/[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 903 by ph10, Sat Jan 21 16:37:17 2012 UTC revision 1397 by ph10, Mon Nov 11 18:33:23 2013 UTC
# Line 6  Line 6 
6  and semantics are as close as possible to those of the Perl 5 language.  and semantics are as close as possible to those of the Perl 5 language.
7    
8                         Written by Philip Hazel                         Written by Philip Hazel
9             Copyright (c) 1997-2012 University of Cambridge             Copyright (c) 1997-2013 University of Cambridge
10    
11  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
12  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 37  POSSIBILITY OF SUCH DAMAGE. Line 37  POSSIBILITY OF SUCH DAMAGE.
37  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
38  */  */
39    
   
40  /* This module contains pcre_exec(), the externally visible function that does  /* This module contains pcre_exec(), the externally visible function that does
41  pattern matching using an NFA algorithm, trying to mimic Perl as closely as  pattern matching using an NFA algorithm, trying to mimic Perl as closely as
42  possible. There are also some static supporting functions. */  possible. There are also some static supporting functions. */
# Line 57  possible. There are also some static sup Line 56  possible. There are also some static sup
56  #undef min  #undef min
57  #undef max  #undef max
58    
59    /* The md->capture_last field uses the lower 16 bits for the last captured
60    substring (which can never be greater than 65535) and a bit in the top half
61    to mean "capture vector overflowed". This odd way of doing things was
62    implemented when it was realized that preserving and restoring the overflow bit
63    whenever the last capture number was saved/restored made for a neater
64    interface, and doing it this way saved on (a) another variable, which would
65    have increased the stack frame size (a big NO-NO in PCRE) and (b) another
66    separate set of save/restore instructions. The following defines are used in
67    implementing this. */
68    
69    #define CAPLMASK    0x0000ffff    /* The bits used for last_capture */
70    #define OVFLMASK    0xffff0000    /* The bits used for the overflow flag */
71    #define OVFLBIT     0x00010000    /* The bit that is set for overflow */
72    
73  /* Values for setting in md->match_function_type to indicate two special types  /* Values for setting in md->match_function_type to indicate two special types
74  of call to match(). We do it this way to save on using another stack variable,  of call to match(). We do it this way to save on using another stack variable,
75  as stack usage is to be discouraged. */  as stack usage is to be discouraged. */
# Line 74  defined PCRE_ERROR_xxx codes, which are Line 87  defined PCRE_ERROR_xxx codes, which are
87  negative to avoid the external error codes. */  negative to avoid the external error codes. */
88    
89  #define MATCH_ACCEPT       (-999)  #define MATCH_ACCEPT       (-999)
90  #define MATCH_COMMIT       (-998)  #define MATCH_KETRPOS      (-998)
91  #define MATCH_KETRPOS      (-997)  #define MATCH_ONCE         (-997)
92  #define MATCH_ONCE         (-996)  /* The next 5 must be kept together and in sequence so that a test that checks
93    for any one of them can use a range. */
94    #define MATCH_COMMIT       (-996)
95  #define MATCH_PRUNE        (-995)  #define MATCH_PRUNE        (-995)
96  #define MATCH_SKIP         (-994)  #define MATCH_SKIP         (-994)
97  #define MATCH_SKIP_ARG     (-993)  #define MATCH_SKIP_ARG     (-993)
98  #define MATCH_THEN         (-992)  #define MATCH_THEN         (-992)
99    #define MATCH_BACKTRACK_MAX MATCH_THEN
100    #define MATCH_BACKTRACK_MIN MATCH_COMMIT
101    
102  /* Maximum number of ints of offset to save on the stack for recursive calls.  /* Maximum number of ints of offset to save on the stack for recursive calls.
103  If the offset vector is bigger, malloc is used. This should be a multiple of 3,  If the offset vector is bigger, malloc is used. This should be a multiple of 3,
# Line 90  because the offset vector is always a mu Line 107  because the offset vector is always a mu
107    
108  /* Min and max values for the common repeats; for the maxima, 0 => infinity */  /* Min and max values for the common repeats; for the maxima, 0 => infinity */
109    
110  static const char rep_min[] = { 0, 0, 1, 1, 0, 0 };  static const char rep_min[] = { 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, };
111  static const char rep_max[] = { 0, 0, 0, 0, 1, 1 };  static const char rep_max[] = { 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, };
   
   
112    
113  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
114  /*************************************************  /*************************************************
# Line 115  Returns:     nothing Line 130  Returns:     nothing
130  static void  static void
131  pchars(const pcre_uchar *p, int length, BOOL is_subject, match_data *md)  pchars(const pcre_uchar *p, int length, BOOL is_subject, match_data *md)
132  {  {
133  unsigned int c;  pcre_uint32 c;
134    BOOL utf = md->utf;
135  if (is_subject && length > md->end_subject - p) length = md->end_subject - p;  if (is_subject && length > md->end_subject - p) length = md->end_subject - p;
136  while (length-- > 0)  while (length-- > 0)
137    if (isprint(c = *(p++))) printf("%c", c); else printf("\\x%02x", c);    if (isprint(c = RAWUCHARINCTEST(p))) printf("%c", (char)c); else printf("\\x{%02x}", c);
138  }  }
139  #endif  #endif
140    
# Line 140  Arguments: Line 156  Arguments:
156    md          points to match data block    md          points to match data block
157    caseless    TRUE if caseless    caseless    TRUE if caseless
158    
159  Returns:      < 0 if not matched, otherwise the number of subject bytes matched  Returns:      >= 0 the number of subject bytes matched
160                  -1 no match
161                  -2 partial match; always given if at end subject
162  */  */
163    
164  static int  static int
# Line 149  match_ref(int offset, register PCRE_PUCH Line 167  match_ref(int offset, register PCRE_PUCH
167  {  {
168  PCRE_PUCHAR eptr_start = eptr;  PCRE_PUCHAR eptr_start = eptr;
169  register PCRE_PUCHAR p = md->start_subject + md->offset_vector[offset];  register PCRE_PUCHAR p = md->start_subject + md->offset_vector[offset];
170    #ifdef SUPPORT_UTF
171    BOOL utf = md->utf;
172    #endif
173    
174  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
175  if (eptr >= md->end_subject)  if (eptr >= md->end_subject)
# Line 163  pchars(p, length, FALSE, md); Line 184  pchars(p, length, FALSE, md);
184  printf("\n");  printf("\n");
185  #endif  #endif
186    
187  /* Always fail if reference not set (and not JavaScript compatible). */  /* Always fail if reference not set (and not JavaScript compatible - in that
188    case the length is passed as zero). */
189    
190  if (length < 0) return -1;  if (length < 0) return -1;
191    
# Line 175  if (caseless) Line 197  if (caseless)
197    {    {
198  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
199  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
200    if (md->utf)    if (utf)
201      {      {
202      /* Match characters up to the end of the reference. NOTE: the number of      /* Match characters up to the end of the reference. NOTE: the number of
203      bytes matched may differ, because there are some characters whose upper and      data units matched may differ, because in UTF-8 there are some characters
204      lower case versions code as different numbers of bytes. For example, U+023A      whose upper and lower case versions code have different numbers of bytes.
205      (2 bytes in UTF-8) is the upper case version of U+2C65 (3 bytes in UTF-8);      For example, U+023A (2 bytes in UTF-8) is the upper case version of U+2C65
206      a sequence of 3 of the former uses 6 bytes, as does a sequence of two of      (3 bytes in UTF-8); a sequence of 3 of the former uses 6 bytes, as does a
207      the latter. It is important, therefore, to check the length along the      sequence of two of the latter. It is important, therefore, to check the
208      reference, not along the subject (earlier code did this wrong). */      length along the reference, not along the subject (earlier code did this
209        wrong). */
210    
211      PCRE_PUCHAR endptr = p + length;      PCRE_PUCHAR endptr = p + length;
212      while (p < endptr)      while (p < endptr)
213        {        {
214        int c, d;        pcre_uint32 c, d;
215        if (eptr >= md->end_subject) return -1;        const ucd_record *ur;
216          if (eptr >= md->end_subject) return -2;   /* Partial match */
217        GETCHARINC(c, eptr);        GETCHARINC(c, eptr);
218        GETCHARINC(d, p);        GETCHARINC(d, p);
219        if (c != d && c != UCD_OTHERCASE(d)) return -1;        ur = GET_UCD(d);
220          if (c != d && c != d + ur->other_case)
221            {
222            const pcre_uint32 *pp = PRIV(ucd_caseless_sets) + ur->caseset;
223            for (;;)
224              {
225              if (c < *pp) return -1;
226              if (c == *pp++) break;
227              }
228            }
229        }        }
230      }      }
231    else    else
# Line 202  if (caseless) Line 235  if (caseless)
235    /* The same code works when not in UTF-8 mode and in UTF-8 mode when there    /* The same code works when not in UTF-8 mode and in UTF-8 mode when there
236    is no UCP support. */    is no UCP support. */
237      {      {
     if (eptr + length > md->end_subject) return -1;  
238      while (length-- > 0)      while (length-- > 0)
239        {        {
240        if (TABLE_GET(*p, md->lcc, *p) != TABLE_GET(*eptr, md->lcc, *eptr)) return -1;        pcre_uint32 cc, cp;
241          if (eptr >= md->end_subject) return -2;   /* Partial match */
242          cc = RAWUCHARTEST(eptr);
243          cp = RAWUCHARTEST(p);
244          if (TABLE_GET(cp, md->lcc, cp) != TABLE_GET(cc, md->lcc, cc)) return -1;
245        p++;        p++;
246        eptr++;        eptr++;
247        }        }
# Line 217  are in UTF-8 mode. */ Line 253  are in UTF-8 mode. */
253    
254  else  else
255    {    {
256    if (eptr + length > md->end_subject) return -1;    while (length-- > 0)
257    while (length-- > 0) if (*p++ != *eptr++) return -1;      {
258        if (eptr >= md->end_subject) return -2;   /* Partial match */
259        if (RAWUCHARINCTEST(p) != RAWUCHARINCTEST(eptr)) return -1;
260        }
261    }    }
262    
263  return (int)(eptr - eptr_start);  return (int)(eptr - eptr_start);
# Line 273  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM Line 312  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM
312         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,
313         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,         RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,
314         RM51,  RM52, RM53, RM54, RM55, RM56, RM57, RM58, RM59, RM60,         RM51,  RM52, RM53, RM54, RM55, RM56, RM57, RM58, RM59, RM60,
315         RM61,  RM62, RM63, RM64, RM65, RM66 };         RM61,  RM62, RM63, RM64, RM65, RM66, RM67, RM68 };
316    
317  /* These versions of the macros use the stack, as normal. There are debugging  /* These versions of the macros use the stack, as normal. There are debugging
318  versions and production versions. Note that the "rw" argument of RMATCH isn't  versions and production versions. Note that the "rw" argument of RMATCH isn't
# Line 291  actually used in this definition. */ Line 330  actually used in this definition. */
330    }    }
331  #define RRETURN(ra) \  #define RRETURN(ra) \
332    { \    { \
333    printf("match() returned %d from line %d ", ra, __LINE__); \    printf("match() returned %d from line %d\n", ra, __LINE__); \
334    return ra; \    return ra; \
335    }    }
336  #else  #else
# Line 311  argument of match(), which never changes Line 350  argument of match(), which never changes
350    
351  #define RMATCH(ra,rb,rc,rd,re,rw)\  #define RMATCH(ra,rb,rc,rd,re,rw)\
352    {\    {\
353    heapframe *newframe = (heapframe *)(PUBL(stack_malloc))(sizeof(heapframe));\    heapframe *newframe = frame->Xnextframe;\
354    if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\    if (newframe == NULL)\
355    frame->Xwhere = rw; \      {\
356        newframe = (heapframe *)(PUBL(stack_malloc))(sizeof(heapframe));\
357        if (newframe == NULL) RRETURN(PCRE_ERROR_NOMEMORY);\
358        newframe->Xnextframe = NULL;\
359        frame->Xnextframe = newframe;\
360        }\
361      frame->Xwhere = rw;\
362    newframe->Xeptr = ra;\    newframe->Xeptr = ra;\
363    newframe->Xecode = rb;\    newframe->Xecode = rb;\
364    newframe->Xmstart = mstart;\    newframe->Xmstart = mstart;\
# Line 332  argument of match(), which never changes Line 377  argument of match(), which never changes
377    {\    {\
378    heapframe *oldframe = frame;\    heapframe *oldframe = frame;\
379    frame = oldframe->Xprevframe;\    frame = oldframe->Xprevframe;\
   if (oldframe != &frame_zero) (PUBL(stack_free))(oldframe);\  
380    if (frame != NULL)\    if (frame != NULL)\
381      {\      {\
382      rrc = ra;\      rrc = ra;\
# Line 346  argument of match(), which never changes Line 390  argument of match(), which never changes
390    
391  typedef struct heapframe {  typedef struct heapframe {
392    struct heapframe *Xprevframe;    struct heapframe *Xprevframe;
393      struct heapframe *Xnextframe;
394    
395    /* Function arguments that may change */    /* Function arguments that may change */
396    
# Line 376  typedef struct heapframe { Line 421  typedef struct heapframe {
421    
422  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
423    int Xprop_type;    int Xprop_type;
424    int Xprop_value;    unsigned int Xprop_value;
425    int Xprop_fail_result;    int Xprop_fail_result;
426    int Xoclength;    int Xoclength;
427    pcre_uchar Xocchars[6];    pcre_uchar Xocchars[6];
# Line 389  typedef struct heapframe { Line 434  typedef struct heapframe {
434    int Xlength;    int Xlength;
435    int Xmax;    int Xmax;
436    int Xmin;    int Xmin;
437    int Xnumber;    unsigned int Xnumber;
438    int Xoffset;    int Xoffset;
439    int Xop;    unsigned int Xop;
440    int Xsave_capture_last;    pcre_int32 Xsave_capture_last;
441    int Xsave_offset1, Xsave_offset2, Xsave_offset3;    int Xsave_offset1, Xsave_offset2, Xsave_offset3;
442    int Xstacksave[REC_STACK_SAVE_MAX];    int Xstacksave[REC_STACK_SAVE_MAX];
443    
# Line 477  so they can be ordinary variables in all Line 522  so they can be ordinary variables in all
522    
523  register int  rrc;         /* Returns from recursive calls */  register int  rrc;         /* Returns from recursive calls */
524  register int  i;           /* Used for loops not involving calls to RMATCH() */  register int  i;           /* Used for loops not involving calls to RMATCH() */
525  register unsigned int c;   /* Character values not kept over RMATCH() calls */  register pcre_uint32 c;    /* Character values not kept over RMATCH() calls */
526  register BOOL utf;         /* Local copy of UTF flag for speed */  register BOOL utf;         /* Local copy of UTF flag for speed */
527    
528  BOOL minimize, possessive; /* Quantifier options */  BOOL minimize, possessive; /* Quantifier options */
# Line 492  the top-level on the stack rather than m Line 537  the top-level on the stack rather than m
537  boost in many cases where there is not much "recursion". */  boost in many cases where there is not much "recursion". */
538    
539  #ifdef NO_RECURSE  #ifdef NO_RECURSE
540  heapframe frame_zero;  heapframe *frame = (heapframe *)md->match_frames_base;
 heapframe *frame = &frame_zero;  
 frame->Xprevframe = NULL;            /* Marks the top level */  
541    
542  /* Copy in the original argument variables */  /* Copy in the original argument variables */
543    
# Line 596  BOOL prev_is_word; Line 639  BOOL prev_is_word;
639    
640  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
641  int prop_type;  int prop_type;
642  int prop_value;  unsigned int prop_value;
643  int prop_fail_result;  int prop_fail_result;
644  int oclength;  int oclength;
645  pcre_uchar occhars[6];  pcre_uchar occhars[6];
# Line 607  int ctype; Line 650  int ctype;
650  int length;  int length;
651  int max;  int max;
652  int min;  int min;
653  int number;  unsigned int number;
654  int offset;  int offset;
655  int op;  unsigned int op;
656  int save_capture_last;  pcre_int32 save_capture_last;
657  int save_offset1, save_offset2, save_offset3;  int save_offset1, save_offset2, save_offset3;
658  int stacksave[REC_STACK_SAVE_MAX];  int stacksave[REC_STACK_SAVE_MAX];
659    
# Line 728  for (;;) Line 771  for (;;)
771      unaltered. */      unaltered. */
772    
773      else if (rrc == MATCH_SKIP_ARG &&      else if (rrc == MATCH_SKIP_ARG &&
774          STRCMP_UC_UC(ecode + 2, md->start_match_ptr) == 0)          STRCMP_UC_UC_TEST(ecode + 2, md->start_match_ptr) == 0)
775        {        {
776        md->start_match_ptr = eptr;        md->start_match_ptr = eptr;
777        RRETURN(MATCH_SKIP);        RRETURN(MATCH_SKIP);
# Line 738  for (;;) Line 781  for (;;)
781      case OP_FAIL:      case OP_FAIL:
782      RRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
783    
     /* COMMIT overrides PRUNE, SKIP, and THEN */  
   
784      case OP_COMMIT:      case OP_COMMIT:
785      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
786        eptrb, RM52);        eptrb, RM52);
787      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE &&      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
         rrc != MATCH_SKIP && rrc != MATCH_SKIP_ARG &&  
         rrc != MATCH_THEN)  
       RRETURN(rrc);  
788      RRETURN(MATCH_COMMIT);      RRETURN(MATCH_COMMIT);
789    
     /* PRUNE overrides THEN */  
   
790      case OP_PRUNE:      case OP_PRUNE:
791      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
792        eptrb, RM51);        eptrb, RM51);
793      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
794      RRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
795    
796      case OP_PRUNE_ARG:      case OP_PRUNE_ARG:
# Line 764  for (;;) Line 800  for (;;)
800        eptrb, RM56);        eptrb, RM56);
801      if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&      if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
