/[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 184 by ph10, Tue Jun 19 13:26:46 2007 UTC revision 435 by ph10, Sat Sep 5 10:20:44 2009 UTC
# Line 6  Line 6 
6  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
7    
8                         Written by Philip Hazel                         Written by Philip Hazel
9             Copyright (c) 1997-2007 University of Cambridge             Copyright (c) 1997-2009 University of Cambridge
10    
11  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
12  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 42  POSSIBILITY OF SUCH DAMAGE. Line 42  POSSIBILITY OF SUCH DAMAGE.
42  pattern matching using an NFA algorithm, trying to mimic Perl as closely as  pattern matching using an NFA algorithm, trying to mimic Perl as closely as
43  possible. There are also some static supporting functions. */  possible. There are also some static supporting functions. */
44    
45    #ifdef HAVE_CONFIG_H
46    #include "config.h"
47    #endif
48    
49  #define NLBLOCK md             /* Block containing newline information */  #define NLBLOCK md             /* Block containing newline information */
50  #define PSSTART start_subject  /* Field containing processed string start */  #define PSSTART start_subject  /* Field containing processed string start */
51  #define PSEND   end_subject    /* Field containing processed string end */  #define PSEND   end_subject    /* Field containing processed string end */
# Line 53  possible. There are also some static sup Line 57  possible. There are also some static sup
57  #undef min  #undef min
58  #undef max  #undef max
59    
 /* The chain of eptrblocks for tail recursions uses memory in stack workspace,  
 obtained at top level, the size of which is defined by EPTR_WORK_SIZE. */  
   
 #define EPTR_WORK_SIZE (1000)  
   
60  /* Flag bits for the match() function */  /* Flag bits for the match() function */
61    
62  #define match_condassert     0x01  /* Called to check a condition assertion */  #define match_condassert     0x01  /* Called to check a condition assertion */
63  #define match_cbegroup       0x02  /* Could-be-empty unlimited repeat group */  #define match_cbegroup       0x02  /* Could-be-empty unlimited repeat group */
 #define match_tail_recursed  0x04  /* Tail recursive call */  
64    
65  /* Non-error returns from the match() function. Error returns are externally  /* Non-error returns from the match() function. Error returns are externally
66  defined PCRE_ERROR_xxx codes, which are all negative. */  defined PCRE_ERROR_xxx codes, which are all negative. */
# Line 70  defined PCRE_ERROR_xxx codes, which are Line 68  defined PCRE_ERROR_xxx codes, which are
68  #define MATCH_MATCH        1  #define MATCH_MATCH        1
69  #define MATCH_NOMATCH      0  #define MATCH_NOMATCH      0
70    
71    /* Special internal returns from the match() function. Make them sufficiently
72    negative to avoid the external error codes. */
73    
74    #define MATCH_COMMIT       (-999)
75    #define MATCH_PRUNE        (-998)
76    #define MATCH_SKIP         (-997)
77    #define MATCH_THEN         (-996)
78    
79  /* Maximum number of ints of offset to save on the stack for recursive calls.  /* Maximum number of ints of offset to save on the stack for recursive calls.
80  If the offset vector is bigger, malloc is used. This should be a multiple of 3,  If the offset vector is bigger, malloc is used. This should be a multiple of 3,
81  because the offset vector is always a multiple of 3 long. */  because the offset vector is always a multiple of 3 long. */
# Line 152  printf("\n"); Line 158  printf("\n");
158    
159  if (length > md->end_subject - eptr) return FALSE;  if (length > md->end_subject - eptr) return FALSE;
160    
161  /* Separate the caselesss case for speed */  /* Separate the caseless case for speed. In UTF-8 mode we can only do this
162    properly if Unicode properties are supported. Otherwise, we can check only
163    ASCII characters. */
164    
165  if ((ims & PCRE_CASELESS) != 0)  if ((ims & PCRE_CASELESS) != 0)
166    {    {
167    #ifdef SUPPORT_UTF8
168    #ifdef SUPPORT_UCP
169      if (md->utf8)
170        {
171        USPTR endptr = eptr + length;
172        while (eptr < endptr)
173          {
174          int c, d;
175          GETCHARINC(c, eptr);
176          GETCHARINC(d, p);
177          if (c != d && c != UCD_OTHERCASE(d)) return FALSE;
178          }
179        }
180      else
181    #endif
182    #endif
183    
184      /* The same code works when not in UTF-8 mode and in UTF-8 mode when there
185      is no UCP support. */
186    
187    while (length-- > 0)    while (length-- > 0)
188      if (md->lcc[*p++] != md->lcc[*eptr++]) return FALSE;      { if (md->lcc[*p++] != md->lcc[*eptr++]) return FALSE; }
189    }    }
190    
191    /* In the caseful case, we can just compare the bytes, whether or not we
192    are in UTF-8 mode. */
193    
194  else  else
195    { while (length-- > 0) if (*p++ != *eptr++) return FALSE; }    { while (length-- > 0) if (*p++ != *eptr++) return FALSE; }
196    
# Line 205  variable instead of being passed in the Line 237  variable instead of being passed in the
237  ****************************************************************************  ****************************************************************************
238  ***************************************************************************/  ***************************************************************************/
239    
240    /* Numbers for RMATCH calls. When this list is changed, the code at HEAP_RETURN
241  /* Numbers for RMATCH calls */  below must be updated in sync.  */
242    
243  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM6,  RM7,  RM8,  RM9,  RM10,  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM6,  RM7,  RM8,  RM9,  RM10,
244         RM11,  RM12, RM13, RM14, RM15, RM16, RM17, RM18, RM19, RM20,         RM11,  RM12, RM13, RM14, RM15, RM16, RM17, RM18, RM19, RM20,
245         RM21,  RM22, RM23, RM24, RM25, RM26, RM27, RM28, RM29, RM30,         RM21,  RM22, RM23, RM24, RM25, RM26, RM27, RM28, RM29, RM30,
246         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,
247         RM41,  RM42, RM43, RM44, RM45, RM46, RM47 };         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,
248           RM51,  RM52, RM53, RM54 };
249    
250  /* These versions of the macros use the stack, as normal. There are debugging  /* These versions of the macros use the stack, as normal. There are debugging
251  versions and production versions. Note that the "rw" argument of RMATCH isn't  versions and production versions. Note that the "rw" argument of RMATCH isn't
# Line 290  typedef struct heapframe { Line 322  typedef struct heapframe {
322    
323    /* Function arguments that may change */    /* Function arguments that may change */
324    
325    const uschar *Xeptr;    USPTR Xeptr;
326    const uschar *Xecode;    const uschar *Xecode;
327    const uschar *Xmstart;    USPTR Xmstart;
328    int Xoffset_top;    int Xoffset_top;
329    long int Xims;    long int Xims;
330    eptrblock *Xeptrb;    eptrblock *Xeptrb;
# Line 301  typedef struct heapframe { Line 333  typedef struct heapframe {
333    
334    /* Function local variables */    /* Function local variables */
335    
336    const uschar *Xcallpat;    USPTR Xcallpat;
337    const uschar *Xcharptr;  #ifdef SUPPORT_UTF8
338    const uschar *Xdata;    USPTR Xcharptr;
339    const uschar *Xnext;  #endif
340    const uschar *Xpp;    USPTR Xdata;
341    const uschar *Xprev;    USPTR Xnext;
342    const uschar *Xsaved_eptr;    USPTR Xpp;
343      USPTR Xprev;
344      USPTR Xsaved_eptr;
345    
346    recursion_info Xnew_recursive;    recursion_info Xnew_recursive;
347    
# Line 328  typedef struct heapframe { Line 362  typedef struct heapframe {
362    uschar Xocchars[8];    uschar Xocchars[8];
363  #endif  #endif
364    
365      int Xcodelink;
366    int Xctype;    int Xctype;
367    unsigned int Xfc;    unsigned int Xfc;
368    int Xfi;    int Xfi;
# Line 363  typedef struct heapframe { Line 398  typedef struct heapframe {
398    
399  /* This function is called recursively in many circumstances. Whenever it  /* This function is called recursively in many circumstances. Whenever it
400  returns a negative (error) response, the outer incarnation must also return the  returns a negative (error) response, the outer incarnation must also return the
401  same response.  same response. */
402    
403    /* These macros pack up tests that are used for partial matching, and which
404    appears several times in the code. We set the "hit end" flag if the pointer is
405    at the end of the subject and also past the start of the subject (i.e.
406    something has been matched). For hard partial matching, we then return
407    immediately. The second one is used when we already know we are past the end of
408    the subject. */
409    
410  Performance note: It might be tempting to extract commonly used fields from the  #define CHECK_PARTIAL()\
411  md structure (e.g. utf8, end_subject) into individual variables to improve    if (md->partial != 0 && eptr >= md->end_subject && eptr > mstart)\
412        {\
413        md->hitend = TRUE;\
414        if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);\
415        }
416    
417    #define SCHECK_PARTIAL()\
418      if (md->partial && eptr > mstart)\
419        {\
420        md->hitend = TRUE;\
421        if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);\
422        }
423    
424    
425    /* Performance note: It might be tempting to extract commonly used fields from
426    the md structure (e.g. utf8, end_subject) into individual variables to improve
427  performance. Tests using gcc on a SPARC disproved this; in the first case, it  performance. Tests using gcc on a SPARC disproved this; in the first case, it
428  made performance worse.  made performance worse.
429    
# Line 384  Arguments: Line 441  Arguments:
441                   match_condassert - this is an assertion condition                   match_condassert - this is an assertion condition
442                   match_cbegroup - this is the start of an unlimited repeat                   match_cbegroup - this is the start of an unlimited repeat
443                     group that can match an empty string                     group that can match an empty string
                  match_tail_recursed - this is a tail_recursed group  
444     rdepth      the recursion depth     rdepth      the recursion depth
445    
446  Returns:       MATCH_MATCH if matched            )  these values are >= 0  Returns:       MATCH_MATCH if matched            )  these values are >= 0
# Line 394  Returns:       MATCH_MATCH if matched Line 450  Returns:       MATCH_MATCH if matched
450  */  */
451    
452  static int  static int
453  match(REGISTER USPTR eptr, REGISTER const uschar *ecode, const uschar *mstart,  match(REGISTER USPTR eptr, REGISTER const uschar *ecode, USPTR mstart,
454    int offset_top, match_data *md, unsigned long int ims, eptrblock *eptrb,    int offset_top, match_data *md, unsigned long int ims, eptrblock *eptrb,
455    int flags, unsigned int rdepth)    int flags, unsigned int rdepth)
456  {  {
# Line 408  register unsigned int c;   /* Character Line 464  register unsigned int c;   /* Character
464  register BOOL utf8;        /* Local copy of UTF-8 flag for speed */  register BOOL utf8;        /* Local copy of UTF-8 flag for speed */
465    
466  BOOL minimize, possessive; /* Quantifier options */  BOOL minimize, possessive; /* Quantifier options */
467    int condcode;
468    
469  /* When recursion is not being used, all "local" variables that have to be  /* When recursion is not being used, all "local" variables that have to be
470  preserved over calls to RMATCH() are part of a "frame" which is obtained from  preserved over calls to RMATCH() are part of a "frame" which is obtained from
# Line 450  HEAP_RECURSE: Line 507  HEAP_RECURSE:
507  #define charptr            frame->Xcharptr  #define charptr            frame->Xcharptr
508  #endif  #endif
509  #define callpat            frame->Xcallpat  #define callpat            frame->Xcallpat
510    #define codelink           frame->Xcodelink
511  #define data               frame->Xdata  #define data               frame->Xdata
512  #define next               frame->Xnext  #define next               frame->Xnext
513  #define pp                 frame->Xpp  #define pp                 frame->Xpp
# Line 530  int oclength; Line 588  int oclength;
588  uschar occhars[8];  uschar occhars[8];
589  #endif  #endif
590    
591    int codelink;
592  int ctype;  int ctype;
593  int length;  int length;
594  int max;  int max;
# Line 586  original_ims = ims;    /* Save for reset Line 645  original_ims = ims;    /* Save for reset
645  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
646  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
647  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.
648  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
649  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
650  already used. */  block that is used is on the stack, so a new one may be required for each
651    match(). */
652    
653  if ((flags & match_cbegroup) != 0)  if ((flags & match_cbegroup) != 0)
654    {    {
655    eptrblock *p;    newptrb.epb_saved_eptr = eptr;
656    if ((flags & match_tail_recursed) != 0)    newptrb.epb_prev = eptrb;
657      {    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;  
658    }    }
659    
660  /* Now start processing the opcodes. */  /* Now start processing the opcodes. */
# Line 610  for (;;) Line 663  for (;;)
663    {    {
664    minimize = possessive = FALSE;    minimize = possessive = FALSE;
665    op = *ecode;    op = *ecode;
666    
   /* For partial matching, remember if we ever hit the end of the subject after  
   matching at least one subject character. */  
   
   if (md->partial &&  
       eptr >= md->end_subject &&  
       eptr > mstart)  
     md->hitend = TRUE;  
   
667    switch(op)    switch(op)
668      {      {
669        case OP_FAIL:
670        RRETURN(MATCH_NOMATCH);
671    
672        case OP_PRUNE:
673        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
674          ims, eptrb, flags, RM51);
675        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
676        RRETURN(MATCH_PRUNE);
677    
678        case OP_COMMIT:
679        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
680          ims, eptrb, flags, RM52);
681        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
682        RRETURN(MATCH_COMMIT);
683    
684        case OP_SKIP:
685        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
686          ims, eptrb, flags, RM53);
687        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
688        md->start_match_ptr = eptr;   /* Pass back current position */
689        RRETURN(MATCH_SKIP);
690    
691        case OP_THEN:
692        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
693          ims, eptrb, flags, RM54);
694        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
695        RRETURN(MATCH_THEN);
696    
697      /* 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
698      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.
