/[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 975 by ph10, Sat Jun 2 11:03:06 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 &&        md->mark = save_mark;
1579            (rrc != MATCH_THEN || md->start_match_ptr != ecode))  
1580          RRETURN(rrc);        /* A COMMIT failure must fail the entire assertion, without trying any
1581          subsequent branches. */
1582    
1583          if (rrc == MATCH_COMMIT) RRETURN(MATCH_NOMATCH);
1584    
1585          /* PCRE does not allow THEN to escape beyond an assertion; it
1586          is treated as NOMATCH. */
1587    
1588          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1589        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1590        }        }
1591      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
1592    
1593      if (*ecode == OP_KET) MRRETURN(MATCH_NOMATCH);      if (*ecode == OP_KET) RRETURN(MATCH_NOMATCH);
1594    
1595      /* If checking an assertion for a condition, return MATCH_MATCH. */      /* If checking an assertion for a condition, return MATCH_MATCH. */
1596    
# Line 1396  for (;;) Line 1610  for (;;)
1610    
1611      case OP_ASSERT_NOT:      case OP_ASSERT_NOT:
1612      case OP_ASSERTBACK_NOT:      case OP_ASSERTBACK_NOT:
1613        save_mark = md->mark;
1614      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1615        {        {
1616        condassert = TRUE;        condassert = TRUE;
# Line 1406  for (;;) Line 1621  for (;;)
1621      do      do
1622        {        {
1623        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);
1624        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) MRRETURN(MATCH_NOMATCH);        md->mark = save_mark;
1625          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) RRETURN(MATCH_NOMATCH);
1626        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)
1627          {          {
1628          do ecode += GET(ecode,1); while (*ecode == OP_ALT);          do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1629          break;          break;
1630          }          }
1631        if (rrc != MATCH_NOMATCH &&  
1632            (rrc != MATCH_THEN || md->start_match_ptr != ecode))        /* PCRE does not allow THEN to escape beyond an assertion; it is treated
1633          RRETURN(rrc);        as NOMATCH. */
1634    
1635          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1636        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1637        }        }
1638      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 1430  for (;;) Line 1648  for (;;)
1648      back a number of characters, not bytes. */      back a number of characters, not bytes. */
1649    
1650      case OP_REVERSE:      case OP_REVERSE:
1651  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
1652      if (utf8)      if (utf)
1653        {        {
1654        i = GET(ecode, 1);        i = GET(ecode, 1);
1655        while (i-- > 0)        while (i-- > 0)
1656          {          {
1657          eptr--;          eptr--;
1658          if (eptr < md->start_subject) MRRETURN(MATCH_NOMATCH);          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1659          BACKCHAR(eptr);          BACKCHAR(eptr);
1660          }          }
1661        }        }
# Line 1448  for (;;) Line 1666  for (;;)
1666    
1667        {        {
1668        eptr -= GET(ecode, 1);        eptr -= GET(ecode, 1);
1669        if (eptr < md->start_subject) MRRETURN(MATCH_NOMATCH);        if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1670        }        }
1671    
1672      /* 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 1680  for (;;)
1680      function is able to force a failure. */      function is able to force a failure. */
1681    
1682      case OP_CALLOUT:      case OP_CALLOUT:
1683      if (pcre_callout != NULL)      if (PUBL(callout) != NULL)
1684        {        {
1685        pcre_callout_block cb;        PUBL(callout_block) cb;
1686        cb.version          = 2;   /* Version 1 of the callout block */        cb.version          = 2;   /* Version 1 of the callout block */
1687        cb.callout_number   = ecode[1];        cb.callout_number   = ecode[1];
1688        cb.offset_vector    = md->offset_vector;        cb.offset_vector    = md->offset_vector;
1689    #ifdef COMPILE_PCRE8
1690        cb.subject          = (PCRE_SPTR)md->start_subject;        cb.subject          = (PCRE_SPTR)md->start_subject;
1691    #else
1692          cb.subject          = (PCRE_SPTR16)md->start_subject;
1693    #endif
1694        cb.subject_length   = (int)(md->end_subject - md->start_subject);        cb.subject_length   = (int)(md->end_subject - md->start_subject);
1695        cb.start_match      = (int)(mstart - md->start_subject);        cb.start_match      = (int)(mstart - md->start_subject);
1696        cb.current_position = (int)(eptr - md->start_subject);        cb.current_position = (int)(eptr - md->start_subject);
# Line 1477  for (;;) Line 1699  for (;;)
1699        cb.capture_top      = offset_top/2;        cb.capture_top      = offset_top/2;
1700        cb.capture_last     = md->capture_last;        cb.capture_last     = md->capture_last;
1701        cb.callout_data     = md->callout_data;        cb.callout_data     = md->callout_data;
1702        cb.mark             = markptr;        cb.mark             = md->nomatch_mark;
1703        if ((rrc = (*pcre_callout)(&cb)) > 0) MRRETURN(MATCH_NOMATCH);        if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1704        if (rrc < 0) RRETURN(rrc);        if (rrc < 0) RRETURN(rrc);
1705        }        }
1706      ecode += 2 + 2*LINK_SIZE;      ecode += 2 + 2*LINK_SIZE;
# Line 1537  for (;;) Line 1759  for (;;)
1759        else        else
1760          {          {
1761          new_recursive.offset_save =          new_recursive.offset_save =
1762            (int *)(pcre_malloc)(new_recursive.saved_max * sizeof(int));            (int *)(PUBL(malloc))(new_recursive.saved_max * sizeof(int));
1763          if (new_recursive.offset_save == NULL) RRETURN(PCRE_ERROR_NOMEMORY);          if (new_recursive.offset_save == NULL) RRETURN(PCRE_ERROR_NOMEMORY);
1764          }          }
1765        memcpy(new_recursive.offset_save, md->offset_vector,        memcpy(new_recursive.offset_save, md->offset_vector,
# Line 1552  for (;;) Line 1774  for (;;)
1774        do        do
1775          {          {
1776          if (cbegroup) md->match_function_type = MATCH_CBEGROUP;          if (cbegroup) md->match_function_type = MATCH_CBEGROUP;
1777          RMATCH(eptr, callpat + _pcre_OP_lengths[*callpat], offset_top,          RMATCH(eptr, callpat + PRIV(OP_lengths)[*callpat], offset_top,
1778            md, eptrb, RM6);            md, eptrb, RM6);
1779          memcpy(md->offset_vector, new_recursive.offset_save,          memcpy(md->offset_vector, new_recursive.offset_save,
1780              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1781            md->recursive = new_recursive.prevrec;
1782          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1783            {            {
1784            DPRINTF(("Recursion matched\n"));            DPRINTF(("Recursion matched\n"));
           md->recursive = new_recursive.prevrec;  
1785            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
1786              (pcre_free)(new_recursive.offset_save);              (PUBL(free))(new_recursive.offset_save);
1787    
1788            /* 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
1789            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 1793  for (;;)
1793            mstart = md->start_match_ptr;            mstart = md->start_match_ptr;
1794            goto RECURSION_MATCHED;        /* Exit loop; end processing */            goto RECURSION_MATCHED;        /* Exit loop; end processing */
1795            }            }
1796          else if (rrc != MATCH_NOMATCH &&  
1797                  (rrc != MATCH_THEN || md->start_match_ptr != ecode))          /* PCRE does not allow THEN or COMMIT to escape beyond a recursion; it
1798            is treated as NOMATCH. */
1799    
1800            else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN &&
1801                     rrc != MATCH_COMMIT)
1802            {            {
1803            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1804            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
1805              (pcre_free)(new_recursive.offset_save);              (PUBL(free))(new_recursive.offset_save);
1806            RRETURN(rrc);            RRETURN(rrc);
1807            }            }
1808    
# Line 1588  for (;;) Line 1814  for (;;)
1814        DPRINTF(("Recursion didn't match\n"));        DPRINTF(("Recursion didn't match\n"));
1815        md->recursive = new_recursive.prevrec;        md->recursive = new_recursive.prevrec;
1816        if (new_recursive.offset_save != stacksave)        if (new_recursive.offset_save != stacksave)
1817          (pcre_free)(new_recursive.offset_save);          (PUBL(free))(new_recursive.offset_save);
1818        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
1819        }        }
1820    
1821      RECURSION_MATCHED:      RECURSION_MATCHED:
# Line 1658  for (;;) Line 1884  for (;;)
1884        }        }
1885      else saved_eptr = NULL;      else saved_eptr = NULL;
1886    
1887      /* 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
1888      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
1889      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
1890      by \K. */      start in case it was changed by \K. */
1891    
1892      if (*prev == OP_ASSERT || *prev == OP_ASSERT_NOT ||      if ((*prev >= OP_ASSERT && *prev <= OP_ASSERTBACK_NOT) ||
1893          *prev == OP_ASSERTBACK || *prev == OP_ASSERTBACK_NOT)           *prev == OP_ONCE_NC)
1894        {        {
1895        md->end_match_ptr = eptr;      /* For ONCE */        md->end_match_ptr = eptr;      /* For ONCE_NC */
1896        md->end_offset_top = offset_top;        md->end_offset_top = offset_top;
1897        md->start_match_ptr = mstart;        md->start_match_ptr = mstart;
1898        MRRETURN(MATCH_MATCH);         /* Sets md->mark */        RRETURN(MATCH_MATCH);         /* Sets md->mark */
1899        }        }
1900    
1901      /* 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 1960  for (;;)
1960      /* For an ordinary non-repeating ket, just continue at this level. This      /* For an ordinary non-repeating ket, just continue at this level. This
1961      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
1962      group. This is the forcible breaking of infinite loops as implemented in      group. This is the forcible breaking of infinite loops as implemented in
1963      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,
1964      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
1965      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
1966      bypassing intermediate backup points, but resetting any captures that      original OP_ONCE level, thereby bypassing intermediate backup points, but
1967      happened along the way. */      resetting any captures that happened along the way. */
1968    
1969      if (*ecode == OP_KET || eptr == saved_eptr)      if (*ecode == OP_KET || eptr == saved_eptr)
1970        {        {
# Line 1811  for (;;) Line 2037  for (;;)
2037      /* Not multiline mode: start of subject assertion, unless notbol. */      /* Not multiline mode: start of subject assertion, unless notbol. */
2038    
2039      case OP_CIRC:      case OP_CIRC:
2040      if (md->notbol && eptr == md->start_subject) MRRETURN(MATCH_NOMATCH);      if (md->notbol && eptr == md->start_subject) RRETURN(MATCH_NOMATCH);
2041    
2042      /* Start of subject assertion */      /* Start of subject assertion */
2043    
2044      case OP_SOD:      case OP_SOD:
2045      if (eptr != md->start_subject) MRRETURN(MATCH_NOMATCH);      if (eptr != md->start_subject) RRETURN(MATCH_NOMATCH);
2046      ecode++;      ecode++;
2047      break;      break;
2048    
2049      /* Multiline mode: start of subject unless notbol, or after any newline. */      /* Multiline mode: start of subject unless notbol, or after any newline. */
2050    
2051      case OP_CIRCM:      case OP_CIRCM:
2052      if (md->notbol && eptr == md->start_subject) MRRETURN(MATCH_NOMATCH);      if (md->notbol && eptr == md->start_subject) RRETURN(MATCH_NOMATCH);
2053      if (eptr != md->start_subject &&      if (eptr != md->start_subject &&
2054          (eptr == md->end_subject || !WAS_NEWLINE(eptr)))          (eptr == md->end_subject || !WAS_NEWLINE(eptr)))
2055        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2056      ecode++;      ecode++;
2057      break;      break;
2058    
2059      /* Start of match assertion */      /* Start of match assertion */
2060    
2061      case OP_SOM:      case OP_SOM:
2062      if (eptr != md->start_subject + md->start_offset) MRRETURN(MATCH_NOMATCH);      if (eptr != md->start_subject + md->start_offset) RRETURN(MATCH_NOMATCH);
2063      ecode++;      ecode++;
2064      break;      break;
2065    
# Line 1849  for (;;) Line 2075  for (;;)
2075    
2076      case OP_DOLLM:      case OP_DOLLM:
2077      if (eptr < md->end_subject)      if (eptr < md->end_subject)
2078        { if (!IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH); }        {
2079          if (!IS_NEWLINE(eptr))
2080            {
2081            if (md->partial != 0 &&
2082                eptr + 1 >= md->end_subject &&
2083                NLBLOCK->nltype == NLTYPE_FIXED &&
2084                NLBLOCK->nllen == 2 &&
2085                *eptr == NLBLOCK->nl[0])
2086              {
2087              md->hitend = TRUE;
2088              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2089              }
2090            RRETURN(MATCH_NOMATCH);
2091            }
2092          }
2093      else      else
2094        {        {
2095        if (md->noteol) MRRETURN(MATCH_NOMATCH);        if (md->noteol) RRETURN(MATCH_NOMATCH);
2096        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2097        }        }
2098      ecode++;      ecode++;
# Line 1862  for (;;) Line 2102  for (;;)
2102      subject unless noteol is set. */      subject unless noteol is set. */
2103    
