/[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 758 by ph10, Mon Nov 21 12:05:36 2011 UTC revision 1057 by chpe, Tue Oct 16 15:53:45 2012 UTC
# Line 7  Line 7 
7  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2011 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
# Line 51  functions whose names all begin with "_p Line 52  functions whose names all begin with "_p
52  #define PCRE_DEBUG  #define PCRE_DEBUG
53  #endif  #endif
54    
55  /* We do not support both EBCDIC and UTF-8 at the same time. The "configure"  /* PCRE is compiled as an 8 bit library if it is not requested otherwise. */
56  script prevents both being selected, but not everybody uses "configure". */  #if !defined COMPILE_PCRE16 && ! defined COMPILE_PCRE32
57    #define COMPILE_PCRE8
 #if defined EBCDIC && defined SUPPORT_UTF8  
 #error The use of both EBCDIC and SUPPORT_UTF8 is not supported.  
58  #endif  #endif
59    
60  /* If SUPPORT_UCP is defined, SUPPORT_UTF8 must also be defined. The  /* If SUPPORT_UCP is defined, SUPPORT_UTF must also be defined. The
61  "configure" script ensures this, but not everybody uses "configure". */  "configure" script ensures this, but not everybody uses "configure". */
62    
63  #if defined SUPPORT_UCP && !defined SUPPORT_UTF8  #if defined SUPPORT_UCP && !(defined SUPPORT_UTF)
64    #define SUPPORT_UTF 1
65    #endif
66    
67    /* We define SUPPORT_UTF if SUPPORT_UTF8 is enabled for compatibility
68    reasons with existing code. */
69    
70    #if defined SUPPORT_UTF8 && !(defined SUPPORT_UTF)
71    #define SUPPORT_UTF 1
72    #endif
73    
74    /* Fixme: SUPPORT_UTF8 should be eventually disappear from the code.
75    Until then we define it if SUPPORT_UTF is defined. */
76    
77    #if defined SUPPORT_UTF && !(defined SUPPORT_UTF8)
78  #define SUPPORT_UTF8 1  #define SUPPORT_UTF8 1
79  #endif  #endif
80    
81    /* We do not support both EBCDIC and UTF-8/16/32 at the same time. The "configure"
82    script prevents both being selected, but not everybody uses "configure". */
83    
84    #if defined EBCDIC && defined SUPPORT_UTF
85    #error The use of both EBCDIC and SUPPORT_UTF is not supported.
86    #endif
87    
88  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef  /* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
89  inline, and there are *still* stupid compilers about that don't like indented  inline, and there are *still* stupid compilers about that don't like indented
90  pre-processor statements, or at least there were when I first wrote this. After  pre-processor statements, or at least there were when I first wrote this. After
# Line 158  set, we ensure here that it has no effec Line 178  set, we ensure here that it has no effec
178  #define PCRE_CALL_CONVENTION  #define PCRE_CALL_CONVENTION
179  #endif  #endif
180    
181  /* We need to have types that specify unsigned 16-bit and 32-bit integers. We  /* We need to have types that specify unsigned 8, 16 and 32-bit integers. We
182  cannot determine these outside the compilation (e.g. by running a program as  cannot determine these outside the compilation (e.g. by running a program as
183  part of "configure") because PCRE is often cross-compiled for use on other  part of "configure") because PCRE is often cross-compiled for use on other
184  systems. Instead we make use of the maximum sizes that are available at  systems. Instead we make use of the maximum sizes that are available at
185  preprocessor time in standard C environments. */  preprocessor time in standard C environments. */
186    
187    typedef unsigned char pcre_uint8;
188    
189  #if USHRT_MAX == 65535  #if USHRT_MAX == 65535
190    typedef unsigned short pcre_uint16;    typedef unsigned short pcre_uint16;
191    typedef short pcre_int16;    typedef short pcre_int16;
# Line 206  by "configure". */ Line 228  by "configure". */
228    
229  /* All character handling must be done as unsigned characters. Otherwise there  /* All character handling must be done as unsigned characters. Otherwise there
230  are problems with top-bit-set characters and functions such as isspace().  are problems with top-bit-set characters and functions such as isspace().
231  However, we leave the interface to the outside world as char *, because that  However, we leave the interface to the outside world as char * or short *,
232  should make things easier for callers. We define a short type for unsigned char  because that should make things easier for callers. This character type is
233  to save lots of typing. I tried "uchar", but it causes problems on Digital  called pcre_uchar.
234  Unix, where it is defined in sys/types, so use "uschar" instead. */  
235    The IN_UCHARS macro multiply its argument with the byte size of the current
236    pcre_uchar type. Useful for memcpy and such operations, whose require the
237    byte size of their input/output buffers.
238    
239    The MAX_255 macro checks whether its pcre_uchar input is less than 256.
240    
241    The TABLE_GET macro is designed for accessing elements of tables whose contain
242    exactly 256 items. When the character is able to contain more than 256
243    items, some check is needed before accessing these tables.
244    */
245    
246    #if defined COMPILE_PCRE8
247    
248    typedef unsigned char pcre_uchar;
249    #define IN_UCHARS(x) (x)
250    #define MAX_255(c) 1
251    #define TABLE_GET(c, table, default) ((table)[c])
252    
253  typedef unsigned char uschar;  #elif defined COMPILE_PCRE16
254    
255    #if USHRT_MAX != 65535
256    /* This is a warning message. Change PCRE_UCHAR16 to a 16 bit data type in
257    pcre.h(.in) and disable (comment out) this message. */
258    #error Warning: PCRE_UCHAR16 is not a 16 bit data type.
259    #endif
260    
261    typedef pcre_uint16 pcre_uchar;
262    #define UCHAR_SHIFT (1)
263    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
264    #define MAX_255(c) ((c) <= 255u)
265    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
266    
267    #elif defined COMPILE_PCRE32
268    
269    typedef pcre_uint32 pcre_uchar;
270    #define UCHAR_SHIFT (2)
271    #define IN_UCHARS(x) ((x) << UCHAR_SHIFT)
272    #define MAX_255(c) ((c) <= 255u)
273    #define TABLE_GET(c, table, default) (MAX_255(c)? ((table)[c]):(default))
274    
275    /* Assert that pcre_uchar32 is a 32-bit type */
276    typedef int __assert_pcre_uchar32_size[sizeof(pcre_uchar) == 4 ? 1 : -1];
277    
278    #else
279    #error Unsupported compiling mode
280    #endif /* COMPILE_PCRE[8|16|32] */
281    
282  /* This is an unsigned int value that no character can ever have. UTF-8  /* This is an unsigned int value that no character can ever have. UTF-8
283  characters only go up to 0x7fffffff (though Unicode doesn't go beyond  characters only go up to 0x7fffffff (though Unicode doesn't go beyond
# Line 234  start/end of string field names are. */ Line 300  start/end of string field names are. */
300  #define IS_NEWLINE(p) \  #define IS_NEWLINE(p) \
301    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
302      ((p) < NLBLOCK->PSEND && \      ((p) < NLBLOCK->PSEND && \
303       _pcre_is_newline((p), NLBLOCK->nltype, NLBLOCK->PSEND, &(NLBLOCK->nllen),\       PRIV(is_newline)((p), NLBLOCK->nltype, NLBLOCK->PSEND, \
304         utf8)) \         &(NLBLOCK->nllen), utf)) \
305      : \      : \
306      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \      ((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
307       (p)[0] == NLBLOCK->nl[0] && \       (p)[0] == NLBLOCK->nl[0] && \
# Line 248  start/end of string field names are. */ Line 314  start/end of string field names are. */
314  #define WAS_NEWLINE(p) \  #define WAS_NEWLINE(p) \
315    ((NLBLOCK->nltype != NLTYPE_FIXED)? \    ((NLBLOCK->nltype != NLTYPE_FIXED)? \
316      ((p) > NLBLOCK->PSSTART && \      ((p) > NLBLOCK->PSSTART && \
317       _pcre_was_newline((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \       PRIV(was_newline)((p), NLBLOCK->nltype, NLBLOCK->PSSTART, \
318         &(NLBLOCK->nllen), utf8)) \         &(NLBLOCK->nllen), utf)) \
319      : \      : \
320      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \      ((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
321       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \       (p)[-NLBLOCK->nllen] == NLBLOCK->nl[0] && \
# Line 267  used for the external interface and appe Line 333  used for the external interface and appe
333  must begin with PCRE_. */  must begin with PCRE_. */
334    
335  #ifdef CUSTOM_SUBJECT_PTR  #ifdef CUSTOM_SUBJECT_PTR
336  #define PCRE_SPTR CUSTOM_SUBJECT_PTR  #define PCRE_PUCHAR CUSTOM_SUBJECT_PTR
 #define USPTR CUSTOM_SUBJECT_PTR  
337  #else  #else
338  #define PCRE_SPTR const char *  #define PCRE_PUCHAR const pcre_uchar *
 #define USPTR const unsigned char *  
339  #endif  #endif
340    
   
   
341  /* Include the public PCRE header and the definitions of UCP character property  /* Include the public PCRE header and the definitions of UCP character property
342  values. */  values. */
343    
# Line 343  The macros are controlled by the value o Line 405  The macros are controlled by the value o
405  the config.h file, but can be overridden by using -D on the command line. This  the config.h file, but can be overridden by using -D on the command line. This
406  is automated on Unix systems via the "configure" command. */  is automated on Unix systems via the "configure" command. */
407    
408    #if defined COMPILE_PCRE8
409    
410  #if LINK_SIZE == 2  #if LINK_SIZE == 2
411    
412  #define PUT(a,n,d)   \  #define PUT(a,n,d)   \
# Line 379  is automated on Unix systems via the "co Line 443  is automated on Unix systems via the "co
443  #define GET(a,n) \  #define GET(a,n) \
444    (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])    (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
445    
446  #define MAX_PATTERN_SIZE (1 << 30)   /* Keep it positive */  /* Keep it positive */
447    #define MAX_PATTERN_SIZE (1 << 30)
448    
449    #else
450    #error LINK_SIZE must be either 2, 3, or 4
451    #endif
452    
453    #elif defined COMPILE_PCRE16
454    
455    #if LINK_SIZE == 2
456    
457    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
458    #undef LINK_SIZE
459    #define LINK_SIZE 1
460    
461    #define PUT(a,n,d)   \
462      (a[n] = (d))
463    
464    #define GET(a,n) \
465      (a[n])
466    
467    #define MAX_PATTERN_SIZE (1 << 16)
468    
469    #elif LINK_SIZE == 3 || LINK_SIZE == 4
470    
471    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
472    #undef LINK_SIZE
473    #define LINK_SIZE 2
