/[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 96 by nigel, Fri Mar 2 13:10:43 2007 UTC revision 1248 by ph10, Wed Feb 13 17:36:38 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) 1997-2012 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 40  POSSIBILITY OF SUCH DAMAGE. Line 40  POSSIBILITY OF SUCH DAMAGE.
40    
41  /* This header contains definitions that are shared between the different  /* This header contains definitions that are shared between the different
42  modules, but which are not relevant to the exported API. This includes some  modules, but which are not relevant to the exported API. This includes some
43  functions whose names all begin with "_pcre_". */  functions whose names all begin with "_pcre_", "_pcre16_" or "_pcre32_"
44    depending on the PRIV macro. */
45    
46  #ifndef PCRE_INTERNAL_H  #ifndef PCRE_INTERNAL_H
47  #define PCRE_INTERNAL_H  #define PCRE_INTERNAL_H
48    
49  /* Define DEBUG to get debugging output on stdout. */  /* Define PCRE_DEBUG to get debugging output on stdout. */
50    
51  #if 0  #if 0
52  #define DEBUG  #define PCRE_DEBUG
53    #endif
54    
55    /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
56    
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
# Line 60  It turns out that the Mac Debugging.h he Line 95  It turns out that the Mac Debugging.h he
95  be absolutely sure we get our version. */  be absolutely sure we get our version. */
96    
97  #undef DPRINTF  #undef DPRINTF
98  #ifdef DEBUG  #ifdef PCRE_DEBUG
99  #define DPRINTF(p) printf p  #define DPRINTF(p) printf p
100  #else  #else
101  #define DPRINTF(p) /* Nothing */  #define DPRINTF(p) /* Nothing */
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
120    
121    /* 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  #endif
173    
174  /* We need to have types that specify unsigned 16-bit and 32-bit integers. We  /* When compiling with the MSVC compiler, it is sometimes necessary to include
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  #elif UINT_MAX == 65535  #elif UINT_MAX == 65535
200    typedef unsigned int pcre_uint16;    typedef unsigned int pcre_uint16;
201      typedef int pcre_int16;
202  #else  #else
203    #error Cannot determine a type for 16-bit unsigned integers  # error Cannot determine a type for 16-bit unsigned integers
204  #endif  #endif
205    
206  #if UINT_MAX == 4294967295  #if UINT_MAX == 4294967295
207    typedef unsigned int pcre_uint32;    typedef unsigned int pcre_uint32;
208      typedef int pcre_int32;
209  #elif ULONG_MAX == 4294967295  #elif ULONG_MAX == 4294967295
210    typedef unsigned long int pcre_uint32;    typedef unsigned long int pcre_uint32;
211      typedef long int pcre_int32;
212    #else
213    # error Cannot determine a type for 32-bit unsigned integers
214    #endif
215    
216    /* When checking for integer overflow in pcre_compile(), we need to handle
217    large integers. If a 64-bit integer type is available, we can use that.
218    Otherwise we have to cast to double, which of course requires floating point
219    arithmetic. Handle this by defining a macro for the appropriate type. If
220    stdint.h is available, include it; it may define INT64_MAX. Systems that do not
221    have stdint.h (e.g. Solaris) may have inttypes.h. The macro int64_t may be set
222    by "configure". */
223    
224    #if defined HAVE_STDINT_H
225    #include <stdint.h>
226    #elif defined HAVE_INTTYPES_H
227    #include <inttypes.h>
228    #endif
229    
230    #if defined INT64_MAX || defined int64_t
231    #define INT64_OR_DOUBLE int64_t
232  #else  #else
233    #error Cannot determine a type for 32-bit unsigned integers  #define INT64_OR_DOUBLE double
234  #endif  #endif
235    
236  /* All character handling must be done as unsigned characters. Otherwise there  /* All character handling must be done as unsigned characters. Otherwise there
237  are problems with top-bit-set characters and functions such as isspace().  are problems with top-bit-set characters and functions such as isspace().
238  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 *,
239  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
240  to save lots of typing. I tried "uchar", but it causes problems on Digital  called pcre_uchar.
241  Unix, where it is defined in sys/types, so use "uschar" instead. */  
242    The IN_UCHARS macro multiply its argument with the byte size of the current
243    pcre_uchar type. Useful for memcpy and such operations, whose require the
244    byte size of their input/output buffers.
245    
246    The MAX_255 macro checks whether its pcre_uchar input is less than 256.
247    
248    The TABLE_GET macro is designed for accessing elements of tables whose contain
249    exactly 256 items. When the character is able to contain more than 256
250    items, some check is needed before accessing these tables.
251    */
252    
253  typedef unsigned char uschar;  #if defined COMPILE_PCRE8
254    
255    typedef unsigned char pcre_uchar;
256    #define IN_UCHARS(x) (x)
257    #define MAX_255(c) 1
258    #define TABLE_GET(c, table, default) ((table)[c])
259    
260    #elif defined COMPILE_PCRE16
261    
262    #if USHRT_MAX != 65535
263    /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
264    pcre.h(.in) and disable (comment out) this message. */
265    #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
266    #endif
267    
268    typedef pcre_uint16 pcre_uchar;
269    #define UCHAR_SHIFT (1)
270    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
271    #define MAX_255(c) ((c) <= 255u)
272    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
273    
274    #elif defined COMPILE_PCRE32
275    
276    typedef pcre_uint32 pcre_uchar;
277    #define UCHAR_SHIFT (2)
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    #else
283    #error Unsupported compiling mode
284    #endif /* COMPILE_PCRE[8|16|32] */
285    
286  /* This is an unsigned int value that no character can ever have. UTF-8  /* This is an unsigned int value that no character can ever have. UTF-8
287  characters only go up to 0x7fffffff (though Unicode doesn't go beyond  characters only go up to 0x7fffffff (though Unicode doesn't go beyond
# Line 125  characters only go up to 0x7fffffff (tho Line 290  characters only go up to 0x7fffffff (tho
290  #define NOTACHAR 0xffffffff  #define NOTACHAR 0xffffffff
291    
292  /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,  /* PCRE is able to support several different kinds of newline (CR, LF, CRLF,
293  and "all" at present). The following macros are used to package up testing for  "any" and "anycrlf" at present). The following macros are used to package up
294  newlines. NLBLOCK, PSSTART, and PSEND are defined in the various modules to  testing for newlines. NLBLOCK, PSSTART, and PSEND are defined in the various
295  indicate in which datablock the parameters exist, and what the start/end of  modules to indicate in which datablock the parameters exist, and what the
296  string field names are. */  start/end of string field names are. */
297    
298  #define NLTYPE_FIXED   0     /* Newline is a fixed length string */  #define NLTYPE_FIXED    0     /* Newline is a fixed length string */
299  #define NLTYPE_ANY     1     /* Newline is any Unicode line ending */  #define NLTYPE_ANY      1     /* Newline is any Unicode line ending */
300    #define NLTYPE_ANYCRLF  2     /* Newline is CR, LF, or CRLF */
301    
302  /* This macro checks for a newline at the given position */  /* This macro checks for a newline at the given position */
303    
304  #define IS_NEWLINE(p) \  #define IS_NEWLINE(p) \
305    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
306      ((p) < NLBLOCK->PSEND && \      ((p) < NLBLOCK->PSEND && \
307       _pcre_is_newline((p), NLBLOCK->PSEND, &(NLBLOCK->nllen), utf8) \       PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
308      ) \         &(NLBLOCK->nllen), utf)) \
309      : \      : \
310      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
311       (p)[0] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
312       (NLBLOCK->nllen == 1 || (p)[1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1])       \
313      ) \      ) \
314    )    )
315    
# Line 152  string field names are. */ Line 318  string field names are. */
318  #define WAS_NEWLINE(p) \  #define WAS_NEWLINE(p) \
319    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
320      ((p) > NLBLOCK->PSSTART && \      ((p) > NLBLOCK->PSSTART && \
321       _pcre_was_newline((p), NLBLOCK->PSSTART, &(NLBLOCK->nllen), utf8) \       PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
322      ) \         &(NLBLOCK->nllen), utf)) \
323      : \      : \
324      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
325       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \       RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] &&              \
326       (NLBLOCK->nllen == 1 || (p)[-NLBLOCK->nllen+1] == NLBLOCK->nl[1]) \       (NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
327      ) \      ) \
328    )    )
329    
# Line 171  used for the external interface and appe Line 337  used for the external interface and appe
337  must begin with PCRE_. */  must begin with PCRE_. */
338    
339  #ifdef CUSTOM_SUBJECT_PTR  #ifdef CUSTOM_SUBJECT_PTR
340  #define PCRE_SPTR CUSTOM_SUBJECT_PTR  #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
 #define USPTR CUSTOM_SUBJECT_PTR  
341  #else  #else
342  #define PCRE_SPTR const char *  #define PCRE_PUCHAR const pcre_uchar *
 #define USPTR const unsigned char *  
343  #endif  #endif
344    
345  /* 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 184  values. */ Line 348  values. */
348  #include "pcre.h"  #include "pcre.h"
349  #include "ucp.h"  #include "ucp.h"
350    
351    #ifdef COMPILE_PCRE32
352    /* Assert that the public PCRE_UCHAR32 is a 32-bit type */
353    typedef int __assert_pcre_uchar32_size[sizeof(PCRE_UCHAR32) == 4 ? 1 : -1];
354    #endif
355    
356  /* When compiling for use with the Virtual Pascal compiler, these functions  /* When compiling for use with the Virtual Pascal compiler, these functions
357  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
358  option on the command line. */  option on the command line. */
359    
360  #ifdef VPCOMPAT  #ifdef VPCOMPAT
361    #define strlen(s)        _strlen(s)
362  #define strncmp(s1,s2,m) _strncmp(s1,s2,m)  #define strncmp(s1,s2,m) _strncmp(s1,s2,m)
363    #define memcmp(s,c,n)    _memcmp(s,c,n)
364  #define memcpy(d,s,n)    _memcpy(d,s,n)  #define memcpy(d,s,n)    _memcpy(d,s,n)
365  #define memmove(d,s,n)   _memmove(d,s,n)  #define memmove(d,s,n)   _memmove(d,s,n)
366  #define memset(s,c,n)    _memset(s,c,n)  #define memset(s,c,n)    _memset(s,c,n)
# Line 198  option on the command line. */ Line 369  option on the command line. */
369  /* 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(),
370  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
371  is set. Otherwise, include an emulating function for those systems that have  is set. Otherwise, include an emulating function for those systems that have
372  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. */  
373    
374  #if ! HAVE_MEMMOVE  #ifndef HAVE_MEMMOVE
375  #undef  memmove        /* some systems may have a macro */  #undef  memmove        /* some systems may have a macro */
376  #if HAVE_BCOPY  #ifdef HAVE_BCOPY
377  #define memmove(a, b, c) bcopy(b, a, c)  #define memmove(a, b, c) bcopy(b, a, c)
378  #else  /* HAVE_BCOPY */  #else  /* HAVE_BCOPY */
379  static void *  static void *
380  pcre_memmove(unsigned char *dest, const unsigned char *src, size_t n)  pcre_memmove(void *d, const void *s, size_t n)
381  {  {
382  size_t i;  size_t i;
383  dest += n;  unsigned char *dest = (unsigned char *)d;
384  src += n;  const unsigned char *src = (const unsigned char *)s;
385  for (i = 0; i < n; ++i) *(--dest) =  *(--src);  if (dest > src)
386  return dest;    {
387      dest += n;
388      src += n;
389      for (i = 0; i < n; ++i) *(--dest) = *(--src);
390      return (void *)dest;
391      }
392    else
393      {
394      for (i = 0; i < n; ++i) *dest++ = *src++;
395      return (void *)(dest - n);
396      }
397  }  }
398  #define memmove(a, b, c) pcre_memmove(a, b, c)  #define memmove(a, b, c) pcre_memmove(a, b, c)
399  #endif   /* not HAVE_BCOPY */  #endif   /* not HAVE_BCOPY */
# Line 235  The macros are controlled by the value o Line 414  The macros are controlled by the value o
414  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
415  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
416    
417    #if defined COMPILE_PCRE8
418    
419  #if LINK_SIZE == 2  #if LINK_SIZE == 2
420    
421  #define PUT(a,n,d)   \  #define PUT(a,n,d)   \
# Line 271  is automated on Unix systems via the "co Line 452  is automated on Unix systems via the "co
452  #define GET(a,n) \  #define GET(a,n) \
453    (((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])
454    
455  #define MAX_PATTERN_SIZE (1 << 30)   /* Keep it positive */  /* Keep it positive */
456    #define MAX_PATTERN_SIZE (1 << 30)
457    
458    #else
459    #error LINK_SIZE must be either 2, 3, or 4
460    #endif
461    
462    #elif defined COMPILE_PCRE16
463    
464    #if LINK_SIZE == 2
465    
466    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
467    #undef LINK_SIZE
468    #define LINK_SIZE 1
469    
470    #define PUT(a,n,d)   \
471      (a[n] = (d))
472    
473    #define GET(a,n) \
474      (a[n])
475    
476    #define MAX_PATTERN_SIZE (1 << 16)
477    
478    #elif LINK_SIZE == 3 || LINK_SIZE == 4
479    
480    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
481    #undef LINK_SIZE
482    #define LINK_SIZE 2
483    
484    #define PUT(a,n,d)   \
485      (a[n] = (d) >> 16), \
486      (a[(n)+1] = (d) & 65535)
487    
488    #define GET(a,n) \
489      (((a)[n] << 16) | (a)[(n)+1])
490    
491    /* Keep it positive */
492    #define MAX_PATTERN_SIZE (1 << 30)
493    
494  #else  #else
495  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
496  #endif  #endif
497    
498    #elif defined COMPILE_PCRE32
499    
500    /* Only supported LINK_SIZE is 4 */
501    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
502    #undef LINK_SIZE
503    #define LINK_SIZE 1
504    
505    #define PUT(a,n,d)   \
506      (a[n] = (d))
507    
508    #define GET(a,n) \
509      (a[n])
510    
511    /* Keep it positive */
512    #define MAX_PATTERN_SIZE (1 << 30)
513    
514    #else
515    #error Unsupported compiling mode
516    #endif /* COMPILE_PCRE[8|16|32] */
517    
518  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
519    
# Line 288  is automated on Unix systems via the "co Line 524  is automated on Unix systems via the "co
524  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
525  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
526    
527    #if defined COMPILE_PCRE8
528    
529    #define IMM2_SIZE 2
530    
531  #define PUT2(a,n,d)   \  #define PUT2(a,n,d)   \
532    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
533    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
534    
535    /* For reasons that I do not understand, the expression in this GET2 macro is
536    treated by gcc as a signed expression, even when a is declared as unsigned. It
537    seems that any kind of arithmetic results in a signed value. */
538    
539  #define GET2(a,n) \  #define GET2(a,n) \
540    (((a)[n] << 8) | (a)[(n)+1])    (unsigned int)(((a)[n] << 8) | (a)[(n)+1])
541    
542  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += 2  #elif defined COMPILE_PCRE16
543    
544    #define IMM2_SIZE 1
545    
546  /* When UTF-8 encoding is being used, a character is no longer just a single  #define PUT2(a,n,d)   \
547  byte. The macros for character handling generate simple sequences when used in     a[n] = d
548  byte-mode, and more complicated ones for UTF-8 characters. */  
549    #define GET2(a,n) \
550       a[n]
551    
552  #ifndef SUPPORT_UTF8  #elif defined COMPILE_PCRE32
553    
554    #define IMM2_SIZE 1
555    
556    #define PUT2(a,n,d)   \
557       a[n] = d
558    
559    #define GET2(a,n) \
560       a[n]
561    
562    #else
563    #error Unsupported compiling mode
564    #endif /* COMPILE_PCRE[8|16|32] */
565    
566    #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
567    
568    /* The maximum length of a MARK name is currently one data unit; it may be
569    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
570    
571    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
572    #define MAX_MARK ((1u << 16) - 1)
573    #else
574    #define MAX_MARK ((1u << 8) - 1)
575    #endif
576    
577    /* When UTF encoding is being used, a character is no longer just a single
578    byte. The macros for character handling generate simple sequences when used in
579    character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
580    and other macros are not used when UTF is not supported, so they are not
581    defined. To make sure they can never even appear when UTF support is omitted,
582    we don't even define them. */
583    
584    #ifndef SUPPORT_UTF
585    
586    /* #define MAX_VALUE_FOR_SINGLE_CHAR */
587    /* #define HAS_EXTRALEN(c) */
588    /* #define GET_EXTRALEN(c) */
