/[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 667 by ph10, Mon Aug 22 14:57:32 2011 UTC revision 937 by ph10, Sun Feb 26 15:58:56 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 82  negative to avoid the external error cod Line 82  negative to avoid the external error cod
82  #define MATCH_SKIP_ARG     (-993)  #define MATCH_SKIP_ARG     (-993)
83  #define MATCH_THEN         (-992)  #define MATCH_THEN         (-992)
84    
 /* This is a convenience macro for code that occurs many times. */  
   
 #define MRRETURN(ra) \  
   { \  
   md->mark = markptr; \  
   RRETURN(ra); \  
   }  
   
85  /* 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.
86  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,
87  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 113  Returns:     nothing
113  */  */
114    
115  static void  static void
116  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)
117  {  {
118  unsigned int c;  unsigned int c;
119  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 140  Arguments:
140    md          points to match data block    md          points to match data block
141    caseless    TRUE if caseless    caseless    TRUE if caseless
142    
143  Returns:      < 0 if not matched, otherwise the number of subject bytes matched  Returns:      >= 0 the number of subject bytes matched
144                  -1 no match
145                  -2 partial match; always given if at end subject
146  */  */
147    
148  static int  static int
149  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,
150    BOOL caseless)    BOOL caseless)
151  {  {
152  USPTR eptr_start = eptr;  PCRE_PUCHAR eptr_start = eptr;
153  register USPTR p = md->start_subject + md->offset_vector[offset];  register PCRE_PUCHAR p = md->start_subject + md->offset_vector[offset];
154    
155  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
156  if (eptr >= md->end_subject)  if (eptr >= md->end_subject)
# Line 171  pchars(p, length, FALSE, md); Line 165  pchars(p, length, FALSE, md);
165  printf("\n");  printf("\n");
166  #endif  #endif
167    
168  /* Always fail if reference not set (and not JavaScript compatible). */  /* Always fail if reference not set (and not JavaScript compatible - in that
169    case the length is passed as zero). */
170    
171  if (length < 0) return -1;  if (length < 0) return -1;
172    
# Line 181  ASCII characters. */ Line 176  ASCII characters. */
176    
177  if (caseless)  if (caseless)
178    {    {
179  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
180  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
181    if (md->utf8)    if (md->utf)
182      {      {
183      /* 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
184      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 188  if (caseless)
188      the latter. It is important, therefore, to check the length along the      the latter. It is important, therefore, to check the length along the
189      reference, not along the subject (earlier code did this wrong). */      reference, not along the subject (earlier code did this wrong). */
190    
191      USPTR endptr = p + length;      PCRE_PUCHAR endptr = p + length;
192      while (p < endptr)      while (p < endptr)
193        {        {
194        int c, d;        int c, d;
195        if (eptr >= md->end_subject) return -1;        if (eptr >= md->end_subject) return -2;   /* Partial match */
196        GETCHARINC(c, eptr);        GETCHARINC(c, eptr);
197        GETCHARINC(d, p);        GETCHARINC(d, p);
198        if (c != d && c != UCD_OTHERCASE(d)) return -1;        if (c != d && c != UCD_OTHERCASE(d)) return -1;
# Line 210  if (caseless) Line 205  if (caseless)
205    /* 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
206    is no UCP support. */    is no UCP support. */
207      {      {
     if (eptr + length > md->end_subject) return -1;  
208      while (length-- > 0)      while (length-- > 0)
209        { if (md->lcc[*p++] != md->lcc[*eptr++]) return -1; }        {
210          if (eptr >= md->end_subject) return -2;   /* Partial match */
211          if (TABLE_GET(*p, md->lcc, *p) != TABLE_GET(*eptr, md->lcc, *eptr)) return -1;
212          p++;
213          eptr++;
214          }
215      }      }
216    }    }
217    
# Line 221  are in UTF-8 mode. */ Line 220  are in UTF-8 mode. */
220    
221  else  else
222    {    {
223    if (eptr + length > md->end_subject) return -1;    while (length-- > 0)
224    while (length-- > 0) if (*p++ != *eptr++) return -1;      {
225        if (eptr >= md->end_subject) return -2;   /* Partial match */
226        if (*p++ != *eptr++) return -1;
227        }
228    }    }
229    
230  return eptr - eptr_start;  return (int)(eptr - eptr_start);
231  }  }
232    
233    
# Line 277  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM Line 279  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM
279         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,
280         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,
281         RM51,  RM52, RM53, RM54, RM55, RM56, RM57, RM58, RM59, RM60,         RM51,  RM52, RM53, RM54, RM55, RM56, RM57, RM58, RM59, RM60,
282         RM61,  RM62, RM63 };         RM61,  RM62, RM63, RM64, RM65, RM66 };
283    
284  /* 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
285  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 292  actually used in this definition. */
292  #define RMATCH(ra,rb,rc,rd,re,rw) \  #define RMATCH(ra,rb,rc,rd,re,rw) \
293    { \    { \
294    printf("match() called in line %d\n", __LINE__); \    printf("match() called in line %d\n", __LINE__); \
295    rrc = match(ra,rb,mstart,markptr,rc,rd,re,rdepth+1); \    rrc = match(ra,rb,mstart,rc,rd,re,rdepth+1); \
296    printf("to line %d\n", __LINE__); \    printf("to line %d\n", __LINE__); \
297    }    }
298  #define RRETURN(ra) \  #define RRETURN(ra) \
# Line 300  actually used in this definition. */ Line 302  actually used in this definition. */
302    }    }
303  #else  #else
304  #define RMATCH(ra,rb,rc,rd,re,rw) \  #define RMATCH(ra,rb,rc,rd,re,rw) \
305    rrc = match(ra,rb,mstart,markptr,rc,rd,re,rdepth+1)    rrc = match(ra,rb,mstart,rc,rd,re,rdepth+1)
306  #define RRETURN(ra) return ra  #define RRETURN(ra) return ra
307  #endif  #endif
308    
# Line 315  argument of match(), which never changes Line 317  argument of match(), which never changes
317    
318  #define RMATCH(ra,rb,rc,rd,re,rw)\  #define RMATCH(ra,rb,rc,rd,re,rw)\
319    {\    {\
320    heapframe *newframe = (heapframe *)(pcre_stack_malloc)(sizeof(heapframe));\    heapframe *newframe = frame->Xnextframe;\
321    if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\    if (newframe == NULL)\
322    frame->Xwhere = rw; \      {\
323        newframe = (heapframe *)(PUBL(stack_malloc))(sizeof(heapframe));\
324        if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\
325        newframe->Xnextframe = NULL;\
326        frame->Xnextframe = newframe;\
327        }\
328      frame->Xwhere = rw;\
329    newframe->Xeptr = ra;\    newframe->Xeptr = ra;\
330    newframe->Xecode = rb;\    newframe->Xecode = rb;\
331    newframe->Xmstart = mstart;\    newframe->Xmstart = mstart;\
   newframe->Xmarkptr = markptr;\  
332    newframe->Xoffset_top = rc;\    newframe->Xoffset_top = rc;\
333    newframe->Xeptrb = re;\    newframe->Xeptrb = re;\
334    newframe->Xrdepth = frame->Xrdepth + 1;\    newframe->Xrdepth = frame->Xrdepth + 1;\
# Line 337  argument of match(), which never changes Line 344  argument of match(), which never changes
344    {\    {\
345    heapframe *oldframe = frame;\    heapframe *oldframe = frame;\
346    frame = oldframe->Xprevframe;\    frame = oldframe->Xprevframe;\
   (pcre_stack_free)(oldframe);\  
347    if (frame != NULL)\    if (frame != NULL)\
348      {\      {\
349      rrc = ra;\      rrc = ra;\
# Line 351  argument of match(), which never changes Line 357  argument of match(), which never changes
357    
358  typedef struct heapframe {  typedef struct heapframe {
359    struct heapframe *Xprevframe;    struct heapframe *Xprevframe;
360      struct heapframe *Xnextframe;
361    
362    /* Function arguments that may change */    /* Function arguments that may change */
363    
364    USPTR Xeptr;    PCRE_PUCHAR Xeptr;
365    const uschar *Xecode;    const pcre_uchar *Xecode;
366    USPTR Xmstart;    PCRE_PUCHAR Xmstart;
   USPTR Xmarkptr;  
367    int Xoffset_top;    int Xoffset_top;
368    eptrblock *Xeptrb;    eptrblock *Xeptrb;
369    unsigned int Xrdepth;    unsigned int Xrdepth;
370    
371    /* Function local variables */    /* Function local variables */
372    
373    USPTR Xcallpat;    PCRE_PUCHAR Xcallpat;
374  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
375    USPTR Xcharptr;    PCRE_PUCHAR Xcharptr;
376  #endif  #endif
377    USPTR Xdata;    PCRE_PUCHAR Xdata;
378    USPTR Xnext;    PCRE_PUCHAR Xnext;
379    USPTR Xpp;    PCRE_PUCHAR Xpp;
380    USPTR Xprev;    PCRE_PUCHAR Xprev;
381    USPTR Xsaved_eptr;    PCRE_PUCHAR Xsaved_eptr;
382    
383    recursion_info Xnew_recursive;    recursion_info Xnew_recursive;
384    
# Line 385  typedef struct heapframe { Line 391  typedef struct heapframe {
391    int Xprop_value;    int Xprop_value;
392    int Xprop_fail_result;    int Xprop_fail_result;
393    int Xoclength;    int Xoclength;
394    uschar Xocchars[8];    pcre_uchar Xocchars[6];
395  #endif  #endif
396    
397    int Xcodelink;    int Xcodelink;
# Line 427  returns a negative (error) response, the Line 433  returns a negative (error) response, the
433  same response. */  same response. */
434    
435  /* 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
436  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
437  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.
438  something has been matched). For hard partial matching, we then return  something has been matched). For hard partial matching, we then return
439  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 444  the subject. */
444        eptr > md->start_used_ptr) \        eptr > md->start_used_ptr) \
445      { \      { \
446      md->hitend = TRUE; \      md->hitend = TRUE; \
447      if (md->partial > 1) MRRETURN(PCRE_ERROR_PARTIAL); \      if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL); \
448      }      }
449    
450  #define SCHECK_PARTIAL()\  #define SCHECK_PARTIAL()\
451    if (md->partial != 0 && eptr > md->start_used_ptr) \    if (md->partial != 0 && eptr > md->start_used_ptr) \
452      { \      { \
453      md->hitend = TRUE; \      md->hitend = TRUE; \
454      if (md->partial > 1) MRRETURN(PCRE_ERROR_PARTIAL); \      if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL); \
455      }      }
456    
457    
458  /* Performance note: It might be tempting to extract commonly used fields from  /* Performance note: It might be tempting to extract commonly used fields from
459  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
460  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
461  made performance worse.  made performance worse.
462    
# Line 459  Arguments: Line 465  Arguments:
465     ecode       pointer to current position in compiled code     ecode       pointer to current position in compiled code
466     mstart      pointer to the current match start position (can be modified     mstart      pointer to the current match start position (can be modified
467                   by encountering \K)                   by encountering \K)
    markptr     pointer to the most recent MARK name, or NULL  
468     offset_top  current top pointer     offset_top  current top pointer
469     md          pointer to "static" info for the match     md          pointer to "static" info for the match
470     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 479  Returns:       MATCH_MATCH if matched
479  */  */
480    
481  static int  static int
482  match(REGISTER USPTR eptr, REGISTER const uschar *ecode, USPTR mstart,  match(REGISTER PCRE_PUCHAR eptr, REGISTER const pcre_uchar *ecode,
483    const uschar *markptr, int offset_top, match_data *md, eptrblock *eptrb,    PCRE_PUCHAR mstart, int offset_top, match_data *md, eptrblock *eptrb,
484    unsigned int rdepth)    unsigned int rdepth)
485  {  {
486  /* 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 490  so they can be ordinary variables in all
490  register int  rrc;         /* Returns from recursive calls */  register int  rrc;         /* Returns from recursive calls */
491  register int  i;           /* Used for loops not involving calls to RMATCH() */  register int  i;           /* Used for loops not involving calls to RMATCH() */
492  register unsigned int c;   /* Character values not kept over RMATCH() calls */  register unsigned int c;   /* Character values not kept over RMATCH() calls */
493  register BOOL utf8;        /* Local copy of UTF-8 flag for speed */  register BOOL utf;         /* Local copy of UTF flag for speed */
494    
495  BOOL minimize, possessive; /* Quantifier options */  BOOL minimize, possessive; /* Quantifier options */
496  BOOL caseless;  BOOL caseless;
497  int condcode;  int condcode;
498    
499  /* 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
500  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
501  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
502  heap whenever RMATCH() does a "recursion". See the macro definitions above. */  whenever RMATCH() does a "recursion". See the macro definitions above. Putting
503    the top-level on the stack rather than malloc-ing them all gives a performance
504    boost in many cases where there is not much "recursion". */
505    
506  #ifdef NO_RECURSE  #ifdef NO_RECURSE
507  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 */  
508    
509  /* Copy in the original argument variables */  /* Copy in the original argument variables */
510    
511  frame->Xeptr = eptr;  frame->Xeptr = eptr;
512  frame->Xecode = ecode;  frame->Xecode = ecode;
513  frame->Xmstart = mstart;  frame->Xmstart = mstart;
 frame->Xmarkptr = markptr;  
514  frame->Xoffset_top = offset_top;  frame->Xoffset_top = offset_top;
515  frame->Xeptrb = eptrb;  frame->Xeptrb = eptrb;
516  frame->Xrdepth = rdepth;  frame->Xrdepth = rdepth;
# Line 520  HEAP_RECURSE: Line 524  HEAP_RECURSE:
524  #define eptr               frame->Xeptr  #define eptr               frame->Xeptr
525  #define ecode              frame->Xecode  #define ecode              frame->Xecode
526  #define mstart             frame->Xmstart  #define mstart             frame->Xmstart
 #define markptr            frame->Xmarkptr  
527  #define offset_top         frame->Xoffset_top  #define offset_top         frame->Xoffset_top
528  #define eptrb              frame->Xeptrb  #define eptrb              frame->Xeptrb
529  #define rdepth             frame->Xrdepth  #define rdepth             frame->Xrdepth
530    
531  /* Ditto for the local variables */  /* Ditto for the local variables */
532    
533  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
534  #define charptr            frame->Xcharptr  #define charptr            frame->Xcharptr
535  #endif  #endif
536  #define callpat            frame->Xcallpat  #define callpat            frame->Xcallpat
