/[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 341 by ph10, Sat Apr 19 16:41:04 2008 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) 1997-2008 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 188  calls by keeping local variables that ne Line 194  calls by keeping local variables that ne
194  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
195  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
196  always used to.  always used to.
197    
198    The original heap-recursive code used longjmp(). However, it seems that this
199    can be very slow on some operating systems. Following a suggestion from Stan
200    Switzer, the use of longjmp() has been abolished, at the cost of having to
201    provide a unique number for each call to RMATCH. There is no way of generating
202    a sequence of numbers at compile time in C. I have given them names, to make
203    them stand out more clearly.
204    
205    Crude tests on x86 Linux show a small speedup of around 5-8%. However, on
206    FreeBSD, avoiding longjmp() more than halves the time taken to run the standard
207    tests. Furthermore, not using longjmp() means that local dynamic variables
208    don't have indeterminate values; this has meant that the frame size can be
209    reduced because the result can be "passed back" by straight setting of the
210    variable instead of being passed in the frame.
211  ****************************************************************************  ****************************************************************************
212  ***************************************************************************/  ***************************************************************************/
213    
214    /* Numbers for RMATCH calls. When this list is changed, the code at HEAP_RETURN
215    below must be updated in sync.  */
216    
217    enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM6,  RM7,  RM8,  RM9,  RM10,
218           RM11,  RM12, RM13, RM14, RM15, RM16, RM17, RM18, RM19, RM20,
219           RM21,  RM22, RM23, RM24, RM25, RM26, RM27, RM28, RM29, RM30,
220           RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,
221           RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,
222           RM51,  RM52, RM53, RM54 };
223    
224  /* 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
225  versions and production versions. */  versions and production versions. Note that the "rw" argument of RMATCH isn't
226    actuall used in this definition. */
227    
228  #ifndef NO_RECURSE  #ifndef NO_RECURSE
229  #define REGISTER register  #define REGISTER register
230    
231  #ifdef DEBUG  #ifdef DEBUG
232  #define RMATCH(rx,ra,rb,rc,rd,re,rf,rg) \  #define RMATCH(ra,rb,rc,rd,re,rf,rg,rw) \
233    { \    { \
234    printf("match() called in line %d\n", __LINE__); \    printf("match() called in line %d\n", __LINE__); \
235    rx = match(ra,rb,rc,rd,re,rf,rg,rdepth+1); \    rrc = match(ra,rb,mstart,rc,rd,re,rf,rg,rdepth+1); \
236    printf("to line %d\n", __LINE__); \    printf("to line %d\n", __LINE__); \
237    }    }
238  #define RRETURN(ra) \  #define RRETURN(ra) \
# Line 210  versions and production versions. */ Line 241  versions and production versions. */
241    return ra; \    return ra; \
242    }    }
243  #else  #else
244  #define RMATCH(rx,ra,rb,rc,rd,re,rf,rg) \  #define RMATCH(ra,rb,rc,rd,re,rf,rg,rw) \
245    rx = match(ra,rb,rc,rd,re,rf,rg,rdepth+1)    rrc = match(ra,rb,mstart,rc,rd,re,rf,rg,rdepth+1)
246  #define RRETURN(ra) return ra  #define RRETURN(ra) return ra
247  #endif  #endif
248    
249  #else  #else
250    
251    
252  /* 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
253  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
254  match(), which never changes. */  argument of match(), which never changes. */
255    
256  #define REGISTER  #define REGISTER
257    
258  #define RMATCH(rx,ra,rb,rc,rd,re,rf,rg)\  #define RMATCH(ra,rb,rc,rd,re,rf,rg,rw)\
259    {\    {\
260    heapframe *newframe = (pcre_stack_malloc)(sizeof(heapframe));\    heapframe *newframe = (pcre_stack_malloc)(sizeof(heapframe));\
261    if (setjmp(frame->Xwhere) == 0)\    frame->Xwhere = rw; \
262      {\    newframe->Xeptr = ra;\
263      newframe->Xeptr = ra;\    newframe->Xecode = rb;\
264      newframe->Xecode = rb;\    newframe->Xmstart = mstart;\
265      newframe->Xoffset_top = rc;\    newframe->Xoffset_top = rc;\
266      newframe->Xims = re;\    newframe->Xims = re;\
267      newframe->Xeptrb = rf;\    newframe->Xeptrb = rf;\
268      newframe->Xflags = rg;\    newframe->Xflags = rg;\
269      newframe->Xrdepth = frame->Xrdepth + 1;\    newframe->Xrdepth = frame->Xrdepth + 1;\
270      newframe->Xprevframe = frame;\    newframe->Xprevframe = frame;\
271      frame = newframe;\    frame = newframe;\
272      DPRINTF(("restarting from line %d\n", __LINE__));\    DPRINTF(("restarting from line %d\n", __LINE__));\
273      goto HEAP_RECURSE;\    goto HEAP_RECURSE;\
274      }\    L_##rw:\
275    else\    DPRINTF(("jumped back to line %d\n", __LINE__));\
     {\  
     DPRINTF(("longjumped back to line %d\n", __LINE__));\  
     frame = md->thisframe;\  
     rx = frame->Xresult;\  
     }\  
276    }    }
277    
278  #define RRETURN(ra)\  #define RRETURN(ra)\
# Line 256  match(), which never changes. */ Line 282  match(), which never changes. */
282    (pcre_stack_free)(newframe);\    (pcre_stack_free)(newframe);\
283    if (frame != NULL)\    if (frame != NULL)\
284      {\      {\
285      frame->Xresult = ra;\      rrc = ra;\
286      md->thisframe = frame;\      goto HEAP_RETURN;\
     longjmp(frame->Xwhere, 1);\  
287      }\      }\
288    return ra;\    return ra;\
289    }    }
# Line 273  typedef struct heapframe { Line 298  typedef struct heapframe {
298    
299    const uschar *Xeptr;    const uschar *Xeptr;
300    const uschar *Xecode;    const uschar *Xecode;
301      const uschar *Xmstart;
302    int Xoffset_top;    int Xoffset_top;
303    long int Xims;    long int Xims;
304    eptrblock *Xeptrb;    eptrblock *Xeptrb;
# Line 323  typedef struct heapframe { Line 349  typedef struct heapframe {
349    
350    eptrblock Xnewptrb;    eptrblock Xnewptrb;
351    
352    /* Place to pass back result, and where to jump back to */    /* Where to jump back to */
353    
354    int  Xresult;    int Xwhere;
   jmp_buf Xwhere;  
355    
356  } heapframe;  } heapframe;
357    
# Line 354  made performance worse. Line 379  made performance worse.
379  Arguments:  Arguments:
380     eptr        pointer to current character in subject     eptr        pointer to current character in subject
381     ecode       pointer to current position in compiled code     ecode       pointer to current position in compiled code
382       mstart      pointer to the current match start position (can be modified
383                     by encountering \K)
384     offset_top  current top pointer     offset_top  current top pointer
385     md          pointer to "static" info for the match     md          pointer to "static" info for the match
386     ims         current /i, /m, and /s options     ims         current /i, /m, and /s options
# Line 363  Arguments: Line 390  Arguments:
390                   match_condassert - this is an assertion condition                   match_condassert - this is an assertion condition
391                   match_cbegroup - this is the start of an unlimited repeat                   match_cbegroup - this is the start of an unlimited repeat
392                     group that can match an empty string                     group that can match an empty string
                  match_tail_recursed - this is a tail_recursed group  
393     rdepth      the recursion depth     rdepth      the recursion depth
394    
395  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 399  Returns:       MATCH_MATCH if matched
399  */  */
400    
401  static int  static int
402  match(REGISTER USPTR eptr, REGISTER const uschar *ecode,  match(REGISTER USPTR eptr, REGISTER const uschar *ecode, const uschar *mstart,
403    int offset_top, match_data *md, unsigned long int ims, eptrblock *eptrb,    int offset_top, match_data *md, unsigned long int ims, eptrblock *eptrb,
404    int flags, unsigned int rdepth)    int flags, unsigned int rdepth)
405  {  {
# Line 401  frame->Xprevframe = NULL;            /* Line 427  frame->Xprevframe = NULL;            /*
427    
428  frame->Xeptr = eptr;  frame->Xeptr = eptr;
429  frame->Xecode = ecode;  frame->Xecode = ecode;
430    frame->Xmstart = mstart;
431  frame->Xoffset_top = offset_top;  frame->Xoffset_top = offset_top;
432  frame->Xims = ims;  frame->Xims = ims;
433  frame->Xeptrb = eptrb;  frame->Xeptrb = eptrb;
# Line 415  HEAP_RECURSE: Line 442  HEAP_RECURSE:
442    
443  #define eptr               frame->Xeptr  #define eptr               frame->Xeptr
444  #define ecode              frame->Xecode  #define ecode              frame->Xecode
445    #define mstart             frame->Xmstart
446  #define offset_top         frame->Xoffset_top  #define offset_top         frame->Xoffset_top
447  #define ims                frame->Xims  #define ims                frame->Xims
448  #define eptrb              frame->Xeptrb  #define eptrb              frame->Xeptrb
# Line 545  defined). However, RMATCH isn't like a f Line 573  defined). However, RMATCH isn't like a f
573  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,
574  however, impact performance when true recursion is being used. */  however, impact performance when true recursion is being used. */
575    
576    #ifdef SUPPORT_UTF8
577    utf8 = md->utf8;       /* Local copy of the flag */
578    #else
579    utf8 = FALSE;
580    #endif
581    
582  /* 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
583  haven't exceeded the recursive call limit. */  haven't exceeded the recursive call limit. */
584    
# Line 553  if (rdepth >= md->match_limit_recursion) Line 587  if (rdepth >= md->match_limit_recursion)
587    
588  original_ims = ims;    /* Save for resetting on ')' */  original_ims = ims;    /* Save for resetting on ')' */
589    
 #ifdef SUPPORT_UTF8  
 utf8 = md->utf8;       /* Local copy of the flag */  
 #else  
 utf8 = FALSE;  
 #endif  
   
590  /* 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
591  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
592  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
593  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.
