/[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 648 by ph10, Mon Aug 1 11:02:08 2011 UTC revision 1041 by ph10, Sun Sep 16 10:16:27 2012 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) 1997-2011 University of Cambridge             Copyright (c) 1997-2012 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 37  POSSIBILITY OF SUCH DAMAGE. Line 37  POSSIBILITY OF SUCH DAMAGE.
37  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
38  */  */
39    
   
40  /* This module contains pcre_exec(), the externally visible function that does  /* This module contains pcre_exec(), the externally visible function that does
41  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
42  possible. There are also some static supporting functions. */  possible. There are also some static supporting functions. */
# Line 82  negative to avoid the external error cod Line 81  negative to avoid the external error cod
81  #define MATCH_SKIP_ARG     (-993)  #define MATCH_SKIP_ARG     (-993)
82  #define MATCH_THEN         (-992)  #define MATCH_THEN         (-992)
83    
 /* This is a convenience macro for code that occurs many times. */  
   
 #define MRRETURN(ra) \  
   { \  
   md->mark = markptr; \  
   RRETURN(ra); \  
   }  
   
84  /* 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.
85  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,
86  because the offset vector is always a multiple of 3 long. */  because the offset vector is always a multiple of 3 long. */
# Line 121  Returns:     nothing Line 112  Returns:     nothing
112  */  */
113    
114  static void  static void
115  pchars(const uschar *p, int length, BOOL is_subject, match_data *md)  pchars(const pcre_uchar *p, int length, BOOL is_subject, match_data *md)
116  {  {
117  unsigned int c;  unsigned int c;
118  if (is_subject && length > md->end_subject - p) length = md->end_subject - p;  if (is_subject && length > md->end_subject - p) length = md->end_subject - p;
# Line 148  Arguments: Line 139  Arguments:
139    md          points to match data block    md          points to match data block
140    caseless    TRUE if caseless    caseless    TRUE if caseless
141    
142  Returns:      < 0 if not matched, otherwise the number of subject bytes matched  Returns:      >= 0 the number of subject bytes matched
143                  -1 no match
144                  -2 partial match; always given if at end subject
145  */  */
146    
147  static int  static int
148  match_ref(int offset, register USPTR eptr, int length, match_data *md,  match_ref(int offset, register PCRE_PUCHAR eptr, int length, match_data *md,
149    BOOL caseless)    BOOL caseless)
150  {  {
151  USPTR eptr_start = eptr;  PCRE_PUCHAR eptr_start = eptr;
152  register USPTR p = md->start_subject + md->offset_vector[offset];  register PCRE_PUCHAR p = md->start_subject + md->offset_vector[offset];
153    
154  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
155  if (eptr >= md->end_subject)  if (eptr >= md->end_subject)
# Line 171  pchars(p, length, FALSE, md); Line 164  pchars(p, length, FALSE, md);
164  printf("\n");  printf("\n");
165  #endif  #endif
166    
167  /* Always fail if reference not set (and not JavaScript compatible). */  /* Always fail if reference not set (and not JavaScript compatible - in that
168    case the length is passed as zero). */
169    
170  if (length < 0) return -1;  if (length < 0) return -1;
171    
# Line 181  ASCII characters. */ Line 175  ASCII characters. */
175    
176  if (caseless)  if (caseless)
177    {    {
178  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
179  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
180    if (md->utf8)    if (md->utf)
181      {      {
182      /* Match characters up to the end of the reference. NOTE: the number of      /* Match characters up to the end of the reference. NOTE: the number of
183      bytes matched may differ, because there are some characters whose upper and      bytes matched may differ, because there are some characters whose upper and
# Line 193  if (caseless) Line 187  if (caseless)
187      the latter. It is important, therefore, to check the length along the      the latter. It is important, therefore, to check the length along the
188      reference, not along the subject (earlier code did this wrong). */      reference, not along the subject (earlier code did this wrong). */
189    
190      USPTR endptr = p + length;      PCRE_PUCHAR endptr = p + length;
191      while (p < endptr)      while (p < endptr)
192        {        {
193        int c, d;        int c, d;
194        if (eptr >= md->end_subject) return -1;        if (eptr >= md->end_subject) return -2;   /* Partial match */
195        GETCHARINC(c, eptr);        GETCHARINC(c, eptr);
196        GETCHARINC(d, p);        GETCHARINC(d, p);
197        if (c != d && c != UCD_OTHERCASE(d)) return -1;        if (c != d && c != UCD_OTHERCASE(d)) return -1;
# Line 210  if (caseless) Line 204  if (caseless)
204    /* The same code works when not in UTF-8 mode and in UTF-8 mode when there    /* The same code works when not in UTF-8 mode and in UTF-8 mode when there
205    is no UCP support. */    is no UCP support. */
206      {      {
     if (eptr + length > md->end_subject) return -1;  
207      while (length-- > 0)      while (length-- > 0)
208        { if (md->lcc[*p++] != md->lcc[*eptr++]) return -1; }        {
209          if (eptr >= md->end_subject) return -2;   /* Partial match */
210          if (TABLE_GET(*p, md->lcc, *p) != TABLE_GET(*eptr, md->lcc, *eptr)) return -1;
211          p++;
212          eptr++;
213          }
214      }      }
215    }    }
216    
# Line 221  are in UTF-8 mode. */ Line 219  are in UTF-8 mode. */
219    
220  else  else
221    {    {
222    if (eptr + length > md->end_subject) return -1;    while (length-- > 0)
223    while (length-- > 0) if (*p++ != *eptr++) return -1;      {
224        if (eptr >= md->end_subject) return -2;   /* Partial match */
225        if (*p++ != *eptr++) return -1;
226        }
227    }    }
228    
229  return eptr - eptr_start;  return (int)(eptr - eptr_start);
230  }  }
231    
232    
# Line 277  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM Line 278  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM
278         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,
279         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,
280         RM51,  RM52, RM53, RM54, RM55, RM56, RM57, RM58, RM59, RM60,         RM51,  RM52, RM53, RM54, RM55, RM56, RM57, RM58, RM59, RM60,
281         RM61,  RM62, RM63 };         RM61,  RM62, RM63, RM64, RM65, RM66 };
282    
283  /* 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
284  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  actually used in this definition. */ Line 291  actually used in this definition. */
291  #define RMATCH(ra,rb,rc,rd,re,rw) \  #define RMATCH(ra,rb,rc,rd,re,rw) \
292    { \    { \
293    printf("match() called in line %d\n", __LINE__); \    printf("match() called in line %d\n", __LINE__); \
294    rrc = match(ra,rb,mstart,markptr,rc,rd,re,rdepth+1); \    rrc = match(ra,rb,mstart,rc,rd,re,rdepth+1); \
295    printf("to line %d\n", __LINE__); \    printf("to line %d\n", __LINE__); \
296    }    }
297  #define RRETURN(ra) \  #define RRETURN(ra) \
# Line 300  actually used in this definition. */ Line 301  actually used in this definition. */
301    }    }
302  #else  #else
303  #define RMATCH(ra,rb,rc,rd,re,rw) \  #define RMATCH(ra,rb,rc,rd,re,rw) \
304    rrc = match(ra,rb,mstart,markptr,rc,rd,re,rdepth+1)    rrc = match(ra,rb,mstart,rc,rd,re,rdepth+1)
305  #define RRETURN(ra) return ra  #define RRETURN(ra) return ra
306  #endif  #endif
307    
# Line 315  argument of match(), which never changes Line 316  argument of match(), which never changes
316    
317  #define RMATCH(ra,rb,rc,rd,re,rw)\  #define RMATCH(ra,rb,rc,rd,re,rw)\
318    {\    {\
319    heapframe *newframe = (heapframe *)(pcre_stack_malloc)(sizeof(heapframe));\    heapframe *newframe = frame->Xnextframe;\
320    if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\    if (newframe == NULL)\
321    frame->Xwhere = rw; \      {\
322        newframe = (heapframe *)(PUBL(stack_malloc))(sizeof(heapframe));\
323        if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\
324        newframe->Xnextframe = NULL;\
325        frame->Xnextframe = newframe;\
326        }\
327      frame->Xwhere = rw;\
328    newframe->Xeptr = ra;\    newframe->Xeptr = ra;\
329    newframe->Xecode = rb;\    newframe->Xecode = rb;\
330    newframe->Xmstart = mstart;\    newframe->Xmstart = mstart;\
   newframe->Xmarkptr = markptr;\  
331    newframe->Xoffset_top = rc;\    newframe->Xoffset_top = rc;\
332    newframe->Xeptrb = re;\    newframe->Xeptrb = re;\
333    newframe->Xrdepth = frame->Xrdepth + 1;\    newframe->Xrdepth = frame->Xrdepth + 1;\
# Line 337  argument of match(), which never changes Line 343  argument of match(), which never changes
343    {\    {\
344    heapframe *oldframe = frame;\    heapframe *oldframe = frame;\
345    frame = oldframe->Xprevframe;\    frame = oldframe->Xprevframe;\
   (pcre_stack_free)(oldframe);\  
346    if (frame != NULL)\    if (frame != NULL)\
347      {\      {\
348      rrc = ra;\      rrc = ra;\
# Line 351  argument of match(), which never changes Line 356  argument of match(), which never changes
356    
357  typedef struct heapframe {  typedef struct heapframe {
358    struct heapframe *Xprevframe;    struct heapframe *Xprevframe;
359      struct heapframe *Xnextframe;
360    
361    /* Function arguments that may change */    /* Function arguments that may change */
362    
363    USPTR Xeptr;    PCRE_PUCHAR Xeptr;
364    const uschar *Xecode;    const pcre_uchar *Xecode;
365    USPTR Xmstart;    PCRE_PUCHAR Xmstart;
   USPTR Xmarkptr;  
366    int Xoffset_top;    int Xoffset_top;
367    eptrblock *Xeptrb;    eptrblock *Xeptrb;
368    unsigned int Xrdepth;    unsigned int Xrdepth;
369    
370    /* Function local variables */    /* Function local variables */
371    
372    USPTR Xcallpat;    PCRE_PUCHAR Xcallpat;
373  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
374    USPTR Xcharptr;    PCRE_PUCHAR Xcharptr;
375  #endif  #endif
376    USPTR Xdata;    PCRE_PUCHAR Xdata;
377    USPTR Xnext;    PCRE_PUCHAR Xnext;
378    USPTR Xpp;    PCRE_PUCHAR Xpp;
379    USPTR Xprev;    PCRE_PUCHAR Xprev;
380    USPTR Xsaved_eptr;    PCRE_PUCHAR Xsaved_eptr;
381    
382    recursion_info Xnew_recursive;    recursion_info Xnew_recursive;
383    
# Line 385  typedef struct heapframe { Line 390  typedef struct heapframe {
390    int Xprop_value;    int Xprop_value;
391    int Xprop_fail_result;    int Xprop_fail_result;
392    int Xoclength;    int Xoclength;
393    uschar Xocchars[8];    pcre_uchar Xocchars[6];
394  #endif  #endif
395    
396    int Xcodelink;    int Xcodelink;
# Line 427  returns a negative (error) response, the Line 432  returns a negative (error) response, the
432  same response. */  same response. */
433    
434  /* These macros pack up tests that are used for partial matching, and which  /* These macros pack up tests that are used for partial matching, and which
435  appears several times in the code. We set the "hit end" flag if the pointer is  appear several times in the code. We set the "hit end" flag if the pointer is
436  at the end of the subject and also past the start of the subject (i.e.  at the end of the subject and also past the start of the subject (i.e.
437  something has been matched). For hard partial matching, we then return  something has been matched). For hard partial matching, we then return
438  immediately. The second one is used when we already know we are past the end of  immediately. The second one is used when we already know we are past the end of
# Line 438  the subject. */ Line 443  the subject. */
443        eptr > md->start_used_ptr) \        eptr > md->start_used_ptr) \
444      { \      { \
445      md->hitend = TRUE; \      md->hitend = TRUE; \
446      if (md->partial > 1) MRRETURN(PCRE_ERROR_PARTIAL); \      if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL); \
447      }      }
448    
449  #define SCHECK_PARTIAL()\  #define SCHECK_PARTIAL()\
450    if (md->partial != 0 && eptr > md->start_used_ptr) \    if (md->partial != 0 && eptr > md->start_used_ptr) \
451      { \      { \
452      md->hitend = TRUE; \      md->hitend = TRUE; \
453      if (md->partial > 1) MRRETURN(PCRE_ERROR_PARTIAL); \      if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL); \
454      }      }
455    
456    
457  /* Performance note: It might be tempting to extract commonly used fields from  /* Performance note: It might be tempting to extract commonly used fields from
458  the md structure (e.g. utf8, end_subject) into individual variables to improve  the md structure (e.g. utf, end_subject) into individual variables to improve
459  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
460  made performance worse.  made performance worse.
461    
# Line 459  Arguments: Line 464  Arguments:
464     ecode       pointer to current position in compiled code     ecode       pointer to current position in compiled code
465     mstart      pointer to the current match start position (can be modified     mstart      pointer to the current match start position (can be modified
466                   by encountering \K)                   by encountering \K)
    markptr     pointer to the most recent MARK name, or NULL  
467     offset_top  current top pointer     offset_top  current top pointer
468     md          pointer to "static" info for the match     md          pointer to "static" info for the match
469     eptrb       pointer to chain of blocks containing eptr at start of     eptrb       pointer to chain of blocks containing eptr at start of
# Line 474  Returns:       MATCH_MATCH if matched Line 478  Returns:       MATCH_MATCH if matched
478  */  */
479    
480  static int  static int
481  match(REGISTER USPTR eptr, REGISTER const uschar *ecode, USPTR mstart,  match(REGISTER PCRE_PUCHAR eptr, REGISTER const pcre_uchar *ecode,
482    const uschar *markptr, int offset_top, match_data *md, eptrblock *eptrb,    PCRE_PUCHAR mstart, int offset_top, match_data *md, eptrblock *eptrb,
483    unsigned int rdepth)    unsigned int rdepth)
484  {  {
485  /* These variables do not need to be preserved over recursion in this function,  /* These variables do not need to be preserved over recursion in this function,
# Line 485  so they can be ordinary variables in all Line 489  so they can be ordinary variables in all
489  register int  rrc;         /* Returns from recursive calls */  register int  rrc;         /* Returns from recursive calls */
490  register int  i;           /* Used for loops not involving calls to RMATCH() */  register int  i;           /* Used for loops not involving calls to RMATCH() */
491  register unsigned int c;   /* Character values not kept over RMATCH() calls */  register unsigned int c;   /* Character values not kept over RMATCH() calls */
492  register BOOL utf8;        /* Local copy of UTF-8 flag for speed */  register BOOL utf;         /* Local copy of UTF flag for speed */
493    
494  BOOL minimize, possessive; /* Quantifier options */  BOOL minimize, possessive; /* Quantifier options */
495  BOOL caseless;  BOOL caseless;
496  int condcode;  int condcode;
497    
498  /* 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
499  preserved over calls to RMATCH() are part of a "frame" which is obtained from  preserved over calls to RMATCH() are part of a "frame". We set up the top-level
500  heap storage. Set up the top-level frame here; others are obtained from the  frame on the stack here; subsequent instantiations are obtained from the heap
501  heap whenever RMATCH() does a "recursion". See the macro definitions above. */  whenever RMATCH() does a "recursion". See the macro definitions above. Putting
502    the top-level on the stack rather than malloc-ing them all gives a performance
503    boost in many cases where there is not much "recursion". */
504    
505  #ifdef NO_RECURSE  #ifdef NO_RECURSE
506  heapframe *frame = (heapframe *)(pcre_stack_malloc)(sizeof(heapframe));  heapframe *frame = (heapframe *)md->match_frames_base;
 if (frame == NULL) RRETURN(PCRE_ERROR_NOMEMORY);  
 frame->Xprevframe = NULL;            /* Marks the top level */  
507    
508  /* Copy in the original argument variables */  /* Copy in the original argument variables */
509    
510  frame->Xeptr = eptr;  frame->Xeptr = eptr;
511  frame->Xecode = ecode;  frame->Xecode = ecode;
512  frame->Xmstart = mstart;  frame->Xmstart = mstart;
 frame->Xmarkptr = markptr;  
513  frame->Xoffset_top = offset_top;  frame->Xoffset_top = offset_top;
514  frame->Xeptrb = eptrb;  frame->Xeptrb = eptrb;
515  frame->Xrdepth = rdepth;  frame->Xrdepth = rdepth;
# Line 520  HEAP_RECURSE: Line 523  HEAP_RECURSE:
523  #define eptr               frame->Xeptr  #define eptr               frame->Xeptr
524  #define ecode              frame->Xecode  #define ecode              frame->Xecode
525  #define mstart             frame->Xmstart  #define mstart             frame->Xmstart
 #define markptr            frame->Xmarkptr  
526  #define offset_top         frame->Xoffset_top  #define offset_top         frame->Xoffset_top
527  #define eptrb              frame->Xeptrb  #define eptrb              frame->Xeptrb
528  #define rdepth             frame->Xrdepth  #define rdepth             frame->Xrdepth
529    
530  /* Ditto for the local variables */  /* Ditto for the local variables */
531    
532  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
533  #define charptr            frame->Xcharptr  #define charptr            frame->Xcharptr
534  #endif  #endif
535  #define callpat            frame->Xcallpat  #define callpat            frame->Xcallpat