802           md->mark == NULL) md->mark = ecode + 2;           md->mark == NULL) md->mark = ecode + 2;
803      if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
804      RRETURN(MATCH_PRUNE);      RRETURN(MATCH_PRUNE);
805    
     /* SKIP overrides PRUNE and THEN */  
   
806      case OP_SKIP:      case OP_SKIP:
807      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
808        eptrb, RM53);        eptrb, RM53);
809      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
       RRETURN(rrc);  
810      md->start_match_ptr = eptr;   /* Pass back current position */      md->start_match_ptr = eptr;   /* Pass back current position */
811      RRETURN(MATCH_SKIP);      RRETURN(MATCH_SKIP);
812    
813      /* Note that, for Perl compatibility, SKIP with an argument does NOT set      /* Note that, for Perl compatibility, SKIP with an argument does NOT set
814      nomatch_mark. There is a flag that disables this opcode when re-matching a      nomatch_mark. When a pattern match ends with a SKIP_ARG for which there was
815      pattern that ended with a SKIP for which there was not a matching MARK. */      not a matching mark, we have to re-run the match, ignoring the SKIP_ARG
816        that failed and any that precede it (either they also failed, or were not
817        triggered). To do this, we maintain a count of executed SKIP_ARGs. If a
818        SKIP_ARG gets to top level, the match is re-run with md->ignore_skip_arg
819        set to the count of the one that failed. */
820    
821      case OP_SKIP_ARG:      case OP_SKIP_ARG:
822      if (md->ignore_skip_arg)      md->skip_arg_count++;
823        if (md->skip_arg_count <= md->ignore_skip_arg)
824        {        {
825        ecode += PRIV(OP_lengths)[*ecode] + ecode[1];        ecode += PRIV(OP_lengths)[*ecode] + ecode[1];
826        break;        break;
827        }        }
828      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
829        eptrb, RM57);        eptrb, RM57);
830      if (rrc != MATCH_NOMATCH && rrc != MATCH_PRUNE && rrc != MATCH_THEN)      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
       RRETURN(rrc);  
831    
832      /* Pass back the current skip name by overloading md->start_match_ptr and      /* Pass back the current skip name by overloading md->start_match_ptr and
833      returning the special MATCH_SKIP_ARG return code. This will either be      returning the special MATCH_SKIP_ARG return code. This will either be
834      caught by a matching MARK, or get to the top, where it causes a rematch      caught by a matching MARK, or get to the top, where it causes a rematch
835      with the md->ignore_skip_arg flag set. */      with md->ignore_skip_arg set to the value of md->skip_arg_count. */
836    
837      md->start_match_ptr = ecode + 2;      md->start_match_ptr = ecode + 2;
838      RRETURN(MATCH_SKIP_ARG);      RRETURN(MATCH_SKIP_ARG);
# Line 897  for (;;) Line 934  for (;;)
934        }        }
935      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
936        {        {
       md->match_function_type = MATCH_CBEGROUP;  
937        RMATCH(eptr, prev, offset_top, md, eptrb, RM66);        RMATCH(eptr, prev, offset_top, md, eptrb, RM66);
938        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
939        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
# Line 1026  for (;;) Line 1062  for (;;)
1062    
1063      for (;;)      for (;;)
1064        {        {
1065        if (op >= OP_SBRA || op == OP_ONCE) md->match_function_type = MATCH_CBEGROUP;        if (op >= OP_SBRA || op == OP_ONCE)
1066            md->match_function_type = MATCH_CBEGROUP;
1067    
1068        /* If this is not a possibly empty group, and there are no (*THEN)s in        /* If this is not a possibly empty group, and there are no (*THEN)s in
1069        the pattern, and this is the final alternative, optimize as described        the pattern, and this is the final alternative, optimize as described
# Line 1041  for (;;) Line 1078  for (;;)
1078        /* In all other cases, we have to make another call to match(). */        /* In all other cases, we have to make another call to match(). */
1079    
1080        save_mark = md->mark;        save_mark = md->mark;
1081          save_capture_last = md->capture_last;
1082        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md, eptrb,        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md, eptrb,
1083          RM2);          RM2);
1084    
# Line 1072  for (;;) Line 1110  for (;;)
1110        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1111        md->mark = save_mark;        md->mark = save_mark;
1112        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1113          md->capture_last = save_capture_last;
1114        }        }
1115    
1116      RRETURN(MATCH_NOMATCH);      RRETURN(MATCH_NOMATCH);
# Line 1134  for (;;) Line 1173  for (;;)
1173            ecode = md->start_code + code_offset;            ecode = md->start_code + code_offset;
1174            save_capture_last = md->capture_last;            save_capture_last = md->capture_last;
1175            matched_once = TRUE;            matched_once = TRUE;
1176              mstart = md->start_match_ptr;    /* In case \K changed it */
1177            continue;            continue;
1178            }            }
1179    
# Line 1193  for (;;) Line 1233  for (;;)
1233      POSSESSIVE_NON_CAPTURE:      POSSESSIVE_NON_CAPTURE:
1234      matched_once = FALSE;      matched_once = FALSE;
1235      code_offset = (int)(ecode - md->start_code);      code_offset = (int)(ecode - md->start_code);
1236        save_capture_last = md->capture_last;
1237    
1238      for (;;)      for (;;)
1239        {        {
# Line 1205  for (;;) Line 1246  for (;;)
1246          eptr = md->end_match_ptr;          eptr = md->end_match_ptr;
1247          ecode = md->start_code + code_offset;          ecode = md->start_code + code_offset;
1248          matched_once = TRUE;          matched_once = TRUE;
1249            mstart = md->start_match_ptr;   /* In case \K reset it */
1250          continue;          continue;
1251          }          }
1252    
# Line 1222  for (;;) Line 1264  for (;;)
1264        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1265        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1266        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1267          md->capture_last = save_capture_last;
1268        }        }
1269    
1270      if (matched_once || allow_zero)      if (matched_once || allow_zero)
# Line 1233  for (;;) Line 1276  for (;;)
1276    
1277      /* Control never reaches here. */      /* Control never reaches here. */
1278    
1279      /* Conditional group: compilation checked that there are no more than      /* Conditional group: compilation checked that there are no more than two
1280      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
1281      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
1282      exactly what going to the ket would do. */      what we want. */
1283    
1284      case OP_COND:      case OP_COND:
1285      case OP_SCOND:      case OP_SCOND:
1286      codelink = GET(ecode, 1);  
1287        /* The variable codelink will be added to ecode when the condition is
1288        false, to get to the second branch. Setting it to the offset to the ALT
1289        or KET, then incrementing ecode achieves this effect. We now have ecode
1290        pointing to the condition or callout. */
1291    
1292        codelink = GET(ecode, 1);   /* Offset to the second branch */
1293        ecode += 1 + LINK_SIZE;     /* From this opcode */
1294    
1295      /* 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
1296      inserted between OP_COND and an assertion condition. */      inserted between OP_COND and an assertion condition. */
1297    
1298      if (ecode[LINK_SIZE+1] == OP_CALLOUT)      if (*ecode == OP_CALLOUT)
1299        {        {
1300        if (PUBL(callout) != NULL)        if (PUBL(callout) != NULL)
1301          {          {
1302          PUBL(callout_block) cb;          PUBL(callout_block) cb;
1303          cb.version          = 2;   /* Version 1 of the callout block */          cb.version          = 2;   /* Version 1 of the callout block */
1304          cb.callout_number   = ecode[LINK_SIZE+2];          cb.callout_number   = ecode[1];
1305          cb.offset_vector    = md->offset_vector;          cb.offset_vector    = md->offset_vector;
1306  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
1307          cb.subject          = (PCRE_SPTR)md->start_subject;          cb.subject          = (PCRE_SPTR)md->start_subject;
1308  #else  #elif defined COMPILE_PCRE16
1309          cb.subject          = (PCRE_SPTR16)md->start_subject;          cb.subject          = (PCRE_SPTR16)md->start_subject;
1310    #elif defined COMPILE_PCRE32
1311            cb.subject          = (PCRE_SPTR32)md->start_subject;
1312  #endif  #endif
1313          cb.subject_length   = (int)(md->end_subject - md->start_subject);          cb.subject_length   = (int)(md->end_subject - md->start_subject);
1314          cb.start_match      = (int)(mstart - md->start_subject);          cb.start_match      = (int)(mstart - md->start_subject);
1315          cb.current_position = (int)(eptr - md->start_subject);          cb.current_position = (int)(eptr - md->start_subject);
1316          cb.pattern_position = GET(ecode, LINK_SIZE + 3);          cb.pattern_position = GET(ecode, 2);
1317          cb.next_item_length = GET(ecode, 3 + 2*LINK_SIZE);          cb.next_item_length = GET(ecode, 2 + LINK_SIZE);
1318          cb.capture_top      = offset_top/2;          cb.capture_top      = offset_top/2;
1319          cb.capture_last     = md->capture_last;          cb.capture_last     = md->capture_last & CAPLMASK;
1320            /* Internal change requires this for API compatibility. */
1321            if (cb.capture_last == 0) cb.capture_last = -1;
1322          cb.callout_data     = md->callout_data;          cb.callout_data     = md->callout_data;
1323          cb.mark             = md->nomatch_mark;          cb.mark             = md->nomatch_mark;
1324          if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);          if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1325          if (rrc < 0) RRETURN(rrc);          if (rrc < 0) RRETURN(rrc);
1326          }          }
1327    
1328          /* Advance ecode past the callout, so it now points to the condition. We
1329          must adjust codelink so that the value of ecode+codelink is unchanged. */
1330    
1331        ecode += PRIV(OP_lengths)[OP_CALLOUT];        ecode += PRIV(OP_lengths)[OP_CALLOUT];
1332          codelink -= PRIV(OP_lengths)[OP_CALLOUT];
1333        }        }
1334    
1335      condcode = ecode[LINK_SIZE+1];      /* Test the various possible conditions */
1336    
1337      /* Now see what the actual condition is */      condition = FALSE;
1338        switch(condcode = *ecode)
     if (condcode == OP_RREF || condcode == OP_NRREF)    /* Recursion test */  
1339        {        {
1340        if (md->recursive == NULL)                /* Not recursing => FALSE */        case OP_RREF:         /* Numbered group recursion test */
1341          {        if (md->recursive != NULL)     /* Not recursing => FALSE */
         condition = FALSE;  
         ecode += GET(ecode, 1);  
         }  
       else  
1342          {          {
1343          int recno = GET2(ecode, LINK_SIZE + 2);   /* Recursion group number*/          unsigned int recno = GET2(ecode, 1);   /* Recursion group number*/
1344          condition = (recno == RREF_ANY || recno == md->recursive->group_num);          condition = (recno == RREF_ANY || recno == md->recursive->group_num);
1345            }
1346          break;
1347    
1348          /* If the test is for recursion into a specific subpattern, and it is        case OP_DNRREF:       /* Duplicate named group recursion test */
1349          false, but the test was set up by name, scan the table to see if the        if (md->recursive != NULL)
1350          name refers to any other numbers, and test them. The condition is true          {
1351          if any one is set. */          int count = GET2(ecode, 1 + IMM2_SIZE);
1352            pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
1353          if (!condition && condcode == OP_NRREF)          while (count-- > 0)
1354            {            {
1355            pcre_uchar *slotA = md->name_table;            unsigned int recno = GET2(slot, 0);
1356            for (i = 0; i < md->name_count; i++)            condition = recno == md->recursive->group_num;
1357              {            if (condition) break;
1358              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;  
                 }  
               }  
             }  
1359            }            }
   
         /* Chose branch according to the condition */  
   
         ecode += condition? 1 + IMM2_SIZE : GET(ecode, 1);  
1360          }          }
1361        }        break;
1362    
1363      else if (condcode == OP_CREF || condcode == OP_NCREF)  /* Group used test */        case OP_CREF:         /* Numbered group used test */
1364        {        offset = GET2(ecode, 1) << 1;  /* Doubled ref number */
       offset = GET2(ecode, LINK_SIZE+2) << 1;  /* Doubled ref number */  
1365        condition = offset < offset_top && md->offset_vector[offset] >= 0;        condition = offset < offset_top && md->offset_vector[offset] >= 0;
1366          break;
1367    
1368        /* 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)  
1369          {          {
1370          int refno = offset >> 1;          int count = GET2(ecode, 1 + IMM2_SIZE);
1371          pcre_uchar *slotA = md->name_table;          pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
1372            while (count-- > 0)
1373          for (i = 0; i < md->name_count; i++)            {
1374            {            offset = GET2(slot, 0) << 1;
1375            if (GET2(slotA, 0) == refno) break;            condition = offset < offset_top && md->offset_vector[offset] >= 0;
1376            slotA += md->name_entry_size;            if (condition) break;
1377            }            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;  
               }  
             }  
1378            }            }
1379          }          }
1380          break;
1381    
1382        /* Chose branch according to the condition */        case OP_DEF:     /* DEFINE - always false */
1383          break;
       ecode += condition? 1 + IMM2_SIZE : GET(ecode, 1);  
       }  
   
     else if (condcode == OP_DEF)     /* DEFINE - always false */  
       {  
       condition = FALSE;  
       ecode += GET(ecode, 1);  
       }  
1384    
1385      /* The condition is an assertion. Call match() to evaluate it - setting        /* The condition is an assertion. Call match() to evaluate it - setting
1386      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
1387      an assertion. */        of an assertion. */
1388    
1389      else        default:
       {  
1390        md->match_function_type = MATCH_CONDASSERT;        md->match_function_type = MATCH_CONDASSERT;
1391        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM3);        RMATCH(eptr, ecode, offset_top, md, NULL, RM3);
1392        if (rrc == MATCH_MATCH)        if (rrc == MATCH_MATCH)
1393          {          {
1394          if (md->end_offset_top > offset_top)          if (md->end_offset_top > offset_top)
1395            offset_top = md->end_offset_top;  /* Captures may have happened */            offset_top = md->end_offset_top;  /* Captures may have happened */
1396          condition = TRUE;          condition = TRUE;
1397          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);  
1398            /* Advance ecode past the assertion to the start of the first branch,
1399            but adjust it so that the general choosing code below works. */
1400    
1401            ecode += GET(ecode, 1);
1402          while (*ecode == OP_ALT) ecode += GET(ecode, 1);          while (*ecode == OP_ALT) ecode += GET(ecode, 1);
1403            ecode += 1 + LINK_SIZE - PRIV(OP_lengths)[condcode];
1404          }          }
1405    
1406        /* PCRE doesn't allow the effect of (*THEN) to escape beyond an        /* PCRE doesn't allow the effect of (*THEN) to escape beyond an
1407        assertion; it is therefore treated as NOMATCH. */        assertion; it is therefore treated as NOMATCH. Any other return is an
1408          error. */
1409    
1410        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
1411          {          {
1412          RRETURN(rrc);         /* Need braces because of following else */          RRETURN(rrc);         /* Need braces because of following else */
1413          }          }
1414        else        break;
         {  
         condition = FALSE;  
         ecode += codelink;  
         }  
1415        }        }
1416    
1417      /* We are now at the branch that is to be obeyed. As there is only one, can      /* Choose branch according to the condition */
     use tail recursion to avoid using another stack frame, except when there is  
     unlimited repeat of a possibly empty group. In the latter case, a recursive  
     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)      ecode += condition? PRIV(OP_lengths)[condcode] : codelink;
1420    
1421        /* We are now at the branch that is to be obeyed. As there is only one, we
1422        can use tail recursion to avoid using another stack frame, except when
1423        there is unlimited repeat of a possibly empty group. In the latter case, a
1424        recursive call to match() is always required, unless the second alternative
1425        doesn't exist, in which case we can just plough on. Note that, for
1426        compatibility with Perl, the | in a conditional group is NOT treated as
1427        creating two alternatives. If a THEN is encountered in the branch, it
1428        propagates out to the enclosing alternative (unless nested in a deeper set
1429        of alternatives, of course). */
1430    
1431        if (condition || ecode[-(1+LINK_SIZE)] == OP_ALT)
1432        {        {
1433        if (op != OP_SCOND)        if (op != OP_SCOND)
1434          {          {
         ecode += 1 + LINK_SIZE;  
1435          goto TAIL_RECURSE;          goto TAIL_RECURSE;
1436          }          }
1437    
1438        md->match_function_type = MATCH_CBEGROUP;        md->match_function_type = MATCH_CBEGROUP;
1439        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM49);        RMATCH(eptr, ecode, offset_top, md, eptrb, RM49);
