/[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 96 by nigel, Fri Mar 2 13:10:43 2007 UTC revision 243 by ph10, Thu Sep 13 09:28:14 2007 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) 1997-2006 University of Cambridge             Copyright (c) 1997-2007 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 */
52    
53  #include "pcre_internal.h"  #include "pcre_internal.h"
54    
55  /* The chain of eptrblocks for tail recursions uses memory in stack workspace,  /* Undefine some potentially clashing cpp symbols */
 obtained at top level, the size of which is defined by EPTR_WORK_SIZE. */  
56    
57  #define EPTR_WORK_SIZE (1000)  #undef min
58    #undef max
59    
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 65  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 183  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 205  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 251  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 268  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 299  typedef struct heapframe { Line 330  typedef struct heapframe {
330    int Xprop_category;    int Xprop_category;
331    int Xprop_chartype;    int Xprop_chartype;
332    int Xprop_script;    int Xprop_script;
333      int Xoclength;
334      uschar Xocchars[8];
335  #endif  #endif
336    
337    int Xctype;    int Xctype;
# Line 316  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 347  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 356  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 366  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 394  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 408  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 441  HEAP_RECURSE: Line 476  HEAP_RECURSE:
476  #define prop_category      frame->Xprop_category  #define prop_category      frame->Xprop_category
477  #define prop_chartype      frame->Xprop_chartype  #define prop_chartype      frame->Xprop_chartype
478  #define prop_script        frame->Xprop_script  #define prop_script        frame->Xprop_script
479    #define oclength           frame->Xoclength
480    #define occhars            frame->Xocchars
481  #endif  #endif
482    
483  #define ctype              frame->Xctype  #define ctype              frame->Xctype
# Line 494  int prop_fail_result; Line 531  int prop_fail_result;
531  int prop_category;  int prop_category;
532  int prop_chartype;  int prop_chartype;
533  int prop_script;  int prop_script;
534    int oclength;
535    uschar occhars[8];
536  #endif  #endif
537    
538  int ctype;  int ctype;
# Line 534  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 542  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 582  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 626  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 643  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 659  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 712  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 732  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 759  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 768  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 791  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 817  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 843  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 874  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 936  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 946  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 956  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 990  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 1036  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 1068  for (;;) Line 1157  for (;;)
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 1079  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        }        }
# Line 1149  for (;;) Line 1238  for (;;)
1238          recursion_info *rec = md->recursive;          recursion_info *rec = md->recursive;
1239          DPRINTF(("Recursion (%d) succeeded - continuing\n", number));          DPRINTF(("Recursion (%d) succeeded - continuing\n", number));
1240          md->recursive = rec->prevrec;          md->recursive = rec->prevrec;
1241          md->start_match = rec->save_start;          mstart = rec->save_start;
1242          memcpy(md->offset_vector, rec->offset_save,          memcpy(md->offset_vector, rec->offset_save,
1243            rec->saved_max * sizeof(int));            rec->saved_max * sizeof(int));
1244          ecode = rec->after_call;          ecode = rec->after_call;
# Line 1178  for (;;) Line 1267  for (;;)
1267    
1268      /* 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
1269      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
1270      tail recursion to avoid using another stack frame. */      tail recursion to avoid using another stack frame, unless we have an
1271        unlimited repeat of a group that can match an empty string. */
1272    
1273      flags = (*prev >= OP_SBRA)? match_cbegroup : 0;      flags = (*prev >= OP_SBRA)? match_cbegroup : 0;
1274    
1275      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1276        {        {
1277        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);
1278        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1279          if (flags != 0)    /* Could match an empty string */
1280            {
1281            RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM50);
1282            RRETURN(rrc);
1283            }
1284        ecode = prev;        ecode = prev;
       flags |= match_tail_recursed;  
1285        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1286        }        }
1287      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
1288        {        {
1289        RMATCH(rrc, eptr, prev, offset_top, md, ims, eptrb, flags);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM13);
1290        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1291        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1292        flags = match_tail_recursed;        flags = 0;
1293        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1294        }        }
1295      /* Control never gets here */      /* Control never gets here */
# Line 1228  for (;;) Line 1322  for (;;)
1322      ecode++;      ecode++;
1323      break;      break;
1324    
1325        /* Reset the start of match point */
1326    
1327        case OP_SET_SOM:
1328        mstart = eptr;
1329        ecode++;
1330        break;
1331    
1332      /* Assert before internal newline if multiline, or before a terminating      /* Assert before internal newline if multiline, or before a terminating
1333      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. */
1334    
# Line 1425  for (;;) Line 1526  for (;;)
1526        case 0x000d:        case 0x000d:
1527        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
1528        break;        break;
1529    
1530        case 0x000a:        case 0x000a:
1531          break;
1532    
1533        case 0x000b:        case 0x000b:
1534        case 0x000c:        case 0x000c:
1535        case 0x0085:        case 0x0085:
1536        case 0x2028:        case 0x2028:
1537        case 0x2029:        case 0x2029:
1538          if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
1539          break;
1540          }
1541        ecode++;
1542        break;
1543    
1544        case OP_NOT_HSPACE:
1545        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1546        GETCHARINCTEST(c, eptr);
1547        switch(c)
1548          {
1549          default: break;
1550          case 0x09:      /* HT */
1551          case 0x20:      /* SPACE */
1552          case 0xa0:      /* NBSP */
1553          case 0x1680:    /* OGHAM SPACE MARK */
1554          case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
1555          case 0x2000:    /* EN QUAD */
1556          case 0x2001:    /* EM QUAD */
1557          case 0x2002:    /* EN SPACE */
1558          case 0x2003:    /* EM SPACE */
1559          case 0x2004:    /* THREE-PER-EM SPACE */
1560          case 0x2005:    /* FOUR-PER-EM SPACE */
1561          case 0x2006:    /* SIX-PER-EM SPACE */
1562          case 0x2007:    /* FIGURE SPACE */
1563          case 0x2008:    /* PUNCTUATION SPACE */
1564          case 0x2009:    /* THIN SPACE */
1565          case 0x200A:    /* HAIR SPACE */
1566          case 0x202f:    /* NARROW NO-BREAK SPACE */
1567          case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
1568          case 0x3000:    /* IDEOGRAPHIC SPACE */
1569          RRETURN(MATCH_NOMATCH);
1570          }
1571        ecode++;
1572        break;
1573    
1574        case OP_HSPACE:
1575        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1576        GETCHARINCTEST(c, eptr);
1577        switch(c)
1578          {
1579          default: RRETURN(MATCH_NOMATCH);
1580          case 0x09:      /* HT */
1581          case 0x20:      /* SPACE */
1582          case 0xa0:      /* NBSP */
1583          case 0x1680:    /* OGHAM SPACE MARK */
1584          case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
1585          case 0x2000:    /* EN QUAD */
1586          case 0x2001:    /* EM QUAD */
1587          case 0x2002:    /* EN SPACE */
1588          case 0x2003:    /* EM SPACE */
1589          case 0x2004:    /* THREE-PER-EM SPACE */
1590          case 0x2005:    /* FOUR-PER-EM SPACE */
1591          case 0x2006:    /* SIX-PER-EM SPACE */
