/[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 926 by ph10, Wed Feb 22 15:01:32 2012 UTC revision 1411 by ph10, Fri Dec 6 17:11:44 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 142  Arguments: Line 158  Arguments:
158    
159  Returns:      >= 0 the number of subject bytes matched  Returns:      >= 0 the number of subject bytes matched
160                -1 no match                -1 no match
161                -2 partial match; always given if at end subject                -2 partial match; always given if at end subject
162  */  */
163    
164  static int  static int
# Line 151  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    #if defined SUPPORT_UTF && defined SUPPORT_UCP
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 165  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 - in that  /* Always fail if reference not set (and not JavaScript compatible - in that
188  case the length is passed as zero). */  case the length is passed as zero). */
189    
190  if (length < 0) return -1;  if (length < 0) return -1;
# Line 176  ASCII characters. */ Line 195  ASCII characters. */
195    
196  if (caseless)  if (caseless)
197    {    {
198  #ifdef SUPPORT_UTF  #if defined SUPPORT_UTF && defined SUPPORT_UCP
199  #ifdef SUPPORT_UCP    if (utf)
   if (md->utf)  
200      {      {
201      /* 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
202      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
203      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.
204      (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
205      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
206      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
207      reference, not along the subject (earlier code did this wrong). */      length along the reference, not along the subject (earlier code did this
208        wrong). */
209    
210      PCRE_PUCHAR endptr = p + length;      PCRE_PUCHAR endptr = p + length;
211      while (p < endptr)      while (p < endptr)
212        {        {
213        int c, d;        pcre_uint32 c, d;
214          const ucd_record *ur;
215        if (eptr >= md->end_subject) return -2;   /* Partial match */        if (eptr >= md->end_subject) return -2;   /* Partial match */
216        GETCHARINC(c, eptr);        GETCHARINC(c, eptr);
217        GETCHARINC(d, p);        GETCHARINC(d, p);
218        if (c != d && c != UCD_OTHERCASE(d)) return -1;        ur = GET_UCD(d);
219          if (c != d && c != d + ur->other_case)
220            {
221            const pcre_uint32 *pp = PRIV(ucd_caseless_sets) + ur->caseset;
222            for (;;)
223              {
224              if (c < *pp) return -1;
225              if (c == *pp++) break;
226              }
227            }
228        }        }
229      }      }
230    else    else
231  #endif  #endif
 #endif  
232    
233    /* 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
234    is no UCP support. */    is no UCP support. */
235      {      {
236      while (length-- > 0)      while (length-- > 0)
237        {        {
238          pcre_uint32 cc, cp;
239        if (eptr >= md->end_subject) return -2;   /* Partial match */        if (eptr >= md->end_subject) return -2;   /* Partial match */
240        if (TABLE_GET(*p, md->lcc, *p) != TABLE_GET(*eptr, md->lcc, *eptr)) return -1;        cc = RAWUCHARTEST(eptr);
241          cp = RAWUCHARTEST(p);
242          if (TABLE_GET(cp, md->lcc, cp) != TABLE_GET(cc, md->lcc, cc)) return -1;
243        p++;        p++;
244        eptr++;        eptr++;
245        }        }
# Line 220  are in UTF-8 mode. */ Line 251  are in UTF-8 mode. */
251    
252  else  else
253    {    {
254    while (length-- > 0)    while (length-- > 0)
255      {      {
256      if (eptr >= md->end_subject) return -2;   /* Partial match */      if (eptr >= md->end_subject) return -2;   /* Partial match */
257      if (*p++ != *eptr++) return -1;      if (RAWUCHARINCTEST(p) != RAWUCHARINCTEST(eptr)) return -1;
258      }      }
259    }    }
260    
261  return (int)(eptr - eptr_start);  return (int)(eptr - eptr_start);
# Line 279  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM Line 310  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM
310         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,
311         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,
312         RM51,  RM52, RM53, RM54, RM55, RM56, RM57, RM58, RM59, RM60,         RM51,  RM52, RM53, RM54, RM55, RM56, RM57, RM58, RM59, RM60,
313         RM61,  RM62, RM63, RM64, RM65, RM66 };         RM61,  RM62, RM63, RM64, RM65, RM66, RM67 };
314    
315  /* 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
316  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 297  actually used in this definition. */ Line 328  actually used in this definition. */
328    }    }
329  #define RRETURN(ra) \  #define RRETURN(ra) \
330    { \    { \
331    printf("match() returned %d from line %d ", ra, __LINE__); \    printf("match() returned %d from line %d\n", ra, __LINE__); \
332    return ra; \    return ra; \
333    }    }
334  #else  #else
# Line 317  argument of match(), which never changes Line 348  argument of match(), which never changes
348    
349  #define RMATCH(ra,rb,rc,rd,re,rw)\  #define RMATCH(ra,rb,rc,rd,re,rw)\
350    {\    {\
351    heapframe *newframe = (heapframe *)(PUBL(stack_malloc))(sizeof(heapframe));\    heapframe *newframe = frame->Xnextframe;\
352    if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\    if (newframe == NULL)\
353    frame->Xwhere = rw; \      {\
354        newframe = (heapframe *)(PUBL(stack_malloc))(sizeof(heapframe));\
355        if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\
356        newframe->Xnextframe = NULL;\
357        frame->Xnextframe = newframe;\
358        }\
359      frame->Xwhere = rw;\
360    newframe->Xeptr = ra;\    newframe->Xeptr = ra;\
361    newframe->Xecode = rb;\    newframe->Xecode = rb;\
362    newframe->Xmstart = mstart;\    newframe->Xmstart = mstart;\
# Line 338  argument of match(), which never changes Line 375  argument of match(), which never changes
375    {\    {\
376    heapframe *oldframe = frame;\    heapframe *oldframe = frame;\
377    frame = oldframe->Xprevframe;\    frame = oldframe->Xprevframe;\
   if (oldframe != &frame_zero) (PUBL(stack_free))(oldframe);\  
378    if (frame != NULL)\    if (frame != NULL)\
379      {\      {\
380      rrc = ra;\      rrc = ra;\
# Line 352  argument of match(), which never changes Line 388  argument of match(), which never changes
388    
389  typedef struct heapframe {  typedef struct heapframe {
390    struct heapframe *Xprevframe;    struct heapframe *Xprevframe;
391      struct heapframe *Xnextframe;
392    
393    /* Function arguments that may change */    /* Function arguments that may change */
394    
# Line 382  typedef struct heapframe { Line 419  typedef struct heapframe {
419    
420  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
421    int Xprop_type;    int Xprop_type;
422    int Xprop_value;    unsigned int Xprop_value;
423    int Xprop_fail_result;    int Xprop_fail_result;
424    int Xoclength;    int Xoclength;
425    pcre_uchar Xocchars[6];    pcre_uchar Xocchars[6];
# Line 395  typedef struct heapframe { Line 432  typedef struct heapframe {
432    int Xlength;    int Xlength;
433    int Xmax;    int Xmax;
434    int Xmin;    int Xmin;
435    int Xnumber;    unsigned int Xnumber;
436    int Xoffset;    int Xoffset;
437    int Xop;    unsigned int Xop;
438    int Xsave_capture_last;    pcre_int32 Xsave_capture_last;
439    int Xsave_offset1, Xsave_offset2, Xsave_offset3;    int Xsave_offset1, Xsave_offset2, Xsave_offset3;
440    int Xstacksave[REC_STACK_SAVE_MAX];    int Xstacksave[REC_STACK_SAVE_MAX];
441    
# Line 483  so they can be ordinary variables in all Line 520  so they can be ordinary variables in all
520    
521  register int  rrc;         /* Returns from recursive calls */  register int  rrc;         /* Returns from recursive calls */
522  register int  i;           /* Used for loops not involving calls to RMATCH() */  register int  i;           /* Used for loops not involving calls to RMATCH() */
523  register unsigned int c;   /* Character values not kept over RMATCH() calls */  register pcre_uint32 c;    /* Character values not kept over RMATCH() calls */
524  register BOOL utf;         /* Local copy of UTF flag for speed */  register BOOL utf;         /* Local copy of UTF flag for speed */
525    
526  BOOL minimize, possessive; /* Quantifier options */  BOOL minimize, possessive; /* Quantifier options */
# Line 498  the top-level on the stack rather than m Line 535  the top-level on the stack rather than m
535  boost in many cases where there is not much "recursion". */  boost in many cases where there is not much "recursion". */
536    
537  #ifdef NO_RECURSE  #ifdef NO_RECURSE
538  heapframe frame_zero;  heapframe *frame = (heapframe *)md->match_frames_base;
 heapframe *frame = &frame_zero;  
 frame->Xprevframe = NULL;            /* Marks the top level */  
539    
540  /* Copy in the original argument variables */  /* Copy in the original argument variables */
541    
# Line 602  BOOL prev_is_word; Line 637  BOOL prev_is_word;
637    
638  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
639  int prop_type;  int prop_type;
640  int prop_value;  unsigned int prop_value;
641  int prop_fail_result;  int prop_fail_result;
642  int oclength;  int oclength;
643  pcre_uchar occhars[6];  pcre_uchar occhars[6];
# Line 613  int ctype; Line 648  int ctype;
648  int length;  int length;
649  int max;  int max;
650  int min;  int min;
651  int number;  unsigned int number;
652  int offset;  int offset;
653  int op;  unsigned int op;
654  int save_capture_last;  pcre_int32 save_capture_last;
655  int save_offset1, save_offset2, save_offset3;  int save_offset1, save_offset2, save_offset3;
656  int stacksave[REC_STACK_SAVE_MAX];  int stacksave[REC_STACK_SAVE_MAX];
657    
# Line 734  for (;;) Line 769  for (;;)
769      unaltered. */      unaltered. */
770    
771      else if (rrc == MATCH_SKIP_ARG &&      else if (rrc == MATCH_SKIP_ARG &&
772          STRCMP_UC_UC(ecode + 2, md->start_match_ptr) == 0)          STRCMP_UC_UC_TEST(ecode + 2, md->start_match_ptr) == 0)
773        {        {
774        md->start_match_ptr = eptr;        md->start_match_ptr = eptr;
775        RRETURN(MATCH_SKIP);        RRETURN(MATCH_SKIP);
# Line 744  for (;;) Line 779  for (;;)
779      case OP_FAIL:      case OP_FAIL:
780      RRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
781    
     /* COMMIT overrides PRUNE, SKIP, and THEN */  
   
782      case OP_COMMIT:      case OP_COMMIT:
783      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
784        eptrb, RM52);        eptrb, RM52);
785      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);  
786      RRETURN(MATCH_COMMIT);      RRETURN(MATCH_COMMIT);
787    
     /* PRUNE overrides THEN */  
   
788      case OP_PRUNE:      case OP_PRUNE:
789      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
790        eptrb, RM51);        eptrb, RM51);
791      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
792      RRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
793    
794      case OP_PRUNE_ARG:      case OP_PRUNE_ARG:
# Line 770  for (;;) Line 798  for (;;)
798        eptrb, RM56);        eptrb, RM56);
799      if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&      if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
800           md->mark == NULL) md->mark = ecode + 2;           md->mark == NULL) md->mark = ecode + 2;
801      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
802      RRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
803    
     /* SKIP overrides PRUNE and THEN */  
   
804      case OP_SKIP:      case OP_SKIP:
805      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
806        eptrb, RM53);        eptrb, RM53);
807      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
       RRETURN(rrc);  
808      md->start_match_ptr = eptr;   /* Pass back current position */      md->start_match_ptr = eptr;   /* Pass back current position */
809      RRETURN(MATCH_SKIP);      RRETURN(MATCH_SKIP);
810    
811      /* Note that, for Perl compatibility, SKIP with an argument does NOT set      /* Note that, for Perl compatibility, SKIP with an argument does NOT set
812      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
813      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
814        that failed and any that precede it (either they also failed, or were not
815        triggered). To do this, we maintain a count of executed SKIP_ARGs. If a
816        SKIP_ARG gets to top level, the match is re-run with md->ignore_skip_arg
817        set to the count of the one that failed. */
818    
819      case OP_SKIP_ARG:      case OP_SKIP_ARG:
820      if (md->ignore_skip_arg)      md->skip_arg_count++;
821        if (md->skip_arg_count <= md->ignore_skip_arg)
822        {        {
823        ecode += PRIV(OP_lengths)[*ecode] + ecode[1];        ecode += PRIV(OP_lengths)[*ecode] + ecode[1];
824        break;        break;
825        }        }
826      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
827        eptrb, RM57);        eptrb, RM57);
828      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
       RRETURN(rrc);  
829    
830      /* 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
831      returning the special MATCH_SKIP_ARG return code. This will either be      returning the special MATCH_SKIP_ARG return code. This will either be
832      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
833      with the md->ignore_skip_arg flag set. */      with md->ignore_skip_arg set to the value of md->skip_arg_count. */
834    
835      md->start_match_ptr = ecode + 2;      md->start_match_ptr = ecode + 2;
836      RRETURN(MATCH_SKIP_ARG);      RRETURN(MATCH_SKIP_ARG);
# Line 903  for (;;) Line 932  for (;;)
932        }        }
933      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
934        {        {
       md->match_function_type = MATCH_CBEGROUP;  
935        RMATCH(eptr, prev, offset_top, md, eptrb, RM66);        RMATCH(eptr, prev, offset_top, md, eptrb, RM66);
936        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
937        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
# Line 1032  for (;;) Line 1060  for (;;)
1060    
1061      for (;;)      for (;;)
1062        {        {
1063        if (op >= OP_SBRA || op == OP_ONCE) md->match_function_type = MATCH_CBEGROUP;        if (op >= OP_SBRA || op == OP_ONCE)
1064            md->match_function_type = MATCH_CBEGROUP;
1065    
1066        /* 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
1067        the pattern, and this is the final alternative, optimize as described        the pattern, and this is the final alternative, optimize as described
# Line 1047  for (;;) Line 1076  for (;;)
1076        /* In all other cases, we have to make another call to match(). */        /* In all other cases, we have to make another call to match(). */
1077    
1078        save_mark = md->mark;        save_mark = md->mark;
1079          save_capture_last = md->capture_last;