1440        RRETURN(rrc);        RRETURN(rrc);
1441        }        }
1442    
# Line 1477  for (;;) Line 1444  for (;;)
1444    
1445      else      else
1446        {        {
       ecode += 1 + LINK_SIZE;  
1447        }        }
1448      break;      break;
1449    
# Line 1486  for (;;) Line 1452  for (;;)
1452      to close any currently open capturing brackets. */      to close any currently open capturing brackets. */
1453    
1454      case OP_CLOSE:      case OP_CLOSE:
1455      number = GET2(ecode, 1);      number = GET2(ecode, 1);   /* Must be less than 65536 */
1456      offset = number << 1;      offset = number << 1;
1457    
1458  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
# Line 1494  for (;;) Line 1460  for (;;)
1460        printf("\n");        printf("\n");
1461  #endif  #endif
1462    
1463      md->capture_last = number;      md->capture_last = (md->capture_last & OVFLMASK) | number;
1464      if (offset >= md->offset_max) md->offset_overflow = TRUE; else      if (offset >= md->offset_max) md->capture_last |= OVFLBIT; else
1465        {        {
1466        md->offset_vector[offset] =        md->offset_vector[offset] =
1467          md->offset_vector[md->offset_end - number];          md->offset_vector[md->offset_end - number];
# Line 1557  for (;;) Line 1523  for (;;)
1523        }        }
1524      else condassert = FALSE;      else condassert = FALSE;
1525    
1526        /* Loop for each branch */
1527    
1528      do      do
1529        {        {
1530        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM4);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM4);
1531    
1532          /* A match means that the assertion is true; break out of the loop
1533          that matches its alternatives. */
1534    
1535        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1536          {          {
1537          mstart = md->start_match_ptr;   /* In case \K reset it */          mstart = md->start_match_ptr;   /* In case \K reset it */
1538          break;          break;
1539          }          }
1540    
1541        /* PCRE does not allow THEN to escape beyond an assertion; it is treated        /* If not matched, restore the previous mark setting. */
       as NOMATCH. */  
1542    
       if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);  
       ecode += GET(ecode, 1);  
1543        md->mark = save_mark;        md->mark = save_mark;
1544    
1545          /* See comment in the code for capturing groups above about handling
1546          THEN. */
1547    
1548          if (rrc == MATCH_THEN)
1549            {
1550            next = ecode + GET(ecode,1);
1551            if (md->start_match_ptr < next &&
1552                (*ecode == OP_ALT || *next == OP_ALT))
1553              rrc = MATCH_NOMATCH;
1554            }
1555    
1556          /* Anything other than NOMATCH causes the entire assertion to fail,
1557          passing back the return code. This includes COMMIT, SKIP, PRUNE and an
1558          uncaptured THEN, which means they take their normal effect. This
1559          consistent approach does not always have exactly the same effect as in
1560          Perl. */
1561    
1562          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1563          ecode += GET(ecode, 1);
1564        }        }
1565      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);   /* Continue for next alternative */
1566    
1567        /* If we have tried all the alternative branches, the assertion has
1568        failed. If not, we broke out after a match. */
1569    
1570      if (*ecode == OP_KET) RRETURN(MATCH_NOMATCH);      if (*ecode == OP_KET) RRETURN(MATCH_NOMATCH);
1571    
# Line 1581  for (;;) Line 1573  for (;;)
1573    
1574      if (condassert) RRETURN(MATCH_MATCH);      if (condassert) RRETURN(MATCH_MATCH);
1575    
1576      /* Continue from after the assertion, updating the offsets high water      /* Continue from after a successful assertion, updating the offsets high
1577      mark, since extracts may have been taken during the assertion. */      water mark, since extracts may have been taken during the assertion. */
1578    
1579      do ecode += GET(ecode,1); while (*ecode == OP_ALT);      do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1580      ecode += 1 + LINK_SIZE;      ecode += 1 + LINK_SIZE;
1581      offset_top = md->end_offset_top;      offset_top = md->end_offset_top;
1582      continue;      continue;
1583    
1584      /* Negative assertion: all branches must fail to match. Encountering SKIP,      /* Negative assertion: all branches must fail to match for the assertion to
1585      PRUNE, or COMMIT means we must assume failure without checking subsequent      succeed. */
     branches. */  
1586    
1587      case OP_ASSERT_NOT:      case OP_ASSERT_NOT:
1588      case OP_ASSERTBACK_NOT:      case OP_ASSERTBACK_NOT:
# Line 1603  for (;;) Line 1594  for (;;)
1594        }        }
1595      else condassert = FALSE;      else condassert = FALSE;
1596    
1597        /* Loop for each alternative branch. */
1598    
1599      do      do
1600        {        {
1601        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);
1602        md->mark = save_mark;        md->mark = save_mark;   /* Always restore the mark setting */
1603        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) RRETURN(MATCH_NOMATCH);  
1604        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)        switch(rrc)
1605          {          {
1606          do ecode += GET(ecode,1); while (*ecode == OP_ALT);          case MATCH_MATCH:            /* A successful match means */
1607            case MATCH_ACCEPT:           /* the assertion has failed. */
1608            RRETURN(MATCH_NOMATCH);
1609    
1610            case MATCH_NOMATCH:          /* Carry on with next branch */
1611          break;          break;
1612    
1613            /* See comment in the code for capturing groups above about handling
1614            THEN. */
1615    
1616            case MATCH_THEN:
1617            next = ecode + GET(ecode,1);
1618            if (md->start_match_ptr < next &&
1619                (*ecode == OP_ALT || *next == OP_ALT))
1620              {
1621              rrc = MATCH_NOMATCH;
1622              break;
1623              }
1624            /* Otherwise fall through. */
1625    
1626            /* COMMIT, SKIP, PRUNE, and an uncaptured THEN cause the whole
1627            assertion to fail to match, without considering any more alternatives.
1628            Failing to match means the assertion is true. This is a consistent
1629            approach, but does not always have the same effect as in Perl. */
1630    
1631            case MATCH_COMMIT:
1632            case MATCH_SKIP:
1633            case MATCH_SKIP_ARG:
1634            case MATCH_PRUNE:
1635            do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1636            goto NEG_ASSERT_TRUE;   /* Break out of alternation loop */
1637    
1638            /* Anything else is an error */
1639    
1640            default:
1641            RRETURN(rrc);
1642          }          }
1643    
1644        /* PCRE does not allow THEN to escape beyond an assertion; it is treated        /* Continue with next branch */
       as NOMATCH. */  
1645    
       if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);  
1646        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1647        }        }
1648      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
1649    
1650      if (condassert) RRETURN(MATCH_MATCH);  /* Condition assertion */      /* All branches in the assertion failed to match. */
1651    
1652      ecode += 1 + LINK_SIZE;      NEG_ASSERT_TRUE:
1653        if (condassert) RRETURN(MATCH_MATCH);  /* Condition assertion */
1654        ecode += 1 + LINK_SIZE;                /* Continue with current branch */
1655      continue;      continue;
1656    
1657      /* Move the subject pointer back. This occurs only at the start of      /* Move the subject pointer back. This occurs only at the start of
# Line 1671  for (;;) Line 1698  for (;;)
1698        cb.version          = 2;   /* Version 1 of the callout block */        cb.version          = 2;   /* Version 1 of the callout block */
1699        cb.callout_number   = ecode[1];        cb.callout_number   = ecode[1];
1700        cb.offset_vector    = md->offset_vector;        cb.offset_vector    = md->offset_vector;
1701  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
1702        cb.subject          = (PCRE_SPTR)md->start_subject;        cb.subject          = (PCRE_SPTR)md->start_subject;
1703  #else  #elif defined COMPILE_PCRE16
1704        cb.subject          = (PCRE_SPTR16)md->start_subject;        cb.subject          = (PCRE_SPTR16)md->start_subject;
1705    #elif defined COMPILE_PCRE32
1706          cb.subject          = (PCRE_SPTR32)md->start_subject;
1707  #endif  #endif
1708        cb.subject_length   = (int)(md->end_subject - md->start_subject);        cb.subject_length   = (int)(md->end_subject - md->start_subject);
1709        cb.start_match      = (int)(mstart - md->start_subject);        cb.start_match      = (int)(mstart - md->start_subject);
# Line 1682  for (;;) Line 1711  for (;;)
1711        cb.pattern_position = GET(ecode, 2);        cb.pattern_position = GET(ecode, 2);
1712        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);
1713        cb.capture_top      = offset_top/2;        cb.capture_top      = offset_top/2;
1714        cb.capture_last     = md->capture_last;        cb.capture_last     = md->capture_last & CAPLMASK;
1715          /* Internal change requires this for API compatibility. */
1716          if (cb.capture_last == 0) cb.capture_last = -1;
1717        cb.callout_data     = md->callout_data;        cb.callout_data     = md->callout_data;
1718        cb.mark             = md->nomatch_mark;        cb.mark             = md->nomatch_mark;
1719        if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);        if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
# Line 1711  for (;;) Line 1742  for (;;)
1742      case OP_RECURSE:      case OP_RECURSE:
1743        {        {
1744        recursion_info *ri;        recursion_info *ri;
1745        int recno;        unsigned int recno;
1746    
1747        callpat = md->start_code + GET(ecode, 1);        callpat = md->start_code + GET(ecode, 1);
1748        recno = (callpat == md->start_code)? 0 :        recno = (callpat == md->start_code)? 0 :
# Line 1728  for (;;) Line 1759  for (;;)
1759        /* Add to "recursing stack" */        /* Add to "recursing stack" */
1760    
1761        new_recursive.group_num = recno;        new_recursive.group_num = recno;
1762          new_recursive.saved_capture_last = md->capture_last;
1763        new_recursive.subject_position = eptr;        new_recursive.subject_position = eptr;
1764        new_recursive.prevrec = md->recursive;        new_recursive.prevrec = md->recursive;
1765        md->recursive = &new_recursive;        md->recursive = &new_recursive;
# Line 1751  for (;;) Line 1783  for (;;)
1783              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1784    
1785        /* OK, now we can do the recursion. After processing each alternative,        /* OK, now we can do the recursion. After processing each alternative,
1786        restore the offset data. If there were nested recursions, md->recursive        restore the offset data and the last captured value. If there were nested
1787        might be changed, so reset it before looping. */        recursions, md->recursive might be changed, so reset it before looping.
1788          */
1789    
1790        DPRINTF(("Recursing into group %d\n", new_recursive.group_num));        DPRINTF(("Recursing into group %d\n", new_recursive.group_num));
1791        cbegroup = (*callpat >= OP_SBRA);        cbegroup = (*callpat >= OP_SBRA);
# Line 1763  for (;;) Line 1796  for (;;)
1796            md, eptrb, RM6);            md, eptrb, RM6);
1797          memcpy(md->offset_vector, new_recursive.offset_save,          memcpy(md->offset_vector, new_recursive.offset_save,
1798              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1799            md->capture_last = new_recursive.saved_capture_last;
1800          md->recursive = new_recursive.prevrec;          md->recursive = new_recursive.prevrec;
1801          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1802            {            {
# Line 1779  for (;;) Line 1813  for (;;)
1813            goto RECURSION_MATCHED;        /* Exit loop; end processing */            goto RECURSION_MATCHED;        /* Exit loop; end processing */
1814            }            }
1815    
1816          /* 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
1817          as NOMATCH. */          recursion; they cause a NOMATCH for the entire recursion. These codes
1818            are defined in a range that can be tested for. */
1819    
1820            if (rrc >= MATCH_BACKTRACK_MIN && rrc <= MATCH_BACKTRACK_MAX)
1821              RRETURN(MATCH_NOMATCH);
1822    
1823            /* Any return code other than NOMATCH is an error. */
1824    
1825          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)          if (rrc != MATCH_NOMATCH)
1826            {            {
1827            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1828            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
# Line 1912  for (;;) Line 1952  for (;;)
1952    
1953        /* Deal with capturing */        /* Deal with capturing */
1954    
1955        md->capture_last = number;        md->capture_last = (md->capture_last & OVFLMASK) | number;
1956        if (offset >= md->offset_max) md->offset_overflow = TRUE; else        if (offset >= md->offset_max) md->capture_last |= OVFLBIT; else
1957          {          {
1958          /* If offset is greater than offset_top, it means that we are          /* If offset is greater than offset_top, it means that we are
1959          "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 1969  for (;;) Line 2009  for (;;)
2009    
2010      if (*ecode == OP_KETRPOS)      if (*ecode == OP_KETRPOS)
2011        {        {
2012          md->start_match_ptr = mstart;    /* In case \K reset it */
2013        md->end_match_ptr = eptr;        md->end_match_ptr = eptr;
2014        md->end_offset_top = offset_top;        md->end_offset_top = offset_top;
2015        RRETURN(MATCH_KETRPOS);        RRETURN(MATCH_KETRPOS);
# Line 1993  for (;;) Line 2034  for (;;)
2034          }          }
2035        if (*prev >= OP_SBRA)    /* Could match an empty string */        if (*prev >= OP_SBRA)    /* Could match an empty string */
2036          {          {
         md->match_function_type = MATCH_CBEGROUP;  
2037          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);
2038          RRETURN(rrc);          RRETURN(rrc);
2039          }          }
# Line 2002  for (;;) Line 2042  for (;;)
2042        }        }
2043      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
2044        {        {
       if (*prev >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;  
2045        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);
2046        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;
2047        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
# Line 2059  for (;;) Line 2098  for (;;)
2098    
2099      case OP_DOLLM:      case OP_DOLLM:
2100      if (eptr < md->end_subject)      if (eptr < md->end_subject)
2101        { if (!IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH); }        {
2102          if (!IS_NEWLINE(eptr))
2103            {
2104            if (md->partial != 0 &&
2105                eptr + 1 >= md->end_subject &&
2106                NLBLOCK->nltype == NLTYPE_FIXED &&
2107                NLBLOCK->nllen == 2 &&
2108                RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2109              {
2110              md->hitend = TRUE;
2111              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2112              }
2113            RRETURN(MATCH_NOMATCH);
2114            }
2115          }
2116      else      else
2117        {        {
2118        if (md->noteol) RRETURN(MATCH_NOMATCH);        if (md->noteol) RRETURN(MATCH_NOMATCH);
# Line 2091  for (;;) Line 2144  for (;;)
2144      ASSERT_NL_OR_EOS:      ASSERT_NL_OR_EOS:
2145      if (eptr < md->end_subject &&      if (eptr < md->end_subject &&
2146          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))
2147          {
2148          if (md->partial != 0 &&
2149              eptr + 1 >= md->end_subject &&
2150              NLBLOCK->nltype == NLTYPE_FIXED &&
2151              NLBLOCK->nllen == 2 &&
2152              RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2153            {
2154            md->hitend = TRUE;
2155            if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2156            }
2157        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2158          }
2159    
2160      /* Either at end of string or \n before end. */      /* Either at end of string or \n before end. */
2161    
# Line 2219  for (;;) Line 2283  for (;;)
2283        }        }
2284      break;      break;
2285    
2286      /* Match a single character type; inline for speed */      /* Match any single character type except newline; have to take care with
2287        CRLF newlines and partial matching. */
2288    
2289      case OP_ANY:      case OP_ANY:
2290      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
2291        if (md->partial != 0 &&
2292            eptr + 1 >= md->end_subject &&
2293            NLBLOCK->nltype == NLTYPE_FIXED &&
2294            NLBLOCK->nllen == 2 &&
2295            RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2296          {
2297          md->hitend = TRUE;
2298          if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2299          }
2300    
2301      /* Fall through */      /* Fall through */
2302    
2303        /* Match any single character whatsoever. */
2304    
2305      case OP_ALLANY:      case OP_ALLANY:
2306      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 */
2307        {                            /* not be updated before SCHECK_PARTIAL. */        {                            /* not be updated before SCHECK_PARTIAL. */
# Line 2364  for (;;) Line 2441  for (;;)
2441        {        {
2442        default: RRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2443    
2444        case 0x000d:        case CHAR_CR:
2445        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr >= md->end_subject)
2446            {
2447            SCHECK_PARTIAL();
2448            }
2449          else if (RAWUCHARTEST(eptr) == CHAR_LF) eptr++;
2450        break;        break;
2451    
2452        case 0x000a:        case CHAR_LF:
2453        break;        break;
2454    
2455        case 0x000b:        case CHAR_VT:
2456        case 0x000c:        case CHAR_FF:
2457        case 0x0085:        case CHAR_NEL:
2458    #ifndef EBCDIC
2459        case 0x2028:        case 0x2028:
2460        case 0x2029:        case 0x2029:
2461    #endif  /* Not EBCDIC */
2462        if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);        if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
2463        break;        break;
2464        }        }
# Line 2391  for (;;) Line 2474  for (;;)
2474      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2475      switch(c)      switch(c)
2476        {        {
2477          HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
2478        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);  
2479        }        }
2480      ecode++;      ecode++;
2481      break;      break;
# Line 2425  for (;;) Line 2489  for (;;)
2489      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2490      switch(c)      switch(c)
2491        {        {
2492          HSPACE_CASES: break;  /* Byte and multibyte cases */
2493        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;  
2494        }        }
2495      ecode++;      ecode++;
2496      break;      break;
# Line 2459  for (;;) Line 2504  for (;;)
2504      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2505      switch(c)      switch(c)
2506        {        {
2507          VSPACE_CASES: RRETURN(MATCH_NOMATCH);
2508        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);  
2509        }        }
2510      ecode++;      ecode++;
2511      break;      break;
# Line 2481  for (;;) Line 2519  for (;;)
2519      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2520      switch(c)      switch(c)
2521        {        {
2522          VSPACE_CASES: break;
2523        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;  
2524        }        }
2525      ecode++;      ecode++;
2526      break;      break;
# Line 2507  for (;;) Line 2538  for (;;)
2538        }        }
2539      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2540        {        {
2541          const pcre_uint32 *cp;
2542        const ucd_record *prop = GET_UCD(c);        const ucd_record *prop = GET_UCD(c);
2543    
2544        switch(ecode[1])        switch(ecode[1])
# Line 2545  for (;;) Line 2577  for (;;)
2577            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2578          break;          break;
2579    
2580          case PT_SPACE:    /* Perl space */          /* Perl space used to exclude VT, but from Perl 5.18 it is included,
2581          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||          which means that Perl space and POSIX space are now identical. PCRE
2582               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;  
2583    
2584            case PT_SPACE:    /* Perl space */
2585          case PT_PXSPACE:  /* POSIX space */          case PT_PXSPACE:  /* POSIX space */
2586          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||          switch(c)
2587               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||            {
2588               c == CHAR_FF || c == CHAR_CR)            HSPACE_CASES:
2589                 == (op == OP_NOTPROP))            VSPACE_CASES:
2590            RRETURN(MATCH_NOMATCH);            if (op == OP_NOTPROP) RRETURN(MATCH_NOMATCH);
2591              break;
2592    
2593              default:
2594              if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z) ==
2595                (op == OP_NOTPROP)) RRETURN(MATCH_NOMATCH);
2596              break;
2597              }
2598          break;          break;
2599    
2600          case PT_WORD:          case PT_WORD:
# Line 2567  for (;;) Line 2604  for (;;)
2604            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2605          break;          break;
2606    
2607            case PT_CLIST:
2608            cp = PRIV(ucd_caseless_sets) + ecode[2];
2609            for (;;)
2610              {
2611              if (c < *cp)
2612                { if (op == OP_PROP) { RRETURN(MATCH_NOMATCH); } else break; }
2613              if (c == *cp++)
2614                { if (op == OP_PROP) break; else { RRETURN(MATCH_NOMATCH); } }
2615              }
2616            break;
2617    
2618            case PT_UCNC:
2619            if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
2620                 c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
2621                 c >= 0xe000) == (op == OP_NOTPROP))
2622              RRETURN(MATCH_NOMATCH);
2623            break;
2624    
2625          /* This should never occur */          /* This should never occur */
2626    
2627          default:          default:
# Line 2586  for (;;) Line 2641  for (;;)
2641        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2642        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2643        }        }
2644      GETCHARINCTEST(c, eptr);      else
     if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
     while (eptr < md->end_subject)  
2645        {        {
2646        int len = 1;        int lgb, rgb;
2647        if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }        GETCHARINCTEST(c, eptr);
2648        if (UCD_CATEGORY(c) != ucp_M) break;        lgb = UCD_GRAPHBREAK(c);
2649        eptr += len;        while (eptr < md->end_subject)
2650            {
2651            int len = 1;
2652            if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
2653            rgb = UCD_GRAPHBREAK(c);
2654            if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
2655            lgb = rgb;
2656            eptr += len;
2657            }
2658        }        }
2659        CHECK_PARTIAL();
2660      ecode++;      ecode++;
2661      break;      break;
2662  #endif  #endif  /* SUPPORT_UCP */
2663    
2664    
2665      /* Match a back reference, possibly repeatedly. Look past the end of the      /* Match a back reference, possibly repeatedly. Look past the end of the
# Line 2606  for (;;) Line 2668  for (;;)
2668      similar code to character type repeats - written out again for speed.      similar code to character type repeats - written out again for speed.
2669      However, if the referenced string is the empty string, always treat      However, if the referenced string is the empty string, always treat
2670      it as matched, any number of times (otherwise there could be infinite      it as matched, any number of times (otherwise there could be infinite
2671      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:  
2672    
2673      (a) In the default, Perl-compatible state, set the length negative;      (a) In the default, Perl-compatible state, set the length negative;
2674      this ensures that every attempt at a match fails. We can't just fail      this ensures that every attempt at a match fails. We can't just fail
# Line 2624  for (;;) Line 2678  for (;;)
2678      so that the back reference matches an empty string.      so that the back reference matches an empty string.
2679    
2680      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
2681      referenced subpattern. */      referenced subpattern.
2682    
2683        The OP_REF and OP_REFI opcodes are used for a reference to a numbered group
2684        or to a non-duplicated named group. For a duplicated named group, OP_DNREF
2685        and OP_DNREFI are used. In this case we must scan the list of groups to
2686        which the name refers, and use the first one that is set. */
2687    
2688        case OP_DNREF:
2689        case OP_DNREFI:
2690        caseless = op == OP_DNREFI;
2691          {
2692          int count = GET2(ecode, 1+IMM2_SIZE);
2693          pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
2694          ecode += 1 + 2*IMM2_SIZE;
2695    
2696          while (count-- > 0)
2697            {
2698            offset = GET2(slot, 0) << 1;
2699            if (offset < offset_top && md->offset_vector[offset] >= 0) break;
2700            slot += md->name_entry_size;
2701            }
2702          if (count < 0)
2703            length = (md->jscript_compat)? 0 : -1;
2704          else
2705            length = md->offset_vector[offset+1] - md->offset_vector[offset];
2706          }
2707        goto REF_REPEAT;
2708    
2709        case OP_REF:
2710        case OP_REFI:
2711        caseless = op == OP_REFI;
2712        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
2713        ecode += 1 + IMM2_SIZE;
2714      if (offset >= offset_top || md->offset_vector[offset] < 0)      if (offset >= offset_top || md->offset_vector[offset] < 0)
2715        length = (md->jscript_compat)? 0 : -1;        length = (md->jscript_compat)? 0 : -1;
2716      else      else
# Line 2633  for (;;) Line 2718  for (;;)
2718    
2719      /* Set up for repetition, or handle the non-repeated case */      /* Set up for repetition, or handle the non-repeated case */
2720    
2721        REF_REPEAT:
2722      switch (*ecode)      switch (*ecode)
2723        {        {
2724        case OP_CRSTAR:        case OP_CRSTAR:
# Line 2660  for (;;) Line 2746  for (;;)
2746        default:               /* No repeat follows */        default:               /* No repeat follows */
2747        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)
2748          {          {
2749            if (length == -2) eptr = md->end_subject;   /* Partial match */
2750          CHECK_PARTIAL();          CHECK_PARTIAL();
2751          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2752          }          }
# Line 2685  for (;;) Line 2772  for (;;)
2772        int slength;        int slength;
2773        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2774          {          {
2775            if (slength == -2) eptr = md->end_subject;   /* Partial match */
2776          CHECK_PARTIAL();          CHECK_PARTIAL();
2777          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2778          }          }
# Line 2708  for (;;) Line 2796  for (;;)
2796          if (fi >= max) RRETURN(MATCH_NOMATCH);          if (fi >= max) RRETURN(MATCH_NOMATCH);
2797          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2798            {            {
2799              if (slength == -2) eptr = md->end_subject;   /* Partial match */
2800            CHECK_PARTIAL();            CHECK_PARTIAL();
2801            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2802            }            }
# Line 2726  for (;;) Line 2815  for (;;)
2815          int slength;          int slength;
2816          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2817            {            {
2818            CHECK_PARTIAL();            /* Can't use CHECK_PARTIAL because we don't want to update eptr in
2819              the soft partial matching case. */
2820    
2821              if (slength == -2 && md->partial != 0 &&
2822                  md->end_subject > md->start_used_ptr)
2823                {
2824                md->hitend = TRUE;
2825                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2826                }
2827            break;            break;
2828            }            }
2829          eptr += slength;          eptr += slength;
2830          }          }
2831    
2832        while (eptr >= pp)        while (eptr >= pp)
2833          {          {
2834          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);
# Line 2769  for (;;) Line 2867  for (;;)
2867          case OP_CRMINPLUS:          case OP_CRMINPLUS:
2868          case OP_CRQUERY:          case OP_CRQUERY:
2869          case OP_CRMINQUERY:          case OP_CRMINQUERY:
2870            case OP_CRPOSSTAR:
2871            case OP_CRPOSPLUS:
2872            case OP_CRPOSQUERY:
2873          c = *ecode++ - OP_CRSTAR;          c = *ecode++ - OP_CRSTAR;
2874          minimize = (c & 1) != 0;          if (c < OP_CRPOSSTAR - OP_CRSTAR) minimize = (c & 1) != 0;
2875            else possessive = TRUE;
2876          min = rep_min[c];                 /* Pick up values from tables; */          min = rep_min[c];                 /* Pick up values from tables; */
2877          max = rep_max[c];                 /* zero for max => infinity */          max = rep_max[c];                 /* zero for max => infinity */
2878          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 2778  for (;;) Line 2880  for (;;)
2880    
2881          case OP_CRRANGE:          case OP_CRRANGE:
2882          case OP_CRMINRANGE:          case OP_CRMINRANGE:
2883            case OP_CRPOSRANGE:
2884          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
2885            possessive = (*ecode == OP_CRPOSRANGE);
2886          min = GET2(ecode, 1);          min = GET2(ecode, 1);
2887          max = GET2(ecode, 1 + IMM2_SIZE);          max = GET2(ecode, 1 + IMM2_SIZE);
2888          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 2920  for (;;) Line 3024  for (;;)
3024                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
3025              eptr += len;              eptr += len;
3026              }              }
3027    
3028              if (possessive) continue;    /* No backtracking */
3029    
3030            for (;;)            for (;;)
3031              {              {
3032              RMATCH(eptr, ecode, offset_top, md, eptrb, RM18);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM18);
# Line 2950  for (;;) Line 3057  for (;;)
3057                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
3058              eptr++;              eptr++;
3059              }              }
3060    
3061              if (possessive) continue;    /* No backtracking */
3062    
3063            while (eptr >= pp)            while (eptr >= pp)
3064              {              {
3065              RMATCH(eptr, ecode, offset_top, md, eptrb, RM19);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM19);
# Line 2983  for (;;) Line 3093  for (;;)
3093          case OP_CRMINPLUS:          case OP_CRMINPLUS:
3094          case OP_CRQUERY:          case OP_CRQUERY:
3095          case OP_CRMINQUERY:          case OP_CRMINQUERY:
3096            case OP_CRPOSSTAR:
3097            case OP_CRPOSPLUS:
3098            case OP_CRPOSQUERY:
3099          c = *ecode++ - OP_CRSTAR;          c = *ecode++ - OP_CRSTAR;
3100          minimize = (c & 1) != 0;          if (c < OP_CRPOSSTAR - OP_CRSTAR) minimize = (c & 1) != 0;
3101            else possessive = TRUE;
3102          min = rep_min[c];                 /* Pick up values from tables; */          min = rep_min[c];                 /* Pick up values from tables; */
3103          max = rep_max[c];                 /* zero for max => infinity */          max = rep_max[c];                 /* zero for max => infinity */
3104          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 2992  for (;;) Line 3106  for (;;)
3106    
3107          case OP_CRRANGE:          case OP_CRRANGE:
3108          case OP_CRMINRANGE:          case OP_CRMINRANGE:
3109            case OP_CRPOSRANGE:
3110          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
3111            possessive = (*ecode == OP_CRPOSRANGE);
3112          min = GET2(ecode, 1);          min = GET2(ecode, 1);
3113          max = GET2(ecode, 1 + IMM2_SIZE);          max = GET2(ecode, 1 + IMM2_SIZE);
3114          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
# Line 3064  for (;;) Line 3180  for (;;)
3180            if (!PRIV(xclass)(c, data, utf)) break;            if (!PRIV(xclass)(c, data, utf)) break;
3181            eptr += len;            eptr += len;
3182            }            }
3183    
3184            if (possessive) continue;    /* No backtracking */
3185    
3186          for(;;)          for(;;)
3187            {            {
3188            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);
# Line 3094  for (;;) Line 3213  for (;;)
3213          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
3214          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3215          }          }
3216        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);        while (length-- > 0) if (*ecode++ != RAWUCHARINC(eptr)) RRETURN(MATCH_NOMATCH);
3217        }        }
3218      else      else
3219  #endif  #endif
# Line 3134  for (;;) Line 3253  for (;;)
3253    
3254        if (fc < 128)        if (fc < 128)
3255          {          {
3256          if (md->lcc[fc]          pcre_uint32 cc = RAWUCHAR(eptr);
3257              != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);          if (md->lcc[fc] != TABLE_GET(cc, md->lcc, cc)) RRETURN(MATCH_NOMATCH);
3258          ecode++;          ecode++;
3259          eptr++;          eptr++;
3260          }          }
# Line 3146  for (;;) Line 3265  for (;;)
3265    
3266        else        else
3267          {          {
3268          unsigned int dc;          pcre_uint32 dc;
3269          GETCHARINC(dc, eptr);          GETCHARINC(dc, eptr);
3270          ecode += length;          ecode += length;
3271    
# Line 3239  for (;;) Line 3358  for (;;)
3358      max = rep_max[c];                 /* zero for max => infinity */      max = rep_max[c];                 /* zero for max => infinity */
3359      if (max == 0) max = INT_MAX;      if (max == 0) max = INT_MAX;
3360    
3361      /* Common code for all repeated single-character matches. */      /* Common code for all repeated single-character matches. We first check
3362        for the minimum number of characters. If the minimum equals the maximum, we
3363        are done. Otherwise, if minimizing, check the rest of the pattern for a
3364        match; if there isn't one, advance up to the maximum, one character at a
3365        time.
3366    
3367        If maximizing, advance up to the maximum number of matching characters,
3368        until eptr is past the end of the maximum run. If possessive, we are
3369        then done (no backing up). Otherwise, match at this position; anything
3370        other than no match is immediately returned. For nomatch, back up one
3371        character, unless we are matching \R and the last thing matched was
3372        \r\n, in which case, back up two bytes. When we reach the first optional
3373        character position, we can save stack by doing a tail recurse.
