/[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 742 by zherczeg, Sun Nov 6 08:05:33 2011 UTC revision 1102 by chpe, Tue Oct 16 15:56:34 2012 UTC
# Line 6  Line 6 
6  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
7    
8                         Written by Philip Hazel                         Written by Philip Hazel
9             Copyright (c) 1997-2011 University of Cambridge             Copyright (c) 1997-2012 University of Cambridge
10    
11  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
12  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 37  POSSIBILITY OF SUCH DAMAGE. Line 37  POSSIBILITY OF SUCH DAMAGE.
37  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
38  */  */
39    
   
40  /* This module contains pcre_exec(), the externally visible function that does  /* This module contains pcre_exec(), the externally visible function that does
41  pattern matching using an NFA algorithm, trying to mimic Perl as closely as  pattern matching using an NFA algorithm, trying to mimic Perl as closely as
42  possible. There are also some static supporting functions. */  possible. There are also some static supporting functions. */
# Line 82  negative to avoid the external error cod Line 81  negative to avoid the external error cod
81  #define MATCH_SKIP_ARG     (-993)  #define MATCH_SKIP_ARG     (-993)
82  #define MATCH_THEN         (-992)  #define MATCH_THEN         (-992)
83    
 /* This is a convenience macro for code that occurs many times. */  
   
 #define MRRETURN(ra) \  
   { \  
   md->mark = markptr; \  
   RRETURN(ra); \  
   }  
   
84  /* Maximum number of ints of offset to save on the stack for recursive calls.  /* Maximum number of ints of offset to save on the stack for recursive calls.
85  If the offset vector is bigger, malloc is used. This should be a multiple of 3,  If the offset vector is bigger, malloc is used. This should be a multiple of 3,
86  because the offset vector is always a multiple of 3 long. */  because the offset vector is always a multiple of 3 long. */
# Line 101  because the offset vector is always a mu Line 92  because the offset vector is always a mu
92  static const char rep_min[] = { 0, 0, 1, 1, 0, 0 };  static const char rep_min[] = { 0, 0, 1, 1, 0, 0 };
93  static const char rep_max[] = { 0, 0, 0, 0, 1, 1 };  static const char rep_max[] = { 0, 0, 0, 0, 1, 1 };
94    
   
   
95  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
96  /*************************************************  /*************************************************
97  *        Debugging function to print chars       *  *        Debugging function to print chars       *
# Line 121  Returns:     nothing Line 110  Returns:     nothing
110  */  */
111    
112  static void  static void
113  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)
114  {  {
115  unsigned int c;  pcre_uint32 c;
116    BOOL utf = md->utf;
117  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;
118  while (length-- > 0)  while (length-- > 0)
119    if (isprint(c = *(p++))) printf("%c", c); else printf("\\x%02x", c);    if (isprint(c = RAWUCHARINCTEST(p))) printf("%c", (char)c); else printf("\\x{%02x}", c);
120  }  }
121  #endif  #endif
122    
# Line 148  Arguments: Line 138  Arguments:
138    md          points to match data block    md          points to match data block
139    caseless    TRUE if caseless    caseless    TRUE if caseless
140    
141  Returns:      < 0 if not matched, otherwise the number of subject bytes matched  Returns:      >= 0 the number of subject bytes matched
142                  -1 no match
143                  -2 partial match; always given if at end subject
144  */  */
145    
146  static int  static int
147  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,
148    BOOL caseless)    BOOL caseless)
149  {  {
150  USPTR eptr_start = eptr;  PCRE_PUCHAR eptr_start = eptr;
151  register USPTR p = md->start_subject + md->offset_vector[offset];  register PCRE_PUCHAR p = md->start_subject + md->offset_vector[offset];
152    BOOL utf = md->utf;
153    
154  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
155  if (eptr >= md->end_subject)  if (eptr >= md->end_subject)
# Line 171  pchars(p, length, FALSE, md); Line 164  pchars(p, length, FALSE, md);
164  printf("\n");  printf("\n");
165  #endif  #endif
166    
167  /* Always fail if reference not set (and not JavaScript compatible). */  /* Always fail if reference not set (and not JavaScript compatible - in that
168    case the length is passed as zero). */
169    
170  if (length < 0) return -1;  if (length < 0) return -1;
171    
# Line 181  ASCII characters. */ Line 175  ASCII characters. */
175    
176  if (caseless)  if (caseless)
177    {    {
178  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
179  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
180    if (md->utf8)    if (utf)
181      {      {
182      /* Match characters up to the end of the reference. NOTE: the number of      /* Match characters up to the end of the reference. NOTE: the number of
183      bytes matched may differ, because there are some characters whose upper and      data units matched may differ, because in UTF-8 there are some characters
184      lower case versions code as different numbers of bytes. For example, U+023A      whose upper and lower case versions code have different numbers of bytes.
185      (2 bytes in UTF-8) is the upper case version of U+2C65 (3 bytes in UTF-8);      For example, U+023A (2 bytes in UTF-8) is the upper case version of U+2C65
186      a sequence of 3 of the former uses 6 bytes, as does a sequence of two of      (3 bytes in UTF-8); a sequence of 3 of the former uses 6 bytes, as does a
187      the latter. It is important, therefore, to check the length along the      sequence of two of the latter. It is important, therefore, to check the
188      reference, not along the subject (earlier code did this wrong). */      length along the reference, not along the subject (earlier code did this
189        wrong). */
190    
191      USPTR endptr = p + length;      PCRE_PUCHAR endptr = p + length;
192      while (p < endptr)      while (p < endptr)
193        {        {
194        int c, d;        pcre_uint32 c, d;
195        if (eptr >= md->end_subject) return -1;        const ucd_record *ur;
196          if (eptr >= md->end_subject) return -2;   /* Partial match */
197        GETCHARINC(c, eptr);        GETCHARINC(c, eptr);
198        GETCHARINC(d, p);        GETCHARINC(d, p);
199        if (c != d && c != UCD_OTHERCASE(d)) return -1;        ur = GET_UCD(d);
200          if (c != d && c != d + ur->other_case)
201            {
202            const pcre_uint32 *pp = PRIV(ucd_caseless_sets) + ur->caseset;
203            for (;;)
204              {
205              if (c < *pp) return -1;
206              if (c == *pp++) break;
207              }
208            }
209        }        }
210      }      }
211    else    else
# Line 210  if (caseless) Line 215  if (caseless)
215    /* 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
216    is no UCP support. */    is no UCP support. */
217      {      {
     if (eptr + length > md->end_subject) return -1;  
218      while (length-- > 0)      while (length-- > 0)
219        { if (md->lcc[*p++] != md->lcc[*eptr++]) return -1; }        {
220          pcre_uchar cc, cp;
221          if (eptr >= md->end_subject) return -2;   /* Partial match */
222          cc = RAWUCHARTEST(eptr);
223          cp = RAWUCHARTEST(p);
224          if (TABLE_GET(cp, md->lcc, cp) != TABLE_GET(cc, md->lcc, cc)) return -1;
225          p++;
226          eptr++;
227          }
228      }      }
229    }    }
230    
# Line 221  are in UTF-8 mode. */ Line 233  are in UTF-8 mode. */
233    
234  else  else
235    {    {
236    if (eptr + length > md->end_subject) return -1;    while (length-- > 0)
237    while (length-- > 0) if (*p++ != *eptr++) return -1;      {
238        if (eptr >= md->end_subject) return -2;   /* Partial match */
239        if (RAWUCHARINCTEST(p) != RAWUCHARINCTEST(eptr)) return -1;
240        }
241    }    }
242    
243  return eptr - eptr_start;  return (int)(eptr - eptr_start);
244  }  }
245    
246    
# Line 290  actually used in this definition. */ Line 305  actually used in this definition. */
305  #define RMATCH(ra,rb,rc,rd,re,rw) \  #define RMATCH(ra,rb,rc,rd,re,rw) \
306    { \    { \
307    printf("match() called in line %d\n", __LINE__); \    printf("match() called in line %d\n", __LINE__); \
308    rrc = match(ra,rb,mstart,markptr,rc,rd,re,rdepth+1); \    rrc = match(ra,rb,mstart,rc,rd,re,rdepth+1); \
309    printf("to line %d\n", __LINE__); \    printf("to line %d\n", __LINE__); \
310    }    }
311  #define RRETURN(ra) \  #define RRETURN(ra) \
312    { \    { \
313    printf("match() returned %d from line %d ", ra, __LINE__); \    printf("match() returned %d from line %d\n", ra, __LINE__); \
314    return ra; \    return ra; \
315    }    }
316  #else  #else
317  #define RMATCH(ra,rb,rc,rd,re,rw) \  #define RMATCH(ra,rb,rc,rd,re,rw) \
318    rrc = match(ra,rb,mstart,markptr,rc,rd,re,rdepth+1)    rrc = match(ra,rb,mstart,rc,rd,re,rdepth+1)
319  #define RRETURN(ra) return ra  #define RRETURN(ra) return ra
320  #endif  #endif
321    
# Line 315  argument of match(), which never changes Line 330  argument of match(), which never changes
330    
331  #define RMATCH(ra,rb,rc,rd,re,rw)\  #define RMATCH(ra,rb,rc,rd,re,rw)\
332    {\    {\
333    heapframe *newframe = (heapframe *)(pcre_stack_malloc)(sizeof(heapframe));\    heapframe *newframe = frame->Xnextframe;\
334    if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\    if (newframe == NULL)\
335    frame->Xwhere = rw; \      {\
336        newframe = (heapframe *)(PUBL(stack_malloc))(sizeof(heapframe));\
337        if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\
338        newframe->Xnextframe = NULL;\
339        frame->Xnextframe = newframe;\
340        }\
341      frame->Xwhere = rw;\
342    newframe->Xeptr = ra;\    newframe->Xeptr = ra;\
343    newframe->Xecode = rb;\    newframe->Xecode = rb;\
344    newframe->Xmstart = mstart;\    newframe->Xmstart = mstart;\
   newframe->Xmarkptr = markptr;\  
345    newframe->Xoffset_top = rc;\    newframe->Xoffset_top = rc;\
346    newframe->Xeptrb = re;\    newframe->Xeptrb = re;\
347    newframe->Xrdepth = frame->Xrdepth + 1;\    newframe->Xrdepth = frame->Xrdepth + 1;\
# Line 337  argument of match(), which never changes Line 357  argument of match(), which never changes
357    {\    {\
358    heapframe *oldframe = frame;\    heapframe *oldframe = frame;\
359    frame = oldframe->Xprevframe;\    frame = oldframe->Xprevframe;\
   (pcre_stack_free)(oldframe);\  
360    if (frame != NULL)\    if (frame != NULL)\
361      {\      {\
362      rrc = ra;\      rrc = ra;\
# Line 351  argument of match(), which never changes Line 370  argument of match(), which never changes
370    
371  typedef struct heapframe {  typedef struct heapframe {
372    struct heapframe *Xprevframe;    struct heapframe *Xprevframe;
373      struct heapframe *Xnextframe;
374    
375    /* Function arguments that may change */    /* Function arguments that may change */
376    
377    USPTR Xeptr;    PCRE_PUCHAR Xeptr;
378    const uschar *Xecode;    const pcre_uchar *Xecode;
379    USPTR Xmstart;    PCRE_PUCHAR Xmstart;
   USPTR Xmarkptr;  
380    int Xoffset_top;    int Xoffset_top;
381    eptrblock *Xeptrb;    eptrblock *Xeptrb;
382    unsigned int Xrdepth;    unsigned int Xrdepth;
383    
384    /* Function local variables */    /* Function local variables */
385    
386    USPTR Xcallpat;    PCRE_PUCHAR Xcallpat;
387  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
388    USPTR Xcharptr;    PCRE_PUCHAR Xcharptr;
389  #endif  #endif
390    USPTR Xdata;    PCRE_PUCHAR Xdata;
391    USPTR Xnext;    PCRE_PUCHAR Xnext;
392    USPTR Xpp;    PCRE_PUCHAR Xpp;
393    USPTR Xprev;    PCRE_PUCHAR Xprev;
394    USPTR Xsaved_eptr;    PCRE_PUCHAR Xsaved_eptr;
395    
396    recursion_info Xnew_recursive;    recursion_info Xnew_recursive;
397    
# Line 385  typedef struct heapframe { Line 404  typedef struct heapframe {
404    int Xprop_value;    int Xprop_value;
405    int Xprop_fail_result;    int Xprop_fail_result;
406    int Xoclength;    int Xoclength;
407    uschar Xocchars[8];    pcre_uchar Xocchars[6];
408  #endif  #endif
409    
410    int Xcodelink;    int Xcodelink;
# Line 427  returns a negative (error) response, the Line 446  returns a negative (error) response, the
446  same response. */  same response. */
447    
448  /* 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
449  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
450  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.
451  something has been matched). For hard partial matching, we then return  something has been matched). For hard partial matching, we then return
452  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 457  the subject. */
457        eptr > md->start_used_ptr) \        eptr > md->start_used_ptr) \
458      { \      { \
459      md->hitend = TRUE; \      md->hitend = TRUE; \
460      if (md->partial > 1) MRRETURN(PCRE_ERROR_PARTIAL); \      if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL); \
461      }      }
462    
463  #define SCHECK_PARTIAL()\  #define SCHECK_PARTIAL()\
464    if (md->partial != 0 && eptr > md->start_used_ptr) \    if (md->partial != 0 && eptr > md->start_used_ptr) \
465      { \      { \
466      md->hitend = TRUE; \      md->hitend = TRUE; \
467      if (md->partial > 1) MRRETURN(PCRE_ERROR_PARTIAL); \      if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL); \
468      }      }
469    
470    
471  /* Performance note: It might be tempting to extract commonly used fields from  /* Performance note: It might be tempting to extract commonly used fields from
472  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
473  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
474  made performance worse.  made performance worse.
475    
# Line 459  Arguments: Line 478  Arguments:
478     ecode       pointer to current position in compiled code     ecode       pointer to current position in compiled code
479     mstart      pointer to the current match start position (can be modified     mstart      pointer to the current match start position (can be modified
480                   by encountering \K)                   by encountering \K)
    markptr     pointer to the most recent MARK name, or NULL  
481     offset_top  current top pointer     offset_top  current top pointer
482     md          pointer to "static" info for the match     md          pointer to "static" info for the match
483     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 492  Returns:       MATCH_MATCH if matched
492  */  */
493    
494  static int  static int
495  match(REGISTER USPTR eptr, REGISTER const uschar *ecode, USPTR mstart,  match(REGISTER PCRE_PUCHAR eptr, REGISTER const pcre_uchar *ecode,
496    const uschar *markptr, int offset_top, match_data *md, eptrblock *eptrb,    PCRE_PUCHAR mstart, int offset_top, match_data *md, eptrblock *eptrb,
497    unsigned int rdepth)    unsigned int rdepth)
498  {  {
499  /* 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 484  so they can be ordinary variables in all Line 502  so they can be ordinary variables in all
502    
503  register int  rrc;         /* Returns from recursive calls */  register int  rrc;         /* Returns from recursive calls */
504  register int  i;           /* Used for loops not involving calls to RMATCH() */  register int  i;           /* Used for loops not involving calls to RMATCH() */
505  register unsigned int c;   /* Character values not kept over RMATCH() calls */  register pcre_uint32 c;    /* Character values not kept over RMATCH() calls */
506  register BOOL utf8;        /* Local copy of UTF-8 flag for speed */  register BOOL utf;         /* Local copy of UTF flag for speed */
507    
508  BOOL minimize, possessive; /* Quantifier options */  BOOL minimize, possessive; /* Quantifier options */
509  BOOL caseless;  BOOL caseless;
510  int condcode;  int condcode;
511    
512  /* 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
513  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
514  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
515  heap whenever RMATCH() does a "recursion". See the macro definitions above. */  whenever RMATCH() does a "recursion". See the macro definitions above. Putting
516    the top-level on the stack rather than malloc-ing them all gives a performance
517    boost in many cases where there is not much "recursion". */
518    
519  #ifdef NO_RECURSE  #ifdef NO_RECURSE
520  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 */  
521    
522  /* Copy in the original argument variables */  /* Copy in the original argument variables */
523    
524  frame->Xeptr = eptr;  frame->Xeptr = eptr;
525  frame->Xecode = ecode;  frame->Xecode = ecode;
526  frame->Xmstart = mstart;  frame->Xmstart = mstart;
 frame->Xmarkptr = markptr;  
527  frame->Xoffset_top = offset_top;  frame->Xoffset_top = offset_top;
528  frame->Xeptrb = eptrb;  frame->Xeptrb = eptrb;
529  frame->Xrdepth = rdepth;  frame->Xrdepth = rdepth;
# Line 520  HEAP_RECURSE: Line 537  HEAP_RECURSE:
537  #define eptr               frame->Xeptr  #define eptr               frame->Xeptr
538  #define ecode              frame->Xecode  #define ecode              frame->Xecode
539  #define mstart             frame->Xmstart  #define mstart             frame->Xmstart
 #define markptr            frame->Xmarkptr  
540  #define offset_top         frame->Xoffset_top  #define offset_top         frame->Xoffset_top
541  #define eptrb              frame->Xeptrb  #define eptrb              frame->Xeptrb
542  #define rdepth             frame->Xrdepth  #define rdepth             frame->Xrdepth
543    
544  /* Ditto for the local variables */  /* Ditto for the local variables */
545    
546  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
547  #define charptr            frame->Xcharptr  #define charptr            frame->Xcharptr
548  #endif  #endif
549  #define callpat            frame->Xcallpat  #define callpat            frame->Xcallpat