1080        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md, eptrb,        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md, eptrb,
1081          RM2);          RM2);
1082    
# Line 1078  for (;;) Line 1108  for (;;)
1108        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1109        md->mark = save_mark;        md->mark = save_mark;
1110        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1111          md->capture_last = save_capture_last;
1112        }        }
1113    
1114      RRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
# Line 1140  for (;;) Line 1171  for (;;)
1171            ecode = md->start_code + code_offset;            ecode = md->start_code + code_offset;
1172            save_capture_last = md->capture_last;            save_capture_last = md->capture_last;
1173            matched_once = TRUE;            matched_once = TRUE;
1174              mstart = md->start_match_ptr;    /* In case \K changed it */
1175            continue;            continue;
1176            }            }
1177    
# Line 1199  for (;;) Line 1231  for (;;)
1231      POSSESSIVE_NON_CAPTURE:      POSSESSIVE_NON_CAPTURE:
1232      matched_once = FALSE;      matched_once = FALSE;
1233      code_offset = (int)(ecode - md->start_code);      code_offset = (int)(ecode - md->start_code);
1234        save_capture_last = md->capture_last;
1235    
1236      for (;;)      for (;;)
1237        {        {
# Line 1211  for (;;) Line 1244  for (;;)
1244          eptr = md->end_match_ptr;          eptr = md->end_match_ptr;
1245          ecode = md->start_code + code_offset;          ecode = md->start_code + code_offset;
1246          matched_once = TRUE;          matched_once = TRUE;
1247            mstart = md->start_match_ptr;   /* In case \K reset it */
1248          continue;          continue;
1249          }          }
1250    
# Line 1228  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 1239  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 1483  for (;;) Line 1442  for (;;)
1442    
1443      else      else
1444        {        {
       ecode += 1 + LINK_SIZE;  
1445        }        }
1446      break;      break;
1447    
# Line 1492  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 1500  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 1563  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    
       if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);  
       ecode += GET(ecode, 1);  
1541        md->mark = save_mark;        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);
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 1587  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:
# Line 1609  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 1677  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 1688  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 1717  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 1734  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 1757  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 1769  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 1785  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 1918  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 1975  for (;;) Line 2007  for (;;)
2007    
2008      if (*ecode == OP_KETRPOS)      if (*ecode == OP_KETRPOS)
2009        {        {
2010          md->start_match_ptr = mstart;    /* In case \K reset it */
2011        md->end_match_ptr = eptr;        md->end_match_ptr = eptr;
2012        md->end_offset_top = offset_top;        md->end_offset_top = offset_top;
2013        RRETURN(MATCH_KETRPOS);        RRETURN(MATCH_KETRPOS);
# Line 1999  for (;;) Line 2032  for (;;)
2032          }          }
2033        if (*prev >= OP_SBRA)    /* Could match an empty string */        if (*prev >= OP_SBRA)    /* Could match an empty string */
2034          {          {
         md->match_function_type = MATCH_CBEGROUP;  
2035          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);
2036          RRETURN(rrc);          RRETURN(rrc);
2037          }          }
# Line 2008  for (;;) Line 2040  for (;;)
2040        }        }
2041      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
2042        {        {
       if (*prev >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;  
2043        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);
2044        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;
2045        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
# Line 2065  for (;;) Line 2096  for (;;)
2096    
2097      case OP_DOLLM:      case OP_DOLLM:
2098      if (eptr < md->end_subject)      if (eptr < md->end_subject)
2099        {        {
2100        if (!IS_NEWLINE(eptr))        if (!IS_NEWLINE(eptr))
2101          {          {
2102          if (md->partial != 0 &&          if (md->partial != 0 &&
2103              eptr + 1 >= md->end_subject &&              eptr + 1 >= md->end_subject &&
2104              NLBLOCK->nltype == NLTYPE_FIXED &&              NLBLOCK->nltype == NLTYPE_FIXED &&
2105              NLBLOCK->nllen == 2 &&              NLBLOCK->nllen == 2 &&
2106              *eptr == NLBLOCK->nl[0])              RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2107            {            {
2108            md->hitend = TRUE;            md->hitend = TRUE;
2109            if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);            if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2110            }            }
2111          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2112          }          }
2113        }        }
2114      else      else
2115        {        {
# Line 2115  for (;;) Line 2146  for (;;)
2146        if (md->partial != 0 &&        if (md->partial != 0 &&
2147            eptr + 1 >= md->end_subject &&            eptr + 1 >= md->end_subject &&
2148            NLBLOCK->nltype == NLTYPE_FIXED &&            NLBLOCK->nltype == NLTYPE_FIXED &&
2149            NLBLOCK->nllen == 2 &&            NLBLOCK->nllen == 2 &&
2150            *eptr == NLBLOCK->nl[0])            RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2151          {          {
2152          md->hitend = TRUE;          md->hitend = TRUE;
2153          if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);          if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2154          }          }
2155        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2156        }        }
2157    
2158      /* Either at end of string or \n before end. */      /* Either at end of string or \n before end. */
2159    
# Line 2258  for (;;) Line 2289  for (;;)
2289      if (md->partial != 0 &&      if (md->partial != 0 &&
2290          eptr + 1 >= md->end_subject &&          eptr + 1 >= md->end_subject &&
2291          NLBLOCK->nltype == NLTYPE_FIXED &&          NLBLOCK->nltype == NLTYPE_FIXED &&
2292          NLBLOCK->nllen == 2 &&          NLBLOCK->nllen == 2 &&
2293          *eptr == NLBLOCK->nl[0])          RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2294        {        {
2295        md->hitend = TRUE;        md->hitend = TRUE;
2296        if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);        if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2297        }        }
2298    
2299      /* Fall through */      /* Fall through */
2300    
2301      /* Match any single character whatsoever. */      /* Match any single character whatsoever. */
2302    
2303      case OP_ALLANY:      case OP_ALLANY:
2304      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 */
2305        {                            /* not be updated before SCHECK_PARTIAL. */        {                            /* not be updated before SCHECK_PARTIAL. */
# Line 2408  for (;;) Line 2439  for (;;)
2439        {        {
2440        default: RRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2441    
2442        case 0x000d:        case CHAR_CR:
2443        if (eptr >= md->end_subject)        if (eptr >= md->end_subject)
2444          {          {
2445          SCHECK_PARTIAL();          SCHECK_PARTIAL();
2446          }          }
2447        else if (*eptr == 0x0a) eptr++;        else if (RAWUCHARTEST(eptr) == CHAR_LF) eptr++;
2448        break;        break;
2449    
2450        case 0x000a:        case CHAR_LF:
2451        break;        break;
2452    
2453        case 0x000b:        case CHAR_VT:
2454        case 0x000c:        case CHAR_FF:
2455        case 0x0085:        case CHAR_NEL:
2456    #ifndef EBCDIC
2457        case 0x2028:        case 0x2028:
2458        case 0x2029:        case 0x2029:
2459    #endif  /* Not EBCDIC */
2460        if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);        if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
2461        break;        break;
2462        }        }
# Line 2439  for (;;) Line 2472  for (;;)
2472      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2473      switch(c)      switch(c)
2474        {        {
2475          HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
2476        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);  
2477        }        }
2478      ecode++;      ecode++;
2479      break;      break;
# Line 2473  for (;;) Line 2487  for (;;)
2487      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2488      switch(c)      switch(c)
2489        {        {
2490          HSPACE_CASES: break;  /* Byte and multibyte cases */
2491        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;  
2492        }        }
2493      ecode++;      ecode++;
2494      break;      break;
# Line 2507  for (;;) Line 2502  for (;;)
2502      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2503      switch(c)      switch(c)
2504        {        {
2505          VSPACE_CASES: RRETURN(MATCH_NOMATCH);
2506        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);  
2507        }        }
2508      ecode++;      ecode++;
2509      break;      break;
# Line 2529  for (;;) Line 2517  for (;;)
2517      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2518      switch(c)      switch(c)
2519        {        {
2520          VSPACE_CASES: break;
2521        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;  
2522        }        }
2523      ecode++;      ecode++;
2524      break;      break;
# Line 2555  for (;;) Line 2536  for (;;)
2536        }        }
2537      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2538        {        {
2539          const pcre_uint32 *cp;
2540        const ucd_record *prop = GET_UCD(c);        const ucd_record *prop = GET_UCD(c);
2541    
2542        switch(ecode[1])        switch(ecode[1])
# Line 2593  for (;;) Line 2575  for (;;)
2575            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2576          break;          break;
2577    
2578          case PT_SPACE:    /* Perl space */          /* Perl space used to exclude VT, but from Perl 5.18 it is included,
2579          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||          which means that Perl space and POSIX space are now identical. PCRE
2580               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;  
2581    
2582            case PT_SPACE:    /* Perl space */
2583          case PT_PXSPACE:  /* POSIX space */          case PT_PXSPACE:  /* POSIX space */
2584          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||          switch(c)
2585               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||            {
2586               c == CHAR_FF || c == CHAR_CR)            HSPACE_CASES:
2587                 == (op == OP_NOTPROP))            VSPACE_CASES:
2588            RRETURN(MATCH_NOMATCH);            if (op == OP_NOTPROP) RRETURN(MATCH_NOMATCH);
2589              break;
2590    
2591              default:
2592              if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z) ==
2593                (op == OP_NOTPROP)) RRETURN(MATCH_NOMATCH);
2594              break;
2595              }
2596          break;          break;
2597    
2598          case PT_WORD:          case PT_WORD:
# Line 2615  for (;;) Line 2602  for (;;)
2602            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2603          break;          break;
2604    
2605            case PT_CLIST:
2606            cp = PRIV(ucd_caseless_sets) + ecode[2];
2607            for (;;)
2608              {
2609              if (c < *cp)
2610                { if (op == OP_PROP) { RRETURN(MATCH_NOMATCH); } else break; }
2611              if (c == *cp++)
2612                { if (op == OP_PROP) break; else { RRETURN(MATCH_NOMATCH); } }
2613              }
2614            break;
2615    
2616            case PT_UCNC:
2617            if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
2618                 c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
2619                 c >= 0xe000) == (op == OP_NOTPROP))
2620              RRETURN(MATCH_NOMATCH);
2621            break;
2622    
2623          /* This should never occur */          /* This should never occur */
2624    
2625          default:          default:
# Line 2634  for (;;) Line 2639  for (;;)
2639        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2640        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2641        }        }
2642      GETCHARINCTEST(c, eptr);      else
     if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
     while (eptr < md->end_subject)  
2643        {        {
2644        int len = 1;        int lgb, rgb;
2645        if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }        GETCHARINCTEST(c, eptr);
2646        if (UCD_CATEGORY(c) != ucp_M) break;        lgb = UCD_GRAPHBREAK(c);
2647        eptr += len;        while (eptr < md->end_subject)
2648            {
2649            int len = 1;
2650            if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
2651            rgb = UCD_GRAPHBREAK(c);
2652            if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
2653            lgb = rgb;
2654            eptr += len;
2655            }
2656        }        }
2657      CHECK_PARTIAL();      CHECK_PARTIAL();
2658      ecode++;      ecode++;
2659      break;      break;
2660  #endif  #endif  /* SUPPORT_UCP */
2661    
2662    
2663      /* 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 2655  for (;;) Line 2666  for (;;)
2666      similar code to character type repeats - written out again for speed.      similar code to character type repeats - written out again for speed.
2667      However, if the referenced string is the empty string, always treat      However, if the referenced string is the empty string, always treat
2668      it as matched, any number of times (otherwise there could be infinite      it as matched, any number of times (otherwise there could be infinite
2669      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:  
2670    
2671      (a) In the default, Perl-compatible state, set the length negative;      (a) In the default, Perl-compatible state, set the length negative;
2672      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 2673  for (;;) Line 2676  for (;;)
2676      so that the back reference matches an empty string.      so that the back reference matches an empty string.
2677    
2678      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
2679      referenced subpattern. */      referenced subpattern.