474    
475    #define PUT(a,n,d)   \
476      (a[n] = (d) >> 16), \
477      (a[(n)+1] = (d) & 65535)
478    
479    #define GET(a,n) \
480      (((a)[n] << 16) | (a)[(n)+1])
481    
482    /* Keep it positive */
483    #define MAX_PATTERN_SIZE (1 << 30)
484    
485  #else  #else
486  #error LINK_SIZE must be either 2, 3, or 4  #error LINK_SIZE must be either 2, 3, or 4
487  #endif  #endif
488    
489    #elif defined COMPILE_PCRE32
490    
491    /* Only supported LINK_SIZE is 4 */
492    /* Redefine LINK_SIZE as a multiple of sizeof(pcre_uchar) */
493    #undef LINK_SIZE
494    #define LINK_SIZE 1
495    
496    #define PUT(a,n,d)   \
497      (a[n] = (d))
498    
499    #define GET(a,n) \
500      (a[n])
501    
502    /* Keep it positive */
503    #define MAX_PATTERN_SIZE (1 << 30)
504    
505    #else
506    #error Unsupported compiling mode
507    #endif /* COMPILE_PCRE[8|16|32] */
508    
509  /* Convenience macro defined in terms of the others */  /* Convenience macro defined in terms of the others */
510    
# Line 396  is automated on Unix systems via the "co Line 515  is automated on Unix systems via the "co
515  offsets changes. There are used for repeat counts and for other things such as  offsets changes. There are used for repeat counts and for other things such as
516  capturing parenthesis numbers in back references. */  capturing parenthesis numbers in back references. */
517    
518    #if defined COMPILE_PCRE8
519    
520    #define IMM2_SIZE 2
521    
522  #define PUT2(a,n,d)   \  #define PUT2(a,n,d)   \
523    a[n] = (d) >> 8; \    a[n] = (d) >> 8; \
524    a[(n)+1] = (d) & 255    a[(n)+1] = (d) & 255
# Line 403  capturing parenthesis numbers in back re Line 526  capturing parenthesis numbers in back re
526  #define GET2(a,n) \  #define GET2(a,n) \
527    (((a)[n] << 8) | (a)[(n)+1])    (((a)[n] << 8) | (a)[(n)+1])
528    
529  #define PUT2INC(a,n,d)  PUT2(a,n,d), a += 2  #elif defined COMPILE_PCRE16
530    
531    #define IMM2_SIZE 1
532    
533  /* When UTF-8 encoding is being used, a character is no longer just a single  #define PUT2(a,n,d)   \
534  byte. The macros for character handling generate simple sequences when used in     a[n] = d
535  byte-mode, and more complicated ones for UTF-8 characters. GETCHARLENTEST is  
536  not used when UTF-8 is not supported, so it is not defined, and BACKCHAR should  #define GET2(a,n) \
537  never be called in byte mode. To make sure they can never even appear when     a[n]
538  UTF-8 support is omitted, we don't even define them. */  
539    #elif defined COMPILE_PCRE32
540    
541    #define IMM2_SIZE 1
542    
543  #ifndef SUPPORT_UTF8  #define PUT2(a,n,d)   \
544       a[n] = d
545    
546    #define GET2(a,n) \
547       a[n]
548    
549    #else
550    #error Unsupported compiling mode
551    #endif /* COMPILE_PCRE[8|16|32] */
552    
553    #define PUT2INC(a,n,d)  PUT2(a,n,d), a += IMM2_SIZE
554    
555    /* The maximum length of a MARK name is currently one data unit; it may be
556    changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
557    
558    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
559    #define MAX_MARK ((1u << 16) - 1)
560    #else
561    #define MAX_MARK ((1u << 8) - 1)
562    #endif
563    
564    /* When UTF encoding is being used, a character is no longer just a single
565    byte. The macros for character handling generate simple sequences when used in
566    character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
567    and other macros are not used when UTF is not supported, so they are not
568    defined. To make sure they can never even appear when UTF support is omitted,
569    we don't even define them. */
570    
571    #ifndef SUPPORT_UTF
572    
573    /* #define MAX_VALUE_FOR_SINGLE_CHAR */
574    /* #define HAS_EXTRALEN(c) */
575    /* #define GET_EXTRALEN(c) */
576    /* #define NOT_FIRSTCHAR(c) */
577  #define GETCHAR(c, eptr) c = *eptr;  #define GETCHAR(c, eptr) c = *eptr;
578  #define GETCHARTEST(c, eptr) c = *eptr;  #define GETCHARTEST(c, eptr) c = *eptr;
579  #define GETCHARINC(c, eptr) c = *eptr++;  #define GETCHARINC(c, eptr) c = *eptr++;
# Line 421  UTF-8 support is omitted, we don't even Line 581  UTF-8 support is omitted, we don't even
581  #define GETCHARLEN(c, eptr, len) c = *eptr;  #define GETCHARLEN(c, eptr, len) c = *eptr;
582  /* #define GETCHARLENTEST(c, eptr, len) */  /* #define GETCHARLENTEST(c, eptr, len) */
583  /* #define BACKCHAR(eptr) */  /* #define BACKCHAR(eptr) */
584    /* #define FORWARDCHAR(eptr) */
585    /* #define ACROSSCHAR(condition, eptr, action) */
586    
587  #else   /* SUPPORT_UTF8 */  #else   /* SUPPORT_UTF */
588    
589    #if defined COMPILE_PCRE8
590    
591  /* These macros were originally written in the form of loops that used data  /* These macros were originally written in the form of loops that used data
592  from the tables whose names start with _pcre_utf8_table. They were rewritten by  from the tables whose names start with PRIV(utf8_table). They were rewritten by
593  a user so as not to use loops, because in some environments this gives a  a user so as not to use loops, because in some environments this gives a
594  significant performance advantage, and it seems never to do any harm. */  significant performance advantage, and it seems never to do any harm. */
595    
596    /* Tells the biggest code point which can be encoded as a single character. */
597    
598    #define MAX_VALUE_FOR_SINGLE_CHAR 127
599    
600    /* Tests whether the code point needs extra characters to decode. */
601    
602    #define HAS_EXTRALEN(c) ((c) >= 0xc0)
603    
604    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
605    Otherwise it has an undefined behaviour. */
606    
607    #define GET_EXTRALEN(c) (PRIV(utf8_table4)[(c) & 0x3f])
608    
609    /* Returns TRUE, if the given character is not the first character
610    of a UTF sequence. */
611    
612    #define NOT_FIRSTCHAR(c) (((c) & 0xc0) == 0x80)
613    
614  /* Base macro to pick up the remaining bytes of a UTF-8 character, not  /* Base macro to pick up the remaining bytes of a UTF-8 character, not
615  advancing the pointer. */  advancing the pointer. */
616    
# Line 463  pointer. */ Line 645  pointer. */
645    
646  #define GETCHARTEST(c, eptr) \  #define GETCHARTEST(c, eptr) \
647    c = *eptr; \    c = *eptr; \
648    if (utf8 && c >= 0xc0) GETUTF8(c, eptr);    if (utf && c >= 0xc0) GETUTF8(c, eptr);
649    
650  /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing  /* Base macro to pick up the remaining bytes of a UTF-8 character, advancing
651  the pointer. */  the pointer. */
# Line 511  This is called when we don't know if we Line 693  This is called when we don't know if we
693    
694  #define GETCHARINCTEST(c, eptr) \  #define GETCHARINCTEST(c, eptr) \
695    c = *eptr++; \    c = *eptr++; \
696    if (utf8 && c >= 0xc0) GETUTF8INC(c, eptr);    if (utf && c >= 0xc0) GETUTF8INC(c, eptr);
697    
698  /* Base macro to pick up the remaining bytes of a UTF-8 character, not  /* Base macro to pick up the remaining bytes of a UTF-8 character, not
699  advancing the pointer, incrementing the length. */  advancing the pointer, incrementing the length. */
# Line 563  do not know if we are in UTF-8 mode. */ Line 745  do not know if we are in UTF-8 mode. */
745    
746  #define GETCHARLENTEST(c, eptr, len) \  #define GETCHARLENTEST(c, eptr, len) \
747    c = *eptr; \    c = *eptr; \
748    if (utf8 && c >= 0xc0) GETUTF8LEN(c, eptr, len);    if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
749    
750  /* If the pointer is not at the start of a character, move it back until  /* If the pointer is not at the start of a character, move it back until
751  it is. This is called only in UTF-8 mode - we don't put a test within the macro  it is. This is called only in UTF-8 mode - we don't put a test within the macro
# Line 571  because almost all calls are already wit Line 753  because almost all calls are already wit
753    
754  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--  #define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--
755    
756  #endif  /* SUPPORT_UTF8 */  /* Same as above, just in the other direction. */
757    #define FORWARDCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr++
758    
759    /* Same as above, but it allows a fully customizable form. */
760    #define ACROSSCHAR(condition, eptr, action) \
761      while((condition) && ((eptr) & 0xc0) == 0x80) action
762    
763  /* In case there is no definition of offsetof() provided - though any proper  #elif defined COMPILE_PCRE16
764  Standard C system should have one. */  
765    /* Tells the biggest code point which can be encoded as a single character. */
766    
767    #define MAX_VALUE_FOR_SINGLE_CHAR 65535
768    
769    /* Tests whether the code point needs extra characters to decode. */
770    
771    #define HAS_EXTRALEN(c) (((c) & 0xfc00) == 0xd800)
772    
773    /* Returns with the additional number of characters if IS_MULTICHAR(c) is TRUE.
774    Otherwise it has an undefined behaviour. */
775    
776    #define GET_EXTRALEN(c) 1
777    
778    /* Returns TRUE, if the given character is not the first character
779    of a UTF sequence. */
780    
781    #define NOT_FIRSTCHAR(c) (((c) & 0xfc00) == 0xdc00)
782    
783    /* Base macro to pick up the low surrogate of a UTF-16 character, not
784    advancing the pointer. */
785    
786    #define GETUTF16(c, eptr) \
787       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; }
788    
789    /* Get the next UTF-16 character, not advancing the pointer. This is called when
790    we know we are in UTF-16 mode. */
791    
792    #define GETCHAR(c, eptr) \
793      c = *eptr; \
794      if ((c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
795    
796    /* Get the next UTF-16 character, testing for UTF-16 mode, and not advancing the
797    pointer. */
798    
799    #define GETCHARTEST(c, eptr) \
800      c = *eptr; \
801      if (utf && (c & 0xfc00) == 0xd800) GETUTF16(c, eptr);
802    
803    /* Base macro to pick up the low surrogate of a UTF-16 character, advancing
804    the pointer. */
805    
806    #define GETUTF16INC(c, eptr) \
807       { c = (((c & 0x3ff) << 10) | (*eptr++ & 0x3ff)) + 0x10000; }
808    
809    /* Get the next UTF-16 character, advancing the pointer. This is called when we
810    know we are in UTF-16 mode. */
811    
812    #define GETCHARINC(c, eptr) \
813      c = *eptr++; \
814      if ((c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
815    
816    /* Get the next character, testing for UTF-16 mode, and advancing the pointer.