# Line 585  declarations can be cut out in a block. Line 601  declarations can be cut out in a block.
601  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
602  to RMATCH(). */  to RMATCH(). */
603    
604  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
605  const uschar *charptr;  const pcre_uchar *charptr;
606  #endif  #endif
607  const uschar *callpat;  const pcre_uchar *callpat;
608  const uschar *data;  const pcre_uchar *data;
609  const uschar *next;  const pcre_uchar *next;
610  USPTR         pp;  PCRE_PUCHAR       pp;
611  const uschar *prev;  const pcre_uchar *prev;
612  USPTR         saved_eptr;  PCRE_PUCHAR       saved_eptr;
613    
614  recursion_info new_recursive;  recursion_info new_recursive;
615    
# Line 606  int prop_type; Line 622  int prop_type;
622  int prop_value;  int prop_value;
623  int prop_fail_result;  int prop_fail_result;
624  int oclength;  int oclength;
625  uschar occhars[8];  pcre_uchar occhars[6];
626  #endif  #endif
627    
628  int codelink;  int codelink;
# Line 616  int max; Line 632  int max;
632  int min;  int min;
633  int number;  int number;
634  int offset;  int offset;
635  int op;  pcre_uchar op;
636  int save_capture_last;  int save_capture_last;
637  int save_offset1, save_offset2, save_offset3;  int save_offset1, save_offset2, save_offset3;
638  int stacksave[REC_STACK_SAVE_MAX];  int stacksave[REC_STACK_SAVE_MAX];
639    
640  eptrblock newptrb;  eptrblock newptrb;
641    
642    /* There is a special fudge for calling match() in a way that causes it to
643    measure the size of its basic stack frame when the stack is being used for
644    recursion. The second argument (ecode) being NULL triggers this behaviour. It
645    cannot normally ever be NULL. The return is the negated value of the frame
646    size. */
647    
648    if (ecode == NULL)
649      {
650      if (rdepth == 0)
651        return match((PCRE_PUCHAR)&rdepth, NULL, NULL, 0, NULL, NULL, 1);
652      else
653        {
654        int len = (char *)&rdepth - (char *)eptr;
655        return (len > 0)? -len : len;
656        }
657      }
658  #endif     /* NO_RECURSE */  #endif     /* NO_RECURSE */
659    
660  /* 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 667  the alternative names that are used. */
667  #define code_offset   codelink  #define code_offset   codelink
668  #define condassert    condition  #define condassert    condition
669  #define matched_once  prev_is_word  #define matched_once  prev_is_word
670    #define foc           number
671    #define save_mark     data
672    
673  /* These statements are here to stop the compiler complaining about unitialized  /* These statements are here to stop the compiler complaining about unitialized
674  variables. */  variables. */
# Line 659  defined). However, RMATCH isn't like a f Line 694  defined). However, RMATCH isn't like a f
694  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,
695  however, impact performance when true recursion is being used. */  however, impact performance when true recursion is being used. */
696    
697  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
698  utf8 = md->utf8;       /* Local copy of the flag */  utf = md->utf;       /* Local copy of the flag */
699  #else  #else
700  utf8 = FALSE;  utf = FALSE;
701  #endif  #endif
702    
703  /* 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 736  for (;;)
736    switch(op)    switch(op)
737      {      {
738      case OP_MARK:      case OP_MARK:
739      markptr = ecode + 2;      md->nomatch_mark = ecode + 2;
740      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      md->mark = NULL;    /* In case previously set by assertion */
741        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
742        eptrb, RM55);        eptrb, RM55);
743        if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
744             md->mark == NULL) md->mark = ecode + 2;
745    
746      /* 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
747      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 750  for (;;)
750      position and return MATCH_SKIP. Otherwise, pass back the return code      position and return MATCH_SKIP. Otherwise, pass back the return code
751      unaltered. */      unaltered. */
752    
753      if (rrc == MATCH_SKIP_ARG &&      else if (rrc == MATCH_SKIP_ARG &&
754          strcmp((char *)markptr, (char *)(md->start_match_ptr)) == 0)          STRCMP_UC_UC_TEST(ecode + 2, md->start_match_ptr) == 0)
755        {        {
756        md->start_match_ptr = eptr;        md->start_match_ptr = eptr;
757        RRETURN(MATCH_SKIP);        RRETURN(MATCH_SKIP);
758        }        }
   
     if (md->mark == NULL) md->mark = markptr;  
759      RRETURN(rrc);      RRETURN(rrc);
760    
761      case OP_FAIL:      case OP_FAIL:
762      MRRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
763    
764      /* COMMIT overrides PRUNE, SKIP, and THEN */      /* COMMIT overrides PRUNE, SKIP, and THEN */
765    
766      case OP_COMMIT:      case OP_COMMIT:
767      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
768        eptrb, RM52);        eptrb, RM52);
769      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE &&      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE &&
770          rrc != MATCH_SKIP && rrc != MATCH_SKIP_ARG &&          rrc != MATCH_SKIP && rrc != MATCH_SKIP_ARG &&
771          rrc != MATCH_THEN)          rrc != MATCH_THEN)
772        RRETURN(rrc);        RRETURN(rrc);
773      MRRETURN(MATCH_COMMIT);      RRETURN(MATCH_COMMIT);
774    
775      /* PRUNE overrides THEN */      /* PRUNE overrides THEN */
776    
777      case OP_PRUNE:      case OP_PRUNE:
778      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
779        eptrb, RM51);        eptrb, RM51);
780      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
781      MRRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
782    
783      case OP_PRUNE_ARG:      case OP_PRUNE_ARG:
784      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      md->nomatch_mark = ecode + 2;
785        md->mark = NULL;    /* In case previously set by assertion */
786        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
787        eptrb, RM56);        eptrb, RM56);
788        if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
789             md->mark == NULL) md->mark = ecode + 2;
790      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
     md->mark = ecode + 2;  
791      RRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
792    
793      /* SKIP overrides PRUNE and THEN */      /* SKIP overrides PRUNE and THEN */
794    
795      case OP_SKIP:      case OP_SKIP:
796      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
797        eptrb, RM53);        eptrb, RM53);
798      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)
799        RRETURN(rrc);        RRETURN(rrc);
800      md->start_match_ptr = eptr;   /* Pass back current position */      md->start_match_ptr = eptr;   /* Pass back current position */
801      MRRETURN(MATCH_SKIP);      RRETURN(MATCH_SKIP);
802    
803        /* Note that, for Perl compatibility, SKIP with an argument does NOT set
804        nomatch_mark. There is a flag that disables this opcode when re-matching a
805        pattern that ended with a SKIP for which there was not a matching MARK. */
806    
807      case OP_SKIP_ARG:      case OP_SKIP_ARG:
808      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      if (md->ignore_skip_arg)
809          {
810          ecode += PRIV(OP_lengths)[*ecode] + ecode[1];
811          break;
812          }
813        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
814        eptrb, RM57);        eptrb, RM57);
815      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)
816        RRETURN(rrc);        RRETURN(rrc);
817    
818      /* 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
819      returning the special MATCH_SKIP_ARG return code. This will either be      returning the special MATCH_SKIP_ARG return code. This will either be
820      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
821      as PRUNE. */      with the md->ignore_skip_arg flag set. */
822    
823      md->start_match_ptr = ecode + 2;      md->start_match_ptr = ecode + 2;
824      RRETURN(MATCH_SKIP_ARG);      RRETURN(MATCH_SKIP_ARG);
# Line 780  for (;;) Line 828  for (;;)
828      match pointer to do this. */      match pointer to do this. */
829    
830      case OP_THEN:      case OP_THEN:
831      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
832        eptrb, RM54);        eptrb, RM54);
833      if (rrc != MATCH_NOMATCH) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
834      md->start_match_ptr = ecode;      md->start_match_ptr = ecode;
835      MRRETURN(MATCH_THEN);      RRETURN(MATCH_THEN);
836    
837      case OP_THEN_ARG:      case OP_THEN_ARG:
838      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top,      md->nomatch_mark = ecode + 2;
839        md->mark = NULL;    /* In case previously set by assertion */
840        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top,
841        md, eptrb, RM58);        md, eptrb, RM58);
842        if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
843             md->mark == NULL) md->mark = ecode + 2;
844      if (rrc != MATCH_NOMATCH) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
845      md->start_match_ptr = ecode;      md->start_match_ptr = ecode;
     md->mark = ecode + 2;  
846      RRETURN(MATCH_THEN);      RRETURN(MATCH_THEN);
847    
848      /* Handle an atomic group that does not contain any capturing parentheses.      /* Handle an atomic group that does not contain any capturing parentheses.
# Line 810  for (;;) Line 861  for (;;)
861      case OP_ONCE_NC:      case OP_ONCE_NC:
862      prev = ecode;      prev = ecode;
863      saved_eptr = eptr;      saved_eptr = eptr;
864        save_mark = md->mark;
865      do      do
866        {        {
867        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM64);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM64);
# Line 828  for (;;) Line 880  for (;;)
880    
881        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
882        ecode += GET(ecode,1);        ecode += GET(ecode,1);
883          md->mark = save_mark;
884        }        }
885      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
886    
# Line 867  for (;;) Line 920  for (;;)
920        }        }
921      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
922        {        {
       md->match_function_type = MATCH_CBEGROUP;  
923        RMATCH(eptr, prev, offset_top, md, eptrb, RM66);        RMATCH(eptr, prev, offset_top, md, eptrb, RM66);
924        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
925        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
# Line 907  for (;;) Line 959  for (;;)
959        save_offset2 = md->offset_vector[offset+1];        save_offset2 = md->offset_vector[offset+1];
960        save_offset3 = md->offset_vector[md->offset_end - number];        save_offset3 = md->offset_vector[md->offset_end - number];
961        save_capture_last = md->capture_last;        save_capture_last = md->capture_last;
962          save_mark = md->mark;
963    
964        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));
965        md->offset_vector[md->offset_end - number] =        md->offset_vector[md->offset_end - number] =
# Line 915  for (;;) Line 968  for (;;)
968        for (;;)        for (;;)
969          {          {
970          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
971          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
972            eptrb, RM1);            eptrb, RM1);
973          if (rrc == MATCH_ONCE) break;  /* Backing up through an atomic group */          if (rrc == MATCH_ONCE) break;  /* Backing up through an atomic group */
974    
# Line 943  for (;;) Line 996  for (;;)
996          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
997          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
998          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
999            md->mark = save_mark;
1000          if (*ecode != OP_ALT) break;          if (*ecode != OP_ALT) break;
1001          }          }
1002    
# Line 953  for (;;) Line 1007  for (;;)
1007    
1008        /* At this point, rrc will be one of MATCH_ONCE or MATCH_NOMATCH. */        /* At this point, rrc will be one of MATCH_ONCE or MATCH_NOMATCH. */
1009    
       if (md->mark == NULL) md->mark = markptr;  
1010        RRETURN(rrc);        RRETURN(rrc);
1011        }        }
1012    
# Line 995  for (;;) Line 1048  for (;;)
1048    
1049      for (;;)      for (;;)
1050        {        {
1051        if (op >= OP_SBRA || op == OP_ONCE) md->match_function_type = MATCH_CBEGROUP;        if (op >= OP_SBRA || op == OP_ONCE)
1052            md->match_function_type = MATCH_CBEGROUP;
1053    
1054        /* If this is not a possibly empty group, and there are no (*THEN)s in        /* If this is not a possibly empty group, and there are no (*THEN)s in
1055        the pattern, and this is the final alternative, optimize as described        the pattern, and this is the final alternative, optimize as described
# Line 1003  for (;;) Line 1057  for (;;)
1057    
1058        else if (!md->hasthen && ecode[GET(ecode, 1)] != OP_ALT)        else if (!md->hasthen && ecode[GET(ecode, 1)] != OP_ALT)
1059          {          {
1060          ecode += _pcre_OP_lengths[*ecode];          ecode += PRIV(OP_lengths)[*ecode];
1061          goto TAIL_RECURSE;          goto TAIL_RECURSE;
1062          }          }
1063    
1064        /* In all other cases, we have to make another call to match(). */        /* In all other cases, we have to make another call to match(). */
1065    
1066        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, eptrb,        save_mark = md->mark;
1067          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md, eptrb,
1068          RM2);          RM2);
1069    
1070        /* See comment in the code for capturing groups above about handling        /* See comment in the code for capturing groups above about handling
# Line 1027  for (;;) Line 1082  for (;;)
1082          {          {
1083          if (rrc == MATCH_ONCE)          if (rrc == MATCH_ONCE)
1084            {            {
1085            const uschar *scode = ecode;            const pcre_uchar *scode = ecode;
1086            if (*scode != OP_ONCE)           /* If not at start, find it */            if (*scode != OP_ONCE)           /* If not at start, find it */
1087              {              {
1088              while (*scode == OP_ALT) scode += GET(scode, 1);              while (*scode == OP_ALT) scode += GET(scode, 1);
# Line 1038  for (;;) Line 1093  for (;;)
1093          RRETURN(rrc);          RRETURN(rrc);
1094          }          }
1095        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1096          md->mark = save_mark;
1097        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1098        }        }
1099    
     if (md->mark == NULL) md->mark = markptr;  
1100      RRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
1101    
1102      /* Handle possessive capturing brackets with an unlimited repeat. We come      /* Handle possessive capturing brackets with an unlimited repeat. We come
# Line 1070  for (;;) Line 1125  for (;;)
1125      if (offset < md->offset_max)      if (offset < md->offset_max)
1126        {        {
1127        matched_once = FALSE;        matched_once = FALSE;
1128        code_offset = ecode - md->start_code;        code_offset = (int)(ecode - md->start_code);
1129    
1130        save_offset1 = md->offset_vector[offset];        save_offset1 = md->offset_vector[offset];
1131        save_offset2 = md->offset_vector[offset+1];        save_offset2 = md->offset_vector[offset+1];
# Line 1093  for (;;) Line 1148  for (;;)
1148          md->offset_vector[md->offset_end - number] =          md->offset_vector[md->offset_end - number] =
1149            (int)(eptr - md->start_subject);            (int)(eptr - md->start_subject);
1150          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
1151          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
1152            eptrb, RM63);            eptrb, RM63);
1153          if (rrc == MATCH_KETRPOS)          if (rrc == MATCH_KETRPOS)
1154            {            {
# Line 1129  for (;;) Line 1184  for (;;)
1184          md->offset_vector[md->offset_end - number] = save_offset3;          md->offset_vector[md->offset_end - number] = save_offset3;
1185          }          }
1186    
       if (md->mark == NULL) md->mark = markptr;  
1187        if (allow_zero || matched_once)        if (allow_zero || matched_once)
1188          {          {
1189          ecode += 1 + LINK_SIZE;          ecode += 1 + LINK_SIZE;
# Line 1161  for (;;) Line 1215  for (;;)
1215    
1216      POSSESSIVE_NON_CAPTURE:      POSSESSIVE_NON_CAPTURE:
1217      matched_once = FALSE;      matched_once = FALSE;
1218      code_offset = ecode - md->start_code;      code_offset = (int)(ecode - md->start_code);
1219    
1220      for (;;)      for (;;)
1221        {        {
1222        if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;        if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
1223        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
1224          eptrb, RM48);          eptrb, RM48);
1225        if (rrc == MATCH_KETRPOS)        if (rrc == MATCH_KETRPOS)
1226          {          {
# Line 1216  for (;;) Line 1270  for (;;)
1270    
1271      if (ecode[LINK_SIZE+1] == OP_CALLOUT)      if (ecode[LINK_SIZE+1] == OP_CALLOUT)
1272        {        {
1273        if (pcre_callout != NULL)        if (PUBL(callout) != NULL)
1274          {          {
1275          pcre_callout_block cb;          PUBL(callout_block) cb;
1276          cb.version          = 2;   /* Version 1 of the callout block */          cb.version          = 2;   /* Version 1 of the callout block */
1277          cb.callout_number   = ecode[LINK_SIZE+2];          cb.callout_number   = ecode[LINK_SIZE+2];
1278          cb.offset_vector    = md->offset_vector;          cb.offset_vector    = md->offset_vector;
1279    #if defined COMPILE_PCRE8
1280          cb.subject          = (PCRE_SPTR)md->start_subject;          cb.subject          = (PCRE_SPTR)md->start_subject;
1281    #elif defined COMPILE_PCRE16
1282            cb.subject          = (PCRE_SPTR16)md->start_subject;
1283    #elif defined COMPILE_PCRE32
1284            cb.subject          = (PCRE_SPTR32)md->start_subject;
1285    #endif
1286          cb.subject_length   = (int)(md->end_subject - md->start_subject);          cb.subject_length   = (int)(md->end_subject - md->start_subject);
1287          cb.start_match      = (int)(mstart - md->start_subject);          cb.start_match      = (int)(mstart - md->start_subject);
1288          cb.current_position = (int)(eptr - md->start_subject);          cb.current_position = (int)(eptr - md->start_subject);
# Line 1231  for (;;) Line 1291  for (;;)
1291          cb.capture_top      = offset_top/2;          cb.capture_top      = offset_top/2;
1292          cb.capture_last     = md->capture_last;          cb.capture_last     = md->capture_last;
1293          cb.callout_data     = md->callout_data;          cb.callout_data     = md->callout_data;
1294          cb.mark             = markptr;          cb.mark             = md->nomatch_mark;
1295          if ((rrc = (*pcre_callout)(&cb)) > 0) MRRETURN(MATCH_NOMATCH);          if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1296          if (rrc < 0) RRETURN(rrc);          if (rrc < 0) RRETURN(rrc);
1297          }          }
1298        ecode += _pcre_OP_lengths[OP_CALLOUT];        ecode += PRIV(OP_lengths)[OP_CALLOUT];
1299        }        }
1300    
1301      condcode = ecode[LINK_SIZE+1];      condcode = ecode[LINK_SIZE+1];
# Line 1252  for (;;) Line 1312  for (;;)
1312        else        else
1313          {          {
1314          int recno = GET2(ecode, LINK_SIZE + 2);   /* Recursion group number*/          int recno = GET2(ecode, LINK_SIZE + 2);   /* Recursion group number*/
1315          condition =  (recno == RREF_ANY || recno == md->recursive->group_num);          condition = (recno == RREF_ANY || recno == md->recursive->group_num);
1316    
1317          /* 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
1318          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
1319          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
1320          if any one is set. */          if any one is set. */
1321    
1322          if (!condition && condcode == OP_NRREF && recno != RREF_ANY)          if (!condition && condcode == OP_NRREF)
1323            {            {
1324            uschar *slotA = md->name_table;            pcre_uchar *slotA = md->name_table;
1325            for (i = 0; i < md->name_count; i++)            for (i = 0; i < md->name_count; i++)
1326              {              {
1327              if (GET2(slotA, 0) == recno) break;              if (GET2(slotA, 0) == recno) break;
# Line 1274  for (;;) Line 1334  for (;;)
1334    
1335            if (i < md->name_count)            if (i < md->name_count)
1336              {              {
1337              uschar *slotB = slotA;              pcre_uchar *slotB = slotA;
1338              while (slotB > md->name_table)              while (slotB > md->name_table)
1339                {                {
1340                slotB -= md->name_entry_size;                slotB -= md->name_entry_size;
1341                if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)                if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1342                  {                  {
1343                  condition = GET2(slotB, 0) == md->recursive->group_num;                  condition = GET2(slotB, 0) == md->recursive->group_num;
1344                  if (condition) break;                  if (condition) break;
# Line 1294  for (;;) Line 1354  for (;;)
1354                for (i++; i < md->name_count; i++)                for (i++; i < md->name_count; i++)
1355                  {                  {
1356                  slotB += md->name_entry_size;                  slotB += md->name_entry_size;
1357                  if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)                  if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1358                    {                    {