3374    
3375        The various UTF/non-UTF and caseful/caseless cases are handled separately,
3376        for speed. */
3377    
3378      REPEATCHAR:      REPEATCHAR:
3379  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
# Line 3256  for (;;) Line 3390  for (;;)
3390        if (length > 1)        if (length > 1)
3391          {          {
3392  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3393          unsigned int othercase;          pcre_uint32 othercase;
3394          if (op >= OP_STARI &&     /* Caseless */          if (op >= OP_STARI &&     /* Caseless */
3395              (othercase = UCD_OTHERCASE(fc)) != fc)              (othercase = UCD_OTHERCASE(fc)) != fc)
3396            oclength = PRIV(ord2utf)(othercase, occhars);            oclength = PRIV(ord2utf)(othercase, occhars);
# Line 3323  for (;;) Line 3457  for (;;)
3457                }                }
3458              }              }
3459    
3460            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
   
3461            for(;;)            for(;;)
3462              {              {
3463                if (eptr == pp) goto TAIL_RECURSE;
3464              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);
3465              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
             if (eptr == pp) { RRETURN(MATCH_NOMATCH); }  
3466  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3467              eptr--;              eptr--;
3468              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 3360  for (;;) Line 3493  for (;;)
3493      maximizing, find the maximum number of characters and work backwards. */      maximizing, find the maximum number of characters and work backwards. */
3494    
3495      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,
3496        max, eptr));        max, (char *)eptr));
3497    
3498      if (op >= OP_STARI)  /* Caseless */      if (op >= OP_STARI)  /* Caseless */
3499        {        {
# Line 3383  for (;;) Line 3516  for (;;)
3516    
3517        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
3518          {          {
3519            pcre_uint32 cc;                 /* Faster than pcre_uchar */
3520          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3521            {            {
3522            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3523            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3524            }            }
3525          if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);          cc = RAWUCHARTEST(eptr);
3526            if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
3527          eptr++;          eptr++;
3528          }          }
3529        if (min == max) continue;        if (min == max) continue;
# Line 3396  for (;;) Line 3531  for (;;)
3531          {          {
3532          for (fi = min;; fi++)          for (fi = min;; fi++)
3533            {            {
3534              pcre_uint32 cc;               /* Faster than pcre_uchar */
3535            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);
3536            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3537            if (fi >= max) RRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
# Line 3404  for (;;) Line 3540  for (;;)
3540              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3541              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3542              }              }
3543            if (fc != *eptr && foc != *eptr) RRETURN(MATCH_NOMATCH);            cc = RAWUCHARTEST(eptr);
3544              if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
3545            eptr++;            eptr++;
3546            }            }
3547          /* Control never gets here */          /* Control never gets here */
# Line 3414  for (;;) Line 3551  for (;;)
3551          pp = eptr;          pp = eptr;
3552          for (i = min; i < max; i++)          for (i = min; i < max; i++)
3553            {            {
3554              pcre_uint32 cc;               /* Faster than pcre_uchar */
3555            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3556              {              {
3557              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3558              break;              break;
3559              }              }
3560            if (fc != *eptr && foc != *eptr) break;            cc = RAWUCHARTEST(eptr);
3561              if (fc != cc && foc != cc) break;
3562            eptr++;            eptr++;
3563            }            }
3564            if (possessive) continue;       /* No backtracking */
3565          if (possessive) continue;          for (;;)
   
         while (eptr >= pp)  
3566            {            {
3567              if (eptr == pp) goto TAIL_RECURSE;
3568            RMATCH(eptr, ecode, offset_top, md, eptrb, RM25);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM25);
3569            eptr--;            eptr--;
3570            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3571            }            }
3572          RRETURN(MATCH_NOMATCH);          /* Control never gets here */
3573          }          }
       /* Control never gets here */  
3574        }        }
3575    
3576      /* Caseful comparisons (includes all multi-byte characters) */      /* Caseful comparisons (includes all multi-byte characters) */
# Line 3447  for (;;) Line 3584  for (;;)
3584            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3585            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3586            }            }
3587          if (fc != *eptr++) RRETURN(MATCH_NOMATCH);          if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
3588          }          }
3589    
3590        if (min == max) continue;        if (min == max) continue;
# Line 3464  for (;;) Line 3601  for (;;)
3601              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3602              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3603              }              }
3604            if (fc != *eptr++) RRETURN(MATCH_NOMATCH);            if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
3605            }            }
3606          /* Control never gets here */          /* Control never gets here */
3607          }          }
# Line 3478  for (;;) Line 3615  for (;;)
3615              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3616              break;              break;
3617              }              }
3618            if (fc != *eptr) break;            if (fc != RAWUCHARTEST(eptr)) break;
3619            eptr++;            eptr++;
3620            }            }
3621          if (possessive) continue;          if (possessive) continue;    /* No backtracking */
3622            for (;;)
         while (eptr >= pp)  
3623            {            {
3624              if (eptr == pp) goto TAIL_RECURSE;
3625            RMATCH(eptr, ecode, offset_top, md, eptrb, RM27);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM27);
3626            eptr--;            eptr--;
3627            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3628            }            }
3629          RRETURN(MATCH_NOMATCH);          /* Control never gets here */
3630          }          }
3631        }        }
3632      /* Control never gets here */      /* Control never gets here */
# Line 3504  for (;;) Line 3641  for (;;)
3641        SCHECK_PARTIAL();        SCHECK_PARTIAL();
3642        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
3643        }        }
     ecode++;  
     GETCHARINCTEST(c, eptr);  
     if (op == OP_NOTI)         /* The caseless case */  
       {  
       register int ch, och;  
       ch = *ecode++;  
 #ifdef COMPILE_PCRE8  
       /* ch must be < 128 if UTF is enabled. */  
       och = md->fcc[ch];  
 #else  
3644  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3645  #ifdef SUPPORT_UCP      if (utf)
3646        if (utf && ch > 127)        {
3647          och = UCD_OTHERCASE(ch);        register pcre_uint32 ch, och;
3648  #else  
3649        if (utf && ch > 127)        ecode++;
3650          och = ch;        GETCHARINC(ch, ecode);
3651  #endif /* SUPPORT_UCP */        GETCHARINC(c, eptr);
3652        else  
3653  #endif /* SUPPORT_UTF */        if (op == OP_NOT)
3654          och = TABLE_GET(ch, md->fcc, ch);          {
3655  #endif /* COMPILE_PCRE8 */          if (ch == c) RRETURN(MATCH_NOMATCH);
3656        if (ch == c || och == c) RRETURN(MATCH_NOMATCH);          }
3657          else
3658            {
3659    #ifdef SUPPORT_UCP
3660            if (ch > 127)
3661              och = UCD_OTHERCASE(ch);
3662    #else
3663            if (ch > 127)
3664              och = ch;
3665    #endif /* SUPPORT_UCP */
3666            else
3667              och = TABLE_GET(ch, md->fcc, ch);
3668            if (ch == c || och == c) RRETURN(MATCH_NOMATCH);
3669            }
3670        }        }
3671      else    /* Caseful */      else
3672    #endif
3673        {        {
3674        if (*ecode++ == c) RRETURN(MATCH_NOMATCH);        register pcre_uint32 ch = ecode[1];
3675          c = *eptr++;
3676          if (ch == c || (op == OP_NOTI && TABLE_GET(ch, md->fcc, ch) == c))
3677            RRETURN(MATCH_NOMATCH);
3678          ecode += 2;
3679        }        }
3680      break;      break;
3681    
# Line 3610  for (;;) Line 3755  for (;;)
3755      /* Common code for all repeated single-byte matches. */      /* Common code for all repeated single-byte matches. */
3756    
3757      REPEATNOTCHAR:      REPEATNOTCHAR:
3758      fc = *ecode++;      GETCHARINCTEST(fc, ecode);
3759    
3760      /* The code is duplicated for the caseless and caseful cases, for speed,      /* The code is duplicated for the caseless and caseful cases, for speed,
3761      since matching characters is likely to be quite common. First, ensure the      since matching characters is likely to be quite common. First, ensure the
# Line 3621  for (;;) Line 3766  for (;;)
3766      characters and work backwards. */      characters and work backwards. */
3767    
3768      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,
3769        max, eptr));        max, (char *)eptr));
3770    
3771      if (op >= OP_NOTSTARI)     /* Caseless */      if (op >= OP_NOTSTARI)     /* Caseless */
3772        {        {
 #ifdef COMPILE_PCRE8  
       /* fc must be < 128 if UTF is enabled. */  
       foc = md->fcc[fc];  
 #else  
3773  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3774  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3775        if (utf && fc > 127)        if (utf && fc > 127)
# Line 3640  for (;;) Line 3781  for (;;)
3781        else        else
3782  #endif /* SUPPORT_UTF */  #endif /* SUPPORT_UTF */
3783          foc = TABLE_GET(fc, md->fcc, fc);          foc = TABLE_GET(fc, md->fcc, fc);
 #endif /* COMPILE_PCRE8 */  
3784    
3785  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3786        if (utf)        if (utf)
3787          {          {
3788          register unsigned int d;          register pcre_uint32 d;
3789          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3790            {            {
3791            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
# Line 3654  for (;;) Line 3794  for (;;)
3794              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3795              }              }
3796            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3797            if (fc == d || foc == d) RRETURN(MATCH_NOMATCH);            if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
3798            }            }
3799          }          }
3800        else        else
3801  #endif  #endif  /* SUPPORT_UTF */
3802        /* Not UTF mode */        /* Not UTF mode */
3803          {          {
3804          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 3680  for (;;) Line 3820  for (;;)
3820  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3821          if (utf)          if (utf)
3822            {            {
3823            register unsigned int d;            register pcre_uint32 d;
3824            for (fi = min;; fi++)            for (fi = min;; fi++)
3825              {              {
3826              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);
# Line 3692  for (;;) Line 3832  for (;;)
3832                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3833                }                }
3834              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3835              if (fc == d || foc == d) RRETURN(MATCH_NOMATCH);              if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
3836              }              }
3837            }            }
3838          else          else
3839  #endif  #endif  /*SUPPORT_UTF */
3840          /* Not UTF mode */          /* Not UTF mode */
3841            {            {
3842            for (fi = min;; fi++)            for (fi = min;; fi++)
# Line 3725  for (;;) Line 3865  for (;;)
3865  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3866          if (utf)          if (utf)
3867            {            {
3868            register unsigned int d;            register pcre_uint32 d;
3869            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3870              {              {
3871              int len = 1;              int len = 1;
# Line 3735  for (;;) Line 3875  for (;;)
3875                break;                break;
3876                }                }
3877              GETCHARLEN(d, eptr, len);              GETCHARLEN(d, eptr, len);
3878              if (fc == d || foc == d) break;              if (fc == d || (unsigned int)foc == d) break;
3879              eptr += len;              eptr += len;
3880              }              }
3881            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
3882            for(;;)            for(;;)
3883              {              {
3884                if (eptr == pp) goto TAIL_RECURSE;
3885              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);
3886              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3887              if (eptr-- == pp) break;        /* Stop if tried at original pos */              eptr--;
3888              BACKCHAR(eptr);              BACKCHAR(eptr);
3889              }              }
3890            }            }
3891          else          else
3892  #endif  #endif  /* SUPPORT_UTF */
3893          /* Not UTF mode */          /* Not UTF mode */
3894            {            {
3895            for (i = min; i < max; i++)            for (i = min; i < max; i++)
# Line 3761  for (;;) Line 3902  for (;;)
3902              if (fc == *eptr || foc == *eptr) break;              if (fc == *eptr || foc == *eptr) break;
3903              eptr++;              eptr++;
3904              }              }
3905            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
3906            while (eptr >= pp)            for (;;)
3907              {              {
3908                if (eptr == pp) goto TAIL_RECURSE;
3909              RMATCH(eptr, ecode, offset_top, md, eptrb, RM31);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM31);
3910              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3911              eptr--;              eptr--;
3912              }              }
3913            }            }
3914            /* Control never gets here */
         RRETURN(MATCH_NOMATCH);  
3915          }          }
       /* Control never gets here */  
3916        }        }
3917    
3918      /* Caseful comparisons */      /* Caseful comparisons */
# Line 3782  for (;;) Line 3922  for (;;)
3922  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3923        if (utf)        if (utf)
3924          {          {
3925          register unsigned int d;          register pcre_uint32 d;
3926          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3927            {            {
3928            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
# Line 3816  for (;;) Line 3956  for (;;)
3956  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3957          if (utf)          if (utf)
3958            {            {
3959            register unsigned int d;            register pcre_uint32 d;
3960            for (fi = min;; fi++)            for (fi = min;; fi++)
3961              {              {
3962              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);
# Line 3860  for (;;) Line 4000  for (;;)
4000  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
4001          if (utf)          if (utf)
4002            {            {
4003            register unsigned int d;            register pcre_uint32 d;
4004            for (i = min; i < max; i++)            for (i = min; i < max; i++)
4005              {              {
4006              int len = 1;              int len = 1;
# Line 3873  for (;;) Line 4013  for (;;)
4013              if (fc == d) break;              if (fc == d) break;
4014              eptr += len;              eptr += len;
4015              }              }
4016            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
4017            for(;;)            for(;;)
4018              {              {
4019                if (eptr == pp) goto TAIL_RECURSE;
4020              RMATCH(eptr, ecode, offset_top, md, eptrb, RM34);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM34);
4021              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4022              if (eptr-- == pp) break;        /* Stop if tried at original pos */              eptr--;
4023              BACKCHAR(eptr);              BACKCHAR(eptr);
4024              }              }
4025            }            }
# Line 3896  for (;;) Line 4037  for (;;)
4037              if (fc == *eptr) break;              if (fc == *eptr) break;
4038              eptr++;              eptr++;
4039              }              }
4040            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
4041            while (eptr >= pp)            for (;;)
4042              {              {
4043                if (eptr == pp) goto TAIL_RECURSE;
4044              RMATCH(eptr, ecode, offset_top, md, eptrb, RM35);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM35);
4045              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4046              eptr--;              eptr--;
4047              }              }
4048            }            }
4049            /* Control never gets here */
         RRETURN(MATCH_NOMATCH);  
4050          }          }
4051        }        }
4052      /* Control never gets here */      /* Control never gets here */
# Line 4087  for (;;) Line 4228  for (;;)
4228              }              }
4229            break;            break;
4230    
4231              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
4232              which means that Perl space and POSIX space are now identical. PCRE
4233              was changed at release 8.34. */
4234    
4235            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
4236              case PT_PXSPACE:  /* POSIX space */
4237            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4238              {              {
4239              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4096  for (;;) Line 4242  for (;;)
4242                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4243                }                }
4244              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4245              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              switch(c)
4246                   c == CHAR_FF || c == CHAR_CR)                {
4247                     == prop_fail_result)                HSPACE_CASES:
4248                RRETURN(MATCH_NOMATCH);                VSPACE_CASES:
4249                  if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4250                  break;
4251    
4252                  default:
4253                  if ((UCD_CATEGORY(c) == ucp_Z) == prop_fail_result)
4254                    RRETURN(MATCH_NOMATCH);
4255                  break;
4256                  }
4257              }              }
4258            break;            break;
4259    
4260            case PT_PXSPACE:  /* POSIX space */            case PT_WORD:
4261            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4262              {              {
4263                int category;
4264              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4265                {                {
4266                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4267                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4268                }                }
4269              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4270              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              category = UCD_CATEGORY(c);
4271                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)
4272                     == prop_fail_result)                     == prop_fail_result)
4273                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4274              }              }
4275            break;            break;
4276    
4277            case PT_WORD:            case PT_CLIST:
4278            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4279              {              {
4280              int category;              const pcre_uint32 *cp;
4281              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4282                {                {
4283                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4284                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4285                }                }
4286              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4287              category = UCD_CATEGORY(c);              cp = PRIV(ucd_caseless_sets) + prop_value;
4288              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)              for (;;)
4289                     == prop_fail_result)                {
4290                  if (c < *cp)
4291                    { if (prop_fail_result) break; else { RRETURN(MATCH_NOMATCH); } }
4292                  if (c == *cp++)
4293                    { if (prop_fail_result) { RRETURN(MATCH_NOMATCH); } else break; }
4294                  }
4295                }
4296              break;
4297    
4298              case PT_UCNC:
4299              for (i = 1; i <= min; i++)
4300                {
4301                if (eptr >= md->end_subject)
4302                  {
4303                  SCHECK_PARTIAL();
4304                  RRETURN(MATCH_NOMATCH);
4305                  }
4306                GETCHARINCTEST(c, eptr);
4307                if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
4308                     c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
4309                     c >= 0xe000) == prop_fail_result)
4310                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4311              }              }
4312            break;            break;
# Line 4155  for (;;) Line 4330  for (;;)
4330              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4331              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4332              }              }
4333            GETCHARINCTEST(c, eptr);            else
           if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
           while (eptr < md->end_subject)  
4334              {              {
4335              int len = 1;              int lgb, rgb;
4336              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
4337              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
4338              eptr += len;             while (eptr < md->end_subject)
4339                  {
4340                  int len = 1;
4341                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
4342                  rgb = UCD_GRAPHBREAK(c);
4343                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
4344                  lgb = rgb;
4345                  eptr += len;
4346                  }
4347              }              }
4348              CHECK_PARTIAL();
4349            }            }
4350          }          }
4351    
# Line 4184  for (;;) Line 4366  for (;;)
4366              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4367              }              }
4368            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4369              if (md->partial != 0 &&
4370                  eptr + 1 >= md->end_subject &&
4371                  NLBLOCK->nltype == NLTYPE_FIXED &&
4372                  NLBLOCK->nllen == 2 &&
4373                  RAWUCHAR(eptr) == NLBLOCK->nl[0])
4374                {
4375                md->hitend = TRUE;
4376                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4377                }
4378            eptr++;            eptr++;
4379            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4380            }            }
# Line 4220  for (;;) Line 4411  for (;;)
4411              {              {