1592          case 0x2007:    /* FIGURE SPACE */
1593          case 0x2008:    /* PUNCTUATION SPACE */
1594          case 0x2009:    /* THIN SPACE */
1595          case 0x200A:    /* HAIR SPACE */
1596          case 0x202f:    /* NARROW NO-BREAK SPACE */
1597          case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
1598          case 0x3000:    /* IDEOGRAPHIC SPACE */
1599          break;
1600          }
1601        ecode++;
1602        break;
1603    
1604        case OP_NOT_VSPACE:
1605        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1606        GETCHARINCTEST(c, eptr);
1607        switch(c)
1608          {
1609          default: break;
1610          case 0x0a:      /* LF */
1611          case 0x0b:      /* VT */
1612          case 0x0c:      /* FF */
1613          case 0x0d:      /* CR */
1614          case 0x85:      /* NEL */
1615          case 0x2028:    /* LINE SEPARATOR */
1616          case 0x2029:    /* PARAGRAPH SEPARATOR */
1617          RRETURN(MATCH_NOMATCH);
1618          }
1619        ecode++;
1620        break;
1621    
1622        case OP_VSPACE:
1623        if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1624        GETCHARINCTEST(c, eptr);
1625        switch(c)
1626          {
1627          default: RRETURN(MATCH_NOMATCH);
1628          case 0x0a:      /* LF */
1629          case 0x0b:      /* VT */
1630          case 0x0c:      /* FF */
1631          case 0x0d:      /* CR */
1632          case 0x85:      /* NEL */
1633          case 0x2028:    /* LINE SEPARATOR */
1634          case 0x2029:    /* PARAGRAPH SEPARATOR */
1635        break;        break;
1636        }        }
1637      ecode++;      ecode++;
# Line 1591  for (;;) Line 1792  for (;;)
1792          {          {
1793          for (fi = min;; fi++)          for (fi = min;; fi++)
1794            {            {
1795            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM14);
1796            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1797            if (fi >= max || !match_ref(offset, eptr, length, md, ims))            if (fi >= max || !match_ref(offset, eptr, length, md, ims))
1798              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
# Line 1612  for (;;) Line 1813  for (;;)
1813            }            }
1814          while (eptr >= pp)          while (eptr >= pp)
1815            {            {
1816            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM15);
1817            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1818            eptr -= length;            eptr -= length;
1819            }            }
# Line 1717  for (;;) Line 1918  for (;;)
1918            {            {
1919            for (fi = min;; fi++)            for (fi = min;; fi++)
1920              {              {
1921              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM16);
1922              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1923              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1924              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 1737  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, RM17);
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              c = *eptr++;              c = *eptr++;
# Line 1774  for (;;) Line 1975  for (;;)
1975              }              }
1976            for (;;)            for (;;)
1977              {              {
1978              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM18);
1979              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1980              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
1981              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 1793  for (;;) Line 1994  for (;;)
1994              }              }
1995            while (eptr >= pp)            while (eptr >= pp)
1996              {              {
1997              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM19);
1998              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1999              eptr--;              eptr--;
2000              }              }
# Line 1864  for (;;) Line 2065  for (;;)
2065          {          {
2066          for (fi = min;; fi++)          for (fi = min;; fi++)
2067            {            {
2068            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM20);
2069            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2070            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2071            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
# Line 1888  for (;;) Line 2089  for (;;)
2089            }            }
2090          for(;;)          for(;;)
2091            {            {
2092            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM21);
2093            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2094            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
2095            BACKCHAR(eptr)            if (utf8) BACKCHAR(eptr);
2096            }            }
2097          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2098          }          }
# Line 2045  for (;;) Line 2246  for (;;)
2246    
2247        if (length > 1)        if (length > 1)
2248          {          {
         int oclength = 0;  
         uschar occhars[8];  
   
2249  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2250          unsigned int othercase;          unsigned int othercase;
2251          if ((ims & PCRE_CASELESS) != 0 &&          if ((ims & PCRE_CASELESS) != 0 &&
2252              (othercase = _pcre_ucp_othercase(fc)) != NOTACHAR)              (othercase = _pcre_ucp_othercase(fc)) != NOTACHAR)
2253            oclength = _pcre_ord2utf8(othercase, occhars);            oclength = _pcre_ord2utf8(othercase, occhars);
2254            else oclength = 0;
2255  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2256    
2257          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2258            {            {
2259            if (memcmp(eptr, charptr, length) == 0) eptr += length;            if (memcmp(eptr, charptr, length) == 0) eptr += length;
2260    #ifdef SUPPORT_UCP
2261            /* Need braces because of following else */            /* Need braces because of following else */
2262            else if (oclength == 0) { RRETURN(MATCH_NOMATCH); }            else if (oclength == 0) { RRETURN(MATCH_NOMATCH); }
2263            else            else
# Line 2065  for (;;) Line 2265  for (;;)
2265              if (memcmp(eptr, occhars, oclength) != 0) RRETURN(MATCH_NOMATCH);              if (memcmp(eptr, occhars, oclength) != 0) RRETURN(MATCH_NOMATCH);
2266              eptr += oclength;              eptr += oclength;
2267              }              }
2268    #else   /* without SUPPORT_UCP */
2269              else { RRETURN(MATCH_NOMATCH); }
2270    #endif  /* SUPPORT_UCP */
2271            }            }
2272    
2273          if (min == max) continue;          if (min == max) continue;
# Line 2073  for (;;) Line 2276  for (;;)
2276            {            {
2277            for (fi = min;; fi++)            for (fi = min;; fi++)
2278              {              {
2279              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM22);
2280              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2281              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2282              if (memcmp(eptr, charptr, length) == 0) eptr += length;              if (memcmp(eptr, charptr, length) == 0) eptr += length;
2283    #ifdef SUPPORT_UCP
2284              /* Need braces because of following else */              /* Need braces because of following else */
2285              else if (oclength == 0) { RRETURN(MATCH_NOMATCH); }              else if (oclength == 0) { RRETURN(MATCH_NOMATCH); }
2286              else              else
# Line 2084  for (;;) Line 2288  for (;;)
2288                if (memcmp(eptr, occhars, oclength) != 0) RRETURN(MATCH_NOMATCH);                if (memcmp(eptr, occhars, oclength) != 0) RRETURN(MATCH_NOMATCH);
2289                eptr += oclength;                eptr += oclength;
2290                }                }
2291    #else   /* without SUPPORT_UCP */
2292                else { RRETURN (MATCH_NOMATCH); }
2293    #endif  /* SUPPORT_UCP */
2294              }              }
2295            /* Control never gets here */            /* Control never gets here */
2296            }            }
# Line 2095  for (;;) Line 2302  for (;;)
2302              {              {
2303              if (eptr > md->end_subject - length) break;              if (eptr > md->end_subject - length) break;
2304              if (memcmp(eptr, charptr, length) == 0) eptr += length;              if (memcmp(eptr, charptr, length) == 0) eptr += length;
2305    #ifdef SUPPORT_UCP
2306              else if (oclength == 0) break;              else if (oclength == 0) break;
2307              else              else
2308                {                {
2309                if (memcmp(eptr, occhars, oclength) != 0) break;                if (memcmp(eptr, occhars, oclength) != 0) break;
2310                eptr += oclength;                eptr += oclength;
2311                }                }
2312    #else   /* without SUPPORT_UCP */
2313                else break;
2314    #endif  /* SUPPORT_UCP */
2315              }              }
2316    
2317            if (possessive) continue;            if (possessive) continue;
2318            while (eptr >= pp)            for(;;)
