/[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 895 by ph10, Fri Jan 20 12:12:03 2012 UTC revision 1379 by ph10, Mon Oct 14 13:54:07 2013 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) 1997-2012 University of Cambridge             Copyright (c) 1997-2013 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 57  possible. There are also some static sup Line 56  possible. There are also some static sup
56  #undef min  #undef min
57  #undef max  #undef max
58    
59    /* The md->capture_last field uses the lower 16 bits for the last captured
60    substring (which can never be greater than 65535) and a bit in the top half
61    to mean "capture vector overflowed". This odd way of doing things was
62    implemented when it was realized that preserving and restoring the overflow bit
63    whenever the last capture number was saved/restored made for a neater
64    interface, and doing it this way saved on (a) another variable, which would
65    have increased the stack frame size (a big NO-NO in PCRE) and (b) another
66    separate set of save/restore instructions. The following defines are used in
67    implementing this. */
68    
69    #define CAPLMASK    0x0000ffff    /* The bits used for last_capture */
70    #define OVFLMASK    0xffff0000    /* The bits used for the overflow flag */
71    #define OVFLBIT     0x00010000    /* The bit that is set for overflow */
72    
73  /* Values for setting in md->match_function_type to indicate two special types  /* Values for setting in md->match_function_type to indicate two special types
74  of call to match(). We do it this way to save on using another stack variable,  of call to match(). We do it this way to save on using another stack variable,
75  as stack usage is to be discouraged. */  as stack usage is to be discouraged. */
# Line 74  defined PCRE_ERROR_xxx codes, which are Line 87  defined PCRE_ERROR_xxx codes, which are
87  negative to avoid the external error codes. */  negative to avoid the external error codes. */
88    
89  #define MATCH_ACCEPT       (-999)  #define MATCH_ACCEPT       (-999)
90  #define MATCH_COMMIT       (-998)  #define MATCH_KETRPOS      (-998)
91  #define MATCH_KETRPOS      (-997)  #define MATCH_ONCE         (-997)
92  #define MATCH_ONCE         (-996)  /* The next 5 must be kept together and in sequence so that a test that checks
93    for any one of them can use a range. */
94    #define MATCH_COMMIT       (-996)
95  #define MATCH_PRUNE        (-995)  #define MATCH_PRUNE        (-995)
96  #define MATCH_SKIP         (-994)  #define MATCH_SKIP         (-994)
97  #define MATCH_SKIP_ARG     (-993)  #define MATCH_SKIP_ARG     (-993)
98  #define MATCH_THEN         (-992)  #define MATCH_THEN         (-992)
99    #define MATCH_BACKTRACK_MAX MATCH_THEN
100    #define MATCH_BACKTRACK_MIN MATCH_COMMIT
101    
102  /* 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.
103  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,
# Line 90  because the offset vector is always a mu Line 107  because the offset vector is always a mu
107    
108  /* Min and max values for the common repeats; for the maxima, 0 => infinity */  /* Min and max values for the common repeats; for the maxima, 0 => infinity */
109    
110  static const char rep_min[] = { 0, 0, 1, 1, 0, 0 };  static const char rep_min[] = { 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, };
111  static const char rep_max[] = { 0, 0, 0, 0, 1, 1 };  static const char rep_max[] = { 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, };
   
   
112    
113  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
114  /*************************************************  /*************************************************
# Line 115  Returns:     nothing Line 130  Returns:     nothing
130  static void  static void
131  pchars(const pcre_uchar *p, int length, BOOL is_subject, match_data *md)  pchars(const pcre_uchar *p, int length, BOOL is_subject, match_data *md)
132  {  {
133  unsigned int c;  pcre_uint32 c;
134    BOOL utf = md->utf;
135  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;
136  while (length-- > 0)  while (length-- > 0)
137    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);
138  }  }
139  #endif  #endif
140    
# Line 140  Arguments: Line 156  Arguments:
156    md          points to match data block    md          points to match data block
157    caseless    TRUE if caseless    caseless    TRUE if caseless
158    
159  Returns:      < 0 if not matched, otherwise the number of subject bytes matched  Returns:      >= 0 the number of subject bytes matched
160                  -1 no match
161                  -2 partial match; always given if at end subject
162  */  */
163    
164  static int  static int
# Line 149  match_ref(int offset, register PCRE_PUCH Line 167  match_ref(int offset, register PCRE_PUCH
167  {  {
168  PCRE_PUCHAR eptr_start = eptr;  PCRE_PUCHAR eptr_start = eptr;
169  register PCRE_PUCHAR p = md->start_subject + md->offset_vector[offset];  register PCRE_PUCHAR p = md->start_subject + md->offset_vector[offset];
170    #ifdef SUPPORT_UTF
171    BOOL utf = md->utf;
172    #endif
173    
174  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
175  if (eptr >= md->end_subject)  if (eptr >= md->end_subject)
# Line 163  pchars(p, length, FALSE, md); Line 184  pchars(p, length, FALSE, md);
184  printf("\n");  printf("\n");
185  #endif  #endif
186    
187  /* Always fail if reference not set (and not JavaScript compatible). */  /* Always fail if reference not set (and not JavaScript compatible - in that
188    case the length is passed as zero). */
189    
190  if (length < 0) return -1;  if (length < 0) return -1;
191    
# Line 175  if (caseless) Line 197  if (caseless)
197    {    {
198  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
199  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
200    if (md->utf)    if (utf)
201      {      {
202      /* 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
203      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
204      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.
205      (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
206      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
207      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
208      reference, not along the subject (earlier code did this wrong). */      length along the reference, not along the subject (earlier code did this
209        wrong). */
210    
211      PCRE_PUCHAR endptr = p + length;      PCRE_PUCHAR endptr = p + length;
212      while (p < endptr)      while (p < endptr)
213        {        {
214        int c, d;        pcre_uint32 c, d;
215        if (eptr >= md->end_subject) return -1;        const ucd_record *ur;
216          if (eptr >= md->end_subject) return -2;   /* Partial match */
217        GETCHARINC(c, eptr);        GETCHARINC(c, eptr);
218        GETCHARINC(d, p);        GETCHARINC(d, p);
219        if (c != d && c != UCD_OTHERCASE(d)) return -1;        ur = GET_UCD(d);
220          if (c != d && c != d + ur->other_case)
221            {
222            const pcre_uint32 *pp = PRIV(ucd_caseless_sets) + ur->caseset;
223            for (;;)
224              {
225              if (c < *pp) return -1;
226              if (c == *pp++) break;
227              }
228            }
229        }        }
230      }      }
231    else    else
# Line 202  if (caseless) Line 235  if (caseless)
235    /* 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
236    is no UCP support. */    is no UCP support. */
237      {      {
     if (eptr + length > md->end_subject) return -1;  
238      while (length-- > 0)      while (length-- > 0)
239        {        {
240        if (TABLE_GET(*p, md->lcc, *p) != TABLE_GET(*eptr, md->lcc, *eptr)) return -1;        pcre_uint32 cc, cp;
241          if (eptr >= md->end_subject) return -2;   /* Partial match */
242          cc = RAWUCHARTEST(eptr);
243          cp = RAWUCHARTEST(p);
244          if (TABLE_GET(cp, md->lcc, cp) != TABLE_GET(cc, md->lcc, cc)) return -1;
245        p++;        p++;
246        eptr++;        eptr++;
247        }        }
# Line 217  are in UTF-8 mode. */ Line 253  are in UTF-8 mode. */
253    
254  else  else
255    {    {
256    if (eptr + length > md->end_subject) return -1;    while (length-- > 0)
257    while (length-- > 0) if (*p++ != *eptr++) return -1;      {
258        if (eptr >= md->end_subject) return -2;   /* Partial match */
259        if (RAWUCHARINCTEST(p) != RAWUCHARINCTEST(eptr)) return -1;
260        }
261    }    }
262    
263  return (int)(eptr - eptr_start);  return (int)(eptr - eptr_start);
# Line 273  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM Line 312  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM
312         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,
313         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,
314         RM51,  RM52, RM53, RM54, RM55, RM56, RM57, RM58, RM59, RM60,         RM51,  RM52, RM53, RM54, RM55, RM56, RM57, RM58, RM59, RM60,
315         RM61,  RM62, RM63, RM64, RM65, RM66 };         RM61,  RM62, RM63, RM64, RM65, RM66, RM67, RM68 };
316    
317  /* These versions of the macros use the stack, as normal. There are debugging  /* These versions of the macros use the stack, as normal. There are debugging
318  versions and production versions. Note that the "rw" argument of RMATCH isn't  versions and production versions. Note that the "rw" argument of RMATCH isn't
# Line 291  actually used in this definition. */ Line 330  actually used in this definition. */
330    }    }
331  #define RRETURN(ra) \  #define RRETURN(ra) \
332    { \    { \
333    printf("match() returned %d from line %d ", ra, __LINE__); \    printf("match() returned %d from line %d\n", ra, __LINE__); \
334    return ra; \    return ra; \
335    }    }
336  #else  #else
# Line 311  argument of match(), which never changes Line 350  argument of match(), which never changes
350    
351  #define RMATCH(ra,rb,rc,rd,re,rw)\  #define RMATCH(ra,rb,rc,rd,re,rw)\
352    {\    {\
353    heapframe *newframe = (heapframe *)(PUBL(stack_malloc))(sizeof(heapframe));\    heapframe *newframe = frame->Xnextframe;\
354    if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\    if (newframe == NULL)\
355    frame->Xwhere = rw; \      {\
356        newframe = (heapframe *)(PUBL(stack_malloc))(sizeof(heapframe));\
357        if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\
358        newframe->Xnextframe = NULL;\
359        frame->Xnextframe = newframe;\
360        }\
361      frame->Xwhere = rw;\
362    newframe->Xeptr = ra;\    newframe->Xeptr = ra;\
363    newframe->Xecode = rb;\    newframe->Xecode = rb;\
364    newframe->Xmstart = mstart;\    newframe->Xmstart = mstart;\
# Line 332  argument of match(), which never changes Line 377  argument of match(), which never changes
377    {\    {\
378    heapframe *oldframe = frame;\    heapframe *oldframe = frame;\
379    frame = oldframe->Xprevframe;\    frame = oldframe->Xprevframe;\
   if (oldframe != &frame_zero) (PUBL(stack_free))(oldframe);\  
380    if (frame != NULL)\    if (frame != NULL)\
381      {\      {\
382      rrc = ra;\      rrc = ra;\
# Line 346  argument of match(), which never changes Line 390  argument of match(), which never changes
390    
391  typedef struct heapframe {  typedef struct heapframe {
392    struct heapframe *Xprevframe;    struct heapframe *Xprevframe;
393      struct heapframe *Xnextframe;
394    
395    /* Function arguments that may change */    /* Function arguments that may change */
396    
# Line 376  typedef struct heapframe { Line 421  typedef struct heapframe {
421    
422  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
423    int Xprop_type;    int Xprop_type;
424    int Xprop_value;    unsigned int Xprop_value;
425    int Xprop_fail_result;    int Xprop_fail_result;
426    int Xoclength;    int Xoclength;
427    pcre_uchar Xocchars[6];    pcre_uchar Xocchars[6];
# Line 389  typedef struct heapframe { Line 434  typedef struct heapframe {
434    int Xlength;    int Xlength;
435    int Xmax;    int Xmax;
436    int Xmin;    int Xmin;
437    int Xnumber;    unsigned int Xnumber;
438    int Xoffset;    int Xoffset;
439    int Xop;    unsigned int Xop;
440    int Xsave_capture_last;    pcre_int32 Xsave_capture_last;
441    int Xsave_offset1, Xsave_offset2, Xsave_offset3;    int Xsave_offset1, Xsave_offset2, Xsave_offset3;
442    int Xstacksave[REC_STACK_SAVE_MAX];    int Xstacksave[REC_STACK_SAVE_MAX];
443    
# Line 477  so they can be ordinary variables in all Line 522  so they can be ordinary variables in all
522    
523  register int  rrc;         /* Returns from recursive calls */  register int  rrc;         /* Returns from recursive calls */
524  register int  i;           /* Used for loops not involving calls to RMATCH() */  register int  i;           /* Used for loops not involving calls to RMATCH() */
525  register unsigned int c;   /* Character values not kept over RMATCH() calls */  register pcre_uint32 c;    /* Character values not kept over RMATCH() calls */
526  register BOOL utf;         /* Local copy of UTF flag for speed */  register BOOL utf;         /* Local copy of UTF flag for speed */
527    
528  BOOL minimize, possessive; /* Quantifier options */  BOOL minimize, possessive; /* Quantifier options */
# Line 487  int condcode; Line 532  int condcode;
532  /* 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
533  preserved over calls to RMATCH() are part of a "frame". We set up the top-level  preserved over calls to RMATCH() are part of a "frame". We set up the top-level
534  frame on the stack here; subsequent instantiations are obtained from the heap  frame on the stack here; subsequent instantiations are obtained from the heap
535  whenever RMATCH() does a "recursion". See the macro definitions above. Putting  whenever RMATCH() does a "recursion". See the macro definitions above. Putting
536  the top-level on the stack rather than malloc-ing them all gives a performance  the top-level on the stack rather than malloc-ing them all gives a performance
537  boost in many cases where there is not much "recursion". */  boost in many cases where there is not much "recursion". */
538    
539  #ifdef NO_RECURSE  #ifdef NO_RECURSE
540  heapframe frame_zero;  heapframe *frame = (heapframe *)md->match_frames_base;
 heapframe *frame = &frame_zero;  
 frame->Xprevframe = NULL;            /* Marks the top level */  
541    
542  /* Copy in the original argument variables */  /* Copy in the original argument variables */
543    
# Line 596  BOOL prev_is_word; Line 639  BOOL prev_is_word;
639    
640  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
641  int prop_type;  int prop_type;
642  int prop_value;  unsigned int prop_value;
643  int prop_fail_result;  int prop_fail_result;
644  int oclength;  int oclength;
645  pcre_uchar occhars[6];  pcre_uchar occhars[6];
# Line 607  int ctype; Line 650  int ctype;
650  int length;  int length;
651  int max;  int max;
652  int min;  int min;
653  int number;  unsigned int number;
654  int offset;  int offset;
655  int op;  unsigned int op;
656  int save_capture_last;  pcre_int32 save_capture_last;
657  int save_offset1, save_offset2, save_offset3;  int save_offset1, save_offset2, save_offset3;
658  int stacksave[REC_STACK_SAVE_MAX];  int stacksave[REC_STACK_SAVE_MAX];
659    
660  eptrblock newptrb;  eptrblock newptrb;
661    
662  /* There is a special fudge for calling match() in a way that causes it to  /* There is a special fudge for calling match() in a way that causes it to
663  measure the size of its basic stack frame when the stack is being used for  measure the size of its basic stack frame when the stack is being used for
664  recursion. The second argument (ecode) being NULL triggers this behaviour. It  recursion. The second argument (ecode) being NULL triggers this behaviour. It
665  cannot normally every be NULL. The return is the negated value of the frame  cannot normally ever be NULL. The return is the negated value of the frame
666  size. */  size. */
667    
668  if (ecode == NULL)  if (ecode == NULL)
# Line 631  if (ecode == NULL) Line 674  if (ecode == NULL)
674      int len = (char *)&rdepth - (char *)eptr;      int len = (char *)&rdepth - (char *)eptr;
675      return (len > 0)? -len : len;      return (len > 0)? -len : len;
676      }      }
677    }    }
678  #endif     /* NO_RECURSE */  #endif     /* NO_RECURSE */
679    
680  /* 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 728  for (;;) Line 771  for (;;)
771      unaltered. */      unaltered. */
772    
773      else if (rrc == MATCH_SKIP_ARG &&      else if (rrc == MATCH_SKIP_ARG &&
774          STRCMP_UC_UC(ecode + 2, md->start_match_ptr) == 0)          STRCMP_UC_UC_TEST(ecode + 2, md->start_match_ptr) == 0)
775        {        {
776        md->start_match_ptr = eptr;        md->start_match_ptr = eptr;
777        RRETURN(MATCH_SKIP);        RRETURN(MATCH_SKIP);
# Line 738  for (;;) Line 781  for (;;)
781      case OP_FAIL:      case OP_FAIL:
782      RRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
783    
     /* COMMIT overrides PRUNE, SKIP, and THEN */  
   
784      case OP_COMMIT:      case OP_COMMIT:
785      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
786        eptrb, RM52);        eptrb, RM52);
787      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE &&      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
         rrc != MATCH_SKIP && rrc != MATCH_SKIP_ARG &&  
         rrc != MATCH_THEN)  
       RRETURN(rrc);  
788      RRETURN(MATCH_COMMIT);      RRETURN(MATCH_COMMIT);
789    
     /* PRUNE overrides THEN */  
   