# Line 585  declarations can be cut out in a block. Line 587  declarations can be cut out in a block.
587  below are for variables that do not have to be preserved over a recursive call  below are for variables that do not have to be preserved over a recursive call
588  to RMATCH(). */  to RMATCH(). */
589    
590  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
591  const uschar *charptr;  const pcre_uchar *charptr;
592  #endif  #endif
593  const uschar *callpat;  const pcre_uchar *callpat;
594  const uschar *data;  const pcre_uchar *data;
595  const uschar *next;  const pcre_uchar *next;
596  USPTR         pp;  PCRE_PUCHAR       pp;
597  const uschar *prev;  const pcre_uchar *prev;
598  USPTR         saved_eptr;  PCRE_PUCHAR       saved_eptr;
599    
600  recursion_info new_recursive;  recursion_info new_recursive;
601    
# Line 606  int prop_type; Line 608  int prop_type;
608  int prop_value;  int prop_value;
609  int prop_fail_result;  int prop_fail_result;
610  int oclength;  int oclength;
611  uschar occhars[8];  pcre_uchar occhars[6];
612  #endif  #endif
613    
614  int codelink;  int codelink;
# Line 622  int save_offset1, save_offset2, save_off Line 624  int save_offset1, save_offset2, save_off
624  int stacksave[REC_STACK_SAVE_MAX];  int stacksave[REC_STACK_SAVE_MAX];
625    
626  eptrblock newptrb;  eptrblock newptrb;
627    
628    /* There is a special fudge for calling match() in a way that causes it to
629    measure the size of its basic stack frame when the stack is being used for
630    recursion. The second argument (ecode) being NULL triggers this behaviour. It
631    cannot normally ever be NULL. The return is the negated value of the frame
632    size. */
633    
634    if (ecode == NULL)
635      {
636      if (rdepth == 0)
637        return match((PCRE_PUCHAR)&rdepth, NULL, NULL, 0, NULL, NULL, 1);
638      else
639        {
640        int len = (char *)&rdepth - (char *)eptr;
641        return (len > 0)? -len : len;
642        }
643      }
644  #endif     /* NO_RECURSE */  #endif     /* NO_RECURSE */
645    
646  /* To save space on the stack and in the heap frame, I have doubled up on some  /* To save space on the stack and in the heap frame, I have doubled up on some
# Line 634  the alternative names that are used. */ Line 653  the alternative names that are used. */
653  #define code_offset   codelink  #define code_offset   codelink
654  #define condassert    condition  #define condassert    condition
655  #define matched_once  prev_is_word  #define matched_once  prev_is_word
656    #define foc           number
657    #define save_mark     data
658    
659  /* These statements are here to stop the compiler complaining about unitialized  /* These statements are here to stop the compiler complaining about unitialized
660  variables. */  variables. */
# Line 659  defined). However, RMATCH isn't like a f Line 680  defined). However, RMATCH isn't like a f
680  complicated macro. It has to be used in one particular way. This shouldn't,  complicated macro. It has to be used in one particular way. This shouldn't,
681  however, impact performance when true recursion is being used. */  however, impact performance when true recursion is being used. */
682    
683  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
684  utf8 = md->utf8;       /* Local copy of the flag */  utf = md->utf;       /* Local copy of the flag */
685  #else  #else
686  utf8 = FALSE;  utf = FALSE;
687  #endif  #endif
688    
689  /* First check that we haven't called match() too many times, or that we  /* First check that we haven't called match() too many times, or that we
# Line 701  for (;;) Line 722  for (;;)
722    switch(op)    switch(op)
723      {      {
724      case OP_MARK:      case OP_MARK:
725      markptr = ecode + 2;      md->nomatch_mark = ecode + 2;
726      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      md->mark = NULL;    /* In case previously set by assertion */
727        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
728        eptrb, RM55);        eptrb, RM55);
729        if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
730             md->mark == NULL) md->mark = ecode + 2;
731    
732      /* A return of MATCH_SKIP_ARG means that matching failed at SKIP with an      /* A return of MATCH_SKIP_ARG means that matching failed at SKIP with an
733      argument, and we must check whether that argument matches this MARK's      argument, and we must check whether that argument matches this MARK's
# Line 712  for (;;) Line 736  for (;;)
736      position and return MATCH_SKIP. Otherwise, pass back the return code      position and return MATCH_SKIP. Otherwise, pass back the return code
737      unaltered. */      unaltered. */
738    
739      if (rrc == MATCH_SKIP_ARG &&      else if (rrc == MATCH_SKIP_ARG &&
740          strcmp((char *)markptr, (char *)(md->start_match_ptr)) == 0)          STRCMP_UC_UC(ecode + 2, md->start_match_ptr) == 0)
741        {        {
742        md->start_match_ptr = eptr;        md->start_match_ptr = eptr;
743        RRETURN(MATCH_SKIP);        RRETURN(MATCH_SKIP);
744        }        }
   
     if (md->mark == NULL) md->mark = markptr;  
745      RRETURN(rrc);      RRETURN(rrc);
746    
747      case OP_FAIL:      case OP_FAIL:
748      MRRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
749    
750      /* COMMIT overrides PRUNE, SKIP, and THEN */      /* COMMIT overrides PRUNE, SKIP, and THEN */
751    
752      case OP_COMMIT:      case OP_COMMIT:
753      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
754        eptrb, RM52);        eptrb, RM52);
755      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE &&      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE &&
756          rrc != MATCH_SKIP && rrc != MATCH_SKIP_ARG &&          rrc != MATCH_SKIP && rrc != MATCH_SKIP_ARG &&
757          rrc != MATCH_THEN)          rrc != MATCH_THEN)
758        RRETURN(rrc);        RRETURN(rrc);
759      MRRETURN(MATCH_COMMIT);      RRETURN(MATCH_COMMIT);
760    
761      /* PRUNE overrides THEN */      /* PRUNE overrides THEN */
762    
763      case OP_PRUNE:      case OP_PRUNE:
764      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
765        eptrb, RM51);        eptrb, RM51);
766      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
767      MRRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
768    
769      case OP_PRUNE_ARG:      case OP_PRUNE_ARG:
770      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      md->nomatch_mark = ecode + 2;
771        md->mark = NULL;    /* In case previously set by assertion */
772        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
773        eptrb, RM56);        eptrb, RM56);
774        if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
775             md->mark == NULL) md->mark = ecode + 2;
776      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
     md->mark = ecode + 2;  
777      RRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
778    
779      /* SKIP overrides PRUNE and THEN */      /* SKIP overrides PRUNE and THEN */
780    
781      case OP_SKIP:      case OP_SKIP:
782      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
783        eptrb, RM53);        eptrb, RM53);
784      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)
785        RRETURN(rrc);        RRETURN(rrc);
786      md->start_match_ptr = eptr;   /* Pass back current position */      md->start_match_ptr = eptr;   /* Pass back current position */
787      MRRETURN(MATCH_SKIP);      RRETURN(MATCH_SKIP);
788    
789        /* Note that, for Perl compatibility, SKIP with an argument does NOT set
790        nomatch_mark. There is a flag that disables this opcode when re-matching a
791        pattern that ended with a SKIP for which there was not a matching MARK. */
792    
793      case OP_SKIP_ARG:      case OP_SKIP_ARG:
794      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      if (md->ignore_skip_arg)
795          {
796          ecode += PRIV(OP_lengths)[*ecode] + ecode[1];
797          break;
798          }
799        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
800        eptrb, RM57);        eptrb, RM57);
801      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)
802        RRETURN(rrc);        RRETURN(rrc);
803    
804      /* Pass back the current skip name by overloading md->start_match_ptr and      /* Pass back the current skip name by overloading md->start_match_ptr and
805      returning the special MATCH_SKIP_ARG return code. This will either be      returning the special MATCH_SKIP_ARG return code. This will either be
806      caught by a matching MARK, or get to the top, where it is treated the same      caught by a matching MARK, or get to the top, where it causes a rematch
807      as PRUNE. */      with the md->ignore_skip_arg flag set. */
808    
809      md->start_match_ptr = ecode + 2;      md->start_match_ptr = ecode + 2;
810      RRETURN(MATCH_SKIP_ARG);      RRETURN(MATCH_SKIP_ARG);
811    
812      /* For THEN (and THEN_ARG) we pass back the address of the bracket or      /* For THEN (and THEN_ARG) we pass back the address of the opcode, so that
813      the alt that is at the start of the current branch. This makes it possible      the branch in which it occurs can be determined. Overload the start of
814      to skip back past alternatives that precede the THEN within the current      match pointer to do this. */
     branch. */  
815    
816      case OP_THEN:      case OP_THEN:
817      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
818        eptrb, RM54);        eptrb, RM54);
819      if (rrc != MATCH_NOMATCH) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
820      md->start_match_ptr = ecode - GET(ecode, 1);      md->start_match_ptr = ecode;
821      MRRETURN(MATCH_THEN);      RRETURN(MATCH_THEN);
822    
823      case OP_THEN_ARG:      case OP_THEN_ARG:
824      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1+LINK_SIZE],      md->nomatch_mark = ecode + 2;
825        offset_top, md, eptrb, RM58);      md->mark = NULL;    /* In case previously set by assertion */
826        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top,
827          md, eptrb, RM58);
828        if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
829             md->mark == NULL) md->mark = ecode + 2;
830      if (rrc != MATCH_NOMATCH) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
831      md->start_match_ptr = ecode - GET(ecode, 1);      md->start_match_ptr = ecode;
     md->mark = ecode + LINK_SIZE + 2;  
832      RRETURN(MATCH_THEN);      RRETURN(MATCH_THEN);
833    
834        /* Handle an atomic group that does not contain any capturing parentheses.
835        This can be handled like an assertion. Prior to 8.13, all atomic groups
836        were handled this way. In 8.13, the code was changed as below for ONCE, so
837        that backups pass through the group and thereby reset captured values.
838        However, this uses a lot more stack, so in 8.20, atomic groups that do not
839        contain any captures generate OP_ONCE_NC, which can be handled in the old,
840        less stack intensive way.
841    
842        Check the alternative branches in turn - the matching won't pass the KET
843        for this kind of subpattern. If any one branch matches, we carry on as at
844        the end of a normal bracket, leaving the subject pointer, but resetting
845        the start-of-match value in case it was changed by \K. */
846    
847        case OP_ONCE_NC:
848        prev = ecode;
849        saved_eptr = eptr;
850        save_mark = md->mark;
851        do
852          {
853          RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM64);
854          if (rrc == MATCH_MATCH)  /* Note: _not_ MATCH_ACCEPT */
855            {
856            mstart = md->start_match_ptr;
857            break;
858            }
859          if (rrc == MATCH_THEN)
860            {
861            next = ecode + GET(ecode,1);
862            if (md->start_match_ptr < next &&
863                (*ecode == OP_ALT || *next == OP_ALT))
864              rrc = MATCH_NOMATCH;
865            }
866    
867          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
868          ecode += GET(ecode,1);
869          md->mark = save_mark;
870          }
871        while (*ecode == OP_ALT);
872    
873        /* If hit the end of the group (which could be repeated), fail */
874    
875        if (*ecode != OP_ONCE_NC && *ecode != OP_ALT) RRETURN(MATCH_NOMATCH);
876    
877        /* Continue as from after the group, updating the offsets high water
878        mark, since extracts may have been taken. */
879    
880        do ecode += GET(ecode, 1); while (*ecode == OP_ALT);
881    
882        offset_top = md->end_offset_top;
883        eptr = md->end_match_ptr;
884    
885        /* For a non-repeating ket, just continue at this level. This also
886        happens for a repeating ket if no characters were matched in the group.
887        This is the forcible breaking of infinite loops as implemented in Perl
888        5.005. */
889    
890        if (*ecode == OP_KET || eptr == saved_eptr)
891          {
892          ecode += 1+LINK_SIZE;
893          break;
894          }
895    
896        /* The repeating kets try the rest of the pattern or restart from the
897        preceding bracket, in the appropriate order. The second "call" of match()
898        uses tail recursion, to avoid using another stack frame. */
899    
900        if (*ecode == OP_KETRMIN)
901          {
902          RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM65);
903          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
904          ecode = prev;
905          goto TAIL_RECURSE;
906          }
907        else  /* OP_KETRMAX */
908          {
909          RMATCH(eptr, prev, offset_top, md, eptrb, RM66);
910          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
911          ecode += 1 + LINK_SIZE;
912          goto TAIL_RECURSE;
913          }
914        /* Control never gets here */
915    
916      /* Handle a capturing bracket, other than those that are possessive with an      /* Handle a capturing bracket, other than those that are possessive with an
917      unlimited repeat. If there is space in the offset vector, save the current      unlimited repeat. If there is space in the offset vector, save the current
918      subject position in the working slot at the top of the vector. We mustn't      subject position in the working slot at the top of the vector. We mustn't
# Line 827  for (;;) Line 945  for (;;)
945        save_offset2 = md->offset_vector[offset+1];        save_offset2 = md->offset_vector[offset+1];
946        save_offset3 = md->offset_vector[md->offset_end - number];        save_offset3 = md->offset_vector[md->offset_end - number];
947        save_capture_last = md->capture_last;        save_capture_last = md->capture_last;
948          save_mark = md->mark;
949    
950        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));
951        md->offset_vector[md->offset_end - number] =        md->offset_vector[md->offset_end - number] =
# Line 835  for (;;) Line 954  for (;;)
954        for (;;)        for (;;)
955          {          {
956          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
957          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
958            eptrb, RM1);            eptrb, RM1);
959          if (rrc == MATCH_ONCE) break;  /* Backing up through an atomic group */          if (rrc == MATCH_ONCE) break;  /* Backing up through an atomic group */
960          if (rrc != MATCH_NOMATCH &&  
961              (rrc != MATCH_THEN || md->start_match_ptr != ecode))          /* If we backed up to a THEN, check whether it is within the current
962            RRETURN(rrc);          branch by comparing the address of the THEN that is passed back with
963            the end of the branch. If it is within the current branch, and the
964            branch is one of two or more alternatives (it either starts or ends
965            with OP_ALT), we have reached the limit of THEN's action, so convert
966            the return code to NOMATCH, which will cause normal backtracking to
967            happen from now on. Otherwise, THEN is passed back to an outer
968            alternative. This implements Perl's treatment of parenthesized groups,
969            where a group not containing | does not affect the current alternative,
970            that is, (X) is NOT the same as (X|(*F)). */
971    
972            if (rrc == MATCH_THEN)
973              {
974              next = ecode + GET(ecode,1);
975              if (md->start_match_ptr < next &&
976                  (*ecode == OP_ALT || *next == OP_ALT))
977                rrc = MATCH_NOMATCH;
978              }
979    
980            /* Anything other than NOMATCH is passed back. */
981    
982            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
983          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
984          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
985            md->mark = save_mark;
986          if (*ecode != OP_ALT) break;          if (*ecode != OP_ALT) break;
987          }          }
988    
# Line 851  for (;;) Line 991  for (;;)
991        md->offset_vector[offset+1] = save_offset2;        md->offset_vector[offset+1] = save_offset2;
992        md->offset_vector[md->offset_end - number] = save_offset3;        md->offset_vector[md->offset_end - number] = save_offset3;
993    
994        /* At this point, rrc will be one of MATCH_ONCE, MATCH_NOMATCH, or        /* At this point, rrc will be one of MATCH_ONCE or MATCH_NOMATCH. */
       MATCH_THEN. */  
995    
996        if (rrc != MATCH_THEN && md->mark == NULL) md->mark = markptr;        RRETURN(rrc);
       RRETURN(((rrc == MATCH_ONCE)? MATCH_ONCE:MATCH_NOMATCH));  
997        }        }
998    
999      /* FALL THROUGH ... Insufficient room for saving captured contents. Treat      /* FALL THROUGH ... Insufficient room for saving captured contents. Treat
# Line 870  for (;;) Line 1008  for (;;)
1008      /* VVVVVVVVVVVVVVVVVVVVVVVVV */      /* VVVVVVVVVVVVVVVVVVVVVVVVV */
1009    
1010      /* Non-capturing or atomic group, except for possessive with unlimited      /* Non-capturing or atomic group, except for possessive with unlimited
1011      repeat. Loop for all the alternatives. When we get to the final alternative      repeat and ONCE group with no captures. Loop for all the alternatives.
1012      within the brackets, we used to return the result of a recursive call to  
1013      match() whatever happened so it was possible to reduce stack usage by      When we get to the final alternative within the brackets, we used to return
1014      turning this into a tail recursion, except in the case of a possibly empty      the result of a recursive call to match() whatever happened so it was
1015      group. However, now that there is the possiblity of (*THEN) occurring in      possible to reduce stack usage by turning this into a tail recursion,
1016      the final alternative, this optimization is no longer possible.      except in the case of a possibly empty group. However, now that there is
1017        the possiblity of (*THEN) occurring in the final alternative, this
1018        optimization is no longer always possible.
1019    
1020        We can optimize if we know there are no (*THEN)s in the pattern; at present
1021        this is the best that can be done.