# Line 585  declarations can be cut out in a block. Line 588  declarations can be cut out in a block.
588  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
589  to RMATCH(). */  to RMATCH(). */
590    
591  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
592  const uschar *charptr;  const pcre_uchar *charptr;
593  #endif  #endif
594  const uschar *callpat;  const pcre_uchar *callpat;
595  const uschar *data;  const pcre_uchar *data;
596  const uschar *next;  const pcre_uchar *next;
597  USPTR         pp;  PCRE_PUCHAR       pp;
598  const uschar *prev;  const pcre_uchar *prev;
599  USPTR         saved_eptr;  PCRE_PUCHAR       saved_eptr;
600    
601  recursion_info new_recursive;  recursion_info new_recursive;
602    
# Line 606  int prop_type; Line 609  int prop_type;
609  int prop_value;  int prop_value;
610  int prop_fail_result;  int prop_fail_result;
611  int oclength;  int oclength;
612  uschar occhars[8];  pcre_uchar occhars[6];
613  #endif  #endif
614    
615  int codelink;  int codelink;
# Line 622  int save_offset1, save_offset2, save_off Line 625  int save_offset1, save_offset2, save_off
625  int stacksave[REC_STACK_SAVE_MAX];  int stacksave[REC_STACK_SAVE_MAX];
626    
627  eptrblock newptrb;  eptrblock newptrb;
628    
629    /* There is a special fudge for calling match() in a way that causes it to
630    measure the size of its basic stack frame when the stack is being used for
631    recursion. The second argument (ecode) being NULL triggers this behaviour. It
632    cannot normally ever be NULL. The return is the negated value of the frame
633    size. */
634    
635    if (ecode == NULL)
636      {
637      if (rdepth == 0)
638        return match((PCRE_PUCHAR)&rdepth, NULL, NULL, 0, NULL, NULL, 1);
639      else
640        {
641        int len = (char *)&rdepth - (char *)eptr;
642        return (len > 0)? -len : len;
643        }
644      }
645  #endif     /* NO_RECURSE */  #endif     /* NO_RECURSE */
646    
647  /* 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 654  the alternative names that are used. */
654  #define code_offset   codelink  #define code_offset   codelink
655  #define condassert    condition  #define condassert    condition
656  #define matched_once  prev_is_word  #define matched_once  prev_is_word
657    #define foc           number
658    #define save_mark     data
659    
660  /* These statements are here to stop the compiler complaining about unitialized  /* These statements are here to stop the compiler complaining about unitialized
661  variables. */  variables. */
# Line 659  defined). However, RMATCH isn't like a f Line 681  defined). However, RMATCH isn't like a f
681  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,
682  however, impact performance when true recursion is being used. */  however, impact performance when true recursion is being used. */
683    
684  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
685  utf8 = md->utf8;       /* Local copy of the flag */  utf = md->utf;       /* Local copy of the flag */
686  #else  #else
687  utf8 = FALSE;  utf = FALSE;
688  #endif  #endif
689    
690  /* 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 723  for (;;)
723    switch(op)    switch(op)
724      {      {
725      case OP_MARK:      case OP_MARK:
726      markptr = ecode + 2;      md->nomatch_mark = ecode + 2;
727      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      md->mark = NULL;    /* In case previously set by assertion */
728        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
729        eptrb, RM55);        eptrb, RM55);
730        if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
731             md->mark == NULL) md->mark = ecode + 2;
732    
733      /* 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
734      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 737  for (;;)
737      position and return MATCH_SKIP. Otherwise, pass back the return code      position and return MATCH_SKIP. Otherwise, pass back the return code
738      unaltered. */      unaltered. */
739    
740      if (rrc == MATCH_SKIP_ARG &&      else if (rrc == MATCH_SKIP_ARG &&
741          strcmp((char *)markptr, (char *)(md->start_match_ptr)) == 0)          STRCMP_UC_UC(ecode + 2, md->start_match_ptr) == 0)
742        {        {
743        md->start_match_ptr = eptr;        md->start_match_ptr = eptr;
744        RRETURN(MATCH_SKIP);        RRETURN(MATCH_SKIP);
745        }        }
   
     if (md->mark == NULL) md->mark = markptr;  
746      RRETURN(rrc);      RRETURN(rrc);
747    
748      case OP_FAIL:      case OP_FAIL:
749      MRRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
750    
751      /* COMMIT overrides PRUNE, SKIP, and THEN */      /* COMMIT overrides PRUNE, SKIP, and THEN */
752    
753      case OP_COMMIT:      case OP_COMMIT:
754      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
755        eptrb, RM52);        eptrb, RM52);
756      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE &&      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE &&
757          rrc != MATCH_SKIP && rrc != MATCH_SKIP_ARG &&          rrc != MATCH_SKIP && rrc != MATCH_SKIP_ARG &&
758          rrc != MATCH_THEN)          rrc != MATCH_THEN)
759        RRETURN(rrc);        RRETURN(rrc);
760      MRRETURN(MATCH_COMMIT);      RRETURN(MATCH_COMMIT);
761    
762      /* PRUNE overrides THEN */      /* PRUNE overrides THEN */
763    
764      case OP_PRUNE:      case OP_PRUNE:
765      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
766        eptrb, RM51);        eptrb, RM51);
767      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
768      MRRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
769    
770      case OP_PRUNE_ARG:      case OP_PRUNE_ARG:
771      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      md->nomatch_mark = ecode + 2;
772        md->mark = NULL;    /* In case previously set by assertion */
773        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
774        eptrb, RM56);        eptrb, RM56);
775        if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
776             md->mark == NULL) md->mark = ecode + 2;
777      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
     md->mark = ecode + 2;  
778      RRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
779    
780      /* SKIP overrides PRUNE and THEN */      /* SKIP overrides PRUNE and THEN */
781    
782      case OP_SKIP:      case OP_SKIP:
783      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
784        eptrb, RM53);        eptrb, RM53);
785      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)
786        RRETURN(rrc);        RRETURN(rrc);
787      md->start_match_ptr = eptr;   /* Pass back current position */      md->start_match_ptr = eptr;   /* Pass back current position */
788      MRRETURN(MATCH_SKIP);      RRETURN(MATCH_SKIP);
789    
790        /* Note that, for Perl compatibility, SKIP with an argument does NOT set
791        nomatch_mark. There is a flag that disables this opcode when re-matching a
792        pattern that ended with a SKIP for which there was not a matching MARK. */
793    
794      case OP_SKIP_ARG:      case OP_SKIP_ARG:
795      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      if (md->ignore_skip_arg)
796          {
797          ecode += PRIV(OP_lengths)[*ecode] + ecode[1];
798          break;
799          }
800        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
801        eptrb, RM57);        eptrb, RM57);
802      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)
803        RRETURN(rrc);        RRETURN(rrc);
804    
805      /* 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
806      returning the special MATCH_SKIP_ARG return code. This will either be      returning the special MATCH_SKIP_ARG return code. This will either be
807      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
808      as PRUNE. */      with the md->ignore_skip_arg flag set. */
809    
810      md->start_match_ptr = ecode + 2;      md->start_match_ptr = ecode + 2;
811      RRETURN(MATCH_SKIP_ARG);      RRETURN(MATCH_SKIP_ARG);
812    
813      /* 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
814      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
815      to skip back past alternatives that precede the THEN within the current      match pointer to do this. */
     branch. */  
816    
817      case OP_THEN:      case OP_THEN:
818      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
819        eptrb, RM54);        eptrb, RM54);
820      if (rrc != MATCH_NOMATCH) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
821      md->start_match_ptr = ecode - GET(ecode, 1);      md->start_match_ptr = ecode;
822      MRRETURN(MATCH_THEN);      RRETURN(MATCH_THEN);
823    
824      case OP_THEN_ARG:      case OP_THEN_ARG:
825      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1+LINK_SIZE],      md->nomatch_mark = ecode + 2;
826        offset_top, md, eptrb, RM58);      md->mark = NULL;    /* In case previously set by assertion */
827        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top,
828          md, eptrb, RM58);
829        if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
830             md->mark == NULL) md->mark = ecode + 2;
831      if (rrc != MATCH_NOMATCH) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
832      md->start_match_ptr = ecode - GET(ecode, 1);      md->start_match_ptr = ecode;
     md->mark = ecode + LINK_SIZE + 2;  
833      RRETURN(MATCH_THEN);      RRETURN(MATCH_THEN);
834    
835        /* Handle an atomic group that does not contain any capturing parentheses.
836        This can be handled like an assertion. Prior to 8.13, all atomic groups
837        were handled this way. In 8.13, the code was changed as below for ONCE, so
838        that backups pass through the group and thereby reset captured values.
839        However, this uses a lot more stack, so in 8.20, atomic groups that do not
840        contain any captures generate OP_ONCE_NC, which can be handled in the old,
841        less stack intensive way.
842    
843        Check the alternative branches in turn - the matching won't pass the KET
844        for this kind of subpattern. If any one branch matches, we carry on as at
845        the end of a normal bracket, leaving the subject pointer, but resetting
846        the start-of-match value in case it was changed by \K. */
847    
848        case OP_ONCE_NC:
849        prev = ecode;
850        saved_eptr = eptr;
851        save_mark = md->mark;
852        do
853          {
854          RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM64);
855          if (rrc == MATCH_MATCH)  /* Note: _not_ MATCH_ACCEPT */
856            {
857            mstart = md->start_match_ptr;
858            break;
859            }
860          if (rrc == MATCH_THEN)
861            {
862            next = ecode + GET(ecode,1);
863            if (md->start_match_ptr < next &&
864                (*ecode == OP_ALT || *next == OP_ALT))
865              rrc = MATCH_NOMATCH;
866            }
867    
868          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
869          ecode += GET(ecode,1);
870          md->mark = save_mark;
871          }
872        while (*ecode == OP_ALT);
873    
874        /* If hit the end of the group (which could be repeated), fail */
875    
876        if (*ecode != OP_ONCE_NC && *ecode != OP_ALT) RRETURN(MATCH_NOMATCH);
877    
878        /* Continue as from after the group, updating the offsets high water
879        mark, since extracts may have been taken. */
880    
881        do ecode += GET(ecode, 1); while (*ecode == OP_ALT);
882    
883        offset_top = md->end_offset_top;
884        eptr = md->end_match_ptr;
885    
886        /* For a non-repeating ket, just continue at this level. This also
887        happens for a repeating ket if no characters were matched in the group.
888        This is the forcible breaking of infinite loops as implemented in Perl
889        5.005. */
890    
891        if (*ecode == OP_KET || eptr == saved_eptr)
892          {
893          ecode += 1+LINK_SIZE;
894          break;
895          }
896    
897        /* The repeating kets try the rest of the pattern or restart from the
898        preceding bracket, in the appropriate order. The second "call" of match()
899        uses tail recursion, to avoid using another stack frame. */
900    
901        if (*ecode == OP_KETRMIN)
902          {
903          RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM65);
904          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
905          ecode = prev;
906          goto TAIL_RECURSE;
907          }
908        else  /* OP_KETRMAX */
909          {
910          md->match_function_type = MATCH_CBEGROUP;
911          RMATCH(eptr, prev, offset_top, md, eptrb, RM66);
912          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
913          ecode += 1 + LINK_SIZE;
914          goto TAIL_RECURSE;
915          }
916        /* Control never gets here */
917    
918      /* Handle a capturing bracket, other than those that are possessive with an      /* Handle a capturing bracket, other than those that are possessive with an
919      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
920      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 947  for (;;)
947        save_offset2 = md->offset_vector[offset+1];        save_offset2 = md->offset_vector[offset+1];
948        save_offset3 = md->offset_vector[md->offset_end - number];        save_offset3 = md->offset_vector[md->offset_end - number];
949        save_capture_last = md->capture_last;        save_capture_last = md->capture_last;
950          save_mark = md->mark;
951    
952        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));
953        md->offset_vector[md->offset_end - number] =        md->offset_vector[md->offset_end - number] =
# Line 835  for (;;) Line 956  for (;;)
956        for (;;)        for (;;)
957          {          {
958          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
959          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
960            eptrb, RM1);            eptrb, RM1);
961          if (rrc == MATCH_ONCE) break;  /* Backing up through an atomic group */          if (rrc == MATCH_ONCE) break;  /* Backing up through an atomic group */
962          if (rrc != MATCH_NOMATCH &&  
963              (rrc != MATCH_THEN || md->start_match_ptr != ecode))          /* If we backed up to a THEN, check whether it is within the current
964            RRETURN(rrc);          branch by comparing the address of the THEN that is passed back with
965            the end of the branch. If it is within the current branch, and the
966            branch is one of two or more alternatives (it either starts or ends
967            with OP_ALT), we have reached the limit of THEN's action, so convert
968            the return code to NOMATCH, which will cause normal backtracking to
969            happen from now on. Otherwise, THEN is passed back to an outer
970            alternative. This implements Perl's treatment of parenthesized groups,
971            where a group not containing | does not affect the current alternative,
972            that is, (X) is NOT the same as (X|(*F)). */
973    
974            if (rrc == MATCH_THEN)
975              {
976              next = ecode + GET(ecode,1);
977              if (md->start_match_ptr < next &&
978                  (*ecode == OP_ALT || *next == OP_ALT))
979                rrc = MATCH_NOMATCH;
980              }
981    
982            /* Anything other than NOMATCH is passed back. */
983    
984            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
985          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
986          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
987            md->mark = save_mark;
988          if (*ecode != OP_ALT) break;          if (*ecode != OP_ALT) break;
989          }          }
990    
# Line 851  for (;;) Line 993  for (;;)
993        md->offset_vector[offset+1] = save_offset2;        md->offset_vector[offset+1] = save_offset2;
994        md->offset_vector[md->offset_end - number] = save_offset3;        md->offset_vector[md->offset_end - number] = save_offset3;
995    
996        /* 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. */  
997    
998        if (rrc != MATCH_THEN && md->mark == NULL) md->mark = markptr;        RRETURN(rrc);
       RRETURN(((rrc == MATCH_ONCE)? MATCH_ONCE:MATCH_NOMATCH));  
999        }        }
1000    
1001      /* FALL THROUGH ... Insufficient room for saving captured contents. Treat      /* FALL THROUGH ... Insufficient room for saving captured contents. Treat
# Line 870  for (;;) Line 1010  for (;;)
1010      /* VVVVVVVVVVVVVVVVVVVVVVVVV */      /* VVVVVVVVVVVVVVVVVVVVVVVVV */
1011    
1012      /* Non-capturing or atomic group, except for possessive with unlimited      /* Non-capturing or atomic group, except for possessive with unlimited
1013      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.