2680    
2681        The OP_REF and OP_REFI opcodes are used for a reference to a numbered group
2682        or to a non-duplicated named group. For a duplicated named group, OP_DNREF
2683        and OP_DNREFI are used. In this case we must scan the list of groups to
2684        which the name refers, and use the first one that is set. */
2685    
2686        case OP_DNREF:
2687        case OP_DNREFI:
2688        caseless = op == OP_DNREFI;
2689          {
2690          int count = GET2(ecode, 1+IMM2_SIZE);
2691          pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
2692          ecode += 1 + 2*IMM2_SIZE;
2693    
2694          while (count-- > 0)
2695            {
2696            offset = GET2(slot, 0) << 1;
2697            if (offset < offset_top && md->offset_vector[offset] >= 0) break;
2698            slot += md->name_entry_size;
2699            }
2700          if (count < 0)
2701            length = (md->jscript_compat)? 0 : -1;
2702          else
2703            length = md->offset_vector[offset+1] - md->offset_vector[offset];
2704          }
2705        goto REF_REPEAT;
2706    
2707        case OP_REF:
2708        case OP_REFI:
2709        caseless = op == OP_REFI;
2710        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
2711        ecode += 1 + IMM2_SIZE;
2712      if (offset >= offset_top || md->offset_vector[offset] < 0)      if (offset >= offset_top || md->offset_vector[offset] < 0)
2713        length = (md->jscript_compat)? 0 : -1;        length = (md->jscript_compat)? 0 : -1;
2714      else      else
# Line 2682  for (;;) Line 2716  for (;;)
2716    
2717      /* Set up for repetition, or handle the non-repeated case */      /* Set up for repetition, or handle the non-repeated case */
2718    
2719        REF_REPEAT:
2720      switch (*ecode)      switch (*ecode)
2721        {        {
2722        case OP_CRSTAR:        case OP_CRSTAR:
# Line 2709  for (;;) Line 2744  for (;;)
2744        default:               /* No repeat follows */        default:               /* No repeat follows */
2745        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)
2746          {          {
2747          if (length == -2) eptr = md->end_subject;   /* Partial match */          if (length == -2) eptr = md->end_subject;   /* Partial match */
2748          CHECK_PARTIAL();          CHECK_PARTIAL();
2749          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2750          }          }
# Line 2735  for (;;) Line 2770  for (;;)
2770        int slength;        int slength;
2771        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2772          {          {
2773          if (slength == -2) eptr = md->end_subject;   /* Partial match */          if (slength == -2) eptr = md->end_subject;   /* Partial match */
2774          CHECK_PARTIAL();          CHECK_PARTIAL();
2775          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2776          }          }
# Line 2759  for (;;) Line 2794  for (;;)
2794          if (fi >= max) RRETURN(MATCH_NOMATCH);          if (fi >= max) RRETURN(MATCH_NOMATCH);
2795          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2796            {            {
2797            if (slength == -2) eptr = md->end_subject;   /* Partial match */            if (slength == -2) eptr = md->end_subject;   /* Partial match */
2798            CHECK_PARTIAL();            CHECK_PARTIAL();
2799            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2800            }            }
# Line 2778  for (;;) Line 2813  for (;;)
2813          int slength;          int slength;
2814          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2815            {            {
2816            /* Can't use CHECK_PARTIAL because we don't want to update eptr in            /* Can't use CHECK_PARTIAL because we don't want to update eptr in
2817            the soft partial matching case. */            the soft partial matching case. */
2818    
2819            if (slength == -2 && md->partial != 0 &&            if (slength == -2 && md->partial != 0 &&
2820                md->end_subject > md->start_used_ptr)                md->end_subject > md->start_used_ptr)
2821              {              {
2822              md->hitend = TRUE;              md->hitend = TRUE;
2823              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2824              }              }
# Line 2791  for (;;) Line 2826  for (;;)
2826            }            }
2827          eptr += slength;          eptr += slength;
2828          }          }
2829    
2830        while (eptr >= pp)        while (eptr >= pp)
2831          {          {
2832          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);
# Line 2830  for (;;) Line 2865  for (;;)
2865          case OP_CRMINPLUS:          case OP_CRMINPLUS:
2866          case OP_CRQUERY:          case OP_CRQUERY:
2867          case OP_CRMINQUERY:          case OP_CRMINQUERY:
2868            case OP_CRPOSSTAR:
2869            case OP_CRPOSPLUS:
2870            case OP_CRPOSQUERY:
2871          c = *ecode++ - OP_CRSTAR;          c = *ecode++ - OP_CRSTAR;
2872          minimize = (c & 1) != 0;          if (c < OP_CRPOSSTAR - OP_CRSTAR) minimize = (c & 1) != 0;
2873            else possessive = TRUE;
2874          min = rep_min[c];                 /* Pick up values from tables; */          min = rep_min[c];                 /* Pick up values from tables; */
2875          max = rep_max[c];                 /* zero for max => infinity */          max = rep_max[c];                 /* zero for max => infinity */
2876          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 2839  for (;;) Line 2878  for (;;)
2878    
2879          case OP_CRRANGE:          case OP_CRRANGE:
2880          case OP_CRMINRANGE:          case OP_CRMINRANGE:
2881            case OP_CRPOSRANGE:
2882          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
2883            possessive = (*ecode == OP_CRPOSRANGE);
2884          min = GET2(ecode, 1);          min = GET2(ecode, 1);
2885          max = GET2(ecode, 1 + IMM2_SIZE);          max = GET2(ecode, 1 + IMM2_SIZE);
2886          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 2981  for (;;) Line 3022  for (;;)
3022                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
3023              eptr += len;              eptr += len;
3024              }              }
3025    
3026              if (possessive) continue;    /* No backtracking */
3027    
3028            for (;;)            for (;;)
3029              {              {
3030              RMATCH(eptr, ecode, offset_top, md, eptrb, RM18);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM18);
# Line 3011  for (;;) Line 3055  for (;;)
3055                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
3056              eptr++;              eptr++;
3057              }              }
3058    
3059              if (possessive) continue;    /* No backtracking */
3060    
3061            while (eptr >= pp)            while (eptr >= pp)
3062              {              {
3063              RMATCH(eptr, ecode, offset_top, md, eptrb, RM19);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM19);
# Line 3026  for (;;) Line 3073  for (;;)
3073      /* Control never gets here */      /* Control never gets here */
3074    
3075    
3076      /* Match an extended character class. This opcode is encountered only      /* Match an extended character class. In the 8-bit library, this opcode is
3077      when UTF-8 mode mode is supported. Nevertheless, we may not be in UTF-8      encountered only when UTF-8 mode mode is supported. In the 16-bit and
3078      mode, because Unicode properties are supported in non-UTF-8 mode. */      32-bit libraries, codepoints greater than 255 may be encountered even when
3079        UTF is not supported. */
3080    
3081  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
3082      case OP_XCLASS:      case OP_XCLASS:
# Line 3044  for (;;) Line 3092  for (;;)
3092          case OP_CRMINPLUS:          case OP_CRMINPLUS:
3093          case OP_CRQUERY:          case OP_CRQUERY:
3094          case OP_CRMINQUERY:          case OP_CRMINQUERY:
3095            case OP_CRPOSSTAR:
3096            case OP_CRPOSPLUS:
3097            case OP_CRPOSQUERY:
3098          c = *ecode++ - OP_CRSTAR;          c = *ecode++ - OP_CRSTAR;
3099          minimize = (c & 1) != 0;          if (c < OP_CRPOSSTAR - OP_CRSTAR) minimize = (c & 1) != 0;
3100            else possessive = TRUE;
3101          min = rep_min[c];                 /* Pick up values from tables; */          min = rep_min[c];                 /* Pick up values from tables; */
3102          max = rep_max[c];                 /* zero for max => infinity */          max = rep_max[c];                 /* zero for max => infinity */
3103          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 3053  for (;;) Line 3105  for (;;)
3105    
3106          case OP_CRRANGE:          case OP_CRRANGE:
3107          case OP_CRMINRANGE:          case OP_CRMINRANGE:
3108            case OP_CRPOSRANGE:
3109          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
3110            possessive = (*ecode == OP_CRPOSRANGE);
3111          min = GET2(ecode, 1);          min = GET2(ecode, 1);
3112          max = GET2(ecode, 1 + IMM2_SIZE);          max = GET2(ecode, 1 + IMM2_SIZE);
3113          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 3125  for (;;) Line 3179  for (;;)
3179            if (!PRIV(xclass)(c, data, utf)) break;            if (!PRIV(xclass)(c, data, utf)) break;
3180            eptr += len;            eptr += len;
3181            }            }
3182    
3183            if (possessive) continue;    /* No backtracking */
3184    
3185          for(;;)          for(;;)
3186            {            {
3187            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);
# Line 3155  for (;;) Line 3212  for (;;)
3212          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
3213          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3214          }          }
3215        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);        while (length-- > 0) if (*ecode++ != RAWUCHARINC(eptr)) RRETURN(MATCH_NOMATCH);
3216        }        }
3217      else      else
3218  #endif  #endif
# Line 3195  for (;;) Line 3252  for (;;)
3252    
3253        if (fc < 128)        if (fc < 128)
3254          {          {
3255          if (md->lcc[fc]          pcre_uint32 cc = RAWUCHAR(eptr);
3256              != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);          if (md->lcc[fc] != TABLE_GET(cc, md->lcc, cc)) RRETURN(MATCH_NOMATCH);
3257          ecode++;          ecode++;
3258          eptr++;          eptr++;
3259          }          }
# Line 3207  for (;;) Line 3264  for (;;)
3264    
3265        else        else
3266          {          {
3267          unsigned int dc;          pcre_uint32 dc;
3268          GETCHARINC(dc, eptr);          GETCHARINC(dc, eptr);
3269          ecode += length;          ecode += length;
3270    
# Line 3300  for (;;) Line 3357  for (;;)
3357      max = rep_max[c];                 /* zero for max => infinity */      max = rep_max[c];                 /* zero for max => infinity */
3358      if (max == 0) max = INT_MAX;      if (max == 0) max = INT_MAX;
3359    
3360      /* Common code for all repeated single-character matches. */      /* Common code for all repeated single-character matches. We first check
3361        for the minimum number of characters. If the minimum equals the maximum, we
3362        are done. Otherwise, if minimizing, check the rest of the pattern for a
3363        match; if there isn't one, advance up to the maximum, one character at a
3364        time.
3365    
3366        If maximizing, advance up to the maximum number of matching characters,
3367        until eptr is past the end of the maximum run. If possessive, we are
3368        then done (no backing up). Otherwise, match at this position; anything
3369        other than no match is immediately returned. For nomatch, back up one
3370        character, unless we are matching \R and the last thing matched was
3371        \r\n, in which case, back up two bytes. When we reach the first optional
3372        character position, we can save stack by doing a tail recurse.