1022    
1023      MATCH_ONCE is returned when the end of an atomic group is successfully      MATCH_ONCE is returned when the end of an atomic group is successfully
1024      reached, but subsequent matching fails. It passes back up the tree (causing      reached, but subsequent matching fails. It passes back up the tree (causing
# Line 891  for (;;) Line 1034  for (;;)
1034    
1035      for (;;)      for (;;)
1036        {        {
1037        if (op >= OP_SBRA || op == OP_ONCE) md->match_function_type = MATCH_CBEGROUP;        if (op >= OP_SBRA || op == OP_ONCE)
1038        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, eptrb,          md->match_function_type = MATCH_CBEGROUP;
1039    
1040          /* If this is not a possibly empty group, and there are no (*THEN)s in
1041          the pattern, and this is the final alternative, optimize as described
1042          above. */
1043    
1044          else if (!md->hasthen && ecode[GET(ecode, 1)] != OP_ALT)
1045            {
1046            ecode += PRIV(OP_lengths)[*ecode];
1047            goto TAIL_RECURSE;
1048            }
1049    
1050          /* In all other cases, we have to make another call to match(). */
1051    
1052          save_mark = md->mark;
1053          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md, eptrb,
1054          RM2);          RM2);
1055        if (rrc != MATCH_NOMATCH &&  
1056            (rrc != MATCH_THEN || md->start_match_ptr != ecode))        /* See comment in the code for capturing groups above about handling
1057          THEN. */
1058    
1059          if (rrc == MATCH_THEN)
1060            {
1061            next = ecode + GET(ecode,1);
1062            if (md->start_match_ptr < next &&
1063                (*ecode == OP_ALT || *next == OP_ALT))
1064              rrc = MATCH_NOMATCH;
1065            }
1066    
1067          if (rrc != MATCH_NOMATCH)
1068          {          {
1069          if (rrc == MATCH_ONCE)          if (rrc == MATCH_ONCE)
1070            {            {
1071            const uschar *scode = ecode;            const pcre_uchar *scode = ecode;
1072            if (*scode != OP_ONCE)           /* If not at start, find it */            if (*scode != OP_ONCE)           /* If not at start, find it */
1073              {              {
1074              while (*scode == OP_ALT) scode += GET(scode, 1);              while (*scode == OP_ALT) scode += GET(scode, 1);
# Line 910  for (;;) Line 1079  for (;;)
1079          RRETURN(rrc);          RRETURN(rrc);
1080          }          }
1081        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1082          md->mark = save_mark;
1083        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1084        }        }
1085      if (rrc != MATCH_THEN && md->mark == NULL) md->mark = markptr;  
1086      RRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
1087    
1088      /* Handle possessive capturing brackets with an unlimited repeat. We come      /* Handle possessive capturing brackets with an unlimited repeat. We come
# Line 941  for (;;) Line 1111  for (;;)
1111      if (offset < md->offset_max)      if (offset < md->offset_max)
1112        {        {
1113        matched_once = FALSE;        matched_once = FALSE;
1114        code_offset = ecode - md->start_code;        code_offset = (int)(ecode - md->start_code);
1115    
1116        save_offset1 = md->offset_vector[offset];        save_offset1 = md->offset_vector[offset];
1117        save_offset2 = md->offset_vector[offset+1];        save_offset2 = md->offset_vector[offset+1];
# Line 964  for (;;) Line 1134  for (;;)
1134          md->offset_vector[md->offset_end - number] =          md->offset_vector[md->offset_end - number] =
1135            (int)(eptr - md->start_subject);            (int)(eptr - md->start_subject);
1136          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
1137          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
1138            eptrb, RM63);            eptrb, RM63);
1139          if (rrc == MATCH_KETRPOS)          if (rrc == MATCH_KETRPOS)
1140            {            {
# Line 975  for (;;) Line 1145  for (;;)
1145            matched_once = TRUE;            matched_once = TRUE;
1146            continue;            continue;
1147            }            }
1148          if (rrc != MATCH_NOMATCH &&  
1149              (rrc != MATCH_THEN || md->start_match_ptr != ecode))          /* See comment in the code for capturing groups above about handling
1150            RRETURN(rrc);          THEN. */
1151    
1152            if (rrc == MATCH_THEN)
1153              {
1154              next = ecode + GET(ecode,1);
1155              if (md->start_match_ptr < next &&
1156                  (*ecode == OP_ALT || *next == OP_ALT))
1157                rrc = MATCH_NOMATCH;
1158              }
1159    
1160            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1161          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
1162          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
1163          if (*ecode != OP_ALT) break;          if (*ecode != OP_ALT) break;
# Line 990  for (;;) Line 1170  for (;;)
1170          md->offset_vector[md->offset_end - number] = save_offset3;          md->offset_vector[md->offset_end - number] = save_offset3;
1171          }          }
1172    
       if (rrc != MATCH_THEN && md->mark == NULL) md->mark = markptr;  
1173        if (allow_zero || matched_once)        if (allow_zero || matched_once)
1174          {          {
1175          ecode += 1 + LINK_SIZE;          ecode += 1 + LINK_SIZE;
# Line 1022  for (;;) Line 1201  for (;;)
1201    
1202      POSSESSIVE_NON_CAPTURE:      POSSESSIVE_NON_CAPTURE:
1203      matched_once = FALSE;      matched_once = FALSE;
1204      code_offset = ecode - md->start_code;      code_offset = (int)(ecode - md->start_code);
1205    
1206      for (;;)      for (;;)
1207        {        {
1208        if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;        if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
1209        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
1210          eptrb, RM48);          eptrb, RM48);
1211        if (rrc == MATCH_KETRPOS)        if (rrc == MATCH_KETRPOS)
1212          {          {
# Line 1037  for (;;) Line 1216  for (;;)
1216          matched_once = TRUE;          matched_once = TRUE;
1217          continue;          continue;
1218          }          }
1219        if (rrc != MATCH_NOMATCH &&  
1220            (rrc != MATCH_THEN || md->start_match_ptr != ecode))        /* See comment in the code for capturing groups above about handling
1221          RRETURN(rrc);        THEN. */
1222    
1223          if (rrc == MATCH_THEN)
1224            {
1225            next = ecode + GET(ecode,1);
1226            if (md->start_match_ptr < next &&
1227                (*ecode == OP_ALT || *next == OP_ALT))
1228              rrc = MATCH_NOMATCH;
1229            }
1230    
1231          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1232        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1233        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1234        }        }
# Line 1067  for (;;) Line 1256  for (;;)
1256    
1257      if (ecode[LINK_SIZE+1] == OP_CALLOUT)      if (ecode[LINK_SIZE+1] == OP_CALLOUT)
1258        {        {
1259        if (pcre_callout != NULL)        if (PUBL(callout) != NULL)
1260          {          {
1261          pcre_callout_block cb;          PUBL(callout_block) cb;
1262          cb.version          = 2;   /* Version 1 of the callout block */          cb.version          = 2;   /* Version 1 of the callout block */
1263          cb.callout_number   = ecode[LINK_SIZE+2];          cb.callout_number   = ecode[LINK_SIZE+2];
1264          cb.offset_vector    = md->offset_vector;          cb.offset_vector    = md->offset_vector;
1265    #ifdef COMPILE_PCRE8
1266          cb.subject          = (PCRE_SPTR)md->start_subject;          cb.subject          = (PCRE_SPTR)md->start_subject;
1267    #else
1268            cb.subject          = (PCRE_SPTR16)md->start_subject;
1269    #endif
1270          cb.subject_length   = (int)(md->end_subject - md->start_subject);          cb.subject_length   = (int)(md->end_subject - md->start_subject);
1271          cb.start_match      = (int)(mstart - md->start_subject);          cb.start_match      = (int)(mstart - md->start_subject);
1272          cb.current_position = (int)(eptr - md->start_subject);          cb.current_position = (int)(eptr - md->start_subject);
# Line 1082  for (;;) Line 1275  for (;;)
1275          cb.capture_top      = offset_top/2;          cb.capture_top      = offset_top/2;
1276          cb.capture_last     = md->capture_last;          cb.capture_last     = md->capture_last;
1277          cb.callout_data     = md->callout_data;          cb.callout_data     = md->callout_data;
1278          cb.mark             = markptr;          cb.mark             = md->nomatch_mark;
1279          if ((rrc = (*pcre_callout)(&cb)) > 0) MRRETURN(MATCH_NOMATCH);          if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1280          if (rrc < 0) RRETURN(rrc);          if (rrc < 0) RRETURN(rrc);
1281          }          }
1282        ecode += _pcre_OP_lengths[OP_CALLOUT];        ecode += PRIV(OP_lengths)[OP_CALLOUT];
1283        }        }
1284    
1285      condcode = ecode[LINK_SIZE+1];      condcode = ecode[LINK_SIZE+1];
# Line 1103  for (;;) Line 1296  for (;;)
1296        else        else
1297          {          {
1298          int recno = GET2(ecode, LINK_SIZE + 2);   /* Recursion group number*/          int recno = GET2(ecode, LINK_SIZE + 2);   /* Recursion group number*/
1299          condition =  (recno == RREF_ANY || recno == md->recursive->group_num);          condition = (recno == RREF_ANY || recno == md->recursive->group_num);
1300    
1301          /* If the test is for recursion into a specific subpattern, and it is          /* If the test is for recursion into a specific subpattern, and it is
1302          false, but the test was set up by name, scan the table to see if the          false, but the test was set up by name, scan the table to see if the
1303          name refers to any other numbers, and test them. The condition is true          name refers to any other numbers, and test them. The condition is true
1304          if any one is set. */          if any one is set. */
1305    
1306          if (!condition && condcode == OP_NRREF && recno != RREF_ANY)          if (!condition && condcode == OP_NRREF)
1307            {            {
1308            uschar *slotA = md->name_table;            pcre_uchar *slotA = md->name_table;
1309            for (i = 0; i < md->name_count; i++)            for (i = 0; i < md->name_count; i++)
1310              {              {
1311              if (GET2(slotA, 0) == recno) break;              if (GET2(slotA, 0) == recno) break;
# Line 1125  for (;;) Line 1318  for (;;)
1318    
1319            if (i < md->name_count)            if (i < md->name_count)
1320              {              {
1321              uschar *slotB = slotA;              pcre_uchar *slotB = slotA;
1322              while (slotB > md->name_table)              while (slotB > md->name_table)
1323                {                {
1324                slotB -= md->name_entry_size;                slotB -= md->name_entry_size;
1325                if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)                if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1326                  {                  {
1327                  condition = GET2(slotB, 0) == md->recursive->group_num;                  condition = GET2(slotB, 0) == md->recursive->group_num;
1328                  if (condition) break;                  if (condition) break;
# Line 1145  for (;;) Line 1338  for (;;)
1338                for (i++; i < md->name_count; i++)                for (i++; i < md->name_count; i++)
1339                  {                  {
1340                  slotB += md->name_entry_size;                  slotB += md->name_entry_size;
1341                  if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)                  if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1342                    {                    {
1343                    condition = GET2(slotB, 0) == md->recursive->group_num;                    condition = GET2(slotB, 0) == md->recursive->group_num;
1344                    if (condition) break;                    if (condition) break;
# Line 1158  for (;;) Line 1351  for (;;)
1351    
1352          /* Chose branch according to the condition */          /* Chose branch according to the condition */
1353    
1354          ecode += condition? 3 : GET(ecode, 1);          ecode += condition? 1 + IMM2_SIZE : GET(ecode, 1);
1355          }          }
1356        }        }
1357    
# Line 1175  for (;;) Line 1368  for (;;)
1368        if (!condition && condcode == OP_NCREF)        if (!condition && condcode == OP_NCREF)
1369          {          {
1370          int refno = offset >> 1;          int refno = offset >> 1;
1371          uschar *slotA = md->name_table;          pcre_uchar *slotA = md->name_table;
1372    
1373          for (i = 0; i < md->name_count; i++)          for (i = 0; i < md->name_count; i++)
1374            {            {
# Line 1189  for (;;) Line 1382  for (;;)
1382    
1383          if (i < md->name_count)          if (i < md->name_count)
1384            {            {
1385            uschar *slotB = slotA;            pcre_uchar *slotB = slotA;
1386            while (slotB > md->name_table)            while (slotB > md->name_table)
1387              {              {
1388              slotB -= md->name_entry_size;              slotB -= md->name_entry_size;
1389              if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)              if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1390                {                {
1391                offset = GET2(slotB, 0) << 1;                offset = GET2(slotB, 0) << 1;
1392                condition = offset < offset_top &&                condition = offset < offset_top &&
# Line 1211  for (;;) Line 1404  for (;;)
1404              for (i++; i < md->name_count; i++)              for (i++; i < md->name_count; i++)
1405                {                {
1406                slotB += md->name_entry_size;                slotB += md->name_entry_size;
1407                if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)                if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1408                  {                  {
1409                  offset = GET2(slotB, 0) << 1;                  offset = GET2(slotB, 0) << 1;
1410                  condition = offset < offset_top &&                  condition = offset < offset_top &&
# Line 1226  for (;;) Line 1419  for (;;)
1419    
1420        /* Chose branch according to the condition */        /* Chose branch according to the condition */
1421    
1422        ecode += condition? 3 : GET(ecode, 1);        ecode += condition? 1 + IMM2_SIZE : GET(ecode, 1);
1423        }        }
1424    
1425      else if (condcode == OP_DEF)     /* DEFINE - always false */      else if (condcode == OP_DEF)     /* DEFINE - always false */
# Line 1251  for (;;) Line 1444  for (;;)
1444          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);
1445          while (*ecode == OP_ALT) ecode += GET(ecode, 1);          while (*ecode == OP_ALT) ecode += GET(ecode, 1);
1446          }          }
1447        else if (rrc != MATCH_NOMATCH &&  
1448                (rrc != MATCH_THEN || md->start_match_ptr != ecode))        /* PCRE doesn't allow the effect of (*THEN) to escape beyond an
1449          assertion; it is therefore treated as NOMATCH. */
1450    
1451          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
1452          {          {
1453          RRETURN(rrc);         /* Need braces because of following else */          RRETURN(rrc);         /* Need braces because of following else */
1454          }          }
# Line 1263  for (;;) Line 1459  for (;;)
1459          }          }
1460        }        }
1461    
1462      /* 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, can
1463      we used to use tail recursion to avoid using another stack frame, except      use tail recursion to avoid using another stack frame, except when there is
1464      when there was unlimited repeat of a possibly empty group. However, that      unlimited repeat of a possibly empty group. In the latter case, a recursive
1465      strategy no longer works because of the possibilty of (*THEN) being      call to match() is always required, unless the second alternative doesn't
1466      encountered in the branch. A recursive call to match() is always required,      exist, in which case we can just plough on. Note that, for compatibility
1467      unless the second alternative doesn't exist, in which case we can just      with Perl, the | in a conditional group is NOT treated as creating two
1468      plough on. */      alternatives. If a THEN is encountered in the branch, it propagates out to
1469        the enclosing alternative (unless nested in a deeper set of alternatives,
1470        of course). */
1471    
1472      if (condition || *ecode == OP_ALT)      if (condition || *ecode == OP_ALT)
1473        {        {
1474        if (op == OP_SCOND) md->match_function_type = MATCH_CBEGROUP;        if (op != OP_SCOND)
1475            {
1476            ecode += 1 + LINK_SIZE;
1477            goto TAIL_RECURSE;
1478            }
1479    
1480          md->match_function_type = MATCH_CBEGROUP;
1481        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM49);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM49);
       if (rrc == MATCH_THEN && md->start_match_ptr == ecode)  
         rrc = MATCH_NOMATCH;  
1482        RRETURN(rrc);        RRETURN(rrc);
1483        }        }
1484      else                         /* Condition false & no alternative */  
1485         /* Condition false & no alternative; continue after the group. */
1486    
1487        else
1488        {        {
1489        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1490        }        }
# Line 1306  for (;;) Line 1511  for (;;)
1511        md->offset_vector[offset+1] = (int)(eptr - md->start_subject);        md->offset_vector[offset+1] = (int)(eptr - md->start_subject);
1512        if (offset_top <= offset) offset_top = offset + 2;        if (offset_top <= offset) offset_top = offset + 2;
1513        }        }
1514      ecode += 3;      ecode += 1 + IMM2_SIZE;
1515      break;      break;
1516    
1517    
# Line 1326  for (;;) Line 1531  for (;;)
1531           (md->notempty ||           (md->notempty ||
1532             (md->notempty_atstart &&             (md->notempty_atstart &&
1533               mstart == md->start_subject + md->start_offset)))               mstart == md->start_subject + md->start_offset)))
1534        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
1535    
1536      /* Otherwise, we have a match. */      /* Otherwise, we have a match. */
1537    
# Line 1335  for (;;) Line 1540  for (;;)
1540      md->start_match_ptr = mstart;       /* and the start (\K can modify) */      md->start_match_ptr = mstart;       /* and the start (\K can modify) */
1541    
1542      /* For some reason, the macros don't work properly if an expression is      /* For some reason, the macros don't work properly if an expression is
1543      given as the argument to MRRETURN when the heap is in use. */      given as the argument to RRETURN when the heap is in use. */
1544    
1545      rrc = (op == OP_END)? MATCH_MATCH : MATCH_ACCEPT;      rrc = (op == OP_END)? MATCH_MATCH : MATCH_ACCEPT;
1546      MRRETURN(rrc);      RRETURN(rrc);
1547    
1548      /* Assertion brackets. Check the alternative branches in turn - the      /* Assertion brackets. Check the alternative branches in turn - the
1549      matching won't pass the KET for an assertion. If any one branch matches,      matching won't pass the KET for an assertion. If any one branch matches,
# Line 1353  for (;;) Line 1558  for (;;)
1558    
1559      case OP_ASSERT:      case OP_ASSERT:
1560      case OP_ASSERTBACK:      case OP_ASSERTBACK:
1561        save_mark = md->mark;
1562      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1563        {        {
1564        condassert = TRUE;        condassert = TRUE;
# Line 1366  for (;;) Line 1572  for (;;)
1572        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1573          {          {
1574          mstart = md->start_match_ptr;   /* In case \K reset it */          mstart = md->start_match_ptr;   /* In case \K reset it */
         markptr = md->mark;  
1575          break;          break;
1576          }          }
1577        if (rrc != MATCH_NOMATCH &&        md->mark = save_mark;
1578            (rrc != MATCH_THEN || md->start_match_ptr != ecode))  
1579          RRETURN(rrc);        /* A COMMIT failure must fail the entire assertion, without trying any