594  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
595  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
596  already used. */  block that is used is on the stack, so a new one may be required for each
597    match(). */
598    
599  if ((flags & match_cbegroup) != 0)  if ((flags & match_cbegroup) != 0)
600    {    {
601    eptrblock *p;    newptrb.epb_saved_eptr = eptr;
602    if ((flags & match_tail_recursed) != 0)    newptrb.epb_prev = eptrb;
603      {    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;  
604    }    }
605    
606  /* Now start processing the opcodes. */  /* Now start processing the opcodes. */
# Line 593  for (;;) Line 615  for (;;)
615    
616    if (md->partial &&    if (md->partial &&
617        eptr >= md->end_subject &&        eptr >= md->end_subject &&
618        eptr > md->start_match)        eptr > mstart)
619      md->hitend = TRUE;      md->hitend = TRUE;
620    
621    switch(op)    switch(op)
622      {      {
623        case OP_FAIL:
624        RRETURN(MATCH_NOMATCH);
625    
626        case OP_PRUNE:
627        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
628          ims, eptrb, flags, RM51);
629        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
630        RRETURN(MATCH_PRUNE);
631    
632        case OP_COMMIT:
633        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
634          ims, eptrb, flags, RM52);
635        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
636        RRETURN(MATCH_COMMIT);
637    
638        case OP_SKIP:
639        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
640          ims, eptrb, flags, RM53);
641        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
642        md->start_match_ptr = eptr;   /* Pass back current position */
643        RRETURN(MATCH_SKIP);
644    
645        case OP_THEN:
646        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
647          ims, eptrb, flags, RM54);
648        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
649        RRETURN(MATCH_THEN);
650    
651      /* 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
652      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.
653      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 687  for (;;)
687        flags = (op == OP_SCBRA)? match_cbegroup : 0;        flags = (op == OP_SCBRA)? match_cbegroup : 0;
688        do        do
689          {          {
690          RMATCH(rrc, eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
691            ims, eptrb, flags);            ims, eptrb, flags, RM1);
692          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
693          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
694          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
695          }          }
# Line 654  for (;;) Line 704  for (;;)
704        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
705        }        }
706    
707      /* Insufficient room for saving captured contents. Treat as a non-capturing      /* FALL THROUGH ... Insufficient room for saving captured contents. Treat
708      bracket. */      as a non-capturing bracket. */
709    
710        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
711        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
712    
713      DPRINTF(("insufficient capture room: treat as non-capturing\n"));      DPRINTF(("insufficient capture room: treat as non-capturing\n"));
714    
715        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
716        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
717    
718      /* 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
719      final alternative within the brackets, we would return the result of a      final alternative within the brackets, we would return the result of a
720      recursive call to match() whatever happened. We can reduce stack usage by      recursive call to match() whatever happened. We can reduce stack usage by
721      turning this into a tail recursion. */      turning this into a tail recursion, except in the case when match_cbegroup
722        is set.*/
723    
724      case OP_BRA:      case OP_BRA:
725      case OP_SBRA:      case OP_SBRA:
# Line 670  for (;;) Line 727  for (;;)
727      flags = (op >= OP_SBRA)? match_cbegroup : 0;      flags = (op >= OP_SBRA)? match_cbegroup : 0;
728      for (;;)      for (;;)
729        {        {
730        if (ecode[GET(ecode, 1)] != OP_ALT)        if (ecode[GET(ecode, 1)] != OP_ALT)   /* Final alternative */
731          {          {
732          ecode += _pcre_OP_lengths[*ecode];          if (flags == 0)    /* Not a possibly empty group */
733          flags |= match_tail_recursed;            {
734          DPRINTF(("bracket 0 tail recursion\n"));            ecode += _pcre_OP_lengths[*ecode];
735          goto TAIL_RECURSE;            DPRINTF(("bracket 0 tail recursion\n"));
736              goto TAIL_RECURSE;
737              }
738    
739            /* Possibly empty group; can't use tail recursion. */
740    
741            RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,
742              eptrb, flags, RM48);
743            RRETURN(rrc);
744          }          }
745    
746        /* For non-final alternatives, continue the loop for a NOMATCH result;        /* For non-final alternatives, continue the loop for a NOMATCH result;
747        otherwise return. */        otherwise return. */
748    
749        RMATCH(rrc, eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,
750          eptrb, flags);          eptrb, flags, RM2);
751        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
752        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
753        }        }
754      /* Control never reaches here. */      /* Control never reaches here. */
# Line 723  for (;;) Line 788  for (;;)
788    
789      else      else
790        {        {
791        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL,
792            match_condassert);            match_condassert, RM3);
793        if (rrc == MATCH_MATCH)        if (rrc == MATCH_MATCH)
794          {          {
795          condition = TRUE;          condition = TRUE;
796          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);
797          while (*ecode == OP_ALT) ecode += GET(ecode, 1);          while (*ecode == OP_ALT) ecode += GET(ecode, 1);
798          }          }
799        else if (rrc != MATCH_NOMATCH)        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
800          {          {
801          RRETURN(rrc);         /* Need braces because of following else */          RRETURN(rrc);         /* Need braces because of following else */
802          }          }
# Line 743  for (;;) Line 808  for (;;)
808        }        }
809    
810      /* 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,
811      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
812      alternative doesn't exist, we can just plough on. */      match_cbegroup is required for an unlimited repeat of a possibly empty
813        group. If the second alternative doesn't exist, we can just plough on. */
814    
815      if (condition || *ecode == OP_ALT)      if (condition || *ecode == OP_ALT)
816        {        {
817        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
818        flags = match_tail_recursed | ((op == OP_SCOND)? match_cbegroup : 0);        if (op == OP_SCOND)        /* Possibly empty group */
819        goto TAIL_RECURSE;          {
820            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, match_cbegroup, RM49);
821            RRETURN(rrc);
822            }
823          else                       /* Group must match something */
824            {
825            flags = 0;
826            goto TAIL_RECURSE;
827            }
828        }        }
829      else      else                         /* Condition false & no 2nd alternative */
830        {        {
831        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
832        }        }
833      break;      break;
834    
835    
836      /* 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
837      restore the offsets appropriately and continue from after the call. */      recursion, we should restore the offsets appropriately and continue from
838        after the call. */
839    
840        case OP_ACCEPT:
841      case OP_END:      case OP_END:
842      if (md->recursive != NULL && md->recursive->group_num == 0)      if (md->recursive != NULL && md->recursive->group_num == 0)
843        {        {
# Line 770  for (;;) Line 846  for (;;)
846        md->recursive = rec->prevrec;        md->recursive = rec->prevrec;
847        memmove(md->offset_vector, rec->offset_save,        memmove(md->offset_vector, rec->offset_save,
848          rec->saved_max * sizeof(int));          rec->saved_max * sizeof(int));
849        md->start_match = rec->save_start;        mstart = rec->save_start;
850        ims = original_ims;        ims = original_ims;
851        ecode = rec->after_call;        ecode = rec->after_call;
852        break;        break;
# Line 779  for (;;) Line 855  for (;;)
855      /* 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
856      string - backtracking will then try other alternatives, if any. */      string - backtracking will then try other alternatives, if any. */
857    
858      if (md->notempty && eptr == md->start_match) RRETURN(MATCH_NOMATCH);      if (md->notempty && eptr == mstart) RRETURN(MATCH_NOMATCH);
859      md->end_match_ptr = eptr;          /* Record where we ended */      md->end_match_ptr = eptr;           /* Record where we ended */
860      md->end_offset_top = offset_top;   /* and how many extracts were taken */      md->end_offset_top = offset_top;    /* and how many extracts were taken */
861        md->start_match_ptr = mstart;       /* and the start (\K can modify) */
862      RRETURN(MATCH_MATCH);      RRETURN(MATCH_MATCH);
863    
864      /* Change option settings */      /* Change option settings */
# Line 802  for (;;) Line 879  for (;;)
879      case OP_ASSERTBACK:      case OP_ASSERTBACK:
880      do      do
881        {        {
882        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,
883            RM4);
884        if (rrc == MATCH_MATCH) break;        if (rrc == MATCH_MATCH) break;
885        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
886        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
887        }        }
888      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 828  for (;;) Line 906  for (;;)
906      case OP_ASSERTBACK_NOT:      case OP_ASSERTBACK_NOT:
907      do      do
908        {        {
909        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,
910            RM5);
911        if (rrc == MATCH_MATCH) RRETURN(MATCH_NOMATCH);        if (rrc == MATCH_MATCH) RRETURN(MATCH_NOMATCH);
912        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
913        ecode += GET(ecode,1);        ecode += GET(ecode,1);
914        }        }
915      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 854  for (;;) Line 933  for (;;)
933          {          {
934          eptr--;          eptr--;
935          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
936          BACKCHAR(eptr)          BACKCHAR(eptr);
937          }          }
938        }        }
939      else      else
# Line 885  for (;;) Line 964  for (;;)
964        cb.offset_vector    = md->offset_vector;        cb.offset_vector    = md->offset_vector;
965        cb.subject          = (PCRE_SPTR)md->start_subject;        cb.subject          = (PCRE_SPTR)md->start_subject;
966        cb.subject_length   = md->end_subject - md->start_subject;        cb.subject_length   = md->end_subject - md->start_subject;
967        cb.start_match      = md->start_match - md->start_subject;        cb.start_match      = mstart - md->start_subject;
968        cb.current_position = eptr - md->start_subject;        cb.current_position = eptr - md->start_subject;
969        cb.pattern_position = GET(ecode, 2);        cb.pattern_position = GET(ecode, 2);
970        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);
# Line 947  for (;;) Line 1026  for (;;)
1026    
1027        memcpy(new_recursive.offset_save, md->offset_vector,        memcpy(new_recursive.offset_save, md->offset_vector,
1028              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1029        new_recursive.save_start = md->start_match;        new_recursive.save_start = mstart;
1030        md->start_match = eptr;        mstart = eptr;
1031    
1032        /* 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
1033        restore the offset and recursion data. */        restore the offset and recursion data. */
# Line 957  for (;;) Line 1036  for (;;)
1036        flags = (*callpat >= OP_SBRA)? match_cbegroup : 0;        flags = (*callpat >= OP_SBRA)? match_cbegroup : 0;
1037        do        do
1038          {          {
1039          RMATCH(rrc, eptr, callpat + _pcre_OP_lengths[*callpat], offset_top,          RMATCH(eptr, callpat + _pcre_OP_lengths[*callpat], offset_top,
