/[pcre]/code/trunk/pcre_exec.c
ViewVC logotype

Diff of /code/trunk/pcre_exec.c

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

revision 144 by ph10, Mon Apr 2 13:32:07 2007 UTC revision 400 by ph10, Sat Mar 21 16:59:40 2009 UTC
# Line 6  Line 6 
6  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.
7    
8                         Written by Philip Hazel                         Written by Philip Hazel
9             Copyright (c) 1997-2007 University of Cambridge             Copyright (c) 1997-2009 University of Cambridge
10    
11  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
12  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 42  POSSIBILITY OF SUCH DAMAGE. Line 42  POSSIBILITY OF SUCH DAMAGE.
42  pattern matching using an NFA algorithm, trying to mimic Perl as closely as  pattern matching using an NFA algorithm, trying to mimic Perl as closely as
43  possible. There are also some static supporting functions. */  possible. There are also some static supporting functions. */
44    
45    #ifdef HAVE_CONFIG_H
46    #include "config.h"
47    #endif
48    
49  #define NLBLOCK md             /* Block containing newline information */  #define NLBLOCK md             /* Block containing newline information */
50  #define PSSTART start_subject  /* Field containing processed string start */  #define PSSTART start_subject  /* Field containing processed string start */
51  #define PSEND   end_subject    /* Field containing processed string end */  #define PSEND   end_subject    /* Field containing processed string end */
# Line 53  possible. There are also some static sup Line 57  possible. There are also some static sup
57  #undef min  #undef min
58  #undef max  #undef max
59    
 /* The chain of eptrblocks for tail recursions uses memory in stack workspace,  
 obtained at top level, the size of which is defined by EPTR_WORK_SIZE. */  
   
 #define EPTR_WORK_SIZE (1000)  
   
60  /* Flag bits for the match() function */  /* Flag bits for the match() function */
61    
62  #define match_condassert     0x01  /* Called to check a condition assertion */  #define match_condassert     0x01  /* Called to check a condition assertion */
63  #define match_cbegroup       0x02  /* Could-be-empty unlimited repeat group */  #define match_cbegroup       0x02  /* Could-be-empty unlimited repeat group */
 #define match_tail_recursed  0x04  /* Tail recursive call */  
64    
65  /* Non-error returns from the match() function. Error returns are externally  /* Non-error returns from the match() function. Error returns are externally
66  defined PCRE_ERROR_xxx codes, which are all negative. */  defined PCRE_ERROR_xxx codes, which are all negative. */
# Line 70  defined PCRE_ERROR_xxx codes, which are Line 68  defined PCRE_ERROR_xxx codes, which are
68  #define MATCH_MATCH        1  #define MATCH_MATCH        1
69  #define MATCH_NOMATCH      0  #define MATCH_NOMATCH      0
70    
71    /* Special internal returns from the match() function. Make them sufficiently
72    negative to avoid the external error codes. */
73    
74    #define MATCH_COMMIT       (-999)
75    #define MATCH_PRUNE        (-998)
76    #define MATCH_SKIP         (-997)
77    #define MATCH_THEN         (-996)
78    
79  /* Maximum number of ints of offset to save on the stack for recursive calls.  /* Maximum number of ints of offset to save on the stack for recursive calls.
80  If the offset vector is bigger, malloc is used. This should be a multiple of 3,  If the offset vector is bigger, malloc is used. This should be a multiple of 3,
81  because the offset vector is always a multiple of 3 long. */  because the offset vector is always a multiple of 3 long. */
# Line 152  printf("\n"); Line 158  printf("\n");
158    
159  if (length > md->end_subject - eptr) return FALSE;  if (length > md->end_subject - eptr) return FALSE;
160    
161  /* Separate the caselesss case for speed */  /* Separate the caseless case for speed. In UTF-8 mode we can only do this
162    properly if Unicode properties are supported. Otherwise, we can check only
163    ASCII characters. */
164    
165  if ((ims & PCRE_CASELESS) != 0)  if ((ims & PCRE_CASELESS) != 0)
166    {    {
167    #ifdef SUPPORT_UTF8
168    #ifdef SUPPORT_UCP
169      if (md->utf8)
170        {
171        USPTR endptr = eptr + length;
172        while (eptr < endptr)
173          {
174          int c, d;
175          GETCHARINC(c, eptr);
176          GETCHARINC(d, p);
177          if (c != d && c != UCD_OTHERCASE(d)) return FALSE;
178          }
179        }
180      else
181    #endif
182    #endif
183    
184      /* The same code works when not in UTF-8 mode and in UTF-8 mode when there
185      is no UCP support. */
186    
187    while (length-- > 0)    while (length-- > 0)
188      if (md->lcc[*p++] != md->lcc[*eptr++]) return FALSE;      { if (md->lcc[*p++] != md->lcc[*eptr++]) return FALSE; }
189    }    }
190    
191    /* In the caseful case, we can just compare the bytes, whether or not we
192    are in UTF-8 mode. */
193    
194  else  else
195    { while (length-- > 0) if (*p++ != *eptr++) return FALSE; }    { while (length-- > 0) if (*p++ != *eptr++) return FALSE; }
196    
# Line 188  calls by keeping local variables that ne Line 220  calls by keeping local variables that ne
220  obtained from malloc() instead instead of on the stack. Macros are used to  obtained from malloc() instead instead of on the stack. Macros are used to
221  achieve this so that the actual code doesn't look very different to what it  achieve this so that the actual code doesn't look very different to what it
222  always used to.  always used to.
223    
224    The original heap-recursive code used longjmp(). However, it seems that this
225    can be very slow on some operating systems. Following a suggestion from Stan
226    Switzer, the use of longjmp() has been abolished, at the cost of having to
227    provide a unique number for each call to RMATCH. There is no way of generating
228    a sequence of numbers at compile time in C. I have given them names, to make
229    them stand out more clearly.
230    
231    Crude tests on x86 Linux show a small speedup of around 5-8%. However, on
232    FreeBSD, avoiding longjmp() more than halves the time taken to run the standard
233    tests. Furthermore, not using longjmp() means that local dynamic variables
234    don't have indeterminate values; this has meant that the frame size can be
235    reduced because the result can be "passed back" by straight setting of the
236    variable instead of being passed in the frame.
237  ****************************************************************************  ****************************************************************************
238  ***************************************************************************/  ***************************************************************************/
239    
240    /* Numbers for RMATCH calls. When this list is changed, the code at HEAP_RETURN
241    below must be updated in sync.  */
242    
243    enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM6,  RM7,  RM8,  RM9,  RM10,
244           RM11,  RM12, RM13, RM14, RM15, RM16, RM17, RM18, RM19, RM20,
245           RM21,  RM22, RM23, RM24, RM25, RM26, RM27, RM28, RM29, RM30,
246           RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,
247           RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,
248           RM51,  RM52, RM53, RM54 };
249    
250  /* These versions of the macros use the stack, as normal. There are debugging  /* These versions of the macros use the stack, as normal. There are debugging
251  versions and production versions. */  versions and production versions. Note that the "rw" argument of RMATCH isn't
252    actuall used in this definition. */
253    
254  #ifndef NO_RECURSE  #ifndef NO_RECURSE
255  #define REGISTER register  #define REGISTER register
256    
257  #ifdef DEBUG  #ifdef DEBUG
258  #define RMATCH(rx,ra,rb,rc,rd,re,rf,rg) \  #define RMATCH(ra,rb,rc,rd,re,rf,rg,rw) \
259    { \    { \
260    printf("match() called in line %d\n", __LINE__); \    printf("match() called in line %d\n", __LINE__); \
261    rx = match(ra,rb,rc,rd,re,rf,rg,rdepth+1); \    rrc = match(ra,rb,mstart,rc,rd,re,rf,rg,rdepth+1); \
262    printf("to line %d\n", __LINE__); \    printf("to line %d\n", __LINE__); \
263    }    }
264  #define RRETURN(ra) \  #define RRETURN(ra) \
# Line 210  versions and production versions. */ Line 267  versions and production versions. */
267    return ra; \    return ra; \
268    }    }
269  #else  #else
270  #define RMATCH(rx,ra,rb,rc,rd,re,rf,rg) \  #define RMATCH(ra,rb,rc,rd,re,rf,rg,rw) \
271    rx = match(ra,rb,rc,rd,re,rf,rg,rdepth+1)    rrc = match(ra,rb,mstart,rc,rd,re,rf,rg,rdepth+1)
272  #define RRETURN(ra) return ra  #define RRETURN(ra) return ra
273  #endif  #endif
274    
275  #else  #else
276    
277    
278  /* These versions of the macros manage a private stack on the heap. Note  /* These versions of the macros manage a private stack on the heap. Note that
279  that the rd argument of RMATCH isn't actually used. It's the md argument of  the "rd" argument of RMATCH isn't actually used in this definition. It's the md
280  match(), which never changes. */  argument of match(), which never changes. */
281    
282  #define REGISTER  #define REGISTER
283    
284  #define RMATCH(rx,ra,rb,rc,rd,re,rf,rg)\  #define RMATCH(ra,rb,rc,rd,re,rf,rg,rw)\
285    {\    {\
286    heapframe *newframe = (pcre_stack_malloc)(sizeof(heapframe));\    heapframe *newframe = (pcre_stack_malloc)(sizeof(heapframe));\
287    if (setjmp(frame->Xwhere) == 0)\    frame->Xwhere = rw; \
288      {\    newframe->Xeptr = ra;\
289      newframe->Xeptr = ra;\    newframe->Xecode = rb;\
290      newframe->Xecode = rb;\    newframe->Xmstart = mstart;\
291      newframe->Xoffset_top = rc;\    newframe->Xoffset_top = rc;\
292      newframe->Xims = re;\    newframe->Xims = re;\
293      newframe->Xeptrb = rf;\    newframe->Xeptrb = rf;\
294      newframe->Xflags = rg;\    newframe->Xflags = rg;\
295      newframe->Xrdepth = frame->Xrdepth + 1;\    newframe->Xrdepth = frame->Xrdepth + 1;\
296      newframe->Xprevframe = frame;\    newframe->Xprevframe = frame;\
297      frame = newframe;\    frame = newframe;\
298      DPRINTF(("restarting from line %d\n", __LINE__));\    DPRINTF(("restarting from line %d\n", __LINE__));\
299      goto HEAP_RECURSE;\    goto HEAP_RECURSE;\
300      }\    L_##rw:\
301    else\    DPRINTF(("jumped back to line %d\n", __LINE__));\
     {\  
     DPRINTF(("longjumped back to line %d\n", __LINE__));\  
     frame = md->thisframe;\  
     rx = frame->Xresult;\  
     }\  
302    }    }
303    
304  #define RRETURN(ra)\  #define RRETURN(ra)\
# Line 256  match(), which never changes. */ Line 308  match(), which never changes. */
308    (pcre_stack_free)(newframe);\    (pcre_stack_free)(newframe);\
309    if (frame != NULL)\    if (frame != NULL)\
310      {\      {\
311      frame->Xresult = ra;\      rrc = ra;\
312      md->thisframe = frame;\      goto HEAP_RETURN;\
     longjmp(frame->Xwhere, 1);\  
313      }\      }\
314    return ra;\    return ra;\
315    }    }
# Line 273  typedef struct heapframe { Line 324  typedef struct heapframe {
324    
325    const uschar *Xeptr;    const uschar *Xeptr;
326    const uschar *Xecode;    const uschar *Xecode;
327      const uschar *Xmstart;
328    int Xoffset_top;    int Xoffset_top;
329    long int Xims;    long int Xims;
330    eptrblock *Xeptrb;    eptrblock *Xeptrb;
# Line 323  typedef struct heapframe { Line 375  typedef struct heapframe {
375    
376    eptrblock Xnewptrb;    eptrblock Xnewptrb;
377    
378    /* Place to pass back result, and where to jump back to */    /* Where to jump back to */
379    
380    int  Xresult;    int Xwhere;
   jmp_buf Xwhere;  
381    
382  } heapframe;  } heapframe;