2319             {             {
2320             RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);             RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM23);
2321             if (rrc != MATCH_NOMATCH) RRETURN(rrc);             if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2322               if (eptr == pp) RRETURN(MATCH_NOMATCH);
2323    #ifdef SUPPORT_UCP
2324               eptr--;
2325               BACKCHAR(eptr);
2326    #else   /* without SUPPORT_UCP */
2327             eptr -= length;             eptr -= length;
2328    #endif  /* SUPPORT_UCP */
2329             }             }
           RRETURN(MATCH_NOMATCH);  
2330            }            }
2331          /* Control never gets here */          /* Control never gets here */
2332          }          }
# Line 2150  for (;;) Line 2366  for (;;)
2366          {          {
2367          for (fi = min;; fi++)          for (fi = min;; fi++)
2368            {            {
2369            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM24);
2370            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2371            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
2372                fc != md->lcc[*eptr++])                fc != md->lcc[*eptr++])
# Line 2169  for (;;) Line 2385  for (;;)
2385          if (possessive) continue;          if (possessive) continue;
2386          while (eptr >= pp)          while (eptr >= pp)
2387            {            {
2388            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM25);
2389            eptr--;            eptr--;
2390            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2391            }            }
# Line 2188  for (;;) Line 2404  for (;;)
2404          {          {
2405          for (fi = min;; fi++)          for (fi = min;; fi++)
2406            {            {
2407            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM26);
2408            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2409            if (fi >= max || eptr >= md->end_subject || fc != *eptr++)            if (fi >= max || eptr >= md->end_subject || fc != *eptr++)
2410              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
# Line 2206  for (;;) Line 2422  for (;;)
2422          if (possessive) continue;          if (possessive) continue;
2423          while (eptr >= pp)          while (eptr >= pp)
2424            {            {
2425            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM27);
2426            eptr--;            eptr--;
2427            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2428            }            }
# Line 2351  for (;;) Line 2567  for (;;)
2567            register unsigned int d;            register unsigned int d;
2568            for (fi = min;; fi++)            for (fi = min;; fi++)
2569              {              {
2570              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM28);
2571              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2572              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
2573              if (d < 256) d = md->lcc[d];              if (d < 256) d = md->lcc[d];
# Line 2365  for (;;) Line 2581  for (;;)
2581            {            {
2582            for (fi = min;; fi++)            for (fi = min;; fi++)
2583              {              {
2584              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM29);
2585              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2586              if (fi >= max || eptr >= md->end_subject || fc == md->lcc[*eptr++])              if (fi >= max || eptr >= md->end_subject || fc == md->lcc[*eptr++])
2587                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2397  for (;;) Line 2613  for (;;)
2613          if (possessive) continue;          if (possessive) continue;
2614          for(;;)          for(;;)
2615              {              {
2616              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM30);
2617              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2618              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
2619              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 2415  for (;;) Line 2631  for (;;)
2631            if (possessive) continue;            if (possessive) continue;
2632            while (eptr >= pp)            while (eptr >= pp)
2633              {              {
2634              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM31);
2635              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2636              eptr--;              eptr--;
2637              }              }
# Line 2460  for (;;) Line 2676  for (;;)
2676            register unsigned int d;            register unsigned int d;
2677            for (fi = min;; fi++)            for (fi = min;; fi++)
2678              {              {
2679              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM32);
2680              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2681              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
2682              if (fi >= max || eptr >= md->end_subject || fc == d)              if (fi >= max || eptr >= md->end_subject || fc == d)
# Line 2473  for (;;) Line 2689  for (;;)
2689            {            {
2690            for (fi = min;; fi++)            for (fi = min;; fi++)
2691              {              {
2692              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM33);
2693              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2694              if (fi >= max || eptr >= md->end_subject || fc == *eptr++)              if (fi >= max || eptr >= md->end_subject || fc == *eptr++)
2695                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2504  for (;;) Line 2720  for (;;)
2720            if (possessive) continue;            if (possessive) continue;
2721            for(;;)            for(;;)
2722              {              {
2723              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM34);
2724              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2725              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
2726              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 2522  for (;;) Line 2738  for (;;)
2738            if (possessive) continue;            if (possessive) continue;
2739            while (eptr >= pp)            while (eptr >= pp)
2740              {              {
2741              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM35);
2742              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2743              eptr--;              eptr--;
2744              }              }
# Line 2629  for (;;) Line 2845  for (;;)
2845            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2846              {              {
2847              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2848              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2849              }              }
2850            break;            break;
2851    
# Line 2637  for (;;) Line 2853  for (;;)
2853            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2854              {              {
2855              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2856              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2857              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);
2858              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
2859                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
# Line 2650  for (;;) Line 2866  for (;;)
2866            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2867              {              {
2868              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2869              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2870              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);
2871              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
2872                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2661  for (;;) Line 2877  for (;;)
2877            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2878              {              {
2879              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2880              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2881              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);
2882              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
2883                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2672  for (;;) Line 2888  for (;;)
2888            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2889              {              {
2890              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2891              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2892              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);
2893              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
2894                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
# Line 2742  for (;;) Line 2958  for (;;)
2958              case 0x000d:              case 0x000d:
2959              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
2960              break;              break;
2961    
2962              case 0x000a:              case 0x000a:
2963                break;
2964    
2965              case 0x000b:              case 0x000b:
2966              case 0x000c:              case 0x000c:
2967              case 0x0085:              case 0x0085:
2968              case 0x2028:              case 0x2028:
2969              case 0x2029:              case 0x2029:
2970                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
2971                break;
2972                }
2973              }
2974            break;
2975    
2976            case OP_NOT_HSPACE:
2977            for (i = 1; i <= min; i++)
2978              {
2979              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2980              GETCHARINC(c, eptr);
2981              switch(c)
2982                {
2983                default: break;
2984                case 0x09:      /* HT */
2985                case 0x20:      /* SPACE */
2986                case 0xa0:      /* NBSP */
2987                case 0x1680:    /* OGHAM SPACE MARK */
2988                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
2989                case 0x2000:    /* EN QUAD */
2990                case 0x2001:    /* EM QUAD */
2991                case 0x2002:    /* EN SPACE */
2992                case 0x2003:    /* EM SPACE */
2993                case 0x2004:    /* THREE-PER-EM SPACE */
2994                case 0x2005:    /* FOUR-PER-EM SPACE */
2995                case 0x2006:    /* SIX-PER-EM SPACE */
2996                case 0x2007:    /* FIGURE SPACE */
2997                case 0x2008:    /* PUNCTUATION SPACE */
2998                case 0x2009:    /* THIN SPACE */
2999                case 0x200A:    /* HAIR SPACE */
3000                case 0x202f:    /* NARROW NO-BREAK SPACE */
3001                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3002                case 0x3000:    /* IDEOGRAPHIC SPACE */
3003                RRETURN(MATCH_NOMATCH);
3004                }
3005              }
3006            break;
3007    
3008            case OP_HSPACE:
3009            for (i = 1; i <= min; i++)
3010              {
3011              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3012              GETCHARINC(c, eptr);
3013              switch(c)
3014                {
3015                default: RRETURN(MATCH_NOMATCH);
3016                case 0x09:      /* HT */
3017                case 0x20:      /* SPACE */
3018                case 0xa0:      /* NBSP */
3019                case 0x1680:    /* OGHAM SPACE MARK */
3020                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3021                case 0x2000:    /* EN QUAD */
3022                case 0x2001:    /* EM QUAD */
3023                case 0x2002:    /* EN SPACE */
3024                case 0x2003:    /* EM SPACE */
3025                case 0x2004:    /* THREE-PER-EM SPACE */
3026                case 0x2005:    /* FOUR-PER-EM SPACE */
3027                case 0x2006:    /* SIX-PER-EM SPACE */
3028                case 0x2007:    /* FIGURE SPACE */
3029                case 0x2008:    /* PUNCTUATION SPACE */
3030                case 0x2009:    /* THIN SPACE */
3031                case 0x200A:    /* HAIR SPACE */
3032                case 0x202f:    /* NARROW NO-BREAK SPACE */
3033                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3034                case 0x3000:    /* IDEOGRAPHIC SPACE */
3035                break;
3036                }
3037              }
3038            break;
3039    
3040            case OP_NOT_VSPACE:
3041            for (i = 1; i <= min; i++)
3042              {
3043              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3044              GETCHARINC(c, eptr);
3045              switch(c)
3046                {
3047                default: break;
3048                case 0x0a:      /* LF */
3049                case 0x0b:      /* VT */
3050                case 0x0c:      /* FF */
3051                case 0x0d:      /* CR */
3052                case 0x85:      /* NEL */
3053                case 0x2028:    /* LINE SEPARATOR */
3054                case 0x2029:    /* PARAGRAPH SEPARATOR */
3055                RRETURN(MATCH_NOMATCH);
3056                }
3057              }
3058            break;
3059    
3060            case OP_VSPACE:
3061            for (i = 1; i <= min; i++)
3062              {
3063              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3064              GETCHARINC(c, eptr);
3065              switch(c)
3066                {
3067                default: RRETURN(MATCH_NOMATCH);
3068                case 0x0a:      /* LF */
3069                case 0x0b:      /* VT */
3070                case 0x0c:      /* FF */
3071                case 0x0d:      /* CR */
3072                case 0x85:      /* NEL */
3073                case 0x2028:    /* LINE SEPARATOR */
3074                case 0x2029:    /* PARAGRAPH SEPARATOR */
3075              break;              break;
3076              }              }
3077            }            }
# Line 2777  for (;;) Line 3101  for (;;)
3101          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3102            {            {
3103            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject ||
3104               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_space) != 0))               (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0))
3105              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3106            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3107            }            }
3108          break;          break;
3109    
# Line 2797  for (;;) Line 3121  for (;;)
3121          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3122            {            {
3123            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject ||
3124               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_word) != 0))               (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0))
3125              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3126            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3127            }            }
3128          break;          break;
3129    
# Line 2856  for (;;) Line 3180  for (;;)
3180              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3181              break;              break;
3182              case 0x000a:              case 0x000a:
3183                break;
3184    
3185              case 0x000b:              case 0x000b:
3186              case 0x000c:              case 0x000c:
3187              case 0x0085:              case 0x0085:
3188                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3189                break;
3190                }
3191              }
3192            break;
3193    
3194            case OP_NOT_HSPACE:
3195            for (i = 1; i <= min; i++)
3196              {
3197              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3198              switch(*eptr++)
3199                {
3200                default: break;
3201                case 0x09:      /* HT */
3202                case 0x20:      /* SPACE */
3203                case 0xa0:      /* NBSP */
3204                RRETURN(MATCH_NOMATCH);
3205                }
3206              }
3207            break;
3208    
3209            case OP_HSPACE:
3210            for (i = 1; i <= min; i++)
3211              {
3212              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3213              switch(*eptr++)
3214                {
3215                default: RRETURN(MATCH_NOMATCH);
3216                case 0x09:      /* HT */
3217                case 0x20:      /* SPACE */
3218                case 0xa0:      /* NBSP */
3219                break;
3220                }
3221              }
3222            break;
3223    
3224            case OP_NOT_VSPACE:
3225            for (i = 1; i <= min; i++)
3226              {
3227              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3228              switch(*eptr++)
3229                {
3230                default: break;
3231                case 0x0a:      /* LF */
3232                case 0x0b:      /* VT */
3233                case 0x0c:      /* FF */
3234                case 0x0d:      /* CR */
3235                case 0x85:      /* NEL */
3236                RRETURN(MATCH_NOMATCH);
3237                }
3238              }
3239            break;
3240    
3241            case OP_VSPACE:
3242            for (i = 1; i <= min; i++)
3243              {
3244              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3245              switch(*eptr++)
3246                {
3247                default: RRETURN(MATCH_NOMATCH);
3248                case 0x0a:      /* LF */
3249                case 0x0b:      /* VT */
3250                case 0x0c:      /* FF */
3251                case 0x0d:      /* CR */
3252                case 0x85:      /* NEL */
3253              break;              break;
3254              }              }
3255            }            }
# Line 2919  for (;;) Line 3310  for (;;)
3310            case PT_ANY:            case PT_ANY:
3311            for (fi = min;; fi++)            for (fi = min;; fi++)
3312              {              {
3313              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM36);
3314              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3315              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3316              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 2930  for (;;) Line 3321  for (;;)
3321            case PT_LAMP:            case PT_LAMP:
3322            for (fi = min;; fi++)            for (fi = min;; fi++)