1040            md, ims, eptrb, flags);            md, ims, eptrb, flags, RM6);
1041          if (rrc == MATCH_MATCH)          if (rrc == MATCH_MATCH)
1042            {            {
1043            DPRINTF(("Recursion matched\n"));            DPRINTF(("Recursion matched\n"));
# Line 967  for (;;) Line 1046  for (;;)
1046              (pcre_free)(new_recursive.offset_save);              (pcre_free)(new_recursive.offset_save);
1047            RRETURN(MATCH_MATCH);            RRETURN(MATCH_MATCH);
1048            }            }
1049          else if (rrc != MATCH_NOMATCH)          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
1050            {            {
1051            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1052            RRETURN(rrc);            RRETURN(rrc);
# Line 1001  for (;;) Line 1080  for (;;)
1080    
1081      do      do
1082        {        {
1083        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);  
1084        if (rrc == MATCH_MATCH) break;        if (rrc == MATCH_MATCH) break;
1085        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1086        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1087        }        }
1088      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 1047  for (;;) Line 1125  for (;;)
1125    
1126      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1127        {        {
1128        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);
1129        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1130        ecode = prev;        ecode = prev;
1131        flags = match_tail_recursed;        flags = 0;
1132        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1133        }        }
1134      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
1135        {        {
1136        RMATCH(rrc, eptr, prev, offset_top, md, ims, eptrb, match_cbegroup);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, match_cbegroup, RM9);
1137        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1138        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1139        flags = match_tail_recursed;        flags = 0;
1140        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1141        }        }
1142      /* Control never gets here */      /* Control never gets here */
# Line 1070  for (;;) Line 1148  for (;;)
1148      do ecode += GET(ecode,1); while (*ecode == OP_ALT);      do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1149      break;      break;
1150    
1151      /* BRAZERO and BRAMINZERO occur just before a bracket group, indicating      /* BRAZERO, BRAMINZERO and SKIPZERO occur just before a bracket group,
1152      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
1153      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
1154      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
1155      preceded by BRAZERO or BRAMINZERO. */      optional ones preceded by BRAZERO or BRAMINZERO. */
1156    
1157      case OP_BRAZERO:      case OP_BRAZERO:
1158        {        {
1159        next = ecode+1;        next = ecode+1;
1160        RMATCH(rrc, eptr, next, offset_top, md, ims, eptrb, 0);        RMATCH(eptr, next, offset_top, md, ims, eptrb, 0, RM10);
1161        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1162        do next += GET(next,1); while (*next == OP_ALT);        do next += GET(next,1); while (*next == OP_ALT);
1163        ecode = next + 1 + LINK_SIZE;        ecode = next + 1 + LINK_SIZE;
# Line 1090  for (;;) Line 1168  for (;;)
1168        {        {
1169        next = ecode+1;        next = ecode+1;
1170        do next += GET(next, 1); while (*next == OP_ALT);        do next += GET(next, 1); while (*next == OP_ALT);
1171        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);
1172        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1173        ecode++;        ecode++;
1174        }        }
1175      break;      break;
1176    
1177        case OP_SKIPZERO:
1178          {
1179          next = ecode+1;
1180          do next += GET(next,1); while (*next == OP_ALT);
1181          ecode = next + 1 + LINK_SIZE;
1182          }
1183        break;
1184    
1185      /* End of a group, repeated or non-repeating. */      /* End of a group, repeated or non-repeating. */
1186    
1187      case OP_KET:      case OP_KET:
# Line 1160  for (;;) Line 1246  for (;;)
1246          recursion_info *rec = md->recursive;          recursion_info *rec = md->recursive;
1247          DPRINTF(("Recursion (%d) succeeded - continuing\n", number));          DPRINTF(("Recursion (%d) succeeded - continuing\n", number));
1248          md->recursive = rec->prevrec;          md->recursive = rec->prevrec;
1249          md->start_match = rec->save_start;          mstart = rec->save_start;
1250          memcpy(md->offset_vector, rec->offset_save,          memcpy(md->offset_vector, rec->offset_save,
1251            rec->saved_max * sizeof(int));            rec->saved_max * sizeof(int));
1252          ecode = rec->after_call;          ecode = rec->after_call;
# Line 1189  for (;;) Line 1275  for (;;)
1275    
1276      /* 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
1277      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
1278      tail recursion to avoid using another stack frame. */      tail recursion to avoid using another stack frame, unless we have an
1279        unlimited repeat of a group that can match an empty string. */
1280    
1281      flags = (*prev >= OP_SBRA)? match_cbegroup : 0;      flags = (*prev >= OP_SBRA)? match_cbegroup : 0;
1282    
1283      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1284        {        {
1285        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);
1286        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1287          if (flags != 0)    /* Could match an empty string */
1288            {
1289            RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM50);
1290            RRETURN(rrc);
1291            }
1292        ecode = prev;        ecode = prev;
       flags |= match_tail_recursed;  
1293        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1294        }        }
1295      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
1296        {        {
1297        RMATCH(rrc, eptr, prev, offset_top, md, ims, eptrb, flags);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM13);
1298        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1299        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1300        flags = match_tail_recursed;        flags = 0;
1301        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1302        }        }
1303      /* Control never gets here */      /* Control never gets here */
# Line 1239  for (;;) Line 1330  for (;;)
1330      ecode++;      ecode++;
1331      break;      break;
1332    
1333        /* Reset the start of match point */
1334    
1335        case OP_SET_SOM:
1336        mstart = eptr;
1337        ecode++;
1338        break;
1339    
1340      /* Assert before internal newline if multiline, or before a terminating      /* Assert before internal newline if multiline, or before a terminating
1341      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. */
1342    
# Line 1335  for (;;) Line 1433  for (;;)
1433        {        {
1434        if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);        if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
1435        }        }
1436        /* Fall through */
1437    
1438        case OP_ALLANY:
1439      if (eptr++ >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr++ >= md->end_subject) RRETURN(MATCH_NOMATCH);
1440      if (utf8)      if (utf8)
1441        while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;        while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
# Line 1436  for (;;) Line 1537  for (;;)
1537        case 0x000d:        case 0x000d:
1538        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
1539        break;        break;
1540    
1541        case 0x000a:        case 0x000a:
1542          break;
1543    
1544        case 0x000b:        case 0x000b:
1545        case 0x000c:        case 0x000c:
1546        case 0x0085:        case 0x0085:
1547        case 0x2028:        case 0x2028:
1548        case 0x2029:        case 0x2029:
1549          if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
1550          break;
1551          }
1552        ecode++;
1553        break;
1554    
1555        case OP_NOT_HSPACE:
1556        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1557        GETCHARINCTEST(c, eptr);
1558        switch(c)
1559          {
1560          default: break;
1561          case 0x09:      /* HT */
1562          case 0x20:      /* SPACE */
1563          case 0xa0:      /* NBSP */
1564          case 0x1680:    /* OGHAM SPACE MARK */
1565          case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
1566          case 0x2000:    /* EN QUAD */
1567          case 0x2001:    /* EM QUAD */
1568          case 0x2002:    /* EN SPACE */
1569          case 0x2003:    /* EM SPACE */
1570          case 0x2004:    /* THREE-PER-EM SPACE */
1571          case 0x2005:    /* FOUR-PER-EM SPACE */
1572          case 0x2006:    /* SIX-PER-EM SPACE */
1573          case 0x2007:    /* FIGURE SPACE */
1574          case 0x2008:    /* PUNCTUATION SPACE */
1575          case 0x2009:    /* THIN SPACE */
1576          case 0x200A:    /* HAIR SPACE */
1577          case 0x202f:    /* NARROW NO-BREAK SPACE */
1578          case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
1579          case 0x3000:    /* IDEOGRAPHIC SPACE */
1580          RRETURN(MATCH_NOMATCH);
1581          }
1582        ecode++;
1583        break;
1584    
1585        case OP_HSPACE:
1586        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1587        GETCHARINCTEST(c, eptr);
1588        switch(c)
1589          {
1590          default: RRETURN(MATCH_NOMATCH);
1591          case 0x09:      /* HT */
1592          case 0x20:      /* SPACE */
1593          case 0xa0:      /* NBSP */
1594          case 0x1680:    /* OGHAM SPACE MARK */
1595          case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
1596          case 0x2000:    /* EN QUAD */
1597          case 0x2001:    /* EM QUAD */
1598          case 0x2002:    /* EN SPACE */
1599          case 0x2003:    /* EM SPACE */
1600          case 0x2004:    /* THREE-PER-EM SPACE */
1601          case 0x2005:    /* FOUR-PER-EM SPACE */
1602          case 0x2006:    /* SIX-PER-EM SPACE */
1603          case 0x2007:    /* FIGURE SPACE */
1604          case 0x2008:    /* PUNCTUATION SPACE */
1605          case 0x2009:    /* THIN SPACE */
1606          case 0x200A:    /* HAIR SPACE */
1607          case 0x202f:    /* NARROW NO-BREAK SPACE */
1608          case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
1609          case 0x3000:    /* IDEOGRAPHIC SPACE */
1610          break;
1611          }
1612        ecode++;
1613        break;
1614    
1615        case OP_NOT_VSPACE:
1616        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1617        GETCHARINCTEST(c, eptr);
1618        switch(c)
1619          {
1620          default: break;
1621          case 0x0a:      /* LF */
1622          case 0x0b:      /* VT */
1623          case 0x0c:      /* FF */
1624          case 0x0d:      /* CR */
1625          case 0x85:      /* NEL */
1626          case 0x2028:    /* LINE SEPARATOR */
1627          case 0x2029:    /* PARAGRAPH SEPARATOR */
1628          RRETURN(MATCH_NOMATCH);
1629          }
1630        ecode++;
1631        break;
1632    
1633        case OP_VSPACE:
1634        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1635        GETCHARINCTEST(c, eptr);
1636        switch(c)
1637          {
1638          default: RRETURN(MATCH_NOMATCH);
1639          case 0x0a:      /* LF */
1640          case 0x0b:      /* VT */
1641          case 0x0c:      /* FF */
1642          case 0x0d:      /* CR */
1643          case 0x85:      /* NEL */
1644          case 0x2028:    /* LINE SEPARATOR */
1645          case 0x2029:    /* PARAGRAPH SEPARATOR */
1646        break;        break;
1647        }        }
1648      ecode++;      ecode++;
# Line 1533  for (;;) Line 1734  for (;;)
1734      case OP_REF:      case OP_REF:
1735        {        {
1736        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
1737        ecode += 3;                                 /* Advance past item */        ecode += 3;
1738    
1739        /* If the reference is unset, set the length to be longer than the amount        /* If the reference is unset, there are two possibilities:
1740        of subject left; this ensures that every attempt at a match fails. We  
1741        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
1742        minima. */        than the amount of subject left; this ensures that every attempt at a
1743          match fails. We can't just fail here, because of the possibility of
1744        length = (offset >= offset_top || md->offset_vector[offset] < 0)?        quantifiers with zero minima.