383    
# Line 354  made performance worse. Line 405  made performance worse.
405  Arguments:  Arguments:
406     eptr        pointer to current character in subject     eptr        pointer to current character in subject
407     ecode       pointer to current position in compiled code     ecode       pointer to current position in compiled code
408       mstart      pointer to the current match start position (can be modified
409                     by encountering \K)
410     offset_top  current top pointer     offset_top  current top pointer
411     md          pointer to "static" info for the match     md          pointer to "static" info for the match
412     ims         current /i, /m, and /s options     ims         current /i, /m, and /s options
# Line 363  Arguments: Line 416  Arguments:
416                   match_condassert - this is an assertion condition                   match_condassert - this is an assertion condition
417                   match_cbegroup - this is the start of an unlimited repeat                   match_cbegroup - this is the start of an unlimited repeat
418                     group that can match an empty string                     group that can match an empty string
                  match_tail_recursed - this is a tail_recursed group  
419     rdepth      the recursion depth     rdepth      the recursion depth
420    
421  Returns:       MATCH_MATCH if matched            )  these values are >= 0  Returns:       MATCH_MATCH if matched            )  these values are >= 0
# Line 373  Returns:       MATCH_MATCH if matched Line 425  Returns:       MATCH_MATCH if matched
425  */  */
426    
427  static int  static int
428  match(REGISTER USPTR eptr, REGISTER const uschar *ecode,  match(REGISTER USPTR eptr, REGISTER const uschar *ecode, const uschar *mstart,
429    int offset_top, match_data *md, unsigned long int ims, eptrblock *eptrb,    int offset_top, match_data *md, unsigned long int ims, eptrblock *eptrb,
430    int flags, unsigned int rdepth)    int flags, unsigned int rdepth)
431  {  {
# Line 401  frame->Xprevframe = NULL;            /* Line 453  frame->Xprevframe = NULL;            /*
453    
454  frame->Xeptr = eptr;  frame->Xeptr = eptr;
455  frame->Xecode = ecode;  frame->Xecode = ecode;
456    frame->Xmstart = mstart;
457  frame->Xoffset_top = offset_top;  frame->Xoffset_top = offset_top;
458  frame->Xims = ims;  frame->Xims = ims;
459  frame->Xeptrb = eptrb;  frame->Xeptrb = eptrb;
# Line 415  HEAP_RECURSE: Line 468  HEAP_RECURSE:
468    
469  #define eptr               frame->Xeptr  #define eptr               frame->Xeptr
470  #define ecode              frame->Xecode  #define ecode              frame->Xecode
471    #define mstart             frame->Xmstart
472  #define offset_top         frame->Xoffset_top  #define offset_top         frame->Xoffset_top
473  #define ims                frame->Xims  #define ims                frame->Xims
474  #define eptrb              frame->Xeptrb  #define eptrb              frame->Xeptrb
# Line 507  int oclength; Line 561  int oclength;
561  uschar occhars[8];  uschar occhars[8];
562  #endif  #endif
563    
564    int codelink;
565    int condcode;
566  int ctype;  int ctype;
567  int length;  int length;
568  int max;  int max;
# Line 545  defined). However, RMATCH isn't like a f Line 601  defined). However, RMATCH isn't like a f
601  complicated macro. It has to be used in one particular way. This shouldn't,  complicated macro. It has to be used in one particular way. This shouldn't,
602  however, impact performance when true recursion is being used. */  however, impact performance when true recursion is being used. */
603    
604    #ifdef SUPPORT_UTF8
605    utf8 = md->utf8;       /* Local copy of the flag */
606    #else
607    utf8 = FALSE;
608    #endif
609    
610  /* First check that we haven't called match() too many times, or that we  /* First check that we haven't called match() too many times, or that we
611  haven't exceeded the recursive call limit. */  haven't exceeded the recursive call limit. */
612    
# Line 553  if (rdepth >= md->match_limit_recursion) Line 615  if (rdepth >= md->match_limit_recursion)
615    
616  original_ims = ims;    /* Save for resetting on ')' */  original_ims = ims;    /* Save for resetting on ')' */
617    
 #ifdef SUPPORT_UTF8  
 utf8 = md->utf8;       /* Local copy of the flag */  
 #else  
 utf8 = FALSE;  
 #endif  
   
618  /* At the start of a group with an unlimited repeat that may match an empty  /* At the start of a group with an unlimited repeat that may match an empty
619  string, the match_cbegroup flag is set. When this is the case, add the current  string, the match_cbegroup flag is set. When this is the case, add the current
620  subject pointer to the chain of such remembered pointers, to be checked when we  subject pointer to the chain of such remembered pointers, to be checked when we
621  hit the closing ket, in order to break infinite loops that match no characters.  hit the closing ket, in order to break infinite loops that match no characters.
622  When match() is called in other circumstances, don't add to the chain. If this  When match() is called in other circumstances, don't add to the chain. The
623  is a tail recursion, use a block from the workspace, as the one on the stack is  match_cbegroup flag must NOT be used with tail recursion, because the memory
624  already used. */  block that is used is on the stack, so a new one may be required for each
625    match(). */
626    
627  if ((flags & match_cbegroup) != 0)  if ((flags & match_cbegroup) != 0)
628    {    {
629    eptrblock *p;    newptrb.epb_saved_eptr = eptr;
630    if ((flags & match_tail_recursed) != 0)    newptrb.epb_prev = eptrb;
631      {    eptrb = &newptrb;
     if (md->eptrn >= EPTR_WORK_SIZE) RRETURN(PCRE_ERROR_NULLWSLIMIT);  
     p = md->eptrchain + md->eptrn++;  
     }  
   else p = &newptrb;  
   p->epb_saved_eptr = eptr;  
   p->epb_prev = eptrb;  
   eptrb = p;  
632    }    }
633    
634  /* Now start processing the opcodes. */  /* Now start processing the opcodes. */
# Line 587  for (;;) Line 637  for (;;)
637    {    {
638    minimize = possessive = FALSE;    minimize = possessive = FALSE;
639    op = *ecode;    op = *ecode;
640    
641    /* For partial matching, remember if we ever hit the end of the subject after    /* For partial matching, remember if we ever hit the end of the subject after
642    matching at least one subject character. */    matching at least one subject character. */
643    
644    if (md->partial &&    if (md->partial &&
645        eptr >= md->end_subject &&        eptr >= md->end_subject &&
646        eptr > md->start_match)        eptr > mstart)
647      md->hitend = TRUE;      md->hitend = TRUE;
648    
649    switch(op)    switch(op)
650      {      {
651        case OP_FAIL:
652        RRETURN(MATCH_NOMATCH);
653    
654        case OP_PRUNE:
655        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
656          ims, eptrb, flags, RM51);
657        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
658        RRETURN(MATCH_PRUNE);
659    
660        case OP_COMMIT:
661        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
662          ims, eptrb, flags, RM52);
663        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
664        RRETURN(MATCH_COMMIT);
665    
666        case OP_SKIP:
667        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
668          ims, eptrb, flags, RM53);
669        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
670        md->start_match_ptr = eptr;   /* Pass back current position */
671        RRETURN(MATCH_SKIP);
672    
673        case OP_THEN:
674        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
675          ims, eptrb, flags, RM54);
676        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
677        RRETURN(MATCH_THEN);
678    
679      /* Handle a capturing bracket. If there is space in the offset vector, save      /* Handle a capturing bracket. If there is space in the offset vector, save
680      the current subject position in the working slot at the top of the vector.      the current subject position in the working slot at the top of the vector.
681      We mustn't change the current values of the data slot, because they may be      We mustn't change the current values of the data slot, because they may be
# Line 637  for (;;) Line 715  for (;;)
715        flags = (op == OP_SCBRA)? match_cbegroup : 0;        flags = (op == OP_SCBRA)? match_cbegroup : 0;
716        do        do
717          {          {
718          RMATCH(rrc, eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
719            ims, eptrb, flags);            ims, eptrb, flags, RM1);
720          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
721          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
722          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
723          }          }
# Line 654  for (;;) Line 732  for (;;)
732        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
733        }        }
734    
735      /* Insufficient room for saving captured contents. Treat as a non-capturing      /* FALL THROUGH ... Insufficient room for saving captured contents. Treat
736      bracket. */      as a non-capturing bracket. */
737    
738        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
739        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
740    
741      DPRINTF(("insufficient capture room: treat as non-capturing\n"));      DPRINTF(("insufficient capture room: treat as non-capturing\n"));
742    
743        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
744        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
745    
746      /* Non-capturing bracket. Loop for all the alternatives. When we get to the      /* Non-capturing bracket. Loop for all the alternatives. When we get to the
747      final alternative within the brackets, we would return the result of a      final alternative within the brackets, we would return the result of a
748      recursive call to match() whatever happened. We can reduce stack usage by      recursive call to match() whatever happened. We can reduce stack usage by
749      turning this into a tail recursion. */      turning this into a tail recursion, except in the case when match_cbegroup
750        is set.*/
751    
752      case OP_BRA:      case OP_BRA:
753      case OP_SBRA:      case OP_SBRA:
# Line 670  for (;;) Line 755  for (;;)
755      flags = (op >= OP_SBRA)? match_cbegroup : 0;      flags = (op >= OP_SBRA)? match_cbegroup : 0;
756      for (;;)      for (;;)
757        {        {
758        if (ecode[GET(ecode, 1)] != OP_ALT)        if (ecode[GET(ecode, 1)] != OP_ALT)   /* Final alternative */
759          {          {
760          ecode += _pcre_OP_lengths[*ecode];          if (flags == 0)    /* Not a possibly empty group */
761          flags |= match_tail_recursed;            {
762          DPRINTF(("bracket 0 tail recursion\n"));            ecode += _pcre_OP_lengths[*ecode];
763          goto TAIL_RECURSE;            DPRINTF(("bracket 0 tail recursion\n"));
764              goto TAIL_RECURSE;
765              }
766    
767            /* Possibly empty group; can't use tail recursion. */
768    
769            RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,
770              eptrb, flags, RM48);
771            RRETURN(rrc);
772          }          }
773    
774        /* For non-final alternatives, continue the loop for a NOMATCH result;        /* For non-final alternatives, continue the loop for a NOMATCH result;
775        otherwise return. */        otherwise return. */
776    
777        RMATCH(rrc, eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,
778          eptrb, flags);          eptrb, flags, RM2);
779        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
780        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
781        }        }
782      /* Control never reaches here. */      /* Control never reaches here. */
# Line 696  for (;;) Line 789  for (;;)
789    
790      case OP_COND:      case OP_COND:
791      case OP_SCOND:      case OP_SCOND:
792      if (ecode[LINK_SIZE+1] == OP_RREF)         /* Recursion test */      codelink= GET(ecode, 1);
793    
794        /* Because of the way auto-callout works during compile, a callout item is
795        inserted between OP_COND and an assertion condition. */
796    
797        if (ecode[LINK_SIZE+1] == OP_CALLOUT)
798          {
799          if (pcre_callout != NULL)
800            {
801            pcre_callout_block cb;
802            cb.version          = 1;   /* Version 1 of the callout block */
803            cb.callout_number   = ecode[LINK_SIZE+2];
804            cb.offset_vector    = md->offset_vector;
805            cb.subject          = (PCRE_SPTR)md->start_subject;
806            cb.subject_length   = md->end_subject - md->start_subject;
807            cb.start_match      = mstart - md->start_subject;
808            cb.current_position = eptr - md->start_subject;
809            cb.pattern_position = GET(ecode, LINK_SIZE + 3);
810            cb.next_item_length = GET(ecode, 3 + 2*LINK_SIZE);
811            cb.capture_top      = offset_top/2;
812            cb.capture_last     = md->capture_last;
813            cb.callout_data     = md->callout_data;
814            if ((rrc = (*pcre_callout)(&cb)) > 0) RRETURN(MATCH_NOMATCH);
815            if (rrc < 0) RRETURN(rrc);
816            }
817          ecode += _pcre_OP_lengths[OP_CALLOUT];
818          }
819    
820        condcode = ecode[LINK_SIZE+1];
821    
822        /* Now see what the actual condition is */
823    
824        if (condcode == OP_RREF)         /* Recursion test */
825        {        {
826        offset = GET2(ecode, LINK_SIZE + 2);     /* Recursion group number*/        offset = GET2(ecode, LINK_SIZE + 2);     /* Recursion group number*/
827        condition = md->recursive != NULL &&        condition = md->recursive != NULL &&
# Line 704  for (;;) Line 829  for (;;)
829        ecode += condition? 3 : GET(ecode, 1);        ecode += condition? 3 : GET(ecode, 1);
830        }        }
831    
832      else if (ecode[LINK_SIZE+1] == OP_CREF)    /* Group used test */      else if (condcode == OP_CREF)    /* Group used test */
833        {        {
834        offset = GET2(ecode, LINK_SIZE+2) << 1;  /* Doubled ref number */        offset = GET2(ecode, LINK_SIZE+2) << 1;  /* Doubled ref number */
835        condition = offset < offset_top && md->offset_vector[offset] >= 0;        condition = offset < offset_top && md->offset_vector[offset] >= 0;
836        ecode += condition? 3 : GET(ecode, 1);        ecode += condition? 3 : GET(ecode, 1);
837        }        }
838    
839      else if (ecode[LINK_SIZE+1] == OP_DEF)     /* DEFINE - always false */      else if (condcode == OP_DEF)     /* DEFINE - always false */
840        {        {
841        condition = FALSE;        condition = FALSE;
842        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
# Line 723  for (;;) Line 848  for (;;)
848    
849      else      else
850        {        {
851        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL,
852            match_condassert);            match_condassert, RM3);
853        if (rrc == MATCH_MATCH)        if (rrc == MATCH_MATCH)
854          {          {
855          condition = TRUE;          condition = TRUE;
856          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);
857          while (*ecode == OP_ALT) ecode += GET(ecode, 1);          while (*ecode == OP_ALT) ecode += GET(ecode, 1);
858          }          }
859        else if (rrc != MATCH_NOMATCH)        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
860          {          {
861          RRETURN(rrc);         /* Need braces because of following else */          RRETURN(rrc);         /* Need braces because of following else */
862          }          }
863        else        else
864          {          {
865          condition = FALSE;          condition = FALSE;
866          ecode += GET(ecode, 1);          ecode += codelink;
867          }          }
868        }        }
869    
870      /* We are now at the branch that is to be obeyed. As there is only one,      /* We are now at the branch that is to be obeyed. As there is only one,
871      we can use tail recursion to avoid using another stack frame. If the second      we can use tail recursion to avoid using another stack frame, except when
872      alternative doesn't exist, we can just plough on. */      match_cbegroup is required for an unlimited repeat of a possibly empty
873        group. If the second alternative doesn't exist, we can just plough on. */
874    
875      if (condition || *ecode == OP_ALT)      if (condition || *ecode == OP_ALT)
876        {        {
877        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
878        flags = match_tail_recursed | ((op == OP_SCOND)? match_cbegroup : 0);        if (op == OP_SCOND)        /* Possibly empty group */
879        goto TAIL_RECURSE;          {
880            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, match_cbegroup, RM49);
881            RRETURN(rrc);
882            }
883          else                       /* Group must match something */
884            {
885            flags = 0;
886            goto TAIL_RECURSE;
887            }
888        }        }
889      else      else                         /* Condition false & no alternative */
890        {        {
891        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
892        }        }
893      break;      break;
894    
895    
896      /* End of the pattern. If we are in a top-level recursion, we should      /* End of the pattern, either real or forced. If we are in a top-level
897      restore the offsets appropriately and continue from after the call. */      recursion, we should restore the offsets appropriately and continue from
898        after the call. */
899    
900        case OP_ACCEPT:
901      case OP_END:      case OP_END:
902      if (md->recursive != NULL && md->recursive->group_num == 0)      if (md->recursive != NULL && md->recursive->group_num == 0)
903        {        {
# Line 770  for (;;) Line 906  for (;;)
906        md->recursive = rec->prevrec;        md->recursive = rec->prevrec;
