/[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 771 by ph10, Tue Nov 29 15:34:12 2011 UTC revision 1365 by ph10, Sun Oct 6 18:33:56 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) 1997-2011 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 93  because the offset vector is always a mu Line 110  because the offset vector is always a mu
110  static const char rep_min[] = { 0, 0, 1, 1, 0, 0 };  static const char rep_min[] = { 0, 0, 1, 1, 0, 0 };
111  static const char rep_max[] = { 0, 0, 0, 0, 1, 1 };  static const char rep_max[] = { 0, 0, 0, 0, 1, 1 };
112    
   
   
113  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
114  /*************************************************  /*************************************************
115  *        Debugging function to print chars       *  *        Debugging function to print chars       *
# Line 113  Returns:     nothing Line 128  Returns:     nothing
128  */  */
129    
130  static void  static void
131  pchars(const uschar *p, int length, BOOL is_subject, match_data *md)  pchars(const pcre_uchar *p, int length, BOOL is_subject, match_data *md)
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
165  match_ref(int offset, register USPTR eptr, int length, match_data *md,  match_ref(int offset, register PCRE_PUCHAR eptr, int length, match_data *md,
166    BOOL caseless)    BOOL caseless)
167  {  {
168  USPTR eptr_start = eptr;  PCRE_PUCHAR eptr_start = eptr;
169  register USPTR 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 173  ASCII characters. */ Line 195  ASCII characters. */
195    
196  if (caseless)  if (caseless)
197    {    {
198  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
199  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
200    if (md->utf8)    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      USPTR 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        { if (md->lcc[*p++] != md->lcc[*eptr++]) return -1; }        {
240          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++;
246          eptr++;
247          }
248      }      }
249    }    }
250    
# Line 213  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 eptr - eptr_start;  return (int)(eptr - eptr_start);
264  }  }
265    
266    
# Line 269  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 287  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 307  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 *)(pcre_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 328  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;\
   (pcre_stack_free)(oldframe);\  
380    if (frame != NULL)\    if (frame != NULL)\
381      {\      {\
382      rrc = ra;\      rrc = ra;\
# Line 342  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    
397    USPTR Xeptr;    PCRE_PUCHAR Xeptr;
398    const uschar *Xecode;    const pcre_uchar *Xecode;
399    USPTR Xmstart;    PCRE_PUCHAR Xmstart;
400    int Xoffset_top;    int Xoffset_top;
401    eptrblock *Xeptrb;    eptrblock *Xeptrb;
402    unsigned int Xrdepth;    unsigned int Xrdepth;
403    
404    /* Function local variables */    /* Function local variables */
405    
406    USPTR Xcallpat;    PCRE_PUCHAR Xcallpat;
407  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
408    USPTR Xcharptr;    PCRE_PUCHAR Xcharptr;
409  #endif  #endif
410    USPTR Xdata;    PCRE_PUCHAR Xdata;
411    USPTR Xnext;    PCRE_PUCHAR Xnext;
412    USPTR Xpp;    PCRE_PUCHAR Xpp;
413    USPTR Xprev;    PCRE_PUCHAR Xprev;
414    USPTR Xsaved_eptr;    PCRE_PUCHAR Xsaved_eptr;
415    
416    recursion_info Xnew_recursive;    recursion_info Xnew_recursive;
417    
# Line 372  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    uschar Xocchars[8];    pcre_uchar Xocchars[6];
428  #endif  #endif
429    
430    int Xcodelink;    int Xcodelink;
# Line 385  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 417  returns a negative (error) response, the Line 466  returns a negative (error) response, the
466  same response. */  same response. */
467    
468  /* These macros pack up tests that are used for partial matching, and which  /* These macros pack up tests that are used for partial matching, and which
469  appears several times in the code. We set the "hit end" flag if the pointer is  appear several times in the code. We set the "hit end" flag if the pointer is
470  at the end of the subject and also past the start of the subject (i.e.  at the end of the subject and also past the start of the subject (i.e.
471  something has been matched). For hard partial matching, we then return  something has been matched). For hard partial matching, we then return
472  immediately. The second one is used when we already know we are past the end of  immediately. The second one is used when we already know we are past the end of
# Line 440  the subject. */ Line 489  the subject. */
489    
490    
491  /* Performance note: It might be tempting to extract commonly used fields from  /* Performance note: It might be tempting to extract commonly used fields from
492  the md structure (e.g. utf8, end_subject) into individual variables to improve  the md structure (e.g. utf, end_subject) into individual variables to improve
493  performance. Tests using gcc on a SPARC disproved this; in the first case, it  performance. Tests using gcc on a SPARC disproved this; in the first case, it
494  made performance worse.  made performance worse.
495    
# Line 463  Returns:       MATCH_MATCH if matched Line 512  Returns:       MATCH_MATCH if matched
512  */  */
513    
514  static int  static int
515  match(REGISTER USPTR eptr, REGISTER const uschar *ecode, USPTR mstart,  match(REGISTER PCRE_PUCHAR eptr, REGISTER const pcre_uchar *ecode,
516    int offset_top, match_data *md, eptrblock *eptrb, unsigned int rdepth)    PCRE_PUCHAR mstart, int offset_top, match_data *md, eptrblock *eptrb,
517      unsigned int rdepth)
518  {  {
519  /* These variables do not need to be preserved over recursion in this function,  /* These variables do not need to be preserved over recursion in this function,
520  so they can be ordinary variables in all cases. Mark some of them with  so they can be ordinary variables in all cases. Mark some of them with
# Line 472  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 utf8;        /* Local copy of UTF-8 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 */
529  BOOL caseless;  BOOL caseless;
530  int condcode;  int condcode;
531    
532  /* When recursion is not being used, all "local" variables that have to be  /* When recursion is not being used, all "local" variables that have to be
533  preserved over calls to RMATCH() are part of a "frame" which is obtained from  preserved over calls to RMATCH() are part of a "frame". We set up the top-level
534  heap storage. Set up the top-level frame here; others are obtained from the  frame on the stack here; subsequent instantiations are obtained from the heap
535  heap whenever RMATCH() does a "recursion". See the macro definitions above. */  whenever RMATCH() does a "recursion". See the macro definitions above. Putting
536    the top-level on the stack rather than malloc-ing them all gives a performance
537    boost in many cases where there is not much "recursion". */
538    
539  #ifdef NO_RECURSE  #ifdef NO_RECURSE
540  heapframe *frame = (heapframe *)(pcre_stack_malloc)(sizeof(heapframe));  heapframe *frame = (heapframe *)md->match_frames_base;
 if (frame == NULL) RRETURN(PCRE_ERROR_NOMEMORY);  
 frame->Xprevframe = NULL;            /* Marks the top level */  
541    
542  /* Copy in the original argument variables */  /* Copy in the original argument variables */
543    
# Line 513  HEAP_RECURSE: Line 563  HEAP_RECURSE:
563    
564  /* Ditto for the local variables */  /* Ditto for the local variables */
565    
566  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
567  #define charptr            frame->Xcharptr  #define charptr            frame->Xcharptr
568  #endif  #endif
569  #define callpat            frame->Xcallpat  #define callpat            frame->Xcallpat
# Line 571  declarations can be cut out in a block. Line 621  declarations can be cut out in a block.
621  below are for variables that do not have to be preserved over a recursive call  below are for variables that do not have to be preserved over a recursive call
622  to RMATCH(). */  to RMATCH(). */
623    
624  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
625  const uschar *charptr;  const pcre_uchar *charptr;
626  #endif  #endif
627  const uschar *callpat;  const pcre_uchar *callpat;
628  const uschar *data;  const pcre_uchar *data;
629  const uschar *next;  const pcre_uchar *next;
630  USPTR         pp;  PCRE_PUCHAR       pp;
631  const uschar *prev;  const pcre_uchar *prev;
632  USPTR         saved_eptr;  PCRE_PUCHAR       saved_eptr;
633    
634  recursion_info new_recursive;  recursion_info new_recursive;
635    
# Line 589  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  uschar occhars[8];  pcre_uchar occhars[6];
646  #endif  #endif
647    
648  int codelink;  int codelink;
# Line 600  int ctype; Line 650  int ctype;
650  int length;  int length;
651  int max;  int max;
652  int min;  int min;
653  int number;  unsigned int number;
654  int offset;  int offset;
655  int op;  unsigned int op;
656  int save_capture_last;  pcre_int32 save_capture_last;
657  int save_offset1, save_offset2, save_offset3;  int save_offset1, save_offset2, save_offset3;
658  int stacksave[REC_STACK_SAVE_MAX];  int stacksave[REC_STACK_SAVE_MAX];
659    
660  eptrblock newptrb;  eptrblock newptrb;
661    
662    /* There is a special fudge for calling match() in a way that causes it to
663    measure the size of its basic stack frame when the stack is being used for
664    recursion. The second argument (ecode) being NULL triggers this behaviour. It
665    cannot normally ever be NULL. The return is the negated value of the frame
666    size. */
667    
668    if (ecode == NULL)
669      {
670      if (rdepth == 0)
671        return match((PCRE_PUCHAR)&rdepth, NULL, NULL, 0, NULL, NULL, 1);
672      else
673        {
674        int len = (char *)&rdepth - (char *)eptr;
675        return (len > 0)? -len : len;
676        }
677      }
678  #endif     /* NO_RECURSE */  #endif     /* NO_RECURSE */
679    
680  /* To save space on the stack and in the heap frame, I have doubled up on some  /* To save space on the stack and in the heap frame, I have doubled up on some
# Line 620  the alternative names that are used. */ Line 687  the alternative names that are used. */
687  #define code_offset   codelink  #define code_offset   codelink
688  #define condassert    condition  #define condassert    condition
689  #define matched_once  prev_is_word  #define matched_once  prev_is_word
690    #define foc           number
691    #define save_mark     data
692    
693  /* These statements are here to stop the compiler complaining about unitialized  /* These statements are here to stop the compiler complaining about unitialized
694  variables. */  variables. */
# Line 645  defined). However, RMATCH isn't like a f Line 714  defined). However, RMATCH isn't like a f
714  complicated macro. It has to be used in one particular way. This shouldn't,  complicated macro. It has to be used in one particular way. This shouldn't,
715  however, impact performance when true recursion is being used. */  however, impact performance when true recursion is being used. */
716    
717  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
718  utf8 = md->utf8;       /* Local copy of the flag */  utf = md->utf;       /* Local copy of the flag */
719  #else  #else
720  utf8 = FALSE;  utf = FALSE;
721  #endif  #endif
722    
723  /* First check that we haven't called match() too many times, or that we  /* First check that we haven't called match() too many times, or that we
# Line 689  for (;;) Line 758  for (;;)
758      case OP_MARK:      case OP_MARK:
759      md->nomatch_mark = ecode + 2;      md->nomatch_mark = ecode + 2;
760      md->mark = NULL;    /* In case previously set by assertion */      md->mark = NULL;    /* In case previously set by assertion */
761      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
762        eptrb, RM55);        eptrb, RM55);
763      if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&      if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
764           md->mark == NULL) md->mark = ecode + 2;           md->mark == NULL) md->mark = ecode + 2;
# Line 702  for (;;) Line 771  for (;;)
771      unaltered. */      unaltered. */
772    
773      else if (rrc == MATCH_SKIP_ARG &&      else if (rrc == MATCH_SKIP_ARG &&
774          strcmp((char *)(ecode + 2), (char *)(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 712  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 + _pcre_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 + _pcre_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:
797      md->nomatch_mark = ecode + 2;      md->nomatch_mark = ecode + 2;
798      md->mark = NULL;    /* In case previously set by assertion */      md->mark = NULL;    /* In case previously set by assertion */
799      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top, md,
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 + _pcre_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 += _pcre_OP_lengths[*ecode] + ecode[1];        ecode += PRIV(OP_lengths)[*ecode] + ecode[1];
826        break;        break;
827        }        }
828      RMATCH(eptr, ecode + _pcre_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 779  for (;;) Line 842  for (;;)
842      match pointer to do this. */      match pointer to do this. */
843    
844      case OP_THEN:      case OP_THEN:
845      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
846        eptrb, RM54);        eptrb, RM54);
847      if (rrc != MATCH_NOMATCH) RRETURN(rrc);      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
848      md->start_match_ptr = ecode;      md->start_match_ptr = ecode;
# Line 788  for (;;) Line 851  for (;;)
851      case OP_THEN_ARG:      case OP_THEN_ARG:
852      md->nomatch_mark = ecode + 2;      md->nomatch_mark = ecode + 2;
853      md->mark = NULL;    /* In case previously set by assertion */      md->mark = NULL;    /* In case previously set by assertion */
854      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode] + ecode[1], offset_top,      RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode] + ecode[1], offset_top,
855        md, eptrb, RM58);        md, eptrb, RM58);
856      if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&      if ((rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) &&
857           md->mark == NULL) md->mark = ecode + 2;           md->mark == NULL) md->mark = ecode + 2;
# Line 812  for (;;) Line 875  for (;;)
875      case OP_ONCE_NC:      case OP_ONCE_NC:
876      prev = ecode;      prev = ecode;
877      saved_eptr = eptr;      saved_eptr = eptr;
878        save_mark = md->mark;
879      do      do
880        {        {
881        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM64);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM64);
# Line 830  for (;;) Line 894  for (;;)
894    
895        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
896        ecode += GET(ecode,1);        ecode += GET(ecode,1);
897          md->mark = save_mark;
898        }        }
899      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
900    
# Line 869  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 909  for (;;) Line 973  for (;;)
973        save_offset2 = md->offset_vector[offset+1];        save_offset2 = md->offset_vector[offset+1];
974        save_offset3 = md->offset_vector[md->offset_end - number];        save_offset3 = md->offset_vector[md->offset_end - number];
975        save_capture_last = md->capture_last;        save_capture_last = md->capture_last;
976          save_mark = md->mark;
977    
978        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));
979        md->offset_vector[md->offset_end - number] =        md->offset_vector[md->offset_end - number] =
# Line 917  for (;;) Line 982  for (;;)
982        for (;;)        for (;;)
983          {          {
984          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
985          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
986            eptrb, RM1);            eptrb, RM1);
987          if (rrc == MATCH_ONCE) break;  /* Backing up through an atomic group */          if (rrc == MATCH_ONCE) break;  /* Backing up through an atomic group */