3323              {              {
3324              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM37);
3325              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3326              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3327              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 2945  for (;;) Line 3336  for (;;)
3336            case PT_GC:            case PT_GC:
3337            for (fi = min;; fi++)            for (fi = min;; fi++)
3338              {              {
3339              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM38);
3340              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3341              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3342              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 2958  for (;;) Line 3349  for (;;)
3349            case PT_PC:            case PT_PC:
3350            for (fi = min;; fi++)            for (fi = min;; fi++)
3351              {              {
3352              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM39);
3353              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3354              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3355              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 2971  for (;;) Line 3362  for (;;)
3362            case PT_SC:            case PT_SC:
3363            for (fi = min;; fi++)            for (fi = min;; fi++)
3364              {              {
3365              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM40);
3366              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3367              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3368              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
# Line 2993  for (;;) Line 3384  for (;;)
3384          {          {
3385          for (fi = min;; fi++)          for (fi = min;; fi++)
3386            {            {
3387            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM41);
3388            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3389            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3390            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
# Line 3022  for (;;) Line 3413  for (;;)
3413          {          {
3414          for (fi = min;; fi++)          for (fi = min;; fi++)
3415            {            {
3416            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM42);
3417            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3418            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
3419                 (ctype == OP_ANY && (ims & PCRE_DOTALL) == 0 &&                 (ctype == OP_ANY && (ims & PCRE_DOTALL) == 0 &&
# Line 3046  for (;;) Line 3437  for (;;)
3437                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3438                break;                break;
3439                case 0x000a:                case 0x000a:
3440                  break;
3441    
3442                case 0x000b:                case 0x000b:
3443                case 0x000c:                case 0x000c:
3444                case 0x0085:                case 0x0085:
3445                case 0x2028:                case 0x2028:
3446                case 0x2029:                case 0x2029:
3447                  if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3448                  break;
3449                  }
3450                break;
3451    
3452                case OP_NOT_HSPACE:
3453                switch(c)
3454                  {
3455                  default: break;
3456                  case 0x09:      /* HT */
3457                  case 0x20:      /* SPACE */
3458                  case 0xa0:      /* NBSP */
3459                  case 0x1680:    /* OGHAM SPACE MARK */
3460                  case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3461                  case 0x2000:    /* EN QUAD */
3462                  case 0x2001:    /* EM QUAD */
3463                  case 0x2002:    /* EN SPACE */
3464                  case 0x2003:    /* EM SPACE */
3465                  case 0x2004:    /* THREE-PER-EM SPACE */
3466                  case 0x2005:    /* FOUR-PER-EM SPACE */
3467                  case 0x2006:    /* SIX-PER-EM SPACE */
3468                  case 0x2007:    /* FIGURE SPACE */
3469                  case 0x2008:    /* PUNCTUATION SPACE */
3470                  case 0x2009:    /* THIN SPACE */
3471                  case 0x200A:    /* HAIR SPACE */
3472                  case 0x202f:    /* NARROW NO-BREAK SPACE */
3473                  case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3474                  case 0x3000:    /* IDEOGRAPHIC SPACE */
3475                  RRETURN(MATCH_NOMATCH);
3476                  }
3477                break;
3478    
3479                case OP_HSPACE:
3480                switch(c)
3481                  {
3482                  default: RRETURN(MATCH_NOMATCH);
3483                  case 0x09:      /* HT */
3484                  case 0x20:      /* SPACE */
3485                  case 0xa0:      /* NBSP */
3486                  case 0x1680:    /* OGHAM SPACE MARK */
3487                  case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3488                  case 0x2000:    /* EN QUAD */
3489                  case 0x2001:    /* EM QUAD */
3490                  case 0x2002:    /* EN SPACE */
3491                  case 0x2003:    /* EM SPACE */
3492                  case 0x2004:    /* THREE-PER-EM SPACE */
3493                  case 0x2005:    /* FOUR-PER-EM SPACE */
3494                  case 0x2006:    /* SIX-PER-EM SPACE */
3495                  case 0x2007:    /* FIGURE SPACE */
3496                  case 0x2008:    /* PUNCTUATION SPACE */
3497                  case 0x2009:    /* THIN SPACE */
3498                  case 0x200A:    /* HAIR SPACE */
3499                  case 0x202f:    /* NARROW NO-BREAK SPACE */
3500                  case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3501                  case 0x3000:    /* IDEOGRAPHIC SPACE */
3502                  break;
3503                  }
3504                break;
3505    
3506                case OP_NOT_VSPACE:
3507                switch(c)
3508                  {
3509                  default: break;
3510                  case 0x0a:      /* LF */
3511                  case 0x0b:      /* VT */
3512                  case 0x0c:      /* FF */
3513                  case 0x0d:      /* CR */
3514                  case 0x85:      /* NEL */
3515                  case 0x2028:    /* LINE SEPARATOR */
3516                  case 0x2029:    /* PARAGRAPH SEPARATOR */
3517                  RRETURN(MATCH_NOMATCH);
3518                  }
3519                break;
3520    
3521                case OP_VSPACE:
3522                switch(c)
3523                  {
3524                  default: RRETURN(MATCH_NOMATCH);
3525                  case 0x0a:      /* LF */
3526                  case 0x0b:      /* VT */
3527                  case 0x0c:      /* FF */
3528                  case 0x0d:      /* CR */
3529                  case 0x85:      /* NEL */
3530                  case 0x2028:    /* LINE SEPARATOR */
3531                  case 0x2029:    /* PARAGRAPH SEPARATOR */
3532                break;                break;
3533                }                }
3534              break;              break;
# Line 3096  for (;;) Line 3574  for (;;)
3574          {          {
3575          for (fi = min;; fi++)          for (fi = min;; fi++)
3576            {            {
3577            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM43);
3578            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3579            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
3580                 ((ims & PCRE_DOTALL) == 0 && IS_NEWLINE(eptr)))                 ((ims & PCRE_DOTALL) == 0 && IS_NEWLINE(eptr)))
# Line 3118  for (;;) Line 3596  for (;;)
3596                case 0x000d:                case 0x000d:
3597                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3598                break;                break;
3599    
3600                case 0x000a:                case 0x000a:
3601                  break;
3602    
3603                case 0x000b:                case 0x000b:
3604                case 0x000c:                case 0x000c:
3605                case 0x0085:                case 0x0085:
3606                  if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3607                  break;
3608                  }
3609                break;
3610    
3611                case OP_NOT_HSPACE:
3612                switch(c)
3613                  {
3614                  default: break;
3615                  case 0x09:      /* HT */
3616                  case 0x20:      /* SPACE */
3617                  case 0xa0:      /* NBSP */
3618                  RRETURN(MATCH_NOMATCH);
3619                  }
3620                break;
3621    
3622                case OP_HSPACE:
3623                switch(c)
3624                  {
3625                  default: RRETURN(MATCH_NOMATCH);
3626                  case 0x09:      /* HT */
3627                  case 0x20:      /* SPACE */
3628                  case 0xa0:      /* NBSP */
3629                  break;
3630                  }
3631                break;
3632    
3633                case OP_NOT_VSPACE:
3634                switch(c)
3635                  {
3636                  default: break;