907        memmove(md->offset_vector, rec->offset_save,        memmove(md->offset_vector, rec->offset_save,
908          rec->saved_max * sizeof(int));          rec->saved_max * sizeof(int));
909        md->start_match = rec->save_start;        mstart = rec->save_start;
910        ims = original_ims;        ims = original_ims;
911        ecode = rec->after_call;        ecode = rec->after_call;
912        break;        break;
# Line 779  for (;;) Line 915  for (;;)
915      /* Otherwise, if PCRE_NOTEMPTY is set, fail if we have matched an empty      /* Otherwise, if PCRE_NOTEMPTY is set, fail if we have matched an empty
916      string - backtracking will then try other alternatives, if any. */      string - backtracking will then try other alternatives, if any. */
917    
918      if (md->notempty && eptr == md->start_match) RRETURN(MATCH_NOMATCH);      if (md->notempty && eptr == mstart) RRETURN(MATCH_NOMATCH);
919      md->end_match_ptr = eptr;          /* Record where we ended */      md->end_match_ptr = eptr;           /* Record where we ended */
920      md->end_offset_top = offset_top;   /* and how many extracts were taken */      md->end_offset_top = offset_top;    /* and how many extracts were taken */
921        md->start_match_ptr = mstart;       /* and the start (\K can modify) */
922      RRETURN(MATCH_MATCH);      RRETURN(MATCH_MATCH);
923    
924      /* Change option settings */      /* Change option settings */
# Line 802  for (;;) Line 939  for (;;)
939      case OP_ASSERTBACK:      case OP_ASSERTBACK:
940      do      do
941        {        {
942        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,
943            RM4);
944        if (rrc == MATCH_MATCH) break;        if (rrc == MATCH_MATCH) break;
945        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
946        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
947        }        }
948      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 828  for (;;) Line 966  for (;;)
966      case OP_ASSERTBACK_NOT:      case OP_ASSERTBACK_NOT:
967      do      do
968        {        {
969        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,
970            RM5);
971        if (rrc == MATCH_MATCH) RRETURN(MATCH_NOMATCH);        if (rrc == MATCH_MATCH) RRETURN(MATCH_NOMATCH);
972        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
973        ecode += GET(ecode,1);        ecode += GET(ecode,1);
974        }        }
975      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 854  for (;;) Line 993  for (;;)
993          {          {
994          eptr--;          eptr--;
995          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
996          BACKCHAR(eptr)          BACKCHAR(eptr);
997          }          }
998        }        }
999      else      else
# Line 885  for (;;) Line 1024  for (;;)
1024        cb.offset_vector    = md->offset_vector;        cb.offset_vector    = md->offset_vector;
1025        cb.subject          = (PCRE_SPTR)md->start_subject;        cb.subject          = (PCRE_SPTR)md->start_subject;
1026        cb.subject_length   = md->end_subject - md->start_subject;        cb.subject_length   = md->end_subject - md->start_subject;
1027        cb.start_match      = md->start_match - md->start_subject;        cb.start_match      = mstart - md->start_subject;
1028        cb.current_position = eptr - md->start_subject;        cb.current_position = eptr - md->start_subject;
1029        cb.pattern_position = GET(ecode, 2);        cb.pattern_position = GET(ecode, 2);
1030        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);
# Line 947  for (;;) Line 1086  for (;;)
1086    
1087        memcpy(new_recursive.offset_save, md->offset_vector,        memcpy(new_recursive.offset_save, md->offset_vector,
1088              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1089        new_recursive.save_start = md->start_match;        new_recursive.save_start = mstart;
1090        md->start_match = eptr;        mstart = eptr;
1091    
1092        /* OK, now we can do the recursion. For each top-level alternative we        /* OK, now we can do the recursion. For each top-level alternative we
1093        restore the offset and recursion data. */        restore the offset and recursion data. */
# Line 957  for (;;) Line 1096  for (;;)
1096        flags = (*callpat >= OP_SBRA)? match_cbegroup : 0;        flags = (*callpat >= OP_SBRA)? match_cbegroup : 0;
1097        do        do
1098          {          {
1099          RMATCH(rrc, eptr, callpat + _pcre_OP_lengths[*callpat], offset_top,          RMATCH(eptr, callpat + _pcre_OP_lengths[*callpat], offset_top,
1100            md, ims, eptrb, flags);            md, ims, eptrb, flags, RM6);
1101          if (rrc == MATCH_MATCH)          if (rrc == MATCH_MATCH)
1102            {            {
1103            DPRINTF(("Recursion matched\n"));            DPRINTF(("Recursion matched\n"));
# Line 967  for (;;) Line 1106  for (;;)
1106              (pcre_free)(new_recursive.offset_save);              (pcre_free)(new_recursive.offset_save);
1107            RRETURN(MATCH_MATCH);            RRETURN(MATCH_MATCH);
1108            }            }
1109          else if (rrc != MATCH_NOMATCH)          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
1110            {            {
1111            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1112              if (new_recursive.offset_save != stacksave)
1113                (pcre_free)(new_recursive.offset_save);
1114            RRETURN(rrc);            RRETURN(rrc);
1115            }            }
1116    
# Line 1001  for (;;) Line 1142  for (;;)
1142    
1143      do      do
1144        {        {
1145        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM7);
         eptrb, 0);  
1146        if (rrc == MATCH_MATCH) break;        if (rrc == MATCH_MATCH) break;
1147        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1148        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1149        }        }
1150      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 1047  for (;;) Line 1187  for (;;)
1187    
1188      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1189        {        {
1190        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM8);
1191        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1192        ecode = prev;        ecode = prev;
1193        flags = match_tail_recursed;        flags = 0;
1194        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1195        }        }
1196      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
1197        {        {
1198        RMATCH(rrc, eptr, prev, offset_top, md, ims, eptrb, match_cbegroup);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, match_cbegroup, RM9);
1199        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1200        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1201        flags = match_tail_recursed;        flags = 0;
1202        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1203        }        }
1204      /* Control never gets here */      /* Control never gets here */
# Line 1070  for (;;) Line 1210  for (;;)
1210      do ecode += GET(ecode,1); while (*ecode == OP_ALT);      do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1211      break;      break;
1212    
1213      /* BRAZERO and BRAMINZERO occur just before a bracket group, indicating      /* BRAZERO, BRAMINZERO and SKIPZERO occur just before a bracket group,
1214      that it may occur zero times. It may repeat infinitely, or not at all -      indicating that it may occur zero times. It may repeat infinitely, or not
1215      i.e. it could be ()* or ()? in the pattern. Brackets with fixed upper      at all - i.e. it could be ()* or ()? or even (){0} in the pattern. Brackets
1216      repeat limits are compiled as a number of copies, with the optional ones      with fixed upper repeat limits are compiled as a number of copies, with the
1217      preceded by BRAZERO or BRAMINZERO. */      optional ones preceded by BRAZERO or BRAMINZERO. */
1218    
1219      case OP_BRAZERO:      case OP_BRAZERO:
1220        {        {
1221        next = ecode+1;        next = ecode+1;
1222        RMATCH(rrc, eptr, next, offset_top, md, ims, eptrb, 0);        RMATCH(eptr, next, offset_top, md, ims, eptrb, 0, RM10);
1223        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1224        do next += GET(next,1); while (*next == OP_ALT);        do next += GET(next,1); while (*next == OP_ALT);
1225        ecode = next + 1 + LINK_SIZE;        ecode = next + 1 + LINK_SIZE;
# Line 1090  for (;;) Line 1230  for (;;)
1230        {        {
1231        next = ecode+1;        next = ecode+1;
1232        do next += GET(next, 1); while (*next == OP_ALT);        do next += GET(next, 1); while (*next == OP_ALT);
1233        RMATCH(rrc, eptr, next + 1+LINK_SIZE, offset_top, md, ims, eptrb, 0);        RMATCH(eptr, next + 1+LINK_SIZE, offset_top, md, ims, eptrb, 0, RM11);
1234        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1235        ecode++;        ecode++;
1236        }        }
1237      break;      break;
1238    
1239        case OP_SKIPZERO:
1240          {
1241          next = ecode+1;
1242          do next += GET(next,1); while (*next == OP_ALT);
1243          ecode = next + 1 + LINK_SIZE;
1244          }
1245        break;
1246    
1247      /* End of a group, repeated or non-repeating. */      /* End of a group, repeated or non-repeating. */
1248    
1249      case OP_KET:      case OP_KET:
# Line 1160  for (;;) Line 1308  for (;;)
1308          recursion_info *rec = md->recursive;          recursion_info *rec = md->recursive;
1309          DPRINTF(("Recursion (%d) succeeded - continuing\n", number));          DPRINTF(("Recursion (%d) succeeded - continuing\n", number));
1310          md->recursive = rec->prevrec;          md->recursive = rec->prevrec;
1311          md->start_match = rec->save_start;          mstart = rec->save_start;
1312          memcpy(md->offset_vector, rec->offset_save,          memcpy(md->offset_vector, rec->offset_save,
1313            rec->saved_max * sizeof(int));            rec->saved_max * sizeof(int));
1314          ecode = rec->after_call;          ecode = rec->after_call;
# Line 1189  for (;;) Line 1337  for (;;)
1337    
1338      /* The repeating kets try the rest of the pattern or restart from the      /* The repeating kets try the rest of the pattern or restart from the
1339      preceding bracket, in the appropriate order. In the second case, we can use      preceding bracket, in the appropriate order. In the second case, we can use
1340      tail recursion to avoid using another stack frame. */      tail recursion to avoid using another stack frame, unless we have an
1341        unlimited repeat of a group that can match an empty string. */
1342    
1343      flags = (*prev >= OP_SBRA)? match_cbegroup : 0;      flags = (*prev >= OP_SBRA)? match_cbegroup : 0;
1344    
1345      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1346        {        {
1347        RMATCH(rrc, eptr, ecode + 1+LINK_SIZE, offset_top, md, ims, eptrb, 0);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM12);
1348        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1349          if (flags != 0)    /* Could match an empty string */
1350            {
1351            RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM50);
1352            RRETURN(rrc);
1353            }
1354        ecode = prev;        ecode = prev;
       flags |= match_tail_recursed;  
1355        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1356        }        }
1357      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
1358        {        {
1359        RMATCH(rrc, eptr, prev, offset_top, md, ims, eptrb, flags);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM13);
1360        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1361        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1362        flags = match_tail_recursed;        flags = 0;
1363        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1364        }        }
1365      /* Control never gets here */      /* Control never gets here */
# Line 1239  for (;;) Line 1392  for (;;)
1392      ecode++;      ecode++;
1393      break;      break;
1394    
1395        /* Reset the start of match point */
1396    
1397        case OP_SET_SOM:
1398        mstart = eptr;
1399        ecode++;
1400        break;
1401    
1402      /* Assert before internal newline if multiline, or before a terminating      /* Assert before internal newline if multiline, or before a terminating
1403      newline unless endonly is set, else end of subject unless noteol is set. */      newline unless endonly is set, else end of subject unless noteol is set. */
1404    
# Line 1331  for (;;) Line 1491  for (;;)
1491      /* Match a single character type; inline for speed */      /* Match a single character type; inline for speed */
1492    
1493      case OP_ANY:      case OP_ANY:
1494      if ((ims & PCRE_DOTALL) == 0)      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
1495        {      /* Fall through */
1496        if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);  
1497        }      case OP_ALLANY:
1498      if (eptr++ >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr++ >= md->end_subject) RRETURN(MATCH_NOMATCH);
1499      if (utf8)      if (utf8) while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
       while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  
1500      ecode++;      ecode++;
1501      break;      break;
1502    
# Line 1436  for (;;) Line 1595  for (;;)
1595        case 0x000d:        case 0x000d:
1596        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
1597        break;        break;
1598    
1599        case 0x000a:        case 0x000a:
1600          break;
1601    
1602        case 0x000b:        case 0x000b:
1603        case 0x000c:        case 0x000c:
1604        case 0x0085:        case 0x0085:
1605        case 0x2028:        case 0x2028:
1606        case 0x2029:        case 0x2029:
1607          if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
1608          break;
1609          }
1610        ecode++;
1611        break;
1612    
1613        case OP_NOT_HSPACE:
1614        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1615        GETCHARINCTEST(c, eptr);
1616        switch(c)
1617          {
1618          default: break;
1619          case 0x09:      /* HT */
1620          case 0x20:      /* SPACE */
1621          case 0xa0:      /* NBSP */
1622          case 0x1680:    /* OGHAM SPACE MARK */
1623          case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
1624          case 0x2000:    /* EN QUAD */
1625          case 0x2001:    /* EM QUAD */
1626          case 0x2002:    /* EN SPACE */
1627          case 0x2003:    /* EM SPACE */
1628          case 0x2004:    /* THREE-PER-EM SPACE */
1629          case 0x2005:    /* FOUR-PER-EM SPACE */
1630          case 0x2006:    /* SIX-PER-EM SPACE */
1631          case 0x2007:    /* FIGURE SPACE */
1632          case 0x2008:    /* PUNCTUATION SPACE */
1633          case 0x2009:    /* THIN SPACE */
1634          case 0x200A:    /* HAIR SPACE */
1635          case 0x202f:    /* NARROW NO-BREAK SPACE */
1636          case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
1637          case 0x3000:    /* IDEOGRAPHIC SPACE */
1638          RRETURN(MATCH_NOMATCH);
1639          }
1640        ecode++;
1641        break;
1642    
1643        case OP_HSPACE:
1644        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1645        GETCHARINCTEST(c, eptr);
1646        switch(c)
1647          {
1648          default: RRETURN(MATCH_NOMATCH);
1649          case 0x09:      /* HT */
1650          case 0x20:      /* SPACE */
1651          case 0xa0:      /* NBSP */
1652          case 0x1680:    /* OGHAM SPACE MARK */
1653          case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
1654          case 0x2000:    /* EN QUAD */
1655          case 0x2001:    /* EM QUAD */
1656          case 0x2002:    /* EN SPACE */
1657          case 0x2003:    /* EM SPACE */
1658          case 0x2004:    /* THREE-PER-EM SPACE */
1659          case 0x2005:    /* FOUR-PER-EM SPACE */
1660          case 0x2006:    /* SIX-PER-EM SPACE */
1661          case 0x2007:    /* FIGURE SPACE */
1662          case 0x2008:    /* PUNCTUATION SPACE */
1663          case 0x2009:    /* THIN SPACE */
1664          case 0x200A:    /* HAIR SPACE */
1665          case 0x202f:    /* NARROW NO-BREAK SPACE */
1666          case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
1667          case 0x3000:    /* IDEOGRAPHIC SPACE */
1668          break;
1669          }
1670        ecode++;
1671        break;
1672    
1673        case OP_NOT_VSPACE:
1674        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1675        GETCHARINCTEST(c, eptr);
1676        switch(c)
1677          {
1678          default: break;
1679          case 0x0a:      /* LF */
1680          case 0x0b:      /* VT */
1681          case 0x0c:      /* FF */
1682          case 0x0d:      /* CR */
1683          case 0x85:      /* NEL */
1684          case 0x2028:    /* LINE SEPARATOR */
1685          case 0x2029:    /* PARAGRAPH SEPARATOR */
1686          RRETURN(MATCH_NOMATCH);
1687          }
1688        ecode++;
1689        break;
1690    
1691        case OP_VSPACE:
1692        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1693        GETCHARINCTEST(c, eptr);
1694        switch(c)
1695          {
1696          default: RRETURN(MATCH_NOMATCH);
1697          case 0x0a:      /* LF */
1698          case 0x0b:      /* VT */
1699          case 0x0c:      /* FF */
1700          case 0x0d:      /* CR */
1701          case 0x85:      /* NEL */
1702          case 0x2028:    /* LINE SEPARATOR */
1703          case 0x2029:    /* PARAGRAPH SEPARATOR */
1704        break;        break;
1705        }        }
1706      ecode++;      ecode++;
# Line 1456  for (;;) Line 1715  for (;;)
1715      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1716      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1717        {        {