2104      case OP_DOLL:      case OP_DOLL:
2105      if (md->noteol) MRRETURN(MATCH_NOMATCH);      if (md->noteol) RRETURN(MATCH_NOMATCH);
2106      if (!md->endonly) goto ASSERT_NL_OR_EOS;      if (!md->endonly) goto ASSERT_NL_OR_EOS;
2107    
2108      /* ... else fall through for endonly */      /* ... else fall through for endonly */
# Line 1870  for (;;) Line 2110  for (;;)
2110      /* End of subject assertion (\z) */      /* End of subject assertion (\z) */
2111    
2112      case OP_EOD:      case OP_EOD:
2113      if (eptr < md->end_subject) MRRETURN(MATCH_NOMATCH);      if (eptr < md->end_subject) RRETURN(MATCH_NOMATCH);
2114      SCHECK_PARTIAL();      SCHECK_PARTIAL();
2115      ecode++;      ecode++;
2116      break;      break;
# Line 1881  for (;;) Line 2121  for (;;)
2121      ASSERT_NL_OR_EOS:      ASSERT_NL_OR_EOS:
2122      if (eptr < md->end_subject &&      if (eptr < md->end_subject &&
2123          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))
2124        MRRETURN(MATCH_NOMATCH);        {
2125          if (md->partial != 0 &&
2126              eptr + 1 >= md->end_subject &&
2127              NLBLOCK->nltype == NLTYPE_FIXED &&
2128              NLBLOCK->nllen == 2 &&
2129              *eptr == NLBLOCK->nl[0])
2130            {
2131            md->hitend = TRUE;
2132            if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2133            }
2134          RRETURN(MATCH_NOMATCH);
2135          }
2136    
2137      /* Either at end of string or \n before end. */      /* Either at end of string or \n before end. */
2138    
# Line 1900  for (;;) Line 2151  for (;;)
2151        be "non-word" characters. Remember the earliest consulted character for        be "non-word" characters. Remember the earliest consulted character for
2152        partial matching. */        partial matching. */
2153    
2154  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2155        if (utf8)        if (utf)
2156          {          {
2157          /* Get status of previous character */          /* Get status of previous character */
2158    
2159          if (eptr == md->start_subject) prev_is_word = FALSE; else          if (eptr == md->start_subject) prev_is_word = FALSE; else
2160            {            {
2161            USPTR lastptr = eptr - 1;            PCRE_PUCHAR lastptr = eptr - 1;
2162            while((*lastptr & 0xc0) == 0x80) lastptr--;            BACKCHAR(lastptr);
2163            if (lastptr < md->start_used_ptr) md->start_used_ptr = lastptr;            if (lastptr < md->start_used_ptr) md->start_used_ptr = lastptr;
2164            GETCHAR(c, lastptr);            GETCHAR(c, lastptr);
2165  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
# Line 1973  for (;;) Line 2224  for (;;)
2224              }              }
2225            else            else
2226  #endif  #endif
2227            prev_is_word = ((md->ctypes[eptr[-1]] & ctype_word) != 0);            prev_is_word = MAX_255(eptr[-1])
2228                && ((md->ctypes[eptr[-1]] & ctype_word) != 0);
2229            }            }
2230    
2231          /* Get status of next character */          /* Get status of next character */
# Line 1996  for (;;) Line 2248  for (;;)
2248            }            }
2249          else          else
2250  #endif  #endif
2251          cur_is_word = ((md->ctypes[*eptr] & ctype_word) != 0);          cur_is_word = MAX_255(*eptr)
2252              && ((md->ctypes[*eptr] & ctype_word) != 0);
2253          }          }
2254    
2255        /* Now see if the situation is what we want */        /* Now see if the situation is what we want */
2256    
2257        if ((*ecode++ == OP_WORD_BOUNDARY)?        if ((*ecode++ == OP_WORD_BOUNDARY)?
2258             cur_is_word == prev_is_word : cur_is_word != prev_is_word)             cur_is_word == prev_is_word : cur_is_word != prev_is_word)
2259          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2260        }        }
2261      break;      break;
2262    
2263      /* Match a single character type; inline for speed */      /* Match any single character type except newline; have to take care with
2264        CRLF newlines and partial matching. */
2265    
2266      case OP_ANY:      case OP_ANY:
2267      if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
2268        if (md->partial != 0 &&
2269            eptr + 1 >= md->end_subject &&
2270            NLBLOCK->nltype == NLTYPE_FIXED &&
2271            NLBLOCK->nllen == 2 &&
2272            *eptr == NLBLOCK->nl[0])
2273          {
2274          md->hitend = TRUE;
2275          if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2276          }
2277    
2278      /* Fall through */      /* Fall through */
2279    
2280        /* Match any single character whatsoever. */
2281    
2282      case OP_ALLANY:      case OP_ALLANY:
2283      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 */
2284        {                            /* not be updated before SCHECK_PARTIAL. */        {                            /* not be updated before SCHECK_PARTIAL. */
2285        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2286        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2287        }        }
2288      eptr++;      eptr++;
2289      if (utf8) while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  #ifdef SUPPORT_UTF
2290        if (utf) ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
2291    #endif
2292      ecode++;      ecode++;
2293      break;      break;
2294    
# Line 2031  for (;;) Line 2299  for (;;)
2299      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 */
2300        {                            /* not be updated before SCHECK_PARTIAL. */        {                            /* not be updated before SCHECK_PARTIAL. */
2301        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2302        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2303        }        }
2304      eptr++;      eptr++;
2305      ecode++;      ecode++;
# Line 2041  for (;;) Line 2309  for (;;)
2309      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2310        {        {
2311        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2312        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2313        }        }
2314      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2315      if (      if (
2316  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2317         c < 256 &&         c < 256 &&
2318  #endif  #endif
2319         (md->ctypes[c] & ctype_digit) != 0         (md->ctypes[c] & ctype_digit) != 0
2320         )         )
2321        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2322      ecode++;      ecode++;
2323      break;      break;
2324    
# Line 2058  for (;;) Line 2326  for (;;)
2326      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2327        {        {
2328        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2329        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2330        }        }
2331      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2332      if (      if (
2333  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2334         c >= 256 ||         c > 255 ||
2335  #endif  #endif
2336         (md->ctypes[c] & ctype_digit) == 0         (md->ctypes[c] & ctype_digit) == 0
2337         )         )
2338        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2339      ecode++;      ecode++;
2340      break;      break;
2341    
# Line 2075  for (;;) Line 2343  for (;;)
2343      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2344        {        {
2345        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2346        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2347        }        }
2348      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2349      if (      if (
2350  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2351         c < 256 &&         c < 256 &&
2352  #endif  #endif
2353         (md->ctypes[c] & ctype_space) != 0         (md->ctypes[c] & ctype_space) != 0
2354         )         )
2355        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2356      ecode++;      ecode++;
2357      break;      break;
2358    
# Line 2092  for (;;) Line 2360  for (;;)
2360      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2361        {        {
2362        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2363        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2364        }        }
2365      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2366      if (      if (
2367  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2368         c >= 256 ||         c > 255 ||
2369  #endif  #endif
2370         (md->ctypes[c] & ctype_space) == 0         (md->ctypes[c] & ctype_space) == 0
2371         )         )
2372        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2373      ecode++;      ecode++;
2374      break;      break;
2375    
# Line 2109  for (;;) Line 2377  for (;;)
2377      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2378        {        {
2379        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2380        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2381        }        }
2382      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2383      if (      if (
2384  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2385         c < 256 &&         c < 256 &&
2386  #endif  #endif
2387         (md->ctypes[c] & ctype_word) != 0         (md->ctypes[c] & ctype_word) != 0
2388         )         )
2389        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2390      ecode++;      ecode++;
2391      break;      break;
2392    
# Line 2126  for (;;) Line 2394  for (;;)
2394      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2395        {        {
2396        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2397        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2398        }        }
2399      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2400      if (      if (
2401  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2402         c >= 256 ||         c > 255 ||
2403  #endif  #endif
2404         (md->ctypes[c] & ctype_word) == 0         (md->ctypes[c] & ctype_word) == 0
2405         )         )
2406        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2407      ecode++;      ecode++;
2408      break;      break;
2409    
# Line 2143  for (;;) Line 2411  for (;;)
2411      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2412        {        {
2413        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2414        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2415        }        }
2416      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2417      switch(c)      switch(c)
2418        {        {
2419        default: MRRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2420    
2421        case 0x000d:        case 0x000d:
2422        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr >= md->end_subject)
2423            {
2424            SCHECK_PARTIAL();
2425            }
2426          else if (*eptr == 0x0a) eptr++;
2427        break;        break;
2428    
2429        case 0x000a:        case 0x000a:
# Line 2162  for (;;) Line 2434  for (;;)
2434        case 0x0085:        case 0x0085:
2435        case 0x2028:        case 0x2028:
2436        case 0x2029:        case 0x2029:
2437        if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);        if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
2438        break;        break;
2439        }        }
2440      ecode++;      ecode++;
# Line 2172  for (;;) Line 2444  for (;;)
2444      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2445        {        {
2446        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2447        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2448        }        }
2449      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2450      switch(c)      switch(c)
# Line 2197  for (;;) Line 2469  for (;;)
2469        case 0x202f:    /* NARROW NO-BREAK SPACE */        case 0x202f:    /* NARROW NO-BREAK SPACE */
2470        case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */        case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
2471        case 0x3000:    /* IDEOGRAPHIC SPACE */        case 0x3000:    /* IDEOGRAPHIC SPACE */
2472        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2473        }        }
2474      ecode++;      ecode++;
2475      break;      break;
# Line 2206  for (;;) Line 2478  for (;;)
2478      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2479        {        {
2480        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2481        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2482        }        }
2483      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2484      switch(c)      switch(c)
2485        {        {
2486        default: MRRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2487        case 0x09:      /* HT */        case 0x09:      /* HT */
2488        case 0x20:      /* SPACE */        case 0x20:      /* SPACE */
2489        case 0xa0:      /* NBSP */        case 0xa0:      /* NBSP */
# Line 2240  for (;;) Line 2512  for (;;)
2512      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2513        {        {
2514        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2515        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2516        }        }
2517      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2518      switch(c)      switch(c)
# Line 2253  for (;;) Line 2525  for (;;)
2525        case 0x85:      /* NEL */        case 0x85:      /* NEL */
2526        case 0x2028:    /* LINE SEPARATOR */        case 0x2028:    /* LINE SEPARATOR */
2527        case 0x2029:    /* PARAGRAPH SEPARATOR */        case 0x2029:    /* PARAGRAPH SEPARATOR */
2528        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2529        }        }
2530      ecode++;      ecode++;
2531      break;      break;
# Line 2262  for (;;) Line 2534  for (;;)
2534      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2535        {        {
2536        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2537        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2538        }        }