4412              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4413    
4414              case 0x000d:              case CHAR_CR:
4415              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && RAWUCHAR(eptr) == CHAR_LF) eptr++;
4416              break;              break;
4417    
4418              case 0x000a:              case CHAR_LF:
4419              break;              break;
4420    
4421              case 0x000b:              case CHAR_VT:
4422              case 0x000c:              case CHAR_FF:
4423              case 0x0085:              case CHAR_NEL:
4424    #ifndef EBCDIC
4425              case 0x2028:              case 0x2028:
4426              case 0x2029:              case 0x2029:
4427    #endif  /* Not EBCDIC */
4428              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4429              break;              break;
4430              }              }
# Line 4249  for (;;) Line 4442  for (;;)
4442            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4443            switch(c)            switch(c)
4444              {              {
4445                HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
4446              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);  
4447              }              }
4448            }            }
4449          break;          break;
# Line 4285  for (;;) Line 4459  for (;;)
4459            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4460            switch(c)            switch(c)
4461              {              {
4462                HSPACE_CASES: break;  /* Byte and multibyte cases */
4463              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;  
4464              }              }
4465            }            }
4466          break;          break;
# Line 4321  for (;;) Line 4476  for (;;)
4476            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4477            switch(c)            switch(c)
4478              {              {
4479                VSPACE_CASES: RRETURN(MATCH_NOMATCH);
4480              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);  
4481              }              }
4482            }            }
4483          break;          break;
# Line 4345  for (;;) Line 4493  for (;;)
4493            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4494            switch(c)            switch(c)
4495              {              {
4496                VSPACE_CASES: break;
4497              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;  
4498              }              }
4499            }            }
4500          break;          break;
# Line 4375  for (;;) Line 4516  for (;;)
4516          case OP_DIGIT:          case OP_DIGIT:
4517          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4518            {            {
4519              pcre_uint32 cc;
4520            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4521              {              {
4522              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4523              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4524              }              }
4525            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_digit) == 0)            cc = RAWUCHAR(eptr);
4526              if (cc >= 128 || (md->ctypes[cc] & ctype_digit) == 0)
4527              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4528            eptr++;            eptr++;
4529            /* No need to skip more bytes - we know it's a 1-byte character */            /* No need to skip more bytes - we know it's a 1-byte character */
# Line 4390  for (;;) Line 4533  for (;;)
4533          case OP_NOT_WHITESPACE:          case OP_NOT_WHITESPACE:
4534          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4535            {            {
4536              pcre_uint32 cc;
4537            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4538              {              {
4539              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4540              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4541              }              }
4542            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)            cc = RAWUCHAR(eptr);
4543              if (cc < 128 && (md->ctypes[cc] & ctype_space) != 0)
4544              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4545            eptr++;            eptr++;
4546            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
# Line 4405  for (;;) Line 4550  for (;;)
4550          case OP_WHITESPACE:          case OP_WHITESPACE:
4551          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4552            {            {
4553              pcre_uint32 cc;
4554            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4555              {              {
4556              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4557              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4558              }              }
4559            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_space) == 0)            cc = RAWUCHAR(eptr);
4560              if (cc >= 128 || (md->ctypes[cc] & ctype_space) == 0)
4561              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4562            eptr++;            eptr++;
4563            /* No need to skip more bytes - we know it's a 1-byte character */            /* No need to skip more bytes - we know it's a 1-byte character */
# Line 4420  for (;;) Line 4567  for (;;)
4567          case OP_NOT_WORDCHAR:          case OP_NOT_WORDCHAR:
4568          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4569            {            {
4570              pcre_uint32 cc;
4571            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4572              {              {
4573              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4574              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4575              }              }
4576            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)            cc = RAWUCHAR(eptr);
4577              if (cc < 128 && (md->ctypes[cc] & ctype_word) != 0)
4578              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4579            eptr++;            eptr++;
4580            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
# Line 4435  for (;;) Line 4584  for (;;)
4584          case OP_WORDCHAR:          case OP_WORDCHAR:
4585          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4586            {            {
4587              pcre_uint32 cc;
4588            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4589              {              {
4590              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4591              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4592              }              }
4593            if (*eptr >= 128 || (md->ctypes[*eptr] & ctype_word) == 0)            cc = RAWUCHAR(eptr);
4594              if (cc >= 128 || (md->ctypes[cc] & ctype_word) == 0)
4595              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4596            eptr++;            eptr++;
4597            /* No need to skip more bytes - we know it's a 1-byte character */            /* No need to skip more bytes - we know it's a 1-byte character */
# Line 4468  for (;;) Line 4619  for (;;)
4619              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4620              }              }
4621            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4622              if (md->partial != 0 &&
4623                  eptr + 1 >= md->end_subject &&
4624                  NLBLOCK->nltype == NLTYPE_FIXED &&
4625                  NLBLOCK->nllen == 2 &&
4626                  *eptr == NLBLOCK->nl[0])
4627                {
4628                md->hitend = TRUE;
4629                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4630                }
4631            eptr++;            eptr++;
4632            }            }
4633          break;          break;
# Line 4502  for (;;) Line 4662  for (;;)
4662              {              {
4663              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4664    
4665              case 0x000d:              case CHAR_CR:
4666              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == CHAR_LF) eptr++;
4667              break;              break;
4668    
4669              case 0x000a:              case CHAR_LF:
4670              break;              break;
4671    
4672              case 0x000b:              case CHAR_VT:
4673              case 0x000c:              case CHAR_FF:
4674              case 0x0085:              case CHAR_NEL:
4675  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4676              case 0x2028:              case 0x2028:
4677              case 0x2029:              case 0x2029:
4678  #endif  #endif
# Line 4533  for (;;) Line 4693  for (;;)
4693            switch(*eptr++)            switch(*eptr++)
4694              {              {
4695              default: break;              default: break;
4696              case 0x09:      /* HT */              HSPACE_BYTE_CASES:
4697              case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4698              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 */  
4699  #endif  #endif
4700              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4701              }              }
# Line 4570  for (;;) Line 4713  for (;;)
4713            switch(*eptr++)            switch(*eptr++)
4714              {              {
4715              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4716              case 0x09:      /* HT */              HSPACE_BYTE_CASES:
4717              case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4718              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 */  
4719  #endif  #endif
4720              break;              break;
4721              }              }
# Line 4606  for (;;) Line 4732  for (;;)
4732              }              }
4733            switch(*eptr++)            switch(*eptr++)
4734              {              {
4735              default: break;              VSPACE_BYTE_CASES:
4736              case 0x0a:      /* LF */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4737              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 */  
4738  #endif  #endif
4739              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4740                default: break;
4741              }              }
4742            }            }
4743          break;          break;
# Line 4632  for (;;) Line 4753  for (;;)
4753            switch(*eptr++)            switch(*eptr++)
4754              {              {
4755              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4756              case 0x0a:      /* LF */              VSPACE_BYTE_CASES:
4757              case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4758              case 0x0c:      /* FF */              VSPACE_MULTIBYTE_CASES:
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
 #ifdef COMPILE_PCRE16  
             case 0x2028:    /* LINE SEPARATOR */  
             case 0x2029:    /* PARAGRAPH SEPARATOR */  
4759  #endif  #endif
4760              break;              break;
4761              }              }
# Line 4857  for (;;) Line 4973  for (;;)
4973              }              }
4974            /* Control never gets here */            /* Control never gets here */
4975    
4976              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
4977              which means that Perl space and POSIX space are now identical. PCRE
4978              was changed at release 8.34. */
4979    
4980            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
4981              case PT_PXSPACE:  /* POSIX space */
4982            for (fi = min;; fi++)            for (fi = min;; fi++)
4983              {              {
4984              RMATCH(eptr, ecode, offset_top, md, eptrb, RM60);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM61);
4985              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4986              if (fi >= max) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4987              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4869  for (;;) Line 4990  for (;;)
4990                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4991                }                }
4992              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4993              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              switch(c)
4994                   c == CHAR_FF || c == CHAR_CR)                {
4995                     == prop_fail_result)                HSPACE_CASES:
4996                RRETURN(MATCH_NOMATCH);                VSPACE_CASES:
4997                  if (prop_fail_result) RRETURN(MATCH_NOMATCH);
4998                  break;
4999    
5000                  default:
5001                  if ((UCD_CATEGORY(c) == ucp_Z) == prop_fail_result)
5002                    RRETURN(MATCH_NOMATCH);
5003                  break;
5004                  }
5005              }              }
5006            /* Control never gets here */            /* Control never gets here */
5007    
5008            case PT_PXSPACE:  /* POSIX space */            case PT_WORD:
5009            for (fi = min;; fi++)            for (fi = min;; fi++)
5010              {              {
5011              RMATCH(eptr, ecode, offset_top, md, eptrb, RM61);              int category;
5012                RMATCH(eptr, ecode, offset_top, md, eptrb, RM62);
5013              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5014              if (fi >= max) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
5015              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4888  for (;;) Line 5018  for (;;)
5018                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5019                }                }
5020              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
5021              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              category = UCD_CATEGORY(c);
5022                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)              if ((category == ucp_L ||
5023                     category == ucp_N ||
5024                     c == CHAR_UNDERSCORE)
5025                     == prop_fail_result)                     == prop_fail_result)
5026                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5027              }              }
5028            /* Control never gets here */            /* Control never gets here */
5029    
5030            case PT_WORD:            case PT_CLIST:
5031            for (fi = min;; fi++)            for (fi = min;; fi++)
5032              {              {
5033              int category;              const pcre_uint32 *cp;
5034              RMATCH(eptr, ecode, offset_top, md, eptrb, RM62);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM67);
5035              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5036              if (fi >= max) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
5037              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 4908  for (;;) Line 5040  for (;;)
5040                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5041                }                }
5042              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
5043              category = UCD_CATEGORY(c);              cp = PRIV(ucd_caseless_sets) + prop_value;
5044              if ((category == ucp_L ||              for (;;)
5045                   category == ucp_N ||                {
5046                   c == CHAR_UNDERSCORE)                if (c < *cp)
5047                     == prop_fail_result)                  { if (prop_fail_result) break; else { RRETURN(MATCH_NOMATCH); } }
5048                  if (c == *cp++)
5049                    { if (prop_fail_result) { RRETURN(MATCH_NOMATCH); } else break; }
5050                  }
5051                }
5052              /* Control never gets here */
5053    
5054              case PT_UCNC:
5055              for (fi = min;; fi++)
5056                {
5057                RMATCH(eptr, ecode, offset_top, md, eptrb, RM68);
5058                if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5059                if (fi >= max) RRETURN(MATCH_NOMATCH);
5060                if (eptr >= md->end_subject)
5061                  {
5062                  SCHECK_PARTIAL();
5063                  RRETURN(MATCH_NOMATCH);
5064                  }
5065                GETCHARINCTEST(c, eptr);
5066                if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
5067                     c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
5068                     c >= 0xe000) == prop_fail_result)
5069                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5070              }              }
5071            /* Control never gets here */            /* Control never gets here */
5072    
5073            /* This should never occur */            /* This should never occur */
   
5074            default:            default:
5075            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
5076            }            }
# Line 4939  for (;;) Line 5091  for (;;)
5091              SCHECK_PARTIAL();              SCHECK_PARTIAL();
5092              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
5093              }              }
5094            GETCHARINCTEST(c, eptr);            else
           if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
           while (eptr < md->end_subject)  
5095              {              {
5096              int len = 1;              int lgb, rgb;
5097              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
5098              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
5099              eptr += len;              while (eptr < md->end_subject)
5100                  {
5101                  int len = 1;
5102                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
5103                  rgb = UCD_GRAPHBREAK(c);
5104                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
5105                  lgb = rgb;
5106                  eptr += len;
5107                  }
5108              }              }
5109              CHECK_PARTIAL();
5110            }            }
5111          }          }
5112        else        else
# Line 4971  for (;;) Line 5130  for (;;)
5130            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
5131            switch(ctype)            switch(ctype)
5132              {              {
5133              case OP_ANY:        /* This is the non-NL case */              case OP_ANY:               /* This is the non-NL case */
5134                if (md->partial != 0 &&    /* Take care with CRLF partial */
5135                    eptr >= md->end_subject &&
5136                    NLBLOCK->nltype == NLTYPE_FIXED &&
5137                    NLBLOCK->nllen == 2 &&
5138                    c == NLBLOCK->nl[0])
5139                  {
5140                  md->hitend = TRUE;
5141                  if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5142                  }
5143                break;
5144    
5145              case OP_ALLANY:              case OP_ALLANY:
5146              case OP_ANYBYTE:              case OP_ANYBYTE:
5147              break;              break;
# Line 4980  for (;;) Line 5150  for (;;)
5150              switch(c)              switch(c)
5151                {                {
5152                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5153                case 0x000d:                case CHAR_CR:
5154                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && RAWUCHAR(eptr) == CHAR_LF) eptr++;
5155                break;                break;
5156                case 0x000a:  
5157                  case CHAR_LF:
5158                break;                break;
5159    
5160                case 0x000b:                case CHAR_VT:
5161                case 0x000c:                case CHAR_FF:
5162                case 0x0085:                case CHAR_NEL:
5163    #ifndef EBCDIC
5164                case 0x2028:                case 0x2028:
5165                case 0x2029:                case 0x2029:
5166    #endif  /* Not EBCDIC */
5167                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
5168                break;                break;
5169                }                }
# Line 4999  for (;;) Line 5172  for (;;)
5172              case OP_NOT_HSPACE:              case OP_NOT_HSPACE:
5173              switch(c)              switch(c)
5174                {                {
5175                  HSPACE_CASES: RRETURN(MATCH_NOMATCH);
5176                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);  
5177                }                }
5178              break;              break;
5179    
5180              case OP_HSPACE:              case OP_HSPACE:
5181              switch(c)              switch(c)
5182                {                {
5183                  HSPACE_CASES: break;
5184                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;  
5185                }                }
5186              break;              break;
5187    
5188              case OP_NOT_VSPACE:              case OP_NOT_VSPACE:
5189              switch(c)              switch(c)
5190                {                {
5191                  VSPACE_CASES: RRETURN(MATCH_NOMATCH);
5192                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);  
5193                }                }
5194              break;              break;
5195    
5196              case OP_VSPACE:              case OP_VSPACE:
5197              switch(c)              switch(c)
5198                {                {
5199                  VSPACE_CASES: break;
5200                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;  
5201                }                }
5202              break;              break;
5203    
# Line 5134  for (;;) Line 5255  for (;;)
5255            c = *eptr++;            c = *eptr++;
5256            switch(ctype)            switch(ctype)
5257              {              {
5258              case OP_ANY:     /* This is the non-NL case */              case OP_ANY:               /* This is the non-NL case */
5259                if (md->partial != 0 &&    /* Take care with CRLF partial */
5260                    eptr >= md->end_subject &&
5261                    NLBLOCK->nltype == NLTYPE_FIXED &&
5262                    NLBLOCK->nllen == 2 &&
5263                    c == NLBLOCK->nl[0])
5264                  {
5265                  md->hitend = TRUE;
5266                  if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5267                  }
5268                break;
5269    
5270              case OP_ALLANY:              case OP_ALLANY:
5271              case OP_ANYBYTE:              case OP_ANYBYTE:
5272              break;              break;
# Line 5143  for (;;) Line 5275  for (;;)
5275              switch(c)              switch(c)
5276                {                {
5277                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5278                case 0x000d:                case CHAR_CR:
5279                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == CHAR_LF) eptr++;
5280                break;                break;
5281    
5282                case 0x000a:                case CHAR_LF:
5283                break;                break;
5284    
5285                case 0x000b:                case CHAR_VT:
5286                case 0x000c:                case CHAR_FF:
5287                case 0x0085:                case CHAR_NEL:
5288  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5289                case 0x2028:                case 0x2028:
5290                case 0x2029:                case 0x2029:
5291  #endif  #endif
# Line 5166  for (;;) Line 5298  for (;;)
5298              switch(c)              switch(c)
5299                {                {
5300                default: break;                default: break;
5301                case 0x09:      /* HT */                HSPACE_BYTE_CASES:
5302                case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5303                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 */  
5304  #endif  #endif
5305                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5306                }                }
# Line 5195  for (;;) Line 5310  for (;;)
5310              switch(c)              switch(c)
5311                {                {
5312                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5313                case 0x09:      /* HT */                HSPACE_BYTE_CASES:
5314                case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5315                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 */  
5316  #endif  #endif
5317                break;                break;
