/[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 91 by nigel, Sat Feb 24 21:41:34 2007 UTC revision 463 by ph10, Sun Oct 18 10:02:46 2009 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) 1997-2006 University of Cambridge             Copyright (c) 1997-2009 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 42  POSSIBILITY OF SUCH DAMAGE. Line 42  POSSIBILITY OF SUCH DAMAGE.
42  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
43  possible. There are also some static supporting functions. */  possible. There are also some static supporting functions. */
44    
45  #define NLBLOCK md           /* The block containing newline information */  #ifdef HAVE_CONFIG_H
46    #include "config.h"
47    #endif
48    
49    #define NLBLOCK md             /* Block containing newline information */
50    #define PSSTART start_subject  /* Field containing processed string start */
51    #define PSEND   end_subject    /* Field containing processed string end */
52    
53  #include "pcre_internal.h"  #include "pcre_internal.h"
54    
55    /* Undefine some potentially clashing cpp symbols */
56    
57  /* Structure for building a chain of data that actually lives on the  #undef min
58  stack, for holding the values of the subject pointer at the start of each  #undef max
 subpattern, so as to detect when an empty string has been matched by a  
 subpattern - to break infinite loops. When NO_RECURSE is set, these blocks  
 are on the heap, not on the stack. */  
   
 typedef struct eptrblock {  
   struct eptrblock *epb_prev;  
   USPTR epb_saved_eptr;  
 } eptrblock;  
59    
60  /* Flag bits for the match() function */  /* Flag bits for the match() function */
61    
62  #define match_condassert   0x01    /* Called to check a condition assertion */  #define match_condassert     0x01  /* Called to check a condition assertion */
63  #define match_isgroup      0x02    /* Set if start of bracketed group */  #define match_cbegroup       0x02  /* Could-be-empty unlimited repeat group */
64    
65  /* Non-error returns from the match() function. Error returns are externally  /* Non-error returns from the match() function. Error returns are externally
66  defined PCRE_ERROR_xxx codes, which are all negative. */  defined PCRE_ERROR_xxx codes, which are all negative. */
# Line 68  defined PCRE_ERROR_xxx codes, which are Line 68  defined PCRE_ERROR_xxx codes, which are
68  #define MATCH_MATCH        1  #define MATCH_MATCH        1
69  #define MATCH_NOMATCH      0  #define MATCH_NOMATCH      0
70    
71    /* Special internal returns from the match() function. Make them sufficiently
72    negative to avoid the external error codes. */
73    
74    #define MATCH_COMMIT       (-999)
75    #define MATCH_PRUNE        (-998)
76    #define MATCH_SKIP         (-997)
77    #define MATCH_THEN         (-996)
78    
79  /* 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.
80  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,
81  because the offset vector is always a multiple of 3 long. */  because the offset vector is always a multiple of 3 long. */
# Line 101  Returns: nothing Line 109  Returns: nothing
109  static void  static void
110  pchars(const uschar *p, int length, BOOL is_subject, match_data *md)  pchars(const uschar *p, int length, BOOL is_subject, match_data *md)
111  {  {
112  int c;  unsigned int c;
113  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;
114  while (length-- > 0)  while (length-- > 0)
115    if (isprint(c = *(p++))) printf("%c", c); else printf("\\x%02x", c);    if (isprint(c = *(p++))) printf("%c", c); else printf("\\x%02x", c);
# Line 150  printf("\n"); Line 158  printf("\n");
158    
159  if (length > md->end_subject - eptr) return FALSE;  if (length > md->end_subject - eptr) return FALSE;
160    
161  /* Separate the caselesss case for speed */  /* Separate the caseless case for speed. In UTF-8 mode we can only do this
162    properly if Unicode properties are supported. Otherwise, we can check only
163    ASCII characters. */
164    
165  if ((ims & PCRE_CASELESS) != 0)  if ((ims & PCRE_CASELESS) != 0)
166    {    {
167    #ifdef SUPPORT_UTF8
168    #ifdef SUPPORT_UCP
169      if (md->utf8)
170        {
171        USPTR endptr = eptr + length;
172        while (eptr < endptr)
173          {
174          int c, d;
175          GETCHARINC(c, eptr);
176          GETCHARINC(d, p);
177          if (c != d && c != UCD_OTHERCASE(d)) return FALSE;
178          }
179        }
180      else
181    #endif
182    #endif
183    
184      /* The same code works when not in UTF-8 mode and in UTF-8 mode when there
185      is no UCP support. */
186    
187    while (length-- > 0)    while (length-- > 0)
188      if (md->lcc[*p++] != md->lcc[*eptr++]) return FALSE;      { if (md->lcc[*p++] != md->lcc[*eptr++]) return FALSE; }
189    }    }
190    
191    /* In the caseful case, we can just compare the bytes, whether or not we
192    are in UTF-8 mode. */
193    
194  else  else
195    { while (length-- > 0) if (*p++ != *eptr++) return FALSE; }    { while (length-- > 0) if (*p++ != *eptr++) return FALSE; }
196    
# Line 186  calls by keeping local variables that ne Line 220  calls by keeping local variables that ne
220  obtained from malloc() instead instead of on the stack. Macros are used to  obtained from malloc() instead instead of on the stack. Macros are used to
221  achieve this so that the actual code doesn't look very different to what it  achieve this so that the actual code doesn't look very different to what it
222  always used to.  always used to.