3373    
3374        The various UTF/non-UTF and caseful/caseless cases are handled separately,
3375        for speed. */
3376    
3377      REPEATCHAR:      REPEATCHAR:
3378  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
# Line 3317  for (;;) Line 3389  for (;;)
3389        if (length > 1)        if (length > 1)
3390          {          {
3391  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3392          unsigned int othercase;          pcre_uint32 othercase;
3393          if (op >= OP_STARI &&     /* Caseless */          if (op >= OP_STARI &&     /* Caseless */
3394              (othercase = UCD_OTHERCASE(fc)) != fc)              (othercase = UCD_OTHERCASE(fc)) != fc)
3395            oclength = PRIV(ord2utf)(othercase, occhars);            oclength = PRIV(ord2utf)(othercase, occhars);
# Line 3384  for (;;) Line 3456  for (;;)
3456                }                }
3457              }              }
3458    
3459            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
   
3460            for(;;)            for(;;)
3461              {              {
3462                if (eptr == pp) goto TAIL_RECURSE;
3463              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);
3464              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
             if (eptr == pp) { RRETURN(MATCH_NOMATCH); }  
3465  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3466              eptr--;              eptr--;
3467              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 3421  for (;;) Line 3492  for (;;)
3492      maximizing, find the maximum number of characters and work backwards. */      maximizing, find the maximum number of characters and work backwards. */
3493    
3494      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,
3495        max, eptr));        max, (char *)eptr));
3496    
3497      if (op >= OP_STARI)  /* Caseless */      if (op >= OP_STARI)  /* Caseless */
3498        {        {
# Line 3444  for (;;) Line 3515  for (;;)
3515    
3516        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
3517          {          {
3518            pcre_uint32 cc;                 /* Faster than pcre_uchar */
3519          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3520            {            {
3521            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3522            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3523            }            }
3524          if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);          cc = RAWUCHARTEST(eptr);
3525            if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
3526          eptr++;          eptr++;
3527          }          }
3528        if (min == max) continue;        if (min == max) continue;
# Line 3457  for (;;) Line 3530  for (;;)
3530          {          {
3531          for (fi = min;; fi++)          for (fi = min;; fi++)
3532            {            {
3533              pcre_uint32 cc;               /* Faster than pcre_uchar */
3534            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);
3535            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3536            if (fi >= max) RRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
# Line 3465  for (;;) Line 3539  for (;;)
3539              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3540              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3541              }              }
3542            if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);            cc = RAWUCHARTEST(eptr);
3543              if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
3544            eptr++;            eptr++;
3545            }            }
3546          /* Control never gets here */          /* Control never gets here */
# Line 3475  for (;;) Line 3550  for (;;)
3550          pp = eptr;          pp = eptr;
3551          for (i = min; i < max; i++)          for (i = min; i < max; i++)
3552            {            {
3553              pcre_uint32 cc;               /* Faster than pcre_uchar */
3554            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3555              {              {
3556              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3557              break;              break;
3558              }              }
3559            if (fc != *eptr && foc != *eptr) break;            cc = RAWUCHARTEST(eptr);
3560              if (fc != cc && foc != cc) break;
3561            eptr++;            eptr++;
3562            }            }
3563            if (possessive) continue;       /* No backtracking */
3564          if (possessive) continue;          for (;;)
   
         while (eptr >= pp)  
3565            {            {
3566              if (eptr == pp) goto TAIL_RECURSE;
3567            RMATCH(eptr, ecode, offset_top, md, eptrb, RM25);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM25);
3568            eptr--;            eptr--;
3569            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3570            }            }
3571          RRETURN(MATCH_NOMATCH);          /* Control never gets here */
3572          }          }
       /* Control never gets here */  
3573        }        }
3574    
3575      /* Caseful comparisons (includes all multi-byte characters) */      /* Caseful comparisons (includes all multi-byte characters) */
# Line 3508  for (;;) Line 3583  for (;;)
3583            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3584            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3585            }            }
3586          if (fc != *eptr++) RRETURN(MATCH_NOMATCH);          if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
3587          }          }
3588    
3589        if (min == max) continue;        if (min == max) continue;
# Line 3525  for (;;) Line 3600  for (;;)
3600              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3601              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3602              }              }
3603            if (fc != *eptr++) RRETURN(MATCH_NOMATCH);            if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
3604            }            }
3605          /* Control never gets here */          /* Control never gets here */
3606          }          }
# Line 3539  for (;;) Line 3614  for (;;)
3614              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3615              break;              break;
3616              }              }
3617            if (fc != *eptr) break;            if (fc != RAWUCHARTEST(eptr)) break;
3618            eptr++;            eptr++;
3619            }            }
3620          if (possessive) continue;          if (possessive) continue;    /* No backtracking */
3621            for (;;)
         while (eptr >= pp)  
3622            {            {
3623              if (eptr == pp) goto TAIL_RECURSE;
3624            RMATCH(eptr, ecode, offset_top, md, eptrb, RM27);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM27);
3625            eptr--;            eptr--;
3626            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3627            }            }
3628          RRETURN(MATCH_NOMATCH);          /* Control never gets here */
3629          }          }
3630        }        }
3631      /* Control never gets here */      /* Control never gets here */
# Line 3568  for (;;) Line 3643  for (;;)
3643  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3644      if (utf)      if (utf)
3645        {        {
3646        register unsigned int ch, och;        register pcre_uint32 ch, och;
3647    
3648        ecode++;        ecode++;
3649        GETCHARINC(ch, ecode);        GETCHARINC(ch, ecode);
# Line 3595  for (;;) Line 3670  for (;;)
3670      else      else
3671  #endif  #endif
3672        {        {
3673        register unsigned int ch = ecode[1];        register pcre_uint32 ch = ecode[1];
3674        c = *eptr++;        c = *eptr++;
3675        if (ch == c || (op == OP_NOTI && TABLE_GET(ch, md->fcc, ch) == c))        if (ch == c || (op == OP_NOTI && TABLE_GET(ch, md->fcc, ch) == c))
3676          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
# Line 3690  for (;;) Line 3765  for (;;)
3765      characters and work backwards. */      characters and work backwards. */
3766    
3767      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,
3768        max, eptr));        max, (char *)eptr));
3769    
3770      if (op >= OP_NOTSTARI)     /* Caseless */      if (op >= OP_NOTSTARI)     /* Caseless */
3771        {        {
# Line 3709  for (;;) Line 3784  for (;;)
3784  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3785        if (utf)        if (utf)
3786          {          {
3787          register unsigned int d;          register pcre_uint32 d;
3788          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3789            {            {
3790            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
# Line 3722  for (;;) Line 3797  for (;;)
3797            }            }
3798          }          }
3799        else        else
3800  #endif  #endif  /* SUPPORT_UTF */
3801        /* Not UTF mode */        /* Not UTF mode */
3802          {          {
3803          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3744  for (;;) Line 3819  for (;;)
3819  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3820          if (utf)          if (utf)
3821            {            {
3822            register unsigned int d;            register pcre_uint32 d;
3823            for (fi = min;; fi++)            for (fi = min;; fi++)
3824              {              {
3825              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);
# Line 3760  for (;;) Line 3835  for (;;)
3835              }              }
3836            }            }
3837          else          else
3838  #endif  #endif  /*SUPPORT_UTF */
3839          /* Not UTF mode */          /* Not UTF mode */
3840            {            {
3841            for (fi = min;; fi++)            for (fi = min;; fi++)
# Line 3789  for (;;) Line 3864  for (;;)
3864  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3865          if (utf)          if (utf)
3866            {            {
3867            register unsigned int d;            register pcre_uint32 d;
3868            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3869              {              {
3870              int len = 1;              int len = 1;
# Line 3802  for (;;) Line 3877  for (;;)
3877              if (fc == d || (unsigned int)foc == d) break;              if (fc == d || (unsigned int)foc == d) break;
3878              eptr += len;              eptr += len;
3879              }              }
3880            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
3881            for(;;)            for(;;)
3882              {              {
3883                if (eptr == pp) goto TAIL_RECURSE;
3884              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);
3885              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3886              if (eptr-- == pp) break;        /* Stop if tried at original pos */              eptr--;
3887              BACKCHAR(eptr);              BACKCHAR(eptr);
3888              }              }
3889            }            }
3890          else          else
3891  #endif  #endif  /* SUPPORT_UTF */
3892          /* Not UTF mode */          /* Not UTF mode */
3893            {            {
3894            for (i = min; i < max; i++)            for (i = min; i < max; i++)
# Line 3825  for (;;) Line 3901  for (;;)
3901              if (fc == *eptr || foc == *eptr) break;              if (fc == *eptr || foc == *eptr) break;
3902              eptr++;              eptr++;
3903              }              }
3904            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
3905            while (eptr >= pp)            for (;;)
3906              {              {
3907                if (eptr == pp) goto TAIL_RECURSE;
3908              RMATCH(eptr, ecode, offset_top, md, eptrb, RM31);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM31);
3909              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3910              eptr--;              eptr--;
3911              }              }
3912            }            }
3913            /* Control never gets here */
         RRETURN(MATCH_NOMATCH);  
3914          }          }
       /* Control never gets here */  
3915        }        }
3916    
3917      /* Caseful comparisons */      /* Caseful comparisons */
# Line 3846  for (;;) Line 3921  for (;;)
3921  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3922        if (utf)        if (utf)
3923          {          {
3924          register unsigned int d;          register pcre_uint32 d;
3925          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3926            {            {
3927            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
# Line 3880  for (;;) Line 3955  for (;;)
3955  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3956          if (utf)          if (utf)
3957            {            {
3958            register unsigned int d;            register pcre_uint32 d;
3959            for (fi = min;; fi++)            for (fi = min;; fi++)
3960              {              {
3961              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);
# Line 3924  for (;;) Line 3999  for (;;)
3999  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
4000          if (utf)          if (utf)
4001            {            {
4002            register unsigned int d;            register pcre_uint32 d;
4003            for (i = min; i < max; i++)            for (i = min; i < max; i++)
4004              {              {
4005              int len = 1;              int len = 1;
# Line 3937  for (;;) Line 4012  for (;;)
4012              if (fc == d) break;              if (fc == d) break;
4013              eptr += len;              eptr += len;
4014              }              }
4015            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
4016            for(;;)            for(;;)
4017              {              {
4018                if (eptr == pp) goto TAIL_RECURSE;
4019              RMATCH(eptr, ecode, offset_top, md, eptrb, RM34);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM34);
4020              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4021              if (eptr-- == pp) break;        /* Stop if tried at original pos */              eptr--;
4022              BACKCHAR(eptr);              BACKCHAR(eptr);
4023              }              }
4024            }            }
# Line 3960  for (;;) Line 4036  for (;;)
4036              if (fc == *eptr) break;              if (fc == *eptr) break;
4037              eptr++;              eptr++;
4038              }              }
4039            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
4040            while (eptr >= pp)            for (;;)
4041              {              {
4042                if (eptr == pp) goto TAIL_RECURSE;
4043              RMATCH(eptr, ecode, offset_top, md, eptrb, RM35);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM35);
4044              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4045              eptr--;              eptr--;
4046              }              }
4047            }            }
4048            /* Control never gets here */
         RRETURN(MATCH_NOMATCH);  
4049          }          }
4050        }        }
4051      /* Control never gets here */      /* Control never gets here */
# Line 4151  for (;;) Line 4227  for (;;)
4227              }              }
4228            break;            break;
4229    
4230              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
4231              which means that Perl space and POSIX space are now identical. PCRE
4232              was changed at release 8.34. */
4233    
4234            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
4235              case PT_PXSPACE:  /* POSIX space */
4236            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4237              {              {
4238              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4160  for (;;) Line 4241  for (;;)
4241                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4242                }                }
4243              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4244              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              switch(c)
4245                   c == CHAR_FF || c == CHAR_CR)                {
4246                     == prop_fail_result)                HSPACE_CASES:
4247                RRETURN(MATCH_NOMATCH);                VSPACE_CASES:
4248                  if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4249                  break;
4250    
4251                  default:
4252                  if ((UCD_CATEGORY(c) == ucp_Z) == prop_fail_result)
4253                    RRETURN(MATCH_NOMATCH);
4254                  break;
4255                  }
4256              }              }
4257            break;            break;
4258    
4259            case PT_PXSPACE:  /* POSIX space */            case PT_WORD:
4260            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4261              {              {
4262                int category;
4263              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4264                {                {
4265                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4266                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4267                }                }
4268              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4269              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              category = UCD_CATEGORY(c);
4270                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)
4271                     == prop_fail_result)                     == prop_fail_result)
4272                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4273              }              }
4274            break;            break;
4275    
4276            case PT_WORD:            case PT_CLIST:
4277            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4278              {              {
4279              int category;              const pcre_uint32 *cp;
4280              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4281                {                {
4282                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4283                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4284                }                }
4285              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4286              category = UCD_CATEGORY(c);              cp = PRIV(ucd_caseless_sets) + prop_value;
4287              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)              for (;;)
4288                     == prop_fail_result)                {
4289                  if (c < *cp)
4290                    { if (prop_fail_result) break; else { RRETURN(MATCH_NOMATCH); } }
4291                  if (c == *cp++)
4292                    { if (prop_fail_result) { RRETURN(MATCH_NOMATCH); } else break; }