790      case OP_PRUNE:      case OP_PRUNE:
791      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
792        eptrb, RM51);        eptrb, RM51);
793      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
794      RRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
795    
796      case OP_PRUNE_ARG:      case OP_PRUNE_ARG:
# Line 764  for (;;) Line 800  for (;;)
800        eptrb, RM56);        eptrb, RM56);
801      if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&      if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
802           md->mark == NULL) md->mark = ecode + 2;           md->mark == NULL) md->mark = ecode + 2;
803      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
804      RRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
805    
     /* SKIP overrides PRUNE and THEN */  
   
806      case OP_SKIP:      case OP_SKIP:
807      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
808        eptrb, RM53);        eptrb, RM53);
809      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
       RRETURN(rrc);  
810      md->start_match_ptr = eptr;   /* Pass back current position */      md->start_match_ptr = eptr;   /* Pass back current position */
811      RRETURN(MATCH_SKIP);      RRETURN(MATCH_SKIP);
812    
813      /* Note that, for Perl compatibility, SKIP with an argument does NOT set      /* Note that, for Perl compatibility, SKIP with an argument does NOT set
814      nomatch_mark. There is a flag that disables this opcode when re-matching a      nomatch_mark. When a pattern match ends with a SKIP_ARG for which there was
815      pattern that ended with a SKIP for which there was not a matching MARK. */      not a matching mark, we have to re-run the match, ignoring the SKIP_ARG
816        that failed and any that precede it (either they also failed, or were not
817        triggered). To do this, we maintain a count of executed SKIP_ARGs. If a
818        SKIP_ARG gets to top level, the match is re-run with md->ignore_skip_arg
819        set to the count of the one that failed. */
820    
821      case OP_SKIP_ARG:      case OP_SKIP_ARG:
822      if (md->ignore_skip_arg)      md->skip_arg_count++;
823        if (md->skip_arg_count <= md->ignore_skip_arg)
824        {        {
825        ecode += PRIV(OP_lengths)[*ecode] + ecode[1];        ecode += PRIV(OP_lengths)[*ecode] + ecode[1];
826        break;        break;
827        }        }
828      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
829        eptrb, RM57);        eptrb, RM57);
830      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
       RRETURN(rrc);  
831    
832      /* 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
833      returning the special MATCH_SKIP_ARG return code. This will either be      returning the special MATCH_SKIP_ARG return code. This will either be
834      caught by a matching MARK, or get to the top, where it causes a rematch      caught by a matching MARK, or get to the top, where it causes a rematch
835      with the md->ignore_skip_arg flag set. */      with md->ignore_skip_arg set to the value of md->skip_arg_count. */
836    
837      md->start_match_ptr = ecode + 2;      md->start_match_ptr = ecode + 2;
838      RRETURN(MATCH_SKIP_ARG);      RRETURN(MATCH_SKIP_ARG);
# Line 838  for (;;) Line 875  for (;;)
875      case OP_ONCE_NC:      case OP_ONCE_NC:
876      prev = ecode;      prev = ecode;
877      saved_eptr = eptr;      saved_eptr = eptr;
878      save_mark = md->mark;      save_mark = md->mark;
879      do      do
880        {        {
881        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM64);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM64);
# Line 857  for (;;) Line 894  for (;;)
894    
895        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
896        ecode += GET(ecode,1);        ecode += GET(ecode,1);
897        md->mark = save_mark;        md->mark = save_mark;
898        }        }
899      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
900    
# Line 897  for (;;) Line 934  for (;;)
934        }        }
935      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
936        {        {
       md->match_function_type = MATCH_CBEGROUP;  
937        RMATCH(eptr, prev, offset_top, md, eptrb, RM66);        RMATCH(eptr, prev, offset_top, md, eptrb, RM66);
938        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
939        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
# Line 937  for (;;) Line 973  for (;;)
973        save_offset2 = md->offset_vector[offset+1];        save_offset2 = md->offset_vector[offset+1];
974        save_offset3 = md->offset_vector[md->offset_end - number];        save_offset3 = md->offset_vector[md->offset_end - number];
975        save_capture_last = md->capture_last;        save_capture_last = md->capture_last;
976        save_mark = md->mark;        save_mark = md->mark;
977    
978        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));
979        md->offset_vector[md->offset_end - number] =        md->offset_vector[md->offset_end - number] =
# Line 1026  for (;;) Line 1062  for (;;)
1062    
1063      for (;;)      for (;;)
1064        {        {
1065        if (op >= OP_SBRA || op == OP_ONCE) md->match_function_type = MATCH_CBEGROUP;        if (op >= OP_SBRA || op == OP_ONCE)
1066            md->match_function_type = MATCH_CBEGROUP;
1067    
1068        /* 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
1069        the pattern, and this is the final alternative, optimize as described        the pattern, and this is the final alternative, optimize as described
# Line 1041  for (;;) Line 1078  for (;;)
1078        /* In all other cases, we have to make another call to match(). */        /* In all other cases, we have to make another call to match(). */
1079    
1080        save_mark = md->mark;        save_mark = md->mark;
1081          save_capture_last = md->capture_last;
1082        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md, eptrb,        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md, eptrb,
1083          RM2);          RM2);
1084    
1085        /* See comment in the code for capturing groups above about handling        /* See comment in the code for capturing groups above about handling
1086        THEN. */        THEN. */
1087    
# Line 1070  for (;;) Line 1108  for (;;)
1108          RRETURN(rrc);          RRETURN(rrc);
1109          }          }
1110        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1111        md->mark = save_mark;        md->mark = save_mark;
1112        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1113          md->capture_last = save_capture_last;
1114        }        }
1115    
1116      RRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
# Line 1193  for (;;) Line 1232  for (;;)
1232      POSSESSIVE_NON_CAPTURE:      POSSESSIVE_NON_CAPTURE:
1233      matched_once = FALSE;      matched_once = FALSE;
1234      code_offset = (int)(ecode - md->start_code);      code_offset = (int)(ecode - md->start_code);
1235        save_capture_last = md->capture_last;
1236    
1237      for (;;)      for (;;)
1238        {        {
# Line 1222  for (;;) Line 1262  for (;;)
1262        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1263        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1264        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1265          md->capture_last = save_capture_last;
1266        }        }
1267    
1268      if (matched_once || allow_zero)      if (matched_once || allow_zero)
# Line 1233  for (;;) Line 1274  for (;;)
1274    
1275      /* Control never reaches here. */      /* Control never reaches here. */
1276    
1277      /* Conditional group: compilation checked that there are no more than      /* Conditional group: compilation checked that there are no more than two
1278      two branches. If the condition is false, skipping the first branch takes us      branches. If the condition is false, skipping the first branch takes us
1279      past the end if there is only one branch, but that's OK because that is      past the end of the item if there is only one branch, but that's exactly
1280      exactly what going to the ket would do. */      what we want. */
1281    
1282      case OP_COND:      case OP_COND:
1283      case OP_SCOND:      case OP_SCOND:
1284      codelink = GET(ecode, 1);  
1285        /* The variable codelink will be added to ecode when the condition is
1286        false, to get to the second branch. Setting it to the offset to the ALT
1287        or KET, then incrementing ecode achieves this effect. We now have ecode
1288        pointing to the condition or callout. */
1289    
1290        codelink = GET(ecode, 1);   /* Offset to the second branch */
1291        ecode += 1 + LINK_SIZE;     /* From this opcode */
1292    
1293      /* Because of the way auto-callout works during compile, a callout item is      /* Because of the way auto-callout works during compile, a callout item is
1294      inserted between OP_COND and an assertion condition. */      inserted between OP_COND and an assertion condition. */
1295    
1296      if (ecode[LINK_SIZE+1] == OP_CALLOUT)      if (*ecode == OP_CALLOUT)
1297        {        {
1298        if (PUBL(callout) != NULL)        if (PUBL(callout) != NULL)
1299          {          {
1300          PUBL(callout_block) cb;          PUBL(callout_block) cb;
1301          cb.version          = 2;   /* Version 1 of the callout block */          cb.version          = 2;   /* Version 1 of the callout block */
1302          cb.callout_number   = ecode[LINK_SIZE+2];          cb.callout_number   = ecode[1];
1303          cb.offset_vector    = md->offset_vector;          cb.offset_vector    = md->offset_vector;
1304  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
1305          cb.subject          = (PCRE_SPTR)md->start_subject;          cb.subject          = (PCRE_SPTR)md->start_subject;
1306  #else  #elif defined COMPILE_PCRE16
1307          cb.subject          = (PCRE_SPTR16)md->start_subject;          cb.subject          = (PCRE_SPTR16)md->start_subject;
1308    #elif defined COMPILE_PCRE32
1309            cb.subject          = (PCRE_SPTR32)md->start_subject;
1310  #endif  #endif
1311          cb.subject_length   = (int)(md->end_subject - md->start_subject);          cb.subject_length   = (int)(md->end_subject - md->start_subject);
1312          cb.start_match      = (int)(mstart - md->start_subject);          cb.start_match      = (int)(mstart - md->start_subject);
1313          cb.current_position = (int)(eptr - md->start_subject);          cb.current_position = (int)(eptr - md->start_subject);
1314          cb.pattern_position = GET(ecode, LINK_SIZE + 3);          cb.pattern_position = GET(ecode, 2);
1315          cb.next_item_length = GET(ecode, 3 + 2*LINK_SIZE);          cb.next_item_length = GET(ecode, 2 + LINK_SIZE);
1316          cb.capture_top      = offset_top/2;          cb.capture_top      = offset_top/2;
1317          cb.capture_last     = md->capture_last;          cb.capture_last     = md->capture_last & CAPLMASK;
1318            /* Internal change requires this for API compatibility. */
1319            if (cb.capture_last == 0) cb.capture_last = -1;
1320          cb.callout_data     = md->callout_data;          cb.callout_data     = md->callout_data;
1321          cb.mark             = md->nomatch_mark;          cb.mark             = md->nomatch_mark;
1322          if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);          if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1323          if (rrc < 0) RRETURN(rrc);          if (rrc < 0) RRETURN(rrc);
1324          }          }
1325    
1326          /* Advance ecode past the callout, so it now points to the condition. We
1327          must adjust codelink so that the value of ecode+codelink is unchanged. */
1328    
1329        ecode += PRIV(OP_lengths)[OP_CALLOUT];        ecode += PRIV(OP_lengths)[OP_CALLOUT];
1330          codelink -= PRIV(OP_lengths)[OP_CALLOUT];
1331        }        }
1332    
1333      condcode = ecode[LINK_SIZE+1];      /* Test the various possible conditions */
   
     /* Now see what the actual condition is */  
1334    
1335      if (condcode == OP_RREF || condcode == OP_NRREF)    /* Recursion test */      condition = FALSE;
1336        switch(condcode = *ecode)
1337        {        {
1338        if (md->recursive == NULL)                /* Not recursing => FALSE */        case OP_RREF:         /* Numbered group recursion test */
1339          {        if (md->recursive != NULL)     /* Not recursing => FALSE */
         condition = FALSE;  
         ecode += GET(ecode, 1);  
         }  
       else  
1340          {          {
1341          int recno = GET2(ecode, LINK_SIZE + 2);   /* Recursion group number*/          unsigned int recno = GET2(ecode, 1);   /* Recursion group number*/
1342          condition = (recno == RREF_ANY || recno == md->recursive->group_num);          condition = (recno == RREF_ANY || recno == md->recursive->group_num);
1343            }
1344          break;
1345    
1346          /* If the test is for recursion into a specific subpattern, and it is        case OP_DNRREF:       /* Duplicate named group recursion test */
1347          false, but the test was set up by name, scan the table to see if the        if (md->recursive != NULL)
1348          name refers to any other numbers, and test them. The condition is true          {
1349          if any one is set. */          int count = GET2(ecode, 1 + IMM2_SIZE);
1350            pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
1351          if (!condition && condcode == OP_NRREF)          while (count-- > 0)
1352            {            {
1353            pcre_uchar *slotA = md->name_table;            unsigned int recno = GET2(slot, 0);
1354            for (i = 0; i < md->name_count; i++)            condition = recno == md->recursive->group_num;
1355              {            if (condition) break;
1356              if (GET2(slotA, 0) == recno) break;            slot += md->name_entry_size;
             slotA += md->name_entry_size;  
             }  
   
           /* Found a name for the number - there can be only one; duplicate  
           names for different numbers are allowed, but not vice versa. First  
           scan down for duplicates. */  
   
           if (i < md->name_count)  
             {  
             pcre_uchar *slotB = slotA;  
             while (slotB > md->name_table)  
               {  
               slotB -= md->name_entry_size;  
               if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)  
                 {  
                 condition = GET2(slotB, 0) == md->recursive->group_num;  
                 if (condition) break;  
                 }  
               else break;  
               }  
   
             /* Scan up for duplicates */  
   
             if (!condition)  
               {  
               slotB = slotA;  
               for (i++; i < md->name_count; i++)  
                 {  
                 slotB += md->name_entry_size;  
                 if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)  
                   {  
                   condition = GET2(slotB, 0) == md->recursive->group_num;  
                   if (condition) break;  
                   }  
                 else break;  
                 }  
               }  
             }  
1357            }            }
   
         /* Chose branch according to the condition */  
   
         ecode += condition? 1 + IMM2_SIZE : GET(ecode, 1);  
1358          }          }
1359        }        break;
1360    
1361      else if (condcode == OP_CREF || condcode == OP_NCREF)  /* Group used test */        case OP_CREF:         /* Numbered group used test */
1362        {        offset = GET2(ecode, 1) << 1;  /* Doubled ref number */
       offset = GET2(ecode, LINK_SIZE+2) << 1;  /* Doubled ref number */  
1363        condition = offset < offset_top && md->offset_vector[offset] >= 0;        condition = offset < offset_top && md->offset_vector[offset] >= 0;
1364          break;
1365    
1366        /* If the numbered capture is unset, but the reference was by name,        case OP_DNCREF:      /* Duplicate named group used test */
       scan the table to see if the name refers to any other numbers, and test  
       them. The condition is true if any one is set. This is tediously similar  
       to the code above, but not close enough to try to amalgamate. */  
   
       if (!condition && condcode == OP_NCREF)  
1367          {          {
1368          int refno = offset >> 1;          int count = GET2(ecode, 1 + IMM2_SIZE);
1369          pcre_uchar *slotA = md->name_table;          pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
1370            while (count-- > 0)
1371          for (i = 0; i < md->name_count; i++)            {
1372            {            offset = GET2(slot, 0) << 1;
1373            if (GET2(slotA, 0) == refno) break;            condition = offset < offset_top && md->offset_vector[offset] >= 0;
1374            slotA += md->name_entry_size;            if (condition) break;
1375            }            slot += md->name_entry_size;
   
         /* Found a name for the number - there can be only one; duplicate names  
         for different numbers are allowed, but not vice versa. First scan down  
         for duplicates. */  
   
         if (i < md->name_count)  
           {  
           pcre_uchar *slotB = slotA;  
           while (slotB > md->name_table)  
             {  
             slotB -= md->name_entry_size;  
             if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)  
               {  
               offset = GET2(slotB, 0) << 1;  
               condition = offset < offset_top &&  
                 md->offset_vector[offset] >= 0;  
               if (condition) break;  
               }  
             else break;  
             }  
   