223    
224    The original heap-recursive code used longjmp(). However, it seems that this
225    can be very slow on some operating systems. Following a suggestion from Stan
226    Switzer, the use of longjmp() has been abolished, at the cost of having to
227    provide a unique number for each call to RMATCH. There is no way of generating
228    a sequence of numbers at compile time in C. I have given them names, to make
229    them stand out more clearly.
230    
231    Crude tests on x86 Linux show a small speedup of around 5-8%. However, on
232    FreeBSD, avoiding longjmp() more than halves the time taken to run the standard
233    tests. Furthermore, not using longjmp() means that local dynamic variables
234    don't have indeterminate values; this has meant that the frame size can be
235    reduced because the result can be "passed back" by straight setting of the
236    variable instead of being passed in the frame.
237  ****************************************************************************  ****************************************************************************
238  ***************************************************************************/  ***************************************************************************/
239    
240    /* Numbers for RMATCH calls. When this list is changed, the code at HEAP_RETURN
241    below must be updated in sync.  */
242    
243    enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM6,  RM7,  RM8,  RM9,  RM10,
244           RM11,  RM12, RM13, RM14, RM15, RM16, RM17, RM18, RM19, RM20,
245           RM21,  RM22, RM23, RM24, RM25, RM26, RM27, RM28, RM29, RM30,
246           RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,
247           RM41,  RM42, RM43, RM44, RM45, RM46, RM47, RM48, RM49, RM50,
248           RM51,  RM52, RM53, RM54 };
249    
250  /* 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
251  versions and production versions. */  versions and production versions. Note that the "rw" argument of RMATCH isn't
252    actuall used in this definition. */
253    
254  #ifndef NO_RECURSE  #ifndef NO_RECURSE
255  #define REGISTER register  #define REGISTER register
256    
257  #ifdef DEBUG  #ifdef DEBUG
258  #define RMATCH(rx,ra,rb,rc,rd,re,rf,rg) \  #define RMATCH(ra,rb,rc,rd,re,rf,rg,rw) \
259    { \    { \
260    printf("match() called in line %d\n", __LINE__); \    printf("match() called in line %d\n", __LINE__); \
261    rx = match(ra,rb,rc,rd,re,rf,rg,rdepth+1); \    rrc = match(ra,rb,mstart,rc,rd,re,rf,rg,rdepth+1); \
262    printf("to line %d\n", __LINE__); \    printf("to line %d\n", __LINE__); \
263    }    }
264  #define RRETURN(ra) \  #define RRETURN(ra) \
# Line 208  versions and production versions. */ Line 267  versions and production versions. */
267    return ra; \    return ra; \
268    }    }
269  #else  #else
270  #define RMATCH(rx,ra,rb,rc,rd,re,rf,rg) \  #define RMATCH(ra,rb,rc,rd,re,rf,rg,rw) \
271    rx = match(ra,rb,rc,rd,re,rf,rg,rdepth+1)    rrc = match(ra,rb,mstart,rc,rd,re,rf,rg,rdepth+1)
272  #define RRETURN(ra) return ra  #define RRETURN(ra) return ra
273  #endif  #endif
274    
275  #else  #else
276    
277    
278  /* These versions of the macros manage a private stack on the heap. Note  /* These versions of the macros manage a private stack on the heap. Note that
279  that the rd argument of RMATCH isn't actually used. It's the md argument of  the "rd" argument of RMATCH isn't actually used in this definition. It's the md
280  match(), which never changes. */  argument of match(), which never changes. */
281    
282  #define REGISTER  #define REGISTER
283    
284  #define RMATCH(rx,ra,rb,rc,rd,re,rf,rg)\  #define RMATCH(ra,rb,rc,rd,re,rf,rg,rw)\
285    {\    {\
286    heapframe *newframe = (pcre_stack_malloc)(sizeof(heapframe));\    heapframe *newframe = (pcre_stack_malloc)(sizeof(heapframe));\
287    if (setjmp(frame->Xwhere) == 0)\    frame->Xwhere = rw; \
288      {\    newframe->Xeptr = ra;\
289      newframe->Xeptr = ra;\    newframe->Xecode = rb;\
290      newframe->Xecode = rb;\    newframe->Xmstart = mstart;\
291      newframe->Xoffset_top = rc;\    newframe->Xoffset_top = rc;\
292      newframe->Xims = re;\    newframe->Xims = re;\
293      newframe->Xeptrb = rf;\    newframe->Xeptrb = rf;\
294      newframe->Xflags = rg;\    newframe->Xflags = rg;\
295      newframe->Xrdepth = frame->Xrdepth + 1;\    newframe->Xrdepth = frame->Xrdepth + 1;\