1359                    condition = GET2(slotB, 0) == md->recursive->group_num;                    condition = GET2(slotB, 0) == md->recursive->group_num;
1360                    if (condition) break;                    if (condition) break;
# Line 1307  for (;;) Line 1367  for (;;)
1367    
1368          /* Chose branch according to the condition */          /* Chose branch according to the condition */
1369    
1370          ecode += condition? 3 : GET(ecode, 1);          ecode += condition? 1 + IMM2_SIZE : GET(ecode, 1);
1371          }          }
1372        }        }
1373    
# Line 1324  for (;;) Line 1384  for (;;)
1384        if (!condition && condcode == OP_NCREF)        if (!condition && condcode == OP_NCREF)
1385          {          {
1386          int refno = offset >> 1;          int refno = offset >> 1;
1387          uschar *slotA = md->name_table;          pcre_uchar *slotA = md->name_table;
1388    
1389          for (i = 0; i < md->name_count; i++)          for (i = 0; i < md->name_count; i++)
1390            {            {
# Line 1338  for (;;) Line 1398  for (;;)
1398    
1399          if (i < md->name_count)          if (i < md->name_count)
1400            {            {
1401            uschar *slotB = slotA;            pcre_uchar *slotB = slotA;
1402            while (slotB > md->name_table)            while (slotB > md->name_table)
1403              {              {
1404              slotB -= md->name_entry_size;              slotB -= md->name_entry_size;
1405              if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)              if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1406                {                {
1407                offset = GET2(slotB, 0) << 1;                offset = GET2(slotB, 0) << 1;
1408                condition = offset < offset_top &&                condition = offset < offset_top &&
# Line 1360  for (;;) Line 1420  for (;;)
1420              for (i++; i < md->name_count; i++)              for (i++; i < md->name_count; i++)
1421                {                {
1422                slotB += md->name_entry_size;                slotB += md->name_entry_size;
1423                if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)                if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)
1424                  {                  {
1425                  offset = GET2(slotB, 0) << 1;                  offset = GET2(slotB, 0) << 1;
1426                  condition = offset < offset_top &&                  condition = offset < offset_top &&
# Line 1375  for (;;) Line 1435  for (;;)
1435    
1436        /* Chose branch according to the condition */        /* Chose branch according to the condition */
1437    
1438        ecode += condition? 3 : GET(ecode, 1);        ecode += condition? 1 + IMM2_SIZE : GET(ecode, 1);
1439        }        }
1440    
1441      else if (condcode == OP_DEF)     /* DEFINE - always false */      else if (condcode == OP_DEF)     /* DEFINE - always false */
# Line 1467  for (;;) Line 1527  for (;;)
1527        md->offset_vector[offset+1] = (int)(eptr - md->start_subject);        md->offset_vector[offset+1] = (int)(eptr - md->start_subject);
1528        if (offset_top <= offset) offset_top = offset + 2;        if (offset_top <= offset) offset_top = offset + 2;
1529        }        }
1530      ecode += 3;      ecode += 1 + IMM2_SIZE;
1531      break;      break;
1532    
1533    
# Line 1487  for (;;) Line 1547  for (;;)
1547           (md->notempty ||           (md->notempty ||
1548             (md->notempty_atstart &&             (md->notempty_atstart &&
1549               mstart == md->start_subject + md->start_offset)))               mstart == md->start_subject + md->start_offset)))
1550        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
1551    
1552      /* Otherwise, we have a match. */      /* Otherwise, we have a match. */
1553    
# Line 1496  for (;;) Line 1556  for (;;)
1556      md->start_match_ptr = mstart;       /* and the start (\K can modify) */      md->start_match_ptr = mstart;       /* and the start (\K can modify) */
1557    
1558      /* 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
1559      given as the argument to MRRETURN when the heap is in use. */      given as the argument to RRETURN when the heap is in use. */
1560    
1561      rrc = (op == OP_END)? MATCH_MATCH : MATCH_ACCEPT;      rrc = (op == OP_END)? MATCH_MATCH : MATCH_ACCEPT;
1562      MRRETURN(rrc);      RRETURN(rrc);
1563    
1564      /* Assertion brackets. Check the alternative branches in turn - the      /* Assertion brackets. Check the alternative branches in turn - the
1565      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 1514  for (;;) Line 1574  for (;;)
1574    
1575      case OP_ASSERT:      case OP_ASSERT:
1576      case OP_ASSERTBACK:      case OP_ASSERTBACK:
1577        save_mark = md->mark;
1578      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1579        {        {
1580        condassert = TRUE;        condassert = TRUE;
# Line 1527  for (;;) Line 1588  for (;;)
1588        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1589          {          {
1590          mstart = md->start_match_ptr;   /* In case \K reset it */          mstart = md->start_match_ptr;   /* In case \K reset it */
         markptr = md->mark;  
1591          break;          break;
1592          }          }
1593          md->mark = save_mark;
1594    
1595        /* PCRE does not allow THEN to escape beyond an assertion; it is treated        /* A COMMIT failure must fail the entire assertion, without trying any
1596        as NOMATCH. */        subsequent branches. */
1597    
1598          if (rrc == MATCH_COMMIT) RRETURN(MATCH_NOMATCH);
1599    
1600          /* PCRE does not allow THEN to escape beyond an assertion; it
1601          is treated as NOMATCH. */
1602    
1603        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1604        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1605        }        }
1606      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
1607    
1608      if (*ecode == OP_KET) MRRETURN(MATCH_NOMATCH);      if (*ecode == OP_KET) RRETURN(MATCH_NOMATCH);
1609    
1610      /* If checking an assertion for a condition, return MATCH_MATCH. */      /* If checking an assertion for a condition, return MATCH_MATCH. */
1611    
# Line 1559  for (;;) Line 1625  for (;;)
1625    
1626      case OP_ASSERT_NOT:      case OP_ASSERT_NOT:
1627      case OP_ASSERTBACK_NOT:      case OP_ASSERTBACK_NOT:
1628        save_mark = md->mark;
1629      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1630        {        {
1631        condassert = TRUE;        condassert = TRUE;
# Line 1569  for (;;) Line 1636  for (;;)
1636      do      do
1637        {        {
1638        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);
1639        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) MRRETURN(MATCH_NOMATCH);        md->mark = save_mark;
1640          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) RRETURN(MATCH_NOMATCH);
1641        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)
1642          {          {
1643          do ecode += GET(ecode,1); while (*ecode == OP_ALT);          do ecode += GET(ecode,1); while (*ecode == OP_ALT);
# Line 1595  for (;;) Line 1663  for (;;)
1663      back a number of characters, not bytes. */      back a number of characters, not bytes. */
1664    
1665      case OP_REVERSE:      case OP_REVERSE:
1666  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
1667      if (utf8)      if (utf)
1668        {        {
1669        i = GET(ecode, 1);        i = GET(ecode, 1);
1670        while (i-- > 0)        while (i-- > 0)
1671          {          {
1672          eptr--;          eptr--;
1673          if (eptr < md->start_subject) MRRETURN(MATCH_NOMATCH);          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1674          BACKCHAR(eptr);          BACKCHAR(eptr);
1675          }          }
1676        }        }
# Line 1613  for (;;) Line 1681  for (;;)
1681    
1682        {        {
1683        eptr -= GET(ecode, 1);        eptr -= GET(ecode, 1);
1684        if (eptr < md->start_subject) MRRETURN(MATCH_NOMATCH);        if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1685        }        }
1686    
1687      /* Save the earliest consulted character, then skip to next op code */      /* Save the earliest consulted character, then skip to next op code */
# Line 1627  for (;;) Line 1695  for (;;)
1695      function is able to force a failure. */      function is able to force a failure. */
1696    
1697      case OP_CALLOUT:      case OP_CALLOUT:
1698      if (pcre_callout != NULL)      if (PUBL(callout) != NULL)
1699        {        {
1700        pcre_callout_block cb;        PUBL(callout_block) cb;
1701        cb.version          = 2;   /* Version 1 of the callout block */        cb.version          = 2;   /* Version 1 of the callout block */
1702        cb.callout_number   = ecode[1];        cb.callout_number   = ecode[1];
1703        cb.offset_vector    = md->offset_vector;        cb.offset_vector    = md->offset_vector;
1704    #if defined COMPILE_PCRE8
1705        cb.subject          = (PCRE_SPTR)md->start_subject;        cb.subject          = (PCRE_SPTR)md->start_subject;
1706    #elif defined COMPILE_PCRE16
1707          cb.subject          = (PCRE_SPTR16)md->start_subject;
1708    #elif defined COMPILE_PCRE32
1709          cb.subject          = (PCRE_SPTR32)md->start_subject;
1710    #endif
1711        cb.subject_length   = (int)(md->end_subject - md->start_subject);        cb.subject_length   = (int)(md->end_subject - md->start_subject);
1712        cb.start_match      = (int)(mstart - md->start_subject);        cb.start_match      = (int)(mstart - md->start_subject);
1713        cb.current_position = (int)(eptr - md->start_subject);        cb.current_position = (int)(eptr - md->start_subject);
# Line 1642  for (;;) Line 1716  for (;;)
1716        cb.capture_top      = offset_top/2;        cb.capture_top      = offset_top/2;
1717        cb.capture_last     = md->capture_last;        cb.capture_last     = md->capture_last;
1718        cb.callout_data     = md->callout_data;        cb.callout_data     = md->callout_data;
1719        cb.mark             = markptr;        cb.mark             = md->nomatch_mark;
1720        if ((rrc = (*pcre_callout)(&cb)) > 0) MRRETURN(MATCH_NOMATCH);        if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1721        if (rrc < 0) RRETURN(rrc);        if (rrc < 0) RRETURN(rrc);
1722        }        }
1723      ecode += 2 + 2*LINK_SIZE;      ecode += 2 + 2*LINK_SIZE;
# Line 1702  for (;;) Line 1776  for (;;)
1776        else        else
1777          {          {
1778          new_recursive.offset_save =          new_recursive.offset_save =
1779            (int *)(pcre_malloc)(new_recursive.saved_max * sizeof(int));            (int *)(PUBL(malloc))(new_recursive.saved_max * sizeof(int));
1780          if (new_recursive.offset_save == NULL) RRETURN(PCRE_ERROR_NOMEMORY);          if (new_recursive.offset_save == NULL) RRETURN(PCRE_ERROR_NOMEMORY);
1781          }          }
1782        memcpy(new_recursive.offset_save, md->offset_vector,        memcpy(new_recursive.offset_save, md->offset_vector,
# Line 1717  for (;;) Line 1791  for (;;)
1791        do        do
1792          {          {
1793          if (cbegroup) md->match_function_type = MATCH_CBEGROUP;          if (cbegroup) md->match_function_type = MATCH_CBEGROUP;
1794          RMATCH(eptr, callpat + _pcre_OP_lengths[*callpat], offset_top,          RMATCH(eptr, callpat + PRIV(OP_lengths)[*callpat], offset_top,
1795            md, eptrb, RM6);            md, eptrb, RM6);
1796          memcpy(md->offset_vector, new_recursive.offset_save,          memcpy(md->offset_vector, new_recursive.offset_save,
1797              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
# Line 1726  for (;;) Line 1800  for (;;)
1800            {            {
1801            DPRINTF(("Recursion matched\n"));            DPRINTF(("Recursion matched\n"));
1802            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
1803              (pcre_free)(new_recursive.offset_save);              (PUBL(free))(new_recursive.offset_save);
1804    
1805            /* 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
1806            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 1737  for (;;) Line 1811  for (;;)
1811            goto RECURSION_MATCHED;        /* Exit loop; end processing */            goto RECURSION_MATCHED;        /* Exit loop; end processing */
1812            }            }
1813    
1814          /* PCRE does not allow THEN to escape beyond a recursion; it is treated          /* PCRE does not allow THEN or COMMIT to escape beyond a recursion; it
1815          as NOMATCH. */          is treated as NOMATCH. */
1816    
1817          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN &&
1818                     rrc != MATCH_COMMIT)
1819            {            {
1820            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1821            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
1822              (pcre_free)(new_recursive.offset_save);              (PUBL(free))(new_recursive.offset_save);
1823            RRETURN(rrc);            RRETURN(rrc);
1824            }            }
1825    
# Line 1756  for (;;) Line 1831  for (;;)
1831        DPRINTF(("Recursion didn't match\n"));        DPRINTF(("Recursion didn't match\n"));
1832        md->recursive = new_recursive.prevrec;        md->recursive = new_recursive.prevrec;
1833        if (new_recursive.offset_save != stacksave)        if (new_recursive.offset_save != stacksave)
1834          (pcre_free)(new_recursive.offset_save);          (PUBL(free))(new_recursive.offset_save);
1835        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
1836        }        }
1837    
1838      RECURSION_MATCHED:      RECURSION_MATCHED:
# Line 1837  for (;;) Line 1912  for (;;)
1912        md->end_match_ptr = eptr;      /* For ONCE_NC */        md->end_match_ptr = eptr;      /* For ONCE_NC */
1913        md->end_offset_top = offset_top;        md->end_offset_top = offset_top;
1914        md->start_match_ptr = mstart;        md->start_match_ptr = mstart;
1915        MRRETURN(MATCH_MATCH);         /* Sets md->mark */        RRETURN(MATCH_MATCH);         /* Sets md->mark */
1916        }        }
1917    
1918      /* 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 1951  for (;;) Line 2026  for (;;)
2026          }          }
2027        if (*prev >= OP_SBRA)    /* Could match an empty string */        if (*prev >= OP_SBRA)    /* Could match an empty string */
2028          {          {
         md->match_function_type = MATCH_CBEGROUP;  
2029          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);
2030          RRETURN(rrc);          RRETURN(rrc);
2031          }          }
# Line 1960  for (;;) Line 2034  for (;;)
2034        }        }
2035      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
2036        {        {
       if (*prev >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;  
2037        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);
2038        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;
2039        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
# Line 1979  for (;;) Line 2052  for (;;)
2052      /* Not multiline mode: start of subject assertion, unless notbol. */      /* Not multiline mode: start of subject assertion, unless notbol. */
2053    
2054      case OP_CIRC:      case OP_CIRC:
2055      if (md->notbol && eptr == md->start_subject) MRRETURN(MATCH_NOMATCH);      if (md->notbol && eptr == md->start_subject) RRETURN(MATCH_NOMATCH);
2056    
2057      /* Start of subject assertion */      /* Start of subject assertion */
2058    
2059      case OP_SOD:      case OP_SOD:
2060      if (eptr != md->start_subject) MRRETURN(MATCH_NOMATCH);      if (eptr != md->start_subject) RRETURN(MATCH_NOMATCH);
2061      ecode++;      ecode++;
2062      break;      break;
2063    
2064      /* Multiline mode: start of subject unless notbol, or after any newline. */      /* Multiline mode: start of subject unless notbol, or after any newline. */
2065    
2066      case OP_CIRCM:      case OP_CIRCM:
2067      if (md->notbol && eptr == md->start_subject) MRRETURN(MATCH_NOMATCH);      if (md->notbol && eptr == md->start_subject) RRETURN(MATCH_NOMATCH);
2068      if (eptr != md->start_subject &&      if (eptr != md->start_subject &&
2069          (eptr == md->end_subject || !WAS_NEWLINE(eptr)))          (eptr == md->end_subject || !WAS_NEWLINE(eptr)))
2070        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2071      ecode++;      ecode++;
2072      break;      break;
2073    
2074      /* Start of match assertion */      /* Start of match assertion */
2075    
2076      case OP_SOM:      case OP_SOM:
2077      if (eptr != md->start_subject + md->start_offset) MRRETURN(MATCH_NOMATCH);      if (eptr != md->start_subject + md->start_offset) RRETURN(MATCH_NOMATCH);
2078      ecode++;      ecode++;
2079      break;      break;
2080    
# Line 2017  for (;;) Line 2090  for (;;)
2090    
2091      case OP_DOLLM:      case OP_DOLLM:
2092      if (eptr < md->end_subject)      if (eptr < md->end_subject)
2093        { if (!IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH); }        {
2094          if (!IS_NEWLINE(eptr))
2095            {
2096            if (md->partial != 0 &&
2097                eptr + 1 >= md->end_subject &&
2098                NLBLOCK->nltype == NLTYPE_FIXED &&
2099                NLBLOCK->nllen == 2 &&
2100                RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2101              {
2102              md->hitend = TRUE;
2103              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2104              }
2105            RRETURN(MATCH_NOMATCH);
2106            }
2107          }
2108      else      else
2109        {        {
2110        if (md->noteol) MRRETURN(MATCH_NOMATCH);        if (md->noteol) RRETURN(MATCH_NOMATCH);
2111        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2112        }        }
2113      ecode++;      ecode++;
# Line 2030  for (;;) Line 2117  for (;;)
2117      subject unless noteol is set. */      subject unless noteol is set. */
2118    
2119      case OP_DOLL:      case OP_DOLL:
2120      if (md->noteol) MRRETURN(MATCH_NOMATCH);      if (md->noteol) RRETURN(MATCH_NOMATCH);
2121      if (!md->endonly) goto ASSERT_NL_OR_EOS;      if (!md->endonly) goto ASSERT_NL_OR_EOS;
2122    
2123      /* ... else fall through for endonly */      /* ... else fall through for endonly */
# Line 2038  for (;;) Line 2125  for (;;)
2125      /* End of subject assertion (\z) */      /* End of subject assertion (\z) */
2126    
2127      case OP_EOD:      case OP_EOD:
2128      if (eptr < md->end_subject) MRRETURN(MATCH_NOMATCH);      if (eptr < md->end_subject) RRETURN(MATCH_NOMATCH);
2129      SCHECK_PARTIAL();      SCHECK_PARTIAL();
2130      ecode++;      ecode++;
2131      break;      break;
# Line 2049  for (;;) Line 2136  for (;;)
2136      ASSERT_NL_OR_EOS:      ASSERT_NL_OR_EOS:
2137      if (eptr < md->end_subject &&      if (eptr < md->end_subject &&
2138          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))
2139        MRRETURN(MATCH_NOMATCH);        {
2140          if (md->partial != 0 &&
2141              eptr + 1 >= md->end_subject &&
2142              NLBLOCK->nltype == NLTYPE_FIXED &&
2143              NLBLOCK->nllen == 2 &&
2144              RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2145            {
2146            md->hitend = TRUE;
2147            if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2148            }
2149          RRETURN(MATCH_NOMATCH);
2150          }
2151    
2152      /* Either at end of string or \n before end. */      /* Either at end of string or \n before end. */
2153    
# Line 2068  for (;;) Line 2166  for (;;)
2166        be "non-word" characters. Remember the earliest consulted character for        be "non-word" characters. Remember the earliest consulted character for
2167        partial matching. */        partial matching. */
2168    
2169  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2170        if (utf8)        if (utf)
2171          {          {
2172          /* Get status of previous character */          /* Get status of previous character */
2173    
2174          if (eptr == md->start_subject) prev_is_word = FALSE; else          if (eptr == md->start_subject) prev_is_word = FALSE; else
2175            {            {
2176            USPTR lastptr = eptr - 1;            PCRE_PUCHAR lastptr = eptr - 1;
2177            while((*lastptr & 0xc0) == 0x80) lastptr--;            BACKCHAR(lastptr);
2178            if (lastptr < md->start_used_ptr) md->start_used_ptr = lastptr;            if (lastptr < md->start_used_ptr) md->start_used_ptr = lastptr;
2179            GETCHAR(c, lastptr);            GETCHAR(c, lastptr);
2180  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
# Line 2141  for (;;) Line 2239  for (;;)
2239              }              }
2240            else            else
2241  #endif  #endif
2242            prev_is_word = ((md->ctypes[eptr[-1]] & ctype_word) != 0);            prev_is_word = MAX_255(eptr[-1])
2243                && ((md->ctypes[eptr[-1]] & ctype_word) != 0);
2244            }            }
2245    
2246          /* Get status of next character */          /* Get status of next character */
# Line 2164  for (;;) Line 2263  for (;;)
2263            }            }
2264          else          else
2265  #endif  #endif
2266          cur_is_word = ((md->ctypes[*eptr] & ctype_word) != 0);          cur_is_word = MAX_255(*eptr)
2267              && ((md->ctypes[*eptr] & ctype_word) != 0);
2268          }          }
2269    
2270        /* Now see if the situation is what we want */        /* Now see if the situation is what we want */
2271    
2272        if ((*ecode++ == OP_WORD_BOUNDARY)?        if ((*ecode++ == OP_WORD_BOUNDARY)?
2273             cur_is_word == prev_is_word : cur_is_word != prev_is_word)             cur_is_word == prev_is_word : cur_is_word != prev_is_word)
2274          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2275        }        }
2276      break;      break;
2277    
2278      /* Match a single character type; inline for speed */      /* Match any single character type except newline; have to take care with
2279        CRLF newlines and partial matching. */
2280    