           /* Scan up for duplicates */  
   
           if (!condition)  
             {  
             slotB = slotA;  
             for (i++; i < md->name_count; i++)  
               {  
               slotB += md->name_entry_size;  
               if (STRCMP_UC_UC(slotA + IMM2_SIZE, slotB + IMM2_SIZE) == 0)  
                 {  
                 offset = GET2(slotB, 0) << 1;  
                 condition = offset < offset_top &&  
                   md->offset_vector[offset] >= 0;  
                 if (condition) break;  
                 }  
               else break;  
               }  
             }  
1376            }            }
1377          }          }
1378          break;
1379    
1380        /* Chose branch according to the condition */        case OP_DEF:     /* DEFINE - always false */
1381          break;
       ecode += condition? 1 + IMM2_SIZE : GET(ecode, 1);  
       }  
   
     else if (condcode == OP_DEF)     /* DEFINE - always false */  
       {  
       condition = FALSE;  
       ecode += GET(ecode, 1);  
       }  
1382    
1383      /* The condition is an assertion. Call match() to evaluate it - setting        /* The condition is an assertion. Call match() to evaluate it - setting
1384      md->match_function_type to MATCH_CONDASSERT causes it to stop at the end of        md->match_function_type to MATCH_CONDASSERT causes it to stop at the end
1385      an assertion. */        of an assertion. */
1386    
1387      else        default:
       {  
1388        md->match_function_type = MATCH_CONDASSERT;        md->match_function_type = MATCH_CONDASSERT;
1389        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM3);        RMATCH(eptr, ecode, offset_top, md, NULL, RM3);
1390        if (rrc == MATCH_MATCH)        if (rrc == MATCH_MATCH)
1391          {          {
1392          if (md->end_offset_top > offset_top)          if (md->end_offset_top > offset_top)
1393            offset_top = md->end_offset_top;  /* Captures may have happened */            offset_top = md->end_offset_top;  /* Captures may have happened */
1394          condition = TRUE;          condition = TRUE;
1395          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);  
1396            /* Advance ecode past the assertion to the start of the first branch,
1397            but adjust it so that the general choosing code below works. */
1398    
1399            ecode += GET(ecode, 1);
1400          while (*ecode == OP_ALT) ecode += GET(ecode, 1);          while (*ecode == OP_ALT) ecode += GET(ecode, 1);
1401            ecode += 1 + LINK_SIZE - PRIV(OP_lengths)[condcode];
1402          }          }
1403    
1404        /* PCRE doesn't allow the effect of (*THEN) to escape beyond an        /* PCRE doesn't allow the effect of (*THEN) to escape beyond an
1405        assertion; it is therefore treated as NOMATCH. */        assertion; it is therefore treated as NOMATCH. Any other return is an
1406          error. */
1407    
1408        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
1409          {          {
1410          RRETURN(rrc);         /* Need braces because of following else */          RRETURN(rrc);         /* Need braces because of following else */
1411          }          }
1412        else        break;
         {  
         condition = FALSE;  
         ecode += codelink;  
         }  
1413        }        }
1414    
1415      /* We are now at the branch that is to be obeyed. As there is only one, can      /* Choose branch according to the condition */
1416      use tail recursion to avoid using another stack frame, except when there is  
1417      unlimited repeat of a possibly empty group. In the latter case, a recursive      ecode += condition? PRIV(OP_lengths)[condcode] : codelink;
     call to match() is always required, unless the second alternative doesn't  
     exist, in which case we can just plough on. Note that, for compatibility  
     with Perl, the | in a conditional group is NOT treated as creating two  
     alternatives. If a THEN is encountered in the branch, it propagates out to  
     the enclosing alternative (unless nested in a deeper set of alternatives,  
     of course). */  
1418    
1419      if (condition || *ecode == OP_ALT)      /* We are now at the branch that is to be obeyed. As there is only one, we
1420        can use tail recursion to avoid using another stack frame, except when
1421        there is unlimited repeat of a possibly empty group. In the latter case, a
1422        recursive call to match() is always required, unless the second alternative
1423        doesn't exist, in which case we can just plough on. Note that, for
1424        compatibility with Perl, the | in a conditional group is NOT treated as
1425        creating two alternatives. If a THEN is encountered in the branch, it
1426        propagates out to the enclosing alternative (unless nested in a deeper set
1427        of alternatives, of course). */
1428    
1429        if (condition || ecode[-(1+LINK_SIZE)] == OP_ALT)
1430        {        {
1431        if (op != OP_SCOND)        if (op != OP_SCOND)
1432          {          {
         ecode += 1 + LINK_SIZE;  
1433          goto TAIL_RECURSE;          goto TAIL_RECURSE;
1434          }          }
1435    
1436        md->match_function_type = MATCH_CBEGROUP;        md->match_function_type = MATCH_CBEGROUP;
1437        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM49);        RMATCH(eptr, ecode, offset_top, md, eptrb, RM49);
1438        RRETURN(rrc);        RRETURN(rrc);
1439        }        }
1440    
# Line 1477  for (;;) Line 1442  for (;;)
1442    
1443      else      else
1444        {        {
       ecode += 1 + LINK_SIZE;  
1445        }        }
1446      break;      break;
1447    
# Line 1486  for (;;) Line 1450  for (;;)
1450      to close any currently open capturing brackets. */      to close any currently open capturing brackets. */
1451    
1452      case OP_CLOSE:      case OP_CLOSE:
1453      number = GET2(ecode, 1);      number = GET2(ecode, 1);   /* Must be less than 65536 */
1454      offset = number << 1;      offset = number << 1;
1455    
1456  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
# Line 1494  for (;;) Line 1458  for (;;)
1458        printf("\n");        printf("\n");
1459  #endif  #endif
1460    
1461      md->capture_last = number;      md->capture_last = (md->capture_last & OVFLMASK) | number;
1462      if (offset >= md->offset_max) md->offset_overflow = TRUE; else      if (offset >= md->offset_max) md->capture_last |= OVFLBIT; else
1463        {        {
1464        md->offset_vector[offset] =        md->offset_vector[offset] =
1465          md->offset_vector[md->offset_end - number];          md->offset_vector[md->offset_end - number];
# Line 1549  for (;;) Line 1513  for (;;)
1513    
1514      case OP_ASSERT:      case OP_ASSERT:
1515      case OP_ASSERTBACK:      case OP_ASSERTBACK:
1516      save_mark = md->mark;      save_mark = md->mark;
1517      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1518        {        {
1519        condassert = TRUE;        condassert = TRUE;
# Line 1557  for (;;) Line 1521  for (;;)
1521        }        }
1522      else condassert = FALSE;      else condassert = FALSE;
1523    
1524        /* Loop for each branch */
1525    
1526      do      do
1527        {        {
1528        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM4);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM4);
1529    
1530          /* A match means that the assertion is true; break out of the loop
1531          that matches its alternatives. */
1532    
1533        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1534          {          {
1535          mstart = md->start_match_ptr;   /* In case \K reset it */          mstart = md->start_match_ptr;   /* In case \K reset it */
1536          break;          break;
1537          }          }
1538    
1539        /* PCRE does not allow THEN to escape beyond an assertion; it is treated        /* If not matched, restore the previous mark setting. */
       as NOMATCH. */  
1540    
1541        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);        md->mark = save_mark;
1542    
1543          /* See comment in the code for capturing groups above about handling
1544          THEN. */
1545    
1546          if (rrc == MATCH_THEN)
1547            {
1548            next = ecode + GET(ecode,1);
1549            if (md->start_match_ptr < next &&
1550                (*ecode == OP_ALT || *next == OP_ALT))
1551              rrc = MATCH_NOMATCH;
1552            }
1553    
1554          /* Anything other than NOMATCH causes the entire assertion to fail,
1555          passing back the return code. This includes COMMIT, SKIP, PRUNE and an
1556          uncaptured THEN, which means they take their normal effect. This
1557          consistent approach does not always have exactly the same effect as in
1558          Perl. */
1559    
1560          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1561        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
       md->mark = save_mark;  
1562        }        }
1563      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);   /* Continue for next alternative */
1564    
1565        /* If we have tried all the alternative branches, the assertion has
1566        failed. If not, we broke out after a match. */
1567    
1568      if (*ecode == OP_KET) RRETURN(MATCH_NOMATCH);      if (*ecode == OP_KET) RRETURN(MATCH_NOMATCH);
1569    
# Line 1581  for (;;) Line 1571  for (;;)
1571    
1572      if (condassert) RRETURN(MATCH_MATCH);      if (condassert) RRETURN(MATCH_MATCH);
1573    
1574      /* Continue from after the assertion, updating the offsets high water      /* Continue from after a successful assertion, updating the offsets high
1575      mark, since extracts may have been taken during the assertion. */      water mark, since extracts may have been taken during the assertion. */
1576    
1577      do ecode += GET(ecode,1); while (*ecode == OP_ALT);      do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1578      ecode += 1 + LINK_SIZE;      ecode += 1 + LINK_SIZE;
1579      offset_top = md->end_offset_top;      offset_top = md->end_offset_top;
1580      continue;      continue;
1581    
1582      /* Negative assertion: all branches must fail to match. Encountering SKIP,      /* Negative assertion: all branches must fail to match for the assertion to
1583      PRUNE, or COMMIT means we must assume failure without checking subsequent      succeed. */
     branches. */  
1584    
1585      case OP_ASSERT_NOT:      case OP_ASSERT_NOT:
1586      case OP_ASSERTBACK_NOT:      case OP_ASSERTBACK_NOT:
1587      save_mark = md->mark;      save_mark = md->mark;
1588      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1589        {        {
1590        condassert = TRUE;        condassert = TRUE;
# Line 1603  for (;;) Line 1592  for (;;)
1592        }        }
1593      else condassert = FALSE;      else condassert = FALSE;
1594    
1595        /* Loop for each alternative branch. */
1596    
1597      do      do
1598        {        {
1599        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);
1600        md->mark = save_mark;        md->mark = save_mark;   /* Always restore the mark setting */
1601        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) RRETURN(MATCH_NOMATCH);  
1602        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)        switch(rrc)
1603          {          {
1604          do ecode += GET(ecode,1); while (*ecode == OP_ALT);          case MATCH_MATCH:            /* A successful match means */
1605            case MATCH_ACCEPT:           /* the assertion has failed. */
1606            RRETURN(MATCH_NOMATCH);
1607    
1608            case MATCH_NOMATCH:          /* Carry on with next branch */
1609          break;          break;
1610    
1611            /* See comment in the code for capturing groups above about handling
1612            THEN. */
1613    
1614            case MATCH_THEN:
1615            next = ecode + GET(ecode,1);
1616            if (md->start_match_ptr < next &&
1617                (*ecode == OP_ALT || *next == OP_ALT))
1618              {
1619              rrc = MATCH_NOMATCH;
1620              break;
1621              }
1622            /* Otherwise fall through. */
1623    
1624            /* COMMIT, SKIP, PRUNE, and an uncaptured THEN cause the whole
1625            assertion to fail to match, without considering any more alternatives.
1626            Failing to match means the assertion is true. This is a consistent
1627            approach, but does not always have the same effect as in Perl. */
1628    
1629            case MATCH_COMMIT:
1630            case MATCH_SKIP:
1631            case MATCH_SKIP_ARG:
1632            case MATCH_PRUNE:
1633            do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1634            goto NEG_ASSERT_TRUE;   /* Break out of alternation loop */
1635    
1636            /* Anything else is an error */
1637    
1638            default:
1639            RRETURN(rrc);
1640          }          }
1641    
1642        /* PCRE does not allow THEN to escape beyond an assertion; it is treated        /* Continue with next branch */
       as NOMATCH. */  
1643    
       if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);  
1644        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1645        }        }
1646      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
1647    
1648      if (condassert) RRETURN(MATCH_MATCH);  /* Condition assertion */      /* All branches in the assertion failed to match. */
1649    
1650      ecode += 1 + LINK_SIZE;      NEG_ASSERT_TRUE:
1651        if (condassert) RRETURN(MATCH_MATCH);  /* Condition assertion */
1652        ecode += 1 + LINK_SIZE;                /* Continue with current branch */
1653      continue;      continue;
1654    
1655      /* Move the subject pointer back. This occurs only at the start of      /* Move the subject pointer back. This occurs only at the start of
# Line 1671  for (;;) Line 1696  for (;;)
1696        cb.version          = 2;   /* Version 1 of the callout block */        cb.version          = 2;   /* Version 1 of the callout block */
1697        cb.callout_number   = ecode[1];        cb.callout_number   = ecode[1];
1698        cb.offset_vector    = md->offset_vector;        cb.offset_vector    = md->offset_vector;
1699  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
1700        cb.subject          = (PCRE_SPTR)md->start_subject;        cb.subject          = (PCRE_SPTR)md->start_subject;
1701  #else  #elif defined COMPILE_PCRE16
1702        cb.subject          = (PCRE_SPTR16)md->start_subject;        cb.subject          = (PCRE_SPTR16)md->start_subject;
1703    #elif defined COMPILE_PCRE32
1704          cb.subject          = (PCRE_SPTR32)md->start_subject;
1705  #endif  #endif
1706        cb.subject_length   = (int)(md->end_subject - md->start_subject);        cb.subject_length   = (int)(md->end_subject - md->start_subject);
1707        cb.start_match      = (int)(mstart - md->start_subject);        cb.start_match      = (int)(mstart - md->start_subject);
# Line 1682  for (;;) Line 1709  for (;;)
1709        cb.pattern_position = GET(ecode, 2);        cb.pattern_position = GET(ecode, 2);
1710        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);
1711        cb.capture_top      = offset_top/2;        cb.capture_top      = offset_top/2;
1712        cb.capture_last     = md->capture_last;        cb.capture_last     = md->capture_last & CAPLMASK;
1713          /* Internal change requires this for API compatibility. */
1714          if (cb.capture_last == 0) cb.capture_last = -1;
1715        cb.callout_data     = md->callout_data;        cb.callout_data     = md->callout_data;
1716        cb.mark             = md->nomatch_mark;        cb.mark             = md->nomatch_mark;
1717        if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);        if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
# Line 1711  for (;;) Line 1740  for (;;)
1740      case OP_RECURSE:      case OP_RECURSE:
1741        {        {
1742        recursion_info *ri;        recursion_info *ri;
1743        int recno;        unsigned int recno;
1744    
1745        callpat = md->start_code + GET(ecode, 1);        callpat = md->start_code + GET(ecode, 1);
1746        recno = (callpat == md->start_code)? 0 :        recno = (callpat == md->start_code)? 0 :
# Line 1728  for (;;) Line 1757  for (;;)
1757        /* Add to "recursing stack" */        /* Add to "recursing stack" */
1758    
1759        new_recursive.group_num = recno;        new_recursive.group_num = recno;
1760          new_recursive.saved_capture_last = md->capture_last;
1761        new_recursive.subject_position = eptr;        new_recursive.subject_position = eptr;
1762        new_recursive.prevrec = md->recursive;        new_recursive.prevrec = md->recursive;
1763        md->recursive = &new_recursive;        md->recursive = &new_recursive;
# Line 1751  for (;;) Line 1781  for (;;)
1781              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1782    
1783        /* OK, now we can do the recursion. After processing each alternative,        /* OK, now we can do the recursion. After processing each alternative,
1784        restore the offset data. If there were nested recursions, md->recursive        restore the offset data and the last captured value. If there were nested
1785        might be changed, so reset it before looping. */        recursions, md->recursive might be changed, so reset it before looping.
1786          */
1787    
1788        DPRINTF(("Recursing into group %d\n", new_recursive.group_num));        DPRINTF(("Recursing into group %d\n", new_recursive.group_num));
1789        cbegroup = (*callpat >= OP_SBRA);        cbegroup = (*callpat >= OP_SBRA);
# Line 1763  for (;;) Line 1794  for (;;)
1794            md, eptrb, RM6);            md, eptrb, RM6);
1795          memcpy(md->offset_vector, new_recursive.offset_save,          memcpy(md->offset_vector, new_recursive.offset_save,
1796              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1797            md->capture_last = new_recursive.saved_capture_last;
1798          md->recursive = new_recursive.prevrec;          md->recursive = new_recursive.prevrec;
1799          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1800            {            {
# Line 1779  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, SKIP, PRUNE or COMMIT to escape beyond a
1815          as NOMATCH. */          recursion; they cause a NOMATCH for the entire recursion. These codes
1816            are defined in a range that can be tested for. */
1817    
1818            if (rrc >= MATCH_BACKTRACK_MIN && rrc <= MATCH_BACKTRACK_MAX)
1819              RRETURN(MATCH_NOMATCH);
1820    
1821            /* Any return code other than NOMATCH is an error. */
1822    
1823          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)          if (rrc != MATCH_NOMATCH)
1824            {            {
1825            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1826            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
# Line 1912  for (;;) Line 1950  for (;;)
1950    
1951        /* Deal with capturing */        /* Deal with capturing */
1952    
1953        md->capture_last = number;        md->capture_last = (md->capture_last & OVFLMASK) | number;
1954        if (offset >= md->offset_max) md->offset_overflow = TRUE; else        if (offset >= md->offset_max) md->capture_last |= OVFLBIT; else
1955          {          {
1956          /* If offset is greater than offset_top, it means that we are          /* If offset is greater than offset_top, it means that we are
1957          "skipping" a capturing group, and that group's offsets must be marked          "skipping" a capturing group, and that group's offsets must be marked
# Line 1993  for (;;) Line 2031  for (;;)
2031          }          }
2032        if (*prev >= OP_SBRA)    /* Could match an empty string */        if (*prev >= OP_SBRA)    /* Could match an empty string */
2033          {          {
         md->match_function_type = MATCH_CBEGROUP;  
2034          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);
2035          RRETURN(rrc);          RRETURN(rrc);
2036          }          }
# Line 2002  for (;;) Line 2039  for (;;)
2039        }        }
2040      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
2041        {        {
       if (*prev >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;  
2042        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);
2043        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;
2044        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
# Line 2059  for (;;) Line 2095  for (;;)
2095    
2096      case OP_DOLLM:      case OP_DOLLM:
2097      if (eptr < md->end_subject)      if (eptr < md->end_subject)
2098        { if (!IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH); }        {
2099          if (!IS_NEWLINE(eptr))
2100            {
2101            if (md->partial != 0 &&
2102                eptr + 1 >= md->end_subject &&
2103                NLBLOCK->nltype == NLTYPE_FIXED &&
2104                NLBLOCK->nllen == 2 &&
2105                RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2106              {
2107              md->hitend = TRUE;
2108              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2109              }
2110            RRETURN(MATCH_NOMATCH);
2111            }
2112          }
2113      else      else
2114        {        {
2115        if (md->noteol) RRETURN(MATCH_NOMATCH);        if (md->noteol) RRETURN(MATCH_NOMATCH);
# Line 2091  for (;;) Line 2141  for (;;)
2141      ASSERT_NL_OR_EOS:      ASSERT_NL_OR_EOS:
2142      if (eptr < md->end_subject &&      if (eptr < md->end_subject &&
2143          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))
2144          {
2145          if (md->partial != 0 &&
2146              eptr + 1 >= md->end_subject &&
2147              NLBLOCK->nltype == NLTYPE_FIXED &&
2148              NLBLOCK->nllen == 2 &&
2149              RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2150            {
2151            md->hitend = TRUE;
2152            if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2153            }
2154        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2155          }
2156    
2157      /* Either at end of string or \n before end. */      /* Either at end of string or \n before end. */
2158    
# Line 2219  for (;;) Line 2280  for (;;)
2280        }        }
2281      break;      break;
2282    
2283      /* Match a single character type; inline for speed */      /* Match any single character type except newline; have to take care with
2284        CRLF newlines and partial matching. */
2285    
2286      case OP_ANY:      case OP_ANY:
2287      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
2288        if (md->partial != 0 &&
2289            eptr + 1 >= md->end_subject &&
2290            NLBLOCK->nltype == NLTYPE_FIXED &&
2291            NLBLOCK->nllen == 2 &&
2292            RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2293          {
2294          md->hitend = TRUE;
2295          if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2296          }
2297    
2298      /* Fall through */      /* Fall through */
2299    
2300        /* Match any single character whatsoever. */
2301    
2302      case OP_ALLANY:      case OP_ALLANY:
2303      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 */
2304        {                            /* not be updated before SCHECK_PARTIAL. */        {                            /* not be updated before SCHECK_PARTIAL. */
# Line 2364  for (;;) Line 2438  for (;;)
2438        {        {
2439        default: RRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2440    
2441        case 0x000d:        case CHAR_CR:
2442        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr >= md->end_subject)
2443            {
2444            SCHECK_PARTIAL();
2445            }
2446          else if (RAWUCHARTEST(eptr) == CHAR_LF) eptr++;
2447        break;        break;
2448    
2449        case 0x000a:        case CHAR_LF:
2450        break;        break;
2451    
2452        case 0x000b:        case CHAR_VT:
2453        case 0x000c:        case CHAR_FF:
2454        case 0x0085:        case CHAR_NEL:
2455    #ifndef EBCDIC
2456        case 0x2028:        case 0x2028:
2457        case 0x2029:        case 0x2029:
2458    #endif  /* Not EBCDIC */
2459        if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);        if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
2460        break;        break;
2461        }        }
# Line 2391  for (;;) Line 2471  for (;;)
2471      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2472      switch(c)      switch(c)
2473        {        {
2474          HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
2475        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 */  
       RRETURN(MATCH_NOMATCH);  
2476        }        }
2477      ecode++;      ecode++;
2478      break;      break;
# Line 2425  for (;;) Line 2486  for (;;)
2486      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2487      switch(c)      switch(c)
2488        {        {
2489          HSPACE_CASES: break;  /* Byte and multibyte cases */
2490        default: RRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
       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 */  
       break;  
2491        }        }
2492      ecode++;      ecode++;
2493      break;      break;
# Line 2459  for (;;) Line 2501  for (;;)
2501      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2502      switch(c)      switch(c)
2503        {        {
2504          VSPACE_CASES: RRETURN(MATCH_NOMATCH);
2505        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 */  
       RRETURN(MATCH_NOMATCH);  
2506        }        }
2507      ecode++;      ecode++;
2508      break;      break;
# Line 2481  for (;;) Line 2516  for (;;)
2516      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2517      switch(c)      switch(c)
2518        {        {
2519          VSPACE_CASES: break;
2520        default: RRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
       case 0x0a:      /* LF */  
       case 0x0b:      /* VT */  
       case 0x0c:      /* FF */  
       case 0x0d:      /* CR */  
       case 0x85:      /* NEL */  
       case 0x2028:    /* LINE SEPARATOR */  
       case 0x2029:    /* PARAGRAPH SEPARATOR */  
       break;  
2521        }        }
2522      ecode++;      ecode++;
2523      break;      break;
# Line 2507  for (;;) Line 2535  for (;;)
2535        }        }
2536      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2537        {        {
2538          const pcre_uint32 *cp;
2539        const ucd_record *prop = GET_UCD(c);        const ucd_record *prop = GET_UCD(c);
2540    
2541        switch(ecode[1])        switch(ecode[1])
# Line 2545  for (;;) Line 2574  for (;;)
2574            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2575          break;          break;
2576    
2577          case PT_SPACE:    /* Perl space */          /* Perl space used to exclude VT, but from Perl 5.18 it is included,
2578          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||          which means that Perl space and POSIX space are now identical. PCRE
2579               c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)          was changed at release 8.34. */
                == (op == OP_NOTPROP))  
           RRETURN(MATCH_NOMATCH);  
         break;  
2580    
2581            case PT_SPACE:    /* Perl space */
2582          case PT_PXSPACE:  /* POSIX space */          case PT_PXSPACE:  /* POSIX space */
2583          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||          switch(c)
2584               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||            {
2585               c == CHAR_FF || c == CHAR_CR)            HSPACE_CASES:
2586                 == (op == OP_NOTPROP))            VSPACE_CASES:
2587            RRETURN(MATCH_NOMATCH);            if (op == OP_NOTPROP) RRETURN(MATCH_NOMATCH);
2588              break;
2589    
2590              default:
2591              if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z) ==
2592                (op == OP_NOTPROP)) RRETURN(MATCH_NOMATCH);
2593              break;
2594              }
2595          break;          break;
2596    
2597          case PT_WORD:          case PT_WORD:
# Line 2567  for (;;) Line 2601  for (;;)
2601            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2602          break;          break;
2603    
2604            case PT_CLIST:
2605            cp = PRIV(ucd_caseless_sets) + ecode[2];
2606            for (;;)
2607              {
2608              if (c < *cp)
2609                { if (op == OP_PROP) { RRETURN(MATCH_NOMATCH); } else break; }
2610              if (c == *cp++)
2611                { if (op == OP_PROP) break; else { RRETURN(MATCH_NOMATCH); } }
2612              }
2613            break;
2614    
2615            case PT_UCNC:
2616            if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
2617                 c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
2618                 c >= 0xe000) == (op == OP_NOTPROP))
2619              RRETURN(MATCH_NOMATCH);
2620            break;
2621    
2622          /* This should never occur */          /* This should never occur */
2623    
2624          default:          default:
# Line 2586  for (;;) Line 2638  for (;;)
2638        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2639        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2640        }        }
2641      GETCHARINCTEST(c, eptr);      else
     if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
     while (eptr < md->end_subject)  
2642        {        {
2643        int len = 1;        int lgb, rgb;
2644        if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }        GETCHARINCTEST(c, eptr);
2645        if (UCD_CATEGORY(c) != ucp_M) break;        lgb = UCD_GRAPHBREAK(c);
2646        eptr += len;        while (eptr < md->end_subject)
2647            {
2648            int len = 1;
2649            if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
2650            rgb = UCD_GRAPHBREAK(c);
2651            if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
2652            lgb = rgb;
2653            eptr += len;
2654            }
2655        }        }
2656        CHECK_PARTIAL();
2657      ecode++;      ecode++;
2658      break;      break;
2659  #endif  #endif  /* SUPPORT_UCP */
2660    
2661    
2662      /* 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 2606  for (;;) Line 2665  for (;;)
2665      similar code to character type repeats - written out again for speed.      similar code to character type repeats - written out again for speed.
2666      However, if the referenced string is the empty string, always treat      However, if the referenced string is the empty string, always treat
2667      it as matched, any number of times (otherwise there could be infinite      it as matched, any number of times (otherwise there could be infinite
2668      loops). */      loops). If the reference is unset, there are two possibilities:
   