589    /* #define NOT_FIRSTCHAR(c) */
590  #define GETCHAR(c, eptr) c = *eptr;  #define GETCHAR(c, eptr) c = *eptr;
591  #define GETCHARTEST(c, eptr) c = *eptr;  #define GETCHARTEST(c, eptr) c = *eptr;
592  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
593  #define GETCHARINCTEST(c, eptr) c = *eptr++;  #define GETCHARINCTEST(c, eptr) c = *eptr++;
594  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
595  #define BACKCHAR(eptr)  #define RAWUCHAR(eptr) (*(eptr))
596    #define RAWUCHARINC(eptr) (*(eptr)++)
597    #define RAWUCHARTEST(eptr) (*(eptr))
598    #define RAWUCHARINCTEST(eptr) (*(eptr)++)
599    /* #define GETCHARLENTEST(c, eptr, len) */
600    /* #define BACKCHAR(eptr) */
601    /* #define FORWARDCHAR(eptr) */
602    /* #define ACROSSCHAR(condition, eptr, action) */
603    
604  #else   /* SUPPORT_UTF8 */  #else   /* SUPPORT_UTF */
605    
606  /* 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. */  
607    
608  #define GETCHAR(c, eptr) \  #define HASUTF8EXTRALEN(c) ((c) >= 0xc0)
609    c = *eptr; \  
610    if (c >= 0xc0) \  /* Base macro to pick up the remaining bytes of a UTF-8 character, not
611    advancing the pointer. */
612    
613    #define GETUTF8(c, eptr) \
614        { \
615        if ((c & 0x20) == 0) \
616          c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
617        else if ((c & 0x10) == 0) \
618          c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
619        else if ((c & 0x08) == 0) \
620          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
621          ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
622        else if ((c & 0x04) == 0) \
623          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
624              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
625              (eptr[4] & 0x3f); \
626        else \
627          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
628              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
629              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
630        }
631    
632    /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
633    the pointer. */
634    
635    #define GETUTF8INC(c, eptr) \
636      { \      { \
637      int gcii; \      if ((c & 0x20) == 0) \
638      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \        c = ((c & 0x1f) << 6) | (*eptr++ & 0x3f); \
639      int gcss = 6*gcaa; \      else if ((c & 0x10) == 0) \
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     for (gcii = 1; gcii <= gcaa; gcii++) \  
640        { \        { \
641        gcss -= 6; \        c = ((c & 0x0f) << 12) | ((*eptr & 0x3f) << 6) | (eptr[1] & 0x3f); \
642        c |= (eptr[gcii] & 0x3f) << gcss; \        eptr += 2; \
643          } \
644        else if ((c & 0x08) == 0) \
645          { \
646          c = ((c & 0x07) << 18) | ((*eptr & 0x3f) << 12) | \
647              ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
648          eptr += 3; \
649          } \
650        else if ((c & 0x04) == 0) \
651          { \
652          c = ((c & 0x03) << 24) | ((*eptr & 0x3f) << 18) | \
653              ((eptr[1] & 0x3f) << 12) | ((eptr[2] & 0x3f) << 6) | \
654              (eptr[3] & 0x3f); \
655          eptr += 4; \
656          } \
657        else \
658          { \
659          c = ((c & 0x01) << 30) | ((*eptr & 0x3f) << 24) | \
660              ((eptr[1] & 0x3f) << 18) | ((eptr[2] & 0x3f) << 12) | \
661              ((eptr[3] & 0x3f) << 6) | (eptr[4] & 0x3f); \
662          eptr += 5; \
663        } \        } \
664      }      }
665    
666    #if defined COMPILE_PCRE8
667    
668    /* These macros were originally written in the form of loops that used data
669    from the tables whose names start with PRIV(utf8_table). They were rewritten by
670    a user so as not to use loops, because in some environments this gives a
671    significant performance advantage, and it seems never to do any harm. */
672    
673    /* Tells the biggest code point which can be encoded as a single character. */
674    
675    #define MAX_VALUE_FOR_SINGLE_CHAR 127
676    
677    /* Tests whether the code point needs extra characters to decode. */
678    
679    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
680    
681    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
682    Otherwise it has an undefined behaviour. */
683    
684    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
685    
686    /* Returns TRUE, if the given character is not the first character
687    of a UTF sequence. */
688    
689    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
690    
691    /* Get the next UTF-8 character, not advancing the pointer. This is called when
692    we know we are in UTF-8 mode. */
693    
694    #define GETCHAR(c, eptr) \
695      c = *eptr; \
696      if (c >= 0xc0) GETUTF8(c, eptr);
697    
698  /* 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
699  pointer. */  pointer. */
700    
701  #define GETCHARTEST(c, eptr) \  #define GETCHARTEST(c, eptr) \
702    c = *eptr; \    c = *eptr; \
703    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8(c, eptr);
     { \  
     int gcii; \  
     int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     for (gcii = 1; gcii <= gcaa; gcii++) \  
       { \  
       gcss -= 6; \  
       c |= (eptr[gcii] & 0x3f) << gcss; \  
       } \  
     }  
704    
705  /* 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
706  know we are in UTF-8 mode. */  know we are in UTF-8 mode. */
707    
708  #define GETCHARINC(c, eptr) \  #define GETCHARINC(c, eptr) \
709    c = *eptr++; \    c = *eptr++; \
710    if (c >= 0xc0) \    if (c >= 0xc0) GETUTF8INC(c, eptr);
     { \  
     int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  
     int gcss = 6*gcaa; \  
     c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
     while (gcaa-- > 0) \  
       { \  
       gcss -= 6; \  
       c |= (*eptr++ & 0x3f) << gcss; \  
       } \  
     }  
711    
712  /* 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.
713    This is called when we don't know if we are in UTF-8 mode. */
714    
715  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
716    c = *eptr++; \    c = *eptr++; \
717    if (utf8 && c >= 0xc0) \    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
718    
719    /* Base macro to pick up the remaining bytes of a UTF-8 character, not
720    advancing the pointer, incrementing the length. */
721    
722    #define GETUTF8LEN(c, eptr, len) \
723      { \      { \
724      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \      if ((c & 0x20) == 0) \
725      int gcss = 6*gcaa; \        { \
726      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \        c = ((c & 0x1f) << 6) | (eptr[1] & 0x3f); \
727      while (gcaa-- > 0) \        len++; \
728          } \
729        else if ((c & 0x10)  == 0) \
730        { \        { \
731        gcss -= 6; \        c = ((c & 0x0f) << 12) | ((eptr[1] & 0x3f) << 6) | (eptr[2] & 0x3f); \
732        c |= (*eptr++ & 0x3f) << gcss; \        len += 2; \
733          } \
734        else if ((c & 0x08)  == 0) \
735          {\
736          c = ((c & 0x07) << 18) | ((eptr[1] & 0x3f) << 12) | \
737              ((eptr[2] & 0x3f) << 6) | (eptr[3] & 0x3f); \
738          len += 3; \
739          } \
740        else if ((c & 0x04)  == 0) \
741          { \
742          c = ((c & 0x03) << 24) | ((eptr[1] & 0x3f) << 18) | \
743              ((eptr[2] & 0x3f) << 12) | ((eptr[3] & 0x3f) << 6) | \
744              (eptr[4] & 0x3f); \
745          len += 4; \
746          } \
747        else \
748          {\
749          c = ((c & 0x01) << 30) | ((eptr[1] & 0x3f) << 24) | \
750              ((eptr[2] & 0x3f) << 18) | ((eptr[3] & 0x3f) << 12) | \
751              ((eptr[4] & 0x3f) << 6) | (eptr[5] & 0x3f); \
752          len += 5; \
753        } \        } \
754      }      }
755    
# Line 386  if there are extra bytes. This is called Line 758  if there are extra bytes. This is called
758    
759  #define GETCHARLEN(c, eptr, len) \  #define GETCHARLEN(c, eptr, len) \
760    c = *eptr; \    c = *eptr; \
761    if (c >= 0xc0) \    if (c >= 0xc0) GETUTF8LEN(c, eptr, len);
762      { \  
763      int gcii; \  /* Get the next UTF-8 character, testing for UTF-8 mode, not advancing the
764      int gcaa = _pcre_utf8_table4[c & 0x3f];  /* Number of additional bytes */ \  pointer, incrementing length if there are extra bytes. This is called when we
765      int gcss = 6*gcaa; \  do not know if we are in UTF-8 mode. */
766      c = (c & _pcre_utf8_table3[gcaa]) << gcss; \  
767      for (gcii = 1; gcii <= gcaa; gcii++) \  #define GETCHARLENTEST(c, eptr, len) \
768        { \    c = *eptr; \
769        gcss -= 6; \    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
770        c |= (eptr[gcii] & 0x3f) << gcss; \  
771        } \  /* Returns the next uchar, not advancing the pointer. This is called when
772      len += gcaa; \  we know we are in UTF mode. */
773      }  
774    #define RAWUCHAR(eptr) \
775      (*(eptr))
776    
777    /* Returns the next uchar, advancing the pointer. This is called when
778    we know we are in UTF mode. */
779    
780    #define RAWUCHARINC(eptr) \
781      (*((eptr)++))
782    
783    /* Returns the next uchar, testing for UTF mode, and not advancing the
784    pointer. */
785    
786    #define RAWUCHARTEST(eptr) \
787      (*(eptr))
788    
789    /* Returns the next uchar, testing for UTF mode, advancing the
790    pointer. */
791    
792    #define RAWUCHARINCTEST(eptr) \
793      (*((eptr)++))
794    
795  /* 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
796  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
797    because almost all calls are already within a block of UTF-8 only code. */
798    
799  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--;  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
800    
801  #endif  /* Same as above, just in the other direction. */
802    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
803    
804    /* Same as above, but it allows a fully customizable form. */
805    #define ACROSSCHAR(condition, eptr, action) \
806      while((condition) && ((eptr) & 0xc0) == 0x80) action
807    
808  /* In case there is no definition of offsetof() provided - though any proper  #elif defined COMPILE_PCRE16
 Standard C system should have one. */  
809    
810  #ifndef offsetof  /* Tells the biggest code point which can be encoded as a single character. */
811  #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
812  #endif  #define MAX_VALUE_FOR_SINGLE_CHAR 65535
813    
814    /* Tests whether the code point needs extra characters to decode. */
815    
816    #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
817    
818    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
819    Otherwise it has an undefined behaviour. */
820    
821    #define GET_EXTRALEN(c) 1
822    
823    /* Returns TRUE, if the given character is not the first character
824    of a UTF sequence. */
825    
826  /* These are the public options that can change during matching. */  #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
827    
828  #define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)  /* Base macro to pick up the low surrogate of a UTF-16 character, not
829    advancing the pointer. */
830    
831    #define GETUTF16(c, eptr) \
832       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
833    
834    /* Get the next UTF-16 character, not advancing the pointer. This is called when
835    we know we are in UTF-16 mode. */
836    
837    #define GETCHAR(c, eptr) \
838      c = *eptr; \
839      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
840    
841    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
842    pointer. */
843    
844    #define GETCHARTEST(c, eptr) \
845      c = *eptr; \
846      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
847    
848    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
849    the pointer. */
850    
851    #define GETUTF16INC(c, eptr) \
852       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
853    
854    /* Get the next UTF-16 character, advancing the pointer. This is called when we
855    know we are in UTF-16 mode. */
856    
857    #define GETCHARINC(c, eptr) \
858      c = *eptr++; \
859      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
860    
861    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
862    This is called when we don't know if we are in UTF-16 mode. */
863    
864    #define GETCHARINCTEST(c, eptr) \
865      c = *eptr++; \
866      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
867    
868    /* Base macro to pick up the low surrogate of a UTF-16 character, not
869    advancing the pointer, incrementing the length. */
870    
871    #define GETUTF16LEN(c, eptr, len) \
872       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
873    
874    /* Get the next UTF-16 character, not advancing the pointer, incrementing
875    length if there is a low surrogate. This is called when we know we are in
876    UTF-16 mode. */
877    
878    #define GETCHARLEN(c, eptr, len) \
879      c = *eptr; \
880      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
881    
882    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
883    pointer, incrementing length if there is a low surrogate. This is called when
884    we do not know if we are in UTF-16 mode. */
885    
886    #define GETCHARLENTEST(c, eptr, len) \
887      c = *eptr; \
888      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
889    
890  /* Private options flags start at the most significant end of the four bytes.  /* Returns the next uchar, not advancing the pointer. This is called when
891  The public options defined in pcre.h start at the least significant end. Make  we know we are in UTF mode. */
 sure they don't overlap! The bits are getting a bit scarce now -- when we run  
 out, there is a dummy word in the structure that could be used for the private  
 bits. */  
892    
893  #define PCRE_NOPARTIAL     0x80000000  /* can't use partial with this regex */  #define RAWUCHAR(eptr) \
894  #define PCRE_FIRSTSET      0x40000000  /* first_byte is set */    (*(eptr))
 #define PCRE_REQCHSET      0x20000000  /* req_byte is set */  
 #define PCRE_STARTLINE     0x10000000  /* start after \n for multiline */  
 #define PCRE_JCHANGED      0x08000000  /* j option changes within regex */  
895    
896  /* Options for the "extra" block produced by pcre_study(). */  /* Returns the next uchar, advancing the pointer. This is called when
897    we know we are in UTF mode. */
898    
899  #define PCRE_STUDY_MAPPED   0x01     /* a map of starting chars exists */  #define RAWUCHARINC(eptr) \
900      (*((eptr)++))
901    
902    /* Returns the next uchar, testing for UTF mode, and not advancing the
903    pointer. */
904    
905    #define RAWUCHARTEST(eptr) \
906      (*(eptr))
907    
908    /* Returns the next uchar, testing for UTF mode, advancing the
909    pointer. */
910    
911    #define RAWUCHARINCTEST(eptr) \
912      (*((eptr)++))
913    
914    /* If the pointer is not at the start of a character, move it back until
915    it is. This is called only in UTF-16 mode - we don't put a test within the
916    macro because almost all calls are already within a block of UTF-16 only
917    code. */
918    
919    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
920    
921    /* Same as above, just in the other direction. */
922    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
923    
924    /* Same as above, but it allows a fully customizable form. */
925    #define ACROSSCHAR(condition, eptr, action) \
926      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
927    
928    #elif defined COMPILE_PCRE32
929    
930    /* These are trivial for the 32-bit library, since all UTF-32 characters fit
931    into one pcre_uchar unit. */
932    #define MAX_VALUE_FOR_SINGLE_CHAR (0x10ffffu)
933    #define HAS_EXTRALEN(c) (0)
934    #define GET_EXTRALEN(c) (0)
935    #define NOT_FIRSTCHAR(c) (0)
936    
937    /* Get the next UTF-32 character, not advancing the pointer. This is called when
938    we know we are in UTF-32 mode. */
939    
940    #define GETCHAR(c, eptr) \
941      c = *(eptr);
942    
943    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
944    pointer. */
945    
946    #define GETCHARTEST(c, eptr) \
947      c = *(eptr);
948    
949    /* Get the next UTF-32 character, advancing the pointer. This is called when we
950    know we are in UTF-32 mode. */
951    
952    #define GETCHARINC(c, eptr) \
953      c = *((eptr)++);
954    
955    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
956    This is called when we don't know if we are in UTF-32 mode. */
957    
958    #define GETCHARINCTEST(c, eptr) \
959      c = *((eptr)++);
960    
961    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
962    length (since all UTF-32 is of length 1). This is called when we know we are in
963    UTF-32 mode. */
964    
965    #define GETCHARLEN(c, eptr, len) \
966      GETCHAR(c, eptr)
967    
968    /* Get the next UTF-32character, testing for UTF-32 mode, not advancing the
969    pointer, not incrementing the length (since all UTF-32 is of length 1).