1580          subsequent branches. */
1581    
1582          if (rrc == MATCH_COMMIT) RRETURN(MATCH_NOMATCH);
1583    
1584          /* PCRE does not allow THEN to escape beyond an assertion; it
1585          is treated as NOMATCH. */
1586    
1587          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1588        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1589        }        }
1590      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
1591    
1592      if (*ecode == OP_KET) MRRETURN(MATCH_NOMATCH);      if (*ecode == OP_KET) RRETURN(MATCH_NOMATCH);
1593    
1594      /* If checking an assertion for a condition, return MATCH_MATCH. */      /* If checking an assertion for a condition, return MATCH_MATCH. */
1595    
# Line 1396  for (;;) Line 1609  for (;;)
1609    
1610      case OP_ASSERT_NOT:      case OP_ASSERT_NOT:
1611      case OP_ASSERTBACK_NOT:      case OP_ASSERTBACK_NOT:
1612        save_mark = md->mark;
1613      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1614        {        {
1615        condassert = TRUE;        condassert = TRUE;
# Line 1406  for (;;) Line 1620  for (;;)
1620      do      do
1621        {        {
1622        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);
1623        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) MRRETURN(MATCH_NOMATCH);        md->mark = save_mark;
1624          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) RRETURN(MATCH_NOMATCH);
1625        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)
1626          {          {
1627          do ecode += GET(ecode,1); while (*ecode == OP_ALT);          do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1628          break;          break;
1629          }          }
1630        if (rrc != MATCH_NOMATCH &&  
1631            (rrc != MATCH_THEN || md->start_match_ptr != ecode))        /* PCRE does not allow THEN to escape beyond an assertion; it is treated
1632          RRETURN(rrc);        as NOMATCH. */
1633    
1634          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1635        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1636        }        }
1637      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 1430  for (;;) Line 1647  for (;;)
1647      back a number of characters, not bytes. */      back a number of characters, not bytes. */
1648    
1649      case OP_REVERSE:      case OP_REVERSE:
1650  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
1651      if (utf8)      if (utf)
1652        {        {
1653        i = GET(ecode, 1);        i = GET(ecode, 1);
1654        while (i-- > 0)        while (i-- > 0)
1655          {          {
1656          eptr--;          eptr--;
1657          if (eptr < md->start_subject) MRRETURN(MATCH_NOMATCH);          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1658          BACKCHAR(eptr);          BACKCHAR(eptr);
1659          }          }
1660        }        }
# Line 1448  for (;;) Line 1665  for (;;)
1665    
1666        {        {
1667        eptr -= GET(ecode, 1);        eptr -= GET(ecode, 1);
1668        if (eptr < md->start_subject) MRRETURN(MATCH_NOMATCH);        if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1669        }        }
1670    
1671      /* Save the earliest consulted character, then skip to next op code */      /* Save the earliest consulted character, then skip to next op code */
# Line 1462  for (;;) Line 1679  for (;;)
1679      function is able to force a failure. */      function is able to force a failure. */
1680    
1681      case OP_CALLOUT:      case OP_CALLOUT:
1682      if (pcre_callout != NULL)      if (PUBL(callout) != NULL)
1683        {        {
1684        pcre_callout_block cb;        PUBL(callout_block) cb;
1685        cb.version          = 2;   /* Version 1 of the callout block */        cb.version          = 2;   /* Version 1 of the callout block */
1686        cb.callout_number   = ecode[1];        cb.callout_number   = ecode[1];
1687        cb.offset_vector    = md->offset_vector;        cb.offset_vector    = md->offset_vector;
1688    #ifdef COMPILE_PCRE8
1689        cb.subject          = (PCRE_SPTR)md->start_subject;        cb.subject          = (PCRE_SPTR)md->start_subject;
1690    #else
1691          cb.subject          = (PCRE_SPTR16)md->start_subject;
1692    #endif
1693        cb.subject_length   = (int)(md->end_subject - md->start_subject);        cb.subject_length   = (int)(md->end_subject - md->start_subject);
1694        cb.start_match      = (int)(mstart - md->start_subject);        cb.start_match      = (int)(mstart - md->start_subject);
1695        cb.current_position = (int)(eptr - md->start_subject);        cb.current_position = (int)(eptr - md->start_subject);
# Line 1477  for (;;) Line 1698  for (;;)
1698        cb.capture_top      = offset_top/2;        cb.capture_top      = offset_top/2;
1699        cb.capture_last     = md->capture_last;        cb.capture_last     = md->capture_last;
1700        cb.callout_data     = md->callout_data;        cb.callout_data     = md->callout_data;
1701        cb.mark             = markptr;        cb.mark             = md->nomatch_mark;
1702        if ((rrc = (*pcre_callout)(&cb)) > 0) MRRETURN(MATCH_NOMATCH);        if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1703        if (rrc < 0) RRETURN(rrc);        if (rrc < 0) RRETURN(rrc);
1704        }        }
1705      ecode += 2 + 2*LINK_SIZE;      ecode += 2 + 2*LINK_SIZE;
# Line 1505  for (;;) Line 1726  for (;;)
1726        {        {
1727        recursion_info *ri;        recursion_info *ri;
1728        int recno;        int recno;
1729    
1730        callpat = md->start_code + GET(ecode, 1);        callpat = md->start_code + GET(ecode, 1);
1731        recno = (callpat == md->start_code)? 0 :        recno = (callpat == md->start_code)? 0 :
1732          GET2(callpat, 1 + LINK_SIZE);          GET2(callpat, 1 + LINK_SIZE);
1733    
1734        /* Check for repeating a recursion without advancing the subject pointer.        /* Check for repeating a recursion without advancing the subject pointer.
1735        This should catch convoluted mutual recursions. (Some simple cases are        This should catch convoluted mutual recursions. (Some simple cases are
1736        caught at compile time.) */        caught at compile time.) */
1737    
1738        for (ri = md->recursive; ri != NULL; ri = ri->prevrec)        for (ri = md->recursive; ri != NULL; ri = ri->prevrec)
1739          if (recno == ri->group_num && eptr == ri->subject_position)          if (recno == ri->group_num && eptr == ri->subject_position)
1740            RRETURN(PCRE_ERROR_RECURSELOOP);            RRETURN(PCRE_ERROR_RECURSELOOP);
1741    
1742        /* Add to "recursing stack" */        /* Add to "recursing stack" */
# Line 1537  for (;;) Line 1758  for (;;)
1758        else        else
1759          {          {
1760          new_recursive.offset_save =          new_recursive.offset_save =
1761            (int *)(pcre_malloc)(new_recursive.saved_max * sizeof(int));            (int *)(PUBL(malloc))(new_recursive.saved_max * sizeof(int));
1762          if (new_recursive.offset_save == NULL) RRETURN(PCRE_ERROR_NOMEMORY);          if (new_recursive.offset_save == NULL) RRETURN(PCRE_ERROR_NOMEMORY);
1763          }          }
1764        memcpy(new_recursive.offset_save, md->offset_vector,        memcpy(new_recursive.offset_save, md->offset_vector,
# Line 1552  for (;;) Line 1773  for (;;)
1773        do        do
1774          {          {
1775          if (cbegroup) md->match_function_type = MATCH_CBEGROUP;          if (cbegroup) md->match_function_type = MATCH_CBEGROUP;
1776          RMATCH(eptr, callpat + _pcre_OP_lengths[*callpat], offset_top,          RMATCH(eptr, callpat + PRIV(OP_lengths)[*callpat], offset_top,
1777            md, eptrb, RM6);            md, eptrb, RM6);
1778          memcpy(md->offset_vector, new_recursive.offset_save,          memcpy(md->offset_vector, new_recursive.offset_save,
1779              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1780            md->recursive = new_recursive.prevrec;
1781          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1782            {            {
1783            DPRINTF(("Recursion matched\n"));            DPRINTF(("Recursion matched\n"));
           md->recursive = new_recursive.prevrec;  
1784            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
1785              (pcre_free)(new_recursive.offset_save);              (PUBL(free))(new_recursive.offset_save);
1786    
1787            /* Set where we got to in the subject, and reset the start in case            /* Set where we got to in the subject, and reset the start in case
1788            it was changed by \K. This *is* propagated back out of a recursion,            it was changed by \K. This *is* propagated back out of a recursion,
# Line 1571  for (;;) Line 1792  for (;;)
1792            mstart = md->start_match_ptr;            mstart = md->start_match_ptr;
1793            goto RECURSION_MATCHED;        /* Exit loop; end processing */            goto RECURSION_MATCHED;        /* Exit loop; end processing */
1794            }            }
1795          else if (rrc != MATCH_NOMATCH &&  
1796                  (rrc != MATCH_THEN || md->start_match_ptr != ecode))          /* PCRE does not allow THEN or COMMIT to escape beyond a recursion; it
1797            is treated as NOMATCH. */
1798    
1799            else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN &&
1800                     rrc != MATCH_COMMIT)
1801            {            {
1802            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1803            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
1804              (pcre_free)(new_recursive.offset_save);              (PUBL(free))(new_recursive.offset_save);
1805            RRETURN(rrc);            RRETURN(rrc);
1806            }            }
1807    
# Line 1588  for (;;) Line 1813  for (;;)
1813        DPRINTF(("Recursion didn't match\n"));        DPRINTF(("Recursion didn't match\n"));
1814        md->recursive = new_recursive.prevrec;        md->recursive = new_recursive.prevrec;
1815        if (new_recursive.offset_save != stacksave)        if (new_recursive.offset_save != stacksave)
1816          (pcre_free)(new_recursive.offset_save);          (PUBL(free))(new_recursive.offset_save);
1817        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
1818        }        }
1819    
1820      RECURSION_MATCHED:      RECURSION_MATCHED:
# Line 1658  for (;;) Line 1883  for (;;)
1883        }        }
1884      else saved_eptr = NULL;      else saved_eptr = NULL;
1885    
1886      /* If we are at the end of an assertion group, stop matching and return      /* If we are at the end of an assertion group or a non-capturing atomic
1887      MATCH_MATCH, but record the current high water mark for use by positive      group, stop matching and return MATCH_MATCH, but record the current high
1888      assertions. We also need to record the match start in case it was changed      water mark for use by positive assertions. We also need to record the match
1889      by \K. */      start in case it was changed by \K. */
1890    
1891      if (*prev == OP_ASSERT || *prev == OP_ASSERT_NOT ||      if ((*prev >= OP_ASSERT && *prev <= OP_ASSERTBACK_NOT) ||
1892          *prev == OP_ASSERTBACK || *prev == OP_ASSERTBACK_NOT)           *prev == OP_ONCE_NC)
1893        {        {
1894        md->end_match_ptr = eptr;      /* For ONCE */        md->end_match_ptr = eptr;      /* For ONCE_NC */
1895        md->end_offset_top = offset_top;        md->end_offset_top = offset_top;
1896        md->start_match_ptr = mstart;        md->start_match_ptr = mstart;
1897        MRRETURN(MATCH_MATCH);         /* Sets md->mark */        RRETURN(MATCH_MATCH);         /* Sets md->mark */
1898        }        }
1899    
1900      /* For capturing groups we have to check the group number back at the start      /* For capturing groups we have to check the group number back at the start
# Line 1734  for (;;) Line 1959  for (;;)
1959      /* For an ordinary non-repeating ket, just continue at this level. This      /* For an ordinary non-repeating ket, just continue at this level. This
1960      also happens for a repeating ket if no characters were matched in the      also happens for a repeating ket if no characters were matched in the
1961      group. This is the forcible breaking of infinite loops as implemented in      group. This is the forcible breaking of infinite loops as implemented in
1962      Perl 5.005. For a non-repeating atomic group, establish a backup point by      Perl 5.005. For a non-repeating atomic group that includes captures,
1963      processing the rest of the pattern at a lower level. If this results in a      establish a backup point by processing the rest of the pattern at a lower
1964      NOMATCH return, pass MATCH_ONCE back to the original OP_ONCE level, thereby      level. If this results in a NOMATCH return, pass MATCH_ONCE back to the
1965      bypassing intermediate backup points, but resetting any captures that      original OP_ONCE level, thereby bypassing intermediate backup points, but
1966      happened along the way. */      resetting any captures that happened along the way. */
1967    
1968      if (*ecode == OP_KET || eptr == saved_eptr)      if (*ecode == OP_KET || eptr == saved_eptr)
1969        {        {
# Line 1783  for (;;) Line 2008  for (;;)
2008          }          }
2009        if (*prev >= OP_SBRA)    /* Could match an empty string */        if (*prev >= OP_SBRA)    /* Could match an empty string */
2010          {          {
         md->match_function_type = MATCH_CBEGROUP;  
2011          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);
2012          RRETURN(rrc);          RRETURN(rrc);
2013          }          }
# Line 1792  for (;;) Line 2016  for (;;)
2016        }        }
2017      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
2018        {        {
       if (*prev >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;  
2019        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);
2020        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;
2021        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
# Line 1811  for (;;) Line 2034  for (;;)
2034      /* Not multiline mode: start of subject assertion, unless notbol. */      /* Not multiline mode: start of subject assertion, unless notbol. */
2035    
2036      case OP_CIRC:      case OP_CIRC:
2037      if (md->notbol && eptr == md->start_subject) MRRETURN(MATCH_NOMATCH);      if (md->notbol && eptr == md->start_subject) RRETURN(MATCH_NOMATCH);
2038    
2039      /* Start of subject assertion */      /* Start of subject assertion */
2040    
2041      case OP_SOD:      case OP_SOD:
2042      if (eptr != md->start_subject) MRRETURN(MATCH_NOMATCH);      if (eptr != md->start_subject) RRETURN(MATCH_NOMATCH);
2043      ecode++;      ecode++;
2044      break;      break;
2045    
2046      /* Multiline mode: start of subject unless notbol, or after any newline. */      /* Multiline mode: start of subject unless notbol, or after any newline. */
2047    
2048      case OP_CIRCM:      case OP_CIRCM:
2049      if (md->notbol && eptr == md->start_subject) MRRETURN(MATCH_NOMATCH);      if (md->notbol && eptr == md->start_subject) RRETURN(MATCH_NOMATCH);
2050      if (eptr != md->start_subject &&      if (eptr != md->start_subject &&
2051          (eptr == md->end_subject || !WAS_NEWLINE(eptr)))          (eptr == md->end_subject || !WAS_NEWLINE(eptr)))
2052        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2053      ecode++;      ecode++;
2054      break;      break;
2055    
2056      /* Start of match assertion */      /* Start of match assertion */
2057    
2058      case OP_SOM:      case OP_SOM:
2059      if (eptr != md->start_subject + md->start_offset) MRRETURN(MATCH_NOMATCH);      if (eptr != md->start_subject + md->start_offset) RRETURN(MATCH_NOMATCH);
2060      ecode++;      ecode++;
2061      break;      break;
2062    
# Line 1849  for (;;) Line 2072  for (;;)
2072    
2073      case OP_DOLLM:      case OP_DOLLM:
2074      if (eptr < md->end_subject)      if (eptr < md->end_subject)
2075        { if (!IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH); }        {
2076          if (!IS_NEWLINE(eptr))
2077            {
2078            if (md->partial != 0 &&
2079                eptr + 1 >= md->end_subject &&
2080                NLBLOCK->nltype == NLTYPE_FIXED &&
2081                NLBLOCK->nllen == 2 &&
2082                *eptr == NLBLOCK->nl[0])
2083              {
2084              md->hitend = TRUE;
2085              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2086              }
2087            RRETURN(MATCH_NOMATCH);
2088            }
2089          }
2090      else      else
2091        {        {
2092        if (md->noteol) MRRETURN(MATCH_NOMATCH);        if (md->noteol) RRETURN(MATCH_NOMATCH);
2093        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2094        }        }
2095      ecode++;      ecode++;
# Line 1862  for (;;) Line 2099  for (;;)
2099      subject unless noteol is set. */      subject unless noteol is set. */
2100    
2101      case OP_DOLL:      case OP_DOLL:
2102      if (md->noteol) MRRETURN(MATCH_NOMATCH);      if (md->noteol) RRETURN(MATCH_NOMATCH);
2103      if (!md->endonly) goto ASSERT_NL_OR_EOS;      if (!md->endonly) goto ASSERT_NL_OR_EOS;
2104    
2105      /* ... else fall through for endonly */      /* ... else fall through for endonly */
# Line 1870  for (;;) Line 2107  for (;;)
2107      /* End of subject assertion (\z) */      /* End of subject assertion (\z) */
2108    
2109      case OP_EOD:      case OP_EOD:
2110      if (eptr < md->end_subject) MRRETURN(MATCH_NOMATCH);      if (eptr < md->end_subject) RRETURN(MATCH_NOMATCH);
2111      SCHECK_PARTIAL();      SCHECK_PARTIAL();
2112      ecode++;      ecode++;
2113      break;      break;
# Line 1881  for (;;) Line 2118  for (;;)
2118      ASSERT_NL_OR_EOS:      ASSERT_NL_OR_EOS:
2119      if (eptr < md->end_subject &&      if (eptr < md->end_subject &&
2120          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))
2121        MRRETURN(MATCH_NOMATCH);        {
2122          if (md->partial != 0 &&
2123              eptr + 1 >= md->end_subject &&
2124              NLBLOCK->nltype == NLTYPE_FIXED &&
2125              NLBLOCK->nllen == 2 &&
2126              *eptr == NLBLOCK->nl[0])
2127            {
2128            md->hitend = TRUE;
2129            if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2130            }
2131          RRETURN(MATCH_NOMATCH);
2132          }
2133    
2134      /* Either at end of string or \n before end. */      /* Either at end of string or \n before end. */
2135    
# Line 1900  for (;;) Line 2148  for (;;)
2148        be "non-word" characters. Remember the earliest consulted character for        be "non-word" characters. Remember the earliest consulted character for
2149        partial matching. */        partial matching. */
2150    
2151  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2152        if (utf8)        if (utf)
2153          {          {
2154          /* Get status of previous character */          /* Get status of previous character */
2155    
2156          if (eptr == md->start_subject) prev_is_word = FALSE; else          if (eptr == md->start_subject) prev_is_word = FALSE; else
2157            {            {
2158            USPTR lastptr = eptr - 1;            PCRE_PUCHAR lastptr = eptr - 1;
2159            while((*lastptr & 0xc0) == 0x80) lastptr--;            BACKCHAR(lastptr);
2160            if (lastptr < md->start_used_ptr) md->start_used_ptr = lastptr;            if (lastptr < md->start_used_ptr) md->start_used_ptr = lastptr;
2161            GETCHAR(c, lastptr);            GETCHAR(c, lastptr);
2162  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
# Line 1973  for (;;) Line 2221  for (;;)
2221              }              }
2222            else            else
2223  #endif  #endif
2224            prev_is_word = ((md->ctypes[eptr[-1]] & ctype_word) != 0);            prev_is_word = MAX_255(eptr[-1])
2225                && ((md->ctypes[eptr[-1]] & ctype_word) != 0);
2226            }            }
2227    
2228          /* Get status of next character */          /* Get status of next character */
# Line 1996  for (;;) Line 2245  for (;;)
2245            }            }
2246          else          else
2247  #endif  #endif
2248          cur_is_word = ((md->ctypes[*eptr] & ctype_word) != 0);          cur_is_word = MAX_255(*eptr)
2249              && ((md->ctypes[*eptr] & ctype_word) != 0);
2250          }          }
2251    
2252        /* Now see if the situation is what we want */        /* Now see if the situation is what we want */
2253    
2254        if ((*ecode++ == OP_WORD_BOUNDARY)?        if ((*ecode++ == OP_WORD_BOUNDARY)?