817    This is called when we don't know if we are in UTF-16 mode. */
818    
819    #define GETCHARINCTEST(c, eptr) \
820      c = *eptr++; \
821      if (utf && (c & 0xfc00) == 0xd800) GETUTF16INC(c, eptr);
822    
823    /* Base macro to pick up the low surrogate of a UTF-16 character, not
824    advancing the pointer, incrementing the length. */
825    
826    #define GETUTF16LEN(c, eptr, len) \
827       { c = (((c & 0x3ff) << 10) | (eptr[1] & 0x3ff)) + 0x10000; len++; }
828    
829    /* Get the next UTF-16 character, not advancing the pointer, incrementing
830    length if there is a low surrogate. This is called when we know we are in
831    UTF-16 mode. */
832    
833    #define GETCHARLEN(c, eptr, len) \
834      c = *eptr; \
835      if ((c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
836    
837    /* Get the next UTF-816character, testing for UTF-16 mode, not advancing the
838    pointer, incrementing length if there is a low surrogate. This is called when
839    we do not know if we are in UTF-16 mode. */
840    
841    #define GETCHARLENTEST(c, eptr, len) \
842      c = *eptr; \
843      if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
844    
845    /* If the pointer is not at the start of a character, move it back until
846    it is. This is called only in UTF-16 mode - we don't put a test within the
847    macro because almost all calls are already within a block of UTF-16 only
848    code. */
849    
850    #define BACKCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr--
851    
852    /* Same as above, just in the other direction. */
853    #define FORWARDCHAR(eptr) if ((*eptr & 0xfc00) == 0xdc00) eptr++
854    
855    /* Same as above, but it allows a fully customizable form. */
856    #define ACROSSCHAR(condition, eptr, action) \
857      if ((condition) && ((eptr) & 0xfc00) == 0xdc00) action
858    
859    #elif defined COMPILE_PCRE32
860    
861    /* These are unnecessary for the 32-bit library */
862    #undef MAX_VALUE_FOR_SINGLE_CHAR
863    #undef HAS_EXTRALEN
864    #undef GET_EXTRALEN
865    #undef NOT_FIRSTCHAR
866    
867    /* Get the next UTF-32 character, not advancing the pointer. This is called when
868    we know we are in UTF-32 mode. */
869    
870    #define GETCHAR(c, eptr) \
871      c = *eptr;
872    
873    /* Get the next UTF-32 character, testing for UTF-32 mode, and not advancing the
874    pointer. */
875    
876    #define GETCHARTEST(c, eptr) \
877      c = *eptr;
878    
879    /* Get the next UTF-32 character, advancing the pointer. This is called when we
880    know we are in UTF-32 mode. */
881    
882    #define GETCHARINC(c, eptr) \
883      c = *eptr++;
884    
885    /* Get the next character, testing for UTF-32 mode, and advancing the pointer.
886    This is called when we don't know if we are in UTF-32 mode. */
887    
888    #define GETCHARINCTEST(c, eptr) \
889      c = *eptr++;
890    
891    /* Get the next UTF-32 character, not advancing the pointer, not incrementing
892    length (since all UTF-32 is of length 1). This is called when we know we are in
893    UTF-32 mode. */
894    
895    #define GETCHARLEN(c, eptr, len) \
896      c = *eptr;
897    
898    /* Get the next UTF-832character, testing for UTF-32 mode, not advancing the
899    pointer, not incrementing the length (since all UTF-32 is of length 1).
900    This is called when we do not know if we are in UTF-32 mode. */
901    
902    #define GETCHARLENTEST(c, eptr, len) \
903      c = *eptr;
904    
905    /* If the pointer is not at the start of a character, move it back until
906    it is. This is called only in UTF-32 mode - we don't put a test within the
907    macro because almost all calls are already within a block of UTF-32 only
908    code. */
909    
910    #define BACKCHAR(eptr) do { } while (0)
911    
912    /* Same as above, just in the other direction. */
913    #define FORWARDCHAR(eptr) do { } while (0)
914    
915    /* Same as above, but it allows a fully customizable form. */
916    #define ACROSSCHAR(condition, eptr, action) do { } while (0)
917    
918    #else
919    #error Unsupported compiling mode
920    #endif /* COMPILE_PCRE[8|16|32] */
921    
922    #endif  /* SUPPORT_UTF */
923    
924    /* Tests for Unicode horizontal and vertical whitespace characters must check a
925    number of different values. Using a switch statement for this generates the
926    fastest code (no loop, no memory access), and there are several places in the
927    interpreter code where this happens. In order to ensure that all the case lists
928    remain in step, we use macros so that there is only one place where the lists
929    are defined.
930    
931    These values are also required as lists in pcre_compile.c when processing \h,
932    \H, \v and \V in a character class. The lists are defined in pcre_tables.c, but
933    macros that define the values are here so that all the definitions are
934    together. The lists must be in ascending character order, terminated by
935    NOTACHAR (which is 0xffffffff).
936    
937    Any changes should ensure that the various macros are kept in step with each
938    other. NOTE: The values also appear in pcre_jit_compile.c. */
939    
940    /* ------ ASCII/Unicode environments ------ */
941    
942    #ifndef EBCDIC
943    
944    #define HSPACE_LIST \
945      CHAR_HT, CHAR_SPACE, 0xa0, \
946      0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
947      0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
948      NOTACHAR
949    
950    #define HSPACE_MULTIBYTE_CASES \
951      case 0x1680:  /* OGHAM SPACE MARK */ \
952      case 0x180e:  /* MONGOLIAN VOWEL SEPARATOR */ \
953      case 0x2000:  /* EN QUAD */ \
954      case 0x2001:  /* EM QUAD */ \
955      case 0x2002:  /* EN SPACE */ \
956      case 0x2003:  /* EM SPACE */ \
957      case 0x2004:  /* THREE-PER-EM SPACE */ \
958      case 0x2005:  /* FOUR-PER-EM SPACE */ \
959      case 0x2006:  /* SIX-PER-EM SPACE */ \
960      case 0x2007:  /* FIGURE SPACE */ \
961      case 0x2008:  /* PUNCTUATION SPACE */ \
962      case 0x2009:  /* THIN SPACE */ \
963      case 0x200A:  /* HAIR SPACE */ \
964      case 0x202f:  /* NARROW NO-BREAK SPACE */ \
965      case 0x205f:  /* MEDIUM MATHEMATICAL SPACE */ \
966      case 0x3000   /* IDEOGRAPHIC SPACE */
967    
968    #define HSPACE_BYTE_CASES \
969      case CHAR_HT: \
970      case CHAR_SPACE: \
971      case 0xa0     /* NBSP */
972    
973    #define HSPACE_CASES \
974      HSPACE_BYTE_CASES: \
975      HSPACE_MULTIBYTE_CASES
976    
977    #define VSPACE_LIST \
978      CHAR_LF, CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, 0x2028, 0x2029, NOTACHAR
979    
980    #define VSPACE_MULTIBYTE_CASES \
981      case 0x2028:    /* LINE SEPARATOR */ \
982      case 0x2029     /* PARAGRAPH SEPARATOR */
983    
984    #define VSPACE_BYTE_CASES \
985      case CHAR_LF: \
986      case CHAR_VT: \
987      case CHAR_FF: \
988      case CHAR_CR: \
989      case CHAR_NEL
990    
991    #define VSPACE_CASES \
992      VSPACE_BYTE_CASES: \
993      VSPACE_MULTIBYTE_CASES
994    
995    /* ------ EBCDIC environments ------ */
996    
997    #else
998    #define HSPACE_LIST CHAR_HT, CHAR_SPACE
999    
1000    #define HSPACE_BYTE_CASES \
1001      case CHAR_HT: \
1002      case CHAR_SPACE
1003    
1004    #define HSPACE_CASES HSPACE_BYTE_CASES
1005    
1006    #ifdef EBCDIC_NL25
1007    #define VSPACE_LIST \
1008      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_NEL, CHAR_LF, NOTACHAR
1009    #else
1010    #define VSPACE_LIST \
1011      CHAR_VT, CHAR_FF, CHAR_CR, CHAR_LF, CHAR_NEL, NOTACHAR
1012    #endif
1013    
1014    #define VSPACE_BYTE_CASES \
1015      case CHAR_LF: \
1016      case CHAR_VT: \
1017      case CHAR_FF: \
1018      case CHAR_CR: \
1019      case CHAR_NEL
1020    
1021    #define VSPACE_CASES VSPACE_BYTE_CASES
1022    #endif  /* EBCDIC */
1023    
1024    /* ------ End of whitespace macros ------ */
1025    
 #ifndef offsetof  
 #define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))  
 #endif  
1026    
1027    
1028  /* Private flags containing information about the compiled regex. They used to  /* Private flags containing information about the compiled regex. They used to
# Line 588  are in a 16-bit flags word. From release Line 1031  are in a 16-bit flags word. From release
1031  the restrictions on partial matching have been lifted. It remains for backwards  the restrictions on partial matching have been lifted. It remains for backwards
1032  compatibility. */  compatibility. */
1033    
1034  #define PCRE_NOPARTIAL     0x0001  /* can't use partial with this regex */  #define PCRE_MODE8         0x0001  /* compiled in 8 bit mode */
1035  #define PCRE_FIRSTSET      0x0002  /* first_byte is set */  #define PCRE_MODE16        0x0002  /* compiled in 16 bit mode */
1036  #define PCRE_REQCHSET      0x0004  /* req_byte is set */  #define PCRE_MODE32        0x0004  /* compiled in 32 bit mode */
1037  #define PCRE_STARTLINE     0x0008  /* start after \n for multiline */  #define PCRE_FIRSTSET      0x0010  /* first_char is set */
1038  #define PCRE_JCHANGED      0x0010  /* j option used in regex */  #define PCRE_FCH_CASELESS  0x0020  /* caseless first char */
1039  #define PCRE_HASCRORLF     0x0020  /* explicit \r or \n in pattern */  #define PCRE_REQCHSET      0x0040  /* req_byte is set */
1040  #define PCRE_HASTHEN       0x0040  /* pattern contains (*THEN) */  #define PCRE_RCH_CASELESS  0x0080  /* caseless requested char */
1041    #define PCRE_STARTLINE     0x0100  /* start after \n for multiline */
1042    #define PCRE_NOPARTIAL     0x0200  /* can't use partial with this regex */
1043    #define PCRE_JCHANGED      0x0400  /* j option used in regex */
1044    #define PCRE_HASCRORLF     0x0800  /* explicit \r or \n in pattern */
1045    #define PCRE_HASTHEN       0x1000  /* pattern contains (*THEN) */
1046    
1047    #if defined COMPILE_PCRE8
1048    #define PCRE_MODE          PCRE_MODE8
1049    #elif defined COMPILE_PCRE16
1050    #define PCRE_MODE          PCRE_MODE16
1051    #elif defined COMPILE_PCRE32
1052    #define PCRE_MODE          PCRE_MODE32
1053    #endif
1054    #define PCRE_MODE_MASK     (PCRE_MODE8 | PCRE_MODE16 | PCRE_MODE32)
1055    
1056  /* Flags for the "extra" block produced by pcre_study(). */  /* Flags for the "extra" block produced by pcre_study(). */
1057    
# Line 626  time, run time, or study time, respectiv Line 1083  time, run time, or study time, respectiv
1083     PCRE_NO_START_OPTIMIZE)     PCRE_NO_START_OPTIMIZE)
1084    
1085  #define PUBLIC_STUDY_OPTIONS \  #define PUBLIC_STUDY_OPTIONS \
1086     PCRE_STUDY_JIT_COMPILE     (PCRE_STUDY_JIT_COMPILE|PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE| \
1087        PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE|PCRE_STUDY_EXTRA_NEEDED)
1088    
1089  /* Magic number to provide a small check against being handed junk. Also used  /* Magic number to provide a small check against being handed junk. */
 to detect whether a pattern was compiled on a host of different endianness. */  
1090    
1091  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */  #define MAGIC_NUMBER  0x50435245UL   /* 'PCRE' */
1092    
1093    /* This variable is used to detect a loaded regular expression
1094    in different endianness. */
1095    
1096    #define REVERSED_MAGIC_NUMBER  0x45524350UL   /* 'ERCP' */
1097    
1098  /* Negative values for the firstchar and reqchar variables */  /* Negative values for the firstchar and reqchar variables */
1099    
1100  #define REQ_UNSET (-2)  #define REQ_UNSET (-2)
# Line 643  req_byte match. */ Line 1105  req_byte match. */
1105    
1106  #define REQ_BYTE_MAX 1000  #define REQ_BYTE_MAX 1000
1107    