2281      case OP_ANY:      case OP_ANY:
2282      if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
2283        if (md->partial != 0 &&
2284            eptr + 1 >= md->end_subject &&
2285            NLBLOCK->nltype == NLTYPE_FIXED &&
2286            NLBLOCK->nllen == 2 &&
2287            RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2288          {
2289          md->hitend = TRUE;
2290          if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2291          }
2292    
2293      /* Fall through */      /* Fall through */
2294    
2295        /* Match any single character whatsoever. */
2296    
2297      case OP_ALLANY:      case OP_ALLANY:
2298      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 */
2299        {                            /* not be updated before SCHECK_PARTIAL. */        {                            /* not be updated before SCHECK_PARTIAL. */
2300        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2301        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2302        }        }
2303      eptr++;      eptr++;
2304      if (utf8) while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  #ifdef SUPPORT_UTF
2305        if (utf) ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
2306    #endif
2307      ecode++;      ecode++;
2308      break;      break;
2309    
# Line 2199  for (;;) Line 2314  for (;;)
2314      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 */
2315        {                            /* not be updated before SCHECK_PARTIAL. */        {                            /* not be updated before SCHECK_PARTIAL. */
2316        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2317        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2318        }        }
2319      eptr++;      eptr++;
2320      ecode++;      ecode++;
# Line 2209  for (;;) Line 2324  for (;;)
2324      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2325        {        {
2326        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2327        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2328        }        }
2329      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2330      if (      if (
2331  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2332         c < 256 &&         c < 256 &&
2333  #endif  #endif
2334         (md->ctypes[c] & ctype_digit) != 0         (md->ctypes[c] & ctype_digit) != 0
2335         )         )
2336        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2337      ecode++;      ecode++;
2338      break;      break;
2339    
# Line 2226  for (;;) Line 2341  for (;;)
2341      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2342        {        {
2343        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2344        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2345        }        }
2346      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2347      if (      if (
2348  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2349         c >= 256 ||         c > 255 ||
2350  #endif  #endif
2351         (md->ctypes[c] & ctype_digit) == 0         (md->ctypes[c] & ctype_digit) == 0
2352         )         )
2353        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2354      ecode++;      ecode++;
2355      break;      break;
2356    
# Line 2243  for (;;) Line 2358  for (;;)
2358      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2359        {        {
2360        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2361        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2362        }        }
2363      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2364      if (      if (
2365  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2366         c < 256 &&         c < 256 &&
2367  #endif  #endif
2368         (md->ctypes[c] & ctype_space) != 0         (md->ctypes[c] & ctype_space) != 0
2369         )         )
2370        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2371      ecode++;      ecode++;
2372      break;      break;
2373    
# Line 2260  for (;;) Line 2375  for (;;)
2375      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2376        {        {
2377        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2378        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2379        }        }
2380      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2381      if (      if (
2382  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2383         c >= 256 ||         c > 255 ||
2384  #endif  #endif
2385         (md->ctypes[c] & ctype_space) == 0         (md->ctypes[c] & ctype_space) == 0
2386         )         )
2387        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2388      ecode++;      ecode++;
2389      break;      break;
2390    
# Line 2277  for (;;) Line 2392  for (;;)
2392      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2393        {        {
2394        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2395        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2396        }        }
2397      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2398      if (      if (
2399  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2400         c < 256 &&         c < 256 &&
2401  #endif  #endif
2402         (md->ctypes[c] & ctype_word) != 0         (md->ctypes[c] & ctype_word) != 0
2403         )         )
2404        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2405      ecode++;      ecode++;
2406      break;      break;
2407    
# Line 2294  for (;;) Line 2409  for (;;)
2409      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2410        {        {
2411        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2412        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2413        }        }
2414      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2415      if (      if (
2416  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2417         c >= 256 ||         c > 255 ||
2418  #endif  #endif
2419         (md->ctypes[c] & ctype_word) == 0         (md->ctypes[c] & ctype_word) == 0
2420         )         )
2421        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2422      ecode++;      ecode++;
2423      break;      break;
2424    
# Line 2311  for (;;) Line 2426  for (;;)
2426      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2427        {        {
2428        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2429        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2430        }        }
2431      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2432      switch(c)      switch(c)
2433        {        {
2434        default: MRRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2435    
2436        case 0x000d:        case CHAR_CR:
2437        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr >= md->end_subject)
2438            {
2439            SCHECK_PARTIAL();
2440            }
2441          else if (RAWUCHARTEST(eptr) == CHAR_LF) eptr++;
2442        break;        break;
2443    
2444        case 0x000a:        case CHAR_LF:
2445        break;        break;
2446    
2447        case 0x000b:        case CHAR_VT:
2448        case 0x000c:        case CHAR_FF:
2449        case 0x0085:        case CHAR_NEL:
2450    #ifndef EBCDIC
2451        case 0x2028:        case 0x2028:
2452        case 0x2029:        case 0x2029:
2453        if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);  #endif  /* Not EBCDIC */
2454          if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
2455        break;        break;
2456        }        }
2457      ecode++;      ecode++;
# Line 2340  for (;;) Line 2461  for (;;)
2461      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2462        {        {
2463        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2464        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2465        }        }
2466      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2467      switch(c)      switch(c)
2468        {        {
2469          HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
2470        default: break;        default: break;
       case 0x09:      /* HT */  
       case 0x20:      /* SPACE */  
       case 0xa0:      /* NBSP */  
       case 0x1680:    /* OGHAM SPACE MARK */  
       case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
       case 0x2000:    /* EN QUAD */  
       case 0x2001:    /* EM QUAD */  
       case 0x2002:    /* EN SPACE */  
       case 0x2003:    /* EM SPACE */  
       case 0x2004:    /* THREE-PER-EM SPACE */  
       case 0x2005:    /* FOUR-PER-EM SPACE */  
       case 0x2006:    /* SIX-PER-EM SPACE */  
       case 0x2007:    /* FIGURE SPACE */  
       case 0x2008:    /* PUNCTUATION SPACE */  
       case 0x2009:    /* THIN SPACE */  
       case 0x200A:    /* HAIR SPACE */  
       case 0x202f:    /* NARROW NO-BREAK SPACE */  
       case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
       case 0x3000:    /* IDEOGRAPHIC SPACE */  
       MRRETURN(MATCH_NOMATCH);  
2471        }        }
2472      ecode++;      ecode++;
2473      break;      break;
# Line 2374  for (;;) Line 2476  for (;;)
2476      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2477        {        {
2478        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2479        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2480        }        }
2481      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2482      switch(c)      switch(c)
2483        {        {
2484        default: MRRETURN(MATCH_NOMATCH);        HSPACE_CASES: break;  /* Byte and multibyte cases */
2485        case 0x09:      /* HT */        default: RRETURN(MATCH_NOMATCH);
       case 0x20:      /* SPACE */  
       case 0xa0:      /* NBSP */  
       case 0x1680:    /* OGHAM SPACE MARK */  
       case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
       case 0x2000:    /* EN QUAD */  
       case 0x2001:    /* EM QUAD */  
       case 0x2002:    /* EN SPACE */  
       case 0x2003:    /* EM SPACE */  
       case 0x2004:    /* THREE-PER-EM SPACE */  
       case 0x2005:    /* FOUR-PER-EM SPACE */  
       case 0x2006:    /* SIX-PER-EM SPACE */  
       case 0x2007:    /* FIGURE SPACE */  
       case 0x2008:    /* PUNCTUATION SPACE */  
       case 0x2009:    /* THIN SPACE */  
       case 0x200A:    /* HAIR SPACE */  
       case 0x202f:    /* NARROW NO-BREAK SPACE */  
       case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
       case 0x3000:    /* IDEOGRAPHIC SPACE */  
       break;  
2486        }        }
2487      ecode++;      ecode++;
2488      break;      break;
# Line 2408  for (;;) Line 2491  for (;;)
2491      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2492        {        {
2493        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2494        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2495        }        }
2496      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2497      switch(c)      switch(c)
2498        {        {
2499          VSPACE_CASES: RRETURN(MATCH_NOMATCH);
2500        default: break;        default: break;
       case 0x0a:      /* LF */  
       case 0x0b:      /* VT */  
       case 0x0c:      /* FF */  
       case 0x0d:      /* CR */  
       case 0x85:      /* NEL */  
       case 0x2028:    /* LINE SEPARATOR */  
       case 0x2029:    /* PARAGRAPH SEPARATOR */  
       MRRETURN(MATCH_NOMATCH);  
2501        }        }
2502      ecode++;      ecode++;
2503      break;      break;
# Line 2430  for (;;) Line 2506  for (;;)
2506      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2507        {        {
2508        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2509        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2510        }        }
2511      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2512      switch(c)      switch(c)
2513        {        {
2514        default: MRRETURN(MATCH_NOMATCH);        VSPACE_CASES: break;
2515        case 0x0a:      /* LF */        default: RRETURN(MATCH_NOMATCH);
       case 0x0b:      /* VT */  
       case 0x0c:      /* FF */  
       case 0x0d:      /* CR */  
       case 0x85:      /* NEL */  
       case 0x2028:    /* LINE SEPARATOR */  
       case 0x2029:    /* PARAGRAPH SEPARATOR */  
       break;  
2516        }        }
2517      ecode++;      ecode++;
2518      break;      break;
# Line 2457  for (;;) Line 2526  for (;;)
2526      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2527        {        {
2528        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2529        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2530        }        }
2531      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2532        {        {
2533          const pcre_uint32 *cp;
2534        const ucd_record *prop = GET_UCD(c);        const ucd_record *prop = GET_UCD(c);
2535    
2536        switch(ecode[1])        switch(ecode[1])
2537          {          {
2538          case PT_ANY:          case PT_ANY:
2539          if (op == OP_NOTPROP) MRRETURN(MATCH_NOMATCH);          if (op == OP_NOTPROP) RRETURN(MATCH_NOMATCH);
2540          break;          break;
2541    
2542          case PT_LAMP:          case PT_LAMP:
2543          if ((prop->chartype == ucp_Lu ||          if ((prop->chartype == ucp_Lu ||
2544               prop->chartype == ucp_Ll ||               prop->chartype == ucp_Ll ||
2545               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))
2546            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2547          break;          break;
2548    
2549          case PT_GC:          case PT_GC:
2550          if ((ecode[2] != _pcre_ucp_gentype[prop->chartype]) == (op == OP_PROP))          if ((ecode[2] != PRIV(ucp_gentype)[prop->chartype]) == (op == OP_PROP))
2551            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2552          break;          break;
2553    
2554          case PT_PC:          case PT_PC:
2555          if ((ecode[2] != prop->chartype) == (op == OP_PROP))          if ((ecode[2] != prop->chartype) == (op == OP_PROP))
2556            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2557          break;          break;
2558    
2559          case PT_SC:          case PT_SC:
2560          if ((ecode[2] != prop->script) == (op == OP_PROP))          if ((ecode[2] != prop->script) == (op == OP_PROP))
2561            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2562          break;          break;
2563    
2564          /* These are specials */          /* These are specials */
2565    
2566          case PT_ALNUM:          case PT_ALNUM:
2567          if ((_pcre_ucp_gentype[prop->chartype] == ucp_L ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2568               _pcre_ucp_gentype[prop->chartype] == ucp_N) == (op == OP_NOTPROP))               PRIV(ucp_gentype)[prop->chartype] == ucp_N) == (op == OP_NOTPROP))
2569            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2570          break;          break;
2571    
2572          case PT_SPACE:    /* Perl space */          case PT_SPACE:    /* Perl space */
2573          if ((_pcre_ucp_gentype[prop->chartype] == ucp_Z ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2574               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)
2575                 == (op == OP_NOTPROP))                 == (op == OP_NOTPROP))
2576            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2577          break;          break;
2578    
2579          case PT_PXSPACE:  /* POSIX space */          case PT_PXSPACE:  /* POSIX space */
2580          if ((_pcre_ucp_gentype[prop->chartype] == ucp_Z ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2581               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
2582               c == CHAR_FF || c == CHAR_CR)               c == CHAR_FF || c == CHAR_CR)
2583                 == (op == OP_NOTPROP))                 == (op == OP_NOTPROP))
2584            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2585          break;          break;
2586    
2587          case PT_WORD:          case PT_WORD:
2588          if ((_pcre_ucp_gentype[prop->chartype] == ucp_L ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2589               _pcre_ucp_gentype[prop->chartype] == ucp_N ||               PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
2590               c == CHAR_UNDERSCORE) == (op == OP_NOTPROP))               c == CHAR_UNDERSCORE) == (op == OP_NOTPROP))
2591            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2592            break;
2593    
2594            case PT_CLIST:
2595            cp = PRIV(ucd_caseless_sets) + prop->caseset;
2596            for (;;)
2597              {
2598              if (c < *cp)
2599                { if (op == OP_PROP) { RRETURN(MATCH_NOMATCH); } else break; }
2600              if (c == *cp++)
2601                { if (op == OP_PROP) break; else { RRETURN(MATCH_NOMATCH); } }
2602              }
2603          break;          break;
2604    
2605          /* This should never occur */          /* This should never occur */
# Line 2538  for (;;) Line 2619  for (;;)
2619      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
2620        {        {
2621        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2622        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2623        }        }
2624      GETCHARINCTEST(c, eptr);      else
     if (UCD_CATEGORY(c) == ucp_M) MRRETURN(MATCH_NOMATCH);  
     while (eptr < md->end_subject)  
2625        {        {
2626        int len = 1;        int lgb, rgb;
2627        if (!utf8) c = *eptr; else { GETCHARLEN(c, eptr, len); }        GETCHARINCTEST(c, eptr);
2628        if (UCD_CATEGORY(c) != ucp_M) break;        lgb = UCD_GRAPHBREAK(c);
2629        eptr += len;        while (eptr < md->end_subject)
2630            {
2631            int len = 1;
2632            if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
2633            rgb = UCD_GRAPHBREAK(c);
2634            if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
2635            lgb = rgb;
2636            eptr += len;
2637            }
2638        }        }
2639        CHECK_PARTIAL();
2640      ecode++;      ecode++;
2641      break;      break;
2642  #endif  #endif  /* SUPPORT_UCP */
2643    
2644    
2645      /* Match a back reference, possibly repeatedly. Look past the end of the      /* Match a back reference, possibly repeatedly. Look past the end of the
# Line 2566  for (;;) Line 2654  for (;;)
2654      case OP_REFI:      case OP_REFI:
2655      caseless = op == OP_REFI;      caseless = op == OP_REFI;
2656      offset = GET2(ecode, 1) << 1;               /* Doubled ref number */      offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
2657      ecode += 3;      ecode += 1 + IMM2_SIZE;
2658    
2659      /* If the reference is unset, there are two possibilities:      /* If the reference is unset, there are two possibilities:
2660    
# Line 2606  for (;;) Line 2694  for (;;)
2694        case OP_CRMINRANGE:        case OP_CRMINRANGE:
2695        minimize = (*ecode == OP_CRMINRANGE);        minimize = (*ecode == OP_CRMINRANGE);
2696        min = GET2(ecode, 1);        min = GET2(ecode, 1);
2697        max = GET2(ecode, 3);        max = GET2(ecode, 1 + IMM2_SIZE);
2698        if (max == 0) max = INT_MAX;        if (max == 0) max = INT_MAX;
2699        ecode += 5;        ecode += 1 + 2 * IMM2_SIZE;
2700        break;        break;
2701    
2702        default:               /* No repeat follows */        default:               /* No repeat follows */
2703        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)
2704          {          {
2705            if (length == -2) eptr = md->end_subject;   /* Partial match */
2706          CHECK_PARTIAL();          CHECK_PARTIAL();
2707          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2708          }          }
2709        eptr += length;        eptr += length;
2710        continue;              /* With the main loop */        continue;              /* With the main loop */
2711        }        }
2712    
2713      /* Handle repeated back references. If the length of the reference is      /* Handle repeated back references. If the length of the reference is
2714      zero, just continue with the main loop. */      zero, just continue with the main loop. If the length is negative, it
2715        means the reference is unset in non-Java-compatible mode. If the minimum is
2716        zero, we can continue at the same level without recursion. For any other
2717        minimum, carrying on will result in NOMATCH. */
2718    
2719      if (length == 0) continue;      if (length == 0) continue;
2720        if (length < 0 && min == 0) continue;
2721    
2722      /* First, ensure the minimum number of matches are present. We get back      /* First, ensure the minimum number of matches are present. We get back
2723      the length of the reference string explicitly rather than passing the      the length of the reference string explicitly rather than passing the
# Line 2635  for (;;) Line 2728  for (;;)
2728        int slength;        int slength;
2729        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2730          {          {
2731            if (slength == -2) eptr = md->end_subject;   /* Partial match */
2732          CHECK_PARTIAL();          CHECK_PARTIAL();
2733          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2734          }          }
2735        eptr += slength;        eptr += slength;
2736        }        }
# Line 2655  for (;;) Line 2749  for (;;)
2749          int slength;          int slength;
2750          RMATCH(eptr, ecode, offset_top, md, eptrb, RM14);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM14);
2751          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2752          if (fi >= max) MRRETURN(MATCH_NOMATCH);          if (fi >= max) RRETURN(MATCH_NOMATCH);
2753          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2754            {            {
2755              if (slength == -2) eptr = md->end_subject;   /* Partial match */
2756            CHECK_PARTIAL();            CHECK_PARTIAL();
2757            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2758            }            }
2759          eptr += slength;          eptr += slength;
2760          }          }
# Line 2676  for (;;) Line 2771  for (;;)
2771          int slength;          int slength;
2772          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2773            {            {