699      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 662  for (;;) Line 735  for (;;)
735          {          {
736          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
737            ims, eptrb, flags, RM1);            ims, eptrb, flags, RM1);
738          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
739          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
740          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
741          }          }
# Line 677  for (;;) Line 750  for (;;)
750        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
751        }        }
752    
753      /* Insufficient room for saving captured contents. Treat as a non-capturing      /* FALL THROUGH ... Insufficient room for saving captured contents. Treat
754      bracket. */      as a non-capturing bracket. */
755    
756        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
757        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
758    
759      DPRINTF(("insufficient capture room: treat as non-capturing\n"));      DPRINTF(("insufficient capture room: treat as non-capturing\n"));
760    
761        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
762        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
763    
764      /* 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
765      final alternative within the brackets, we would return the result of a      final alternative within the brackets, we would return the result of a
766      recursive call to match() whatever happened. We can reduce stack usage by      recursive call to match() whatever happened. We can reduce stack usage by
767      turning this into a tail recursion. */      turning this into a tail recursion, except in the case when match_cbegroup
768        is set.*/
769    
770      case OP_BRA:      case OP_BRA:
771      case OP_SBRA:      case OP_SBRA:
# Line 693  for (;;) Line 773  for (;;)
773      flags = (op >= OP_SBRA)? match_cbegroup : 0;      flags = (op >= OP_SBRA)? match_cbegroup : 0;
774      for (;;)      for (;;)
775        {        {
776        if (ecode[GET(ecode, 1)] != OP_ALT)        if (ecode[GET(ecode, 1)] != OP_ALT)   /* Final alternative */
777          {          {
778          ecode += _pcre_OP_lengths[*ecode];          if (flags == 0)    /* Not a possibly empty group */
779          flags |= match_tail_recursed;            {
780          DPRINTF(("bracket 0 tail recursion\n"));            ecode += _pcre_OP_lengths[*ecode];
781          goto TAIL_RECURSE;            DPRINTF(("bracket 0 tail recursion\n"));
782              goto TAIL_RECURSE;
783              }
784    
785            /* Possibly empty group; can't use tail recursion. */
786    
787            RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,
788              eptrb, flags, RM48);
789            RRETURN(rrc);
790          }          }
791    
792        /* For non-final alternatives, continue the loop for a NOMATCH result;        /* For non-final alternatives, continue the loop for a NOMATCH result;
# Line 706  for (;;) Line 794  for (;;)
794    
795        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,
796          eptrb, flags, RM2);          eptrb, flags, RM2);
797        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
798        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
799        }        }
800      /* Control never reaches here. */      /* Control never reaches here. */
# Line 719  for (;;) Line 807  for (;;)
807    
808      case OP_COND:      case OP_COND:
809      case OP_SCOND:      case OP_SCOND:
810      if (ecode[LINK_SIZE+1] == OP_RREF)         /* Recursion test */      codelink= GET(ecode, 1);
811    
812        /* Because of the way auto-callout works during compile, a callout item is
813        inserted between OP_COND and an assertion condition. */
814    
815        if (ecode[LINK_SIZE+1] == OP_CALLOUT)
816          {
817          if (pcre_callout != NULL)
818            {
819            pcre_callout_block cb;
820            cb.version          = 1;   /* Version 1 of the callout block */
821            cb.callout_number   = ecode[LINK_SIZE+2];
822            cb.offset_vector    = md->offset_vector;
823            cb.subject          = (PCRE_SPTR)md->start_subject;
824            cb.subject_length   = md->end_subject - md->start_subject;
825            cb.start_match      = mstart - md->start_subject;
826            cb.current_position = eptr - md->start_subject;
827            cb.pattern_position = GET(ecode, LINK_SIZE + 3);
828            cb.next_item_length = GET(ecode, 3 + 2*LINK_SIZE);
829            cb.capture_top      = offset_top/2;
830            cb.capture_last     = md->capture_last;
831            cb.callout_data     = md->callout_data;
832            if ((rrc = (*pcre_callout)(&cb)) > 0) RRETURN(MATCH_NOMATCH);
833            if (rrc < 0) RRETURN(rrc);
834            }
835          ecode += _pcre_OP_lengths[OP_CALLOUT];
836          }
837    
838        condcode = ecode[LINK_SIZE+1];
839    
840        /* Now see what the actual condition is */
841    
842        if (condcode == OP_RREF)         /* Recursion test */
843        {        {
844        offset = GET2(ecode, LINK_SIZE + 2);     /* Recursion group number*/        offset = GET2(ecode, LINK_SIZE + 2);     /* Recursion group number*/
845        condition = md->recursive != NULL &&        condition = md->recursive != NULL &&
# Line 727  for (;;) Line 847  for (;;)
847        ecode += condition? 3 : GET(ecode, 1);        ecode += condition? 3 : GET(ecode, 1);
848        }        }
849    
850      else if (ecode[LINK_SIZE+1] == OP_CREF)    /* Group used test */      else if (condcode == OP_CREF)    /* Group used test */
851        {        {
852        offset = GET2(ecode, LINK_SIZE+2) << 1;  /* Doubled ref number */        offset = GET2(ecode, LINK_SIZE+2) << 1;  /* Doubled ref number */
853        condition = offset < offset_top && md->offset_vector[offset] >= 0;        condition = offset < offset_top && md->offset_vector[offset] >= 0;
854        ecode += condition? 3 : GET(ecode, 1);        ecode += condition? 3 : GET(ecode, 1);
855        }        }
856    
857      else if (ecode[LINK_SIZE+1] == OP_DEF)     /* DEFINE - always false */      else if (condcode == OP_DEF)     /* DEFINE - always false */
858        {        {
859        condition = FALSE;        condition = FALSE;
860        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
# Line 754  for (;;) Line 874  for (;;)
874          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);
875          while (*ecode == OP_ALT) ecode += GET(ecode, 1);          while (*ecode == OP_ALT) ecode += GET(ecode, 1);
876          }          }
877        else if (rrc != MATCH_NOMATCH)        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
878          {          {
879          RRETURN(rrc);         /* Need braces because of following else */          RRETURN(rrc);         /* Need braces because of following else */
880          }          }
881        else        else
882          {          {
883          condition = FALSE;          condition = FALSE;
884          ecode += GET(ecode, 1);          ecode += codelink;
885          }          }
886        }        }
887    
888      /* 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,
889      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
890      alternative doesn't exist, we can just plough on. */      match_cbegroup is required for an unlimited repeat of a possibly empty
891        group. If the second alternative doesn't exist, we can just plough on. */
892    
893      if (condition || *ecode == OP_ALT)      if (condition || *ecode == OP_ALT)
894        {        {
895        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
896        flags = match_tail_recursed | ((op == OP_SCOND)? match_cbegroup : 0);        if (op == OP_SCOND)        /* Possibly empty group */
897        goto TAIL_RECURSE;          {
898            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, match_cbegroup, RM49);
899            RRETURN(rrc);
900            }
901          else                       /* Group must match something */
902            {
903            flags = 0;
904            goto TAIL_RECURSE;
905            }
906        }        }
907      else      else                         /* Condition false & no alternative */
908        {        {
909        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
910        }        }
911      break;      break;
912    
913    
914      /* 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
915      restore the offsets appropriately and continue from after the call. */      recursion, we should restore the offsets appropriately and continue from
916        after the call. */
917    
918        case OP_ACCEPT:
919      case OP_END:      case OP_END:
920      if (md->recursive != NULL && md->recursive->group_num == 0)      if (md->recursive != NULL && md->recursive->group_num == 0)
921        {        {
# Line 805  for (;;) Line 936  for (;;)
936      if (md->notempty && eptr == mstart) RRETURN(MATCH_NOMATCH);      if (md->notempty && eptr == mstart) RRETURN(MATCH_NOMATCH);
937      md->end_match_ptr = eptr;           /* Record where we ended */      md->end_match_ptr = eptr;           /* Record where we ended */
938      md->end_offset_top = offset_top;    /* and how many extracts were taken */      md->end_offset_top = offset_top;    /* and how many extracts were taken */
939      md->start_match_ptr = mstart;  /* and the start (\K can modify) */      md->start_match_ptr = mstart;       /* and the start (\K can modify) */
940      RRETURN(MATCH_MATCH);      RRETURN(MATCH_MATCH);
941    
942      /* Change option settings */      /* Change option settings */
# Line 829  for (;;) Line 960  for (;;)
960        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,
961          RM4);          RM4);
962        if (rrc == MATCH_MATCH) break;        if (rrc == MATCH_MATCH) break;
963        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
964        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
965        }        }
966      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 856  for (;;) Line 987  for (;;)
987        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,
988          RM5);          RM5);
989        if (rrc == MATCH_MATCH) RRETURN(MATCH_NOMATCH);        if (rrc == MATCH_MATCH) RRETURN(MATCH_NOMATCH);
990        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
991        ecode += GET(ecode,1);        ecode += GET(ecode,1);
992        }        }
993      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 880  for (;;) Line 1011  for (;;)
1011          {          {
1012          eptr--;          eptr--;
1013          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1014          BACKCHAR(eptr)          BACKCHAR(eptr);
1015          }          }
1016        }        }
1017      else      else
# Line 893  for (;;) Line 1024  for (;;)
1024        if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);        if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1025        }        }
1026    
1027      /* Skip to next op code */      /* Save the earliest consulted character, then skip to next op code */
1028    
1029        if (eptr < md->start_used_ptr) md->start_used_ptr = eptr;
1030      ecode += 1 + LINK_SIZE;      ecode += 1 + LINK_SIZE;
1031      break;      break;
1032    
# Line 993  for (;;) Line 1125  for (;;)
1125              (pcre_free)(new_recursive.offset_save);              (pcre_free)(new_recursive.offset_save);
1126            RRETURN(MATCH_MATCH);            RRETURN(MATCH_MATCH);
1127            }            }
1128          else if (rrc != MATCH_NOMATCH)          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
1129            {            {
1130            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1131              if (new_recursive.offset_save != stacksave)
1132                (pcre_free)(new_recursive.offset_save);
1133            RRETURN(rrc);            RRETURN(rrc);
1134            }            }
1135    
# Line 1027  for (;;) Line 1161  for (;;)
1161    
1162      do      do
1163        {        {
1164        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM7);
         eptrb, 0, RM7);  
1165        if (rrc == MATCH_MATCH) break;        if (rrc == MATCH_MATCH) break;
1166        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1167        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1168        }        }
1169      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 1073  for (;;) Line 1206  for (;;)
1206    
1207      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1208        {        {
1209        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM8);
         RM8);  
1210        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1211        ecode = prev;        ecode = prev;
1212        flags = match_tail_recursed;        flags = 0;
1213        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1214        }        }
1215      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
# Line 1085  for (;;) Line 1217  for (;;)
1217        RMATCH(eptr, prev, offset_top, md, ims, eptrb, match_cbegroup, RM9);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, match_cbegroup, RM9);
1218        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1219        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1220        flags = match_tail_recursed;        flags = 0;
1221        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1222        }        }
1223      /* Control never gets here */      /* Control never gets here */
# Line 1097  for (;;) Line 1229  for (;;)
1229      do ecode += GET(ecode,1); while (*ecode == OP_ALT);      do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1230      break;      break;
1231    
1232      /* BRAZERO and BRAMINZERO occur just before a bracket group, indicating      /* BRAZERO, BRAMINZERO and SKIPZERO occur just before a bracket group,
1233      that it may occur zero times. It may repeat infinitely, or not at all -      indicating that it may occur zero times. It may repeat infinitely, or not
1234      i.e. it could be ()* or ()? in the pattern. Brackets with fixed upper      at all - i.e. it could be ()* or ()? or even (){0} in the pattern. Brackets
1235      repeat limits are compiled as a number of copies, with the optional ones      with fixed upper repeat limits are compiled as a number of copies, with the
1236      preceded by BRAZERO or BRAMINZERO. */      optional ones preceded by BRAZERO or BRAMINZERO. */
1237    
1238      case OP_BRAZERO:      case OP_BRAZERO:
1239        {        {
# Line 1123  for (;;) Line 1255  for (;;)
1255        }        }
1256      break;      break;
1257    
1258        case OP_SKIPZERO:
1259          {
1260          next = ecode+1;
1261          do next += GET(next,1); while (*next == OP_ALT);
1262          ecode = next + 1 + LINK_SIZE;
1263          }
1264        break;
1265    
1266      /* End of a group, repeated or non-repeating. */      /* End of a group, repeated or non-repeating. */
1267    
1268      case OP_KET:      case OP_KET:
# Line 1216  for (;;) Line 1356  for (;;)
1356    
1357      /* 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
1358      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
1359      tail recursion to avoid using another stack frame. */      tail recursion to avoid using another stack frame, unless we have an
1360        unlimited repeat of a group that can match an empty string. */
1361    
1362      flags = (*prev >= OP_SBRA)? match_cbegroup : 0;      flags = (*prev >= OP_SBRA)? match_cbegroup : 0;
1363    
1364      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1365        {        {
1366        RMATCH(eptr, ecode + 1+LINK_SIZE, offset_top, md, ims, eptrb, 0,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM12);
         RM12);  
1367        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1368          if (flags != 0)    /* Could match an empty string */
1369            {
1370            RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM50);
1371            RRETURN(rrc);
1372            }
1373        ecode = prev;        ecode = prev;
       flags |= match_tail_recursed;  
1374        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1375        }        }
1376      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
# Line 1234  for (;;) Line 1378  for (;;)
1378        RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM13);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM13);
1379        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1380        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1381        flags = match_tail_recursed;        flags = 0;
1382        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1383        }        }
1384      /* Control never gets here */      /* Control never gets here */
# Line 1325  for (;;) Line 1469  for (;;)
1469    
1470        /* Find out if the previous and current characters are "word" characters.        /* Find out if the previous and current characters are "word" characters.