 /* Flags added to firstbyte or reqbyte; a "non-literal" item is either a  
 variable-length repeat, or a anything other than literal characters. */  
   
 #define REQ_CASELESS 0x0100    /* indicates caselessness */  
 #define REQ_VARY     0x0200    /* reqbyte followed non-literal item */  
   
1108  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in  /* Miscellaneous definitions. The #ifndef is to pacify compiler warnings in
1109  environments where these macros are defined elsewhere. Unfortunately, there  environments where these macros are defined elsewhere. Unfortunately, there
1110  is no way to do the same for the typedef. */  is no way to do the same for the typedef. */
# Line 677  for) in a minority area (EBCDIC platform Line 1133  for) in a minority area (EBCDIC platform
1133  application that did need both could compile two versions of the library, using  application that did need both could compile two versions of the library, using
1134  macros to give the functions distinct names. */  macros to give the functions distinct names. */
1135    
1136  #ifndef SUPPORT_UTF8  #ifndef SUPPORT_UTF
1137    
1138  /* UTF-8 support is not enabled; use the platform-dependent character literals  /* UTF-8 support is not enabled; use the platform-dependent character literals
1139  so that PCRE works on both ASCII and EBCDIC platforms, in non-UTF-mode only. */  so that PCRE works in both ASCII and EBCDIC environments, but only in non-UTF
1140    mode. Newline characters are problematic in EBCDIC. Though it has CR and LF
1141    characters, a common practice has been to use its NL (0x15) character as the
1142    line terminator in C-like processing environments. However, sometimes the LF
1143    (0x25) character is used instead, according to this Unicode document:
1144    
1145    http://unicode.org/standard/reports/tr13/tr13-5.html
1146    
1147    PCRE defaults EBCDIC NL to 0x15, but has a build-time option to select 0x25
1148    instead. Whichever is *not* chosen is defined as NEL.
1149    
1150    In both ASCII and EBCDIC environments, CHAR_NL and CHAR_LF are synonyms for the
1151    same code point. */
1152    
1153    #ifdef EBCDIC
1154    
1155    #ifndef EBCDIC_NL25
1156    #define CHAR_NL                     '\x15'
1157    #define CHAR_NEL                    '\x25'
1158    #define STR_NL                      "\x15"
1159    #define STR_NEL                     "\x25"
1160    #else
1161    #define CHAR_NL                     '\x25'
1162    #define CHAR_NEL                    '\x15'
1163    #define STR_NL                      "\x25"
1164    #define STR_NEL                     "\x15"
1165    #endif
1166    
1167    #define CHAR_LF                     CHAR_NL
1168    #define STR_LF                      STR_NL
1169    
1170    #define CHAR_ESC                    '\047'
1171    #define CHAR_DEL                    '\007'
1172    #define STR_ESC                     "\047"
1173    #define STR_DEL                     "\007"
1174    
1175    #else  /* Not EBCDIC */
1176    
1177    /* In ASCII/Unicode, linefeed is '\n' and we equate this to NL for
1178    compatibility. NEL is the Unicode newline character; make sure it is
1179    a positive value. */
1180    
1181    #define CHAR_LF                     '\n'
1182    #define CHAR_NL                     CHAR_LF
1183    #define CHAR_NEL                    ((unsigned char)'\x85')
1184    #define CHAR_ESC                    '\033'
1185    #define CHAR_DEL                    '\177'
1186    
1187    #define STR_LF                      "\n"
1188    #define STR_NL                      STR_LF
1189    #define STR_NEL                     "\x85"
1190    #define STR_ESC                     "\033"
1191    #define STR_DEL                     "\177"
1192    
1193    #endif  /* EBCDIC */
1194    
1195    /* The remaining definitions work in both environments. */
1196    
1197  #define CHAR_HT                     '\t'  #define CHAR_HT                     '\t'
1198  #define CHAR_VT                     '\v'  #define CHAR_VT                     '\v'
1199  #define CHAR_FF                     '\f'  #define CHAR_FF                     '\f'
1200  #define CHAR_CR                     '\r'  #define CHAR_CR                     '\r'
 #define CHAR_NL                     '\n'  
1201  #define CHAR_BS                     '\b'  #define CHAR_BS                     '\b'
1202  #define CHAR_BEL                    '\a'  #define CHAR_BEL                    '\a'
 #ifdef EBCDIC  
 #define CHAR_ESC                    '\047'  
 #define CHAR_DEL                    '\007'  
 #else  
 #define CHAR_ESC                    '\033'  
 #define CHAR_DEL                    '\177'  
 #endif  
1203    
1204  #define CHAR_SPACE                  ' '  #define CHAR_SPACE                  ' '
1205  #define CHAR_EXCLAMATION_MARK       '!'  #define CHAR_EXCLAMATION_MARK       '!'
# Line 797  so that PCRE works on both ASCII and EBC Line 1301  so that PCRE works on both ASCII and EBC
1301  #define STR_VT                      "\v"  #define STR_VT                      "\v"
1302  #define STR_FF                      "\f"  #define STR_FF                      "\f"
1303  #define STR_CR                      "\r"  #define STR_CR                      "\r"
 #define STR_NL                      "\n"  
1304  #define STR_BS                      "\b"  #define STR_BS                      "\b"
1305  #define STR_BEL                     "\a"  #define STR_BEL                     "\a"
 #ifdef EBCDIC  
 #define STR_ESC                     "\047"  
 #define STR_DEL                     "\007"  
 #else  
 #define STR_ESC                     "\033"  
 #define STR_DEL                     "\177"  
 #endif  
1306    
1307  #define STR_SPACE                   " "  #define STR_SPACE                   " "
1308  #define STR_EXCLAMATION_MARK        "!"  #define STR_EXCLAMATION_MARK        "!"
# Line 937  so that PCRE works on both ASCII and EBC Line 1433  so that PCRE works on both ASCII and EBC
1433  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"  #define STRING_ANYCRLF_RIGHTPAR        "ANYCRLF)"
1434  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"  #define STRING_BSR_ANYCRLF_RIGHTPAR    "BSR_ANYCRLF)"
1435  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"  #define STRING_BSR_UNICODE_RIGHTPAR    "BSR_UNICODE)"
1436  #define STRING_UTF8_RIGHTPAR           "UTF8)"  #ifdef COMPILE_PCRE8
1437    #define STRING_UTF_RIGHTPAR            "UTF8)"
1438    #endif
1439    #ifdef COMPILE_PCRE16
1440    #define STRING_UTF_RIGHTPAR            "UTF16)"
1441    #endif
1442    #ifdef COMPILE_PCRE32
1443    #define STRING_UTF_RIGHTPAR            "UTF32)"
1444    #endif
1445  #define STRING_UCP_RIGHTPAR            "UCP)"  #define STRING_UCP_RIGHTPAR            "UCP)"
1446  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"  #define STRING_NO_START_OPT_RIGHTPAR   "NO_START_OPT)"
1447    
1448  #else  /* SUPPORT_UTF8 */  #else  /* SUPPORT_UTF */
1449    
1450  /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This  /* UTF-8 support is enabled; always use UTF-8 (=ASCII) character codes. This
1451  works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode  works in both modes non-EBCDIC platforms, and on EBCDIC platforms in UTF-8 mode
# Line 951  only. */ Line 1455  only. */
1455  #define CHAR_VT                     '\013'  #define CHAR_VT                     '\013'
1456  #define CHAR_FF                     '\014'  #define CHAR_FF                     '\014'
1457  #define CHAR_CR                     '\015'  #define CHAR_CR                     '\015'
1458  #define CHAR_NL                     '\012'  #define CHAR_LF                     '\012'
1459    #define CHAR_NL                     CHAR_LF
1460    #define CHAR_NEL                    ((unsigned char)'\x85')
1461  #define CHAR_BS                     '\010'  #define CHAR_BS                     '\010'
1462  #define CHAR_BEL                    '\007'  #define CHAR_BEL                    '\007'
1463  #define CHAR_ESC                    '\033'  #define CHAR_ESC                    '\033'
# Line 1192  only. */ Line 1698  only. */
1698  #define STRING_ANYCRLF_RIGHTPAR        STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS  #define STRING_ANYCRLF_RIGHTPAR        STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1699  #define STRING_BSR_ANYCRLF_RIGHTPAR    STR_B STR_S STR_R STR_UNDERSCORE STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS  #define STRING_BSR_ANYCRLF_RIGHTPAR    STR_B STR_S STR_R STR_UNDERSCORE STR_A STR_N STR_Y STR_C STR_R STR_L STR_F STR_RIGHT_PARENTHESIS
1700  #define STRING_BSR_UNICODE_RIGHTPAR    STR_B STR_S STR_R STR_UNDERSCORE STR_U STR_N STR_I STR_C STR_O STR_D STR_E STR_RIGHT_PARENTHESIS  #define STRING_BSR_UNICODE_RIGHTPAR    STR_B STR_S STR_R STR_UNDERSCORE STR_U STR_N STR_I STR_C STR_O STR_D STR_E STR_RIGHT_PARENTHESIS
1701  #define STRING_UTF8_RIGHTPAR           STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS  #ifdef COMPILE_PCRE8
1702    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_8 STR_RIGHT_PARENTHESIS
1703    #endif
1704    #ifdef COMPILE_PCRE16
1705    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_1 STR_6 STR_RIGHT_PARENTHESIS
1706    #endif
1707    #ifdef COMPILE_PCRE32
1708    #define STRING_UTF_RIGHTPAR            STR_U STR_T STR_F STR_3 STR_2 STR_RIGHT_PARENTHESIS
1709    #endif
1710  #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS  #define STRING_UCP_RIGHTPAR            STR_U STR_C STR_P STR_RIGHT_PARENTHESIS
1711  #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  #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
1712    
1713  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
1714    
1715  /* Escape items that are just an encoding of a particular data value. */  /* Escape items that are just an encoding of a particular data value. */
1716    
# Line 1209  only. */ Line 1723  only. */
1723  #endif  #endif
1724    
1725  #ifndef ESC_n  #ifndef ESC_n
1726  #define ESC_n CHAR_NL  #define ESC_n CHAR_LF
1727  #endif  #endif
1728    
1729  #ifndef ESC_r  #ifndef ESC_r
# Line 1234  only. */ Line 1748  only. */
1748  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */  #define PT_SPACE      6    /* Perl space - Z plus 9,10,12,13 */
1749  #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */  #define PT_PXSPACE    7    /* POSIX space - Z plus 9,10,11,12,13 */
1750  #define PT_WORD       8    /* Word - L plus N plus underscore */  #define PT_WORD       8    /* Word - L plus N plus underscore */
1751    #define PT_CLIST      9    /* Pseudo-property: match character list */
1752    
1753  /* 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
1754  contain UTF-8 characters with values greater than 255. */  contain characters with values greater than 255. */
1755    
1756  #define XCL_NOT    0x01    /* Flag: this is a negative class */  #define XCL_NOT    0x01    /* Flag: this is a negative class */
1757  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */  #define XCL_MAP    0x02    /* Flag: a 32-byte map is present */
# Line 1252  value such as \n. They must have non-zer Line 1767  value such as \n. They must have non-zer
1767  their negation. Also, they must appear in the same order as in the opcode  their negation. Also, they must appear in the same order as in the opcode
1768  definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it  definitions below, up to ESC_z. There's a dummy for OP_ALLANY because it
1769  corresponds to "." in DOTALL mode rather than an escape sequence. It is also  corresponds to "." in DOTALL mode rather than an escape sequence. It is also
1770  used for [^] in JavaScript compatibility mode, and for \C in non-utf8 mode. In  used for [^] in JavaScript compatibility mode, and for \C in non-utf mode. In
1771  non-DOTALL mode, "." behaves like \N.  non-DOTALL mode, "." behaves like \N.