296      newframe->Xprevframe = frame;\    newframe->Xprevframe = frame;\
297      frame = newframe;\    frame = newframe;\
298      DPRINTF(("restarting from line %d\n", __LINE__));\    DPRINTF(("restarting from line %d\n", __LINE__));\
299      goto HEAP_RECURSE;\    goto HEAP_RECURSE;\
300      }\    L_##rw:\
301    else\    DPRINTF(("jumped back to line %d\n", __LINE__));\
     {\  
     DPRINTF(("longjumped back to line %d\n", __LINE__));\  
     frame = md->thisframe;\  
     rx = frame->Xresult;\  
     }\  
302    }    }
303    
304  #define RRETURN(ra)\  #define RRETURN(ra)\
# Line 254  match(), which never changes. */ Line 308  match(), which never changes. */
308    (pcre_stack_free)(newframe);\    (pcre_stack_free)(newframe);\
309    if (frame != NULL)\    if (frame != NULL)\
310      {\      {\
311      frame->Xresult = ra;\      rrc = ra;\
312      md->thisframe = frame;\      goto HEAP_RETURN;\
     longjmp(frame->Xwhere, 1);\  
313      }\      }\
314    return ra;\    return ra;\
315    }    }
# Line 269  typedef struct heapframe { Line 322  typedef struct heapframe {
322    
323    /* Function arguments that may change */    /* Function arguments that may change */
324    
325    const uschar *Xeptr;    USPTR Xeptr;
326    const uschar *Xecode;    const uschar *Xecode;
327      USPTR Xmstart;
328    int Xoffset_top;    int Xoffset_top;
329    long int Xims;    long int Xims;
330    eptrblock *Xeptrb;    eptrblock *Xeptrb;
# Line 279  typedef struct heapframe { Line 333  typedef struct heapframe {
333    
334    /* Function local variables */    /* Function local variables */
335    
336    const uschar *Xcallpat;    USPTR Xcallpat;
337    const uschar *Xcharptr;  #ifdef SUPPORT_UTF8
338    const uschar *Xdata;    USPTR Xcharptr;
339    const uschar *Xnext;  #endif
340    const uschar *Xpp;    USPTR Xdata;
341    const uschar *Xprev;    USPTR Xnext;
342    const uschar *Xsaved_eptr;    USPTR Xpp;
343      USPTR Xprev;
344      USPTR Xsaved_eptr;
345    
346    recursion_info Xnew_recursive;    recursion_info Xnew_recursive;
347    
348    BOOL Xcur_is_word;    BOOL Xcur_is_word;
349    BOOL Xcondition;    BOOL Xcondition;
   BOOL Xminimize;  
350    BOOL Xprev_is_word;    BOOL Xprev_is_word;
351    
352    unsigned long int Xoriginal_ims;    unsigned long int Xoriginal_ims;
# Line 303  typedef struct heapframe { Line 358  typedef struct heapframe {
358    int Xprop_category;    int Xprop_category;
359    int Xprop_chartype;    int Xprop_chartype;
360    int Xprop_script;    int Xprop_script;
361    int *Xprop_test_variable;    int Xoclength;
362      uschar Xocchars[8];
363  #endif  #endif
364    
365      int Xcodelink;
366    int Xctype;    int Xctype;
367    int Xfc;    unsigned int Xfc;
368    int Xfi;    int Xfi;
369    int Xlength;    int Xlength;
370    int Xmax;    int Xmax;
# Line 321  typedef struct heapframe { Line 378  typedef struct heapframe {
378    
379    eptrblock Xnewptrb;    eptrblock Xnewptrb;
380    
381    /* Place to pass back result, and where to jump back to */    /* Where to jump back to */
382    
383    int  Xresult;    int Xwhere;
   jmp_buf Xwhere;  
384    
385  } heapframe;  } heapframe;
386    
# Line 340  typedef struct heapframe { Line 396  typedef struct heapframe {
396  *         Match from current position            *  *         Match from current position            *
397  *************************************************/  *************************************************/
398    
399  /* On entry ecode points to the first opcode, and eptr to the first character  /* This function is called recursively in many circumstances. Whenever it
 in the subject string, while eptrb holds the value of eptr at the start of the  
 last bracketed group - used for breaking infinite loops matching zero-length  
 strings. This function is called recursively in many circumstances. Whenever it  
400  returns a negative (error) response, the outer incarnation must also return the  returns a negative (error) response, the outer incarnation must also return the
401  same response.  same response. */
402    
403    /* These macros pack up tests that are used for partial matching, and which
404    appears several times in the code. We set the "hit end" flag if the pointer is
405    at the end of the subject and also past the start of the subject (i.e.