988    
# Line 945  for (;;) Line 1010  for (;;)
1010          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1011          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
1012          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
1013            md->mark = save_mark;
1014          if (*ecode != OP_ALT) break;          if (*ecode != OP_ALT) break;
1015          }          }
1016    
# Line 996  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 1004  for (;;) Line 1071  for (;;)
1071    
1072        else if (!md->hasthen && ecode[GET(ecode, 1)] != OP_ALT)        else if (!md->hasthen && ecode[GET(ecode, 1)] != OP_ALT)
1073          {          {
1074          ecode += _pcre_OP_lengths[*ecode];          ecode += PRIV(OP_lengths)[*ecode];
1075          goto TAIL_RECURSE;          goto TAIL_RECURSE;
1076          }          }
1077    
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        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, eptrb,        save_mark = md->mark;
1081          save_capture_last = md->capture_last;
1082          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md, eptrb,
1083          RM2);          RM2);
1084    
1085        /* See comment in the code for capturing groups above about handling        /* See comment in the code for capturing groups above about handling
# Line 1028  for (;;) Line 1097  for (;;)
1097          {          {
1098          if (rrc == MATCH_ONCE)          if (rrc == MATCH_ONCE)
1099            {            {
1100            const uschar *scode = ecode;            const pcre_uchar *scode = ecode;
1101            if (*scode != OP_ONCE)           /* If not at start, find it */            if (*scode != OP_ONCE)           /* If not at start, find it */
1102              {              {
1103              while (*scode == OP_ALT) scode += GET(scode, 1);              while (*scode == OP_ALT) scode += GET(scode, 1);
# Line 1039  for (;;) Line 1108  for (;;)
1108          RRETURN(rrc);          RRETURN(rrc);
1109          }          }
1110        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1111          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 1070  for (;;) Line 1141  for (;;)
1141      if (offset < md->offset_max)      if (offset < md->offset_max)
1142        {        {
1143        matched_once = FALSE;        matched_once = FALSE;
1144        code_offset = ecode - md->start_code;        code_offset = (int)(ecode - md->start_code);
1145    
1146        save_offset1 = md->offset_vector[offset];        save_offset1 = md->offset_vector[offset];
1147        save_offset2 = md->offset_vector[offset+1];        save_offset2 = md->offset_vector[offset+1];
# Line 1093  for (;;) Line 1164  for (;;)
1164          md->offset_vector[md->offset_end - number] =          md->offset_vector[md->offset_end - number] =
1165            (int)(eptr - md->start_subject);            (int)(eptr - md->start_subject);
1166          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;          if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
1167          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
1168            eptrb, RM63);            eptrb, RM63);
1169          if (rrc == MATCH_KETRPOS)          if (rrc == MATCH_KETRPOS)
1170            {            {
# Line 1160  for (;;) Line 1231  for (;;)
1231    
1232      POSSESSIVE_NON_CAPTURE:      POSSESSIVE_NON_CAPTURE:
1233      matched_once = FALSE;      matched_once = FALSE;
1234      code_offset = ecode - md->start_code;      code_offset = (int)(ecode - md->start_code);
1235        save_capture_last = md->capture_last;
1236    
1237      for (;;)      for (;;)
1238        {        {
1239        if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;        if (op >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;
1240        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,        RMATCH(eptr, ecode + PRIV(OP_lengths)[*ecode], offset_top, md,
1241          eptrb, RM48);          eptrb, RM48);
1242        if (rrc == MATCH_KETRPOS)        if (rrc == MATCH_KETRPOS)
1243          {          {
# Line 1190  for (;;) Line 1262  for (;;)
1262        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1263        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1264        if (*ecode != OP_ALT) break;        if (*ecode != OP_ALT) break;
1265          md->capture_last = save_capture_last;
1266        }        }
1267    
1268      if (matched_once || allow_zero)      if (matched_once || allow_zero)
# Line 1201  for (;;) Line 1274  for (;;)
1274    
1275      /* Control never reaches here. */      /* Control never reaches here. */
1276    
1277      /* Conditional group: compilation checked that there are no more than      /* Conditional group: compilation checked that there are no more than two
1278      two branches. If the condition is false, skipping the first branch takes us      branches. If the condition is false, skipping the first branch takes us
1279      past the end if there is only one branch, but that's OK because that is      past the end of the item if there is only one branch, but that's exactly
1280      exactly what going to the ket would do. */      what we want. */
1281    
1282      case OP_COND:      case OP_COND:
1283      case OP_SCOND:      case OP_SCOND:
1284      codelink = GET(ecode, 1);  
1285        /* The variable codelink will be added to ecode when the condition is
1286        false, to get to the second branch. Setting it to the offset to the ALT
1287        or KET, then incrementing ecode achieves this effect. We now have ecode
1288        pointing to the condition or callout. */
1289    
1290        codelink = GET(ecode, 1);   /* Offset to the second branch */
1291        ecode += 1 + LINK_SIZE;     /* From this opcode */
1292    
1293      /* Because of the way auto-callout works during compile, a callout item is      /* Because of the way auto-callout works during compile, a callout item is
1294      inserted between OP_COND and an assertion condition. */      inserted between OP_COND and an assertion condition. */
1295    
1296      if (ecode[LINK_SIZE+1] == OP_CALLOUT)      if (*ecode == OP_CALLOUT)
1297        {        {
1298        if (pcre_callout != NULL)        if (PUBL(callout) != NULL)
1299          {          {
1300          pcre_callout_block cb;          PUBL(callout_block) cb;
1301          cb.version          = 2;   /* Version 1 of the callout block */          cb.version          = 2;   /* Version 1 of the callout block */
1302          cb.callout_number   = ecode[LINK_SIZE+2];          cb.callout_number   = ecode[1];
1303          cb.offset_vector    = md->offset_vector;          cb.offset_vector    = md->offset_vector;
1304    #if defined COMPILE_PCRE8
1305          cb.subject          = (PCRE_SPTR)md->start_subject;          cb.subject          = (PCRE_SPTR)md->start_subject;
1306    #elif defined COMPILE_PCRE16
1307            cb.subject          = (PCRE_SPTR16)md->start_subject;
1308    #elif defined COMPILE_PCRE32
1309            cb.subject          = (PCRE_SPTR32)md->start_subject;
1310    #endif
1311          cb.subject_length   = (int)(md->end_subject - md->start_subject);          cb.subject_length   = (int)(md->end_subject - md->start_subject);
1312          cb.start_match      = (int)(mstart - md->start_subject);          cb.start_match      = (int)(mstart - md->start_subject);
1313          cb.current_position = (int)(eptr - md->start_subject);          cb.current_position = (int)(eptr - md->start_subject);
1314          cb.pattern_position = GET(ecode, LINK_SIZE + 3);          cb.pattern_position = GET(ecode, 2);
1315          cb.next_item_length = GET(ecode, 3 + 2*LINK_SIZE);          cb.next_item_length = GET(ecode, 2 + LINK_SIZE);
1316          cb.capture_top      = offset_top/2;          cb.capture_top      = offset_top/2;
1317          cb.capture_last     = md->capture_last;          cb.capture_last     = md->capture_last & CAPLMASK;
1318            /* Internal change requires this for API compatibility. */
1319            if (cb.capture_last == 0) cb.capture_last = -1;
1320          cb.callout_data     = md->callout_data;          cb.callout_data     = md->callout_data;
1321          cb.mark             = md->nomatch_mark;          cb.mark             = md->nomatch_mark;
1322          if ((rrc = (*pcre_callout)(&cb)) > 0) RRETURN(MATCH_NOMATCH);          if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1323          if (rrc < 0) RRETURN(rrc);          if (rrc < 0) RRETURN(rrc);
1324          }          }
1325        ecode += _pcre_OP_lengths[OP_CALLOUT];  
1326        }        /* Advance ecode past the callout, so it now points to the condition. We
1327          must adjust codelink so that the value of ecode+codelink is unchanged. */
1328      condcode = ecode[LINK_SIZE+1];  
1329          ecode += PRIV(OP_lengths)[OP_CALLOUT];
1330      /* Now see what the actual condition is */        codelink -= PRIV(OP_lengths)[OP_CALLOUT];
1331          }
1332      if (condcode == OP_RREF || condcode == OP_NRREF)    /* Recursion test */  
1333        {      /* Test the various possible conditions */
1334        if (md->recursive == NULL)                /* Not recursing => FALSE */  
1335          {      condition = FALSE;
1336          condition = FALSE;      switch(condcode = *ecode)
1337          ecode += GET(ecode, 1);        {
1338          }        case OP_RREF:         /* Numbered group recursion test */
1339        else        if (md->recursive != NULL)     /* Not recursing => FALSE */
1340          {          {
1341          int recno = GET2(ecode, LINK_SIZE + 2);   /* Recursion group number*/          unsigned int recno = GET2(ecode, 1);   /* Recursion group number*/
1342          condition = (recno == RREF_ANY || recno == md->recursive->group_num);          condition = (recno == RREF_ANY || recno == md->recursive->group_num);
1343            }
1344          break;
1345    
1346          /* If the test is for recursion into a specific subpattern, and it is        case OP_DNRREF:       /* Duplicate named group recursion test */
1347          false, but the test was set up by name, scan the table to see if the        if (md->recursive != NULL)
1348          name refers to any other numbers, and test them. The condition is true          {
1349          if any one is set. */          int count = GET2(ecode, 1 + IMM2_SIZE);
1350            pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
1351          if (!condition && condcode == OP_NRREF)          while (count-- > 0)
1352            {            {
1353            uschar *slotA = md->name_table;            unsigned int recno = GET2(slot, 0);
1354            for (i = 0; i < md->name_count; i++)            condition = recno == md->recursive->group_num;
1355              {            if (condition) break;
1356              if (GET2(slotA, 0) == recno) break;            slot += md->name_entry_size;
             slotA += md->name_entry_size;  
             }  
   
           /* Found a name for the number - there can be only one; duplicate  
           names for different numbers are allowed, but not vice versa. First  
           scan down for duplicates. */  
   
           if (i < md->name_count)  
             {  
             uschar *slotB = slotA;  
             while (slotB > md->name_table)  
               {  
               slotB -= md->name_entry_size;  
               if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 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((char *)slotA + 2, (char *)slotB + 2) == 0)  
                   {  
                   condition = GET2(slotB, 0) == md->recursive->group_num;  
                   if (condition) break;  
                   }  
                 else break;  
                 }  
               }  
             }  
1357            }            }
1358            }
1359          break;
1360    
1361          /* Chose branch according to the condition */        case OP_CREF:         /* Numbered group used test */
1362          offset = GET2(ecode, 1) << 1;  /* Doubled ref number */
         ecode += condition? 3 : GET(ecode, 1);  
         }  
       }  
   
     else if (condcode == OP_CREF || condcode == OP_NCREF)  /* Group used test */  
       {  
       offset = GET2(ecode, LINK_SIZE+2) << 1;  /* Doubled ref number */  
1363        condition = offset < offset_top && md->offset_vector[offset] >= 0;        condition = offset < offset_top && md->offset_vector[offset] >= 0;
1364          break;
1365    
1366        /* If the numbered capture is unset, but the reference was by name,        case OP_DNCREF:      /* Duplicate named group used test */
       scan the table to see if the name refers to any other numbers, and test  
       them. The condition is true if any one is set. This is tediously similar  
       to the code above, but not close enough to try to amalgamate. */  
   
       if (!condition && condcode == OP_NCREF)  
1367          {          {
1368          int refno = offset >> 1;          int count = GET2(ecode, 1 + IMM2_SIZE);
1369          uschar *slotA = md->name_table;          pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
1370            while (count-- > 0)
         for (i = 0; i < md->name_count; i++)  
           {  
           if (GET2(slotA, 0) == refno) break;  
           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)  
1371            {            {
1372            uschar *slotB = slotA;            offset = GET2(slot, 0) << 1;
1373            while (slotB > md->name_table)            condition = offset < offset_top && md->offset_vector[offset] >= 0;
1374              {            if (condition) break;
1375              slotB -= md->name_entry_size;            slot += md->name_entry_size;
             if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 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((char *)slotA + 2, (char *)slotB + 2) == 0)  
                 {  
                 offset = GET2(slotB, 0) << 1;  
                 condition = offset < offset_top &&  
                   md->offset_vector[offset] >= 0;  
                 if (condition) break;  
                 }  
               else break;  
               }  
             }  
1376            }            }
1377          }          }
1378          break;
1379    
1380        /* Chose branch according to the condition */        case OP_DEF:     /* DEFINE - always false */
1381          break;
       ecode += condition? 3 : GET(ecode, 1);  
       }  
   
     else if (condcode == OP_DEF)     /* DEFINE - always false */  
       {  
       condition = FALSE;  
       ecode += GET(ecode, 1);  
       }  
   
     /* The condition is an assertion. Call match() to evaluate it - setting  
     md->match_function_type to MATCH_CONDASSERT causes it to stop at the end of  
     an assertion. */  
1382    
1383      else        /* The condition is an assertion. Call match() to evaluate it - setting
1384        {        md->match_function_type to MATCH_CONDASSERT causes it to stop at the end
1385          of an assertion. */
1386    
1387          default:
1388        md->match_function_type = MATCH_CONDASSERT;        md->match_function_type = MATCH_CONDASSERT;
1389        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM3);        RMATCH(eptr, ecode, offset_top, md, NULL, RM3);
1390        if (rrc == MATCH_MATCH)        if (rrc == MATCH_MATCH)
1391          {          {
1392          if (md->end_offset_top > offset_top)          if (md->end_offset_top > offset_top)
1393            offset_top = md->end_offset_top;  /* Captures may have happened */            offset_top = md->end_offset_top;  /* Captures may have happened */
1394          condition = TRUE;          condition = TRUE;
1395          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);  
1396            /* Advance ecode past the assertion to the start of the first branch,
1397            but adjust it so that the general choosing code below works. */
1398    
1399            ecode += GET(ecode, 1);
1400          while (*ecode == OP_ALT) ecode += GET(ecode, 1);          while (*ecode == OP_ALT) ecode += GET(ecode, 1);
1401            ecode += 1 + LINK_SIZE - PRIV(OP_lengths)[condcode];
1402          }          }
1403    
1404        /* PCRE doesn't allow the effect of (*THEN) to escape beyond an        /* PCRE doesn't allow the effect of (*THEN) to escape beyond an
1405        assertion; it is therefore treated as NOMATCH. */        assertion; it is therefore treated as NOMATCH. Any other return is an