1772    
1773  The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.  The special values ESC_DU, ESC_du, etc. are used instead of ESC_D, ESC_d, etc.
1774  when PCRE_UCP is set, when replacement of \d etc by \p sequences is required.  when PCRE_UCP is set and replacement of \d etc by \p sequences is required.
1775  They must be contiguous, and remain in order so that the replacements can be  They must be contiguous, and remain in order so that the replacements can be
1776  looked up from a table.  looked up from a table.
1777    
# Line 1298  enum { Line 1813  enum {
1813    OP_NOT_WORDCHAR,       /* 10 \W */    OP_NOT_WORDCHAR,       /* 10 \W */
1814    OP_WORDCHAR,           /* 11 \w */    OP_WORDCHAR,           /* 11 \w */
1815    
1816    OP_ANY,            /* 12 Match any character except newline */    OP_ANY,            /* 12 Match any character except newline (\N) */
1817    OP_ALLANY,         /* 13 Match any character */    OP_ALLANY,         /* 13 Match any character */
1818    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */    OP_ANYBYTE,        /* 14 Match any byte (\C); different to OP_ANY for UTF-8 */
1819    OP_NOTPROP,        /* 15 \P (not Unicode property) */    OP_NOTPROP,        /* 15 \P (not Unicode property) */
# Line 1309  enum { Line 1824  enum {
1824    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */    OP_NOT_VSPACE,     /* 20 \V (not vertical whitespace) */
1825    OP_VSPACE,         /* 21 \v (vertical whitespace) */    OP_VSPACE,         /* 21 \v (vertical whitespace) */
1826    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */    OP_EXTUNI,         /* 22 \X (extended Unicode sequence */
1827    OP_EODN,           /* 23 End of data or \n at end of data: \Z. */    OP_EODN,           /* 23 End of data or \n at end of data (\Z) */
1828    OP_EOD,            /* 24 End of data: \z */    OP_EOD,            /* 24 End of data (\z) */
1829    
1830    OP_CIRC,           /* 25 Start of line - not multiline */    OP_CIRC,           /* 25 Start of line - not multiline */
1831    OP_CIRCM,          /* 26 Start of line - multiline */    OP_CIRCM,          /* 26 Start of line - multiline */
# Line 1433  enum { Line 1948  enum {
1948    OP_CLASS,          /* 106 Match a character class, chars < 256 only */    OP_CLASS,          /* 106 Match a character class, chars < 256 only */
1949    OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative    OP_NCLASS,         /* 107 Same, but the bitmap was created from a negative
1950                                class - the difference is relevant only when a                                class - the difference is relevant only when a
1951                                UTF-8 character > 255 is encountered. */                                character > 255 is encountered. */
1952    OP_XCLASS,         /* 108 Extended class for handling UTF-8 chars within the    OP_XCLASS,         /* 108 Extended class for handling > 255 chars within the
1953                                class. This does both positive and negative. */                                class. This does both positive and negative. */
1954    OP_REF,            /* 109 Match a back reference, casefully */    OP_REF,            /* 109 Match a back reference, casefully */
1955    OP_REFI,           /* 110 Match a back reference, caselessly */    OP_REFI,           /* 110 Match a back reference, caselessly */
# Line 1591  in UTF-8 mode. The code that uses this t Line 2106  in UTF-8 mode. The code that uses this t
2106    2,                             /* noti                                   */ \    2,                             /* noti                                   */ \
2107    /* Positive single-char repeats                             ** These are */ \    /* Positive single-char repeats                             ** These are */ \
2108    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \    2, 2, 2, 2, 2, 2,              /* *, *?, +, +?, ?, ??       ** minima in */ \
2109    4, 4, 4,                       /* upto, minupto, exact      ** mode      */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto, minupto             ** mode      */ \
2110    2, 2, 2, 4,                    /* *+, ++, ?+, upto+                      */ \    2+IMM2_SIZE,                   /* exact                                  */ \
2111      2, 2, 2, 2+IMM2_SIZE,          /* *+, ++, ?+, upto+                      */ \
2112    2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \    2, 2, 2, 2, 2, 2,              /* *I, *?I, +I, +?I, ?I, ??I ** UTF-8     */ \
2113    4, 4, 4,                       /* upto I, minupto I, exact I             */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* upto I, minupto I                      */ \
2114    2, 2, 2, 4,                    /* *+I, ++I, ?+I, upto+I                  */ \    2+IMM2_SIZE,                   /* exact I                                */ \
2115      2, 2, 2, 2+IMM2_SIZE,          /* *+I, ++I, ?+I, upto+I                  */ \
2116    /* Negative single-char repeats - only for chars < 256                   */ \    /* Negative single-char repeats - only for chars < 256                   */ \
2117    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \    2, 2, 2, 2, 2, 2,              /* NOT *, *?, +, +?, ?, ??                */ \
2118    4, 4, 4,                       /* NOT upto, minupto, exact               */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto, minupto                      */ \
2119    2, 2, 2, 4,                    /* Possessive NOT *, +, ?, upto           */ \    2+IMM2_SIZE,                   /* NOT exact                              */ \
2120      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *, +, ?, upto           */ \
2121    2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \    2, 2, 2, 2, 2, 2,              /* NOT *I, *?I, +I, +?I, ?I, ??I          */ \
2122    4, 4, 4,                       /* NOT upto I, minupto I, exact I         */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* NOT upto I, minupto I                  */ \
2123    2, 2, 2, 4,                    /* Possessive NOT *I, +I, ?I, upto I      */ \    2+IMM2_SIZE,                   /* NOT exact I                            */ \
2124      2, 2, 2, 2+IMM2_SIZE,          /* Possessive NOT *I, +I, ?I, upto I      */ \
2125    /* Positive type repeats                                                 */ \    /* Positive type repeats                                                 */ \
2126    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \    2, 2, 2, 2, 2, 2,              /* Type *, *?, +, +?, ?, ??               */ \
2127    4, 4, 4,                       /* Type upto, minupto, exact              */ \    2+IMM2_SIZE, 2+IMM2_SIZE,      /* Type upto, minupto                     */ \
2128    2, 2, 2, 4,                    /* Possessive *+, ++, ?+, upto+           */ \    2+IMM2_SIZE,                   /* Type exact                             */ \
2129      2, 2, 2, 2+IMM2_SIZE,          /* Possessive *+, ++, ?+, upto+           */ \
2130    /* Character class & ref repeats                                         */ \    /* Character class & ref repeats                                         */ \
2131    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \    1, 1, 1, 1, 1, 1,              /* *, *?, +, +?, ?, ??                    */ \
2132    5, 5,                          /* CRRANGE, CRMINRANGE                    */ \    1+2*IMM2_SIZE, 1+2*IMM2_SIZE,  /* CRRANGE, CRMINRANGE                    */ \
2133   33,                             /* CLASS                                  */ \    1+(32/sizeof(pcre_uchar)),     /* CLASS                                  */ \
2134   33,                             /* NCLASS                                 */ \    1+(32/sizeof(pcre_uchar)),     /* NCLASS                                 */ \
2135    0,                             /* XCLASS - variable length               */ \    0,                             /* XCLASS - variable length               */ \
2136    3,                             /* REF                                    */ \    1+IMM2_SIZE,                   /* REF                                    */ \
2137    3,                             /* REFI                                   */ \    1+IMM2_SIZE,                   /* REFI                                   */ \
2138    1+LINK_SIZE,                   /* RECURSE                                */ \    1+LINK_SIZE,                   /* RECURSE                                */ \
2139    2+2*LINK_SIZE,                 /* CALLOUT                                */ \    2+2*LINK_SIZE,                 /* CALLOUT                                */ \
2140    1+LINK_SIZE,                   /* Alt                                    */ \    1+LINK_SIZE,                   /* Alt                                    */ \
# Line 1631  in UTF-8 mode. The code that uses this t Line 2151  in UTF-8 mode. The code that uses this t
2151    1+LINK_SIZE,                   /* ONCE_NC                                */ \    1+LINK_SIZE,                   /* ONCE_NC                                */ \
2152    1+LINK_SIZE,                   /* BRA                                    */ \    1+LINK_SIZE,                   /* BRA                                    */ \
2153    1+LINK_SIZE,                   /* BRAPOS                                 */ \    1+LINK_SIZE,                   /* BRAPOS                                 */ \
2154    3+LINK_SIZE,                   /* CBRA                                   */ \    1+LINK_SIZE+IMM2_SIZE,         /* CBRA                                   */ \
2155    3+LINK_SIZE,                   /* CBRAPOS                                */ \    1+LINK_SIZE+IMM2_SIZE,         /* CBRAPOS                                */ \
2156    1+LINK_SIZE,                   /* COND                                   */ \    1+LINK_SIZE,                   /* COND                                   */ \
2157    1+LINK_SIZE,                   /* SBRA                                   */ \    1+LINK_SIZE,                   /* SBRA                                   */ \
2158    1+LINK_SIZE,                   /* SBRAPOS                                */ \    1+LINK_SIZE,                   /* SBRAPOS                                */ \
2159    3+LINK_SIZE,                   /* SCBRA                                  */ \    1+LINK_SIZE+IMM2_SIZE,         /* SCBRA                                  */ \
2160    3+LINK_SIZE,                   /* SCBRAPOS                               */ \    1+LINK_SIZE+IMM2_SIZE,         /* SCBRAPOS                               */ \
2161    1+LINK_SIZE,                   /* SCOND                                  */ \    1+LINK_SIZE,                   /* SCOND                                  */ \