     case OP_REF:  
     case OP_REFI:  
     caseless = op == OP_REFI;  
     offset = GET2(ecode, 1) << 1;               /* Doubled ref number */  
     ecode += 1 + IMM2_SIZE;  
   
     /* If the reference is unset, there are two possibilities:  
2669    
2670      (a) In the default, Perl-compatible state, set the length negative;      (a) In the default, Perl-compatible state, set the length negative;
2671      this ensures that every attempt at a match fails. We can't just fail      this ensures that every attempt at a match fails. We can't just fail
# Line 2624  for (;;) Line 2675  for (;;)
2675      so that the back reference matches an empty string.      so that the back reference matches an empty string.
2676    
2677      Otherwise, set the length to the length of what was matched by the      Otherwise, set the length to the length of what was matched by the
2678      referenced subpattern. */      referenced subpattern.
2679    
2680        The OP_REF and OP_REFI opcodes are used for a reference to a numbered group
2681        or to a non-duplicated named group. For a duplicated named group, OP_DNREF
2682        and OP_DNREFI are used. In this case we must scan the list of groups to
2683        which the name refers, and use the first one that is set. */
2684    
2685        case OP_DNREF:
2686        case OP_DNREFI:
2687        caseless = op == OP_DNREFI;
2688          {
2689          int count = GET2(ecode, 1+IMM2_SIZE);
2690          pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
2691          ecode += 1 + 2*IMM2_SIZE;
2692    
2693          while (count-- > 0)
2694            {
2695            offset = GET2(slot, 0) << 1;
2696            if (offset < offset_top && md->offset_vector[offset] >= 0) break;
2697            slot += md->name_entry_size;
2698            }
2699          if (count < 0)
2700            length = (md->jscript_compat)? 0 : -1;
2701          else
2702            length = md->offset_vector[offset+1] - md->offset_vector[offset];
2703          }
2704        goto REF_REPEAT;
2705    
2706        case OP_REF:
2707        case OP_REFI:
2708        caseless = op == OP_REFI;
2709        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
2710        ecode += 1 + IMM2_SIZE;
2711      if (offset >= offset_top || md->offset_vector[offset] < 0)      if (offset >= offset_top || md->offset_vector[offset] < 0)
2712        length = (md->jscript_compat)? 0 : -1;        length = (md->jscript_compat)? 0 : -1;
2713      else      else
# Line 2633  for (;;) Line 2715  for (;;)
2715    
2716      /* Set up for repetition, or handle the non-repeated case */      /* Set up for repetition, or handle the non-repeated case */
2717    
2718        REF_REPEAT:
2719      switch (*ecode)      switch (*ecode)
2720        {        {
2721        case OP_CRSTAR:        case OP_CRSTAR:
# Line 2660  for (;;) Line 2743  for (;;)
2743        default:               /* No repeat follows */        default:               /* No repeat follows */
2744        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)
2745          {          {
2746            if (length == -2) eptr = md->end_subject;   /* Partial match */
2747          CHECK_PARTIAL();          CHECK_PARTIAL();
2748          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2749          }          }
# Line 2685  for (;;) Line 2769  for (;;)
2769        int slength;        int slength;
2770        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2771          {          {
2772            if (slength == -2) eptr = md->end_subject;   /* Partial match */
2773          CHECK_PARTIAL();          CHECK_PARTIAL();
2774          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2775          }          }
# Line 2708  for (;;) Line 2793  for (;;)
2793          if (fi >= max) RRETURN(MATCH_NOMATCH);          if (fi >= max) RRETURN(MATCH_NOMATCH);
2794          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2795            {            {
2796              if (slength == -2) eptr = md->end_subject;   /* Partial match */
2797            CHECK_PARTIAL();            CHECK_PARTIAL();
2798            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2799            }            }
# Line 2726  for (;;) Line 2812  for (;;)
2812          int slength;          int slength;
2813          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2814            {            {
2815            CHECK_PARTIAL();            /* Can't use CHECK_PARTIAL because we don't want to update eptr in
2816              the soft partial matching case. */
2817    
2818              if (slength == -2 && md->partial != 0 &&
2819                  md->end_subject > md->start_used_ptr)
2820                {
2821                md->hitend = TRUE;
2822                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2823                }
2824            break;            break;
2825            }            }
2826          eptr += slength;          eptr += slength;
2827          }          }
2828    
2829        while (eptr >= pp)        while (eptr >= pp)
2830          {          {
2831          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);
# Line 2769  for (;;) Line 2864  for (;;)
2864          case OP_CRMINPLUS:          case OP_CRMINPLUS:
2865          case OP_CRQUERY:          case OP_CRQUERY:
2866          case OP_CRMINQUERY:          case OP_CRMINQUERY:
2867            case OP_CRPOSSTAR:
2868            case OP_CRPOSPLUS:
2869            case OP_CRPOSQUERY:
2870          c = *ecode++ - OP_CRSTAR;          c = *ecode++ - OP_CRSTAR;
2871          minimize = (c & 1) != 0;          if (c < OP_CRPOSSTAR - OP_CRSTAR) minimize = (c & 1) != 0;
2872            else possessive = TRUE;
2873          min = rep_min[c];                 /* Pick up values from tables; */          min = rep_min[c];                 /* Pick up values from tables; */
2874          max = rep_max[c];                 /* zero for max => infinity */          max = rep_max[c];                 /* zero for max => infinity */
2875          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 2778  for (;;) Line 2877  for (;;)
2877    
2878          case OP_CRRANGE:          case OP_CRRANGE:
2879          case OP_CRMINRANGE:          case OP_CRMINRANGE:
2880            case OP_CRPOSRANGE:
2881          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
2882            possessive = (*ecode == OP_CRPOSRANGE);
2883          min = GET2(ecode, 1);          min = GET2(ecode, 1);
2884          max = GET2(ecode, 1 + IMM2_SIZE);          max = GET2(ecode, 1 + IMM2_SIZE);
2885          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 2920  for (;;) Line 3021  for (;;)
3021                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
3022              eptr += len;              eptr += len;
3023              }              }
3024    
3025              if (possessive) continue;    /* No backtracking */
3026    
3027            for (;;)            for (;;)
3028              {              {
3029              RMATCH(eptr, ecode, offset_top, md, eptrb, RM18);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM18);
# Line 2950  for (;;) Line 3054  for (;;)
3054                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
3055              eptr++;              eptr++;
3056              }              }
3057    
3058              if (possessive) continue;    /* No backtracking */
3059    
3060            while (eptr >= pp)            while (eptr >= pp)
3061              {              {
3062              RMATCH(eptr, ecode, offset_top, md, eptrb, RM19);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM19);
# Line 2983  for (;;) Line 3090  for (;;)
3090          case OP_CRMINPLUS:          case OP_CRMINPLUS:
3091          case OP_CRQUERY:          case OP_CRQUERY:
3092          case OP_CRMINQUERY:          case OP_CRMINQUERY:
3093            case OP_CRPOSSTAR:
3094            case OP_CRPOSPLUS:
3095            case OP_CRPOSQUERY:
3096          c = *ecode++ - OP_CRSTAR;          c = *ecode++ - OP_CRSTAR;
3097          minimize = (c & 1) != 0;          if (c < OP_CRPOSSTAR - OP_CRSTAR) minimize = (c & 1) != 0;
3098            else possessive = TRUE;
3099          min = rep_min[c];                 /* Pick up values from tables; */          min = rep_min[c];                 /* Pick up values from tables; */
3100          max = rep_max[c];                 /* zero for max => infinity */          max = rep_max[c];                 /* zero for max => infinity */
3101          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 2992  for (;;) Line 3103  for (;;)
3103    
3104          case OP_CRRANGE:          case OP_CRRANGE:
3105          case OP_CRMINRANGE:          case OP_CRMINRANGE:
3106            case OP_CRPOSRANGE:
3107          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
3108            possessive = (*ecode == OP_CRPOSRANGE);
3109          min = GET2(ecode, 1);          min = GET2(ecode, 1);
3110          max = GET2(ecode, 1 + IMM2_SIZE);          max = GET2(ecode, 1 + IMM2_SIZE);
3111          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 3064  for (;;) Line 3177  for (;;)
3177            if (!PRIV(xclass)(c, data, utf)) break;            if (!PRIV(xclass)(c, data, utf)) break;
3178            eptr += len;            eptr += len;
3179            }            }
3180    
3181            if (possessive) continue;    /* No backtracking */
3182    
3183          for(;;)          for(;;)
3184            {            {
3185            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);
# Line 3094  for (;;) Line 3210  for (;;)
3210          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
3211          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3212          }          }
3213        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);        while (length-- > 0) if (*ecode++ != RAWUCHARINC(eptr)) RRETURN(MATCH_NOMATCH);
3214        }        }
3215      else      else
3216  #endif  #endif
# Line 3134  for (;;) Line 3250  for (;;)
3250    
3251        if (fc < 128)        if (fc < 128)
3252          {          {
3253          if (md->lcc[fc]          pcre_uint32 cc = RAWUCHAR(eptr);
3254              != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);          if (md->lcc[fc] != TABLE_GET(cc, md->lcc, cc)) RRETURN(MATCH_NOMATCH);
3255          ecode++;          ecode++;
3256          eptr++;          eptr++;
3257          }          }
# Line 3146  for (;;) Line 3262  for (;;)
3262    
3263        else        else
3264          {          {
3265          unsigned int dc;          pcre_uint32 dc;
3266          GETCHARINC(dc, eptr);          GETCHARINC(dc, eptr);
3267          ecode += length;          ecode += length;
3268    
# Line 3239  for (;;) Line 3355  for (;;)
3355      max = rep_max[c];                 /* zero for max => infinity */      max = rep_max[c];                 /* zero for max => infinity */
3356      if (max == 0) max = INT_MAX;      if (max == 0) max = INT_MAX;
3357    
3358      /* Common code for all repeated single-character matches. */      /* Common code for all repeated single-character matches. We first check
3359        for the minimum number of characters. If the minimum equals the maximum, we
3360        are done. Otherwise, if minimizing, check the rest of the pattern for a
3361        match; if there isn't one, advance up to the maximum, one character at a
3362        time.
3363    
3364        If maximizing, advance up to the maximum number of matching characters,
3365        until eptr is past the end of the maximum run. If possessive, we are
3366        then done (no backing up). Otherwise, match at this position; anything
3367        other than no match is immediately returned. For nomatch, back up one
3368        character, unless we are matching \R and the last thing matched was
3369        \r\n, in which case, back up two bytes. When we reach the first optional
3370        character position, we can save stack by doing a tail recurse.
3371    
3372        The various UTF/non-UTF and caseful/caseless cases are handled separately,
3373        for speed. */
3374    
3375      REPEATCHAR:      REPEATCHAR:
3376  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
# Line 3256  for (;;) Line 3387  for (;;)
3387        if (length > 1)        if (length > 1)
3388          {          {
3389  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3390          unsigned int othercase;          pcre_uint32 othercase;
3391          if (op >= OP_STARI &&     /* Caseless */          if (op >= OP_STARI &&     /* Caseless */
3392              (othercase = UCD_OTHERCASE(fc)) != fc)              (othercase = UCD_OTHERCASE(fc)) != fc)
3393            oclength = PRIV(ord2utf)(othercase, occhars);            oclength = PRIV(ord2utf)(othercase, occhars);
# Line 3323  for (;;) Line 3454  for (;;)
3454                }                }
3455              }              }
3456    
3457            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
   