970    This is called when we do not know if we are in UTF-32 mode. */
971    
972    #define GETCHARLENTEST(c, eptr, len) \
973      GETCHARTEST(c, eptr)
974    
975    /* Returns the next uchar, not advancing the pointer. This is called when
976    we know we are in UTF mode. */
977    
978    #define RAWUCHAR(eptr) \
979      (*(eptr))
980    
981    /* Returns the next uchar, advancing the pointer. This is called when
982    we know we are in UTF mode. */
983    
984    #define RAWUCHARINC(eptr) \
985      (*((eptr)++))
986    
987    /* Returns the next uchar, testing for UTF mode, and not advancing the
988    pointer. */
989    
990    #define RAWUCHARTEST(eptr) \
991      (*(eptr))
992    
993    /* Returns the next uchar, testing for UTF mode, advancing the
994    pointer. */
995    
996    #define RAWUCHARINCTEST(eptr) \
997      (*((eptr)++))
998    
999    /* If the pointer is not at the start of a character, move it back until
1000    it is. This is called only in UTF-32 mode - we don't put a test within the
1001    macro because almost all calls are already within a block of UTF-32 only
1002    code.
1003    These are all no-ops since all UTF-32 characters fit into one pcre_uchar. */
1004    
1005    #define BACKCHAR(eptr) do { } while (0)
1006    
1007    /* Same as above, just in the other direction. */
1008    #define FORWARDCHAR(eptr) do { } while (0)
1009    
1010    /* Same as above, but it allows a fully customizable form. */
1011    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
1012    
1013    #else
1014    #error Unsupported compiling mode
1015    #endif /* COMPILE_PCRE[8|16|32] */
1016    
1017    #endif  /* SUPPORT_UTF */
1018    
1019    /* Tests for Unicode horizontal and vertical whitespace characters must check a
1020    number of different values. Using a switch statement for this generates the
1021    fastest code (no loop, no memory access), and there are several places in the
1022    interpreter code where this happens. In order to ensure that all the case lists
1023    remain in step, we use macros so that there is only one place where the lists
1024    are defined.
1025    
1026    These values are also required as lists in pcre_compile.c when processing \h,
1027    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
1028    macros that define the values are here so that all the definitions are
1029    together. The lists must be in ascending character order, terminated by
1030    NOTACHAR (which is 0xffffffff).
1031    
1032    Any changes should ensure that the various macros are kept in step with each
1033    other. NOTE: The values also appear in pcre_jit_compile.c. */
1034    
1035    /* ------ ASCII/Unicode environments ------ */
1036    
1037    #ifndef EBCDIC
1038    
1039    #define HSPACE_LIST \
1040      CHAR_HT, CHAR_SPACE, 0xa0, \
1041      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
1042      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
1043      NOTACHAR
1044    
1045    #define HSPACE_MULTIBYTE_CASES \
1046      case 0x1680:  /* OGHAM SPACE MARK */ \
1047      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
1048      case 0x2000:  /* EN QUAD */ \
1049      case 0x2001:  /* EM QUAD */ \
1050      case 0x2002:  /* EN SPACE */ \
1051      case 0x2003:  /* EM SPACE */ \
1052      case 0x2004:  /* THREE-PER-EM SPACE */ \
1053      case 0x2005:  /* FOUR-PER-EM SPACE */ \
1054      case 0x2006:  /* SIX-PER-EM SPACE */ \
1055      case 0x2007:  /* FIGURE SPACE */ \
1056      case 0x2008:  /* PUNCTUATION SPACE */ \
1057      case 0x2009:  /* THIN SPACE */ \
1058      case 0x200A:  /* HAIR SPACE */ \
1059      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
1060      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
1061      case 0x3000   /* IDEOGRAPHIC SPACE */
1062    
1063    #define HSPACE_BYTE_CASES \
1064      case CHAR_HT: \
1065      case CHAR_SPACE: \
1066      case 0xa0     /* NBSP */
1067    
1068    #define HSPACE_CASES \
1069      HSPACE_BYTE_CASES: \
1070      HSPACE_MULTIBYTE_CASES
1071    
1072    #define VSPACE_LIST \
1073      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
1074    
1075    #define VSPACE_MULTIBYTE_CASES \
1076      case 0x2028:    /* LINE SEPARATOR */ \
1077      case 0x2029     /* PARAGRAPH SEPARATOR */
1078    
1079    #define VSPACE_BYTE_CASES \
1080      case CHAR_LF: \
1081      case CHAR_VT: \
1082      case CHAR_FF: \
1083      case CHAR_CR: \
1084      case CHAR_NEL
1085    
1086    #define VSPACE_CASES \
1087      VSPACE_BYTE_CASES: \
1088      VSPACE_MULTIBYTE_CASES
1089    
1090    /* ------ EBCDIC environments ------ */
1091    
1092    #else
1093    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
1094    
1095    #define HSPACE_BYTE_CASES \
1096      case CHAR_HT: \
1097      case CHAR_SPACE
1098    
1099    #define HSPACE_CASES HSPACE_BYTE_CASES
1100    
1101    #ifdef EBCDIC_NL25
1102    #define VSPACE_LIST \
1103      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1104    #else
1105    #define VSPACE_LIST \
1106      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1107    #endif
1108    
1109    #define VSPACE_BYTE_CASES \
1110      case CHAR_LF: \
1111      case CHAR_VT: \
1112      case CHAR_FF: \
1113      case CHAR_CR: \
1114      case CHAR_NEL
1115    
1116    #define VSPACE_CASES VSPACE_BYTE_CASES
1117    #endif  /* EBCDIC */
1118    
1119    /* ------ End of whitespace macros ------ */
1120    
1121    
1122    
1123    /* Private flags containing information about the compiled regex. They used to
1124    live at the top end of the options word, but that got almost full, so now they
1125    are in a 16-bit flags word. From release 8.00, PCRE_NOPARTIAL is unused, as
1126    the restrictions on partial matching have been lifted. It remains for backwards
1127    compatibility. */
1128    
1129    #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1130    #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1131    #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
1132    #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1133    #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1134    #define PCRE_REQCHSET      0x0040  /* req_byte is set */
1135    #define PCRE_RCH_CASELESS  0x0080  /* caseless requested char */
1136    #define PCRE_STARTLINE     0x0100  /* start after \n for multiline */
1137    #define PCRE_NOPARTIAL     0x0200  /* can't use partial with this regex */
1138    #define PCRE_JCHANGED      0x0400  /* j option used in regex */
1139    #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1140    #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1141    
1142    #if defined COMPILE_PCRE8
1143    #define PCRE_MODE          PCRE_MODE8
1144    #elif defined COMPILE_PCRE16
1145    #define PCRE_MODE          PCRE_MODE16
1146    #elif defined COMPILE_PCRE32
1147    #define PCRE_MODE          PCRE_MODE32
1148    #endif
1149    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1150    
1151    /* Flags for the "extra" block produced by pcre_study(). */
1152    
1153    #define PCRE_STUDY_MAPPED  0x0001  /* a map of starting chars exists */
1154    #define PCRE_STUDY_MINLEN  0x0002  /* a minimum length field exists */
1155    
1156  /* Masks for identifying the public options that are permitted at compile  /* Masks for identifying the public options that are permitted at compile
1157  time, run time, or study time, respectively. */  time, run time, or study time, respectively. */
1158    
1159  #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY)  #define PCRE_NEWLINE_BITS (PCRE_NEWLINE_CR|PCRE_NEWLINE_LF|PCRE_NEWLINE_ANY| \
1160                               PCRE_NEWLINE_ANYCRLF)
1161    
1162  #define PUBLIC_OPTIONS \  #define PUBLIC_COMPILE_OPTIONS \
1163    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \    (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
1164     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \     PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
1165     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| \
1166     PCRE_DUPNAMES|PCRE_NEWLINE_BITS)     PCRE_DUPNAMES|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1167       PCRE_JAVASCRIPT_COMPAT|PCRE_UCP|PCRE_NO_START_OPTIMIZE)
1168    
1169  #define PUBLIC_EXEC_OPTIONS \  #define PUBLIC_EXEC_OPTIONS \
1170    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1171     PCRE_PARTIAL|PCRE_NEWLINE_BITS)     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_NEWLINE_BITS| \
1172       PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE|PCRE_NO_START_OPTIMIZE)
1173    
1174  #define PUBLIC_DFA_EXEC_OPTIONS \  #define PUBLIC_DFA_EXEC_OPTIONS \
1175    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \    (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NOTEMPTY_ATSTART| \
1176     PCRE_PARTIAL|PCRE_DFA_SHORTEST|PCRE_DFA_RESTART|PCRE_NEWLINE_BITS)     PCRE_NO_UTF8_CHECK|PCRE_PARTIAL_HARD|PCRE_PARTIAL_SOFT|PCRE_DFA_SHORTEST| \
1177       PCRE_DFA_RESTART|PCRE_NEWLINE_BITS|PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE| \
1178       PCRE_NO_START_OPTIMIZE)
1179    
1180    #define PUBLIC_STUDY_OPTIONS \
1181       (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1182        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1183    
1184    #define PUBLIC_JIT_EXEC_OPTIONS \
1185       (PCRE_NO_UTF8_CHECK|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|\
1186        PCRE_NOTEMPTY_ATSTART|PCRE_PARTIAL_SOFT|PCRE_PARTIAL_HARD)
1187    
1188  #define PUBLIC_STUDY_OPTIONS 0   /* None defined */  /* Magic number to provide a small check against being handed junk. */
   
 /* Magic number to provide a small check against being handed junk. Also used  
 to detect whether a pattern was compiled on a host of different endianness. */  
1189    
1190  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1191    
1192  /* Negative values for the firstchar and reqchar variables */  /* This variable is used to detect a loaded regular expression
1193    in different endianness. */
1194    
1195  #define REQ_UNSET (-2)  #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
 #define REQ_NONE  (-1)  
1196    
1197  /* 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
1198  req_byte match. */  req_byte match. */
1199    
1200  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1201    
1202  /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1203  variable-length repeat, or a anything other than literal characters. */  environments where these macros are defined elsewhere. Unfortunately, there
1204    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 */  
1205    
1206  typedef int BOOL;  typedef int BOOL;
1207    
1208    #ifndef FALSE
1209  #define FALSE   0  #define FALSE   0
1210  #define TRUE    1  #define TRUE    1
1211    #endif
1212    
1213    /* If PCRE is to support UTF-8 on EBCDIC platforms, we cannot use normal
1214    character constants like '*' because the compiler would emit their EBCDIC code,
1215    which is different from their ASCII/UTF-8 code. Instead we define macros for
1216    the characters so that they always use the ASCII/UTF-8 code when UTF-8 support
1217    is enabled. When UTF-8 support is not enabled, the definitions use character
1218    literals. Both character and string versions of each character are needed, and
1219    there are some longer strings as well.
1220    
1221    This means that, on EBCDIC platforms, the PCRE library can handle either
1222    EBCDIC, or UTF-8, but not both. To support both in the same compiled library
1223    would need different lookups depending on whether PCRE_UTF8 was set or not.
1224    This would make it impossible to use characters in switch/case statements,
1225    which would reduce performance. For a theoretical use (which nobody has asked
1226    for) in a minority area (EBCDIC platforms), this is not sensible. Any
1227    application that did need both could compile two versions of the library, using
1228    macros to give the functions distinct names. */
1229    
1230    #ifndef SUPPORT_UTF
1231    
1232    /* UTF-8 support is not enabled; use the platform-dependent character literals
1233    so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1234    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1235    characters, a common practice has been to use its NL (0x15) character as the
1236    line terminator in C-like processing environments. However, sometimes the LF
1237    (0x25) character is used instead, according to this Unicode document:
1238    
1239    http://unicode.org/standard/reports/tr13/tr13-5.html
1240    
1241    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1242    instead. Whichever is *not* chosen is defined as NEL.
1243    
1244    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1245    same code point. */
1246    
1247    #ifdef EBCDIC
1248    
1249    #ifndef EBCDIC_NL25
1250    #define CHAR_NL                     '\x15'
1251    #define CHAR_NEL                    '\x25'
1252    #define STR_NL                      "\x15"
1253    #define STR_NEL                     "\x25"
1254    #else
1255    #define CHAR_NL                     '\x25'
1256    #define CHAR_NEL                    '\x15'
1257    #define STR_NL                      "\x25"
1258    #define STR_NEL                     "\x15"
1259    #endif
1260    
1261    #define CHAR_LF                     CHAR_NL
1262    #define STR_LF                      STR_NL
1263    
1264    #define CHAR_ESC                    '\047'
1265    #define CHAR_DEL                    '\007'
1266    #define STR_ESC                     "\047"
1267    #define STR_DEL                     "\007"
1268    
1269    #else  /* Not EBCDIC */
1270    
1271    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1272    compatibility. NEL is the Unicode newline character; make sure it is
1273    a positive value. */
1274    
1275    #define CHAR_LF                     '\n'
1276    #define CHAR_NL                     CHAR_LF
1277    #define CHAR_NEL                    ((unsigned char)'\x85')
1278    #define CHAR_ESC                    '\033'
1279    #define CHAR_DEL                    '\177'
1280    
1281    #define STR_LF                      "\n"
1282    #define STR_NL                      STR_LF
1283    #define STR_NEL                     "\x85"
1284    #define STR_ESC                     "\033"
1285    #define STR_DEL                     "\177"
1286    
1287    #endif  /* EBCDIC */
1288    
1289    /* The remaining definitions work in both environments. */
1290    
1291    #define CHAR_NULL                   '\0'
1292    #define CHAR_HT                     '\t'
1293    #define CHAR_VT                     '\v'
1294    #define CHAR_FF                     '\f'
1295    #define CHAR_CR                     '\r'
1296    #define CHAR_BS                     '\b'
1297    #define CHAR_BEL                    '\a'
1298    
1299    #define CHAR_SPACE                  ' '
1300    #define CHAR_EXCLAMATION_MARK       '!'
1301    #define CHAR_QUOTATION_MARK         '"'
1302    #define CHAR_NUMBER_SIGN            '#'
1303    #define CHAR_DOLLAR_SIGN            '$'
1304    #define CHAR_PERCENT_SIGN           '%'
1305    #define CHAR_AMPERSAND              '&'
1306    #define CHAR_APOSTROPHE             '\''
1307    #define CHAR_LEFT_PARENTHESIS       '('
1308    #define CHAR_RIGHT_PARENTHESIS      ')'
1309    #define CHAR_ASTERISK               '*'
1310    #define CHAR_PLUS                   '+'
1311    #define CHAR_COMMA                  ','
1312    #define CHAR_MINUS                  '-'
1313    #define CHAR_DOT                    '.'
1314    #define CHAR_SLASH                  '/'
1315    #define CHAR_0                      '0'
1316    #define CHAR_1                      '1'
1317    #define CHAR_2                      '2'
1318    #define CHAR_3                      '3'
1319    #define CHAR_4                      '4'
1320    #define CHAR_5                      '5'
1321    #define CHAR_6                      '6'
1322    #define CHAR_7                      '7'
1323    #define CHAR_8                      '8'
1324    #define CHAR_9                      '9'
1325    #define CHAR_COLON                  ':'
1326    #define CHAR_SEMICOLON              ';'
1327    #define CHAR_LESS_THAN_SIGN         '<'
1328    #define CHAR_EQUALS_SIGN            '='
1329    #define CHAR_GREATER_THAN_SIGN      '>'
1330    #define CHAR_QUESTION_MARK          '?'