2162    3, 3,                          /* CREF, NCREF                            */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* CREF, NCREF                            */ \
2163    3, 3,                          /* RREF, NRREF                            */ \    1+IMM2_SIZE, 1+IMM2_SIZE,      /* RREF, NRREF                            */ \
2164    1,                             /* DEF                                    */ \    1,                             /* DEF                                    */ \
2165    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \    1, 1, 1,                       /* BRAZERO, BRAMINZERO, BRAPOSZERO        */ \
2166    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \    3, 1, 3,                       /* MARK, PRUNE, PRUNE_ARG                 */ \
2167    1, 3,                          /* SKIP, SKIP_ARG                         */ \    1, 3,                          /* SKIP, SKIP_ARG                         */ \
2168    1, 3,                          /* THEN, THEN_ARG                         */ \    1, 3,                          /* THEN, THEN_ARG                         */ \
2169    1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \    1, 1, 1, 1,                    /* COMMIT, FAIL, ACCEPT, ASSERT_ACCEPT    */ \
2170    3, 1                           /* CLOSE, SKIPZERO  */    1+IMM2_SIZE, 1                 /* CLOSE, SKIPZERO                        */
2171    
2172  /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"  /* A magic value for OP_RREF and OP_NRREF to indicate the "any recursion"
2173  condition. */  condition. */
# Line 1665  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4, Line 2185  enum { ERR0,  ERR1,  ERR2,  ERR3,  ERR4,
2185         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,         ERR40, ERR41, ERR42, ERR43, ERR44, ERR45, ERR46, ERR47, ERR48, ERR49,
2186         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,         ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
2187         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,         ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
2188         ERR70, ERR71, ERRCOUNT };         ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERRCOUNT };
2189    
2190    /* JIT compiling modes. The function list is indexed by them. */
2191    enum { JIT_COMPILE, JIT_PARTIAL_SOFT_COMPILE, JIT_PARTIAL_HARD_COMPILE,
2192           JIT_NUMBER_OF_COMPILE_MODES };
2193    
2194  /* 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
2195  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 1684  fields are present. Currently PCRE alway Line 2208  fields are present. Currently PCRE alway
2208  NOTE NOTE NOTE  NOTE NOTE NOTE
2209  */  */
2210    
2211  typedef struct real_pcre {  #if defined COMPILE_PCRE8
2212    #define REAL_PCRE real_pcre
2213    #elif defined COMPILE_PCRE16
2214    #define REAL_PCRE real_pcre16
2215    #elif defined COMPILE_PCRE32
2216    #define REAL_PCRE real_pcre32
2217    #endif
2218    
2219    /* It is necessary to fork the struct for 32 bit, since it needs to use
2220     * pcre_uchar for first_char and req_char. Can't put an ifdef inside the
2221     * typedef since pcretest needs access to  the struct of the 8-, 16-
2222     * and 32-bit variants. */
2223    
2224    typedef struct real_pcre8_or_16 {
2225    pcre_uint32 magic_number;    pcre_uint32 magic_number;
2226    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2227    pcre_uint32 options;            /* Public options */    pcre_uint32 options;            /* Public options */
2228    pcre_uint16 flags;              /* Private flags */    pcre_uint16 flags;              /* Private flags */
2229    pcre_uint16 dummy1;             /* For future use */    pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2230    pcre_uint16 top_bracket;    pcre_uint16 top_bracket;        /* Highest numbered group */
2231    pcre_uint16 top_backref;    pcre_uint16 top_backref;        /* Highest numbered back reference */
2232    pcre_uint16 first_byte;    pcre_uint16 first_char;         /* Starting character */
2233    pcre_uint16 req_byte;    pcre_uint16 req_char;           /* This character must be seen */
2234    pcre_uint16 name_table_offset;  /* Offset to name table that follows */    pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2235    pcre_uint16 name_entry_size;    /* Size of any name items */    pcre_uint16 name_entry_size;    /* Size of any name items */
2236    pcre_uint16 name_count;         /* Number of name items */    pcre_uint16 name_count;         /* Number of name items */
2237    pcre_uint16 ref_count;          /* Reference count */    pcre_uint16 ref_count;          /* Reference count */
2238      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2239      const pcre_uint8 *nullpad;      /* NULL padding */
2240    } real_pcre8_or_16;
2241    
2242    typedef struct real_pcre8_or_16 real_pcre;
2243    typedef struct real_pcre8_or_16 real_pcre16;
2244    
2245    const unsigned char *tables;    /* Pointer to tables or NULL for std */  typedef struct real_pcre32 {
2246    const unsigned char *nullpad;   /* NULL padding */    pcre_uint32 magic_number;
2247  } real_pcre;    pcre_uint32 size;               /* Total that was malloced */
2248      pcre_uint32 options;            /* Public options */
2249      pcre_uint16 flags;              /* Private flags */
2250      pcre_uint16 max_lookbehind;     /* Longest lookbehind (characters) */
2251      pcre_uint16 top_bracket;        /* Highest numbered group */
2252      pcre_uint16 top_backref;        /* Highest numbered back reference */
2253      pcre_uint32 first_char;         /* Starting character */
2254      pcre_uint32 req_char;           /* This character must be seen */
2255      pcre_uint16 name_table_offset;  /* Offset to name table that follows */
2256      pcre_uint16 name_entry_size;    /* Size of any name items */
2257      pcre_uint16 name_count;         /* Number of name items */
2258      pcre_uint16 ref_count;          /* Reference count */
2259      pcre_uint16 dummy1;             /* for later expansion */
2260      pcre_uint16 dummy2;             /* for later expansion */
2261      const pcre_uint8 *tables;       /* Pointer to tables or NULL for std */
2262      void *nullpad;                  /* for later expansion */
2263    } real_pcre32;
2264    
2265    /* Assert that the size of REAL_PCRE is divisible by 8 */
2266    typedef int __assert_real_pcre_size_divisible_8[(sizeof(REAL_PCRE) % 8) == 0 ? 1 : -1];
2267    
2268    /* Needed in pcretest to access some fields in the real_pcre* structures
2269     * directly. They're unified for 8/16/32 bits since the structs only differ
2270     * after these fields; if that ever changes, need to fork those defines into
2271     * 8/16 and 32 bit versions. */
2272    #define REAL_PCRE_MAGIC(re)     (((REAL_PCRE*)re)->magic_number)
2273    #define REAL_PCRE_SIZE(re)      (((REAL_PCRE*)re)->size)
2274    #define REAL_PCRE_OPTIONS(re)   (((REAL_PCRE*)re)->options)
2275    #define REAL_PCRE_FLAGS(re)     (((REAL_PCRE*)re)->flags)
2276    
2277  /* 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
2278  remark (see NOTE above) about extending this structure applies. */  remark (see NOTE above) about extending this structure applies. */
# Line 1709  remark (see NOTE above) about extending Line 2280  remark (see NOTE above) about extending
2280  typedef struct pcre_study_data {  typedef struct pcre_study_data {
2281    pcre_uint32 size;               /* Total that was malloced */    pcre_uint32 size;               /* Total that was malloced */
2282    pcre_uint32 flags;              /* Private flags */    pcre_uint32 flags;              /* Private flags */
2283    uschar start_bits[32];          /* Starting char bits */    pcre_uint8 start_bits[32];      /* Starting char bits */
2284    pcre_uint32 minlength;          /* Minimum subject length */    pcre_uint32 minlength;          /* Minimum subject length */
2285  } pcre_study_data;  } pcre_study_data;
2286    
# Line 1728  typedef struct open_capitem { Line 2299  typedef struct open_capitem {
2299  doing the compiling, so that they are thread-safe. */  doing the compiling, so that they are thread-safe. */
2300    
2301  typedef struct compile_data {  typedef struct compile_data {
2302    const uschar *lcc;            /* Points to lower casing table */    const pcre_uint8 *lcc;            /* Points to lower casing table */
2303    const uschar *fcc;            /* Points to case-flipping table */    const pcre_uint8 *fcc;            /* Points to case-flipping table */
2304    const uschar *cbits;          /* Points to character type table */    const pcre_uint8 *cbits;          /* Points to character type table */
2305    const uschar *ctypes;         /* Points to table of type maps */    const pcre_uint8 *ctypes;         /* Points to table of type maps */
2306    const uschar *start_workspace;/* The start of working space */    const pcre_uchar *start_workspace;/* The start of working space */
2307    const uschar *start_code;     /* The start of the compiled code */    const pcre_uchar *start_code;     /* The start of the compiled code */
2308    const uschar *start_pattern;  /* The start of the pattern */    const pcre_uchar *start_pattern;  /* The start of the pattern */
2309    const uschar *end_pattern;    /* The end of the pattern */    const pcre_uchar *end_pattern;    /* The end of the pattern */
2310    open_capitem *open_caps;      /* Chain of open capture items */    open_capitem *open_caps;          /* Chain of open capture items */
2311    uschar *hwm;                  /* High watermark of workspace */    pcre_uchar *hwm;                  /* High watermark of workspace */
2312    uschar *name_table;           /* The name/number table */    pcre_uchar *name_table;           /* The name/number table */
2313    int  names_found;             /* Number of entries so far */    int  names_found;                 /* Number of entries so far */
2314    int  name_entry_size;         /* Size of each entry */    int  name_entry_size;             /* Size of each entry */
2315    int  bracount;                /* Count of capturing parens as we compile */    int  workspace_size;              /* Size of workspace */
2316    int  final_bracount;          /* Saved value after first pass */    int  bracount;                    /* Count of capturing parens as we compile */
2317    int  top_backref;             /* Maximum back reference */    int  final_bracount;              /* Saved value after first pass */
2318    unsigned int backref_map;     /* Bitmap of low back refs */    int  max_lookbehind;              /* Maximum lookbehind (characters) */
2319    int  assert_depth;            /* Depth of nested assertions */    int  top_backref;                 /* Maximum back reference */