1014      within the brackets, we used to return the result of a recursive call to  
1015      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
1016      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
1017      group. However, now that there is the possiblity of (*THEN) occurring in      possible to reduce stack usage by turning this into a tail recursion,
1018      the final alternative, this optimization is no longer possible.      except in the case of a possibly empty group. However, now that there is
1019        the possiblity of (*THEN) occurring in the final alternative, this
1020        optimization is no longer always possible.
1021    
1022        We can optimize if we know there are no (*THEN)s in the pattern; at present
1023        this is the best that can be done.
1024    
1025      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
1026      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 892  for (;;) Line 1037  for (;;)
1037      for (;;)      for (;;)
1038        {        {
1039        if (op >= OP_SBRA || op == OP_ONCE) md->match_function_type = MATCH_CBEGROUP;        if (op >= OP_SBRA || op == OP_ONCE) md->match_function_type = MATCH_CBEGROUP;
1040        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, eptrb,  
1041          /* If this is not a possibly empty group, and there are no (*THEN)s in
1042          the pattern, and this is the final alternative, optimize as described
1043          above. */
1044    
1045          else if (!md->hasthen && ecode[GET(ecode, 1)] != OP_ALT)
1046            {
1047            ecode += PRIV(OP_lengths)[*ecode];
1048            goto TAIL_RECURSE;
1049            }
1050    
1051          /* In all other cases, we have to make another call to match(). */
1052    
1053          save_mark = md->mark;
1054          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md, eptrb,
1055          RM2);          RM2);
1056        if (rrc != MATCH_NOMATCH &&  
1057            (rrc != MATCH_THEN || md->start_match_ptr != ecode))        /* See comment in the code for capturing groups above about handling
1058          THEN. */
1059    
1060          if (rrc == MATCH_THEN)
1061            {
1062            next = ecode + GET(ecode,1);
1063            if (md->start_match_ptr < next &&
1064                (*ecode == OP_ALT || *next == OP_ALT))
1065              rrc = MATCH_NOMATCH;
1066            }
1067    
1068          if (rrc != MATCH_NOMATCH)
1069          {          {
1070          if (rrc == MATCH_ONCE)          if (rrc == MATCH_ONCE)
1071            {            {
1072            const uschar *scode = ecode;            const pcre_uchar *scode = ecode;
1073            if (*scode != OP_ONCE)           /* If not at start, find it */            if (*scode != OP_ONCE)           /* If not at start, find it */
1074              {              {
1075              while (*scode == OP_ALT) scode += GET(scode, 1);              while (*scode == OP_ALT) scode += GET(scode, 1);
# Line 910  for (;;) Line 1080  for (;;)
1080          RRETURN(rrc);          RRETURN(rrc);
1081          }          }
1082        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1083          md->mark = save_mark;
1084        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1085        }        }
1086      if (rrc != MATCH_THEN && md->mark == NULL) md->mark = markptr;  
1087      RRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
1088    
1089      /* Handle possessive capturing brackets with an unlimited repeat. We come      /* Handle possessive capturing brackets with an unlimited repeat. We come
# Line 941  for (;;) Line 1112  for (;;)
1112      if (offset < md->offset_max)      if (offset < md->offset_max)
1113        {        {
1114        matched_once = FALSE;        matched_once = FALSE;
1115        code_offset = ecode - md->start_code;        code_offset = (int)(ecode - md->start_code);
1116    
1117        save_offset1 = md->offset_vector[offset];        save_offset1 = md->offset_vector[offset];
1118        save_offset2 = md->offset_vector[offset+1];        save_offset2 = md->offset_vector[offset+1];
# Line 964  for (;;) Line 1135  for (;;)
1135          md->offset_vector[md->offset_end - number] =          md->offset_vector[md->offset_end - number] =
1136            (int)(eptr - md->start_subject);            (int)(eptr - md->start_subject);
1137          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
1138          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
1139            eptrb, RM63);            eptrb, RM63);
1140          if (rrc == MATCH_KETRPOS)          if (rrc == MATCH_KETRPOS)
1141            {            {
# Line 975  for (;;) Line 1146  for (;;)
1146            matched_once = TRUE;            matched_once = TRUE;
1147            continue;            continue;
1148            }            }
1149          if (rrc != MATCH_NOMATCH &&  
1150              (rrc != MATCH_THEN || md->start_match_ptr != ecode))          /* See comment in the code for capturing groups above about handling
1151            RRETURN(rrc);          THEN. */
1152    
1153            if (rrc == MATCH_THEN)
1154              {
1155              next = ecode + GET(ecode,1);
1156              if (md->start_match_ptr < next &&
1157                  (*ecode == OP_ALT || *next == OP_ALT))
1158                rrc = MATCH_NOMATCH;
1159              }
1160    
1161            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1162          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
1163          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
1164          if (*ecode != OP_ALT) break;          if (*ecode != OP_ALT) break;
# Line 990  for (;;) Line 1171  for (;;)
1171          md->offset_vector[md->offset_end - number] = save_offset3;          md->offset_vector[md->offset_end - number] = save_offset3;
1172          }          }
1173    
       if (rrc != MATCH_THEN && md->mark == NULL) md->mark = markptr;  
1174        if (allow_zero || matched_once)        if (allow_zero || matched_once)
1175          {          {
1176          ecode += 1 + LINK_SIZE;          ecode += 1 + LINK_SIZE;
# Line 1022  for (;;) Line 1202  for (;;)
1202    
1203      POSSESSIVE_NON_CAPTURE:      POSSESSIVE_NON_CAPTURE:
1204      matched_once = FALSE;      matched_once = FALSE;
1205      code_offset = ecode - md->start_code;      code_offset = (int)(ecode - md->start_code);
1206    
1207      for (;;)      for (;;)
1208        {        {
1209        if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;        if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
1210        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
1211          eptrb, RM48);          eptrb, RM48);
1212        if (rrc == MATCH_KETRPOS)        if (rrc == MATCH_KETRPOS)
1213          {          {
# Line 1037  for (;;) Line 1217  for (;;)
1217          matched_once = TRUE;          matched_once = TRUE;
1218          continue;          continue;
1219          }          }
1220        if (rrc != MATCH_NOMATCH &&  
1221            (rrc != MATCH_THEN || md->start_match_ptr != ecode))        /* See comment in the code for capturing groups above about handling
1222          RRETURN(rrc);        THEN. */
1223    
1224          if (rrc == MATCH_THEN)
1225            {
1226            next = ecode + GET(ecode,1);
1227            if (md->start_match_ptr < next &&
1228                (*ecode == OP_ALT || *next == OP_ALT))
1229              rrc = MATCH_NOMATCH;
1230            }
1231    
1232          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1233        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1234        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1235        }        }
# Line 1067  for (;;) Line 1257  for (;;)
1257    
1258      if (ecode[LINK_SIZE+1] == OP_CALLOUT)      if (ecode[LINK_SIZE+1] == OP_CALLOUT)
1259        {        {
1260        if (pcre_callout != NULL)        if (PUBL(callout) != NULL)
1261          {          {
1262          pcre_callout_block cb;          PUBL(callout_block) cb;
1263          cb.version          = 2;   /* Version 1 of the callout block */          cb.version          = 2;   /* Version 1 of the callout block */
1264          cb.callout_number   = ecode[LINK_SIZE+2];          cb.callout_number   = ecode[LINK_SIZE+2];
1265          cb.offset_vector    = md->offset_vector;          cb.offset_vector    = md->offset_vector;
1266    #ifdef COMPILE_PCRE8
1267          cb.subject          = (PCRE_SPTR)md->start_subject;          cb.subject          = (PCRE_SPTR)md->start_subject;
1268    #else
1269            cb.subject          = (PCRE_SPTR16)md->start_subject;
1270    #endif
1271          cb.subject_length   = (int)(md->end_subject - md->start_subject);          cb.subject_length   = (int)(md->end_subject - md->start_subject);
1272          cb.start_match      = (int)(mstart - md->start_subject);          cb.start_match      = (int)(mstart - md->start_subject);
1273          cb.current_position = (int)(eptr - md->start_subject);          cb.current_position = (int)(eptr - md->start_subject);
# Line 1082  for (;;) Line 1276  for (;;)
1276          cb.capture_top      = offset_top/2;          cb.capture_top      = offset_top/2;
1277          cb.capture_last     = md->capture_last;          cb.capture_last     = md->capture_last;
1278          cb.callout_data     = md->callout_data;          cb.callout_data     = md->callout_data;
1279          cb.mark             = markptr;          cb.mark             = md->nomatch_mark;
1280          if ((rrc = (*pcre_callout)(&cb)) > 0) MRRETURN(MATCH_NOMATCH);          if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1281          if (rrc < 0) RRETURN(rrc);          if (rrc < 0) RRETURN(rrc);
1282          }          }
1283        ecode += _pcre_OP_lengths[OP_CALLOUT];        ecode += PRIV(OP_lengths)[OP_CALLOUT];
1284        }        }
1285    
1286      condcode = ecode[LINK_SIZE+1];      condcode = ecode[LINK_SIZE+1];
# Line 1103  for (;;) Line 1297  for (;;)
1297        else        else
1298          {          {
1299          int recno = GET2(ecode, LINK_SIZE + 2);   /* Recursion group number*/          int recno = GET2(ecode, LINK_SIZE + 2);   /* Recursion group number*/
1300          condition =  (recno == RREF_ANY || recno == md->recursive->group_num);          condition = (recno == RREF_ANY || recno == md->recursive->group_num);
1301    
1302          /* 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
1303          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
1304          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
1305          if any one is set. */          if any one is set. */
1306    
1307          if (!condition && condcode == OP_NRREF && recno != RREF_ANY)          if (!condition && condcode == OP_NRREF)
1308            {            {
1309            uschar *slotA = md->name_table;            pcre_uchar *slotA = md->name_table;
1310            for (i = 0; i < md->name_count; i++)            for (i = 0; i < md->name_count; i++)
1311              {              {
1312              if (GET2(slotA, 0) == recno) break;              if (GET2(slotA, 0) == recno) break;
# Line 1125  for (;;) Line 1319  for (;;)
1319    
1320            if (i < md->name_count)            if (i < md->name_count)
1321              {              {
1322              uschar *slotB = slotA;              pcre_uchar *slotB = slotA;
1323              while (slotB > md->name_table)              while (slotB > md->name_table)
1324                {                {
1325                slotB -= md->name_entry_size;                slotB -= md->name_entry_size;
1326                if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)                if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1327                  {                  {
1328                  condition = GET2(slotB, 0) == md->recursive->group_num;                  condition = GET2(slotB, 0) == md->recursive->group_num;
1329                  if (condition) break;                  if (condition) break;
# Line 1145  for (;;) Line 1339  for (;;)
1339                for (i++; i < md->name_count; i++)                for (i++; i < md->name_count; i++)
1340                  {                  {
1341                  slotB += md->name_entry_size;                  slotB += md->name_entry_size;
1342                  if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)                  if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1343                    {                    {
1344                    condition = GET2(slotB, 0) == md->recursive->group_num;                    condition = GET2(slotB, 0) == md->recursive->group_num;
1345                    if (condition) break;                    if (condition) break;
# Line 1158  for (;;) Line 1352  for (;;)
1352    
1353          /* Chose branch according to the condition */          /* Chose branch according to the condition */
1354    
1355          ecode += condition? 3 : GET(ecode, 1);          ecode += condition? 1 + IMM2_SIZE : GET(ecode, 1);
1356          }          }
1357        }        }
1358    
# Line 1175  for (;;) Line 1369  for (;;)
1369        if (!condition && condcode == OP_NCREF)        if (!condition && condcode == OP_NCREF)
1370          {          {
1371          int refno = offset >> 1;          int refno = offset >> 1;
1372          uschar *slotA = md->name_table;          pcre_uchar *slotA = md->name_table;
1373    
1374          for (i = 0; i < md->name_count; i++)          for (i = 0; i < md->name_count; i++)
1375            {            {
# Line 1189  for (;;) Line 1383  for (;;)
1383    
1384          if (i < md->name_count)          if (i < md->name_count)
1385            {            {
1386            uschar *slotB = slotA;            pcre_uchar *slotB = slotA;
1387            while (slotB > md->name_table)            while (slotB > md->name_table)
1388              {              {
1389              slotB -= md->name_entry_size;              slotB -= md->name_entry_size;
1390              if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)              if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1391                {                {
1392                offset = GET2(slotB, 0) << 1;                offset = GET2(slotB, 0) << 1;
1393                condition = offset < offset_top &&                condition = offset < offset_top &&
# Line 1211  for (;;) Line 1405  for (;;)
1405              for (i++; i < md->name_count; i++)              for (i++; i < md->name_count; i++)
1406                {                {
1407                slotB += md->name_entry_size;                slotB += md->name_entry_size;
1408                if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)                if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1409                  {                  {
1410                  offset = GET2(slotB, 0) << 1;                  offset = GET2(slotB, 0) << 1;
1411                  condition = offset < offset_top &&                  condition = offset < offset_top &&
# Line 1226  for (;;) Line 1420  for (;;)
1420    
1421        /* Chose branch according to the condition */        /* Chose branch according to the condition */
1422    
1423        ecode += condition? 3 : GET(ecode, 1);        ecode += condition? 1 + IMM2_SIZE : GET(ecode, 1);
1424        }        }
1425    
1426      else if (condcode == OP_DEF)     /* DEFINE - always false */      else if (condcode == OP_DEF)     /* DEFINE - always false */
# Line 1251  for (;;) Line 1445  for (;;)
1445          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);
1446          while (*ecode == OP_ALT) ecode += GET(ecode, 1);          while (*ecode == OP_ALT) ecode += GET(ecode, 1);
1447          }          }
1448        else if (rrc != MATCH_NOMATCH &&  
1449                (rrc != MATCH_THEN || md->start_match_ptr != ecode))        /* PCRE doesn't allow the effect of (*THEN) to escape beyond an
1450          assertion; it is therefore treated as NOMATCH. */
1451    
1452          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
1453          {          {
1454          RRETURN(rrc);         /* Need braces because of following else */          RRETURN(rrc);         /* Need braces because of following else */
1455          }          }
# Line 1263  for (;;) Line 1460  for (;;)
1460          }          }
1461        }        }
1462    
1463      /* 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
1464      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
1465      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
1466      strategy no longer works because of the possibilty of (*THEN) being      call to match() is always required, unless the second alternative doesn't
1467      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
1468      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
1469      plough on. */      alternatives. If a THEN is encountered in the branch, it propagates out to
1470        the enclosing alternative (unless nested in a deeper set of alternatives,
1471        of course). */
1472    
1473      if (condition || *ecode == OP_ALT)      if (condition || *ecode == OP_ALT)
1474        {        {
1475        if (op == OP_SCOND) md->match_function_type = MATCH_CBEGROUP;        if (op != OP_SCOND)
1476            {
1477            ecode += 1 + LINK_SIZE;
1478            goto TAIL_RECURSE;
1479            }
1480    
1481          md->match_function_type = MATCH_CBEGROUP;
1482        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;  
1483        RRETURN(rrc);        RRETURN(rrc);
1484        }        }
1485      else                         /* Condition false & no alternative */  
1486         /* Condition false & no alternative; continue after the group. */
1487    
1488        else
1489        {        {
1490        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1491        }        }
# Line 1306  for (;;) Line 1512  for (;;)
1512        md->offset_vector[offset+1] = (int)(eptr - md->start_subject);        md->offset_vector[offset+1] = (int)(eptr - md->start_subject);
1513        if (offset_top <= offset) offset_top = offset + 2;        if (offset_top <= offset) offset_top = offset + 2;
1514        }        }
1515      ecode += 3;      ecode += 1 + IMM2_SIZE;
1516      break;      break;
1517    
1518    
# Line 1326  for (;;) Line 1532  for (;;)
1532           (md->notempty ||           (md->notempty ||
1533             (md->notempty_atstart &&             (md->notempty_atstart &&
1534               mstart == md->start_subject + md->start_offset)))               mstart == md->start_subject + md->start_offset)))
1535        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
1536    
1537      /* Otherwise, we have a match. */      /* Otherwise, we have a match. */
1538    
# Line 1335  for (;;) Line 1541  for (;;)
1541      md->start_match_ptr = mstart;       /* and the start (\K can modify) */      md->start_match_ptr = mstart;       /* and the start (\K can modify) */
1542    
1543      /* 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
1544      given as the argument to MRRETURN when the heap is in use. */      given as the argument to RRETURN when the heap is in use. */
1545    
1546      rrc = (op == OP_END)? MATCH_MATCH : MATCH_ACCEPT;      rrc = (op == OP_END)? MATCH_MATCH : MATCH_ACCEPT;
1547      MRRETURN(rrc);      RRETURN(rrc);
1548    
1549      /* Assertion brackets. Check the alternative branches in turn - the      /* Assertion brackets. Check the alternative branches in turn - the
1550      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 1559  for (;;)
1559    
1560      case OP_ASSERT:      case OP_ASSERT:
1561      case OP_ASSERTBACK:      case OP_ASSERTBACK:
1562        save_mark = md->mark;
1563      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1564        {        {
1565        condassert = TRUE;        condassert = TRUE;
# Line 1366  for (;;) Line 1573  for (;;)
1573        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1574          {          {
1575          mstart = md->start_match_ptr;   /* In case \K reset it */          mstart = md->start_match_ptr;   /* In case \K reset it */
         markptr = md->mark;  
1576          break;          break;
1577          }          }
1578        if (rrc != MATCH_NOMATCH &&  
1579            (rrc != MATCH_THEN || md->start_match_ptr != ecode))        /* PCRE does not allow THEN or COMMIT to escape beyond an assertion; it
1580          RRETURN(rrc);        is treated as NOMATCH. */
1581    
1582          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN &&
1583              rrc != MATCH_COMMIT) RRETURN(rrc);
1584    
1585        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1586          md->mark = save_mark;
1587        }        }
1588      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
1589    
1590      if (*ecode == OP_KET) MRRETURN(MATCH_NOMATCH);      if (*ecode == OP_KET) RRETURN(MATCH_NOMATCH);
1591    
1592      /* If checking an assertion for a condition, return MATCH_MATCH. */      /* If checking an assertion for a condition, return MATCH_MATCH. */
1593    
# Line 1396  for (;;) Line 1607  for (;;)
1607    
1608      case OP_ASSERT_NOT:      case OP_ASSERT_NOT:
1609      case OP_ASSERTBACK_NOT:      case OP_ASSERTBACK_NOT:
1610        save_mark = md->mark;
1611      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1612        {        {
1613        condassert = TRUE;        condassert = TRUE;
# Line 1406  for (;;) Line 1618  for (;;)
1618      do      do
1619        {        {
1620        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);
1621        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) MRRETURN(MATCH_NOMATCH);        md->mark = save_mark;
1622          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) RRETURN(MATCH_NOMATCH);
1623        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)
1624          {          {
1625          do ecode += GET(ecode,1); while (*ecode == OP_ALT);          do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1626          break;          break;
1627          }          }
1628        if (rrc != MATCH_NOMATCH &&  
1629            (rrc != MATCH_THEN || md->start_match_ptr != ecode))        /* PCRE does not allow THEN to escape beyond an assertion; it is treated
1630          RRETURN(rrc);        as NOMATCH. */
1631    
1632          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1633        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1634        }        }
1635      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 1430  for (;;) Line 1645  for (;;)
1645      back a number of characters, not bytes. */      back a number of characters, not bytes. */
1646    
1647      case OP_REVERSE:      case OP_REVERSE:
1648  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
1649      if (utf8)      if (utf)
1650        {        {
1651        i = GET(ecode, 1);        i = GET(ecode, 1);
1652        while (i-- > 0)        while (i-- > 0)
1653          {          {
1654          eptr--;          eptr--;
1655          if (eptr < md->start_subject) MRRETURN(MATCH_NOMATCH);          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1656          BACKCHAR(eptr);          BACKCHAR(eptr);
1657          }          }
1658        }        }
# Line 1448  for (;;) Line 1663  for (;;)
1663    
1664        {        {
1665        eptr -= GET(ecode, 1);        eptr -= GET(ecode, 1);
1666        if (eptr < md->start_subject) MRRETURN(MATCH_NOMATCH);        if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1667        }        }
1668    
1669      /* 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 1677  for (;;)
1677      function is able to force a failure. */      function is able to force a failure. */
1678    
1679      case OP_CALLOUT:      case OP_CALLOUT:
1680      if (pcre_callout != NULL)      if (PUBL(callout) != NULL)
1681        {        {
1682        pcre_callout_block cb;        PUBL(callout_block) cb;
1683        cb.version          = 2;   /* Version 1 of the callout block */        cb.version          = 2;   /* Version 1 of the callout block */
1684        cb.callout_number   = ecode[1];        cb.callout_number   = ecode[1];
1685        cb.offset_vector    = md->offset_vector;        cb.offset_vector    = md->offset_vector;
1686    #ifdef COMPILE_PCRE8
1687        cb.subject          = (PCRE_SPTR)md->start_subject;        cb.subject          = (PCRE_SPTR)md->start_subject;
1688    #else
1689          cb.subject          = (PCRE_SPTR16)md->start_subject;
1690    #endif
1691        cb.subject_length   = (int)(md->end_subject - md->start_subject);        cb.subject_length   = (int)(md->end_subject - md->start_subject);
1692        cb.start_match      = (int)(mstart - md->start_subject);        cb.start_match      = (int)(mstart - md->start_subject);
1693        cb.current_position = (int)(eptr - md->start_subject);        cb.current_position = (int)(eptr - md->start_subject);
# Line 1477  for (;;) Line 1696  for (;;)
1696        cb.capture_top      = offset_top/2;        cb.capture_top      = offset_top/2;
1697        cb.capture_last     = md->capture_last;        cb.capture_last     = md->capture_last;
1698        cb.callout_data     = md->callout_data;        cb.callout_data     = md->callout_data;
1699        cb.mark             = markptr;        cb.mark             = md->nomatch_mark;
1700        if ((rrc = (*pcre_callout)(&cb)) > 0) MRRETURN(MATCH_NOMATCH);        if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1701        if (rrc < 0) RRETURN(rrc);        if (rrc < 0) RRETURN(rrc);
1702        }        }
1703      ecode += 2 + 2*LINK_SIZE;      ecode += 2 + 2*LINK_SIZE;
# Line 1537  for (;;) Line 1756  for (;;)
1756        else        else
1757          {          {
1758          new_recursive.offset_save =          new_recursive.offset_save =
1759            (int *)(pcre_malloc)(new_recursive.saved_max * sizeof(int));            (int *)(PUBL(malloc))(new_recursive.saved_max * sizeof(int));
1760          if (new_recursive.offset_save == NULL) RRETURN(PCRE_ERROR_NOMEMORY);          if (new_recursive.offset_save == NULL) RRETURN(PCRE_ERROR_NOMEMORY);
1761          }          }
1762        memcpy(new_recursive.offset_save, md->offset_vector,        memcpy(new_recursive.offset_save, md->offset_vector,
# Line 1552  for (;;) Line 1771  for (;;)
1771        do        do
1772          {          {
1773          if (cbegroup) md->match_function_type = MATCH_CBEGROUP;          if (cbegroup) md->match_function_type = MATCH_CBEGROUP;
1774          RMATCH(eptr, callpat + _pcre_OP_lengths[*callpat], offset_top,          RMATCH(eptr, callpat + PRIV(OP_lengths)[*callpat], offset_top,
1775            md, eptrb, RM6);            md, eptrb, RM6);
1776          memcpy(md->offset_vector, new_recursive.offset_save,          memcpy(md->offset_vector, new_recursive.offset_save,
1777              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1778            md->recursive = new_recursive.prevrec;
1779          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1780            {            {
1781            DPRINTF(("Recursion matched\n"));            DPRINTF(("Recursion matched\n"));
           md->recursive = new_recursive.prevrec;  
1782            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
1783              (pcre_free)(new_recursive.offset_save);              (PUBL(free))(new_recursive.offset_save);
1784    
1785            /* 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
1786            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 1790  for (;;)
1790            mstart = md->start_match_ptr;            mstart = md->start_match_ptr;
1791            goto RECURSION_MATCHED;        /* Exit loop; end processing */            goto RECURSION_MATCHED;        /* Exit loop; end processing */
1792            }            }
1793          else if (rrc != MATCH_NOMATCH &&  
1794                  (rrc != MATCH_THEN || md->start_match_ptr != ecode))          /* PCRE does not allow THEN or COMMIT to escape beyond a recursion; it
1795            is treated as NOMATCH. */
1796    
1797            else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN &&
1798                     rrc != MATCH_COMMIT)
1799            {            {
1800            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1801            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
1802              (pcre_free)(new_recursive.offset_save);              (PUBL(free))(new_recursive.offset_save);
1803            RRETURN(rrc);            RRETURN(rrc);
1804            }            }
1805    
# Line 1588  for (;;) Line 1811  for (;;)
1811        DPRINTF(("Recursion didn't match\n"));        DPRINTF(("Recursion didn't match\n"));
1812        md->recursive = new_recursive.prevrec;        md->recursive = new_recursive.prevrec;
1813        if (new_recursive.offset_save != stacksave)        if (new_recursive.offset_save != stacksave)
1814          (pcre_free)(new_recursive.offset_save);          (PUBL(free))(new_recursive.offset_save);
1815        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
1816        }        }
1817    
1818      RECURSION_MATCHED:      RECURSION_MATCHED:
# Line 1658  for (;;) Line 1881  for (;;)
1881        }        }
1882      else saved_eptr = NULL;      else saved_eptr = NULL;
1883    
1884      /* 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
1885      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
1886      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
1887      by \K. */      start in case it was changed by \K. */
1888    
1889      if (*prev == OP_ASSERT || *prev == OP_ASSERT_NOT ||      if ((*prev >= OP_ASSERT && *prev <= OP_ASSERTBACK_NOT) ||
1890          *prev == OP_ASSERTBACK || *prev == OP_ASSERTBACK_NOT)           *prev == OP_ONCE_NC)
1891        {        {
1892        md->end_match_ptr = eptr;      /* For ONCE */        md->end_match_ptr = eptr;      /* For ONCE_NC */
1893        md->end_offset_top = offset_top;        md->end_offset_top = offset_top;
1894        md->start_match_ptr = mstart;        md->start_match_ptr = mstart;
1895        MRRETURN(MATCH_MATCH);         /* Sets md->mark */        RRETURN(MATCH_MATCH);         /* Sets md->mark */
1896        }        }
1897    
1898      /* 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 1957  for (;;)
1957      /* For an ordinary non-repeating ket, just continue at this level. This      /* For an ordinary non-repeating ket, just continue at this level. This
1958      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
1959      group. This is the forcible breaking of infinite loops as implemented in      group. This is the forcible breaking of infinite loops as implemented in
1960      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,
1961      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
1962      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
1963      bypassing intermediate backup points, but resetting any captures that      original OP_ONCE level, thereby bypassing intermediate backup points, but
1964      happened along the way. */      resetting any captures that happened along the way. */
1965    
1966      if (*ecode == OP_KET || eptr == saved_eptr)      if (*ecode == OP_KET || eptr == saved_eptr)
1967        {        {
# 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++;
# 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 0x000d:
2419        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr >= md->end_subject)
2420            {
2421            SCHECK_PARTIAL();
2422            }
2423          else if (*eptr == 0x0a) eptr++;
2424        break;        break;
2425    
2426        case 0x000a:        case 0x000a:
# Line 2162  for (;;) Line 2431  for (;;)
2431        case 0x0085:        case 0x0085:
2432        case 0x2028:        case 0x2028:
2433        case 0x2029:        case 0x2029:
2434        if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);        if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
2435        break;        break;
2436        }        }
2437      ecode++;      ecode++;
# Line 2172  for (;;) Line 2441  for (;;)
2441      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2442        {        {
2443        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2444        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2445        }        }
2446      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2447      switch(c)      switch(c)
# Line 2197  for (;;) Line 2466  for (;;)
2466        case 0x202f:    /* NARROW NO-BREAK SPACE */        case 0x202f:    /* NARROW NO-BREAK SPACE */
2467        case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */        case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
2468        case 0x3000:    /* IDEOGRAPHIC SPACE */        case 0x3000:    /* IDEOGRAPHIC SPACE */
2469        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2470        }        }
2471      ecode++;      ecode++;
2472      break;      break;
# Line 2206  for (;;) Line 2475  for (;;)
2475      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2476        {        {
2477        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2478        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2479        }        }
2480      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2481      switch(c)      switch(c)
2482        {        {
2483        default: MRRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2484        case 0x09:      /* HT */        case 0x09:      /* HT */
2485        case 0x20:      /* SPACE */        case 0x20:      /* SPACE */
2486        case 0xa0:      /* NBSP */        case 0xa0:      /* NBSP */
# Line 2240  for (;;) Line 2509  for (;;)
2509      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2510        {        {
2511        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2512        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2513        }        }
2514      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2515      switch(c)      switch(c)
# Line 2253  for (;;) Line 2522  for (;;)
2522        case 0x85:      /* NEL */        case 0x85:      /* NEL */
2523        case 0x2028:    /* LINE SEPARATOR */        case 0x2028:    /* LINE SEPARATOR */
2524        case 0x2029:    /* PARAGRAPH SEPARATOR */        case 0x2029:    /* PARAGRAPH SEPARATOR */
2525        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2526        }        }
2527      ecode++;      ecode++;
2528      break;      break;
# Line 2262  for (;;) Line 2531  for (;;)
2531      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2532        {        {
2533        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2534        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2535        }        }
2536      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2537      switch(c)      switch(c)
2538        {        {
2539        default: MRRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2540        case 0x0a:      /* LF */        case 0x0a:      /* LF */
2541        case 0x0b:      /* VT */        case 0x0b:      /* VT */
2542        case 0x0c:      /* FF */        case 0x0c:      /* FF */
# Line 2289  for (;;) Line 2558  for (;;)
2558      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2559        {        {
2560        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2561        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2562        }        }
2563      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2564        {        {
# Line 2298  for (;;) Line 2567  for (;;)
2567        switch(ecode[1])        switch(ecode[1])
2568          {          {
2569          case PT_ANY:          case PT_ANY:
2570          if (op == OP_NOTPROP) MRRETURN(MATCH_NOMATCH);          if (op == OP_NOTPROP) RRETURN(MATCH_NOMATCH);
2571          break;          break;
2572    
2573          case PT_LAMP:          case PT_LAMP:
2574          if ((prop->chartype == ucp_Lu ||          if ((prop->chartype == ucp_Lu ||
2575               prop->chartype == ucp_Ll ||               prop->chartype == ucp_Ll ||
2576               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))
2577            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2578          break;          break;
2579    
2580          case PT_GC:          case PT_GC:
2581          if ((ecode[2] != _pcre_ucp_gentype[prop->chartype]) == (op == OP_PROP))          if ((ecode[2] != PRIV(ucp_gentype)[prop->chartype]) == (op == OP_PROP))
2582            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2583          break;          break;
2584    
2585          case PT_PC:          case PT_PC:
2586          if ((ecode[2] != prop->chartype) == (op == OP_PROP))          if ((ecode[2] != prop->chartype) == (op == OP_PROP))
2587            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2588          break;          break;
2589    
2590          case PT_SC:          case PT_SC:
2591          if ((ecode[2] != prop->script) == (op == OP_PROP))          if ((ecode[2] != prop->script) == (op == OP_PROP))
2592            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2593          break;          break;
2594    
2595          /* These are specials */          /* These are specials */
2596    
2597          case PT_ALNUM:          case PT_ALNUM:
2598          if ((_pcre_ucp_gentype[prop->chartype] == ucp_L ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2599               _pcre_ucp_gentype[prop->chartype] == ucp_N) == (op == OP_NOTPROP))               PRIV(ucp_gentype)[prop->chartype] == ucp_N) == (op == OP_NOTPROP))
2600            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2601          break;          break;
2602    
2603          case PT_SPACE:    /* Perl space */          case PT_SPACE:    /* Perl space */
2604          if ((_pcre_ucp_gentype[prop->chartype] == ucp_Z ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2605               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)
2606                 == (op == OP_NOTPROP))                 == (op == OP_NOTPROP))
2607            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2608          break;          break;
2609    
2610          case PT_PXSPACE:  /* POSIX space */          case PT_PXSPACE:  /* POSIX space */
2611          if ((_pcre_ucp_gentype[prop->chartype] == ucp_Z ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2612               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
2613               c == CHAR_FF || c == CHAR_CR)               c == CHAR_FF || c == CHAR_CR)
2614                 == (op == OP_NOTPROP))                 == (op == OP_NOTPROP))
2615            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2616          break;          break;
2617    
2618          case PT_WORD:          case PT_WORD:
2619          if ((_pcre_ucp_gentype[prop->chartype] == ucp_L ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2620               _pcre_ucp_gentype[prop->chartype] == ucp_N ||               PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
2621               c == CHAR_UNDERSCORE) == (op == OP_NOTPROP))               c == CHAR_UNDERSCORE) == (op == OP_NOTPROP))
2622            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2623          break;          break;
2624    
2625          /* This should never occur */          /* This should never occur */