2774            CHECK_PARTIAL();            /* Can't use CHECK_PARTIAL because we don't want to update eptr in
2775              the soft partial matching case. */
2776    
2777              if (slength == -2 && md->partial != 0 &&
2778                  md->end_subject > md->start_used_ptr)
2779                {
2780                md->hitend = TRUE;
2781                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2782                }
2783            break;            break;
2784            }            }
2785          eptr += slength;          eptr += slength;
2786          }          }
2787    
2788        while (eptr >= pp)        while (eptr >= pp)
2789          {          {
2790          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);
2791          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2792          eptr -= length;          eptr -= length;
2793          }          }
2794        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2795        }        }
2796      /* Control never gets here */      /* Control never gets here */
2797    
# Line 2705  for (;;) Line 2809  for (;;)
2809      case OP_NCLASS:      case OP_NCLASS:
2810      case OP_CLASS:      case OP_CLASS:
2811        {        {
2812          /* The data variable is saved across frames, so the byte map needs to
2813          be stored there. */
2814    #define BYTE_MAP ((pcre_uint8 *)data)
2815        data = ecode + 1;                /* Save for matching */        data = ecode + 1;                /* Save for matching */
2816        ecode += 33;                     /* Advance past the item */        ecode += 1 + (32 / sizeof(pcre_uchar)); /* Advance past the item */
2817    
2818        switch (*ecode)        switch (*ecode)
2819          {          {
# Line 2727  for (;;) Line 2834  for (;;)
2834          case OP_CRMINRANGE:          case OP_CRMINRANGE:
2835          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
2836          min = GET2(ecode, 1);          min = GET2(ecode, 1);
2837          max = GET2(ecode, 3);          max = GET2(ecode, 1 + IMM2_SIZE);
2838          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
2839          ecode += 5;          ecode += 1 + 2 * IMM2_SIZE;
2840          break;          break;
2841    
2842          default:               /* No repeat follows */          default:               /* No repeat follows */
# Line 2739  for (;;) Line 2846  for (;;)
2846    
2847        /* First, ensure the minimum number of matches are present. */        /* First, ensure the minimum number of matches are present. */
2848    
2849  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2850        /* UTF-8 mode */        if (utf)
       if (utf8)  
2851          {          {
2852          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2853            {            {
2854            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
2855              {              {
2856              SCHECK_PARTIAL();              SCHECK_PARTIAL();
2857              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2858              }              }
2859            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
2860            if (c > 255)            if (c > 255)
2861              {              {
2862              if (op == OP_CLASS) MRRETURN(MATCH_NOMATCH);              if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2863              }              }
2864            else            else
2865              {              if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
             if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  
             }  
2866            }            }
2867          }          }
2868        else        else
2869  #endif  #endif
2870        /* Not UTF-8 mode */        /* Not UTF mode */
2871          {          {
2872          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2873            {            {
2874            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
2875              {              {
2876              SCHECK_PARTIAL();              SCHECK_PARTIAL();
2877              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2878              }              }
2879            c = *eptr++;            c = *eptr++;
2880            if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  #ifndef COMPILE_PCRE8
2881              if (c > 255)
2882                {
2883                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2884                }
2885              else
2886    #endif
2887                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2888            }            }
2889          }          }
2890    
# Line 2787  for (;;) Line 2898  for (;;)
2898    
2899        if (minimize)        if (minimize)
2900          {          {
2901  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2902          /* UTF-8 mode */          if (utf)
         if (utf8)  
2903            {            {
2904            for (fi = min;; fi++)            for (fi = min;; fi++)
2905              {              {
2906              RMATCH(eptr, ecode, offset_top, md, eptrb, RM16);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM16);
2907              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2908              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2909              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
2910                {                {
2911                SCHECK_PARTIAL();                SCHECK_PARTIAL();
2912                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2913                }                }
2914              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
2915              if (c > 255)              if (c > 255)
2916                {                {
2917                if (op == OP_CLASS) MRRETURN(MATCH_NOMATCH);                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2918                }                }
2919              else              else
2920                {                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
               if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  
               }  
2921              }              }
2922            }            }
2923          else          else
2924  #endif  #endif
2925          /* Not UTF-8 mode */          /* Not UTF mode */
2926            {            {
2927            for (fi = min;; fi++)            for (fi = min;; fi++)
2928              {              {
2929              RMATCH(eptr, ecode, offset_top, md, eptrb, RM17);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM17);
2930              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2931              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2932              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
2933                {                {
2934                SCHECK_PARTIAL();                SCHECK_PARTIAL();
2935                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2936                }                }
2937              c = *eptr++;              c = *eptr++;
2938              if ((data[c/8] & (1 << (c&7))) == 0) MRRETURN(MATCH_NOMATCH);  #ifndef COMPILE_PCRE8
2939                if (c > 255)
2940                  {
2941                  if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2942                  }
2943                else
2944    #endif
2945                  if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2946              }              }
2947            }            }
2948          /* Control never gets here */          /* Control never gets here */
# Line 2839  for (;;) Line 2954  for (;;)
2954          {          {
2955          pp = eptr;          pp = eptr;
2956    
2957  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2958          /* UTF-8 mode */          if (utf)
         if (utf8)  
2959            {            {
2960            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2961              {              {
# Line 2857  for (;;) Line 2971  for (;;)
2971                if (op == OP_CLASS) break;                if (op == OP_CLASS) break;
2972                }                }
2973              else              else
2974                {                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
               if ((data[c/8] & (1 << (c&7))) == 0) break;  
               }  
2975              eptr += len;              eptr += len;
2976              }              }
2977            for (;;)            for (;;)
# Line 2872  for (;;) Line 2984  for (;;)
2984            }            }
2985          else          else
2986  #endif  #endif
2987            /* Not UTF-8 mode */            /* Not UTF mode */
2988            {            {
2989            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2990              {              {
# Line 2882  for (;;) Line 2994  for (;;)
2994                break;                break;
2995                }                }
2996              c = *eptr;              c = *eptr;
2997              if ((data[c/8] & (1 << (c&7))) == 0) break;  #ifndef COMPILE_PCRE8
2998                if (c > 255)
2999                  {
3000                  if (op == OP_CLASS) break;
3001                  }
3002                else
3003    #endif
3004                  if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
3005              eptr++;              eptr++;
3006              }              }
3007            while (eptr >= pp)            while (eptr >= pp)
# Line 2893  for (;;) Line 3012  for (;;)
3012              }              }
3013            }            }
3014    
3015          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3016          }          }
3017    #undef BYTE_MAP
3018        }        }
3019      /* Control never gets here */      /* Control never gets here */
3020    
# Line 2903  for (;;) Line 3023  for (;;)
3023      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
3024      mode, because Unicode properties are supported in non-UTF-8 mode. */      mode, because Unicode properties are supported in non-UTF-8 mode. */
3025    
3026  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
3027      case OP_XCLASS:      case OP_XCLASS:
3028        {        {
3029        data = ecode + 1 + LINK_SIZE;                /* Save for matching */        data = ecode + 1 + LINK_SIZE;                /* Save for matching */
# Line 2928  for (;;) Line 3048  for (;;)
3048          case OP_CRMINRANGE:          case OP_CRMINRANGE:
3049          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
3050          min = GET2(ecode, 1);          min = GET2(ecode, 1);
3051          max = GET2(ecode, 3);          max = GET2(ecode, 1 + IMM2_SIZE);
3052          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
3053          ecode += 5;          ecode += 1 + 2 * IMM2_SIZE;
3054          break;          break;
3055    
3056          default:               /* No repeat follows */          default:               /* No repeat follows */
# Line 2945  for (;;) Line 3065  for (;;)
3065          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3066            {            {
3067            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3068            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3069            }            }
3070          GETCHARINCTEST(c, eptr);          GETCHARINCTEST(c, eptr);
3071          if (!_pcre_xclass(c, data)) MRRETURN(MATCH_NOMATCH);          if (!PRIV(xclass)(c, data, utf)) RRETURN(MATCH_NOMATCH);
3072          }          }
3073    
3074        /* 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 2965  for (;;) Line 3085  for (;;)
3085            {            {
3086            RMATCH(eptr, ecode, offset_top, md, eptrb, RM20);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM20);
3087            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3088            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3089            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3090              {              {
3091              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3092              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3093              }              }
3094            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3095            if (!_pcre_xclass(c, data)) MRRETURN(MATCH_NOMATCH);            if (!PRIV(xclass)(c, data, utf)) RRETURN(MATCH_NOMATCH);
3096            }            }
3097          /* Control never gets here */          /* Control never gets here */
3098          }          }
# Line 2990  for (;;) Line 3110  for (;;)
3110              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3111              break;              break;
3112              }              }
3113    #ifdef SUPPORT_UTF
3114            GETCHARLENTEST(c, eptr, len);            GETCHARLENTEST(c, eptr, len);
3115            if (!_pcre_xclass(c, data)) break;  #else
3116              c = *eptr;
3117    #endif
3118              if (!PRIV(xclass)(c, data, utf)) break;
3119            eptr += len;            eptr += len;
3120            }            }
3121          for(;;)          for(;;)
# Line 2999  for (;;) Line 3123  for (;;)
3123            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);
3124            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3125            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3126            if (utf8) BACKCHAR(eptr);  #ifdef SUPPORT_UTF
3127              if (utf) BACKCHAR(eptr);
3128    #endif
3129            }            }
3130          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3131          }          }
3132    
3133        /* Control never gets here */        /* Control never gets here */
# Line 3011  for (;;) Line 3137  for (;;)
3137      /* Match a single character, casefully */      /* Match a single character, casefully */
3138    
3139      case OP_CHAR:      case OP_CHAR:
3140  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3141      if (utf8)      if (utf)
3142        {        {
3143        length = 1;        length = 1;
3144        ecode++;        ecode++;
# Line 3020  for (;;) Line 3146  for (;;)
3146        if (length > md->end_subject - eptr)        if (length > md->end_subject - eptr)
3147          {          {
3148          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
3149          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3150          }          }
3151        while (length-- > 0) if (*ecode++ != *eptr++) MRRETURN(MATCH_NOMATCH);        while (length-- > 0) if (*ecode++ != RAWUCHARINC(eptr)) RRETURN(MATCH_NOMATCH);
3152        }        }
3153      else      else
3154  #endif  #endif
3155        /* Not UTF mode */
     /* Non-UTF-8 mode */  
3156        {        {
3157        if (md->end_subject - eptr < 1)        if (md->end_subject - eptr < 1)
3158          {          {
3159          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */
3160          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3161          }          }
3162        if (ecode[1] != *eptr++) MRRETURN(MATCH_NOMATCH);        if (ecode[1] != *eptr++) RRETURN(MATCH_NOMATCH);
3163        ecode += 2;        ecode += 2;
3164        }        }
3165      break;      break;
3166    
3167      /* Match a single character, caselessly */      /* Match a single character, caselessly. If we are at the end of the
3168        subject, give up immediately. */
3169    
3170      case OP_CHARI:      case OP_CHARI:
3171  #ifdef SUPPORT_UTF8      if (eptr >= md->end_subject)
3172      if (utf8)        {
3173          SCHECK_PARTIAL();
3174          RRETURN(MATCH_NOMATCH);
3175          }
3176    
3177    #ifdef SUPPORT_UTF
3178        if (utf)
3179        {        {
3180        length = 1;        length = 1;
3181        ecode++;        ecode++;
3182        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
3183    
       if (length > md->end_subject - eptr)  
         {  
         CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */  
         MRRETURN(MATCH_NOMATCH);  
         }  
   
3184        /* 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
3185        can use the fast lookup table. */        we know that its other case must also be one byte long, so we can use the
3186          fast lookup table. We know that there is at least one byte left in the
3187          subject. */
3188    
3189        if (fc < 128)        if (fc < 128)
3190          {          {
3191          if (md->lcc[*ecode++] != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);          pcre_uchar cc = RAWUCHAR(eptr);
3192            if (md->lcc[fc] != TABLE_GET(cc, md->lcc, cc)) RRETURN(MATCH_NOMATCH);
3193            ecode++;
3194            eptr++;
3195          }          }
3196    
3197        /* Otherwise we must pick up the subject character */        /* Otherwise we must pick up the subject character. Note that we cannot
3198          use the value of "length" to check for sufficient bytes left, because the
3199          other case of the character may have more or fewer bytes.  */
3200    
3201        else        else
3202          {          {
3203          unsigned int dc;          pcre_uint32 dc;
3204          GETCHARINC(dc, eptr);          GETCHARINC(dc, eptr);
3205          ecode += length;          ecode += length;
3206    
# Line 3079  for (;;) Line 3212  for (;;)
3212  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3213            if (dc != UCD_OTHERCASE(fc))            if (dc != UCD_OTHERCASE(fc))
3214  #endif  #endif
3215              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3216            }            }
3217          }          }
3218        }        }
3219      else      else
3220  #endif   /* SUPPORT_UTF8 */  #endif   /* SUPPORT_UTF */
3221    
3222      /* Non-UTF-8 mode */      /* Not UTF mode */
3223        {        {
3224        if (md->end_subject - eptr < 1)        if (TABLE_GET(ecode[1], md->lcc, ecode[1])
3225          {            != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);
3226          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */        eptr++;
         MRRETURN(MATCH_NOMATCH);  
         }  
       if (md->lcc[ecode[1]] != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);  
3227        ecode += 2;        ecode += 2;
3228        }        }
3229      break;      break;
# Line 3103  for (;;) Line 3233  for (;;)
3233      case OP_EXACT:      case OP_EXACT:
3234      case OP_EXACTI:      case OP_EXACTI:
3235      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3236      ecode += 3;      ecode += 1 + IMM2_SIZE;
3237      goto REPEATCHAR;      goto REPEATCHAR;
3238    
3239      case OP_POSUPTO:      case OP_POSUPTO:
# Line 3118  for (;;) Line 3248  for (;;)
3248      min = 0;      min = 0;
3249      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3250      minimize = *ecode == OP_MINUPTO || *ecode == OP_MINUPTOI;      minimize = *ecode == OP_MINUPTO || *ecode == OP_MINUPTOI;
3251      ecode += 3;      ecode += 1 + IMM2_SIZE;
3252      goto REPEATCHAR;      goto REPEATCHAR;
3253    
3254      case OP_POSSTAR:      case OP_POSSTAR:
# Line 3166  for (;;) Line 3296  for (;;)
3296      /* Common code for all repeated single-character matches. */      /* Common code for all repeated single-character matches. */
3297    
3298      REPEATCHAR:      REPEATCHAR:
3299  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3300      if (utf8)      if (utf)
3301        {        {
3302        length = 1;        length = 1;
3303        charptr = ecode;        charptr = ecode;
# Line 3180  for (;;) Line 3310  for (;;)
3310        if (length > 1)        if (length > 1)
3311          {          {
3312  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3313          unsigned int othercase;          pcre_uint32 othercase;
3314          if (op >= OP_STARI &&     /* Caseless */          if (op >= OP_STARI &&     /* Caseless */
3315              (othercase = UCD_OTHERCASE(fc)) != fc)              (othercase = UCD_OTHERCASE(fc)) != fc)
3316            oclength = _pcre_ord2utf8(othercase, occhars);            oclength = PRIV(ord2utf)(othercase, occhars);
3317          else oclength = 0;          else oclength = 0;
3318  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3319    
3320          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3321            {            {
3322            if (eptr <= md->end_subject - length &&            if (eptr <= md->end_subject - length &&
3323              memcmp(eptr, charptr, length) == 0) eptr += length;              memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3324  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3325            else if (oclength > 0 &&            else if (oclength > 0 &&
3326                     eptr <= md->end_subject - oclength &&                     eptr <= md->end_subject - oclength &&
3327                     memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                     memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3328  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3329            else            else
3330              {              {
3331              CHECK_PARTIAL();              CHECK_PARTIAL();
3332              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3333              }              }
3334            }            }
3335    
# Line 3211  for (;;) Line 3341  for (;;)
3341              {              {
3342              RMATCH(eptr, ecode, offset_top, md, eptrb, RM22);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM22);
3343              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3344              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3345              if (eptr <= md->end_subject - length &&              if (eptr <= md->end_subject - length &&
3346                memcmp(eptr, charptr, length) == 0) eptr += length;                memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3347  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3348              else if (oclength > 0 &&              else if (oclength > 0 &&
3349                       eptr <= md->end_subject - oclength &&                       eptr <= md->end_subject - oclength &&
3350                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                       memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3351  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3352              else              else