1471        It takes a bit more work in UTF-8 mode. Characters > 255 are assumed to        It takes a bit more work in UTF-8 mode. Characters > 255 are assumed to
1472        be "non-word" characters. */        be "non-word" characters. Remember the earliest consulted character for
1473          partial matching. */
1474    
1475  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
1476        if (utf8)        if (utf8)
1477          {          {
1478          if (eptr == md->start_subject) prev_is_word = FALSE; else          if (eptr == md->start_subject) prev_is_word = FALSE; else
1479            {            {
1480            const uschar *lastptr = eptr - 1;            USPTR lastptr = eptr - 1;
1481            while((*lastptr & 0xc0) == 0x80) lastptr--;            while((*lastptr & 0xc0) == 0x80) lastptr--;
1482              if (lastptr < md->start_used_ptr) md->start_used_ptr = lastptr;
1483            GETCHAR(c, lastptr);            GETCHAR(c, lastptr);
1484            prev_is_word = c < 256 && (md->ctypes[c] & ctype_word) != 0;            prev_is_word = c < 256 && (md->ctypes[c] & ctype_word) != 0;
1485            }            }
1486          if (eptr >= md->end_subject) cur_is_word = FALSE; else          if (eptr >= md->end_subject)
1487              {
1488              SCHECK_PARTIAL();
1489              cur_is_word = FALSE;
1490              }
1491            else
1492            {            {
1493            GETCHAR(c, eptr);            GETCHAR(c, eptr);
1494            cur_is_word = c < 256 && (md->ctypes[c] & ctype_word) != 0;            cur_is_word = c < 256 && (md->ctypes[c] & ctype_word) != 0;
# Line 1346  for (;;) Line 1497  for (;;)
1497        else        else
1498  #endif  #endif
1499    
1500        /* More streamlined when not in UTF-8 mode */        /* Not in UTF-8 mode */
1501    
1502          {          {
1503          prev_is_word = (eptr != md->start_subject) &&          if (eptr == md->start_subject) prev_is_word = FALSE; else
1504            ((md->ctypes[eptr[-1]] & ctype_word) != 0);            {
1505          cur_is_word = (eptr < md->end_subject) &&            if (eptr <= md->start_used_ptr) md->start_used_ptr = eptr - 1;
1506            ((md->ctypes[*eptr] & ctype_word) != 0);            prev_is_word = ((md->ctypes[eptr[-1]] & ctype_word) != 0);
1507              }
1508            if (eptr >= md->end_subject)
1509              {
1510              SCHECK_PARTIAL();
1511              cur_is_word = FALSE;
1512              }
1513            else cur_is_word = ((md->ctypes[*eptr] & ctype_word) != 0);
1514          }          }
1515    
1516        /* Now see if the situation is what we want */        /* Now see if the situation is what we want */
# Line 1366  for (;;) Line 1524  for (;;)
1524      /* Match a single character type; inline for speed */      /* Match a single character type; inline for speed */
1525    
1526      case OP_ANY:      case OP_ANY:
1527      if ((ims & PCRE_DOTALL) == 0)      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
1528        /* Fall through */
1529    
1530        case OP_ALLANY:
1531        if (eptr++ >= md->end_subject)
1532        {        {
1533        if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);        SCHECK_PARTIAL();
1534        }        RRETURN(MATCH_NOMATCH);
1535      if (eptr++ >= md->end_subject) RRETURN(MATCH_NOMATCH);        }
1536      if (utf8)      if (utf8) while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
       while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  
1537      ecode++;      ecode++;
1538      break;      break;
1539    
# Line 1380  for (;;) Line 1541  for (;;)
1541      any byte, even newline, independent of the setting of PCRE_DOTALL. */      any byte, even newline, independent of the setting of PCRE_DOTALL. */
1542    
1543      case OP_ANYBYTE:      case OP_ANYBYTE:
1544      if (eptr++ >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr++ >= md->end_subject)
1545          {
1546          SCHECK_PARTIAL();
1547          RRETURN(MATCH_NOMATCH);
1548          }
1549      ecode++;      ecode++;
1550      break;      break;
1551    
1552      case OP_NOT_DIGIT:      case OP_NOT_DIGIT:
1553      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1554          {
1555          SCHECK_PARTIAL();
1556          RRETURN(MATCH_NOMATCH);
1557          }
1558      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1559      if (      if (
1560  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1398  for (;;) Line 1567  for (;;)
1567      break;      break;
1568    
1569      case OP_DIGIT:      case OP_DIGIT:
1570      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1571          {
1572          SCHECK_PARTIAL();
1573          RRETURN(MATCH_NOMATCH);
1574          }
1575      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1576      if (      if (
1577  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1411  for (;;) Line 1584  for (;;)
1584      break;      break;
1585    
1586      case OP_NOT_WHITESPACE:      case OP_NOT_WHITESPACE:
1587      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1588          {
1589          SCHECK_PARTIAL();
1590          RRETURN(MATCH_NOMATCH);
1591          }
1592      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1593      if (      if (
1594  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1424  for (;;) Line 1601  for (;;)
1601      break;      break;
1602    
1603      case OP_WHITESPACE:      case OP_WHITESPACE:
1604      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1605          {
1606          SCHECK_PARTIAL();
1607          RRETURN(MATCH_NOMATCH);
1608          }
1609      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1610      if (      if (
1611  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1437  for (;;) Line 1618  for (;;)
1618      break;      break;
1619    
1620      case OP_NOT_WORDCHAR:      case OP_NOT_WORDCHAR:
1621      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1622          {
1623          SCHECK_PARTIAL();
1624          RRETURN(MATCH_NOMATCH);
1625          }
1626      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1627      if (      if (
1628  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1450  for (;;) Line 1635  for (;;)
1635      break;      break;
1636    
1637      case OP_WORDCHAR:      case OP_WORDCHAR:
1638      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1639          {
1640          SCHECK_PARTIAL();
1641          RRETURN(MATCH_NOMATCH);
1642          }
1643      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1644      if (      if (
1645  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1463  for (;;) Line 1652  for (;;)
1652      break;      break;
1653    
1654      case OP_ANYNL:      case OP_ANYNL:
1655      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1656          {
1657          SCHECK_PARTIAL();
1658          RRETURN(MATCH_NOMATCH);
1659          }
1660      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1661      switch(c)      switch(c)
1662        {        {
# Line 1471  for (;;) Line 1664  for (;;)
1664        case 0x000d:        case 0x000d:
1665        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
1666        break;        break;
1667    
1668        case 0x000a:        case 0x000a:
1669          break;
1670    
1671        case 0x000b:        case 0x000b:
1672        case 0x000c:        case 0x000c:
1673        case 0x0085:        case 0x0085:
1674        case 0x2028:        case 0x2028:
1675        case 0x2029:        case 0x2029:
1676          if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
1677        break;        break;
1678        }        }
1679      ecode++;      ecode++;
1680      break;      break;
1681    
1682      case OP_NOT_HSPACE:      case OP_NOT_HSPACE:
1683      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1684          {
1685          SCHECK_PARTIAL();
1686          RRETURN(MATCH_NOMATCH);
1687          }
1688      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1689      switch(c)      switch(c)
1690        {        {
# Line 1513  for (;;) Line 1714  for (;;)
1714      break;      break;
1715    
1716      case OP_HSPACE:      case OP_HSPACE:
1717      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1718          {
1719          SCHECK_PARTIAL();
1720          RRETURN(MATCH_NOMATCH);
1721          }
1722      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1723      switch(c)      switch(c)
1724        {        {
# Line 1543  for (;;) Line 1748  for (;;)
1748      break;      break;
1749    
1750      case OP_NOT_VSPACE:      case OP_NOT_VSPACE:
1751      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1752          {
1753          SCHECK_PARTIAL();
1754          RRETURN(MATCH_NOMATCH);
1755          }
1756      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1757      switch(c)      switch(c)
1758        {        {
# Line 1561  for (;;) Line 1770  for (;;)
1770      break;      break;
1771    
1772      case OP_VSPACE:      case OP_VSPACE:
1773      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1774          {
1775          SCHECK_PARTIAL();
1776          RRETURN(MATCH_NOMATCH);
1777          }
1778      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1779      switch(c)      switch(c)
1780        {        {
# Line 1584  for (;;) Line 1797  for (;;)
1797    
1798      case OP_PROP:      case OP_PROP:
1799      case OP_NOTPROP:      case OP_NOTPROP:
1800      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1801          {
1802          SCHECK_PARTIAL();
1803          RRETURN(MATCH_NOMATCH);
1804          }
1805      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1806        {        {
1807        int chartype, script;        const ucd_record *prop = GET_UCD(c);
       int category = _pcre_ucp_findprop(c, &chartype, &script);  
1808    
1809        switch(ecode[1])        switch(ecode[1])
1810          {          {
# Line 1597  for (;;) Line 1813  for (;;)
1813          break;          break;
1814    
1815          case PT_LAMP:          case PT_LAMP:
1816          if ((chartype == ucp_Lu ||          if ((prop->chartype == ucp_Lu ||
1817               chartype == ucp_Ll ||               prop->chartype == ucp_Ll ||
1818               chartype == ucp_Lt) == (op == OP_NOTPROP))               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))
1819            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1820           break;           break;
1821    
1822          case PT_GC:          case PT_GC:
1823          if ((ecode[2] != category) == (op == OP_PROP))          if ((ecode[2] != _pcre_ucp_gentype[prop->chartype]) == (op == OP_PROP))
1824            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1825          break;          break;
1826    
1827          case PT_PC:          case PT_PC:
1828          if ((ecode[2] != chartype) == (op == OP_PROP))          if ((ecode[2] != prop->chartype) == (op == OP_PROP))
1829            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1830          break;          break;
1831    
1832          case PT_SC:          case PT_SC:
1833          if ((ecode[2] != script) == (op == OP_PROP))          if ((ecode[2] != prop->script) == (op == OP_PROP))
1834            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1835          break;          break;
1836    
# Line 1630  for (;;) Line 1846  for (;;)
1846      is in the binary; otherwise a compile-time error occurs. */      is in the binary; otherwise a compile-time error occurs. */
1847    
1848      case OP_EXTUNI:      case OP_EXTUNI:
1849      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1850          {
1851          SCHECK_PARTIAL();
1852          RRETURN(MATCH_NOMATCH);
1853          }
1854      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1855        {        {
1856        int chartype, script;        int category = UCD_CATEGORY(c);
       int category = _pcre_ucp_findprop(c, &chartype, &script);  
1857        if (category == ucp_M) RRETURN(MATCH_NOMATCH);        if (category == ucp_M) RRETURN(MATCH_NOMATCH);
1858        while (eptr < md->end_subject)        while (eptr < md->end_subject)
1859          {          {
# Line 1643  for (;;) Line 1862  for (;;)
1862            {            {
1863            GETCHARLEN(c, eptr, len);            GETCHARLEN(c, eptr, len);
1864            }            }
1865          category = _pcre_ucp_findprop(c, &chartype, &script);          category = UCD_CATEGORY(c);
1866          if (category != ucp_M) break;          if (category != ucp_M) break;
1867          eptr += len;          eptr += len;
1868          }          }
# Line 1664  for (;;) Line 1883  for (;;)
1883      case OP_REF:      case OP_REF:
1884        {        {
1885        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
1886        ecode += 3;                                 /* Advance past item */        ecode += 3;
1887    
1888          /* If the reference is unset, there are two possibilities:
1889    
1890          (a) In the default, Perl-compatible state, set the length to be longer
1891          than the amount of subject left; this ensures that every attempt at a
1892          match fails. We can't just fail here, because of the possibility of
1893          quantifiers with zero minima.
1894    
1895        /* If the reference is unset, set the length to be longer than the amount        (b) If the JavaScript compatibility flag is set, set the length to zero
1896        of subject left; this ensures that every attempt at a match fails. We        so that the back reference matches an empty string.
1897        can't just fail here, because of the possibility of quantifiers with zero  
1898        minima. */        Otherwise, set the length to the length of what was matched by the
1899          referenced subpattern. */
1900        length = (offset >= offset_top || md->offset_vector[offset] < 0)?  