# Line 2370  for (;;) Line 2639  for (;;)
2639      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2640        {        {
2641        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2642        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2643        }        }
2644      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2645      if (UCD_CATEGORY(c) == ucp_M) MRRETURN(MATCH_NOMATCH);      if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);
2646      while (eptr < md->end_subject)      while (eptr < md->end_subject)
2647        {        {
2648        int len = 1;        int len = 1;
2649        if (!utf8) c = *eptr; else { GETCHARLEN(c, eptr, len); }        if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
2650        if (UCD_CATEGORY(c) != ucp_M) break;        if (UCD_CATEGORY(c) != ucp_M) break;
2651        eptr += len;        eptr += len;
2652        }        }
2653        CHECK_PARTIAL();
2654      ecode++;      ecode++;
2655      break;      break;
2656  #endif  #endif
# Line 2398  for (;;) Line 2668  for (;;)
2668      case OP_REFI:      case OP_REFI:
2669      caseless = op == OP_REFI;      caseless = op == OP_REFI;
2670      offset = GET2(ecode, 1) << 1;               /* Doubled ref number */      offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
2671      ecode += 3;      ecode += 1 + IMM2_SIZE;
2672    
2673      /* If the reference is unset, there are two possibilities:      /* If the reference is unset, there are two possibilities:
2674    
# Line 2438  for (;;) Line 2708  for (;;)
2708        case OP_CRMINRANGE:        case OP_CRMINRANGE:
2709        minimize = (*ecode == OP_CRMINRANGE);        minimize = (*ecode == OP_CRMINRANGE);
2710        min = GET2(ecode, 1);        min = GET2(ecode, 1);
2711        max = GET2(ecode, 3);        max = GET2(ecode, 1 + IMM2_SIZE);
2712        if (max == 0) max = INT_MAX;        if (max == 0) max = INT_MAX;
2713        ecode += 5;        ecode += 1 + 2 * IMM2_SIZE;
2714        break;        break;
2715    
2716        default:               /* No repeat follows */        default:               /* No repeat follows */
2717        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)
2718          {          {
2719            if (length == -2) eptr = md->end_subject;   /* Partial match */
2720          CHECK_PARTIAL();          CHECK_PARTIAL();
2721          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2722          }          }
2723        eptr += length;        eptr += length;
2724        continue;              /* With the main loop */        continue;              /* With the main loop */
2725        }        }
2726    
2727      /* Handle repeated back references. If the length of the reference is      /* Handle repeated back references. If the length of the reference is
2728      zero, just continue with the main loop. */      zero, just continue with the main loop. If the length is negative, it
2729        means the reference is unset in non-Java-compatible mode. If the minimum is
2730        zero, we can continue at the same level without recursion. For any other
2731        minimum, carrying on will result in NOMATCH. */
2732    
2733      if (length == 0) continue;      if (length == 0) continue;
2734        if (length < 0 && min == 0) continue;
2735    
2736      /* First, ensure the minimum number of matches are present. We get back      /* First, ensure the minimum number of matches are present. We get back
2737      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 2742  for (;;)
2742        int slength;        int slength;
2743        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2744          {          {
2745            if (slength == -2) eptr = md->end_subject;   /* Partial match */
2746          CHECK_PARTIAL();          CHECK_PARTIAL();
2747          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2748          }          }
2749        eptr += slength;        eptr += slength;
2750        }        }
# Line 2487  for (;;) Line 2763  for (;;)
2763          int slength;          int slength;
2764          RMATCH(eptr, ecode, offset_top, md, eptrb, RM14);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM14);
2765          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2766          if (fi >= max) MRRETURN(MATCH_NOMATCH);          if (fi >= max) RRETURN(MATCH_NOMATCH);
2767          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2768            {            {
2769              if (slength == -2) eptr = md->end_subject;   /* Partial match */
2770            CHECK_PARTIAL();            CHECK_PARTIAL();
2771            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2772            }            }
2773          eptr += slength;          eptr += slength;
2774          }          }
# Line 2508  for (;;) Line 2785  for (;;)
2785          int slength;          int slength;
2786          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2787            {            {
2788            CHECK_PARTIAL();            /* Can't use CHECK_PARTIAL because we don't want to update eptr in
2789              the soft partial matching case. */
2790    
2791              if (slength == -2 && md->partial != 0 &&
2792                  md->end_subject > md->start_used_ptr)
2793                {
2794                md->hitend = TRUE;
2795                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2796                }
2797            break;            break;
2798            }            }
2799          eptr += slength;          eptr += slength;
2800          }          }
2801    
2802        while (eptr >= pp)        while (eptr >= pp)
2803          {          {
2804          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);
2805          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2806          eptr -= length;          eptr -= length;
2807          }          }
2808        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2809        }        }
2810      /* Control never gets here */      /* Control never gets here */
2811    
# Line 2537  for (;;) Line 2823  for (;;)
2823      case OP_NCLASS:      case OP_NCLASS:
2824      case OP_CLASS:      case OP_CLASS:
2825        {        {
2826          /* The data variable is saved across frames, so the byte map needs to
2827          be stored there. */
2828    #define BYTE_MAP ((pcre_uint8 *)data)
2829        data = ecode + 1;                /* Save for matching */        data = ecode + 1;                /* Save for matching */
2830        ecode += 33;                     /* Advance past the item */        ecode += 1 + (32 / sizeof(pcre_uchar)); /* Advance past the item */
2831    
2832        switch (*ecode)        switch (*ecode)
2833          {          {
# Line 2559  for (;;) Line 2848  for (;;)
2848          case OP_CRMINRANGE:          case OP_CRMINRANGE:
2849          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
2850          min = GET2(ecode, 1);          min = GET2(ecode, 1);
2851          max = GET2(ecode, 3);          max = GET2(ecode, 1 + IMM2_SIZE);
2852          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
2853          ecode += 5;          ecode += 1 + 2 * IMM2_SIZE;
2854          break;          break;
2855    
2856          default:               /* No repeat follows */          default:               /* No repeat follows */
# Line 2571  for (;;) Line 2860  for (;;)
2860    
2861        /* First, ensure the minimum number of matches are present. */        /* First, ensure the minimum number of matches are present. */
2862    
2863  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2864        /* UTF-8 mode */        if (utf)
       if (utf8)  
2865          {          {
2866          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2867            {            {
2868            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
2869              {              {
2870              SCHECK_PARTIAL();              SCHECK_PARTIAL();
2871              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2872              }              }
2873            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
2874            if (c > 255)            if (c > 255)
2875              {              {
2876              if (op == OP_CLASS) MRRETURN(MATCH_NOMATCH);              if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2877              }              }
2878            else            else
2879              {              if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
             if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  
             }  
2880            }            }
2881          }          }
2882        else        else
2883  #endif  #endif
2884        /* Not UTF-8 mode */        /* Not UTF mode */
2885          {          {
2886          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2887            {            {
2888            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
2889              {              {
2890              SCHECK_PARTIAL();              SCHECK_PARTIAL();
2891              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2892              }              }
2893            c = *eptr++;            c = *eptr++;
2894            if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  #ifndef COMPILE_PCRE8
2895              if (c > 255)
2896                {
2897                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2898                }
2899              else
2900    #endif
2901                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2902            }            }
2903          }          }
2904    
# Line 2619  for (;;) Line 2912  for (;;)
2912    
2913        if (minimize)        if (minimize)
2914          {          {
2915  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2916          /* UTF-8 mode */          if (utf)
         if (utf8)  
2917            {            {
2918            for (fi = min;; fi++)            for (fi = min;; fi++)
2919              {              {
2920              RMATCH(eptr, ecode, offset_top, md, eptrb, RM16);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM16);
2921              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2922              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2923              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
2924                {                {
2925                SCHECK_PARTIAL();                SCHECK_PARTIAL();
2926                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2927                }                }
2928              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
2929              if (c > 255)              if (c > 255)
2930                {                {
2931                if (op == OP_CLASS) MRRETURN(MATCH_NOMATCH);                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2932                }                }
2933              else              else
2934                {                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
               if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  
               }  
2935              }              }
2936            }            }
2937          else          else
2938  #endif  #endif
2939          /* Not UTF-8 mode */          /* Not UTF mode */
2940            {            {
2941            for (fi = min;; fi++)            for (fi = min;; fi++)
2942              {              {
2943              RMATCH(eptr, ecode, offset_top, md, eptrb, RM17);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM17);
2944              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2945              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2946              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
2947                {                {
2948                SCHECK_PARTIAL();                SCHECK_PARTIAL();
2949                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2950                }                }
2951              c = *eptr++;              c = *eptr++;
2952              if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  #ifndef COMPILE_PCRE8
2953                if (c > 255)
2954                  {
2955                  if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2956                  }
2957                else
2958    #endif
2959                  if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2960              }              }
2961            }            }
2962          /* Control never gets here */          /* Control never gets here */
# Line 2671  for (;;) Line 2968  for (;;)
2968          {          {
2969          pp = eptr;          pp = eptr;
2970    
2971  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2972          /* UTF-8 mode */          if (utf)
         if (utf8)  
2973            {            {
2974            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2975              {              {
# Line 2689  for (;;) Line 2985  for (;;)
2985                if (op == OP_CLASS) break;                if (op == OP_CLASS) break;
2986                }                }
2987              else              else
2988                {                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
               if ((data[c/8] & (1 << (c&7))) == 0) break;  
               }  
2989              eptr += len;              eptr += len;
2990              }              }
2991            for (;;)            for (;;)
# Line 2704  for (;;) Line 2998  for (;;)
2998            }            }
2999          else          else
3000  #endif  #endif
3001            /* Not UTF-8 mode */            /* Not UTF mode */
3002            {            {
3003            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3004              {              {
# Line 2714  for (;;) Line 3008  for (;;)
3008                break;                break;
3009                }                }
3010              c = *eptr;              c = *eptr;
3011              if ((data[c/8] & (1 << (c&7))) == 0) break;  #ifndef COMPILE_PCRE8
3012                if (c > 255)
3013                  {
3014                  if (op == OP_CLASS) break;
3015                  }
3016                else
3017    #endif
3018                  if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
3019              eptr++;              eptr++;
3020              }              }
3021            while (eptr >= pp)            while (eptr >= pp)
# Line 2725  for (;;) Line 3026  for (;;)
3026              }              }
3027            }            }
3028    
3029          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3030          }          }
3031    #undef BYTE_MAP
3032        }        }
3033      /* Control never gets here */      /* Control never gets here */
3034    
# Line 2735  for (;;) Line 3037  for (;;)
3037      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
3038      mode, because Unicode properties are supported in non-UTF-8 mode. */      mode, because Unicode properties are supported in non-UTF-8 mode. */
3039    
3040  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
3041      case OP_XCLASS:      case OP_XCLASS:
3042        {        {
3043        data = ecode + 1 + LINK_SIZE;                /* Save for matching */        data = ecode + 1 + LINK_SIZE;                /* Save for matching */
# Line 2760  for (;;) Line 3062  for (;;)
3062          case OP_CRMINRANGE:          case OP_CRMINRANGE:
3063          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
3064          min = GET2(ecode, 1);          min = GET2(ecode, 1);
3065          max = GET2(ecode, 3);          max = GET2(ecode, 1 + IMM2_SIZE);
3066          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
3067          ecode += 5;          ecode += 1 + 2 * IMM2_SIZE;
3068          break;          break;
3069    
3070          default:               /* No repeat follows */          default:               /* No repeat follows */
# Line 2777  for (;;) Line 3079  for (;;)
3079          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3080            {            {
3081            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3082            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3083            }            }
3084          GETCHARINCTEST(c, eptr);          GETCHARINCTEST(c, eptr);
3085          if (!_pcre_xclass(c, data)) MRRETURN(MATCH_NOMATCH);          if (!PRIV(xclass)(c, data, utf)) RRETURN(MATCH_NOMATCH);
3086          }          }
3087    
3088        /* 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 3099  for (;;)
3099            {            {
3100            RMATCH(eptr, ecode, offset_top, md, eptrb, RM20);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM20);
3101            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3102            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3103            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3104              {              {
3105              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3106              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3107              }              }
3108            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3109            if (!_pcre_xclass(c, data)) MRRETURN(MATCH_NOMATCH);            if (!PRIV(xclass)(c, data, utf)) RRETURN(MATCH_NOMATCH);
3110            }            }
3111          /* Control never gets here */          /* Control never gets here */
3112          }          }
# Line 2822  for (;;) Line 3124  for (;;)
3124              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3125              break;              break;
3126              }              }
3127    #ifdef SUPPORT_UTF
3128            GETCHARLENTEST(c, eptr, len);            GETCHARLENTEST(c, eptr, len);
3129            if (!_pcre_xclass(c, data)) break;  #else
3130              c = *eptr;
3131    #endif
3132              if (!PRIV(xclass)(c, data, utf)) break;
3133            eptr += len;            eptr += len;
3134            }            }
3135          for(;;)          for(;;)
# Line 2831  for (;;) Line 3137  for (;;)
3137            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);
3138            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3139            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3140            if (utf8) BACKCHAR(eptr);  #ifdef SUPPORT_UTF
3141              if (utf) BACKCHAR(eptr);
3142    #endif
3143            }            }
3144          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3145          }          }
3146    
3147        /* Control never gets here */        /* Control never gets here */
# Line 2843  for (;;) Line 3151  for (;;)
3151      /* Match a single character, casefully */      /* Match a single character, casefully */
3152    
3153      case OP_CHAR:      case OP_CHAR:
3154  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3155      if (utf8)      if (utf)
3156        {        {
3157        length = 1;        length = 1;
3158        ecode++;        ecode++;
# Line 2852  for (;;) Line 3160  for (;;)
3160        if (length > md->end_subject - eptr)        if (length > md->end_subject - eptr)
3161          {          {
3162          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
3163          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3164          }          }
3165        while (length-- > 0) if (*ecode++ != *eptr++) MRRETURN(MATCH_NOMATCH);        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);
3166        }        }
3167      else      else
3168  #endif  #endif
3169        /* Not UTF mode */
     /* Non-UTF-8 mode */  
3170        {        {
3171        if (md->end_subject - eptr < 1)        if (md->end_subject - eptr < 1)