3353                {                {
3354                CHECK_PARTIAL();                CHECK_PARTIAL();
3355                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3356                }                }
3357              }              }
3358            /* Control never gets here */            /* Control never gets here */
# Line 3234  for (;;) Line 3364  for (;;)
3364            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3365              {              {
3366              if (eptr <= md->end_subject - length &&              if (eptr <= md->end_subject - length &&
3367                  memcmp(eptr, charptr, length) == 0) eptr += length;                  memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3368  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3369              else if (oclength > 0 &&              else if (oclength > 0 &&
3370                       eptr <= md->end_subject - oclength &&                       eptr <= md->end_subject - oclength &&
3371                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                       memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3372  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3373              else              else
3374                {                {
# Line 3253  for (;;) Line 3383  for (;;)
3383              {              {
3384              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);
3385              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3386              if (eptr == pp) { MRRETURN(MATCH_NOMATCH); }              if (eptr == pp) { RRETURN(MATCH_NOMATCH); }
3387  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3388              eptr--;              eptr--;
3389              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 3270  for (;;) Line 3400  for (;;)
3400        value of fc will always be < 128. */        value of fc will always be < 128. */
3401        }        }
3402      else      else
3403  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
3404          /* When not in UTF-8 mode, load a single-byte character. */
3405      /* When not in UTF-8 mode, load a single-byte character. */        fc = *ecode++;
3406    
3407      fc = *ecode++;      /* The value of fc at this point is always one character, though we may
3408        or may not be in UTF mode. The code is duplicated for the caseless and
     /* The value of fc at this point is always less than 256, though we may or  
     may not be in UTF-8 mode. The code is duplicated for the caseless and  
3409      caseful cases, for speed, since matching characters is likely to be quite      caseful cases, for speed, since matching characters is likely to be quite
3410      common. First, ensure the minimum number of matches are present. If min =      common. First, ensure the minimum number of matches are present. If min =
3411      max, continue at the same level without recursing. Otherwise, if      max, continue at the same level without recursing. Otherwise, if
# Line 3286  for (;;) Line 3414  for (;;)
3414      maximizing, find the maximum number of characters and work backwards. */      maximizing, find the maximum number of characters and work backwards. */
3415    
3416      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,
3417        max, eptr));        max, (char *)eptr));
3418    
3419      if (op >= OP_STARI)  /* Caseless */      if (op >= OP_STARI)  /* Caseless */
3420        {        {
3421        fc = md->lcc[fc];  #ifdef COMPILE_PCRE8
3422          /* fc must be < 128 if UTF is enabled. */
3423          foc = md->fcc[fc];
3424    #else
3425    #ifdef SUPPORT_UTF
3426    #ifdef SUPPORT_UCP
3427          if (utf && fc > 127)
3428            foc = UCD_OTHERCASE(fc);
3429    #else
3430          if (utf && fc > 127)
3431            foc = fc;
3432    #endif /* SUPPORT_UCP */
3433          else
3434    #endif /* SUPPORT_UTF */
3435            foc = TABLE_GET(fc, md->fcc, fc);
3436    #endif /* COMPILE_PCRE8 */
3437    
3438        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
3439          {          {
3440            pcre_uchar cc;
3441    
3442          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3443            {            {
3444            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3445            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3446            }            }
3447          if (fc != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);          cc = RAWUCHARTEST(eptr);
3448            if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
3449            eptr++;
3450          }          }
3451        if (min == max) continue;        if (min == max) continue;
3452        if (minimize)        if (minimize)
3453          {          {
3454          for (fi = min;; fi++)          for (fi = min;; fi++)
3455            {            {
3456              pcre_uchar cc;
3457    
3458            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);
3459            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3460            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3461            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3462              {              {
3463              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3464              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3465              }              }
3466            if (fc != md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);            cc = RAWUCHARTEST(eptr);
3467              if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
3468              eptr++;
3469            }            }
3470          /* Control never gets here */          /* Control never gets here */
3471          }          }
# Line 3322  for (;;) Line 3474  for (;;)
3474          pp = eptr;          pp = eptr;
3475          for (i = min; i < max; i++)          for (i = min; i < max; i++)
3476            {            {
3477              pcre_uchar cc;
3478    
3479            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3480              {              {
3481              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3482              break;              break;
3483              }              }
3484            if (fc != md->lcc[*eptr]) break;            cc = RAWUCHARTEST(eptr);
3485              if (fc != cc && foc != cc) break;
3486            eptr++;            eptr++;
3487            }            }
3488    
# Line 3339  for (;;) Line 3494  for (;;)
3494            eptr--;            eptr--;
3495            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3496            }            }
3497          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3498          }          }
3499        /* Control never gets here */        /* Control never gets here */
3500        }        }
# Line 3353  for (;;) Line 3508  for (;;)
3508          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3509            {            {
3510            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3511            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3512            }            }
3513          if (fc != *eptr++) MRRETURN(MATCH_NOMATCH);          if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
3514          }          }
3515    
3516        if (min == max) continue;        if (min == max) continue;
# Line 3366  for (;;) Line 3521  for (;;)
3521            {            {
3522            RMATCH(eptr, ecode, offset_top, md, eptrb, RM26);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM26);
3523            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3524            if (fi >= max) MRRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
3525            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3526              {              {
3527              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3528              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3529              }              }
3530            if (fc != *eptr++) MRRETURN(MATCH_NOMATCH);            if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
3531            }            }
3532          /* Control never gets here */          /* Control never gets here */
3533          }          }
# Line 3386  for (;;) Line 3541  for (;;)
3541              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3542              break;              break;
3543              }              }
3544            if (fc != *eptr) break;            if (fc != RAWUCHARTEST(eptr)) break;
3545            eptr++;            eptr++;
3546            }            }
3547          if (possessive) continue;          if (possessive) continue;
# Line 3397  for (;;) Line 3552  for (;;)
3552            eptr--;            eptr--;
3553            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3554            }            }
3555          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3556          }          }
3557        }        }
3558      /* Control never gets here */      /* Control never gets here */
# Line 3410  for (;;) Line 3565  for (;;)
3565      if (eptr >= md->end_subject)      if (eptr >= md->end_subject)
3566        {        {
3567        SCHECK_PARTIAL();        SCHECK_PARTIAL();
3568        MRRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
3569        }        }
3570      ecode++;  #ifdef SUPPORT_UTF
3571      GETCHARINCTEST(c, eptr);      if (utf)
     if (op == OP_NOTI)         /* The caseless case */  
3572        {        {
3573  #ifdef SUPPORT_UTF8        register pcre_uint32 ch, och;
3574        if (c < 256)  
3575  #endif        ecode++;
3576        c = md->lcc[c];        GETCHARINC(ch, ecode);
3577        if (md->lcc[*ecode++] == c) MRRETURN(MATCH_NOMATCH);        GETCHARINC(c, eptr);
3578    
3579          if (op == OP_NOT)
3580            {
3581            if (ch == c) RRETURN(MATCH_NOMATCH);
3582            }
3583          else
3584            {
3585    #ifdef SUPPORT_UCP
3586            if (ch > 127)
3587              och = UCD_OTHERCASE(ch);
3588    #else
3589            if (ch > 127)
3590              och = ch;
3591    #endif /* SUPPORT_UCP */
3592            else
3593              och = TABLE_GET(ch, md->fcc, ch);
3594            if (ch == c || och == c) RRETURN(MATCH_NOMATCH);
3595            }
3596        }        }
3597      else    /* Caseful */      else
3598    #endif
3599        {        {
3600        if (*ecode++ == c) MRRETURN(MATCH_NOMATCH);        register pcre_uint32 ch = ecode[1];
3601          c = *eptr++;
3602          if (ch == c || (op == OP_NOTI && TABLE_GET(ch, md->fcc, ch) == c))
3603            RRETURN(MATCH_NOMATCH);
3604          ecode += 2;
3605        }        }
3606      break;      break;
3607    
# Line 3438  for (;;) Line 3615  for (;;)
3615      case OP_NOTEXACT:      case OP_NOTEXACT:
3616      case OP_NOTEXACTI:      case OP_NOTEXACTI:
3617      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3618      ecode += 3;      ecode += 1 + IMM2_SIZE;
3619      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3620    
3621      case OP_NOTUPTO:      case OP_NOTUPTO:
# Line 3448  for (;;) Line 3625  for (;;)
3625      min = 0;      min = 0;
3626      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3627      minimize = *ecode == OP_NOTMINUPTO || *ecode == OP_NOTMINUPTOI;      minimize = *ecode == OP_NOTMINUPTO || *ecode == OP_NOTMINUPTOI;
3628      ecode += 3;      ecode += 1 + IMM2_SIZE;
3629      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3630    
3631      case OP_NOTPOSSTAR:      case OP_NOTPOSSTAR:
# Line 3480  for (;;) Line 3657  for (;;)
3657      possessive = TRUE;      possessive = TRUE;
3658      min = 0;      min = 0;
3659      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3660      ecode += 3;      ecode += 1 + IMM2_SIZE;
3661      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3662    
3663      case OP_NOTSTAR:      case OP_NOTSTAR:
# Line 3504  for (;;) Line 3681  for (;;)
3681      /* Common code for all repeated single-byte matches. */      /* Common code for all repeated single-byte matches. */
3682    
3683      REPEATNOTCHAR:      REPEATNOTCHAR:
3684      fc = *ecode++;      GETCHARINCTEST(fc, ecode);
3685    
3686      /* The code is duplicated for the caseless and caseful cases, for speed,      /* The code is duplicated for the caseless and caseful cases, for speed,
3687      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 3515  for (;;) Line 3692  for (;;)
3692      characters and work backwards. */      characters and work backwards. */
3693    
3694      DPRINTF(("negative matching %c{%d,%d} against subject %.*s\n", fc, min, max,      DPRINTF(("negative matching %c{%d,%d} against subject %.*s\n", fc, min, max,
3695        max, eptr));        max, (char *)eptr));
3696    
3697      if (op >= OP_NOTSTARI)     /* Caseless */      if (op >= OP_NOTSTARI)     /* Caseless */
3698        {        {
3699        fc = md->lcc[fc];  #ifdef SUPPORT_UTF
3700    #ifdef SUPPORT_UCP
3701          if (utf && fc > 127)
3702            foc = UCD_OTHERCASE(fc);
3703    #else
3704          if (utf && fc > 127)
3705            foc = fc;
3706    #endif /* SUPPORT_UCP */
3707          else
3708    #endif /* SUPPORT_UTF */
3709            foc = TABLE_GET(fc, md->fcc, fc);
3710    
3711  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3712        /* UTF-8 mode */        if (utf)
       if (utf8)  
3713          {          {
3714          register unsigned int d;          register pcre_uint32 d;
3715          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3716            {            {
3717            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3718              {              {
3719              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3720              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3721              }              }
3722            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3723            if (d < 256) d = md->lcc[d];            if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
           if (fc == d) MRRETURN(MATCH_NOMATCH);  
3724            }            }
3725          }          }
3726        else        else
3727  #endif  #endif
3728          /* Not UTF mode */
       /* Not UTF-8 mode */  
3729          {          {
3730          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3731            {            {
3732            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3733              {              {
3734              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3735              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3736              }              }
3737            if (fc == md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);            if (fc == *eptr || foc == *eptr) RRETURN(MATCH_NOMATCH);
3738              eptr++;
3739            }            }
3740          }          }
3741    
# Line 3558  for (;;) Line 3743  for (;;)
3743    
3744        if (minimize)        if (minimize)
3745          {          {
3746  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3747          /* UTF-8 mode */          if (utf)
         if (utf8)  
3748            {            {
3749            register unsigned int d;            register pcre_uint32 d;
3750            for (fi = min;; fi++)            for (fi = min;; fi++)
3751              {              {
3752              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);
3753              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3754              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3755              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3756                {                {
3757                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3758                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3759                }                }
3760              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3761              if (d < 256) d = md->lcc[d];              if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
             if (fc == d) MRRETURN(MATCH_NOMATCH);  
3762              }              }
3763            }            }
3764          else          else
3765  #endif  #endif
3766          /* Not UTF-8 mode */          /* Not UTF mode */
3767            {            {
3768            for (fi = min;; fi++)            for (fi = min;; fi++)
3769              {              {
3770              RMATCH(eptr, ecode, offset_top, md, eptrb, RM29);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM29);
3771              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3772              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3773              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3774                {                {
3775                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3776                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3777                }                }
3778              if (fc == md->lcc[*eptr++]) MRRETURN(MATCH_NOMATCH);              if (fc == *eptr || foc == *eptr) RRETURN(MATCH_NOMATCH);
3779                eptr++;
3780              }              }
3781            }            }
3782          /* Control never gets here */          /* Control never gets here */
# Line 3604  for (;;) Line 3788  for (;;)
3788          {          {
3789          pp = eptr;          pp = eptr;
3790    
3791  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3792          /* UTF-8 mode */          if (utf)
         if (utf8)  
3793            {            {
3794            register unsigned int d;            register pcre_uint32 d;
3795            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3796              {              {
3797              int len = 1;              int len = 1;
# Line 3618  for (;;) Line 3801  for (;;)
3801                break;                break;
3802                }                }
3803              GETCHARLEN(d, eptr, len);              GETCHARLEN(d, eptr, len);
3804              if (d < 256) d = md->lcc[d];              if (fc == d || (unsigned int)foc == d) break;
             if (fc == d) break;  
3805              eptr += len;              eptr += len;
3806              }              }
3807          if (possessive) continue;            if (possessive) continue;
3808          for(;;)            for(;;)
3809              {              {
3810              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);
3811              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
# Line 3633  for (;;) Line 3815  for (;;)
3815            }            }
3816          else          else
3817  #endif  #endif
3818          /* Not UTF-8 mode */          /* Not UTF mode */
3819            {            {
3820            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3821              {              {
# Line 3642  for (;;) Line 3824  for (;;)
3824                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3825                break;                break;
3826                }                }
3827              if (fc == md->lcc[*eptr]) break;              if (fc == *eptr || foc == *eptr) break;
3828              eptr++;              eptr++;
3829              }              }
3830            if (possessive) continue;            if (possessive) continue;
# Line 3654  for (;;) Line 3836  for (;;)
3836              }              }
3837            }            }
3838    
3839          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3840          }          }
3841        /* Control never gets here */        /* Control never gets here */
3842        }        }
# Line 3663  for (;;) Line 3845  for (;;)
3845    
3846      else      else
3847        {        {
3848  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3849        /* UTF-8 mode */        if (utf)
       if (utf8)  
3850          {          {
3851          register unsigned int d;          register pcre_uint32 d;
3852          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3853            {            {
3854            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3855              {              {
3856              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3857              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3858              }              }
3859            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3860            if (fc == d) MRRETURN(MATCH_NOMATCH);            if (fc == d) RRETURN(MATCH_NOMATCH);
3861            }            }
3862          }          }
3863        else        else
3864  #endif  #endif
3865        /* Not UTF-8 mode */        /* Not UTF mode */
3866          {          {
3867          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3868            {            {
3869            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3870              {              {
3871              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3872              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3873              }              }
3874            if (fc == *eptr++) MRRETURN(MATCH_NOMATCH);            if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
3875            }            }
3876          }          }
3877    
# Line 3698  for (;;) Line 3879  for (;;)
3879    
3880        if (minimize)        if (minimize)
3881          {          {
3882  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3883          /* UTF-8 mode */          if (utf)
         if (utf8)  
3884            {            {
3885            register unsigned int d;            register pcre_uint32 d;
3886            for (fi = min;; fi++)            for (fi = min;; fi++)
3887              {              {
3888              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);
3889              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3890              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3891              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3892                {                {
3893                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3894                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3895                }                }
3896              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3897              if (fc == d) MRRETURN(MATCH_NOMATCH);              if (fc == d) RRETURN(MATCH_NOMATCH);