406    something has been matched). For hard partial matching, we then return
407    immediately. The second one is used when we already know we are past the end of
408    the subject. */
409    
410    #define CHECK_PARTIAL()\
411      if (md->partial != 0 && eptr >= md->end_subject && eptr > mstart)\
412        {\
413        md->hitend = TRUE;\
414        if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);\
415        }
416    
417    #define SCHECK_PARTIAL()\
418      if (md->partial != 0 && eptr > mstart)\
419        {\
420        md->hitend = TRUE;\
421        if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);\
422        }
423    
424  Performance note: It might be tempting to extract commonly used fields from the  
425  md structure (e.g. utf8, end_subject) into individual variables to improve  /* Performance note: It might be tempting to extract commonly used fields from
426    the md structure (e.g. utf8, end_subject) into individual variables to improve
427  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
428  made performance worse.  made performance worse.
429    
430  Arguments:  Arguments:
431     eptr        pointer in subject     eptr        pointer to current character in subject
432     ecode       position in code     ecode       pointer to current position in compiled code
433       mstart      pointer to the current match start position (can be modified
434                     by encountering \K)
435     offset_top  current top pointer     offset_top  current top pointer
436     md          pointer to "static" info for the match     md          pointer to "static" info for the match
437     ims         current /i, /m, and /s options     ims         current /i, /m, and /s options
# Line 362  Arguments: Line 439  Arguments:
439                   brackets - for testing for empty matches                   brackets - for testing for empty matches
440     flags       can contain     flags       can contain
441                   match_condassert - this is an assertion condition                   match_condassert - this is an assertion condition
442                   match_isgroup - this is the start of a bracketed group                   match_cbegroup - this is the start of an unlimited repeat
443                       group that can match an empty string
444     rdepth      the recursion depth     rdepth      the recursion depth
445    
446  Returns:       MATCH_MATCH if matched            )  these values are >= 0  Returns:       MATCH_MATCH if matched            )  these values are >= 0
# Line 372  Returns: MATCH_MATCH if matched Line 450  Returns: MATCH_MATCH if matched
450  */  */
451    
452  static int  static int
453  match(REGISTER USPTR eptr, REGISTER const uschar *ecode,  match(REGISTER USPTR eptr, REGISTER const uschar *ecode, USPTR mstart,
454    int offset_top, match_data *md, unsigned long int ims, eptrblock *eptrb,    int offset_top, match_data *md, unsigned long int ims, eptrblock *eptrb,
455    int flags, unsigned int rdepth)    int flags, unsigned int rdepth)
456  {  {
457  /* 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,
458  so they can be ordinary variables in all cases. Mark them with "register"  so they can be ordinary variables in all cases. Mark some of them with
459  because they are used a lot in loops. */  "register" because they are used a lot in loops. */
460    
461  register int  rrc;         /* Returns from recursive calls */  register int  rrc;         /* Returns from recursive calls */
462  register int  i;           /* Used for loops not involving calls to RMATCH() */  register int  i;           /* Used for loops not involving calls to RMATCH() */
463  register unsigned int  c;  /* Character values not kept over RMATCH() calls */  register unsigned int c;   /* Character values not kept over RMATCH() calls */
464  register BOOL utf8;        /* Local copy of UTF-8 flag for speed */  register BOOL utf8;        /* Local copy of UTF-8 flag for speed */
465    
466    BOOL minimize, possessive; /* Quantifier options */
467    int condcode;
468    
469  /* 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
470  preserved over calls to RMATCH() are part of a "frame" which is obtained from  preserved over calls to RMATCH() are part of a "frame" which is obtained from
471  heap storage. Set up the top-level frame here; others are obtained from the  heap storage. Set up the top-level frame here; others are obtained from the
# Line 398  frame->Xprevframe = NULL; /* Line 479  frame->Xprevframe = NULL; /*
479    
480  frame->Xeptr = eptr;  frame->Xeptr = eptr;
481  frame->Xecode = ecode;  frame->Xecode = ecode;
482    frame->Xmstart = mstart;
483  frame->Xoffset_top = offset_top;  frame->Xoffset_top = offset_top;
484  frame->Xims = ims;  frame->Xims = ims;
485  frame->Xeptrb = eptrb;  frame->Xeptrb = eptrb;
# Line 412  HEAP_RECURSE: Line 494  HEAP_RECURSE:
494    