1406          error. */
1407    
1408        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
1409          {          {
1410          RRETURN(rrc);         /* Need braces because of following else */          RRETURN(rrc);         /* Need braces because of following else */
1411          }          }
1412        else        break;
         {  
         condition = FALSE;  
         ecode += codelink;  
         }  
1413        }        }
1414    
1415      /* We are now at the branch that is to be obeyed. As there is only one, can      /* Choose branch according to the condition */
1416      use tail recursion to avoid using another stack frame, except when there is  
1417      unlimited repeat of a possibly empty group. In the latter case, a recursive      ecode += condition? PRIV(OP_lengths)[condcode] : codelink;
1418      call to match() is always required, unless the second alternative doesn't  
1419      exist, in which case we can just plough on. Note that, for compatibility      /* We are now at the branch that is to be obeyed. As there is only one, we
1420      with Perl, the | in a conditional group is NOT treated as creating two      can use tail recursion to avoid using another stack frame, except when
1421      alternatives. If a THEN is encountered in the branch, it propagates out to      there is unlimited repeat of a possibly empty group. In the latter case, a
1422      the enclosing alternative (unless nested in a deeper set of alternatives,      recursive call to match() is always required, unless the second alternative
1423      of course). */      doesn't exist, in which case we can just plough on. Note that, for
1424        compatibility with Perl, the | in a conditional group is NOT treated as
1425      if (condition || *ecode == OP_ALT)      creating two alternatives. If a THEN is encountered in the branch, it
1426        propagates out to the enclosing alternative (unless nested in a deeper set
1427        of alternatives, of course). */
1428    
1429        if (condition || ecode[-(1+LINK_SIZE)] == OP_ALT)
1430        {        {
1431        if (op != OP_SCOND)        if (op != OP_SCOND)
1432          {          {
         ecode += 1 + LINK_SIZE;  
1433          goto TAIL_RECURSE;          goto TAIL_RECURSE;
1434          }          }
1435    
1436        md->match_function_type = MATCH_CBEGROUP;        md->match_function_type = MATCH_CBEGROUP;
1437        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, eptrb, RM49);        RMATCH(eptr, ecode, offset_top, md, eptrb, RM49);
1438        RRETURN(rrc);        RRETURN(rrc);
1439        }        }
1440    
# Line 1441  for (;;) Line 1442  for (;;)
1442    
1443      else      else
1444        {        {
       ecode += 1 + LINK_SIZE;  
1445        }        }
1446      break;      break;
1447    
# Line 1450  for (;;) Line 1450  for (;;)
1450      to close any currently open capturing brackets. */      to close any currently open capturing brackets. */
1451    
1452      case OP_CLOSE:      case OP_CLOSE:
1453      number = GET2(ecode, 1);      number = GET2(ecode, 1);   /* Must be less than 65536 */
1454      offset = number << 1;      offset = number << 1;
1455    
1456  #ifdef PCRE_DEBUG  #ifdef PCRE_DEBUG
# Line 1458  for (;;) Line 1458  for (;;)
1458        printf("\n");        printf("\n");
1459  #endif  #endif
1460    
1461      md->capture_last = number;      md->capture_last = (md->capture_last & OVFLMASK) | number;
1462      if (offset >= md->offset_max) md->offset_overflow = TRUE; else      if (offset >= md->offset_max) md->capture_last |= OVFLBIT; else
1463        {        {
1464        md->offset_vector[offset] =        md->offset_vector[offset] =
1465          md->offset_vector[md->offset_end - number];          md->offset_vector[md->offset_end - number];
1466        md->offset_vector[offset+1] = (int)(eptr - md->start_subject);        md->offset_vector[offset+1] = (int)(eptr - md->start_subject);
1467        if (offset_top <= offset) offset_top = offset + 2;        if (offset_top <= offset) offset_top = offset + 2;
1468        }        }
1469      ecode += 3;      ecode += 1 + IMM2_SIZE;
1470      break;      break;
1471    
1472    
# Line 1513  for (;;) Line 1513  for (;;)
1513    
1514      case OP_ASSERT:      case OP_ASSERT:
1515      case OP_ASSERTBACK:      case OP_ASSERTBACK:
1516        save_mark = md->mark;
1517      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1518        {        {
1519        condassert = TRUE;        condassert = TRUE;
# Line 1520  for (;;) Line 1521  for (;;)
1521        }        }
1522      else condassert = FALSE;      else condassert = FALSE;
1523    
1524        /* Loop for each branch */
1525    
1526      do      do
1527        {        {
1528        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM4);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM4);
1529    
1530          /* A match means that the assertion is true; break out of the loop
1531          that matches its alternatives. */
1532    
1533        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1534          {          {
1535          mstart = md->start_match_ptr;   /* In case \K reset it */          mstart = md->start_match_ptr;   /* In case \K reset it */
1536          break;          break;
1537          }          }
1538    
1539        /* PCRE does not allow THEN to escape beyond an assertion; it is treated        /* If not matched, restore the previous mark setting. */
1540        as NOMATCH. */  
1541          md->mark = save_mark;
1542    
1543          /* See comment in the code for capturing groups above about handling
1544          THEN. */
1545    
1546          if (rrc == MATCH_THEN)
1547            {
1548            next = ecode + GET(ecode,1);
1549            if (md->start_match_ptr < next &&
1550                (*ecode == OP_ALT || *next == OP_ALT))
1551              rrc = MATCH_NOMATCH;
1552            }
1553    
1554          /* Anything other than NOMATCH causes the entire assertion to fail,
1555          passing back the return code. This includes COMMIT, SKIP, PRUNE and an
1556          uncaptured THEN, which means they take their normal effect. This
1557          consistent approach does not always have exactly the same effect as in
1558          Perl. */
1559    
1560        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1561        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1562        }        }
1563      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);   /* Continue for next alternative */
1564    
1565        /* If we have tried all the alternative branches, the assertion has
1566        failed. If not, we broke out after a match. */
1567    
1568      if (*ecode == OP_KET) RRETURN(MATCH_NOMATCH);      if (*ecode == OP_KET) RRETURN(MATCH_NOMATCH);
1569    
# Line 1543  for (;;) Line 1571  for (;;)
1571    
1572      if (condassert) RRETURN(MATCH_MATCH);      if (condassert) RRETURN(MATCH_MATCH);
1573    
1574      /* Continue from after the assertion, updating the offsets high water      /* Continue from after a successful assertion, updating the offsets high
1575      mark, since extracts may have been taken during the assertion. */      water mark, since extracts may have been taken during the assertion. */
1576    
1577      do ecode += GET(ecode,1); while (*ecode == OP_ALT);      do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1578      ecode += 1 + LINK_SIZE;      ecode += 1 + LINK_SIZE;
1579      offset_top = md->end_offset_top;      offset_top = md->end_offset_top;
1580      continue;      continue;
1581    
1582      /* Negative assertion: all branches must fail to match. Encountering SKIP,      /* Negative assertion: all branches must fail to match for the assertion to
1583      PRUNE, or COMMIT means we must assume failure without checking subsequent      succeed. */
     branches. */  
1584    
1585      case OP_ASSERT_NOT:      case OP_ASSERT_NOT:
1586      case OP_ASSERTBACK_NOT:      case OP_ASSERTBACK_NOT:
1587        save_mark = md->mark;
1588      if (md->match_function_type == MATCH_CONDASSERT)      if (md->match_function_type == MATCH_CONDASSERT)
1589        {        {
1590        condassert = TRUE;        condassert = TRUE;
# Line 1564  for (;;) Line 1592  for (;;)
1592        }        }
1593      else condassert = FALSE;      else condassert = FALSE;
1594    
1595        /* Loop for each alternative branch. */
1596    
1597      do      do
1598        {        {
1599        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, NULL, RM5);
1600        if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT) RRETURN(MATCH_NOMATCH);        md->mark = save_mark;   /* Always restore the mark setting */
1601        if (rrc == MATCH_SKIP || rrc == MATCH_PRUNE || rrc == MATCH_COMMIT)  
1602          switch(rrc)
1603          {          {
1604          do ecode += GET(ecode,1); while (*ecode == OP_ALT);          case MATCH_MATCH:            /* A successful match means */
1605            case MATCH_ACCEPT:           /* the assertion has failed. */
1606            RRETURN(MATCH_NOMATCH);
1607    
1608            case MATCH_NOMATCH:          /* Carry on with next branch */
1609          break;          break;
1610    
1611            /* See comment in the code for capturing groups above about handling
1612            THEN. */
1613    
1614            case MATCH_THEN:
1615            next = ecode + GET(ecode,1);
1616            if (md->start_match_ptr < next &&
1617                (*ecode == OP_ALT || *next == OP_ALT))
1618              {
1619              rrc = MATCH_NOMATCH;
1620              break;
1621              }
1622            /* Otherwise fall through. */
1623    
1624            /* COMMIT, SKIP, PRUNE, and an uncaptured THEN cause the whole
1625            assertion to fail to match, without considering any more alternatives.
1626            Failing to match means the assertion is true. This is a consistent
1627            approach, but does not always have the same effect as in Perl. */
1628    
1629            case MATCH_COMMIT:
1630            case MATCH_SKIP:
1631            case MATCH_SKIP_ARG:
1632            case MATCH_PRUNE:
1633            do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1634            goto NEG_ASSERT_TRUE;   /* Break out of alternation loop */
1635    
1636            /* Anything else is an error */
1637    
1638            default:
1639            RRETURN(rrc);
1640          }          }
1641    
1642        /* PCRE does not allow THEN to escape beyond an assertion; it is treated        /* Continue with next branch */
       as NOMATCH. */  
1643    
       if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);  
1644        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1645        }        }
1646      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
1647    
1648      if (condassert) RRETURN(MATCH_MATCH);  /* Condition assertion */      /* All branches in the assertion failed to match. */
1649    
1650      ecode += 1 + LINK_SIZE;      NEG_ASSERT_TRUE:
1651        if (condassert) RRETURN(MATCH_MATCH);  /* Condition assertion */
1652        ecode += 1 + LINK_SIZE;                /* Continue with current branch */
1653      continue;      continue;
1654    
1655      /* Move the subject pointer back. This occurs only at the start of      /* Move the subject pointer back. This occurs only at the start of
# Line 1593  for (;;) Line 1658  for (;;)
1658      back a number of characters, not bytes. */      back a number of characters, not bytes. */
1659    
1660      case OP_REVERSE:      case OP_REVERSE:
1661  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
1662      if (utf8)      if (utf)
1663        {        {
1664        i = GET(ecode, 1);        i = GET(ecode, 1);
1665        while (i-- > 0)        while (i-- > 0)
# Line 1625  for (;;) Line 1690  for (;;)
1690      function is able to force a failure. */      function is able to force a failure. */
1691    
1692      case OP_CALLOUT:      case OP_CALLOUT:
1693      if (pcre_callout != NULL)      if (PUBL(callout) != NULL)
1694        {        {
1695        pcre_callout_block cb;        PUBL(callout_block) cb;
1696        cb.version          = 2;   /* Version 1 of the callout block */        cb.version          = 2;   /* Version 1 of the callout block */
1697        cb.callout_number   = ecode[1];        cb.callout_number   = ecode[1];
1698        cb.offset_vector    = md->offset_vector;        cb.offset_vector    = md->offset_vector;
1699    #if defined COMPILE_PCRE8
1700        cb.subject          = (PCRE_SPTR)md->start_subject;        cb.subject          = (PCRE_SPTR)md->start_subject;
1701    #elif defined COMPILE_PCRE16
1702          cb.subject          = (PCRE_SPTR16)md->start_subject;
1703    #elif defined COMPILE_PCRE32
1704          cb.subject          = (PCRE_SPTR32)md->start_subject;
1705    #endif
1706        cb.subject_length   = (int)(md->end_subject - md->start_subject);        cb.subject_length   = (int)(md->end_subject - md->start_subject);
1707        cb.start_match      = (int)(mstart - md->start_subject);        cb.start_match      = (int)(mstart - md->start_subject);
1708        cb.current_position = (int)(eptr - md->start_subject);        cb.current_position = (int)(eptr - md->start_subject);
1709        cb.pattern_position = GET(ecode, 2);        cb.pattern_position = GET(ecode, 2);
1710        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);
1711        cb.capture_top      = offset_top/2;        cb.capture_top      = offset_top/2;
1712        cb.capture_last     = md->capture_last;        cb.capture_last     = md->capture_last & CAPLMASK;
1713          /* Internal change requires this for API compatibility. */
1714          if (cb.capture_last == 0) cb.capture_last = -1;
1715        cb.callout_data     = md->callout_data;        cb.callout_data     = md->callout_data;
1716        cb.mark             = md->nomatch_mark;        cb.mark             = md->nomatch_mark;
1717        if ((rrc = (*pcre_callout)(&cb)) > 0) RRETURN(MATCH_NOMATCH);        if ((rrc = (*PUBL(callout))(&cb)) > 0) RRETURN(MATCH_NOMATCH);
1718        if (rrc < 0) RRETURN(rrc);        if (rrc < 0) RRETURN(rrc);
1719        }        }
1720      ecode += 2 + 2*LINK_SIZE;      ecode += 2 + 2*LINK_SIZE;
# Line 1667  for (;;) Line 1740  for (;;)
1740      case OP_RECURSE:      case OP_RECURSE:
1741        {        {
1742        recursion_info *ri;        recursion_info *ri;
1743        int recno;        unsigned int recno;
1744    
1745        callpat = md->start_code + GET(ecode, 1);        callpat = md->start_code + GET(ecode, 1);
1746        recno = (callpat == md->start_code)? 0 :        recno = (callpat == md->start_code)? 0 :
# Line 1684  for (;;) Line 1757  for (;;)
1757        /* Add to "recursing stack" */        /* Add to "recursing stack" */
1758    
1759        new_recursive.group_num = recno;        new_recursive.group_num = recno;
1760          new_recursive.saved_capture_last = md->capture_last;
1761        new_recursive.subject_position = eptr;        new_recursive.subject_position = eptr;
1762        new_recursive.prevrec = md->recursive;        new_recursive.prevrec = md->recursive;
1763        md->recursive = &new_recursive;        md->recursive = &new_recursive;
# Line 1700  for (;;) Line 1774  for (;;)
1774        else        else
1775          {          {
1776          new_recursive.offset_save =          new_recursive.offset_save =
1777            (int *)(pcre_malloc)(new_recursive.saved_max * sizeof(int));            (int *)(PUBL(malloc))(new_recursive.saved_max * sizeof(int));
1778          if (new_recursive.offset_save == NULL) RRETURN(PCRE_ERROR_NOMEMORY);          if (new_recursive.offset_save == NULL) RRETURN(PCRE_ERROR_NOMEMORY);
1779          }          }
1780        memcpy(new_recursive.offset_save, md->offset_vector,        memcpy(new_recursive.offset_save, md->offset_vector,
1781              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1782    
1783        /* OK, now we can do the recursion. After processing each alternative,        /* OK, now we can do the recursion. After processing each alternative,
1784        restore the offset data. If there were nested recursions, md->recursive        restore the offset data and the last captured value. If there were nested
1785        might be changed, so reset it before looping. */        recursions, md->recursive might be changed, so reset it before looping.