1718        int chartype, script;        const ucd_record *prop = GET_UCD(c);
       int category = _pcre_ucp_findprop(c, &chartype, &script);  
1719    
1720        switch(ecode[1])        switch(ecode[1])
1721          {          {
# Line 1466  for (;;) Line 1724  for (;;)
1724          break;          break;
1725    
1726          case PT_LAMP:          case PT_LAMP:
1727          if ((chartype == ucp_Lu ||          if ((prop->chartype == ucp_Lu ||
1728               chartype == ucp_Ll ||               prop->chartype == ucp_Ll ||
1729               chartype == ucp_Lt) == (op == OP_NOTPROP))               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))
1730            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1731           break;           break;
1732    
1733          case PT_GC:          case PT_GC:
1734          if ((ecode[2] != category) == (op == OP_PROP))          if ((ecode[2] != _pcre_ucp_gentype[prop->chartype]) == (op == OP_PROP))
1735            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1736          break;          break;
1737    
1738          case PT_PC:          case PT_PC:
1739          if ((ecode[2] != chartype) == (op == OP_PROP))          if ((ecode[2] != prop->chartype) == (op == OP_PROP))
1740            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1741          break;          break;
1742    
1743          case PT_SC:          case PT_SC:
1744          if ((ecode[2] != script) == (op == OP_PROP))          if ((ecode[2] != prop->script) == (op == OP_PROP))
1745            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1746          break;          break;
1747    
# Line 1502  for (;;) Line 1760  for (;;)
1760      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1761      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1762        {        {
1763        int chartype, script;        int category = UCD_CATEGORY(c);
       int category = _pcre_ucp_findprop(c, &chartype, &script);  
1764        if (category == ucp_M) RRETURN(MATCH_NOMATCH);        if (category == ucp_M) RRETURN(MATCH_NOMATCH);
1765        while (eptr < md->end_subject)        while (eptr < md->end_subject)
1766          {          {
# Line 1512  for (;;) Line 1769  for (;;)
1769            {            {
1770            GETCHARLEN(c, eptr, len);            GETCHARLEN(c, eptr, len);
1771            }            }
1772          category = _pcre_ucp_findprop(c, &chartype, &script);          category = UCD_CATEGORY(c);
1773          if (category != ucp_M) break;          if (category != ucp_M) break;
1774          eptr += len;          eptr += len;
1775          }          }
# Line 1533  for (;;) Line 1790  for (;;)
1790      case OP_REF:      case OP_REF:
1791        {        {
1792        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
1793        ecode += 3;                                 /* Advance past item */        ecode += 3;
1794    
1795        /* If the reference is unset, set the length to be longer than the amount        /* If the reference is unset, there are two possibilities:
1796        of subject left; this ensures that every attempt at a match fails. We  
1797        can't just fail here, because of the possibility of quantifiers with zero        (a) In the default, Perl-compatible state, set the length to be longer
1798        minima. */        than the amount of subject left; this ensures that every attempt at a
1799          match fails. We can't just fail here, because of the possibility of
1800        length = (offset >= offset_top || md->offset_vector[offset] < 0)?        quantifiers with zero minima.
1801          md->end_subject - eptr + 1 :  
1802          md->offset_vector[offset+1] - md->offset_vector[offset];        (b) If the JavaScript compatibility flag is set, set the length to zero
1803          so that the back reference matches an empty string.
1804    
1805          Otherwise, set the length to the length of what was matched by the
1806          referenced subpattern. */
1807    
1808          if (offset >= offset_top || md->offset_vector[offset] < 0)
1809            length = (md->jscript_compat)? 0 : md->end_subject - eptr + 1;
1810          else
1811            length = md->offset_vector[offset+1] - md->offset_vector[offset];
1812    
1813        /* Set up for repetition, or handle the non-repeated case */        /* Set up for repetition, or handle the non-repeated case */
1814    
# Line 1602  for (;;) Line 1868  for (;;)
1868          {          {
1869          for (fi = min;; fi++)          for (fi = min;; fi++)
1870            {            {
1871            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM14);
1872            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1873            if (fi >= max || !match_ref(offset, eptr, length, md, ims))            if (fi >= max || !match_ref(offset, eptr, length, md, ims))
1874              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
# Line 1623  for (;;) Line 1889  for (;;)
1889            }            }
1890          while (eptr >= pp)          while (eptr >= pp)
1891            {            {
1892            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM15);
1893            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1894            eptr -= length;            eptr -= length;
1895            }            }
# Line 1728  for (;;) Line 1994  for (;;)
1994            {            {
1995            for (fi = min;; fi++)            for (fi = min;; fi++)
1996              {              {
1997              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM16);
1998              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1999              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2000              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 1748  for (;;) Line 2014  for (;;)
2014            {            {
2015            for (fi = min;; fi++)            for (fi = min;; fi++)
2016              {              {
2017              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM17);
2018              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2019              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2020              c = *eptr++;              c = *eptr++;
# Line 1785  for (;;) Line 2051  for (;;)
2051              }              }
2052            for (;;)            for (;;)
2053              {              {
2054              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM18);
2055              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2056              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
2057              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 1804  for (;;) Line 2070  for (;;)
2070              }              }
2071            while (eptr >= pp)            while (eptr >= pp)
2072              {              {
2073              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM19);
2074              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2075              eptr--;              eptr--;
2076              }              }
# Line 1817  for (;;) Line 2083  for (;;)
2083    
2084    
2085      /* Match an extended character class. This opcode is encountered only      /* Match an extended character class. This opcode is encountered only
2086      in UTF-8 mode, because that's the only time it is compiled. */      when UTF-8 mode mode is supported. Nevertheless, we may not be in UTF-8
2087        mode, because Unicode properties are supported in non-UTF-8 mode. */
2088    
2089  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
2090      case OP_XCLASS:      case OP_XCLASS:
# Line 1859  for (;;) Line 2126  for (;;)
2126        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
2127          {          {
2128          if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);          if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2129          GETCHARINC(c, eptr);          GETCHARINCTEST(c, eptr);
2130          if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);          if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);
2131          }          }
2132    
# Line 1875  for (;;) Line 2142  for (;;)
2142          {          {
2143          for (fi = min;; fi++)          for (fi = min;; fi++)
2144            {            {
2145            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM20);
2146            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2147            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2148            GETCHARINC(c, eptr);            GETCHARINCTEST(c, eptr);
2149            if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);            if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);
2150            }            }
2151          /* Control never gets here */          /* Control never gets here */
# Line 1893  for (;;) Line 2160  for (;;)
2160            {            {
2161            int len = 1;            int len = 1;
2162            if (eptr >= md->end_subject) break;            if (eptr >= md->end_subject) break;
2163            GETCHARLEN(c, eptr, len);            GETCHARLENTEST(c, eptr, len);
2164            if (!_pcre_xclass(c, data)) break;            if (!_pcre_xclass(c, data)) break;
2165            eptr += len;            eptr += len;
2166            }            }
2167          for(;;)          for(;;)
2168            {            {
2169            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM21);
2170            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2171            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
2172            BACKCHAR(eptr)            if (utf8) BACKCHAR(eptr);
2173            }            }
2174          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2175          }          }
# Line 1968  for (;;) Line 2235  for (;;)
2235          if (fc != dc)          if (fc != dc)
2236            {            {
2237  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2238            if (dc != _pcre_ucp_othercase(fc))            if (dc != UCD_OTHERCASE(fc))
2239  #endif  #endif
2240              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2241            }            }
# Line 2059  for (;;) Line 2326  for (;;)
2326  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2327          unsigned int othercase;          unsigned int othercase;
2328          if ((ims & PCRE_CASELESS) != 0 &&          if ((ims & PCRE_CASELESS) != 0 &&
2329              (othercase = _pcre_ucp_othercase(fc)) != NOTACHAR)              (othercase = UCD_OTHERCASE(fc)) != fc)
2330            oclength = _pcre_ord2utf8(othercase, occhars);            oclength = _pcre_ord2utf8(othercase, occhars);
2331          else oclength = 0;          else oclength = 0;
2332  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
# Line 2086  for (;;) Line 2353  for (;;)
2353            {            {
2354            for (fi = min;; fi++)            for (fi = min;; fi++)
2355              {              {
2356              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM22);
2357              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2358              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2359              if (memcmp(eptr, charptr, length) == 0) eptr += length;              if (memcmp(eptr, charptr, length) == 0) eptr += length;
# Line 2127  for (;;) Line 2394  for (;;)
2394            if (possessive) continue;            if (possessive) continue;
2395            for(;;)            for(;;)
2396             {             {
2397             RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);             RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM23);
2398             if (rrc != MATCH_NOMATCH) RRETURN(rrc);             if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2399             if (eptr == pp) RRETURN(MATCH_NOMATCH);             if (eptr == pp) RRETURN(MATCH_NOMATCH);
2400  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
# Line 2176  for (;;) Line 2443  for (;;)
2443          {          {
2444          for (fi = min;; fi++)          for (fi = min;; fi++)
2445            {            {
2446            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM24);
2447            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2448            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
2449                fc != md->lcc[*eptr++])                fc != md->lcc[*eptr++])
# Line 2195  for (;;) Line 2462  for (;;)
2462          if (possessive) continue;          if (possessive) continue;
2463          while (eptr >= pp)          while (eptr >= pp)
2464            {            {
2465            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM25);
2466            eptr--;            eptr--;
2467            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2468            }            }
# Line 2214  for (;;) Line 2481  for (;;)
2481          {          {
2482          for (fi = min;; fi++)          for (fi = min;; fi++)
2483            {            {
2484            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM26);
2485            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2486            if (fi >= max || eptr >= md->end_subject || fc != *eptr++)            if (fi >= max || eptr >= md->end_subject || fc != *eptr++)
2487              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
# Line 2232  for (;;) Line 2499  for (;;)
2499          if (possessive) continue;          if (possessive) continue;
2500          while (eptr >= pp)          while (eptr >= pp)
2501            {            {
2502            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM27);
2503            eptr--;            eptr--;
2504            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2505            }            }
# Line 2377  for (;;) Line 2644  for (;;)
2644            register unsigned int d;            register unsigned int d;
2645            for (fi = min;; fi++)            for (fi = min;; fi++)
2646              {              {
2647              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM28);
2648              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2649                if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2650              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
2651              if (d < 256) d = md->lcc[d];              if (d < 256) d = md->lcc[d];
2652              if (fi >= max || eptr >= md->end_subject || fc == d)              if (fc == d) RRETURN(MATCH_NOMATCH);
2653                RRETURN(MATCH_NOMATCH);  
2654              }              }
2655            }            }
2656          else          else
# Line 2391  for (;;) Line 2659  for (;;)
2659            {            {
2660            for (fi = min;; fi++)            for (fi = min;; fi++)
2661              {              {
2662              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM29);
2663              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2664              if (fi >= max || eptr >= md->end_subject || fc == md->lcc[*eptr++])              if (fi >= max || eptr >= md->end_subject || fc == md->lcc[*eptr++])
2665                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2423  for (;;) Line 2691  for (;;)
2691          if (possessive) continue;          if (possessive) continue;
2692          for(;;)          for(;;)
2693              {              {
2694              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM30);
2695              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2696              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
2697              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 2441  for (;;) Line 2709  for (;;)
2709            if (possessive) continue;            if (possessive) continue;
2710            while (eptr >= pp)            while (eptr >= pp)
2711              {              {
2712              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM31);
2713              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2714              eptr--;              eptr--;
2715              }              }
# Line 2486  for (;;) Line 2754  for (;;)
2754            register unsigned int d;            register unsigned int d;
2755            for (fi = min;; fi++)            for (fi = min;; fi++)
2756              {              {
2757              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM32);
2758              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2759                if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2760              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
2761              if (fi >= max || eptr >= md->end_subject || fc == d)              if (fc == d) RRETURN(MATCH_NOMATCH);
               RRETURN(MATCH_NOMATCH);  
2762              }              }
2763            }            }
2764          else          else
# Line 2499  for (;;) Line 2767  for (;;)
2767            {            {
2768            for (fi = min;; fi++)            for (fi = min;; fi++)
2769              {              {
2770              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM33);
2771              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2772              if (fi >= max || eptr >= md->end_subject || fc == *eptr++)              if (fi >= max || eptr >= md->end_subject || fc == *eptr++)
2773                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2530  for (;;) Line 2798  for (;;)
2798            if (possessive) continue;            if (possessive) continue;
2799            for(;;)            for(;;)
2800              {              {
2801              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM34);
2802              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2803              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
2804              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 2548  for (;;) Line 2816  for (;;)
2816            if (possessive) continue;            if (possessive) continue;
2817            while (eptr >= pp)            while (eptr >= pp)
2818              {              {
2819              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM35);