2539      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2540      switch(c)      switch(c)
2541        {        {
2542        default: MRRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2543        case 0x0a:      /* LF */        case 0x0a:      /* LF */
2544        case 0x0b:      /* VT */        case 0x0b:      /* VT */
2545        case 0x0c:      /* FF */        case 0x0c:      /* FF */
# Line 2289  for (;;) Line 2561  for (;;)
2561      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2562        {        {
2563        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2564        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2565        }        }
2566      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2567        {        {
# Line 2298  for (;;) Line 2570  for (;;)
2570        switch(ecode[1])        switch(ecode[1])
2571          {          {
2572          case PT_ANY:          case PT_ANY:
2573          if (op == OP_NOTPROP) MRRETURN(MATCH_NOMATCH);          if (op == OP_NOTPROP) RRETURN(MATCH_NOMATCH);
2574          break;          break;
2575    
2576          case PT_LAMP:          case PT_LAMP:
2577          if ((prop->chartype == ucp_Lu ||          if ((prop->chartype == ucp_Lu ||
2578               prop->chartype == ucp_Ll ||               prop->chartype == ucp_Ll ||
2579               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))
2580            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2581          break;          break;
2582    
2583          case PT_GC:          case PT_GC:
2584          if ((ecode[2] != _pcre_ucp_gentype[prop->chartype]) == (op == OP_PROP))          if ((ecode[2] != PRIV(ucp_gentype)[prop->chartype]) == (op == OP_PROP))
2585            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2586          break;          break;
2587    
2588          case PT_PC:          case PT_PC:
2589          if ((ecode[2] != prop->chartype) == (op == OP_PROP))          if ((ecode[2] != prop->chartype) == (op == OP_PROP))
2590            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2591          break;          break;
2592    
2593          case PT_SC:          case PT_SC:
2594          if ((ecode[2] != prop->script) == (op == OP_PROP))          if ((ecode[2] != prop->script) == (op == OP_PROP))
2595            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2596          break;          break;
2597    
2598          /* These are specials */          /* These are specials */
2599    
2600          case PT_ALNUM:          case PT_ALNUM:
2601          if ((_pcre_ucp_gentype[prop->chartype] == ucp_L ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2602               _pcre_ucp_gentype[prop->chartype] == ucp_N) == (op == OP_NOTPROP))               PRIV(ucp_gentype)[prop->chartype] == ucp_N) == (op == OP_NOTPROP))
2603            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2604          break;          break;
2605    
2606          case PT_SPACE:    /* Perl space */          case PT_SPACE:    /* Perl space */
2607          if ((_pcre_ucp_gentype[prop->chartype] == ucp_Z ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2608               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)
2609                 == (op == OP_NOTPROP))                 == (op == OP_NOTPROP))
2610            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2611          break;          break;
2612    
2613          case PT_PXSPACE:  /* POSIX space */          case PT_PXSPACE:  /* POSIX space */
2614          if ((_pcre_ucp_gentype[prop->chartype] == ucp_Z ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2615               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
2616               c == CHAR_FF || c == CHAR_CR)               c == CHAR_FF || c == CHAR_CR)
2617                 == (op == OP_NOTPROP))                 == (op == OP_NOTPROP))
2618            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2619          break;          break;
2620    
2621          case PT_WORD:          case PT_WORD:
2622          if ((_pcre_ucp_gentype[prop->chartype] == ucp_L ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2623               _pcre_ucp_gentype[prop->chartype] == ucp_N ||               PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
2624               c == CHAR_UNDERSCORE) == (op == OP_NOTPROP))               c == CHAR_UNDERSCORE) == (op == OP_NOTPROP))
2625            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2626          break;          break;
2627    
2628          /* This should never occur */          /* This should never occur */
# Line 2370  for (;;) Line 2642  for (;;)
2642      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2643        {        {
2644        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2645        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2646        }        }
2647      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2648      if (UCD_CATEGORY(c) == ucp_M) MRRETURN(MATCH_NOMATCH);      if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);
2649      while (eptr < md->end_subject)      while (eptr < md->end_subject)
2650        {        {
2651        int len = 1;        int len = 1;
2652        if (!utf8) c = *eptr; else { GETCHARLEN(c, eptr, len); }        if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
2653        if (UCD_CATEGORY(c) != ucp_M) break;        if (UCD_CATEGORY(c) != ucp_M) break;
2654        eptr += len;        eptr += len;
2655        }        }
2656        CHECK_PARTIAL();
2657      ecode++;      ecode++;
2658      break;      break;
2659  #endif  #endif
# Line 2398  for (;;) Line 2671  for (;;)
2671      case OP_REFI:      case OP_REFI:
2672      caseless = op == OP_REFI;      caseless = op == OP_REFI;
2673      offset = GET2(ecode, 1) << 1;               /* Doubled ref number */      offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
2674      ecode += 3;      ecode += 1 + IMM2_SIZE;
2675    
2676      /* If the reference is unset, there are two possibilities:      /* If the reference is unset, there are two possibilities:
2677    
# Line 2438  for (;;) Line 2711  for (;;)
2711        case OP_CRMINRANGE:        case OP_CRMINRANGE:
2712        minimize = (*ecode == OP_CRMINRANGE);        minimize = (*ecode == OP_CRMINRANGE);
2713        min = GET2(ecode, 1);        min = GET2(ecode, 1);
2714        max = GET2(ecode, 3);        max = GET2(ecode, 1 + IMM2_SIZE);
2715        if (max == 0) max = INT_MAX;        if (max == 0) max = INT_MAX;
2716        ecode += 5;        ecode += 1 + 2 * IMM2_SIZE;
2717        break;        break;
2718    
2719        default:               /* No repeat follows */        default:               /* No repeat follows */
2720        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)
2721          {          {
2722            if (length == -2) eptr = md->end_subject;   /* Partial match */
2723          CHECK_PARTIAL();          CHECK_PARTIAL();
2724          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2725          }          }
2726        eptr += length;        eptr += length;
2727        continue;              /* With the main loop */        continue;              /* With the main loop */
2728        }        }
2729    
2730      /* Handle repeated back references. If the length of the reference is      /* Handle repeated back references. If the length of the reference is
2731      zero, just continue with the main loop. */      zero, just continue with the main loop. If the length is negative, it
2732        means the reference is unset in non-Java-compatible mode. If the minimum is
2733        zero, we can continue at the same level without recursion. For any other
2734        minimum, carrying on will result in NOMATCH. */
2735    
2736      if (length == 0) continue;      if (length == 0) continue;
2737        if (length < 0 && min == 0) continue;
2738    
2739      /* First, ensure the minimum number of matches are present. We get back      /* First, ensure the minimum number of matches are present. We get back
2740      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 2745  for (;;)
2745        int slength;        int slength;
2746        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2747          {          {
2748            if (slength == -2) eptr = md->end_subject;   /* Partial match */
2749          CHECK_PARTIAL();          CHECK_PARTIAL();
2750          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2751          }          }
2752        eptr += slength;        eptr += slength;
2753        }        }
# Line 2487  for (;;) Line 2766  for (;;)
2766          int slength;          int slength;
2767          RMATCH(eptr, ecode, offset_top, md, eptrb, RM14);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM14);
2768          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2769          if (fi >= max) MRRETURN(MATCH_NOMATCH);          if (fi >= max) RRETURN(MATCH_NOMATCH);
2770          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2771            {            {
2772              if (slength == -2) eptr = md->end_subject;   /* Partial match */
2773            CHECK_PARTIAL();            CHECK_PARTIAL();
2774            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2775            }            }
2776          eptr += slength;          eptr += slength;
2777          }          }
# Line 2508  for (;;) Line 2788  for (;;)
2788          int slength;          int slength;
2789          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2790            {            {
2791            CHECK_PARTIAL();            /* Can't use CHECK_PARTIAL because we don't want to update eptr in
2792              the soft partial matching case. */
2793    
2794              if (slength == -2 && md->partial != 0 &&
2795                  md->end_subject > md->start_used_ptr)
2796                {
2797                md->hitend = TRUE;
2798                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2799                }
2800            break;            break;
2801            }            }
2802          eptr += slength;          eptr += slength;
2803          }          }
2804    
2805        while (eptr >= pp)        while (eptr >= pp)
2806          {          {
2807          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);
2808          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2809          eptr -= length;          eptr -= length;
2810          }          }
2811        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2812        }        }
2813      /* Control never gets here */      /* Control never gets here */
2814    
# Line 2537  for (;;) Line 2826  for (;;)
2826      case OP_NCLASS:      case OP_NCLASS:
2827      case OP_CLASS:      case OP_CLASS:
2828        {        {
2829          /* The data variable is saved across frames, so the byte map needs to
2830          be stored there. */
2831    #define BYTE_MAP ((pcre_uint8 *)data)
2832        data = ecode + 1;                /* Save for matching */        data = ecode + 1;                /* Save for matching */
2833        ecode += 33;                     /* Advance past the item */        ecode += 1 + (32 / sizeof(pcre_uchar)); /* Advance past the item */
2834    
2835        switch (*ecode)        switch (*ecode)
2836          {          {
# Line 2559  for (;;) Line 2851  for (;;)
2851          case OP_CRMINRANGE:          case OP_CRMINRANGE:
2852          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
2853          min = GET2(ecode, 1);          min = GET2(ecode, 1);
2854          max = GET2(ecode, 3);          max = GET2(ecode, 1 + IMM2_SIZE);
2855          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
2856          ecode += 5;          ecode += 1 + 2 * IMM2_SIZE;
2857          break;          break;
2858    
2859          default:               /* No repeat follows */          default:               /* No repeat follows */
# Line 2571  for (;;) Line 2863  for (;;)
2863    
2864        /* First, ensure the minimum number of matches are present. */        /* First, ensure the minimum number of matches are present. */
2865    
2866  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2867        /* UTF-8 mode */        if (utf)
       if (utf8)  
2868          {          {
2869          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2870            {            {
2871            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
2872              {              {
2873              SCHECK_PARTIAL();              SCHECK_PARTIAL();
2874              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2875              }              }
2876            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
2877            if (c > 255)            if (c > 255)
2878              {              {
2879              if (op == OP_CLASS) MRRETURN(MATCH_NOMATCH);              if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2880              }              }
2881            else            else
2882              {              if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
             if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  
             }  
2883            }            }
2884          }          }
2885        else        else
2886  #endif  #endif
2887        /* Not UTF-8 mode */        /* Not UTF mode */
2888          {          {
2889          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2890            {            {
2891            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
2892              {              {
2893              SCHECK_PARTIAL();              SCHECK_PARTIAL();
2894              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2895              }              }
2896            c = *eptr++;            c = *eptr++;
2897            if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  #ifndef COMPILE_PCRE8
2898              if (c > 255)
2899                {
2900                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2901                }
2902              else
2903    #endif
2904                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2905            }            }
