CONVENTION
186    #endif
187    
188    /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
189  cannot determine these outside the compilation (e.g. by running a program as  cannot determine these outside the compilation (e.g. by running a program as
190  part of "configure") because PCRE is often cross-compiled for use on other  part of "configure") because PCRE is often cross-compiled for use on other
191  systems. Instead we make use of the maximum sizes that are available at  systems. Instead we make use of the maximum sizes that are available at
192  preprocessor time in standard C environments. */  preprocessor time in standard C environments. */
193    
194    typedef unsigned char pcre_uint8;
195    
196  #if USHRT_MAX == 65535  #if USHRT_MAX == 65535
197    typedef unsigned short pcre_uint16;  typedef unsigned short pcre_uint16;
198    typedef short pcre_int16;
199    #define PCRE_UINT16_MAX USHRT_MAX
200    #define PCRE_INT16_MAX SHRT_MAX
201  #elif UINT_MAX == 65535  #elif UINT_MAX == 65535
202    typedef unsigned int pcre_uint16;  typedef unsigned int pcre_uint16;
203    typedef int pcre_int16;
204    #define PCRE_UINT16_MAX UINT_MAX
205    #define PCRE_INT16_MAX INT_MAX
206  #else  #else
207    #error Cannot determine a type for 16-bit unsigned integers  #error Cannot determine a type for 16-bit integers
208    #endif
209    
210    #if UINT_MAX == 4294967295U
211    typedef unsigned int pcre_uint32;
212    typedef int pcre_int32;
213    #define PCRE_UINT32_MAX UINT_MAX
214    #define PCRE_INT32_MAX INT_MAX
215    #elif ULONG_MAX == 4294967295UL
216    typedef unsigned long int pcre_uint32;
217    typedef long int pcre_int32;
218    #define PCRE_UINT32_MAX ULONG_MAX
219    #define PCRE_INT32_MAX LONG_MAX
220    #else
221    #error Cannot determine a type for 32-bit integers
222    #endif
223    
224    /* When checking for integer overflow in pcre_compile(), we need to handle
225    large integers. If a 64-bit integer type is available, we can use that.
226    Otherwise we have to cast to double, which of course requires floating point
227    arithmetic. Handle this by defining a macro for the appropriate type. If
228    stdint.h is available, include it; it may define INT64_MAX. Systems that do not
229    have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
230    by "configure". */
231    
232    #if defined HAVE_STDINT_H
233    #include <stdint.h>
234    #elif defined HAVE_INTTYPES_H
235    #include <inttypes.h>
236  #endif  #endif
237    
238  #if UINT_MAX == 4294967295  #if defined INT64_MAX || defined int64_t
239    typedef unsigned int pcre_uint32;  #define INT64_OR_DOUBLE int64_t
 #elif ULONG_MAX == 4294967295  
   typedef unsigned long int pcre_uint32;  
240  #else  #else
241    #error Cannot determine a type for 32-bit unsigned integers  #define INT64_OR_DOUBLE double
242  #endif  #endif
243    
244  /* All character handling must be done as unsigned characters. Otherwise there  /* All character handling must be done as unsigned characters. Otherwise there
245  are problems with top-bit-set characters and functions such as isspace().  are problems with top-bit-set characters and functions such as isspace().
246  However, we leave the interface to the outside world as char *, because that  However, we leave the interface to the outside world as char * or short *,
247  should make things easier for callers. We define a short type for unsigned char  because that should make things easier for callers. This character type is
248  to save lots of typing. I tried "uchar", but it causes problems on Digital  called pcre_uchar.
249  Unix, where it is defined in sys/types, so use "uschar" instead. */  
250    The IN_UCHARS macro multiply its argument with the byte size of the current
251    pcre_uchar type. Useful for memcpy and such operations, whose require the
252    byte size of their input/output buffers.
253    
254    The MAX_255 macro checks whether its pcre_uchar input is less than 256.
255    
256    The TABLE_GET macro is designed for accessing elements of tables whose contain
257    exactly 256 items. When the character is able to contain more than 256
258    items, some check is needed before accessing these tables.
259    */
260    
261    #if defined COMPILE_PCRE8
262    
263  typedef unsigned char uschar;  typedef unsigned char pcre_uchar;
264    #define IN_UCHARS(x) (x)
265    #define MAX_255(c) 1
266    #define TABLE_GET(c, table, default) ((table)[c])
267    
268    #elif defined COMPILE_PCRE16
269    
270    #if USHRT_MAX != 65535
271    /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
272    pcre.h(.in) and disable (comment out) this message. */
273    #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
274    #endif
275    
276    typedef pcre_uint16 pcre_uchar;
277    #define UCHAR_SHIFT (1)
278    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
279    #define MAX_255(c) ((c) <= 255u)
280    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
281    
282    #elif defined COMPILE_PCRE32
283    
284    typedef pcre_uint32 pcre_uchar;
285    #define UCHAR_SHIFT (2)
286    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
287    #define MAX_255(c) ((c) <= 255u)
288    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
289    
290    #else
291    #error Unsupported compiling mode
292    #endif /* COMPILE_PCRE[8|16|32] */
293    
294    /* This is an unsigned int value that no character can ever have. UTF-8
295    characters only go up to 0x7fffffff (though Unicode doesn't go beyond
296    0x0010ffff). */
297    
298    #define NOTACHAR 0xffffffff
299    
300    /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
301    "any" and "anycrlf" at present). The following macros are used to package up
302    testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
303    modules to indicate in which datablock the parameters exist, and what the
304    start/end of string field names are. */
305    
306    #define NLTYPE_FIXED    0     /* Newline is a fixed length string */
307    #define NLTYPE_ANY      1     /* Newline is any Unicode line ending */
308    #define NLTYPE_ANYCRLF  2     /* Newline is CR, LF, or CRLF */
309    
310    /* This macro checks for a newline at the given position */
311    
312    #define IS_NEWLINE(p) \
313      ((NLBLOCK->nltype != NLTYPE_FIXED)? \
314        ((p) < NLBLOCK->PSEND && \
315         PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
316           &(NLBLOCK->nllen), utf)) \
317        : \
318        ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
319         RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
320         (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
321        ) \
322      )
323    
324    /* This macro checks for a newline immediately preceding the given position */
325    
326    #define WAS_NEWLINE(p) \
327      ((NLBLOCK->nltype != NLTYPE_FIXED)? \
328        ((p) > NLBLOCK->PSSTART && \
329         PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
330           &(NLBLOCK->nllen), utf)) \
331        : \
332        ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
333         RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
334         (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
335        ) \
336      )
337    
338  /* When PCRE is compiled as a C++ library, the subject pointer can be replaced  /* When PCRE is compiled as a C++ library, the subject pointer can be replaced
339  with a custom type. This makes it possible, for example, to allow pcre_exec()  with a custom type. This makes it possible, for example, to allow pcre_exec()
# Line 124  used for the external interface and appe Line 345  used for the external interface and appe
345  must begin with PCRE_. */  must begin with PCRE_. */
346    
347  #ifdef CUSTOM_SUBJECT_PTR  #ifdef CUSTOM_SUBJECT_PTR
348  #define PCRE_SPTR CUSTOM_SUBJECT_PTR  #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
 #define USPTR CUSTOM_SUBJECT_PTR  
349  #else  #else
350  #define PCRE_SPTR const char *  #define PCRE_PUCHAR const pcre_uchar *
 #define USPTR const unsigned char *  
351  #endif  #endif
352    
353  /* Include the public PCRE header and the definitions of UCP character property  /* Include the public PCRE header and the definitions of UCP character property
# Line 137  values. */ Line 356  values. */
356  #include "pcre.h"  #include "pcre.h"
357  #include "ucp.h"  #include "ucp.h"
358    
359    #ifdef COMPILE_PCRE32
360    /* Assert that the public PCRE_UCHAR32 is a 32-bit type */
361    typedef int __assert_pcre_uchar32_size[sizeof(PCRE_UCHAR32) == 4 ? 1 : -1];
362    #endif
363    
364  /* When compiling for use with the Virtual Pascal compiler, these functions  /* When compiling for use with the Virtual Pascal compiler, these functions
365  need to have their names changed. PCRE must be compiled with the -DVPCOMPAT  need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
366  option on the command line. */  option on the command line. */
367    
368  #ifdef VPCOMPAT  #ifdef VPCOMPAT
369    #define strlen(s)        _strlen(s)
370  #define strncmp(s1,s2,m) _strncmp(s1,s2,m)  #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
371    #define memcmp(s,c,n)    _memcmp(s,c,n)
372  #define memcpy(d,s,n)    _memcpy(d,s,n)  #define memcpy(d,s,n)    _memcpy(d,s,n)
373  #define memmove(d,s,n)   _memmove(d,s,n)  #define memmove(d,s,n)   _memmove(d,s,n)
374  #define memset(s,c,n)    _memset(s,c,n)  #define memset(s,c,n)    _memset(s,c,n)
# Line 151  option on the command line. */ Line 377  option on the command line. */
377  /* 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(),
378  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
379  is set. Otherwise, include an emulating function for those systems that have  is set. Otherwise, include an emulating function for those systems that have
380  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. */  
381    
382  #if ! HAVE_MEMMOVE  #ifndef HAVE_MEMMOVE
383  #undef  memmove        /* some systems may have a macro */  #undef  memmove        /* some systems may have a macro */
384  #if HAVE_BCOPY  #ifdef HAVE_BCOPY
385  #define memmove(a, b, c) bcopy(b, a, c)  #define memmove(a, b, c) bcopy(b, a, c)
386  #else  /* HAVE_BCOPY */  #else  /* HAVE_BCOPY */
387  void *  static void *
388  pcre_memmove(unsigned char *dest, const unsigned char *src, size_t n)  pcre_memmove(void *d, const void *s, size_t n)
389  {  {
390  size_t i;  size_t i;
391  dest += n;  unsigned char *dest = (unsigned char *)d;
392  src += n;  const unsigned char *src = (const unsigned char *)s;
393  for (i = 0; i < n; ++i) *(--dest) =  *(--src);  if (dest > src)
394  return dest;    {
395      dest += n;
396      src += n;
397      for (i = 0; i < n; ++i) *(--dest) = *(--src);
398      return (void *)dest;
399      }
400    else
401      {
402      for (i = 0; i < n; ++i) *dest++ = *src++;
403      return (void *)(dest - n);
404      }
405  }  }
406  #define memmove(a, b, c) pcre_memmove(a, b, c)  #define memmove(a, b, c) pcre_memmove(a, b, c)
407  #endif   /* not HAVE_BCOPY */  #endif   /* not HAVE_BCOPY */
# Line 188  The macros are controlled by the value o Line 422  The macros are controlled by the value o
422  the config.h file, but can be overridden by using -D on the command line. This  the config.h file, but can be overridden by using -D on the command line. This
423  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
424    
425    #if defined COMPILE_PCRE8
426    
427  #if LINK_SIZE == 2  #if LINK_SIZE == 2
428    
429  #define PUT(a,n,d)   \  #define PUT(a,n,d)   \
# Line 224  is automated on Unix systems via the "co Line 460  is automated on Unix systems via the "co
460  #define GET(a,n) \  #define GET(a,n) \
461    (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])    (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
462    
463  #define MAX_PATTERN_SIZE (1 << 30)   /* Keep it positive */  /* Keep it positive */
464    #define MAX_PATTERN_SIZE (1 << 30)
465    
466    #else
467    #error LINK_SIZE must be either 2, 3, or 4
468    #endif
469    
470    #elif defined COMPILE_PCRE16
471    
472    #if LINK_SIZE == 2
473    
474    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
475    #undef LINK_SIZE
476    #define LINK_SIZE 1
477    
478    #define PUT(a,n,d)   \
479      (a[n] = (d))
480    
481    #define GET(a,n) \
482      (a[n])
483    
484    #define MAX_PATTERN_SIZE (1 << 16)
485    
486    #elif LINK_SIZE == 3 || LINK_SIZE == 4
487    
488    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
489    #undef LINK_SIZE
490    #define LINK_SIZE 2
491    
492    #define PUT(a,n,d)   \
493      (a[n] = (d) >> 16), \
494      (a[(n)+1] = (d) & 65535)
495    
496    #define GET(a,n) \
497      (((a)[n] << 16) | (a)[(n)+1])
498    
499    /* Keep it positive */
500    #define MAX_PATTERN_SIZE (1 << 30)
501    
502  #else  #else
503  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
504  #endif  #endif
505    
506    #elif defined COMPILE_PCRE32
507    
508    /* Only supported LINK_SIZE is 4 */
509    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
510    #undef LINK_SIZE
511    #define LINK_SIZE 1
512    
513    #define PUT(a,n,d)   \
514      (a[n] = (d))
515    
516    #define GET(a,n) \
517      (a[n])
518    
519    /* Keep it positive */
520    #define MAX_PATTERN_SIZE (1 << 30)
521    
522    #else
523    #error Unsupported compiling mode
524    #endif /* COMPILE_PCRE[8|16|32] */
525    
526  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
527    
# Line 241  is automated on Unix systems via the "co Line 532  is automated on Unix systems via the "co
532  offsets changes. There are used for repeat counts and for other things such as  offsets changes. There are used for repeat counts and for other things such as
533  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
534    
535    #if defined COMPILE_PCRE8
536    
537    #define IMM2_SIZE 2
538    
539  #define PUT2(a,n,d)   \  #define PUT2(a,n,d)   \
540    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
541    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
542    
543    /* For reasons that I do not understand, the expression in this GET2 macro is
544    treated by gcc as a signed expression, even when a is declared as unsigned. It
545    seems that any kind of arithmetic results in a signed value. */
546    
547  #define GET2(a,n) \  #define GET2(a,n) \
548    (((a)[n] << 8) | (a)[(n)+1])    (unsigned int)(((a)[n] << 8) | (a)[(n)+1])
549    
550  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += 2  #elif defined COMPILE_PCRE16
551    
552    #define IMM2_SIZE 1
553    
554  /* When UTF-8 encoding is being used, a character is no longer just a single  #define PUT2(a,n,d)   \
555  byte. The macros for character handling generate simple sequences when used in     a[n] = d
556  byte-mode, and more complicated ones for UTF-8 characters. */  
557    #define GET2(a,n) \
558       a[n]
559    
560    #elif defined COMPILE_PCRE32
561    
562  #ifndef SUPPORT_UTF8  #define IMM2_SIZE 1
563    
564    #define PUT2(a,n,d)   \
565       a[n] = d
566    
567    #define GET2(a,n) \
568       a[n]
569    
570    #else
571    #error Unsupported compiling mode
572    #endif /* COMPILE_PCRE[8|16|32] */
573    
574    #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
575    
576    /* The maximum length of a MARK name is currently one data unit; it may be
577    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
578    
579    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
580    #define MAX_MARK ((1u << 16) - 1)
581    #else
582    #define MAX_MARK ((1u << 8) - 1)
583    #endif
584    
585    /* When UTF encoding is being used, a character is no longer just a single
586    byte. The macros for character handling generate simple sequences when used in
587    character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
588    and other macros are not used when UTF is not supported, so they are not
589    defined. To make sure they can never even appear when UTF support is omitted,
590    we don't even define them. */
591    
592    #ifndef SUPPORT_UTF
593    
594    /* #define MAX_VALUE_FOR_SINGLE_CHAR */
595    /* #define HAS_EXTRALEN(c) */
596    /* #define GET_EXTRALEN(c) */
597    /* #define NOT_FIRSTCHAR(c) */
598  #define GETCHAR(c, eptr) c = *eptr;  #define GETCHAR(c, eptr) c = *eptr;
599  #define GETCHARTEST(c, eptr) c = *eptr;  #define GETCHARTEST(c, eptr) c = *eptr;
600  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
601  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
602  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
603  #define BACKCHAR(eptr)  #define RAWUCHAR(eptr) (*(eptr))
604    #define RAWUCHARINC(eptr) (*(eptr)++)
605    #define RAWUCHARTEST(eptr) (*(eptr))
606    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
607    /* #define GETCHARLENTEST(c, eptr, len) */
608    /* #define BACKCHAR(eptr) */
609    /* #define FORWARDCHAR(eptr) */