3172          {          {
3173          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */
3174          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3175          }          }
3176        if (ecode[1] != *eptr++) MRRETURN(MATCH_NOMATCH);        if (ecode[1] != *eptr++) RRETURN(MATCH_NOMATCH);
3177        ecode += 2;        ecode += 2;
3178        }        }
3179      break;      break;
3180    
3181      /* Match a single character, caselessly */      /* Match a single character, caselessly. If we are at the end of the
3182        subject, give up immediately. */
3183    
3184      case OP_CHARI:      case OP_CHARI:
3185  #ifdef SUPPORT_UTF8      if (eptr >= md->end_subject)
3186      if (utf8)        {
3187          SCHECK_PARTIAL();
3188          RRETURN(MATCH_NOMATCH);
3189          }
3190    
3191    #ifdef SUPPORT_UTF
3192        if (utf)
3193        {        {
3194        length = 1;        length = 1;
3195        ecode++;        ecode++;
3196        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
3197    
       if (length > md->end_subject - eptr)  
         {  
         CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */  
         MRRETURN(MATCH_NOMATCH);  
         }  
   
3198        /* 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
3199        can use the fast lookup table. */        we know that its other case must also be one byte long, so we can use the
3200          fast lookup table. We know that there is at least one byte left in the
3201          subject. */
3202    
3203        if (fc < 128)        if (fc < 128)
3204          {          {
3205          if (md->lcc[*ecode++] != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);          if (md->lcc[fc]
3206                != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);
3207            ecode++;
3208            eptr++;
3209          }          }
3210    
3211        /* Otherwise we must pick up the subject character */        /* Otherwise we must pick up the subject character. Note that we cannot
3212          use the value of "length" to check for sufficient bytes left, because the
3213          other case of the character may have more or fewer bytes.  */
3214    
3215        else        else
3216          {          {
# Line 2911  for (;;) Line 3226  for (;;)
3226  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3227            if (dc != UCD_OTHERCASE(fc))            if (dc != UCD_OTHERCASE(fc))
3228  #endif  #endif
3229              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3230            }            }
3231          }          }
3232        }        }
3233      else      else
3234  #endif   /* SUPPORT_UTF8 */  #endif   /* SUPPORT_UTF */
3235    
3236      /* Non-UTF-8 mode */      /* Not UTF mode */
3237        {        {
3238        if (md->end_subject - eptr < 1)        if (TABLE_GET(ecode[1], md->lcc, ecode[1])
3239          {            != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);
3240          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */        eptr++;
         MRRETURN(MATCH_NOMATCH);  
         }  
       if (md->lcc[ecode[1]] != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);  
3241        ecode += 2;        ecode += 2;
3242        }        }
3243      break;      break;
# Line 2935  for (;;) Line 3247  for (;;)
3247      case OP_EXACT:      case OP_EXACT:
3248      case OP_EXACTI:      case OP_EXACTI:
3249      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3250      ecode += 3;      ecode += 1 + IMM2_SIZE;
3251      goto REPEATCHAR;      goto REPEATCHAR;
3252    
3253      case OP_POSUPTO:      case OP_POSUPTO:
# Line 2950  for (;;) Line 3262  for (;;)
3262      min = 0;      min = 0;
3263      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3264      minimize = *ecode == OP_MINUPTO || *ecode == OP_MINUPTOI;      minimize = *ecode == OP_MINUPTO || *ecode == OP_MINUPTOI;
3265      ecode += 3;      ecode += 1 + IMM2_SIZE;
3266      goto REPEATCHAR;      goto REPEATCHAR;
3267    
3268      case OP_POSSTAR:      case OP_POSSTAR:
# Line 2998  for (;;) Line 3310  for (;;)
3310      /* Common code for all repeated single-character matches. */      /* Common code for all repeated single-character matches. */
3311    
3312      REPEATCHAR:      REPEATCHAR:
3313  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3314      if (utf8)      if (utf)
3315        {        {
3316        length = 1;        length = 1;
3317        charptr = ecode;        charptr = ecode;
# Line 3015  for (;;) Line 3327  for (;;)
3327          unsigned int othercase;          unsigned int othercase;
3328          if (op >= OP_STARI &&     /* Caseless */          if (op >= OP_STARI &&     /* Caseless */
3329              (othercase = UCD_OTHERCASE(fc)) != fc)              (othercase = UCD_OTHERCASE(fc)) != fc)
3330            oclength = _pcre_ord2utf8(othercase, occhars);            oclength = PRIV(ord2utf)(othercase, occhars);
3331          else oclength = 0;          else oclength = 0;
3332  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3333    
3334          for (i = 1; i <= min; i++)          for (i = 1; i <= min; 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              {              {
3345              CHECK_PARTIAL();              CHECK_PARTIAL();
3346              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3347              }              }
3348            }            }
3349    
# Line 3043  for (;;) Line 3355  for (;;)
3355              {              {
3356              RMATCH(eptr, ecode, offset_top, md, eptrb, RM22);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM22);
3357              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3358              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3359              if (eptr <= md->end_subject - length &&              if (eptr <= md->end_subject - length &&
3360                memcmp(eptr, charptr, length) == 0) eptr += length;                memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3361  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3362              else if (oclength > 0 &&              else if (oclength > 0 &&
3363                       eptr <= md->end_subject - oclength &&                       eptr <= md->end_subject - oclength &&
3364                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                       memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3365  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3366              else              else
3367                {                {
3368                CHECK_PARTIAL();                CHECK_PARTIAL();
3369                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3370                }                }
3371              }              }
3372            /* Control never gets here */            /* Control never gets here */
# Line 3066  for (;;) Line 3378  for (;;)
3378            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3379              {              {
3380              if (eptr <= md->end_subject - length &&              if (eptr <= md->end_subject - length &&
3381                  memcmp(eptr, charptr, length) == 0) eptr += length;                  memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3382  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3383              else if (oclength > 0 &&              else if (oclength > 0 &&
3384                       eptr <= md->end_subject - oclength &&                       eptr <= md->end_subject - oclength &&
3385                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                       memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3386  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3387              else              else
3388                {                {
# Line 3085  for (;;) Line 3397  for (;;)
3397              {              {
3398              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);
3399              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3400              if (eptr == pp) { MRRETURN(MATCH_NOMATCH); }              if (eptr == pp) { RRETURN(MATCH_NOMATCH); }
3401  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3402              eptr--;              eptr--;
3403              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 3102  for (;;) Line 3414  for (;;)
3414        value of fc will always be < 128. */        value of fc will always be < 128. */
3415        }        }
3416      else      else
3417  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
3418          /* When not in UTF-8 mode, load a single-byte character. */
3419      /* When not in UTF-8 mode, load a single-byte character. */        fc = *ecode++;
3420    
3421      fc = *ecode++;      /* The value of fc at this point is always one character, though we may
3422        or may not be in UTF mode. The code is duplicated for the caseless and
     /* 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  
3423      caseful cases, for speed, since matching characters is likely to be quite      caseful cases, for speed, since matching characters is likely to be quite
3424      common. First, ensure the minimum number of matches are present. If min =      common. First, ensure the minimum number of matches are present. If min =
3425      max, continue at the same level without recursing. Otherwise, if      max, continue at the same level without recursing. Otherwise, if
# Line 3122  for (;;) Line 3432  for (;;)
3432    
3433      if (op >= OP_STARI)  /* Caseless */      if (op >= OP_STARI)  /* Caseless */
3434        {        {
3435        fc = md->lcc[fc];  #ifdef COMPILE_PCRE8
3436          /* fc must be < 128 if UTF is enabled. */
3437          foc = md->fcc[fc];
3438    #else
3439    #ifdef SUPPORT_UTF
3440    #ifdef SUPPORT_UCP
3441          if (utf && fc > 127)
3442            foc = UCD_OTHERCASE(fc);
3443    #else
3444          if (utf && fc > 127)
3445            foc = fc;
3446    #endif /* SUPPORT_UCP */
3447          else
3448    #endif /* SUPPORT_UTF */
3449            foc = TABLE_GET(fc, md->fcc, fc);
3450    #endif /* COMPILE_PCRE8 */
3451    
3452        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
3453          {          {
3454          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3455            {            {
3456            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3457            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3458            }            }
3459          if (fc != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);          if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);
3460            eptr++;
3461          }          }
3462        if (min == max) continue;        if (min == max) continue;
3463        if (minimize)        if (minimize)
# Line 3139  for (;;) Line 3466  for (;;)
3466            {            {
3467            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);
3468            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3469            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3470            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3471              {              {
3472              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3473              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3474              }              }
3475            if (fc != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);            if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);
3476              eptr++;
3477            }            }
3478          /* Control never gets here */          /* Control never gets here */
3479          }          }
# Line 3159  for (;;) Line 3487  for (;;)
3487              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3488              break;              break;
3489              }              }
3490            if (fc != md->lcc[*eptr]) break;            if (fc != *eptr && foc != *eptr) break;
3491            eptr++;            eptr++;
3492            }            }
3493    
# Line 3171  for (;;) Line 3499  for (;;)
3499            eptr--;            eptr--;
3500            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3501            }            }
3502          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3503          }          }
3504        /* Control never gets here */        /* Control never gets here */
3505        }        }
# Line 3185  for (;;) Line 3513  for (;;)
3513          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3514            {            {
3515            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3516            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3517            }            }
3518          if (fc != *eptr++) MRRETURN(MATCH_NOMATCH);          if (fc != *eptr++) RRETURN(MATCH_NOMATCH);
3519          }          }
3520    
3521        if (min == max) continue;        if (min == max) continue;
# Line 3198  for (;;) Line 3526  for (;;)
3526            {            {
3527            RMATCH(eptr, ecode, offset_top, md, eptrb, RM26);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM26);
3528            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3529            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3530            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3531              {              {
3532              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3533              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3534              }              }
3535            if (fc != *eptr++) MRRETURN(MATCH_NOMATCH);            if (fc != *eptr++) RRETURN(MATCH_NOMATCH);
3536            }            }
3537          /* Control never gets here */          /* Control never gets here */
3538          }          }
# Line 3229  for (;;) Line 3557  for (;;)
3557            eptr--;            eptr--;
3558            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3559            }            }
3560          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3561          }          }
3562        }        }
3563      /* Control never gets here */      /* Control never gets here */
# Line 3242  for (;;) Line 3570  for (;;)
3570      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
3571        {        {
3572        SCHECK_PARTIAL();        SCHECK_PARTIAL();
3573        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
3574        }        }
3575      ecode++;  #ifdef SUPPORT_UTF
3576      GETCHARINCTEST(c, eptr);      if (utf)
     if (op == OP_NOTI)         /* The caseless case */  
3577        {        {
3578  #ifdef SUPPORT_UTF8        register unsigned int ch, och;
3579        if (c < 256)  
3580  #endif        ecode++;
3581        c = md->lcc[c];        GETCHARINC(ch, ecode);
3582        if (md->lcc[*ecode++] == c) MRRETURN(MATCH_NOMATCH);        GETCHARINC(c, eptr);
3583    
3584          if (op == OP_NOT)
3585            {
3586            if (ch == c) RRETURN(MATCH_NOMATCH);
3587            }
3588          else
3589            {
3590    #ifdef SUPPORT_UCP
3591            if (ch > 127)
3592              och = UCD_OTHERCASE(ch);
3593    #else
3594            if (ch > 127)
3595              och = ch;
3596    #endif /* SUPPORT_UCP */
3597            else
3598              och = TABLE_GET(ch, md->fcc, ch);
3599            if (ch == c || och == c) RRETURN(MATCH_NOMATCH);
3600            }
3601        }        }
3602      else    /* Caseful */      else
3603    #endif
3604        {        {
3605        if (*ecode++ == c) MRRETURN(MATCH_NOMATCH);        register unsigned int ch = ecode[1];
3606          c = *eptr++;
3607          if (ch == c || (op == OP_NOTI && TABLE_GET(ch, md->fcc, ch) == c))
3608            RRETURN(MATCH_NOMATCH);
3609          ecode += 2;
3610        }        }
3611      break;      break;
3612    
# Line 3270  for (;;) Line 3620  for (;;)
3620      case OP_NOTEXACT:      case OP_NOTEXACT:
3621      case OP_NOTEXACTI:      case OP_NOTEXACTI:
3622      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3623      ecode += 3;      ecode += 1 + IMM2_SIZE;
3624      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3625    
3626      case OP_NOTUPTO:      case OP_NOTUPTO:
# Line 3280  for (;;) Line 3630  for (;;)
3630      min = 0;      min = 0;
3631      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3632      minimize = *ecode == OP_NOTMINUPTO || *ecode == OP_NOTMINUPTOI;      minimize = *ecode == OP_NOTMINUPTO || *ecode == OP_NOTMINUPTOI;
3633      ecode += 3;      ecode += 1 + IMM2_SIZE;
3634      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3635    
3636      case OP_NOTPOSSTAR:      case OP_NOTPOSSTAR:
# Line 3312  for (;;) Line 3662  for (;;)
3662      possessive = TRUE;      possessive = TRUE;
3663      min = 0;      min = 0;
3664      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3665      ecode += 3;      ecode += 1 + IMM2_SIZE;
3666      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3667    
3668      case OP_NOTSTAR:      case OP_NOTSTAR:
# Line 3336  for (;;) Line 3686  for (;;)
3686      /* Common code for all repeated single-byte matches. */      /* Common code for all repeated single-byte matches. */
3687    
3688      REPEATNOTCHAR:      REPEATNOTCHAR:
3689      fc = *ecode++;      GETCHARINCTEST(fc, ecode);
3690    
3691      /* The code is duplicated for the caseless and caseful cases, for speed,      /* The code is duplicated for the caseless and caseful cases, for speed,
3692      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 3351  for (;;) Line 3701  for (;;)
3701    
3702      if (op >= OP_NOTSTARI)     /* Caseless */      if (op >= OP_NOTSTARI)     /* Caseless */
3703        {        {
3704        fc = md->lcc[fc];  #ifdef SUPPORT_UTF
3705    #ifdef SUPPORT_UCP
3706          if (utf && fc > 127)
3707            foc = UCD_OTHERCASE(fc);
3708    #else
3709          if (utf && fc > 127)
3710            foc = fc;
3711    #endif /* SUPPORT_UCP */
3712          else
3713    #endif /* SUPPORT_UTF */
3714            foc = TABLE_GET(fc, md->fcc, fc);
3715    
3716  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3717        /* UTF-8 mode */        if (utf)
       if (utf8)  
3718          {          {
3719          register unsigned int d;          register unsigned int d;
3720          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3363  for (;;) Line 3722  for (;;)
3722            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3723              {              {
3724              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3725              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3726              }              }
3727            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3728            if (d < 256) d = md->lcc[d];            if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
           if (fc == d) MRRETURN(MATCH_NOMATCH);  
3729            }            }
3730          }          }
3731        else        else
3732  #endif  #endif
3733          /* Not UTF mode */
       /* Not UTF-8 mode */  
3734          {          {
3735          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3736            {            {
3737            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3738              {              {
3739              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3740              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3741              }              }
3742            if (fc == md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);            if (fc == *eptr || foc == *eptr) RRETURN(MATCH_NOMATCH);
3743              eptr++;
3744            }            }