1331    #define CHAR_COMMERCIAL_AT          '@'
1332    #define CHAR_A                      'A'
1333    #define CHAR_B                      'B'
1334    #define CHAR_C                      'C'
1335    #define CHAR_D                      'D'
1336    #define CHAR_E                      'E'
1337    #define CHAR_F                      'F'
1338    #define CHAR_G                      'G'
1339    #define CHAR_H                      'H'
1340    #define CHAR_I                      'I'
1341    #define CHAR_J                      'J'
1342    #define CHAR_K                      'K'
1343    #define CHAR_L                      'L'
1344    #define CHAR_M                      'M'
1345    #define CHAR_N                      'N'
1346    #define CHAR_O                      'O'
1347    #define CHAR_P                      'P'
1348    #define CHAR_Q                      'Q'
1349    #define CHAR_R                      'R'
1350    #define CHAR_S                      'S'
1351    #define CHAR_T                      'T'
1352    #define CHAR_U                      'U'
1353    #define CHAR_V                      'V'
1354    #define CHAR_W                      'W'
1355    #define CHAR_X                      'X'
1356    #define CHAR_Y                      'Y'
1357    #define CHAR_Z                      'Z'
1358    #define CHAR_LEFT_SQUARE_BRACKET    '['
1359    #define CHAR_BACKSLASH              '\\'
1360    #define CHAR_RIGHT_SQUARE_BRACKET   ']'
1361    #define CHAR_CIRCUMFLEX_ACCENT      '^'
1362    #define CHAR_UNDERSCORE             '_'
1363    #define CHAR_GRAVE_ACCENT           '`'
1364    #define CHAR_a                      'a'
1365    #define CHAR_b                      'b'
1366    #define CHAR_c                      'c'
1367    #define CHAR_d                      'd'
1368    #define CHAR_e                      'e'
1369    #define CHAR_f                      'f'
1370    #define CHAR_g                      'g'
1371    #define CHAR_h                      'h'
1372    #define CHAR_i                      'i'
1373    #define CHAR_j                      'j'
1374    #define CHAR_k                      'k'
1375    #define CHAR_l                      'l'
1376    #define CHAR_m                      'm'
1377    #define CHAR_n                      'n'
1378    #define CHAR_o                      'o'
1379    #define CHAR_p                      'p'
1380    #define CHAR_q                      'q'
1381    #define CHAR_r                      'r'
1382    #define CHAR_s                      's'
1383    #define CHAR_t                      't'
1384    #define CHAR_u                      'u'
1385    #define CHAR_v                      'v'
1386    #define CHAR_w                      'w'
1387    #define CHAR_x                      'x'
1388    #define CHAR_y                      'y'
1389    #define CHAR_z                      'z'
1390    #define CHAR_LEFT_CURLY_BRACKET     '{'
1391    #define CHAR_VERTICAL_LINE          '|'
1392    #define CHAR_RIGHT_CURLY_BRACKET    '}'
1393    #define CHAR_TILDE                  '~'
1394    
1395    #define STR_HT                      "\t"
1396    #define STR_VT                      "\v"
1397    #define STR_FF                      "\f"
1398    #define STR_CR                      "\r"
1399    #define STR_BS                      "\b"
1400    #define STR_BEL                     "\a"
1401    
1402    #define STR_SPACE                   " "
1403    #define STR_EXCLAMATION_MARK        "!"
1404    #define STR_QUOTATION_MARK          "\""
1405    #define STR_NUMBER_SIGN             "#"
1406    #define STR_DOLLAR_SIGN             "$"
1407    #define STR_PERCENT_SIGN            "%"
1408    #define STR_AMPERSAND               "&"
1409    #define STR_APOSTROPHE              "'"
1410    #define STR_LEFT_PARENTHESIS        "("
1411    #define STR_RIGHT_PARENTHESIS       ")"
1412    #define STR_ASTERISK                "*"
1413    #define STR_PLUS                    "+"
1414    #define STR_COMMA                   ","
1415    #define STR_MINUS                   "-"
1416    #define STR_DOT                     "."
1417    #define STR_SLASH                   "/"
1418    #define STR_0                       "0"
1419    #define STR_1                       "1"
1420    #define STR_2                       "2"
1421    #define STR_3                       "3"
1422    #define STR_4                       "4"
1423    #define STR_5                       "5"
1424    #define STR_6                       "6"
1425    #define STR_7                       "7"
1426    #define STR_8                       "8"
1427    #define STR_9                       "9"
1428    #define STR_COLON                   ":"
1429    #define STR_SEMICOLON               ";"
1430    #define STR_LESS_THAN_SIGN          "<"
1431    #define STR_EQUALS_SIGN             "="
1432    #define STR_GREATER_THAN_SIGN       ">"
1433    #define STR_QUESTION_MARK           "?"
1434    #define STR_COMMERCIAL_AT           "@"
1435    #define STR_A                       "A"
1436    #define STR_B                       "B"
1437    #define STR_C                       "C"
1438    #define STR_D                       "D"
1439    #define STR_E                       "E"
1440    #define STR_F                       "F"
1441    #define STR_G                       "G"
1442    #define STR_H                       "H"
1443    #define STR_I                       "I"
1444    #define STR_J                       "J"
1445    #define STR_K                       "K"
1446    #define STR_L                       "L"
1447    #define STR_M                       "M"
1448    #define STR_N                       "N"
1449    #define STR_O                       "O"
1450    #define STR_P                       "P"
1451    #define STR_Q                       "Q"
1452    #define STR_R                       "R"
1453    #define STR_S                       "S"
1454    #define STR_T                       "T"
1455    #define STR_U                       "U"
1456    #define STR_V                       "V"
1457    #define STR_W                       "W"
1458    #define STR_X                       "X"
1459    #define STR_Y                       "Y"
1460    #define STR_Z                       "Z"
1461    #define STR_LEFT_SQUARE_BRACKET     "["
1462    #define STR_BACKSLASH               "\\"
1463    #define STR_RIGHT_SQUARE_BRACKET    "]"
1464    #define STR_CIRCUMFLEX_ACCENT       "^"
1465    #define STR_UNDERSCORE              "_"
1466    #define STR_GRAVE_ACCENT            "`"
1467    #define STR_a                       "a"
1468    #define STR_b                       "b"
1469    #define STR_c                       "c"
1470    #define STR_d                       "d"
1471    #define STR_e                       "e"
1472    #define STR_f                       "f"
1473    #define STR_g                       "g"
1474    #define STR_h                       "h"
1475    #define STR_i                       "i"
1476    #define STR_j                       "j"
1477    #define STR_k                       "k"
1478    #define STR_l                       "l"
1479    #define STR_m                       "m"
1480    #define STR_n                       "n"
1481    #define STR_o                       "o"
1482    #define STR_p                       "p"
1483    #define STR_q                       "q"
1484    #define STR_r                       "r"
1485    #define STR_s                       "s"
1486    #define STR_t                       "t"
1487    #define STR_u                       "u"
1488    #define STR_v                       "v"
1489    #define STR_w                       "w"
1490    #define STR_x                       "x"
1491    #define STR_y                       "y"
1492    #define STR_z                       "z"
1493    #define STR_LEFT_CURLY_BRACKET      "{"
1494    #define STR_VERTICAL_LINE           "|"
1495    #define STR_RIGHT_CURLY_BRACKET     "}"
1496    #define STR_TILDE                   "~"
1497    
1498    #define STRING_ACCEPT0              "ACCEPT\0"
1499    #define STRING_COMMIT0              "COMMIT\0"
1500    #define STRING_F0                   "F\0"
1501    #define STRING_FAIL0                "FAIL\0"
1502    #define STRING_MARK0                "MARK\0"
1503    #define STRING_PRUNE0               "PRUNE\0"
1504    #define STRING_SKIP0                "SKIP\0"
1505    #define STRING_THEN                 "THEN"
1506    
1507    #define STRING_alpha0               "alpha\0"
1508    #define STRING_lower0               "lower\0"
1509    #define STRING_upper0               "upper\0"
1510    #define STRING_alnum0               "alnum\0"
1511    #define STRING_ascii0               "ascii\0"
1512    #define STRING_blank0               "blank\0"
1513    #define STRING_cntrl0               "cntrl\0"
1514    #define STRING_digit0               "digit\0"
1515    #define STRING_graph0               "graph\0"
1516    #define STRING_print0               "print\0"
1517    #define STRING_punct0               "punct\0"
1518    #define STRING_space0               "space\0"
1519    #define STRING_word0                "word\0"
1520    #define STRING_xdigit               "xdigit"
1521    
1522    #define STRING_DEFINE               "DEFINE"
1523    
1524    #define STRING_CR_RIGHTPAR             "CR)"
1525    #define STRING_LF_RIGHTPAR             "LF)"
1526    #define STRING_CRLF_RIGHTPAR           "CRLF)"
1527    #define STRING_ANY_RIGHTPAR            "ANY)"
1528    #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1529    #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1530    #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1531    #define STRING_UTF8_RIGHTPAR           "UTF8)"
1532    #define STRING_UTF16_RIGHTPAR          "UTF16)"
1533    #define STRING_UTF32_RIGHTPAR          "UTF32)"
1534    #define STRING_UTF_RIGHTPAR            "UTF)"
1535    #define STRING_UCP_RIGHTPAR            "UCP)"
1536    #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1537    
1538    #else  /* SUPPORT_UTF */
1539    
1540    /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1541    works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
1542    only. */
1543    
1544    #define CHAR_HT                     '\011'
1545    #define CHAR_VT                     '\013'
1546    #define CHAR_FF                     '\014'
1547    #define CHAR_CR                     '\015'
1548    #define CHAR_LF                     '\012'
1549    #define CHAR_NL                     CHAR_LF
1550    #define CHAR_NEL                    ((unsigned char)'\x85')
1551    #define CHAR_BS                     '\010'
1552    #define CHAR_BEL                    '\007'
1553    #define CHAR_ESC                    '\033'
1554    #define CHAR_DEL                    '\177'
1555    
1556    #define CHAR_NULL                   '\0'
1557    #define CHAR_SPACE                  '\040'
1558    #define CHAR_EXCLAMATION_MARK       '\041'
1559    #define CHAR_QUOTATION_MARK         '\042'
1560    #define CHAR_NUMBER_SIGN            '\043'
1561    #define CHAR_DOLLAR_SIGN            '\044'
1562    #define CHAR_PERCENT_SIGN           '\045'
1563    #define CHAR_AMPERSAND              '\046'
1564    #define CHAR_APOSTROPHE             '\047'
1565    #define CHAR_LEFT_PARENTHESIS       '\050'
1566    #define CHAR_RIGHT_PARENTHESIS      '\051'
1567    #define CHAR_ASTERISK               '\052'
1568    #define CHAR_PLUS                   '\053'
1569    #define CHAR_COMMA                  '\054'
1570    #define CHAR_MINUS                  '\055'
1571    #define CHAR_DOT                    '\056'
1572    #define CHAR_SLASH                  '\057'
1573    #define CHAR_0                      '\060'
1574    #define CHAR_1                      '\061'
1575    #define CHAR_2                      '\062'
1576    #define CHAR_3                      '\063'
1577    #define CHAR_4                      '\064'
1578    #define CHAR_5                      '\065'
1579    #define CHAR_6                      '\066'
1580    #define CHAR_7                      '\067'
1581    #define CHAR_8                      '\070'
1582    #define CHAR_9                      '\071'
1583    #define CHAR_COLON                  '\072'
1584    #define CHAR_SEMICOLON              '\073'
1585    #define CHAR_LESS_THAN_SIGN         '\074'
1586    #define CHAR_EQUALS_SIGN            '\075'
1587    #define CHAR_GREATER_THAN_SIGN      '\076'
1588    #define CHAR_QUESTION_MARK          '\077'
1589    #define CHAR_COMMERCIAL_AT          '\100'
1590    #define CHAR_A                      '\101'
1591    #define CHAR_B                      '\102'
1592    #define CHAR_C                      '\103'
1593    #define CHAR_D                      '\104'
1594    #define CHAR_E                      '\105'
1595    #define CHAR_F                      '\106'
1596    #define CHAR_G                      '\107'
1597    #define CHAR_H                      '\110'
1598    #define CHAR_I                      '\111'
1599    #define CHAR_J                      '\112'
1600    #define CHAR_K                      '\113'
1601    #define CHAR_L                      '\114'
1602    #define CHAR_M                      '\115'
1603    #define CHAR_N                      '\116'
1604    #define CHAR_O                      '\117'
1605    #define CHAR_P                      '\120'
1606    #define CHAR_Q                      '\121'
1607    #define CHAR_R                      '\122'
1608    #define CHAR_S                      '\123'
1609    #define CHAR_T                      '\124'
1610    #define CHAR_U                      '\125'
1611    #define CHAR_V                      '\126'
1612    #define CHAR_W                      '\127'
1613    #define CHAR_X                      '\130'
1614    #define CHAR_Y                      '\131'
1615    #define CHAR_Z                      '\132'
1616    #define CHAR_LEFT_SQUARE_BRACKET    '\133'
1617    #define CHAR_BACKSLASH              '\134'
1618    #define CHAR_RIGHT_SQUARE_BRACKET   '\135'
1619    #define CHAR_CIRCUMFLEX_ACCENT      '\136'
1620    #define CHAR_UNDERSCORE             '\137'
1621    #define CHAR_GRAVE_ACCENT           '\140'
1622    #define CHAR_a                      '\141'
1623    #define CHAR_b                      '\142'
1624    #define CHAR_c                      '\143'