5318                }                }
# Line 5224  for (;;) Line 5322  for (;;)
5322              switch(c)              switch(c)
5323                {                {
5324                default: break;                default: break;
5325                case 0x0a:      /* LF */                VSPACE_BYTE_CASES:
5326                case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5327                case 0x0c:      /* FF */                VSPACE_MULTIBYTE_CASES:
               case 0x0d:      /* CR */  
               case 0x85:      /* NEL */  
 #ifdef COMPILE_PCRE16  
               case 0x2028:    /* LINE SEPARATOR */  
               case 0x2029:    /* PARAGRAPH SEPARATOR */  
5328  #endif  #endif
5329                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
5330                }                }
# Line 5241  for (;;) Line 5334  for (;;)
5334              switch(c)              switch(c)
5335                {                {
5336                default: RRETURN(MATCH_NOMATCH);                default: RRETURN(MATCH_NOMATCH);
5337                case 0x0a:      /* LF */                VSPACE_BYTE_CASES:
5338                case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
5339                case 0x0c:      /* FF */                VSPACE_MULTIBYTE_CASES:
               case 0x0d:      /* CR */  
               case 0x85:      /* NEL */  
 #ifdef COMPILE_PCRE16  
               case 0x2028:    /* LINE SEPARATOR */  
               case 0x2029:    /* PARAGRAPH SEPARATOR */  
5340  #endif  #endif
5341                break;                break;
5342                }                }
# Line 5397  for (;;) Line 5485  for (;;)
5485              }              }
5486            break;            break;
5487    
5488              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
5489              which means that Perl space and POSIX space are now identical. PCRE
5490              was changed at release 8.34. */
5491    
5492            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
5493              case PT_PXSPACE:  /* POSIX space */
5494            for (i = min; i < max; i++)            for (i = min; i < max; i++)
5495              {              {
5496              int len = 1;              int len = 1;
# Line 5407  for (;;) Line 5500  for (;;)
5500                break;                break;
5501                }                }
5502              GETCHARLENTEST(c, eptr, len);              GETCHARLENTEST(c, eptr, len);
5503              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              switch(c)
5504                   c == CHAR_FF || c == CHAR_CR)                {
5505                   == prop_fail_result)                HSPACE_CASES:
5506                  VSPACE_CASES:
5507                  if (prop_fail_result) goto ENDLOOP99;  /* Break the loop */
5508                break;                break;
5509    
5510                  default:
5511                  if ((UCD_CATEGORY(c) == ucp_Z) == prop_fail_result)
5512                    goto ENDLOOP99;   /* Break the loop */
5513                  break;
5514                  }
5515              eptr+= len;              eptr+= len;
5516              }              }
5517              ENDLOOP99:
5518            break;            break;
5519    
5520            case PT_PXSPACE:  /* POSIX space */            case PT_WORD:
5521            for (i = min; i < max; i++)            for (i = min; i < max; i++)
5522              {              {
5523                int category;
5524              int len = 1;              int len = 1;
5525              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
5526                {                {
# Line 5425  for (;;) Line 5528  for (;;)
5528                break;                break;
5529                }                }
5530              GETCHARLENTEST(c, eptr, len);              GETCHARLENTEST(c, eptr, len);
5531              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              category = UCD_CATEGORY(c);
5532                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)              if ((category == ucp_L || category == ucp_N ||
5533                   == prop_fail_result)                   c == CHAR_UNDERSCORE) == prop_fail_result)
5534                break;                break;
5535              eptr+= len;              eptr+= len;
5536              }              }
5537            break;            break;
5538    
5539            case PT_WORD:            case PT_CLIST:
5540            for (i = min; i < max; i++)            for (i = min; i < max; i++)
5541              {              {
5542              int category;              const pcre_uint32 *cp;
5543              int len = 1;              int len = 1;
5544              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
5545                {                {
# Line 5444  for (;;) Line 5547  for (;;)
5547                break;                break;
5548                }                }
5549              GETCHARLENTEST(c, eptr, len);              GETCHARLENTEST(c, eptr, len);
5550              category = UCD_CATEGORY(c);              cp = PRIV(ucd_caseless_sets) + prop_value;
5551              if ((category == ucp_L || category == ucp_N ||              for (;;)
5552                   c == CHAR_UNDERSCORE) == prop_fail_result)                {
5553                  if (c < *cp)
5554                    { if (prop_fail_result) break; else goto GOT_MAX; }
5555                  if (c == *cp++)
5556                    { if (prop_fail_result) goto GOT_MAX; else break; }
5557                  }
5558                eptr += len;
5559                }
5560              GOT_MAX:
5561              break;
5562    
5563              case PT_UCNC:
5564              for (i = min; i < max; i++)
5565                {
5566                int len = 1;
5567                if (eptr >= md->end_subject)
5568                  {
5569                  SCHECK_PARTIAL();
5570                break;                break;
5571              eptr+= len;                }
5572                GETCHARLENTEST(c, eptr, len);
5573                if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
5574                     c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
5575                     c >= 0xe000) == prop_fail_result)
5576                  break;
5577                eptr += len;
5578              }              }
5579            break;            break;
5580    
# Line 5458  for (;;) Line 5584  for (;;)
5584    
5585          /* eptr is now past the end of the maximum run */          /* eptr is now past the end of the maximum run */
5586    
5587          if (possessive) continue;          if (possessive) continue;    /* No backtracking */
5588          for(;;)          for(;;)
5589            {            {
5590              if (eptr == pp) goto TAIL_RECURSE;
5591            RMATCH(eptr, ecode, offset_top, md, eptrb, RM44);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM44);
5592            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5593            if (eptr-- == pp) break;        /* Stop if tried at original pos */            eptr--;
5594            if (utf) BACKCHAR(eptr);            if (utf) BACKCHAR(eptr);
5595            }            }
5596          }          }
5597    
5598        /* Match extended Unicode sequences. We will get here only if the        /* Match extended Unicode grapheme clusters. We will get here only if the
5599        support is in the binary; otherwise a compile-time error occurs. */        support is in the binary; otherwise a compile-time error occurs. */
5600    
5601        else if (ctype == OP_EXTUNI)        else if (ctype == OP_EXTUNI)
5602          {          {
5603          for (i = min; i < max; i++)          for (i = min; i < max; i++)
5604            {            {
           int len = 1;  
5605            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
5606              {              {
5607              SCHECK_PARTIAL();              SCHECK_PARTIAL();
5608              break;              break;
5609              }              }
5610            if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }            else
           if (UCD_CATEGORY(c) == ucp_M) break;  
           eptr += len;  
           while (eptr < md->end_subject)  
5611              {              {
5612              len = 1;              int lgb, rgb;
5613              if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
5614              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
5615              eptr += len;              while (eptr < md->end_subject)
5616                  {
5617                  int len = 1;
5618                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
5619                  rgb = UCD_GRAPHBREAK(c);
5620                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
5621                  lgb = rgb;
5622                  eptr += len;
5623                  }
5624              }              }
5625              CHECK_PARTIAL();
5626            }            }
5627    
5628          /* eptr is now past the end of the maximum run */          /* eptr is now past the end of the maximum run */
5629    
5630          if (possessive) continue;          if (possessive) continue;    /* No backtracking */
5631    
5632          for(;;)          for(;;)
5633            {            {
5634              int lgb, rgb;
5635              PCRE_PUCHAR fptr;
5636    
5637              if (eptr == pp) goto TAIL_RECURSE;   /* At start of char run */
5638            RMATCH(eptr, ecode, offset_top, md, eptrb, RM45);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM45);
5639            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5640            if (eptr-- == pp) break;        /* Stop if tried at original pos */  
5641            for (;;)                        /* Move back over one extended */            /* Backtracking over an extended grapheme cluster involves inspecting
5642              the previous two characters (if present) to see if a break is
5643              permitted between them. */
5644    
5645              eptr--;
5646              if (!utf) c = *eptr; else
5647              {              {
5648              if (!utf) c = *eptr; else              BACKCHAR(eptr);
5649                {              GETCHAR(c, eptr);
5650                BACKCHAR(eptr);              }
5651                GETCHAR(c, eptr);            rgb = UCD_GRAPHBREAK(c);
5652                }  
5653              if (UCD_CATEGORY(c) != ucp_M) break;            for (;;)
5654              eptr--;              {
5655                if (eptr == pp) goto TAIL_RECURSE;   /* At start of char run */
5656                fptr = eptr - 1;
5657                if (!utf) c = *fptr; else
5658                  {
5659                  BACKCHAR(fptr);
5660                  GETCHAR(c, fptr);
5661                  }
5662                lgb = UCD_GRAPHBREAK(c);
5663                if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
5664                eptr = fptr;
5665                rgb = lgb;
5666              }              }
5667            }            }
5668          }          }
# Line 5534  for (;;) Line 5686  for (;;)
5686                  break;                  break;
5687                  }                  }
5688                if (IS_NEWLINE(eptr)) break;                if (IS_NEWLINE(eptr)) break;
5689                  if (md->partial != 0 &&    /* Take care with CRLF partial */
5690                      eptr + 1 >= md->end_subject &&
5691                      NLBLOCK->nltype == NLTYPE_FIXED &&
5692                      NLBLOCK->nllen == 2 &&
5693                      RAWUCHAR(eptr) == NLBLOCK->nl[0])
5694                    {
5695                    md->hitend = TRUE;
5696                    if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5697                    }
5698                eptr++;                eptr++;
5699                ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);                ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
5700                }                }
# Line 5551  for (;;) Line 5712  for (;;)
5712                  break;                  break;
5713                  }                  }
5714                if (IS_NEWLINE(eptr)) break;                if (IS_NEWLINE(eptr)) break;
5715                  if (md->partial != 0 &&    /* Take care with CRLF partial */
5716                      eptr + 1 >= md->end_subject &&
5717                      NLBLOCK->nltype == NLTYPE_FIXED &&
5718                      NLBLOCK->nllen == 2 &&
5719                      RAWUCHAR(eptr) == NLBLOCK->nl[0])
5720                    {
5721                    md->hitend = TRUE;
5722                    if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5723                    }
5724                eptr++;                eptr++;
5725                ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);                ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
5726                }                }
# Line 5600  for (;;) Line 5770  for (;;)
5770                break;                break;
5771                }                }
5772              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
5773              if (c == 0x000d)              if (c == CHAR_CR)
5774                {                {
5775                if (++eptr >= md->end_subject) break;                if (++eptr >= md->end_subject) break;
5776                if (*eptr == 0x000a) eptr++;                if (RAWUCHAR(eptr) == CHAR_LF) eptr++;
5777                }                }
5778              else              else
5779                {                {
5780                if (c != 0x000a &&                if (c != CHAR_LF &&
5781                    (md->bsr_anycrlf ||                    (md->bsr_anycrlf ||
5782                     (c != 0x000b && c != 0x000c &&                     (c != CHAR_VT && c != CHAR_FF && c != CHAR_NEL
5783                      c != 0x0085 && c != 0x2028 && c != 0x2029)))  #ifndef EBCDIC
5784                        && c != 0x2028 && c != 0x2029
5785    #endif  /* Not EBCDIC */
5786                        )))
5787                  break;                  break;
5788                eptr += len;                eptr += len;
5789                }                }
# Line 5631  for (;;) Line 5804  for (;;)
5804              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
5805              switch(c)              switch(c)
5806                {                {
5807                  HSPACE_CASES: gotspace = TRUE; break;
5808                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;  
5809                }                }
5810              if (gotspace == (ctype == OP_NOT_HSPACE)) break;              if (gotspace == (ctype == OP_NOT_HSPACE)) break;
5811              eptr += len;              eptr += len;
# Line 5673  for (;;) Line 5826  for (;;)
5826              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
5827              switch(c)              switch(c)
5828                {                {
5829                  VSPACE_CASES: gotspace = TRUE; break;
5830                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;  
5831                }                }
5832              if (gotspace == (ctype == OP_NOT_VSPACE)) break;              if (gotspace == (ctype == OP_NOT_VSPACE)) break;
5833              eptr += len;              eptr += len;
# Line 5783  for (;;) Line 5928  for (;;)
5928            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
5929            }            }
5930    
5931          /* 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;  
5932          for(;;)          for(;;)
5933            {            {
5934              if (eptr == pp) goto TAIL_RECURSE;
5935            RMATCH(eptr, ecode, offset_top, md, eptrb, RM46);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM46);
5936            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
5937            if (eptr-- == pp) break;        /* Stop if tried at original pos */            eptr--;
5938            BACKCHAR(eptr);            BACKCHAR(eptr);
5939            if (ctype == OP_ANYNL && eptr > pp  && *eptr == '\n' &&            if (ctype == OP_ANYNL && eptr > pp  && RAWUCHAR(eptr) == CHAR_NL &&
5940                eptr[-1] == '\r') eptr--;                RAWUCHAR(eptr - 1) == CHAR_CR) eptr--;
5941            }            }
5942          }          }
5943        else        else
# Line 5815  for (;;) Line 5955  for (;;)
5955                break;                break;
5956                }                }
5957              if (IS_NEWLINE(eptr)) break;              if (IS_NEWLINE(eptr)) break;
5958                if (md->partial != 0 &&    /* Take care with CRLF partial */
5959                    eptr + 1 >= md->end_subject &&
5960                    NLBLOCK->nltype == NLTYPE_FIXED &&
5961                    NLBLOCK->nllen == 2 &&
5962                    *eptr == NLBLOCK->nl[0])
5963                  {
5964                  md->hitend = TRUE;
5965                  if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
5966                  }
5967              eptr++;              eptr++;
5968              }              }
5969            break;            break;
# Line 5839  for (;;) Line 5988  for (;;)
5988                break;                break;
5989                }                }
5990              c = *eptr;              c = *eptr;
5991              if (c == 0x000d)              if (c == CHAR_CR)
5992                {                {
5993                if (++eptr >= md->end_subject) break;                if (++eptr >= md->end_subject) break;
5994                if (*eptr == 0x000a) eptr++;                if (*eptr == CHAR_LF) eptr++;
5995                }                }
5996              else              else
5997                {                {
5998                if (c != 0x000a && (md->bsr_anycrlf ||                if (c != CHAR_LF && (md->bsr_anycrlf ||
5999                  (c != 0x000b && c != 0x000c && c != 0x0085                   (c != CHAR_VT && c != CHAR_FF && c != CHAR_NEL
6000  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6001                  && c != 0x2028 && c != 0x2029                   && c != 0x2028 && c != 0x2029
6002  #endif  #endif
6003                  ))) break;                   ))) break;
6004                eptr++;                eptr++;
6005                }                }
6006              }              }
# Line 5865  for (;;) Line 6014  for (;;)
6014                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6015                break;                break;
6016                }                }
6017              c = *eptr;              switch(*eptr)
6018              if (c == 0x09 || c == 0x20 || c == 0xa0                {
6019  #ifdef COMPILE_PCRE16                default: eptr++; break;
6020                || c == 0x1680 || c == 0x180e || (c >= 0x2000 && c <= 0x200A)                HSPACE_BYTE_CASES:
6021                || c == 0x202f || c == 0x205f || c == 0x3000  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6022                  HSPACE_MULTIBYTE_CASES:
6023  #endif  #endif
6024                ) break;                goto ENDLOOP00;
6025              eptr++;                }
6026              }              }
6027              ENDLOOP00:
6028            break;            break;
6029    
6030            case OP_HSPACE:            case OP_HSPACE:
# Line 5884  for (;;) Line 6035  for (;;)
6035                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6036                break;                break;
6037                }                }
6038              c = *eptr;              switch(*eptr)
6039              if (c != 0x09 && c != 0x20 && c != 0xa0                {
6040  #ifdef COMPILE_PCRE16                default: goto ENDLOOP01;
6041                && c != 0x1680 && c != 0x180e && (c < 0x2000 || c > 0x200A)                HSPACE_BYTE_CASES:
6042                && c != 0x202f && c != 0x205f && c != 0x3000  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6043                  HSPACE_MULTIBYTE_CASES:
6044  #endif  #endif
6045                ) break;                eptr++; break;
6046              eptr++;                }
6047              }              }
6048              ENDLOOP01:
6049            break;            break;
6050    
6051            case OP_NOT_VSPACE:            case OP_NOT_VSPACE:
# Line 5903  for (;;) Line 6056  for (;;)
6056                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6057                break;                break;
6058                }                }
6059              c = *eptr;              switch(*eptr)
6060              if (c == 0x0a || c == 0x0b || c == 0x0c || c == 0x0d || c == 0x85                {
6061  #ifdef COMPILE_PCRE16                default: eptr++; break;
6062                || c == 0x2028 || c == 0x2029                VSPACE_BYTE_CASES:
6063    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6064                  VSPACE_MULTIBYTE_CASES:
6065  #endif  #endif
6066                ) break;                goto ENDLOOP02;
6067              eptr++;                }
6068              }              }
6069              ENDLOOP02:
6070            break;            break;
6071    
6072            case OP_VSPACE:            case OP_VSPACE:
# Line 5921  for (;;) Line 6077  for (;;)
6077                SCHECK_PARTIAL();                SCHECK_PARTIAL();
6078                break;                break;
6079                }                }
6080              c = *eptr;              switch(*eptr)
6081              if (c != 0x0a && c != 0x0b && c != 0x0c && c != 0x0d && c != 0x85                {
6082  #ifdef COMPILE_PCRE16                default: goto ENDLOOP03;
6083                && c != 0x2028 && c != 0x2029                VSPACE_BYTE_CASES:
6084    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
6085                  VSPACE_MULTIBYTE_CASES:
6086  #endif  #endif
6087                ) break;                eptr++; break;
6088              eptr++;                }
6089              }              }
6090              ENDLOOP03:
6091            break;            break;
6092    
6093            case OP_NOT_DIGIT:            case OP_NOT_DIGIT:
# Line 6013  for (;;) Line 6172  for (;;)
6172            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
6173            }            }
6174    
6175          /* eptr is now past the end of the maximum run. If possessive, we are          if (possessive) continue;    /* No backtracking */
6176          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)  
6177            {            {
6178              if (eptr == pp) goto TAIL_RECURSE;
6179            RMATCH(eptr, ecode, offset_top, md, eptrb, RM47);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM47);
6180            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
6181            eptr--;            eptr--;
6182            if (ctype == OP_ANYNL && eptr > pp  && *eptr == '\n' &&            if (ctype == OP_ANYNL && eptr > pp  && *eptr == CHAR_LF &&
6183                eptr[-1] == '\r') eptr--;                eptr[-1] == CHAR_CR) eptr--;
6184            }            }
6185          }          }
6186    
6187        /* Get here if we can't make it match with any permitted repetitions */        /* Control never gets here */
   
       RRETURN(MATCH_NOMATCH);  
6188        }        }
     /* Control never gets here */  
6189    
6190      /* 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
6191      something seriously wrong in the code above or the OP_xxx definitions. */      something seriously wrong in the code above or the OP_xxx definitions. */
# Line 6076  switch (frame->Xwhere) Line 6227  switch (frame->Xwhere)
6227    LBL(32) LBL(34) LBL(42) LBL(46)    LBL(32) LBL(34) LBL(42) LBL(46)
6228  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
6229    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)
6230    LBL(59) LBL(60) LBL(61) LBL(62)    LBL(59) LBL(60) LBL(61) LBL(62) LBL(67) LBL(68)
6231  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
6232  #endif  /* SUPPORT_UTF */  #endif  /* SUPPORT_UTF */
6233    default:    default:
6234    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);  
   
6235    return PCRE_ERROR_INTERNAL;    return PCRE_ERROR_INTERNAL;
6236    }    }
6237  #undef LBL  #undef LBL
# Line 6145  Undefine all the macros that were define Line 6293  Undefine all the macros that were define
6293  ***************************************************************************/  ***************************************************************************/
6294    
6295    
6296    #ifdef NO_RECURSE
6297    /*************************************************
6298    *          Release allocated heap frames         *
6299    *************************************************/
6300    
6301    /* This function releases all the allocated frames. The base frame is on the
6302    machine stack, and so must not be freed.