610    /* #define ACROSSCHAR(condition, eptr, action) */
611    
612  #else   /* SUPPORT_UTF8 */  #else   /* SUPPORT_UTF */
613    
614  /* Get the next UTF-8 character, not advancing the pointer. This is called when  /* Tests whether the code point needs extra characters to decode. */
 we know we are in UTF-8 mode. */  
615    
616  #define GETCHAR(c, eptr) \  #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
617    c = *eptr; \  
618    if ((c & 0xc0) == 0xc0) \  /* Base macro to pick up the remaining bytes of a UTF-8 character, not
619    advancing the pointer. */
620    
621    #define GETUTF8(c, eptr) \
622      { \      { \
623      int gcii; \      if ((c & 0x20) == 0) \
624      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
625      int gcss = 6*gcaa; \      else if ((c & 0x10) == 0) \
626      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
627      for (gcii = 1; gcii <= gcaa; gcii++) \      else if ((c & 0x08) == 0) \
628          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
629          ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
630        else if ((c & 0x04) == 0) \
631          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
632              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
633              (eptr[4] & 0x3f); \
634        else \
635          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
636              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
637              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
638        }
639    
640    /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
641    the pointer. */
642    
643    #define GETUTF8INC(c, eptr) \
644        { \
645        if ((c & 0x20) == 0) \
646          c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
647        else if ((c & 0x10) == 0) \
648          { \
649          c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
650          eptr += 2; \
651          } \
652        else if ((c & 0x08) == 0) \
653        { \        { \
654        gcss -= 6; \        c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
655        c |= (eptr[gcii] & 0x3f) << gcss; \            ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
656          eptr += 3; \
657          } \
658        else if ((c & 0x04) == 0) \
659          { \
660          c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
661              ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
662              (eptr[3] & 0x3f); \
663          eptr += 4; \
664          } \
665        else \
666          { \
667          c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
668              ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
669              ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
670          eptr += 5; \
671        } \        } \
672      }      }
673    
674    #if defined COMPILE_PCRE8
675    
676    /* These macros were originally written in the form of loops that used data
677    from the tables whose names start with PRIV(utf8_table). They were rewritten by
678    a user so as not to use loops, because in some environments this gives a
679    significant performance advantage, and it seems never to do any harm. */
680    
681    /* Tells the biggest code point which can be encoded as a single character. */
682    
683    #define MAX_VALUE_FOR_SINGLE_CHAR 127
684    
685    /* Tests whether the code point needs extra characters to decode. */
686    
687    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
688    
689    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
690    Otherwise it has an undefined behaviour. */
691    
692    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
693    
694    /* Returns TRUE, if the given character is not the first character
695    of a UTF sequence. */
696    
697    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
698    
699    /* Get the next UTF-8 character, not advancing the pointer. This is called when
700    we know we are in UTF-8 mode. */
701    
702    #define GETCHAR(c, eptr) \
703      c = *eptr; \
704      if (c >= 0xc0) GETUTF8(c, eptr);
705    
706  /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the  /* Get the next UTF-8 character, testing for UTF-8 mode, and not advancing the
707  pointer. */  pointer. */
708    
709  #define GETCHARTEST(c, eptr) \  #define GETCHARTEST(c, eptr) \
710    c = *eptr; \    c = *eptr; \
711    if (utf8 && (c & 0xc0) == 0xc0) \    if (utf && c >= 0xc0) GETUTF8(c, eptr);
     { \  
     int gcii; \  
     int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     for (gcii = 1; gcii <= gcaa; gcii++) \  
       { \  
       gcss -= 6; \  
       c |= (eptr[gcii] & 0x3f) << gcss; \  
       } \  
     }  
712    
713  /* Get the next UTF-8 character, advancing the pointer. This is called when we  /* Get the next UTF-8 character, advancing the pointer. This is called when we
714  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
715    
716  #define GETCHARINC(c, eptr) \  #define GETCHARINC(c, eptr) \
717    c = *eptr++; \    c = *eptr++; \
718    if ((c & 0xc0) == 0xc0) \    if (c >= 0xc0) GETUTF8INC(c, eptr);
     { \  
     int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     while (gcaa-- > 0) \  
       { \  
       gcss -= 6; \  
       c |= (*eptr++ & 0x3f) << gcss; \  
       } \  
     }  
719    
720  /* Get the next character, testing for UTF-8 mode, and advancing the pointer */  /* Get the next character, testing for UTF-8 mode, and advancing the pointer.
721    This is called when we don't know if we are in UTF-8 mode. */
722    
723  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
724    c = *eptr++; \    c = *eptr++; \
725    if (utf8 && (c & 0xc0) == 0xc0) \    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
726    
727    /* Base macro to pick up the remaining bytes of a UTF-8 character, not
728    advancing the pointer, incrementing the length. */
729    
730    #define GETUTF8LEN(c, eptr, len) \
731      { \      { \
732      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \      if ((c & 0x20) == 0) \
733      int gcss = 6*gcaa; \        { \
734      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
735      while (gcaa-- > 0) \        len++; \
736          } \
737        else if ((c & 0x10)  == 0) \
738          { \
739          c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
740          len += 2; \
741          } \
742        else if ((c & 0x08)  == 0) \
743          {\
744          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
745              ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
746          len += 3; \
747          } \
748        else if ((c & 0x04)  == 0) \
749        { \        { \
750        gcss -= 6; \        c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
751        c |= (*eptr++ & 0x3f) << gcss; \            ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
752              (eptr[4] & 0x3f); \
753          len += 4; \
754          } \
755        else \
756          {\
757          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
758              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
759              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
760          len += 5; \
761        } \        } \
762      }      }
763    
# Line 339  if there are extra bytes. This is called Line 766  if there are extra bytes. This is called
766    
767  #define GETCHARLEN(c, eptr, len) \  #define GETCHARLEN(c, eptr, len) \
768    c = *eptr; \    c = *eptr; \
769    if ((c & 0xc0) == 0xc0) \    if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
770      { \  
771      int gcii; \  /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
772      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  pointer, incrementing length if there are extra bytes. This is called when we
773      int gcss = 6*gcaa; \  do not know if we are in UTF-8 mode. */
774      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
775      for (gcii = 1; gcii <= gcaa; gcii++) \  #define GETCHARLENTEST(c, eptr, len) \
776        { \    c = *eptr; \
777        gcss -= 6; \    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
778        c |= (eptr[gcii] & 0x3f) << gcss; \  
779        } \  /* Returns the next uchar, not advancing the pointer. This is called when
780      len += gcaa; \  we know we are in UTF mode. */
781      }  
782    #define RAWUCHAR(eptr) \
783      (*(eptr))
784    
785    /* Returns the next uchar, advancing the pointer. This is called when
786    we know we are in UTF mode. */
787    
788    #define RAWUCHARINC(eptr) \
789      (*((eptr)++))
790    
791    /* Returns the next uchar, testing for UTF mode, and not advancing the
792    pointer. */
793    
794    #define RAWUCHARTEST(eptr) \
795      (*(eptr))
796    
797    /* Returns the next uchar, testing for UTF mode, advancing the
798    pointer. */
799    
800    #define RAWUCHARINCTEST(eptr) \
801      (*((eptr)++))
802    
803  /* If the pointer is not at the start of a character, move it back until  /* If the pointer is not at the start of a character, move it back until
804  it is. 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
805    because almost all calls are already within a block of UTF-8 only code. */
806    
807  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--;  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
808    
809  #endif  /* Same as above, just in the other direction. */
810    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
811    
812    /* Same as above, but it allows a fully customizable form. */
813    #define ACROSSCHAR(condition, eptr, action) \
814      while((condition) && ((eptr) & 0xc0) == 0x80) action
815    
816  /* In case there is no definition of offsetof() provided - though any proper  #elif defined COMPILE_PCRE16
 Standard C system should have one. */  
817    
818  #ifndef offsetof  /* Tells the biggest code point which can be encoded as a single character. */
819  #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
820  #endif  #define MAX_VALUE_FOR_SINGLE_CHAR 65535
821    
822    /* Tests whether the code point needs extra characters to decode. */
823    
824    #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
825    
826    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
827    Otherwise it has an undefined behaviour. */
828    
829    #define GET_EXTRALEN(c) 1
830    
831    /* Returns TRUE, if the given character is not the first character
832    of a UTF sequence. */
833    
834    #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
835    
836    /* Base macro to pick up the low surrogate of a UTF-16 character, not
837    advancing the pointer. */
838    
839    #define GETUTF16(c, eptr) \
840       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
841    
842    /* Get the next UTF-16 character, not advancing the pointer. This is called when
843    we know we are in UTF-16 mode. */
844    
845    #define GETCHAR(c, eptr) \
846      c = *eptr; \
847      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
848    
849    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
850    pointer. */
851    
852    #define GETCHARTEST(c, eptr) \
853      c = *eptr; \
854      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
855    
856    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
857    the pointer. */
858    
859    #define GETUTF16INC(c, eptr) \
860       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
861    
862    /* Get the next UTF-16 character, advancing the pointer. This is called when we
863    know we are in UTF-16 mode. */
864    
865    #define GETCHARINC(c, eptr) \
866      c = *eptr++; \
867      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
868    
869    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
870    This is called when we don't know if we are in UTF-16 mode. */
871    
872    #define GETCHARINCTEST(c, eptr) \
873      c = *eptr++; \
874      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
875    
876    /* Base macro to pick up the low surrogate of a UTF-16 character, not
877    advancing the pointer, incrementing the length. */
878    
879    #define GETUTF16LEN(c, eptr, len) \
880       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
881    
882    /* Get the next UTF-16 character, not advancing the pointer, incrementing
883    length if there is a low surrogate. This is called when we know we are in
884    UTF-16 mode. */
885    
886    #define GETCHARLEN(c, eptr, len) \
887      c = *eptr; \
888      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
889    
890    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
891    pointer, incrementing length if there is a low surrogate. This is called when
892    we do not know if we are in UTF-16 mode. */
893    
894    #define GETCHARLENTEST(c, eptr, len) \
895      c = *eptr; \
896      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
897    
898    /* Returns the next uchar, not advancing the pointer. This is called when
899    we know we are in UTF mode. */
900    
901    #define RAWUCHAR(eptr) \
902      (*(eptr))
903    
904    /* Returns the next uchar, advancing the pointer. This is called when
905    we know we are in UTF mode. */
906    
907    #define RAWUCHARINC(eptr) \
908      (*((eptr)++))
909    
910    /* Returns the next uchar, testing for UTF mode, and not advancing the
911    pointer. */
912    
913    #define RAWUCHARTEST(eptr) \
914      (*(eptr))
915    
916    /* Returns the next uchar, testing for UTF mode, advancing the
917    pointer. */
918    
919    #define RAWUCHARINCTEST(eptr) \
920      (*((eptr)++))
921    
922    /* If the pointer is not at the start of a character, move it back until
923    it is. This is called only in UTF-16 mode - we don't put a test within the
924    macro because almost all calls are already within a block of UTF-16 only
925    code. */
926    
927    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
928    
929    /* Same as above, just in the other direction. */
930    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
931    
932    /* Same as above, but it allows a fully customizable form. */
933    #define ACROSSCHAR(condition, eptr, action) \
934      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
935    
936    #elif defined COMPILE_PCRE32
937    
938    /* These are trivial for the 32-bit library, since all UTF-32 characters fit
939    into one pcre_uchar unit. */
940    #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
941    #define HAS_EXTRALEN(c) (0)
942    #define GET_EXTRALEN(c) (0)
943    #define NOT_FIRSTCHAR(c) (0)
944    
945    /* Get the next UTF-32 character, not advancing the pointer. This is called when
946    we know we are in UTF-32 mode. */
947    
948    #define GETCHAR(c, eptr) \
949      c = *(eptr);
950    
951    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
952    pointer. */
953    
954    #define GETCHARTEST(c, eptr) \
955      c = *(eptr);
956    
957    /* Get the next UTF-32 character, advancing the pointer. This is called when we
958    know we are in UTF-32 mode. */
959    
960    #define GETCHARINC(c, eptr) \
961      c = *((eptr)++);
962    
963    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
964    This is called when we don't know if we are in UTF-32 mode. */
965    
966    #define GETCHARINCTEST(c, eptr) \
967      c = *((eptr)++);
968    
969    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
970    length (since all UTF-32 is of length 1). This is called when we know we are in
971    UTF-32 mode. */
972    
973    #define GETCHARLEN(c, eptr, len) \
974      GETCHAR(c, eptr)
975    
976    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
977    pointer, not incrementing the length (since all UTF-32 is of length 1).
978    This is called when we do not know if we are in UTF-32 mode. */
979    
980    #define GETCHARLENTEST(c, eptr, len) \
981      GETCHARTEST(c, eptr)
982    
983    /* Returns the next uchar, not advancing the pointer. This is called when
984    we know we are in UTF mode. */
985    
986    #define RAWUCHAR(eptr) \
987      (*(eptr))
988    
989    /* Returns the next uchar, advancing the pointer. This is called when
990    we know we are in UTF mode. */
991    
992    #define RAWUCHARINC(eptr) \
993      (*((eptr)++))
994    
995    /* Returns the next uchar, testing for UTF mode, and not advancing the
996    pointer. */
997    
998    #define RAWUCHARTEST(eptr) \
999      (*(eptr))
1000    
1001    /* Returns the next uchar, testing for UTF mode, advancing the
1002    pointer. */
1003    
1004    #define RAWUCHARINCTEST(eptr) \
1005      (*((eptr)++))
1006    
1007    /* If the pointer is not at the start of a character, move it back until
1008    it is. This is called only in UTF-32 mode - we don't put a test within the
1009    macro because almost all calls are already within a block of UTF-32 only
1010    code.