1625    #define CHAR_d                      '\144'
1626    #define CHAR_e                      '\145'
1627    #define CHAR_f                      '\146'
1628    #define CHAR_g                      '\147'
1629    #define CHAR_h                      '\150'
1630    #define CHAR_i                      '\151'
1631    #define CHAR_j                      '\152'
1632    #define CHAR_k                      '\153'
1633    #define CHAR_l                      '\154'
1634    #define CHAR_m                      '\155'
1635    #define CHAR_n                      '\156'
1636    #define CHAR_o                      '\157'
1637    #define CHAR_p                      '\160'
1638    #define CHAR_q                      '\161'
1639    #define CHAR_r                      '\162'
1640    #define CHAR_s                      '\163'
1641    #define CHAR_t                      '\164'
1642    #define CHAR_u                      '\165'
1643    #define CHAR_v                      '\166'
1644    #define CHAR_w                      '\167'
1645    #define CHAR_x                      '\170'
1646    #define CHAR_y                      '\171'
1647    #define CHAR_z                      '\172'
1648    #define CHAR_LEFT_CURLY_BRACKET     '\173'
1649    #define CHAR_VERTICAL_LINE          '\174'
1650    #define CHAR_RIGHT_CURLY_BRACKET    '\175'
1651    #define CHAR_TILDE                  '\176'
1652    
1653    #define STR_HT                      "\011"
1654    #define STR_VT                      "\013"
1655    #define STR_FF                      "\014"
1656    #define STR_CR                      "\015"
1657    #define STR_NL                      "\012"
1658    #define STR_BS                      "\010"
1659    #define STR_BEL                     "\007"
1660    #define STR_ESC                     "\033"
1661    #define STR_DEL                     "\177"
1662    
1663    #define STR_SPACE                   "\040"
1664    #define STR_EXCLAMATION_MARK        "\041"
1665    #define STR_QUOTATION_MARK          "\042"
1666    #define STR_NUMBER_SIGN             "\043"
1667    #define STR_DOLLAR_SIGN             "\044"
1668    #define STR_PERCENT_SIGN            "\045"
1669    #define STR_AMPERSAND               "\046"
1670    #define STR_APOSTROPHE              "\047"
1671    #define STR_LEFT_PARENTHESIS        "\050"
1672    #define STR_RIGHT_PARENTHESIS       "\051"
1673    #define STR_ASTERISK                "\052"
1674    #define STR_PLUS                    "\053"
1675    #define STR_COMMA                   "\054"
1676    #define STR_MINUS                   "\055"
1677    #define STR_DOT                     "\056"
1678    #define STR_SLASH                   "\057"
1679    #define STR_0                       "\060"
1680    #define STR_1                       "\061"
1681    #define STR_2                       "\062"
1682    #define STR_3                       "\063"
1683    #define STR_4                       "\064"
1684    #define STR_5                       "\065"
1685    #define STR_6                       "\066"
1686    #define STR_7                       "\067"
1687    #define STR_8                       "\070"
1688    #define STR_9                       "\071"
1689    #define STR_COLON                   "\072"
1690    #define STR_SEMICOLON               "\073"
1691    #define STR_LESS_THAN_SIGN          "\074"
1692    #define STR_EQUALS_SIGN             "\075"
1693    #define STR_GREATER_THAN_SIGN       "\076"
1694    #define STR_QUESTION_MARK           "\077"
1695    #define STR_COMMERCIAL_AT           "\100"
1696    #define STR_A                       "\101"
1697    #define STR_B                       "\102"
1698    #define STR_C                       "\103"
1699    #define STR_D                       "\104"
1700    #define STR_E                       "\105"
1701    #define STR_F                       "\106"
1702    #define STR_G                       "\107"
1703    #define STR_H                       "\110"
1704    #define STR_I                       "\111"
1705    #define STR_J                       "\112"
1706    #define STR_K                       "\113"
1707    #define STR_L                       "\114"
1708    #define STR_M                       "\115"
1709    #define STR_N                       "\116"
1710    #define STR_O                       "\117"
1711    #define STR_P                       "\120"
1712    #define STR_Q                       "\121"
1713    #define STR_R                       "\122"
1714    #define STR_S                       "\123"
1715    #define STR_T                       "\124"
1716    #define STR_U                       "\125"
1717    #define STR_V                       "\126"
1718    #define STR_W                       "\127"
1719    #define STR_X                       "\130"
1720    #define STR_Y                       "\131"
1721    #define STR_Z                       "\132"
1722    #define STR_LEFT_SQUARE_BRACKET     "\133"
1723    #define STR_BACKSLASH               "\134"
1724    #define STR_RIGHT_SQUARE_BRACKET    "\135"
1725    #define STR_CIRCUMFLEX_ACCENT       "\136"
1726    #define STR_UNDERSCORE              "\137"
1727    #define STR_GRAVE_ACCENT            "\140"
1728    #define STR_a                       "\141"
1729    #define STR_b                       "\142"
1730    #define STR_c                       "\143"
1731    #define STR_d                       "\144"
1732    #define STR_e                       "\145"
1733    #define STR_f                       "\146"
1734    #define STR_g                       "\147"
1735    #define STR_h                       "\150"
1736    #define STR_i                       "\151"
1737    #define STR_j                       "\152"
1738    #define STR_k                       "\153"
1739    #define STR_l                       "\154"
1740    #define STR_m                       "\155"
1741    #define STR_n                       "\156"
1742    #define STR_o                       "\157"
1743    #define STR_p                       "\160"
1744    #define STR_q                       "\161"
1745    #define STR_r                       "\162"
1746    #define STR_s                       "\163"
1747    #define STR_t                       "\164"
1748    #define STR_u                       "\165"
1749    #define STR_v                       "\166"
1750    #define STR_w                       "\167"
1751    #define STR_x                       "\170"
1752    #define STR_y                       "\171"
1753    #define STR_z                       "\172"
1754    #define STR_LEFT_CURLY_BRACKET      "\173"
1755    #define STR_VERTICAL_LINE           "\174"
1756    #define STR_RIGHT_CURLY_BRACKET     "\175"
1757    #define STR_TILDE                   "\176"
1758    
1759    #define STRING_ACCEPT0              STR_A STR_C STR_C STR_E STR_P STR_T "\0"
1760    #define STRING_COMMIT0              STR_C STR_O STR_M STR_M STR_I STR_T "\0"
1761    #define STRING_F0                   STR_F "\0"
1762    #define STRING_FAIL0                STR_F STR_A STR_I STR_L "\0"
1763    #define STRING_MARK0                STR_M STR_A STR_R STR_K "\0"
1764    #define STRING_PRUNE0               STR_P STR_R STR_U STR_N STR_E "\0"
1765    #define STRING_SKIP0                STR_S STR_K STR_I STR_P "\0"
1766    #define STRING_THEN                 STR_T STR_H STR_E STR_N
1767    
1768    #define STRING_alpha0               STR_a STR_l STR_p STR_h STR_a "\0"
1769    #define STRING_lower0               STR_l STR_o STR_w STR_e STR_r "\0"
1770    #define STRING_upper0               STR_u STR_p STR_p STR_e STR_r "\0"
1771    #define STRING_alnum0               STR_a STR_l STR_n STR_u STR_m "\0"
1772    #define STRING_ascii0               STR_a STR_s STR_c STR_i STR_i "\0"
1773    #define STRING_blank0               STR_b STR_l STR_a STR_n STR_k "\0"
1774    #define STRING_cntrl0               STR_c STR_n STR_t STR_r STR_l "\0"
1775    #define STRING_digit0               STR_d STR_i STR_g STR_i STR_t "\0"
1776    #define STRING_graph0               STR_g STR_r STR_a STR_p STR_h "\0"
1777    #define STRING_print0               STR_p STR_r STR_i STR_n STR_t "\0"
1778    #define STRING_punct0               STR_p STR_u STR_n STR_c STR_t "\0"
1779    #define STRING_space0               STR_s STR_p STR_a STR_c STR_e "\0"
1780    #define STRING_word0                STR_w STR_o STR_r STR_d       "\0"
1781    #define STRING_xdigit               STR_x STR_d STR_i STR_g STR_i STR_t
1782    
1783    #define STRING_DEFINE               STR_D STR_E STR_F STR_I STR_N STR_E
1784    
1785    #define STRING_CR_RIGHTPAR             STR_C STR_R STR_RIGHT_PARENTHESIS
1786    #define STRING_LF_RIGHTPAR             STR_L STR_F STR_RIGHT_PARENTHESIS
1787    #define STRING_CRLF_RIGHTPAR           STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1788    #define STRING_ANY_RIGHTPAR            STR_A STR_N STR_Y STR_RIGHT_PARENTHESIS
1789    #define STRING_ANYCRLF_RIGHTPAR        STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1790    #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
1791    #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
1792    #define STRING_UTF8_RIGHTPAR           STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1793    #define STRING_UTF16_RIGHTPAR          STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1794    #define STRING_UTF32_RIGHTPAR          STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1795    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_RIGHT_PARENTHESIS
1796    #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1797    #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
1798    
1799    #endif  /* SUPPORT_UTF */
1800    
1801  /* Escape items that are just an encoding of a particular data value. */  /* Escape items that are just an encoding of a particular data value. */
1802    
1803  #ifndef ESC_e  #ifndef ESC_e
1804  #define ESC_e 27  #define ESC_e CHAR_ESC
1805  #endif  #endif
1806    
1807  #ifndef ESC_f  #ifndef ESC_f
1808  #define ESC_f '\f'  #define ESC_f CHAR_FF
1809  #endif  #endif
1810    
1811  #ifndef ESC_n  #ifndef ESC_n
1812  #define ESC_n '\n'  #define ESC_n CHAR_LF
1813  #endif  #endif
1814    
1815  #ifndef ESC_r  #ifndef ESC_r
1816  #define ESC_r '\r'  #define ESC_r CHAR_CR
1817  #endif  #endif
1818    
1819  /* 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
1820  (presumably because of all the others like size_t). */  (presumably because of all the others like size_t). */
1821    
1822  #ifndef ESC_tee  #ifndef ESC_tee
1823  #define ESC_tee '\t'  #define ESC_tee CHAR_HT
1824  #endif  #endif
1825    
1826  /* Codes for different types of Unicode property */  /* Codes for different types of Unicode property */
1827    
1828  #define PT_ANY        0    /* Any property - matches all chars */  #define PT_ANY        0    /* Any property - matches all chars */
1829  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */  #define PT_LAMP       1    /* L& - the union of Lu, Ll, Lt */
1830  #define PT_GC         2    /* General characteristic (e.g. L) */  #define PT_GC         2    /* Specified general characteristic (e.g. L) */
1831  #define PT_PC         3    /* Particular characteristic (e.g. Lu) */  #define PT_PC         3    /* Specified particular characteristic (e.g. Lu) */
1832  #define PT_SC         4    /* Script (e.g. Han) */  #define PT_SC         4    /* Script (e.g. Han) */
1833    #define PT_ALNUM      5    /* Alphanumeric - the union of L and N */
1834    #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1835    #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1836    #define PT_WORD       8    /* Word - L plus N plus underscore */
1837    #define PT_CLIST      9    /* Pseudo-property: match character list */
1838    
1839  /* 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
1840  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1841    
1842  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1843  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */
# Line 532  contain UTF-8 characters with values gre Line 1850  contain UTF-8 characters with values gre
1850    
1851  /* 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
1852  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
1853  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
1854  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
1855  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
1856  ESC_REF as subsequent values are used for backreferences (\1, \2, \3, etc).  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1857  There are two tests in the code for an escape greater than ESC_b and less than  non-DOTALL mode, "." behaves like \N.
1858  ESC_Z to detect the types that may be repeated. These are the types that  
1859  consume 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.
1860  character, that code will have to change. */  when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1861    They must be contiguous, and remain in order so that the replacements can be
1862  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.
1863         ESC_w, ESC_dum1, ESC_C, ESC_P, ESC_p, ESC_R, ESC_X, ESC_Z, ESC_z,  
1864         ESC_E, ESC_Q, ESC_k, ESC_REF };  Negative numbers are used to encode a backreference (\1, \2, \3, etc.) in
1865    check_escape(). There are two tests in the code for an escape
1866    greater than ESC_b and less than ESC_Z to detect the types that may be
1867    repeated. These are the types that consume characters. If any new escapes are
1868    put in between that don't consume a character, that code will have to change.
1869    */
1870    
1871    enum { ESC_A = 1, ESC_G, ESC_K, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s,
1872           ESC_W, ESC_w, ESC_N, ESC_dum, ESC_C, ESC_P, ESC_p, ESC_R, ESC_H,
1873           ESC_h, ESC_V, ESC_v, ESC_X, ESC_Z, ESC_z,
1874           ESC_E, ESC_Q, ESC_g, ESC_k,
1875           ESC_DU, ESC_du, ESC_SU, ESC_su, ESC_WU, ESC_wu };
1876    
1877  /* 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  
1878  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.
1879    
 To keep stored, compiled patterns compatible, new opcodes should be added  
 immediately before OP_BRA, where (since release 7.0) a gap is left for this  
 purpose.  