3898              }              }
3899            }            }
3900          else          else
3901  #endif  #endif
3902          /* Not UTF-8 mode */          /* Not UTF mode */
3903            {            {
3904            for (fi = min;; fi++)            for (fi = min;; fi++)
3905              {              {
3906              RMATCH(eptr, ecode, offset_top, md, eptrb, RM33);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM33);
3907              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3908              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3909              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
3910                {                {
3911                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3912                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3913                }                }
3914              if (fc == *eptr++) MRRETURN(MATCH_NOMATCH);              if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
3915              }              }
3916            }            }
3917          /* Control never gets here */          /* Control never gets here */
# Line 3743  for (;;) Line 3923  for (;;)
3923          {          {
3924          pp = eptr;          pp = eptr;
3925    
3926  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3927          /* UTF-8 mode */          if (utf)
         if (utf8)  
3928            {            {
3929            register unsigned int d;            register pcre_uint32 d;
3930            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3931              {              {
3932              int len = 1;              int len = 1;
# Line 3771  for (;;) Line 3950  for (;;)
3950            }            }
3951          else          else
3952  #endif  #endif
3953          /* Not UTF-8 mode */          /* Not UTF mode */
3954            {            {
3955            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3956              {              {
# Line 3792  for (;;) Line 3971  for (;;)
3971              }              }
3972            }            }
3973    
3974          MRRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3975          }          }
3976        }        }
3977      /* Control never gets here */      /* Control never gets here */
# Line 3804  for (;;) Line 3983  for (;;)
3983      case OP_TYPEEXACT:      case OP_TYPEEXACT:
3984      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3985      minimize = TRUE;      minimize = TRUE;
3986      ecode += 3;      ecode += 1 + IMM2_SIZE;
3987      goto REPEATTYPE;      goto REPEATTYPE;
3988    
3989      case OP_TYPEUPTO:      case OP_TYPEUPTO:
# Line 3812  for (;;) Line 3991  for (;;)
3991      min = 0;      min = 0;
3992      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3993      minimize = *ecode == OP_TYPEMINUPTO;      minimize = *ecode == OP_TYPEMINUPTO;
3994      ecode += 3;      ecode += 1 + IMM2_SIZE;
3995      goto REPEATTYPE;      goto REPEATTYPE;
3996    
3997      case OP_TYPEPOSSTAR:      case OP_TYPEPOSSTAR:
# Line 3840  for (;;) Line 4019  for (;;)
4019      possessive = TRUE;      possessive = TRUE;
4020      min = 0;      min = 0;
4021      max = GET2(ecode, 1);      max = GET2(ecode, 1);
4022      ecode += 3;      ecode += 1 + IMM2_SIZE;
4023      goto REPEATTYPE;      goto REPEATTYPE;
4024    
4025      case OP_TYPESTAR:      case OP_TYPESTAR:
# Line 3886  for (;;) Line 4065  for (;;)
4065          switch(prop_type)          switch(prop_type)
4066            {            {
4067            case PT_ANY:            case PT_ANY:
4068            if (prop_fail_result) MRRETURN(MATCH_NOMATCH);            if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4069            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4070              {              {
4071              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4072                {                {
4073                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4074                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4075                }                }
4076              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4077              }              }
# Line 3905  for (;;) Line 4084  for (;;)
4084              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4085                {                {
4086                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4087                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4088                }                }
4089              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4090              chartype = UCD_CHARTYPE(c);              chartype = UCD_CHARTYPE(c);
4091              if ((chartype == ucp_Lu ||              if ((chartype == ucp_Lu ||
4092                   chartype == ucp_Ll ||                   chartype == ucp_Ll ||
4093                   chartype == ucp_Lt) == prop_fail_result)                   chartype == ucp_Lt) == prop_fail_result)
4094                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4095              }              }
4096            break;            break;
4097    
# Line 3922  for (;;) Line 4101  for (;;)
4101              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4102                {                {
4103                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4104                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4105                }                }
4106              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4107              if ((UCD_CATEGORY(c) == prop_value) == prop_fail_result)              if ((UCD_CATEGORY(c) == prop_value) == prop_fail_result)
4108                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4109              }              }
4110            break;            break;
4111    
# Line 3936  for (;;) Line 4115  for (;;)
4115              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4116                {                {
4117                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4118                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4119                }                }
4120              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4121              if ((UCD_CHARTYPE(c) == prop_value) == prop_fail_result)              if ((UCD_CHARTYPE(c) == prop_value) == prop_fail_result)
4122                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4123              }              }
4124            break;            break;
4125    
# Line 3950  for (;;) Line 4129  for (;;)
4129              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4130                {                {
4131                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4132                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4133                }                }
4134              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4135              if ((UCD_SCRIPT(c) == prop_value) == prop_fail_result)              if ((UCD_SCRIPT(c) == prop_value) == prop_fail_result)
4136                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4137              }              }
4138            break;            break;
4139    
# Line 3965  for (;;) Line 4144  for (;;)
4144              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4145                {                {
4146                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4147                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4148                }                }
4149              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4150              category = UCD_CATEGORY(c);              category = UCD_CATEGORY(c);
4151              if ((category == ucp_L || category == ucp_N) == prop_fail_result)              if ((category == ucp_L || category == ucp_N) == prop_fail_result)
4152                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4153              }              }
4154            break;            break;
4155    
# Line 3980  for (;;) Line 4159  for (;;)
4159              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4160                {                {
4161                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4162                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4163                }                }
4164              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4165              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||
4166                   c == CHAR_FF || c == CHAR_CR)                   c == CHAR_FF || c == CHAR_CR)
4167                     == prop_fail_result)                     == prop_fail_result)
4168                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4169              }              }
4170            break;            break;
4171    
# Line 3996  for (;;) Line 4175  for (;;)
4175              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4176                {                {
4177                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4178                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4179                }                }
4180              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4181              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||
4182                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)
4183                     == prop_fail_result)                     == prop_fail_result)
4184                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4185              }              }
4186            break;            break;
4187    
# Line 4013  for (;;) Line 4192  for (;;)
4192              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4193                {                {
4194                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4195                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4196                }                }
4197              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4198              category = UCD_CATEGORY(c);              category = UCD_CATEGORY(c);
4199              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)
4200                     == prop_fail_result)                     == prop_fail_result)
4201                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4202              }              }
4203            break;            break;
4204    
4205              case PT_CLIST:
4206              for (i = 1; i <= min; i++)
4207                {
4208                const pcre_uint32 *cp;
4209                if (eptr >= md->end_subject)
4210                  {
4211                  SCHECK_PARTIAL();
4212                  RRETURN(MATCH_NOMATCH);
4213                  }
4214                GETCHARINCTEST(c, eptr);
4215                cp = PRIV(ucd_caseless_sets) + UCD_CASESET(c);
4216                for (;;)
4217                  {
4218                  if (c < *cp)
4219                    { if (prop_fail_result) break; else { RRETURN(MATCH_NOMATCH); } }
4220                  if (c == *cp++)
4221                    { if (prop_fail_result) { RRETURN(MATCH_NOMATCH); } else break; }
4222                  }
4223                }
4224              break;
4225    
4226            /* This should not occur */            /* This should not occur */
4227    
4228            default:            default:
# Line 4040  for (;;) Line 4240  for (;;)
4240            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4241              {              {
4242              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4243              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4244              }              }
4245            GETCHARINCTEST(c, eptr);            else
           if (UCD_CATEGORY(c) == ucp_M) MRRETURN(MATCH_NOMATCH);  
           while (eptr < md->end_subject)  
4246              {              {
4247              int len = 1;              int lgb, rgb;
4248              if (!utf8) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
4249              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
4250              eptr += len;             while (eptr < md->end_subject)
4251                  {
4252                  int len = 1;
4253                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
4254                  rgb = UCD_GRAPHBREAK(c);
4255                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
4256                  lgb = rgb;
4257                  eptr += len;
4258                  }
4259              }              }
4260              CHECK_PARTIAL();
4261            }            }
4262          }          }
4263    
# Line 4059  for (;;) Line 4266  for (;;)
4266    
4267  /* Handle all other cases when the coding is UTF-8 */  /* Handle all other cases when the coding is UTF-8 */
4268    
4269  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4270        if (utf8) switch(ctype)        if (utf) switch(ctype)
4271          {          {
4272          case OP_ANY:          case OP_ANY:
4273          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 4068  for (;;) Line 4275  for (;;)
4275            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4276              {              {
4277              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4278              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4279                }
4280              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4281              if (md->partial != 0 &&
4282                  eptr + 1 >= md->end_subject &&
4283                  NLBLOCK->nltype == NLTYPE_FIXED &&
4284                  NLBLOCK->nllen == 2 &&
4285                  RAWUCHAR(eptr) == NLBLOCK->nl[0])
4286                {
4287                md->hitend = TRUE;
4288                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4289              }              }
           if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);  
4290            eptr++;            eptr++;
4291            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4292            }            }
4293          break;          break;
4294    
# Line 4082  for (;;) Line 4298  for (;;)
4298            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4299              {              {
4300              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4301              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4302              }              }
4303            eptr++;            eptr++;
4304            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4305            }            }
4306          break;          break;
4307    
4308          case OP_ANYBYTE:          case OP_ANYBYTE:
4309          if (eptr > md->end_subject - min) MRRETURN(MATCH_NOMATCH);          if (eptr > md->end_subject - min) RRETURN(MATCH_NOMATCH);
4310          eptr += min;          eptr += min;
4311          break;          break;
4312    
# Line 4100  for (;;) Line 4316  for (;;)
4316            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4317              {              {
4318              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4319              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4320              }              }
4321            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4322            switch(c)            switch(c)
4323              {              {
4324              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4325    
4326              case 0x000d:              case CHAR_CR:
4327              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && RAWUCHAR(eptr) == CHAR_LF) eptr++;
4328              break;              break;
4329    
4330              case 0x000a:              case CHAR_LF:
4331              break;              break;
4332    
4333              case 0x000b:              case CHAR_VT:
4334              case 0x000c:              case CHAR_FF:
4335              case 0x0085:              case CHAR_NEL:
4336    #ifndef EBCDIC
4337              case 0x2028:              case 0x2028:
4338              case 0x2029:              case 0x2029:
4339              if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);  #endif  /* Not EBCDIC */
4340                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4341              break;              break;
4342              }              }
4343            }            }
# Line 4131  for (;;) Line 4349  for (;;)
4349            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4350              {              {
4351              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4352              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4353              }              }
4354            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4355            switch(c)            switch(c)
4356              {              {
4357                HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
4358              default: break;              default: break;
             case 0x09:      /* HT */  
             case 0x20:      /* SPACE */  
             case 0xa0:      /* NBSP */  
             case 0x1680:    /* OGHAM SPACE MARK */  
             case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
             case 0x2000:    /* EN QUAD */  
             case 0x2001:    /* EM QUAD */  
             case 0x2002:    /* EN SPACE */  
             case 0x2003:    /* EM SPACE */  
             case 0x2004:    /* THREE-PER-EM SPACE */  
             case 0x2005:    /* FOUR-PER-EM SPACE */  
             case 0x2006:    /* SIX-PER-EM SPACE */  
             case 0x2007:    /* FIGURE SPACE */  
             case 0x2008:    /* PUNCTUATION SPACE */  
             case 0x2009:    /* THIN SPACE */  
             case 0x200A:    /* HAIR SPACE */  
             case 0x202f:    /* NARROW NO-BREAK SPACE */  
             case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
             case 0x3000:    /* IDEOGRAPHIC SPACE */  
             MRRETURN(MATCH_NOMATCH);  
4359              }              }
4360            }            }
4361          break;          break;
# Line 4167  for (;;) Line 4366  for (;;)
4366            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4367              {              {
4368              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4369              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4370              }              }
4371            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4372            switch(c)            switch(c)
4373              {              {
4374              default: MRRETURN(MATCH_NOMATCH);              HSPACE_CASES: break;  /* Byte and multibyte cases */
4375              case 0x09:      /* HT */              default: RRETURN(MATCH_NOMATCH);
             case 0x20:      /* SPACE */  
             case 0xa0:      /* NBSP */  
             case 0x1680:    /* OGHAM SPACE MARK */  
             case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
             case 0x2000:    /* EN QUAD */  
             case 0x2001:    /* EM QUAD */  
             case 0x2002:    /* EN SPACE */  
             case 0x2003:    /* EM SPACE */  
             case 0x2004:    /* THREE-PER-EM SPACE */  
             case 0x2005:    /* FOUR-PER-EM SPACE */  
             case 0x2006:    /* SIX-PER-EM SPACE */  
             case 0x2007:    /* FIGURE SPACE */  
             case 0x2008:    /* PUNCTUATION SPACE */  
             case 0x2009:    /* THIN SPACE */  
             case 0x200A:    /* HAIR SPACE */  
             case 0x202f:    /* NARROW NO-BREAK SPACE */  
             case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
             case 0x3000:    /* IDEOGRAPHIC SPACE */  
             break;  
4376              }              }
4377            }            }
4378          break;          break;
# Line 4203  for (;;) Line 4383  for (;;)
4383            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4384              {              {
4385              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4386              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4387              }              }
4388            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4389            switch(c)            switch(c)
4390              {              {
4391                VSPACE_CASES: RRETURN(MATCH_NOMATCH);
4392              default: break;              default: break;
             case 0x0a:      /* LF */  
             case 0x0b:      /* VT */  
             case 0x0c:      /* FF */  
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
             case 0x2028:    /* LINE SEPARATOR */  
             case 0x2029:    /* PARAGRAPH SEPARATOR */  
             MRRETURN(MATCH_NOMATCH);  
4393              }              }
4394            }            }
4395          break;          break;
# Line 4227  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)
4407              {              {
4408              default: MRRETURN(MATCH_NOMATCH);              VSPACE_CASES: break;
4409              case 0x0a:      /* LF */              default: RRETURN(MATCH_NOMATCH);
             case 0x0b:      /* VT */  
             case 0x0c:      /* FF */  
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
             case 0x2028:    /* LINE SEPARATOR */  
             case 0x2029:    /* PARAGRAPH SEPARATOR */  
             break;  
4410              }              }
4411            }            }
4412          break;          break;
# Line 4251  for (;;) Line 4417  for (;;)
4417            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4418              {              {
4419              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4420              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4421              }              }
4422            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4423            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)
4424              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4425            }            }
4426          break;          break;
4427    
4428          case OP_DIGIT:          case OP_DIGIT:
4429          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4430            {            {
4431              pcre_uchar cc;
4432    
4433            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4434              {              {