2320    int  external_options;        /* External (initial) options */    unsigned int backref_map;         /* Bitmap of low back refs */
2321    int  external_flags;          /* External flag bits to be set */    int  assert_depth;                /* Depth of nested assertions */
2322    int  req_varyopt;             /* "After variable item" flag for reqbyte */    int  external_options;            /* External (initial) options */
2323    BOOL had_accept;              /* (*ACCEPT) encountered */    int  external_flags;              /* External flag bits to be set */
2324    BOOL check_lookbehind;        /* Lookbehinds need later checking */    int  req_varyopt;                 /* "After variable item" flag for reqbyte */
2325    int  nltype;                  /* Newline type */    BOOL had_accept;                  /* (*ACCEPT) encountered */
2326    int  nllen;                   /* Newline string length */    BOOL had_pruneorskip;             /* (*PRUNE) or (*SKIP) encountered */
2327    uschar nl[4];                 /* Newline string when fixed length */    BOOL check_lookbehind;            /* Lookbehinds need later checking */
2328      int  nltype;                      /* Newline type */
2329      int  nllen;                       /* Newline string length */
2330      pcre_uchar nl[4];                 /* Newline string when fixed length */
2331  } compile_data;  } compile_data;
2332    
2333  /* Structure for maintaining a chain of pointers to the currently incomplete  /* Structure for maintaining a chain of pointers to the currently incomplete
# Line 1761  branches, for testing for left recursion Line 2335  branches, for testing for left recursion
2335    
2336  typedef struct branch_chain {  typedef struct branch_chain {
2337    struct branch_chain *outer;    struct branch_chain *outer;
2338    uschar *current_branch;    pcre_uchar *current_branch;
2339  } branch_chain;  } branch_chain;
2340    
2341  /* Structure for items in a linked list that represents an explicit recursive  /* Structure for items in a linked list that represents an explicit recursive
# Line 1772  typedef struct recursion_info { Line 2346  typedef struct recursion_info {
2346    int group_num;                  /* Number of group that was called */    int group_num;                  /* Number of group that was called */
2347    int *offset_save;               /* Pointer to start of saved offsets */    int *offset_save;               /* Pointer to start of saved offsets */
2348    int saved_max;                  /* Number of saved offsets */    int saved_max;                  /* Number of saved offsets */
2349    USPTR subject_position;         /* Position at start of recursion */    PCRE_PUCHAR subject_position;   /* Position at start of recursion */
2350  } recursion_info;  } recursion_info;
2351    
2352  /* A similar structure for pcre_dfa_exec(). */  /* A similar structure for pcre_dfa_exec(). */
# Line 1780  typedef struct recursion_info { Line 2354  typedef struct recursion_info {
2354  typedef struct dfa_recursion_info {  typedef struct dfa_recursion_info {
2355    struct dfa_recursion_info *prevrec;    struct dfa_recursion_info *prevrec;
2356    int group_num;    int group_num;
2357    USPTR subject_position;    PCRE_PUCHAR subject_position;
2358  } dfa_recursion_info;  } dfa_recursion_info;
2359    
2360  /* 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
# Line 1790  pcre_exec(). */ Line 2364  pcre_exec(). */
2364    
2365  typedef struct eptrblock {  typedef struct eptrblock {
2366    struct eptrblock *epb_prev;    struct eptrblock *epb_prev;
2367    USPTR epb_saved_eptr;    PCRE_PUCHAR epb_saved_eptr;
2368  } eptrblock;  } eptrblock;
2369    
2370    
# Line 1801  typedef struct match_data { Line 2375  typedef struct match_data {
2375    unsigned long int match_call_count;      /* As it says */    unsigned long int match_call_count;      /* As it says */
2376    unsigned long int match_limit;           /* As it says */    unsigned long int match_limit;           /* As it says */
2377    unsigned long int match_limit_recursion; /* As it says */    unsigned long int match_limit_recursion; /* As it says */
2378    int   *offset_vector;         /* Offset vector */    int   *offset_vector;           /* Offset vector */
2379    int    offset_end;            /* One past the end */    int    offset_end;              /* One past the end */
2380    int    offset_max;            /* The maximum usable for return data */    int    offset_max;              /* The maximum usable for return data */
2381    int    nltype;                /* Newline type */    int    nltype;                  /* Newline type */
2382    int    nllen;                 /* Newline string length */    int    nllen;                   /* Newline string length */
2383    int    name_count;            /* Number of names in name table */    int    name_count;              /* Number of names in name table */
2384    int    name_entry_size;       /* Size of entry in names table */    int    name_entry_size;         /* Size of entry in names table */
2385    uschar *name_table;           /* Table of names */    pcre_uchar *name_table;         /* Table of names */
2386    uschar nl[4];                 /* Newline string when fixed */    pcre_uchar nl[4];               /* Newline string when fixed */
2387    const  uschar *lcc;           /* Points to lower casing table */    const  pcre_uint8 *lcc;         /* Points to lower casing table */
2388    const  uschar *ctypes;        /* Points to table of type maps */    const  pcre_uint8 *fcc;         /* Points to case-flipping table */
2389    BOOL   offset_overflow;       /* Set if too many extractions */    const  pcre_uint8 *ctypes;      /* Points to table of type maps */
2390    BOOL   notbol;                /* NOTBOL flag */    BOOL   offset_overflow;         /* Set if too many extractions */
2391    BOOL   noteol;                /* NOTEOL flag */    BOOL   notbol;                  /* NOTBOL flag */
2392    BOOL   utf8;                  /* UTF8 flag */    BOOL   noteol;                  /* NOTEOL flag */
2393    BOOL   jscript_compat;        /* JAVASCRIPT_COMPAT flag */    BOOL   utf;                     /* UTF-8 / UTF-16 flag */
2394    BOOL   use_ucp;               /* PCRE_UCP flag */    BOOL   jscript_compat;          /* JAVASCRIPT_COMPAT flag */
2395    BOOL   endonly;               /* Dollar not before final \n */    BOOL   use_ucp;                 /* PCRE_UCP flag */
2396    BOOL   notempty;              /* Empty string match not wanted */    BOOL   endonly;                 /* Dollar not before final \n */
2397    BOOL   notempty_atstart;      /* Empty string match at start not wanted */    BOOL   notempty;                /* Empty string match not wanted */
2398    BOOL   hitend;                /* Hit the end of the subject at some point */    BOOL   notempty_atstart;        /* Empty string match at start not wanted */
2399    BOOL   bsr_anycrlf;           /* \R is just any CRLF, not full Unicode */    BOOL   hitend;                  /* Hit the end of the subject at some point */
2400    BOOL   hasthen;               /* Pattern contains (*THEN) */    BOOL   bsr_anycrlf;             /* \R is just any CRLF, not full Unicode */
2401    const  uschar *start_code;    /* For use when recursing */    BOOL   hasthen;                 /* Pattern contains (*THEN) */
2402    USPTR  start_subject;         /* Start of the subject string */    BOOL   ignore_skip_arg;         /* For re-run when SKIP name not found */
2403    USPTR  end_subject;           /* End of the subject string */    const  pcre_uchar *start_code;  /* For use when recursing */
2404    USPTR  start_match_ptr;       /* Start of matched string */    PCRE_PUCHAR start_subject;      /* Start of the subject string */
2405    USPTR  end_match_ptr;         /* Subject position at end match */    PCRE_PUCHAR end_subject;        /* End of the subject string */
2406    USPTR  start_used_ptr;        /* Earliest consulted character */    PCRE_PUCHAR start_match_ptr;    /* Start of matched string */
2407    int    partial;               /* PARTIAL options */    PCRE_PUCHAR end_match_ptr;      /* Subject position at end match */
2408    int    end_offset_top;        /* Highwater mark at end of match */    PCRE_PUCHAR start_used_ptr;     /* Earliest consulted character */
2409    int    capture_last;          /* Most recent capture number */    int    partial;                 /* PARTIAL options */
2410    int    start_offset;          /* The start offset value */    int    end_offset_top;          /* Highwater mark at end of match */
2411    int    match_function_type;   /* Set for certain special calls of MATCH() */    int    capture_last;            /* Most recent capture number */
2412    eptrblock *eptrchain;         /* Chain of eptrblocks for tail recursions */    int    start_offset;            /* The start offset value */
2413    int    eptrn;                 /* Next free eptrblock */    int    match_function_type;     /* Set for certain special calls of MATCH() */
2414    recursion_info *recursive;    /* Linked list of recursion data */    eptrblock *eptrchain;           /* Chain of eptrblocks for tail recursions */
2415    void  *callout_data;          /* To pass back to callouts */    int    eptrn;                   /* Next free eptrblock */
2416    const  uschar *mark;          /* Mark pointer to pass back */    recursion_info *recursive;      /* Linked list of recursion data */
2417    const  uschar *once_target;   /* Where to back up to for atomic groups */    void  *callout_data;            /* To pass back to callouts */
2418      const  pcre_uchar *mark;        /* Mark pointer to pass back on success */
2419      const  pcre_uchar *nomatch_mark;/* Mark pointer to pass back on failure */
2420      const  pcre_uchar *once_target; /* Where to back up to for atomic groups */
2421    #ifdef NO_RECURSE
2422      void  *match_frames_base;       /* For remembering malloc'd frames */
2423    #endif
2424  } match_data;  } match_data;
2425    
2426  /* 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
2427  functions. */  functions. */
2428    
2429  typedef struct dfa_match_data {  typedef struct dfa_match_data {
2430    const uschar *start_code;      /* Start of the compiled pattern */    const pcre_uchar *start_code;     /* Start of the compiled pattern */
2431    const uschar *start_subject;   /* Start of the subject string */    const pcre_uchar *start_subject ; /* Start of the subject string */
2432    const uschar *end_subject;     /* End of subject string */    const pcre_uchar *end_subject;    /* End of subject string */
2433    const uschar *start_used_ptr;  /* Earliest consulted character */    const pcre_uchar *start_used_ptr; /* Earliest consulted character */