   
1880  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions  *** NOTE NOTE NOTE *** Whenever this list is updated, the two macro definitions
1881  that follow must also be updated to match. There is also a table called  that follow must also be updated to match. There are also tables called
1882  "coptable" in pcre_dfa_exec.c that must be updated. */  "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
1883    
1884  enum {  enum {
1885    OP_END,            /* 0 End of pattern */    OP_END,            /* 0 End of pattern */
# Line 565  enum { Line 1888  enum {
1888    
1889    OP_SOD,            /* 1 Start of data: \A */    OP_SOD,            /* 1 Start of data: \A */
1890    OP_SOM,            /* 2 Start of match (subject + offset): \G */    OP_SOM,            /* 2 Start of match (subject + offset): \G */
1891    OP_NOT_WORD_BOUNDARY,  /*  3 \B */    OP_SET_SOM,        /* 3 Set start of match (\K) */
1892    OP_WORD_BOUNDARY,      /*  4 \b */    OP_NOT_WORD_BOUNDARY,  /*  4 \B */
1893    OP_NOT_DIGIT,          /*  5 \D */    OP_WORD_BOUNDARY,      /*  5 \b */
1894    OP_DIGIT,              /*  6 \d */    OP_NOT_DIGIT,          /*  6 \D */
1895    OP_NOT_WHITESPACE,     /*  7 \S */    OP_DIGIT,              /*  7 \d */
1896    OP_WHITESPACE,         /*  8 \s */    OP_NOT_WHITESPACE,     /*  8 \S */
1897    OP_NOT_WORDCHAR,       /*  9 \W */    OP_WHITESPACE,         /*  9 \s */
1898    OP_WORDCHAR,           /* 10 \w */    OP_NOT_WORDCHAR,       /* 10 \W */
1899    OP_ANY,            /* 11 Match any character */    OP_WORDCHAR,           /* 11 \w */
1900    OP_ANYBYTE,        /* 12 Match any byte (\C); different to OP_ANY for UTF-8 */  
1901    OP_NOTPROP,        /* 13 \P (not Unicode property) */    OP_ANY,            /* 12 Match any character except newline (\N) */
1902    OP_PROP,           /* 14 \p (Unicode property) */    OP_ALLANY,         /* 13 Match any character */
1903    OP_ANYNL,          /* 15 \R (any newline sequence) */    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1904    OP_EXTUNI,         /* 16 \X (extended Unicode sequence */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
1905    OP_EODN,           /* 17 End of data or \n at end of data: \Z. */    OP_PROP,           /* 16 \p (Unicode property) */
1906    OP_EOD,            /* 18 End of data: \z */    OP_ANYNL,          /* 17 \R (any newline sequence) */
1907      OP_NOT_HSPACE,     /* 18 \H (not horizontal whitespace) */
1908    OP_OPT,            /* 19 Set runtime options */    OP_HSPACE,         /* 19 \h (horizontal whitespace) */
1909    OP_CIRC,           /* 20 Start of line - varies with multiline switch */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1910    OP_DOLL,           /* 21 End of line - varies with multiline switch */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1911    OP_CHAR,           /* 22 Match one character, casefully */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1912    OP_CHARNC,         /* 23 Match one character, caselessly */    OP_EODN,           /* 23 End of data or \n at end of data (\Z) */
1913    OP_NOT,            /* 24 Match one character, not the following one */    OP_EOD,            /* 24 End of data (\z) */
1914    
1915    OP_STAR,           /* 25 The maximizing and minimizing versions of */    OP_CIRC,           /* 25 Start of line - not multiline */
1916    OP_MINSTAR,        /* 26 these six opcodes must come in pairs, with */    OP_CIRCM,          /* 26 Start of line - multiline */
1917    OP_PLUS,           /* 27 the minimizing one second. */    OP_DOLL,           /* 27 End of line - not multiline */
1918    OP_MINPLUS,        /* 28 This first set applies to single characters.*/    OP_DOLLM,          /* 28 End of line - multiline */
1919    OP_QUERY,          /* 29 */    OP_CHAR,           /* 29 Match one character, casefully */
1920    OP_MINQUERY,       /* 30 */    OP_CHARI,          /* 30 Match one character, caselessly */
1921      OP_NOT,            /* 31 Match one character, not the given one, casefully */
1922    OP_UPTO,           /* 31 From 0 to n matches */    OP_NOTI,           /* 32 Match one character, not the given one, caselessly */
1923    OP_MINUPTO,        /* 32 */  
1924    OP_EXACT,          /* 33 Exactly n matches */    /* The following sets of 13 opcodes must always be kept in step because
1925      the offset from the first one is used to generate the others. */
1926    OP_POSSTAR,        /* 34 Possessified star */  
1927    OP_POSPLUS,        /* 35 Possessified plus */    /**** Single characters, caseful, must precede the caseless ones ****/
1928    OP_POSQUERY,       /* 36 Posesssified query */  
1929    OP_POSUPTO,        /* 37 Possessified upto */    OP_STAR,           /* 33 The maximizing and minimizing versions of */
1930      OP_MINSTAR,        /* 34 these six opcodes must come in pairs, with */
1931    OP_NOTSTAR,        /* 38 The maximizing and minimizing versions of */    OP_PLUS,           /* 35 the minimizing one second. */
1932    OP_NOTMINSTAR,     /* 39 these six opcodes must come in pairs, with */    OP_MINPLUS,        /* 36 */
1933    OP_NOTPLUS,        /* 40 the minimizing one second. They must be in */    OP_QUERY,          /* 37 */
1934    OP_NOTMINPLUS,     /* 41 exactly the same order as those above. */    OP_MINQUERY,       /* 38 */
1935    OP_NOTQUERY,       /* 42 This set applies to "not" single characters. */  
1936    OP_NOTMINQUERY,    /* 43 */    OP_UPTO,           /* 39 From 0 to n matches of one character, caseful*/
1937      OP_MINUPTO,        /* 40 */
1938    OP_NOTUPTO,        /* 44 From 0 to n matches */    OP_EXACT,          /* 41 Exactly n matches */
1939    OP_NOTMINUPTO,     /* 45 */  
1940    OP_NOTEXACT,       /* 46 Exactly n matches */    OP_POSSTAR,        /* 42 Possessified star, caseful */
1941      OP_POSPLUS,        /* 43 Possessified plus, caseful */
1942    OP_NOTPOSSTAR,     /* 47 Possessified versions */    OP_POSQUERY,       /* 44 Posesssified query, caseful */
1943    OP_NOTPOSPLUS,     /* 48 */    OP_POSUPTO,        /* 45 Possessified upto, caseful */
1944    OP_NOTPOSQUERY,    /* 49 */  
1945    OP_NOTPOSUPTO,     /* 50 */    /**** Single characters, caseless, must follow the caseful ones */
1946    
1947    OP_TYPESTAR,       /* 51 The maximizing and minimizing versions of */    OP_STARI,          /* 46 */
1948    OP_TYPEMINSTAR,    /* 52 these six opcodes must come in pairs, with */    OP_MINSTARI,       /* 47 */
1949    OP_TYPEPLUS,       /* 53 the minimizing one second. These codes must */    OP_PLUSI,          /* 48 */
1950    OP_TYPEMINPLUS,    /* 54 be in exactly the same order as those above. */    OP_MINPLUSI,       /* 49 */
1951    OP_TYPEQUERY,      /* 55 This set applies to character types such as \d */    OP_QUERYI,         /* 50 */
1952    OP_TYPEMINQUERY,   /* 56 */    OP_MINQUERYI,      /* 51 */
1953    
1954    OP_TYPEUPTO,       /* 57 From 0 to n matches */    OP_UPTOI,          /* 52 From 0 to n matches of one character, caseless */
1955    OP_TYPEMINUPTO,    /* 58 */    OP_MINUPTOI,       /* 53 */
1956    OP_TYPEEXACT,      /* 59 Exactly n matches */    OP_EXACTI,         /* 54 */
1957    
1958    OP_TYPEPOSSTAR,    /* 60 Possessified versions */    OP_POSSTARI,       /* 55 Possessified star, caseless */
1959    OP_TYPEPOSPLUS,    /* 61 */    OP_POSPLUSI,       /* 56 Possessified plus, caseless */
1960    OP_TYPEPOSQUERY,   /* 62 */    OP_POSQUERYI,      /* 57 Posesssified query, caseless */
1961    OP_TYPEPOSUPTO,    /* 63 */    OP_POSUPTOI,       /* 58 Possessified upto, caseless */
1962    
1963    OP_CRSTAR,         /* 64 The maximizing and minimizing versions of */    /**** The negated ones must follow the non-negated ones, and match them ****/
1964    OP_CRMINSTAR,      /* 65 all these opcodes must come in pairs, with */    /**** Negated single character, caseful; must precede the caseless ones ****/
1965    OP_CRPLUS,         /* 66 the minimizing one second. These codes must */  
1966    OP_CRMINPLUS,      /* 67 be in exactly the same order as those above. */    OP_NOTSTAR,        /* 59 The maximizing and minimizing versions of */
1967    OP_CRQUERY,        /* 68 These are for character classes and back refs */    OP_NOTMINSTAR,     /* 60 these six opcodes must come in pairs, with */
1968    OP_CRMINQUERY,     /* 69 */    OP_NOTPLUS,        /* 61 the minimizing one second. They must be in */
1969    OP_CRRANGE,        /* 70 These are different to the three sets above. */    OP_NOTMINPLUS,     /* 62 exactly the same order as those above. */
1970    OP_CRMINRANGE,     /* 71 */    OP_NOTQUERY,       /* 63 */
1971      OP_NOTMINQUERY,    /* 64 */
1972    OP_CLASS,          /* 72 Match a character class, chars < 256 only */  
1973    OP_NCLASS,         /* 73 Same, but the bitmap was created from a negative    OP_NOTUPTO,        /* 65 From 0 to n matches, caseful */
1974                             class - the difference is relevant only when a UTF-8    OP_NOTMINUPTO,     /* 66 */
1975                             character > 255 is encountered. */    OP_NOTEXACT,       /* 67 Exactly n matches */
1976    
1977    OP_XCLASS,         /* 74 Extended class for handling UTF-8 chars within the    OP_NOTPOSSTAR,     /* 68 Possessified versions, caseful */
1978                             class. This does both positive and negative. */    OP_NOTPOSPLUS,     /* 69 */
1979      OP_NOTPOSQUERY,    /* 70 */
1980    OP_REF,            /* 75 Match a back reference */    OP_NOTPOSUPTO,     /* 71 */
1981    OP_RECURSE,        /* 76 Match a numbered subpattern (possibly recursive) */  
1982    OP_CALLOUT,        /* 77 Call out to external function if provided */    /**** Negated single character, caseless; must follow the caseful ones ****/
1983    
1984    OP_ALT,            /* 78 Start of alternation */    OP_NOTSTARI,       /* 72 */
1985    OP_KET,            /* 79 End of group that doesn't have an unbounded repeat */    OP_NOTMINSTARI,    /* 73 */
1986    OP_KETRMAX,        /* 80 These two must remain together and in this */    OP_NOTPLUSI,       /* 74 */
1987    OP_KETRMIN,        /* 81 order. They are for groups the repeat for ever. */    OP_NOTMINPLUSI,    /* 75 */
1988      OP_NOTQUERYI,      /* 76 */
1989    /* The assertions must come before BRA, CBRA, ONCE, and COND.*/    OP_NOTMINQUERYI,   /* 77 */
1990    
1991    OP_ASSERT,         /* 82 Positive lookahead */    OP_NOTUPTOI,       /* 78 From 0 to n matches, caseless */
1992    OP_ASSERT_NOT,     /* 83 Negative lookahead */    OP_NOTMINUPTOI,    /* 79 */
1993    OP_ASSERTBACK,     /* 84 Positive lookbehind */    OP_NOTEXACTI,      /* 80 Exactly n matches */
1994    OP_ASSERTBACK_NOT, /* 85 Negative lookbehind */  
1995    OP_REVERSE,        /* 86 Move pointer back - used in lookbehind assertions */    OP_NOTPOSSTARI,    /* 81 Possessified versions, caseless */
1996      OP_NOTPOSPLUSI,    /* 82 */
1997    /* ONCE, BRA, CBRA, and COND must come after the assertions, with ONCE first,    OP_NOTPOSQUERYI,   /* 83 */
1998    as there's a test for >= ONCE for a subpattern that isn't an assertion. */    OP_NOTPOSUPTOI,    /* 84 */
1999    
2000    OP_ONCE,           /* 87 Atomic group */    /**** Character types ****/
2001    OP_BRA,            /* 88 Start of non-capturing bracket */  
2002    OP_CBRA,           /* 89 Start of capturing bracket */    OP_TYPESTAR,       /* 85 The maximizing and minimizing versions of */
2003    OP_COND,           /* 90 Conditional group */    OP_TYPEMINSTAR,    /* 86 these six opcodes must come in pairs, with */
2004      OP_TYPEPLUS,       /* 87 the minimizing one second. These codes must */
2005      OP_TYPEMINPLUS,    /* 88 be in exactly the same order as those above. */
2006      OP_TYPEQUERY,      /* 89 */
2007      OP_TYPEMINQUERY,   /* 90 */
2008    
2009      OP_TYPEUPTO,       /* 91 From 0 to n matches */
2010      OP_TYPEMINUPTO,    /* 92 */
2011      OP_TYPEEXACT,      /* 93 Exactly n matches */
2012    
2013      OP_TYPEPOSSTAR,    /* 94 Possessified versions */
2014      OP_TYPEPOSPLUS,    /* 95 */
2015      OP_TYPEPOSQUERY,   /* 96 */
2016      OP_TYPEPOSUPTO,    /* 97 */
2017    
2018      /* These are used for character classes and back references; only the
2019      first six are the same as the sets above. */
2020    
2021      OP_CRSTAR,         /* 98 The maximizing and minimizing versions of */
2022      OP_CRMINSTAR,      /* 99 all these opcodes must come in pairs, with */
2023      OP_CRPLUS,         /* 100 the minimizing one second. These codes must */
2024      OP_CRMINPLUS,      /* 101 be in exactly the same order as those above. */
2025      OP_CRQUERY,        /* 102 */
2026      OP_CRMINQUERY,     /* 103 */
2027    
2028      OP_CRRANGE,        /* 104 These are different to the three sets above. */
2029      OP_CRMINRANGE,     /* 105 */
2030    
2031      /* End of quantifier opcodes */
2032    
2033      OP_CLASS,          /* 106 Match a character class, chars < 256 only */
2034      OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
2035                                  class - the difference is relevant only when a
2036                                  character > 255 is encountered. */
2037      OP_XCLASS,         /* 108 Extended class for handling > 255 chars within the
2038                                  class. This does both positive and negative. */
2039      OP_REF,            /* 109 Match a back reference, casefully */
2040      OP_REFI,           /* 110 Match a back reference, caselessly */
2041      OP_RECURSE,        /* 111 Match a numbered subpattern (possibly recursive) */
2042      OP_CALLOUT,        /* 112 Call out to external function if provided */
2043    
2044      OP_ALT,            /* 113 Start of alternation */
2045      OP_KET,            /* 114 End of group that doesn't have an unbounded repeat */
2046      OP_KETRMAX,        /* 115 These two must remain together and in this */
2047      OP_KETRMIN,        /* 116 order. They are for groups the repeat for ever. */
2048      OP_KETRPOS,        /* 117 Possessive unlimited repeat. */
2049    
2050      /* The assertions must come before BRA, CBRA, ONCE, and COND, and the four
2051      asserts must remain in order. */
2052    
2053      OP_REVERSE,        /* 118 Move pointer back - used in lookbehind assertions */
2054      OP_ASSERT,         /* 119 Positive lookahead */
2055      OP_ASSERT_NOT,     /* 120 Negative lookahead */
2056      OP_ASSERTBACK,     /* 121 Positive lookbehind */
2057      OP_ASSERTBACK_NOT, /* 122 Negative lookbehind */
2058    
2059      /* ONCE, ONCE_NC, BRA, BRAPOS, CBRA, CBRAPOS, and COND must come immediately
2060      after the assertions, with ONCE first, as there's a test for >= ONCE for a
2061      subpattern that isn't an assertion. The POS versions must immediately follow
2062      the non-POS versions in each case. */
2063    
2064      OP_ONCE,           /* 123 Atomic group, contains captures */
2065      OP_ONCE_NC,        /* 124 Atomic group containing no captures */
2066      OP_BRA,            /* 125 Start of non-capturing bracket */
2067      OP_BRAPOS,         /* 126 Ditto, with unlimited, possessive repeat */
2068      OP_CBRA,           /* 127 Start of capturing bracket */
2069      OP_CBRAPOS,        /* 128 Ditto, with unlimited, possessive repeat */
2070      OP_COND,           /* 129 Conditional group */
2071    
2072    /* These three must follow the previous three, in the same order. There's a    /* These five must follow the previous five, in the same order. There's a
2073    check for >= SBRA to distinguish the two sets. */    check for >= SBRA to distinguish the two sets. */
2074    
2075    OP_SBRA,           /* 91 Start of non-capturing bracket, check empty  */    OP_SBRA,           /* 130 Start of non-capturing bracket, check empty  */
2076    OP_SCBRA,          /* 92 Start of capturing bracket, check empty */    OP_SBRAPOS,        /* 131 Ditto, with unlimited, possessive repeat */
2077    OP_SCOND,          /* 93 Conditional group, check empty */    OP_SCBRA,          /* 132 Start of capturing bracket, check empty */
2078      OP_SCBRAPOS,       /* 133 Ditto, with unlimited, possessive repeat */
2079    OP_CREF,           /* 94 Used to hold a capture number as condition */    OP_SCOND,          /* 134 Conditional group, check empty */
2080    OP_RREF,           /* 95 Used to hold a recursion number as condition */  
2081    OP_DEF,            /* 96 The DEFINE condition */    /* The next two pairs must (respectively) be kept together. */
2082    
2083      OP_CREF,           /* 135 Used to hold a capture number as condition */
2084      OP_NCREF,          /* 136 Same, but generated by a name reference*/
2085      OP_RREF,           /* 137 Used to hold a recursion number as condition */
2086      OP_NRREF,          /* 138 Same, but generated by a name reference*/
2087      OP_DEF,            /* 139 The DEFINE condition */
2088    
2089      OP_BRAZERO,        /* 140 These two must remain together and in this */
2090      OP_BRAMINZERO,     /* 141 order. */
2091      OP_BRAPOSZERO,     /* 142 */
2092    
2093      /* These are backtracking control verbs */
2094    
2095      OP_MARK,           /* 143 always has an argument */
2096      OP_PRUNE,          /* 144 */
2097      OP_PRUNE_ARG,      /* 145 same, but with argument */
2098      OP_SKIP,           /* 146 */
2099      OP_SKIP_ARG,       /* 147 same, but with argument */
2100      OP_THEN,           /* 148 */
2101      OP_THEN_ARG,       /* 149 same, but with argument */
2102      OP_COMMIT,         /* 150 */
2103    
2104      /* These are forced failure and success verbs */
2105    
2106      OP_FAIL,           /* 151 */
2107      OP_ACCEPT,         /* 152 */
2108      OP_ASSERT_ACCEPT,  /* 153 Used inside assertions */
2109      OP_CLOSE,          /* 154 Used before OP_ACCEPT to close open captures */
2110    
2111      /* This is used to skip a subpattern with a {0} quantifier */
2112    
2113      OP_SKIPZERO,       /* 155 */
2114    
2115      /* This is not an opcode, but is used to check that tables indexed by opcode
2116      are the correct length, in order to catch updating errors - there have been
2117      some in the past. */
2118    
2119    OP_BRAZERO,        /* 97 These two must remain together and in this */    OP_TABLE_LENGTH
   OP_BRAMINZERO      /* 98 order. */  
2120  };  };
2121    
2122    /* *** NOTE NOTE NOTE *** Whenever the list above is updated, the two macro
2123    definitions that follow must also be updated to match. There are also tables
2124    called "coptable" and "poptable" in pcre_dfa_exec.c that must be updated. */
2125    
2126    
2127  /* 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
2128  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
2129    only in pcre_printint.c, which fills out the full names in many cases (and in
2130    some cases doesn't actually use these names at all). */
2131    
2132  #define OP_NAME_LIST \  #define OP_NAME_LIST \
2133    "End", "\\A", "\\G", "\\B", "\\b", "\\D", "\\d",                \    "End", "\\A", "\\G", "\\K", "\\B", "\\b", "\\D", "\\d",         \
2134    "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte",                   \    "\\S", "\\s", "\\W", "\\w", "Any", "AllAny", "Anybyte",         \
2135    "notprop", "prop", "anynl", "extuni",                           \    "notprop", "prop", "\\R", "\\H", "\\h", "\\V", "\\v",           \
2136    "\\Z", "\\z",                                                   \    "extuni",  "\\Z", "\\z",                                        \
2137    "Opt", "^", "$", "char", "charnc", "not",                       \    "^", "^", "$", "$", "char", "chari", "not", "noti",             \
2138    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2139      "{", "{", "{",                                                  \
2140    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2141    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??",                                \
2142      "{", "{", "{",                                                  \
2143      "*+","++", "?+", "{",                                           \
2144      "*", "*?", "+", "+?", "?", "??",                                \
2145      "{", "{", "{",                                                  \
2146      "*+","++", "?+", "{",                                           \