1745          md->end_subject - eptr + 1 :  
1746          md->offset_vector[offset+1] - md->offset_vector[offset];        (b) If the JavaScript compatibility flag is set, set the length to zero
1747          so that the back reference matches an empty string.
1748    
1749          Otherwise, set the length to the length of what was matched by the
1750          referenced subpattern. */
1751    
1752          if (offset >= offset_top || md->offset_vector[offset] < 0)
1753            length = (md->jscript_compat)? 0 : md->end_subject - eptr + 1;
1754          else
1755            length = md->offset_vector[offset+1] - md->offset_vector[offset];
1756    
1757        /* Set up for repetition, or handle the non-repeated case */        /* Set up for repetition, or handle the non-repeated case */
1758    
# Line 1602  for (;;) Line 1812  for (;;)
1812          {          {
1813          for (fi = min;; fi++)          for (fi = min;; fi++)
1814            {            {
1815            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM14);
1816            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1817            if (fi >= max || !match_ref(offset, eptr, length, md, ims))            if (fi >= max || !match_ref(offset, eptr, length, md, ims))
1818              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
# Line 1623  for (;;) Line 1833  for (;;)
1833            }            }
1834          while (eptr >= pp)          while (eptr >= pp)
1835            {            {
1836            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM15);
1837            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1838            eptr -= length;            eptr -= length;
1839            }            }
# Line 1728  for (;;) Line 1938  for (;;)
1938            {            {
1939            for (fi = min;; fi++)            for (fi = min;; fi++)
1940              {              {
1941              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM16);
1942              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1943              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1944              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 1748  for (;;) Line 1958  for (;;)
1958            {            {
1959            for (fi = min;; fi++)            for (fi = min;; fi++)
1960              {              {
1961              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM17);
1962              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1963              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1964              c = *eptr++;              c = *eptr++;
# Line 1785  for (;;) Line 1995  for (;;)
1995              }              }
1996            for (;;)            for (;;)
1997              {              {
1998              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM18);
1999              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2000              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
2001              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 1804  for (;;) Line 2014  for (;;)
2014              }              }
2015            while (eptr >= pp)            while (eptr >= pp)
2016              {              {
2017              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM19);
2018              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2019              eptr--;              eptr--;
2020              }              }
# Line 1875  for (;;) Line 2085  for (;;)
2085          {          {
2086          for (fi = min;; fi++)          for (fi = min;; fi++)
2087            {            {
2088            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM20);
2089            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2090            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2091            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
# Line 1899  for (;;) Line 2109  for (;;)
2109            }            }
2110          for(;;)          for(;;)
2111            {            {
2112            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM21);
2113            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2114            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
2115            BACKCHAR(eptr)            if (utf8) BACKCHAR(eptr);
2116            }            }
2117          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2118          }          }
# Line 2086  for (;;) Line 2296  for (;;)
2296            {            {
2297            for (fi = min;; fi++)            for (fi = min;; fi++)
2298              {              {
2299              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM22);
2300              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2301              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2302              if (memcmp(eptr, charptr, length) == 0) eptr += length;              if (memcmp(eptr, charptr, length) == 0) eptr += length;
# Line 2127  for (;;) Line 2337  for (;;)
2337            if (possessive) continue;            if (possessive) continue;
2338            for(;;)            for(;;)
2339             {             {
2340             RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);             RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM23);
2341             if (rrc != MATCH_NOMATCH) RRETURN(rrc);             if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2342             if (eptr == pp) RRETURN(MATCH_NOMATCH);             if (eptr == pp) RRETURN(MATCH_NOMATCH);
2343  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
# Line 2176  for (;;) Line 2386  for (;;)
2386          {          {
2387          for (fi = min;; fi++)          for (fi = min;; fi++)
2388            {            {
2389            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM24);
2390            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2391            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
2392                fc != md->lcc[*eptr++])                fc != md->lcc[*eptr++])
# Line 2195  for (;;) Line 2405  for (;;)
2405          if (possessive) continue;          if (possessive) continue;
2406          while (eptr >= pp)          while (eptr >= pp)
2407            {            {
2408            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM25);
2409            eptr--;            eptr--;
2410            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2411            }            }
# Line 2214  for (;;) Line 2424  for (;;)
2424          {          {
2425          for (fi = min;; fi++)          for (fi = min;; fi++)
2426            {            {
2427            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM26);
2428            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2429            if (fi >= max || eptr >= md->end_subject || fc != *eptr++)            if (fi >= max || eptr >= md->end_subject || fc != *eptr++)
2430              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
# Line 2232  for (;;) Line 2442  for (;;)
2442          if (possessive) continue;          if (possessive) continue;
2443          while (eptr >= pp)          while (eptr >= pp)
2444            {            {
2445            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM27);
2446            eptr--;            eptr--;
2447            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2448            }            }
# Line 2377  for (;;) Line 2587  for (;;)
2587            register unsigned int d;            register unsigned int d;
2588            for (fi = min;; fi++)            for (fi = min;; fi++)
2589              {              {
2590              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM28);
2591              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2592              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
2593              if (d < 256) d = md->lcc[d];              if (d < 256) d = md->lcc[d];
# Line 2391  for (;;) Line 2601  for (;;)
2601            {            {
2602            for (fi = min;; fi++)            for (fi = min;; fi++)
2603              {              {
2604              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM29);
2605              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2606              if (fi >= max || eptr >= md->end_subject || fc == md->lcc[*eptr++])              if (fi >= max || eptr >= md->end_subject || fc == md->lcc[*eptr++])
2607                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2423  for (;;) Line 2633  for (;;)
2633          if (possessive) continue;          if (possessive) continue;
2634          for(;;)          for(;;)
2635              {              {
2636              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM30);
2637              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2638              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
2639              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 2441  for (;;) Line 2651  for (;;)
2651            if (possessive) continue;            if (possessive) continue;
2652            while (eptr >= pp)            while (eptr >= pp)
2653              {              {
2654              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM31);
2655              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2656              eptr--;              eptr--;
2657              }              }
# Line 2486  for (;;) Line 2696  for (;;)
2696            register unsigned int d;            register unsigned int d;
2697            for (fi = min;; fi++)            for (fi = min;; fi++)
2698              {              {
2699              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM32);
2700              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2701              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
2702              if (fi >= max || eptr >= md->end_subject || fc == d)              if (fi >= max || eptr >= md->end_subject || fc == d)
# Line 2499  for (;;) Line 2709  for (;;)
2709            {            {
2710            for (fi = min;; fi++)            for (fi = min;; fi++)
2711              {              {
2712              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM33);
2713              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2714              if (fi >= max || eptr >= md->end_subject || fc == *eptr++)              if (fi >= max || eptr >= md->end_subject || fc == *eptr++)
2715                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2530  for (;;) Line 2740  for (;;)
2740            if (possessive) continue;            if (possessive) continue;
2741            for(;;)            for(;;)
2742              {              {
2743              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM34);
2744              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2745              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
2746              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 2548  for (;;) Line 2758  for (;;)
2758            if (possessive) continue;            if (possessive) continue;
2759            while (eptr >= pp)            while (eptr >= pp)
2760              {              {
2761              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM35);
2762              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2763              eptr--;              eptr--;
2764              }              }
# Line 2655  for (;;) Line 2865  for (;;)
2865            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2866              {              {
2867              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2868              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2869              }              }
2870            break;            break;
2871    
# Line 2663  for (;;) Line 2873  for (;;)
2873            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2874              {              {
2875              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2876              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2877              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);
2878              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
2879                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
# Line 2676  for (;;) Line 2886  for (;;)
2886            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2887              {              {
2888              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2889              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2890              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);