3745          }          }
3746    
# Line 3390  for (;;) Line 3748  for (;;)
3748    
3749        if (minimize)        if (minimize)
3750          {          {
3751  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3752          /* UTF-8 mode */          if (utf)
         if (utf8)  
3753            {            {
3754            register unsigned int d;            register unsigned int d;
3755            for (fi = min;; fi++)            for (fi = min;; fi++)
3756              {              {
3757              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);
3758              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3759              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3760              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3761                {                {
3762                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3763                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3764                }                }
3765              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3766              if (d < 256) d = md->lcc[d];              if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
             if (fc == d) MRRETURN(MATCH_NOMATCH);  
3767              }              }
3768            }            }
3769          else          else
3770  #endif  #endif
3771          /* Not UTF-8 mode */          /* Not UTF mode */
3772            {            {
3773            for (fi = min;; fi++)            for (fi = min;; fi++)
3774              {              {
3775              RMATCH(eptr, ecode, offset_top, md, eptrb, RM29);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM29);
3776              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3777              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3778              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3779                {                {
3780                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3781                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3782                }                }
3783              if (fc == md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);              if (fc == *eptr || foc == *eptr) RRETURN(MATCH_NOMATCH);
3784                eptr++;
3785              }              }
3786            }            }
3787          /* Control never gets here */          /* Control never gets here */
# Line 3436  for (;;) Line 3793  for (;;)
3793          {          {
3794          pp = eptr;          pp = eptr;
3795    
3796  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3797          /* UTF-8 mode */          if (utf)
         if (utf8)  
3798            {            {
3799            register unsigned int d;            register unsigned int d;
3800            for (i = min; i < max; i++)            for (i = min; i < max; i++)
# Line 3450  for (;;) Line 3806  for (;;)
3806                break;                break;
3807                }                }
3808              GETCHARLEN(d, eptr, len);              GETCHARLEN(d, eptr, len);
3809              if (d < 256) d = md->lcc[d];              if (fc == d || (unsigned int)foc == d) break;
             if (fc == d) break;  
3810              eptr += len;              eptr += len;
3811              }              }
3812          if (possessive) continue;            if (possessive) continue;
3813          for(;;)            for(;;)
3814              {              {
3815              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);
3816              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
# Line 3465  for (;;) Line 3820  for (;;)
3820            }            }
3821          else          else
3822  #endif  #endif
3823          /* Not UTF-8 mode */          /* Not UTF mode */
3824            {            {
3825            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3826              {              {
# Line 3474  for (;;) Line 3829  for (;;)
3829                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3830                break;                break;
3831                }                }
3832              if (fc == md->lcc[*eptr]) break;              if (fc == *eptr || foc == *eptr) break;
3833              eptr++;              eptr++;
3834              }              }
3835            if (possessive) continue;            if (possessive) continue;
# Line 3486  for (;;) Line 3841  for (;;)
3841              }              }
3842            }            }
3843    
3844          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3845          }          }
3846        /* Control never gets here */        /* Control never gets here */
3847        }        }
# Line 3495  for (;;) Line 3850  for (;;)
3850    
3851      else      else
3852        {        {
3853  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3854        /* UTF-8 mode */        if (utf)
       if (utf8)  
3855          {          {
3856          register unsigned int d;          register unsigned int d;
3857          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3505  for (;;) Line 3859  for (;;)
3859            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3860              {              {
3861              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3862              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3863              }              }
3864            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3865            if (fc == d) MRRETURN(MATCH_NOMATCH);            if (fc == d) RRETURN(MATCH_NOMATCH);
3866            }            }
3867          }          }
3868        else        else
3869  #endif  #endif
3870        /* Not UTF-8 mode */        /* Not UTF mode */
3871          {          {
3872          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3873            {            {
3874            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3875              {              {
3876              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3877              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3878              }              }
3879            if (fc == *eptr++) MRRETURN(MATCH_NOMATCH);            if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
3880            }            }
3881          }          }
3882    
# Line 3530  for (;;) Line 3884  for (;;)
3884    
3885        if (minimize)        if (minimize)
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 (fi = min;; fi++)            for (fi = min;; fi++)
3892              {              {
3893              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);
3894              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3895              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3896              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3897                {                {
3898                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3899                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3900                }                }
3901              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3902              if (fc == d) MRRETURN(MATCH_NOMATCH);              if (fc == d) RRETURN(MATCH_NOMATCH);
3903              }              }
3904            }            }
3905          else          else
3906  #endif  #endif
3907          /* Not UTF-8 mode */          /* Not UTF mode */
3908            {            {
3909            for (fi = min;; fi++)            for (fi = min;; fi++)
3910              {              {
3911              RMATCH(eptr, ecode, offset_top, md, eptrb, RM33);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM33);
3912              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3913              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3914              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3915                {                {
3916                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3917                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3918                }                }
3919              if (fc == *eptr++) MRRETURN(MATCH_NOMATCH);              if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
3920              }              }
3921            }            }
3922          /* Control never gets here */          /* Control never gets here */
# Line 3575  for (;;) Line 3928  for (;;)
3928          {          {
3929          pp = eptr;          pp = eptr;
3930    
3931  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3932          /* UTF-8 mode */          if (utf)
         if (utf8)  
3933            {            {
3934            register unsigned int d;            register unsigned int d;
3935            for (i = min; i < max; i++)            for (i = min; i < max; i++)
# Line 3603  for (;;) Line 3955  for (;;)
3955            }            }
3956          else          else
3957  #endif  #endif
3958          /* Not UTF-8 mode */          /* Not UTF mode */
3959            {            {
3960            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3961              {              {
# Line 3624  for (;;) Line 3976  for (;;)
3976              }              }
3977            }            }
3978    
3979          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3980          }          }
3981        }        }
3982      /* Control never gets here */      /* Control never gets here */
# Line 3636  for (;;) Line 3988  for (;;)
3988      case OP_TYPEEXACT:      case OP_TYPEEXACT:
3989      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3990      minimize = TRUE;      minimize = TRUE;
3991      ecode += 3;      ecode += 1 + IMM2_SIZE;
3992      goto REPEATTYPE;      goto REPEATTYPE;
3993    
3994      case OP_TYPEUPTO:      case OP_TYPEUPTO:
# Line 3644  for (;;) Line 3996  for (;;)
3996      min = 0;      min = 0;
3997      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3998      minimize = *ecode == OP_TYPEMINUPTO;      minimize = *ecode == OP_TYPEMINUPTO;
3999      ecode += 3;      ecode += 1 + IMM2_SIZE;
4000      goto REPEATTYPE;      goto REPEATTYPE;
4001    
4002      case OP_TYPEPOSSTAR:      case OP_TYPEPOSSTAR:
# Line 3672  for (;;) Line 4024  for (;;)
4024      possessive = TRUE;      possessive = TRUE;
4025      min = 0;      min = 0;
4026      max = GET2(ecode, 1);      max = GET2(ecode, 1);
4027      ecode += 3;      ecode += 1 + IMM2_SIZE;
4028      goto REPEATTYPE;      goto REPEATTYPE;
4029    
4030      case OP_TYPESTAR:      case OP_TYPESTAR:
# Line 3718  for (;;) Line 4070  for (;;)
4070          switch(prop_type)          switch(prop_type)
4071            {            {
4072            case PT_ANY:            case PT_ANY:
4073            if (prop_fail_result) MRRETURN(MATCH_NOMATCH);            if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4074            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4075              {              {
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              }              }
# Line 3737  for (;;) Line 4089  for (;;)
4089              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4090                {                {
4091                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4092                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4093                }                }
4094              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4095              chartype = UCD_CHARTYPE(c);              chartype = UCD_CHARTYPE(c);
4096              if ((chartype == ucp_Lu ||              if ((chartype == ucp_Lu ||
4097                   chartype == ucp_Ll ||                   chartype == ucp_Ll ||
4098                   chartype == ucp_Lt) == prop_fail_result)                   chartype == ucp_Lt) == prop_fail_result)
4099                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4100              }              }
4101            break;            break;
4102    
# Line 3754  for (;;) Line 4106  for (;;)
4106              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4107                {                {
4108                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4109                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4110                }                }
4111              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4112              if ((UCD_CATEGORY(c) == prop_value) == prop_fail_result)              if ((UCD_CATEGORY(c) == prop_value) == prop_fail_result)
4113                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4114              }              }
4115            break;            break;
4116    
# Line 3768  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_CHARTYPE(c) == prop_value) == prop_fail_result)              if ((UCD_CHARTYPE(c) == prop_value) == prop_fail_result)
4127                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4128              }              }
4129            break;            break;
4130    
# Line 3782  for (;;) Line 4134  for (;;)
4134              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4135                {                {
4136                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4137                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4138                }                }
4139              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4140              if ((UCD_SCRIPT(c) == prop_value) == prop_fail_result)              if ((UCD_SCRIPT(c) == prop_value) == prop_fail_result)
4141                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4142              }              }
4143            break;            break;
4144    
# Line 3797  for (;;) Line 4149  for (;;)
4149              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4150                {                {
4151                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4152                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4153                }                }
4154              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4155              category = UCD_CATEGORY(c);              category = UCD_CATEGORY(c);
4156              if ((category == ucp_L || category == ucp_N) == prop_fail_result)              if ((category == ucp_L || category == ucp_N) == prop_fail_result)
4157                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4158              }              }
4159            break;            break;
4160    
# Line 3812  for (;;) Line 4164  for (;;)
4164              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4165                {                {
4166                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4167                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4168                }                }
4169              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4170              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||
4171                   c == CHAR_FF || c == CHAR_CR)                   c == CHAR_FF || c == CHAR_CR)
4172                     == prop_fail_result)                     == prop_fail_result)
4173                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4174              }              }
4175            break;            break;
4176    
# Line 3828  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);              GETCHARINCTEST(c, eptr);
4186              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||
4187                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)
4188                     == prop_fail_result)                     == prop_fail_result)
4189                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4190              }              }
4191            break;            break;
4192    
# Line 3845  for (;;) Line 4197  for (;;)
4197              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4198                {                {
4199                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4200                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4201                }                }
4202              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4203              category = UCD_CATEGORY(c);              category = UCD_CATEGORY(c);
4204              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)
4205                     == prop_fail_result)                     == prop_fail_result)
4206                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4207              }              }
4208            break;            break;
4209    
# Line 3872  for (;;) Line 4224  for (;;)
4224            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4225              {              {
4226              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4227              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4228              }              }
4229            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
4230            if (UCD_CATEGORY(c) == ucp_M) MRRETURN(MATCH_NOMATCH);            if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);
4231            while (eptr < md->end_subject)            while (eptr < md->end_subject)
4232              {              {
4233              int len = 1;              int len = 1;
4234              if (!utf8) c = *eptr; else { GETCHARLEN(c, eptr, len); }              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
4235              if (UCD_CATEGORY(c) != ucp_M) break;              if (UCD_CATEGORY(c) != ucp_M) break;
4236              eptr += len;              eptr += len;
4237              }              }
4238              CHECK_PARTIAL();
4239            }            }
4240          }          }
4241    
# Line 3891  for (;;) Line 4244  for (;;)
4244    
4245  /* Handle all other cases when the coding is UTF-8 */  /* Handle all other cases when the coding is UTF-8 */
4246    
4247  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4248        if (utf8) switch(ctype)        if (utf) switch(ctype)
4249          {          {
4250          case OP_ANY:          case OP_ANY:
4251          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3900  for (;;) Line 4253  for (;;)
4253            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4254              {              {
4255              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4256              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4257                }
4258              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4259              if (md->partial != 0 &&
4260                  eptr + 1 >= md->end_subject &&
4261                  NLBLOCK->nltype == NLTYPE_FIXED &&
4262                  NLBLOCK->nllen == 2 &&
4263                  *eptr == NLBLOCK->nl[0])
4264                {
4265                md->hitend = TRUE;
4266                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4267              }              }
           if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);  
4268            eptr++;            eptr++;
4269            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4270            }            }
4271          break;          break;
4272    
# Line 3914  for (;;) Line 4276  for (;;)
4276            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4277              {              {
4278              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4279              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4280              }              }
4281            eptr++;            eptr++;