1901          md->end_subject - eptr + 1 :        if (offset >= offset_top || md->offset_vector[offset] < 0)
1902          md->offset_vector[offset+1] - md->offset_vector[offset];          length = (md->jscript_compat)? 0 : md->end_subject - eptr + 1;
1903          else
1904            length = md->offset_vector[offset+1] - md->offset_vector[offset];
1905    
1906        /* Set up for repetition, or handle the non-repeated case */        /* Set up for repetition, or handle the non-repeated case */
1907    
# Line 1702  for (;;) Line 1930  for (;;)
1930          break;          break;
1931    
1932          default:               /* No repeat follows */          default:               /* No repeat follows */
1933          if (!match_ref(offset, eptr, length, md, ims)) RRETURN(MATCH_NOMATCH);          if (!match_ref(offset, eptr, length, md, ims))
1934              {
1935              CHECK_PARTIAL();
1936              RRETURN(MATCH_NOMATCH);
1937              }
1938          eptr += length;          eptr += length;
1939          continue;              /* With the main loop */          continue;              /* With the main loop */
1940          }          }
1941    
1942        /* If the length of the reference is zero, just continue with the        /* If the length of the reference is zero, just continue with the
1943        main loop. */        main loop. */
1944    
1945        if (length == 0) continue;        if (length == 0) continue;
1946    
1947        /* First, ensure the minimum number of matches are present. We get back        /* First, ensure the minimum number of matches are present. We get back
# Line 1718  for (;;) Line 1950  for (;;)
1950    
1951        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
1952          {          {
1953          if (!match_ref(offset, eptr, length, md, ims)) RRETURN(MATCH_NOMATCH);          if (!match_ref(offset, eptr, length, md, ims))
1954              {
1955              CHECK_PARTIAL();
1956              RRETURN(MATCH_NOMATCH);
1957              }
1958          eptr += length;          eptr += length;
1959          }          }
1960    
# Line 1735  for (;;) Line 1971  for (;;)
1971            {            {
1972            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM14);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM14);
1973            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1974            if (fi >= max || !match_ref(offset, eptr, length, md, ims))            if (fi >= max) RRETURN(MATCH_NOMATCH);
1975              if (!match_ref(offset, eptr, length, md, ims))
1976                {
1977                CHECK_PARTIAL();
1978              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
1979                }
1980            eptr += length;            eptr += length;
1981            }            }
1982          /* Control never gets here */          /* Control never gets here */
# Line 1763  for (;;) Line 2003  for (;;)
2003        }        }
2004      /* Control never gets here */      /* Control never gets here */
2005    
   
   
2006      /* Match a bit-mapped character class, possibly repeatedly. This op code is      /* Match a bit-mapped character class, possibly repeatedly. This op code is
2007      used when all the characters in the class have values in the range 0-255,      used when all the characters in the class have values in the range 0-255,
2008      and either the matching is caseful, or the characters are in the range      and either the matching is caseful, or the characters are in the range
# Line 1819  for (;;) Line 2057  for (;;)
2057          {          {
2058          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2059            {            {
2060            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
2061                {
2062                SCHECK_PARTIAL();
2063                RRETURN(MATCH_NOMATCH);
2064                }
2065            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
2066            if (c > 255)            if (c > 255)
2067              {              {
# Line 1837  for (;;) Line 2079  for (;;)
2079          {          {
2080          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2081            {            {
2082            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
2083                {
2084                SCHECK_PARTIAL();
2085                RRETURN(MATCH_NOMATCH);
2086                }
2087            c = *eptr++;            c = *eptr++;
2088            if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);            if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2089            }            }
# Line 1861  for (;;) Line 2107  for (;;)
2107              {              {
2108              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM16);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM16);
2109              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2110              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2111                if (eptr >= md->end_subject)
2112                  {
2113                  SCHECK_PARTIAL();
2114                  RRETURN(MATCH_NOMATCH);
2115                  }
2116              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
2117              if (c > 255)              if (c > 255)
2118                {                {
# Line 1881  for (;;) Line 2132  for (;;)
2132              {              {
2133              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM17);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM17);
2134              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2135              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2136                if (eptr >= md->end_subject)
2137                  {
2138                  SCHECK_PARTIAL();
2139                  RRETURN(MATCH_NOMATCH);
2140                  }
2141              c = *eptr++;              c = *eptr++;
2142              if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);              if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2143              }              }
# Line 1948  for (;;) Line 2204  for (;;)
2204    
2205    
2206      /* Match an extended character class. This opcode is encountered only      /* Match an extended character class. This opcode is encountered only
2207      in UTF-8 mode, because that's the only time it is compiled. */      when UTF-8 mode mode is supported. Nevertheless, we may not be in UTF-8
2208        mode, because Unicode properties are supported in non-UTF-8 mode. */
2209    
2210  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
2211      case OP_XCLASS:      case OP_XCLASS:
# Line 1989  for (;;) Line 2246  for (;;)
2246    
2247        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
2248          {          {
2249          if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);          if (eptr >= md->end_subject)
2250          GETCHARINC(c, eptr);            {
2251              SCHECK_PARTIAL();
2252              RRETURN(MATCH_NOMATCH);
2253              }
2254            GETCHARINCTEST(c, eptr);
2255          if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);          if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);
2256          }          }
2257    
# Line 2008  for (;;) Line 2269  for (;;)
2269            {            {
2270            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM20);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM20);
2271            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2272            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
2273            GETCHARINC(c, eptr);            if (eptr >= md->end_subject)
2274                {
2275                SCHECK_PARTIAL();
2276                RRETURN(MATCH_NOMATCH);
2277                }
2278              GETCHARINCTEST(c, eptr);
2279            if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);            if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);
2280            }            }
2281          /* Control never gets here */          /* Control never gets here */
# Line 2024  for (;;) Line 2290  for (;;)
2290            {            {
2291            int len = 1;            int len = 1;
2292            if (eptr >= md->end_subject) break;            if (eptr >= md->end_subject) break;
2293            GETCHARLEN(c, eptr, len);            GETCHARLENTEST(c, eptr, len);
2294            if (!_pcre_xclass(c, data)) break;            if (!_pcre_xclass(c, data)) break;
2295            eptr += len;            eptr += len;
2296            }            }
# Line 2033  for (;;) Line 2299  for (;;)
2299            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM21);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM21);
2300            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2301            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
2302            BACKCHAR(eptr)            if (utf8) BACKCHAR(eptr);
2303            }            }
2304          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2305          }          }
# Line 2051  for (;;) Line 2317  for (;;)
2317        length = 1;        length = 1;
2318        ecode++;        ecode++;
2319        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
2320        if (length > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);        if (length > md->end_subject - eptr)
2321            {
2322            CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
2323            RRETURN(MATCH_NOMATCH);
2324            }
2325        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);
2326        }        }
2327      else      else
# Line 2059  for (;;) Line 2329  for (;;)
2329    
2330      /* Non-UTF-8 mode */      /* Non-UTF-8 mode */
2331        {        {
2332        if (md->end_subject - eptr < 1) RRETURN(MATCH_NOMATCH);        if (md->end_subject - eptr < 1)
2333            {
2334            SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */
2335            RRETURN(MATCH_NOMATCH);
2336            }
2337        if (ecode[1] != *eptr++) RRETURN(MATCH_NOMATCH);        if (ecode[1] != *eptr++) RRETURN(MATCH_NOMATCH);
2338        ecode += 2;        ecode += 2;
2339        }        }
# Line 2075  for (;;) Line 2349  for (;;)
2349        ecode++;        ecode++;
2350        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
2351    
2352        if (length > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);        if (length > md->end_subject - eptr)
2353            {
2354            CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
2355            RRETURN(MATCH_NOMATCH);
2356            }
2357    
2358        /* If the pattern character's value is < 128, we have only one byte, and        /* If the pattern character's value is < 128, we have only one byte, and
2359        can use the fast lookup table. */        can use the fast lookup table. */
# Line 2099  for (;;) Line 2377  for (;;)
2377          if (fc != dc)          if (fc != dc)
2378            {            {
2379  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2380            if (dc != _pcre_ucp_othercase(fc))            if (dc != UCD_OTHERCASE(fc))
2381  #endif  #endif
2382              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2383            }            }
# Line 2110  for (;;) Line 2388  for (;;)
2388    
2389      /* Non-UTF-8 mode */      /* Non-UTF-8 mode */
2390        {        {
2391        if (md->end_subject - eptr < 1) RRETURN(MATCH_NOMATCH);        if (md->end_subject - eptr < 1)
2392            {
2393            SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */
2394            RRETURN(MATCH_NOMATCH);
2395            }
2396        if (md->lcc[ecode[1]] != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);        if (md->lcc[ecode[1]] != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);
2397        ecode += 2;        ecode += 2;
2398        }        }
# Line 2164  for (;;) Line 2446  for (;;)
2446      case OP_MINQUERY:      case OP_MINQUERY:
2447      c = *ecode++ - OP_STAR;      c = *ecode++ - OP_STAR;
2448      minimize = (c & 1) != 0;      minimize = (c & 1) != 0;
2449    
2450      min = rep_min[c];                 /* Pick up values from tables; */      min = rep_min[c];                 /* Pick up values from tables; */
2451      max = rep_max[c];                 /* zero for max => infinity */      max = rep_max[c];                 /* zero for max => infinity */
2452      if (max == 0) max = INT_MAX;      if (max == 0) max = INT_MAX;
2453    
2454      /* Common code for all repeated single-character matches. We can give      /* Common code for all repeated single-character matches. */
     up quickly if there are fewer than the minimum number of characters left in  
     the subject. */  
2455    
2456      REPEATCHAR:      REPEATCHAR:
2457  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 2179  for (;;) Line 2460  for (;;)
2460        length = 1;        length = 1;
2461        charptr = ecode;        charptr = ecode;
2462        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
       if (min * length > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);  
2463        ecode += length;        ecode += length;
2464    
2465        /* Handle multibyte character matching specially here. There is        /* Handle multibyte character matching specially here. There is
# Line 2190  for (;;) Line 2470  for (;;)
2470  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2471          unsigned int othercase;          unsigned int othercase;
2472          if ((ims & PCRE_CASELESS) != 0 &&          if ((ims & PCRE_CASELESS) != 0 &&
2473              (othercase = _pcre_ucp_othercase(fc)) != NOTACHAR)              (othercase = UCD_OTHERCASE(fc)) != fc)
2474            oclength = _pcre_ord2utf8(othercase, occhars);            oclength = _pcre_ord2utf8(othercase, occhars);
2475          else oclength = 0;          else oclength = 0;
2476  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2477    
2478          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2479            {            {
2480            if (memcmp(eptr, charptr, length) == 0) eptr += length;            if (eptr <= md->end_subject - length &&
2481                memcmp(eptr, charptr, length) == 0) eptr += length;
2482  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2483            /* Need braces because of following else */            else if (oclength > 0 &&
2484            else if (oclength == 0) { RRETURN(MATCH_NOMATCH); }                     eptr <= md->end_subject - oclength &&
2485                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;
2486    #endif  /* SUPPORT_UCP */
2487            else            else
2488              {              {
2489              if (memcmp(eptr, occhars, oclength) != 0) RRETURN(MATCH_NOMATCH);              CHECK_PARTIAL();
2490              eptr += oclength;              RRETURN(MATCH_NOMATCH);
2491              }              }
 #else   /* without SUPPORT_UCP */  
           else { RRETURN(MATCH_NOMATCH); }  
 #endif  /* SUPPORT_UCP */  
2492            }            }
2493    
2494          if (min == max) continue;          if (min == max) continue;
# Line 2219  for (;;) Line 2499  for (;;)
2499              {              {
2500              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM22);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM22);
2501              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2502              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2503              if (memcmp(eptr, charptr, length) == 0) eptr += length;              if (eptr <= md->end_subject - length &&
2504                  memcmp(eptr, charptr, length) == 0) eptr += length;
2505  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2506              /* Need braces because of following else */              else if (oclength > 0 &&
2507              else if (oclength == 0) { RRETURN(MATCH_NOMATCH); }                       eptr <= md->end_subject - oclength &&
2508                         memcmp(eptr, occhars, oclength) == 0) eptr += oclength;
2509    #endif  /* SUPPORT_UCP */
2510              else              else
2511                {                {
2512                if (memcmp(eptr, occhars, oclength) != 0) RRETURN(MATCH_NOMATCH);                CHECK_PARTIAL();
2513                eptr += oclength;                RRETURN(MATCH_NOMATCH);
2514                }                }
 #else   /* without SUPPORT_UCP */  
             else { RRETURN (MATCH_NOMATCH); }  
 #endif  /* SUPPORT_UCP */  
2515              }              }
2516            /* Control never gets here */            /* Control never gets here */
2517            }            }
# Line 2241  for (;;) Line 2521  for (;;)
2521            pp = eptr;            pp = eptr;
2522            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2523              {              {
2524              if (eptr > md->end_subject - length) break;              if (eptr <= md->end_subject - length &&
2525              if (memcmp(eptr, charptr, length) == 0) eptr += length;                  memcmp(eptr, charptr, length) == 0) eptr += length;
2526  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2527              else if (oclength == 0) break;              else if (oclength > 0 &&
2528              else                       eptr <= md->end_subject - oclength &&
2529                {                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;
               if (memcmp(eptr, occhars, oclength) != 0) break;  
               eptr += oclength;  
               }  
 #else   /* without SUPPORT_UCP */  
             else break;  
2530  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2531                else break;
2532              }              }
2533    
2534            if (possessive) continue;            if (possessive) continue;
2535    
2536            for(;;)            for(;;)
2537             {              {
2538             RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM23);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM23);
2539             if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2540             if (eptr == pp) RRETURN(MATCH_NOMATCH);              if (eptr == pp) { RRETURN(MATCH_NOMATCH); }
2541  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2542             eptr--;              eptr--;
2543             BACKCHAR(eptr);              BACKCHAR(eptr);
2544  #else   /* without SUPPORT_UCP */  #else   /* without SUPPORT_UCP */
2545             eptr -= length;              eptr -= length;
2546  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2547             }              }
2548            }            }
2549          /* Control never gets here */          /* Control never gets here */
2550          }          }
# Line 2280  for (;;) Line 2557  for (;;)
2557  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF8 */
2558    
2559      /* When not in UTF-8 mode, load a single-byte character. */      /* When not in UTF-8 mode, load a single-byte character. */
       {  
       if (min > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);  
       fc = *ecode++;  
       }  
2560    
2561        fc = *ecode++;
2562    
2563      /* The value of fc at this point is always less than 256, though we may or      /* The value of fc at this point is always less than 256, though we may or
2564      may not be in UTF-8 mode. The code is duplicated for the caseless and      may not be in UTF-8 mode. The code is duplicated for the caseless and
2565      caseful cases, for speed, since matching characters is likely to be quite      caseful cases, for speed, since matching characters is likely to be quite
# Line 2301  for (;;) Line 2576  for (;;)
2576        {        {
2577        fc = md->lcc[fc];        fc = md->lcc[fc];
2578        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
2579            {
2580            if (eptr >= md->end_subject)
2581              {
2582              SCHECK_PARTIAL();
2583              RRETURN(MATCH_NOMATCH);
2584              }
2585          if (fc != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);          if (fc != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);
2586            }
2587        if (min == max) continue;        if (min == max) continue;
2588        if (minimize)        if (minimize)
2589          {          {
# Line 2309  for (;;) Line 2591  for (;;)
2591            {            {
2592            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM24);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM24);
2593            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2594            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max) RRETURN(MATCH_NOMATCH);
2595                fc != md->lcc[*eptr++])            if (eptr >= md->end_subject)
2596                {
2597                SCHECK_PARTIAL();
2598              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2599                }
2600              if (fc != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);
2601            }            }
2602          /* Control never gets here */          /* Control never gets here */