2906          }          }
2907    
# Line 2619  for (;;) Line 2915  for (;;)
2915    
2916        if (minimize)        if (minimize)
2917          {          {
2918  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2919          /* UTF-8 mode */          if (utf)
         if (utf8)  
2920            {            {
2921            for (fi = min;; fi++)            for (fi = min;; fi++)
2922              {              {
2923              RMATCH(eptr, ecode, offset_top, md, eptrb, RM16);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM16);
2924              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2925              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2926              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
2927                {                {
2928                SCHECK_PARTIAL();                SCHECK_PARTIAL();
2929                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2930                }                }
2931              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
2932              if (c > 255)              if (c > 255)
2933                {                {
2934                if (op == OP_CLASS) MRRETURN(MATCH_NOMATCH);                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2935                }                }
2936              else              else
2937                {                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
               if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  
               }  
2938              }              }
2939            }            }
2940          else          else
2941  #endif  #endif
2942          /* Not UTF-8 mode */          /* Not UTF mode */
2943            {            {
2944            for (fi = min;; fi++)            for (fi = min;; fi++)
2945              {              {
2946              RMATCH(eptr, ecode, offset_top, md, eptrb, RM17);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM17);
2947              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2948              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2949              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
2950                {                {
2951                SCHECK_PARTIAL();                SCHECK_PARTIAL();
2952                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2953                }                }
2954              c = *eptr++;              c = *eptr++;
2955              if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  #ifndef COMPILE_PCRE8
2956                if (c > 255)
2957                  {
2958                  if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2959                  }
2960                else
2961    #endif
2962                  if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2963              }              }
2964            }            }
2965          /* Control never gets here */          /* Control never gets here */
# Line 2671  for (;;) Line 2971  for (;;)
2971          {          {
2972          pp = eptr;          pp = eptr;
2973    
2974  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2975          /* UTF-8 mode */          if (utf)
         if (utf8)  
2976            {            {
2977            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2978              {              {
# Line 2689  for (;;) Line 2988  for (;;)
2988                if (op == OP_CLASS) break;                if (op == OP_CLASS) break;
2989                }                }
2990              else              else
2991                {                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
               if ((data[c/8] & (1 << (c&7))) == 0) break;  
               }  
2992              eptr += len;              eptr += len;
2993              }              }
2994            for (;;)            for (;;)
# Line 2704  for (;;) Line 3001  for (;;)
3001            }            }
3002          else          else
3003  #endif  #endif
3004            /* Not UTF-8 mode */            /* Not UTF mode */
3005            {            {
3006            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3007              {              {
# Line 2714  for (;;) Line 3011  for (;;)
3011                break;                break;
3012                }                }
3013              c = *eptr;              c = *eptr;
3014              if ((data[c/8] & (1 << (c&7))) == 0) break;  #ifndef COMPILE_PCRE8
3015                if (c > 255)
3016                  {
3017                  if (op == OP_CLASS) break;
3018                  }
3019                else
3020    #endif
3021                  if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
3022              eptr++;              eptr++;
3023              }              }
3024            while (eptr >= pp)            while (eptr >= pp)
# Line 2725  for (;;) Line 3029  for (;;)
3029              }              }
3030            }            }
3031    
3032          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3033          }          }
3034    #undef BYTE_MAP
3035        }        }
3036      /* Control never gets here */      /* Control never gets here */
3037    
# Line 2735  for (;;) Line 3040  for (;;)
3040      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
3041      mode, because Unicode properties are supported in non-UTF-8 mode. */      mode, because Unicode properties are supported in non-UTF-8 mode. */
3042    
3043  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
3044      case OP_XCLASS:      case OP_XCLASS:
3045        {        {
3046        data = ecode + 1 + LINK_SIZE;                /* Save for matching */        data = ecode + 1 + LINK_SIZE;                /* Save for matching */
# Line 2760  for (;;) Line 3065  for (;;)
3065          case OP_CRMINRANGE:          case OP_CRMINRANGE:
3066          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
3067          min = GET2(ecode, 1);          min = GET2(ecode, 1);
3068          max = GET2(ecode, 3);          max = GET2(ecode, 1 + IMM2_SIZE);
3069          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
3070          ecode += 5;          ecode += 1 + 2 * IMM2_SIZE;
3071          break;          break;
3072    
3073          default:               /* No repeat follows */          default:               /* No repeat follows */
# Line 2777  for (;;) Line 3082  for (;;)
3082          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3083            {            {
3084            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3085            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3086            }            }
3087          GETCHARINCTEST(c, eptr);          GETCHARINCTEST(c, eptr);
3088          if (!_pcre_xclass(c, data)) MRRETURN(MATCH_NOMATCH);          if (!PRIV(xclass)(c, data, utf)) RRETURN(MATCH_NOMATCH);
3089          }          }
3090    
3091        /* 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 3102  for (;;)
3102            {            {
3103            RMATCH(eptr, ecode, offset_top, md, eptrb, RM20);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM20);
3104            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3105            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3106            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3107              {              {
3108              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3109              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3110              }              }
3111            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3112            if (!_pcre_xclass(c, data)) MRRETURN(MATCH_NOMATCH);            if (!PRIV(xclass)(c, data, utf)) RRETURN(MATCH_NOMATCH);
3113            }            }
3114          /* Control never gets here */          /* Control never gets here */
3115          }          }
# Line 2822  for (;;) Line 3127  for (;;)
3127              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3128              break;              break;
3129              }              }
3130    #ifdef SUPPORT_UTF
3131            GETCHARLENTEST(c, eptr, len);            GETCHARLENTEST(c, eptr, len);
3132            if (!_pcre_xclass(c, data)) break;  #else
3133              c = *eptr;
3134    #endif
3135              if (!PRIV(xclass)(c, data, utf)) break;
3136            eptr += len;            eptr += len;
3137            }            }
3138          for(;;)          for(;;)
# Line 2831  for (;;) Line 3140  for (;;)
3140            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);
3141            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3142            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3143            if (utf8) BACKCHAR(eptr);  #ifdef SUPPORT_UTF
3144              if (utf) BACKCHAR(eptr);
3145    #endif
3146            }            }
3147          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3148          }          }
3149    
3150        /* Control never gets here */        /* Control never gets here */
# Line 2843  for (;;) Line 3154  for (;;)
3154      /* Match a single character, casefully */      /* Match a single character, casefully */
3155    
3156      case OP_CHAR:      case OP_CHAR:
3157  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3158      if (utf8)      if (utf)
3159        {        {
3160        length = 1;        length = 1;
3161        ecode++;        ecode++;
# Line 2852  for (;;) Line 3163  for (;;)
3163        if (length > md->end_subject - eptr)        if (length > md->end_subject - eptr)
3164          {          {
3165          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
3166          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3167          }          }
3168        while (length-- > 0) if (*ecode++ != *eptr++) MRRETURN(MATCH_NOMATCH);        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);
3169        }        }
3170      else      else
3171  #endif  #endif
3172        /* Not UTF mode */
     /* Non-UTF-8 mode */  
3173        {        {
3174        if (md->end_subject - eptr < 1)        if (md->end_subject - eptr < 1)
3175          {          {
3176          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */
3177          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3178          }          }
3179        if (ecode[1] != *eptr++) MRRETURN(MATCH_NOMATCH);        if (ecode[1] != *eptr++) RRETURN(MATCH_NOMATCH);
3180        ecode += 2;        ecode += 2;
3181        }        }
3182      break;      break;
3183    
3184      /* Match a single character, caselessly */      /* Match a single character, caselessly. If we are at the end of the
3185        subject, give up immediately. */
3186    
3187      case OP_CHARI:      case OP_CHARI:
3188  #ifdef SUPPORT_UTF8      if (eptr >= md->end_subject)
3189      if (utf8)        {
3190          SCHECK_PARTIAL();
3191          RRETURN(MATCH_NOMATCH);
3192          }
3193    
3194    #ifdef SUPPORT_UTF
3195        if (utf)
3196        {        {
3197        length = 1;        length = 1;
3198        ecode++;        ecode++;
3199        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
3200    
       if (length > md->end_subject - eptr)  
         {  
         CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */  
         MRRETURN(MATCH_NOMATCH);  
         }  
   
3201        /* 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
3202        can use the fast lookup table. */        we know that its other case must also be one byte long, so we can use the
3203          fast lookup table. We know that there is at least one byte left in the
3204          subject. */
3205    
3206        if (fc < 128)        if (fc < 128)
3207          {          {
3208          if (md->lcc[*ecode++] != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);          if (md->lcc[fc]
3209                != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);
3210            ecode++;
3211            eptr++;
3212          }          }
3213    
3214        /* Otherwise we must pick up the subject character */        /* Otherwise we must pick up the subject character. Note that we cannot
3215          use the value of "length" to check for sufficient bytes left, because the
3216          other case of the character may have more or fewer bytes.  */
3217    
3218        else        else
3219          {          {
# Line 2911  for (;;) Line 3229  for (;;)
3229  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3230            if (dc != UCD_OTHERCASE(fc))            if (dc != UCD_OTHERCASE(fc))
3231  #endif  #endif
3232              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3233            }            }
3234          }          }
3235        }        }
3236      else      else
3237  #endif   /* SUPPORT_UTF8 */  #endif   /* SUPPORT_UTF */
3238    
3239      /* Non-UTF-8 mode */      /* Not UTF mode */
3240        {        {
3241        if (md->end_subject - eptr < 1)        if (TABLE_GET(ecode[1], md->lcc, ecode[1])
3242          {            != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);
3243          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */        eptr++;
         MRRETURN(MATCH_NOMATCH);  
         }  
       if (md->lcc[ecode[1]] != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);  
3244        ecode += 2;        ecode += 2;
3245        }        }
3246      break;      break;
# Line 2935  for (;;) Line 3250  for (;;)
3250      case OP_EXACT:      case OP_EXACT:
3251      case OP_EXACTI:      case OP_EXACTI:
3252      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3253      ecode += 3;      ecode += 1 + IMM2_SIZE;
3254      goto REPEATCHAR;      goto REPEATCHAR;
3255    
3256      case OP_POSUPTO:      case OP_POSUPTO:
# Line 2950  for (;;) Line 3265  for (;;)
3265      min = 0;      min = 0;
3266      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3267      minimize = *ecode == OP_MINUPTO || *ecode == OP_MINUPTOI;      minimize = *ecode == OP_MINUPTO || *ecode == OP_MINUPTOI;
3268      ecode += 3;      ecode += 1 + IMM2_SIZE;
3269      goto REPEATCHAR;      goto REPEATCHAR;
3270    
3271      case OP_POSSTAR:      case OP_POSSTAR:
# Line 2998  for (;;) Line 3313  for (;;)
3313      /* Common code for all repeated single-character matches. */      /* Common code for all repeated single-character matches. */
3314    
3315      REPEATCHAR:      REPEATCHAR:
3316  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3317      if (utf8)      if (utf)
3318        {        {
3319        length = 1;        length = 1;
3320        charptr = ecode;        charptr = ecode;
# Line 3015  for (;;) Line 3330  for (;;)
3330          unsigned int othercase;          unsigned int othercase;
3331          if (op >= OP_STARI &&     /* Caseless */          if (op >= OP_STARI &&     /* Caseless */
3332              (othercase = UCD_OTHERCASE(fc)) != fc)              (othercase = UCD_OTHERCASE(fc)) != fc)
3333            oclength = _pcre_ord2utf8(othercase, occhars);            oclength = PRIV(ord2utf)(othercase, occhars);
3334          else oclength = 0;          else oclength = 0;
3335  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3336    
3337          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3338            {            {
3339            if (eptr <= md->end_subject - length &&            if (eptr <= md->end_subject - length &&
3340              memcmp(eptr, charptr, length) == 0) eptr += length;              memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3341  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3342            else if (oclength > 0 &&            else if (oclength > 0 &&
3343                     eptr <= md->end_subject - oclength &&                     eptr <= md->end_subject - oclength &&
3344                     memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                     memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3345  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3346            else            else
3347              {              {
3348              CHECK_PARTIAL();              CHECK_PARTIAL();
3349              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3350              }              }
3351            }            }
3352    
# Line 3043  for (;;) Line 3358  for (;;)
3358              {              {
3359              RMATCH(eptr, ecode, offset_top, md, eptrb, RM22);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM22);
3360              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3361              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3362              if (eptr <= md->end_subject - length &&              if (eptr <= md->end_subject - length &&
3363                memcmp(eptr, charptr, length) == 0) eptr += length;                memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3364  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3365              else if (oclength > 0 &&              else if (oclength > 0 &&
3366                       eptr <= md->end_subject - oclength &&                       eptr <= md->end_subject - oclength &&
3367                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                       memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3368  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3369              else              else
3370                {                {
3371                CHECK_PARTIAL();                CHECK_PARTIAL();
3372                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3373                }                }
3374              }              }
3375            /* Control never gets here */            /* Control never gets here */
# Line 3066  for (;;) Line 3381  for (;;)
3381            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3382              {              {
3383              if (eptr <= md->end_subject - length &&              if (eptr <= md->end_subject - length &&
3384                  memcmp(eptr, charptr, length) == 0) eptr += length;                  memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3385  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3386              else if (oclength > 0 &&              else if (oclength > 0 &&
3387                       eptr <= md->end_subject - oclength &&                       eptr <= md->end_subject - oclength &&
3388                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                       memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3389  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3390              else              else
3391                {                {
# Line 3085  for (;;) Line 3400  for (;;)
3400              {              {
3401              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);
3402              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3403              if (eptr == pp) { MRRETURN(MATCH_NOMATCH); }              if (eptr == pp) { RRETURN(MATCH_NOMATCH); }
3404  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3405              eptr--;              eptr--;
3406              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 3102  for (;;) Line 3417  for (;;)
3417        value of fc will always be < 128. */        value of fc will always be < 128. */
3418        }        }
3419      else      else
3420  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
3421          /* When not in UTF-8 mode, load a single-byte character. */
3422      /* When not in UTF-8 mode, load a single-byte character. */        fc = *ecode++;
   
     fc = *ecode++;  
3423    
3424      /* The value of fc at this point is always less than 256, though we may or      /* The value of fc at this point is always one character, though we may
3425      may not be in UTF-8 mode. The code is duplicated for the caseless and      or may not be in UTF mode. The code is duplicated for the caseless and
3426      caseful cases, for speed, since matching characters is likely to be quite      caseful cases, for speed, since matching characters is likely to be quite
3427      common. First, ensure the minimum number of matches are present. If min =      common. First, ensure the minimum number of matches are present. If min =
3428      max, continue at the same level without recursing. Otherwise, if      max, continue at the same level without recursing. Otherwise, if
# Line 3122  for (;;) Line 3435  for (;;)
3435    
3436      if (op >= OP_STARI)  /* Caseless */      if (op >= OP_STARI)  /* Caseless */
3437        {        {
3438        fc = md->lcc[fc];  #ifdef COMPILE_PCRE8
3439          /* fc must be < 128 if UTF is enabled. */
3440          foc = md->fcc[fc];
3441    #else
3442    #ifdef SUPPORT_UTF
3443    #ifdef SUPPORT_UCP
3444          if (utf && fc > 127)
3445            foc = UCD_OTHERCASE(fc);
3446    #else
3447          if (utf && fc > 127)
3448            foc = fc;
3449    #endif /* SUPPORT_UCP */
3450          else
3451    #endif /* SUPPORT_UTF */
3452            foc = TABLE_GET(fc, md->fcc, fc);
3453    #endif /* COMPILE_PCRE8 */
3454    
3455        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
3456          {          {
3457          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3458            {            {
3459            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3460            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3461            }            }
3462          if (fc != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);          if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);
3463            eptr++;
3464          }          }
3465        if (min == max) continue;        if (min == max) continue;
3466        if (minimize)        if (minimize)
# Line 3139  for (;;) Line 3469  for (;;)
3469            {            {
3470            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);
3471            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3472            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3473            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3474              {              {
3475              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3476              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3477              }              }
3478            if (fc != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);            if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);
3479              eptr++;
3480            }            }
3481          /* Control never gets here */          /* Control never gets here */
3482          }          }
# Line 3159  for (;;) Line 3490  for (;;)
3490              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3491              break;              break;
3492              }              }
3493            if (fc != md->lcc[*eptr]) break;            if (fc != *eptr && foc != *eptr) break;
3494            eptr++;            eptr++;
3495            }            }
3496    
# Line 3171  for (;;) Line 3502  for (;;)
3502            eptr--;            eptr--;
3503            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3504            }            }
3505          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3506          }          }
3507        /* Control never gets here */        /* Control never gets here */
3508        }        }
# Line 3185  for (;;) Line 3516  for (;;)
3516          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3517            {            {
3518            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3519            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3520            }            }
3521          if (fc != *eptr++) MRRETURN(MATCH_NOMATCH);          if (fc != *eptr++) RRETURN(MATCH_NOMATCH);
3522          }          }
3523    
3524        if (min == max) continue;        if (min == max) continue;
# Line 3198  for (;;) Line 3529  for (;;)
3529            {            {
3530            RMATCH(eptr, ecode, offset_top, md, eptrb, RM26);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM26);
3531            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3532            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3533            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3534              {              {
3535              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3536              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3537              }              }
3538            if (fc != *eptr++) MRRETURN(MATCH_NOMATCH);            if (fc != *eptr++) RRETURN(MATCH_NOMATCH);
3539            }            }
3540          /* Control never gets here */          /* Control never gets here */
3541          }          }
# Line 3229  for (;;) Line 3560  for (;;)
3560            eptr--;            eptr--;
3561            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3562            }            }
3563          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3564          }          }
3565        }        }
3566      /* Control never gets here */      /* Control never gets here */
# Line 3242  for (;;) Line 3573  for (;;)
3573      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
3574        {        {
3575        SCHECK_PARTIAL();        SCHECK_PARTIAL();
3576        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
3577        }        }
3578      ecode++;  #ifdef SUPPORT_UTF
3579      GETCHARINCTEST(c, eptr);      if (utf)
     if (op == OP_NOTI)         /* The caseless case */  
3580        {        {
3581  #ifdef SUPPORT_UTF8        register unsigned int ch, och;
3582        if (c < 256)  
3583  #endif        ecode++;
3584        c = md->lcc[c];        GETCHARINC(ch, ecode);
3585        if (md->lcc[*ecode++] == c) MRRETURN(MATCH_NOMATCH);        GETCHARINC(c, eptr);
3586    
3587          if (op == OP_NOT)
3588            {
3589            if (ch == c) RRETURN(MATCH_NOMATCH);
3590            }
3591          else
3592            {
3593    #ifdef SUPPORT_UCP
3594            if (ch > 127)
3595              och = UCD_OTHERCASE(ch);
3596    #else
3597            if (ch > 127)
3598              och = ch;
3599    #endif /* SUPPORT_UCP */
3600            else
3601              och = TABLE_GET(ch, md->fcc, ch);
3602            if (ch == c || och == c) RRETURN(MATCH_NOMATCH);
3603            }
3604        }        }
3605      else    /* Caseful */      else
3606    #endif
3607        {        {
3608        if (*ecode++ == c) MRRETURN(MATCH_NOMATCH);        register unsigned int ch = ecode[1];
3609          c = *eptr++;
3610          if (ch == c || (op == OP_NOTI && TABLE_GET(ch, md->fcc, ch) == c))
3611            RRETURN(MATCH_NOMATCH);
3612          ecode += 2;
3613        }        }
3614      break;      break;
3615    
# Line 3270  for (;;) Line 3623  for (;;)
3623      case OP_NOTEXACT:      case OP_NOTEXACT:
3624      case OP_NOTEXACTI:      case OP_NOTEXACTI:
3625      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3626      ecode += 3;      ecode += 1 + IMM2_SIZE;
3627      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3628    
3629      case OP_NOTUPTO:      case OP_NOTUPTO:
# Line 3280  for (;;) Line 3633  for (;;)
3633      min = 0;      min = 0;
3634      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3635      minimize = *ecode == OP_NOTMINUPTO || *ecode == OP_NOTMINUPTOI;      minimize = *ecode == OP_NOTMINUPTO || *ecode == OP_NOTMINUPTOI;
3636      ecode += 3;      ecode += 1 + IMM2_SIZE;
3637      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3638    
3639      case OP_NOTPOSSTAR:      case OP_NOTPOSSTAR:
# Line 3312  for (;;) Line 3665  for (;;)
3665      possessive = TRUE;      possessive = TRUE;
3666      min = 0;      min = 0;
3667      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3668      ecode += 3;      ecode += 1 + IMM2_SIZE;
3669      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3670    
3671      case OP_NOTSTAR:      case OP_NOTSTAR:
# Line 3336  for (;;) Line 3689  for (;;)
3689      /* Common code for all repeated single-byte matches. */      /* Common code for all repeated single-byte matches. */
3690    
3691      REPEATNOTCHAR:      REPEATNOTCHAR:
3692      fc = *ecode++;      GETCHARINCTEST(fc, ecode);
3693    
3694      /* The code is duplicated for the caseless and caseful cases, for speed,      /* The code is duplicated for the caseless and caseful cases, for speed,