2820              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2821              eptr--;              eptr--;
2822              }              }
# Line 2655  for (;;) Line 2923  for (;;)
2923            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2924              {              {
2925              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2926              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2927              }              }
2928            break;            break;
2929    
# Line 2663  for (;;) Line 2931  for (;;)
2931            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2932              {              {
2933              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2934              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2935              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
2936              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
2937                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
2938                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
# Line 2676  for (;;) Line 2944  for (;;)
2944            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2945              {              {
2946              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2947              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2948              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
2949              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
2950                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2951              }              }
# Line 2687  for (;;) Line 2955  for (;;)
2955            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2956              {              {
2957              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2958              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2959              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
2960              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
2961                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2962              }              }
# Line 2698  for (;;) Line 2966  for (;;)
2966            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2967              {              {
2968              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2969              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2970              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_script = UCD_SCRIPT(c);
2971              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
2972                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2973              }              }
# Line 2718  for (;;) Line 2986  for (;;)
2986          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2987            {            {
2988            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
2989            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
2990            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);
2991            while (eptr < md->end_subject)            while (eptr < md->end_subject)
2992              {              {
# Line 2727  for (;;) Line 2995  for (;;)
2995                {                {
2996                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
2997                }                }
2998              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
2999              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
3000              eptr += len;              eptr += len;
3001              }              }
# Line 2745  for (;;) Line 3013  for (;;)
3013          case OP_ANY:          case OP_ANY:
3014          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3015            {            {
3016            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject || IS_NEWLINE(eptr))
                ((ims & PCRE_DOTALL) == 0 && IS_NEWLINE(eptr)))  
3017              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3018            eptr++;            eptr++;
3019            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
3020            }            }
3021          break;          break;
3022    
3023            case OP_ALLANY:
3024            for (i = 1; i <= min; i++)
3025              {
3026              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3027              eptr++;
3028              while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
3029              }
3030            break;
3031    
3032          case OP_ANYBYTE:          case OP_ANYBYTE:
3033          eptr += min;          eptr += min;
3034          break;          break;
# Line 2768  for (;;) Line 3044  for (;;)
3044              case 0x000d:              case 0x000d:
3045              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3046              break;              break;
3047    
3048              case 0x000a:              case 0x000a:
3049                break;
3050    
3051              case 0x000b:              case 0x000b:
3052              case 0x000c:              case 0x000c:
3053              case 0x0085:              case 0x0085:
3054              case 0x2028:              case 0x2028:
3055              case 0x2029:              case 0x2029:
3056                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3057                break;
3058                }
3059              }
3060            break;
3061    
3062            case OP_NOT_HSPACE:
3063            for (i = 1; i <= min; i++)
3064              {
3065              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3066              GETCHARINC(c, eptr);
3067              switch(c)
3068                {
3069                default: break;
3070                case 0x09:      /* HT */
3071                case 0x20:      /* SPACE */
3072                case 0xa0:      /* NBSP */
3073                case 0x1680:    /* OGHAM SPACE MARK */
3074                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3075                case 0x2000:    /* EN QUAD */
3076                case 0x2001:    /* EM QUAD */
3077                case 0x2002:    /* EN SPACE */
3078                case 0x2003:    /* EM SPACE */
3079                case 0x2004:    /* THREE-PER-EM SPACE */
3080                case 0x2005:    /* FOUR-PER-EM SPACE */
3081                case 0x2006:    /* SIX-PER-EM SPACE */
3082                case 0x2007:    /* FIGURE SPACE */
3083                case 0x2008:    /* PUNCTUATION SPACE */
3084                case 0x2009:    /* THIN SPACE */
3085                case 0x200A:    /* HAIR SPACE */
3086                case 0x202f:    /* NARROW NO-BREAK SPACE */
3087                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3088                case 0x3000:    /* IDEOGRAPHIC SPACE */
3089                RRETURN(MATCH_NOMATCH);
3090                }
3091              }
3092            break;
3093    
3094            case OP_HSPACE:
3095            for (i = 1; i <= min; i++)
3096              {
3097              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3098              GETCHARINC(c, eptr);
3099              switch(c)
3100                {
3101                default: RRETURN(MATCH_NOMATCH);
3102                case 0x09:      /* HT */
3103                case 0x20:      /* SPACE */
3104                case 0xa0:      /* NBSP */
3105                case 0x1680:    /* OGHAM SPACE MARK */
3106                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3107                case 0x2000:    /* EN QUAD */
3108                case 0x2001:    /* EM QUAD */
3109                case 0x2002:    /* EN SPACE */
3110                case 0x2003:    /* EM SPACE */
3111                case 0x2004:    /* THREE-PER-EM SPACE */
3112                case 0x2005:    /* FOUR-PER-EM SPACE */
3113                case 0x2006:    /* SIX-PER-EM SPACE */
3114                case 0x2007:    /* FIGURE SPACE */
3115                case 0x2008:    /* PUNCTUATION SPACE */
3116                case 0x2009:    /* THIN SPACE */
3117                case 0x200A:    /* HAIR SPACE */
3118                case 0x202f:    /* NARROW NO-BREAK SPACE */
3119                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3120                case 0x3000:    /* IDEOGRAPHIC SPACE */
3121                break;
3122                }
3123              }
3124            break;
3125    
3126            case OP_NOT_VSPACE:
3127            for (i = 1; i <= min; i++)
3128              {
3129              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3130              GETCHARINC(c, eptr);
3131              switch(c)
3132                {
3133                default: break;
3134                case 0x0a:      /* LF */
3135                case 0x0b:      /* VT */
3136                case 0x0c:      /* FF */
3137                case 0x0d:      /* CR */
3138                case 0x85:      /* NEL */
3139                case 0x2028:    /* LINE SEPARATOR */
3140                case 0x2029:    /* PARAGRAPH SEPARATOR */
3141                RRETURN(MATCH_NOMATCH);
3142                }
3143              }
3144            break;
3145    
3146            case OP_VSPACE:
3147            for (i = 1; i <= min; i++)
3148              {
3149              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3150              GETCHARINC(c, eptr);
3151              switch(c)
3152                {
3153                default: RRETURN(MATCH_NOMATCH);
3154                case 0x0a:      /* LF */
3155                case 0x0b:      /* VT */
3156                case 0x0c:      /* FF */
3157                case 0x0d:      /* CR */
3158                case 0x85:      /* NEL */
3159                case 0x2028:    /* LINE SEPARATOR */
3160                case 0x2029:    /* PARAGRAPH SEPARATOR */
3161              break;              break;
3162              }              }
3163            }            }
# Line 2803  for (;;) Line 3187  for (;;)
3187          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3188            {            {
3189            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject ||
3190               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_space) != 0))               (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0))
3191              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3192            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3193            }            }
3194          break;          break;
3195    
# Line 2823  for (;;) Line 3207  for (;;)
3207          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3208            {            {
3209            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject ||
3210               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_word) != 0))               (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0))
3211              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3212            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3213            }            }
3214          break;          break;
3215    
# Line 2853  for (;;) Line 3237  for (;;)
3237        switch(ctype)        switch(ctype)
3238          {          {
3239          case OP_ANY:          case OP_ANY:
3240          if ((ims & PCRE_DOTALL) == 0)          for (i = 1; i <= min; i++)
3241            {            {
3242            for (i = 1; i <= min; i++)            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
3243              {            eptr++;
             if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);  
             eptr++;  
             }  
3244            }            }
3245          else eptr += min;          break;
3246    
3247            case OP_ALLANY:
3248            eptr += min;
3249          break;          break;
3250    
3251          case OP_ANYBYTE:          case OP_ANYBYTE:
# Line 2882  for (;;) Line 3266  for (;;)
3266              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3267              break;              break;
3268              case 0x000a:              case 0x000a:
3269                break;
3270    
3271              case 0x000b:              case 0x000b:
3272              case 0x000c:              case 0x000c:
3273              case 0x0085:              case 0x0085:
3274                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3275                break;
3276                }
3277              }
3278            break;
3279    
3280            case OP_NOT_HSPACE:
3281            for (i = 1; i <= min; i++)
3282              {
3283              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3284              switch(*eptr++)
3285                {
3286                default: break;
3287                case 0x09:      /* HT */
3288                case 0x20:      /* SPACE */
3289                case 0xa0:      /* NBSP */
3290                RRETURN(MATCH_NOMATCH);
3291                }
3292              }
3293            break;
3294    
3295            case OP_HSPACE:
3296            for (i = 1; i <= min; i++)
3297              {
3298              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3299              switch(*eptr++)
3300                {
3301                default: RRETURN(MATCH_NOMATCH);
3302                case 0x09:      /* HT */
3303                case 0x20:      /* SPACE */
3304                case 0xa0:      /* NBSP */
3305                break;
3306                }
3307              }
3308            break;
3309    
3310            case OP_NOT_VSPACE:
3311            for (i = 1; i <= min; i++)
3312              {
3313              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3314              switch(*eptr++)
3315                {
3316                default: break;
3317                case 0x0a:      /* LF */
3318                case 0x0b:      /* VT */
3319                case 0x0c:      /* FF */
3320                case 0x0d:      /* CR */
3321                case 0x85:      /* NEL */
3322                RRETURN(MATCH_NOMATCH);
3323                }
3324              }
3325            break;
3326    
3327            case OP_VSPACE:
3328            for (i = 1; i <= min; i++)
3329              {
3330              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3331              switch(*eptr++)
3332                {
3333                default: RRETURN(MATCH_NOMATCH);
3334                case 0x0a:      /* LF */
3335                case 0x0b:      /* VT */
3336                case 0x0c:      /* FF */
3337                case 0x0d:      /* CR */
3338                case 0x85:      /* NEL */
3339              break;              break;
3340              }              }
3341            }            }
# Line 2945  for (;;) Line 3396  for (;;)
3396            case PT_ANY:            case PT_ANY:
3397            for (fi = min;; fi++)            for (fi = min;; fi++)
3398              {              {
3399              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM36);
3400              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3401              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3402              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 2956  for (;;) Line 3407  for (;;)
3407            case PT_LAMP:            case PT_LAMP:
3408            for (fi = min;; fi++)            for (fi = min;; fi++)
3409              {              {
3410              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM37);
3411              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3412              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3413              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3414              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3415              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
3416                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
3417                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
# Line 2971  for (;;) Line 3422  for (;;)
3422            case PT_GC:            case PT_GC:
3423            for (fi = min;; fi++)            for (fi = min;; fi++)
3424              {              {
3425              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM38);
3426              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3427              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3428              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3429              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3430              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
3431                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3432              }              }
# Line 2984  for (;;) Line 3435  for (;;)
3435            case PT_PC:            case PT_PC:
3436            for (fi = min;; fi++)            for (fi = min;; fi++)
3437              {              {
3438              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM39);
3439              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3440              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3441              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3442              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3443              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
3444                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3445              }              }
# Line 2997  for (;;) Line 3448  for (;;)
3448            case PT_SC:            case PT_SC:
3449            for (fi = min;; fi++)            for (fi = min;; fi++)
3450              {              {
3451              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM40);
3452              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3453              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3454              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3455              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_script = UCD_SCRIPT(c);
3456              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
3457                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3458              }              }