2603          }          }
# Line 2323  for (;;) Line 2609  for (;;)
2609            if (eptr >= md->end_subject || fc != md->lcc[*eptr]) break;            if (eptr >= md->end_subject || fc != md->lcc[*eptr]) break;
2610            eptr++;            eptr++;
2611            }            }
2612    
2613          if (possessive) continue;          if (possessive) continue;
2614    
2615          while (eptr >= pp)          while (eptr >= pp)
2616            {            {
2617            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM25);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM25);
# Line 2339  for (;;) Line 2627  for (;;)
2627    
2628      else      else
2629        {        {
2630        for (i = 1; i <= min; i++) if (fc != *eptr++) RRETURN(MATCH_NOMATCH);        for (i = 1; i <= min; i++)
2631            {
2632            if (eptr >= md->end_subject)
2633              {
2634              SCHECK_PARTIAL();
2635              RRETURN(MATCH_NOMATCH);
2636              }
2637            if (fc != *eptr++) RRETURN(MATCH_NOMATCH);
2638            }
2639    
2640        if (min == max) continue;        if (min == max) continue;
2641    
2642        if (minimize)        if (minimize)
2643          {          {
2644          for (fi = min;; fi++)          for (fi = min;; fi++)
2645            {            {
2646            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM26);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM26);
2647            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2648            if (fi >= max || eptr >= md->end_subject || fc != *eptr++)            if (fi >= max) RRETURN(MATCH_NOMATCH);
2649              if (eptr >= md->end_subject)
2650                {
2651                SCHECK_PARTIAL();
2652              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2653                }
2654              if (fc != *eptr++) RRETURN(MATCH_NOMATCH);
2655            }            }
2656          /* Control never gets here */          /* Control never gets here */
2657          }          }
# Line 2361  for (;;) Line 2664  for (;;)
2664            eptr++;            eptr++;
2665            }            }
2666          if (possessive) continue;          if (possessive) continue;
2667    
2668          while (eptr >= pp)          while (eptr >= pp)
2669            {            {
2670            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM27);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM27);
# Line 2376  for (;;) Line 2680  for (;;)
2680      checking can be multibyte. */      checking can be multibyte. */
2681    
2682      case OP_NOT:      case OP_NOT:
2683      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
2684          {
2685          SCHECK_PARTIAL();
2686          RRETURN(MATCH_NOMATCH);
2687          }
2688      ecode++;      ecode++;
2689      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2690      if ((ims & PCRE_CASELESS) != 0)      if ((ims & PCRE_CASELESS) != 0)
# Line 2453  for (;;) Line 2761  for (;;)
2761      max = rep_max[c];                 /* zero for max => infinity */      max = rep_max[c];                 /* zero for max => infinity */
2762      if (max == 0) max = INT_MAX;      if (max == 0) max = INT_MAX;
2763    
2764      /* Common code for all repeated single-byte matches. We can give up quickly      /* Common code for all repeated single-byte matches. */
     if there are fewer than the minimum number of bytes left in the  
     subject. */  
2765    
2766      REPEATNOTCHAR:      REPEATNOTCHAR:
     if (min > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);  
2767      fc = *ecode++;      fc = *ecode++;
2768    
2769      /* The code is duplicated for the caseless and caseful cases, for speed,      /* The code is duplicated for the caseless and caseful cases, for speed,
# Line 2483  for (;;) Line 2788  for (;;)
2788          register unsigned int d;          register unsigned int d;
2789          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2790            {            {
2791              if (eptr >= md->end_subject)
2792                {
2793                SCHECK_PARTIAL();
2794                RRETURN(MATCH_NOMATCH);
2795                }
2796            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
2797            if (d < 256) d = md->lcc[d];            if (d < 256) d = md->lcc[d];
2798            if (fc == d) RRETURN(MATCH_NOMATCH);            if (fc == d) RRETURN(MATCH_NOMATCH);
# Line 2494  for (;;) Line 2804  for (;;)
2804        /* Not UTF-8 mode */        /* Not UTF-8 mode */
2805          {          {
2806          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2807              {
2808              if (eptr >= md->end_subject)
2809                {
2810                SCHECK_PARTIAL();
2811                RRETURN(MATCH_NOMATCH);
2812                }
2813            if (fc == md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);            if (fc == md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);
2814              }
2815          }          }
2816    
2817        if (min == max) continue;        if (min == max) continue;
# Line 2510  for (;;) Line 2827  for (;;)
2827              {              {
2828              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM28);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM28);
2829              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2830                if (fi >= max) RRETURN(MATCH_NOMATCH);
2831                if (eptr >= md->end_subject)
2832                  {
2833                  SCHECK_PARTIAL();
2834                  RRETURN(MATCH_NOMATCH);
2835                  }
2836              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
2837              if (d < 256) d = md->lcc[d];              if (d < 256) d = md->lcc[d];
2838              if (fi >= max || eptr >= md->end_subject || fc == d)              if (fc == d) RRETURN(MATCH_NOMATCH);
               RRETURN(MATCH_NOMATCH);  
2839              }              }
2840            }            }
2841          else          else
# Line 2524  for (;;) Line 2846  for (;;)
2846              {              {
2847              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM29);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM29);
2848              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2849              if (fi >= max || eptr >= md->end_subject || fc == md->lcc[*eptr++])              if (fi >= max) RRETURN(MATCH_NOMATCH);
2850                if (eptr >= md->end_subject)
2851                  {
2852                  SCHECK_PARTIAL();
2853                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2854                  }
2855                if (fc == md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);
2856              }              }
2857            }            }
2858          /* Control never gets here */          /* Control never gets here */
# Line 2594  for (;;) Line 2921  for (;;)
2921          register unsigned int d;          register unsigned int d;
2922          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2923            {            {
2924              if (eptr >= md->end_subject)
2925                {
2926                SCHECK_PARTIAL();
2927                RRETURN(MATCH_NOMATCH);
2928                }
2929            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
2930            if (fc == d) RRETURN(MATCH_NOMATCH);            if (fc == d) RRETURN(MATCH_NOMATCH);
2931            }            }
# Line 2603  for (;;) Line 2935  for (;;)
2935        /* Not UTF-8 mode */        /* Not UTF-8 mode */
2936          {          {
2937          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2938              {
2939              if (eptr >= md->end_subject)
2940                {
2941                SCHECK_PARTIAL();
2942                RRETURN(MATCH_NOMATCH);
2943                }
2944            if (fc == *eptr++) RRETURN(MATCH_NOMATCH);            if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
2945              }
2946          }          }
2947    
2948        if (min == max) continue;        if (min == max) continue;
# Line 2619  for (;;) Line 2958  for (;;)
2958              {              {
2959              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM32);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM32);
2960              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2961              GETCHARINC(d, eptr);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2962              if (fi >= max || eptr >= md->end_subject || fc == d)              if (eptr >= md->end_subject)
2963                  {
2964                  SCHECK_PARTIAL();
2965                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2966                  }
2967                GETCHARINC(d, eptr);
2968                if (fc == d) RRETURN(MATCH_NOMATCH);
2969              }              }
2970            }            }
2971          else          else
# Line 2632  for (;;) Line 2976  for (;;)
2976              {              {
2977              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM33);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM33);
2978              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2979              if (fi >= max || eptr >= md->end_subject || fc == *eptr++)              if (fi >= max) RRETURN(MATCH_NOMATCH);
2980                if (eptr >= md->end_subject)
2981                  {
2982                  SCHECK_PARTIAL();
2983                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2984                  }
2985                if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
2986              }              }
2987            }            }
2988          /* Control never gets here */          /* Control never gets here */
# Line 2767  for (;;) Line 3116  for (;;)
3116    
3117      /* First, ensure the minimum number of matches are present. Use inline      /* First, ensure the minimum number of matches are present. Use inline
3118      code for maximizing the speed, and do the type test once at the start      code for maximizing the speed, and do the type test once at the start
3119      (i.e. keep it out of the loop). Also we can test that there are at least      (i.e. keep it out of the loop). Separate the UTF-8 code completely as that
     the minimum number of bytes before we start. This isn't as effective in  
     UTF-8 mode, but it does no harm. Separate the UTF-8 code completely as that  
3120      is tidier. Also separate the UCP code, which can be the same for both UTF-8      is tidier. Also separate the UCP code, which can be the same for both UTF-8
3121      and single-bytes. */      and single-bytes. */
3122    
     if (min > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);  
3123      if (min > 0)      if (min > 0)
3124        {        {
3125  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
# Line 2785  for (;;) Line 3131  for (;;)
3131            if (prop_fail_result) RRETURN(MATCH_NOMATCH);            if (prop_fail_result) RRETURN(MATCH_NOMATCH);
3132            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
3133              {              {
3134              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject)
3135                  {
3136                  SCHECK_PARTIAL();
3137                  RRETURN(MATCH_NOMATCH);
3138                  }
3139              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
3140              }              }
3141            break;            break;
# Line 2793  for (;;) Line 3143  for (;;)
3143            case PT_LAMP:            case PT_LAMP:
3144            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
3145              {              {
3146              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject)
3147                  {
3148                  SCHECK_PARTIAL();
3149                  RRETURN(MATCH_NOMATCH);
3150                  }
3151              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
3152              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3153              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
3154                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
3155                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
# Line 2806  for (;;) Line 3160  for (;;)
3160            case PT_GC:            case PT_GC:
3161            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
3162              {              {
3163              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject)
3164                  {
3165                  SCHECK_PARTIAL();
3166                  RRETURN(MATCH_NOMATCH);
3167                  }
3168              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
3169              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3170              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
3171                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3172              }              }
# Line 2817  for (;;) Line 3175  for (;;)
3175            case PT_PC:            case PT_PC:
3176            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
3177              {              {
3178              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject)
3179                  {
3180                  SCHECK_PARTIAL();
3181                  RRETURN(MATCH_NOMATCH);
3182                  }
3183              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
3184              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3185              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
3186                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3187              }              }
# Line 2828  for (;;) Line 3190  for (;;)
3190            case PT_SC:            case PT_SC:
3191            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
3192              {              {
3193              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject)
3194                  {
3195                  SCHECK_PARTIAL();
3196                  RRETURN(MATCH_NOMATCH);
3197                  }
3198              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
3199              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_script = UCD_SCRIPT(c);
3200              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
3201                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3202              }              }
# Line 2848  for (;;) Line 3214  for (;;)
3214          {          {
3215          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3216            {            {
3217              if (eptr >= md->end_subject)
3218                {
3219                SCHECK_PARTIAL();
3220                RRETURN(MATCH_NOMATCH);
3221                }
3222            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3223            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
3224            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);
3225            while (eptr < md->end_subject)            while (eptr < md->end_subject)
3226              {              {
3227              int len = 1;              int len = 1;
3228              if (!utf8) c = *eptr; else              if (!utf8) c = *eptr;
3229                {                else { GETCHARLEN(c, eptr, len); }
3230                GETCHARLEN(c, eptr, len);              prop_category = UCD_CATEGORY(c);
               }  
             prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);  
3231              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
3232              eptr += len;              eptr += len;
3233              }              }
# Line 2876  for (;;) Line 3245  for (;;)
3245          case OP_ANY:          case OP_ANY:
3246          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3247            {            {
3248            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject)
3249                 ((ims & PCRE_DOTALL) == 0 && IS_NEWLINE(eptr)))              {
3250                SCHECK_PARTIAL();
3251                RRETURN(MATCH_NOMATCH);
3252                }
3253              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
3254              eptr++;
3255              while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
3256              }
3257            break;
3258    
3259            case OP_ALLANY:
3260            for (i = 1; i <= min; i++)
3261              {
3262              if (eptr >= md->end_subject)
3263                {
3264                SCHECK_PARTIAL();
3265              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3266                }
3267            eptr++;            eptr++;
3268            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
3269            }            }
3270          break;          break;
3271    
3272          case OP_ANYBYTE:          case OP_ANYBYTE:
3273            if (eptr > md->end_subject - min) RRETURN(MATCH_NOMATCH);
3274          eptr += min;          eptr += min;
3275          break;          break;
3276    
3277          case OP_ANYNL:          case OP_ANYNL:
3278          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3279            {            {
3280            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3281                {
3282                SCHECK_PARTIAL();
3283                RRETURN(MATCH_NOMATCH);
3284                }
3285            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
3286            switch(c)            switch(c)
3287              {              {
# Line 2899  for (;;) Line 3289  for (;;)
3289              case 0x000d:              case 0x000d:
3290              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3291              break;              break;
3292    
3293              case 0x000a:              case 0x000a:
3294                break;
3295    
3296              case 0x000b:              case 0x000b:
3297              case 0x000c:              case 0x000c:
3298              case 0x0085:              case 0x0085:
3299              case 0x2028:              case 0x2028:
3300              case 0x2029:              case 0x2029:
3301                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3302              break;              break;
3303              }              }
3304            }            }
# Line 2913  for (;;) Line 3307  for (;;)
3307          case OP_NOT_HSPACE:          case OP_NOT_HSPACE:
3308          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3309            {            {
3310            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3311                {
3312                SCHECK_PARTIAL();
3313                RRETURN(MATCH_NOMATCH);
3314                }
3315            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
3316            switch(c)            switch(c)
3317              {              {
# Line 2945  for (;;) Line 3343  for (;;)
3343          case OP_HSPACE:          case OP_HSPACE:
3344          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3345            {            {
3346            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3347                {
3348                SCHECK_PARTIAL();
3349                RRETURN(MATCH_NOMATCH);
3350                }
3351            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
3352            switch(c)            switch(c)
3353              {              {
# Line 2977  for (;;) Line 3379  for (;;)
3379          case OP_NOT_VSPACE:          case OP_NOT_VSPACE:
3380          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3381            {            {
3382            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3383                {
3384                SCHECK_PARTIAL();
3385                RRETURN(MATCH_NOMATCH);
3386                }
3387            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
3388            switch(c)            switch(c)
3389              {              {
# Line 2997  for (;;) Line 3403  for (;;)
3403          case OP_VSPACE:          case OP_VSPACE:
3404          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3405            {            {
3406            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3407                {
3408                SCHECK_PARTIAL();
3409                RRETURN(MATCH_NOMATCH);
3410                }
3411            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
3412            switch(c)            switch(c)
3413              {              {
# Line 3017  for (;;) Line 3427  for (;;)
3427          case OP_NOT_DIGIT:          case OP_NOT_DIGIT:
3428          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3429            {            {
3430            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3431                {
3432                SCHECK_PARTIAL();
3433                RRETURN(MATCH_NOMATCH);
3434                }
3435            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
3436            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)