1786          */
1787    
1788        DPRINTF(("Recursing into group %d\n", new_recursive.group_num));        DPRINTF(("Recursing into group %d\n", new_recursive.group_num));
1789        cbegroup = (*callpat >= OP_SBRA);        cbegroup = (*callpat >= OP_SBRA);
1790        do        do
1791          {          {
1792          if (cbegroup) md->match_function_type = MATCH_CBEGROUP;          if (cbegroup) md->match_function_type = MATCH_CBEGROUP;
1793          RMATCH(eptr, callpat + _pcre_OP_lengths[*callpat], offset_top,          RMATCH(eptr, callpat + PRIV(OP_lengths)[*callpat], offset_top,
1794            md, eptrb, RM6);            md, eptrb, RM6);
1795          memcpy(md->offset_vector, new_recursive.offset_save,          memcpy(md->offset_vector, new_recursive.offset_save,
1796              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1797            md->capture_last = new_recursive.saved_capture_last;
1798          md->recursive = new_recursive.prevrec;          md->recursive = new_recursive.prevrec;
1799          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)          if (rrc == MATCH_MATCH || rrc == MATCH_ACCEPT)
1800            {            {
1801            DPRINTF(("Recursion matched\n"));            DPRINTF(("Recursion matched\n"));
1802            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
1803              (pcre_free)(new_recursive.offset_save);              (PUBL(free))(new_recursive.offset_save);
1804    
1805            /* Set where we got to in the subject, and reset the start in case            /* Set where we got to in the subject, and reset the start in case
1806            it was changed by \K. This *is* propagated back out of a recursion,            it was changed by \K. This *is* propagated back out of a recursion,
# Line 1735  for (;;) Line 1811  for (;;)
1811            goto RECURSION_MATCHED;        /* Exit loop; end processing */            goto RECURSION_MATCHED;        /* Exit loop; end processing */
1812            }            }
1813    
1814          /* PCRE does not allow THEN to escape beyond a recursion; it is treated          /* PCRE does not allow THEN, SKIP, PRUNE or COMMIT to escape beyond a
1815          as NOMATCH. */          recursion; they cause a NOMATCH for the entire recursion. These codes
1816            are defined in a range that can be tested for. */
1817    
1818          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)          if (rrc >= MATCH_BACKTRACK_MIN && rrc <= MATCH_BACKTRACK_MAX)
1819              RRETURN(MATCH_NOMATCH);
1820    
1821            /* Any return code other than NOMATCH is an error. */
1822    
1823            if (rrc != MATCH_NOMATCH)
1824            {            {
1825            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1826            if (new_recursive.offset_save != stacksave)            if (new_recursive.offset_save != stacksave)
1827              (pcre_free)(new_recursive.offset_save);              (PUBL(free))(new_recursive.offset_save);
1828            RRETURN(rrc);            RRETURN(rrc);
1829            }            }
1830    
# Line 1754  for (;;) Line 1836  for (;;)
1836        DPRINTF(("Recursion didn't match\n"));        DPRINTF(("Recursion didn't match\n"));
1837        md->recursive = new_recursive.prevrec;        md->recursive = new_recursive.prevrec;
1838        if (new_recursive.offset_save != stacksave)        if (new_recursive.offset_save != stacksave)
1839          (pcre_free)(new_recursive.offset_save);          (PUBL(free))(new_recursive.offset_save);
1840        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
1841        }        }
1842    
# Line 1868  for (;;) Line 1950  for (;;)
1950    
1951        /* Deal with capturing */        /* Deal with capturing */
1952    
1953        md->capture_last = number;        md->capture_last = (md->capture_last & OVFLMASK) | number;
1954        if (offset >= md->offset_max) md->offset_overflow = TRUE; else        if (offset >= md->offset_max) md->capture_last |= OVFLBIT; else
1955          {          {
1956          /* If offset is greater than offset_top, it means that we are          /* If offset is greater than offset_top, it means that we are
1957          "skipping" a capturing group, and that group's offsets must be marked          "skipping" a capturing group, and that group's offsets must be marked
# Line 1949  for (;;) Line 2031  for (;;)
2031          }          }
2032        if (*prev >= OP_SBRA)    /* Could match an empty string */        if (*prev >= OP_SBRA)    /* Could match an empty string */
2033          {          {
         md->match_function_type = MATCH_CBEGROUP;  
2034          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);          RMATCH(eptr, prev, offset_top, md, eptrb, RM50);
2035          RRETURN(rrc);          RRETURN(rrc);
2036          }          }
# Line 1958  for (;;) Line 2039  for (;;)
2039        }        }
2040      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
2041        {        {
       if (*prev >= OP_SBRA) md->match_function_type = MATCH_CBEGROUP;  
2042        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);        RMATCH(eptr, prev, offset_top, md, eptrb, RM13);
2043        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;        if (rrc == MATCH_ONCE && md->once_target == prev) rrc = MATCH_NOMATCH;
2044        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
# Line 2015  for (;;) Line 2095  for (;;)
2095    
2096      case OP_DOLLM:      case OP_DOLLM:
2097      if (eptr < md->end_subject)      if (eptr < md->end_subject)
2098        { if (!IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH); }        {
2099          if (!IS_NEWLINE(eptr))
2100            {
2101            if (md->partial != 0 &&
2102                eptr + 1 >= md->end_subject &&
2103                NLBLOCK->nltype == NLTYPE_FIXED &&
2104                NLBLOCK->nllen == 2 &&
2105                RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2106              {
2107              md->hitend = TRUE;
2108              if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2109              }
2110            RRETURN(MATCH_NOMATCH);
2111            }
2112          }
2113      else      else
2114        {        {
2115        if (md->noteol) RRETURN(MATCH_NOMATCH);        if (md->noteol) RRETURN(MATCH_NOMATCH);
# Line 2047  for (;;) Line 2141  for (;;)
2141      ASSERT_NL_OR_EOS:      ASSERT_NL_OR_EOS:
2142      if (eptr < md->end_subject &&      if (eptr < md->end_subject &&
2143          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))
2144          {
2145          if (md->partial != 0 &&
2146              eptr + 1 >= md->end_subject &&
2147              NLBLOCK->nltype == NLTYPE_FIXED &&
2148              NLBLOCK->nllen == 2 &&
2149              RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2150            {
2151            md->hitend = TRUE;
2152            if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2153            }
2154        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2155          }
2156    
2157      /* Either at end of string or \n before end. */      /* Either at end of string or \n before end. */
2158    
# Line 2066  for (;;) Line 2171  for (;;)
2171        be "non-word" characters. Remember the earliest consulted character for        be "non-word" characters. Remember the earliest consulted character for
2172        partial matching. */        partial matching. */
2173    
2174  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2175        if (utf8)        if (utf)
2176          {          {
2177          /* Get status of previous character */          /* Get status of previous character */
2178    
2179          if (eptr == md->start_subject) prev_is_word = FALSE; else          if (eptr == md->start_subject) prev_is_word = FALSE; else
2180            {            {
2181            USPTR lastptr = eptr - 1;            PCRE_PUCHAR lastptr = eptr - 1;
2182            while((*lastptr & 0xc0) == 0x80) lastptr--;            BACKCHAR(lastptr);
2183            if (lastptr < md->start_used_ptr) md->start_used_ptr = lastptr;            if (lastptr < md->start_used_ptr) md->start_used_ptr = lastptr;
2184            GETCHAR(c, lastptr);            GETCHAR(c, lastptr);
2185  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
# Line 2139  for (;;) Line 2244  for (;;)
2244              }              }
2245            else            else
2246  #endif  #endif
2247            prev_is_word = ((md->ctypes[eptr[-1]] & ctype_word) != 0);            prev_is_word = MAX_255(eptr[-1])
2248                && ((md->ctypes[eptr[-1]] & ctype_word) != 0);
2249            }            }
2250    
2251          /* Get status of next character */          /* Get status of next character */
# Line 2162  for (;;) Line 2268  for (;;)
2268            }            }
2269          else          else
2270  #endif  #endif
2271          cur_is_word = ((md->ctypes[*eptr] & ctype_word) != 0);          cur_is_word = MAX_255(*eptr)
2272              && ((md->ctypes[*eptr] & ctype_word) != 0);
2273          }          }
2274    
2275        /* Now see if the situation is what we want */        /* Now see if the situation is what we want */
# Line 2173  for (;;) Line 2280  for (;;)
2280        }        }
2281      break;      break;
2282    
2283      /* Match a single character type; inline for speed */      /* Match any single character type except newline; have to take care with
2284        CRLF newlines and partial matching. */
2285    
2286      case OP_ANY:      case OP_ANY:
2287      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
2288        if (md->partial != 0 &&
2289            eptr + 1 >= md->end_subject &&
2290            NLBLOCK->nltype == NLTYPE_FIXED &&
2291            NLBLOCK->nllen == 2 &&
2292            RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
2293          {
2294          md->hitend = TRUE;
2295          if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2296          }
2297    
2298      /* Fall through */      /* Fall through */
2299    
2300        /* Match any single character whatsoever. */
2301    
2302      case OP_ALLANY:      case OP_ALLANY:
2303      if (eptr >= md->end_subject)   /* DO NOT merge the eptr++ here; it must */      if (eptr >= md->end_subject)   /* DO NOT merge the eptr++ here; it must */
2304        {                            /* not be updated before SCHECK_PARTIAL. */        {                            /* not be updated before SCHECK_PARTIAL. */
# Line 2186  for (;;) Line 2306  for (;;)
2306        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2307        }        }
2308      eptr++;      eptr++;
2309      if (utf8) while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  #ifdef SUPPORT_UTF
2310        if (utf) ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
2311    #endif
2312      ecode++;      ecode++;
2313      break;      break;
2314    
# Line 2211  for (;;) Line 2333  for (;;)
2333        }        }
2334      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2335      if (      if (
2336  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2337         c < 256 &&         c < 256 &&
2338  #endif  #endif
2339         (md->ctypes[c] & ctype_digit) != 0         (md->ctypes[c] & ctype_digit) != 0
# Line 2228  for (;;) Line 2350  for (;;)
2350        }        }
2351      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2352      if (      if (
2353  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2354         c >= 256 ||         c > 255 ||
2355  #endif  #endif
2356         (md->ctypes[c] & ctype_digit) == 0         (md->ctypes[c] & ctype_digit) == 0
2357         )         )
# Line 2245  for (;;) Line 2367  for (;;)
2367        }        }
2368      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2369      if (      if (
2370  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2371         c < 256 &&         c < 256 &&
2372  #endif  #endif
2373         (md->ctypes[c] & ctype_space) != 0         (md->ctypes[c] & ctype_space) != 0
# Line 2262  for (;;) Line 2384  for (;;)
2384        }        }
2385      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2386      if (      if (
2387  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2388         c >= 256 ||         c > 255 ||
2389  #endif  #endif
2390         (md->ctypes[c] & ctype_space) == 0         (md->ctypes[c] & ctype_space) == 0
2391         )         )
# Line 2279  for (;;) Line 2401  for (;;)
2401        }        }
2402      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2403      if (      if (
2404  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2405         c < 256 &&         c < 256 &&
2406  #endif  #endif
2407         (md->ctypes[c] & ctype_word) != 0         (md->ctypes[c] & ctype_word) != 0
# Line 2296  for (;;) Line 2418  for (;;)
2418        }        }
2419      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2420      if (      if (
2421  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
2422         c >= 256 ||         c > 255 ||
2423  #endif  #endif
2424         (md->ctypes[c] & ctype_word) == 0         (md->ctypes[c] & ctype_word) == 0
2425         )         )
# Line 2316  for (;;) Line 2438  for (;;)
2438        {        {
2439        default: RRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
2440    
2441        case 0x000d:        case CHAR_CR:
2442        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr >= md->end_subject)
2443            {
2444            SCHECK_PARTIAL();
2445            }
2446          else if (RAWUCHARTEST(eptr) == CHAR_LF) eptr++;
2447        break;        break;
2448    
2449        case 0x000a:        case CHAR_LF:
2450        break;        break;
2451    
2452        case 0x000b:        case CHAR_VT:
2453        case 0x000c:        case CHAR_FF:
2454        case 0x0085:        case CHAR_NEL:
2455    #ifndef EBCDIC
2456        case 0x2028:        case 0x2028:
2457        case 0x2029:        case 0x2029:
2458    #endif  /* Not EBCDIC */
2459        if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);        if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
2460        break;        break;
2461        }        }
# Line 2343  for (;;) Line 2471  for (;;)
2471      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2472      switch(c)      switch(c)
2473        {        {
2474          HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
2475        default: break;        default: break;
       case 0x09:      /* HT */  
       case 0x20:      /* SPACE */  
       case 0xa0:      /* NBSP */  
       case 0x1680:    /* OGHAM SPACE MARK */  
       case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
       case 0x2000:    /* EN QUAD */  
       case 0x2001:    /* EM QUAD */  
       case 0x2002:    /* EN SPACE */  
       case 0x2003:    /* EM SPACE */  
       case 0x2004:    /* THREE-PER-EM SPACE */  
       case 0x2005:    /* FOUR-PER-EM SPACE */  
       case 0x2006:    /* SIX-PER-EM SPACE */  
       case 0x2007:    /* FIGURE SPACE */  
       case 0x2008:    /* PUNCTUATION SPACE */  
       case 0x2009:    /* THIN SPACE */  
       case 0x200A:    /* HAIR SPACE */  
       case 0x202f:    /* NARROW NO-BREAK SPACE */  
       case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
       case 0x3000:    /* IDEOGRAPHIC SPACE */  
       RRETURN(MATCH_NOMATCH);  
2476        }        }
2477      ecode++;      ecode++;
2478      break;      break;
# Line 2377  for (;;) Line 2486  for (;;)
2486      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2487      switch(c)      switch(c)
2488        {        {
2489          HSPACE_CASES: break;  /* Byte and multibyte cases */
2490        default: RRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
       case 0x09:      /* HT */  
       case 0x20:      /* SPACE */  
       case 0xa0:      /* NBSP */  
       case 0x1680:    /* OGHAM SPACE MARK */  
       case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */  
       case 0x2000:    /* EN QUAD */  
       case 0x2001:    /* EM QUAD */  
       case 0x2002:    /* EN SPACE */  
       case 0x2003:    /* EM SPACE */  
       case 0x2004:    /* THREE-PER-EM SPACE */  
       case 0x2005:    /* FOUR-PER-EM SPACE */  
       case 0x2006:    /* SIX-PER-EM SPACE */  
       case 0x2007:    /* FIGURE SPACE */  
       case 0x2008:    /* PUNCTUATION SPACE */  
       case 0x2009:    /* THIN SPACE */  
       case 0x200A:    /* HAIR SPACE */  
       case 0x202f:    /* NARROW NO-BREAK SPACE */  
       case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */  
       case 0x3000:    /* IDEOGRAPHIC SPACE */  
       break;  
2491        }        }
2492      ecode++;      ecode++;
2493      break;      break;
# Line 2411  for (;;) Line 2501  for (;;)
2501      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2502      switch(c)      switch(c)
2503        {        {
2504          VSPACE_CASES: RRETURN(MATCH_NOMATCH);
2505        default: break;        default: break;
       case 0x0a:      /* LF */  
       case 0x0b:      /* VT */  
       case 0x0c:      /* FF */  
       case 0x0d:      /* CR */  
       case 0x85:      /* NEL */  
       case 0x2028:    /* LINE SEPARATOR */  
       case 0x2029:    /* PARAGRAPH SEPARATOR */  
       RRETURN(MATCH_NOMATCH);  
2506        }        }
2507      ecode++;      ecode++;
2508      break;      break;
# Line 2433  for (;;) Line 2516  for (;;)
2516      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2517      switch(c)      switch(c)
2518        {        {
2519          VSPACE_CASES: break;
2520        default: RRETURN(MATCH_NOMATCH);        default: RRETURN(MATCH_NOMATCH);
       case 0x0a:      /* LF */  
       case 0x0b:      /* VT */  
       case 0x0c:      /* FF */  
       case 0x0d:      /* CR */  
       case 0x85:      /* NEL */  
       case 0x2028:    /* LINE SEPARATOR */  
       case 0x2029:    /* PARAGRAPH SEPARATOR */  
       break;  
2521        }        }
2522      ecode++;      ecode++;
2523      break;      break;
# Line 2459  for (;;) Line 2535  for (;;)
2535        }        }
2536      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2537        {        {
2538          const pcre_uint32 *cp;
2539        const ucd_record *prop = GET_UCD(c);        const ucd_record *prop = GET_UCD(c);
2540    
2541        switch(ecode[1])        switch(ecode[1])
# Line 2475  for (;;) Line 2552  for (;;)
2552          break;          break;
2553    
2554          case PT_GC:          case PT_GC:
2555          if ((ecode[2] != _pcre_ucp_gentype[prop->chartype]) == (op == OP_PROP))          if ((ecode[2] != PRIV(ucp_gentype)[prop->chartype]) == (op == OP_PROP))
2556            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2557          break;          break;
2558    
# Line 2492  for (;;) Line 2569  for (;;)
2569          /* These are specials */          /* These are specials */
2570    
2571          case PT_ALNUM:          case PT_ALNUM:
2572          if ((_pcre_ucp_gentype[prop->chartype] == ucp_L ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2573               _pcre_ucp_gentype[prop->chartype] == ucp_N) == (op == OP_NOTPROP))               PRIV(ucp_gentype)[prop->chartype] == ucp_N) == (op == OP_NOTPROP))
2574            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2575          break;          break;
2576    
2577            /* Perl space used to exclude VT, but from Perl 5.18 it is included,
2578            which means that Perl space and POSIX space are now identical. PCRE
2579            was changed at release 8.34. */
2580    
2581          case PT_SPACE:    /* Perl space */          case PT_SPACE:    /* Perl space */
         if ((_pcre_ucp_gentype[prop->chartype] == ucp_Z ||  
              c == CHAR_HT || c == CHAR_NL || c == CHAR_FF || c == CHAR_CR)  
                == (op == OP_NOTPROP))  
           RRETURN(MATCH_NOMATCH);  
         break;  
   
2582          case PT_PXSPACE:  /* POSIX space */          case PT_PXSPACE:  /* POSIX space */
2583          if ((_pcre_ucp_gentype[prop->chartype] == ucp_Z ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z ||
2584               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||               c == CHAR_HT || c == CHAR_NL || c == CHAR_VT ||
2585               c == CHAR_FF || c == CHAR_CR)               c == CHAR_FF || c == CHAR_CR)
2586                 == (op == OP_NOTPROP))                 == (op == OP_NOTPROP))
# Line 2513  for (;;) Line 2588  for (;;)
2588          break;          break;
2589    
2590          case PT_WORD:          case PT_WORD:
2591          if ((_pcre_ucp_gentype[prop->chartype] == ucp_L ||          if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L ||
2592               _pcre_ucp_gentype[prop->chartype] == ucp_N ||               PRIV(ucp_gentype)[prop->chartype] == ucp_N ||
2593               c == CHAR_UNDERSCORE) == (op == OP_NOTPROP))               c == CHAR_UNDERSCORE) == (op == OP_NOTPROP))
2594            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2595          break;          break;
2596    
2597            case PT_CLIST:
2598            cp = PRIV(ucd_caseless_sets) + ecode[2];
2599            for (;;)
2600              {
2601              if (c < *cp)
2602                { if (op == OP_PROP) { RRETURN(MATCH_NOMATCH); } else break; }
2603              if (c == *cp++)
2604                { if (op == OP_PROP) break; else { RRETURN(MATCH_NOMATCH); } }
2605              }
2606            break;
2607    
2608            case PT_UCNC:
2609            if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
2610                 c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
2611                 c >= 0xe000) == (op == OP_NOTPROP))
2612              RRETURN(MATCH_NOMATCH);
2613            break;
2614    
2615          /* This should never occur */          /* This should never occur */
2616    
2617          default:          default:
# Line 2538  for (;;) Line 2631  for (;;)
2631        SCHECK_PARTIAL();        SCHECK_PARTIAL();
2632        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
2633        }        }
2634      GETCHARINCTEST(c, eptr);      else
     if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
     while (eptr < md->end_subject)  
2635        {        {
2636        int len = 1;        int lgb, rgb;
2637        if (!utf8) c = *eptr; else { GETCHARLEN(c, eptr, len); }        GETCHARINCTEST(c, eptr);
2638        if (UCD_CATEGORY(c) != ucp_M) break;        lgb = UCD_GRAPHBREAK(c);
2639        eptr += len;        while (eptr < md->end_subject)
2640            {
2641            int len = 1;
2642            if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
2643            rgb = UCD_GRAPHBREAK(c);
2644            if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
2645            lgb = rgb;
2646            eptr += len;
2647            }
2648        }        }
2649        CHECK_PARTIAL();
2650      ecode++;      ecode++;
2651      break;      break;
2652  #endif  #endif  /* SUPPORT_UCP */
2653    
2654    
2655      /* 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 2558  for (;;) Line 2658  for (;;)
2658      similar code to character type repeats - written out again for speed.      similar code to character type repeats - written out again for speed.