3458            for(;;)            for(;;)
3459              {              {
3460                if (eptr == pp) goto TAIL_RECURSE;
3461              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);
3462              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
             if (eptr == pp) { RRETURN(MATCH_NOMATCH); }  
3463  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3464              eptr--;              eptr--;
3465              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 3360  for (;;) Line 3490  for (;;)
3490      maximizing, find the maximum number of characters and work backwards. */      maximizing, find the maximum number of characters and work backwards. */
3491    
3492      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,
3493        max, eptr));        max, (char *)eptr));
3494    
3495      if (op >= OP_STARI)  /* Caseless */      if (op >= OP_STARI)  /* Caseless */
3496        {        {
# Line 3383  for (;;) Line 3513  for (;;)
3513    
3514        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
3515          {          {
3516            pcre_uint32 cc;                 /* Faster than pcre_uchar */
3517          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3518            {            {
3519            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3520            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3521            }            }
3522          if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);          cc = RAWUCHARTEST(eptr);
3523            if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
3524          eptr++;          eptr++;
3525          }          }
3526        if (min == max) continue;        if (min == max) continue;
# Line 3396  for (;;) Line 3528  for (;;)
3528          {          {
3529          for (fi = min;; fi++)          for (fi = min;; fi++)
3530            {            {
3531              pcre_uint32 cc;               /* Faster than pcre_uchar */
3532            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);
3533            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3534            if (fi >= max) RRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
# Line 3404  for (;;) Line 3537  for (;;)
3537              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3538              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3539              }              }
3540            if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);            cc = RAWUCHARTEST(eptr);
3541              if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
3542            eptr++;            eptr++;
3543            }            }
3544          /* Control never gets here */          /* Control never gets here */
# Line 3414  for (;;) Line 3548  for (;;)
3548          pp = eptr;          pp = eptr;
3549          for (i = min; i < max; i++)          for (i = min; i < max; i++)
3550            {            {
3551              pcre_uint32 cc;               /* Faster than pcre_uchar */
3552            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3553              {              {
3554              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3555              break;              break;
3556              }              }
3557            if (fc != *eptr && foc != *eptr) break;            cc = RAWUCHARTEST(eptr);
3558              if (fc != cc && foc != cc) break;
3559            eptr++;            eptr++;
3560            }            }
3561            if (possessive) continue;       /* No backtracking */
3562          if (possessive) continue;          for (;;)
   