4293                  }
4294                }
4295              break;
4296    
4297              case PT_UCNC:
4298              for (i = 1; i <= min; i++)
4299                {
4300                if (eptr >= md->end_subject)
4301                  {
4302                  SCHECK_PARTIAL();
4303                  RRETURN(MATCH_NOMATCH);
4304                  }
4305                GETCHARINCTEST(c, eptr);
4306                if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
4307                     c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
4308                     c >= 0xe000) == prop_fail_result)
4309                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4310              }              }
4311            break;            break;
# Line 4219  for (;;) Line 4329  for (;;)
4329              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4330              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4331              }              }
4332            GETCHARINCTEST(c, eptr);            else
           if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
           while (eptr < md->end_subject)  
4333              {              {
4334              int len = 1;              int lgb, rgb;
4335              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
4336              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
4337              eptr += len;             while (eptr < md->end_subject)
4338                  {
4339                  int len = 1;
4340                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
4341                  rgb = UCD_GRAPHBREAK(c);
4342                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
4343                  lgb = rgb;
4344                  eptr += len;
4345                  }
4346              }              }
4347            CHECK_PARTIAL();            CHECK_PARTIAL();
4348            }            }
4349          }          }
4350    
# Line 4252  for (;;) Line 4368  for (;;)
4368            if (md->partial != 0 &&            if (md->partial != 0 &&
4369                eptr + 1 >= md->end_subject &&                eptr + 1 >= md->end_subject &&
4370                NLBLOCK->nltype == NLTYPE_FIXED &&                NLBLOCK->nltype == NLTYPE_FIXED &&
4371                NLBLOCK->nllen == 2 &&                NLBLOCK->nllen == 2 &&
4372                *eptr == NLBLOCK->nl[0])                RAWUCHAR(eptr) == NLBLOCK->nl[0])
4373              {              {
4374              md->hitend = TRUE;              md->hitend = TRUE;
4375              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4376              }              }
# Line 4294  for (;;) Line 4410  for (;;)
4410              {              {
4411              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4412    
4413              case 0x000d:              case CHAR_CR:
4414              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && RAWUCHAR(eptr) == CHAR_LF) eptr++;
4415              break;              break;
4416    
4417              case 0x000a:              case CHAR_LF:
4418              break;              break;
4419    
4420              case 0x000b:              case CHAR_VT:
4421              case 0x000c:              case CHAR_FF:
4422              case 0x0085:              case CHAR_NEL:
4423    #ifndef EBCDIC
4424              case 0x2028:              case 0x2028:
4425              case 0x2029:              case 0x2029:
4426    #endif  /* Not EBCDIC */
4427              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4428              break;              break;
4429              }              }
# Line 4323  for (;;) Line 4441  for (;;)
4441            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4442            switch(c)            switch(c)
4443              {              {
4444                HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
4445              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);  
4446              }              }
4447            }            }
4448          break;          break;
# Line 4359  for (;;) Line 4458  for (;;)
4458            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4459            switch(c)            switch(c)
4460              {              {
4461                HSPACE_CASES: break;  /* Byte and multibyte cases */
4462              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;  
4463              }              }
4464            }            }
4465          break;          break;
# Line 4395  for (;;) Line 4475  for (;;)
4475            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4476            switch(c)            switch(c)
4477              {              {
4478                VSPACE_CASES: RRETURN(MATCH_NOMATCH);
4479              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);  
4480              }              }
4481            }            }
4482          break;          break;
# Line 4419  for (;;) Line 4492  for (;;)
4492            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4493            switch(c)            switch(c)
4494              {              {
4495                VSPACE_CASES: break;
4496              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;  
4497              }              }
4498            }            }
4499          break;          break;
# Line 4449  for (;;) Line 4515  for (;;)
4515          case OP_DIGIT:          case OP_DIGIT:
4516          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4517            {            {
4518              pcre_uint32 cc;
4519            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4520              {              {
4521              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4522              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4523              }              }
4524            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_digit) == 0)            cc = RAWUCHAR(eptr);
4525              if (cc >= 128 || (md->ctypes[cc] & ctype_digit) == 0)
4526              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4527            eptr++;            eptr++;
4528            /* 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 4464  for (;;) Line 4532  for (;;)
4532          case OP_NOT_WHITESPACE:          case OP_NOT_WHITESPACE:
4533          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4534            {            {
4535              pcre_uint32 cc;
4536            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4537              {              {
4538              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4539              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4540              }              }
4541            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)            cc = RAWUCHAR(eptr);
4542              if (cc < 128 && (md->ctypes[cc] & ctype_space) != 0)
4543              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4544            eptr++;            eptr++;
4545            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
# Line 4479  for (;;) Line 4549  for (;;)
4549          case OP_WHITESPACE:          case OP_WHITESPACE:
4550          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4551            {            {
4552              pcre_uint32 cc;
4553            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4554              {              {
4555              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4556              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4557              }              }
4558            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_space) == 0)            cc = RAWUCHAR(eptr);
4559              if (cc >= 128 || (md->ctypes[cc] & ctype_space) == 0)
4560              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4561            eptr++;            eptr++;
4562            /* 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 4494  for (;;) Line 4566  for (;;)
4566          case OP_NOT_WORDCHAR:          case OP_NOT_WORDCHAR:
4567          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4568            {            {
4569              pcre_uint32 cc;
4570            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4571              {              {
4572              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4573              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4574              }              }
4575            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)            cc = RAWUCHAR(eptr);
4576              if (cc < 128 && (md->ctypes[cc] & ctype_word) != 0)
4577              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4578            eptr++;            eptr++;
4579            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
# Line 4509  for (;;) Line 4583  for (;;)
4583          case OP_WORDCHAR:          case OP_WORDCHAR:
4584          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4585            {            {
4586              pcre_uint32 cc;
4587            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4588              {              {
4589              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4590              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4591              }              }
4592            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_word) == 0)            cc = RAWUCHAR(eptr);
4593              if (cc >= 128 || (md->ctypes[cc] & ctype_word) == 0)
4594              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4595            eptr++;            eptr++;
4596            /* 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 4545  for (;;) Line 4621  for (;;)
4621            if (md->partial != 0 &&            if (md->partial != 0 &&
4622                eptr + 1 >= md->end_subject &&                eptr + 1 >= md->end_subject &&
4623                NLBLOCK->nltype == NLTYPE_FIXED &&                NLBLOCK->nltype == NLTYPE_FIXED &&
4624                NLBLOCK->nllen == 2 &&                NLBLOCK->nllen == 2 &&
4625                *eptr == NLBLOCK->nl[0])                *eptr == NLBLOCK->nl[0])
4626              {              {
4627              md->hitend = TRUE;              md->hitend = TRUE;
4628              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4629              }              }
# Line 4585  for (;;) Line 4661  for (;;)
4661              {              {
4662              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4663    
4664              case 0x000d:              case CHAR_CR:
4665              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == CHAR_LF) eptr++;
4666              break;              break;
4667    
4668              case 0x000a:              case CHAR_LF:
4669              break;              break;
4670    
4671              case 0x000b:              case CHAR_VT:
4672              case 0x000c:              case CHAR_FF:
4673              case 0x0085:              case CHAR_NEL:
4674  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4675              case 0x2028:              case 0x2028:
4676              case 0x2029:              case 0x2029:
4677  #endif  #endif
# Line 4616  for (;;) Line 4692  for (;;)
4692            switch(*eptr++)            switch(*eptr++)
4693              {              {
4694              default: break;              default: break;
4695              case 0x09:      /* HT */              HSPACE_BYTE_CASES:
4696              case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4697              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 */  
4698  #endif  #endif
4699              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4700              }              }
# Line 4653  for (;;) Line 4712  for (;;)
4712            switch(*eptr++)            switch(*eptr++)
4713              {              {
4714              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4715              case 0x09:      /* HT */              HSPACE_BYTE_CASES:
4716              case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4717              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 */  
4718  #endif  #endif
4719              break;              break;
4720              }              }
# Line 4689  for (;;) Line 4731  for (;;)
4731              }              }
4732            switch(*eptr++)            switch(*eptr++)
4733              {              {
4734              default: break;              VSPACE_BYTE_CASES:
4735              case 0x0a:      /* LF */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4736              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 */  
4737  #endif  #endif
4738              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4739                default: break;
4740              }              }
4741            }            }
4742          break;          break;
# Line 4715  for (;;) Line 4752  for (;;)
4752            switch(*eptr++)            switch(*eptr++)
4753              {              {
4754              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4755              case 0x0a:      /* LF */              VSPACE_BYTE_CASES:
4756              case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4757              case 0x0c:      /* FF */              VSPACE_MULTIBYTE_CASES:
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
 #ifdef COMPILE_PCRE16  
             case 0x2028:    /* LINE SEPARATOR */  
             case 0x2029:    /* PARAGRAPH SEPARATOR */  
4758  #endif  #endif
4759              break;              break;
4760              }              }
# Line 4940  for (;;) Line 4972  for (;;)
4972              }              }
4973            /* Control never gets here */            /* Control never gets here */
4974    
4975              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
4976              which means that Perl space and POSIX space are now identical. PCRE
4977              was changed at release 8.34. */
4978    
4979            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
4980              case PT_PXSPACE:  /* POSIX space */
4981            for (fi = min;; fi++)            for (fi = min;; fi++)
4982              {              {
4983              RMATCH(eptr, ecode, offset_top, md, eptrb, RM60);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM61);
4984              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4985              if (fi >= max) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4986              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4952  for (;;) Line 4989  for (;;)
4989                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4990                }                }
4991              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4992              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              switch(c)
4993                   c == CHAR_FF || c == CHAR_CR)                {
4994                     == prop_fail_result)                HSPACE_CASES:
4995                RRETURN(MATCH_NOMATCH);                VSPACE_CASES:
4996                  if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4997                  break;
4998    
4999                  default:
5000                  if ((UCD_CATEGORY(c) == ucp_Z) == prop_fail_result)
5001                    RRETURN(MATCH_NOMATCH);
5002                  break;
5003                  }
5004              }              }
5005            /* Control never gets here */            /* Control never gets here */
5006    
5007            case PT_PXSPACE:  /* POSIX space */            case PT_WORD:
5008            for (fi = min;; fi++)            for (fi = min;; fi++)
5009              {              {
5010              RMATCH(eptr, ecode, offset_top, md, eptrb, RM61);              int category;
5011                RMATCH(eptr, ecode, offset_top, md, eptrb, RM62);
5012              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5013              if (fi >= max) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
5014              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4971  for (;;) Line 5017  for (;;)
5017                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5018                }                }
5019              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
5020              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              category = UCD_CATEGORY(c);
5021                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)              if ((category == ucp_L ||
5022                     category == ucp_N ||
5023                     c == CHAR_UNDERSCORE)
5024                     == prop_fail_result)                     == prop_fail_result)
5025                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5026              }              }
5027            /* Control never gets here */            /* Control never gets here */
5028    
5029            case PT_WORD:            case PT_CLIST:
5030            for (fi = min;; fi++)            for (fi = min;; fi++)
5031              {              {
5032              int category;              const pcre_uint32 *cp;
5033              RMATCH(eptr, ecode, offset_top, md, eptrb, RM62);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM67);
5034              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5035              if (fi >= max) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
5036              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4991  for (;;) Line 5039  for (;;)
5039                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5040                }                }
5041              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
5042              category = UCD_CATEGORY(c);              cp = PRIV(ucd_caseless_sets) + prop_value;
5043              if ((category == ucp_L ||              for (;;)
5044                   category == ucp_N ||                {
5045                   c == CHAR_UNDERSCORE)                if (c < *cp)
5046                     == prop_fail_result)                  { if (prop_fail_result) break; else { RRETURN(MATCH_NOMATCH); } }
5047                  if (c == *cp++)
5048                    { if (prop_fail_result) { RRETURN(MATCH_NOMATCH); } else break; }
5049                  }
5050                }
5051              /* Control never gets here */
5052    
5053              case PT_UCNC:
5054              for (fi = min;; fi++)
5055                {
5056                RMATCH(eptr, ecode, offset_top, md, eptrb, RM60);
5057                if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5058                if (fi >= max) RRETURN(MATCH_NOMATCH);
5059                if (eptr >= md->end_subject)
5060                  {
5061                  SCHECK_PARTIAL();
5062                  RRETURN(MATCH_NOMATCH);
5063                  }
5064                GETCHARINCTEST(c, eptr);
5065                if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
5066                     c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