495  #define eptr               frame->Xeptr  #define eptr               frame->Xeptr
496  #define ecode              frame->Xecode  #define ecode              frame->Xecode
497    #define mstart             frame->Xmstart
498  #define offset_top         frame->Xoffset_top  #define offset_top         frame->Xoffset_top
499  #define ims                frame->Xims  #define ims                frame->Xims
500  #define eptrb              frame->Xeptrb  #define eptrb              frame->Xeptrb
# Line 424  HEAP_RECURSE: Line 507  HEAP_RECURSE:
507  #define charptr            frame->Xcharptr  #define charptr            frame->Xcharptr
508  #endif  #endif
509  #define callpat            frame->Xcallpat  #define callpat            frame->Xcallpat
510    #define codelink           frame->Xcodelink
511  #define data               frame->Xdata  #define data               frame->Xdata
512  #define next               frame->Xnext  #define next               frame->Xnext
513  #define pp                 frame->Xpp  #define pp                 frame->Xpp
# Line 434  HEAP_RECURSE: Line 518  HEAP_RECURSE:
518    
519  #define cur_is_word        frame->Xcur_is_word  #define cur_is_word        frame->Xcur_is_word
520  #define condition          frame->Xcondition  #define condition          frame->Xcondition
 #define minimize           frame->Xminimize  
521  #define prev_is_word       frame->Xprev_is_word  #define prev_is_word       frame->Xprev_is_word
522    
523  #define original_ims       frame->Xoriginal_ims  #define original_ims       frame->Xoriginal_ims
# Line 446  HEAP_RECURSE: Line 529  HEAP_RECURSE:
529  #define prop_category      frame->Xprop_category  #define prop_category      frame->Xprop_category
530  #define prop_chartype      frame->Xprop_chartype  #define prop_chartype      frame->Xprop_chartype
531  #define prop_script        frame->Xprop_script  #define prop_script        frame->Xprop_script
532  #define prop_test_variable frame->Xprop_test_variable  #define oclength           frame->Xoclength
533    #define occhars            frame->Xocchars
534  #endif  #endif
535    
536  #define ctype              frame->Xctype  #define ctype              frame->Xctype
# Line 470  HEAP_RECURSE: Line 554  HEAP_RECURSE:
554  get preserved during recursion in the normal way. In this environment, fi and  get preserved during recursion in the normal way. In this environment, fi and
555  i, and fc and c, can be the same variables. */  i, and fc and c, can be the same variables. */
556    
557  #else  #else         /* NO_RECURSE not defined */
558  #define fi i  #define fi i
559  #define fc c  #define fc c
560    
# Line 489  recursion_info new_recursive; /* wi Line 573  recursion_info new_recursive; /* wi
573                                     /* that do not have to be preserved over  */                                     /* that do not have to be preserved over  */
574  BOOL cur_is_word;                  /* a recursive call to RMATCH().          */  BOOL cur_is_word;                  /* a recursive call to RMATCH().          */
575  BOOL condition;  BOOL condition;
 BOOL minimize;  
576  BOOL prev_is_word;  BOOL prev_is_word;
577    
578  unsigned long int original_ims;  unsigned long int original_ims;
# Line 501  int prop_fail_result; Line 584  int prop_fail_result;
584  int prop_category;  int prop_category;
585  int prop_chartype;  int prop_chartype;
586  int prop_script;  int prop_script;
587  int *prop_test_variable;  int oclength;
588    uschar occhars[8];
589  #endif  #endif
590    
591    int codelink;
592  int ctype;  int ctype;
593  int length;  int length;
594  int max;  int max;
# Line 516  int save_offset1, save_offset2, save_off Line 601  int save_offset1, save_offset2, save_off
601  int stacksave[REC_STACK_SAVE_MAX];  int stacksave[REC_STACK_SAVE_MAX];
602    
603  eptrblock newptrb;  eptrblock newptrb;
604  #endif  #endif     /* NO_RECURSE */
605    
606  /* These statements are here to stop the compiler complaining about unitialized  /* These statements are here to stop the compiler complaining about unitialized
607  variables. */  variables. */
# Line 524  variables. */ Line 609  variables. */
609  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
610  prop_value = 0;  prop_value = 0;
611  prop_fail_result = 0;  prop_fail_result = 0;
 prop_test_variable = NULL;  
612  #endif  #endif
613    
614    
615  /* This label is used for tail recursion, which is used in a few cases even  /* This label is used for tail recursion, which is used in a few cases even
616  when NO_RECURSE is not defined, in order to reduce the amount of stack that is  when NO_RECURSE is not defined, in order to reduce the amount of stack that is
617  used. Thanks to Ian Taylor for noticing this possibility and sending the  used. Thanks to Ian Taylor for noticing this possibility and sending the