3437              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
# Line 3027  for (;;) Line 3441  for (;;)
3441          case OP_DIGIT:          case OP_DIGIT:
3442          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3443            {            {
3444            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject)
3445               *eptr >= 128 || (md->ctypes[*eptr++] & ctype_digit) == 0)              {
3446                SCHECK_PARTIAL();
3447                RRETURN(MATCH_NOMATCH);
3448                }
3449              if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_digit) == 0)
3450              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3451            /* No need to skip more bytes - we know it's a 1-byte character */            /* No need to skip more bytes - we know it's a 1-byte character */
3452            }            }
# Line 3037  for (;;) Line 3455  for (;;)
3455          case OP_NOT_WHITESPACE:          case OP_NOT_WHITESPACE:
3456          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3457            {            {
3458            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject)
3459               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_space) != 0))              {
3460                SCHECK_PARTIAL();
3461              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3462            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;              }
3463              if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)
3464                RRETURN(MATCH_NOMATCH);
3465              while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3466            }            }
3467          break;          break;
3468    
3469          case OP_WHITESPACE:          case OP_WHITESPACE:
3470          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3471            {            {
3472            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject)
3473               *eptr >= 128 || (md->ctypes[*eptr++] & ctype_space) == 0)              {
3474                SCHECK_PARTIAL();
3475                RRETURN(MATCH_NOMATCH);
3476                }
3477              if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_space) == 0)
3478              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3479            /* No need to skip more bytes - we know it's a 1-byte character */            /* No need to skip more bytes - we know it's a 1-byte character */
3480            }            }
# Line 3058  for (;;) Line 3484  for (;;)
3484          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3485            {            {
3486            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject ||
3487               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_word) != 0))               (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0))
3488              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3489            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3490            }            }
3491          break;          break;
3492    
3493          case OP_WORDCHAR:          case OP_WORDCHAR:
3494          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3495            {            {
3496            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject)
3497               *eptr >= 128 || (md->ctypes[*eptr++] & ctype_word) == 0)              {
3498                SCHECK_PARTIAL();
3499                RRETURN(MATCH_NOMATCH);
3500                }
3501              if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_word) == 0)
3502              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3503            /* No need to skip more bytes - we know it's a 1-byte character */            /* No need to skip more bytes - we know it's a 1-byte character */
3504            }            }
# Line 3082  for (;;) Line 3512  for (;;)
3512  #endif     /* SUPPORT_UTF8 */  #endif     /* SUPPORT_UTF8 */
3513    
3514        /* Code for the non-UTF-8 case for minimum matching of operators other        /* Code for the non-UTF-8 case for minimum matching of operators other
3515        than OP_PROP and OP_NOTPROP. We can assume that there are the minimum        than OP_PROP and OP_NOTPROP. */
       number of bytes present, as this was tested above. */  
3516    
3517        switch(ctype)        switch(ctype)
3518          {          {
3519          case OP_ANY:          case OP_ANY:
3520          if ((ims & PCRE_DOTALL) == 0)          for (i = 1; i <= min; i++)
3521            {            {
3522            for (i = 1; i <= min; i++)            if (eptr >= md->end_subject)
3523              {              {
3524              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);              SCHECK_PARTIAL();
3525              eptr++;              RRETURN(MATCH_NOMATCH);
3526              }              }
3527              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
3528              eptr++;
3529            }            }
         else eptr += min;  
3530          break;          break;
3531    
3532          case OP_ANYBYTE:          case OP_ALLANY:
3533            if (eptr > md->end_subject - min)
3534              {
3535              SCHECK_PARTIAL();
3536              RRETURN(MATCH_NOMATCH);
3537              }
3538          eptr += min;          eptr += min;
3539          break;          break;
3540    
3541          /* Because of the CRLF case, we can't assume the minimum number of          case OP_ANYBYTE:
3542          bytes are present in this case. */          if (eptr > md->end_subject - min)
3543              {
3544              SCHECK_PARTIAL();
3545              RRETURN(MATCH_NOMATCH);
3546              }
3547            eptr += min;
3548            break;
3549    
3550          case OP_ANYNL:          case OP_ANYNL:
3551          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3552            {            {
3553            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3554                {
3555                SCHECK_PARTIAL();
3556                RRETURN(MATCH_NOMATCH);
3557                }
3558            switch(*eptr++)            switch(*eptr++)
3559              {              {
3560              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
# Line 3117  for (;;) Line 3562  for (;;)
3562              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3563              break;              break;
3564              case 0x000a:              case 0x000a:
3565                break;
3566    
3567              case 0x000b:              case 0x000b:
3568              case 0x000c:              case 0x000c:
3569              case 0x0085:              case 0x0085:
3570                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3571              break;              break;
3572              }              }
3573            }            }
# Line 3128  for (;;) Line 3576  for (;;)
3576          case OP_NOT_HSPACE:          case OP_NOT_HSPACE:
3577          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3578            {            {
3579            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3580                {
3581                SCHECK_PARTIAL();
3582                RRETURN(MATCH_NOMATCH);
3583                }
3584            switch(*eptr++)            switch(*eptr++)
3585              {              {
3586              default: break;              default: break;
# Line 3143  for (;;) Line 3595  for (;;)
3595          case OP_HSPACE:          case OP_HSPACE:
3596          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3597            {            {
3598            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3599                {
3600                SCHECK_PARTIAL();
3601                RRETURN(MATCH_NOMATCH);
3602                }
3603            switch(*eptr++)            switch(*eptr++)
3604              {              {
3605              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
# Line 3158  for (;;) Line 3614  for (;;)
3614          case OP_NOT_VSPACE:          case OP_NOT_VSPACE:
3615          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3616            {            {
3617            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3618                {
3619                SCHECK_PARTIAL();
3620                RRETURN(MATCH_NOMATCH);
3621                }
3622            switch(*eptr++)            switch(*eptr++)
3623              {              {
3624              default: break;              default: break;
# Line 3175  for (;;) Line 3635  for (;;)
3635          case OP_VSPACE:          case OP_VSPACE:
3636          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3637            {            {
3638            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3639                {
3640                SCHECK_PARTIAL();
3641                RRETURN(MATCH_NOMATCH);
3642                }
3643            switch(*eptr++)            switch(*eptr++)
3644              {              {
3645              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
# Line 3191  for (;;) Line 3655  for (;;)
3655    
3656          case OP_NOT_DIGIT:          case OP_NOT_DIGIT:
3657          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3658              {
3659              if (eptr >= md->end_subject)
3660                {
3661                SCHECK_PARTIAL();
3662                RRETURN(MATCH_NOMATCH);
3663                }
3664            if ((md->ctypes[*eptr++] & ctype_digit) != 0) RRETURN(MATCH_NOMATCH);            if ((md->ctypes[*eptr++] & ctype_digit) != 0) RRETURN(MATCH_NOMATCH);
3665              }
3666          break;          break;
3667    
3668          case OP_DIGIT:          case OP_DIGIT:
3669          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3670              {
3671              if (eptr >= md->end_subject)
3672                {
3673                SCHECK_PARTIAL();
3674                RRETURN(MATCH_NOMATCH);
3675                }
3676            if ((md->ctypes[*eptr++] & ctype_digit) == 0) RRETURN(MATCH_NOMATCH);            if ((md->ctypes[*eptr++] & ctype_digit) == 0) RRETURN(MATCH_NOMATCH);
3677              }
3678          break;          break;
3679    
3680          case OP_NOT_WHITESPACE:          case OP_NOT_WHITESPACE:
3681          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3682              {
3683              if (eptr >= md->end_subject)
3684                {
3685                SCHECK_PARTIAL();
3686                RRETURN(MATCH_NOMATCH);
3687                }
3688            if ((md->ctypes[*eptr++] & ctype_space) != 0) RRETURN(MATCH_NOMATCH);            if ((md->ctypes[*eptr++] & ctype_space) != 0) RRETURN(MATCH_NOMATCH);
3689              }
3690          break;          break;
3691    
3692          case OP_WHITESPACE:          case OP_WHITESPACE:
3693          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3694              {
3695              if (eptr >= md->end_subject)
3696                {
3697                SCHECK_PARTIAL();
3698                RRETURN(MATCH_NOMATCH);
3699                }
3700            if ((md->ctypes[*eptr++] & ctype_space) == 0) RRETURN(MATCH_NOMATCH);            if ((md->ctypes[*eptr++] & ctype_space) == 0) RRETURN(MATCH_NOMATCH);
3701              }
3702          break;          break;
3703    
3704          case OP_NOT_WORDCHAR:          case OP_NOT_WORDCHAR:
3705          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3706              {
3707              if (eptr >= md->end_subject)
3708                {
3709                SCHECK_PARTIAL();
3710                RRETURN(MATCH_NOMATCH);
3711                }
3712            if ((md->ctypes[*eptr++] & ctype_word) != 0)            if ((md->ctypes[*eptr++] & ctype_word) != 0)
3713              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3714              }
3715          break;          break;
3716    
3717          case OP_WORDCHAR:          case OP_WORDCHAR:
3718          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3719              {
3720              if (eptr >= md->end_subject)
3721                {
3722                SCHECK_PARTIAL();
3723                RRETURN(MATCH_NOMATCH);
3724                }
3725            if ((md->ctypes[*eptr++] & ctype_word) == 0)            if ((md->ctypes[*eptr++] & ctype_word) == 0)
3726              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3727              }
3728          break;          break;
3729    
3730          default:          default:
# Line 3246  for (;;) Line 3752  for (;;)
3752              {              {
3753              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM36);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM36);
3754              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3755              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3756                if (eptr >= md->end_subject)
3757                  {
3758                  SCHECK_PARTIAL();
3759                  RRETURN(MATCH_NOMATCH);
3760                  }
3761              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3762              if (prop_fail_result) RRETURN(MATCH_NOMATCH);              if (prop_fail_result) RRETURN(MATCH_NOMATCH);
3763              }              }
# Line 3257  for (;;) Line 3768  for (;;)
3768              {              {
3769              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM37);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM37);
3770              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3771              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3772                if (eptr >= md->end_subject)
3773                  {
3774                  SCHECK_PARTIAL();
3775                  RRETURN(MATCH_NOMATCH);
3776                  }
3777              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3778              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3779              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
3780                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
3781                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
# Line 3272  for (;;) Line 3788  for (;;)
3788              {              {
3789              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM38);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM38);
3790              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3791              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3792                if (eptr >= md->end_subject)
3793                  {
3794                  SCHECK_PARTIAL();
3795                  RRETURN(MATCH_NOMATCH);
3796                  }
3797              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3798              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3799              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
3800                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3801              }              }
# Line 3285  for (;;) Line 3806  for (;;)
3806              {              {
3807              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM39);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM39);
3808              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3809              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3810                if (eptr >= md->end_subject)
3811                  {
3812                  SCHECK_PARTIAL();
3813                  RRETURN(MATCH_NOMATCH);
3814                  }
3815              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3816              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3817              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
3818                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3819              }              }
# Line 3298  for (;;) Line 3824  for (;;)
3824              {              {
3825              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM40);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM40);
3826              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3827              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3828                if (eptr >= md->end_subject)
3829                  {
3830                  SCHECK_PARTIAL();
3831                  RRETURN(MATCH_NOMATCH);
3832                  }
3833              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3834              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_script = UCD_SCRIPT(c);
3835              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
3836                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3837              }              }
# Line 3320  for (;;) Line 3851  for (;;)
3851            {            {
3852            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM41);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM41);
3853            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3854            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3855              if (eptr >= md->end_subject)
3856                {
3857                SCHECK_PARTIAL();
3858                RRETURN(MATCH_NOMATCH);
3859                }
3860            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3861            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
3862            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);
3863            while (eptr < md->end_subject)            while (eptr < md->end_subject)
3864              {              {
3865              int len = 1;              int len = 1;
3866              if (!utf8) c = *eptr; else              if (!utf8) c = *eptr;
3867                {                else { GETCHARLEN(c, eptr, len); }
3868                GETCHARLEN(c, eptr, len);              prop_category = UCD_CATEGORY(c);
               }  
             prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);  
3869              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
3870              eptr += len;              eptr += len;
3871              }              }
# Line 3349  for (;;) Line 3883  for (;;)
3883            {            {
3884            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM42);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM42);
3885            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3886            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max) RRETURN(MATCH_NOMATCH);
3887                 (ctype == OP_ANY && (ims & PCRE_DOTALL) == 0 &&            if (eptr >= md->end_subject)
3888                  IS_NEWLINE(eptr)))              {
3889                SCHECK_PARTIAL();
3890                RRETURN(MATCH_NOMATCH);
3891                }
3892              if (ctype == OP_ANY && IS_NEWLINE(eptr))
3893              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
   
3894            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
3895            switch(ctype)            switch(ctype)
3896              {              {
3897              case OP_ANY:        /* This is the DOTALL case */              case OP_ANY:        /* This is the non-NL case */
3898              break;              case OP_ALLANY:
   
3899              case OP_ANYBYTE:              case OP_ANYBYTE:
3900              break;              break;
3901    
# Line 3371  for (;;) Line 3907  for (;;)
3907                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3908                break;                break;
3909                case 0x000a:                case 0x000a:
3910                  break;
3911    
3912                case 0x000b:                case 0x000b:
3913                case 0x000c:                case 0x000c:
3914                case 0x0085:                case 0x0085:
3915                case 0x2028:                case 0x2028:
3916                case 0x2029:                case 0x2029:
3917                  if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3918                break;                break;
3919                }                }
3920              break;              break;
# Line 3507  for (;;) Line 4046  for (;;)
4046            {            {
4047            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM43);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM43);
4048            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4049            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max) RRETURN(MATCH_NOMATCH);
4050                 ((ims & PCRE_DOTALL) == 0 && IS_NEWLINE(eptr)))            if (eptr >= md->end_subject)
4051                {
4052                SCHECK_PARTIAL();
4053                RRETURN(MATCH_NOMATCH);
4054                }
4055              if (ctype == OP_ANY && IS_NEWLINE(eptr))
4056              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
   
4057            c = *eptr++;            c = *eptr++;
4058            switch(ctype)            switch(ctype)
4059              {              {
4060              case OP_ANY:   /* This is the DOTALL case */              case OP_ANY:     /* This is the non-NL case */