# Line 3019  for (;;) Line 3470  for (;;)
3470          {          {
3471          for (fi = min;; fi++)          for (fi = min;; fi++)
3472            {            {
3473            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM41);
3474            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3475            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3476            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3477            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
3478            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);
3479            while (eptr < md->end_subject)            while (eptr < md->end_subject)
3480              {              {
# Line 3032  for (;;) Line 3483  for (;;)
3483                {                {
3484                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
3485                }                }
3486              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3487              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
3488              eptr += len;              eptr += len;
3489              }              }
# Line 3048  for (;;) Line 3499  for (;;)
3499          {          {
3500          for (fi = min;; fi++)          for (fi = min;; fi++)
3501            {            {
3502            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM42);
3503            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3504            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
3505                 (ctype == OP_ANY && (ims & PCRE_DOTALL) == 0 &&                 (ctype == OP_ANY && IS_NEWLINE(eptr)))
                 IS_NEWLINE(eptr)))  
3506              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3507    
3508            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
3509            switch(ctype)            switch(ctype)
3510              {              {
3511              case OP_ANY:        /* This is the DOTALL case */              case OP_ANY:        /* This is the non-NL case */
3512              break;              case OP_ALLANY:
   
3513              case OP_ANYBYTE:              case OP_ANYBYTE:
3514              break;              break;
3515    
# Line 3072  for (;;) Line 3521  for (;;)
3521                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3522                break;                break;
3523                case 0x000a:                case 0x000a:
3524                  break;
3525    
3526                case 0x000b:                case 0x000b:
3527                case 0x000c:                case 0x000c:
3528                case 0x0085:                case 0x0085:
3529                case 0x2028:                case 0x2028:
3530                case 0x2029:                case 0x2029:
3531                  if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3532                  break;
3533                  }
3534                break;
3535    
3536                case OP_NOT_HSPACE:
3537                switch(c)
3538                  {
3539                  default: break;
3540                  case 0x09:      /* HT */
3541                  case 0x20:      /* SPACE */
3542                  case 0xa0:      /* NBSP */
3543                  case 0x1680:    /* OGHAM SPACE MARK */
3544                  case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3545                  case 0x2000:    /* EN QUAD */
3546                  case 0x2001:    /* EM QUAD */
3547                  case 0x2002:    /* EN SPACE */
3548                  case 0x2003:    /* EM SPACE */
3549                  case 0x2004:    /* THREE-PER-EM SPACE */
3550                  case 0x2005:    /* FOUR-PER-EM SPACE */
3551                  case 0x2006:    /* SIX-PER-EM SPACE */
3552                  case 0x2007:    /* FIGURE SPACE */
3553                  case 0x2008:    /* PUNCTUATION SPACE */
3554                  case 0x2009:    /* THIN SPACE */
3555                  case 0x200A:    /* HAIR SPACE */
3556                  case 0x202f:    /* NARROW NO-BREAK SPACE */
3557                  case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3558                  case 0x3000:    /* IDEOGRAPHIC SPACE */
3559                  RRETURN(MATCH_NOMATCH);
3560                  }
3561                break;
3562    
3563                case OP_HSPACE:
3564                switch(c)
3565                  {
3566                  default: RRETURN(MATCH_NOMATCH);
3567                  case 0x09:      /* HT */
3568                  case 0x20:      /* SPACE */
3569                  case 0xa0:      /* NBSP */
3570                  case 0x1680:    /* OGHAM SPACE MARK */
3571                  case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3572                  case 0x2000:    /* EN QUAD */
3573                  case 0x2001:    /* EM QUAD */
3574                  case 0x2002:    /* EN SPACE */
3575                  case 0x2003:    /* EM SPACE */
3576                  case 0x2004:    /* THREE-PER-EM SPACE */
3577                  case 0x2005:    /* FOUR-PER-EM SPACE */
3578                  case 0x2006:    /* SIX-PER-EM SPACE */
3579                  case 0x2007:    /* FIGURE SPACE */
3580                  case 0x2008:    /* PUNCTUATION SPACE */
3581                  case 0x2009:    /* THIN SPACE */
3582                  case 0x200A:    /* HAIR SPACE */
3583                  case 0x202f:    /* NARROW NO-BREAK SPACE */
3584                  case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3585                  case 0x3000:    /* IDEOGRAPHIC SPACE */
3586                  break;
3587                  }
3588                break;
3589    
3590                case OP_NOT_VSPACE:
3591                switch(c)
3592                  {
3593                  default: break;
3594                  case 0x0a:      /* LF */
3595                  case 0x0b:      /* VT */
3596                  case 0x0c:      /* FF */
3597                  case 0x0d:      /* CR */
3598                  case 0x85:      /* NEL */
3599                  case 0x2028:    /* LINE SEPARATOR */
3600                  case 0x2029:    /* PARAGRAPH SEPARATOR */
3601                  RRETURN(MATCH_NOMATCH);
3602                  }
3603                break;
3604    
3605                case OP_VSPACE:
3606                switch(c)
3607                  {
3608                  default: RRETURN(MATCH_NOMATCH);
3609                  case 0x0a:      /* LF */
3610                  case 0x0b:      /* VT */
3611                  case 0x0c:      /* FF */
3612                  case 0x0d:      /* CR */
3613                  case 0x85:      /* NEL */
3614                  case 0x2028:    /* LINE SEPARATOR */
3615                  case 0x2029:    /* PARAGRAPH SEPARATOR */
3616                break;                break;
3617                }                }
3618              break;              break;
# Line 3122  for (;;) Line 3658  for (;;)
3658          {          {
3659          for (fi = min;; fi++)          for (fi = min;; fi++)
3660            {            {
3661            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM43);
3662            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3663            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
3664                 ((ims & PCRE_DOTALL) == 0 && IS_NEWLINE(eptr)))                 (ctype == OP_ANY && IS_NEWLINE(eptr)))
3665              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3666    
3667            c = *eptr++;            c = *eptr++;
3668            switch(ctype)            switch(ctype)
3669              {              {
3670              case OP_ANY:   /* This is the DOTALL case */              case OP_ANY:     /* This is the non-NL case */
3671              break;              case OP_ALLANY:
   
3672              case OP_ANYBYTE:              case OP_ANYBYTE:
3673              break;              break;
3674    
# Line 3144  for (;;) Line 3679  for (;;)
3679                case 0x000d:                case 0x000d:
3680                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3681                break;                break;
3682    
3683                case 0x000a:                case 0x000a:
3684                  break;
3685    
3686                case 0x000b:                case 0x000b:
3687                case 0x000c:                case 0x000c:
3688                case 0x0085:                case 0x0085:
3689                  if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3690                  break;
3691                  }
3692                break;
3693    
3694                case OP_NOT_HSPACE:
3695                switch(c)
3696                  {
3697                  default: break;
3698                  case 0x09:      /* HT */
3699                  case 0x20:      /* SPACE */
3700                  case 0xa0:      /* NBSP */
3701                  RRETURN(MATCH_NOMATCH);
3702                  }
3703                break;
3704    
3705                case OP_HSPACE:
3706                switch(c)
3707                  {
3708                  default: RRETURN(MATCH_NOMATCH);
3709                  case 0x09:      /* HT */
3710                  case 0x20:      /* SPACE */
3711                  case 0xa0:      /* NBSP */
3712                  break;
3713                  }
3714                break;
3715    
3716                case OP_NOT_VSPACE:
3717                switch(c)
3718                  {
3719                  default: break;
3720                  case 0x0a:      /* LF */
3721                  case 0x0b:      /* VT */
3722                  case 0x0c:      /* FF */
3723                  case 0x0d:      /* CR */
3724                  case 0x85:      /* NEL */
3725                  RRETURN(MATCH_NOMATCH);
3726                  }
3727                break;
3728    
3729                case OP_VSPACE:
3730                switch(c)
3731                  {
3732                  default: RRETURN(MATCH_NOMATCH);
3733                  case 0x0a:      /* LF */
3734                  case 0x0b:      /* VT */
3735                  case 0x0c:      /* FF */
3736                  case 0x0d:      /* CR */
3737                  case 0x85:      /* NEL */
3738                break;                break;
3739                }                }
3740              break;              break;
# Line 3214  for (;;) Line 3801  for (;;)
3801              int len = 1;              int len = 1;
3802              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
3803              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
3804              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3805              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
3806                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
3807                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
# Line 3229  for (;;) Line 3816  for (;;)
3816              int len = 1;              int len = 1;
3817              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
3818              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
3819              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3820              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
3821                break;                break;
3822              eptr+= len;              eptr+= len;
# Line 3242  for (;;) Line 3829  for (;;)
3829              int len = 1;              int len = 1;
3830              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
3831              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
3832              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3833              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
3834                break;                break;
3835              eptr+= len;              eptr+= len;
# Line 3255  for (;;) Line 3842  for (;;)
3842              int len = 1;              int len = 1;
3843              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
3844              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
3845              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_script = UCD_SCRIPT(c);
3846              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
3847                break;                break;
3848              eptr+= len;              eptr+= len;
# Line 3268  for (;;) Line 3855  for (;;)
3855          if (possessive) continue;          if (possessive) continue;
3856          for(;;)          for(;;)
3857            {            {
3858            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM44);
3859            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3860            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3861            BACKCHAR(eptr);            if (utf8) BACKCHAR(eptr);
3862            }            }
3863          }          }
3864    
# Line 3284  for (;;) Line 3871  for (;;)
3871            {            {
3872            if (eptr >= md->end_subject) break;            if (eptr >= md->end_subject) break;
3873            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3874            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
3875            if (prop_category == ucp_M) break;            if (prop_category == ucp_M) break;
3876            while (eptr < md->end_subject)            while (eptr < md->end_subject)
3877              {              {
# Line 3293  for (;;) Line 3880  for (;;)
3880                {                {
3881                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
3882                }                }
3883              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3884              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
3885              eptr += len;              eptr += len;
3886              }              }
# Line 3304  for (;;) Line 3891  for (;;)
3891          if (possessive) continue;          if (possessive) continue;
3892          for(;;)          for(;;)
3893            {            {
3894            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM45);
3895            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3896            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3897            for (;;)                        /* Move back over one extended */            for (;;)                        /* Move back over one extended */
3898              {              {
3899              int len = 1;              int len = 1;
             BACKCHAR(eptr);  
3900              if (!utf8) c = *eptr; else              if (!utf8) c = *eptr; else
3901                {                {
3902                  BACKCHAR(eptr);
3903                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
3904                }                }
3905              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3906              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
3907              eptr--;              eptr--;
3908              }              }
# Line 3333  for (;;) Line 3920  for (;;)
3920          switch(ctype)          switch(ctype)
3921            {            {
3922            case OP_ANY:            case OP_ANY:
   
           /* Special code is required for UTF8, but when the maximum is  
           unlimited we don't need it, so we repeat the non-UTF8 code. This is  
           probably worth it, because .* is quite a common idiom. */  
   
3923            if (max < INT_MAX)            if (max < INT_MAX)
3924              {              {
3925              if ((ims & PCRE_DOTALL) == 0)              for (i = min; i < max; i++)
               {  
               for (i = min; i < max; i++)  
                 {  
                 if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;  
                 eptr++;  
                 while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  
                 }  
               }  
             else  
3926                {                {
3927                for (i = min; i < max; i++)                if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
3928                  {                eptr++;
3929                  if (eptr >= md->end_subject) break;                while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
                 eptr++;  
                 while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  
                 }  
3930                }                }
3931              }              }
3932    
# Line 3364  for (;;) Line 3934  for (;;)
3934    
3935            else            else
3936              {              {
3937              if ((ims & PCRE_DOTALL) == 0)              for (i = min; i < max; i++)
3938                {                {
3939                for (i = min; i < max; i++)                if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
3940                  {                eptr++;
3941                  if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;                while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
                 eptr++;  
                 }  
               break;  
3942                }                }
3943              else              }
3944              break;
3945    
3946              case OP_ALLANY:
3947              if (max < INT_MAX)
3948                {
3949                for (i = min; i < max; i++)
3950                {                {
3951                c = max - min;                if (eptr >= md->end_subject) break;
3952                if (c > (unsigned int)(md->end_subject - eptr))                eptr++;
3953                  c = md->end_subject - eptr;                while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
               eptr += c;  
3954                }                }
3955              }              }
3956              else eptr = md->end_subject;   /* Unlimited UTF-8 repeat */
3957            break;            break;
3958    
3959            /* The byte case is the same as non-UTF8 */            /* The byte case is the same as non-UTF8 */
# Line 3405  for (;;) Line 3978  for (;;)
3978                }                }
3979              else              else
3980                {                {
3981                if (c != 0x000a && c != 0x000b && c != 0x000c &&                if (c != 0x000a &&