1011    These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
1012    
1013    #define BACKCHAR(eptr) do { } while (0)
1014    
1015    /* Same as above, just in the other direction. */
1016    #define FORWARDCHAR(eptr) do { } while (0)
1017    
1018    /* Same as above, but it allows a fully customizable form. */
1019    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1020    
1021    #else
1022    #error Unsupported compiling mode
1023    #endif /* COMPILE_PCRE[8|16|32] */
1024    
1025    #endif  /* SUPPORT_UTF */
1026    
1027  /* These are the public options that can change during matching. */  /* Tests for Unicode horizontal and vertical whitespace characters must check a
1028    number of different values. Using a switch statement for this generates the
1029    fastest code (no loop, no memory access), and there are several places in the
1030    interpreter code where this happens. In order to ensure that all the case lists
1031    remain in step, we use macros so that there is only one place where the lists
1032    are defined.
1033    
1034    These values are also required as lists in pcre_compile.c when processing \h,
1035    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1036    macros that define the values are here so that all the definitions are
1037    together. The lists must be in ascending character order, terminated by
1038    NOTACHAR (which is 0xffffffff).
1039    
1040    Any changes should ensure that the various macros are kept in step with each
1041    other. NOTE: The values also appear in pcre_jit_compile.c. */
1042    
1043    /* ------ ASCII/Unicode environments ------ */
1044    
1045    #ifndef EBCDIC
1046    
1047    #define HSPACE_LIST \
1048      CHAR_HT, CHAR_SPACE, 0xa0, \
1049      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1050      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1051      NOTACHAR
1052    
1053    #define HSPACE_MULTIBYTE_CASES \
1054      case 0x1680:  /* OGHAM SPACE MARK */ \
1055      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1056      case 0x2000:  /* EN QUAD */ \
1057      case 0x2001:  /* EM QUAD */ \
1058      case 0x2002:  /* EN SPACE */ \
1059      case 0x2003:  /* EM SPACE */ \
1060      case 0x2004:  /* THREE-PER-EM SPACE */ \
1061      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1062      case 0x2006:  /* SIX-PER-EM SPACE */ \
1063      case 0x2007:  /* FIGURE SPACE */ \
1064      case 0x2008:  /* PUNCTUATION SPACE */ \
1065      case 0x2009:  /* THIN SPACE */ \
1066      case 0x200A:  /* HAIR SPACE */ \
1067      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1068      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1069      case 0x3000   /* IDEOGRAPHIC SPACE */
1070    
1071    #define HSPACE_BYTE_CASES \
1072      case CHAR_HT: \
1073      case CHAR_SPACE: \
1074      case 0xa0     /* NBSP */
1075    
1076    #define HSPACE_CASES \
1077      HSPACE_BYTE_CASES: \
1078      HSPACE_MULTIBYTE_CASES
1079    
1080    #define VSPACE_LIST \
1081      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1082    
1083    #define VSPACE_MULTIBYTE_CASES \
1084      case 0x2028:    /* LINE SEPARATOR */ \
1085      case 0x2029     /* PARAGRAPH SEPARATOR */
1086    
1087    #define VSPACE_BYTE_CASES \
1088      case CHAR_LF: \
1089      case CHAR_VT: \
1090      case CHAR_FF: \
1091      case CHAR_CR: \
1092      case CHAR_NEL
1093    
1094    #define VSPACE_CASES \
1095      VSPACE_BYTE_CASES: \
1096      VSPACE_MULTIBYTE_CASES
1097    
1098  #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)  /* ------ EBCDIC environments ------ */
1099    
1100    #else
1101    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1102    
1103    #define HSPACE_BYTE_CASES \
1104      case CHAR_HT: \
1105      case CHAR_SPACE
1106    
1107    #define HSPACE_CASES HSPACE_BYTE_CASES
1108    
1109    #ifdef EBCDIC_NL25
1110    #define VSPACE_LIST \
1111      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1112    #else
1113    #define VSPACE_LIST \
1114      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1115    #endif
1116    
1117    #define VSPACE_BYTE_CASES \
1118      case CHAR_LF: \
1119      case CHAR_VT: \
1120      case CHAR_FF: \
1121      case CHAR_CR: \
1122      case CHAR_NEL
1123    
1124    #define VSPACE_CASES VSPACE_BYTE_CASES
1125    #endif  /* EBCDIC */
1126    
1127    /* ------ End of whitespace macros ------ */
1128    
 /* Private options flags start at the most significant end of the four bytes,  
 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! */  
1129    
 #define PCRE_FIRSTSET      0x40000000  /* first_byte is set */  
 #define PCRE_REQCHSET      0x20000000  /* req_byte is set */  
 #define PCRE_STARTLINE     0x10000000  /* start after \n for multiline */  
 #define PCRE_ICHANGED      0x08000000  /* i option changes within regex */  
 #define PCRE_NOPARTIAL     0x04000000  /* can't use partial with this regex */  
1130    
1131  /* Options for the "extra" block produced by pcre_study(). */  /* Private flags containing information about the compiled regex. They used to
1132    live at the top end of the options word, but that got almost full, so they were
1133    moved to a 16-bit flags word - which got almost full, so now they are in a
1134    32-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as the
1135    restrictions on partial matching have been lifted. It remains for backwards
1136    compatibility. */
1137    
1138    #define PCRE_MODE8         0x00000001  /* compiled in 8 bit mode */
1139    #define PCRE_MODE16        0x00000002  /* compiled in 16 bit mode */
1140    #define PCRE_MODE32        0x00000004  /* compiled in 32 bit mode */
1141    #define PCRE_FIRSTSET      0x00000010  /* first_char is set */
1142    #define PCRE_FCH_CASELESS  0x00000020  /* caseless first char */
1143    #define PCRE_REQCHSET      0x00000040  /* req_byte is set */
1144    #define PCRE_RCH_CASELESS  0x00000080  /* caseless requested char */
1145    #define PCRE_STARTLINE     0x00000100  /* start after \n for multiline */
1146    #define PCRE_NOPARTIAL     0x00000200  /* can't use partial with this regex */
1147    #define PCRE_JCHANGED      0x00000400  /* j option used in regex */
1148    #define PCRE_HASCRORLF     0x00000800  /* explicit \r or \n in pattern */
1149    #define PCRE_HASTHEN       0x00001000  /* pattern contains (*THEN) */
1150    #define PCRE_MLSET         0x00002000  /* match limit set by regex */
1151    #define PCRE_RLSET         0x00004000  /* recursion limit set by regex */
1152    
1153    #if defined COMPILE_PCRE8
1154    #define PCRE_MODE          PCRE_MODE8
1155    #elif defined COMPILE_PCRE16
1156    #define PCRE_MODE          PCRE_MODE16
1157    #elif defined COMPILE_PCRE32
1158    #define PCRE_MODE          PCRE_MODE32
1159    #endif
1160    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1161    
1162    /* Flags for the "extra" block produced by pcre_study(). */
1163    
1164  #define PCRE_STUDY_MAPPED   0x01     /* a map of starting chars exists */  #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
1165    #define PCRE_STUDY_MINLEN  0x0002  /* a minimum length field exists */
1166    
1167  /* Masks for identifying the public options that are permitted at compile  /* Masks for identifying the public options that are permitted at compile
1168  time, run time, or study time, respectively. */  time, run time, or study time, respectively. */
1169    
1170  #define PUBLIC_OPTIONS \  #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
1171                               PCRE_NEWLINE_ANYCRLF)
1172    
1173    #define PUBLIC_COMPILE_OPTIONS \
1174    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
1175     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1176     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| \
1177       PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1178       PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE|PCRE_NEVER_UTF)
1179    
1180  #define PUBLIC_EXEC_OPTIONS \  #define PUBLIC_EXEC_OPTIONS \
1181    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1182     PCRE_PARTIAL)     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
1183       PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
1184    
1185  #define PUBLIC_DFA_EXEC_OPTIONS \  #define PUBLIC_DFA_EXEC_OPTIONS \
1186    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1187     PCRE_PARTIAL|PCRE_DFA_SHORTEST|PCRE_DFA_RESTART)     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
1188       PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1189  #define PUBLIC_STUDY_OPTIONS 0   /* None defined */     PCRE_NO_START_OPTIMIZE)
1190    
1191    #define PUBLIC_STUDY_OPTIONS \
1192       (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1193        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1194    
1195    #define PUBLIC_JIT_EXEC_OPTIONS \
1196       (PCRE_NO_UTF8_CHECK|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|\
1197        PCRE_NOTEMPTY_ATSTART|PCRE_PARTIAL_SOFT|PCRE_PARTIAL_HARD)
1198    
1199  /* 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. */  
1200    
1201  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1202    
1203  /* Negative values for the firstchar and reqchar variables */  /* This variable is used to detect a loaded regular expression
1204    in different endianness. */
1205    
1206  #define REQ_UNSET (-2)  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
 #define REQ_NONE  (-1)  
1207    
1208  /* 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
1209  req_byte match. */  req_byte match. */
1210    
1211  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1212    
1213  /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1214  variable-length repeat, or a anything other than literal characters. */  environments where these macros are defined elsewhere. Unfortunately, there
1215    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 */  
1216    
1217  typedef int BOOL;  typedef int BOOL;
1218    
1219    #ifndef FALSE
1220  #define FALSE   0  #define FALSE   0
1221  #define TRUE    1  #define TRUE    1
1222    #endif
1223    
1224    /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
1225    character constants like '*' because the compiler would emit their EBCDIC code,
1226    which is different from their ASCII/UTF-8 code. Instead we define macros for
1227    the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
1228    is enabled. When UTF-8 support is not enabled, the definitions use character
1229    literals. Both character and string versions of each character are needed, and
1230    there are some longer strings as well.
1231    
1232    This means that, on EBCDIC platforms, the PCRE library can handle either
1233    EBCDIC, or UTF-8, but not both. To support both in the same compiled library
1234    would need different lookups depending on whether PCRE_UTF8 was set or not.
1235    This would make it impossible to use characters in switch/case statements,
1236    which would reduce performance. For a theoretical use (which nobody has asked
1237    for) in a minority area (EBCDIC platforms), this is not sensible. Any
1238    application that did need both could compile two versions of the library, using
1239    macros to give the functions distinct names. */
1240    
1241    #ifndef SUPPORT_UTF
1242    
1243    /* UTF-8 support is not enabled; use the platform-dependent character literals
1244    so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1245    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1246    characters, a common practice has been to use its NL (0x15) character as the
1247    line terminator in C-like processing environments. However, sometimes the LF
1248    (0x25) character is used instead, according to this Unicode document:
1249    
1250    http://unicode.org/standard/reports/tr13/tr13-5.html
1251    
1252    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1253    instead. Whichever is *not* chosen is defined as NEL.
1254    
1255    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1256    same code point. */
1257    
1258    #ifdef EBCDIC
1259    
1260    #ifndef EBCDIC_NL25
1261    #define CHAR_NL                     '\x15'
1262    #define CHAR_NEL                    '\x25'
1263    #define STR_NL                      "\x15"
1264    #define STR_NEL                     "\x25"
1265    #else
1266    #define CHAR_NL                     '\x25'
1267    #define CHAR_NEL                    '\x15'
1268    #define STR_NL                      "\x25"
1269    #define STR_NEL                     "\x15"
1270    #endif
1271    
1272    #define CHAR_LF                     CHAR_NL
1273    #define STR_LF                      STR_NL
1274    
1275    #define CHAR_ESC                    '\047'
1276    #define CHAR_DEL                    '\007'
1277    #define STR_ESC                     "\047"
1278    #define STR_DEL                     "\007"
1279    
1280    #else  /* Not EBCDIC */
1281    
1282    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1283    compatibility. NEL is the Unicode newline character; make sure it is
1284    a positive value. */
1285    
1286    #define CHAR_LF                     '\n'
1287    #define CHAR_NL                     CHAR_LF
1288    #define CHAR_NEL                    ((unsigned char)'\x85')
1289    #define CHAR_ESC                    '\033'
1290    #define CHAR_DEL                    '\177'
1291    
1292    #define STR_LF                      "\n"
1293    #define STR_NL                      STR_LF
1294    #define STR_NEL                     "\x85"
1295    #define STR_ESC                     "\033"
1296    #define STR_DEL                     "\177"
1297    
1298    #endif  /* EBCDIC */
1299    
1300    /* The remaining definitions work in both environments. */
1301    
1302    #define CHAR_NULL                   '\0'
1303    #define CHAR_HT                     '\t'
1304    #define CHAR_VT                     '\v'
1305    #define CHAR_FF                     '\f'
1306    #define CHAR_CR                     '\r'
1307    #define CHAR_BS                     '\b'
1308    #define CHAR_BEL                    '\a'
1309    
1310    #define CHAR_SPACE                  ' '
1311    #define CHAR_EXCLAMATION_MARK       '!'
1312    #define CHAR_QUOTATION_MARK         '"'
1313    #define CHAR_NUMBER_SIGN            '#'
1314    #define CHAR_DOLLAR_SIGN            '$'
1315    #define CHAR_PERCENT_SIGN           '%'
1316    #define CHAR_AMPERSAND              '&'
1317    #define CHAR_APOSTROPHE             '\''
1318    #define CHAR_LEFT_PARENTHESIS       '('
1319    #define CHAR_RIGHT_PARENTHESIS      ')'
1320    #define CHAR_ASTERISK               '*'
1321    #define CHAR_PLUS                   '+'
1322    #define CHAR_COMMA                  ','
1323    #define CHAR_MINUS                  '-'
1324    #define CHAR_DOT                    '.'
1325    #define CHAR_SLASH                  '/'
1326    #define CHAR_0                      '0'
1327    #define CHAR_1                      '1'
1328    #define CHAR_2                      '2'
1329    #define CHAR_3                      '3'
1330    #define CHAR_4                      '4'
1331    #define CHAR_5                      '5'
1332    #define CHAR_6                      '6'
1333    #define CHAR_7                      '7'
1334    #define CHAR_8                      '8'
1335    #define CHAR_9                      '9'
1336    #define CHAR_COLON                  ':'
1337    #define CHAR_SEMICOLON              ';'
1338    #define CHAR_LESS_THAN_SIGN         '<'
1339    #define CHAR_EQUALS_SIGN            '='
1340    #define CHAR_GREATER_THAN_SIGN      '>'
1341    #define CHAR_QUESTION_MARK          '?'