2434    const uschar *tables;          /* Character tables */    const pcre_uint8 *tables;         /* Character tables */
2435    int   start_offset;            /* The start offset value */    int   start_offset;               /* The start offset value */
2436    int   moptions;                /* Match options */    int   moptions;                   /* Match options */
2437    int   poptions;                /* Pattern options */    int   poptions;                   /* Pattern options */
2438    int    nltype;                 /* Newline type */    int   nltype;                     /* Newline type */
2439    int    nllen;                  /* Newline string length */    int   nllen;                      /* Newline string length */
2440    uschar nl[4];                  /* Newline string when fixed */    pcre_uchar nl[4];                 /* Newline string when fixed */
2441    void  *callout_data;           /* To pass back to callouts */    void *callout_data;               /* To pass back to callouts */
2442    dfa_recursion_info *recursive; /* Linked list of recursion data */    dfa_recursion_info *recursive;    /* Linked list of recursion data */
2443  } dfa_match_data;  } dfa_match_data;
2444    
2445  /* Bit definitions for entries in the pcre_ctypes table. */  /* Bit definitions for entries in the pcre_ctypes table. */
# Line 1895  total length. */ Line 2475  total length. */
2475  #define ctypes_offset (cbits_offset + cbit_length)  #define ctypes_offset (cbits_offset + cbit_length)
2476  #define tables_length (ctypes_offset + 256)  #define tables_length (ctypes_offset + 256)
2477    
2478    /* Internal function and data prefixes. */
2479    
2480    #if defined COMPILE_PCRE8
2481    #ifndef PUBL
2482    #define PUBL(name) pcre_##name
2483    #endif
2484    #ifndef PRIV
2485    #define PRIV(name) _pcre_##name
2486    #endif
2487    #elif defined COMPILE_PCRE16
2488    #ifndef PUBL
2489    #define PUBL(name) pcre16_##name
2490    #endif
2491    #ifndef PRIV
2492    #define PRIV(name) _pcre16_##name
2493    #endif
2494    #elif defined COMPILE_PCRE32
2495    #ifndef PUBL
2496    #define PUBL(name) pcre32_##name
2497    #endif
2498    #ifndef PRIV
2499    #define PRIV(name) _pcre32_##name
2500    #endif
2501    #else
2502    #error Unsupported compiling mode
2503    #endif /* COMPILE_PCRE[8|16|32] */
2504    
2505  /* 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
2506  codes. Each entry used to point directly to a name, but to reduce the number of  codes. Each entry used to point directly to a name, but to reduce the number of
2507  relocations in shared libraries, it now has an offset into a single string  relocations in shared libraries, it now has an offset into a single string
# Line 1912  of the exported public functions. They h Line 2519  of the exported public functions. They h
2519  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
2520  pcre_tables.c module. */  pcre_tables.c module. */
2521    
2522  extern const int    _pcre_utf8_table1[];  #ifdef COMPILE_PCRE8
2523  extern const int    _pcre_utf8_table2[];  extern const int            PRIV(utf8_table1)[];
2524  extern const int    _pcre_utf8_table3[];  extern const int            PRIV(utf8_table1_size);
2525  extern const uschar _pcre_utf8_table4[];  extern const int            PRIV(utf8_table2)[];
2526    extern const int            PRIV(utf8_table3)[];
2527    extern const pcre_uint8     PRIV(utf8_table4)[];
2528    #endif /* COMPILE_PCRE8 */
2529    
2530    extern const char           PRIV(utt_names)[];
2531    extern const ucp_type_table PRIV(utt)[];
2532    extern const int            PRIV(utt_size);
2533    
2534  #ifdef SUPPORT_JIT  extern const pcre_uint8     PRIV(OP_lengths)[];
2535  extern const uschar _pcre_utf8_char_sizes[];  extern const pcre_uint8     PRIV(default_tables)[];
 #endif  
   
 extern const int    _pcre_utf8_table1_size;  
   
 extern const char   _pcre_utt_names[];  
 extern const ucp_type_table _pcre_utt[];  
 extern const int _pcre_utt_size;  
2536    
2537  extern const uschar _pcre_default_tables[];  extern const pcre_uint32    PRIV(hspace_list)[];
2538    extern const pcre_uint32    PRIV(vspace_list)[];
 extern const uschar _pcre_OP_lengths[];  
2539    
2540    
2541  /* Internal shared functions. These are functions that are used by more than  /* Internal shared functions. These are functions that are used by more than
2542  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
2543  sense, but are not part of the PCRE public API. */  sense, but are not part of the PCRE public API. */
2544    
2545  extern const uschar *_pcre_find_bracket(const uschar *, BOOL, int);  /* String comparison functions. */
2546  extern BOOL          _pcre_is_newline(USPTR, int, USPTR, int *, BOOL);  #if defined COMPILE_PCRE8
2547  extern int           _pcre_ord2utf8(int, uschar *);  
2548  extern real_pcre    *_pcre_try_flipped(const real_pcre *, real_pcre *,  #define STRCMP_UC_UC(str1, str2) \
2549                         const pcre_study_data *, pcre_study_data *);    strcmp((char *)(str1), (char *)(str2))
2550  extern int           _pcre_valid_utf8(USPTR, int, int *);  #define STRCMP_UC_C8(str1, str2) \
2551  extern BOOL          _pcre_was_newline(USPTR, int, USPTR, int *, BOOL);    strcmp((char *)(str1), (str2))
2552  extern BOOL          _pcre_xclass(int, const uschar *);  #define STRNCMP_UC_UC(str1, str2, num) \
2553      strncmp((char *)(str1), (char *)(str2), (num))
2554    #define STRNCMP_UC_C8(str1, str2, num) \
2555      strncmp((char *)(str1), (str2), (num))
2556    #define STRLEN_UC(str) strlen((const char *)str)
2557    
2558    #elif defined COMPILE_PCRE16 || defined COMPILE_PCRE32
2559    
2560    extern int               PRIV(strcmp_uc_uc)(const pcre_uchar *,
2561                               const pcre_uchar *);
2562    extern int               PRIV(strcmp_uc_c8)(const pcre_uchar *,
2563                               const char *);
2564    extern int               PRIV(strncmp_uc_uc)(const pcre_uchar *,
2565                               const pcre_uchar *, unsigned int num);
2566    extern int               PRIV(strncmp_uc_c8)(const pcre_uchar *,
2567                               const char *, unsigned int num);
2568    extern unsigned int      PRIV(strlen_uc)(const pcre_uchar *str);
2569    
2570    #define STRCMP_UC_UC(str1, str2) \
2571      PRIV(strcmp_uc_uc)((str1), (str2))
2572    #define STRCMP_UC_C8(str1, str2) \
2573      PRIV(strcmp_uc_c8)((str1), (str2))
2574    #define STRNCMP_UC_UC(str1, str2, num) \
2575      PRIV(strncmp_uc_uc)((str1), (str2), (num))
2576    #define STRNCMP_UC_C8(str1, str2, num) \
2577      PRIV(strncmp_uc_c8)((str1), (str2), (num))
2578    #define STRLEN_UC(str) PRIV(strlen_uc)(str)
2579    
2580    #endif /* COMPILE_PCRE[8|16|32] */
2581    
2582    extern const pcre_uchar *PRIV(find_bracket)(const pcre_uchar *, BOOL, int);
2583    extern BOOL              PRIV(is_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2584                               int *, BOOL);
2585    extern int               PRIV(ord2utf)(pcre_uint32, pcre_uchar *);
2586    extern int               PRIV(valid_utf)(PCRE_PUCHAR, int, int *);
2587    extern BOOL              PRIV(was_newline)(PCRE_PUCHAR, int, PCRE_PUCHAR,
2588                               int *, BOOL);
2589    extern BOOL              PRIV(xclass)(pcre_uint32, const pcre_uchar *, BOOL);
2590    
2591  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2592  extern void          _pcre_jit_compile(const real_pcre *, pcre_extra *);  extern void              PRIV(jit_compile)(const REAL_PCRE *,
2593  extern int           _pcre_jit_exec(const real_pcre *, void *, PCRE_SPTR,                             PUBL(extra) *, int);
2594                          int, int, int, int, int *, int);  extern int               PRIV(jit_exec)(const REAL_PCRE *, const PUBL(extra) *,
2595  extern void          _pcre_jit_free(void *);                             const pcre_uchar *, int, int, int, int *, int);
2596    extern void              PRIV(jit_free)(void *);
2597    extern int               PRIV(jit_get_size)(void *);
2598    extern const char*       PRIV(jit_get_target)(void);
2599  #endif  #endif
2600    
2601  /* Unicode character database (UCD) */  /* Unicode character database (UCD) */
2602    
2603  typedef struct {  typedef struct {
2604    uschar script;    pcre_uint8 script;     /* ucp_Arabic, etc. */
2605    uschar chartype;    pcre_uint8 chartype;   /* ucp_Cc, etc. (general categories) */
2606    pcre_int32 other_case;    pcre_uint8 gbprop;     /* ucp_gbControl, etc. (grapheme break property) */
2607      pcre_uint8 caseset;    /* offset to multichar other cases or zero */
2608      pcre_int32 other_case; /* offset to other case, or zero if none */
2609  } ucd_record;  } ucd_record;
2610    
2611  extern const ucd_record  _pcre_ucd_records[];  extern const pcre_uint32 PRIV(ucd_caseless_sets)[];
2612  extern const uschar      _pcre_ucd_stage1[];  extern const ucd_record  PRIV(ucd_records)[];
2613  extern const pcre_uint16 _pcre_ucd_stage2[];  extern const pcre_uint8  PRIV(ucd_stage1)[];
2614  extern const int         _pcre_ucp_gentype[];  extern const pcre_uint16 PRIV(ucd_stage2)[];
2615    extern const int         PRIV(ucp_gentype)[];
2616    extern const pcre_uint32 PRIV(ucp_gbtable)[];
2617  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
2618  extern const int         _pcre_ucp_typerange[];  extern const int         PRIV(ucp_typerange)[];
2619  #endif  #endif
2620    
2621    #ifdef SUPPORT_UCP
2622  /* UCD access macros */  /* UCD access macros */
2623    
2624  #define UCD_BLOCK_SIZE 128  #define UCD_BLOCK_SIZE 128
2625  #define GET_UCD(ch) (_pcre_ucd_records + \  #define GET_UCD(ch) (PRIV(ucd_records) + \
2626          _pcre_ucd_stage2[_pcre_ucd_stage1[(ch) / UCD_BLOCK_SIZE] * \          PRIV(ucd_stage2)[PRIV(ucd_stage1)[(ch) / UCD_BLOCK_SIZE] * \
2627          UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])          UCD_BLOCK_SIZE + (ch) % UCD_BLOCK_SIZE])
2628    
2629  #define UCD_CHARTYPE(ch)  GET_UCD(ch)->chartype  #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
2630  #define UCD_SCRIPT(ch)    GET_UCD(ch)->script  #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
2631  #define UCD_CATEGORY(ch)  _pcre_ucp_gentype[UCD_CHARTYPE(ch)]  #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
2632  #define UCD_OTHERCASE(ch) (ch + GET_UCD(ch)->other_case)  #define UCD_GRAPHBREAK(ch)  GET_UCD(ch)->gbprop
2633    #define UCD_CASESET(ch)     GET_UCD(ch)->caseset
2634    #define UCD_OTHERCASE(ch)   (ch + GET_UCD(ch)->other_case)
2635    
2636    #endif /* SUPPORT_UCP */
2637    
2638  #endif  #endif
2639    

Legend:
Removed from v.758  
changed lines
  Added in v.1057

  ViewVC Help
Powered by ViewVC 1.1.5