5067                     c >= 0xe000) == prop_fail_result)
5068                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5069              }              }
5070            /* Control never gets here */            /* Control never gets here */
5071    
5072            /* This should never occur */            /* This should never occur */
   
5073            default:            default:
5074            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
5075            }            }
# Line 5022  for (;;) Line 5090  for (;;)
5090              SCHECK_PARTIAL();              SCHECK_PARTIAL();
5091              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
5092              }              }
5093            GETCHARINCTEST(c, eptr);            else
           if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
           while (eptr < md->end_subject)  
5094              {              {
5095              int len = 1;              int lgb, rgb;
5096              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
5097              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
5098              eptr += len;              while (eptr < md->end_subject)
5099                  {
5100                  int len = 1;
5101                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
5102                  rgb = UCD_GRAPHBREAK(c);
5103                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
5104                  lgb = rgb;
5105                  eptr += len;
5106                  }
5107              }              }
5108            CHECK_PARTIAL();            CHECK_PARTIAL();
5109            }            }
5110          }          }
5111        else        else
# Line 5059  for (;;) Line 5133  for (;;)
5133              if (md->partial != 0 &&    /* Take care with CRLF partial */              if (md->partial != 0 &&    /* Take care with CRLF partial */
5134                  eptr >= md->end_subject &&                  eptr >= md->end_subject &&
5135                  NLBLOCK->nltype == NLTYPE_FIXED &&                  NLBLOCK->nltype == NLTYPE_FIXED &&
5136                  NLBLOCK->nllen == 2 &&                  NLBLOCK->nllen == 2 &&
5137                  c == NLBLOCK->nl[0])                  c == NLBLOCK->nl[0])
5138                {                {
5139                md->hitend = TRUE;                md->hitend = TRUE;
5140                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5141                }                }
5142              break;              break;
5143    
5144              case OP_ALLANY:              case OP_ALLANY:
5145              case OP_ANYBYTE:              case OP_ANYBYTE:
5146              break;              break;
# Line 5075  for (;;) Line 5149  for (;;)
5149              switch(c)              switch(c)
5150                {                {
5151                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5152                case 0x000d:                case CHAR_CR:
5153                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && RAWUCHAR(eptr) == CHAR_LF) eptr++;
5154                break;                break;
5155                case 0x000a:  
5156                  case CHAR_LF:
5157                break;                break;
5158    
5159                case 0x000b:                case CHAR_VT:
5160                case 0x000c:                case CHAR_FF:
5161                case 0x0085:                case CHAR_NEL:
5162    #ifndef EBCDIC
5163                case 0x2028:                case 0x2028:
5164                case 0x2029:                case 0x2029:
5165    #endif  /* Not EBCDIC */
5166                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
5167                break;                break;
5168                }                }
# Line 5094  for (;;) Line 5171  for (;;)
5171              case OP_NOT_HSPACE:              case OP_NOT_HSPACE:
5172              switch(c)              switch(c)
5173                {                {
5174                  HSPACE_CASES: RRETURN(MATCH_NOMATCH);
5175                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);  
5176                }                }
5177              break;              break;
5178    
5179              case OP_HSPACE:              case OP_HSPACE:
5180              switch(c)              switch(c)
5181                {                {
5182                  HSPACE_CASES: break;
5183                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;  
5184                }                }
5185              break;              break;
5186    
5187              case OP_NOT_VSPACE:              case OP_NOT_VSPACE:
5188              switch(c)              switch(c)
5189                {                {
5190                  VSPACE_CASES: RRETURN(MATCH_NOMATCH);
5191                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);  
5192                }                }
5193              break;              break;
5194    
5195              case OP_VSPACE:              case OP_VSPACE:
5196              switch(c)              switch(c)
5197                {                {
5198                  VSPACE_CASES: break;
5199                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;  
5200                }                }
5201              break;              break;
5202    
# Line 5233  for (;;) Line 5258  for (;;)
5258              if (md->partial != 0 &&    /* Take care with CRLF partial */              if (md->partial != 0 &&    /* Take care with CRLF partial */
5259                  eptr >= md->end_subject &&                  eptr >= md->end_subject &&
5260                  NLBLOCK->nltype == NLTYPE_FIXED &&                  NLBLOCK->nltype == NLTYPE_FIXED &&
5261                  NLBLOCK->nllen == 2 &&                  NLBLOCK->nllen == 2 &&
5262                  c == NLBLOCK->nl[0])                  c == NLBLOCK->nl[0])
5263                {                {
5264                md->hitend = TRUE;                md->hitend = TRUE;
5265                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5266                }                }
5267              break;              break;
5268    
5269              case OP_ALLANY:              case OP_ALLANY:
5270              case OP_ANYBYTE:              case OP_ANYBYTE:
5271              break;              break;
# Line 5249  for (;;) Line 5274  for (;;)
5274              switch(c)              switch(c)
5275                {                {
5276                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5277                case 0x000d:                case CHAR_CR:
5278                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == CHAR_LF) eptr++;
5279                break;                break;
5280    
5281                case 0x000a:                case CHAR_LF:
5282                break;                break;
5283    
5284                case 0x000b:                case CHAR_VT:
5285                case 0x000c:                case CHAR_FF:
5286                case 0x0085:                case CHAR_NEL:
5287  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5288                case 0x2028:                case 0x2028:
5289                case 0x2029:                case 0x2029:
5290  #endif  #endif
# Line 5272  for (;;) Line 5297  for (;;)
5297              switch(c)              switch(c)
5298                {                {
5299                default: break;                default: break;
5300                case 0x09:      /* HT */                HSPACE_BYTE_CASES:
5301                case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5302                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 */  
5303  #endif  #endif
5304                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5305                }                }
# Line 5301  for (;;) Line 5309  for (;;)
5309              switch(c)              switch(c)
5310                {                {
5311                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5312                case 0x09:      /* HT */                HSPACE_BYTE_CASES:
5313                case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5314                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 */  
5315  #endif  #endif
5316                break;                break;
5317                }                }
# Line 5330  for (;;) Line 5321  for (;;)
5321              switch(c)              switch(c)
5322                {                {
5323                default: break;                default: break;
5324                case 0x0a:      /* LF */                VSPACE_BYTE_CASES:
5325                case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5326                case 0x0c:      /* FF */                VSPACE_MULTIBYTE_CASES:
               case 0x0d:      /* CR */  
               case 0x85:      /* NEL */  
 #ifdef COMPILE_PCRE16  
               case 0x2028:    /* LINE SEPARATOR */  
               case 0x2029:    /* PARAGRAPH SEPARATOR */  
5327  #endif  #endif
5328                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5329                }                }
# Line 5347  for (;;) Line 5333  for (;;)
5333              switch(c)              switch(c)
5334                {                {
5335                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5336                case 0x0a:      /* LF */                VSPACE_BYTE_CASES:
5337                case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5338                case 0x0c:      /* FF */                VSPACE_MULTIBYTE_CASES:
               case 0x0d:      /* CR */  
               case 0x85:      /* NEL */  
 #ifdef COMPILE_PCRE16  
               case 0x2028:    /* LINE SEPARATOR */  
               case 0x2029:    /* PARAGRAPH SEPARATOR */  
5339  #endif  #endif
5340                break;                break;
5341                }                }
# Line 5503  for (;;) Line 5484  for (;;)
5484              }              }
5485            break;            break;
5486    
5487              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
5488              which means that Perl space and POSIX space are now identical. PCRE
5489              was changed at release 8.34. */
5490    
5491            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
5492              case PT_PXSPACE:  /* POSIX space */
5493            for (i = min; i < max; i++)            for (i = min; i < max; i++)
5494              {              {
5495              int len = 1;              int len = 1;
# Line 5513  for (;;) Line 5499  for (;;)
5499                break;                break;
5500                }                }
5501              GETCHARLENTEST(c, eptr, len);              GETCHARLENTEST(c, eptr, len);
5502              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              switch(c)
5503                   c == CHAR_FF || c == CHAR_CR)                {
5504                   == prop_fail_result)                HSPACE_CASES:
5505                  VSPACE_CASES:
5506                  if (prop_fail_result) goto ENDLOOP99;  /* Break the loop */
5507                  break;
5508    
5509                  default:
5510                  if ((UCD_CATEGORY(c) == ucp_Z) == prop_fail_result)
5511                    goto ENDLOOP99;   /* Break the loop */
5512                break;                break;
5513                  }
5514              eptr+= len;              eptr+= len;
5515              }              }
5516              ENDLOOP99:
5517            break;            break;
5518    
5519            case PT_PXSPACE:  /* POSIX space */            case PT_WORD:
5520            for (i = min; i < max; i++)            for (i = min; i < max; i++)
5521              {              {
5522                int category;
5523              int len = 1;              int len = 1;
5524              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
5525                {                {
# Line 5531  for (;;) Line 5527  for (;;)
5527                break;                break;
5528                }                }
5529              GETCHARLENTEST(c, eptr, len);              GETCHARLENTEST(c, eptr, len);
5530              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              category = UCD_CATEGORY(c);
5531                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)              if ((category == ucp_L || category == ucp_N ||
5532                   == prop_fail_result)                   c == CHAR_UNDERSCORE) == prop_fail_result)
5533                break;                break;
5534              eptr+= len;              eptr+= len;
5535              }              }
5536            break;            break;
5537    
5538            case PT_WORD:            case PT_CLIST:
5539            for (i = min; i < max; i++)            for (i = min; i < max; i++)
5540              {              {
5541              int category;              const pcre_uint32 *cp;
5542              int len = 1;              int len = 1;
5543              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
5544                {                {
# Line 5550  for (;;) Line 5546  for (;;)
5546                break;                break;
5547                }                }
5548              GETCHARLENTEST(c, eptr, len);              GETCHARLENTEST(c, eptr, len);
5549              category = UCD_CATEGORY(c);              cp = PRIV(ucd_caseless_sets) + prop_value;
5550              if ((category == ucp_L || category == ucp_N ||              for (;;)
5551                   c == CHAR_UNDERSCORE) == prop_fail_result)                {
5552                  if (c < *cp)
5553                    { if (prop_fail_result) break; else goto GOT_MAX; }
5554                  if (c == *cp++)
5555                    { if (prop_fail_result) goto GOT_MAX; else break; }
5556                  }
5557                eptr += len;
5558                }
5559              GOT_MAX:
5560              break;
5561    
5562              case PT_UCNC:
5563              for (i = min; i < max; i++)
5564                {
5565                int len = 1;
5566                if (eptr >= md->end_subject)
5567                  {
5568                  SCHECK_PARTIAL();
5569                break;                break;
5570              eptr+= len;                }
5571                GETCHARLENTEST(c, eptr, len);
5572                if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
5573                     c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
5574                     c >= 0xe000) == prop_fail_result)
5575                  break;
5576                eptr += len;
5577              }              }
5578            break;            break;
5579    
# Line 5564  for (;;) Line 5583  for (;;)
5583    
5584          /* eptr is now past the end of the maximum run */          /* eptr is now past the end of the maximum run */
5585    
5586          if (possessive) continue;          if (possessive) continue;    /* No backtracking */
5587          for(;;)          for(;;)
5588            {            {
5589              if (eptr == pp) goto TAIL_RECURSE;
5590            RMATCH(eptr, ecode, offset_top, md, eptrb, RM44);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM44);
5591            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5592            if (eptr-- == pp) break;        /* Stop if tried at original pos */            eptr--;
5593            if (utf) BACKCHAR(eptr);            if (utf) BACKCHAR(eptr);
5594            }            }
5595          }          }
5596    
5597        /* Match extended Unicode sequences. We will get here only if the        /* Match extended Unicode grapheme clusters. We will get here only if the
5598        support is in the binary; otherwise a compile-time error occurs. */        support is in the binary; otherwise a compile-time error occurs. */
5599    
5600        else if (ctype == OP_EXTUNI)        else if (ctype == OP_EXTUNI)
5601          {          {
5602          for (i = min; i < max; i++)          for (i = min; i < max; i++)
5603            {            {
           int len = 1;  
5604            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
5605              {              {
5606              SCHECK_PARTIAL();              SCHECK_PARTIAL();
5607              break;              break;
5608              }              }
5609            if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }            else
           if (UCD_CATEGORY(c) == ucp_M) break;  
           eptr += len;  
           while (eptr < md->end_subject)  
5610              {              {
5611              len = 1;              int lgb, rgb;
5612              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
5613              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
5614              eptr += len;              while (eptr < md->end_subject)
5615                  {
5616                  int len = 1;
5617                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
5618                  rgb = UCD_GRAPHBREAK(c);
5619                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
5620                  lgb = rgb;
5621                  eptr += len;
5622                  }
5623              }              }
5624            CHECK_PARTIAL();            CHECK_PARTIAL();
5625            }            }
5626    
5627          /* eptr is now past the end of the maximum run */          /* eptr is now past the end of the maximum run */
5628    
5629          if (possessive) continue;          if (possessive) continue;    /* No backtracking */
5630    
5631          for(;;)          for(;;)
5632            {            {
5633              int lgb, rgb;
5634              PCRE_PUCHAR fptr;
5635    
5636              if (eptr == pp) goto TAIL_RECURSE;   /* At start of char run */
5637            RMATCH(eptr, ecode, offset_top, md, eptrb, RM45);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM45);
5638            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5639            if (eptr-- == pp) break;        /* Stop if tried at original pos */  
5640            for (;;)                        /* Move back over one extended */            /* Backtracking over an extended grapheme cluster involves inspecting
5641              the previous two characters (if present) to see if a break is
5642              permitted between them. */
5643    
5644              eptr--;
5645              if (!utf) c = *eptr; else
5646              {              {
5647              if (!utf) c = *eptr; else              BACKCHAR(eptr);
5648                {              GETCHAR(c, eptr);
5649                BACKCHAR(eptr);              }
5650                GETCHAR(c, eptr);            rgb = UCD_GRAPHBREAK(c);
5651                }  
5652              if (UCD_CATEGORY(c) != ucp_M) break;            for (;;)
5653              eptr--;              {
5654                if (eptr == pp) goto TAIL_RECURSE;   /* At start of char run */
5655                fptr = eptr - 1;
5656                if (!utf) c = *fptr; else
5657                  {