6303    
6304    Argument: the address of the base frame
6305    Returns:  nothing
6306    */
6307    
6308    static void
6309    release_match_heapframes (heapframe *frame_base)
6310    {
6311    heapframe *nextframe = frame_base->Xnextframe;
6312    while (nextframe != NULL)
6313      {
6314      heapframe *oldframe = nextframe;
6315      nextframe = nextframe->Xnextframe;
6316      (PUBL(stack_free))(oldframe);
6317      }
6318    }
6319    #endif
6320    
6321    
6322  /*************************************************  /*************************************************
6323  *         Execute a Regular Expression           *  *         Execute a Regular Expression           *
# Line 6170  Returns:          > 0 => success; value Line 6343  Returns:          > 0 => success; value
6343                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
6344  */  */
6345    
6346  #ifdef COMPILE_PCRE8  #if defined COMPILE_PCRE8
6347  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
6348  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,
6349    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,
6350    int offsetcount)    int offsetcount)
6351  #else  #elif defined COMPILE_PCRE16
6352  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
6353  pcre16_exec(const pcre16 *argument_re, const pcre16_extra *extra_data,  pcre16_exec(const pcre16 *argument_re, const pcre16_extra *extra_data,
6354    PCRE_SPTR16 subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR16 subject, int length, int start_offset, int options, int *offsets,
6355    int offsetcount)    int offsetcount)
6356    #elif defined COMPILE_PCRE32
6357    PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
6358    pcre32_exec(const pcre32 *argument_re, const pcre32_extra *extra_data,
6359      PCRE_SPTR32 subject, int length, int start_offset, int options, int *offsets,
6360      int offsetcount)
6361  #endif  #endif
6362  {  {
6363  int rc, ocount, arg_offset_max;  int rc, ocount, arg_offset_max;
# Line 6202  const pcre_uint8 *start_bits = NULL; Line 6380  const pcre_uint8 *start_bits = NULL;
6380  PCRE_PUCHAR start_match = (PCRE_PUCHAR)subject + start_offset;  PCRE_PUCHAR start_match = (PCRE_PUCHAR)subject + start_offset;
6381  PCRE_PUCHAR end_subject;  PCRE_PUCHAR end_subject;
6382  PCRE_PUCHAR start_partial = NULL;  PCRE_PUCHAR start_partial = NULL;
6383    PCRE_PUCHAR match_partial = NULL;
6384  PCRE_PUCHAR req_char_ptr = start_match - 1;  PCRE_PUCHAR req_char_ptr = start_match - 1;
6385    
6386  const pcre_study_data *study;  const pcre_study_data *study;
6387  const REAL_PCRE *re = (const REAL_PCRE *)argument_re;  const REAL_PCRE *re = (const REAL_PCRE *)argument_re;
6388    
6389    #ifdef NO_RECURSE
6390    heapframe frame_zero;
6391    frame_zero.Xprevframe = NULL;            /* Marks the top level */
6392    frame_zero.Xnextframe = NULL;            /* None are allocated yet */
6393    md->match_frames_base = &frame_zero;
6394    #endif
6395    
6396  /* 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
6397  per recursive call of match(). */  per recursive call of match(). Without the funny casting for sizeof, a Windows
6398    compiler gave this error: "unary minus operator applied to unsigned type,
6399    result still unsigned". Hopefully the cast fixes that. */
6400    
6401  if (re == NULL && extra_data == NULL && subject == NULL && length == -999 &&  if (re == NULL && extra_data == NULL && subject == NULL && length == -999 &&
6402      start_offset == -999)      start_offset == -999)
6403  #ifdef NO_RECURSE  #ifdef NO_RECURSE
6404    return -sizeof(heapframe);    return -((int)sizeof(heapframe));
6405  #else  #else
6406    return match(NULL, NULL, NULL, 0, NULL, NULL, 0);    return match(NULL, NULL, NULL, 0, NULL, NULL, 0);
6407  #endif  #endif
# Line 6224  if ((options & ~PUBLIC_EXEC_OPTIONS) != Line 6412  if ((options & ~PUBLIC_EXEC_OPTIONS) !=
6412  if (re == NULL || subject == NULL || (offsets == NULL && offsetcount > 0))  if (re == NULL || subject == NULL || (offsets == NULL && offsetcount > 0))
6413    return PCRE_ERROR_NULL;    return PCRE_ERROR_NULL;
6414  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;  if (offsetcount < 0) return PCRE_ERROR_BADCOUNT;
6415    if (length < 0) return PCRE_ERROR_BADLENGTH;
6416  if (start_offset < 0 || start_offset > length) return PCRE_ERROR_BADOFFSET;  if (start_offset < 0 || start_offset > length) return PCRE_ERROR_BADOFFSET;
6417    
6418  /* 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 6261  if (utf && (options & PCRE_NO_UTF8_CHECK Line 6450  if (utf && (options & PCRE_NO_UTF8_CHECK
6450        offsets[0] = erroroffset;        offsets[0] = erroroffset;
6451        offsets[1] = errorcode;        offsets[1] = errorcode;
6452        }        }
6453  #ifdef COMPILE_PCRE16  #if defined COMPILE_PCRE8
     return (errorcode <= PCRE_UTF16_ERR1 && md->partial > 1)?  
       PCRE_ERROR_SHORTUTF16 : PCRE_ERROR_BADUTF16;  
 #else  
6454      return (errorcode <= PCRE_UTF8_ERR5 && md->partial > 1)?      return (errorcode <= PCRE_UTF8_ERR5 && md->partial > 1)?
6455        PCRE_ERROR_SHORTUTF8 : PCRE_ERROR_BADUTF8;        PCRE_ERROR_SHORTUTF8 : PCRE_ERROR_BADUTF8;
6456    #elif defined COMPILE_PCRE16
6457        return (errorcode <= PCRE_UTF16_ERR1 && md->partial > 1)?
6458          PCRE_ERROR_SHORTUTF16 : PCRE_ERROR_BADUTF16;
6459    #elif defined COMPILE_PCRE32
6460        return PCRE_ERROR_BADUTF32;
6461  #endif  #endif
6462      }      }
6463    #if defined COMPILE_PCRE8 || defined COMPILE_PCRE16
6464    /* 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. */
6465    if (start_offset > 0 && start_offset < length &&    if (start_offset > 0 && start_offset < length &&
6466        NOT_FIRSTCHAR(((PCRE_PUCHAR)subject)[start_offset]))        NOT_FIRSTCHAR(((PCRE_PUCHAR)subject)[start_offset]))
6467      return PCRE_ERROR_BADUTF8_OFFSET;      return PCRE_ERROR_BADUTF8_OFFSET;
6468    #endif
6469    }    }
6470  #endif  #endif
6471    
6472  /* If the pattern was successfully studied with JIT support, run the JIT  /* If the pattern was successfully studied with JIT support, run the JIT
6473  executable instead of the rest of this function. Most options must be set at  executable instead of the rest of this function. Most options must be set at
6474  compile time for the JIT code to be usable. Fallback to the normal code path if  compile time for the JIT code to be usable. Fallback to the normal code path if
6475  an unsupported flag is set. In particular, JIT does not support partial  an unsupported flag is set. */
 matching. */  
6476    
6477  #ifdef SUPPORT_JIT  #ifdef SUPPORT_JIT
6478  if (extra_data != NULL  if (extra_data != NULL
6479      && (extra_data->flags & PCRE_EXTRA_EXECUTABLE_JIT) != 0      && (extra_data->flags & (PCRE_EXTRA_EXECUTABLE_JIT |
6480                                 PCRE_EXTRA_TABLES)) == PCRE_EXTRA_EXECUTABLE_JIT
6481      && extra_data->executable_jit != NULL      && extra_data->executable_jit != NULL
6482      && (extra_data->flags & PCRE_EXTRA_TABLES) == 0      && (options & ~PUBLIC_JIT_EXEC_OPTIONS) == 0)
6483      && (options & ~(PCRE_NO_UTF8_CHECK | PCRE_NOTBOL | PCRE_NOTEOL |    {
6484                      PCRE_NOTEMPTY | PCRE_NOTEMPTY_ATSTART)) == 0)    rc = PRIV(jit_exec)(extra_data, (const pcre_uchar *)subject, length,
6485    return PRIV(jit_exec)(re, extra_data->executable_jit,         start_offset, options, offsets, offsetcount);
6486      (const pcre_uchar *)subject, length, start_offset, options,  
6487      ((extra_data->flags & PCRE_EXTRA_MATCH_LIMIT) == 0)    /* PCRE_ERROR_NULL means that the selected normal or partial matching
6488      ? MATCH_LIMIT : extra_data->match_limit, offsets, offsetcount);    mode is not compiled. In this case we simply fallback to interpreter. */
6489    
6490      if (rc != PCRE_ERROR_JIT_BADOPTION) return rc;
6491      }
6492  #endif  #endif
6493    
6494  /* Carry on with non-JIT matching. This information is for finding all the  /* Carry on with non-JIT matching. This information is for finding all the
# Line 6315  md->callout_data = NULL; Line 6510  md->callout_data = NULL;
6510    
6511  tables = re->tables;  tables = re->tables;
6512    
6513    /* The two limit values override the defaults, whatever their value. */
6514    
6515  if (extra_data != NULL)  if (extra_data != NULL)
6516    {    {
6517    register unsigned int flags = extra_data->flags;    register unsigned int flags = extra_data->flags;
# Line 6329  if (extra_data != NULL) Line 6526  if (extra_data != NULL)
6526    if ((flags & PCRE_EXTRA_TABLES) != 0) tables = extra_data->tables;    if ((flags & PCRE_EXTRA_TABLES) != 0) tables = extra_data->tables;
6527    }    }
6528    
6529    /* Limits in the regex override only if they are smaller. */
6530    
6531    if ((re->flags & PCRE_MLSET) != 0 && re->limit_match < md->match_limit)
6532      md->match_limit = re->limit_match;
6533    
6534    if ((re->flags & PCRE_RLSET) != 0 &&
6535        re->limit_recursion < md->match_limit_recursion)
6536      md->match_limit_recursion = re->limit_recursion;
6537    
6538  /* 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
6539  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
6540  in other programs later. */  in other programs later. */
# Line 6354  end_subject = md->end_subject; Line 6560  end_subject = md->end_subject;
6560  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;
6561  md->use_ucp = (re->options & PCRE_UCP) != 0;  md->use_ucp = (re->options & PCRE_UCP) != 0;
6562  md->jscript_compat = (re->options & PCRE_JAVASCRIPT_COMPAT) != 0;  md->jscript_compat = (re->options & PCRE_JAVASCRIPT_COMPAT) != 0;
6563  md->ignore_skip_arg = FALSE;  md->ignore_skip_arg = 0;
6564    
6565  /* Some options are unpacked into BOOL variables in the hope that testing  /* Some options are unpacked into BOOL variables in the hope that testing
6566  them will be faster than individual option bits. */  them will be faster than individual option bits. */
# Line 6464  if (re->top_backref > 0 && re->top_backr Line 6670  if (re->top_backref > 0 && re->top_backr
6670    DPRINTF(("Got memory to hold back references\n"));    DPRINTF(("Got memory to hold back references\n"));
6671    }    }
6672  else md->offset_vector = offsets;  else md->offset_vector = offsets;
   
6673  md->offset_end = ocount;  md->offset_end = ocount;
6674  md->offset_max = (2*ocount)/3;  md->offset_max = (2*ocount)/3;
6675  md->offset_overflow = FALSE;  md->capture_last = 0;
 md->capture_last = -1;  
6676    
6677  /* Reset the working variable associated with each extraction. These should  /* Reset the working variable associated with each extraction. These should
6678  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 6496  if (!anchored) Line 6700  if (!anchored)
6700    if ((re->flags & PCRE_FIRSTSET) != 0)    if ((re->flags & PCRE_FIRSTSET) != 0)
6701      {      {
6702      has_first_char = TRUE;      has_first_char = TRUE;
6703      first_char = first_char2 = re->first_char;      first_char = first_char2 = (pcre_uchar)(re->first_char);
6704      if ((re->flags & PCRE_FCH_CASELESS) != 0)      if ((re->flags & PCRE_FCH_CASELESS) != 0)
6705        {        {
6706        first_char2 = TABLE_GET(first_char, md->fcc, first_char);        first_char2 = TABLE_GET(first_char, md->fcc, first_char);
# Line 6518  character" set. */ Line 6722  character" set. */
6722  if ((re->flags & PCRE_REQCHSET) != 0)  if ((re->flags & PCRE_REQCHSET) != 0)
6723    {    {