3695      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 3704  for (;;)
3704    
3705      if (op >= OP_NOTSTARI)     /* Caseless */      if (op >= OP_NOTSTARI)     /* Caseless */
3706        {        {
3707        fc = md->lcc[fc];  #ifdef SUPPORT_UTF
3708    #ifdef SUPPORT_UCP
3709          if (utf && fc > 127)
3710            foc = UCD_OTHERCASE(fc);
3711    #else
3712          if (utf && fc > 127)
3713            foc = fc;
3714    #endif /* SUPPORT_UCP */
3715          else
3716    #endif /* SUPPORT_UTF */
3717            foc = TABLE_GET(fc, md->fcc, fc);
3718    
3719  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3720        /* UTF-8 mode */        if (utf)
       if (utf8)  
3721          {          {
3722          register unsigned int d;          register unsigned int d;
3723          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3363  for (;;) Line 3725  for (;;)
3725            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3726              {              {
3727              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3728              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3729              }              }
3730            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3731            if (d < 256) d = md->lcc[d];            if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
           if (fc == d) MRRETURN(MATCH_NOMATCH);  
3732            }            }
3733          }          }
3734        else        else
3735  #endif  #endif
3736          /* Not UTF mode */
       /* Not UTF-8 mode */  
3737          {          {
3738          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3739            {            {
3740            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3741              {              {
3742              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3743              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3744              }              }
3745            if (fc == md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);            if (fc == *eptr || foc == *eptr) RRETURN(MATCH_NOMATCH);
3746              eptr++;
3747            }            }
3748          }          }
3749    
# Line 3390  for (;;) Line 3751  for (;;)
3751    
3752        if (minimize)        if (minimize)
3753          {          {
3754  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3755          /* UTF-8 mode */          if (utf)
         if (utf8)  
3756            {            {
3757            register unsigned int d;            register unsigned int d;
3758            for (fi = min;; fi++)            for (fi = min;; fi++)
3759              {              {
3760              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);
3761              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3762              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3763              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3764                {                {
3765                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3766                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3767                }                }
3768              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3769              if (d < 256) d = md->lcc[d];              if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
             if (fc == d) MRRETURN(MATCH_NOMATCH);  
3770              }              }
3771            }            }
3772          else          else
3773  #endif  #endif
3774          /* Not UTF-8 mode */          /* Not UTF mode */
3775            {            {
3776            for (fi = min;; fi++)            for (fi = min;; fi++)
3777              {              {
3778              RMATCH(eptr, ecode, offset_top, md, eptrb, RM29);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM29);
3779              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3780              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3781              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3782                {                {
3783                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3784                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3785                }                }
3786              if (fc == md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);              if (fc == *eptr || foc == *eptr) RRETURN(MATCH_NOMATCH);
3787                eptr++;
3788              }              }
3789            }            }
3790          /* Control never gets here */          /* Control never gets here */
# Line 3436  for (;;) Line 3796  for (;;)
3796          {          {
3797          pp = eptr;          pp = eptr;
3798    
3799  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3800          /* UTF-8 mode */          if (utf)
         if (utf8)  
3801            {            {
3802            register unsigned int d;            register unsigned int d;
3803            for (i = min; i < max; i++)            for (i = min; i < max; i++)
# Line 3450  for (;;) Line 3809  for (;;)
3809                break;                break;
3810                }                }
3811              GETCHARLEN(d, eptr, len);              GETCHARLEN(d, eptr, len);
3812              if (d < 256) d = md->lcc[d];              if (fc == d || (unsigned int)foc == d) break;
             if (fc == d) break;  
3813              eptr += len;              eptr += len;
3814              }              }
3815          if (possessive) continue;            if (possessive) continue;
3816          for(;;)            for(;;)
3817              {              {
3818              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);
3819              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
# Line 3465  for (;;) Line 3823  for (;;)
3823            }            }
3824          else          else
3825  #endif  #endif
3826          /* Not UTF-8 mode */          /* Not UTF mode */
3827            {            {
3828            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3829              {              {
# Line 3474  for (;;) Line 3832  for (;;)
3832                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3833                break;                break;
3834                }                }
3835              if (fc == md->lcc[*eptr]) break;              if (fc == *eptr || foc == *eptr) break;
3836              eptr++;              eptr++;
3837              }              }
3838            if (possessive) continue;            if (possessive) continue;
# Line 3486  for (;;) Line 3844  for (;;)
3844              }              }
3845            }            }
3846    
3847          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3848          }          }
3849        /* Control never gets here */        /* Control never gets here */
3850        }        }
# Line 3495  for (;;) Line 3853  for (;;)
3853    
3854      else      else
3855        {        {
3856  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3857        /* UTF-8 mode */        if (utf)
       if (utf8)  
3858          {          {
3859          register unsigned int d;          register unsigned int d;
3860          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3505  for (;;) Line 3862  for (;;)
3862            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3863              {              {
3864              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3865              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3866              }              }
3867            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3868            if (fc == d) MRRETURN(MATCH_NOMATCH);            if (fc == d) RRETURN(MATCH_NOMATCH);
3869            }            }
3870          }          }
3871        else        else
3872  #endif  #endif
3873        /* Not UTF-8 mode */        /* Not UTF mode */
3874          {          {
3875          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3876            {            {
3877            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3878              {              {
3879              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3880              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3881              }              }
3882            if (fc == *eptr++) MRRETURN(MATCH_NOMATCH);            if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
3883            }            }
3884          }          }
3885    
# Line 3530  for (;;) Line 3887  for (;;)
3887    
3888        if (minimize)        if (minimize)
3889          {          {
3890  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3891          /* UTF-8 mode */          if (utf)
         if (utf8)  
3892            {            {
3893            register unsigned int d;            register unsigned int d;
3894            for (fi = min;; fi++)            for (fi = min;; fi++)
3895              {              {
3896              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);
3897              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3898              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3899              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3900                {                {
3901                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3902                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3903                }                }
3904              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3905              if (fc == d) MRRETURN(MATCH_NOMATCH);              if (fc == d) RRETURN(MATCH_NOMATCH);
3906              }              }
3907            }            }
3908          else          else
3909  #endif  #endif
3910          /* Not UTF-8 mode */          /* Not UTF mode */
3911            {            {
3912            for (fi = min;; fi++)            for (fi = min;; fi++)
3913              {              {
3914              RMATCH(eptr, ecode, offset_top, md, eptrb, RM33);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM33);
3915              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3916              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3917              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3918                {                {
3919                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3920                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3921                }                }
3922              if (fc == *eptr++) MRRETURN(MATCH_NOMATCH);              if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
3923              }              }
3924            }            }
3925          /* Control never gets here */          /* Control never gets here */
# Line 3575  for (;;) Line 3931  for (;;)
3931          {          {
3932          pp = eptr;          pp = eptr;
3933    
3934  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3935          /* UTF-8 mode */          if (utf)
         if (utf8)  
3936            {            {
3937            register unsigned int d;            register unsigned int d;
3938            for (i = min; i < max; i++)            for (i = min; i < max; i++)
# Line 3603  for (;;) Line 3958  for (;;)
3958            }            }
3959          else          else
3960  #endif  #endif
3961          /* Not UTF-8 mode */          /* Not UTF mode */
3962            {            {
3963            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3964              {              {
# Line 3624  for (;;) Line 3979  for (;;)
3979              }              }
3980            }            }
3981    
3982          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3983          }          }
3984        }        }
3985      /* Control never gets here */      /* Control never gets here */
# Line 3636  for (;;) Line 3991  for (;;)
3991      case OP_TYPEEXACT:      case OP_TYPEEXACT:
3992      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3993      minimize = TRUE;      minimize = TRUE;
3994      ecode += 3;      ecode += 1 + IMM2_SIZE;
3995      goto REPEATTYPE;      goto REPEATTYPE;
3996    
3997      case OP_TYPEUPTO:      case OP_TYPEUPTO:
# Line 3644  for (;;) Line 3999  for (;;)
3999      min = 0;      min = 0;
4000      max = GET2(ecode, 1);      max = GET2(ecode, 1);
4001      minimize = *ecode == OP_TYPEMINUPTO;      minimize = *ecode == OP_TYPEMINUPTO;
4002      ecode += 3;      ecode += 1 + IMM2_SIZE;
4003      goto REPEATTYPE;      goto REPEATTYPE;
4004    
4005      case OP_TYPEPOSSTAR:      case OP_TYPEPOSSTAR:
# Line 3672  for (;;) Line 4027  for (;;)
4027      possessive = TRUE;      possessive = TRUE;
4028      min = 0;      min = 0;
4029      max = GET2(ecode, 1);      max = GET2(ecode, 1);
4030      ecode += 3;      ecode += 1 + IMM2_SIZE;
4031      goto REPEATTYPE;      goto REPEATTYPE;
4032    
4033      case OP_TYPESTAR:      case OP_TYPESTAR:
# Line 3718  for (;;) Line 4073  for (;;)
4073          switch(prop_type)          switch(prop_type)
4074            {            {
4075            case PT_ANY:            case PT_ANY:
4076            if (prop_fail_result) MRRETURN(MATCH_NOMATCH);            if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4077            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4078              {              {
4079              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4080                {                {
4081                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4082                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4083                }                }
4084              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4085              }              }
# Line 3737  for (;;) Line 4092  for (;;)
4092              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4093                {                {
4094                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4095                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4096                }                }
4097              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4098              chartype = UCD_CHARTYPE(c);              chartype = UCD_CHARTYPE(c);