4282            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4283            }            }
4284          break;          break;
4285    
4286          case OP_ANYBYTE:          case OP_ANYBYTE:
4287          if (eptr > md->end_subject - min) MRRETURN(MATCH_NOMATCH);          if (eptr > md->end_subject - min) RRETURN(MATCH_NOMATCH);
4288          eptr += min;          eptr += min;
4289          break;          break;
4290    
# Line 3932  for (;;) Line 4294  for (;;)
4294            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4295              {              {
4296              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4297              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4298              }              }
4299            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4300            switch(c)            switch(c)
4301              {              {
4302              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4303    
4304              case 0x000d:              case 0x000d:
4305              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
# Line 3951  for (;;) Line 4313  for (;;)
4313              case 0x0085:              case 0x0085:
4314              case 0x2028:              case 0x2028:
4315              case 0x2029:              case 0x2029:
4316              if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4317              break;              break;
4318              }              }
4319            }            }
# Line 3963  for (;;) Line 4325  for (;;)
4325            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4326              {              {
4327              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4328              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4329              }              }
4330            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4331            switch(c)            switch(c)
# Line 3988  for (;;) Line 4350  for (;;)
4350              case 0x202f:    /* NARROW NO-BREAK SPACE */              case 0x202f:    /* NARROW NO-BREAK SPACE */
4351              case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */              case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
4352              case 0x3000:    /* IDEOGRAPHIC SPACE */              case 0x3000:    /* IDEOGRAPHIC SPACE */
4353              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4354              }              }
4355            }            }
4356          break;          break;
# Line 3999  for (;;) Line 4361  for (;;)
4361            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4362              {              {
4363              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4364              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4365              }              }
4366            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4367            switch(c)            switch(c)
4368              {              {
4369              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4370              case 0x09:      /* HT */              case 0x09:      /* HT */
4371              case 0x20:      /* SPACE */              case 0x20:      /* SPACE */
4372              case 0xa0:      /* NBSP */              case 0xa0:      /* NBSP */
# Line 4035  for (;;) Line 4397  for (;;)
4397            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4398              {              {
4399              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4400              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4401              }              }
4402            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4403            switch(c)            switch(c)
# Line 4048  for (;;) Line 4410  for (;;)
4410              case 0x85:      /* NEL */              case 0x85:      /* NEL */
4411              case 0x2028:    /* LINE SEPARATOR */              case 0x2028:    /* LINE SEPARATOR */
4412              case 0x2029:    /* PARAGRAPH SEPARATOR */              case 0x2029:    /* PARAGRAPH SEPARATOR */
4413              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4414              }              }
4415            }            }
4416          break;          break;
# Line 4059  for (;;) Line 4421  for (;;)
4421            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4422              {              {
4423              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4424              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4425              }              }
4426            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4427            switch(c)            switch(c)
4428              {              {
4429              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4430              case 0x0a:      /* LF */              case 0x0a:      /* LF */
4431              case 0x0b:      /* VT */              case 0x0b:      /* VT */
4432              case 0x0c:      /* FF */              case 0x0c:      /* FF */
# Line 4083  for (;;) Line 4445  for (;;)
4445            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4446              {              {
4447              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4448              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4449              }              }
4450            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4451            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)
4452              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4453            }            }
4454          break;          break;
4455    
# Line 4097  for (;;) Line 4459  for (;;)
4459            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4460              {              {
4461              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4462              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4463              }              }
4464            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_digit) == 0)            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_digit) == 0)
4465              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4466              eptr++;
4467            /* 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 */
4468            }            }
4469          break;          break;
# Line 4111  for (;;) Line 4474  for (;;)
4474            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4475              {              {
4476              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4477              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4478              }              }
4479            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)
4480              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4481            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);            eptr++;
4482              ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4483            }            }
4484          break;          break;
4485    
# Line 4125  for (;;) Line 4489  for (;;)
4489            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4490              {              {
4491              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4492              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4493              }              }
4494            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_space) == 0)            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_space) == 0)
4495              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4496              eptr++;
4497            /* 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 */
4498            }            }
4499          break;          break;
# Line 4139  for (;;) Line 4504  for (;;)
4504            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4505              {              {
4506              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4507              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4508              }              }
4509            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)
4510              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4511            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);            eptr++;
4512              ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4513            }            }
4514          break;          break;
4515    
# Line 4153  for (;;) Line 4519  for (;;)
4519            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4520              {              {
4521              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4522              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4523              }              }
4524            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_word) == 0)            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_word) == 0)
4525              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4526              eptr++;
4527            /* 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 */
4528            }            }
4529          break;          break;
# Line 4166  for (;;) Line 4533  for (;;)
4533          }  /* End switch(ctype) */          }  /* End switch(ctype) */
4534    
4535        else        else
4536  #endif     /* SUPPORT_UTF8 */  #endif     /* SUPPORT_UTF */
4537    
4538        /* 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
4539        than OP_PROP and OP_NOTPROP. */        than OP_PROP and OP_NOTPROP. */
# Line 4179  for (;;) Line 4546  for (;;)
4546            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4547              {              {
4548              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4549              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4550                }
4551              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4552              if (md->partial != 0 &&
4553                  eptr + 1 >= md->end_subject &&
4554                  NLBLOCK->nltype == NLTYPE_FIXED &&
4555                  NLBLOCK->nllen == 2 &&
4556                  *eptr == NLBLOCK->nl[0])
4557                {
4558                md->hitend = TRUE;
4559                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4560              }              }
           if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);  
4561            eptr++;            eptr++;
4562            }            }
4563          break;          break;
# Line 4190  for (;;) Line 4566  for (;;)
4566          if (eptr > md->end_subject - min)          if (eptr > md->end_subject - min)
4567            {            {
4568            SCHECK_PARTIAL();            SCHECK_PARTIAL();
4569            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
4570            }            }
4571          eptr += min;          eptr += min;
4572          break;          break;
# Line 4199  for (;;) Line 4575  for (;;)
4575          if (eptr > md->end_subject - min)          if (eptr > md->end_subject - min)
4576            {            {
4577            SCHECK_PARTIAL();            SCHECK_PARTIAL();
4578            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
4579            }            }
4580          eptr += min;          eptr += min;
4581          break;          break;
# Line 4210  for (;;) Line 4586  for (;;)
4586            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4587              {              {
4588              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4589              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4590              }              }
4591            switch(*eptr++)            switch(*eptr++)
4592              {              {
4593              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4594    
4595              case 0x000d:              case 0x000d:
4596              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
# Line 4226  for (;;) Line 4602  for (;;)
4602              case 0x000b:              case 0x000b:
4603              case 0x000c:              case 0x000c:
4604              case 0x0085:              case 0x0085:
4605              if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);  #ifdef COMPILE_PCRE16
4606                case 0x2028:
4607                case 0x2029:
4608    #endif
4609                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4610              break;              break;
4611              }              }
4612            }            }
# Line 4238  for (;;) Line 4618  for (;;)
4618            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4619              {              {
4620              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4621              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4622              }              }
4623            switch(*eptr++)            switch(*eptr++)
4624              {              {
# Line 4246  for (;;) Line 4626  for (;;)
4626              case 0x09:      /* HT */              case 0x09:      /* HT */
4627              case 0x20:      /* SPACE */              case 0x20:      /* SPACE */
4628              case 0xa0:      /* NBSP */              case 0xa0:      /* NBSP */
4629              MRRETURN(MATCH_NOMATCH);  #ifdef COMPILE_PCRE16
4630                case 0x1680:    /* OGHAM SPACE MARK */
4631                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
4632                case 0x2000:    /* EN QUAD */
4633                case 0x2001:    /* EM QUAD */
4634                case 0x2002:    /* EN SPACE */
4635                case 0x2003:    /* EM SPACE */
4636                case 0x2004:    /* THREE-PER-EM SPACE */
4637                case 0x2005:    /* FOUR-PER-EM SPACE */
4638                case 0x2006:    /* SIX-PER-EM SPACE */
4639                case 0x2007:    /* FIGURE SPACE */
4640                case 0x2008:    /* PUNCTUATION SPACE */
4641                case 0x2009:    /* THIN SPACE */
4642                case 0x200A:    /* HAIR SPACE */
4643                case 0x202f:    /* NARROW NO-BREAK SPACE */
4644                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
4645                case 0x3000:    /* IDEOGRAPHIC SPACE */
4646    #endif
4647                RRETURN(MATCH_NOMATCH);
4648              }              }
4649            }            }
4650          break;          break;
# Line 4257  for (;;) Line 4655  for (;;)
4655            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4656              {              {
4657              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4658              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4659              }              }
4660            switch(*eptr++)            switch(*eptr++)
4661              {              {
4662              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4663              case 0x09:      /* HT */              case 0x09:      /* HT */
4664              case 0x20:      /* SPACE */              case 0x20:      /* SPACE */
4665              case 0xa0:      /* NBSP */              case 0xa0:      /* NBSP */
4666    #ifdef COMPILE_PCRE16
4667                case 0x1680:    /* OGHAM SPACE MARK */
4668                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
4669                case 0x2000:    /* EN QUAD */
4670                case 0x2001:    /* EM QUAD */
4671                case 0x2002:    /* EN SPACE */
4672                case 0x2003:    /* EM SPACE */
4673                case 0x2004:    /* THREE-PER-EM SPACE */
4674                case 0x2005:    /* FOUR-PER-EM SPACE */
4675                case 0x2006:    /* SIX-PER-EM SPACE */
4676                case 0x2007:    /* FIGURE SPACE */
4677                case 0x2008:    /* PUNCTUATION SPACE */
4678                case 0x2009:    /* THIN SPACE */
4679                case 0x200A:    /* HAIR SPACE */
4680                case 0x202f:    /* NARROW NO-BREAK SPACE */
4681                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
4682                case 0x3000:    /* IDEOGRAPHIC SPACE */
4683    #endif
4684              break;              break;
4685              }              }
4686            }            }
# Line 4276  for (;;) Line 4692  for (;;)
4692            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4693              {              {
4694              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4695              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4696              }              }
4697            switch(*eptr++)            switch(*eptr++)
4698              {              {
# Line 4286  for (;;) Line 4702  for (;;)
4702              case 0x0c:      /* FF */              case 0x0c:      /* FF */
4703              case 0x0d:      /* CR */              case 0x0d:      /* CR */
4704              case 0x85:      /* NEL */              case 0x85:      /* NEL */
4705              MRRETURN(MATCH_NOMATCH);  #ifdef COMPILE_PCRE16
4706                case 0x2028:    /* LINE SEPARATOR */
4707                case 0x2029:    /* PARAGRAPH SEPARATOR */
4708    #endif
4709                RRETURN(MATCH_NOMATCH);
4710              }              }
4711            }            }
4712          break;          break;
# Line 4297  for (;;) Line 4717  for (;;)
4717            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4718              {              {
4719              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4720              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4721              }              }
4722            switch(*eptr++)            switch(*eptr++)
4723              {              {
4724              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4725              case 0x0a:      /* LF */              case 0x0a:      /* LF */
4726              case 0x0b:      /* VT */              case 0x0b:      /* VT */
4727              case 0x0c:      /* FF */              case 0x0c:      /* FF */
4728              case 0x0d:      /* CR */              case 0x0d:      /* CR */
4729              case 0x85:      /* NEL */              case 0x85:      /* NEL */
4730    #ifdef COMPILE_PCRE16
4731                case 0x2028:    /* LINE SEPARATOR */
4732                case 0x2029:    /* PARAGRAPH SEPARATOR */
4733    #endif
4734              break;              break;
4735              }              }
4736            }            }
# Line 4318  for (;;) Line 4742  for (;;)
4742            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4743              {              {
4744              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4745              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4746              }              }
4747            if ((md->ctypes[*eptr++] & ctype_digit) != 0) MRRETURN(MATCH_NOMATCH);            if (MAX_255(*eptr) && (md->ctypes[*eptr] & ctype_digit) != 0)
4748                RRETURN(MATCH_NOMATCH);
4749              eptr++;
4750            }            }
4751          break;          break;
4752    
# Line 4330  for (;;) Line 4756  for (;;)
4756            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4757              {              {
4758              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4759              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4760              }              }
4761            if ((md->ctypes[*eptr++] & ctype_digit) == 0) MRRETURN(MATCH_NOMATCH);            if (!MAX_255(*eptr) || (md->ctypes[*eptr] & ctype_digit) == 0)
4762                RRETURN(MATCH_NOMATCH);
4763              eptr++;
4764            }            }
4765          break;          break;
4766    
# Line 4342  for (;;) Line 4770  for (;;)