3982                    c != 0x0085 && c != 0x2028 && c != 0x2029)                    (md->bsr_anycrlf ||
3983                       (c != 0x000b && c != 0x000c &&
3984                        c != 0x0085 && c != 0x2028 && c != 0x2029)))
3985                  break;                  break;
3986                eptr += len;                eptr += len;
3987                }                }
3988              }              }
3989            break;            break;
3990    
3991              case OP_NOT_HSPACE:
3992              case OP_HSPACE:
3993              for (i = min; i < max; i++)
3994                {
3995                BOOL gotspace;
3996                int len = 1;
3997                if (eptr >= md->end_subject) break;
3998                GETCHARLEN(c, eptr, len);
3999                switch(c)
4000                  {
4001                  default: gotspace = FALSE; break;
4002                  case 0x09:      /* HT */
4003                  case 0x20:      /* SPACE */
4004                  case 0xa0:      /* NBSP */
4005                  case 0x1680:    /* OGHAM SPACE MARK */
4006                  case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
4007                  case 0x2000:    /* EN QUAD */
4008                  case 0x2001:    /* EM QUAD */
4009                  case 0x2002:    /* EN SPACE */
4010                  case 0x2003:    /* EM SPACE */
4011                  case 0x2004:    /* THREE-PER-EM SPACE */
4012                  case 0x2005:    /* FOUR-PER-EM SPACE */
4013                  case 0x2006:    /* SIX-PER-EM SPACE */
4014                  case 0x2007:    /* FIGURE SPACE */
4015                  case 0x2008:    /* PUNCTUATION SPACE */
4016                  case 0x2009:    /* THIN SPACE */
4017                  case 0x200A:    /* HAIR SPACE */
4018                  case 0x202f:    /* NARROW NO-BREAK SPACE */
4019                  case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
4020                  case 0x3000:    /* IDEOGRAPHIC SPACE */
4021                  gotspace = TRUE;
4022                  break;
4023                  }
4024                if (gotspace == (ctype == OP_NOT_HSPACE)) break;
4025                eptr += len;
4026                }
4027              break;
4028    
4029              case OP_NOT_VSPACE:
4030              case OP_VSPACE:
4031              for (i = min; i < max; i++)
4032                {
4033                BOOL gotspace;
4034                int len = 1;
4035                if (eptr >= md->end_subject) break;
4036                GETCHARLEN(c, eptr, len);
4037                switch(c)
4038                  {
4039                  default: gotspace = FALSE; break;
4040                  case 0x0a:      /* LF */
4041                  case 0x0b:      /* VT */
4042                  case 0x0c:      /* FF */
4043                  case 0x0d:      /* CR */
4044                  case 0x85:      /* NEL */
4045                  case 0x2028:    /* LINE SEPARATOR */
4046                  case 0x2029:    /* PARAGRAPH SEPARATOR */
4047                  gotspace = TRUE;
4048                  break;
4049                  }
4050                if (gotspace == (ctype == OP_NOT_VSPACE)) break;
4051                eptr += len;
4052                }
4053              break;
4054    
4055            case OP_NOT_DIGIT:            case OP_NOT_DIGIT:
4056            for (i = min; i < max; i++)            for (i = min; i < max; i++)
4057              {              {
# Line 3488  for (;;) Line 4127  for (;;)
4127          if (possessive) continue;          if (possessive) continue;
4128          for(;;)          for(;;)
4129            {            {
4130            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM46);
4131            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4132            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
4133            BACKCHAR(eptr);            BACKCHAR(eptr);
4134            }            }
4135          }          }
4136        else        else
4137  #endif  #endif  /* SUPPORT_UTF8 */
4138    
4139        /* Not UTF-8 mode */        /* Not UTF-8 mode */
4140          {          {
4141          switch(ctype)          switch(ctype)
4142            {            {
4143            case OP_ANY:            case OP_ANY:
4144            if ((ims & PCRE_DOTALL) == 0)            for (i = min; i < max; i++)
4145              {              {
4146              for (i = min; i < max; i++)              if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
4147                {              eptr++;
               if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;  
               eptr++;  
               }  
             break;  
4148              }              }
4149            /* For DOTALL case, fall through and treat as \C */            break;
4150    
4151              case OP_ALLANY:
4152            case OP_ANYBYTE:            case OP_ANYBYTE:
4153            c = max - min;            c = max - min;
4154            if (c > (unsigned int)(md->end_subject - eptr))            if (c > (unsigned int)(md->end_subject - eptr))
# Line 3532  for (;;) Line 4168  for (;;)
4168                }                }
4169              else              else
4170                {                {
4171                if (c != 0x000a && c != 0x000b && c != 0x000c && c != 0x0085)                if (c != 0x000a &&
4172                      (md->bsr_anycrlf ||
4173                        (c != 0x000b && c != 0x000c && c != 0x0085)))
4174                  break;                  break;
4175                eptr++;                eptr++;
4176                }                }
4177              }              }
4178            break;            break;
4179    
4180              case OP_NOT_HSPACE:
4181              for (i = min; i < max; i++)
4182                {
4183                if (eptr >= md->end_subject) break;
4184                c = *eptr;
4185                if (c == 0x09 || c == 0x20 || c == 0xa0) break;
4186                eptr++;
4187                }
4188              break;
4189    
4190              case OP_HSPACE:
4191              for (i = min; i < max; i++)
4192                {
4193                if (eptr >= md->end_subject) break;
4194                c = *eptr;
4195                if (c != 0x09 && c != 0x20 && c != 0xa0) break;
4196                eptr++;
4197                }
4198              break;
4199    
4200              case OP_NOT_VSPACE:
4201              for (i = min; i < max; i++)
4202                {
4203                if (eptr >= md->end_subject) break;
4204                c = *eptr;
4205                if (c == 0x0a || c == 0x0b || c == 0x0c || c == 0x0d || c == 0x85)
4206                  break;
4207                eptr++;
4208                }
4209              break;
4210    
4211              case OP_VSPACE:
4212              for (i = min; i < max; i++)
4213                {
4214                if (eptr >= md->end_subject) break;
4215                c = *eptr;
4216                if (c != 0x0a && c != 0x0b && c != 0x0c && c != 0x0d && c != 0x85)
4217                  break;
4218                eptr++;
4219                }
4220              break;
4221    
4222            case OP_NOT_DIGIT:            case OP_NOT_DIGIT:
4223            for (i = min; i < max; i++)            for (i = min; i < max; i++)
4224              {              {
# Line 3602  for (;;) Line 4282  for (;;)
4282          if (possessive) continue;          if (possessive) continue;
4283          while (eptr >= pp)          while (eptr >= pp)
4284            {            {
4285            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM47);
4286            eptr--;            eptr--;
4287            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4288            }            }
# Line 3628  for (;;) Line 4308  for (;;)
4308    
4309    }             /* End of main loop */    }             /* End of main loop */
4310  /* Control never reaches here */  /* Control never reaches here */
4311    
4312    
4313    /* When compiling to use the heap rather than the stack for recursive calls to
4314    match(), the RRETURN() macro jumps here. The number that is saved in
4315    frame->Xwhere indicates which label we actually want to return to. */
4316    
4317    #ifdef NO_RECURSE
4318    #define LBL(val) case val: goto L_RM##val;
4319    HEAP_RETURN:
4320    switch (frame->Xwhere)
4321      {
4322      LBL( 1) LBL( 2) LBL( 3) LBL( 4) LBL( 5) LBL( 6) LBL( 7) LBL( 8)
4323      LBL( 9) LBL(10) LBL(11) LBL(12) LBL(13) LBL(14) LBL(15) LBL(17)
4324      LBL(19) LBL(24) LBL(25) LBL(26) LBL(27) LBL(29) LBL(31) LBL(33)
4325      LBL(35) LBL(43) LBL(47) LBL(48) LBL(49) LBL(50) LBL(51) LBL(52)
4326      LBL(53) LBL(54)
4327    #ifdef SUPPORT_UTF8
4328      LBL(16) LBL(18) LBL(20) LBL(21) LBL(22) LBL(23) LBL(28) LBL(30)
4329      LBL(32) LBL(34) LBL(42) LBL(46)
4330    #ifdef SUPPORT_UCP
4331      LBL(36) LBL(37) LBL(38) LBL(39) LBL(40) LBL(41) LBL(44) LBL(45)
4332    #endif  /* SUPPORT_UCP */
4333    #endif  /* SUPPORT_UTF8 */
4334      default:
4335      DPRINTF(("jump error in pcre match: label %d non-existent\n", frame->Xwhere));
4336      return PCRE_ERROR_INTERNAL;
4337      }
4338    #undef LBL
4339    #endif  /* NO_RECURSE */
4340  }  }
4341    
4342    
# Line 3640  Undefine all the macros that were define Line 4349  Undefine all the macros that were define
4349  #ifdef NO_RECURSE  #ifdef NO_RECURSE
4350  #undef eptr  #undef eptr
4351  #undef ecode  #undef ecode
4352    #undef mstart
4353  #undef offset_top  #undef offset_top
4354  #undef ims  #undef ims
4355  #undef eptrb  #undef eptrb
# Line 3712  Returns:          > 0 => success; value Line 4422  Returns:          > 0 => success; value
4422                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
4423  */  */
4424    
4425  PCRE_DATA_SCOPE int  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
4426  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,
4427    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,
4428    int offsetcount)    int offsetcount)
# Line 3737  const uschar *start_bits = NULL; Line 4447  const uschar *start_bits = NULL;
4447  USPTR start_match = (USPTR)subject + start_offset;  USPTR start_match = (USPTR)subject + start_offset;
4448  USPTR end_subject;  USPTR end_subject;
4449  USPTR req_byte_ptr = start_match - 1;  USPTR req_byte_ptr = start_match - 1;
 eptrblock eptrchain[EPTR_WORK_SIZE];  
4450    
4451  pcre_study_data internal_study;  pcre_study_data internal_study;
4452  const pcre_study_data *study;  const pcre_study_data *study;
# Line 3800  if (re->magic_number != MAGIC_NUMBER) Line 4509  if (re->magic_number != MAGIC_NUMBER)
4509  /* Set up other data */  /* Set up other data */
4510    
4511  anchored = ((re->options | options) & PCRE_ANCHORED) != 0;  anchored = ((re->options | options) & PCRE_ANCHORED) != 0;
4512  startline = (re->options & PCRE_STARTLINE) != 0;  startline = (re->flags & PCRE_STARTLINE) != 0;
4513  firstline = (re->options & PCRE_FIRSTLINE) != 0;  firstline = (re->options & PCRE_FIRSTLINE) != 0;
4514    
4515  /* The code starts after the real_pcre block and the capture name table. */  /* The code starts after the real_pcre block and the capture name table. */
# Line 3815  end_subject = md->end_subject; Line 4524  end_subject = md->end_subject;
4524    
4525  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;
4526  utf8 = md->utf8 = (re->options & PCRE_UTF8) != 0;  utf8 = md->utf8 = (re->options & PCRE_UTF8) != 0;
4527    md->jscript_compat = (re->options & PCRE_JAVASCRIPT_COMPAT) != 0;
4528    
4529  md->notbol = (options & PCRE_NOTBOL) != 0;  md->notbol = (options & PCRE_NOTBOL) != 0;
4530  md->noteol = (options & PCRE_NOTEOL) != 0;  md->noteol = (options & PCRE_NOTEOL) != 0;
# Line 3823  md->partial = (options & PCRE_PARTIAL) ! Line 4533  md->partial = (options & PCRE_PARTIAL) !
4533  md->hitend = FALSE;  md->hitend = FALSE;
4534    
4535  md->recursive = NULL;                   /* No recursion at top level */  md->recursive = NULL;                   /* No recursion at top level */
 md->eptrchain = eptrchain;              /* Make workspace generally available */  
4536    
4537  md->lcc = tables + lcc_offset;  md->lcc = tables + lcc_offset;
4538  md->ctypes = tables + ctypes_offset;  md->ctypes = tables + ctypes_offset;
4539    
4540    /* Handle different \R options. */
4541    
4542    switch (options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE))
4543      {
4544      case 0:
4545      if ((re->options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) != 0)
4546        md->bsr_anycrlf = (re->options & PCRE_BSR_ANYCRLF) != 0;
4547      else
4548    #ifdef BSR_ANYCRLF
4549      md->bsr_anycrlf = TRUE;
4550    #else
4551      md->bsr_anycrlf = FALSE;
4552    #endif
4553      break;
4554    
4555      case PCRE_BSR_ANYCRLF:
4556      md->bsr_anycrlf = TRUE;
4557      break;
4558    
4559      case PCRE_BSR_UNICODE:
4560      md->bsr_anycrlf = FALSE;
4561      break;
4562    
4563      default: return PCRE_ERROR_BADNEWLINE;
4564      }
4565    
4566  /* Handle different types of newline. The three bits give eight cases. If  /* Handle different types of newline. The three bits give eight cases. If
4567  nothing is set at run time, whatever was used at compile time applies. */  nothing is set at run time, whatever was used at compile time applies. */
4568    
4569  switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options : (pcre_uint32)options) &  switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options :
4570         PCRE_NEWLINE_BITS)          (pcre_uint32)options) & PCRE_NEWLINE_BITS)
4571    {    {
4572    case 0: newline = NEWLINE; break;   /* Compile-time default */    case 0: newline = NEWLINE; break;   /* Compile-time default */
4573    case PCRE_NEWLINE_CR: newline = '\r'; break;    case PCRE_NEWLINE_CR: newline = CHAR_CR; break;
4574    case PCRE_NEWLINE_LF: newline = '\n'; break;    case PCRE_NEWLINE_LF: newline = CHAR_NL; break;
4575    case PCRE_NEWLINE_CR+    case PCRE_NEWLINE_CR+
4576         PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break;         PCRE_NEWLINE_LF: newline = (CHAR_CR << 8) | CHAR_NL; break;
4577    case PCRE_NEWLINE_ANY: newline = -1; break;    case PCRE_NEWLINE_ANY: newline = -1; break;
4578      case PCRE_NEWLINE_ANYCRLF: newline = -2; break;
4579    default: return PCRE_ERROR_BADNEWLINE;    default: return PCRE_ERROR_BADNEWLINE;
4580    }    }
4581    
4582  if (newline < 0)  if (newline == -2)
4583      {
4584      md->nltype = NLTYPE_ANYCRLF;
4585      }
4586    else if (newline < 0)
4587    {    {
4588    md->nltype = NLTYPE_ANY;    md->nltype = NLTYPE_ANY;
4589    }    }
# Line 3866  else Line 4606  else
4606  /* Partial matching is supported only for a restricted set of regexes at the  /* Partial matching is supported only for a restricted set of regexes at the
4607  moment. */  moment. */
4608    
4609  if (md->partial && (re->options & PCRE_NOPARTIAL) != 0)  if (md->partial && (re->flags & PCRE_NOPARTIAL) != 0)
4610    return PCRE_ERROR_BADPARTIAL;    return PCRE_ERROR_BADPARTIAL;
4611    
4612  /* Check a UTF-8 string if required. Unfortunately there's no way of passing  /* Check a UTF-8 string if required. Unfortunately there's no way of passing
# Line 3943  studied, there may be a bitmap of possib Line 4683  studied, there may be a bitmap of possib
4683    
4684  if (!anchored)  if (!anchored)
4685    {    {
4686    if ((re->options & PCRE_FIRSTSET) != 0)    if ((re->flags & PCRE_FIRSTSET) != 0)
4687      {      {
4688      first_byte = re->first_byte & 255;      first_byte = re->first_byte & 255;
4689      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)
# Line 3958  if (!anchored) Line 4698  if (!anchored)
4698  /* For anchored or unanchored matches, there may be a "last known required  /* For anchored or unanchored matches, there may be a "last known required
4699  character" set. */  character" set. */
4700    
4701  if ((re->options & PCRE_REQCHSET) != 0)  if ((re->flags & PCRE_REQCHSET) != 0)
4702    {    {
4703    req_byte = re->req_byte & 255;    req_byte = re->req_byte & 255;
4704    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;
# Line 3974  the loop runs just once. */ Line 4714  the loop runs just once. */
4714  for(;;)  for(;;)
4715    {    {
4716    USPTR save_end_subject = end_subject;    USPTR save_end_subject = end_subject;
4717      USPTR new_start_match;
4718    
4719    /* Reset the maximum number of extractions we might see. */    /* Reset the maximum number of extractions we might see. */
4720    
# Line 3984  for(;;) Line 4725  for(;;)
4725      while (iptr < iend) *iptr++ = -1;      while (iptr < iend) *iptr++ = -1;
4726      }      }
4727    
4728    /* Advance to a unique first char if possible. If firstline is TRUE, the    /* If firstline is TRUE, the start of the match is constrained to the first
4729    start of the match is constrained to the first line of a multiline string.    line of a multiline string. That is, the match must be before or at the first
4730    That is, the match must be before or at the first newline. Implement this by    newline. Implement this by temporarily adjusting end_subject so that we stop
4731    temporarily adjusting end_subject so that we stop scanning at a newline. If    scanning at a newline. If the match fails at the newline, later code breaks
4732    the match fails at the newline, later code breaks this loop. */    this loop. */
4733    
4734    if (firstline)    if (firstline)
4735      {      {
4736      USPTR t = start_match;      USPTR t = start_match;
4737    #ifdef SUPPORT_UTF8
4738        if (utf8)
4739          {
4740          while (t < md->end_subject && !IS_NEWLINE(t))
4741            {
4742            t++;
4743            while (t < end_subject && (*t & 0xc0) == 0x80) t++;
4744            }
4745          }
4746        else
4747    #endif
4748      while (t < md->end_subject && !IS_NEWLINE(t)) t++;      while (t < md->end_subject && !IS_NEWLINE(t)) t++;
4749      end_subject = t;      end_subject = t;
4750      }      }
4751    
4752    /* Now test for a unique first byte */    /* There are some optimizations that avoid running the match if a known
4753      starting point is not found, or if a known later character is not present.