5658                  BACKCHAR(fptr);
5659                  GETCHAR(c, fptr);
5660                  }
5661                lgb = UCD_GRAPHBREAK(c);
5662                if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
5663                eptr = fptr;
5664                rgb = lgb;
5665              }              }
5666            }            }
5667          }          }
# Line 5644  for (;;) Line 5688  for (;;)
5688                if (md->partial != 0 &&    /* Take care with CRLF partial */                if (md->partial != 0 &&    /* Take care with CRLF partial */
5689                    eptr + 1 >= md->end_subject &&                    eptr + 1 >= md->end_subject &&
5690                    NLBLOCK->nltype == NLTYPE_FIXED &&                    NLBLOCK->nltype == NLTYPE_FIXED &&
5691                    NLBLOCK->nllen == 2 &&                    NLBLOCK->nllen == 2 &&
5692                    *eptr == NLBLOCK->nl[0])                    RAWUCHAR(eptr) == NLBLOCK->nl[0])
5693                  {                  {
5694                  md->hitend = TRUE;                  md->hitend = TRUE;
5695                  if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);                  if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5696                  }                  }
# Line 5670  for (;;) Line 5714  for (;;)
5714                if (md->partial != 0 &&    /* Take care with CRLF partial */                if (md->partial != 0 &&    /* Take care with CRLF partial */
5715                    eptr + 1 >= md->end_subject &&                    eptr + 1 >= md->end_subject &&
5716                    NLBLOCK->nltype == NLTYPE_FIXED &&                    NLBLOCK->nltype == NLTYPE_FIXED &&
5717                    NLBLOCK->nllen == 2 &&                    NLBLOCK->nllen == 2 &&
5718                    *eptr == NLBLOCK->nl[0])                    RAWUCHAR(eptr) == NLBLOCK->nl[0])
5719                  {                  {
5720                  md->hitend = TRUE;                  md->hitend = TRUE;
5721                  if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);                  if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5722                  }                  }
# Line 5725  for (;;) Line 5769  for (;;)
5769                break;                break;
5770                }                }
5771              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
5772              if (c == 0x000d)              if (c == CHAR_CR)
5773                {                {
5774                if (++eptr >= md->end_subject) break;                if (++eptr >= md->end_subject) break;
5775                if (*eptr == 0x000a) eptr++;                if (RAWUCHAR(eptr) == CHAR_LF) eptr++;
5776                }                }
5777              else              else
5778                {                {
5779                if (c != 0x000a &&                if (c != CHAR_LF &&
5780                    (md->bsr_anycrlf ||                    (md->bsr_anycrlf ||
5781                     (c != 0x000b && c != 0x000c &&                     (c != CHAR_VT && c != CHAR_FF && c != CHAR_NEL
5782                      c != 0x0085 && c != 0x2028 && c != 0x2029)))  #ifndef EBCDIC
5783                        && c != 0x2028 && c != 0x2029
5784    #endif  /* Not EBCDIC */
5785                        )))
5786                  break;                  break;
5787                eptr += len;                eptr += len;
5788                }                }
# Line 5756  for (;;) Line 5803  for (;;)
5803              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
5804              switch(c)              switch(c)
5805                {                {
5806                  HSPACE_CASES: gotspace = TRUE; break;
5807                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;  
5808                }                }
5809              if (gotspace == (ctype == OP_NOT_HSPACE)) break;              if (gotspace == (ctype == OP_NOT_HSPACE)) break;
5810              eptr += len;              eptr += len;
# Line 5798  for (;;) Line 5825  for (;;)
5825              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
5826              switch(c)              switch(c)
5827                {                {
5828                  VSPACE_CASES: gotspace = TRUE; break;
5829                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;  
5830                }                }
5831              if (gotspace == (ctype == OP_NOT_VSPACE)) break;              if (gotspace == (ctype == OP_NOT_VSPACE)) break;
5832              eptr += len;              eptr += len;
# Line 5908  for (;;) Line 5927  for (;;)
5927            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
5928            }            }
5929    
5930          /* 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;  
5931          for(;;)          for(;;)
5932            {            {
5933              if (eptr == pp) goto TAIL_RECURSE;
5934            RMATCH(eptr, ecode, offset_top, md, eptrb, RM46);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM46);
5935            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5936            if (eptr-- == pp) break;        /* Stop if tried at original pos */            eptr--;
5937            BACKCHAR(eptr);            BACKCHAR(eptr);
5938            if (ctype == OP_ANYNL && eptr > pp  && *eptr == '\n' &&            if (ctype == OP_ANYNL && eptr > pp  && RAWUCHAR(eptr) == CHAR_NL &&
5939                eptr[-1] == '\r') eptr--;                RAWUCHAR(eptr - 1) == CHAR_CR) eptr--;
5940            }            }
5941          }          }
5942        else        else
# Line 5943  for (;;) Line 5957  for (;;)
5957              if (md->partial != 0 &&    /* Take care with CRLF partial */              if (md->partial != 0 &&    /* Take care with CRLF partial */
5958                  eptr + 1 >= md->end_subject &&                  eptr + 1 >= md->end_subject &&
5959                  NLBLOCK->nltype == NLTYPE_FIXED &&                  NLBLOCK->nltype == NLTYPE_FIXED &&
5960                  NLBLOCK->nllen == 2 &&                  NLBLOCK->nllen == 2 &&
5961                  *eptr == NLBLOCK->nl[0])                  *eptr == NLBLOCK->nl[0])
5962                {                {
5963                md->hitend = TRUE;                md->hitend = TRUE;
5964                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5965                }                }
# Line 5973  for (;;) Line 5987  for (;;)
5987                break;                break;
5988                }                }
5989              c = *eptr;              c = *eptr;
5990              if (c == 0x000d)              if (c == CHAR_CR)
5991                {                {
5992                if (++eptr >= md->end_subject) break;                if (++eptr >= md->end_subject) break;
5993                if (*eptr == 0x000a) eptr++;                if (*eptr == CHAR_LF) eptr++;
5994                }                }
5995              else              else
5996                {                {
5997                if (c != 0x000a && (md->bsr_anycrlf ||                if (c != CHAR_LF && (md->bsr_anycrlf ||
5998                  (c != 0x000b && c != 0x000c && c != 0x0085                   (c != CHAR_VT && c != CHAR_FF && c != CHAR_NEL
5999  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6000                  && c != 0x2028 && c != 0x2029                   && c != 0x2028 && c != 0x2029
6001  #endif  #endif
6002                  ))) break;                   ))) break;
6003                eptr++;                eptr++;
6004                }                }
6005              }              }
# Line 5999  for (;;) Line 6013  for (;;)
6013                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6014                break;                break;
6015                }                }
6016              c = *eptr;              switch(*eptr)
6017              if (c == 0x09 || c == 0x20 || c == 0xa0                {
6018  #ifdef COMPILE_PCRE16                default: eptr++; break;
6019                || c == 0x1680 || c == 0x180e || (c >= 0x2000 && c <= 0x200A)                HSPACE_BYTE_CASES:
6020                || c == 0x202f || c == 0x205f || c == 0x3000  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6021                  HSPACE_MULTIBYTE_CASES:
6022  #endif  #endif
6023                ) break;                goto ENDLOOP00;
6024              eptr++;                }
6025              }              }
6026              ENDLOOP00:
6027            break;            break;
6028    
6029            case OP_HSPACE:            case OP_HSPACE:
# Line 6018  for (;;) Line 6034  for (;;)
6034                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6035                break;                break;
6036                }                }
6037              c = *eptr;              switch(*eptr)
6038              if (c != 0x09 && c != 0x20 && c != 0xa0                {
6039  #ifdef COMPILE_PCRE16                default: goto ENDLOOP01;
6040                && c != 0x1680 && c != 0x180e && (c < 0x2000 || c > 0x200A)                HSPACE_BYTE_CASES:
6041                && c != 0x202f && c != 0x205f && c != 0x3000  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6042                  HSPACE_MULTIBYTE_CASES:
6043  #endif  #endif
6044                ) break;                eptr++; break;
6045              eptr++;                }
6046              }              }
6047              ENDLOOP01:
6048            break;            break;
6049    
6050            case OP_NOT_VSPACE:            case OP_NOT_VSPACE:
# Line 6037  for (;;) Line 6055  for (;;)
6055                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6056                break;                break;
6057                }                }
6058              c = *eptr;              switch(*eptr)
6059              if (c == 0x0a || c == 0x0b || c == 0x0c || c == 0x0d || c == 0x85                {
6060  #ifdef COMPILE_PCRE16                default: eptr++; break;
6061                || c == 0x2028 || c == 0x2029                VSPACE_BYTE_CASES:
6062    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6063                  VSPACE_MULTIBYTE_CASES:
6064  #endif  #endif
6065                ) break;                goto ENDLOOP02;
6066              eptr++;                }
6067              }              }
6068              ENDLOOP02:
6069            break;            break;
6070    
6071            case OP_VSPACE:            case OP_VSPACE:
# Line 6055  for (;;) Line 6076  for (;;)
6076                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6077                break;                break;
6078                }                }
6079              c = *eptr;              switch(*eptr)
6080              if (c != 0x0a && c != 0x0b && c != 0x0c && c != 0x0d && c != 0x85                {
6081  #ifdef COMPILE_PCRE16                default: goto ENDLOOP03;
6082                && c != 0x2028 && c != 0x2029                VSPACE_BYTE_CASES:
6083    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6084                  VSPACE_MULTIBYTE_CASES:
6085  #endif  #endif
6086                ) break;                eptr++; break;
6087              eptr++;                }
6088              }              }
6089              ENDLOOP03:
6090            break;            break;
6091    
6092            case OP_NOT_DIGIT:            case OP_NOT_DIGIT:
# Line 6147  for (;;) Line 6171  for (;;)
6171            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
6172            }            }
6173    
6174          /* eptr is now past the end of the maximum run. If possessive, we are          if (possessive) continue;    /* No backtracking */
6175          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)  
6176            {            {
6177              if (eptr == pp) goto TAIL_RECURSE;
6178            RMATCH(eptr, ecode, offset_top, md, eptrb, RM47);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM47);
6179            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
6180            eptr--;            eptr--;
6181            if (ctype == OP_ANYNL && eptr > pp  && *eptr == '\n' &&            if (ctype == OP_ANYNL && eptr > pp  && *eptr == CHAR_LF &&
6182                eptr[-1] == '\r') eptr--;                eptr[-1] == CHAR_CR) eptr--;
6183            }            }
6184          }          }
6185    
6186        /* Get here if we can't make it match with any permitted repetitions */        /* Control never gets here */
   
       RRETURN(MATCH_NOMATCH);  
6187        }        }
     /* Control never gets here */  
6188    
6189      /* 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
6190      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 6202  switch (frame->Xwhere) Line 6218  switch (frame->Xwhere)
6218    LBL(53) LBL(54) LBL(55) LBL(56) LBL(57) LBL(58) LBL(63) LBL(64)    LBL(53) LBL(54) LBL(55) LBL(56) LBL(57) LBL(58) LBL(63) LBL(64)
6219    LBL(65) LBL(66)    LBL(65) LBL(66)
6220  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
6221    LBL(21)    LBL(20) LBL(21)
6222  #endif  #endif
6223  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
6224    LBL(16) LBL(18) LBL(20)    LBL(16) LBL(18)
6225    LBL(22) LBL(23) LBL(28) LBL(30)    LBL(22) LBL(23) LBL(28) LBL(30)
6226    LBL(32) LBL(34) LBL(42) LBL(46)    LBL(32) LBL(34) LBL(42) LBL(46)
6227  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
6228    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)
6229    LBL(59) LBL(60) LBL(61) LBL(62)    LBL(59) LBL(60) LBL(61) LBL(62) LBL(67)
6230  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
6231  #endif  /* SUPPORT_UTF */  #endif  /* SUPPORT_UTF */
6232    default:    default:
6233    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);  
   
6234    return PCRE_ERROR_INTERNAL;    return PCRE_ERROR_INTERNAL;
6235    }    }
6236  #undef LBL  #undef LBL
# Line 6279  Undefine all the macros that were define Line 6292  Undefine all the macros that were define
6292  ***************************************************************************/  ***************************************************************************/
6293    
6294    
6295    #ifdef NO_RECURSE
6296    /*************************************************
6297    *          Release allocated heap frames         *
6298    *************************************************/
6299    
6300    /* This function releases all the allocated frames. The base frame is on the
6301    machine stack, and so must not be freed.
6302    
6303    Argument: the address of the base frame
6304    Returns:  nothing
6305    */
6306    
6307    static void
6308    release_match_heapframes (heapframe *frame_base)
6309    {
6310    heapframe *nextframe = frame_base->Xnextframe;
6311    while (nextframe != NULL)
6312      {
6313      heapframe *oldframe = nextframe;
6314      nextframe = nextframe->Xnextframe;
6315      (PUBL(stack_free))(oldframe);
6316      }
6317    }
6318    #endif
6319    
6320    
6321  /*************************************************  /*************************************************
6322  *         Execute a Regular Expression           *  *         Execute a Regular Expression           *
# Line 6304  Returns:          > 0 => success; value Line 6342  Returns:          > 0 => success; value
6342                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
6343  */  */
6344    
6345  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
6346  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
6347  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,
6348    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,
6349    int offsetcount)    int offsetcount)
6350  #else  #elif defined COMPILE_PCRE16
6351  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
6352  pcre16_exec(const pcre16 *argument_re, const pcre16_extra *extra_data,  pcre16_exec(const pcre16 *argument_re, const pcre16_extra *extra_data,
6353    PCRE_SPTR16 subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR16 subject, int length, int start_offset, int options, int *offsets,
6354    int offsetcount)    int offsetcount)
6355    #elif defined COMPILE_PCRE32
6356    PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
6357    pcre32_exec(const pcre32 *argument_re, const pcre32_extra *extra_data,
6358      PCRE_SPTR32 subject, int length, int start_offset, int options, int *offsets,
6359      int offsetcount)
6360  #endif  #endif
6361  {  {
6362  int rc, ocount, arg_offset_max;  int rc, ocount, arg_offset_max;
# Line 6336  const pcre_uint8 *start_bits = NULL; Line 6379  const pcre_uint8 *start_bits = NULL;
6379  PCRE_PUCHAR start_match = (PCRE_PUCHAR)subject + start_offset;  PCRE_PUCHAR start_match = (PCRE_PUCHAR)subject + start_offset;
6380  PCRE_PUCHAR end_subject;  PCRE_PUCHAR end_subject;
6381  PCRE_PUCHAR start_partial = NULL;  PCRE_PUCHAR start_partial = NULL;
6382    PCRE_PUCHAR match_partial = NULL;
6383  PCRE_PUCHAR req_char_ptr = start_match - 1;  PCRE_PUCHAR req_char_ptr = start_match - 1;
6384    
6385  const pcre_study_data *study;  const pcre_study_data *study;
6386  const REAL_PCRE *re = (const REAL_PCRE *)argument_re;  const REAL_PCRE *re = (const REAL_PCRE *)argument_re;
6387    
6388    #ifdef NO_RECURSE
6389    heapframe frame_zero;
6390    frame_zero.Xprevframe = NULL;            /* Marks the top level */
6391    frame_zero.Xnextframe = NULL;            /* None are allocated yet */
6392    md->match_frames_base = &frame_zero;
6393    #endif
6394    
6395  /* Check for the special magic call that measures the size of the stack used  /* Check for the special magic call that measures the size of the stack used
6396  per recursive call of match(). Without the funny casting for sizeof, a Windows  per recursive call of match(). Without the funny casting for sizeof, a Windows
6397  compiler gave this error: "unary minus operator applied to unsigned type,  compiler gave this error: "unary minus operator applied to unsigned type,
6398  result still unsigned". Hopefully the cast fixes that. */  result still unsigned". Hopefully the cast fixes that. */
6399    
# Line 6360  if ((options & ~PUBLIC_EXEC_OPTIONS) != Line 6411  if ((options & ~PUBLIC_EXEC_OPTIONS) !=
6411  if (re == NULL || subject == NULL || (offsets == NULL && offsetcount > 0))  if (re == NULL || subject == NULL || (offsets == NULL && offsetcount > 0))
6412    return PCRE_ERROR_NULL;    return PCRE_ERROR_NULL;
6413  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;
6414    if (length < 0) return PCRE_ERROR_BADLENGTH;
6415  if (start_offset < 0 || start_offset > length) return PCRE_ERROR_BADOFFSET;  if (start_offset < 0 || start_offset > length) return PCRE_ERROR_BADOFFSET;
6416    
6417  /* 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 6397  if (utf && (options & PCRE_NO_UTF8_CHECK Line 6449  if (utf && (options & PCRE_NO_UTF8_CHECK
6449        offsets[0] = erroroffset;        offsets[0] = erroroffset;
6450        offsets[1] = errorcode;        offsets[1] = errorcode;
6451        }        }
6452  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE8
     return (errorcode <= PCRE_UTF16_ERR1 && md->partial > 1)?  