1342    #define CHAR_COMMERCIAL_AT          '@'
1343    #define CHAR_A                      'A'
1344    #define CHAR_B                      'B'
1345    #define CHAR_C                      'C'
1346    #define CHAR_D                      'D'
1347    #define CHAR_E                      'E'
1348    #define CHAR_F                      'F'
1349    #define CHAR_G                      'G'
1350    #define CHAR_H                      'H'
1351    #define CHAR_I                      'I'
1352    #define CHAR_J                      'J'
1353    #define CHAR_K                      'K'
1354    #define CHAR_L                      'L'
1355    #define CHAR_M                      'M'
1356    #define CHAR_N                      'N'
1357    #define CHAR_O                      'O'
1358    #define CHAR_P                      'P'
1359    #define CHAR_Q                      'Q'
1360    #define CHAR_R                      'R'
1361    #define CHAR_S                      'S'
1362    #define CHAR_T                      'T'
1363    #define CHAR_U                      'U'
1364    #define CHAR_V                      'V'
1365    #define CHAR_W                      'W'
1366    #define CHAR_X                      'X'
1367    #define CHAR_Y                      'Y'
1368    #define CHAR_Z                      'Z'
1369    #define CHAR_LEFT_SQUARE_BRACKET    '['
1370    #define CHAR_BACKSLASH              '\\'
1371    #define CHAR_RIGHT_SQUARE_BRACKET   ']'
1372    #define CHAR_CIRCUMFLEX_ACCENT      '^'
1373    #define CHAR_UNDERSCORE             '_'
1374    #define CHAR_GRAVE_ACCENT           '`'
1375    #define CHAR_a                      'a'
1376    #define CHAR_b                      'b'
1377    #define CHAR_c                      'c'
1378    #define CHAR_d                      'd'
1379    #define CHAR_e                      'e'
1380    #define CHAR_f                      'f'
1381    #define CHAR_g                      'g'
1382    #define CHAR_h                      'h'
1383    #define CHAR_i                      'i'
1384    #define CHAR_j                      'j'
1385    #define CHAR_k                      'k'
1386    #define CHAR_l                      'l'
1387    #define CHAR_m                      'm'
1388    #define CHAR_n                      'n'
1389    #define CHAR_o                      'o'
1390    #define CHAR_p                      'p'
1391    #define CHAR_q                      'q'
1392    #define CHAR_r                      'r'
1393    #define CHAR_s                      's'
1394    #define CHAR_t                      't'
1395    #define CHAR_u                      'u'
1396    #define CHAR_v                      'v'
1397    #define CHAR_w                      'w'
1398    #define CHAR_x                      'x'
1399    #define CHAR_y                      'y'
1400    #define CHAR_z                      'z'
1401    #define CHAR_LEFT_CURLY_BRACKET     '{'
1402    #define CHAR_VERTICAL_LINE          '|'
1403    #define CHAR_RIGHT_CURLY_BRACKET    '}'
1404    #define CHAR_TILDE                  '~'
1405    
1406    #define STR_HT                      "\t"
1407    #define STR_VT                      "\v"
1408    #define STR_FF                      "\f"
1409    #define STR_CR                      "\r"
1410    #define STR_BS                      "\b"
1411    #define STR_BEL                     "\a"
1412    
1413    #define STR_SPACE                   " "
1414    #define STR_EXCLAMATION_MARK        "!"
1415    #define STR_QUOTATION_MARK          "\""
1416    #define STR_NUMBER_SIGN             "#"
1417    #define STR_DOLLAR_SIGN             "$"
1418    #define STR_PERCENT_SIGN            "%"
1419    #define STR_AMPERSAND               "&"
1420    #define STR_APOSTROPHE              "'"
1421    #define STR_LEFT_PARENTHESIS        "("
1422    #define STR_RIGHT_PARENTHESIS       ")"
1423    #define STR_ASTERISK                "*"
1424    #define STR_PLUS                    "+"
1425    #define STR_COMMA                   ","
1426    #define STR_MINUS                   "-"
1427    #define STR_DOT                     "."
1428    #define STR_SLASH                   "/"
1429    #define STR_0                       "0"
1430    #define STR_1                       "1"
1431    #define STR_2                       "2"
1432    #define STR_3                       "3"
1433    #define STR_4                       "4"
1434    #define STR_5                       "5"
1435    #define STR_6                       "6"
1436    #define STR_7                       "7"
1437    #define STR_8                       "8"
1438    #define STR_9                       "9"
1439    #define STR_COLON                   ":"
1440    #define STR_SEMICOLON               ";"
1441    #define STR_LESS_THAN_SIGN          "<"
1442    #define STR_EQUALS_SIGN             "="
1443    #define STR_GREATER_THAN_SIGN       ">"
1444    #define STR_QUESTION_MARK           "?"
1445    #define STR_COMMERCIAL_AT           "@"
1446    #define STR_A                       "A"
1447    #define STR_B                       "B"
1448    #define STR_C                       "C"
1449    #define STR_D                       "D"
1450    #define STR_E                       "E"
1451    #define STR_F                       "F"
1452    #define STR_G                       "G"
1453    #define STR_H                       "H"
1454    #define STR_I                       "I"
1455    #define STR_J                       "J"
1456    #define STR_K                       "K"
1457    #define STR_L                       "L"
1458    #define STR_M                       "M"
1459    #define STR_N                       "N"
1460    #define STR_O                       "O"
1461    #define STR_P                       "P"
1462    #define STR_Q                       "Q"
1463    #define STR_R                       "R"
1464    #define STR_S                       "S"
1465    #define STR_T                       "T"
1466    #define STR_U                       "U"
1467    #define STR_V                       "V"
1468    #define STR_W                       "W"
1469    #define STR_X                       "X"
1470    #define STR_Y                       "Y"
1471    #define STR_Z                       "Z"
1472    #define STR_LEFT_SQUARE_BRACKET     "["
1473    #define STR_BACKSLASH               "\\"
1474    #define STR_RIGHT_SQUARE_BRACKET    "]"
1475    #define STR_CIRCUMFLEX_ACCENT       "^"
1476    #define STR_UNDERSCORE              "_"
1477    #define STR_GRAVE_ACCENT            "`"
1478    #define STR_a                       "a"
1479    #define STR_b                       "b"
1480    #define STR_c                       "c"
1481    #define STR_d                       "d"
1482    #define STR_e                       "e"
1483    #define STR_f                       "f"
1484    #define STR_g                       "g"
1485    #define STR_h                       "h"
1486    #define STR_i                       "i"
1487    #define STR_j                       "j"
1488    #define STR_k                       "k"
1489    #define STR_l                       "l"
1490    #define STR_m                       "m"
1491    #define STR_n                       "n"
1492    #define STR_o                       "o"
1493    #define STR_p                       "p"
1494    #define STR_q                       "q"
1495    #define STR_r                       "r"
1496    #define STR_s                       "s"
1497    #define STR_t                       "t"
1498    #define STR_u                       "u"
1499    #define STR_v                       "v"
1500    #define STR_w                       "w"
1501    #define STR_x                       "x"
1502    #define STR_y                       "y"
1503    #define STR_z                       "z"
1504    #define STR_LEFT_CURLY_BRACKET      "{"
1505    #define STR_VERTICAL_LINE           "|"
1506    #define STR_RIGHT_CURLY_BRACKET     "}"
1507    #define STR_TILDE                   "~"
1508    
1509    #define STRING_ACCEPT0              "ACCEPT\0"
1510    #define STRING_COMMIT0              "COMMIT\0"
1511    #define STRING_F0                   "F\0"
1512    #define STRING_FAIL0                "FAIL\0"
1513    #define STRING_MARK0                "MARK\0"
1514    #define STRING_PRUNE0               "PRUNE\0"
1515    #define STRING_SKIP0                "SKIP\0"
1516    #define STRING_THEN                 "THEN"
1517    
1518    #define STRING_alpha0               "alpha\0"
1519    #define STRING_lower0               "lower\0"
1520    #define STRING_upper0               "upper\0"
1521    #define STRING_alnum0               "alnum\0"
1522    #define STRING_ascii0               "ascii\0"
1523    #define STRING_blank0               "blank\0"
1524    #define STRING_cntrl0               "cntrl\0"
1525    #define STRING_digit0               "digit\0"
1526    #define STRING_graph0               "graph\0"
1527    #define STRING_print0               "print\0"
1528    #define STRING_punct0               "punct\0"
1529    #define STRING_space0               "space\0"
1530    #define STRING_word0                "word\0"
1531    #define STRING_xdigit               "xdigit"
1532    
1533    #define STRING_DEFINE               "DEFINE"
1534    
1535    #define STRING_CR_RIGHTPAR             "CR)"
1536    #define STRING_LF_RIGHTPAR             "LF)"
1537    #define STRING_CRLF_RIGHTPAR           "CRLF)"
1538    #define STRING_ANY_RIGHTPAR            "ANY)"
1539    #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1540    #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1541    #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1542    #define STRING_UTF8_RIGHTPAR           "UTF8)"
1543    #define STRING_UTF16_RIGHTPAR          "UTF16)"
1544    #define STRING_UTF32_RIGHTPAR          "UTF32)"
1545    #define STRING_UTF_RIGHTPAR            "UTF)"
1546    #define STRING_UCP_RIGHTPAR            "UCP)"
1547    #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1548    #define STRING_LIMIT_MATCH_EQ          "LIMIT_MATCH="
1549    #define STRING_LIMIT_RECURSION_EQ      "LIMIT_RECURSION="
1550    
1551    #else  /* SUPPORT_UTF */
1552    
1553    /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1554    works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
1555    only. */
1556    
1557    #define CHAR_HT                     '\011'
1558    #define CHAR_VT                     '\013'
1559    #define CHAR_FF                     '\014'
1560    #define CHAR_CR                     '\015'
1561    #define CHAR_LF                     '\012'
1562    #define CHAR_NL                     CHAR_LF
1563    #define CHAR_NEL                    ((unsigned char)'\x85')
1564    #define CHAR_BS                     '\010'
1565    #define CHAR_BEL                    '\007'
1566    #define CHAR_ESC                    '\033'
1567    #define CHAR_DEL                    '\177'
1568    
1569    #define CHAR_NULL                   '\0'
1570    #define CHAR_SPACE                  '\040'
1571    #define CHAR_EXCLAMATION_MARK       '\041'
1572    #define CHAR_QUOTATION_MARK         '\042'
1573    #define CHAR_NUMBER_SIGN            '\043'
1574    #define CHAR_DOLLAR_SIGN            '\044'
1575    #define CHAR_PERCENT_SIGN           '\045'
1576    #define CHAR_AMPERSAND              '\046'
1577    #define CHAR_APOSTROPHE             '\047'
1578    #define CHAR_LEFT_PARENTHESIS       '\050'
1579    #define CHAR_RIGHT_PARENTHESIS      '\051'
1580    #define CHAR_ASTERISK               '\052'
1581    #define CHAR_PLUS                   '\053'
1582    #define CHAR_COMMA                  '\054'
1583    #define CHAR_MINUS                  '\055'
1584    #define CHAR_DOT                    '\056'
1585    #define CHAR_SLASH                  '\057'
1586    #define CHAR_0                      '\060'
1587    #define CHAR_1                      '\061'
1588    #define CHAR_2                      '\062'
1589    #define CHAR_3                      '\063'
1590    #define CHAR_4                      '\064'
1591    #define CHAR_5                      '\065'
1592    #define CHAR_6                      '\066'
1593    #define CHAR_7                      '\067'
1594    #define CHAR_8                      '\070'
1595    #define CHAR_9                      '\071'
1596    #define CHAR_COLON                  '\072'
1597    #define CHAR_SEMICOLON              '\073'
1598    #define CHAR_LESS_THAN_SIGN         '\074'
1599    #define CHAR_EQUALS_SIGN            '\075'
1600    #define CHAR_GREATER_THAN_SIGN      '\076'
1601    #define CHAR_QUESTION_MARK          '\077'
1602    #define CHAR_COMMERCIAL_AT          '\100'
1603    #define CHAR_A                      '\101'
1604    #define CHAR_B                      '\102'
1605    #define CHAR_C                      '\103'
1606    #define CHAR_D                      '\104'
1607    #define CHAR_E                      '\105'
1608    #define CHAR_F                      '\106'
1609    #define CHAR_G                      '\107'
1610    #define CHAR_H                      '\110'
1611    #define CHAR_I                      '\111'
1612    #define CHAR_J                      '\112'
1613    #define CHAR_K                      '\113'
1614    #define CHAR_L                      '\114'
1615    #define CHAR_M                      '\115'
1616    #define CHAR_N                      '\116'
1617    #define CHAR_O                      '\117'
1618    #define CHAR_P                      '\120'
1619    #define CHAR_Q                      '\121'
1620    #define CHAR_R                      '\122'
1621    #define CHAR_S                      '\123'
1622    #define CHAR_T                      '\124'
1623    #define CHAR_U                      '\125'
1624    #define CHAR_V                      '\126'
1625    #define CHAR_W                      '\127'
1626    #define CHAR_X                      '\130'
1627    #define CHAR_Y                      '\131'
1628    #define CHAR_Z                      '\132'
1629    #define CHAR_LEFT_SQUARE_BRACKET    '\133'
1630    #define CHAR_BACKSLASH              '\134'
1631    #define CHAR_RIGHT_SQUARE_BRACKET   '\135'
1632    #define CHAR_CIRCUMFLEX_ACCENT      '\136'
1633    #define CHAR_UNDERSCORE             '\137'
1634    #define CHAR_GRAVE_ACCENT           '\140'
1635    #define CHAR_a                      '\141'
1636    #define CHAR_b                      '\142'
1637    #define CHAR_c                      '\143'
1638    #define CHAR_d                      '\144'
1639    #define CHAR_e                      '\145'
1640    #define CHAR_f                      '\146'
1641    #define CHAR_g                      '\147'
1642    #define CHAR_h                      '\150'
1643    #define CHAR_i                      '\151'
1644    #define CHAR_j                      '\152'
1645    #define CHAR_k                      '\153'
1646    #define CHAR_l                      '\154'