# Line 542  defined). However, RMATCH isn't like a f Line 627  defined). However, RMATCH isn't like a f
627  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,
628  however, impact performance when true recursion is being used. */  however, impact performance when true recursion is being used. */
629    
630    #ifdef SUPPORT_UTF8
631    utf8 = md->utf8;       /* Local copy of the flag */
632    #else
633    utf8 = FALSE;
634    #endif
635    
636  /* 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
637  haven't exceeded the recursive call limit. */  haven't exceeded the recursive call limit. */
638    
# Line 550  if (rdepth >= md->match_limit_recursion) Line 641  if (rdepth >= md->match_limit_recursion)
641    
642  original_ims = ims;    /* Save for resetting on ')' */  original_ims = ims;    /* Save for resetting on ')' */
643    
644  #ifdef SUPPORT_UTF8  /* At the start of a group with an unlimited repeat that may match an empty
645  utf8 = md->utf8;       /* Local copy of the flag */  string, the match_cbegroup flag is set. When this is the case, add the current
646  #else  subject pointer to the chain of such remembered pointers, to be checked when we
647  utf8 = FALSE;  hit the closing ket, in order to break infinite loops that match no characters.
648  #endif  When match() is called in other circumstances, don't add to the chain. The
649    match_cbegroup flag must NOT be used with tail recursion, because the memory
650  /* At the start of a bracketed group, add the current subject pointer to the  block that is used is on the stack, so a new one may be required for each
651  stack of such pointers, to be re-instated at the end of the group when we hit  match(). */
 the closing ket. When match() is called in other circumstances, we don't add to  
 this stack. */  
652    
653  if ((flags & match_isgroup) != 0)  if ((flags & match_cbegroup) != 0)
654    {    {
   newptrb.epb_prev = eptrb;  
655    newptrb.epb_saved_eptr = eptr;    newptrb.epb_saved_eptr = eptr;
656      newptrb.epb_prev = eptrb;
657    eptrb = &newptrb;    eptrb = &newptrb;
658    }    }
659    
660  /* Now start processing the operations. */  /* Now start processing the opcodes. */
661    
662  for (;;)  for (;;)
663    {    {
664      minimize = possessive = FALSE;
665    op = *ecode;    op = *ecode;
   minimize = FALSE;  
   
   /* For partial matching, remember if we ever hit the end of the subject after  
   matching at least one subject character. */  
666    
667    if (md->partial &&    switch(op)
       eptr >= md->end_subject &&  
       eptr > md->start_match)  
     md->hitend = TRUE;  
   
   /* Opening capturing bracket. If there is space in the offset vector, save  
   the current subject position in the working slot at the top of the vector. We  
   mustn't change the current values of the data slot, because they may be set  
   from a previous iteration of this group, and be referred to by a reference  
   inside the group.  
   
   If the bracket fails to match, we need to restore this value and also the  
   values of the final offsets, in case they were set by a previous iteration of  
   the same bracket.  
   
   If there isn't enough space in the offset vector, treat this as if it were a  
   non-capturing bracket. Don't worry about setting the flag for the error case  
   here; that is handled in the code for KET. */  
   
   if (op > OP_BRA)  
668      {      {
669      number = op - OP_BRA;      case OP_FAIL:
670        RRETURN(MATCH_NOMATCH);
671    
672      /* For extended extraction brackets (large number), we have to fish out the      case OP_PRUNE:
673      number from a dummy opcode at the start. */      RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
674          ims, eptrb, flags, RM51);
675      if (number > EXTRACT_BASIC_MAX)      if (rrc != MATCH_NOMATCH) RRETURN(rrc);
676        number = GET2(ecode, 2+LINK_SIZE);      RRETURN(MATCH_PRUNE);
677    
678        case OP_COMMIT:
679        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
680          ims, eptrb, flags, RM52);
681        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
682        RRETURN(MATCH_COMMIT);
683    
684        case OP_SKIP:
685        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
686          ims, eptrb, flags, RM53);
687        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
688        md->start_match_ptr = eptr;   /* Pass back current position */
689        RRETURN(MATCH_SKIP);
690    
691        case OP_THEN:
692        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
693          ims, eptrb, flags, RM54);
694        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
695        RRETURN(MATCH_THEN);
696    
697        /* Handle a capturing bracket. If there is space in the offset vector, save
698        the current subject position in the working slot at the top of the vector.