         while (eptr >= pp)  
3563            {            {
3564              if (eptr == pp) goto TAIL_RECURSE;
3565            RMATCH(eptr, ecode, offset_top, md, eptrb, RM25);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM25);
3566            eptr--;            eptr--;
3567            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3568            }            }
3569          RRETURN(MATCH_NOMATCH);          /* Control never gets here */
3570          }          }
       /* Control never gets here */  
3571        }        }
3572    
3573      /* Caseful comparisons (includes all multi-byte characters) */      /* Caseful comparisons (includes all multi-byte characters) */
# Line 3447  for (;;) Line 3581  for (;;)
3581            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3582            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3583            }            }
3584          if (fc != *eptr++) RRETURN(MATCH_NOMATCH);          if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
3585          }          }
3586    
3587        if (min == max) continue;        if (min == max) continue;
# Line 3464  for (;;) Line 3598  for (;;)
3598              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3599              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3600              }              }
3601            if (fc != *eptr++) RRETURN(MATCH_NOMATCH);            if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
3602            }            }
3603          /* Control never gets here */          /* Control never gets here */
3604          }          }
# Line 3478  for (;;) Line 3612  for (;;)
3612              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3613              break;              break;
3614              }              }
3615            if (fc != *eptr) break;            if (fc != RAWUCHARTEST(eptr)) break;
3616            eptr++;            eptr++;
3617            }            }
3618          if (possessive) continue;          if (possessive) continue;    /* No backtracking */
3619            for (;;)
         while (eptr >= pp)  
3620            {            {
3621              if (eptr == pp) goto TAIL_RECURSE;
3622            RMATCH(eptr, ecode, offset_top, md, eptrb, RM27);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM27);
3623            eptr--;            eptr--;
3624            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3625            }            }
3626          RRETURN(MATCH_NOMATCH);          /* Control never gets here */
3627          }          }
3628        }        }
3629      /* Control never gets here */      /* Control never gets here */
# Line 3504  for (;;) Line 3638  for (;;)
3638        SCHECK_PARTIAL();        SCHECK_PARTIAL();
3639        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
3640        }        }
     ecode++;  
     GETCHARINCTEST(c, eptr);  
     if (op == OP_NOTI)         /* The caseless case */  
       {  
       register int ch, och;  
       ch = *ecode++;  
 #ifdef COMPILE_PCRE8  
       /* ch must be < 128 if UTF is enabled. */  
       och = md->fcc[ch];  
 #else  
3641  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3642  #ifdef SUPPORT_UCP      if (utf)
3643        if (utf && ch > 127)        {
3644          och = UCD_OTHERCASE(ch);        register pcre_uint32 ch, och;
3645  #else  
3646        if (utf && ch > 127)        ecode++;
3647          och = ch;        GETCHARINC(ch, ecode);
3648  #endif /* SUPPORT_UCP */        GETCHARINC(c, eptr);
3649        else  
3650  #endif /* SUPPORT_UTF */        if (op == OP_NOT)
3651          och = TABLE_GET(ch, md->fcc, ch);          {
3652  #endif /* COMPILE_PCRE8 */          if (ch == c) RRETURN(MATCH_NOMATCH);
3653        if (ch == c || och == c) RRETURN(MATCH_NOMATCH);          }
3654        }        else
3655      else    /* Caseful */          {
3656    #ifdef SUPPORT_UCP
3657            if (ch > 127)
3658              och = UCD_OTHERCASE(ch);
3659    #else
3660            if (ch > 127)
3661              och = ch;
3662    #endif /* SUPPORT_UCP */
3663            else
3664              och = TABLE_GET(ch, md->fcc, ch);
3665            if (ch == c || och == c) RRETURN(MATCH_NOMATCH);
3666            }
3667          }
3668        else
3669    #endif
3670        {        {
3671        if (*ecode++ == c) RRETURN(MATCH_NOMATCH);        register pcre_uint32 ch = ecode[1];
3672          c = *eptr++;
3673          if (ch == c || (op == OP_NOTI && TABLE_GET(ch, md->fcc, ch) == c))
3674            RRETURN(MATCH_NOMATCH);
3675          ecode += 2;
3676        }        }
3677      break;      break;
3678    
# Line 3610  for (;;) Line 3752  for (;;)
3752      /* Common code for all repeated single-byte matches. */      /* Common code for all repeated single-byte matches. */
3753    
3754      REPEATNOTCHAR:      REPEATNOTCHAR:
3755      fc = *ecode++;      GETCHARINCTEST(fc, ecode);
3756    
3757      /* The code is duplicated for the caseless and caseful cases, for speed,      /* The code is duplicated for the caseless and caseful cases, for speed,
3758      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 3621  for (;;) Line 3763  for (;;)
3763      characters and work backwards. */      characters and work backwards. */
3764    
3765      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,
3766        max, eptr));        max, (char *)eptr));
3767    
3768      if (op >= OP_NOTSTARI)     /* Caseless */      if (op >= OP_NOTSTARI)     /* Caseless */
3769        {        {
 #ifdef COMPILE_PCRE8  
       /* fc must be < 128 if UTF is enabled. */  
       foc = md->fcc[fc];  
 #else  
3770  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3771  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3772        if (utf && fc > 127)        if (utf && fc > 127)
# Line 3640  for (;;) Line 3778  for (;;)
3778        else        else
3779  #endif /* SUPPORT_UTF */  #endif /* SUPPORT_UTF */
3780          foc = TABLE_GET(fc, md->fcc, fc);          foc = TABLE_GET(fc, md->fcc, fc);
 #endif /* COMPILE_PCRE8 */  
3781    
3782  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3783        if (utf)        if (utf)
3784          {          {
3785          register unsigned int d;          register pcre_uint32 d;
3786          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3787            {            {
3788            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
# Line 3654  for (;;) Line 3791  for (;;)
3791              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3792              }              }
3793            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3794            if (fc == d || foc == d) RRETURN(MATCH_NOMATCH);            if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
3795            }            }
3796          }          }
3797        else        else
3798  #endif  #endif  /* SUPPORT_UTF */
3799        /* Not UTF mode */        /* Not UTF mode */
3800          {          {
3801          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3680  for (;;) Line 3817  for (;;)
3817  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3818          if (utf)          if (utf)
3819            {            {
3820            register unsigned int d;            register pcre_uint32 d;
3821            for (fi = min;; fi++)            for (fi = min;; fi++)
3822              {              {
3823              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);
# Line 3692  for (;;) Line 3829  for (;;)
3829                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3830                }                }
3831              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3832              if (fc == d || foc == d) RRETURN(MATCH_NOMATCH);              if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
3833              }              }
3834            }            }
3835          else          else
3836  #endif  #endif  /*SUPPORT_UTF */
3837          /* Not UTF mode */          /* Not UTF mode */
3838            {            {
3839            for (fi = min;; fi++)            for (fi = min;; fi++)
# Line 3725  for (;;) Line 3862  for (;;)
3862  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3863          if (utf)          if (utf)
3864            {            {
3865            register unsigned int d;            register pcre_uint32 d;
3866            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3867              {              {
3868              int len = 1;              int len = 1;
# Line 3735  for (;;) Line 3872  for (;;)
3872                break;                break;
3873                }                }
3874              GETCHARLEN(d, eptr, len);              GETCHARLEN(d, eptr, len);
3875              if (fc == d || foc == d) break;              if (fc == d || (unsigned int)foc == d) break;
3876              eptr += len;              eptr += len;
3877              }              }
3878            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
3879            for(;;)            for(;;)
3880              {              {
3881                if (eptr == pp) goto TAIL_RECURSE;
3882              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);
3883              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3884              if (eptr-- == pp) break;        /* Stop if tried at original pos */              eptr--;
3885              BACKCHAR(eptr);              BACKCHAR(eptr);
3886              }              }
3887            }            }
3888          else          else
3889  #endif  #endif  /* SUPPORT_UTF */
3890          /* Not UTF mode */          /* Not UTF mode */
3891            {            {
3892            for (i = min; i < max; i++)            for (i = min; i < max; i++)
# Line 3761  for (;;) Line 3899  for (;;)
3899              if (fc == *eptr || foc == *eptr) break;              if (fc == *eptr || foc == *eptr) break;
3900              eptr++;              eptr++;
3901              }              }
3902            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
3903            while (eptr >= pp)            for (;;)
3904              {              {
3905                if (eptr == pp) goto TAIL_RECURSE;
3906              RMATCH(eptr, ecode, offset_top, md, eptrb, RM31);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM31);
3907              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3908              eptr--;              eptr--;
3909              }              }
3910            }            }
3911            /* Control never gets here */
         RRETURN(MATCH_NOMATCH);  
3912          }          }
       /* Control never gets here */  
3913        }        }
3914    
3915      /* Caseful comparisons */      /* Caseful comparisons */
# Line 3782  for (;;) Line 3919  for (;;)
3919  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3920        if (utf)        if (utf)
3921          {          {
3922          register unsigned int d;          register pcre_uint32 d;
3923          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3924            {            {
3925            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
# Line 3816  for (;;) Line 3953  for (;;)
3953  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3954          if (utf)          if (utf)
3955            {            {
3956            register unsigned int d;            register pcre_uint32 d;
3957            for (fi = min;; fi++)            for (fi = min;; fi++)
3958              {              {
3959              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);
# Line 3860  for (;;) Line 3997  for (;;)
3997  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3998          if (utf)          if (utf)
3999            {            {
4000            register unsigned int d;            register pcre_uint32 d;
4001            for (i = min; i < max; i++)            for (i = min; i < max; i++)
4002              {              {
4003              int len = 1;              int len = 1;
# Line 3873  for (;;) Line 4010  for (;;)
4010              if (fc == d) break;              if (fc == d) break;
4011              eptr += len;              eptr += len;
4012              }              }
4013            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
4014            for(;;)            for(;;)
4015              {              {
4016                if (eptr == pp) goto TAIL_RECURSE;
4017              RMATCH(eptr, ecode, offset_top, md, eptrb, RM34);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM34);
4018              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4019              if (eptr-- == pp) break;        /* Stop if tried at original pos */              eptr--;
4020              BACKCHAR(eptr);              BACKCHAR(eptr);
4021              }              }
4022            }            }
# Line 3896  for (;;) Line 4034  for (;;)
4034              if (fc == *eptr) break;              if (fc == *eptr) break;
4035              eptr++;              eptr++;
4036              }              }
4037            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
4038            while (eptr >= pp)            for (;;)
4039              {              {
4040                if (eptr == pp) goto TAIL_RECURSE;
4041              RMATCH(eptr, ecode, offset_top, md, eptrb, RM35);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM35);
4042              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4043              eptr--;              eptr--;
4044              }              }
4045            }            }
4046            /* Control never gets here */
         RRETURN(MATCH_NOMATCH);  
4047          }          }
4048        }        }
4049      /* Control never gets here */      /* Control never gets here */
# Line 4087  for (;;) Line 4225  for (;;)
4225              }              }
4226            break;            break;
4227    
4228              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
4229              which means that Perl space and POSIX space are now identical. PCRE
4230              was changed at release 8.34. */
4231    
4232            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
4233              case PT_PXSPACE:  /* POSIX space */
4234            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4235              {              {
4236              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4096  for (;;) Line 4239  for (;;)
4239                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4240                }                }
4241              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4242              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              switch(c)
4243                   c == CHAR_FF || c == CHAR_CR)                {
4244                     == prop_fail_result)                HSPACE_CASES:
4245                RRETURN(MATCH_NOMATCH);                VSPACE_CASES:
4246                  if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4247                  break;
4248    
4249                  default:
4250                  if ((UCD_CATEGORY(c) == ucp_Z) == prop_fail_result)
4251                    RRETURN(MATCH_NOMATCH);
4252                  break;
4253                  }
4254              }              }
4255            break;            break;
4256    
4257            case PT_PXSPACE:  /* POSIX space */            case PT_WORD:
4258            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4259              {              {
4260                int category;
4261              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4262                {                {
4263                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4264                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4265                }                }
4266              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4267              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              category = UCD_CATEGORY(c);
4268                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)
4269                     == prop_fail_result)                     == prop_fail_result)
4270                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4271              }              }
4272            break;            break;
4273    
4274            case PT_WORD:            case PT_CLIST:
4275            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4276              {              {
4277              int category;              const pcre_uint32 *cp;
4278              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4279                {                {
4280                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4281                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4282                }                }
4283              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4284              category = UCD_CATEGORY(c);              cp = PRIV(ucd_caseless_sets) + prop_value;
4285              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)              for (;;)
4286                     == prop_fail_result)                {
4287                  if (c < *cp)
4288                    { if (prop_fail_result) break; else { RRETURN(MATCH_NOMATCH); } }
4289                  if (c == *cp++)
4290                    { if (prop_fail_result) { RRETURN(MATCH_NOMATCH); } else break; }
4291                  }
4292                }
4293              break;
4294    
4295              case PT_UCNC:
4296              for (i = 1; i <= min; i++)
4297                {
4298                if (eptr >= md->end_subject)
4299                  {
4300                  SCHECK_PARTIAL();
4301                  RRETURN(MATCH_NOMATCH);
4302                  }
4303                GETCHARINCTEST(c, eptr);
4304                if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
4305                     c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
4306                     c >= 0xe000) == prop_fail_result)
4307                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4308              }              }
4309            break;            break;
# Line 4155  for (;;) Line 4327  for (;;)
4327              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4328              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4329              }              }
4330            GETCHARINCTEST(c, eptr);            else
           if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
           while (eptr < md->end_subject)  
4331              {              {
4332              int len = 1;              int lgb, rgb;
4333              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
4334              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
4335              eptr += len;             while (eptr < md->end_subject)
4336                  {
4337                  int len = 1;
4338                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
4339                  rgb = UCD_GRAPHBREAK(c);
4340                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
4341                  lgb = rgb;
4342                  eptr += len;
4343                  }
4344              }              }
4345              CHECK_PARTIAL();
4346            }            }
4347          }          }
4348    
# Line 4184  for (;;) Line 4363  for (;;)
4363              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4364              }              }
4365            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4366              if (md->partial != 0 &&
4367                  eptr + 1 >= md->end_subject &&
4368                  NLBLOCK->nltype == NLTYPE_FIXED &&
4369                  NLBLOCK->nllen == 2 &&
4370                  RAWUCHAR(eptr) == NLBLOCK->nl[0])
4371                {
4372                md->hitend = TRUE;
4373                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4374                }
4375            eptr++;            eptr++;
4376            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4377            }            }
# Line 4220  for (;;) Line 4408  for (;;)
4408              {              {
4409              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4410    
4411              case 0x000d:              case CHAR_CR:
4412              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && RAWUCHAR(eptr) == CHAR_LF) eptr++;
4413              break;              break;
4414    
4415              case 0x000a:              case CHAR_LF:
4416              break;              break;
4417    
4418              case 0x000b:              case CHAR_VT:
4419              case 0x000c:              case CHAR_FF:
4420              case 0x0085:              case CHAR_NEL:
4421    #ifndef EBCDIC
4422              case 0x2028:              case 0x2028:
4423              case 0x2029:              case 0x2029:
4424    #endif  /* Not EBCDIC */
4425              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4426              break;              break;
4427              }              }
# Line 4249  for (;;) Line 4439  for (;;)
4439            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4440            switch(c)            switch(c)
4441              {              {
4442                HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
4443              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 */  
             RRETURN(MATCH_NOMATCH);  
4444              }              }
4445            }            }
4446          break;          break;
# Line 4285  for (;;) Line 4456  for (;;)
4456            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4457            switch(c)            switch(c)
4458              {              {
4459                HSPACE_CASES: break;  /* Byte and multibyte cases */
4460              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
             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 */  
             break;  
4461              }              }
4462            }            }
4463          break;          break;
# Line 4321  for (;;) Line 4473  for (;;)
4473            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4474            switch(c)            switch(c)
4475              {              {
4476                VSPACE_CASES: RRETURN(MATCH_NOMATCH);
4477              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 */  
             RRETURN(MATCH_NOMATCH);  
4478              }              }
4479            }            }
4480          break;          break;
# Line 4345  for (;;) Line 4490  for (;;)
4490            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4491            switch(c)            switch(c)
4492              {              {
4493                VSPACE_CASES: break;
4494              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
             case 0x0a:      /* LF */  
             case 0x0b:      /* VT */  
             case 0x0c:      /* FF */  
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
             case 0x2028:    /* LINE SEPARATOR */  
             case 0x2029:    /* PARAGRAPH SEPARATOR */  
             break;  
4495              }              }
4496            }            }
4497          break;          break;
# Line 4375  for (;;) Line 4513  for (;;)
4513          case OP_DIGIT:          case OP_DIGIT:
4514          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4515            {            {
4516              pcre_uint32 cc;
4517            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4518              {              {
4519              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4520              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4521              }              }
4522            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_digit) == 0)            cc = RAWUCHAR(eptr);
4523              if (cc >= 128 || (md->ctypes[cc] & ctype_digit) == 0)
4524              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4525            eptr++;            eptr++;
4526            /* 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 */
# Line 4390  for (;;) Line 4530  for (;;)
4530          case OP_NOT_WHITESPACE:          case OP_NOT_WHITESPACE:
4531          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4532            {            {
4533              pcre_uint32 cc;
4534            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4535              {              {
4536              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4537              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4538              }              }
4539            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)            cc = RAWUCHAR(eptr);
4540              if (cc < 128 && (md->ctypes[cc] & ctype_space) != 0)
4541              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4542            eptr++;            eptr++;
4543            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
# Line 4405  for (;;) Line 4547  for (;;)
4547          case OP_WHITESPACE:          case OP_WHITESPACE:
4548          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4549            {            {
4550              pcre_uint32 cc;
4551            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4552              {              {
4553              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4554              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4555              }              }
4556            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_space) == 0)            cc = RAWUCHAR(eptr);
4557              if (cc >= 128 || (md->ctypes[cc] & ctype_space) == 0)
4558              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4559            eptr++;            eptr++;
4560            /* 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 */
# Line 4420  for (;;) Line 4564  for (;;)
4564          case OP_NOT_WORDCHAR:          case OP_NOT_WORDCHAR:
4565          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4566            {            {
4567              pcre_uint32 cc;
4568            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4569              {              {
4570              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4571              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4572              }              }
4573            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)            cc = RAWUCHAR(eptr);
4574              if (cc < 128 && (md->ctypes[cc] & ctype_word) != 0)
4575              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4576            eptr++;            eptr++;
4577            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
# Line 4435  for (;;) Line 4581  for (;;)
4581          case OP_WORDCHAR:          case OP_WORDCHAR:
4582          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4583            {            {
4584              pcre_uint32 cc;
4585            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4586              {              {
4587              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4588              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4589              }              }
4590            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_word) == 0)            cc = RAWUCHAR(eptr);
4591              if (cc >= 128 || (md->ctypes[cc] & ctype_word) == 0)
4592              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4593            eptr++;            eptr++;
4594            /* 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 */
# Line 4468  for (;;) Line 4616  for (;;)
4616              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4617              }              }
4618            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4619              if (md->partial != 0 &&
4620                  eptr + 1 >= md->end_subject &&
4621                  NLBLOCK->nltype == NLTYPE_FIXED &&
4622                  NLBLOCK->nllen == 2 &&
4623                  *eptr == NLBLOCK->nl[0])
4624                {
4625                md->hitend = TRUE;
4626                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4627                }
4628            eptr++;            eptr++;
4629            }            }
4630          break;          break;
# Line 4502  for (;;) Line 4659  for (;;)
4659              {              {
4660              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4661    
4662              case 0x000d:              case CHAR_CR:
4663              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == CHAR_LF) eptr++;
4664              break;              break;
4665    
4666              case 0x000a:              case CHAR_LF:
4667              break;              break;
4668    
4669              case 0x000b:              case CHAR_VT:
4670              case 0x000c:              case CHAR_FF:
4671              case 0x0085:              case CHAR_NEL:
4672  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4673              case 0x2028:              case 0x2028:
4674              case 0x2029:              case 0x2029:
4675  #endif  #endif
# Line 4533  for (;;) Line 4690  for (;;)
4690            switch(*eptr++)            switch(*eptr++)
4691              {              {
4692              default: break;              default: break;
4693              case 0x09:      /* HT */              HSPACE_BYTE_CASES:
4694              case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4695              case 0xa0:      /* NBSP */              HSPACE_MULTIBYTE_CASES:
 #ifdef COMPILE_PCRE16  
             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 */  
4696  #endif  #endif
4697              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4698              }              }
# Line 4570  for (;;) Line 4710  for (;;)
4710            switch(*eptr++)            switch(*eptr++)
4711              {              {
4712              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4713              case 0x09:      /* HT */              HSPACE_BYTE_CASES:
4714              case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4715              case 0xa0:      /* NBSP */              HSPACE_MULTIBYTE_CASES:
 #ifdef COMPILE_PCRE16  
             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 */  
4716  #endif  #endif
4717              break;              break;
4718              }              }
# Line 4606  for (;;) Line 4729  for (;;)
4729              }              }
4730            switch(*eptr++)            switch(*eptr++)
4731              {              {
4732              default: break;              VSPACE_BYTE_CASES:
4733              case 0x0a:      /* LF */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4734              case 0x0b:      /* VT */              VSPACE_MULTIBYTE_CASES:
             case 0x0c:      /* FF */  
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
 #ifdef COMPILE_PCRE16  
             case 0x2028:    /* LINE SEPARATOR */  
             case 0x2029:    /* PARAGRAPH SEPARATOR */  
4735  #endif  #endif
4736              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4737                default: break;
4738              }              }
4739            }            }
4740          break;          break;
# Line 4632  for (;;) Line 4750  for (;;)
4750            switch(*eptr++)            switch(*eptr++)
4751              {              {
4752              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4753              case 0x0a:      /* LF */              VSPACE_BYTE_CASES:
4754              case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4755              case 0x0c:      /* FF */              VSPACE_MULTIBYTE_CASES:
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
 #ifdef COMPILE_PCRE16  
             case 0x2028:    /* LINE SEPARATOR */  
             case 0x2029:    /* PARAGRAPH SEPARATOR */  
4756  #endif  #endif
4757              break;              break;
4758              }              }
# Line 4857  for (;;) Line 4970  for (;;)
4970              }              }
4971            /* Control never gets here */            /* Control never gets here */
4972    
4973              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
4974              which means that Perl space and POSIX space are now identical. PCRE
4975              was changed at release 8.34. */
4976    
4977            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
4978              case PT_PXSPACE:  /* POSIX space */
4979            for (fi = min;; fi++)            for (fi = min;; fi++)
4980              {              {
4981              RMATCH(eptr, ecode, offset_top, md, eptrb, RM60);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM61);
4982              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4983              if (fi >= max) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4984              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4869  for (;;) Line 4987  for (;;)
4987                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4988                }                }
4989              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4990              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              switch(c)
4991                   c == CHAR_FF || c == CHAR_CR)                {
4992                     == prop_fail_result)                HSPACE_CASES:
4993                RRETURN(MATCH_NOMATCH);                VSPACE_CASES:
4994                  if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4995                  break;
4996    
4997                  default:
4998                  if ((UCD_CATEGORY(c) == ucp_Z) == prop_fail_result)
4999                    RRETURN(MATCH_NOMATCH);
5000                  break;
5001                  }
5002              }              }
5003            /* Control never gets here */            /* Control never gets here */
5004    
5005            case PT_PXSPACE:  /* POSIX space */            case PT_WORD:
5006            for (fi = min;; fi++)            for (fi = min;; fi++)
5007              {              {
5008              RMATCH(eptr, ecode, offset_top, md, eptrb, RM61);              int category;
5009                RMATCH(eptr, ecode, offset_top, md, eptrb, RM62);
5010              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5011              if (fi >= max) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
5012              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4888  for (;;) Line 5015  for (;;)
5015                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5016                }                }
5017              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
5018              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              category = UCD_CATEGORY(c);
5019                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)              if ((category == ucp_L ||
5020                     category == ucp_N ||
5021                     c == CHAR_UNDERSCORE)
5022                     == prop_fail_result)                     == prop_fail_result)
5023                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5024              }              }
5025            /* Control never gets here */            /* Control never gets here */
5026    
5027            case PT_WORD:            case PT_CLIST:
5028            for (fi = min;; fi++)            for (fi = min;; fi++)
5029              {              {
5030              int category;              const pcre_uint32 *cp;
5031              RMATCH(eptr, ecode, offset_top, md, eptrb, RM62);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM67);
5032              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5033              if (fi >= max) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
5034              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4908  for (;;) Line 5037  for (;;)
5037                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5038                }                }
5039              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
5040              category = UCD_CATEGORY(c);              cp = PRIV(ucd_caseless_sets) + prop_value;
5041              if ((category == ucp_L ||              for (;;)
5042                   category == ucp_N ||                {
5043                   c == CHAR_UNDERSCORE)                if (c < *cp)
5044                     == prop_fail_result)                  { if (prop_fail_result) break; else { RRETURN(MATCH_NOMATCH); } }
5045                  if (c == *cp++)
5046                    { if (prop_fail_result) { RRETURN(MATCH_NOMATCH); } else break; }
5047                  }
5048                }
5049              /* Control never gets here */
5050    
5051              case PT_UCNC:
5052              for (fi = min;; fi++)
5053                {
5054                RMATCH(eptr, ecode, offset_top, md, eptrb, RM68);
5055                if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5056                if (fi >= max) RRETURN(MATCH_NOMATCH);
5057                if (eptr >= md->end_subject)
5058                  {
5059                  SCHECK_PARTIAL();
5060                  RRETURN(MATCH_NOMATCH);
5061                  }
5062                GETCHARINCTEST(c, eptr);
5063                if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
5064                     c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
5065                     c >= 0xe000) == prop_fail_result)
5066                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5067              }              }
5068            /* Control never gets here */            /* Control never gets here */
5069    
5070            /* This should never occur */            /* This should never occur */
   