2659      However, if the referenced string is the empty string, always treat      However, if the referenced string is the empty string, always treat
2660      it as matched, any number of times (otherwise there could be infinite      it as matched, any number of times (otherwise there could be infinite
2661      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 += 3;  
   
     /* If the reference is unset, there are two possibilities:  
2662    
2663      (a) In the default, Perl-compatible state, set the length negative;      (a) In the default, Perl-compatible state, set the length negative;
2664      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 2576  for (;;) Line 2668  for (;;)
2668      so that the back reference matches an empty string.      so that the back reference matches an empty string.
2669    
2670      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
2671      referenced subpattern. */      referenced subpattern.
2672    
2673        The OP_REF and OP_REFI opcodes are used for a reference to a numbered group
2674        or to a non-duplicated named group. For a duplicated named group, OP_DNREF
2675        and OP_DNREFI are used. In this case we must scan the list of groups to
2676        which the name refers, and use the first one that is set. */
2677    
2678        case OP_DNREF:
2679        case OP_DNREFI:
2680        caseless = op == OP_DNREFI;
2681          {
2682          int count = GET2(ecode, 1+IMM2_SIZE);
2683          pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
2684          ecode += 1 + 2*IMM2_SIZE;
2685    
2686          while (count-- > 0)
2687            {
2688            offset = GET2(slot, 0) << 1;
2689            if (offset < offset_top && md->offset_vector[offset] >= 0) break;
2690            slot += md->name_entry_size;
2691            }
2692          if (count < 0)
2693            length = (md->jscript_compat)? 0 : -1;
2694          else
2695            length = md->offset_vector[offset+1] - md->offset_vector[offset];
2696          }
2697        goto REF_REPEAT;
2698    
2699        case OP_REF:
2700        case OP_REFI:
2701        caseless = op == OP_REFI;
2702        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
2703        ecode += 1 + IMM2_SIZE;
2704      if (offset >= offset_top || md->offset_vector[offset] < 0)      if (offset >= offset_top || md->offset_vector[offset] < 0)
2705        length = (md->jscript_compat)? 0 : -1;        length = (md->jscript_compat)? 0 : -1;
2706      else      else
# Line 2585  for (;;) Line 2708  for (;;)
2708    
2709      /* Set up for repetition, or handle the non-repeated case */      /* Set up for repetition, or handle the non-repeated case */
2710    
2711        REF_REPEAT:
2712      switch (*ecode)      switch (*ecode)
2713        {        {
2714        case OP_CRSTAR:        case OP_CRSTAR:
# Line 2604  for (;;) Line 2728  for (;;)
2728        case OP_CRMINRANGE:        case OP_CRMINRANGE:
2729        minimize = (*ecode == OP_CRMINRANGE);        minimize = (*ecode == OP_CRMINRANGE);
2730        min = GET2(ecode, 1);        min = GET2(ecode, 1);
2731        max = GET2(ecode, 3);        max = GET2(ecode, 1 + IMM2_SIZE);
2732        if (max == 0) max = INT_MAX;        if (max == 0) max = INT_MAX;
2733        ecode += 5;        ecode += 1 + 2 * IMM2_SIZE;
2734        break;        break;
2735    
2736        default:               /* No repeat follows */        default:               /* No repeat follows */
2737        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((length = match_ref(offset, eptr, length, md, caseless)) < 0)
2738          {          {
2739            if (length == -2) eptr = md->end_subject;   /* Partial match */
2740          CHECK_PARTIAL();          CHECK_PARTIAL();
2741          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2742          }          }
# Line 2620  for (;;) Line 2745  for (;;)
2745        }        }
2746    
2747      /* Handle repeated back references. If the length of the reference is      /* Handle repeated back references. If the length of the reference is
2748      zero, just continue with the main loop. */      zero, just continue with the main loop. If the length is negative, it
2749        means the reference is unset in non-Java-compatible mode. If the minimum is
2750        zero, we can continue at the same level without recursion. For any other
2751        minimum, carrying on will result in NOMATCH. */
2752    
2753      if (length == 0) continue;      if (length == 0) continue;
2754        if (length < 0 && min == 0) continue;
2755    
2756      /* First, ensure the minimum number of matches are present. We get back      /* First, ensure the minimum number of matches are present. We get back
2757      the length of the reference string explicitly rather than passing the      the length of the reference string explicitly rather than passing the
# Line 2633  for (;;) Line 2762  for (;;)
2762        int slength;        int slength;
2763        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)        if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2764          {          {
2765            if (slength == -2) eptr = md->end_subject;   /* Partial match */
2766          CHECK_PARTIAL();          CHECK_PARTIAL();
2767          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2768          }          }
# Line 2656  for (;;) Line 2786  for (;;)
2786          if (fi >= max) RRETURN(MATCH_NOMATCH);          if (fi >= max) RRETURN(MATCH_NOMATCH);
2787          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2788            {            {
2789              if (slength == -2) eptr = md->end_subject;   /* Partial match */
2790            CHECK_PARTIAL();            CHECK_PARTIAL();
2791            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
2792            }            }
# Line 2674  for (;;) Line 2805  for (;;)
2805          int slength;          int slength;
2806          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)          if ((slength = match_ref(offset, eptr, length, md, caseless)) < 0)
2807            {            {
2808            CHECK_PARTIAL();            /* Can't use CHECK_PARTIAL because we don't want to update eptr in
2809              the soft partial matching case. */
2810    
2811              if (slength == -2 && md->partial != 0 &&
2812                  md->end_subject > md->start_used_ptr)
2813                {
2814                md->hitend = TRUE;
2815                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
2816                }
2817            break;            break;
2818            }            }
2819          eptr += slength;          eptr += slength;
2820          }          }
2821    
2822        while (eptr >= pp)        while (eptr >= pp)
2823          {          {
2824          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);          RMATCH(eptr, ecode, offset_top, md, eptrb, RM15);
# Line 2703  for (;;) Line 2843  for (;;)
2843      case OP_NCLASS:      case OP_NCLASS:
2844      case OP_CLASS:      case OP_CLASS:
2845        {        {
2846          /* The data variable is saved across frames, so the byte map needs to
2847          be stored there. */
2848    #define BYTE_MAP ((pcre_uint8 *)data)
2849        data = ecode + 1;                /* Save for matching */        data = ecode + 1;                /* Save for matching */
2850        ecode += 33;                     /* Advance past the item */        ecode += 1 + (32 / sizeof(pcre_uchar)); /* Advance past the item */
2851    
2852        switch (*ecode)        switch (*ecode)
2853          {          {
# Line 2725  for (;;) Line 2868  for (;;)
2868          case OP_CRMINRANGE:          case OP_CRMINRANGE:
2869          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
2870          min = GET2(ecode, 1);          min = GET2(ecode, 1);
2871          max = GET2(ecode, 3);          max = GET2(ecode, 1 + IMM2_SIZE);
2872          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
2873          ecode += 5;          ecode += 1 + 2 * IMM2_SIZE;
2874          break;          break;
2875    
2876          default:               /* No repeat follows */          default:               /* No repeat follows */
# Line 2737  for (;;) Line 2880  for (;;)
2880    
2881        /* First, ensure the minimum number of matches are present. */        /* First, ensure the minimum number of matches are present. */
2882    
2883  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2884        /* UTF-8 mode */        if (utf)
       if (utf8)  
2885          {          {
2886          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2887            {            {
# Line 2754  for (;;) Line 2896  for (;;)
2896              if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);              if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2897              }              }
2898            else            else
2899              {              if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
             if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);  
             }  
2900            }            }
2901          }          }
2902        else        else
2903  #endif  #endif
2904        /* Not UTF-8 mode */        /* Not UTF mode */
2905          {          {
2906          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2907            {            {
# Line 2771  for (;;) Line 2911  for (;;)
2911              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2912              }              }
2913            c = *eptr++;            c = *eptr++;
2914            if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);  #ifndef COMPILE_PCRE8
2915              if (c > 255)
2916                {
2917                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2918                }
2919              else
2920    #endif
2921                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2922            }            }
2923          }          }
2924    
# Line 2785  for (;;) Line 2932  for (;;)
2932    
2933        if (minimize)        if (minimize)
2934          {          {
2935  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2936          /* UTF-8 mode */          if (utf)
         if (utf8)  
2937            {            {
2938            for (fi = min;; fi++)            for (fi = min;; fi++)
2939              {              {
# Line 2805  for (;;) Line 2951  for (;;)
2951                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);                if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2952                }                }
2953              else              else
2954                {                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
               if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);  
               }  
2955              }              }
2956            }            }
2957          else          else
2958  #endif  #endif
2959          /* Not UTF-8 mode */          /* Not UTF mode */
2960            {            {
2961            for (fi = min;; fi++)            for (fi = min;; fi++)
2962              {              {
# Line 2825  for (;;) Line 2969  for (;;)
2969                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2970                }                }
2971              c = *eptr++;              c = *eptr++;
2972              if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);  #ifndef COMPILE_PCRE8
2973                if (c > 255)
2974                  {
2975                  if (op == OP_CLASS) RRETURN(MATCH_NOMATCH);
2976                  }
2977                else
2978    #endif
2979                  if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2980              }              }
2981            }            }
2982          /* Control never gets here */          /* Control never gets here */
# Line 2837  for (;;) Line 2988  for (;;)
2988          {          {
2989          pp = eptr;          pp = eptr;
2990    
2991  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2992          /* UTF-8 mode */          if (utf)
         if (utf8)  
2993            {            {
2994            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2995              {              {
# Line 2855  for (;;) Line 3005  for (;;)
3005                if (op == OP_CLASS) break;                if (op == OP_CLASS) break;
3006                }                }
3007              else              else
3008                {                if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
               if ((data[c/8] & (1 << (c&7))) == 0) break;  
               }  
3009              eptr += len;              eptr += len;
3010              }              }
3011            for (;;)            for (;;)
# Line 2870  for (;;) Line 3018  for (;;)
3018            }            }
3019          else          else
3020  #endif  #endif
3021            /* Not UTF-8 mode */            /* Not UTF mode */
3022            {            {
3023            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3024              {              {
# Line 2880  for (;;) Line 3028  for (;;)
3028                break;                break;
3029                }                }
3030              c = *eptr;              c = *eptr;
3031              if ((data[c/8] & (1 << (c&7))) == 0) break;  #ifndef COMPILE_PCRE8
3032                if (c > 255)
3033                  {
3034                  if (op == OP_CLASS) break;
3035                  }
3036                else
3037    #endif
3038                  if ((BYTE_MAP[c/8] & (1 << (c&7))) == 0) break;
3039              eptr++;              eptr++;
3040              }              }
3041            while (eptr >= pp)            while (eptr >= pp)
# Line 2893  for (;;) Line 3048  for (;;)
3048    
3049          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3050          }          }
3051    #undef BYTE_MAP
3052        }        }
3053      /* Control never gets here */      /* Control never gets here */
3054    
# Line 2901  for (;;) Line 3057  for (;;)
3057      when UTF-8 mode mode is supported. Nevertheless, we may not be in UTF-8      when UTF-8 mode mode is supported. Nevertheless, we may not be in UTF-8
3058      mode, because Unicode properties are supported in non-UTF-8 mode. */      mode, because Unicode properties are supported in non-UTF-8 mode. */
3059    
3060  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
3061      case OP_XCLASS:      case OP_XCLASS:
3062        {        {
3063        data = ecode + 1 + LINK_SIZE;                /* Save for matching */        data = ecode + 1 + LINK_SIZE;                /* Save for matching */
# Line 2926  for (;;) Line 3082  for (;;)
3082          case OP_CRMINRANGE:          case OP_CRMINRANGE:
3083          minimize = (*ecode == OP_CRMINRANGE);          minimize = (*ecode == OP_CRMINRANGE);
3084          min = GET2(ecode, 1);          min = GET2(ecode, 1);
3085          max = GET2(ecode, 3);          max = GET2(ecode, 1 + IMM2_SIZE);
3086          if (max == 0) max = INT_MAX;          if (max == 0) max = INT_MAX;
3087          ecode += 5;          ecode += 1 + 2 * IMM2_SIZE;
3088          break;          break;
3089    
3090          default:               /* No repeat follows */          default:               /* No repeat follows */
# Line 2946  for (;;) Line 3102  for (;;)
3102            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3103            }            }
3104          GETCHARINCTEST(c, eptr);          GETCHARINCTEST(c, eptr);
3105          if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);          if (!PRIV(xclass)(c, data, utf)) RRETURN(MATCH_NOMATCH);
3106          }          }
3107    
3108        /* If max == min we can continue with the main loop without the        /* If max == min we can continue with the main loop without the
# Line 2970  for (;;) Line 3126  for (;;)
3126              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3127              }              }
3128            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3129            if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);            if (!PRIV(xclass)(c, data, utf)) RRETURN(MATCH_NOMATCH);
3130            }            }
3131          /* Control never gets here */          /* Control never gets here */
3132          }          }
# Line 2988  for (;;) Line 3144  for (;;)
3144              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3145              break;              break;
3146              }              }
3147    #ifdef SUPPORT_UTF
3148            GETCHARLENTEST(c, eptr, len);            GETCHARLENTEST(c, eptr, len);
3149            if (!_pcre_xclass(c, data)) break;  #else
3150              c = *eptr;
3151    #endif
3152              if (!PRIV(xclass)(c, data, utf)) break;
3153            eptr += len;            eptr += len;
3154            }            }
3155          for(;;)          for(;;)
# Line 2997  for (;;) Line 3157  for (;;)
3157            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM21);
3158            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3159            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3160            if (utf8) BACKCHAR(eptr);  #ifdef SUPPORT_UTF
3161              if (utf) BACKCHAR(eptr);
3162    #endif
3163            }            }
3164          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3165          }          }
# Line 3009  for (;;) Line 3171  for (;;)
3171      /* Match a single character, casefully */      /* Match a single character, casefully */
3172    
3173      case OP_CHAR:      case OP_CHAR:
3174  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3175      if (utf8)      if (utf)
3176        {        {
3177        length = 1;        length = 1;
3178        ecode++;        ecode++;
# Line 3020  for (;;) Line 3182  for (;;)
3182          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
3183          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
3184          }          }
3185        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);        while (length-- > 0) if (*ecode++ != RAWUCHARINC(eptr)) RRETURN(MATCH_NOMATCH);
3186        }        }
3187      else      else
3188  #endif  #endif
3189        /* Not UTF mode */
     /* Non-UTF-8 mode */  
3190        {        {
3191        if (md->end_subject - eptr < 1)        if (md->end_subject - eptr < 1)
3192          {          {
# Line 3037  for (;;) Line 3198  for (;;)
3198        }        }
3199      break;      break;
3200    
3201      /* Match a single character, caselessly */      /* Match a single character, caselessly. If we are at the end of the