699        We mustn't change the current values of the data slot, because they may be
700        set from a previous iteration of this group, and be referred to by a
701        reference inside the group.
702    
703        If the bracket fails to match, we need to restore this value and also the
704        values of the final offsets, in case they were set by a previous iteration
705        of the same bracket.
706    
707        If there isn't enough space in the offset vector, treat this as if it were
708        a non-capturing bracket. Don't worry about setting the flag for the error
709        case here; that is handled in the code for KET. */
710    
711        case OP_CBRA:
712        case OP_SCBRA:
713        number = GET2(ecode, 1+LINK_SIZE);
714      offset = number << 1;      offset = number << 1;
715    
716  #ifdef DEBUG  #ifdef DEBUG
717      printf("start bracket %d subject=", number);      printf("start bracket %d\n", number);
718        printf("subject=");
719      pchars(eptr, 16, TRUE, md);      pchars(eptr, 16, TRUE, md);
720      printf("\n");      printf("\n");
721  #endif  #endif
# Line 624  for (;;) Line 730  for (;;)
730        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));        DPRINTF(("saving %d %d %d\n", save_offset1, save_offset2, save_offset3));
731        md->offset_vector[md->offset_end - number] = eptr - md->start_subject;        md->offset_vector[md->offset_end - number] = eptr - md->start_subject;
732    
733          flags = (op == OP_SCBRA)? match_cbegroup : 0;
734        do        do
735          {          {
736          RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb,          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
737            match_isgroup);            ims, eptrb, flags, RM1);
738          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
739          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
740          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
741          }          }
# Line 643  for (;;) Line 750  for (;;)
750        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
751        }        }
752    
753      /* Insufficient room for saving captured contents */      /* FALL THROUGH ... Insufficient room for saving captured contents. Treat
754        as a non-capturing bracket. */
     else op = OP_BRA;  
     }  
   
   /* Other types of node can be handled by a switch */  
755    
756    switch(op)      /* VVVVVVVVVVVVVVVVVVVVVVVVV */
757      {      /* VVVVVVVVVVVVVVVVVVVVVVVVV */
     case OP_BRA:     /* Non-capturing bracket: optimized */  
     DPRINTF(("start bracket 0\n"));  
758    
759      /* Loop for all the alternatives */      DPRINTF(("insufficient capture room: treat as non-capturing\n"));
760    
761        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
762        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
763    
764        /* Non-capturing bracket. Loop for all the alternatives. When we get to the
765        final alternative within the brackets, we would return the result of a
766        recursive call to match() whatever happened. We can reduce stack usage by
767        turning this into a tail recursion, except in the case when match_cbegroup
768        is set.*/
769    
770        case OP_BRA:
771        case OP_SBRA:
772        DPRINTF(("start non-capturing bracket\n"));
773        flags = (op >= OP_SBRA)? match_cbegroup : 0;
774      for (;;)      for (;;)
775        {        {
776        /* When we get to the final alternative within the brackets, we would        if (ecode[GET(ecode, 1)] != OP_ALT)   /* Final alternative */
777        return the result of a recursive call to match() whatever happened. We          {
778        can reduce stack usage by turning this into a tail recursion. */          if (flags == 0)    /* Not a possibly empty group */
779              {
780        if (ecode[GET(ecode, 1)] != OP_ALT)            ecode += _pcre_OP_lengths[*ecode];
781         {            DPRINTF(("bracket 0 tail recursion\n"));
782         ecode += 1 + LINK_SIZE;            goto TAIL_RECURSE;
783         flags = match_isgroup;            }
784         DPRINTF(("bracket 0 tail recursion\n"));  
785         goto TAIL_RECURSE;          /* Possibly empty group; can't use tail recursion. */
786         }  
787            RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,
788              eptrb, flags, RM48);
789            RRETURN(rrc);
790            }
791    
792        /* For non-final alternatives, continue the loop for a NOMATCH result;        /* For non-final alternatives, continue the loop for a NOMATCH result;
793        otherwise return. */        otherwise return. */
794    
795        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb,        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,
796          match_isgroup);          eptrb, flags, RM2);
797        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
798        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
799        }        }
800      /* Control never reaches here. */      /* Control never reaches here. */