2147      "*", "*?", "+", "+?", "?", "??",                                \
2148      "{", "{", "{",                                                  \
2149    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2150    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \    "*", "*?", "+", "+?", "?", "??", "{", "{", "{",                 \
2151    "*+","++", "?+", "{",                                           \    "*+","++", "?+", "{",                                           \
2152    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \    "*", "*?", "+", "+?", "?", "??", "{", "{",                      \
2153    "class", "nclass", "xclass", "Ref", "Recurse", "Callout",       \    "class", "nclass", "xclass", "Ref", "Refi",                     \
2154    "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not",     \    "Recurse", "Callout",                                           \
2155    "AssertB", "AssertB not", "Reverse",                            \    "Alt", "Ket", "KetRmax", "KetRmin", "KetRpos",                  \
2156    "Once", "Bra 0", "Bra", "Cond", "SBra 0", "SBra", "SCond",      \    "Reverse", "Assert", "Assert not", "AssertB", "AssertB not",    \
2157    "Cond ref", "Cond rec", "Cond def", "Brazero", "Braminzero"    "Once", "Once_NC",                                              \
2158      "Bra", "BraPos", "CBra", "CBraPos",                             \
2159      "Cond",                                                         \
2160      "SBra", "SBraPos", "SCBra", "SCBraPos",                         \
2161      "SCond",                                                        \
2162      "Cond ref", "Cond nref", "Cond rec", "Cond nrec", "Cond def",   \
2163      "Brazero", "Braminzero", "Braposzero",                          \
2164      "*MARK", "*PRUNE", "*PRUNE", "*SKIP", "*SKIP",                  \
2165      "*THEN", "*THEN", "*COMMIT", "*FAIL",                           \
2166      "*ACCEPT", "*ASSERT_ACCEPT",                                    \
2167      "Close", "Skip zero"
2168    
2169    
2170  /* 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 729  in UTF-8 mode. The code that uses this t Line 2178  in UTF-8 mode. The code that uses this t
2178    
2179  #define OP_LENGTHS \  #define OP_LENGTHS \
2180    1,                             /* End                                    */ \    1,                             /* End                                    */ \
2181    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                     */ \
2182    1, 1,                          /* Any, Anybyte                           */ \    1, 1, 1, 1, 1, 1,              /* \D, \d, \S, \s, \W, \w                 */ \
2183    3, 3, 1, 1,                    /* NOTPROP, PROP, EXTUNI, ANYNL           */ \    1, 1, 1,                       /* Any, AllAny, Anybyte                   */ \
2184    1, 1, 2, 1, 1,                 /* \Z, \z, Opt, ^, $                      */ \    3, 3,                          /* \P, \p                                 */ \
2185      1, 1, 1, 1, 1,                 /* \R, \H, \h, \V, \v                     */ \
2186      1,                             /* \X                                     */ \
2187      1, 1, 1, 1, 1, 1,              /* \Z, \z, ^, ^M, $, $M                   */ \
2188    2,                             /* Char  - the minimum length             */ \    2,                             /* Char  - the minimum length             */ \
2189    2,                             /* Charnc  - the minimum length           */ \    2,                             /* Chari  - the minimum length            */ \
2190    2,                             /* not                                    */ \    2,                             /* not                                    */ \
2191    /* Positive single-char repeats                            ** These are  */ \    2,                             /* noti                                   */ \
2192    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??      ** minima in  */ \    /* Positive single-char repeats                             ** These are */ \
2193    4, 4, 4,                       /* upto, minupto, exact     ** UTF-8 mode */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2194    2, 2, 2, 4,                    /* *+, ++, ?+, upto+                      */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2195      2+IMM2_SIZE,                   /* exact                                  */ \
2196      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2197      2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2198      2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2199      2+IMM2_SIZE,                   /* exact I                                */ \
2200      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2201    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2202    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2203    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2204    2, 2, 2, 4,                    /* Possessive *, +, ?, upto               */ \    2+IMM2_SIZE,                   /* NOT exact                              */ \
2205      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2206      2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2207      2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2208      2+IMM2_SIZE,                   /* NOT exact I                            */ \
2209      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2210    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2211    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2212    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2213    2, 2, 2, 4,                    /* Possessive *+, ++, ?+, upto+           */ \    2+IMM2_SIZE,                   /* Type exact                             */ \
2214      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2215    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2216    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2217    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2218   33,                             /* CLASS                                  */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2219   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2220    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2221    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2222      1+IMM2_SIZE,                   /* REFI                                   */ \
2223    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2224    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2225    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
2226    1+LINK_SIZE,                   /* Ket                                    */ \    1+LINK_SIZE,                   /* Ket                                    */ \
2227    1+LINK_SIZE,                   /* KetRmax                                */ \    1+LINK_SIZE,                   /* KetRmax                                */ \
2228    1+LINK_SIZE,                   /* KetRmin                                */ \    1+LINK_SIZE,                   /* KetRmin                                */ \
2229      1+LINK_SIZE,                   /* KetRpos                                */ \
2230      1+LINK_SIZE,                   /* Reverse                                */ \
2231    1+LINK_SIZE,                   /* Assert                                 */ \    1+LINK_SIZE,                   /* Assert                                 */ \
2232    1+LINK_SIZE,                   /* Assert not                             */ \    1+LINK_SIZE,                   /* Assert not                             */ \
2233    1+LINK_SIZE,                   /* Assert behind                          */ \    1+LINK_SIZE,                   /* Assert behind                          */ \
2234    1+LINK_SIZE,                   /* Assert behind not                      */ \    1+LINK_SIZE,                   /* Assert behind not                      */ \
   1+LINK_SIZE,                   /* Reverse                                */ \  
2235    1+LINK_SIZE,                   /* ONCE                                   */ \    1+LINK_SIZE,                   /* ONCE                                   */ \
2236      1+LINK_SIZE,                   /* ONCE_NC                                */ \
2237    1+LINK_SIZE,                   /* BRA                                    */ \    1+LINK_SIZE,                   /* BRA                                    */ \
2238    3+LINK_SIZE,                   /* CBRA                                   */ \    1+LINK_SIZE,                   /* BRAPOS                                 */ \
2239      1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2240      1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2241    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2242    1+LINK_SIZE,                   /* SBRA                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2243    3+LINK_SIZE,                   /* SCBRA                                  */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2244      1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2245      1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2246    1+LINK_SIZE,                   /* SCOND                                  */ \    1+LINK_SIZE,                   /* SCOND                                  */ \
2247    3,                             /* CREF                                   */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* CREF, NCREF                            */ \
2248    3,                             /* RREF                                   */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* RREF, NRREF                            */ \
2249    1,                             /* DEF                                    */ \    1,                             /* DEF                                    */ \
2250    1, 1,                          /* BRAZERO, BRAMINZERO                    */ \    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2251      3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2252      1, 3,                          /* SKIP, SKIP_ARG                         */ \
2253      1, 3,                          /* THEN, THEN_ARG                         */ \
2254      1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2255      1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2256    
2257    /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
2258  /* A magic value for OP_RREF to indicate the "any recursion" condition. */  condition. */
2259    
2260  #define RREF_ANY  0xffff  #define RREF_ANY  0xffff
2261    
2262  /* Error code numbers. They are given names so that they can more easily be  /* Compile time error code numbers. They are given names so that they can more
2263  tracked. */  easily be tracked. When a new number is added, the table called eint in
2264    pcreposix.c must be updated. */
2265    
2266  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,  ERR5,  ERR6,  ERR7,  ERR8,  ERR9,
2267         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,         ERR10, ERR11, ERR12, ERR13, ERR14, ERR15, ERR16, ERR17, ERR18, ERR19,
2268         ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,         ERR20, ERR21, ERR22, ERR23, ERR24, ERR25, ERR26, ERR27, ERR28, ERR29,
2269         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,         ERR30, ERR31, ERR32, ERR33, ERR34, ERR35, ERR36, ERR37, ERR38, ERR39,
2270         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2271         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57 };         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2272           ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2273           ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2274    
2275    /* JIT compiling modes. The function list is indexed by them. */
2276    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2277           JIT_NUMBER_OF_COMPILE_MODES };
2278    
2279  /* 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
2280  code vector run on as long as necessary after the end. We store an explicit  code vector run on as long as necessary after the end. We store an explicit
# Line 807  Because people can now save and re-use c Line 2290  Because people can now save and re-use c
2290  structure should be made at the end, and something earlier (e.g. a new  structure should be made at the end, and something earlier (e.g. a new
2291  flag in the options or one of the dummy fields) should indicate that the new  flag in the options or one of the dummy fields) should indicate that the new
2292  fields are present. Currently PCRE always sets the dummy fields to zero.  fields are present. Currently PCRE always sets the dummy fields to zero.
2293  NOTE NOTE NOTE:  NOTE NOTE NOTE
2294  */  */
2295    
2296  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2297    #define REAL_PCRE real_pcre
2298    #elif defined COMPILE_PCRE16
2299    #define REAL_PCRE real_pcre16
2300    #elif defined COMPILE_PCRE32
2301    #define REAL_PCRE real_pcre32
2302    #endif
2303    
2304    /* It is necessary to fork the struct for 32 bit, since it needs to use
2305     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2306     * typedef since pcretest needs access to  the struct of the 8-, 16-
2307     * and 32-bit variants. */
2308    
2309    typedef struct real_pcre8_or_16 {
2310    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2311    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2312    pcre_uint32 options;    pcre_uint32 options;            /* Public options */
2313    pcre_uint32 dummy1;             /* For future use, maybe */    pcre_uint16 flags;              /* Private flags */
2314      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2315    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2316    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2317    pcre_uint16 first_byte;    pcre_uint16 first_char;         /* Starting character */
2318    pcre_uint16 req_byte;    pcre_uint16 req_char;           /* This character must be seen */
2319    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2320    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2321    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2322    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2323      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2324      const pcre_uint8 *nullpad;      /* NULL padding */
2325    } real_pcre8_or_16;
2326    
2327    const unsigned char *tables;    /* Pointer to tables or NULL for std */  typedef struct real_pcre8_or_16 real_pcre;
2328    const unsigned char *nullpad;   /* NULL padding */  typedef struct real_pcre8_or_16 real_pcre16;
2329  } real_pcre;  
2330    typedef struct real_pcre32 {
2331      pcre_uint32 magic_number;
2332      pcre_uint32 size;               /* Total that was malloced */
2333      pcre_uint32 options;            /* Public options */
2334      pcre_uint16 flags;              /* Private flags */
2335      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2336      pcre_uint16 top_bracket;        /* Highest numbered group */
2337      pcre_uint16 top_backref;        /* Highest numbered back reference */
2338      pcre_uint32 first_char;         /* Starting character */
2339      pcre_uint32 req_char;           /* This character must be seen */
2340      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2341      pcre_uint16 name_entry_size;    /* Size of any name items */
2342      pcre_uint16 name_count;         /* Number of name items */
2343      pcre_uint16 ref_count;          /* Reference count */
2344      pcre_uint16 dummy1;             /* for later expansion */
2345      pcre_uint16 dummy2;             /* for later expansion */
2346      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2347      void *nullpad;                  /* for later expansion */
2348    } real_pcre32;
2349    
2350    /* Assert that the size of REAL_PCRE is divisible by 8 */
2351    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2352    
2353    /* Needed in pcretest to access some fields in the real_pcre* structures
2354     * directly. They're unified for 8/16/32 bits since the structs only differ
2355     * after these fields; if that ever changes, need to fork those defines into
2356     * 8/16 and 32 bit versions. */
2357    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2358    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2359    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2360    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2361    
2362  /* 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
2363  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
2364    
2365  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2366    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2367    pcre_uint32 options;    pcre_uint32 flags;              /* Private flags */
2368    uschar start_bits[32];    pcre_uint8 start_bits[32];      /* Starting char bits */
2369      pcre_uint32 minlength;          /* Minimum subject length */
2370  } pcre_study_data;  } pcre_study_data;
2371    
2372    /* Structure for building a chain of open capturing subpatterns during
2373    compiling, so that instructions to close them can be compiled when (*ACCEPT) is
2374    encountered. This is also used to identify subpatterns that contain recursive
2375    back references to themselves, so that they can be made atomic. */
2376    
2377    typedef struct open_capitem {
2378      struct open_capitem *next;    /* Chain link */
2379      pcre_uint16 number;           /* Capture number */
2380      pcre_uint16 flag;             /* Set TRUE if recursive back ref */
2381    } open_capitem;
2382    
2383  /* Structure for passing "static" information around between the functions  /* Structure for passing "static" information around between the functions
2384  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2385    
2386  typedef struct compile_data {  typedef struct compile_data {
2387    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2388    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2389    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2390    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2391    const uschar *start_workspace;/* The start of working space */    const pcre_uchar *start_workspace;/* The start of working space */
2392    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_code;     /* The start of the compiled code */
2393    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2394    const uschar *end_pattern;    /* The end of the pattern */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2395    uschar *hwm;                  /* High watermark of workspace */    open_capitem *open_caps;          /* Chain of open capture items */
2396    uschar *name_table;           /* The name/number table */    pcre_uchar *hwm;                  /* High watermark of workspace */
2397    int  names_found;             /* Number of entries so far */    pcre_uchar *name_table;           /* The name/number table */
2398    int  name_entry_size;         /* Size of each entry */    int  names_found;                 /* Number of entries so far */
2399    int  bracount;                /* Count of capturing parens */    int  name_entry_size;             /* Size of each entry */
2400    int  top_backref;             /* Maximum back reference */    int  workspace_size;              /* Size of workspace */
2401    unsigned int backref_map;     /* Bitmap of low back refs */    unsigned int  bracount;           /* Count of capturing parens as we compile */
2402    int  external_options;        /* External (initial) options */    int  final_bracount;              /* Saved value after first pass */
2403    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  max_lookbehind;              /* Maximum lookbehind (characters) */
2404    BOOL nopartial;               /* Set TRUE if partial won't work */    int  top_backref;                 /* Maximum back reference */
2405    int  nltype;                  /* Newline type */    unsigned int backref_map;         /* Bitmap of low back refs */