3637                  case 0x0a:      /* LF */
3638                  case 0x0b:      /* VT */
3639                  case 0x0c:      /* FF */
3640                  case 0x0d:      /* CR */
3641                  case 0x85:      /* NEL */
3642                  RRETURN(MATCH_NOMATCH);
3643                  }
3644                break;
3645    
3646                case OP_VSPACE:
3647                switch(c)
3648                  {
3649                  default: RRETURN(MATCH_NOMATCH);
3650                  case 0x0a:      /* LF */
3651                  case 0x0b:      /* VT */
3652                  case 0x0c:      /* FF */
3653                  case 0x0d:      /* CR */
3654                  case 0x85:      /* NEL */
3655                break;                break;
3656                }                }
3657              break;              break;
# Line 3242  for (;;) Line 3772  for (;;)
3772          if (possessive) continue;          if (possessive) continue;
3773          for(;;)          for(;;)
3774            {            {
3775            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM44);
3776            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3777            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3778            BACKCHAR(eptr);            if (utf8) BACKCHAR(eptr);
3779            }            }
3780          }          }
3781    
# Line 3278  for (;;) Line 3808  for (;;)
3808          if (possessive) continue;          if (possessive) continue;
3809          for(;;)          for(;;)
3810            {            {
3811            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM45);
3812            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3813            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3814            for (;;)                        /* Move back over one extended */            for (;;)                        /* Move back over one extended */
3815              {              {
3816              int len = 1;              int len = 1;
             BACKCHAR(eptr);  
3817              if (!utf8) c = *eptr; else              if (!utf8) c = *eptr; else
3818                {                {
3819                  BACKCHAR(eptr);
3820                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
3821                }                }
3822              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);
# Line 3307  for (;;) Line 3837  for (;;)
3837          switch(ctype)          switch(ctype)
3838            {            {
3839            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. */  
   
3840            if (max < INT_MAX)            if (max < INT_MAX)
3841              {              {
3842              if ((ims & PCRE_DOTALL) == 0)              if ((ims & PCRE_DOTALL) == 0)
# Line 3344  for (;;) Line 3869  for (;;)
3869                  {                  {
3870                  if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;                  if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
3871                  eptr++;                  eptr++;
3872                    while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
3873                  }                  }
               break;  
3874                }                }
3875              else              else
3876                {                {
3877                c = max - min;                eptr = md->end_subject;
               if (c > (unsigned int)(md->end_subject - eptr))  
                 c = md->end_subject - eptr;  
               eptr += c;  
3878                }                }
3879              }              }
3880            break;            break;
# Line 3379  for (;;) Line 3901  for (;;)
3901                }                }
3902              else              else
3903                {                {
3904                if (c != 0x000a && c != 0x000b && c != 0x000c &&                if (c != 0x000a &&
3905                    c != 0x0085 && c != 0x2028 && c != 0x2029)                    (md->bsr_anycrlf ||
3906                       (c != 0x000b && c != 0x000c &&
3907                        c != 0x0085 && c != 0x2028 && c != 0x2029)))
3908                  break;                  break;
3909                eptr += len;                eptr += len;
3910                }                }
3911              }              }
3912            break;            break;
3913    
3914              case OP_NOT_HSPACE:
3915              case OP_HSPACE:
3916              for (i = min; i < max; i++)
3917                {
3918                BOOL gotspace;
3919                int len = 1;
3920                if (eptr >= md->end_subject) break;
3921                GETCHARLEN(c, eptr, len);
3922                switch(c)
3923                  {
3924                  default: gotspace = FALSE; break;
3925                  case 0x09:      /* HT */
3926                  case 0x20:      /* SPACE */
3927                  case 0xa0:      /* NBSP */
3928                  case 0x1680:    /* OGHAM SPACE MARK */
3929                  case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3930                  case 0x2000:    /* EN QUAD */
3931                  case 0x2001:    /* EM QUAD */
3932                  case 0x2002:    /* EN SPACE */
3933                  case 0x2003:    /* EM SPACE */
3934                  case 0x2004:    /* THREE-PER-EM SPACE */
3935                  case 0x2005:    /* FOUR-PER-EM SPACE */
3936                  case 0x2006:    /* SIX-PER-EM SPACE */
3937                  case 0x2007:    /* FIGURE SPACE */
3938                  case 0x2008:    /* PUNCTUATION SPACE */
3939                  case 0x2009:    /* THIN SPACE */
3940                  case 0x200A:    /* HAIR SPACE */
3941                  case 0x202f:    /* NARROW NO-BREAK SPACE */
3942                  case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3943                  case 0x3000:    /* IDEOGRAPHIC SPACE */
3944                  gotspace = TRUE;
3945                  break;
3946                  }
3947                if (gotspace == (ctype == OP_NOT_HSPACE)) break;
3948                eptr += len;
3949                }
3950              break;
3951    
3952              case OP_NOT_VSPACE:
3953              case OP_VSPACE:
3954              for (i = min; i < max; i++)
3955                {
3956                BOOL gotspace;
3957                int len = 1;
3958                if (eptr >= md->end_subject) break;
3959                GETCHARLEN(c, eptr, len);
3960                switch(c)
3961                  {
3962                  default: gotspace = FALSE; break;
3963                  case 0x0a:      /* LF */
3964                  case 0x0b:      /* VT */
3965                  case 0x0c:      /* FF */
3966                  case 0x0d:      /* CR */
3967                  case 0x85:      /* NEL */
3968                  case 0x2028:    /* LINE SEPARATOR */
3969                  case 0x2029:    /* PARAGRAPH SEPARATOR */
3970                  gotspace = TRUE;
3971                  break;
3972                  }
3973                if (gotspace == (ctype == OP_NOT_VSPACE)) break;
3974                eptr += len;
3975                }
3976              break;
3977    
3978            case OP_NOT_DIGIT:            case OP_NOT_DIGIT:
3979            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3980              {              {
# Line 3462  for (;;) Line 4050  for (;;)
4050          if (possessive) continue;          if (possessive) continue;
4051          for(;;)          for(;;)
4052            {            {
4053            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM46);
4054            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4055            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
4056            BACKCHAR(eptr);            BACKCHAR(eptr);
4057            }            }
4058          }          }
4059        else        else
4060  #endif  #endif  /* SUPPORT_UTF8 */
4061    
4062        /* Not UTF-8 mode */        /* Not UTF-8 mode */
4063          {          {
# Line 3506  for (;;) Line 4094  for (;;)
4094                }                }
4095              else              else
4096                {                {
4097                if (c != 0x000a && c != 0x000b && c != 0x000c && c != 0x0085)                if (c != 0x000a &&
4098                      (md->bsr_anycrlf ||
4099                        (c != 0x000b && c != 0x000c && c != 0x0085)))
4100                  break;                  break;
4101                eptr++;                eptr++;
4102                }                }
4103              }              }
4104            break;            break;
4105    
4106              case OP_NOT_HSPACE:
4107              for (i = min; i < max; i++)
4108                {
4109                if (eptr >= md->end_subject) break;
4110                c = *eptr;
4111                if (c == 0x09 || c == 0x20 || c == 0xa0) break;
4112                eptr++;
4113                }
4114              break;
4115    
4116              case OP_HSPACE:
4117              for (i = min; i < max; i++)
4118                {
4119                if (eptr >= md->end_subject) break;
4120                c = *eptr;
4121                if (c != 0x09 && c != 0x20 && c != 0xa0) break;
4122                eptr++;
4123                }
4124              break;
4125    
4126              case OP_NOT_VSPACE:
4127              for (i = min; i < max; i++)
4128                {