5071            default:            default:
5072            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
5073            }            }
# Line 4939  for (;;) Line 5088  for (;;)
5088              SCHECK_PARTIAL();              SCHECK_PARTIAL();
5089              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
5090              }              }
5091            GETCHARINCTEST(c, eptr);            else
           if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
           while (eptr < md->end_subject)  
5092              {              {
5093              int len = 1;              int lgb, rgb;
5094              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
5095              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
5096              eptr += len;              while (eptr < md->end_subject)
5097                  {
5098                  int len = 1;
5099                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
5100                  rgb = UCD_GRAPHBREAK(c);
5101                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
5102                  lgb = rgb;
5103                  eptr += len;
5104                  }
5105              }              }
5106              CHECK_PARTIAL();
5107            }            }
5108          }          }
5109        else        else
# Line 4971  for (;;) Line 5127  for (;;)
5127            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
5128            switch(ctype)            switch(ctype)
5129              {              {
5130              case OP_ANY:        /* This is the non-NL case */              case OP_ANY:               /* This is the non-NL case */
5131                if (md->partial != 0 &&    /* Take care with CRLF partial */
5132                    eptr >= md->end_subject &&
5133                    NLBLOCK->nltype == NLTYPE_FIXED &&
5134                    NLBLOCK->nllen == 2 &&
5135                    c == NLBLOCK->nl[0])
5136                  {
5137                  md->hitend = TRUE;
5138                  if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5139                  }
5140                break;
5141    
5142              case OP_ALLANY:              case OP_ALLANY:
5143              case OP_ANYBYTE:              case OP_ANYBYTE:
5144              break;              break;
# Line 4980  for (;;) Line 5147  for (;;)
5147              switch(c)              switch(c)
5148                {                {
5149                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5150                case 0x000d:                case CHAR_CR:
5151                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && RAWUCHAR(eptr) == CHAR_LF) eptr++;
5152                break;                break;
5153                case 0x000a:  
5154                  case CHAR_LF:
5155                break;                break;
5156    
5157                case 0x000b:                case CHAR_VT:
5158                case 0x000c:                case CHAR_FF:
5159                case 0x0085:                case CHAR_NEL:
5160    #ifndef EBCDIC
5161                case 0x2028:                case 0x2028:
5162                case 0x2029:                case 0x2029:
5163    #endif  /* Not EBCDIC */
5164                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
5165                break;                break;
5166                }                }
# Line 4999  for (;;) Line 5169  for (;;)
5169              case OP_NOT_HSPACE:              case OP_NOT_HSPACE:
5170              switch(c)              switch(c)
5171                {                {
5172                  HSPACE_CASES: RRETURN(MATCH_NOMATCH);
5173                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 */  
               RRETURN(MATCH_NOMATCH);  
5174                }                }
5175              break;              break;
5176    
5177              case OP_HSPACE:              case OP_HSPACE:
5178              switch(c)              switch(c)
5179                {                {
5180                  HSPACE_CASES: break;
5181                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
               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 */  
               break;  
5182                }                }
5183              break;              break;
5184    
5185              case OP_NOT_VSPACE:              case OP_NOT_VSPACE:
5186              switch(c)              switch(c)
5187                {                {
5188                  VSPACE_CASES: RRETURN(MATCH_NOMATCH);
5189                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 */  
               RRETURN(MATCH_NOMATCH);  
5190                }                }
5191              break;              break;
5192    
5193              case OP_VSPACE:              case OP_VSPACE:
5194              switch(c)              switch(c)
5195                {                {
5196                  VSPACE_CASES: break;
5197                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
               case 0x0a:      /* LF */  
               case 0x0b:      /* VT */  
               case 0x0c:      /* FF */  
               case 0x0d:      /* CR */  
               case 0x85:      /* NEL */  
               case 0x2028:    /* LINE SEPARATOR */  
               case 0x2029:    /* PARAGRAPH SEPARATOR */  
               break;  
5198                }                }
5199              break;              break;
5200    
# Line 5134  for (;;) Line 5252  for (;;)
5252            c = *eptr++;            c = *eptr++;
5253            switch(ctype)            switch(ctype)
5254              {              {
5255              case OP_ANY:     /* This is the non-NL case */              case OP_ANY:               /* This is the non-NL case */
5256                if (md->partial != 0 &&    /* Take care with CRLF partial */
5257                    eptr >= md->end_subject &&
5258                    NLBLOCK->nltype == NLTYPE_FIXED &&
5259                    NLBLOCK->nllen == 2 &&
5260                    c == NLBLOCK->nl[0])
5261                  {
5262                  md->hitend = TRUE;
5263                  if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5264                  }
5265                break;
5266    
5267              case OP_ALLANY:              case OP_ALLANY:
5268              case OP_ANYBYTE:              case OP_ANYBYTE:
5269              break;              break;
# Line 5143  for (;;) Line 5272  for (;;)
5272              switch(c)              switch(c)
5273                {                {
5274                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5275                case 0x000d:                case CHAR_CR:
5276                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == CHAR_LF) eptr++;
5277                break;                break;
5278    
5279                case 0x000a:                case CHAR_LF:
5280                break;                break;
5281    
5282                case 0x000b:                case CHAR_VT:
5283                case 0x000c:                case CHAR_FF:
5284                case 0x0085:                case CHAR_NEL:
5285  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5286                case 0x2028:                case 0x2028:
5287                case 0x2029:                case 0x2029:
5288  #endif  #endif
# Line 5166  for (;;) Line 5295  for (;;)
5295              switch(c)              switch(c)
5296                {                {
5297                default: break;                default: break;
5298                case 0x09:      /* HT */                HSPACE_BYTE_CASES:
5299                case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5300                case 0xa0:      /* NBSP */                HSPACE_MULTIBYTE_CASES:
 #ifdef COMPILE_PCRE16  
               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 */  
5301  #endif  #endif
5302                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5303                }                }
# Line 5195  for (;;) Line 5307  for (;;)
5307              switch(c)              switch(c)
5308                {                {
5309                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5310                case 0x09:      /* HT */                HSPACE_BYTE_CASES:
5311                case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5312                case 0xa0:      /* NBSP */                HSPACE_MULTIBYTE_CASES:
 #ifdef COMPILE_PCRE16  
               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 */  
5313  #endif  #endif
5314                break;                break;
5315                }                }
# Line 5224  for (;;) Line 5319  for (;;)
5319              switch(c)              switch(c)
5320                {                {
5321                default: break;                default: break;
5322                case 0x0a:      /* LF */                VSPACE_BYTE_CASES:
5323                case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5324                case 0x0c:      /* FF */                VSPACE_MULTIBYTE_CASES:
               case 0x0d:      /* CR */  
               case 0x85:      /* NEL */  
 #ifdef COMPILE_PCRE16  
               case 0x2028:    /* LINE SEPARATOR */  
               case 0x2029:    /* PARAGRAPH SEPARATOR */  
5325  #endif  #endif
5326                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5327                }                }
# Line 5241  for (;;) Line 5331  for (;;)
5331              switch(c)              switch(c)
5332                {                {
5333                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5334                case 0x0a:      /* LF */                VSPACE_BYTE_CASES:
5335                case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5336                case 0x0c:      /* FF */                VSPACE_MULTIBYTE_CASES:
               case 0x0d:      /* CR */  
               case 0x85:      /* NEL */  
 #ifdef COMPILE_PCRE16  
               case 0x2028:    /* LINE SEPARATOR */  
               case 0x2029:    /* PARAGRAPH SEPARATOR */  
5337  #endif  #endif
5338                break;                break;
5339                }                }
# Line 5397  for (;;) Line 5482  for (;;)
5482              }              }
5483            break;            break;
5484    
5485              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
5486              which means that Perl space and POSIX space are now identical. PCRE
5487              was changed at release 8.34. */
5488    
5489            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
5490              case PT_PXSPACE:  /* POSIX space */
5491            for (i = min; i < max; i++)            for (i = min; i < max; i++)
5492              {              {
5493              int len = 1;              int len = 1;
# Line 5407  for (;;) Line 5497  for (;;)
5497                break;                break;
5498                }                }
5499              GETCHARLENTEST(c, eptr, len);              GETCHARLENTEST(c, eptr, len);
5500              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              switch(c)
5501                   c == CHAR_FF || c == CHAR_CR)                {
5502                   == prop_fail_result)                HSPACE_CASES:
5503                  VSPACE_CASES:
5504                  if (prop_fail_result) goto ENDLOOP99;  /* Break the loop */
5505                break;                break;
5506    
5507                  default:
5508                  if ((UCD_CATEGORY(c) == ucp_Z) == prop_fail_result)
5509                    goto ENDLOOP99;   /* Break the loop */
5510                  break;
5511                  }
5512              eptr+= len;              eptr+= len;
5513              }              }
5514              ENDLOOP99:
5515            break;            break;
5516    
5517            case PT_PXSPACE:  /* POSIX space */            case PT_WORD:
5518            for (i = min; i < max; i++)            for (i = min; i < max; i++)
5519              {              {
5520                int category;
5521              int len = 1;              int len = 1;
5522              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
5523                {                {
# Line 5425  for (;;) Line 5525  for (;;)
5525                break;                break;
5526                }                }
5527              GETCHARLENTEST(c, eptr, len);              GETCHARLENTEST(c, eptr, len);
5528              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              category = UCD_CATEGORY(c);
5529                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)              if ((category == ucp_L || category == ucp_N ||
5530                   == prop_fail_result)                   c == CHAR_UNDERSCORE) == prop_fail_result)
5531                break;                break;
5532              eptr+= len;              eptr+= len;
5533              }              }
5534            break;            break;
5535    
5536            case PT_WORD:            case PT_CLIST:
5537            for (i = min; i < max; i++)            for (i = min; i < max; i++)
5538              {              {
5539              int category;              const pcre_uint32 *cp;
5540              int len = 1;              int len = 1;
5541              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
5542                {                {
# Line 5444  for (;;) Line 5544  for (;;)
5544                break;                break;
5545                }                }
5546              GETCHARLENTEST(c, eptr, len);              GETCHARLENTEST(c, eptr, len);
5547              category = UCD_CATEGORY(c);              cp = PRIV(ucd_caseless_sets) + prop_value;
5548              if ((category == ucp_L || category == ucp_N ||              for (;;)
5549                   c == CHAR_UNDERSCORE) == prop_fail_result)                {
5550                  if (c < *cp)
5551                    { if (prop_fail_result) break; else goto GOT_MAX; }
5552                  if (c == *cp++)
5553                    { if (prop_fail_result) goto GOT_MAX; else break; }
5554                  }
5555                eptr += len;
5556                }
5557              GOT_MAX:
5558              break;
5559    
5560              case PT_UCNC:
5561              for (i = min; i < max; i++)
5562                {
5563                int len = 1;
5564                if (eptr >= md->end_subject)
5565                  {
5566                  SCHECK_PARTIAL();
5567                break;                break;
5568              eptr+= len;                }
5569                GETCHARLENTEST(c, eptr, len);
5570                if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
5571                     c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
5572                     c >= 0xe000) == prop_fail_result)
5573                  break;
5574                eptr += len;
5575              }              }
5576            break;            break;
5577    
# Line 5458  for (;;) Line 5581  for (;;)
5581    
5582          /* eptr is now past the end of the maximum run */          /* eptr is now past the end of the maximum run */
5583    
5584          if (possessive) continue;          if (possessive) continue;    /* No backtracking */
5585          for(;;)          for(;;)
5586            {            {
5587              if (eptr == pp) goto TAIL_RECURSE;
5588            RMATCH(eptr, ecode, offset_top, md, eptrb, RM44);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM44);
5589            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5590            if (eptr-- == pp) break;        /* Stop if tried at original pos */            eptr--;
5591            if (utf) BACKCHAR(eptr);            if (utf) BACKCHAR(eptr);
5592            }            }
5593          }          }
5594    
5595        /* Match extended Unicode sequences. We will get here only if the        /* Match extended Unicode grapheme clusters. We will get here only if the
5596        support is in the binary; otherwise a compile-time error occurs. */        support is in the binary; otherwise a compile-time error occurs. */
5597    
5598        else if (ctype == OP_EXTUNI)        else if (ctype == OP_EXTUNI)
5599          {          {
5600          for (i = min; i < max; i++)          for (i = min; i < max; i++)
5601            {            {
           int len = 1;  
5602            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
5603              {              {
5604              SCHECK_PARTIAL();              SCHECK_PARTIAL();
5605              break;              break;
5606              }              }
5607            if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }            else
           if (UCD_CATEGORY(c) == ucp_M) break;  
           eptr += len;  
           while (eptr < md->end_subject)  
5608              {              {
5609              len = 1;              int lgb, rgb;
5610              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
5611              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
5612              eptr += len;              while (eptr < md->end_subject)
5613                  {
5614                  int len = 1;
5615                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
5616                  rgb = UCD_GRAPHBREAK(c);
5617                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
5618                  lgb = rgb;
5619                  eptr += len;
5620                  }
5621              }              }
5622              CHECK_PARTIAL();
5623            }            }
5624    
5625          /* eptr is now past the end of the maximum run */          /* eptr is now past the end of the maximum run */
5626    
5627          if (possessive) continue;          if (possessive) continue;    /* No backtracking */
5628    
5629          for(;;)          for(;;)
5630            {            {
5631              int lgb, rgb;
5632              PCRE_PUCHAR fptr;
5633    
5634              if (eptr == pp) goto TAIL_RECURSE;   /* At start of char run */
5635            RMATCH(eptr, ecode, offset_top, md, eptrb, RM45);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM45);
5636            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5637            if (eptr-- == pp) break;        /* Stop if tried at original pos */  
5638            for (;;)                        /* Move back over one extended */            /* Backtracking over an extended grapheme cluster involves inspecting
5639              the previous two characters (if present) to see if a break is
5640              permitted between them. */
5641    
5642              eptr--;
5643              if (!utf) c = *eptr; else
5644              {              {
5645              if (!utf) c = *eptr; else              BACKCHAR(eptr);
5646                {              GETCHAR(c, eptr);
5647                BACKCHAR(eptr);              }
5648                GETCHAR(c, eptr);            rgb = UCD_GRAPHBREAK(c);
5649                }  
5650              if (UCD_CATEGORY(c) != ucp_M) break;            for (;;)
5651              eptr--;              {
5652                if (eptr == pp) goto TAIL_RECURSE;   /* At start of char run */
5653                fptr = eptr - 1;
5654                if (!utf) c = *fptr; else
5655                  {
5656                  BACKCHAR(fptr);
5657                  GETCHAR(c, fptr);
5658                  }
5659                lgb = UCD_GRAPHBREAK(c);
5660                if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
5661                eptr = fptr;
5662                rgb = lgb;
5663              }              }
5664            }            }
5665          }          }
# Line 5534  for (;;) Line 5683  for (;;)
5683                  break;                  break;
5684                  }                  }
5685                if (IS_NEWLINE(eptr)) break;                if (IS_NEWLINE(eptr)) break;
5686                  if (md->partial != 0 &&    /* Take care with CRLF partial */
5687                      eptr + 1 >= md->end_subject &&
5688                      NLBLOCK->nltype == NLTYPE_FIXED &&
5689                      NLBLOCK->nllen == 2 &&
5690                      RAWUCHAR(eptr) == NLBLOCK->nl[0])
5691                    {
5692                    md->hitend = TRUE;
5693                    if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5694                    }
5695                eptr++;                eptr++;
5696                ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);                ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
5697                }                }
# Line 5551  for (;;) Line 5709  for (;;)
5709                  break;                  break;
5710                  }                  }
5711                if (IS_NEWLINE(eptr)) break;                if (IS_NEWLINE(eptr)) break;
5712                  if (md->partial != 0 &&    /* Take care with CRLF partial */
5713                      eptr + 1 >= md->end_subject &&
5714                      NLBLOCK->nltype == NLTYPE_FIXED &&
5715                      NLBLOCK->nllen == 2 &&
5716                      RAWUCHAR(eptr) == NLBLOCK->nl[0])
5717                    {
5718                    md->hitend = TRUE;
5719                    if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5720                    }
5721                eptr++;                eptr++;
5722                ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);                ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
5723                }                }
# Line 5600  for (;;) Line 5767  for (;;)
5767                break;                break;
5768                }                }
5769              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
5770              if (c == 0x000d)              if (c == CHAR_CR)
5771                {                {
5772                if (++eptr >= md->end_subject) break;                if (++eptr >= md->end_subject) break;
5773                if (*eptr == 0x000a) eptr++;                if (RAWUCHAR(eptr) == CHAR_LF) eptr++;
5774                }                }
5775              else              else
5776                {                {
5777                if (c != 0x000a &&                if (c != CHAR_LF &&
5778                    (md->bsr_anycrlf ||                    (md->bsr_anycrlf ||
5779                     (c != 0x000b && c != 0x000c &&                     (c != CHAR_VT && c != CHAR_FF && c != CHAR_NEL
5780                      c != 0x0085 && c != 0x2028 && c != 0x2029)))  #ifndef EBCDIC
5781                        && c != 0x2028 && c != 0x2029
5782    #endif  /* Not EBCDIC */
5783                        )))
5784                  break;                  break;
5785                eptr += len;                eptr += len;
5786                }                }
# Line 5631  for (;;) Line 5801  for (;;)
5801              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
5802              switch(c)              switch(c)
5803                {                {
5804                  HSPACE_CASES: gotspace = TRUE; break;
5805                default: gotspace = FALSE; break;                default: gotspace = FALSE; 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 */  
               gotspace = TRUE;  
               break;  
5806                }                }
5807              if (gotspace == (ctype == OP_NOT_HSPACE)) break;              if (gotspace == (ctype == OP_NOT_HSPACE)) break;
5808              eptr += len;              eptr += len;
# Line 5673  for (;;) Line 5823  for (;;)
5823              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
5824              switch(c)              switch(c)
5825                {                {
5826                  VSPACE_CASES: gotspace = TRUE; break;
5827                default: gotspace = FALSE; break;                default: gotspace = FALSE; break;
               case 0x0a:      /* LF */  
               case 0x0b:      /* VT */  
               case 0x0c:      /* FF */  
               case 0x0d:      /* CR */  
               case 0x85:      /* NEL */  
               case 0x2028:    /* LINE SEPARATOR */  
               case 0x2029:    /* PARAGRAPH SEPARATOR */  
               gotspace = TRUE;  
               break;  
5828                }                }
5829              if (gotspace == (ctype == OP_NOT_VSPACE)) break;              if (gotspace == (ctype == OP_NOT_VSPACE)) break;
5830              eptr += len;              eptr += len;
# Line 5783  for (;;) Line 5925  for (;;)
5925            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
5926            }            }
5927    
5928          /* eptr is now past the end of the maximum run. If possessive, we are          if (possessive) continue;    /* No backtracking */
         done (no backing up). Otherwise, match at this position; anything other  
         than no match is immediately returned. For nomatch, back up one  
         character, unless we are matching \R and the last thing matched was  
         \r\n, in which case, back up two bytes. */  
   