2255             cur_is_word == prev_is_word : cur_is_word != prev_is_word)             cur_is_word == prev_is_word : cur_is_word != prev_is_word)
2256          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2257        }        }
2258      break;      break;
2259    
2260      /* Match a single character type; inline for speed */      /* Match any single character type except newline; have to take care with
2261        CRLF newlines and partial matching. */
2262    
2263      case OP_ANY:      case OP_ANY:
2264      if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
2265        if (md->partial != 0 &&
2266            eptr + 1 >= md->end_subject &&
2267            NLBLOCK->nltype == NLTYPE_FIXED &&
2268            NLBLOCK->nllen == 2 &&
2269            *eptr == NLBLOCK->nl[0])
2270          {
2271          md->hitend = TRUE;
2272          if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2273          }
2274    
2275      /* Fall through */      /* Fall through */
2276    
2277        /* Match any single character whatsoever. */
2278    
2279      case OP_ALLANY:      case OP_ALLANY:
2280      if (eptr >= md->end_subject)   /* DO NOT merge the eptr++ here; it must */      if (eptr >= md->end_subject)   /* DO NOT merge the eptr++ here; it must */
2281        {                            /* not be updated before SCHECK_PARTIAL. */        {                            /* not be updated before SCHECK_PARTIAL. */
2282        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2283        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2284        }        }
2285      eptr++;      eptr++;
2286      if (utf8) while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  #ifdef SUPPORT_UTF
2287        if (utf) ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
2288    #endif
2289      ecode++;      ecode++;
2290      break;      break;
2291    
# Line 2031  for (;;) Line 2296  for (;;)
2296      if (eptr >= md->end_subject)   /* DO NOT merge the eptr++ here; it must */      if (eptr >= md->end_subject)   /* DO NOT merge the eptr++ here; it must */
2297        {                            /* not be updated before SCHECK_PARTIAL. */        {                            /* not be updated before SCHECK_PARTIAL. */
2298        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2299        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2300        }        }
2301      eptr++;      eptr++;
2302      ecode++;      ecode++;
2303      break;      break;
2304    
# Line 2041  for (;;) Line 2306  for (;;)
2306      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2307        {        {
2308        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2309        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2310        }        }
2311      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2312      if (      if (
2313  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2314         c < 256 &&         c < 256 &&
2315  #endif  #endif
2316         (md->ctypes[c] & ctype_digit) != 0         (md->ctypes[c] & ctype_digit) != 0
2317         )         )
2318        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2319      ecode++;      ecode++;
2320      break;      break;
2321    
# Line 2058  for (;;) Line 2323  for (;;)
2323      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2324        {        {
2325        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2326        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2327        }        }
2328      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2329      if (      if (
2330  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2331         c >= 256 ||         c > 255 ||
2332  #endif  #endif
2333         (md->ctypes[c] & ctype_digit) == 0         (md->ctypes[c] & ctype_digit) == 0
2334         )         )
2335        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2336      ecode++;      ecode++;
2337      break;      break;
2338    
# Line 2075  for (;;) Line 2340  for (;;)
2340      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2341        {        {
2342        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2343        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2344        }        }
2345      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2346      if (      if (
2347  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2348         c < 256 &&         c < 256 &&
2349  #endif  #endif
2350         (md->ctypes[c] & ctype_space) != 0         (md->ctypes[c] & ctype_space) != 0
2351         )         )
2352        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2353      ecode++;      ecode++;
2354      break;      break;
2355    
# Line 2092  for (;;) Line 2357  for (;;)
2357      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2358        {        {
2359        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2360        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2361        }        }
2362      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2363      if (      if (
2364  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2365         c >= 256 ||         c > 255 ||
2366  #endif  #endif
2367         (md->ctypes[c] & ctype_space) == 0         (md->ctypes[c] & ctype_space) == 0
2368         )         )
2369        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2370      ecode++;      ecode++;
2371      break;      break;
2372    
# Line 2109  for (;;) Line 2374  for (;;)
2374      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2375        {        {
2376        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2377        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2378        }        }
2379      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2380      if (      if (
2381  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2382         c < 256 &&         c < 256 &&
2383  #endif  #endif
2384         (md->ctypes[c] & ctype_word) != 0         (md->ctypes[c] & ctype_word) != 0
2385         )         )
2386        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2387      ecode++;      ecode++;
2388      break;      break;
2389    
# Line 2126  for (;;) Line 2391  for (;;)
2391      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2392        {        {
2393        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2394        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2395        }        }
2396      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2397      if (      if (
2398  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2399         c >= 256 ||         c > 255 ||
2400  #endif  #endif
2401         (md->ctypes[c] & ctype_word) == 0         (md->ctypes[c] & ctype_word) == 0
2402         )         )
2403        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2404      ecode++;      ecode++;
2405      break;      break;
2406    
# Line 2143  for (;;) Line 2408  for (;;)
2408      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2409        {        {
2410        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2411        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2412        }        }
2413      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2414      switch(c)      switch(c)
2415        {        {
2416        default: MRRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2417    
2418        case 0x000d:        case CHAR_CR:
2419        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr >= md->end_subject)
2420            {
2421            SCHECK_PARTIAL();
2422            }
2423          else if (*eptr == CHAR_LF) eptr++;
2424        break;        break;
2425    
2426        case 0x000a:        case CHAR_LF:
2427        break;        break;
2428    
2429        case 0x000b:        case CHAR_VT:
2430        case 0x000c:        case CHAR_FF:
2431        case 0x0085:        case CHAR_NEL:
2432    #ifndef EBCDIC
2433        case 0x2028:        case 0x2028:
2434        case 0x2029:        case 0x2029:
2435        if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);  #endif  /* Not EBCDIC */
2436          if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
2437        break;        break;
2438        }        }
2439      ecode++;      ecode++;
# Line 2172  for (;;) Line 2443  for (;;)
2443      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2444        {        {
2445        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2446        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2447        }        }
2448      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2449      switch(c)      switch(c)
2450        {        {
2451          HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
2452        default: break;        default: break;
       case 0x09:      /* HT */  
       case 0x20:      /* SPACE */  
       case 0xa0:      /* NBSP */  
       case 0x1680:    /* OGHAM SPACE MARK */  
       case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
       case 0x2000:    /* EN QUAD */  
       case 0x2001:    /* EM QUAD */  
       case 0x2002:    /* EN SPACE */  
       case 0x2003:    /* EM SPACE */  
       case 0x2004:    /* THREE-PER-EM SPACE */  
       case 0x2005:    /* FOUR-PER-EM SPACE */  
       case 0x2006:    /* SIX-PER-EM SPACE */  
       case 0x2007:    /* FIGURE SPACE */  
       case 0x2008:    /* PUNCTUATION SPACE */  
       case 0x2009:    /* THIN SPACE */  
       case 0x200A:    /* HAIR SPACE */  
       case 0x202f:    /* NARROW NO-BREAK SPACE */  
       case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
       case 0x3000:    /* IDEOGRAPHIC SPACE */  
       MRRETURN(MATCH_NOMATCH);  
2453        }        }
2454      ecode++;      ecode++;
2455      break;      break;
# Line 2206  for (;;) Line 2458  for (;;)
2458      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2459        {        {
2460        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2461        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2462        }        }
2463      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2464      switch(c)      switch(c)
2465        {        {
2466        default: MRRETURN(MATCH_NOMATCH);        HSPACE_CASES: break;  /* Byte and multibyte cases */
2467        case 0x09:      /* HT */        default: RRETURN(MATCH_NOMATCH);
       case 0x20:      /* SPACE */  
       case 0xa0:      /* NBSP */  
       case 0x1680:    /* OGHAM SPACE MARK */  
       case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
       case 0x2000:    /* EN QUAD */  
       case 0x2001:    /* EM QUAD */  
       case 0x2002:    /* EN SPACE */  
       case 0x2003:    /* EM SPACE */  
       case 0x2004:    /* THREE-PER-EM SPACE */  
       case 0x2005:    /* FOUR-PER-EM SPACE */  
       case 0x2006:    /* SIX-PER-EM SPACE */  
       case 0x2007:    /* FIGURE SPACE */  
       case 0x2008:    /* PUNCTUATION SPACE */  
       case 0x2009:    /* THIN SPACE */  
       case 0x200A:    /* HAIR SPACE */  
       case 0x202f:    /* NARROW NO-BREAK SPACE */  
       case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
       case 0x3000:    /* IDEOGRAPHIC SPACE */  
       break;  
2468        }        }
2469      ecode++;      ecode++;
2470      break;      break;
# Line 2240  for (;;) Line 2473  for (;;)
2473      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2474        {        {
2475        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2476        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2477        }        }
2478      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2479      switch(c)      switch(c)
2480        {        {
2481          VSPACE_CASES: RRETURN(MATCH_NOMATCH);
2482        default: break;        default: break;
       case 0x0a:      /* LF */  
       case 0x0b:      /* VT */  
       case 0x0c:      /* FF */  
       case 0x0d:      /* CR */  
       case 0x85:      /* NEL */  
       case 0x2028:    /* LINE SEPARATOR */  
       case 0x2029:    /* PARAGRAPH SEPARATOR */  
       MRRETURN(MATCH_NOMATCH);  
2483        }        }
2484      ecode++;      ecode++;
2485      break;      break;
# Line 2262  for (;;) Line 2488  for (;;)
2488      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2489        {        {
2490        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2491        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2492        }        }
2493      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2494      switch(c)      switch(c)
2495        {        {
2496        default: MRRETURN(MATCH_NOMATCH);        VSPACE_CASES: break;
2497        case 0x0a:      /* LF */        default: RRETURN(MATCH_NOMATCH);
       case 0x0b:      /* VT */  
       case 0x0c:      /* FF */  
       case 0x0d:      /* CR */  
       case 0x85:      /* NEL */  
       case 0x2028:    /* LINE SEPARATOR */  
       case 0x2029:    /* PARAGRAPH SEPARATOR */  
       break;  
2498        }        }
2499      ecode++;      ecode++;
2500      break;      break;
# Line 2289  for (;;) Line 2508  for (;;)
2508      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2509        {        {
2510        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2511        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2512        }        }
2513      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2514        {        {
# Line 2298  for (;;) Line 2517  for (;;)
2517        switch(ecode[1])        switch(ecode[1])
2518          {          {
2519          case PT_ANY:          case PT_ANY:
2520          if (op == OP_NOTPROP) MRRETURN(MATCH_NOMATCH);          if (op == OP_NOTPROP) RRETURN(MATCH_NOMATCH);
2521          break;          break;
2522    
2523          case PT_LAMP:          case PT_LAMP:
2524          if ((prop->chartype == ucp_Lu ||          if ((prop->chartype == ucp_Lu ||
2525               prop->chartype == ucp_Ll ||               prop->chartype == ucp_Ll ||
2526               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))
2527            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2528          break;          break;
2529    
2530          case PT_GC:          case PT_GC:
2531          if ((ecode[2] != _pcre_ucp_gentype[prop->chartype]) == (op == OP_PROP))          if ((ecode[2] != PRIV(ucp_gentype)[prop->chartype]) == (op == OP_PROP))
2532            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2533          break;          break;
2534    
2535          case PT_PC:          case PT_PC:
2536          if ((ecode[2] != prop->chartype) == (op == OP_PROP))          if ((ecode[2] != prop->chartype) == (op == OP_PROP))
2537            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2538          break;          break;
2539    
2540          case PT_SC:          case PT_SC:
2541          if ((ecode[2] != prop->script) == (op == OP_PROP))          if ((ecode[2] != prop->script) == (op == OP_PROP))
2542            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2543          break;          break;
2544    
2545          /* These are specials */          /* These are specials */
2546    
2547          case PT_ALNUM:          case PT_ALNUM:
2548          if ((_pcre_ucp_gentype[prop->chartype] == ucp_L ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2549               _pcre_ucp_gentype[prop->chartype] == ucp_N) == (op == OP_NOTPROP))               PRIV(ucp_gentype)[prop->chartype] == ucp_N) == (op == OP_NOTPROP))
2550            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2551          break;          break;
2552    
2553          case PT_SPACE:    /* Perl space */          case PT_SPACE:    /* Perl space */
2554          if ((_pcre_ucp_gentype[prop->chartype] == ucp_Z ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2555               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)
2556                 == (op == OP_NOTPROP))                 == (op == OP_NOTPROP))
2557            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2558          break;          break;
2559    
2560          case PT_PXSPACE:  /* POSIX space */          case PT_PXSPACE:  /* POSIX space */
2561          if ((_pcre_ucp_gentype[prop->chartype] == ucp_Z ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2562               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
2563               c == CHAR_FF || c == CHAR_CR)               c == CHAR_FF || c == CHAR_CR)
2564                 == (op == OP_NOTPROP))                 == (op == OP_NOTPROP))
2565            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2566          break;          break;
2567    
2568          case PT_WORD:          case PT_WORD:
2569          if ((_pcre_ucp_gentype[prop->chartype] == ucp_L ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2570               _pcre_ucp_gentype[prop->chartype] == ucp_N ||               PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
2571               c == CHAR_UNDERSCORE) == (op == OP_NOTPROP))               c == CHAR_UNDERSCORE) == (op == OP_NOTPROP))
2572            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2573          break;          break;
2574    
2575          /* This should never occur */          /* This should never occur */
# Line 2370  for (;;) Line 2589  for (;;)
2589      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2590        {        {
2591        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2592        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2593        }        }
2594      GETCHARINCTEST(c, eptr);      else
     if (UCD_CATEGORY(c) == ucp_M) MRRETURN(MATCH_NOMATCH);  
     while (eptr < md->end_subject)  
2595        {        {
2596        int len = 1;        int lgb, rgb;
2597        if (!utf8) c = *eptr; else { GETCHARLEN(c, eptr, len); }        GETCHARINCTEST(c, eptr);
2598        if (UCD_CATEGORY(c) != ucp_M) break;        lgb = UCD_GRAPHBREAK(c);
2599        eptr += len;        while (eptr < md->end_subject)
2600            {
2601            int len = 1;
2602            if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
2603            rgb = UCD_GRAPHBREAK(c);
2604            if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
2605            lgb = rgb;
2606            eptr += len;
2607            }
2608        }        }
2609        CHECK_PARTIAL();
2610      ecode++;      ecode++;
2611      break;      break;
2612  #endif  #endif
# Line 2398  for (;;) Line 2624  for (;;)
2624      case OP_REFI:      case OP_REFI:
2625      caseless = op == OP_REFI;      caseless = op == OP_REFI;
2626      offset = GET2(ecode, 1) << 1;               /* Doubled ref number */      offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
2627      ecode += 3;      ecode += 1 + IMM2_SIZE;
2628    
2629      /* If the reference is unset, there are two possibilities:      /* If the reference is unset, there are two possibilities:
2630    
# Line 2438  for (;;) Line 2664  for (;;)
2664        case OP_CRMINRANGE:        case OP_CRMINRANGE:
2665        minimize = (*ecode == OP_CRMINRANGE);        minimize = (*ecode == OP_CRMINRANGE);
2666        min = GET2(ecode, 1);        min = GET2(ecode, 1);
2667        max = GET2(ecode, 3);        max = GET2(ecode, 1 + IMM2_SIZE);
2668        if (max == 0) max = INT_MAX;        if (max == 0) max = INT_MAX;
2669        ecode += 5;        ecode += 1 + 2 * IMM2_SIZE;
2670        break;        break;
2671    
2672        default:               /* No repeat follows */        default:               /* No repeat follows */
2673        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)
2674          {          {
2675            if (length == -2) eptr = md->end_subject;   /* Partial match */
2676          CHECK_PARTIAL();          CHECK_PARTIAL();
2677          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2678          }          }
2679        eptr += length;        eptr += length;
2680        continue;              /* With the main loop */        continue;              /* With the main loop */
2681        }        }
2682    
2683      /* Handle repeated back references. If the length of the reference is      /* Handle repeated back references. If the length of the reference is
2684      zero, just continue with the main loop. */      zero, just continue with the main loop. If the length is negative, it
2685        means the reference is unset in non-Java-compatible mode. If the minimum is
2686        zero, we can continue at the same level without recursion. For any other
2687        minimum, carrying on will result in NOMATCH. */
2688    
2689      if (length == 0) continue;      if (length == 0) continue;
2690        if (length < 0 && min == 0) continue;
2691    
2692      /* First, ensure the minimum number of matches are present. We get back      /* First, ensure the minimum number of matches are present. We get back
2693      the length of the reference string explicitly rather than passing the      the length of the reference string explicitly rather than passing the
# Line 2467  for (;;) Line 2698  for (;;)
2698        int slength;        int slength;
2699        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2700          {          {
2701            if (slength == -2) eptr = md->end_subject;   /* Partial match */
2702          CHECK_PARTIAL();          CHECK_PARTIAL();
2703          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2704          }          }
2705        eptr += slength;        eptr += slength;
2706        }        }
# Line 2487  for (;;) Line 2719  for (;;)
2719          int slength;          int slength;
2720          RMATCH(eptr, ecode, offset_top, md, eptrb, RM14);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM14);
2721          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2722          if (fi >= max) MRRETURN(MATCH_NOMATCH);          if (fi >= max) RRETURN(MATCH_NOMATCH);
2723          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2724            {            {
2725              if (slength == -2) eptr = md->end_subject;   /* Partial match */
2726            CHECK_PARTIAL();            CHECK_PARTIAL();
2727            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2728            }            }
2729          eptr += slength;          eptr += slength;
2730          }          }
# Line 2508  for (;;) Line 2741  for (;;)
2741          int slength;          int slength;
2742          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2743            {            {
2744            CHECK_PARTIAL();            /* Can't use CHECK_PARTIAL because we don't want to update eptr in
2745              the soft partial matching case. */
2746    
2747              if (slength == -2 && md->partial != 0 &&
2748                  md->end_subject > md->start_used_ptr)
2749                {
2750                md->hitend = TRUE;
2751                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2752                }
2753            break;            break;
2754            }            }
2755          eptr += slength;          eptr += slength;
2756          }          }
2757    
2758        while (eptr >= pp)        while (eptr >= pp)
2759          {          {
2760          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);
2761          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2762          eptr -= length;          eptr -= length;
2763          }          }
2764        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2765        }        }
2766      /* Control never gets here */      /* Control never gets here */
2767    
# Line 2537  for (;;) Line 2779  for (;;)
2779      case OP_NCLASS:      case OP_NCLASS:
2780      case OP_CLASS:      case OP_CLASS:
2781        {        {
2782          /* The data variable is saved across frames, so the byte map needs to
2783          be stored there. */
2784    #define BYTE_MAP ((pcre_uint8 *)data)
2785        data = ecode + 1;                /* Save for matching */        data = ecode + 1;                /* Save for matching */
2786        ecode += 33;                     /* Advance past the item */        ecode += 1 + (32 / sizeof(pcre_uchar)); /* Advance past the item */
2787    
2788        switch (*ecode)        switch (*ecode)
2789          {          {
# Line 2559  for (;;) Line 2804  for (;;)
2804          case OP_CRMINRANGE:          case OP_CRMINRANGE:
2805          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
2806          min = GET2(ecode, 1);          min = GET2(ecode, 1);
2807          max = GET2(ecode, 3);          max = GET2(ecode, 1 + IMM2_SIZE);
2808          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
2809          ecode += 5;          ecode += 1 + 2 * IMM2_SIZE;
2810          break;          break;
2811    
2812          default:               /* No repeat follows */          default:               /* No repeat follows */
# Line 2571  for (;;) Line 2816  for (;;)
2816    
2817        /* First, ensure the minimum number of matches are present. */        /* First, ensure the minimum number of matches are present. */
2818    
2819  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2820        /* UTF-8 mode */        if (utf)
       if (utf8)  
2821          {          {
2822          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2823            {            {
2824            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
2825              {              {
2826              SCHECK_PARTIAL();              SCHECK_PARTIAL();
2827              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2828              }              }
2829            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
2830            if (c > 255)            if (c > 255)
2831              {              {
2832              if (op == OP_CLASS) MRRETURN(MATCH_NOMATCH);              if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2833              }              }
2834            else            else
2835              {              if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
             if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  
             }  
2836            }            }
2837          }          }
2838        else        else
2839  #endif  #endif
2840        /* Not UTF-8 mode */        /* Not UTF mode */
2841          {          {
2842          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2843            {            {
2844            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
2845              {              {
2846              SCHECK_PARTIAL();              SCHECK_PARTIAL();
2847              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2848              }              }
2849            c = *eptr++;            c = *eptr++;
2850            if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  #ifndef COMPILE_PCRE8
2851              if (c > 255)
2852                {
2853                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2854                }
2855              else
2856    #endif
2857                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2858            }            }
2859          }          }
2860    
# Line 2619  for (;;) Line 2868  for (;;)
2868    
2869        if (minimize)        if (minimize)
2870          {          {
2871  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2872          /* UTF-8 mode */          if (utf)
         if (utf8)  
2873            {            {
2874            for (fi = min;; fi++)            for (fi = min;; fi++)
2875              {              {
2876              RMATCH(eptr, ecode, offset_top, md, eptrb, RM16);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM16);
2877              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2878              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2879              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
2880                {                {
2881                SCHECK_PARTIAL();                SCHECK_PARTIAL();
2882                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2883                }                }
2884              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
2885              if (c > 255)              if (c > 255)
2886                {                {
2887                if (op == OP_CLASS) MRRETURN(MATCH_NOMATCH);                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2888                }                }
2889              else              else
2890                {                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
               if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  
               }  
2891              }              }
2892            }            }
2893          else          else
2894  #endif  #endif
2895          /* Not UTF-8 mode */          /* Not UTF mode */
2896            {            {
2897            for (fi = min;; fi++)            for (fi = min;; fi++)
2898              {              {
2899              RMATCH(eptr, ecode, offset_top, md, eptrb, RM17);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM17);
2900              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2901              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2902              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
2903                {                {
2904                SCHECK_PARTIAL();                SCHECK_PARTIAL();
2905                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2906                }                }
2907              c = *eptr++;              c = *eptr++;
2908              if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  #ifndef COMPILE_PCRE8
2909                if (c > 255)
2910                  {
2911                  if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2912                  }
2913                else
2914    #endif
2915                  if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2916              }              }
2917            }            }
2918          /* Control never gets here */          /* Control never gets here */
# Line 2671  for (;;) Line 2924  for (;;)
2924          {          {
2925          pp = eptr;          pp = eptr;
2926    
2927  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2928          /* UTF-8 mode */          if (utf)
         if (utf8)  
2929            {            {
2930            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2931              {              {
# Line 2689  for (;;) Line 2941  for (;;)
2941                if (op == OP_CLASS) break;                if (op == OP_CLASS) break;
2942                }                }
2943              else              else
2944                {                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
               if ((data[c/8] & (1 << (c&7))) == 0) break;  
               }  
2945              eptr += len;              eptr += len;
2946              }              }
2947            for (;;)            for (;;)
# Line 2704  for (;;) Line 2954  for (;;)
2954            }            }
2955          else          else
2956  #endif  #endif
2957            /* Not UTF-8 mode */            /* Not UTF mode */
2958            {            {
2959            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2960              {              {
# Line 2714  for (;;) Line 2964  for (;;)
2964                break;                break;
2965                }                }
2966              c = *eptr;              c = *eptr;
2967              if ((data[c/8] & (1 << (c&7))) == 0) break;  #ifndef COMPILE_PCRE8
2968                if (c > 255)
2969                  {
2970                  if (op == OP_CLASS) break;
2971                  }
2972                else
2973    #endif
2974                  if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
2975              eptr++;              eptr++;
2976              }              }
2977            while (eptr >= pp)            while (eptr >= pp)
# Line 2725  for (;;) Line 2982  for (;;)
2982              }              }
2983            }            }
2984    
2985          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2986          }          }
2987    #undef BYTE_MAP
2988        }        }
2989      /* Control never gets here */      /* Control never gets here */
2990    
# Line 2735  for (;;) Line 2993  for (;;)
2993      when UTF-8 mode mode is supported. Nevertheless, we may not be in UTF-8      when UTF-8 mode mode is supported. Nevertheless, we may not be in UTF-8
2994      mode, because Unicode properties are supported in non-UTF-8 mode. */      mode, because Unicode properties are supported in non-UTF-8 mode. */
2995    
2996  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
2997      case OP_XCLASS:      case OP_XCLASS:
2998        {        {
2999        data = ecode + 1 + LINK_SIZE;                /* Save for matching */        data = ecode + 1 + LINK_SIZE;                /* Save for matching */
# Line 2760  for (;;) Line 3018  for (;;)
3018          case OP_CRMINRANGE:          case OP_CRMINRANGE:
3019          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
3020          min = GET2(ecode, 1);          min = GET2(ecode, 1);
3021          max = GET2(ecode, 3);          max = GET2(ecode, 1 + IMM2_SIZE);
3022          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
3023          ecode += 5;          ecode += 1 + 2 * IMM2_SIZE;
3024          break;          break;
3025    
3026          default:               /* No repeat follows */          default:               /* No repeat follows */
# Line 2777  for (;;) Line 3035  for (;;)
3035          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3036            {            {
3037            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3038            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3039            }            }
3040          GETCHARINCTEST(c, eptr);          GETCHARINCTEST(c, eptr);
3041          if (!_pcre_xclass(c, data)) MRRETURN(MATCH_NOMATCH);          if (!PRIV(xclass)(c, data, utf)) RRETURN(MATCH_NOMATCH);
3042          }          }
3043    
3044        /* If max == min we can continue with the main loop without the        /* If max == min we can continue with the main loop without the
# Line 2797  for (;;) Line 3055  for (;;)
3055            {            {
3056            RMATCH(eptr, ecode, offset_top, md, eptrb, RM20);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM20);
3057            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3058            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3059            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3060              {              {
3061              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3062              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3063              }              }
3064            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3065            if (!_pcre_xclass(c, data)) MRRETURN(MATCH_NOMATCH);            if (!PRIV(xclass)(c, data, utf)) RRETURN(MATCH_NOMATCH);
3066            }            }
3067          /* Control never gets here */          /* Control never gets here */
3068          }          }
# Line 2822  for (;;) Line 3080  for (;;)
3080              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3081              break;              break;
3082              }              }
3083    #ifdef SUPPORT_UTF
3084            GETCHARLENTEST(c, eptr, len);            GETCHARLENTEST(c, eptr, len);
3085            if (!_pcre_xclass(c, data)) break;  #else
3086              c = *eptr;
3087    #endif
3088              if (!PRIV(xclass)(c, data, utf)) break;
3089            eptr += len;            eptr += len;
3090            }            }
3091          for(;;)          for(;;)
# Line 2831  for (;;) Line 3093  for (;;)
3093            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);
3094            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3095            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3096            if (utf8) BACKCHAR(eptr);  #ifdef SUPPORT_UTF
3097              if (utf) BACKCHAR(eptr);
3098    #endif
3099            }            }
3100          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3101          }          }
3102    
3103        /* Control never gets here */        /* Control never gets here */
# Line 2843  for (;;) Line 3107  for (;;)
3107      /* Match a single character, casefully */      /* Match a single character, casefully */
3108    
3109      case OP_CHAR:      case OP_CHAR:
3110  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3111      if (utf8)      if (utf)
3112        {        {
3113        length = 1;        length = 1;
3114        ecode++;        ecode++;
# Line 2852  for (;;) Line 3116  for (;;)
3116        if (length > md->end_subject - eptr)        if (length > md->end_subject - eptr)
3117          {          {
3118          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
3119          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3120          }          }
3121        while (length-- > 0) if (*ecode++ != *eptr++) MRRETURN(MATCH_NOMATCH);        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);
3122        }        }
3123      else      else
3124  #endif  #endif
3125        /* Not UTF mode */
     /* Non-UTF-8 mode */  
3126        {        {
3127        if (md->end_subject - eptr < 1)        if (md->end_subject - eptr < 1)
3128          {          {
3129          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */
3130          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3131          }          }
3132        if (ecode[1] != *eptr++) MRRETURN(MATCH_NOMATCH);        if (ecode[1] != *eptr++) RRETURN(MATCH_NOMATCH);
3133        ecode += 2;        ecode += 2;
3134        }        }
3135      break;      break;
3136    
3137      /* Match a single character, caselessly */      /* Match a single character, caselessly. If we are at the end of the
3138        subject, give up immediately. */
3139    
3140      case OP_CHARI:      case OP_CHARI:
3141  #ifdef SUPPORT_UTF8      if (eptr >= md->end_subject)
3142      if (utf8)        {
3143          SCHECK_PARTIAL();
3144          RRETURN(MATCH_NOMATCH);
3145          }
3146    
3147    #ifdef SUPPORT_UTF
3148        if (utf)
3149        {        {
3150        length = 1;        length = 1;
3151        ecode++;        ecode++;
3152        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
3153    
       if (length > md->end_subject - eptr)  
         {  
         CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */  
         MRRETURN(MATCH_NOMATCH);  
         }  
   
3154        /* 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
3155        can use the fast lookup table. */        we know that its other case must also be one byte long, so we can use the
3156          fast lookup table. We know that there is at least one byte left in the
3157          subject. */
3158    
3159        if (fc < 128)        if (fc < 128)
3160          {          {
3161          if (md->lcc[*ecode++] != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);          if (md->lcc[fc]
3162                != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);
3163            ecode++;
3164            eptr++;
3165          }          }
3166    
3167        /* Otherwise we must pick up the subject character */        /* Otherwise we must pick up the subject character. Note that we cannot
3168          use the value of "length" to check for sufficient bytes left, because the
3169          other case of the character may have more or fewer bytes.  */
3170    
3171        else        else
3172          {          {
# Line 2911  for (;;) Line 3182  for (;;)
3182  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3183            if (dc != UCD_OTHERCASE(fc))            if (dc != UCD_OTHERCASE(fc))
3184  #endif  #endif
3185              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3186            }            }
3187          }          }
3188        }        }
3189      else      else
3190  #endif   /* SUPPORT_UTF8 */  #endif   /* SUPPORT_UTF */
3191    
3192      /* Non-UTF-8 mode */      /* Not UTF mode */
3193        {        {
3194        if (md->end_subject - eptr < 1)        if (TABLE_GET(ecode[1], md->lcc, ecode[1])
3195          {            != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);
3196          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */        eptr++;
         MRRETURN(MATCH_NOMATCH);  
         }  
       if (md->lcc[ecode[1]] != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);  
3197        ecode += 2;        ecode += 2;
3198        }        }
3199      break;      break;
# Line 2935  for (;;) Line 3203  for (;;)
3203      case OP_EXACT:      case OP_EXACT:
3204      case OP_EXACTI:      case OP_EXACTI:
3205      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3206      ecode += 3;      ecode += 1 + IMM2_SIZE;
3207      goto REPEATCHAR;      goto REPEATCHAR;
3208    
3209      case OP_POSUPTO:      case OP_POSUPTO:
# Line 2950  for (;;) Line 3218  for (;;)
3218      min = 0;      min = 0;
3219      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3220      minimize = *ecode == OP_MINUPTO || *ecode == OP_MINUPTOI;      minimize = *ecode == OP_MINUPTO || *ecode == OP_MINUPTOI;
3221      ecode += 3;      ecode += 1 + IMM2_SIZE;
3222      goto REPEATCHAR;      goto REPEATCHAR;
3223    
3224      case OP_POSSTAR:      case OP_POSSTAR:
# Line 2998  for (;;) Line 3266  for (;;)
3266      /* Common code for all repeated single-character matches. */      /* Common code for all repeated single-character matches. */
3267    
3268      REPEATCHAR:      REPEATCHAR:
3269  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3270      if (utf8)      if (utf)
3271        {        {
3272        length = 1;        length = 1;
3273        charptr = ecode;        charptr = ecode;
# Line 3015  for (;;) Line 3283  for (;;)
3283          unsigned int othercase;          unsigned int othercase;
3284          if (op >= OP_STARI &&     /* Caseless */          if (op >= OP_STARI &&     /* Caseless */
3285              (othercase = UCD_OTHERCASE(fc)) != fc)              (othercase = UCD_OTHERCASE(fc)) != fc)
3286            oclength = _pcre_ord2utf8(othercase, occhars);            oclength = PRIV(ord2utf)(othercase, occhars);
3287          else oclength = 0;          else oclength = 0;
3288  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3289    
3290          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3291            {            {
3292            if (eptr <= md->end_subject - length &&            if (eptr <= md->end_subject - length &&
3293              memcmp(eptr, charptr, length) == 0) eptr += length;              memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3294  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3295            else if (oclength > 0 &&            else if (oclength > 0 &&
3296                     eptr <= md->end_subject - oclength &&                     eptr <= md->end_subject - oclength &&
3297                     memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                     memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3298  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3299            else            else
3300              {              {
3301              CHECK_PARTIAL();              CHECK_PARTIAL();
3302              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3303              }              }
3304            }            }
3305    
# Line 3043  for (;;) Line 3311  for (;;)
3311              {              {
3312              RMATCH(eptr, ecode, offset_top, md, eptrb, RM22);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM22);
3313              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3314              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3315              if (eptr <= md->end_subject - length &&              if (eptr <= md->end_subject - length &&
3316                memcmp(eptr, charptr, length) == 0) eptr += length;                memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3317  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3318              else if (oclength > 0 &&              else if (oclength > 0 &&
3319                       eptr <= md->end_subject - oclength &&                       eptr <= md->end_subject - oclength &&
3320                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                       memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3321  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3322              else              else
3323                {                {
3324                CHECK_PARTIAL();                CHECK_PARTIAL();
3325                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3326                }                }
3327              }              }
3328            /* Control never gets here */            /* Control never gets here */
# Line 3066  for (;;) Line 3334  for (;;)
3334            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3335              {              {
3336              if (eptr <= md->end_subject - length &&              if (eptr <= md->end_subject - length &&
3337                  memcmp(eptr, charptr, length) == 0) eptr += length;                  memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3338  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3339              else if (oclength > 0 &&              else if (oclength > 0 &&
3340                       eptr <= md->end_subject - oclength &&                       eptr <= md->end_subject - oclength &&
3341                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                       memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3342  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3343              else              else
3344                {                {
# Line 3085  for (;;) Line 3353  for (;;)
3353              {              {
3354              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);
3355              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3356              if (eptr == pp) { MRRETURN(MATCH_NOMATCH); }              if (eptr == pp) { RRETURN(MATCH_NOMATCH); }
3357  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3358              eptr--;              eptr--;
3359              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 3102  for (;;) Line 3370  for (;;)
3370        value of fc will always be < 128. */        value of fc will always be < 128. */
3371        }        }
3372      else      else
3373  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
3374          /* When not in UTF-8 mode, load a single-byte character. */
3375          fc = *ecode++;
3376    
3377      /* When not in UTF-8 mode, load a single-byte character. */      /* The value of fc at this point is always one character, though we may
3378        or may not be in UTF mode. The code is duplicated for the caseless and
     fc = *ecode++;  
   
     /* The value of fc at this point is always less than 256, though we may or  
     may not be in UTF-8 mode. The code is duplicated for the caseless and  