3202        subject, give up immediately. */
3203    
3204      case OP_CHARI:      case OP_CHARI:
3205  #ifdef SUPPORT_UTF8      if (eptr >= md->end_subject)
3206      if (utf8)        {
3207          SCHECK_PARTIAL();
3208          RRETURN(MATCH_NOMATCH);
3209          }
3210    
3211    #ifdef SUPPORT_UTF
3212        if (utf)
3213        {        {
3214        length = 1;        length = 1;
3215        ecode++;        ecode++;
3216        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
3217    
3218        if (length > md->end_subject - eptr)        /* If the pattern character's value is < 128, we have only one byte, and
3219          we know that its other case must also be one byte long, so we can use the
3220          fast lookup table. We know that there is at least one byte left in the
3221          subject. */
3222    
3223          if (fc < 128)
3224          {          {
3225          CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */          pcre_uint32 cc = RAWUCHAR(eptr);
3226          RRETURN(MATCH_NOMATCH);          if (md->lcc[fc] != TABLE_GET(cc, md->lcc, cc)) RRETURN(MATCH_NOMATCH);
3227          }          ecode++;
3228            eptr++;
       /* If the pattern character's value is < 128, we have only one byte, and  
       can use the fast lookup table. */  
   
       if (fc < 128)  
         {  
         if (md->lcc[*ecode++] != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);  
3229          }          }
3230    
3231        /* Otherwise we must pick up the subject character */        /* Otherwise we must pick up the subject character. Note that we cannot
3232          use the value of "length" to check for sufficient bytes left, because the
3233          other case of the character may have more or fewer bytes.  */
3234    
3235        else        else
3236          {          {
3237          unsigned int dc;          pcre_uint32 dc;
3238          GETCHARINC(dc, eptr);          GETCHARINC(dc, eptr);
3239          ecode += length;          ecode += length;
3240    
# Line 3082  for (;;) Line 3251  for (;;)
3251          }          }
3252        }        }
3253      else      else
3254  #endif   /* SUPPORT_UTF8 */  #endif   /* SUPPORT_UTF */
3255    
3256      /* Non-UTF-8 mode */      /* Not UTF mode */
3257        {        {
3258        if (md->end_subject - eptr < 1)        if (TABLE_GET(ecode[1], md->lcc, ecode[1])
3259          {            != TABLE_GET(*eptr, md->lcc, *eptr)) RRETURN(MATCH_NOMATCH);
3260          SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */        eptr++;
         RRETURN(MATCH_NOMATCH);  
         }  
       if (md->lcc[ecode[1]] != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);  
3261        ecode += 2;        ecode += 2;
3262        }        }
3263      break;      break;
# Line 3101  for (;;) Line 3267  for (;;)
3267      case OP_EXACT:      case OP_EXACT:
3268      case OP_EXACTI:      case OP_EXACTI:
3269      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3270      ecode += 3;      ecode += 1 + IMM2_SIZE;
3271      goto REPEATCHAR;      goto REPEATCHAR;
3272    
3273      case OP_POSUPTO:      case OP_POSUPTO:
# Line 3116  for (;;) Line 3282  for (;;)
3282      min = 0;      min = 0;
3283      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3284      minimize = *ecode == OP_MINUPTO || *ecode == OP_MINUPTOI;      minimize = *ecode == OP_MINUPTO || *ecode == OP_MINUPTOI;
3285      ecode += 3;      ecode += 1 + IMM2_SIZE;
3286      goto REPEATCHAR;      goto REPEATCHAR;
3287    
3288      case OP_POSSTAR:      case OP_POSSTAR:
# Line 3161  for (;;) Line 3327  for (;;)
3327      max = rep_max[c];                 /* zero for max => infinity */      max = rep_max[c];                 /* zero for max => infinity */
3328      if (max == 0) max = INT_MAX;      if (max == 0) max = INT_MAX;
3329    
3330      /* Common code for all repeated single-character matches. */      /* Common code for all repeated single-character matches. We first check
3331        for the minimum number of characters. If the minimum equals the maximum, we
3332        are done. Otherwise, if minimizing, check the rest of the pattern for a
3333        match; if there isn't one, advance up to the maximum, one character at a
3334        time.
3335    
3336        If maximizing, advance up to the maximum number of matching characters,
3337        until eptr is past the end of the maximum run. If possessive, we are
3338        then done (no backing up). Otherwise, match at this position; anything
3339        other than no match is immediately returned. For nomatch, back up one
3340        character, unless we are matching \R and the last thing matched was
3341        \r\n, in which case, back up two bytes. When we reach the first optional
3342        character position, we can save stack by doing a tail recurse.
3343    
3344        The various UTF/non-UTF and caseful/caseless cases are handled separately,
3345        for speed. */
3346    
3347      REPEATCHAR:      REPEATCHAR:
3348  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3349      if (utf8)      if (utf)
3350        {        {
3351        length = 1;        length = 1;
3352        charptr = ecode;        charptr = ecode;
# Line 3178  for (;;) Line 3359  for (;;)
3359        if (length > 1)        if (length > 1)
3360          {          {
3361  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3362          unsigned int othercase;          pcre_uint32 othercase;
3363          if (op >= OP_STARI &&     /* Caseless */          if (op >= OP_STARI &&     /* Caseless */
3364              (othercase = UCD_OTHERCASE(fc)) != fc)              (othercase = UCD_OTHERCASE(fc)) != fc)
3365            oclength = _pcre_ord2utf8(othercase, occhars);            oclength = PRIV(ord2utf)(othercase, occhars);
3366          else oclength = 0;          else oclength = 0;
3367  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3368    
3369          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3370            {            {
3371            if (eptr <= md->end_subject - length &&            if (eptr <= md->end_subject - length &&
3372              memcmp(eptr, charptr, length) == 0) eptr += length;              memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3373  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3374            else if (oclength > 0 &&            else if (oclength > 0 &&
3375                     eptr <= md->end_subject - oclength &&                     eptr <= md->end_subject - oclength &&
3376                     memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                     memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3377  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3378            else            else
3379              {              {
# Line 3211  for (;;) Line 3392  for (;;)
3392              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3393              if (fi >= max) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3394              if (eptr <= md->end_subject - length &&              if (eptr <= md->end_subject - length &&
3395                memcmp(eptr, charptr, length) == 0) eptr += length;                memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3396  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3397              else if (oclength > 0 &&              else if (oclength > 0 &&
3398                       eptr <= md->end_subject - oclength &&                       eptr <= md->end_subject - oclength &&
3399                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                       memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3400  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3401              else              else
3402                {                {
# Line 3232  for (;;) Line 3413  for (;;)
3413            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3414              {              {
3415              if (eptr <= md->end_subject - length &&              if (eptr <= md->end_subject - length &&
3416                  memcmp(eptr, charptr, length) == 0) eptr += length;                  memcmp(eptr, charptr, IN_UCHARS(length)) == 0) eptr += length;
3417  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3418              else if (oclength > 0 &&              else if (oclength > 0 &&
3419                       eptr <= md->end_subject - oclength &&                       eptr <= md->end_subject - oclength &&
3420                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;                       memcmp(eptr, occhars, IN_UCHARS(oclength)) == 0) eptr += oclength;
3421  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
3422              else              else
3423                {                {
# Line 3245  for (;;) Line 3426  for (;;)
3426                }                }
3427              }              }
3428    
3429            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
   
3430            for(;;)            for(;;)
3431              {              {
3432                if (eptr == pp) goto TAIL_RECURSE;
3433              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM23);
3434              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
             if (eptr == pp) { RRETURN(MATCH_NOMATCH); }  
3435  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
3436              eptr--;              eptr--;
3437              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 3268  for (;;) Line 3448  for (;;)
3448        value of fc will always be < 128. */        value of fc will always be < 128. */
3449        }        }
3450      else      else
3451  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
3452          /* When not in UTF-8 mode, load a single-byte character. */
3453      /* When not in UTF-8 mode, load a single-byte character. */        fc = *ecode++;
3454    
3455      fc = *ecode++;      /* The value of fc at this point is always one character, though we may
3456        or may not be in UTF mode. The code is duplicated for the caseless and
     /* The value of fc at this point is always less than 256, though we may or  
     may not be in UTF-8 mode. The code is duplicated for the caseless and  
3457      caseful cases, for speed, since matching characters is likely to be quite      caseful cases, for speed, since matching characters is likely to be quite
3458      common. First, ensure the minimum number of matches are present. If min =      common. First, ensure the minimum number of matches are present. If min =
3459      max, continue at the same level without recursing. Otherwise, if      max, continue at the same level without recursing. Otherwise, if
# Line 3284  for (;;) Line 3462  for (;;)
3462      maximizing, find the maximum number of characters and work backwards. */      maximizing, find the maximum number of characters and work backwards. */
3463    
3464      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,      DPRINTF(("matching %c{%d,%d} against subject %.*s\n", fc, min, max,
3465        max, eptr));        max, (char *)eptr));
3466    
3467      if (op >= OP_STARI)  /* Caseless */      if (op >= OP_STARI)  /* Caseless */
3468        {        {
3469        fc = md->lcc[fc];  #ifdef COMPILE_PCRE8
3470          /* fc must be < 128 if UTF is enabled. */
3471          foc = md->fcc[fc];
3472    #else
3473    #ifdef SUPPORT_UTF
3474    #ifdef SUPPORT_UCP
3475          if (utf && fc > 127)
3476            foc = UCD_OTHERCASE(fc);
3477    #else
3478          if (utf && fc > 127)
3479            foc = fc;
3480    #endif /* SUPPORT_UCP */
3481          else
3482    #endif /* SUPPORT_UTF */
3483            foc = TABLE_GET(fc, md->fcc, fc);
3484    #endif /* COMPILE_PCRE8 */
3485    
3486        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
3487          {          {
3488            pcre_uint32 cc;                 /* Faster than pcre_uchar */
3489          if (eptr >= md->end_subject)          if (eptr >= md->end_subject)
3490            {            {
3491            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3492            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3493            }            }
3494          if (fc != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);          cc = RAWUCHARTEST(eptr);
3495            if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
3496            eptr++;
3497          }          }
3498        if (min == max) continue;        if (min == max) continue;
3499        if (minimize)        if (minimize)
3500          {          {
3501          for (fi = min;; fi++)          for (fi = min;; fi++)
3502            {            {
3503              pcre_uint32 cc;               /* Faster than pcre_uchar */
3504            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM24);
3505            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3506            if (fi >= max) RRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
# Line 3311  for (;;) Line 3509  for (;;)
3509              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3510              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3511              }              }
3512            if (fc != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);            cc = RAWUCHARTEST(eptr);
3513              if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
3514              eptr++;
3515            }            }
3516          /* Control never gets here */          /* Control never gets here */
3517          }          }
# Line 3320  for (;;) Line 3520  for (;;)
3520          pp = eptr;          pp = eptr;
3521          for (i = min; i < max; i++)          for (i = min; i < max; i++)
3522            {            {
3523              pcre_uint32 cc;               /* Faster than pcre_uchar */
3524            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
3525              {              {
3526              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3527              break;              break;
3528              }              }
3529            if (fc != md->lcc[*eptr]) break;            cc = RAWUCHARTEST(eptr);
3530              if (fc != cc && foc != cc) break;
3531            eptr++;            eptr++;
3532            }            }
3533            if (possessive) continue;       /* No backtracking */
3534          if (possessive) continue;          for (;;)
   
         while (eptr >= pp)  
3535            {            {
3536              if (eptr == pp) goto TAIL_RECURSE;
3537            RMATCH(eptr, ecode, offset_top, md, eptrb, RM25);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM25);
3538            eptr--;            eptr--;
3539            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3540            }            }
3541          RRETURN(MATCH_NOMATCH);          /* Control never gets here */
3542          }          }
       /* Control never gets here */  
3543        }        }
3544    
3545      /* Caseful comparisons (includes all multi-byte characters) */      /* Caseful comparisons (includes all multi-byte characters) */
# Line 3353  for (;;) Line 3553  for (;;)
3553            SCHECK_PARTIAL();            SCHECK_PARTIAL();
3554            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
3555            }            }
3556          if (fc != *eptr++) RRETURN(MATCH_NOMATCH);          if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
3557          }          }
3558    
3559        if (min == max) continue;        if (min == max) continue;
# Line 3370  for (;;) Line 3570  for (;;)
3570              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3571              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3572              }              }
3573            if (fc != *eptr++) RRETURN(MATCH_NOMATCH);            if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
3574            }            }
3575          /* Control never gets here */          /* Control never gets here */
3576          }          }
# Line 3384  for (;;) Line 3584  for (;;)
3584              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3585              break;              break;
3586              }              }
3587            if (fc != *eptr) break;            if (fc != RAWUCHARTEST(eptr)) break;
3588            eptr++;            eptr++;
3589            }            }
3590          if (possessive) continue;          if (possessive) continue;    /* No backtracking */
3591            for (;;)
         while (eptr >= pp)  
3592            {            {
3593              if (eptr == pp) goto TAIL_RECURSE;
3594            RMATCH(eptr, ecode, offset_top, md, eptrb, RM27);            RMATCH(eptr, ecode, offset_top, md, eptrb, RM27);
3595            eptr--;            eptr--;
3596            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3597            }            }
3598          RRETURN(MATCH_NOMATCH);          /* Control never gets here */
3599          }          }
3600        }        }
3601      /* Control never gets here */      /* Control never gets here */
# Line 3410  for (;;) Line 3610  for (;;)
3610        SCHECK_PARTIAL();        SCHECK_PARTIAL();
3611        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
3612        }        }
3613      ecode++;  #ifdef SUPPORT_UTF
3614      GETCHARINCTEST(c, eptr);      if (utf)
     if (op == OP_NOTI)         /* The caseless case */  
3615        {        {
3616  #ifdef SUPPORT_UTF8        register pcre_uint32 ch, och;
3617        if (c < 256)  
3618  #endif        ecode++;
3619        c = md->lcc[c];        GETCHARINC(ch, ecode);
3620        if (md->lcc[*ecode++] == c) RRETURN(MATCH_NOMATCH);        GETCHARINC(c, eptr);
3621    
3622          if (op == OP_NOT)
3623            {
3624            if (ch == c) RRETURN(MATCH_NOMATCH);
3625            }
3626          else
3627            {
3628    #ifdef SUPPORT_UCP
3629            if (ch > 127)
3630              och = UCD_OTHERCASE(ch);
3631    #else
3632            if (ch > 127)
3633              och = ch;
3634    #endif /* SUPPORT_UCP */
3635            else
3636              och = TABLE_GET(ch, md->fcc, ch);
3637            if (ch == c || och == c) RRETURN(MATCH_NOMATCH);
3638            }
3639        }        }
3640      else    /* Caseful */      else
3641    #endif
3642        {        {
3643        if (*ecode++ == c) RRETURN(MATCH_NOMATCH);        register pcre_uint32 ch = ecode[1];
3644          c = *eptr++;
3645          if (ch == c || (op == OP_NOTI && TABLE_GET(ch, md->fcc, ch) == c))
3646            RRETURN(MATCH_NOMATCH);
3647          ecode += 2;
3648        }        }
3649      break;      break;
3650    