1647    #define CHAR_m                      '\155'
1648    #define CHAR_n                      '\156'
1649    #define CHAR_o                      '\157'
1650    #define CHAR_p                      '\160'
1651    #define CHAR_q                      '\161'
1652    #define CHAR_r                      '\162'
1653    #define CHAR_s                      '\163'
1654    #define CHAR_t                      '\164'
1655    #define CHAR_u                      '\165'
1656    #define CHAR_v                      '\166'
1657    #define CHAR_w                      '\167'
1658    #define CHAR_x                      '\170'
1659    #define CHAR_y                      '\171'
1660    #define CHAR_z                      '\172'
1661    #define CHAR_LEFT_CURLY_BRACKET     '\173'
1662    #define CHAR_VERTICAL_LINE          '\174'
1663    #define CHAR_RIGHT_CURLY_BRACKET    '\175'
1664    #define CHAR_TILDE                  '\176'
1665    
1666    #define STR_HT                      "\011"
1667    #define STR_VT                      "\013"
1668    #define STR_FF                      "\014"
1669    #define STR_CR                      "\015"
1670    #define STR_NL                      "\012"
1671    #define STR_BS                      "\010"
1672    #define STR_BEL                     "\007"
1673    #define STR_ESC                     "\033"
1674    #define STR_DEL                     "\177"
1675    
1676    #define STR_SPACE                   "\040"
1677    #define STR_EXCLAMATION_MARK        "\041"
1678    #define STR_QUOTATION_MARK          "\042"
1679    #define STR_NUMBER_SIGN             "\043"
1680    #define STR_DOLLAR_SIGN             "\044"
1681    #define STR_PERCENT_SIGN            "\045"
1682    #define STR_AMPERSAND               "\046"
1683    #define STR_APOSTROPHE              "\047"
1684    #define STR_LEFT_PARENTHESIS        "\050"
1685    #define STR_RIGHT_PARENTHESIS       "\051"
1686    #define STR_ASTERISK                "\052"
1687    #define STR_PLUS                    "\053"
1688    #define STR_COMMA                   "\054"
1689    #define STR_MINUS                   "\055"
1690    #define STR_DOT                     "\056"
1691    #define STR_SLASH                   "\057"
1692    #define STR_0                       "\060"
1693    #define STR_1                       "\061"
1694    #define STR_2                       "\062"
1695    #define STR_3                       "\063"
1696    #define STR_4                       "\064"
1697    #define STR_5                       "\065"
1698    #define STR_6                       "\066"
1699    #define STR_7                       "\067"
1700    #define STR_8                       "\070"
1701    #define STR_9                       "\071"
1702    #define STR_COLON                   "\072"
1703    #define STR_SEMICOLON               "\073"
1704    #define STR_LESS_THAN_SIGN          "\074"
1705    #define STR_EQUALS_SIGN             "\075"
1706    #define STR_GREATER_THAN_SIGN       "\076"
1707    #define STR_QUESTION_MARK           "\077"
1708    #define STR_COMMERCIAL_AT           "\100"
1709    #define STR_A                       "\101"
1710    #define STR_B                       "\102"
1711    #define STR_C                       "\103"
1712    #define STR_D                       "\104"
1713    #define STR_E                       "\105"
1714    #define STR_F                       "\106"
1715    #define STR_G                       "\107"
1716    #define STR_H                       "\110"
1717    #define STR_I                       "\111"
1718    #define STR_J                       "\112"
1719    #define STR_K                       "\113"
1720    #define STR_L                       "\114"
1721    #define STR_M                       "\115"
1722    #define STR_N                       "\116"
1723    #define STR_O                       "\117"
1724    #define STR_P                       "\120"
1725    #define STR_Q                       "\121"
1726    #define STR_R                       "\122"
1727    #define STR_S                       "\123"
1728    #define STR_T                       "\124"
1729    #define STR_U                       "\125"
1730    #define STR_V                       "\126"
1731    #define STR_W                       "\127"
1732    #define STR_X                       "\130"
1733    #define STR_Y                       "\131"
1734    #define STR_Z                       "\132"
1735    #define STR_LEFT_SQUARE_BRACKET     "\133"
1736    #define STR_BACKSLASH               "\134"
1737    #define STR_RIGHT_SQUARE_BRACKET    "\135"
1738    #define STR_CIRCUMFLEX_ACCENT       "\136"
1739    #define STR_UNDERSCORE              "\137"
1740    #define STR_GRAVE_ACCENT            "\140"
1741    #define STR_a                       "\141"
1742    #define STR_b                       "\142"
1743    #define STR_c                       "\143"
1744    #define STR_d                       "\144"
1745    #define STR_e                       "\145"
1746    #define STR_f                       "\146"
1747    #define STR_g                       "\147"
1748    #define STR_h                       "\150"
1749    #define STR_i                       "\151"
1750    #define STR_j                       "\152"
1751    #define STR_k                       "\153"
1752    #define STR_l                       "\154"
1753    #define STR_m                       "\155"
1754    #define STR_n                       "\156"
1755    #define STR_o                       "\157"
1756    #define STR_p                       "\160"
1757    #define STR_q                       "\161"
1758    #define STR_r                       "\162"
1759    #define STR_s                       "\163"
1760    #define STR_t                       "\164"
1761    #define STR_u                       "\165"
1762    #define STR_v                       "\166"
1763    #define STR_w                       "\167"
1764    #define STR_x                       "\170"
1765    #define STR_y                       "\171"
1766    #define STR_z                       "\172"
1767    #define STR_LEFT_CURLY_BRACKET      "\173"
1768    #define STR_VERTICAL_LINE           "\174"
1769    #define STR_RIGHT_CURLY_BRACKET     "\175"
1770    #define STR_TILDE                   "\176"
1771    
1772    #define STRING_ACCEPT0              STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1773    #define STRING_COMMIT0              STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1774    #define STRING_F0                   STR_F "\0"
1775    #define STRING_FAIL0                STR_F STR_A STR_I STR_L "\0"
1776    #define STRING_MARK0                STR_M STR_A STR_R STR_K "\0"
1777    #define STRING_PRUNE0               STR_P STR_R STR_U STR_N STR_E "\0"
1778    #define STRING_SKIP0                STR_S STR_K STR_I STR_P "\0"
1779    #define STRING_THEN                 STR_T STR_H STR_E STR_N
1780    
1781    #define STRING_alpha0               STR_a STR_l STR_p STR_h STR_a "\0"
1782    #define STRING_lower0               STR_l STR_o STR_w STR_e STR_r "\0"
1783    #define STRING_upper0               STR_u STR_p STR_p STR_e STR_r "\0"
1784    #define STRING_alnum0               STR_a STR_l STR_n STR_u STR_m "\0"
1785    #define STRING_ascii0               STR_a STR_s STR_c STR_i STR_i "\0"
1786    #define STRING_blank0               STR_b STR_l STR_a STR_n STR_k "\0"
1787    #define STRING_cntrl0               STR_c STR_n STR_t STR_r STR_l "\0"
1788    #define STRING_digit0               STR_d STR_i STR_g STR_i STR_t "\0"
1789    #define STRING_graph0               STR_g STR_r STR_a STR_p STR_h "\0"
1790    #define STRING_print0               STR_p STR_r STR_i STR_n STR_t "\0"
1791    #define STRING_punct0               STR_p STR_u STR_n STR_c STR_t "\0"
1792    #define STRING_space0               STR_s STR_p STR_a STR_c STR_e "\0"
1793    #define STRING_word0                STR_w STR_o STR_r STR_d       "\0"
1794    #define STRING_xdigit               STR_x STR_d STR_i STR_g STR_i STR_t
1795    
1796    #define STRING_DEFINE               STR_D STR_E STR_F STR_I STR_N STR_E
1797    
1798    #define STRING_CR_RIGHTPAR             STR_C STR_R STR_RIGHT_PARENTHESIS
1799    #define STRING_LF_RIGHTPAR             STR_L STR_F STR_RIGHT_PARENTHESIS
1800    #define STRING_CRLF_RIGHTPAR           STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1801    #define STRING_ANY_RIGHTPAR            STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1802    #define STRING_ANYCRLF_RIGHTPAR        STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1803    #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
1804    #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
1805    #define STRING_UTF8_RIGHTPAR           STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1806    #define STRING_UTF16_RIGHTPAR          STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1807    #define STRING_UTF32_RIGHTPAR          STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1808    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_RIGHT_PARENTHESIS
1809    #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1810    #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
1811    #define STRING_LIMIT_MATCH_EQ          STR_L STR_I STR_M STR_I STR_T STR_UNDERSCORE STR_M STR_A STR_T STR_C STR_H STR_EQUALS_SIGN
1812    #define STRING_LIMIT_RECURSION_EQ      STR_L STR_I STR_M STR_I STR_T STR_UNDERSCORE STR_R STR_E STR_C STR_U STR_R STR_S STR_I STR_O STR_N STR_EQUALS_SIGN
1813    
1814  /* Escape items that are just an encoding of a particular data value. Note that  #endif  /* SUPPORT_UTF */
1815  ESC_n is defined as yet another macro, which is set in config.h to either \n  
1816  (the default) or \r (which some people want). */  /* Escape items that are just an encoding of a particular data value. */
1817    
1818  #ifndef ESC_e  #ifndef ESC_e
1819  #define ESC_e 27  #define ESC_e CHAR_ESC
1820  #endif  #endif
1821    
1822  #ifndef ESC_f  #ifndef ESC_f
1823  #define ESC_f '\f'  #define ESC_f CHAR_FF
1824  #endif  #endif
1825    
1826  #ifndef ESC_n  #ifndef ESC_n
1827  #define ESC_n NEWLINE  #define ESC_n CHAR_LF
1828  #endif  #endif
1829    
1830  #ifndef ESC_r  #ifndef ESC_r
1831  #define ESC_r '\r'  #define ESC_r CHAR_CR
1832  #endif  #endif
1833    
1834  /* 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
1835  (presumably because of all the others like size_t). */  (presumably because of all the others like size_t). */
1836    
1837  #ifndef ESC_tee  #ifndef ESC_tee
1838  #define ESC_tee '\t'  #define ESC_tee CHAR_HT
1839  #endif  #endif
1840    
1841  /* Codes for different types of Unicode property */  /* Codes for different types of Unicode property */
1842    
1843  #define PT_ANY        0    /* Any property - matches all chars */  #define PT_ANY        0    /* Any property - matches all chars */
1844  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */
1845  #define PT_GC         2    /* General characteristic (e.g. L) */  #define PT_GC         2    /* Specified general characteristic (e.g. L) */
1846  #define PT_PC         3    /* Particular characteristic (e.g. Lu) */  #define PT_PC         3    /* Specified particular characteristic (e.g. Lu) */
1847  #define PT_SC         4    /* Script (e.g. Han) */  #define PT_SC         4    /* Script (e.g. Han) */
1848    #define PT_ALNUM      5    /* Alphanumeric - the union of L and N */
1849    #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1850    #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1851    #define PT_WORD       8    /* Word - L plus N plus underscore */
1852    #define PT_CLIST      9    /* Pseudo-property: match character list */
1853    #define PT_UCNC      10    /* Universal Character nameable character */
1854    
1855  /* 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
1856  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1857    
1858  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1859  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */
# Line 483  contain UTF-8 characters with values gre Line 1866  contain UTF-8 characters with values gre
1866    
1867  /* These are escaped items that aren't just an encoding of a particular data  /* These are escaped items that aren't just an encoding of a particular data
1868  value such as \n. They must have non-zero values, as check_escape() returns  value such as \n. They must have non-zero values, as check_escape() returns
1869  their negation. Also, they must appear in the same order as in the opcode  0 for a data character.  Also, they must appear in the same order as in the opcode
1870  definitions below, up to ESC_z. There's a dummy for OP_ANY because it  definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1871  corresponds to "." rather than an escape sequence. The final one must be  corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1872  ESC_REF as subsequent values are used for \1, \2, \3, etc. There is are two  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1873  tests in the code for an escape greater than ESC_b and less than ESC_Z to  non-DOTALL mode, "." behaves like \N.
1874  detect the types that may be repeated. These are the types that consume  
1875  characters. If any new escapes are put in between that don't consume a  The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1876  character, that code will have to change. */  when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1877    They must be contiguous, and remain in order so that the replacements can be
1878  enum { ESC_A = 1, ESC_G, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, ESC_W,  looked up from a table.
1879         ESC_w, ESC_dum1, ESC_C, ESC_P, ESC_p, ESC_X, ESC_Z, ESC_z, ESC_E,  
1880         ESC_Q, ESC_REF };  Negative numbers are used to encode a backreference (\1, \2, \3, etc.) in
1881    check_escape(). There are two tests in the code for an escape
1882    greater than ESC_b and less than ESC_Z to detect the types that may be
1883    repeated. These are the types that consume characters. If any new escapes are
1884    put in between that don't consume a character, that code will have to change.
1885    */
1886    
1887    enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1888           ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1889           ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1890           ESC_E, ESC_Q, ESC_g, ESC_k,
1891           ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1892    
1893  /* Opcode table: OP_BRA must be last, as all values >= it are used for brackets  /* Opcode table: Starting from 1 (i.e. after OP_END), the values up to
 that extract substrings. Starting from 1 (i.e. after OP_END), the values up to  
1894  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.