3379      caseful cases, for speed, since matching characters is likely to be quite      caseful cases, for speed, since matching characters is likely to be quite
3380      common. First, ensure the minimum number of matches are present. If min =      common. First, ensure the minimum number of matches are present. If min =
3381      max, continue at the same level without recursing. Otherwise, if      max, continue at the same level without recursing. Otherwise, if
# Line 3118  for (;;) Line 3384  for (;;)
3384      maximizing, find the maximum number of characters and work backwards. */      maximizing, find the maximum number of characters and work backwards. */
3385    
3386      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,
3387        max, eptr));        max, (char *)eptr));
3388    
3389      if (op >= OP_STARI)  /* Caseless */      if (op >= OP_STARI)  /* Caseless */
3390        {        {
3391        fc = md->lcc[fc];  #ifdef COMPILE_PCRE8
3392          /* fc must be < 128 if UTF is enabled. */
3393          foc = md->fcc[fc];
3394    #else
3395    #ifdef SUPPORT_UTF
3396    #ifdef SUPPORT_UCP
3397          if (utf && fc > 127)
3398            foc = UCD_OTHERCASE(fc);
3399    #else
3400          if (utf && fc > 127)
3401            foc = fc;
3402    #endif /* SUPPORT_UCP */
3403          else
3404    #endif /* SUPPORT_UTF */
3405            foc = TABLE_GET(fc, md->fcc, fc);
3406    #endif /* COMPILE_PCRE8 */
3407    
3408        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
3409          {          {
3410          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3411            {            {
3412            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3413            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3414            }            }
3415          if (fc != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);          if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);
3416            eptr++;
3417          }          }
3418        if (min == max) continue;        if (min == max) continue;
3419        if (minimize)        if (minimize)
# Line 3139  for (;;) Line 3422  for (;;)
3422            {            {
3423            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);
3424            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3425            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3426            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3427              {              {
3428              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3429              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3430              }              }
3431            if (fc != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);            if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);
3432              eptr++;
3433            }            }
3434          /* Control never gets here */          /* Control never gets here */
3435          }          }
# Line 3159  for (;;) Line 3443  for (;;)
3443              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3444              break;              break;
3445              }              }
3446            if (fc != md->lcc[*eptr]) break;            if (fc != *eptr && foc != *eptr) break;
3447            eptr++;            eptr++;
3448            }            }
3449    
# Line 3171  for (;;) Line 3455  for (;;)
3455            eptr--;            eptr--;
3456            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3457            }            }
3458          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3459          }          }
3460        /* Control never gets here */        /* Control never gets here */
3461        }        }
# Line 3185  for (;;) Line 3469  for (;;)
3469          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3470            {            {
3471            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3472            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3473            }            }
3474          if (fc != *eptr++) MRRETURN(MATCH_NOMATCH);          if (fc != *eptr++) RRETURN(MATCH_NOMATCH);
3475          }          }
3476    
3477        if (min == max) continue;        if (min == max) continue;
# Line 3198  for (;;) Line 3482  for (;;)
3482            {            {
3483            RMATCH(eptr, ecode, offset_top, md, eptrb, RM26);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM26);
3484            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3485            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3486            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3487              {              {
3488              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3489              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3490              }              }
3491            if (fc != *eptr++) MRRETURN(MATCH_NOMATCH);            if (fc != *eptr++) RRETURN(MATCH_NOMATCH);
3492            }            }
3493          /* Control never gets here */          /* Control never gets here */
3494          }          }
# Line 3229  for (;;) Line 3513  for (;;)
3513            eptr--;            eptr--;
3514            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3515            }            }
3516          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3517          }          }
3518        }        }
3519      /* Control never gets here */      /* Control never gets here */
# Line 3242  for (;;) Line 3526  for (;;)
3526      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
3527        {        {
3528        SCHECK_PARTIAL();        SCHECK_PARTIAL();
3529        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
3530        }        }
3531      ecode++;  #ifdef SUPPORT_UTF
3532      GETCHARINCTEST(c, eptr);      if (utf)
     if (op == OP_NOTI)         /* The caseless case */  
3533        {        {
3534  #ifdef SUPPORT_UTF8        register unsigned int ch, och;
3535        if (c < 256)  
3536  #endif        ecode++;
3537        c = md->lcc[c];        GETCHARINC(ch, ecode);
3538        if (md->lcc[*ecode++] == c) MRRETURN(MATCH_NOMATCH);        GETCHARINC(c, eptr);
3539    
3540          if (op == OP_NOT)
3541            {
3542            if (ch == c) RRETURN(MATCH_NOMATCH);
3543            }
3544          else
3545            {
3546    #ifdef SUPPORT_UCP
3547            if (ch > 127)
3548              och = UCD_OTHERCASE(ch);
3549    #else
3550            if (ch > 127)
3551              och = ch;
3552    #endif /* SUPPORT_UCP */
3553            else
3554              och = TABLE_GET(ch, md->fcc, ch);
3555            if (ch == c || och == c) RRETURN(MATCH_NOMATCH);
3556            }
3557        }        }
3558      else    /* Caseful */      else
3559    #endif
3560        {        {
3561        if (*ecode++ == c) MRRETURN(MATCH_NOMATCH);        register unsigned int ch = ecode[1];
3562          c = *eptr++;
3563          if (ch == c || (op == OP_NOTI && TABLE_GET(ch, md->fcc, ch) == c))
3564            RRETURN(MATCH_NOMATCH);
3565          ecode += 2;
3566        }        }
3567      break;      break;
3568    
# Line 3270  for (;;) Line 3576  for (;;)
3576      case OP_NOTEXACT:      case OP_NOTEXACT:
3577      case OP_NOTEXACTI:      case OP_NOTEXACTI:
3578      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3579      ecode += 3;      ecode += 1 + IMM2_SIZE;
3580      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3581    
3582      case OP_NOTUPTO:      case OP_NOTUPTO:
# Line 3280  for (;;) Line 3586  for (;;)
3586      min = 0;      min = 0;
3587      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3588      minimize = *ecode == OP_NOTMINUPTO || *ecode == OP_NOTMINUPTOI;      minimize = *ecode == OP_NOTMINUPTO || *ecode == OP_NOTMINUPTOI;
3589      ecode += 3;      ecode += 1 + IMM2_SIZE;
3590      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3591    
3592      case OP_NOTPOSSTAR:      case OP_NOTPOSSTAR:
# Line 3312  for (;;) Line 3618  for (;;)
3618      possessive = TRUE;      possessive = TRUE;
3619      min = 0;      min = 0;
3620      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3621      ecode += 3;      ecode += 1 + IMM2_SIZE;
3622      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3623    
3624      case OP_NOTSTAR:      case OP_NOTSTAR:
# Line 3336  for (;;) Line 3642  for (;;)
3642      /* Common code for all repeated single-byte matches. */      /* Common code for all repeated single-byte matches. */
3643    
3644      REPEATNOTCHAR:      REPEATNOTCHAR:
3645      fc = *ecode++;      GETCHARINCTEST(fc, ecode);
3646    
3647      /* The code is duplicated for the caseless and caseful cases, for speed,      /* The code is duplicated for the caseless and caseful cases, for speed,
3648      since matching characters is likely to be quite common. First, ensure the      since matching characters is likely to be quite common. First, ensure the
# Line 3347  for (;;) Line 3653  for (;;)
3653      characters and work backwards. */      characters and work backwards. */
3654    
3655      DPRINTF(("negative matching %c{%d,%d} against subject %.*s\n", fc, min, max,      DPRINTF(("negative matching %c{%d,%d} against subject %.*s\n", fc, min, max,
3656        max, eptr));        max, (char *)eptr));
3657    
3658      if (op >= OP_NOTSTARI)     /* Caseless */      if (op >= OP_NOTSTARI)     /* Caseless */
3659        {        {
3660        fc = md->lcc[fc];  #ifdef SUPPORT_UTF
3661    #ifdef SUPPORT_UCP
3662          if (utf && fc > 127)
3663            foc = UCD_OTHERCASE(fc);
3664    #else
3665          if (utf && fc > 127)
3666            foc = fc;
3667    #endif /* SUPPORT_UCP */
3668          else
3669    #endif /* SUPPORT_UTF */
3670            foc = TABLE_GET(fc, md->fcc, fc);
3671    
3672  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3673        /* UTF-8 mode */        if (utf)
       if (utf8)  
3674          {          {
3675          register unsigned int d;          register unsigned int d;
3676          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3363  for (;;) Line 3678  for (;;)
3678            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3679              {              {
3680              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3681              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3682              }              }
3683            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3684            if (d < 256) d = md->lcc[d];            if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
           if (fc == d) MRRETURN(MATCH_NOMATCH);  
3685            }            }
3686          }          }
3687        else        else
3688  #endif  #endif
3689          /* Not UTF mode */
       /* Not UTF-8 mode */  
3690          {          {
3691          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3692            {            {
3693            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3694              {              {
3695              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3696              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3697              }              }
3698            if (fc == md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);            if (fc == *eptr || foc == *eptr) RRETURN(MATCH_NOMATCH);
3699              eptr++;
3700            }            }
3701          }          }
3702    
# Line 3390  for (;;) Line 3704  for (;;)
3704    
3705        if (minimize)        if (minimize)
3706          {          {
3707  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3708          /* UTF-8 mode */          if (utf)
         if (utf8)  
3709            {            {
3710            register unsigned int d;            register unsigned int d;
3711            for (fi = min;; fi++)            for (fi = min;; fi++)
3712              {              {
3713              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);
3714              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3715              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3716              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3717                {                {
3718                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3719                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3720                }                }
3721              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3722              if (d < 256) d = md->lcc[d];              if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
             if (fc == d) MRRETURN(MATCH_NOMATCH);  
3723              }              }
3724            }            }
3725          else          else
3726  #endif  #endif
3727          /* Not UTF-8 mode */          /* Not UTF mode */
3728            {            {
3729            for (fi = min;; fi++)            for (fi = min;; fi++)
3730              {              {
3731              RMATCH(eptr, ecode, offset_top, md, eptrb, RM29);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM29);
3732              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3733              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3734              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3735                {                {
3736                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3737                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3738                }                }
3739              if (fc == md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);              if (fc == *eptr || foc == *eptr) RRETURN(MATCH_NOMATCH);
3740                eptr++;
3741              }              }
3742            }            }
3743          /* Control never gets here */          /* Control never gets here */
# Line 3436  for (;;) Line 3749  for (;;)
3749          {          {
3750          pp = eptr;          pp = eptr;
3751    
3752  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3753          /* UTF-8 mode */          if (utf)
         if (utf8)  
3754            {            {
3755            register unsigned int d;            register unsigned int d;
3756            for (i = min; i < max; i++)            for (i = min; i < max; i++)
# Line 3450  for (;;) Line 3762  for (;;)
3762                break;                break;
3763                }                }
3764              GETCHARLEN(d, eptr, len);              GETCHARLEN(d, eptr, len);
3765              if (d < 256) d = md->lcc[d];              if (fc == d || (unsigned int)foc == d) break;
             if (fc == d) break;  
3766              eptr += len;              eptr += len;
3767              }              }
3768          if (possessive) continue;            if (possessive) continue;
3769          for(;;)            for(;;)
3770              {              {
3771              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);
3772              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
# Line 3465  for (;;) Line 3776  for (;;)
3776            }            }
3777          else          else
3778  #endif  #endif
3779          /* Not UTF-8 mode */          /* Not UTF mode */
3780            {            {
3781            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3782              {              {
# Line 3474  for (;;) Line 3785  for (;;)
3785                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3786                break;                break;
3787                }                }
3788              if (fc == md->lcc[*eptr]) break;              if (fc == *eptr || foc == *eptr) break;
3789              eptr++;              eptr++;
3790              }              }
3791            if (possessive) continue;            if (possessive) continue;
# Line 3486  for (;;) Line 3797  for (;;)
3797              }              }
3798            }            }
3799    
3800          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3801          }          }
3802        /* Control never gets here */        /* Control never gets here */
3803        }        }
# Line 3495  for (;;) Line 3806  for (;;)
3806    
3807      else      else
3808        {        {
3809  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3810        /* UTF-8 mode */        if (utf)
       if (utf8)  
3811          {          {
3812          register unsigned int d;          register unsigned int d;
3813          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3505  for (;;) Line 3815  for (;;)
3815            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3816              {              {
3817              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3818              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3819              }              }
3820            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3821            if (fc == d) MRRETURN(MATCH_NOMATCH);            if (fc == d) RRETURN(MATCH_NOMATCH);
3822            }            }
3823          }          }
3824        else        else
3825  #endif  #endif
3826        /* Not UTF-8 mode */        /* Not UTF mode */
3827          {          {
3828          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3829            {            {
3830            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3831              {              {
3832              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3833              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3834              }              }
3835            if (fc == *eptr++) MRRETURN(MATCH_NOMATCH);            if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
3836            }            }
3837          }          }
3838    
# Line 3530  for (;;) Line 3840  for (;;)
3840    
3841        if (minimize)        if (minimize)
3842          {          {
3843  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3844          /* UTF-8 mode */          if (utf)
         if (utf8)  
3845            {            {
3846            register unsigned int d;            register unsigned int d;
3847            for (fi = min;; fi++)            for (fi = min;; fi++)
3848              {              {
3849              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);
3850              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3851              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3852              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3853                {                {
3854                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3855                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3856                }                }
3857              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3858              if (fc == d) MRRETURN(MATCH_NOMATCH);              if (fc == d) RRETURN(MATCH_NOMATCH);
3859              }              }
3860            }            }
3861          else          else
3862  #endif  #endif
3863          /* Not UTF-8 mode */          /* Not UTF mode */
3864            {            {
3865            for (fi = min;; fi++)            for (fi = min;; fi++)
3866              {              {
3867              RMATCH(eptr, ecode, offset_top, md, eptrb, RM33);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM33);
3868              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3869              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3870              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3871                {                {
3872                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3873                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3874                }                }
3875              if (fc == *eptr++) MRRETURN(MATCH_NOMATCH);              if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
3876              }              }
3877            }            }
3878          /* Control never gets here */          /* Control never gets here */
# Line 3575  for (;;) Line 3884  for (;;)
3884          {          {
3885          pp = eptr;          pp = eptr;
3886    
3887  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3888          /* UTF-8 mode */          if (utf)
         if (utf8)  
3889            {            {
3890            register unsigned int d;            register unsigned int d;
3891            for (i = min; i < max; i++)            for (i = min; i < max; i++)
# Line 3603  for (;;) Line 3911  for (;;)
3911            }            }
3912          else          else
3913  #endif  #endif
3914          /* Not UTF-8 mode */          /* Not UTF mode */
3915            {            {
3916            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3917              {              {
# Line 3624  for (;;) Line 3932  for (;;)
3932              }              }
3933            }            }
3934    
3935          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3936          }          }
3937        }        }
3938      /* Control never gets here */      /* Control never gets here */
# Line 3636  for (;;) Line 3944  for (;;)
3944      case OP_TYPEEXACT:      case OP_TYPEEXACT:
3945      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3946      minimize = TRUE;      minimize = TRUE;
3947      ecode += 3;      ecode += 1 + IMM2_SIZE;
3948      goto REPEATTYPE;      goto REPEATTYPE;
3949    
3950      case OP_TYPEUPTO:      case OP_TYPEUPTO:
# Line 3644  for (;;) Line 3952  for (;;)
3952      min = 0;      min = 0;
3953      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3954      minimize = *ecode == OP_TYPEMINUPTO;      minimize = *ecode == OP_TYPEMINUPTO;
3955      ecode += 3;      ecode += 1 + IMM2_SIZE;
3956      goto REPEATTYPE;      goto REPEATTYPE;
3957    
3958      case OP_TYPEPOSSTAR:      case OP_TYPEPOSSTAR:
# Line 3672  for (;;) Line 3980  for (;;)
3980      possessive = TRUE;      possessive = TRUE;
3981      min = 0;      min = 0;
3982      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3983      ecode += 3;      ecode += 1 + IMM2_SIZE;
3984      goto REPEATTYPE;      goto REPEATTYPE;
3985    
3986      case OP_TYPESTAR:      case OP_TYPESTAR:
# Line 3718  for (;;) Line 4026  for (;;)
4026          switch(prop_type)          switch(prop_type)
4027            {            {
4028            case PT_ANY:            case PT_ANY:
4029            if (prop_fail_result) MRRETURN(MATCH_NOMATCH);            if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4030            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4031              {              {
4032              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4033                {                {
4034                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4035                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4036                }                }
4037              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4038              }              }
# Line 3737  for (;;) Line 4045  for (;;)
4045              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4046                {                {
4047                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4048                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4049                }                }
4050              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4051              chartype = UCD_CHARTYPE(c);              chartype = UCD_CHARTYPE(c);
4052              if ((chartype == ucp_Lu ||              if ((chartype == ucp_Lu ||