2891              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
2892                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2687  for (;;) Line 2897  for (;;)
2897            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2898              {              {
2899              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2900              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2901              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);
2902              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
2903                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2698  for (;;) Line 2908  for (;;)
2908            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2909              {              {
2910              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2911              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2912              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);
2913              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
2914                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2753  for (;;) Line 2963  for (;;)
2963            }            }
2964          break;          break;
2965    
2966            case OP_ALLANY:
2967            for (i = 1; i <= min; i++)
2968              {
2969              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2970              eptr++;
2971              while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
2972              }
2973            break;
2974    
2975          case OP_ANYBYTE:          case OP_ANYBYTE:
2976          eptr += min;          eptr += min;
2977          break;          break;
# Line 2768  for (;;) Line 2987  for (;;)
2987              case 0x000d:              case 0x000d:
2988              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
2989              break;              break;
2990    
2991              case 0x000a:              case 0x000a:
2992                break;
2993    
2994              case 0x000b:              case 0x000b:
2995              case 0x000c:              case 0x000c:
2996              case 0x0085:              case 0x0085:
2997              case 0x2028:              case 0x2028:
2998              case 0x2029:              case 0x2029:
2999                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3000                break;
3001                }
3002              }
3003            break;
3004    
3005            case OP_NOT_HSPACE:
3006            for (i = 1; i <= min; i++)
3007              {
3008              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3009              GETCHARINC(c, eptr);
3010              switch(c)
3011                {
3012                default: break;
3013                case 0x09:      /* HT */
3014                case 0x20:      /* SPACE */
3015                case 0xa0:      /* NBSP */
3016                case 0x1680:    /* OGHAM SPACE MARK */
3017                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3018                case 0x2000:    /* EN QUAD */
3019                case 0x2001:    /* EM QUAD */
3020                case 0x2002:    /* EN SPACE */
3021                case 0x2003:    /* EM SPACE */
3022                case 0x2004:    /* THREE-PER-EM SPACE */
3023                case 0x2005:    /* FOUR-PER-EM SPACE */
3024                case 0x2006:    /* SIX-PER-EM SPACE */
3025                case 0x2007:    /* FIGURE SPACE */
3026                case 0x2008:    /* PUNCTUATION SPACE */
3027                case 0x2009:    /* THIN SPACE */
3028                case 0x200A:    /* HAIR SPACE */
3029                case 0x202f:    /* NARROW NO-BREAK SPACE */
3030                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3031                case 0x3000:    /* IDEOGRAPHIC SPACE */
3032                RRETURN(MATCH_NOMATCH);
3033                }
3034              }
3035            break;
3036    
3037            case OP_HSPACE:
3038            for (i = 1; i <= min; i++)
3039              {
3040              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3041              GETCHARINC(c, eptr);
3042              switch(c)
3043                {
3044                default: RRETURN(MATCH_NOMATCH);
3045                case 0x09:      /* HT */
3046                case 0x20:      /* SPACE */
3047                case 0xa0:      /* NBSP */
3048                case 0x1680:    /* OGHAM SPACE MARK */
3049                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3050                case 0x2000:    /* EN QUAD */
3051                case 0x2001:    /* EM QUAD */
3052                case 0x2002:    /* EN SPACE */
3053                case 0x2003:    /* EM SPACE */
3054                case 0x2004:    /* THREE-PER-EM SPACE */
3055                case 0x2005:    /* FOUR-PER-EM SPACE */
3056                case 0x2006:    /* SIX-PER-EM SPACE */
3057                case 0x2007:    /* FIGURE SPACE */
3058                case 0x2008:    /* PUNCTUATION SPACE */
3059                case 0x2009:    /* THIN SPACE */
3060                case 0x200A:    /* HAIR SPACE */
3061                case 0x202f:    /* NARROW NO-BREAK SPACE */
3062                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3063                case 0x3000:    /* IDEOGRAPHIC SPACE */
3064                break;
3065                }
3066              }
3067            break;
3068    
3069            case OP_NOT_VSPACE:
3070            for (i = 1; i <= min; i++)
3071              {
3072              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3073              GETCHARINC(c, eptr);
3074              switch(c)
3075                {
3076                default: break;
3077                case 0x0a:      /* LF */
3078                case 0x0b:      /* VT */
3079                case 0x0c:      /* FF */
3080                case 0x0d:      /* CR */
3081                case 0x85:      /* NEL */
3082                case 0x2028:    /* LINE SEPARATOR */
3083                case 0x2029:    /* PARAGRAPH SEPARATOR */
3084                RRETURN(MATCH_NOMATCH);
3085                }
3086              }
3087            break;
3088    
3089            case OP_VSPACE:
3090            for (i = 1; i <= min; i++)
3091              {
3092              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3093              GETCHARINC(c, eptr);
3094              switch(c)
3095                {
3096                default: RRETURN(MATCH_NOMATCH);
3097                case 0x0a:      /* LF */
3098                case 0x0b:      /* VT */
3099                case 0x0c:      /* FF */
3100                case 0x0d:      /* CR */
3101                case 0x85:      /* NEL */
3102                case 0x2028:    /* LINE SEPARATOR */
3103                case 0x2029:    /* PARAGRAPH SEPARATOR */
3104              break;              break;
3105              }              }
3106            }            }
# Line 2803  for (;;) Line 3130  for (;;)
3130          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3131            {            {
3132            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject ||
3133               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_space) != 0))               (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0))
3134              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3135            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3136            }            }
3137          break;          break;
3138    
# Line 2823  for (;;) Line 3150  for (;;)
3150          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3151            {            {
3152            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject ||
3153               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_word) != 0))               (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0))
3154              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3155            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3156            }            }
3157          break;          break;
3158    
# Line 2864  for (;;) Line 3191  for (;;)
3191          else eptr += min;          else eptr += min;
3192          break;          break;
3193    
3194            case OP_ALLANY:
3195            eptr += min;
3196            break;
3197    
3198          case OP_ANYBYTE:          case OP_ANYBYTE:
3199          eptr += min;          eptr += min;
3200          break;          break;
# Line 2882  for (;;) Line 3213  for (;;)
3213              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3214              break;              break;
3215              case 0x000a:              case 0x000a:
3216                break;
3217    
3218              case 0x000b:              case 0x000b:
3219              case 0x000c:              case 0x000c:
3220              case 0x0085:              case 0x0085:
3221                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3222                break;
3223                }
3224              }
3225            break;
3226    
3227            case OP_NOT_HSPACE:
3228            for (i = 1; i <= min; i++)
3229              {
3230              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3231              switch(*eptr++)
3232                {
3233                default: break;
3234                case 0x09:      /* HT */
3235                case 0x20:      /* SPACE */
3236                case 0xa0:      /* NBSP */
3237                RRETURN(MATCH_NOMATCH);
3238                }
3239              }
3240            break;
3241    
3242            case OP_HSPACE:
3243            for (i = 1; i <= min; i++)
3244              {
3245              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3246              switch(*eptr++)
3247                {
3248                default: RRETURN(MATCH_NOMATCH);
3249                case 0x09:      /* HT */
3250                case 0x20:      /* SPACE */
3251                case 0xa0:      /* NBSP */
3252                break;
3253                }
3254              }
3255            break;
3256    
3257            case OP_NOT_VSPACE:
3258            for (i = 1; i <= min; i++)
3259              {
3260              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3261              switch(*eptr++)
3262                {
3263                default: break;
3264                case 0x0a:      /* LF */
3265                case 0x0b:      /* VT */
3266                case 0x0c:      /* FF */
3267                case 0x0d:      /* CR */
3268                case 0x85:      /* NEL */
3269                RRETURN(MATCH_NOMATCH);
3270                }
3271              }
3272            break;
3273    
3274            case OP_VSPACE:
3275            for (i = 1; i <= min; i++)
3276              {
3277              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3278              switch(*eptr++)
3279                {
3280                default: RRETURN(MATCH_NOMATCH);
3281                case 0x0a:      /* LF */
3282                case 0x0b:      /* VT */
3283                case 0x0c:      /* FF */
3284                case 0x0d:      /* CR */
3285                case 0x85:      /* NEL */
3286              break;              break;
3287              }              }
3288            }            }
# Line 2945  for (;;) Line 3343  for (;;)
3343            case PT_ANY:            case PT_ANY:
3344            for (fi = min;; fi++)            for (fi = min;; fi++)
3345              {              {
3346              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM36);
3347              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3348              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3349              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 2956  for (;;) Line 3354  for (;;)
3354            case PT_LAMP:            case PT_LAMP:
3355            for (fi = min;; fi++)            for (fi = min;; fi++)
3356              {              {
3357              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM37);
3358              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3359              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3360              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 2971  for (;;) Line 3369  for (;;)
3369            case PT_GC:            case PT_GC:
3370            for (fi = min;; fi++)            for (fi = min;; fi++)
3371              {              {
3372              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM38);