1895  Note that whenever this list is updated, the two macro definitions that follow  
1896  must also be updated to match. */  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1897    that follow must also be updated to match. There are also tables called
1898    "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1899    
1900  enum {  enum {
1901    OP_END,            /* 0 End of pattern */    OP_END,            /* 0 End of pattern */
# Line 509  enum { Line 1904  enum {
1904    
1905    OP_SOD,            /* 1 Start of data: \A */    OP_SOD,            /* 1 Start of data: \A */
1906    OP_SOM,            /* 2 Start of match (subject + offset): \G */    OP_SOM,            /* 2 Start of match (subject + offset): \G */
1907    OP_NOT_WORD_BOUNDARY,  /*  3 \B */    OP_SET_SOM,        /* 3 Set start of match (\K) */
1908    OP_WORD_BOUNDARY,      /*  4 \b */    OP_NOT_WORD_BOUNDARY,  /*  4 \B */
1909    OP_NOT_DIGIT,          /*  5 \D */    OP_WORD_BOUNDARY,      /*  5 \b */
1910    OP_DIGIT,              /*  6 \d */    OP_NOT_DIGIT,          /*  6 \D */
1911    OP_NOT_WHITESPACE,     /*  7 \S */    OP_DIGIT,              /*  7 \d */
1912    OP_WHITESPACE,         /*  8 \s */    OP_NOT_WHITESPACE,     /*  8 \S */
1913    OP_NOT_WORDCHAR,       /*  9 \W */    OP_WHITESPACE,         /*  9 \s */
1914    OP_WORDCHAR,           /* 10 \w */    OP_NOT_WORDCHAR,       /* 10 \W */
1915    OP_ANY,            /* 11 Match any character */    OP_WORDCHAR,           /* 11 \w */
1916    OP_ANYBYTE,        /* 12 Match any byte (\C); different to OP_ANY for UTF-8 */  
1917    OP_NOTPROP,        /* 13 \P (not Unicode property) */    OP_ANY,            /* 12 Match any character except newline (\N) */
1918    OP_PROP,           /* 14 \p (Unicode property) */    OP_ALLANY,         /* 13 Match any character */
1919    OP_EXTUNI,         /* 15 \X (extended Unicode sequence */    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1920    OP_EODN,           /* 16 End of data or \n at end of data: \Z. */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
1921    OP_EOD,            /* 17 End of data: \z */    OP_PROP,           /* 16 \p (Unicode property) */
1922      OP_ANYNL,          /* 17 \R (any newline sequence) */
1923    OP_OPT,            /* 18 Set runtime options */    OP_NOT_HSPACE,     /* 18 \H (not horizontal whitespace) */
1924    OP_CIRC,           /* 19 Start of line - varies with multiline switch */    OP_HSPACE,         /* 19 \h (horizontal whitespace) */
1925    OP_DOLL,           /* 20 End of line - varies with multiline switch */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1926    OP_CHAR,           /* 21 Match one character, casefully */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1927    OP_CHARNC,         /* 22 Match one character, caselessly */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1928    OP_NOT,            /* 23 Match anything but the following char */    OP_EODN,           /* 23 End of data or \n at end of data (\Z) */
1929      OP_EOD,            /* 24 End of data (\z) */
1930    OP_STAR,           /* 24 The maximizing and minimizing versions of */  
1931    OP_MINSTAR,        /* 25 all these opcodes must come in pairs, with */    OP_CIRC,           /* 25 Start of line - not multiline */
1932    OP_PLUS,           /* 26 the minimizing one second. */    OP_CIRCM,          /* 26 Start of line - multiline */
1933    OP_MINPLUS,        /* 27 This first set applies to single characters */    OP_DOLL,           /* 27 End of line - not multiline */
1934    OP_QUERY,          /* 28 */    OP_DOLLM,          /* 28 End of line - multiline */
1935    OP_MINQUERY,       /* 29 */    OP_CHAR,           /* 29 Match one character, casefully */
1936    OP_UPTO,           /* 30 From 0 to n matches */    OP_CHARI,          /* 30 Match one character, caselessly */
1937    OP_MINUPTO,        /* 31 */    OP_NOT,            /* 31 Match one character, not the given one, casefully */
1938    OP_EXACT,          /* 32 Exactly n matches */    OP_NOTI,           /* 32 Match one character, not the given one, caselessly */
1939    
1940    OP_NOTSTAR,        /* 33 The maximizing and minimizing versions of */    /* The following sets of 13 opcodes must always be kept in step because
1941    OP_NOTMINSTAR,     /* 34 all these opcodes must come in pairs, with */    the offset from the first one is used to generate the others. */
1942    OP_NOTPLUS,        /* 35 the minimizing one second. */  
1943    OP_NOTMINPLUS,     /* 36 This set applies to "not" single characters */    /**** Single characters, caseful, must precede the caseless ones ****/
1944    OP_NOTQUERY,       /* 37 */  
1945    OP_NOTMINQUERY,    /* 38 */    OP_STAR,           /* 33 The maximizing and minimizing versions of */
1946    OP_NOTUPTO,        /* 39 From 0 to n matches */    OP_MINSTAR,        /* 34 these six opcodes must come in pairs, with */
1947    OP_NOTMINUPTO,     /* 40 */    OP_PLUS,           /* 35 the minimizing one second. */
1948    OP_NOTEXACT,       /* 41 Exactly n matches */    OP_MINPLUS,        /* 36 */
1949      OP_QUERY,          /* 37 */
1950    OP_TYPESTAR,       /* 42 The maximizing and minimizing versions of */    OP_MINQUERY,       /* 38 */
1951    OP_TYPEMINSTAR,    /* 43 all these opcodes must come in pairs, with */  
1952    OP_TYPEPLUS,       /* 44 the minimizing one second. These codes must */    OP_UPTO,           /* 39 From 0 to n matches of one character, caseful*/
1953    OP_TYPEMINPLUS,    /* 45 be in exactly the same order as those above. */    OP_MINUPTO,        /* 40 */
1954    OP_TYPEQUERY,      /* 46 This set applies to character types such as \d */    OP_EXACT,          /* 41 Exactly n matches */
1955    OP_TYPEMINQUERY,   /* 47 */  
1956    OP_TYPEUPTO,       /* 48 From 0 to n matches */    OP_POSSTAR,        /* 42 Possessified star, caseful */
1957    OP_TYPEMINUPTO,    /* 49 */    OP_POSPLUS,        /* 43 Possessified plus, caseful */
1958    OP_TYPEEXACT,      /* 50 Exactly n matches */    OP_POSQUERY,       /* 44 Posesssified query, caseful */
1959      OP_POSUPTO,        /* 45 Possessified upto, caseful */
1960    OP_CRSTAR,         /* 51 The maximizing and minimizing versions of */  
1961    OP_CRMINSTAR,      /* 52 all these opcodes must come in pairs, with */    /**** Single characters, caseless, must follow the caseful ones */
1962    OP_CRPLUS,         /* 53 the minimizing one second. These codes must */  
1963    OP_CRMINPLUS,      /* 54 be in exactly the same order as those above. */    OP_STARI,          /* 46 */
1964    OP_CRQUERY,        /* 55 These are for character classes and back refs */    OP_MINSTARI,       /* 47 */
1965    OP_CRMINQUERY,     /* 56 */    OP_PLUSI,          /* 48 */
1966    OP_CRRANGE,        /* 57 These are different to the three sets above. */    OP_MINPLUSI,       /* 49 */
1967    OP_CRMINRANGE,     /* 58 */    OP_QUERYI,         /* 50 */
1968      OP_MINQUERYI,      /* 51 */
1969    OP_CLASS,          /* 59 Match a character class, chars < 256 only */  
1970    OP_NCLASS,         /* 60 Same, but the bitmap was created from a negative    OP_UPTOI,          /* 52 From 0 to n matches of one character, caseless */
1971                             class - the difference is relevant only when a UTF-8    OP_MINUPTOI,       /* 53 */
1972                             character > 255 is encountered. */    OP_EXACTI,         /* 54 */
1973    
1974    OP_XCLASS,         /* 61 Extended class for handling UTF-8 chars within the    OP_POSSTARI,       /* 55 Possessified star, caseless */
1975                             class. This does both positive and negative. */    OP_POSPLUSI,       /* 56 Possessified plus, caseless */
1976      OP_POSQUERYI,      /* 57 Posesssified query, caseless */
1977    OP_REF,            /* 62 Match a back reference */    OP_POSUPTOI,       /* 58 Possessified upto, caseless */
1978    OP_RECURSE,        /* 63 Match a numbered subpattern (possibly recursive) */  
1979    OP_CALLOUT,        /* 64 Call out to external function if provided */    /**** The negated ones must follow the non-negated ones, and match them ****/
1980      /**** Negated single character, caseful; must precede the caseless ones ****/
1981    OP_ALT,            /* 65 Start of alternation */  
1982    OP_KET,            /* 66 End of group that doesn't have an unbounded repeat */    OP_NOTSTAR,        /* 59 The maximizing and minimizing versions of */
1983    OP_KETRMAX,        /* 67 These two must remain together and in this */    OP_NOTMINSTAR,     /* 60 these six opcodes must come in pairs, with */
1984    OP_KETRMIN,        /* 68 order. They are for groups the repeat for ever. */    OP_NOTPLUS,        /* 61 the minimizing one second. They must be in */
1985      OP_NOTMINPLUS,     /* 62 exactly the same order as those above. */
1986    /* The assertions must come before ONCE and COND */    OP_NOTQUERY,       /* 63 */
1987      OP_NOTMINQUERY,    /* 64 */
1988    OP_ASSERT,         /* 69 Positive lookahead */  
1989    OP_ASSERT_NOT,     /* 70 Negative lookahead */    OP_NOTUPTO,        /* 65 From 0 to n matches, caseful */
1990    OP_ASSERTBACK,     /* 71 Positive lookbehind */    OP_NOTMINUPTO,     /* 66 */
1991    OP_ASSERTBACK_NOT, /* 72 Negative lookbehind */    OP_NOTEXACT,       /* 67 Exactly n matches */
1992    OP_REVERSE,        /* 73 Move pointer back - used in lookbehind assertions */  
1993      OP_NOTPOSSTAR,     /* 68 Possessified versions, caseful */
1994    /* ONCE and COND must come after the assertions, with ONCE first, as there's    OP_NOTPOSPLUS,     /* 69 */
1995    a test for >= ONCE for a subpattern that isn't an assertion. */    OP_NOTPOSQUERY,    /* 70 */
1996      OP_NOTPOSUPTO,     /* 71 */
1997    OP_ONCE,           /* 74 Once matched, don't back up into the subpattern */  
1998    OP_COND,           /* 75 Conditional group */    /**** Negated single character, caseless; must follow the caseful ones ****/
1999    OP_CREF,           /* 76 Used to hold an extraction string number (cond ref) */  
2000      OP_NOTSTARI,       /* 72 */
2001    OP_BRAZERO,        /* 77 These two must remain together and in this */    OP_NOTMINSTARI,    /* 73 */
2002    OP_BRAMINZERO,     /* 78 order. */    OP_NOTPLUSI,       /* 74 */
2003      OP_NOTMINPLUSI,    /* 75 */
2004    OP_BRANUMBER,      /* 79 Used for extracting brackets whose number is greater    OP_NOTQUERYI,      /* 76 */
2005                             than can fit into an opcode. */    OP_NOTMINQUERYI,   /* 77 */
2006    
2007    OP_BRA             /* 80 This and greater values are used for brackets that    OP_NOTUPTOI,       /* 78 From 0 to n matches, caseless */
2008                             extract substrings up to EXTRACT_BASIC_MAX. After    OP_NOTMINUPTOI,    /* 79 */
2009                             that, use is made of OP_BRANUMBER. */    OP_NOTEXACTI,      /* 80 Exactly n matches */
2010  };  
2011      OP_NOTPOSSTARI,    /* 81 Possessified versions, caseless */
2012      OP_NOTPOSPLUSI,    /* 82 */
2013      OP_NOTPOSQUERYI,   /* 83 */
2014      OP_NOTPOSUPTOI,    /* 84 */
2015    
2016      /**** Character types ****/
2017    
2018      OP_TYPESTAR,       /* 85 The maximizing and minimizing versions of */
2019      OP_TYPEMINSTAR,    /* 86 these six opcodes must come in pairs, with */
2020      OP_TYPEPLUS,       /* 87 the minimizing one second. These codes must */
2021      OP_TYPEMINPLUS,    /* 88 be in exactly the same order as those above. */
2022      OP_TYPEQUERY,      /* 89 */
2023      OP_TYPEMINQUERY,   /* 90 */
2024    
2025      OP_TYPEUPTO,       /* 91 From 0 to n matches */
2026      OP_TYPEMINUPTO,    /* 92 */
2027      OP_TYPEEXACT,      /* 93 Exactly n matches */
2028    
2029      OP_TYPEPOSSTAR,    /* 94 Possessified versions */
2030      OP_TYPEPOSPLUS,    /* 95 */
2031      OP_TYPEPOSQUERY,   /* 96 */
2032      OP_TYPEPOSUPTO,    /* 97 */
2033    
2034      /* These are used for character classes and back references; only the
2035      first six are the same as the sets above. */
2036    
2037      OP_CRSTAR,         /* 98 The maximizing and minimizing versions of */
2038      OP_CRMINSTAR,      /* 99 all these opcodes must come in pairs, with */
2039      OP_CRPLUS,         /* 100 the minimizing one second. These codes must */
2040      OP_CRMINPLUS,      /* 101 be in exactly the same order as those above. */
2041      OP_CRQUERY,        /* 102 */
2042      OP_CRMINQUERY,     /* 103 */
2043    
2044      OP_CRRANGE,        /* 104 These are different to the three sets above. */
2045      OP_CRMINRANGE,     /* 105 */
2046    
2047      /* End of quantifier opcodes */
2048    
2049      OP_CLASS,          /* 106 Match a character class, chars < 256 only */
2050      OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
2051                                  class - the difference is relevant only when a
2052                                  character > 255 is encountered. */
2053      OP_XCLASS,         /* 108 Extended class for handling > 255 chars within the
2054                                  class. This does both positive and negative. */
2055      OP_REF,            /* 109 Match a back reference, casefully */
2056      OP_REFI,           /* 110 Match a back reference, caselessly */
2057      OP_RECURSE,        /* 111 Match a numbered subpattern (possibly recursive) */
2058      OP_CALLOUT,        /* 112 Call out to external function if provided */
2059    
2060      OP_ALT,            /* 113 Start of alternation */
2061      OP_KET,            /* 114 End of group that doesn't have an unbounded repeat */
2062      OP_KETRMAX,        /* 115 These two must remain together and in this */
2063      OP_KETRMIN,        /* 116 order. They are for groups the repeat for ever. */
2064      OP_KETRPOS,        /* 117 Possessive unlimited repeat. */
2065    
2066      /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
2067      asserts must remain in order. */
2068    
2069      OP_REVERSE,        /* 118 Move pointer back - used in lookbehind assertions */
2070      OP_ASSERT,         /* 119 Positive lookahead */
2071      OP_ASSERT_NOT,     /* 120 Negative lookahead */
2072      OP_ASSERTBACK,     /* 121 Positive lookbehind */
2073      OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
2074    
2075      /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
2076      after the assertions, with ONCE first, as there's a test for >= ONCE for a
2077      subpattern that isn't an assertion. The POS versions must immediately follow
2078      the non-POS versions in each case. */
2079    
2080      OP_ONCE,           /* 123 Atomic group, contains captures */
2081      OP_ONCE_NC,        /* 124 Atomic group containing no captures */
2082      OP_BRA,            /* 125 Start of non-capturing bracket */
2083      OP_BRAPOS,         /* 126 Ditto, with unlimited, possessive repeat */
2084      OP_CBRA,           /* 127 Start of capturing bracket */
2085      OP_CBRAPOS,        /* 128 Ditto, with unlimited, possessive repeat */
2086      OP_COND,           /* 129 Conditional group */
2087    
2088      /* These five must follow the previous five, in the same order. There's a
2089      check for >= SBRA to distinguish the two sets. */
2090    
2091      OP_SBRA,           /* 130 Start of non-capturing bracket, check empty  */
2092      OP_SBRAPOS,        /* 131 Ditto, with unlimited, possessive repeat */
2093      OP_SCBRA,          /* 132 Start of capturing bracket, check empty */
2094      OP_SCBRAPOS,       /* 133 Ditto, with unlimited, possessive repeat */
2095      OP_SCOND,          /* 134 Conditional group, check empty */
2096    
2097      /* The next two pairs must (respectively) be kept together. */
2098    
2099      OP_CREF,           /* 135 Used to hold a capture number as condition */
2100      OP_NCREF,          /* 136 Same, but generated by a name reference*/
2101      OP_RREF,           /* 137 Used to hold a recursion number as condition */
2102      OP_NRREF,          /* 138 Same, but generated by a name reference*/
2103      OP_DEF,            /* 139 The DEFINE condition */
2104    
2105      OP_BRAZERO,        /* 140 These two must remain together and in this */
2106      OP_BRAMINZERO,     /* 141 order. */
2107      OP_BRAPOSZERO,     /* 142 */
2108    
2109      /* These are backtracking control verbs */
2110    
2111      OP_MARK,           /* 143 always has an argument */
2112      OP_PRUNE,          /* 144 */
2113      OP_PRUNE_ARG,      /* 145 same, but with argument */
2114      OP_SKIP,           /* 146 */
2115      OP_SKIP_ARG,       /* 147 same, but with argument */
2116      OP_THEN,           /* 148 */
2117      OP_THEN_ARG,       /* 149 same, but with argument */
2118      OP_COMMIT,         /* 150 */
2119    
2120      /* These are forced failure and success verbs */
2121    
2122      OP_FAIL,           /* 151 */
2123      OP_ACCEPT,         /* 152 */
2124      OP_ASSERT_ACCEPT,  /* 153 Used inside assertions */
2125      OP_CLOSE,          /* 154 Used before OP_ACCEPT to close open captures */
2126    
2127      /* This is used to skip a subpattern with a {0} quantifier */
2128    
2129      OP_SKIPZERO,       /* 155 */
2130    
2131      /* This is not an opcode, but is used to check that tables indexed by opcode
2132      are the correct length, in order to catch updating errors - there have been
2133      some in the past. */
2134    
2135  /* WARNING WARNING WARNING: There is an implicit assumption in pcre.c and    OP_TABLE_LENGTH
2136  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. */  
2137    
2138  #define EXTRACT_BASIC_MAX  100  /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2139    definitions that follow must also be updated to match. There are also tables
2140    called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
2141    
2142    
2143  /* 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
2144  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
2145    only in pcre_printint.c, which fills out the full names in many cases (and in
2146    some cases doesn't actually use these names at all). */
2147    
2148  #define OP_NAME_LIST \  #define OP_NAME_LIST \
2149    "End", "\\A", "\\G", "\\B", "\\b", "\\D", "\\d",                \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
2150    "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte",                   \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
2151    "notprop", "prop", "extuni",                                    \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
2152    "\\Z", "\\z",                                                   \    "extuni",  "\\Z", "\\z",                                        \
2153    "Opt", "^", "$", "char", "charnc", "not",                       \    "^", "^", "$", "$", "char", "chari", "not", "noti",             \
2154    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2155    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "{", "{", "{",                                                  \
2156      "*+","++", "?+", "{",                                           \
2157      "*", "*?", "+", "+?", "?", "??",                                \
2158      "{", "{", "{",                                                  \
2159      "*+","++", "?+", "{",                                           \
2160      "*", "*?", "+", "+?", "?", "??",                                \
2161      "{", "{", "{",                                                  \
2162      "*+","++", "?+", "{",                                           \
2163      "*", "*?", "+", "+?", "?", "??",                                \