4129                if (eptr >= md->end_subject) break;
4130                c = *eptr;
4131                if (c == 0x0a || c == 0x0b || c == 0x0c || c == 0x0d || c == 0x85)
4132                  break;
4133                eptr++;
4134                }
4135              break;
4136    
4137              case OP_VSPACE:
4138              for (i = min; i < max; i++)
4139                {
4140                if (eptr >= md->end_subject) break;
4141                c = *eptr;
4142                if (c != 0x0a && c != 0x0b && c != 0x0c && c != 0x0d && c != 0x85)
4143                  break;
4144                eptr++;
4145                }
4146              break;
4147    
4148            case OP_NOT_DIGIT:            case OP_NOT_DIGIT:
4149            for (i = min; i < max; i++)            for (i = min; i < max; i++)
4150              {              {
# Line 3576  for (;;) Line 4208  for (;;)
4208          if (possessive) continue;          if (possessive) continue;
4209          while (eptr >= pp)          while (eptr >= pp)
4210            {            {
4211            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM47);
4212            eptr--;            eptr--;
4213            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4214            }            }
# Line 3602  for (;;) Line 4234  for (;;)
4234    
4235    }             /* End of main loop */    }             /* End of main loop */
4236  /* Control never reaches here */  /* Control never reaches here */
4237    
4238    
4239    /* When compiling to use the heap rather than the stack for recursive calls to
4240    match(), the RRETURN() macro jumps here. The number that is saved in
4241    frame->Xwhere indicates which label we actually want to return to. */
4242    
4243    #ifdef NO_RECURSE
4244    #define LBL(val) case val: goto L_RM##val;
4245    HEAP_RETURN:
4246    switch (frame->Xwhere)
4247      {
4248      LBL( 1) LBL( 2) LBL( 3) LBL( 4) LBL( 5) LBL( 6) LBL( 7) LBL( 8)
4249      LBL( 9) LBL(10) LBL(11) LBL(12) LBL(13) LBL(14) LBL(15) LBL(16)
4250      LBL(17) LBL(18) LBL(19) LBL(20) LBL(21) LBL(22) LBL(23) LBL(24)
4251      LBL(25) LBL(26) LBL(27) LBL(28) LBL(29) LBL(30) LBL(31) LBL(32)
4252      LBL(33) LBL(34) LBL(35) LBL(36) LBL(37) LBL(38) LBL(39) LBL(40)
4253      LBL(41) LBL(42) LBL(43) LBL(44) LBL(45) LBL(46) LBL(47) LBL(48)
4254      LBL(49) LBL(50) LBL(51) LBL(52) LBL(53) LBL(54)
4255      default:
4256      DPRINTF(("jump error in pcre match: label %d non-existent\n", frame->Xwhere));
4257      return PCRE_ERROR_INTERNAL;
4258      }
4259    #undef LBL
4260    #endif  /* NO_RECURSE */
4261  }  }
4262    
4263    
# Line 3614  Undefine all the macros that were define Line 4270  Undefine all the macros that were define
4270  #ifdef NO_RECURSE  #ifdef NO_RECURSE
4271  #undef eptr  #undef eptr
4272  #undef ecode  #undef ecode
4273    #undef mstart
4274  #undef offset_top  #undef offset_top
4275  #undef ims  #undef ims
4276  #undef eptrb  #undef eptrb
# Line 3686  Returns: > 0 => success; value Line 4343  Returns: > 0 => success; value
4343                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
4344  */  */
4345    
4346  PCRE_DATA_SCOPE int  PCRE_EXP_DEFN int
4347  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,
4348    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,
4349    int offsetcount)    int offsetcount)
# Line 3711  const uschar *start_bits = NULL; Line 4368  const uschar *start_bits = NULL;
4368  USPTR start_match = (USPTR)subject + start_offset;  USPTR start_match = (USPTR)subject + start_offset;
4369  USPTR end_subject;  USPTR end_subject;
4370  USPTR req_byte_ptr = start_match - 1;  USPTR req_byte_ptr = start_match - 1;
 eptrblock eptrchain[EPTR_WORK_SIZE];  
4371    
4372  pcre_study_data internal_study;  pcre_study_data internal_study;
4373  const pcre_study_data *study;  const pcre_study_data *study;
# Line 3774  if (re->magic_number != MAGIC_NUMBER) Line 4430  if (re->magic_number != MAGIC_NUMBER)
4430  /* Set up other data */  /* Set up other data */
4431    
4432  anchored = ((re->options | options) & PCRE_ANCHORED) != 0;  anchored = ((re->options | options) & PCRE_ANCHORED) != 0;
4433  startline = (re->options & PCRE_STARTLINE) != 0;  startline = (re->flags & PCRE_STARTLINE) != 0;
4434  firstline = (re->options & PCRE_FIRSTLINE) != 0;  firstline = (re->options & PCRE_FIRSTLINE) != 0;
4435    
4436  /* 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 3797  md->partial = (options & PCRE_PARTIAL) ! Line 4453  md->partial = (options & PCRE_PARTIAL) !
4453  md->hitend = FALSE;  md->hitend = FALSE;
4454    
4455  md->recursive = NULL;                   /* No recursion at top level */  md->recursive = NULL;                   /* No recursion at top level */
 md->eptrchain = eptrchain;              /* Make workspace generally available */  
4456    
4457  md->lcc = tables + lcc_offset;  md->lcc = tables + lcc_offset;
4458  md->ctypes = tables + ctypes_offset;  md->ctypes = tables + ctypes_offset;
4459    
4460  /* Handle different types of newline. The two bits give four cases. If nothing  /* Handle different \R options. */
4461  is set at run time, whatever was used at compile time applies. */  
4462    switch (options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE))
4463      {
4464      case 0:
4465      if ((re->options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) != 0)
4466        md->bsr_anycrlf = (re->options & PCRE_BSR_ANYCRLF) != 0;
4467      else
4468    #ifdef BSR_ANYCRLF
4469      md->bsr_anycrlf = TRUE;
4470    #else
4471      md->bsr_anycrlf = FALSE;
4472    #endif
4473      break;
4474    
4475      case PCRE_BSR_ANYCRLF:
4476      md->bsr_anycrlf = TRUE;
4477      break;
4478    
4479      case PCRE_BSR_UNICODE:
4480      md->bsr_anycrlf = FALSE;
4481      break;
4482    
4483      default: return PCRE_ERROR_BADNEWLINE;
4484      }
4485    
4486    /* Handle different types of newline. The three bits give eight cases. If
4487    nothing is set at run time, whatever was used at compile time applies. */
4488    
4489  switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options : options) &  switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options :
4490         PCRE_NEWLINE_BITS)          (pcre_uint32)options) & PCRE_NEWLINE_BITS)
4491    {    {
4492    case 0: newline = NEWLINE; break;   /* Compile-time default */    case 0: newline = NEWLINE; break;   /* Compile-time default */
4493    case PCRE_NEWLINE_CR: newline = '\r'; break;    case PCRE_NEWLINE_CR: newline = '\r'; break;
# Line 3814  switch ((((options & PCRE_NEWLINE_BITS) Line 4495  switch ((((options & PCRE_NEWLINE_BITS)
4495    case PCRE_NEWLINE_CR+    case PCRE_NEWLINE_CR+
4496         PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break;         PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break;
4497    case PCRE_NEWLINE_ANY: newline = -1; break;    case PCRE_NEWLINE_ANY: newline = -1; break;
4498      case PCRE_NEWLINE_ANYCRLF: newline = -2; break;
4499    default: return PCRE_ERROR_BADNEWLINE;    default: return PCRE_ERROR_BADNEWLINE;
4500    }    }
4501    
4502  if (newline < 0)  if (newline == -2)
4503      {
4504      md->nltype = NLTYPE_ANYCRLF;
4505      }
4506    else if (newline < 0)
4507    {    {
4508    md->nltype = NLTYPE_ANY;    md->nltype = NLTYPE_ANY;
4509    }    }
# Line 3840  else Line 4526  else
4526  /* 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
4527  moment. */  moment. */
4528    
4529  if (md->partial && (re->options & PCRE_NOPARTIAL) != 0)  if (md->partial && (re->flags & PCRE_NOPARTIAL) != 0)
4530    return PCRE_ERROR_BADPARTIAL;    return PCRE_ERROR_BADPARTIAL;
4531    
4532  /* 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 3917  studied, there may be a bitmap of possib Line 4603  studied, there may be a bitmap of possib
4603    
4604  if (!anchored)  if (!anchored)
4605    {    {
4606    if ((re->options & PCRE_FIRSTSET) != 0)    if ((re->flags & PCRE_FIRSTSET) != 0)
4607      {      {
4608      first_byte = re->first_byte & 255;      first_byte = re->first_byte & 255;
4609      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)
# Line 3932  if (!anchored) Line 4618  if (!anchored)
4618  /* For anchored or unanchored matches, there may be a "last known required  /* For anchored or unanchored matches, there may be a "last known required
4619  character" set. */  character" set. */
4620    
4621  if ((re->options & PCRE_REQCHSET) != 0)  if ((re->flags & PCRE_REQCHSET) != 0)
4622    {    {
4623    req_byte = re->req_byte & 255;    req_byte = re->req_byte & 255;
4624    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;
# Line 3948  the loop runs just once. */ Line 4634  the loop runs just once. */
4634  for(;;)  for(;;)
4635    {    {
4636    USPTR save_end_subject = end_subject;    USPTR save_end_subject = end_subject;
4637      USPTR new_start_match;
4638    
4639    /* Reset the maximum number of extractions we might see. */    /* Reset the maximum number of extractions we might see. */
4640    
# Line 3992  for(;;) Line 4679  for(;;)
4679        {        {
4680        while (start_match <= end_subject && !WAS_NEWLINE(start_match))        while (start_match <= end_subject && !WAS_NEWLINE(start_match))
4681          start_match++;          start_match++;
4682    
4683          /* If we have just passed a CR and the newline option is ANY or ANYCRLF,
4684          and we are now at a LF, advance the match position by one more character.