3373              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3374              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3375              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 2984  for (;;) Line 3382  for (;;)
3382            case PT_PC:            case PT_PC:
3383            for (fi = min;; fi++)            for (fi = min;; fi++)
3384              {              {
3385              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM39);
3386              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3387              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3388              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 2997  for (;;) Line 3395  for (;;)
3395            case PT_SC:            case PT_SC:
3396            for (fi = min;; fi++)            for (fi = min;; fi++)
3397              {              {
3398              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM40);
3399              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3400              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3401              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 3019  for (;;) Line 3417  for (;;)
3417          {          {
3418          for (fi = min;; fi++)          for (fi = min;; fi++)
3419            {            {
3420            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM41);
3421            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3422            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3423            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
# Line 3048  for (;;) Line 3446  for (;;)
3446          {          {
3447          for (fi = min;; fi++)          for (fi = min;; fi++)
3448            {            {
3449            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM42);
3450            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3451            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
3452                 (ctype == OP_ANY && (ims & PCRE_DOTALL) == 0 &&                 (ctype == OP_ANY && (ims & PCRE_DOTALL) == 0 &&
# Line 3059  for (;;) Line 3457  for (;;)
3457            switch(ctype)            switch(ctype)
3458              {              {
3459              case OP_ANY:        /* This is the DOTALL case */              case OP_ANY:        /* This is the DOTALL case */
3460              break;              case OP_ALLANY:
   
3461              case OP_ANYBYTE:              case OP_ANYBYTE:
3462              break;              break;
3463    
# Line 3072  for (;;) Line 3469  for (;;)
3469                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3470                break;                break;
3471                case 0x000a:                case 0x000a:
3472                  break;
3473    
3474                case 0x000b:                case 0x000b:
3475                case 0x000c:                case 0x000c:
3476                case 0x0085:                case 0x0085:
3477                case 0x2028:                case 0x2028:
3478                case 0x2029:                case 0x2029:
3479                  if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3480                  break;
3481                  }
3482                break;
3483    
3484                case OP_NOT_HSPACE:
3485                switch(c)
3486                  {
3487                  default: break;
3488                  case 0x09:      /* HT */
3489                  case 0x20:      /* SPACE */
3490                  case 0xa0:      /* NBSP */
3491                  case 0x1680:    /* OGHAM SPACE MARK */
3492                  case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3493                  case 0x2000:    /* EN QUAD */
3494                  case 0x2001:    /* EM QUAD */
3495                  case 0x2002:    /* EN SPACE */
3496                  case 0x2003:    /* EM SPACE */
3497                  case 0x2004:    /* THREE-PER-EM SPACE */
3498                  case 0x2005:    /* FOUR-PER-EM SPACE */
3499                  case 0x2006:    /* SIX-PER-EM SPACE */
3500                  case 0x2007:    /* FIGURE SPACE */
3501                  case 0x2008:    /* PUNCTUATION SPACE */
3502                  case 0x2009:    /* THIN SPACE */
3503                  case 0x200A:    /* HAIR SPACE */
3504                  case 0x202f:    /* NARROW NO-BREAK SPACE */
3505                  case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3506                  case 0x3000:    /* IDEOGRAPHIC SPACE */
3507                  RRETURN(MATCH_NOMATCH);
3508                  }
3509                break;
3510    
3511                case OP_HSPACE:
3512                switch(c)
3513                  {
3514                  default: RRETURN(MATCH_NOMATCH);
3515                  case 0x09:      /* HT */
3516                  case 0x20:      /* SPACE */
3517                  case 0xa0:      /* NBSP */
3518                  case 0x1680:    /* OGHAM SPACE MARK */
3519                  case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3520                  case 0x2000:    /* EN QUAD */
3521                  case 0x2001:    /* EM QUAD */
3522                  case 0x2002:    /* EN SPACE */
3523                  case 0x2003:    /* EM SPACE */
3524                  case 0x2004:    /* THREE-PER-EM SPACE */
3525                  case 0x2005:    /* FOUR-PER-EM SPACE */
3526                  case 0x2006:    /* SIX-PER-EM SPACE */
3527                  case 0x2007:    /* FIGURE SPACE */
3528                  case 0x2008:    /* PUNCTUATION SPACE */
3529                  case 0x2009:    /* THIN SPACE */
3530                  case 0x200A:    /* HAIR SPACE */
3531                  case 0x202f:    /* NARROW NO-BREAK SPACE */
3532                  case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3533                  case 0x3000:    /* IDEOGRAPHIC SPACE */
3534                  break;
3535                  }
3536                break;
3537    
3538                case OP_NOT_VSPACE:
3539                switch(c)
3540                  {
3541                  default: break;
3542                  case 0x0a:      /* LF */
3543                  case 0x0b:      /* VT */
3544                  case 0x0c:      /* FF */
3545                  case 0x0d:      /* CR */
3546                  case 0x85:      /* NEL */
3547                  case 0x2028:    /* LINE SEPARATOR */
3548                  case 0x2029:    /* PARAGRAPH SEPARATOR */
3549                  RRETURN(MATCH_NOMATCH);
3550                  }
3551                break;
3552    
3553                case OP_VSPACE:
3554                switch(c)
3555                  {
3556                  default: RRETURN(MATCH_NOMATCH);
3557                  case 0x0a:      /* LF */
3558                  case 0x0b:      /* VT */
3559                  case 0x0c:      /* FF */
3560                  case 0x0d:      /* CR */
3561                  case 0x85:      /* NEL */
3562                  case 0x2028:    /* LINE SEPARATOR */
3563                  case 0x2029:    /* PARAGRAPH SEPARATOR */
3564                break;                break;
3565                }                }
3566              break;              break;
# Line 3122  for (;;) Line 3606  for (;;)
3606          {          {
3607          for (fi = min;; fi++)          for (fi = min;; fi++)
3608            {            {
3609            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM43);
3610            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3611            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
3612                 ((ims & PCRE_DOTALL) == 0 && IS_NEWLINE(eptr)))                 ((ims & PCRE_DOTALL) == 0 && IS_NEWLINE(eptr)))
# Line 3131  for (;;) Line 3615  for (;;)
3615            c = *eptr++;            c = *eptr++;
3616            switch(ctype)            switch(ctype)
3617              {              {
3618              case OP_ANY:   /* This is the DOTALL case */              case OP_ANY:     /* This is the DOTALL case */
3619              break;              case OP_ALLANY:
   
3620              case OP_ANYBYTE:              case OP_ANYBYTE:
3621              break;              break;
3622    
# Line 3144  for (;;) Line 3627  for (;;)
3627                case 0x000d:                case 0x000d:
3628                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3629                break;                break;
3630    
3631                case 0x000a:                case 0x000a:
3632                  break;
3633    
3634                case 0x000b:                case 0x000b:
3635                case 0x000c:                case 0x000c:
3636                case 0x0085:                case 0x0085:
3637                  if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3638                  break;
3639                  }
3640                break;
3641    
3642                case OP_NOT_HSPACE:
3643                switch(c)
3644                  {
3645                  default: break;
3646                  case 0x09:      /* HT */
3647                  case 0x20:      /* SPACE */
3648                  case 0xa0:      /* NBSP */
3649                  RRETURN(MATCH_NOMATCH);
3650                  }
3651                break;
3652    
3653                case OP_HSPACE:
3654                switch(c)
3655                  {
3656                  default: RRETURN(MATCH_NOMATCH);
3657                  case 0x09:      /* HT */
3658                  case 0x20:      /* SPACE */
3659                  case 0xa0:      /* NBSP */
3660                  break;
3661                  }
3662                break;
3663    
3664                case OP_NOT_VSPACE:
3665                switch(c)
3666                  {
3667                  default: break;
3668                  case 0x0a:      /* LF */
3669                  case 0x0b:      /* VT */
3670                  case 0x0c:      /* FF */
3671                  case 0x0d:      /* CR */
3672                  case 0x85:      /* NEL */
3673                  RRETURN(MATCH_NOMATCH);
3674                  }
3675                break;
3676    
3677                case OP_VSPACE:
3678                switch(c)
3679                  {
3680                  default: RRETURN(MATCH_NOMATCH);
3681                  case 0x0a:      /* LF */
3682                  case 0x0b:      /* VT */
3683                  case 0x0c:      /* FF */
3684                  case 0x0d:      /* CR */
3685                  case 0x85:      /* NEL */
3686                break;                break;
3687                }                }
3688              break;              break;
# Line 3268  for (;;) Line 3803  for (;;)
3803          if (possessive) continue;          if (possessive) continue;
3804          for(;;)          for(;;)
3805            {            {
3806            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM44);
3807            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3808            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3809            BACKCHAR(eptr);            if (utf8) BACKCHAR(eptr);
3810            }            }
3811          }          }
3812    
# Line 3304  for (;;) Line 3839  for (;;)
3839          if (possessive) continue;          if (possessive) continue;
3840          for(;;)          for(;;)
3841            {            {
3842            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM45);
3843            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3844            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3845            for (;;)                        /* Move back over one extended */            for (;;)                        /* Move back over one extended */
3846              {              {
3847              int len = 1;              int len = 1;
             BACKCHAR(eptr);  
3848              if (!utf8) c = *eptr; else              if (!utf8) c = *eptr; else
3849                {                {
3850                  BACKCHAR(eptr);
3851                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
3852                }                }
3853              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);
# Line 3333  for (;;) Line 3868  for (;;)
3868          switch(ctype)          switch(ctype)
3869            {            {
3870            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. */  
   
3871            if (max < INT_MAX)            if (max < INT_MAX)
3872              {              {
3873              if ((ims & PCRE_DOTALL) == 0)              if ((ims & PCRE_DOTALL) == 0)
# Line 3370  for (;;) Line 3900  for (;;)
3900                  {                  {
3901                  if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;                  if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