2164      "{", "{", "{",                                                  \
2165      "*+","++", "?+", "{",                                           \
2166    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
2167      "*+","++", "?+", "{",                                           \
2168    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \
2169    "class", "nclass", "xclass", "Ref", "Recurse", "Callout",       \    "class", "nclass", "xclass", "Ref", "Refi",                     \
2170    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",     \    "Recurse", "Callout",                                           \
2171    "AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cond ref",\    "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos",                  \
2172    "Brazero", "Braminzero", "Branumber", "Bra"    "Reverse", "Assert", "Assert not", "AssertB", "AssertB not",    \
2173      "Once", "Once_NC",                                              \
2174      "Bra", "BraPos", "CBra", "CBraPos",                             \
2175      "Cond",                                                         \
2176      "SBra", "SBraPos", "SCBra", "SCBraPos",                         \
2177      "SCond",                                                        \
2178      "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \
2179      "Brazero", "Braminzero", "Braposzero",                          \
2180      "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \
2181      "*THEN", "*THEN", "*COMMIT", "*FAIL",                           \
2182      "*ACCEPT", "*ASSERT_ACCEPT",                                    \
2183      "Close", "Skip zero"
2184    
2185    
2186  /* This macro defines the length of fixed length operations in the compiled  /* This macro defines the length of fixed length operations in the compiled
# Line 657  in UTF-8 mode. The code that uses this t Line 2194  in UTF-8 mode. The code that uses this t
2194    
2195  #define OP_LENGTHS \  #define OP_LENGTHS \
2196    1,                             /* End                                    */ \    1,                             /* End                                    */ \
2197    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                     */ \
2198    1, 1,                          /* Any, Anybyte                           */ \    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
2199    3, 3, 1,                       /* NOTPROP, PROP, EXTUNI                  */ \    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \
2200    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    3, 3,                          /* \P, \p                                 */ \
2201      1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
2202      1,                             /* \X                                     */ \
2203      1, 1, 1, 1, 1, 1,              /* \Z, \z, ^, ^M, $, $M                   */ \
2204    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
2205    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Chari  - the minimum length            */ \
2206    2,                             /* not                                    */ \    2,                             /* not                                    */ \
2207    /* Positive single-char repeats                            ** These are  */ \    2,                             /* noti                                   */ \
2208    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??      ** minima in  */ \    /* Positive single-char repeats                             ** These are */ \
2209    4, 4, 4,                       /* upto, minupto, exact     ** UTF-8 mode */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2210      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2211      2+IMM2_SIZE,                   /* exact                                  */ \
2212      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2213      2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2214      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2215      2+IMM2_SIZE,                   /* exact I                                */ \
2216      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2217    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2218    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2219    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2220      2+IMM2_SIZE,                   /* NOT exact                              */ \
2221      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2222      2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2223      2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2224      2+IMM2_SIZE,                   /* NOT exact I                            */ \
2225      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2226    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2227    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2228    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2229      2+IMM2_SIZE,                   /* Type exact                             */ \
2230      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2231    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2232    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2233    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2234   33,                             /* CLASS                                  */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2235   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2236    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2237    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2238      1+IMM2_SIZE,                   /* REFI                                   */ \
2239    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2240    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2241    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
2242    1+LINK_SIZE,                   /* Ket                                    */ \    1+LINK_SIZE,                   /* Ket                                    */ \
2243    1+LINK_SIZE,                   /* KetRmax                                */ \    1+LINK_SIZE,                   /* KetRmax                                */ \
2244    1+LINK_SIZE,                   /* KetRmin                                */ \    1+LINK_SIZE,                   /* KetRmin                                */ \
2245      1+LINK_SIZE,                   /* KetRpos                                */ \
2246      1+LINK_SIZE,                   /* Reverse                                */ \
2247    1+LINK_SIZE,                   /* Assert                                 */ \    1+LINK_SIZE,                   /* Assert                                 */ \
2248    1+LINK_SIZE,                   /* Assert not                             */ \    1+LINK_SIZE,                   /* Assert not                             */ \
2249    1+LINK_SIZE,                   /* Assert behind                          */ \    1+LINK_SIZE,                   /* Assert behind                          */ \
2250    1+LINK_SIZE,                   /* Assert behind not                      */ \    1+LINK_SIZE,                   /* Assert behind not                      */ \
2251    1+LINK_SIZE,                   /* Reverse                                */ \    1+LINK_SIZE,                   /* ONCE                                   */ \
2252    1+LINK_SIZE,                   /* Once                                   */ \    1+LINK_SIZE,                   /* ONCE_NC                                */ \
2253      1+LINK_SIZE,                   /* BRA                                    */ \
2254      1+LINK_SIZE,                   /* BRAPOS                                 */ \
2255      1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2256      1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2257    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2258    3,                             /* CREF                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2259    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2260    3,                             /* BRANUMBER                              */ \    1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2261    1+LINK_SIZE                    /* BRA                                    */ \    1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2262      1+LINK_SIZE,                   /* SCOND                                  */ \
2263      1+IMM2_SIZE, 1+IMM2_SIZE,      /* CREF, NCREF                            */ \
2264  /* A magic value for OP_CREF to indicate the "in recursion" condition. */    1+IMM2_SIZE, 1+IMM2_SIZE,      /* RREF, NRREF                            */ \
2265      1,                             /* DEF                                    */ \
2266  #define CREF_RECURSE  0xffff    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2267      3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2268  /* Error code numbers. They are given names so that they can more easily be    1, 3,                          /* SKIP, SKIP_ARG                         */ \
2269  tracked. */    1, 3,                          /* THEN, THEN_ARG                         */ \
2270      1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2271      1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2272    
2273    /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
2274    condition. */
2275    
2276    #define RREF_ANY  0xffff
2277    
2278    /* Compile time error code numbers. They are given names so that they can more
2279    easily be tracked. When a new number is added, the table called eint in
2280    pcreposix.c must be updated. */
2281    
2282  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,
2283         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
2284         ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,         ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
2285         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2286         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47 };         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2287           ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2288           ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2289           ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERR78, ERRCOUNT };
2290    
2291    /* JIT compiling modes. The function list is indexed by them. */
2292    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2293           JIT_NUMBER_OF_COMPILE_MODES };
2294    
2295  /* 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
2296  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
2297  offset to the name table so that if a regex is compiled on one host, saved, and  offset to the name table so that if a regex is compiled on one host, saved, and
2298  then run on another where the size of pointers is different, all might still  then run on another where the size of pointers is different, all might still
2299  be well. For the case of compiled-on-4 and run-on-8, we include an extra  be well.
2300  pointer that is always NULL. For future-proofing, a few dummy fields were  
2301  originally included - even though you can never get this planning right - but  The size of the structure must be a multiple of 8 bytes. For the case of
2302  there is only one left now.  compiled-on-4 and run-on-8, we include an extra pointer that is always NULL so
2303    that there are an even number of pointers which therefore are a multiple of 8
2304  NOTE NOTE NOTE:  bytes.
2305  Because people can now save and re-use compiled patterns, any additions to this  
2306  structure should be made at the end, and something earlier (e.g. a new  It is necessary to fork the struct for the 32 bit library, since it needs to
2307  flag in the options or one of the dummy fields) should indicate that the new  use pcre_uint32 for first_char and req_char. We can't put an ifdef inside the
2308  fields are present. Currently PCRE always sets the dummy fields to zero.  typedef because pcretest needs access to the struct of the 8-, 16- and 32-bit
2309  NOTE NOTE NOTE:  variants.
2310    
2311    *** WARNING ***
2312    When new fields are added to these structures, remember to adjust the code in
2313    pcre_byte_order.c that is concerned with swapping the byte order of the fields
2314    when a compiled regex is reloaded on a host with different endianness.
2315    *** WARNING ***
2316    There is also similar byte-flipping code in pcretest.c, which is used for
2317    testing the byte-flipping features. It must also be kept in step.
2318    *** WARNING ***
2319  */  */
2320    
2321  typedef struct real_pcre {  typedef struct real_pcre8_or_16 {
2322    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2323    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2324    pcre_uint32 options;    pcre_uint32 options;            /* Public options */
2325    pcre_uint32 dummy1;             /* For future use, maybe */    pcre_uint32 flags;              /* Private flags */
2326      pcre_uint32 limit_match;        /* Limit set from regex */
2327      pcre_uint32 limit_recursion;    /* Limit set from regex */
2328      pcre_uint16 first_char;         /* Starting character */
2329      pcre_uint16 req_char;           /* This character must be seen */
2330      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2331      pcre_uint16 top_bracket;        /* Highest numbered group */
2332      pcre_uint16 top_backref;        /* Highest numbered back reference */
2333      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2334      pcre_uint16 name_entry_size;    /* Size of any name items */
2335      pcre_uint16 name_count;         /* Number of name items */
2336      pcre_uint16 ref_count;          /* Reference count */
2337      pcre_uint16 dummy1;             /* To ensure size is a multiple of 8 */
2338      pcre_uint16 dummy2;             /* To ensure size is a multiple of 8 */
2339      pcre_uint16 dummy3;             /* To ensure size is a multiple of 8 */
2340      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2341      void             *nullpad;      /* NULL padding */
2342    } real_pcre8_or_16;
2343    
2344    pcre_uint16 top_bracket;  typedef struct real_pcre8_or_16 real_pcre;
2345    pcre_uint16 top_backref;  typedef struct real_pcre8_or_16 real_pcre16;
2346    pcre_uint16 first_byte;  
2347    pcre_uint16 req_byte;  typedef struct real_pcre32 {
2348      pcre_uint32 magic_number;
2349      pcre_uint32 size;               /* Total that was malloced */
2350      pcre_uint32 options;            /* Public options */
2351      pcre_uint32 flags;              /* Private flags */
2352      pcre_uint32 limit_match;        /* Limit set from regex */
2353      pcre_uint32 limit_recursion;    /* Limit set from regex */
2354      pcre_uint32 first_char;         /* Starting character */
2355      pcre_uint32 req_char;           /* This character must be seen */
2356      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2357      pcre_uint16 top_bracket;        /* Highest numbered group */
2358      pcre_uint16 top_backref;        /* Highest numbered back reference */
2359    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2360    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2361    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2362    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2363      pcre_uint16 dummy;              /* To ensure size is a multiple of 8 */
2364      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2365      void             *nullpad;      /* NULL padding */
2366    } real_pcre32;
2367    
2368    #if defined COMPILE_PCRE8
2369    #define REAL_PCRE real_pcre
2370    #elif defined COMPILE_PCRE16
2371    #define REAL_PCRE real_pcre16
2372    #elif defined COMPILE_PCRE32
2373    #define REAL_PCRE real_pcre32
2374    #endif
2375    
2376    /* Assert that the size of REAL_PCRE is divisible by 8 */
2377    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2378    
2379    const unsigned char *tables;    /* Pointer to tables or NULL for std */  /* Needed in pcretest to access some fields in the real_pcre* structures
2380    const unsigned char *nullpad;   /* NULL padding */   * directly. They're unified for 8/16/32 bits since the structs only differ
2381  } real_pcre;   * after these fields; if that ever changes, need to fork those defines into
2382     * 8/16 and 32 bit versions. */
2383    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2384    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2385    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2386    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2387    
2388  /* 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
2389  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
2390    
2391  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2392    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2393    pcre_uint32 options;    pcre_uint32 flags;              /* Private flags */
2394    uschar start_bits[32];    pcre_uint8 start_bits[32];      /* Starting char bits */
2395      pcre_uint32 minlength;          /* Minimum subject length */
2396  } pcre_study_data;  } pcre_study_data;
2397    
2398    /* Structure for building a chain of open capturing subpatterns during
2399    compiling, so that instructions to close them can be compiled when (*ACCEPT) is
2400    encountered. This is also used to identify subpatterns that contain recursive
2401    back references to themselves, so that they can be made atomic. */
2402    
2403    typedef struct open_capitem {
2404      struct open_capitem *next;    /* Chain link */
2405      pcre_uint16 number;           /* Capture number */
2406      pcre_uint16 flag;             /* Set TRUE if recursive back ref */
2407    } open_capitem;
2408    
2409  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2410  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2411    
2412  typedef struct compile_data {  typedef struct compile_data {
2413    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2414    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2415    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2416    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2417    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_workspace;/* The start of working space */
2418    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_code;     /* The start of the compiled code */
2419    uschar *name_table;           /* The name/number table */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2420    int  names_found;             /* Number of entries so far */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2421    int  name_entry_size;         /* Size of each entry */    open_capitem *open_caps;          /* Chain of open capture items */
2422    int  top_backref;             /* Maximum back reference */    pcre_uchar *hwm;                  /* High watermark of workspace */
2423    unsigned int backref_map;     /* Bitmap of low back refs */    pcre_uchar *name_table;           /* The name/number table */
2424    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  names_found;                 /* Number of entries so far */
2425    BOOL nopartial;               /* Set TRUE if partial won't work */    int  name_entry_size;             /* Size of each entry */
2426      int  workspace_size;              /* Size of workspace */
2427      unsigned int bracount;            /* Count of capturing parens as we compile */