4685          */
4686    
4687          if (start_match[-1] == '\r' &&
4688               (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&
4689               start_match < end_subject &&
4690               *start_match == '\n')
4691            start_match++;
4692        }        }
4693      }      }
4694    
# Line 4078  for(;;) Line 4775  for(;;)
4775    
4776    /* OK, we can now run the match. */    /* OK, we can now run the match. */
4777    
4778    md->start_match = start_match;    md->start_match_ptr = start_match;
4779    md->match_call_count = 0;    md->match_call_count = 0;
4780    md->eptrn = 0;                          /* Next free eptrchain slot */    rc = match(start_match, md->start_code, start_match, 2, md, ims, NULL, 0, 0);
4781    rc = match(start_match, md->start_code, 2, md, ims, NULL, 0, 0);  
4782      switch(rc)
4783        {
4784        /* NOMATCH and PRUNE advance by one character. THEN at this level acts
4785        exactly like PRUNE. */
4786    
4787    /* Any return other than MATCH_NOMATCH breaks the loop. */      case MATCH_NOMATCH:
4788        case MATCH_PRUNE:
4789        case MATCH_THEN:
4790        new_start_match = start_match + 1;
4791    #ifdef SUPPORT_UTF8
4792        if (utf8)
4793          while(new_start_match < end_subject && (*new_start_match & 0xc0) == 0x80)
4794            new_start_match++;
4795    #endif
4796        break;
4797    
4798        /* SKIP passes back the next starting point explicitly. */
4799    
4800    if (rc != MATCH_NOMATCH) break;      case MATCH_SKIP:
4801        new_start_match = md->start_match_ptr;
4802        break;
4803    
4804        /* COMMIT disables the bumpalong, but otherwise behaves as NOMATCH. */
4805    
4806        case MATCH_COMMIT:
4807        rc = MATCH_NOMATCH;
4808        goto ENDLOOP;
4809    
4810        /* Any other return is some kind of error. */
4811    
4812        default:
4813        goto ENDLOOP;
4814        }
4815    
4816      /* Control reaches here for the various types of "no match at this point"
4817      result. Reset the code to MATCH_NOMATCH for subsequent checking. */
4818    
4819      rc = MATCH_NOMATCH;
4820    
4821    /* 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
4822    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 4093  for(;;) Line 4824  for(;;)
4824    
4825    if (firstline && IS_NEWLINE(start_match)) break;    if (firstline && IS_NEWLINE(start_match)) break;
4826    
4827    /* Advance the match position by one character. */    /* Advance to new matching position */
4828    
4829    start_match++;    start_match = new_start_match;
 #ifdef SUPPORT_UTF8  
   if (utf8)  
     while(start_match < end_subject && (*start_match & 0xc0) == 0x80)  
       start_match++;  
 #endif  
4830    
4831    /* 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
4832    the subject. */    the subject. */
4833    
4834    if (anchored || start_match > end_subject) break;    if (anchored || start_match > end_subject) break;
4835    
4836    /* 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
4837    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
4838      or ANY or ANYCRLF, advance the match position by one more character. */
4839    
4840    if (start_match[-1] == '\r' &&    if (start_match[-1] == '\r' &&
4841         (md->nltype == NLTYPE_ANY || md->nllen == 2) &&        start_match < end_subject &&
4842         start_match < end_subject &&        *start_match == '\n' &&
4843         *start_match == '\n')        (re->flags & PCRE_HASCRORLF) == 0 &&
4844            (md->nltype == NLTYPE_ANY ||
4845             md->nltype == NLTYPE_ANYCRLF ||
4846             md->nllen == 2))
4847      start_match++;      start_match++;
4848    
4849    }   /* End of for(;;) "bumpalong" loop */    }   /* End of for(;;) "bumpalong" loop */
# Line 4123  for(;;) Line 4853  for(;;)
4853  /* 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
4854  conditions is true:  conditions is true:
4855    
4856  (1) The pattern is anchored;  (1) The pattern is anchored or the match was failed by (*COMMIT);
4857    
4858  (2) We are past the end of the subject;  (2) We are past the end of the subject;
4859    
# Line 4138  processing, copy those that we can. In t Line 4868  processing, copy those that we can. In t
4868  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
4869  capturing parentheses than vector slots. */  capturing parentheses than vector slots. */
4870    
4871    ENDLOOP:
4872    
4873  if (rc == MATCH_MATCH)  if (rc == MATCH_MATCH)
4874    {    {
4875    if (using_temporary_offsets)    if (using_temporary_offsets)
# Line 4158  if (rc == MATCH_MATCH) Line 4890  if (rc == MATCH_MATCH)
4890    
4891    rc = md->offset_overflow? 0 : md->end_offset_top/2;    rc = md->offset_overflow? 0 : md->end_offset_top/2;
4892    
4893    /* 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
4894      md->start_match_ptr might be modified if \K was encountered on the success
4895      matching path. */
4896    
4897    if (offsetcount < 2) rc = 0; else    if (offsetcount < 2) rc = 0; else
4898      {      {
4899      offsets[0] = start_match - md->start_subject;      offsets[0] = md->start_match_ptr - md->start_subject;
4900      offsets[1] = md->end_match_ptr - md->start_subject;      offsets[1] = md->end_match_ptr - md->start_subject;
4901      }      }
4902    

Legend:
Removed from v.96  
changed lines
  Added in v.243

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12