         if (possessive) continue;  
5929          for(;;)          for(;;)
5930            {            {
5931              if (eptr == pp) goto TAIL_RECURSE;
5932            RMATCH(eptr, ecode, offset_top, md, eptrb, RM46);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM46);
5933            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5934            if (eptr-- == pp) break;        /* Stop if tried at original pos */            eptr--;
5935            BACKCHAR(eptr);            BACKCHAR(eptr);
5936            if (ctype == OP_ANYNL && eptr > pp  && *eptr == '\n' &&            if (ctype == OP_ANYNL && eptr > pp  && RAWUCHAR(eptr) == CHAR_NL &&
5937                eptr[-1] == '\r') eptr--;                RAWUCHAR(eptr - 1) == CHAR_CR) eptr--;
5938            }            }
5939          }          }
5940        else        else
# Line 5815  for (;;) Line 5952  for (;;)
5952                break;                break;
5953                }                }
5954              if (IS_NEWLINE(eptr)) break;              if (IS_NEWLINE(eptr)) break;
5955                if (md->partial != 0 &&    /* Take care with CRLF partial */
5956                    eptr + 1 >= md->end_subject &&
5957                    NLBLOCK->nltype == NLTYPE_FIXED &&
5958                    NLBLOCK->nllen == 2 &&
5959                    *eptr == NLBLOCK->nl[0])
5960                  {
5961                  md->hitend = TRUE;
5962                  if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5963                  }
5964              eptr++;              eptr++;
5965              }              }
5966            break;            break;
# Line 5839  for (;;) Line 5985  for (;;)
5985                break;                break;
5986                }                }
5987              c = *eptr;              c = *eptr;
5988              if (c == 0x000d)              if (c == CHAR_CR)
5989                {                {
5990                if (++eptr >= md->end_subject) break;                if (++eptr >= md->end_subject) break;
5991                if (*eptr == 0x000a) eptr++;                if (*eptr == CHAR_LF) eptr++;
5992                }                }
5993              else              else
5994                {                {
5995                if (c != 0x000a && (md->bsr_anycrlf ||                if (c != CHAR_LF && (md->bsr_anycrlf ||
5996                  (c != 0x000b && c != 0x000c && c != 0x0085                   (c != CHAR_VT && c != CHAR_FF && c != CHAR_NEL
5997  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5998                  && c != 0x2028 && c != 0x2029                   && c != 0x2028 && c != 0x2029
5999  #endif  #endif
6000                  ))) break;                   ))) break;
6001                eptr++;                eptr++;
6002                }                }
6003              }              }
# Line 5865  for (;;) Line 6011  for (;;)
6011                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6012                break;                break;
6013                }                }
6014              c = *eptr;              switch(*eptr)
6015              if (c == 0x09 || c == 0x20 || c == 0xa0                {
6016  #ifdef COMPILE_PCRE16                default: eptr++; break;
6017                || c == 0x1680 || c == 0x180e || (c >= 0x2000 && c <= 0x200A)                HSPACE_BYTE_CASES:
6018                || c == 0x202f || c == 0x205f || c == 0x3000  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6019                  HSPACE_MULTIBYTE_CASES:
6020  #endif  #endif
6021                ) break;                goto ENDLOOP00;
6022              eptr++;                }
6023              }              }
6024              ENDLOOP00:
6025            break;            break;
6026    
6027            case OP_HSPACE:            case OP_HSPACE:
# Line 5884  for (;;) Line 6032  for (;;)
6032                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6033                break;                break;
6034                }                }
6035              c = *eptr;              switch(*eptr)
6036              if (c != 0x09 && c != 0x20 && c != 0xa0                {
6037  #ifdef COMPILE_PCRE16                default: goto ENDLOOP01;
6038                && c != 0x1680 && c != 0x180e && (c < 0x2000 || c > 0x200A)                HSPACE_BYTE_CASES:
6039                && c != 0x202f && c != 0x205f && c != 0x3000  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6040                  HSPACE_MULTIBYTE_CASES:
6041  #endif  #endif
6042                ) break;                eptr++; break;
6043              eptr++;                }
6044              }              }
6045              ENDLOOP01:
6046            break;            break;
6047    
6048            case OP_NOT_VSPACE:            case OP_NOT_VSPACE:
# Line 5903  for (;;) Line 6053  for (;;)
6053                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6054                break;                break;
6055                }                }
6056              c = *eptr;              switch(*eptr)
6057              if (c == 0x0a || c == 0x0b || c == 0x0c || c == 0x0d || c == 0x85                {
6058  #ifdef COMPILE_PCRE16                default: eptr++; break;
6059                || c == 0x2028 || c == 0x2029                VSPACE_BYTE_CASES:
6060    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6061                  VSPACE_MULTIBYTE_CASES:
6062  #endif  #endif
6063                ) break;                goto ENDLOOP02;
6064              eptr++;                }
6065              }              }
6066              ENDLOOP02:
6067            break;            break;
6068    
6069            case OP_VSPACE:            case OP_VSPACE:
# Line 5921  for (;;) Line 6074  for (;;)
6074                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6075                break;                break;
6076                }                }
6077              c = *eptr;              switch(*eptr)
6078              if (c != 0x0a && c != 0x0b && c != 0x0c && c != 0x0d && c != 0x85                {
6079  #ifdef COMPILE_PCRE16                default: goto ENDLOOP03;
6080                && c != 0x2028 && c != 0x2029                VSPACE_BYTE_CASES:
6081    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6082                  VSPACE_MULTIBYTE_CASES:
6083  #endif  #endif
6084                ) break;                eptr++; break;
6085              eptr++;                }
6086              }              }
6087              ENDLOOP03:
6088            break;            break;
6089    
6090            case OP_NOT_DIGIT:            case OP_NOT_DIGIT:
# Line 6013  for (;;) Line 6169  for (;;)
6169            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
6170            }            }
6171    
6172          /* eptr is now past the end of the maximum run. If possessive, we are          if (possessive) continue;    /* No backtracking */
6173          done (no backing up). Otherwise, match at this position; anything other          for (;;)
         than no match is immediately returned. For nomatch, back up one  
         character (byte), unless we are matching \R and the last thing matched  
         was \r\n, in which case, back up two bytes. */  
   
         if (possessive) continue;  
         while (eptr >= pp)  
6174            {            {
6175              if (eptr == pp) goto TAIL_RECURSE;
6176            RMATCH(eptr, ecode, offset_top, md, eptrb, RM47);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM47);
6177            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
6178            eptr--;            eptr--;
6179            if (ctype == OP_ANYNL && eptr > pp  && *eptr == '\n' &&            if (ctype == OP_ANYNL && eptr > pp  && *eptr == CHAR_LF &&
6180                eptr[-1] == '\r') eptr--;                eptr[-1] == CHAR_CR) eptr--;
6181            }            }
6182          }          }
6183    
6184        /* Get here if we can't make it match with any permitted repetitions */        /* Control never gets here */
   
       RRETURN(MATCH_NOMATCH);  
6185        }        }
     /* Control never gets here */  
6186    
6187      /* There's been some horrible disaster. Arrival here can only mean there is      /* There's been some horrible disaster. Arrival here can only mean there is
6188      something seriously wrong in the code above or the OP_xxx definitions. */      something seriously wrong in the code above or the OP_xxx definitions. */
# Line 6076  switch (frame->Xwhere) Line 6224  switch (frame->Xwhere)
6224    LBL(32) LBL(34) LBL(42) LBL(46)    LBL(32) LBL(34) LBL(42) LBL(46)
6225  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
6226    LBL(36) LBL(37) LBL(38) LBL(39) LBL(40) LBL(41) LBL(44) LBL(45)    LBL(36) LBL(37) LBL(38) LBL(39) LBL(40) LBL(41) LBL(44) LBL(45)
6227    LBL(59) LBL(60) LBL(61) LBL(62)    LBL(59) LBL(60) LBL(61) LBL(62) LBL(67) LBL(68)
6228  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
6229  #endif  /* SUPPORT_UTF */  #endif  /* SUPPORT_UTF */
6230    default:    default:
6231    DPRINTF(("jump error in pcre match: label %d non-existent\n", frame->Xwhere));    DPRINTF(("jump error in pcre match: label %d non-existent\n", frame->Xwhere));
   
 printf("+++jump error in pcre match: label %d non-existent\n", frame->Xwhere);  
   
6232    return PCRE_ERROR_INTERNAL;    return PCRE_ERROR_INTERNAL;
6233    }    }
6234  #undef LBL  #undef LBL
# Line 6145  Undefine all the macros that were define Line 6290  Undefine all the macros that were define
6290  ***************************************************************************/  ***************************************************************************/
6291    
6292    
6293    #ifdef NO_RECURSE
6294    /*************************************************
6295    *          Release allocated heap frames         *
6296    *************************************************/
6297    
6298    /* This function releases all the allocated frames. The base frame is on the
6299    machine stack, and so must not be freed.
6300    
6301    Argument: the address of the base frame
6302    Returns:  nothing
6303    */
6304    
6305    static void
6306    release_match_heapframes (heapframe *frame_base)
6307    {
6308    heapframe *nextframe = frame_base->Xnextframe;
6309    while (nextframe != NULL)
6310      {
6311      heapframe *oldframe = nextframe;
6312      nextframe = nextframe->Xnextframe;
6313      (PUBL(stack_free))(oldframe);
6314      }
6315    }
6316    #endif
6317    
6318    
6319  /*************************************************  /*************************************************
6320  *         Execute a Regular Expression           *  *         Execute a Regular Expression           *
# Line 6170  Returns:          > 0 => success; value Line 6340  Returns:          > 0 => success; value
6340                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
6341  */  */
6342    
6343  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
6344  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
6345  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,
6346    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,
6347    int offsetcount)    int offsetcount)
6348  #else  #elif defined COMPILE_PCRE16
6349  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
6350  pcre16_exec(const pcre16 *argument_re, const pcre16_extra *extra_data,  pcre16_exec(const pcre16 *argument_re, const pcre16_extra *extra_data,
6351    PCRE_SPTR16 subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR16 subject, int length, int start_offset, int options, int *offsets,
6352    int offsetcount)    int offsetcount)
6353    #elif defined COMPILE_PCRE32
6354    PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
6355    pcre32_exec(const pcre32 *argument_re, const pcre32_extra *extra_data,
6356      PCRE_SPTR32 subject, int length, int start_offset, int options, int *offsets,
6357      int offsetcount)
6358  #endif  #endif
6359  {  {
6360  int rc, ocount, arg_offset_max;  int rc, ocount, arg_offset_max;
# Line 6202  const pcre_uint8 *start_bits = NULL; Line 6377  const pcre_uint8 *start_bits = NULL;
6377  PCRE_PUCHAR start_match = (PCRE_PUCHAR)subject + start_offset;  PCRE_PUCHAR start_match = (PCRE_PUCHAR)subject + start_offset;
6378  PCRE_PUCHAR end_subject;  PCRE_PUCHAR end_subject;
6379  PCRE_PUCHAR start_partial = NULL;  PCRE_PUCHAR start_partial = NULL;
6380    PCRE_PUCHAR match_partial = NULL;
6381  PCRE_PUCHAR req_char_ptr = start_match - 1;  PCRE_PUCHAR req_char_ptr = start_match - 1;
6382    
6383  const pcre_study_data *study;  const pcre_study_data *study;
6384  const REAL_PCRE *re = (const REAL_PCRE *)argument_re;  const REAL_PCRE *re = (const REAL_PCRE *)argument_re;
6385    
6386  /* Check for the special magic call that measures the size of the stack used  #ifdef NO_RECURSE
6387  per recursive call of match(). */  heapframe frame_zero;
6388    frame_zero.Xprevframe = NULL;            /* Marks the top level */
6389    frame_zero.Xnextframe = NULL;            /* None are allocated yet */
6390    md->match_frames_base = &frame_zero;
6391    #endif
6392    
6393  if (re == NULL && extra_data == NULL && subject == NULL && length == -1)  /* Check for the special magic call that measures the size of the stack used
6394    per recursive call of match(). Without the funny casting for sizeof, a Windows
6395    compiler gave this error: "unary minus operator applied to unsigned type,
6396    result still unsigned". Hopefully the cast fixes that. */
6397    
6398    if (re == NULL && extra_data == NULL && subject == NULL && length == -999 &&
6399        start_offset == -999)
6400  #ifdef NO_RECURSE  #ifdef NO_RECURSE
6401    return -sizeof(heapframe);    return -((int)sizeof(heapframe));
6402  #else  #else
6403    return match(NULL, NULL, NULL, 0, NULL, NULL, 0);    return match(NULL, NULL, NULL, 0, NULL, NULL, 0);
6404  #endif  #endif
6405    
6406  /* Plausibility checks */  /* Plausibility checks */
6407    
6408  if ((options & ~PUBLIC_EXEC_OPTIONS) != 0) return PCRE_ERROR_BADOPTION;  if ((options & ~PUBLIC_EXEC_OPTIONS) != 0) return PCRE_ERROR_BADOPTION;
6409  if (re == NULL || subject == NULL || (offsets == NULL && offsetcount > 0))  if (re == NULL || subject == NULL || (offsets == NULL && offsetcount > 0))
6410    return PCRE_ERROR_NULL;    return PCRE_ERROR_NULL;
6411  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;
6412    if (length < 0) return PCRE_ERROR_BADLENGTH;
6413  if (start_offset < 0 || start_offset > length) return PCRE_ERROR_BADOFFSET;  if (start_offset < 0 || start_offset > length) return PCRE_ERROR_BADOFFSET;
6414    
6415  /* Check that the first field in the block is the magic number. If it is not,  /* Check that the first field in the block is the magic number. If it is not,
# Line 6260  if (utf && (options & PCRE_NO_UTF8_CHECK Line 6447  if (utf && (options & PCRE_NO_UTF8_CHECK
6447        offsets[0] = erroroffset;        offsets[0] = erroroffset;
6448        offsets[1] = errorcode;        offsets[1] = errorcode;
6449        }        }
6450  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE8
     return (errorcode <= PCRE_UTF16_ERR1 && md->partial > 1)?  
       PCRE_ERROR_SHORTUTF16 : PCRE_ERROR_BADUTF16;  
 #else  
6451      return (errorcode <= PCRE_UTF8_ERR5 && md->partial > 1)?      return (errorcode <= PCRE_UTF8_ERR5 && md->partial > 1)?
6452        PCRE_ERROR_SHORTUTF8 : PCRE_ERROR_BADUTF8;        PCRE_ERROR_SHORTUTF8 : PCRE_ERROR_BADUTF8;
6453    #elif defined COMPILE_PCRE16
6454        return (errorcode <= PCRE_UTF16_ERR1 && md->partial > 1)?
6455          PCRE_ERROR_SHORTUTF16 : PCRE_ERROR_BADUTF16;
6456    #elif defined COMPILE_PCRE32
6457        return PCRE_ERROR_BADUTF32;
6458  #endif  #endif
6459      }      }
6460    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
6461    /* Check that a start_offset points to the start of a UTF character. */    /* Check that a start_offset points to the start of a UTF character. */
6462    if (start_offset > 0 && start_offset < length &&    if (start_offset > 0 && start_offset < length &&
6463        NOT_FIRSTCHAR(((PCRE_PUCHAR)subject)[start_offset]))        NOT_FIRSTCHAR(((PCRE_PUCHAR)subject)[start_offset]))
6464      return PCRE_ERROR_BADUTF8_OFFSET;      return PCRE_ERROR_BADUTF8_OFFSET;
6465    #endif
6466    }    }
6467  #endif  #endif
6468    
6469  /* If the pattern was successfully studied with JIT support, run the JIT  /* If the pattern was successfully studied with JIT support, run the JIT
6470  executable instead of the rest of this function. Most options must be set at  executable instead of the rest of this function. Most options must be set at
6471  compile time for the JIT code to be usable. Fallback to the normal code path if  compile time for the JIT code to be usable. Fallback to the normal code path if
6472  an unsupported flag is set. In particular, JIT does not support partial  an unsupported flag is set. */
 matching. */  
6473    
6474  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
6475  if (extra_data != NULL  if (extra_data != NULL
6476      && (extra_data->flags & PCRE_EXTRA_EXECUTABLE_JIT) != 0      && (extra_data->flags & (PCRE_EXTRA_EXECUTABLE_JIT |
6477                                 PCRE_EXTRA_TABLES)) == PCRE_EXTRA_EXECUTABLE_JIT
6478      && extra_data->executable_jit != NULL      && extra_data->executable_jit != NULL
6479      && (extra_data->flags & PCRE_EXTRA_TABLES) == 0      && (options & ~PUBLIC_JIT_EXEC_OPTIONS) == 0)
6480      && (options & ~(PCRE_NO_UTF8_CHECK | PCRE_NOTBOL | PCRE_NOTEOL |    {
6481                      PCRE_NOTEMPTY | PCRE_NOTEMPTY_ATSTART)) == 0)    rc = PRIV(jit_exec)(extra_data, (const pcre_uchar *)subject, length,
6482    return PRIV(jit_exec)(re, extra_data->executable_jit,         start_offset, options, offsets, offsetcount);
6483      (const pcre_uchar *)subject, length, start_offset, options,  
6484      ((extra_data->flags & PCRE_EXTRA_MATCH_LIMIT) == 0)    /* PCRE_ERROR_NULL means that the selected normal or partial matching
6485      ? MATCH_LIMIT : extra_data->match_limit, offsets, offsetcount);    mode is not compiled. In this case we simply fallback to interpreter. */
6486    
6487      if (rc != PCRE_ERROR_JIT_BADOPTION) return rc;
6488      }
6489  #endif  #endif
6490    
6491  /* Carry on with non-JIT matching. This information is for finding all the  /* Carry on with non-JIT matching. This information is for finding all the
# Line 6314  md->callout_data = NULL; Line 6507  md->callout_data = NULL;
6507    
6508  tables = re->tables;  tables = re->tables;
6509    
6510    /* The two limit values override the defaults, whatever their value. */
6511    
6512  if (extra_data != NULL)  if (extra_data != NULL)
6513    {    {
6514    register unsigned int flags = extra_data->flags;    register unsigned int flags = extra_data->flags;
# Line 6328  if (extra_data != NULL) Line 6523  if (extra_data != NULL)
6523    if ((flags & PCRE_EXTRA_TABLES) != 0) tables = extra_data->tables;    if ((flags & PCRE_EXTRA_TABLES) != 0) tables = extra_data->tables;
6524    }    }