2428      int  final_bracount;              /* Saved value after first pass */
2429      int  max_lookbehind;              /* Maximum lookbehind (characters) */
2430      int  top_backref;                 /* Maximum back reference */
2431      unsigned int backref_map;         /* Bitmap of low back refs */
2432      int  assert_depth;                /* Depth of nested assertions */
2433      pcre_uint32 external_options;     /* External (initial) options */
2434      pcre_uint32 external_flags;       /* External flag bits to be set */
2435      int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2436      BOOL had_accept;                  /* (*ACCEPT) encountered */
2437      BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2438      BOOL check_lookbehind;            /* Lookbehinds need later checking */
2439      int  nltype;                      /* Newline type */
2440      int  nllen;                       /* Newline string length */
2441      pcre_uchar nl[4];                 /* Newline string when fixed length */
2442  } compile_data;  } compile_data;
2443    
2444  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
2445  branches, for testing for left recursion. */  branches, for testing for left recursion while compiling. */
2446    
2447  typedef struct branch_chain {  typedef struct branch_chain {
2448    struct branch_chain *outer;    struct branch_chain *outer;
2449    uschar *current;    pcre_uchar *current_branch;
2450  } branch_chain;  } branch_chain;
2451    
2452  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
2453  call within the pattern. */  call within the pattern; used by pcre_exec(). */
2454    
2455  typedef struct recursion_info {  typedef struct recursion_info {
2456    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2457    int group_num;                /* Number of group that was called */    unsigned int group_num;         /* Number of group that was called */
2458    const uschar *after_call;     /* "Return value": points after the call in the expr */    int *offset_save;               /* Pointer to start of saved offsets */
2459    USPTR save_start;             /* Old value of md->start_match */    int saved_max;                  /* Number of saved offsets */
2460    int *offset_save;             /* Pointer to start of saved offsets */    int saved_capture_last;         /* Last capture number */
2461    int saved_max;                /* Number of saved offsets */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
2462  } recursion_info;  } recursion_info;
2463    
2464  /* When compiling in a mode that doesn't use recursive calls to match(),  /* A similar structure for pcre_dfa_exec(). */
2465  a structure is used to remember local variables on the heap. It is defined in  
2466  pcre.c, close to the match() function, so that it is easy to keep it in step  typedef struct dfa_recursion_info {
2467  with any changes of local variable. However, the pointer to the current frame    struct dfa_recursion_info *prevrec;
2468  must be saved in some "static" place over a longjmp(). We declare the    int group_num;
2469  structure here so that we can put a pointer in the match_data structure.    PCRE_PUCHAR subject_position;
2470  NOTE: This isn't used for a "normal" compilation of pcre. */  } dfa_recursion_info;
2471    
2472    /* Structure for building a chain of data for holding the values of the subject
2473    pointer at the start of each subpattern, so as to detect when an empty string
2474    has been matched by a subpattern - to break infinite loops; used by
2475    pcre_exec(). */
2476    
2477    typedef struct eptrblock {
2478      struct eptrblock *epb_prev;
2479      PCRE_PUCHAR epb_saved_eptr;
2480    } eptrblock;
2481    
 struct heapframe;  
2482    
2483  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2484  doing traditional NFA matching, so that they are thread-safe. */  doing traditional NFA matching, so that they are thread-safe. */
# Line 813  typedef struct match_data { Line 2487  typedef struct match_data {
2487    unsigned long int match_call_count;      /* As it says */    unsigned long int match_call_count;      /* As it says */
2488    unsigned long int match_limit;           /* As it says */    unsigned long int match_limit;           /* As it says */
2489    unsigned long int match_limit_recursion; /* As it says */    unsigned long int match_limit_recursion; /* As it says */
2490    int   *offset_vector;         /* Offset vector */    int   *offset_vector;           /* Offset vector */
2491    int    offset_end;            /* One past the end */    int    offset_end;              /* One past the end */
2492    int    offset_max;            /* The maximum usable for return data */    int    offset_max;              /* The maximum usable for return data */
2493    const uschar *lcc;            /* Points to lower casing table */    int    nltype;                  /* Newline type */
2494    const uschar *ctypes;         /* Points to table of type maps */    int    nllen;                   /* Newline string length */
2495    BOOL   offset_overflow;       /* Set if too many extractions */    int    name_count;              /* Number of names in name table */
2496    BOOL   notbol;                /* NOTBOL flag */    int    name_entry_size;         /* Size of entry in names table */
2497    BOOL   noteol;                /* NOTEOL flag */    unsigned int skip_arg_count;    /* For counting SKIP_ARGs */
2498    BOOL   utf8;                  /* UTF8 flag */    unsigned int ignore_skip_arg;   /* For re-run when SKIP arg name not found */
2499    BOOL   endonly;               /* Dollar not before final \n */    pcre_uchar *name_table;         /* Table of names */
2500    BOOL   notempty;              /* Empty string match not wanted */    pcre_uchar nl[4];               /* Newline string when fixed */
2501    BOOL   partial;               /* PARTIAL flag */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2502    BOOL   hitend;                /* Hit the end of the subject at some point */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2503    const uschar *start_code;     /* For use when recursing */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2504    USPTR  start_subject;         /* Start of the subject string */    BOOL   notbol;                  /* NOTBOL flag */
2505    USPTR  end_subject;           /* End of the subject string */    BOOL   noteol;                  /* NOTEOL flag */
2506    USPTR  start_match;           /* Start of this match attempt */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2507    USPTR  end_match_ptr;         /* Subject position at end match */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2508    int    end_offset_top;        /* Highwater mark at end of match */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2509    int    capture_last;          /* Most recent capture number */    BOOL   endonly;                 /* Dollar not before final \n */
2510    int    start_offset;          /* The start offset value */    BOOL   notempty;                /* Empty string match not wanted */
2511    recursion_info *recursive;    /* Linked list of recursion data */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2512    void  *callout_data;          /* To pass back to callouts */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2513    struct heapframe *thisframe;  /* Used only when compiling for no recursion */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2514      BOOL   hasthen;                 /* Pattern contains (*THEN) */
2515      const  pcre_uchar *start_code;  /* For use when recursing */
2516      PCRE_PUCHAR start_subject;      /* Start of the subject string */
2517      PCRE_PUCHAR end_subject;        /* End of the subject string */
2518      PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2519      PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2520      PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2521      int    partial;                 /* PARTIAL options */
2522      int    end_offset_top;          /* Highwater mark at end of match */
2523      pcre_int32 capture_last;        /* Most recent capture number + overflow flag */
2524      int    start_offset;            /* The start offset value */
2525      int    match_function_type;     /* Set for certain special calls of MATCH() */
2526      eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2527      int    eptrn;                   /* Next free eptrblock */
2528      recursion_info *recursive;      /* Linked list of recursion data */
2529      void  *callout_data;            /* To pass back to callouts */
2530      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2531      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2532      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2533    #ifdef NO_RECURSE
2534      void  *match_frames_base;       /* For remembering malloc'd frames */
2535    #endif
2536  } match_data;  } match_data;
2537    
2538  /* 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
2539  functions. */  functions. */
2540    
2541  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2542    const uschar *start_code;     /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2543    const uschar *start_subject;  /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2544    const uschar *end_subject;    /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2545    const uschar *tables;         /* Character tables */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2546    int   moptions;               /* Match options */    const pcre_uint8 *tables;         /* Character tables */
2547    int   poptions;               /* Pattern options */    int   start_offset;               /* The start offset value */
2548    void  *callout_data;          /* To pass back to callouts */    int   moptions;                   /* Match options */
2549      int   poptions;                   /* Pattern options */
2550      int   nltype;                     /* Newline type */
2551      int   nllen;                      /* Newline string length */
2552      pcre_uchar nl[4];                 /* Newline string when fixed */
2553      void *callout_data;               /* To pass back to callouts */
2554      dfa_recursion_info *recursive;    /* Linked list of recursion data */
2555  } dfa_match_data;  } dfa_match_data;
2556    
2557  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 858  typedef struct dfa_match_data { Line 2560  typedef struct dfa_match_data {
2560  #define ctype_letter  0x02  #define ctype_letter  0x02
2561  #define ctype_digit   0x04  #define ctype_digit   0x04
2562  #define ctype_xdigit  0x08  #define ctype_xdigit  0x08
2563  #define ctype_word    0x10   /* alphameric or '_' */  #define ctype_word    0x10   /* alphanumeric or '_' */
2564  #define ctype_meta    0x80   /* regexp meta char or zero (end pattern) */  #define ctype_meta    0x80   /* regexp meta char or zero (end pattern) */
2565    
2566  /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set  /* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
# Line 885  total length. */ Line 2587  total length. */
2587  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2588  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2589    
2590    /* Internal function and data prefixes. */
2591    
2592    #if defined COMPILE_PCRE8
2593    #ifndef PUBL
2594    #define PUBL(name) pcre_##name
2595    #endif
2596    #ifndef PRIV
2597    #define PRIV(name) _pcre_##name
2598    #endif
2599    #elif defined COMPILE_PCRE16
2600    #ifndef PUBL
2601    #define PUBL(name) pcre16_##name
2602    #endif
2603    #ifndef PRIV
2604    #define PRIV(name) _pcre16_##name
2605    #endif
2606    #elif defined COMPILE_PCRE32
2607    #ifndef PUBL
2608    #define PUBL(name) pcre32_##name
2609    #endif
2610    #ifndef PRIV
2611    #define PRIV(name) _pcre32_##name
2612    #endif
2613    #else
2614    #error Unsupported compiling mode
2615    #endif /* COMPILE_PCRE[8|16|32] */
2616    
2617  /* Layout of the UCP type table that translates property names into types and  /* Layout of the UCP type table that translates property names into types and
2618  codes. */  codes. Each entry used to point directly to a name, but to reduce the number of
2619    relocations in shared libraries, it now has an offset into a single string
2620    instead. */
2621    
2622  typedef struct {  typedef struct {
2623    const char *name;    pcre_uint16 name_offset;
2624    pcre_uint16 type;    pcre_uint16 type;
2625    pcre_uint16 value;    pcre_uint16 value;
2626  } ucp_type_table;  } ucp_type_table;
# Line 900  of the exported public functions. They h Line 2631  of the exported public functions. They h
2631  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
2632  pcre_tables.c module. */  pcre_tables.c module. */
2633    
2634  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2635  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2636  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2637  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2638    extern const int            PRIV(utf8_table3)[];
2639  extern const int    _pcre_utf8_table1_size;  extern const pcre_uint8     PRIV(utf8_table4)[];
2640    #endif /* COMPILE_PCRE8 */
2641  extern const ucp_type_table _pcre_utt[];  
2642  extern const int _pcre_utt_size;  extern const char           PRIV(utt_names)[];
2643    extern const ucp_type_table PRIV(utt)[];
2644    extern const int            PRIV(utt_size);
2645    
2646  extern const uschar _pcre_default_tables[];  extern const pcre_uint8     PRIV(OP_lengths)[];
2647    extern const pcre_uint8     PRIV(default_tables)[];
2648    
2649  extern const uschar _pcre_OP_lengths[];  extern const pcre_uint32    PRIV(hspace_list)[];
2650    extern const pcre_uint32    PRIV(vspace_list)[];
2651    
2652    
2653  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2654  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
2655  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2656    
2657  extern int         _pcre_ord2utf8(int, uschar *);  /* String comparison functions. */
2658  extern real_pcre * _pcre_try_flipped(const real_pcre *, real_pcre *,  #if defined COMPILE_PCRE8
2659                       const pcre_study_data *, pcre_study_data *);  
2660  extern int         _pcre_ucp_findprop(const int, int *, int *);  #define STRCMP_UC_UC(str1, str2) \
2661  extern int         _pcre_ucp_othercase(const int);    strcmp((char *)(str1), (char *)(str2))
2662  extern int         _pcre_valid_utf8(const uschar *, int);  #define STRCMP_UC_C8(str1, str2) \
2663  extern BOOL        _pcre_xclass(int, const uschar *);    strcmp((char *)(str1), (str2))
2664    #define STRNCMP_UC_UC(str1, str2, num) \
2665      strncmp((char *)(str1), (char *)(str2), (num))
2666    #define STRNCMP_UC_C8(str1, str2, num) \
2667      strncmp((char *)(str1), (str2), (num))
2668    #define STRLEN_UC(str) strlen((const char *)str)
2669    
2670    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2671    
2672    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2673                               const pcre_uchar *);
2674    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2675                               const char *);
2676    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2677                               const pcre_uchar *, unsigned int num);
2678    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2679                               const char *, unsigned int num);
2680    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2681    
2682    #define STRCMP_UC_UC(str1, str2) \
2683      PRIV(strcmp_uc_uc)((str1), (str2))
2684    #define STRCMP_UC_C8(str1, str2) \
2685      PRIV(strcmp_uc_c8)((str1), (str2))
2686    #define STRNCMP_UC_UC(str1, str2, num) \
2687      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2688    #define STRNCMP_UC_C8(str1, str2, num) \
2689      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2690    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2691    
2692    #endif /* COMPILE_PCRE[8|16|32] */
2693    
2694    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2695    
2696    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2697    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2698    
2699    #elif defined COMPILE_PCRE32
2700    
2701    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2702                               const pcre_uchar *);
2703    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2704                               const char *);
2705    
2706    #define STRCMP_UC_UC_TEST(str1, str2) \
2707      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2708    #define STRCMP_UC_C8_TEST(str1, str2) \
2709      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2710    
2711    #endif /* COMPILE_PCRE[8|16|32] */
2712    
2713    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2714    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2715                               int *, BOOL);
2716    extern unsigned int      PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2717    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2718    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2719                               int *, BOOL);
2720    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2721    
2722    #ifdef SUPPORT_JIT
2723    extern void              PRIV(jit_compile)(const REAL_PCRE *,
2724                               PUBL(extra) *, int);
2725    extern int               PRIV(jit_exec)(const PUBL(extra) *,
2726                               const pcre_uchar *, int, int, int, int *, int);
2727    extern void              PRIV(jit_free)(void *);
2728    extern int               PRIV(jit_get_size)(void *);
2729    extern const char*       PRIV(jit_get_target)(void);
2730    #endif
2731    
2732    /* Unicode character database (UCD) */
2733    
2734    typedef struct {
2735      pcre_uint8 script;     /* ucp_Arabic, etc. */
2736      pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2737      pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2738      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2739      pcre_int32 other_case; /* offset to other case, or zero if none */
2740    } ucd_record;
2741    
2742    extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2743    extern const ucd_record  PRIV(ucd_records)[];
2744    extern const pcre_uint8  PRIV(ucd_stage1)[];
2745    extern const pcre_uint16 PRIV(ucd_stage2)[];
2746    extern const pcre_uint32 PRIV(ucp_gentype)[];
2747    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2748    #ifdef SUPPORT_JIT
2749    extern const int         PRIV(ucp_typerange)[];
2750    #endif
2751    
2752    #ifdef SUPPORT_UCP
2753    /* UCD access macros */
2754    
2755    #define UCD_BLOCK_SIZE 128
2756    #define GET_UCD(ch) (PRIV(ucd_records) + \
2757            PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2758            UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2759    
2760    #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2761    #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2762    #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2763    #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2764    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2765    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2766    
2767    #endif /* SUPPORT_UCP */
2768    
2769  #endif  #endif
2770    

Legend:
Removed from v.87  
changed lines
  Added in v.1320

  ViewVC Help
Powered by ViewVC 1.1.5