2406    int  nllen;                   /* Newline string length */    int  assert_depth;                /* Depth of nested assertions */
2407    uschar nl[4];                 /* Newline string when fixed length */    int  external_options;            /* External (initial) options */
2408      int  external_flags;              /* External flag bits to be set */
2409      int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2410      BOOL had_accept;                  /* (*ACCEPT) encountered */
2411      BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2412      BOOL check_lookbehind;            /* Lookbehinds need later checking */
2413      int  nltype;                      /* Newline type */
2414      int  nllen;                       /* Newline string length */
2415      pcre_uchar nl[4];                 /* Newline string when fixed length */
2416  } compile_data;  } compile_data;
2417    
2418  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
2419  branches, for testing for left recursion. */  branches, for testing for left recursion while compiling. */
2420    
2421  typedef struct branch_chain {  typedef struct branch_chain {
2422    struct branch_chain *outer;    struct branch_chain *outer;
2423    uschar *current;    pcre_uchar *current_branch;
2424  } branch_chain;  } branch_chain;
2425    
2426  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
2427  call within the pattern. */  call within the pattern; used by pcre_exec(). */
2428    
2429  typedef struct recursion_info {  typedef struct recursion_info {
2430    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */    struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
2431    int group_num;                /* Number of group that was called */    unsigned int group_num;         /* Number of group that was called */
2432    const uschar *after_call;     /* "Return value": points after the call in the expr */    int *offset_save;               /* Pointer to start of saved offsets */
2433    USPTR save_start;             /* Old value of md->start_match */    int saved_max;                  /* Number of saved offsets */
2434    int *offset_save;             /* Pointer to start of saved offsets */    int saved_capture_last;         /* Last capture number */
2435    int saved_max;                /* Number of saved offsets */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
2436  } recursion_info;  } recursion_info;
2437    
2438  /* When compiling in a mode that doesn't use recursive calls to match(),  /* A similar structure for pcre_dfa_exec(). */
 a structure is used to remember local variables on the heap. It is defined in  
 pcre_exec.c, close to the match() function, so that it is easy to keep it in  
 step with any changes of local variable. However, the pointer to the current  
 frame must be saved in some "static" place over a longjmp(). We declare the  
 structure here so that we can put a pointer in the match_data structure. NOTE:  
 This isn't used for a "normal" compilation of pcre. */  
2439    
2440  struct heapframe;  typedef struct dfa_recursion_info {
2441      struct dfa_recursion_info *prevrec;
2442      int group_num;
2443      PCRE_PUCHAR subject_position;
2444    } dfa_recursion_info;
2445    
2446  /* Structure for building a chain of data for holding the values of the subject  /* Structure for building a chain of data for holding the values of the subject
2447  pointer at the start of each subpattern, so as to detect when an empty string  pointer at the start of each subpattern, so as to detect when an empty string
2448  has been matched by a subpattern - to break infinite loops. */  has been matched by a subpattern - to break infinite loops; used by
2449    pcre_exec(). */
2450    
2451  typedef struct eptrblock {  typedef struct eptrblock {
2452    struct eptrblock *epb_prev;    struct eptrblock *epb_prev;
2453    USPTR epb_saved_eptr;    PCRE_PUCHAR epb_saved_eptr;
2454  } eptrblock;  } eptrblock;
2455    
2456    
# Line 912  typedef struct match_data { Line 2461  typedef struct match_data {
2461    unsigned long int match_call_count;      /* As it says */    unsigned long int match_call_count;      /* As it says */
2462    unsigned long int match_limit;           /* As it says */    unsigned long int match_limit;           /* As it says */
2463    unsigned long int match_limit_recursion; /* As it says */    unsigned long int match_limit_recursion; /* As it says */
2464    int   *offset_vector;         /* Offset vector */    int   *offset_vector;           /* Offset vector */
2465    int    offset_end;            /* One past the end */    int    offset_end;              /* One past the end */
2466    int    offset_max;            /* The maximum usable for return data */    int    offset_max;              /* The maximum usable for return data */
2467    int    nltype;                /* Newline type */    int    nltype;                  /* Newline type */
2468    int    nllen;                 /* Newline string length */    int    nllen;                   /* Newline string length */
2469    uschar nl[4];                 /* Newline string when fixed */    int    name_count;              /* Number of names in name table */
2470    const uschar *lcc;            /* Points to lower casing table */    int    name_entry_size;         /* Size of entry in names table */
2471    const uschar *ctypes;         /* Points to table of type maps */    pcre_uchar *name_table;         /* Table of names */
2472    BOOL   offset_overflow;       /* Set if too many extractions */    pcre_uchar nl[4];               /* Newline string when fixed */
2473    BOOL   notbol;                /* NOTBOL flag */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2474    BOOL   noteol;                /* NOTEOL flag */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2475    BOOL   utf8;                  /* UTF8 flag */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2476    BOOL   endonly;               /* Dollar not before final \n */    BOOL   notbol;                  /* NOTBOL flag */
2477    BOOL   notempty;              /* Empty string match not wanted */    BOOL   noteol;                  /* NOTEOL flag */
2478    BOOL   partial;               /* PARTIAL flag */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2479    BOOL   hitend;                /* Hit the end of the subject at some point */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2480    const uschar *start_code;     /* For use when recursing */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2481    USPTR  start_subject;         /* Start of the subject string */    BOOL   endonly;                 /* Dollar not before final \n */
2482    USPTR  end_subject;           /* End of the subject string */    BOOL   notempty;                /* Empty string match not wanted */
2483    USPTR  start_match;           /* Start of this match attempt */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2484    USPTR  end_match_ptr;         /* Subject position at end match */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2485    int    end_offset_top;        /* Highwater mark at end of match */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2486    int    capture_last;          /* Most recent capture number */    BOOL   hasthen;                 /* Pattern contains (*THEN) */
2487    int    start_offset;          /* The start offset value */    BOOL   ignore_skip_arg;         /* For re-run when SKIP name not found */
2488    eptrblock *eptrchain;         /* Chain of eptrblocks for tail recursions */    const  pcre_uchar *start_code;  /* For use when recursing */
2489    int    eptrn;                 /* Next free eptrblock */    PCRE_PUCHAR start_subject;      /* Start of the subject string */
2490    recursion_info *recursive;    /* Linked list of recursion data */    PCRE_PUCHAR end_subject;        /* End of the subject string */
2491    void  *callout_data;          /* To pass back to callouts */    PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2492    struct heapframe *thisframe;  /* Used only when compiling for no recursion */    PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2493      PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2494      int    partial;                 /* PARTIAL options */
2495      int    end_offset_top;          /* Highwater mark at end of match */
2496      pcre_int32 capture_last;        /* Most recent capture number + overflow flag */
2497      int    start_offset;            /* The start offset value */
2498      int    match_function_type;     /* Set for certain special calls of MATCH() */
2499      eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2500      int    eptrn;                   /* Next free eptrblock */
2501      recursion_info *recursive;      /* Linked list of recursion data */
2502      void  *callout_data;            /* To pass back to callouts */
2503      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2504      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2505      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2506    #ifdef NO_RECURSE
2507      void  *match_frames_base;       /* For remembering malloc'd frames */
2508    #endif
2509  } match_data;  } match_data;
2510    
2511  /* 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
2512  functions. */  functions. */
2513    
2514  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2515    const uschar *start_code;     /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2516    const uschar *start_subject;  /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2517    const uschar *end_subject;    /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2518    const uschar *tables;         /* Character tables */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2519    int   moptions;               /* Match options */    const pcre_uint8 *tables;         /* Character tables */
2520    int   poptions;               /* Pattern options */    int   start_offset;               /* The start offset value */
2521    int    nltype;                /* Newline type */    int   moptions;                   /* Match options */
2522    int    nllen;                 /* Newline string length */    int   poptions;                   /* Pattern options */
2523    uschar nl[4];                 /* Newline string when fixed */    int   nltype;                     /* Newline type */
2524    void  *callout_data;          /* To pass back to callouts */    int   nllen;                      /* Newline string length */
2525      pcre_uchar nl[4];                 /* Newline string when fixed */
2526      void *callout_data;               /* To pass back to callouts */
2527      dfa_recursion_info *recursive;    /* Linked list of recursion data */
2528  } dfa_match_data;  } dfa_match_data;
2529    
2530  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 965  typedef struct dfa_match_data { Line 2533  typedef struct dfa_match_data {
2533  #define ctype_letter  0x02  #define ctype_letter  0x02
2534  #define ctype_digit   0x04  #define ctype_digit   0x04
2535  #define ctype_xdigit  0x08  #define ctype_xdigit  0x08
2536  #define ctype_word    0x10   /* alphameric or '_' */  #define ctype_word    0x10   /* alphanumeric or '_' */
2537  #define ctype_meta    0x80   /* regexp meta char or zero (end pattern) */  #define ctype_meta    0x80   /* regexp meta char or zero (end pattern) */
2538    
2539  /* 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 992  total length. */ Line 2560  total length. */
2560  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2561  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2562    
2563    /* Internal function and data prefixes. */
2564    
2565    #if defined COMPILE_PCRE8
2566    #ifndef PUBL
2567    #define PUBL(name) pcre_##name
2568    #endif
2569    #ifndef PRIV
2570    #define PRIV(name) _pcre_##name
2571    #endif
2572    #elif defined COMPILE_PCRE16
2573    #ifndef PUBL
2574    #define PUBL(name) pcre16_##name
2575    #endif
2576    #ifndef PRIV
2577    #define PRIV(name) _pcre16_##name
2578    #endif
2579    #elif defined COMPILE_PCRE32
2580    #ifndef PUBL
2581    #define PUBL(name) pcre32_##name
2582    #endif
2583    #ifndef PRIV
2584    #define PRIV(name) _pcre32_##name
2585    #endif
2586    #else
2587    #error Unsupported compiling mode
2588    #endif /* COMPILE_PCRE[8|16|32] */
2589    
2590  /* 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
2591  codes. */  codes. Each entry used to point directly to a name, but to reduce the number of
2592    relocations in shared libraries, it now has an offset into a single string
2593    instead. */
2594    
2595  typedef struct {  typedef struct {
2596    const char *name;    pcre_uint16 name_offset;
2597    pcre_uint16 type;    pcre_uint16 type;
2598    pcre_uint16 value;    pcre_uint16 value;
2599  } ucp_type_table;  } ucp_type_table;
# Line 1007  of the exported public functions. They h Line 2604  of the exported public functions. They h
2604  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
2605  pcre_tables.c module. */  pcre_tables.c module. */
2606    
2607  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2608  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2609  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2610  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2611    extern const int            PRIV(utf8_table3)[];
2612  extern const int    _pcre_utf8_table1_size;  extern const pcre_uint8     PRIV(utf8_table4)[];
2613    #endif /* COMPILE_PCRE8 */
2614  extern const ucp_type_table _pcre_utt[];  
2615  extern const int _pcre_utt_size;  extern const char           PRIV(utt_names)[];
2616    extern const ucp_type_table PRIV(utt)[];
2617    extern const int            PRIV(utt_size);
2618    
2619  extern const uschar _pcre_default_tables[];  extern const pcre_uint8     PRIV(OP_lengths)[];
2620    extern const pcre_uint8     PRIV(default_tables)[];
2621    
2622  extern const uschar _pcre_OP_lengths[];  extern const pcre_uint32    PRIV(hspace_list)[];
2623    extern const pcre_uint32    PRIV(vspace_list)[];
2624    
2625    
2626  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2627  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
2628  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2629    
2630  extern BOOL         _pcre_is_newline(const uschar *, const uschar *, int *,  /* String comparison functions. */
2631                        BOOL);  #if defined COMPILE_PCRE8
2632  extern int          _pcre_ord2utf8(int, uschar *);  
2633  extern real_pcre   *_pcre_try_flipped(const real_pcre *, real_pcre *,  #define STRCMP_UC_UC(str1, str2) \
2634                        const pcre_study_data *, pcre_study_data *);    strcmp((char *)(str1), (char *)(str2))
2635  extern int          _pcre_ucp_findprop(const unsigned int, int *, int *);  #define STRCMP_UC_C8(str1, str2) \
2636  extern unsigned int _pcre_ucp_othercase(const unsigned int);    strcmp((char *)(str1), (str2))
2637  extern int          _pcre_valid_utf8(const uschar *, int);  #define STRNCMP_UC_UC(str1, str2, num) \
2638  extern BOOL         _pcre_was_newline(const uschar *, const uschar *, int *,    strncmp((char *)(str1), (char *)(str2), (num))
2639                        BOOL);  #define STRNCMP_UC_C8(str1, str2, num) \
2640  extern BOOL         _pcre_xclass(int, const uschar *);    strncmp((char *)(str1), (str2), (num))
2641    #define STRLEN_UC(str) strlen((const char *)str)
2642    
2643    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2644    
2645    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2646                               const pcre_uchar *);
2647    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2648                               const char *);
2649    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2650                               const pcre_uchar *, unsigned int num);
2651    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2652                               const char *, unsigned int num);
2653    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2654    
2655    #define STRCMP_UC_UC(str1, str2) \
2656      PRIV(strcmp_uc_uc)((str1), (str2))
2657    #define STRCMP_UC_C8(str1, str2) \
2658      PRIV(strcmp_uc_c8)((str1), (str2))
2659    #define STRNCMP_UC_UC(str1, str2, num) \
2660      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2661    #define STRNCMP_UC_C8(str1, str2, num) \
2662      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2663    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2664    
2665    #endif /* COMPILE_PCRE[8|16|32] */
2666    
2667    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
2668    
2669    #define STRCMP_UC_UC_TEST(str1, str2) STRCMP_UC_UC(str1, str2)
2670    #define STRCMP_UC_C8_TEST(str1, str2) STRCMP_UC_C8(str1, str2)
2671    
2672    #elif defined COMPILE_PCRE32
2673    
2674    extern int               PRIV(strcmp_uc_uc_utf)(const pcre_uchar *,
2675                               const pcre_uchar *);
2676    extern int               PRIV(strcmp_uc_c8_utf)(const pcre_uchar *,
2677                               const char *);
2678    
2679    #define STRCMP_UC_UC_TEST(str1, str2) \
2680      (utf ? PRIV(strcmp_uc_uc_utf)((str1), (str2)) : PRIV(strcmp_uc_uc)((str1), (str2)))
2681    #define STRCMP_UC_C8_TEST(str1, str2) \
2682      (utf ? PRIV(strcmp_uc_c8_utf)((str1), (str2)) : PRIV(strcmp_uc_c8)((str1), (str2)))
2683    
2684    #endif /* COMPILE_PCRE[8|16|32] */
2685    
2686    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2687    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2688                               int *, BOOL);
2689    extern unsigned int      PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2690    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2691    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2692                               int *, BOOL);
2693    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2694    
2695    #ifdef SUPPORT_JIT
2696    extern void              PRIV(jit_compile)(const REAL_PCRE *,
2697                               PUBL(extra) *, int);
2698    extern int               PRIV(jit_exec)(const PUBL(extra) *,
2699                               const pcre_uchar *, int, int, int, int *, int);
2700    extern void              PRIV(jit_free)(void *);
2701    extern int               PRIV(jit_get_size)(void *);
2702    extern const char*       PRIV(jit_get_target)(void);
2703    #endif
2704    
2705    /* Unicode character database (UCD) */
2706    
2707    typedef struct {
2708      pcre_uint8 script;     /* ucp_Arabic, etc. */
2709      pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2710      pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2711      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2712      pcre_int32 other_case; /* offset to other case, or zero if none */
2713    } ucd_record;
2714    
2715    extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2716    extern const ucd_record  PRIV(ucd_records)[];
2717    extern const pcre_uint8  PRIV(ucd_stage1)[];
2718    extern const pcre_uint16 PRIV(ucd_stage2)[];
2719    extern const pcre_uint32 PRIV(ucp_gentype)[];
2720    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2721    #ifdef SUPPORT_JIT
2722    extern const int         PRIV(ucp_typerange)[];
2723    #endif
2724    
2725    #ifdef SUPPORT_UCP
2726    /* UCD access macros */
2727    
2728    #define UCD_BLOCK_SIZE 128
2729    #define GET_UCD(ch) (PRIV(ucd_records) + \
2730            PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
2731            UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
2732    
2733    #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2734    #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2735    #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2736    #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2737    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2738    #define UCD_OTHERCASE(ch)   ((pcre_uint32)((int)ch + (int)(GET_UCD(ch)->other_case)))
2739    
2740    #endif /* SUPPORT_UCP */
2741    
2742  #endif  #endif
2743    

Legend:
Removed from v.96  
changed lines
  Added in v.1248

  ViewVC Help
Powered by ViewVC 1.1.5