4053                   chartype == ucp_Ll ||                   chartype == ucp_Ll ||
4054                   chartype == ucp_Lt) == prop_fail_result)                   chartype == ucp_Lt) == prop_fail_result)
4055                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4056              }              }
4057            break;            break;
4058    
# Line 3754  for (;;) Line 4062  for (;;)
4062              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4063                {                {
4064                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4065                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4066                }                }
4067              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4068              if ((UCD_CATEGORY(c) == prop_value) == prop_fail_result)              if ((UCD_CATEGORY(c) == prop_value) == prop_fail_result)
4069                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4070              }              }
4071            break;            break;
4072    
# Line 3768  for (;;) Line 4076  for (;;)
4076              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4077                {                {
4078                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4079                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4080                }                }
4081              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4082              if ((UCD_CHARTYPE(c) == prop_value) == prop_fail_result)              if ((UCD_CHARTYPE(c) == prop_value) == prop_fail_result)
4083                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4084              }              }
4085            break;            break;
4086    
# Line 3782  for (;;) Line 4090  for (;;)
4090              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4091                {                {
4092                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4093                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4094                }                }
4095              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4096              if ((UCD_SCRIPT(c) == prop_value) == prop_fail_result)              if ((UCD_SCRIPT(c) == prop_value) == prop_fail_result)
4097                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4098              }              }
4099            break;            break;
4100    
# Line 3797  for (;;) Line 4105  for (;;)
4105              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4106                {                {
4107                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4108                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4109                }                }
4110              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4111              category = UCD_CATEGORY(c);              category = UCD_CATEGORY(c);
4112              if ((category == ucp_L || category == ucp_N) == prop_fail_result)              if ((category == ucp_L || category == ucp_N) == prop_fail_result)
4113                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4114              }              }
4115            break;            break;
4116    
# Line 3812  for (;;) Line 4120  for (;;)
4120              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4121                {                {
4122                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4123                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4124                }                }
4125              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4126              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||
4127                   c == CHAR_FF || c == CHAR_CR)                   c == CHAR_FF || c == CHAR_CR)
4128                     == prop_fail_result)                     == prop_fail_result)
4129                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4130              }              }
4131            break;            break;
4132    
# Line 3828  for (;;) Line 4136  for (;;)
4136              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4137                {                {
4138                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4139                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4140                }                }
4141              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4142              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||
4143                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)
4144                     == prop_fail_result)                     == prop_fail_result)
4145                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4146              }              }
4147            break;            break;
4148    
# Line 3845  for (;;) Line 4153  for (;;)
4153              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4154                {                {
4155                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4156                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4157                }                }
4158              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4159              category = UCD_CATEGORY(c);              category = UCD_CATEGORY(c);
4160              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)
4161                     == prop_fail_result)                     == prop_fail_result)
4162                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4163              }              }
4164            break;            break;
4165    
# Line 3872  for (;;) Line 4180  for (;;)
4180            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4181              {              {
4182              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4183              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4184              }              }
4185            GETCHARINCTEST(c, eptr);            else
           if (UCD_CATEGORY(c) == ucp_M) MRRETURN(MATCH_NOMATCH);  
           while (eptr < md->end_subject)  
4186              {              {
4187              int len = 1;              int lgb, rgb;
4188              if (!utf8) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
4189              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
4190              eptr += len;             while (eptr < md->end_subject)
4191                  {
4192                  int len = 1;
4193                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
4194                  rgb = UCD_GRAPHBREAK(c);
4195                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
4196                  lgb = rgb;
4197                  eptr += len;
4198                  }
4199              }              }
4200              CHECK_PARTIAL();
4201            }            }
4202          }          }
4203    
# Line 3891  for (;;) Line 4206  for (;;)
4206    
4207  /* Handle all other cases when the coding is UTF-8 */  /* Handle all other cases when the coding is UTF-8 */
4208    
4209  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4210        if (utf8) switch(ctype)        if (utf) switch(ctype)
4211          {          {
4212          case OP_ANY:          case OP_ANY:
4213          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3900  for (;;) Line 4215  for (;;)
4215            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4216              {              {
4217              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4218              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4219                }
4220              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4221              if (md->partial != 0 &&
4222                  eptr + 1 >= md->end_subject &&
4223                  NLBLOCK->nltype == NLTYPE_FIXED &&
4224                  NLBLOCK->nllen == 2 &&
4225                  *eptr == NLBLOCK->nl[0])
4226                {
4227                md->hitend = TRUE;
4228                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4229              }              }
           if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);  
4230            eptr++;            eptr++;
4231            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4232            }            }
4233          break;          break;
4234    
# Line 3914  for (;;) Line 4238  for (;;)
4238            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4239              {              {
4240              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4241              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4242              }              }
4243            eptr++;            eptr++;
4244            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4245            }            }
4246          break;          break;
4247    
4248          case OP_ANYBYTE:          case OP_ANYBYTE:
4249          if (eptr > md->end_subject - min) MRRETURN(MATCH_NOMATCH);          if (eptr > md->end_subject - min) RRETURN(MATCH_NOMATCH);
4250          eptr += min;          eptr += min;
4251          break;          break;
4252    
# Line 3932  for (;;) Line 4256  for (;;)
4256            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4257              {              {
4258              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4259              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4260              }              }
4261            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4262            switch(c)            switch(c)
4263              {              {
4264              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4265    
4266              case 0x000d:              case CHAR_CR:
4267              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == CHAR_LF) eptr++;
4268              break;              break;
4269    
4270              case 0x000a:              case CHAR_LF:
4271              break;              break;
4272    
4273              case 0x000b:              case CHAR_VT:
4274              case 0x000c:              case CHAR_FF:
4275              case 0x0085:              case CHAR_NEL:
4276    #ifndef EBCDIC
4277              case 0x2028:              case 0x2028:
4278              case 0x2029:              case 0x2029:
4279              if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);  #endif  /* Not EBCDIC */
4280                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4281              break;              break;
4282              }              }
4283            }            }
# Line 3963  for (;;) Line 4289  for (;;)
4289            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4290              {              {
4291              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4292              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4293              }              }
4294            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4295            switch(c)            switch(c)
4296              {              {
4297                HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
4298              default: break;              default: break;
             case 0x09:      /* HT */  
             case 0x20:      /* SPACE */  
             case 0xa0:      /* NBSP */  
             case 0x1680:    /* OGHAM SPACE MARK */  
             case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
             case 0x2000:    /* EN QUAD */  
             case 0x2001:    /* EM QUAD */  
             case 0x2002:    /* EN SPACE */  
             case 0x2003:    /* EM SPACE */  
             case 0x2004:    /* THREE-PER-EM SPACE */  
             case 0x2005:    /* FOUR-PER-EM SPACE */  
             case 0x2006:    /* SIX-PER-EM SPACE */  
             case 0x2007:    /* FIGURE SPACE */  
             case 0x2008:    /* PUNCTUATION SPACE */  
             case 0x2009:    /* THIN SPACE */  
             case 0x200A:    /* HAIR SPACE */  
             case 0x202f:    /* NARROW NO-BREAK SPACE */  
             case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
             case 0x3000:    /* IDEOGRAPHIC SPACE */  
             MRRETURN(MATCH_NOMATCH);  
4299              }              }
4300            }            }
4301          break;          break;
# Line 3999  for (;;) Line 4306  for (;;)
4306            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4307              {              {
4308              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4309              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4310              }              }
4311            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4312            switch(c)            switch(c)
4313              {              {
4314              default: MRRETURN(MATCH_NOMATCH);              HSPACE_CASES: break;  /* Byte and multibyte cases */
4315              case 0x09:      /* HT */              default: RRETURN(MATCH_NOMATCH);
             case 0x20:      /* SPACE */  
             case 0xa0:      /* NBSP */  
             case 0x1680:    /* OGHAM SPACE MARK */  
             case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
             case 0x2000:    /* EN QUAD */  
             case 0x2001:    /* EM QUAD */  
             case 0x2002:    /* EN SPACE */  
             case 0x2003:    /* EM SPACE */  
             case 0x2004:    /* THREE-PER-EM SPACE */  
             case 0x2005:    /* FOUR-PER-EM SPACE */  
             case 0x2006:    /* SIX-PER-EM SPACE */  
             case 0x2007:    /* FIGURE SPACE */  
             case 0x2008:    /* PUNCTUATION SPACE */  
             case 0x2009:    /* THIN SPACE */  
             case 0x200A:    /* HAIR SPACE */  
             case 0x202f:    /* NARROW NO-BREAK SPACE */  
             case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
             case 0x3000:    /* IDEOGRAPHIC SPACE */  
             break;  
4316              }              }
4317            }            }
4318          break;          break;
# Line 4035  for (;;) Line 4323  for (;;)
4323            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4324              {              {
4325              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4326              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4327              }              }
4328            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4329            switch(c)            switch(c)
4330              {              {
4331                VSPACE_CASES: RRETURN(MATCH_NOMATCH);
4332              default: break;              default: break;
             case 0x0a:      /* LF */  
             case 0x0b:      /* VT */  
             case 0x0c:      /* FF */  
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
             case 0x2028:    /* LINE SEPARATOR */  
             case 0x2029:    /* PARAGRAPH SEPARATOR */  
             MRRETURN(MATCH_NOMATCH);  
4333              }              }
4334            }            }
4335          break;          break;
# Line 4059  for (;;) Line 4340  for (;;)
4340            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4341              {              {
4342              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4343              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4344              }              }
4345            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4346            switch(c)            switch(c)
4347              {              {
4348              default: MRRETURN(MATCH_NOMATCH);              VSPACE_CASES: break;
4349              case 0x0a:      /* LF */              default: RRETURN(MATCH_NOMATCH);
             case 0x0b:      /* VT */  
             case 0x0c:      /* FF */  
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
             case 0x2028:    /* LINE SEPARATOR */  
             case 0x2029:    /* PARAGRAPH SEPARATOR */  
             break;  
4350              }              }
4351            }            }
4352          break;          break;
# Line 4083  for (;;) Line 4357  for (;;)
4357            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4358              {              {
4359              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4360              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4361              }              }
4362            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4363            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)
4364              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4365            }            }
4366          break;          break;
4367    
# Line 4097  for (;;) Line 4371  for (;;)
4371            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4372              {              {
4373              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4374              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4375              }              }
4376            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_digit) == 0)            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_digit) == 0)
4377              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4378              eptr++;
4379            /* 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 */
4380            }            }
4381          break;          break;
# Line 4111  for (;;) Line 4386  for (;;)
4386            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4387              {              {
4388              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4389              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4390              }              }
4391            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)
4392              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4393            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);            eptr++;
4394              ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4395            }            }
4396          break;          break;
4397    
# Line 4125  for (;;) Line 4401  for (;;)
4401            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4402              {              {
4403              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4404              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4405              }              }
4406            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_space) == 0)            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_space) == 0)
4407              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4408              eptr++;
4409            /* 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 */
4410            }            }
4411          break;          break;
# Line 4139  for (;;) Line 4416  for (;;)
4416            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4417              {              {
4418              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4419              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4420              }              }
4421            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)
4422              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4423            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);            eptr++;
4424              ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4425            }            }
4426          break;          break;
4427    
# Line 4153  for (;;) Line 4431  for (;;)
4431            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4432              {              {
4433              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4434              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4435              }              }
4436            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_word) == 0)            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_word) == 0)
4437              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4438              eptr++;
4439            /* 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 */
4440            }            }
4441          break;          break;
# Line 4166  for (;;) Line 4445  for (;;)
4445          }  /* End switch(ctype) */          }  /* End switch(ctype) */
4446    
4447        else        else
4448  #endif     /* SUPPORT_UTF8 */  #endif     /* SUPPORT_UTF */
4449    
4450        /* 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
4451        than OP_PROP and OP_NOTPROP. */        than OP_PROP and OP_NOTPROP. */
# Line 4179  for (;;) Line 4458  for (;;)
4458            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4459              {              {
4460              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4461              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4462                }
4463              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4464              if (md->partial != 0 &&
4465                  eptr + 1 >= md->end_subject &&
4466                  NLBLOCK->nltype == NLTYPE_FIXED &&
4467                  NLBLOCK->nllen == 2 &&
4468                  *eptr == NLBLOCK->nl[0])
4469                {
4470                md->hitend = TRUE;
4471                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4472              }              }
           if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);  
4473            eptr++;            eptr++;
4474            }            }
4475          break;          break;
# Line 4190  for (;;) Line 4478  for (;;)
4478          if (eptr > md->end_subject - min)          if (eptr > md->end_subject - min)
4479            {            {
4480            SCHECK_PARTIAL();            SCHECK_PARTIAL();
4481            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
4482            }            }
4483          eptr += min;          eptr += min;
4484          break;          break;
# Line 4199  for (;;) Line 4487  for (;;)
4487          if (eptr > md->end_subject - min)          if (eptr > md->end_subject - min)
4488            {            {
4489            SCHECK_PARTIAL();            SCHECK_PARTIAL();
4490            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
4491            }            }
4492          eptr += min;          eptr += min;
4493          break;          break;
# Line 4210  for (;;) Line 4498  for (;;)
4498            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4499              {              {
4500              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4501              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4502              }              }
4503            switch(*eptr++)            switch(*eptr++)
4504              {              {
4505              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4506    
4507              case 0x000d:              case CHAR_CR:
4508              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == CHAR_LF) eptr++;
4509              break;              break;
4510    
4511              case 0x000a:              case CHAR_LF:
4512              break;              break;
4513    
4514              case 0x000b:              case CHAR_VT:
4515              case 0x000c:              case CHAR_FF:
4516              case 0x0085:              case CHAR_NEL:
4517              if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);  #ifdef COMPILE_PCRE16
4518                case 0x2028:
4519                case 0x2029:
4520