# Line 3436  for (;;) Line 3658  for (;;)
3658      case OP_NOTEXACT:      case OP_NOTEXACT:
3659      case OP_NOTEXACTI:      case OP_NOTEXACTI:
3660      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
3661      ecode += 3;      ecode += 1 + IMM2_SIZE;
3662      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3663    
3664      case OP_NOTUPTO:      case OP_NOTUPTO:
# Line 3446  for (;;) Line 3668  for (;;)
3668      min = 0;      min = 0;
3669      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3670      minimize = *ecode == OP_NOTMINUPTO || *ecode == OP_NOTMINUPTOI;      minimize = *ecode == OP_NOTMINUPTO || *ecode == OP_NOTMINUPTOI;
3671      ecode += 3;      ecode += 1 + IMM2_SIZE;
3672      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3673    
3674      case OP_NOTPOSSTAR:      case OP_NOTPOSSTAR:
# Line 3478  for (;;) Line 3700  for (;;)
3700      possessive = TRUE;      possessive = TRUE;
3701      min = 0;      min = 0;
3702      max = GET2(ecode, 1);      max = GET2(ecode, 1);
3703      ecode += 3;      ecode += 1 + IMM2_SIZE;
3704      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
3705    
3706      case OP_NOTSTAR:      case OP_NOTSTAR:
# Line 3502  for (;;) Line 3724  for (;;)
3724      /* Common code for all repeated single-byte matches. */      /* Common code for all repeated single-byte matches. */
3725    
3726      REPEATNOTCHAR:      REPEATNOTCHAR:
3727      fc = *ecode++;      GETCHARINCTEST(fc, ecode);
3728    
3729      /* The code is duplicated for the caseless and caseful cases, for speed,      /* The code is duplicated for the caseless and caseful cases, for speed,
3730      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 3513  for (;;) Line 3735  for (;;)
3735      characters and work backwards. */      characters and work backwards. */
3736    
3737      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,
3738        max, eptr));        max, (char *)eptr));
3739    
3740      if (op >= OP_NOTSTARI)     /* Caseless */      if (op >= OP_NOTSTARI)     /* Caseless */
3741        {        {
3742        fc = md->lcc[fc];  #ifdef SUPPORT_UTF
3743    #ifdef SUPPORT_UCP
3744          if (utf && fc > 127)
3745            foc = UCD_OTHERCASE(fc);
3746    #else
3747          if (utf && fc > 127)
3748            foc = fc;
3749    #endif /* SUPPORT_UCP */
3750          else
3751    #endif /* SUPPORT_UTF */
3752            foc = TABLE_GET(fc, md->fcc, fc);
3753    
3754  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3755        /* UTF-8 mode */        if (utf)
       if (utf8)  
3756          {          {
3757          register unsigned int d;          register pcre_uint32 d;
3758          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3759            {            {
3760            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
# Line 3532  for (;;) Line 3763  for (;;)
3763              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3764              }              }
3765            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3766            if (d < 256) d = md->lcc[d];            if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
           if (fc == d) RRETURN(MATCH_NOMATCH);  
3767            }            }
3768          }          }
3769        else        else
3770  #endif  #endif  /* SUPPORT_UTF */
3771          /* Not UTF mode */
       /* Not UTF-8 mode */  
3772          {          {
3773          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3774            {            {
# Line 3548  for (;;) Line 3777  for (;;)
3777              SCHECK_PARTIAL();              SCHECK_PARTIAL();
3778              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3779              }              }
3780            if (fc == md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);            if (fc == *eptr || foc == *eptr) RRETURN(MATCH_NOMATCH);
3781              eptr++;
3782            }            }
3783          }          }
3784    
# Line 3556  for (;;) Line 3786  for (;;)
3786    
3787        if (minimize)        if (minimize)
3788          {          {
3789  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3790          /* UTF-8 mode */          if (utf)
         if (utf8)  
3791            {            {
3792            register unsigned int d;            register pcre_uint32 d;
3793            for (fi = min;; fi++)            for (fi = min;; fi++)
3794              {              {
3795              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM28);
# Line 3572  for (;;) Line 3801  for (;;)
3801                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3802                }                }
3803              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3804              if (d < 256) d = md->lcc[d];              if (fc == d || (unsigned int)foc == d) RRETURN(MATCH_NOMATCH);
             if (fc == d) RRETURN(MATCH_NOMATCH);  
3805              }              }
3806            }            }
3807          else          else
3808  #endif  #endif  /*SUPPORT_UTF */
3809          /* Not UTF-8 mode */          /* Not UTF mode */
3810            {            {
3811            for (fi = min;; fi++)            for (fi = min;; fi++)
3812              {              {
# Line 3590  for (;;) Line 3818  for (;;)
3818                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3819                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3820                }                }
3821              if (fc == md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);              if (fc == *eptr || foc == *eptr) RRETURN(MATCH_NOMATCH);
3822                eptr++;
3823              }              }
3824            }            }
3825          /* Control never gets here */          /* Control never gets here */
# Line 3602  for (;;) Line 3831  for (;;)
3831          {          {
3832          pp = eptr;          pp = eptr;
3833    
3834  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3835          /* UTF-8 mode */          if (utf)
         if (utf8)  
3836            {            {
3837            register unsigned int d;            register pcre_uint32 d;
3838            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3839              {              {
3840              int len = 1;              int len = 1;
# Line 3616  for (;;) Line 3844  for (;;)
3844                break;                break;
3845                }                }
3846              GETCHARLEN(d, eptr, len);              GETCHARLEN(d, eptr, len);
3847              if (d < 256) d = md->lcc[d];              if (fc == d || (unsigned int)foc == d) break;
             if (fc == d) break;  
3848              eptr += len;              eptr += len;
3849              }              }
3850          if (possessive) continue;            if (possessive) continue;    /* No backtracking */
3851          for(;;)            for(;;)
3852              {              {
3853                if (eptr == pp) goto TAIL_RECURSE;
3854              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM30);
3855              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3856              if (eptr-- == pp) break;        /* Stop if tried at original pos */              eptr--;
3857              BACKCHAR(eptr);              BACKCHAR(eptr);
3858              }              }
3859            }            }
3860          else          else
3861  #endif  #endif  /* SUPPORT_UTF */
3862          /* Not UTF-8 mode */          /* Not UTF mode */
3863            {            {
3864            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3865              {              {
# Line 3640  for (;;) Line 3868  for (;;)
3868                SCHECK_PARTIAL();                SCHECK_PARTIAL();
3869                break;                break;
3870                }                }
3871              if (fc == md->lcc[*eptr]) break;              if (fc == *eptr || foc == *eptr) break;
3872              eptr++;              eptr++;
3873              }              }
3874            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
3875            while (eptr >= pp)            for (;;)
3876              {              {
3877                if (eptr == pp) goto TAIL_RECURSE;
3878              RMATCH(eptr, ecode, offset_top, md, eptrb, RM31);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM31);
3879              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3880              eptr--;              eptr--;
3881              }              }
3882            }            }
3883            /* Control never gets here */
         RRETURN(MATCH_NOMATCH);  
3884          }          }
       /* Control never gets here */  
3885        }        }
3886    
3887      /* Caseful comparisons */      /* Caseful comparisons */
3888    
3889      else      else
3890        {        {
3891  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3892        /* UTF-8 mode */        if (utf)
       if (utf8)  
3893          {          {
3894          register unsigned int d;          register pcre_uint32 d;
3895          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3896            {            {
3897            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
# Line 3679  for (;;) Line 3905  for (;;)
3905          }          }
3906        else        else
3907  #endif  #endif
3908        /* Not UTF-8 mode */        /* Not UTF mode */
3909          {          {
3910          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3911            {            {
# Line 3696  for (;;) Line 3922  for (;;)
3922    
3923        if (minimize)        if (minimize)
3924          {          {
3925  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3926          /* UTF-8 mode */          if (utf)
         if (utf8)  
3927            {            {
3928            register unsigned int d;            register pcre_uint32 d;
3929            for (fi = min;; fi++)            for (fi = min;; fi++)
3930              {              {
3931              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM32);
# Line 3717  for (;;) Line 3942  for (;;)
3942            }            }
3943          else          else
3944  #endif  #endif
3945          /* Not UTF-8 mode */          /* Not UTF mode */
3946            {            {
3947            for (fi = min;; fi++)            for (fi = min;; fi++)
3948              {              {
# Line 3741  for (;;) Line 3966  for (;;)
3966          {          {
3967          pp = eptr;          pp = eptr;
3968    
3969  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3970          /* UTF-8 mode */          if (utf)
         if (utf8)  
3971            {            {
3972            register unsigned int d;            register pcre_uint32 d;
3973            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3974              {              {
3975              int len = 1;              int len = 1;
# Line 3758  for (;;) Line 3982  for (;;)
3982              if (fc == d) break;              if (fc == d) break;
3983              eptr += len;              eptr += len;
3984              }              }
3985            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
3986            for(;;)            for(;;)
3987              {              {
3988                if (eptr == pp) goto TAIL_RECURSE;
3989              RMATCH(eptr, ecode, offset_top, md, eptrb, RM34);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM34);
3990              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3991              if (eptr-- == pp) break;        /* Stop if tried at original pos */              eptr--;
3992              BACKCHAR(eptr);              BACKCHAR(eptr);
3993              }              }
3994            }            }
3995          else          else
3996  #endif  #endif
3997          /* Not UTF-8 mode */          /* Not UTF mode */
3998            {            {
3999            for (i = min; i < max; i++)            for (i = min; i < max; i++)
4000              {              {
# Line 3781  for (;;) Line 4006  for (;;)
4006              if (fc == *eptr) break;              if (fc == *eptr) break;
4007              eptr++;              eptr++;
4008              }              }
4009            if (possessive) continue;            if (possessive) continue;    /* No backtracking */
4010            while (eptr >= pp)            for (;;)
4011              {              {
4012                if (eptr == pp) goto TAIL_RECURSE;
4013              RMATCH(eptr, ecode, offset_top, md, eptrb, RM35);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM35);
4014              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4015              eptr--;              eptr--;
4016              }              }
4017            }            }
4018            /* Control never gets here */
         RRETURN(MATCH_NOMATCH);  
4019          }          }
4020        }        }
4021      /* Control never gets here */      /* Control never gets here */
# Line 3802  for (;;) Line 4027  for (;;)
4027      case OP_TYPEEXACT:      case OP_TYPEEXACT:
4028      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
4029      minimize = TRUE;      minimize = TRUE;
4030      ecode += 3;      ecode += 1 + IMM2_SIZE;
4031      goto REPEATTYPE;      goto REPEATTYPE;
4032    
4033      case OP_TYPEUPTO:      case OP_TYPEUPTO:
# Line 3810  for (;;) Line 4035  for (;;)
4035      min = 0;      min = 0;
4036      max = GET2(ecode, 1);      max = GET2(ecode, 1);
4037      minimize = *ecode == OP_TYPEMINUPTO;      minimize = *ecode == OP_TYPEMINUPTO;
4038      ecode += 3;      ecode += 1 + IMM2_SIZE;
4039      goto REPEATTYPE;      goto REPEATTYPE;
4040    
4041      case OP_TYPEPOSSTAR:      case OP_TYPEPOSSTAR:
# Line 3838  for (;;) Line 4063  for (;;)
4063      possessive = TRUE;      possessive = TRUE;
4064      min = 0;      min = 0;
4065      max = GET2(ecode, 1);      max = GET2(ecode, 1);
4066      ecode += 3;      ecode += 1 + IMM2_SIZE;
4067      goto REPEATTYPE;      goto REPEATTYPE;
4068    
4069      case OP_TYPESTAR:      case OP_TYPESTAR:
# Line 3972  for (;;) Line 4197  for (;;)
4197              }              }
4198            break;            break;
4199    
4200              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
4201              which means that Perl space and POSIX space are now identical. PCRE
4202              was changed at release 8.34. */
4203    
4204            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
4205              case PT_PXSPACE:  /* POSIX space */
4206            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4207              {              {
4208              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
# Line 3982  for (;;) Line 4212  for (;;)
4212                }                }
4213              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4214              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||
4215                   c == CHAR_FF || c == CHAR_CR)                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)
4216                     == prop_fail_result)                     == prop_fail_result)
4217                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4218              }              }
4219            break;            break;
4220    
4221            case PT_PXSPACE:  /* POSIX space */            case PT_WORD:
4222            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4223              {              {
4224                int category;
4225              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4226                {                {
4227                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4228                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4229                }                }
4230              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4231              if ((UCD_CATEGORY(c) == ucp_Z || c == CHAR_HT || c == CHAR_NL ||              category = UCD_CATEGORY(c);
4232                   c == CHAR_VT || c == CHAR_FF || c == CHAR_CR)              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)
4233                     == prop_fail_result)                     == prop_fail_result)
4234                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4235              }              }
4236            break;            break;
4237    
4238            case PT_WORD:            case PT_CLIST:
4239            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
4240              {              {
4241              int category;              const pcre_uint32 *cp;
4242              if (eptr >= md->end_subject)              if (eptr >= md->end_subject)
4243                {                {
4244                SCHECK_PARTIAL();                SCHECK_PARTIAL();
4245                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4246                }                }
4247              GETCHARINCTEST(c, eptr);              GETCHARINCTEST(c, eptr);
4248              category = UCD_CATEGORY(c);              cp = PRIV(ucd_caseless_sets) + prop_value;
4249              if ((category == ucp_L || category == ucp_N || c == CHAR_UNDERSCORE)              for (;;)
4250                     == prop_fail_result)                {
4251                  if (c < *cp)
4252                    { if (prop_fail_result) break; else { RRETURN(MATCH_NOMATCH); } }
4253                  if (c == *cp++)
4254                    { if (prop_fail_result) { RRETURN(MATCH_NOMATCH); } else break; }
4255                  }
4256                }
4257              break;
4258    
4259              case PT_UCNC:
4260              for (i = 1; i <= min; i++)
4261                {
4262                if (eptr >= md->end_subject)
4263                  {
4264                  SCHECK_PARTIAL();
4265                  RRETURN(MATCH_NOMATCH);
4266                  }
4267                GETCHARINCTEST(c, eptr);
4268                if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT ||
4269                     c == CHAR_GRAVE_ACCENT || (c >= 0xa0 && c <= 0xd7ff) ||
4270                     c >= 0xe000) == prop_fail_result)
4271                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4272              }              }
4273            break;            break;
# Line 4040  for (;;) Line 4291  for (;;)
4291              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4292              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4293              }              }
4294            GETCHARINCTEST(c, eptr);            else
           if (UCD_CATEGORY(c) == ucp_M) RRETURN(MATCH_NOMATCH);  
           while (eptr < md->end_subject)  