4754      However, there is an option that disables these, for testing and for ensuring
4755      that all callouts do actually occur. */
4756    
4757    if (first_byte >= 0)    if ((options & PCRE_NO_START_OPTIMIZE) == 0)
4758      {      {
4759      if (first_byte_caseless)      /* Advance to a unique first byte if there is one. */
       while (start_match < end_subject &&  
              md->lcc[*start_match] != first_byte)  
         start_match++;  
     else  
       while (start_match < end_subject && *start_match != first_byte)  
         start_match++;  
     }  
4760    
4761    /* Or to just after a linebreak for a multiline match if possible */      if (first_byte >= 0)
4762          {
4763          if (first_byte_caseless)
4764            while (start_match < end_subject && md->lcc[*start_match] != first_byte)
4765              start_match++;
4766          else
4767            while (start_match < end_subject && *start_match != first_byte)
4768              start_match++;
4769          }
4770    
4771    else if (startline)      /* Or to just after a linebreak for a multiline match */
4772      {  
4773      if (start_match > md->start_subject + start_offset)      else if (startline)
4774        {        {
4775        while (start_match <= end_subject && !WAS_NEWLINE(start_match))        if (start_match > md->start_subject + start_offset)
4776          start_match++;          {
4777    #ifdef SUPPORT_UTF8
4778            if (utf8)
4779              {
4780              while (start_match < end_subject && !WAS_NEWLINE(start_match))
4781                {
4782                start_match++;
4783                while(start_match < end_subject && (*start_match & 0xc0) == 0x80)
4784                  start_match++;
4785                }
4786              }
4787            else
4788    #endif
4789            while (start_match < end_subject && !WAS_NEWLINE(start_match))
4790              start_match++;
4791    
4792        /* If we have just passed a CR and the newline option is ANY, and we are          /* If we have just passed a CR and the newline option is ANY or ANYCRLF,
4793        now at a LF, advance the match position by one more character. */          and we are now at a LF, advance the match position by one more character.
4794            */
4795    
4796        if (start_match[-1] == '\r' &&          if (start_match[-1] == CHAR_CR &&
4797             md->nltype == NLTYPE_ANY &&               (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&
4798             start_match < end_subject &&               start_match < end_subject &&
4799             *start_match == '\n')               *start_match == CHAR_NL)
4800          start_match++;            start_match++;
4801            }
4802        }        }
     }  
4803    
4804    /* Or to a non-unique first char after study */      /* Or to a non-unique first byte after study */
4805    
4806    else if (start_bits != NULL)      else if (start_bits != NULL)
     {  
     while (start_match < end_subject)  
4807        {        {
4808        register unsigned int c = *start_match;        while (start_match < end_subject)
4809        if ((start_bits[c/8] & (1 << (c&7))) == 0) start_match++; else break;          {
4810            register unsigned int c = *start_match;
4811            if ((start_bits[c/8] & (1 << (c&7))) == 0) start_match++;
4812              else break;
4813            }
4814        }        }
4815      }      }   /* Starting optimizations */
4816    
4817    /* Restore fudged end_subject */    /* Restore fudged end_subject */
4818    
# Line 4051  for(;;) Line 4824  for(;;)
4824    printf("\n");    printf("\n");
4825  #endif  #endif
4826    
4827    /* If req_byte is set, we know that that character must appear in the subject    /* If req_byte is set, we know that that character must appear in the
4828    for the match to succeed. If the first character is set, req_byte must be    subject for the match to succeed. If the first character is set, req_byte
4829    later in the subject; otherwise the test starts at the match point. This    must be later in the subject; otherwise the test starts at the match point.
4830    optimization can save a huge amount of backtracking in patterns with nested    This optimization can save a huge amount of backtracking in patterns with
4831    unlimited repeats that aren't going to match. Writing separate code for    nested unlimited repeats that aren't going to match. Writing separate code
4832    cased/caseless versions makes it go faster, as does using an autoincrement    for cased/caseless versions makes it go faster, as does using an
4833    and backing off on a match.    autoincrement and backing off on a match.
4834    
4835    HOWEVER: when the subject string is very, very long, searching to its end can    HOWEVER: when the subject string is very, very long, searching to its end
4836    take a long time, and give bad performance on quite ordinary patterns. This    can take a long time, and give bad performance on quite ordinary patterns.
4837    showed up when somebody was matching something like /^\d+C/ on a 32-megabyte    This showed up when somebody was matching something like /^\d+C/ on a
4838    string... so we don't do this when the string is sufficiently long.    32-megabyte string... so we don't do this when the string is sufficiently
4839      long.
4840    
4841    ALSO: this processing is disabled when partial matching is requested.    ALSO: this processing is disabled when partial matching is requested, or if
4842    */    disabling is explicitly requested. */
4843    
4844    if (req_byte >= 0 &&    if ((options & PCRE_NO_START_OPTIMIZE) == 0 &&
4845          req_byte >= 0 &&
4846        end_subject - start_match < REQ_BYTE_MAX &&        end_subject - start_match < REQ_BYTE_MAX &&
4847        !md->partial)        !md->partial)
4848      {      {
# Line 4113  for(;;) Line 4888  for(;;)
4888    
4889    /* OK, we can now run the match. */    /* OK, we can now run the match. */
4890    
4891    md->start_match = start_match;    md->start_match_ptr = start_match;
4892    md->match_call_count = 0;    md->match_call_count = 0;
4893    md->eptrn = 0;                          /* Next free eptrchain slot */    rc = match(start_match, md->start_code, start_match, 2, md, ims, NULL, 0, 0);
4894    rc = match(start_match, md->start_code, 2, md, ims, NULL, 0, 0);  
4895      switch(rc)
4896        {
4897        /* NOMATCH and PRUNE advance by one character. THEN at this level acts
4898        exactly like PRUNE. */
4899    
4900        case MATCH_NOMATCH:
4901        case MATCH_PRUNE:
4902        case MATCH_THEN:
4903        new_start_match = start_match + 1;
4904    #ifdef SUPPORT_UTF8
4905        if (utf8)
4906          while(new_start_match < end_subject && (*new_start_match & 0xc0) == 0x80)
4907            new_start_match++;
4908    #endif
4909        break;
4910    
4911        /* SKIP passes back the next starting point explicitly. */
4912    
4913        case MATCH_SKIP:
4914        new_start_match = md->start_match_ptr;
4915        break;
4916    
4917        /* COMMIT disables the bumpalong, but otherwise behaves as NOMATCH. */
4918    
4919    /* Any return other than MATCH_NOMATCH breaks the loop. */      case MATCH_COMMIT:
4920        rc = MATCH_NOMATCH;
4921        goto ENDLOOP;
4922    
4923    if (rc != MATCH_NOMATCH) break;      /* Any other return is some kind of error. */
4924    
4925        default:
4926        goto ENDLOOP;
4927        }
4928    
4929      /* Control reaches here for the various types of "no match at this point"
4930      result. Reset the code to MATCH_NOMATCH for subsequent checking. */
4931    
4932      rc = MATCH_NOMATCH;
4933    
4934    /* If PCRE_FIRSTLINE is set, the match must happen before or at the first    /* If PCRE_FIRSTLINE is set, the match must happen before or at the first
4935    newline in the subject (though it may continue over the newline). Therefore,    newline in the subject (though it may continue over the newline). Therefore,
# Line 4128  for(;;) Line 4937  for(;;)
4937    
4938    if (firstline && IS_NEWLINE(start_match)) break;    if (firstline && IS_NEWLINE(start_match)) break;
4939    
4940    /* Advance the match position by one character. */    /* Advance to new matching position */
4941    
4942    start_match++;    start_match = new_start_match;
 #ifdef SUPPORT_UTF8  
   if (utf8)  
     while(start_match < end_subject && (*start_match & 0xc0) == 0x80)  
       start_match++;  
 #endif  
4943    
4944    /* Break the loop if the pattern is anchored or if we have passed the end of    /* Break the loop if the pattern is anchored or if we have passed the end of
4945    the subject. */    the subject. */
4946    
4947    if (anchored || start_match > end_subject) break;    if (anchored || start_match > end_subject) break;
4948    
4949    /* If we have just passed a CR and the newline option is CRLF or ANY, and we    /* If we have just passed a CR and we are now at a LF, and the pattern does
4950    are now at a LF, advance the match position by one more character. */    not contain any explicit matches for \r or \n, and the newline option is CRLF
4951      or ANY or ANYCRLF, advance the match position by one more character. */
4952    if (start_match[-1] == '\r' &&  
4953         (md->nltype == NLTYPE_ANY || md->nllen == 2) &&    if (start_match[-1] == CHAR_CR &&
4954         start_match < end_subject &&        start_match < end_subject &&
4955         *start_match == '\n')        *start_match == CHAR_NL &&
4956          (re->flags & PCRE_HASCRORLF) == 0 &&
4957            (md->nltype == NLTYPE_ANY ||
4958             md->nltype == NLTYPE_ANYCRLF ||
4959             md->nllen == 2))
4960      start_match++;      start_match++;
4961    
4962    }   /* End of for(;;) "bumpalong" loop */    }   /* End of for(;;) "bumpalong" loop */
# Line 4158  for(;;) Line 4966  for(;;)
4966  /* We reach here when rc is not MATCH_NOMATCH, or if one of the stopping  /* We reach here when rc is not MATCH_NOMATCH, or if one of the stopping
4967  conditions is true:  conditions is true:
4968    
4969  (1) The pattern is anchored;  (1) The pattern is anchored or the match was failed by (*COMMIT);
4970    
4971  (2) We are past the end of the subject;  (2) We are past the end of the subject;
4972    
# Line 4173  processing, copy those that we can. In t Line 4981  processing, copy those that we can. In t
4981  certain parts of the pattern were not used, even though there are more  certain parts of the pattern were not used, even though there are more
4982  capturing parentheses than vector slots. */  capturing parentheses than vector slots. */
4983    
4984    ENDLOOP:
4985    
4986  if (rc == MATCH_MATCH)  if (rc == MATCH_MATCH)
4987    {    {
4988    if (using_temporary_offsets)    if (using_temporary_offsets)
# Line 4193  if (rc == MATCH_MATCH) Line 5003  if (rc == MATCH_MATCH)
5003    
5004    rc = md->offset_overflow? 0 : md->end_offset_top/2;    rc = md->offset_overflow? 0 : md->end_offset_top/2;
5005    
5006    /* If there is space, set up the whole thing as substring 0. */    /* If there is space, set up the whole thing as substring 0. The value of
5007      md->start_match_ptr might be modified if \K was encountered on the success
5008      matching path. */
5009    
5010    if (offsetcount < 2) rc = 0; else    if (offsetcount < 2) rc = 0; else
5011      {      {
5012      offsets[0] = start_match - md->start_subject;      offsets[0] = md->start_match_ptr - md->start_subject;
5013      offsets[1] = md->end_match_ptr - md->start_subject;      offsets[1] = md->end_match_ptr - md->start_subject;
5014      }      }
5015    

Legend:
Removed from v.144  
changed lines
  Added in v.400

  ViewVC Help
Powered by ViewVC 1.1.5