3902                  eptr++;                  eptr++;
3903                    while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
3904                  }                  }
               break;  
3905                }                }
3906              else              else
3907                {                {
3908                c = max - min;                eptr = md->end_subject;
3909                if (c > (unsigned int)(md->end_subject - eptr))                }
3910                  c = md->end_subject - eptr;              }
3911                eptr += c;            break;
3912    
3913              case OP_ALLANY:
3914              if (max < INT_MAX)
3915                {
3916                for (i = min; i < max; i++)
3917                  {
3918                  if (eptr >= md->end_subject) break;
3919                  eptr++;
3920                  while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
3921                }                }
3922              }              }
3923              else eptr = md->end_subject;   /* Unlimited UTF-8 repeat */
3924            break;            break;
3925    
3926            /* The byte case is the same as non-UTF8 */            /* The byte case is the same as non-UTF8 */
# Line 3405  for (;;) Line 3945  for (;;)
3945                }                }
3946              else              else
3947                {                {
3948                if (c != 0x000a && c != 0x000b && c != 0x000c &&                if (c != 0x000a &&
3949                    c != 0x0085 && c != 0x2028 && c != 0x2029)                    (md->bsr_anycrlf ||
3950                       (c != 0x000b && c != 0x000c &&
3951                        c != 0x0085 && c != 0x2028 && c != 0x2029)))
3952                  break;                  break;
3953                eptr += len;                eptr += len;
3954                }                }
3955              }              }
3956            break;            break;
3957    
3958              case OP_NOT_HSPACE:
3959              case OP_HSPACE:
3960              for (i = min; i < max; i++)
3961                {
3962                BOOL gotspace;
3963                int len = 1;
3964                if (eptr >= md->end_subject) break;
3965                GETCHARLEN(c, eptr, len);
3966                switch(c)
3967                  {
3968                  default: gotspace = FALSE; break;
3969                  case 0x09:      /* HT */
3970                  case 0x20:      /* SPACE */
3971                  case 0xa0:      /* NBSP */
3972                  case 0x1680:    /* OGHAM SPACE MARK */
3973                  case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3974                  case 0x2000:    /* EN QUAD */
3975                  case 0x2001:    /* EM QUAD */
3976                  case 0x2002:    /* EN SPACE */
3977                  case 0x2003:    /* EM SPACE */
3978                  case 0x2004:    /* THREE-PER-EM SPACE */
3979                  case 0x2005:    /* FOUR-PER-EM SPACE */
3980                  case 0x2006:    /* SIX-PER-EM SPACE */
3981                  case 0x2007:    /* FIGURE SPACE */
3982                  case 0x2008:    /* PUNCTUATION SPACE */
3983                  case 0x2009:    /* THIN SPACE */
3984                  case 0x200A:    /* HAIR SPACE */
3985                  case 0x202f:    /* NARROW NO-BREAK SPACE */
3986                  case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3987                  case 0x3000:    /* IDEOGRAPHIC SPACE */
3988                  gotspace = TRUE;
3989                  break;
3990                  }
3991                if (gotspace == (ctype == OP_NOT_HSPACE)) break;
3992                eptr += len;
3993                }
3994              break;
3995    
3996              case OP_NOT_VSPACE:
3997              case OP_VSPACE:
3998              for (i = min; i < max; i++)
3999                {
4000                BOOL gotspace;
4001                int len = 1;
4002                if (eptr >= md->end_subject) break;
4003                GETCHARLEN(c, eptr, len);
4004                switch(c)
4005                  {
4006                  default: gotspace = FALSE; break;
4007                  case 0x0a:      /* LF */
4008                  case 0x0b:      /* VT */
4009                  case 0x0c:      /* FF */
4010                  case 0x0d:      /* CR */
4011                  case 0x85:      /* NEL */
4012                  case 0x2028:    /* LINE SEPARATOR */
4013                  case 0x2029:    /* PARAGRAPH SEPARATOR */
4014                  gotspace = TRUE;
4015                  break;
4016                  }
4017                if (gotspace == (ctype == OP_NOT_VSPACE)) break;
4018                eptr += len;
4019                }
4020              break;
4021    
4022            case OP_NOT_DIGIT:            case OP_NOT_DIGIT:
4023            for (i = min; i < max; i++)            for (i = min; i < max; i++)
4024              {              {
# Line 3488  for (;;) Line 4094  for (;;)
4094          if (possessive) continue;          if (possessive) continue;
4095          for(;;)          for(;;)
4096            {            {
4097            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM46);
4098            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4099            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
4100            BACKCHAR(eptr);            BACKCHAR(eptr);
4101            }            }
4102          }          }
4103        else        else
4104  #endif  #endif  /* SUPPORT_UTF8 */
4105    
4106        /* Not UTF-8 mode */        /* Not UTF-8 mode */
4107          {          {
# Line 3511  for (;;) Line 4117  for (;;)
4117                }                }
4118              break;              break;
4119              }              }
4120            /* For DOTALL case, fall through and treat as \C */            /* For DOTALL case, fall through */
4121    
4122              case OP_ALLANY:
4123            case OP_ANYBYTE:            case OP_ANYBYTE:
4124            c = max - min;            c = max - min;
4125            if (c > (unsigned int)(md->end_subject - eptr))            if (c > (unsigned int)(md->end_subject - eptr))
# Line 3532  for (;;) Line 4139  for (;;)
4139                }                }
4140              else              else
4141                {                {
4142                if (c != 0x000a && c != 0x000b && c != 0x000c && c != 0x0085)                if (c != 0x000a &&
4143                      (md->bsr_anycrlf ||
4144                        (c != 0x000b && c != 0x000c && c != 0x0085)))
4145                  break;                  break;
4146                eptr++;                eptr++;
4147                }                }
4148              }              }
4149            break;            break;
4150    
4151              case OP_NOT_HSPACE:
4152              for (i = min; i < max; i++)
4153                {
4154                if (eptr >= md->end_subject) break;
4155                c = *eptr;
4156                if (c == 0x09 || c == 0x20 || c == 0xa0) break;
4157                eptr++;
4158                }
4159              break;
4160    
4161              case OP_HSPACE:
4162              for (i = min; i < max; i++)
4163                {
4164                if (eptr >= md->end_subject) break;
4165                c = *eptr;
4166                if (c != 0x09 && c != 0x20 && c != 0xa0) break;
4167                eptr++;
4168                }
4169              break;
4170    
4171              case OP_NOT_VSPACE:
4172              for (i = min; i < max; i++)
4173                {
4174                if (eptr >= md->end_subject) break;
4175                c = *eptr;
4176                if (c == 0x0a || c == 0x0b || c == 0x0c || c == 0x0d || c == 0x85)
4177                  break;
4178                eptr++;
4179                }
4180              break;
4181    
4182              case OP_VSPACE:
4183              for (i = min; i < max; i++)
4184                {
4185                if (eptr >= md->end_subject) break;
4186                c = *eptr;
4187                if (c != 0x0a && c != 0x0b && c != 0x0c && c != 0x0d && c != 0x85)
4188                  break;
4189                eptr++;
4190                }
4191              break;
4192    
4193            case OP_NOT_DIGIT:            case OP_NOT_DIGIT:
4194            for (i = min; i < max; i++)            for (i = min; i < max; i++)
4195              {              {
# Line 3602  for (;;) Line 4253  for (;;)
4253          if (possessive) continue;          if (possessive) continue;
4254          while (eptr >= pp)          while (eptr >= pp)
4255            {            {
4256            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM47);
4257            eptr--;            eptr--;
4258            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4259            }            }
# Line 3628  for (;;) Line 4279  for (;;)
4279    
4280    }             /* End of main loop */    }             /* End of main loop */
4281  /* Control never reaches here */  /* Control never reaches here */
4282    
4283    
4284    /* When compiling to use the heap rather than the stack for recursive calls to
4285    match(), the RRETURN() macro jumps here. The number that is saved in
4286    frame->Xwhere indicates which label we actually want to return to. */
4287    
4288    #ifdef NO_RECURSE
4289    #define LBL(val) case val: goto L_RM##val;
4290    HEAP_RETURN:
4291    switch (frame->Xwhere)
4292      {
4293      LBL( 1) LBL( 2) LBL( 3) LBL( 4) LBL( 5) LBL( 6) LBL( 7) LBL( 8)
4294      LBL( 9) LBL(10) LBL(11) LBL(12) LBL(13) LBL(14) LBL(15) LBL(17)
4295      LBL(19) LBL(24) LBL(25) LBL(26) LBL(27) LBL(29) LBL(31) LBL(33)
4296      LBL(35) LBL(43) LBL(47) LBL(48) LBL(49) LBL(50) LBL(51) LBL(52)
4297      LBL(53) LBL(54)
4298    #ifdef SUPPORT_UTF8
4299      LBL(16) LBL(18) LBL(20) LBL(21) LBL(22) LBL(23) LBL(28) LBL(30)
4300      LBL(32) LBL(34) LBL(42) LBL(46)
4301    #ifdef SUPPORT_UCP
4302      LBL(36) LBL(37) LBL(38) LBL(39) LBL(40) LBL(41) LBL(44) LBL(45)
4303    #endif  /* SUPPORT_UCP */
4304    #endif  /* SUPPORT_UTF8 */
4305      default:
4306      DPRINTF(("jump error in pcre match: label %d non-existent\n", frame->Xwhere));
4307      return PCRE_ERROR_INTERNAL;
4308      }
4309    #undef LBL
4310    #endif  /* NO_RECURSE */
4311  }  }
4312    
4313    
# Line 3640  Undefine all the macros that were define Line 4320  Undefine all the macros that were define
4320  #ifdef NO_RECURSE  #ifdef NO_RECURSE
4321  #undef eptr  #undef eptr
4322  #undef ecode  #undef ecode
4323    #undef mstart
4324  #undef offset_top  #undef offset_top
4325  #undef ims  #undef ims
4326  #undef eptrb  #undef eptrb
# Line 3712  Returns:          > 0 => success; value Line 4393  Returns:          > 0 => success; value
4393                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
4394  */  */
4395    
4396  PCRE_DATA_SCOPE int  PCRE_EXP_DEFN int
4397  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,
4398    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,
4399    int offsetcount)    int offsetcount)
# Line 3737  const uschar *start_bits = NULL; Line 4418  const uschar *start_bits = NULL;
4418  USPTR start_match = (USPTR)subject + start_offset;  USPTR start_match = (USPTR)subject + start_offset;
4419  USPTR end_subject;  USPTR end_subject;
4420  USPTR req_byte_ptr = start_match - 1;  USPTR req_byte_ptr = start_match - 1;
 eptrblock eptrchain[EPTR_WORK_SIZE];  
4421    
4422  pcre_study_data internal_study;  pcre_study_data internal_study;
4423  const pcre_study_data *study;  const pcre_study_data *study;
# Line 3800  if (re->magic_number != MAGIC_NUMBER) Line 4480  if (re->magic_number != MAGIC_NUMBER)
4480  /* Set up other data */  /* Set up other data */
4481    
4482  anchored = ((re->options | options) & PCRE_ANCHORED) != 0;  anchored = ((re->options | options) & PCRE_ANCHORED) != 0;
4483  startline = (re->options & PCRE_STARTLINE) != 0;  startline = (re->flags & PCRE_STARTLINE) != 0;
4484  firstline = (re->options & PCRE_FIRSTLINE) != 0;  firstline = (re->options & PCRE_FIRSTLINE) != 0;
4485    
4486  /* 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 4495  end_subject = md->end_subject;
4495    
4496  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;
4497  utf8 = md->utf8 = (re->options & PCRE_UTF8) != 0;  utf8 = md->utf8 = (re->options & PCRE_UTF8) != 0;
4498    md->jscript_compat = (re->options & PCRE_JAVASCRIPT_COMPAT) != 0;
4499    
4500  md->notbol = (options & PCRE_NOTBOL) != 0;  md->notbol = (options & PCRE_NOTBOL) != 0;
4501  md->noteol = (options & PCRE_NOTEOL) != 0;  md->noteol = (options & PCRE_NOTEOL) != 0;