4061              break;              case OP_ALLANY:
   
4062              case OP_ANYBYTE:              case OP_ANYBYTE:
4063              break;              break;
4064    
# Line 3527  for (;;) Line 4069  for (;;)
4069                case 0x000d:                case 0x000d:
4070                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
4071                break;                break;
4072    
4073                case 0x000a:                case 0x000a:
4074                  break;
4075    
4076                case 0x000b:                case 0x000b:
4077                case 0x000c:                case 0x000c:
4078                case 0x0085:                case 0x0085:
4079                  if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4080                break;                break;
4081                }                }
4082              break;              break;
# Line 3645  for (;;) Line 4191  for (;;)
4191              int len = 1;              int len = 1;
4192              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
4193              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
4194              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
4195              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
4196                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
4197                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
# Line 3660  for (;;) Line 4206  for (;;)
4206              int len = 1;              int len = 1;
4207              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
4208              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
4209              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
4210              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
4211                break;                break;
4212              eptr+= len;              eptr+= len;
# Line 3673  for (;;) Line 4219  for (;;)
4219              int len = 1;              int len = 1;
4220              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
4221              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
4222              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
4223              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
4224                break;                break;
4225              eptr+= len;              eptr+= len;
# Line 3686  for (;;) Line 4232  for (;;)
4232              int len = 1;              int len = 1;
4233              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
4234              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
4235              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_script = UCD_SCRIPT(c);
4236              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
4237                break;                break;
4238              eptr+= len;              eptr+= len;
# Line 3702  for (;;) Line 4248  for (;;)
4248            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM44);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM44);
4249            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4250            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
4251            BACKCHAR(eptr);            if (utf8) BACKCHAR(eptr);
4252            }            }
4253          }          }
4254    
# Line 3715  for (;;) Line 4261  for (;;)
4261            {            {
4262            if (eptr >= md->end_subject) break;            if (eptr >= md->end_subject) break;
4263            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
4264            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
4265            if (prop_category == ucp_M) break;            if (prop_category == ucp_M) break;
4266            while (eptr < md->end_subject)            while (eptr < md->end_subject)
4267              {              {
# Line 3724  for (;;) Line 4270  for (;;)
4270                {                {
4271                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
4272                }                }
4273              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
4274              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
4275              eptr += len;              eptr += len;
4276              }              }
# Line 3741  for (;;) Line 4287  for (;;)
4287            for (;;)                        /* Move back over one extended */            for (;;)                        /* Move back over one extended */
4288              {              {
4289              int len = 1;              int len = 1;
             BACKCHAR(eptr);  
4290              if (!utf8) c = *eptr; else              if (!utf8) c = *eptr; else
4291                {                {
4292                  BACKCHAR(eptr);
4293                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
4294                }                }
4295              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
4296              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
4297              eptr--;              eptr--;
4298              }              }
# Line 3764  for (;;) Line 4310  for (;;)
4310          switch(ctype)          switch(ctype)
4311            {            {
4312            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. */  
   
4313            if (max < INT_MAX)            if (max < INT_MAX)
4314              {              {
4315              if ((ims & PCRE_DOTALL) == 0)              for (i = min; i < max; i++)
               {  
               for (i = min; i < max; i++)  
                 {  
                 if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;  
                 eptr++;  
                 while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  
                 }  
               }  
             else  
4316                {                {
4317                for (i = min; i < max; i++)                if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
4318                  {                eptr++;
4319                  if (eptr >= md->end_subject) break;                while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
                 eptr++;  
                 while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  
                 }  
4320                }                }
4321              }              }
4322    
# Line 3795  for (;;) Line 4324  for (;;)
4324    
4325            else            else
4326              {              {
4327              if ((ims & PCRE_DOTALL) == 0)              for (i = min; i < max; i++)
4328                {                {
4329                for (i = min; i < max; i++)                if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
4330                  {                eptr++;
4331                  if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;                while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
                 eptr++;  
                 }  
               break;  
4332                }                }
4333              else              }
4334              break;
4335    
4336              case OP_ALLANY:
4337              if (max < INT_MAX)
4338                {
4339                for (i = min; i < max; i++)
4340                {                {
4341                c = max - min;                if (eptr >= md->end_subject) break;
4342                if (c > (unsigned int)(md->end_subject - eptr))                eptr++;
4343                  c = md->end_subject - eptr;                while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
               eptr += c;  
4344                }                }
4345              }              }
4346              else eptr = md->end_subject;   /* Unlimited UTF-8 repeat */
4347            break;            break;
4348    
4349            /* The byte case is the same as non-UTF8 */            /* The byte case is the same as non-UTF8 */
# Line 3836  for (;;) Line 4368  for (;;)
4368                }                }
4369              else              else
4370                {                {
4371                if (c != 0x000a && c != 0x000b && c != 0x000c &&                if (c != 0x000a &&
4372                    c != 0x0085 && c != 0x2028 && c != 0x2029)                    (md->bsr_anycrlf ||
4373                       (c != 0x000b && c != 0x000c &&
4374                        c != 0x0085 && c != 0x2028 && c != 0x2029)))
4375                  break;                  break;
4376                eptr += len;                eptr += len;
4377                }                }
# Line 3990  for (;;) Line 4524  for (;;)
4524            }            }
4525          }          }
4526        else        else
4527  #endif  #endif  /* SUPPORT_UTF8 */
4528    
4529        /* Not UTF-8 mode */        /* Not UTF-8 mode */
4530          {          {
4531          switch(ctype)          switch(ctype)
4532            {            {
4533            case OP_ANY:            case OP_ANY:
4534            if ((ims & PCRE_DOTALL) == 0)            for (i = min; i < max; i++)
4535              {              {
4536              for (i = min; i < max; i++)              if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
4537                {              eptr++;
               if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;  
               eptr++;  
               }  
             break;  
4538              }              }
4539            /* For DOTALL case, fall through and treat as \C */            break;
4540    
4541              case OP_ALLANY:
4542            case OP_ANYBYTE:            case OP_ANYBYTE:
4543            c = max - min;            c = max - min;
4544            if (c > (unsigned int)(md->end_subject - eptr))            if (c > (unsigned int)(md->end_subject - eptr))
# Line 4027  for (;;) Line 4558  for (;;)
4558                }                }
4559              else              else
4560                {                {
4561                if (c != 0x000a && c != 0x000b && c != 0x000c && c != 0x0085)                if (c != 0x000a &&
4562                      (md->bsr_anycrlf ||
4563                        (c != 0x000b && c != 0x000c && c != 0x0085)))
4564                  break;                  break;
4565                eptr++;                eptr++;
4566                }                }
# Line 4177  HEAP_RETURN: Line 4710  HEAP_RETURN:
4710  switch (frame->Xwhere)  switch (frame->Xwhere)
4711    {    {
4712    LBL( 1) LBL( 2) LBL( 3) LBL( 4) LBL( 5) LBL( 6) LBL( 7) LBL( 8)    LBL( 1) LBL( 2) LBL( 3) LBL( 4) LBL( 5) LBL( 6) LBL( 7) LBL( 8)
4713    LBL( 9) LBL(10) LBL(11) LBL(12) LBL(13) LBL(14) LBL(15) LBL(16)    LBL( 9) LBL(10) LBL(11) LBL(12) LBL(13) LBL(14) LBL(15) LBL(17)
4714    LBL(17) LBL(18) LBL(19) LBL(20) LBL(21) LBL(22) LBL(23) LBL(24)    LBL(19) LBL(24) LBL(25) LBL(26) LBL(27) LBL(29) LBL(31) LBL(33)
4715    LBL(25) LBL(26) LBL(27) LBL(28) LBL(29) LBL(30) LBL(31) LBL(32)    LBL(35) LBL(43) LBL(47) LBL(48) LBL(49) LBL(50) LBL(51) LBL(52)
4716    LBL(33) LBL(34) LBL(35) LBL(36) LBL(37) LBL(38) LBL(39) LBL(40)    LBL(53) LBL(54)
4717    LBL(41) LBL(42) LBL(43) LBL(44) LBL(45) LBL(46) LBL(47)  #ifdef SUPPORT_UTF8
4718      LBL(16) LBL(18) LBL(20) LBL(21) LBL(22) LBL(23) LBL(28) LBL(30)
4719      LBL(32) LBL(34) LBL(42) LBL(46)
4720    #ifdef SUPPORT_UCP
4721      LBL(36) LBL(37) LBL(38) LBL(39) LBL(40) LBL(41) LBL(44) LBL(45)
4722    #endif  /* SUPPORT_UCP */
4723    #endif  /* SUPPORT_UTF8 */
4724    default:    default:
4725    DPRINTF(("jump error in pcre match: label %d non-existent\n", frame->Xwhere));    DPRINTF(("jump error in pcre match: label %d non-existent\n", frame->Xwhere));
4726    return PCRE_ERROR_INTERNAL;    return PCRE_ERROR_INTERNAL;
# Line 4273  Returns:          > 0 => success; value Line 4812  Returns:          > 0 => success; value
4812                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
4813  */  */
4814    
4815  PCRE_EXP_DEFN int  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
4816  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,
4817    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,
4818    int offsetcount)    int offsetcount)
# Line 4297  const uschar *tables; Line 4836  const uschar *tables;
4836  const uschar *start_bits = NULL;  const uschar *start_bits = NULL;
4837  USPTR start_match = (USPTR)subject + start_offset;  USPTR start_match = (USPTR)subject + start_offset;
4838  USPTR end_subject;  USPTR end_subject;
4839    USPTR start_partial = NULL;
4840  USPTR req_byte_ptr = start_match - 1;  USPTR req_byte_ptr = start_match - 1;
 eptrblock eptrchain[EPTR_WORK_SIZE];  
4841    
4842  pcre_study_data internal_study;  pcre_study_data internal_study;
4843  const pcre_study_data *study;  const pcre_study_data *study;
# Line 4361  if (re->magic_number != MAGIC_NUMBER) Line 4900  if (re->magic_number != MAGIC_NUMBER)
4900  /* Set up other data */  /* Set up other data */
4901    
4902  anchored = ((re->options | options) & PCRE_ANCHORED) != 0;  anchored = ((re->options | options) & PCRE_ANCHORED) != 0;
4903  startline = (re->options & PCRE_STARTLINE) != 0;  startline = (re->flags & PCRE_STARTLINE) != 0;
4904  firstline = (re->options & PCRE_FIRSTLINE) != 0;  firstline = (re->options & PCRE_FIRSTLINE) != 0;
4905    
4906  /* 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 4376  end_subject = md->end_subject; Line 4915  end_subject = md->end_subject;
4915    
4916  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;
4917  utf8 = md->utf8 = (re->options & PCRE_UTF8) != 0;  utf8 = md->utf8 = (re->options & PCRE_UTF8) != 0;
4918    md->jscript_compat = (re->options & PCRE_JAVASCRIPT_COMPAT) != 0;
4919    
4920  md->notbol = (options & PCRE_NOTBOL) != 0;  md->notbol = (options & PCRE_NOTBOL) != 0;
4921  md->noteol = (options & PCRE_NOTEOL) != 0;  md->noteol = (options & PCRE_NOTEOL) != 0;
4922  md->notempty = (options & PCRE_NOTEMPTY) != 0;  md->notempty = (options & PCRE_NOTEMPTY) != 0;
4923  md->partial = (options & PCRE_PARTIAL) != 0;  md->partial = ((options & PCRE_PARTIAL_HARD) != 0)? 2 :
4924                  ((options & PCRE_PARTIAL_SOFT) != 0)? 1 : 0;
4925  md->hitend = FALSE;  md->hitend = FALSE;
4926    
4927  md->recursive = NULL;                   /* No recursion at top level */  md->recursive = NULL;                   /* No recursion at top level */
 md->eptrchain = eptrchain;              /* Make workspace generally available */  
4928    
4929  md->lcc = tables + lcc_offset;  md->lcc = tables + lcc_offset;
4930  md->ctypes = tables + ctypes_offset;  md->ctypes = tables + ctypes_offset;
4931    
4932    /* Handle different \R options. */
4933    
4934    switch (options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE))
4935      {
4936      case 0:
4937      if ((re->options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) != 0)
4938        md->bsr_anycrlf = (re->options & PCRE_BSR_ANYCRLF) != 0;
4939      else
4940    #ifdef BSR_ANYCRLF
4941      md->bsr_anycrlf = TRUE;
4942    #else
4943      md->bsr_anycrlf = FALSE;
4944    #endif
4945      break;
4946    
4947      case PCRE_BSR_ANYCRLF:
4948      md->bsr_anycrlf = TRUE;
4949      break;
4950    
4951      case PCRE_BSR_UNICODE:
4952      md->bsr_anycrlf = FALSE;
4953      break;
4954    
4955      default: return PCRE_ERROR_BADNEWLINE;
4956      }
4957    
4958  /* Handle different types of newline. The three bits give eight cases. If  /* Handle different types of newline. The three bits give eight cases. If
4959  nothing is set at run time, whatever was used at compile time applies. */  nothing is set at run time, whatever was used at compile time applies. */
4960    
4961  switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options : (pcre_uint32)options) &  switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options :
4962         PCRE_NEWLINE_BITS)          (pcre_uint32)options) & PCRE_NEWLINE_BITS)
4963    {    {
4964    case 0: newline = NEWLINE; break;   /* Compile-time default */    case 0: newline = NEWLINE; break;   /* Compile-time default */
4965    case PCRE_NEWLINE_CR: newline = '\r'; break;    case PCRE_NEWLINE_CR: newline = CHAR_CR; break;
4966    case PCRE_NEWLINE_LF: newline = '\n'; break;    case PCRE_NEWLINE_LF: newline = CHAR_NL; break;
4967    case PCRE_NEWLINE_CR+    case PCRE_NEWLINE_CR+
4968         PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break;         PCRE_NEWLINE_LF: newline = (CHAR_CR << 8) | CHAR_NL; break;
4969    case PCRE_NEWLINE_ANY: newline = -1; break;    case PCRE_NEWLINE_ANY: newline = -1; break;
4970    case PCRE_NEWLINE_ANYCRLF: newline = -2; break;    case PCRE_NEWLINE_ANYCRLF: newline = -2; break;
4971    default: return PCRE_ERROR_BADNEWLINE;    default: return PCRE_ERROR_BADNEWLINE;
# Line 4429  else Line 4995  else
4995      }      }
4996    }    }
4997    
4998  /* Partial matching is supported only for a restricted set of regexes at the  /* Partial matching was originally supported only for a restricted set of
4999  moment. */  regexes; from release 8.00 there are no restrictions, but the bits are still
5000    defined (though never set). So there's no harm in leaving this code. */
5001    
5002  if (md->partial && (re->options & PCRE_NOPARTIAL) != 0)  if (md->partial && (re->flags & PCRE_NOPARTIAL) != 0)
5003    return PCRE_ERROR_BADPARTIAL;    return PCRE_ERROR_BADPARTIAL;
5004    
5005  /* 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 4441  back the character offset. */ Line 5008  back the character offset. */
5008  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
5009  if (utf8 && (options & PCRE_NO_UTF8_CHECK) == 0)  if (utf8 && (options & PCRE_NO_UTF8_CHECK) == 0)
5010    {    {
5011    if (_pcre_valid_utf8((uschar *)subject, length) >= 0)    if (_pcre_valid_utf8((USPTR)subject, length) >= 0)
5012      return PCRE_ERROR_BADUTF8;      return PCRE_ERROR_BADUTF8;
5013    if (start_offset > 0 && start_offset < length)    if (start_offset > 0 && start_offset < length)
5014      {      {
5015      int tb = ((uschar *)subject)[start_offset];      int tb = ((USPTR)subject)[start_offset];
5016      if (tb > 127)      if (tb > 127)
5017        {        {
5018        tb &= 0xc0;        tb &= 0xc0;
# Line 4509  studied, there may be a bitmap of possib Line 5076  studied, there may be a bitmap of possib
5076    
5077  if (!anchored)  if (!anchored)
5078    {    {
5079    if ((re->options & PCRE_FIRSTSET) != 0)    if ((re->flags & PCRE_FIRSTSET) != 0)
5080      {      {
5081      first_byte = re->first_byte & 255;      first_byte = re->first_byte & 255;
5082      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)
# Line 4524  if (!anchored) Line 5091  if (!anchored)
5091  /* For anchored or unanchored matches, there may be a "last known required  /* For anchored or unanchored matches, there may be a "last known required
5092  character" set. */  character" set. */
5093    
5094  if ((re->options & PCRE_REQCHSET) != 0)  if ((re->flags & PCRE_REQCHSET) != 0)
5095    {    {
5096    req_byte = re->req_byte & 255;    req_byte = re->req_byte & 255;
5097    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;
# Line 4540  the loop runs just once. */ Line 5107  the loop runs just once. */
5107  for(;;)  for(;;)
5108    {    {
5109    USPTR save_end_subject = end_subject;    USPTR save_end_subject = end_subject;
5110      USPTR new_start_match;
5111    
5112    /* Reset the maximum number of extractions we might see. */    /* Reset the maximum number of extractions we might see. */
5113    
# Line 4550  for(;;) Line 5118  for(;;)
5118      while (iptr < iend) *iptr++ = -1;      while (iptr < iend) *iptr++ = -1;
5119      }      }
5120    
5121    /* Advance to a unique first char if possible. If firstline is TRUE, the    /* If firstline is TRUE, the start of the match is constrained to the first
5122    start of the match is constrained to the first line of a multiline string.    line of a multiline string. That is, the match must be before or at the first
5123    That is, the match must be before or at the first newline. Implement this by    newline. Implement this by temporarily adjusting end_subject so that we stop
5124    temporarily adjusting end_subject so that we stop scanning at a newline. If    scanning at a newline. If the match fails at the newline, later code breaks
5125    the match fails at the newline, later code breaks this loop. */    this loop. */
5126    
5127    if (firstline)    if (firstline)
5128      {      {
5129      USPTR t = start_match;      USPTR t = start_match;
5130    #ifdef SUPPORT_UTF8
5131        if (utf8)
5132          {
5133          while (t < md->end_subject && !IS_NEWLINE(t))
5134            {
5135            t++;
5136            while (t < end_subject && (*t & 0xc0) == 0x80) t++;
5137            }
5138          }
5139        else
5140    #endif
5141      while (t < md->end_subject && !IS_NEWLINE(t)) t++;      while (t < md->end_subject && !IS_NEWLINE(t)) t++;
5142      end_subject = t;      end_subject = t;
5143      }      }
5144    
5145    /* Now test for a unique first byte */    /* There are some optimizations that avoid running the match if a known
5146      starting point is not found, or if a known later character is not present.