4295              {              {
4296              int len = 1;              int lgb, rgb;
4297              if (!utf8) c = *eptr; else { GETCHARLEN(c, eptr, len); }              GETCHARINCTEST(c, eptr);
4298              if (UCD_CATEGORY(c) != ucp_M) break;              lgb = UCD_GRAPHBREAK(c);
4299              eptr += len;             while (eptr < md->end_subject)
4300                  {
4301                  int len = 1;
4302                  if (!utf) c = *eptr; else { GETCHARLEN(c, eptr, len); }
4303                  rgb = UCD_GRAPHBREAK(c);
4304                  if ((PRIV(ucp_gbtable)[lgb] & (1 << rgb)) == 0) break;
4305                  lgb = rgb;
4306                  eptr += len;
4307                  }
4308              }              }
4309              CHECK_PARTIAL();
4310            }            }
4311          }          }
4312    
# Line 4057  for (;;) Line 4315  for (;;)
4315    
4316  /* Handle all other cases when the coding is UTF-8 */  /* Handle all other cases when the coding is UTF-8 */
4317    
4318  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4319        if (utf8) switch(ctype)        if (utf) switch(ctype)
4320          {          {
4321          case OP_ANY:          case OP_ANY:
4322          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
# Line 4069  for (;;) Line 4327  for (;;)
4327              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4328              }              }
4329            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4330              if (md->partial != 0 &&
4331                  eptr + 1 >= md->end_subject &&
4332                  NLBLOCK->nltype == NLTYPE_FIXED &&
4333                  NLBLOCK->nllen == 2 &&
4334                  RAWUCHAR(eptr) == NLBLOCK->nl[0])
4335                {
4336                md->hitend = TRUE;
4337                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4338                }
4339            eptr++;            eptr++;
4340            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4341            }            }
4342          break;          break;
4343    
# Line 4083  for (;;) Line 4350  for (;;)
4350              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4351              }              }
4352            eptr++;            eptr++;
4353            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4354            }            }
4355          break;          break;
4356    
# Line 4105  for (;;) Line 4372  for (;;)
4372              {              {
4373              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4374    
4375              case 0x000d:              case CHAR_CR:
4376              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && RAWUCHAR(eptr) == CHAR_LF) eptr++;
4377              break;              break;
4378    
4379              case 0x000a:              case CHAR_LF:
4380              break;              break;
4381    
4382              case 0x000b:              case CHAR_VT:
4383              case 0x000c:              case CHAR_FF:
4384              case 0x0085:              case CHAR_NEL:
4385    #ifndef EBCDIC
4386              case 0x2028:              case 0x2028:
4387              case 0x2029:              case 0x2029:
4388    #endif  /* Not EBCDIC */
4389              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4390              break;              break;
4391              }              }
# Line 4134  for (;;) Line 4403  for (;;)
4403            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4404            switch(c)            switch(c)
4405              {              {
4406                HSPACE_CASES: RRETURN(MATCH_NOMATCH);  /* Byte and multibyte cases */
4407              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);  
4408              }              }
4409            }            }
4410          break;          break;
# Line 4170  for (;;) Line 4420  for (;;)
4420            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4421            switch(c)            switch(c)
4422              {              {
4423                HSPACE_CASES: break;  /* Byte and multibyte cases */
4424              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;  
4425              }              }
4426            }            }
4427          break;          break;
# Line 4206  for (;;) Line 4437  for (;;)
4437            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4438            switch(c)            switch(c)
4439              {              {
4440                VSPACE_CASES: RRETURN(MATCH_NOMATCH);
4441              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);  
4442              }              }
4443            }            }
4444          break;          break;
# Line 4230  for (;;) Line 4454  for (;;)
4454            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
4455            switch(c)            switch(c)
4456              {              {
4457                VSPACE_CASES: break;
4458              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;  
4459              }              }
4460            }            }
4461          break;          break;
# Line 4260  for (;;) Line 4477  for (;;)
4477          case OP_DIGIT:          case OP_DIGIT:
4478          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4479            {            {
4480              pcre_uint32 cc;
4481            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4482              {              {
4483              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4484              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4485              }              }
4486            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_digit) == 0)            cc = RAWUCHAR(eptr);
4487              if (cc >= 128 || (md->ctypes[cc] & ctype_digit) == 0)
4488              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4489              eptr++;
4490            /* 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 */
4491            }            }
4492          break;          break;
# Line 4274  for (;;) Line 4494  for (;;)
4494          case OP_NOT_WHITESPACE:          case OP_NOT_WHITESPACE:
4495          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4496            {            {
4497              pcre_uint32 cc;
4498            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4499              {              {
4500              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4501              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4502              }              }
4503            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)            cc = RAWUCHAR(eptr);
4504              if (cc < 128 && (md->ctypes[cc] & ctype_space) != 0)
4505              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4506            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);            eptr++;
4507              ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4508            }            }
4509          break;          break;
4510    
4511          case OP_WHITESPACE:          case OP_WHITESPACE:
4512          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4513            {            {
4514              pcre_uint32 cc;
4515            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4516              {              {
4517              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4518              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4519              }              }
4520            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_space) == 0)            cc = RAWUCHAR(eptr);
4521              if (cc >= 128 || (md->ctypes[cc] & ctype_space) == 0)
4522              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4523              eptr++;
4524            /* 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 */
4525            }            }
4526          break;          break;
# Line 4302  for (;;) Line 4528  for (;;)
4528          case OP_NOT_WORDCHAR:          case OP_NOT_WORDCHAR:
4529          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4530            {            {
4531              pcre_uint32 cc;
4532            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4533              {              {
4534              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4535              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4536              }              }
4537            if (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0)            cc = RAWUCHAR(eptr);
4538              if (cc < 128 && (md->ctypes[cc] & ctype_word) != 0)
4539              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4540            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);            eptr++;
4541              ACROSSCHAR(eptr < md->end_subject, *eptr, eptr++);
4542            }            }
4543          break;          break;
4544    
4545          case OP_WORDCHAR:          case OP_WORDCHAR:
4546          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
4547            {            {
4548              pcre_uint32 cc;
4549            if (eptr >= md->end_subject)            if (eptr >= md->end_subject)
4550              {              {
4551              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4552              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4553              }              }
4554            if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_word) == 0)            cc = RAWUCHAR(eptr);
4555              if (cc >= 128 || (md->ctypes[cc] & ctype_word) == 0)
4556              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4557              eptr++;
4558            /* 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 */
4559            }            }
4560          break;          break;
# Line 4332  for (;;) Line 4564  for (;;)
4564          }  /* End switch(ctype) */          }  /* End switch(ctype) */
4565    
4566        else        else
4567  #endif     /* SUPPORT_UTF8 */  #endif     /* SUPPORT_UTF */
4568    
4569        /* Code for the non-UTF-8 case for minimum matching of operators other        /* Code for the non-UTF-8 case for minimum matching of operators other
4570        than OP_PROP and OP_NOTPROP. */        than OP_PROP and OP_NOTPROP. */
# Line 4348  for (;;) Line 4580  for (;;)
4580              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4581              }              }
4582            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
4583              if (md->partial != 0 &&
4584                  eptr + 1 >= md->end_subject &&
4585                  NLBLOCK->nltype == NLTYPE_FIXED &&
4586                  NLBLOCK->nllen == 2 &&
4587                  *eptr == NLBLOCK->nl[0])
4588                {
4589                md->hitend = TRUE;
4590                if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
4591                }
4592            eptr++;            eptr++;
4593            }            }
4594          break;          break;
# Line 4382  for (;;) Line 4623  for (;;)
4623              {              {
4624              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4625    
4626              case 0x000d:              case CHAR_CR:
4627              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == CHAR_LF) eptr++;
4628              break;              break;
4629    
4630              case 0x000a:              case CHAR_LF:
4631              break;              break;
4632    
4633              case 0x000b:              case CHAR_VT:
4634              case 0x000c:              case CHAR_FF:
4635              case 0x0085:              case CHAR_NEL:
4636    #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4637                case 0x2028:
4638                case 0x2029:
4639    #endif
4640              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);              if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
4641              break;              break;
4642              }              }
# Line 4409  for (;;) Line 4654  for (;;)
4654            switch(*eptr++)            switch(*eptr++)
4655              {              {
4656              default: break;              default: break;
4657              case 0x09:      /* HT */              HSPACE_BYTE_CASES:
4658              case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4659              case 0xa0:      /* NBSP */              HSPACE_MULTIBYTE_CASES:
4660    #endif
4661              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4662              }              }
4663            }            }
# Line 4428  for (;;) Line 4674  for (;;)
4674            switch(*eptr++)            switch(*eptr++)
4675              {              {
4676              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4677              case 0x09:      /* HT */              HSPACE_BYTE_CASES:
4678              case 0x20:      /* SPACE */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4679              case 0xa0:      /* NBSP */              HSPACE_MULTIBYTE_CASES:
4680    #endif
4681              break;              break;
4682              }              }
4683            }            }
# Line 4446  for (;;) Line 4693  for (;;)
4693              }              }
4694            switch(*eptr++)            switch(*eptr++)
4695              {              {
4696              default: break;              VSPACE_BYTE_CASES:
4697              case 0x0a:      /* LF */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4698              case 0x0b:      /* VT */              VSPACE_MULTIBYTE_CASES:
4699              case 0x0c:      /* FF */  #endif
             case 0x0d:      /* CR */  
             case 0x85:      /* NEL */  
4700              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4701                default: break;
4702              }              }
4703            }            }
4704          break;          break;
# Line 4468  for (;;) Line 4714  for (;;)
4714            switch(*eptr++)            switch(*eptr++)
4715              {              {
4716              default: RRETURN(MATCH_NOMATCH);              default: RRETURN(MATCH_NOMATCH);
4717              case 0x0a:      /* LF */              VSPACE_BYTE_CASES:
4718              case 0x0b:      /* VT */  #if defined COMPILE_PCRE16 || defined COMPILE_PCRE32
4719              case 0x0c:      /* FF */              VSPACE_MULTIBYTE_CASES:
4720              case 0x0d:      /* CR */  #endif
             case 0x85:      /* NEL */  
4721              break;              break;
4722              }              }
4723            }            }
# Line 4486  for (;;) Line 4731  for (;;)
4731              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4732              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4733              }              }
4734            if ((md->ctypes[*eptr++] & ctype_digit) != 0) RRETURN(MATCH_NOMATCH);            if (MAX_255(*eptr) && (md->ctypes[*eptr] & ctype_digit) != 0)
4735                RRETURN(MATCH_NOMATCH);
4736              eptr++;
4737            }            }
4738          break;          break;
4739    
# Line 4498  for (;;) Line 4745  for (;;)
4745              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4746              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4747              }              }
4748            if ((md->ctypes[*eptr++] & ctype_digit) == 0) RRETURN(MATCH_NOMATCH);            if (!MAX_255(*eptr) || (md->ctypes[*eptr] & ctype_digit) == 0)
4749                RRETURN(MATCH_NOMATCH);
4750              eptr++;
4751            }            }
4752          break;          break;
4753    
# Line 4510  for (;;) Line 4759  for (;;)
4759              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4760              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4761              }              }
4762            if ((md->ctypes[*eptr++] & ctype_space) != 0) RRETURN(MATCH_NOMATCH);            if (MAX_255(*eptr) && (md->ctypes[*eptr] & ctype_space) != 0)
4763                RRETURN(MATCH_NOMATCH);
4764              eptr++;
4765            }            }
4766          break;          break;
4767    
# Line 4522  for (;;) Line 4773  for (;;)
4773              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4774              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4775              }              }
4776            if ((md->ctypes[*eptr++] & ctype_space) == 0) RRETURN(MATCH_NOMATCH);            if (!MAX_255(*eptr) || (md->ctypes[*eptr] & ctype_space) == 0)
4777                RRETURN(MATCH_NOMATCH);
4778              eptr++;
4779            }            }
4780          break;          break;
4781    
# Line 4534  for (;;) Line 4787  for (;;)
4787              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4788              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4789              }              }
4790            if ((md->ctypes[*eptr++] & ctype_word) != 0)            if (MAX_255(*eptr) && (md->ctypes[*eptr] & ctype_word) != 0)
4791              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4792              eptr++;
4793            }            }
4794          break;          break;
4795    
# Line 4547  for (;;) Line 4801  for (;;)
4801              SCHECK_PARTIAL();              SCHECK_PARTIAL();
4802              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4803              }              }
4804            if ((md->ctypes[*eptr++] & ctype_word) == 0)            if (!MAX_255(*eptr) || (md->ctypes[*eptr] & ctype_word) == 0)
4805              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
4806              eptr++;
4807            }            }
4808          break;          break;
4809    
# Line 4679  for (;;) Line 4934  for (;;)
4934              }              }
4935            /* Control never gets here */            /* Control never gets here */
4936    
4937              /* Perl space used to exclude VT, but from Perl 5.18 it is included,
4938              which means that Perl space and POSIX space are now identical. PCRE
4939              was changed at release 8.34. */
4940    
4941            case PT_SPACE:    /* Perl space */            case PT_SPACE:    /* Perl space */
4942              case PT_PXSPACE:  /* POSIX space */
4943            for (fi = min;; fi++)            for (fi = min;; fi++)
4944              {              {
4945              RMATCH(eptr, ecode, offset_top, md, eptrb, RM60);              RMATCH(eptr, ecode, offset_top, md, eptrb, RM61);
4946              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);