# Line 3823  md->partial = (options & PCRE_PARTIAL) ! Line 4504  md->partial = (options & PCRE_PARTIAL) !
4504  md->hitend = FALSE;  md->hitend = FALSE;
4505    
4506  md->recursive = NULL;                   /* No recursion at top level */  md->recursive = NULL;                   /* No recursion at top level */
 md->eptrchain = eptrchain;              /* Make workspace generally available */  
4507    
4508  md->lcc = tables + lcc_offset;  md->lcc = tables + lcc_offset;
4509  md->ctypes = tables + ctypes_offset;  md->ctypes = tables + ctypes_offset;
4510    
4511    /* Handle different \R options. */
4512    
4513    switch (options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE))
4514      {
4515      case 0:
4516      if ((re->options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) != 0)
4517        md->bsr_anycrlf = (re->options & PCRE_BSR_ANYCRLF) != 0;
4518      else
4519    #ifdef BSR_ANYCRLF
4520      md->bsr_anycrlf = TRUE;
4521    #else
4522      md->bsr_anycrlf = FALSE;
4523    #endif
4524      break;
4525    
4526      case PCRE_BSR_ANYCRLF:
4527      md->bsr_anycrlf = TRUE;
4528      break;
4529    
4530      case PCRE_BSR_UNICODE:
4531      md->bsr_anycrlf = FALSE;
4532      break;
4533    
4534      default: return PCRE_ERROR_BADNEWLINE;
4535      }
4536    
4537  /* Handle different types of newline. The three bits give eight cases. If  /* Handle different types of newline. The three bits give eight cases. If
4538  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. */
4539    
4540  switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options : (pcre_uint32)options) &  switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options :
4541         PCRE_NEWLINE_BITS)          (pcre_uint32)options) & PCRE_NEWLINE_BITS)
4542    {    {
4543    case 0: newline = NEWLINE; break;   /* Compile-time default */    case 0: newline = NEWLINE; break;   /* Compile-time default */
4544    case PCRE_NEWLINE_CR: newline = '\r'; break;    case PCRE_NEWLINE_CR: newline = '\r'; break;
# Line 3840  switch ((((options & PCRE_NEWLINE_BITS) Line 4546  switch ((((options & PCRE_NEWLINE_BITS)
4546    case PCRE_NEWLINE_CR+    case PCRE_NEWLINE_CR+
4547         PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break;         PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break;
4548    case PCRE_NEWLINE_ANY: newline = -1; break;    case PCRE_NEWLINE_ANY: newline = -1; break;
4549      case PCRE_NEWLINE_ANYCRLF: newline = -2; break;
4550    default: return PCRE_ERROR_BADNEWLINE;    default: return PCRE_ERROR_BADNEWLINE;
4551    }    }
4552    
4553  if (newline < 0)  if (newline == -2)
4554      {
4555      md->nltype = NLTYPE_ANYCRLF;
4556      }
4557    else if (newline < 0)
4558    {    {
4559    md->nltype = NLTYPE_ANY;    md->nltype = NLTYPE_ANY;
4560    }    }
# Line 3866  else Line 4577  else
4577  /* 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
4578  moment. */  moment. */
4579    
4580  if (md->partial && (re->options & PCRE_NOPARTIAL) != 0)  if (md->partial && (re->flags & PCRE_NOPARTIAL) != 0)
4581    return PCRE_ERROR_BADPARTIAL;    return PCRE_ERROR_BADPARTIAL;
4582    
4583  /* 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 4654  studied, there may be a bitmap of possib
4654    
4655  if (!anchored)  if (!anchored)
4656    {    {
4657    if ((re->options & PCRE_FIRSTSET) != 0)    if ((re->flags & PCRE_FIRSTSET) != 0)
4658      {      {
4659      first_byte = re->first_byte & 255;      first_byte = re->first_byte & 255;
4660      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 4669  if (!anchored)
4669  /* For anchored or unanchored matches, there may be a "last known required  /* For anchored or unanchored matches, there may be a "last known required
4670  character" set. */  character" set. */
4671    
4672  if ((re->options & PCRE_REQCHSET) != 0)  if ((re->flags & PCRE_REQCHSET) != 0)
4673    {    {
4674    req_byte = re->req_byte & 255;    req_byte = re->req_byte & 255;
4675    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 4685  the loop runs just once. */
4685  for(;;)  for(;;)
4686    {    {
4687    USPTR save_end_subject = end_subject;    USPTR save_end_subject = end_subject;
4688      USPTR new_start_match;
4689    
4690    /* Reset the maximum number of extractions we might see. */    /* Reset the maximum number of extractions we might see. */
4691    
# Line 4004  for(;;) Line 4716  for(;;)
4716      if (first_byte_caseless)      if (first_byte_caseless)
4717        while (start_match < end_subject &&        while (start_match < end_subject &&
4718               md->lcc[*start_match] != first_byte)               md->lcc[*start_match] != first_byte)
4719          start_match++;          { NEXTCHAR(start_match); }
4720      else      else
4721        while (start_match < end_subject && *start_match != first_byte)        while (start_match < end_subject && *start_match != first_byte)
4722          start_match++;          { NEXTCHAR(start_match); }
4723      }      }
4724    
4725    /* Or to just after a linebreak for a multiline match if possible */    /* Or to just after a linebreak for a multiline match if possible */
# Line 4017  for(;;) Line 4729  for(;;)
4729      if (start_match > md->start_subject + start_offset)      if (start_match > md->start_subject + start_offset)
4730        {        {
4731        while (start_match <= end_subject && !WAS_NEWLINE(start_match))        while (start_match <= end_subject && !WAS_NEWLINE(start_match))
4732          start_match++;          { NEXTCHAR(start_match); }
4733    
4734        /* 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,
4735        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.
4736          */
4737    
4738        if (start_match[-1] == '\r' &&        if (start_match[-1] == '\r' &&
4739             md->nltype == NLTYPE_ANY &&             (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&
4740             start_match < end_subject &&             start_match < end_subject &&
4741             *start_match == '\n')             *start_match == '\n')
4742          start_match++;          start_match++;
# Line 4037  for(;;) Line 4750  for(;;)
4750      while (start_match < end_subject)      while (start_match < end_subject)
4751        {        {
4752        register unsigned int c = *start_match;        register unsigned int c = *start_match;
4753        if ((start_bits[c/8] & (1 << (c&7))) == 0) start_match++; else break;        if ((start_bits[c/8] & (1 << (c&7))) == 0)
4754            { NEXTCHAR(start_match); }
4755          else break;
4756        }        }
4757      }      }
4758    
# Line 4113  for(;;) Line 4828  for(;;)
4828    
4829    /* OK, we can now run the match. */    /* OK, we can now run the match. */
4830    
4831    md->start_match = start_match;    md->start_match_ptr = start_match;
4832    md->match_call_count = 0;    md->match_call_count = 0;
4833    md->eptrn = 0;                          /* Next free eptrchain slot */    rc = match(start_match, md->start_code, start_match, 2, md, ims, NULL, 0, 0);
4834    rc = match(start_match, md->start_code, 2, md, ims, NULL, 0, 0);  
4835      switch(rc)
4836        {
4837        /* NOMATCH and PRUNE advance by one character. THEN at this level acts
4838        exactly like PRUNE. */
4839    
4840        case MATCH_NOMATCH:
4841        case MATCH_PRUNE:
4842        case MATCH_THEN:
4843        new_start_match = start_match + 1;
4844    #ifdef SUPPORT_UTF8
4845        if (utf8)
4846          while(new_start_match < end_subject && (*new_start_match & 0xc0) == 0x80)
4847            new_start_match++;
4848    #endif
4849        break;
4850    
4851    /* Any return other than MATCH_NOMATCH breaks the loop. */      /* SKIP passes back the next starting point explicitly. */
4852    
4853    if (rc != MATCH_NOMATCH) break;      case MATCH_SKIP:
4854        new_start_match = md->start_match_ptr;
4855        break;
4856    
4857        /* COMMIT disables the bumpalong, but otherwise behaves as NOMATCH. */
4858    
4859        case MATCH_COMMIT:
4860        rc = MATCH_NOMATCH;
4861        goto ENDLOOP;
4862    
4863        /* Any other return is some kind of error. */
4864    
4865        default:
4866        goto ENDLOOP;
4867        }
4868    
4869      /* Control reaches here for the various types of "no match at this point"
4870      result. Reset the code to MATCH_NOMATCH for subsequent checking. */
4871    
4872      rc = MATCH_NOMATCH;
4873    
4874    /* 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
4875    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 4877  for(;;)
4877    
4878    if (firstline && IS_NEWLINE(start_match)) break;    if (firstline && IS_NEWLINE(start_match)) break;
4879    
4880    /* Advance the match position by one character. */    /* Advance to new matching position */
4881    
4882    start_match++;    start_match = new_start_match;
 #ifdef SUPPORT_UTF8  
   if (utf8)  
     while(start_match < end_subject && (*start_match & 0xc0) == 0x80)  
       start_match++;  
 #endif  
4883    
4884    /* 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
4885    the subject. */    the subject. */
4886    
4887    if (anchored || start_match > end_subject) break;    if (anchored || start_match > end_subject) break;
4888    
4889    /* 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
4890    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
4891      or ANY or ANYCRLF, advance the match position by one more character. */
4892    
4893    if (start_match[-1] == '\r' &&    if (start_match[-1] == '\r' &&
4894         (md->nltype == NLTYPE_ANY || md->nllen == 2) &&        start_match < end_subject &&
4895         start_match < end_subject &&        *start_match == '\n' &&
4896         *start_match == '\n')        (re->flags & PCRE_HASCRORLF) == 0 &&
4897            (md->nltype == NLTYPE_ANY ||
4898             md->nltype == NLTYPE_ANYCRLF ||
4899             md->nllen == 2))
4900      start_match++;      start_match++;
4901    
4902    }   /* End of for(;;) "bumpalong" loop */    }   /* End of for(;;) "bumpalong" loop */
# Line 4158  for(;;) Line 4906  for(;;)
4906  /* 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
4907  conditions is true:  conditions is true:
4908    
4909  (1) The pattern is anchored;  (1) The pattern is anchored or the match was failed by (*COMMIT);
4910    
4911  (2) We are past the end of the subject;  (2) We are past the end of the subject;
4912    
# Line 4173  processing, copy those that we can. In t Line 4921  processing, copy those that we can. In t
4921  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
4922  capturing parentheses than vector slots. */  capturing parentheses than vector slots. */
4923    
4924    ENDLOOP:
4925    
4926  if (rc == MATCH_MATCH)  if (rc == MATCH_MATCH)
4927    {    {
4928    if (using_temporary_offsets)    if (using_temporary_offsets)
# Line 4193  if (rc == MATCH_MATCH) Line 4943  if (rc == MATCH_MATCH)
4943    
4944    rc = md->offset_overflow? 0 : md->end_offset_top/2;    rc = md->offset_overflow? 0 : md->end_offset_top/2;
4945    
4946    /* 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
4947      md->start_match_ptr might be modified if \K was encountered on the success
4948      matching path. */
4949    
4950    if (offsetcount < 2) rc = 0; else    if (offsetcount < 2) rc = 0; else
4951      {      {
4952      offsets[0] = start_match - md->start_subject;      offsets[0] = md->start_match_ptr - md->start_subject;
4953      offsets[1] = md->end_match_ptr - md->start_subject;      offsets[1] = md->end_match_ptr - md->start_subject;
4954      }      }
4955    

Legend:
Removed from v.144  
changed lines
  Added in v.341

  ViewVC Help
Powered by ViewVC 1.1.5