5147      However, there is an option that disables these, for testing and for ensuring
5148      that all callouts do actually occur. */
5149    
5150    if (first_byte >= 0)    if ((options & PCRE_NO_START_OPTIMIZE) == 0)
5151      {      {
5152      if (first_byte_caseless)      /* Advance to a unique first byte if there is one. */
5153        while (start_match < end_subject &&  
5154               md->lcc[*start_match] != first_byte)      if (first_byte >= 0)
5155          start_match++;        {
5156      else        if (first_byte_caseless)
5157        while (start_match < end_subject && *start_match != first_byte)          while (start_match < end_subject && md->lcc[*start_match] != first_byte)
5158          start_match++;            start_match++;
5159      }        else
5160            while (start_match < end_subject && *start_match != first_byte)
5161              start_match++;
5162          }
5163    
5164    /* Or to just after a linebreak for a multiline match if possible */      /* Or to just after a linebreak for a multiline match */
5165    
5166    else if (startline)      else if (startline)
     {  
     if (start_match > md->start_subject + start_offset)  
5167        {        {
5168        while (start_match <= end_subject && !WAS_NEWLINE(start_match))        if (start_match > md->start_subject + start_offset)
5169          start_match++;          {
5170    #ifdef SUPPORT_UTF8
5171            if (utf8)
5172              {
5173              while (start_match < end_subject && !WAS_NEWLINE(start_match))
5174                {
5175                start_match++;
5176                while(start_match < end_subject && (*start_match & 0xc0) == 0x80)
5177                  start_match++;
5178                }
5179              }
5180            else
5181    #endif
5182            while (start_match < end_subject && !WAS_NEWLINE(start_match))
5183              start_match++;
5184    
5185            /* If we have just passed a CR and the newline option is ANY or ANYCRLF,
5186            and we are now at a LF, advance the match position by one more character.
5187            */
5188    
5189        /* If we have just passed a CR and the newline option is ANY or ANYCRLF,          if (start_match[-1] == CHAR_CR &&
5190        and we are now at a LF, advance the match position by one more character.               (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&
5191        */               start_match < end_subject &&
5192                 *start_match == CHAR_NL)
5193        if (start_match[-1] == '\r' &&            start_match++;
5194             (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&          }
            start_match < end_subject &&  
            *start_match == '\n')  
         start_match++;  
5195        }        }
     }  
5196    
5197    /* Or to a non-unique first char after study */      /* Or to a non-unique first byte after study */
5198    
5199    else if (start_bits != NULL)      else if (start_bits != NULL)
     {  
     while (start_match < end_subject)  
5200        {        {
5201        register unsigned int c = *start_match;        while (start_match < end_subject)
5202        if ((start_bits[c/8] & (1 << (c&7))) == 0) start_match++; else break;          {
5203            register unsigned int c = *start_match;
5204            if ((start_bits[c/8] & (1 << (c&7))) == 0) start_match++;
5205              else break;
5206            }
5207        }        }
5208      }      }   /* Starting optimizations */
5209    
5210    /* Restore fudged end_subject */    /* Restore fudged end_subject */
5211    
# Line 4618  for(;;) Line 5217  for(;;)
5217    printf("\n");    printf("\n");
5218  #endif  #endif
5219    
5220    /* If req_byte is set, we know that that character must appear in the subject    /* If req_byte is set, we know that that character must appear in the
5221    for the match to succeed. If the first character is set, req_byte must be    subject for the match to succeed. If the first character is set, req_byte
5222    later in the subject; otherwise the test starts at the match point. This    must be later in the subject; otherwise the test starts at the match point.
5223    optimization can save a huge amount of backtracking in patterns with nested    This optimization can save a huge amount of backtracking in patterns with
5224    unlimited repeats that aren't going to match. Writing separate code for    nested unlimited repeats that aren't going to match. Writing separate code
5225    cased/caseless versions makes it go faster, as does using an autoincrement    for cased/caseless versions makes it go faster, as does using an
5226    and backing off on a match.    autoincrement and backing off on a match.
5227    
5228    HOWEVER: when the subject string is very, very long, searching to its end can    HOWEVER: when the subject string is very, very long, searching to its end
5229    take a long time, and give bad performance on quite ordinary patterns. This    can take a long time, and give bad performance on quite ordinary patterns.
5230    showed up when somebody was matching something like /^\d+C/ on a 32-megabyte    This showed up when somebody was matching something like /^\d+C/ on a
5231    string... so we don't do this when the string is sufficiently long.    32-megabyte string... so we don't do this when the string is sufficiently
5232      long.
5233    
5234    ALSO: this processing is disabled when partial matching is requested.    ALSO: this processing is disabled when partial matching is requested, or if
5235    */    disabling is explicitly requested. */
5236    
5237    if (req_byte >= 0 &&    if ((options & PCRE_NO_START_OPTIMIZE) == 0 &&
5238          req_byte >= 0 &&
5239        end_subject - start_match < REQ_BYTE_MAX &&        end_subject - start_match < REQ_BYTE_MAX &&
5240        !md->partial)        !md->partial)
5241      {      {
# Line 4678  for(;;) Line 5279  for(;;)
5279        }        }
5280      }      }
5281    
5282    /* OK, we can now run the match. */    /* OK, we can now run the match. If "hitend" is set afterwards, remember the
5283      first starting point for which a partial match was found. */
5284    
5285    md->start_match_ptr = start_match;      /* Insurance */    md->start_match_ptr = start_match;
5286      md->start_used_ptr = start_match;
5287    md->match_call_count = 0;    md->match_call_count = 0;
5288    md->eptrn = 0;                          /* Next free eptrchain slot */    rc = match(start_match, md->start_code, start_match, 2, md, ims, NULL, 0, 0);
5289    rc = match(start_match, md->start_code, start_match, 2, md,    if (md->hitend && start_partial == NULL) start_partial = md->start_used_ptr;
5290      ims, NULL, 0, 0);  
5291      switch(rc)
5292        {
5293        /* NOMATCH and PRUNE advance by one character. THEN at this level acts
5294        exactly like PRUNE. */
5295    
5296    /* Any return other than MATCH_NOMATCH breaks the loop. */      case MATCH_NOMATCH:
5297        case MATCH_PRUNE:
5298        case MATCH_THEN:
5299        new_start_match = start_match + 1;
5300    #ifdef SUPPORT_UTF8
5301        if (utf8)
5302          while(new_start_match < end_subject && (*new_start_match & 0xc0) == 0x80)
5303            new_start_match++;
5304    #endif
5305        break;
5306    
5307    if (rc != MATCH_NOMATCH) break;      /* SKIP passes back the next starting point explicitly. */
5308    
5309        case MATCH_SKIP:
5310        new_start_match = md->start_match_ptr;
5311        break;
5312    
5313        /* COMMIT disables the bumpalong, but otherwise behaves as NOMATCH. */
5314    
5315        case MATCH_COMMIT:
5316        rc = MATCH_NOMATCH;
5317        goto ENDLOOP;
5318    
5319        /* Any other return is some kind of error. */
5320    
5321        default:
5322        goto ENDLOOP;
5323        }
5324    
5325      /* Control reaches here for the various types of "no match at this point"
5326      result. Reset the code to MATCH_NOMATCH for subsequent checking. */
5327    
5328      rc = MATCH_NOMATCH;
5329    
5330    /* 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
5331    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 4696  for(;;) Line 5333  for(;;)
5333    
5334    if (firstline && IS_NEWLINE(start_match)) break;    if (firstline && IS_NEWLINE(start_match)) break;
5335    
5336    /* Advance the match position by one character. */    /* Advance to new matching position */
5337    
5338    start_match++;    start_match = new_start_match;
 #ifdef SUPPORT_UTF8  
   if (utf8)  
     while(start_match < end_subject && (*start_match & 0xc0) == 0x80)  
       start_match++;  
 #endif  
5339    
5340    /* 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
5341    the subject. */    the subject. */
5342    
5343    if (anchored || start_match > end_subject) break;    if (anchored || start_match > end_subject) break;
5344    
5345    /* If we have just passed a CR and the newline option is CRLF or ANY or    /* If we have just passed a CR and we are now at a LF, and the pattern does
5346    ANYCRLF, and we are now at a LF, advance the match position by one more    not contain any explicit matches for \r or \n, and the newline option is CRLF
5347    character. */    or ANY or ANYCRLF, advance the match position by one more character. */
5348    
5349    if (start_match[-1] == '\r' &&    if (start_match[-1] == CHAR_CR &&
5350         (md->nltype == NLTYPE_ANY ||        start_match < end_subject &&
5351          md->nltype == NLTYPE_ANYCRLF ||        *start_match == CHAR_NL &&
5352          md->nllen == 2) &&        (re->flags & PCRE_HASCRORLF) == 0 &&
5353         start_match < end_subject &&          (md->nltype == NLTYPE_ANY ||
5354         *start_match == '\n')           md->nltype == NLTYPE_ANYCRLF ||
5355             md->nllen == 2))
5356      start_match++;      start_match++;
5357    
5358    }   /* End of for(;;) "bumpalong" loop */    }   /* End of for(;;) "bumpalong" loop */
# Line 4729  for(;;) Line 5362  for(;;)
5362  /* 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
5363  conditions is true:  conditions is true:
5364    
5365  (1) The pattern is anchored;  (1) The pattern is anchored or the match was failed by (*COMMIT);
5366    
5367  (2) We are past the end of the subject;  (2) We are past the end of the subject;
5368    
# Line 4744  processing, copy those that we can. In t Line 5377  processing, copy those that we can. In t
5377  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
5378  capturing parentheses than vector slots. */  capturing parentheses than vector slots. */
5379    
5380    ENDLOOP:
5381    
5382  if (rc == MATCH_MATCH)  if (rc == MATCH_MATCH)
5383    {    {
5384    if (using_temporary_offsets)    if (using_temporary_offsets)
# Line 4787  if (using_temporary_offsets) Line 5422  if (using_temporary_offsets)
5422    (pcre_free)(md->offset_vector);    (pcre_free)(md->offset_vector);
5423    }    }
5424    
5425  if (rc != MATCH_NOMATCH)  if (rc != MATCH_NOMATCH && rc != PCRE_ERROR_PARTIAL)
5426    {    {
5427    DPRINTF((">>>> error: returning %d\n", rc));    DPRINTF((">>>> error: returning %d\n", rc));
5428    return rc;    return rc;
5429    }    }
5430  else if (md->partial && md->hitend)  else if (start_partial != NULL)
5431    {    {
5432    DPRINTF((">>>> returning PCRE_ERROR_PARTIAL\n"));    DPRINTF((">>>> returning PCRE_ERROR_PARTIAL\n"));
5433      if (offsetcount > 1)
5434        {
5435        offsets[0] = start_partial - (USPTR)subject;
5436        offsets[1] = end_subject - (USPTR)subject;
5437        }
5438    return PCRE_ERROR_PARTIAL;    return PCRE_ERROR_PARTIAL;
5439    }    }
5440  else  else

Legend:
Removed from v.184  
changed lines
  Added in v.435

  ViewVC Help
Powered by ViewVC 1.1.5