4099              if ((chartype == ucp_Lu ||              if ((chartype == ucp_Lu ||
4100                   chartype == ucp_Ll ||                   chartype == ucp_Ll ||
4101                   chartype == ucp_Lt) == prop_fail_result)                   chartype == ucp_Lt) == prop_fail_result)
4102                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4103              }              }
4104            break;            break;
4105    
# Line 3754  for (;;) Line 4109  for (;;)
4109              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4110                {                {
4111                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4112                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4113                }                }
4114              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4115              if ((UCD_CATEGORY(c) == prop_value) == prop_fail_result)              if ((UCD_CATEGORY(c) == prop_value) == prop_fail_result)
4116                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4117              }              }
4118            break;            break;
4119    
# Line 3768  for (;;) Line 4123  for (;;)
4123              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4124                {                {
4125                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4126                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4127                }                }
4128              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4129              if ((UCD_CHARTYPE(c) == prop_value) == prop_fail_result)              if ((UCD_CHARTYPE(c) == prop_value) == prop_fail_result)
4130                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4131              }              }
4132            break;            break;
4133    
# Line 3782  for (;;) Line 4137  for (;;)
4137              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4138                {                {
4139                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4140                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4141                }                }
4142              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4143              if ((UCD_SCRIPT(c) == prop_value) == prop_fail_result)              if ((UCD_SCRIPT(c) == prop_value) == prop_fail_result)
4144                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4145              }              }
4146            break;            break;
4147    
# Line 3797  for (;;) Line 4152  for (;;)
4152              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4153                {                {
4154                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4155                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4156                }                }
4157              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4158              category = UCD_CATEGORY(c);              category = UCD_CATEGORY(c);
4159              if ((category == ucp_L || category == ucp_N) == prop_fail_result)              if ((category == ucp_L || category == ucp_N) == prop_fail_result)
4160                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4161              }              }
4162            break;            break;
4163    
# Line 3812  for (;;) Line 4167  for (;;)
4167              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4168                {                {
4169                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4170                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4171                }                }
4172              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4173              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||
4174                   c == CHAR_FF || c == CHAR_CR)                   c == CHAR_FF || c == CHAR_CR)
4175                     == prop_fail_result)                     == prop_fail_result)
4176                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4177              }              }
4178            break;            break;
4179    
# Line 3828  for (;;) Line 4183  for (;;)
4183              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4184                {                {
4185                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4186                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4187                }                }
4188              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4189              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||
4190                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)
4191                     == prop_fail_result)                     == prop_fail_result)
4192                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4193              }              }
4194            break;            break;
4195    
# Line 3845  for (;;) Line 4200  for (;;)
4200              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4201                {                {
4202                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4203                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4204                }                }
4205              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4206              category = UCD_CATEGORY(c);              category = UCD_CATEGORY(c);
4207              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)
4208                     == prop_fail_result)                     == prop_fail_result)
4209                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4210              }              }
4211            break;            break;
4212    
# Line 3872  for (;;) Line 4227  for (;;)
4227            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4228              {              {
4229              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4230              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4231              }              }
4232            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
4233            if (UCD_CATEGORY(c) == ucp_M) MRRETURN(MATCH_NOMATCH);            if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);
4234            while (eptr < md->end_subject)            while (eptr < md->end_subject)
4235              {              {
4236              int len = 1;              int len = 1;
4237              if (!utf8) c = *eptr; else { GETCHARLEN(c, eptr, len); }              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
4238              if (UCD_CATEGORY(c) != ucp_M) break;              if (UCD_CATEGORY(c) != ucp_M) break;
4239              eptr += len;              eptr += len;
4240              }              }
4241              CHECK_PARTIAL();
4242            }            }
4243          }          }
4244    
# Line 3891  for (;;) Line 4247  for (;;)
4247    
4248  /* Handle all other cases when the coding is UTF-8 */  /* Handle all other cases when the coding is UTF-8 */
4249    
4250  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4251        if (utf8) switch(ctype)        if (utf) switch(ctype)
4252          {          {
4253          case OP_ANY:          case OP_ANY:
4254          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3900  for (;;) Line 4256  for (;;)
4256            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4257              {              {
4258              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4259              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4260                }
4261              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4262              if (md->partial != 0 &&
4263                  eptr + 1 >= md->end_subject &&
4264                  NLBLOCK->nltype == NLTYPE_FIXED &&
4265                  NLBLOCK->nllen == 2 &&
4266                  *eptr == NLBLOCK->nl[0])
4267                {
4268                md->hitend = TRUE;
4269                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4270              }              }
           if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);  
4271            eptr++;            eptr++;
4272            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4273            }            }
4274          break;          break;
4275    
# Line 3914  for (;;) Line 4279  for (;;)
4279            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4280              {              {
4281              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4282              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4283              }              }
4284            eptr++;            eptr++;
4285            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4286            }            }
4287          break;          break;
4288    
4289          case OP_ANYBYTE:          case OP_ANYBYTE:
4290          if (eptr > md->end_subject - min) MRRETURN(MATCH_NOMATCH);          if (eptr > md->end_subject - min) RRETURN(MATCH_NOMATCH);
4291          eptr += min;          eptr += min;
4292          break;          break;
4293    
# Line 3932  for (;;) Line 4297  for (;;)
4297            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4298              {              {
4299              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4300              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4301              }              }
4302            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4303            switch(c)            switch(c)
4304              {              {
4305              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4306    
4307              case 0x000d:              case 0x000d:
4308              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
# Line 3951  for (;;) Line 4316  for (;;)
4316              case 0x0085:              case 0x0085:
4317              case 0x2028:              case 0x2028:
4318              case 0x2029:              case 0x2029:
4319              if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4320              break;              break;
4321              }              }
4322            }            }
# Line 3963  for (;;) Line 4328  for (;;)
4328            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4329              {              {
4330              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4331              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4332              }              }
4333            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4334            switch(c)            switch(c)
# Line 3988  for (;;) Line 4353  for (;;)
4353              case 0x202f:    /* NARROW NO-BREAK SPACE */              case 0x202f:    /* NARROW NO-BREAK SPACE */
4354              case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */              case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
4355              case 0x3000:    /* IDEOGRAPHIC SPACE */              case 0x3000:    /* IDEOGRAPHIC SPACE */
4356              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4357              }              }
4358            }            }
4359          break;          break;
# Line 3999  for (;;) Line 4364  for (;;)
4364            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4365              {              {
4366              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4367              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4368              }              }
4369            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4370            switch(c)            switch(c)
4371              {              {
4372              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4373              case 0x09:      /* HT */              case 0x09:      /* HT */
4374              case 0x20:      /* SPACE */              case 0x20:      /* SPACE */
4375              case 0xa0:      /* NBSP */              case 0xa0:      /* NBSP */
# Line 4035  for (;;) Line 4400  for (;;)
4400            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4401              {              {
4402              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4403              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4404              }              }
4405            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4406            switch(c)            switch(c)
# Line 4048  for (;;) Line 4413  for (;;)
4413              case 0x85:      /* NEL */              case 0x85:      /* NEL */
4414              case 0x2028:    /* LINE SEPARATOR */              case 0x2028:    /* LINE SEPARATOR */
4415              case 0x2029:    /* PARAGRAPH SEPARATOR */              case 0x2029:    /* PARAGRAPH SEPARATOR */
4416              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4417              }              }
4418            }            }
4419          break;          break;
# Line 4059  for (;;) Line 4424  for (;;)
4424            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4425              {              {
4426              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4427              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4428              }              }
4429            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4430            switch(c)            switch(c)
4431              {              {
4432              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4433              case 0x0a:      /* LF */              case 0x0a:      /* LF */
4434              case 0x0b:      /* VT */              case 0x0b:      /* VT */
4435              case 0x0c:      /* FF */              case 0x0c:      /* FF */
# Line 4083  for (;;) Line 4448  for (;;)
4448            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4449              {              {
4450              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4451              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4452              }              }