       PCRE_ERROR_SHORTUTF16 : PCRE_ERROR_BADUTF16;  
 #else  
6453      return (errorcode <= PCRE_UTF8_ERR5 && md->partial > 1)?      return (errorcode <= PCRE_UTF8_ERR5 && md->partial > 1)?
6454        PCRE_ERROR_SHORTUTF8 : PCRE_ERROR_BADUTF8;        PCRE_ERROR_SHORTUTF8 : PCRE_ERROR_BADUTF8;
6455    #elif defined COMPILE_PCRE16
6456        return (errorcode <= PCRE_UTF16_ERR1 && md->partial > 1)?
6457          PCRE_ERROR_SHORTUTF16 : PCRE_ERROR_BADUTF16;
6458    #elif defined COMPILE_PCRE32
6459        return PCRE_ERROR_BADUTF32;
6460  #endif  #endif
6461      }      }
6462    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
6463    /* 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. */
6464    if (start_offset > 0 && start_offset < length &&    if (start_offset > 0 && start_offset < length &&
6465        NOT_FIRSTCHAR(((PCRE_PUCHAR)subject)[start_offset]))        NOT_FIRSTCHAR(((PCRE_PUCHAR)subject)[start_offset]))
6466      return PCRE_ERROR_BADUTF8_OFFSET;      return PCRE_ERROR_BADUTF8_OFFSET;
6467    #endif
6468    }    }
6469  #endif  #endif
6470    
# Line 6423  if (extra_data != NULL Line 6478  if (extra_data != NULL
6478      && (extra_data->flags & (PCRE_EXTRA_EXECUTABLE_JIT |      && (extra_data->flags & (PCRE_EXTRA_EXECUTABLE_JIT |
6479                               PCRE_EXTRA_TABLES)) == PCRE_EXTRA_EXECUTABLE_JIT                               PCRE_EXTRA_TABLES)) == PCRE_EXTRA_EXECUTABLE_JIT
6480      && extra_data->executable_jit != NULL      && extra_data->executable_jit != NULL
6481      && (options & ~(PCRE_NO_UTF8_CHECK | PCRE_NOTBOL | PCRE_NOTEOL |      && (options & ~PUBLIC_JIT_EXEC_OPTIONS) == 0)
                     PCRE_NOTEMPTY | PCRE_NOTEMPTY_ATSTART |  
                     PCRE_PARTIAL_SOFT | PCRE_PARTIAL_HARD)) == 0)  
6482    {    {
6483    rc = PRIV(jit_exec)(re, extra_data->executable_jit,    rc = PRIV(jit_exec)(extra_data, (const pcre_uchar *)subject, length,
6484      (const pcre_uchar *)subject, length, start_offset, options,         start_offset, options, offsets, offsetcount);
6485      ((extra_data->flags & PCRE_EXTRA_MATCH_LIMIT) == 0)  
     ? MATCH_LIMIT : extra_data->match_limit, offsets, offsetcount);  
   
6486    /* PCRE_ERROR_NULL means that the selected normal or partial matching    /* PCRE_ERROR_NULL means that the selected normal or partial matching
6487    mode is not compiled. In this case we simply fallback to interpreter. */    mode is not compiled. In this case we simply fallback to interpreter. */
6488    
6489    if (rc != PCRE_ERROR_NULL) return rc;    if (rc != PCRE_ERROR_JIT_BADOPTION) return rc;
6490    }    }
6491  #endif  #endif
6492    
# Line 6458  md->callout_data = NULL; Line 6509  md->callout_data = NULL;
6509    
6510  tables = re->tables;  tables = re->tables;
6511    
6512    /* The two limit values override the defaults, whatever their value. */
6513    
6514  if (extra_data != NULL)  if (extra_data != NULL)
6515    {    {
6516    register unsigned int flags = extra_data->flags;    register unsigned int flags = extra_data->flags;
# Line 6472  if (extra_data != NULL) Line 6525  if (extra_data != NULL)
6525    if ((flags & PCRE_EXTRA_TABLES) != 0) tables = extra_data->tables;    if ((flags & PCRE_EXTRA_TABLES) != 0) tables = extra_data->tables;
6526    }    }
6527    
6528    /* Limits in the regex override only if they are smaller. */
6529    
6530    if ((re->flags & PCRE_MLSET) != 0 && re->limit_match < md->match_limit)
6531      md->match_limit = re->limit_match;
6532    
6533    if ((re->flags & PCRE_RLSET) != 0 &&
6534        re->limit_recursion < md->match_limit_recursion)
6535      md->match_limit_recursion = re->limit_recursion;
6536    
6537  /* If the exec call supplied NULL for tables, use the inbuilt ones. This  /* If the exec call supplied NULL for tables, use the inbuilt ones. This
6538  is a feature that makes it possible to save compiled regex and re-use them  is a feature that makes it possible to save compiled regex and re-use them
6539  in other programs later. */  in other programs later. */
# Line 6497  end_subject = md->end_subject; Line 6559  end_subject = md->end_subject;
6559  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;
6560  md->use_ucp = (re->options & PCRE_UCP) != 0;  md->use_ucp = (re->options & PCRE_UCP) != 0;
6561  md->jscript_compat = (re->options & PCRE_JAVASCRIPT_COMPAT) != 0;  md->jscript_compat = (re->options & PCRE_JAVASCRIPT_COMPAT) != 0;
6562  md->ignore_skip_arg = FALSE;  md->ignore_skip_arg = 0;
6563    
6564  /* Some options are unpacked into BOOL variables in the hope that testing  /* Some options are unpacked into BOOL variables in the hope that testing
6565  them will be faster than individual option bits. */  them will be faster than individual option bits. */
# Line 6607  if (re->top_backref > 0 && re->top_backr Line 6669  if (re->top_backref > 0 && re->top_backr
6669    DPRINTF(("Got memory to hold back references\n"));    DPRINTF(("Got memory to hold back references\n"));
6670    }    }
6671  else md->offset_vector = offsets;  else md->offset_vector = offsets;
   
6672  md->offset_end = ocount;  md->offset_end = ocount;
6673  md->offset_max = (2*ocount)/3;  md->offset_max = (2*ocount)/3;
6674  md->offset_overflow = FALSE;  md->capture_last = 0;
 md->capture_last = -1;  
6675    
6676  /* Reset the working variable associated with each extraction. These should  /* Reset the working variable associated with each extraction. These should
6677  never be used unless previously set, but they get saved and restored, and so we  never be used unless previously set, but they get saved and restored, and so we
# Line 6719  for(;;) Line 6779  for(;;)
6779    
6780      if (has_first_char)      if (has_first_char)
6781        {        {
6782          pcre_uchar smc;
6783    
6784        if (first_char != first_char2)        if (first_char != first_char2)
6785          while (start_match < end_subject &&          while (start_match < end_subject &&
6786              *start_match != first_char && *start_match != first_char2)            (smc = RAWUCHARTEST(start_match)) != first_char && smc != first_char2)
6787            start_match++;            start_match++;
6788        else        else
6789          while (start_match < end_subject && *start_match != first_char)          while (start_match < end_subject && RAWUCHARTEST(start_match) != first_char)
6790            start_match++;            start_match++;
6791        }        }
6792    
# Line 6756  for(;;) Line 6818  for(;;)
6818          if (start_match[-1] == CHAR_CR &&          if (start_match[-1] == CHAR_CR &&
6819               (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&               (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&
6820               start_match < end_subject &&               start_match < end_subject &&
6821               *start_match == CHAR_NL)               RAWUCHARTEST(start_match) == CHAR_NL)
6822            start_match++;            start_match++;
6823          }          }
6824        }        }
# Line 6767  for(;;) Line 6829  for(;;)
6829        {        {
6830        while (start_match < end_subject)        while (start_match < end_subject)
6831          {          {
6832          register unsigned int c = *start_match;          register pcre_uint32 c = RAWUCHARTEST(start_match);
6833  #ifndef COMPILE_PCRE8  #ifndef COMPILE_PCRE8
6834          if (c > 255) c = 255;          if (c > 255) c = 255;
6835  #endif  #endif
# Line 6835  for(;;) Line 6897  for(;;)
6897            {            {
6898            while (p < end_subject)            while (p < end_subject)
6899              {              {
6900              register int pp = *p++;              register pcre_uint32 pp = RAWUCHARINCTEST(p);
6901              if (pp == req_char || pp == req_char2) { p--; break; }              if (pp == req_char || pp == req_char2) { p--; break; }
6902              }              }
6903            }            }
# Line 6843  for(;;) Line 6905  for(;;)
6905            {            {
6906            while (p < end_subject)            while (p < end_subject)
6907              {              {
6908              if (*p++ == req_char) { p--; break; }              if (RAWUCHARINCTEST(p) == req_char) { p--; break; }
6909              }              }
6910            }            }
6911    
# Line 6879  for(;;) Line 6941  for(;;)
6941    md->match_call_count = 0;    md->match_call_count = 0;
6942    md->match_function_type = 0;    md->match_function_type = 0;
6943    md->end_offset_top = 0;    md->end_offset_top = 0;
6944      md->skip_arg_count = 0;
6945    rc = match(start_match, md->start_code, start_match, 2, md, NULL, 0);    rc = match(start_match, md->start_code, start_match, 2, md, NULL, 0);
6946    if (md->hitend && start_partial == NULL) start_partial = md->start_used_ptr;    if (md->hitend && start_partial == NULL)
6947        {
6948        start_partial = md->start_used_ptr;
6949        match_partial = start_match;
6950        }
6951    
6952    switch(rc)    switch(rc)
6953      {      {
# Line 6893  for(;;) Line 6960  for(;;)
6960    
6961      case MATCH_SKIP_ARG:      case MATCH_SKIP_ARG:
6962      new_start_match = start_match;      new_start_match = start_match;
6963      md->ignore_skip_arg = TRUE;      md->ignore_skip_arg = md->skip_arg_count;
6964      break;      break;
6965    
6966      /* SKIP passes back the next starting point explicitly, but if it is the      /* SKIP passes back the next starting point explicitly, but if it is no
6967      same as the match we have just done, treat it as NOMATCH. */      greater than the match we have just done, treat it as NOMATCH. */
6968    
6969      case MATCH_SKIP:      case MATCH_SKIP:
6970      if (md->start_match_ptr != start_match)      if (md->start_match_ptr > start_match)
6971        {        {
6972        new_start_match = md->start_match_ptr;        new_start_match = md->start_match_ptr;
6973        break;        break;
# Line 6908  for(;;) Line 6975  for(;;)
6975      /* Fall through */      /* Fall through */
6976    
6977      /* NOMATCH and PRUNE advance by one character. THEN at this level acts      /* NOMATCH and PRUNE advance by one character. THEN at this level acts
6978      exactly like PRUNE. Unset the ignore SKIP-with-argument flag. */      exactly like PRUNE. Unset ignore SKIP-with-argument. */
6979    
6980      case MATCH_NOMATCH:      case MATCH_NOMATCH:
6981      case MATCH_PRUNE:      case MATCH_PRUNE:
6982      case MATCH_THEN:      case MATCH_THEN:
6983      md->ignore_skip_arg = FALSE;      md->ignore_skip_arg = 0;
6984      new_start_match = start_match + 1;      new_start_match = start_match + 1;
6985  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
6986      if (utf)      if (utf)
#