4435              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4436              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4437              }              }
4438            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_digit) == 0)            cc = RAWUCHAR(eptr);
4439              MRRETURN(MATCH_NOMATCH);            if (cc >= 128 || (md->ctypes[cc] & ctype_digit) == 0)
4440                RRETURN(MATCH_NOMATCH);
4441              eptr++;
4442            /* 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 */
4443            }            }
4444          break;          break;
# Line 4276  for (;;) Line 4446  for (;;)
4446          case OP_NOT_WHITESPACE:          case OP_NOT_WHITESPACE:
4447          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4448            {            {
4449              pcre_uchar cc;
4450    
4451            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4452              {              {
4453              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4454              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4455              }              }
4456            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)            cc = RAWUCHAR(eptr);
4457              MRRETURN(MATCH_NOMATCH);            if (cc < 128 && (md->ctypes[cc] & ctype_space) != 0)
4458            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);              RRETURN(MATCH_NOMATCH);
4459              eptr++;
4460              ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4461            }            }
4462          break;          break;
4463    
4464          case OP_WHITESPACE:          case OP_WHITESPACE:
4465          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4466            {            {
4467              pcre_uchar cc;
4468    
4469            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4470              {              {
4471              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4472              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4473              }              }
4474            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_space) == 0)            cc = RAWUCHAR(eptr);
4475              MRRETURN(MATCH_NOMATCH);            if (cc >= 128 || (md->ctypes[cc] & ctype_space) == 0)
4476                RRETURN(MATCH_NOMATCH);
4477              eptr++;
4478            /* 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 */
4479            }            }
4480          break;          break;
# Line 4304  for (;;) Line 4482  for (;;)
4482          case OP_NOT_WORDCHAR:          case OP_NOT_WORDCHAR:
4483          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4484            {            {
4485              pcre_uchar cc;
4486    
4487            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4488              {              {
4489              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4490              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4491              }              }
4492            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)            cc = RAWUCHAR(eptr);
4493              MRRETURN(MATCH_NOMATCH);            if (cc < 128 && (md->ctypes[cc] & ctype_word) != 0)
4494            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);              RRETURN(MATCH_NOMATCH);
4495              eptr++;
4496              ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4497            }            }
4498          break;          break;
4499    
4500          case OP_WORDCHAR:          case OP_WORDCHAR:
4501          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4502            {            {
4503              pcre_uchar cc;
4504    
4505            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4506              {              {
4507              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4508              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4509              }              }
4510            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_word) == 0)            cc = RAWUCHAR(eptr);
4511              MRRETURN(MATCH_NOMATCH);            if (cc >= 128 || (md->ctypes[cc] & ctype_word) == 0)
4512                RRETURN(MATCH_NOMATCH);
4513              eptr++;
4514            /* 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 */
4515            }            }
4516          break;          break;
# Line 4334  for (;;) Line 4520  for (;;)
4520          }  /* End switch(ctype) */          }  /* End switch(ctype) */
4521    
4522        else        else
4523  #endif     /* SUPPORT_UTF8 */  #endif     /* SUPPORT_UTF */
4524    
4525        /* 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
4526        than OP_PROP and OP_NOTPROP. */        than OP_PROP and OP_NOTPROP. */
# Line 4347  for (;;) Line 4533  for (;;)
4533            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4534              {              {
4535              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4536              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4537                }
4538              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4539              if (md->partial != 0 &&
4540                  eptr + 1 >= md->end_subject &&
4541                  NLBLOCK->nltype == NLTYPE_FIXED &&
4542                  NLBLOCK->nllen == 2 &&
4543                  *eptr == NLBLOCK->nl[0])
4544                {
4545                md->hitend = TRUE;
4546                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4547              }              }
           if (IS_NEWLINE(eptr)) MRRETURN(MATCH_NOMATCH);  
4548            eptr++;            eptr++;
4549            }            }
4550          break;          break;
# Line 4358  for (;;) Line 4553  for (;;)
4553          if (eptr > md->end_subject - min)          if (eptr > md->end_subject - min)
4554            {            {
4555            SCHECK_PARTIAL();            SCHECK_PARTIAL();
4556            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
4557            }            }
4558          eptr += min;          eptr += min;
4559          break;          break;
# Line 4367  for (;;) Line 4562  for (;;)
4562          if (eptr > md->end_subject - min)          if (eptr > md->end_subject - min)
4563            {            {
4564            SCHECK_PARTIAL();            SCHECK_PARTIAL();
4565            MRRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
4566            }            }
4567          eptr += min;          eptr += min;
4568          break;          break;
# Line 4378  for (;;) Line 4573  for (;;)
4573            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4574              {              {
4575              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4576              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4577              }              }
4578            switch(*eptr++)            switch(*eptr++)
4579              {              {
4580              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4581    
4582              case 0x000d:              case CHAR_CR:
4583              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == CHAR_LF) eptr++;
4584              break;              break;
4585    
4586              case 0x000a:              case CHAR_LF:
4587              break;              break;
4588    
4589              case 0x000b:              case CHAR_VT:
4590              case 0x000c:              case CHAR_FF:
4591              case 0x0085:              case CHAR_NEL:
4592              if (md->bsr_anycrlf) MRRETURN(MATCH_NOMATCH);  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4593                case 0x2028:
4594                case 0x2029:
4595    #endif
4596                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4597              break;              break;
4598              }              }
4599            }            }
# Line 4406  for (;;) Line 4605  for (;;)
4605            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4606              {              {
4607              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4608              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4609              }              }
4610            switch(*eptr++)            switch(*eptr++)
4611              {              {
4612              default: break;              default: break;
4613              case 0x09:      /* HT */              HSPACE_BYTE_CASES:
4614              case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4615              case 0xa0:      /* NBSP */              HSPACE_MULTIBYTE_CASES:
4616              MRRETURN(MATCH_NOMATCH);  #endif
4617                RRETURN(MATCH_NOMATCH);
4618              }              }
4619            }            }
4620          break;          break;
# Line 4425  for (;;) Line 4625  for (;;)
4625            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4626              {              {
4627              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4628              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4629              }              }
4630            switch(*eptr++)            switch(*eptr++)
4631              {              {
4632              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4633              case 0x09:      /* HT */              HSPACE_BYTE_CASES:
4634              case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4635              case 0xa0:      /* NBSP */              HSPACE_MULTIBYTE_CASES:
4636    #endif
4637              break;              break;
4638              }              }
4639            }            }
# Line 4444  for (;;) Line 4645  for (;;)
4645            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4646              {              {
4647              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4648              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4649              }              }
4650            switch(*eptr++)            switch(*eptr++)
4651              {              {
4652                VSPACE_BYTE_CASES:
4653    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4654                VSPACE_MULTIBYTE_CASES:
4655    #endif
4656                RRETURN(MATCH_NOMATCH);
4657              default: break;              default: break;
             case 0x0a:      /* LF */  
             case 0x0b:      /* VT */  
             case 0x0c:      /* FF */  
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
             MRRETURN(MATCH_NOMATCH);  
4658              }              }
4659            }            }
4660          break;          break;
# Line 4465  for (;;) Line 4665  for (;;)
4665            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4666              {              {
4667              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4668              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4669              }              }
4670            switch(*eptr++)            switch(*eptr++)
4671              {              {
4672              default: MRRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4673              case 0x0a:      /* LF */              VSPACE_BYTE_CASES:
4674              case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4675              case 0x0c:      /* FF */              VSPACE_MULTIBYTE_CASES:
4676              case 0x0d:      /* CR */  #endif
             case 0x85:      /* NEL */  
4677              break;              break;
4678              }              }
4679            }            }
# Line 4486  for (;;) Line 4685  for (;;)
4685            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4686              {              {
4687              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4688              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4689              }              }
4690            if ((md->ctypes[*eptr++] & ctype_digit) != 0) MRRETURN(MATCH_NOMATCH);            if (MAX_255(*eptr) && (md->ctypes[*eptr] & ctype_digit) != 0)
4691                RRETURN(MATCH_NOMATCH);
4692              eptr++;
4693            }            }
4694          break;          break;
4695    
# Line 4498  for (;;) Line 4699  for (;;)
4699            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4700              {              {
4701              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4702              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4703              }              }
4704            if ((md->ctypes[*eptr++] & ctype_digit) == 0) MRRETURN(MATCH_NOMATCH);            if (!MAX_255(*eptr) || (md->ctypes[*eptr] & ctype_digit) == 0)
4705                RRETURN(MATCH_NOMATCH);
4706              eptr++;
4707            }            }
4708          break;          break;
4709    
# Line 4510  for (;;) Line 4713  for (;;)
4713            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4714              {              {
4715              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4716              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4717              }              }
4718            if ((md->ctypes[*eptr++] & ctype_space) != 0) MRRETURN(MATCH_NOMATCH);            if (MAX_255(*eptr) && (md->ctypes[*eptr] & ctype_space) != 0)
4719                RRETURN(MATCH_NOMATCH);
4720              eptr++;
4721            }            }
4722          break;          break;
4723    
# Line 4522  for (;;) Line 4727  for (;;)
4727            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4728              {              {
4729              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4730              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4731              }              }
4732            if ((md->ctypes[*eptr++] & ctype_space) == 0) MRRETURN(MATCH_NOMATCH);            if (!MAX_255(*eptr) || (md->ctypes[*eptr] & ctype_space) == 0)
4733                RRETURN(MATCH_NOMATCH);
4734              eptr++;
4735            }            }
4736          break;          break;
4737    
# Line 4534  for (;;) Line 4741  for (;;)
4741            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4742              {              {
4743              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4744              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4745              }              }
4746            if ((md->ctypes[*eptr++] & ctype_word) != 0)            if (MAX_255(*eptr) && (md->ctypes[*eptr] & ctype_word) != 0)
4747              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4748              eptr++;
4749            }            }
4750          break;          break;
4751    
# Line 4547  for (;;) Line 4755  for (;;)
4755            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4756              {              {
4757              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4758              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4759              }              }
4760            if ((md->ctypes[*eptr++] & ctype_word) == 0)            if (!MAX_255(*eptr) || (md->ctypes[*eptr] & ctype_word) == 0)
4761              MRRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4762              eptr++;
4763            }            }
4764          break;          break;
4765    
# Line 4579  for (;;) Line 4788  for (;;)
4788              {              {
4789              RMATCH(eptr, ecode, offset_top, md, eptrb, RM36);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM36);
4790              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4791              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4792              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4793                {                {
4794                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4795                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4796                }                }
4797              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4798              if (prop_fail_result) MRRETURN(MATCH_NOMATCH);              if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4799              }              }
4800            /* Control never gets here */            /* Control never gets here */
4801    
# Line 4596  for (;;) Line 4805  for (;;)
4805              int chartype;              int chartype;
4806              RMATCH(eptr, ecode, offset_top, md, eptrb, RM37);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM37);
4807              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4808              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4809              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4810                {                {
4811                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4812                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4813                }                }
4814              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4815              chartype = UCD_CHARTYPE(c);              chartype = UCD_CHARTYPE(c);
4816              if ((chartype == ucp_Lu ||              if ((chartype == ucp_Lu ||
4817                   chartype == ucp_Ll ||                   chartype == ucp_Ll ||
4818                   chartype == ucp_Lt) == prop_fail_result)                   chartype == ucp_Lt) == prop_fail_result)
4819                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4820              }              }
4821            /* Control never gets here */            /* Control never gets here */
4822    
# Line 4616  for (;;) Line 4825  for (;;)
4825              {              {
4826              RMATCH(eptr, ecode, offset_top, md, eptrb, RM38);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM38);
4827              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4828              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4829              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4830                {                {
4831                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4832                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4833                }                }
4834              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4835              if ((UCD_CATEGORY(c) == prop_value) == prop_fail_result)              if ((UCD_CATEGORY(c) == prop_value) == prop_fail_result)
4836                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4837              }              }
4838            /* Control never gets here */            /* Control never gets here */
4839    
# Line 4633  for (;;) Line 4842  for (;;)
4842              {              {
4843              RMATCH(eptr, ecode, offset_top, md, eptrb, RM39);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM39);
4844              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4845              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4846              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4847                {                {
4848                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4849                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4850                }                }
4851              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4852              if ((UCD_CHARTYPE(c) == prop_value) == prop_fail_result)              if ((UCD_CHARTYPE(c) == prop_value) == prop_fail_result)
4853                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4854              }              }
4855            /* Control never gets here */            /* Control never gets here */
4856    
# Line 4650  for (;;) Line 4859  for (;;)
4859              {              {
4860              RMATCH(eptr, ecode, offset_top, md, eptrb, RM40);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM40);
4861              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4862              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4863              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4864                {                {
4865                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4866                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4867                }                }
4868              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4869              if ((UCD_SCRIPT(c) == prop_value) == prop_fail_result)              if ((UCD_SCRIPT(c) == prop_value) == prop_fail_result)
4870                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4871              }              }
4872            /* Control never gets here */            /* Control never gets here */
4873    
# Line 4668  for (;;) Line 4877  for (;;)
4877              int category;              int category;
4878              RMATCH(eptr, ecode, offset_top, md, eptrb, RM59);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM59);
4879              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4880              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4881              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4882                {                {
4883                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4884                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4885                }                }
4886              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4887              category = UCD_CATEGORY(c);              category = UCD_CATEGORY(c);
4888              if ((category == ucp_L || category == ucp_N) == prop_fail_result)              if ((category == ucp_L || category == ucp_N) == prop_fail_result)
4889                MRRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4890              }              }
4891            /* Control never gets here */            /* Control never gets here */
4892    
# Line 4686  for (;;) Line 4895  for (;;)
4895              {              {
4896              RMATCH(eptr, ecode, offset_top, md, eptrb, RM60);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM60);
4897              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4898              if (fi >= max) MRRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4899              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4900                {                {