4453            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4454            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)
4455              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4456            }            }
4457          break;          break;
4458    
# Line 4097  for (;;) Line 4462  for (;;)
4462            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4463              {              {
4464              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4465              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4466              }              }
4467            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_digit) == 0)            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_digit) == 0)
4468              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4469              eptr++;
4470            /* 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 */
4471            }            }
4472          break;          break;
# Line 4111  for (;;) Line 4477  for (;;)
4477            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4478              {              {
4479              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4480              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4481              }              }
4482            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)
4483              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4484            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);            eptr++;
4485              ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4486            }            }
4487          break;          break;
4488    
# Line 4125  for (;;) Line 4492  for (;;)
4492            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4493              {              {
4494              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4495              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4496              }              }
4497            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_space) == 0)            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_space) == 0)
4498              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4499              eptr++;
4500            /* 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 */
4501            }            }
4502          break;          break;
# Line 4139  for (;;) Line 4507  for (;;)
4507            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4508              {              {
4509              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4510              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4511              }              }
4512            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)
4513              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4514            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);            eptr++;
4515              ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4516            }            }
4517          break;          break;
4518    
# Line 4153  for (;;) Line 4522  for (;;)
4522            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4523              {              {
4524              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4525              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4526              }              }
4527            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_word) == 0)            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_word) == 0)
4528              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4529              eptr++;
4530            /* 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 */
4531            }            }
4532          break;          break;
# Line 4166  for (;;) Line 4536  for (;;)
4536          }  /* End switch(ctype) */          }  /* End switch(ctype) */
4537    
4538        else        else
4539  #endif     /* SUPPORT_UTF8 */  #endif     /* SUPPORT_UTF */
4540    
4541        /* 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
4542        than OP_PROP and OP_NOTPROP. */        than OP_PROP and OP_NOTPROP. */
# Line 4179  for (;;) Line 4549  for (;;)
4549            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4550              {              {
4551              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4552              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4553                }
4554              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4555              if (md->partial != 0 &&
4556                  eptr + 1 >= md->end_subject &&
4557                  NLBLOCK->nltype == NLTYPE_FIXED &&
4558                  NLBLOCK->nllen == 2 &&
4559                  *eptr == NLBLOCK->nl[0])
4560                {
4561                md->hitend = TRUE;
4562                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4563              }              }
           if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);  
4564            eptr++;            eptr++;
4565            }            }
4566          break;          break;
# Line 4190  for (;;) Line 4569  for (;;)
4569          if (eptr > md->end_subject - min)          if (eptr > md->end_subject - min)
4570            {            {
4571            SCHECK_PARTIAL();            SCHECK_PARTIAL();
4572            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
4573            }            }
4574          eptr += min;          eptr += min;
4575          break;          break;
# Line 4199  for (;;) Line 4578  for (;;)
4578          if (eptr > md->end_subject - min)          if (eptr > md->end_subject - min)
4579            {            {
4580            SCHECK_PARTIAL();            SCHECK_PARTIAL();
4581            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
4582            }            }
4583          eptr += min;          eptr += min;
4584          break;          break;
# Line 4210  for (;;) Line 4589  for (;;)
4589            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4590              {              {
4591              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4592              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4593              }              }
4594            switch(*eptr++)            switch(*eptr++)
4595              {              {
4596              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4597    
4598              case 0x000d:              case 0x000d:
4599              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
# Line 4226  for (;;) Line 4605  for (;;)
4605              case 0x000b:              case 0x000b:
4606              case 0x000c:              case 0x000c:
4607              case 0x0085:              case 0x0085:
4608              if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);  #ifdef COMPILE_PCRE16
4609                case 0x2028:
4610                case 0x2029:
4611    #endif
4612                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4613              break;              break;
4614              }              }
4615            }            }
# Line 4238  for (;;) Line 4621  for (;;)
4621            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4622              {              {
4623              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4624              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4625              }              }
4626            switch(*eptr++)            switch(*eptr++)
4627              {              {
# Line 4246  for (;;) Line 4629  for (;;)
4629              case 0x09:      /* HT */              case 0x09:      /* HT */
4630              case 0x20:      /* SPACE */              case 0x20:      /* SPACE */
4631              case 0xa0:      /* NBSP */              case 0xa0:      /* NBSP */
4632              MRRETURN(MATCH_NOMATCH);  #ifdef COMPILE_PCRE16
4633                case 0x1680:    /* OGHAM SPACE MARK */
4634                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
4635                case 0x2000:    /* EN QUAD */
4636                case 0x2001:    /* EM QUAD */
4637                case 0x2002:    /* EN SPACE */
4638                case 0x2003:    /* EM SPACE */
4639                case 0x2004:    /* THREE-PER-EM SPACE */
4640                case 0x2005:    /* FOUR-PER-EM SPACE */
4641                case 0x2006:    /* SIX-PER-EM SPACE */
4642                case 0x2007:    /* FIGURE SPACE */
4643                case 0x2008:    /* PUNCTUATION SPACE */
4644                case 0x2009:    /* THIN SPACE */
4645                case 0x200A:    /* HAIR SPACE */
4646                case 0x202f:    /* NARROW NO-BREAK SPACE */
4647                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
4648                case 0x3000:    /* IDEOGRAPHIC SPACE */
4649    #endif
4650                RRETURN(MATCH_NOMATCH);
4651              }              }
4652            }            }
4653          break;          break;
# Line 4257  for (;;) Line 4658  for (;;)
4658            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4659              {              {
4660              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4661              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4662              }              }
4663            switch(*eptr++)            switch(*eptr++)
4664              {              {
4665              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4666              case 0x09:      /* HT */              case 0x09:      /* HT */
4667              case 0x20:      /* SPACE */              case 0x20:      /* SPACE */
4668              case 0xa0:      /* NBSP */              case 0xa0:      /* NBSP */
4669    #ifdef COMPILE_PCRE16
4670                case 0x1680:    /* OGHAM SPACE MARK */
4671                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
4672                case 0x2000:    /* EN QUAD */
4673                case 0x2001:    /* EM QUAD */
4674                case 0x2002:    /* EN SPACE */
4675                case 0x2003:    /* EM SPACE */
4676                case 0x2004:    /* THREE-PER-EM SPACE */
4677                case 0x2005:    /* FOUR-PER-EM SPACE */
4678                case 0x2006:    /* SIX-PER-EM SPACE */
4679                case 0x2007:    /* FIGURE SPACE */
4680                case 0x2008:    /* PUNCTUATION SPACE */
4681                case 0x2009:    /* THIN SPACE */
4682                case 0x200A:    /* HAIR SPACE */
4683                case 0x202f:    /* NARROW NO-BREAK SPACE */
4684                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
4685                case 0x3000:    /* IDEOGRAPHIC SPACE */
4686    #endif
4687              break;              break;
4688              }              }
4689            }            }
# Line 4276  for (;;) Line 4695  for (;;)
4695            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4696              {              {
4697              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4698              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4699              }              }
4700            switch(*eptr++)            switch(*eptr++)
4701              {              {
# Line 4286  for (;;) Line 4705  for (;;)
4705              case 0x0c:      /* FF */              case 0x0c:      /* FF */
4706              case 0x0d:      /* CR */              case 0x0d:      /* CR */
4707              case 0x85:      /* NEL */              case 0x85:      /* NEL */
4708              MRRETURN(MATCH_NOMATCH);  #ifdef COMPILE_PCRE16
4709                case 0x2028:    /* LINE SEPARATOR */
4710                case 0x2029:    /* PARAGRAPH SEPARATOR */
4711    #endif
4712                RRETURN(MATCH_NOMATCH);
4713              }              }
4714            }            }
4715          break;          break;
# Line 4297  for (;;) Line 4720  for (;;)
4720            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4721              {              {
4722              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4723              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4724              }              }
4725            switch(*eptr++)            switch(*eptr++)
4726              {              {
4727              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4728              case 0x0a:      /* LF */              case 0x0a:      /* LF */
4729              case 0x0b:      /* VT */              case 0x0b:      /* VT */
4730              case 0x0c:      /* FF */              case 0x0c:      /* FF */
4731              case 0x0d:      /* CR */              case 0x0d:      /* CR */
4732              case 0x85:      /* NEL */              case 0x85:      /* NEL */
4733    #ifdef COMPILE_PCRE16
4734                case 0x2028:    /* LINE SEPARATOR */
4735                case 0x2029:    /* PARAGRAPH SEPARATOR */
4736    #endif
4737              break;              break;
4738              }              }
4739            }            }
# Line 4318  for (;;) Line 4745  for (;;)
4745            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4746              {              {
4747              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4748              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4749              }              }
4750            if ((md->ctypes[*eptr++] & ctype_digit) != 0) MRRETURN(MATCH_NOMATCH);            if (MAX_255(*eptr) && (md->ctypes[*eptr] & ctype_digit) != 0)
4751                RRETURN(MATCH_NOMATCH);
4752              eptr++;
4753            }            }
4754          break;          break;
4755    
# Line 4330  for (;;) Line 4759  for (;;)
4759            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4760              {              {
4761              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4762              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4763              }              }
4764            if ((md->ctypes[*eptr++] & ctype_digit) == 0) MRRETURN(MATCH_NOMATCH);            if (!MAX_255(*eptr) || (md->ctypes[*eptr] & ctype_digit) == 0)
4765                RRETURN(MATCH_NOMATCH);
4766              eptr++;
4767            }            }