# Line 688  for (;;) Line 806  for (;;)
806      obeyed, we can use tail recursion to avoid using another stack frame. */      obeyed, we can use tail recursion to avoid using another stack frame. */
807    
808      case OP_COND:      case OP_COND:
809      if (ecode[LINK_SIZE+1] == OP_CREF) /* Condition extract or recurse test */      case OP_SCOND:
810        codelink= GET(ecode, 1);
811    
812        /* Because of the way auto-callout works during compile, a callout item is
813        inserted between OP_COND and an assertion condition. */
814    
815        if (ecode[LINK_SIZE+1] == OP_CALLOUT)
816          {
817          if (pcre_callout != NULL)
818            {
819            pcre_callout_block cb;
820            cb.version          = 1;   /* Version 1 of the callout block */
821            cb.callout_number   = ecode[LINK_SIZE+2];
822            cb.offset_vector    = md->offset_vector;
823            cb.subject          = (PCRE_SPTR)md->start_subject;
824            cb.subject_length   = md->end_subject - md->start_subject;
825            cb.start_match      = mstart - md->start_subject;
826            cb.current_position = eptr - md->start_subject;
827            cb.pattern_position = GET(ecode, LINK_SIZE + 3);
828            cb.next_item_length = GET(ecode, 3 + 2*LINK_SIZE);
829            cb.capture_top      = offset_top/2;
830            cb.capture_last     = md->capture_last;
831            cb.callout_data     = md->callout_data;
832            if ((rrc = (*pcre_callout)(&cb)) > 0) RRETURN(MATCH_NOMATCH);
833            if (rrc < 0) RRETURN(rrc);
834            }
835          ecode += _pcre_OP_lengths[OP_CALLOUT];
836          }
837    
838        condcode = ecode[LINK_SIZE+1];
839    
840        /* Now see what the actual condition is */
841    
842        if (condcode == OP_RREF || condcode == OP_NRREF)    /* Recursion test */
843          {
844          if (md->recursive == NULL)                /* Not recursing => FALSE */
845            {
846            condition = FALSE;
847            ecode += GET(ecode, 1);
848            }
849          else
850            {
851            int recno = GET2(ecode, LINK_SIZE + 2);   /* Recursion group number*/
852            condition =  (recno == RREF_ANY || recno == md->recursive->group_num);
853    
854            /* If the test is for recursion into a specific subpattern, and it is
855            false, but the test was set up by name, scan the table to see if the
856            name refers to any other numbers, and test them. The condition is true
857            if any one is set. */
858    
859            if (!condition && condcode == OP_NRREF && recno != RREF_ANY)
860              {
861              uschar *slotA = md->name_table;
862              for (i = 0; i < md->name_count; i++)
863                {
864                if (GET2(slotA, 0) == recno) break;
865                slotA += md->name_entry_size;
866                }
867    
868              /* Found a name for the number - there can be only one; duplicate
869              names for different numbers are allowed, but not vice versa. First
870              scan down for duplicates. */
871    
872              if (i < md->name_count)
873                {
874                uschar *slotB = slotA;
875                while (slotB > md->name_table)
876                  {
877                  slotB -= md->name_entry_size;
878                  if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)
879                    {
880                    condition = GET2(slotB, 0) == md->recursive->group_num;
881                    if (condition) break;
882                    }
883                  else break;
884                  }
885    
886                /* Scan up for duplicates */
887    
888                if (!condition)
889                  {
890                  slotB = slotA;
891                  for (i++; i < md->name_count; i++)
892                    {
893                    slotB += md->name_entry_size;
894                    if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)
895                      {
896                      condition = GET2(slotB, 0) == md->recursive->group_num;
897                      if (condition) break;
898                      }
899                    else break;
900                    }
901                  }
902                }
903              }
904    
905            /* Chose branch according to the condition */
906    
907            ecode += condition? 3 : GET(ecode, 1);
908            }
909          }
910    
911        else if (condcode == OP_CREF || condcode == OP_NCREF)  /* Group used test */
912        {        {
913        offset = GET2(ecode, LINK_SIZE+2) << 1;  /* Doubled ref number */        offset = GET2(ecode, LINK_SIZE+2) << 1;  /* Doubled ref number */
914        condition = (offset == CREF_RECURSE * 2)?        condition = offset < offset_top && md->offset_vector[offset] >= 0;
915          (md->recursive != NULL) :  
916          (offset < offset_top && md->offset_vector[offset] >= 0);        /* If the numbered capture is unset, but the reference was by name,
917        ecode += condition? (LINK_SIZE + 4) : (LINK_SIZE + 1 + GET(ecode, 1));        scan the table to see if the name refers to any other numbers, and test
918        flags = match_isgroup;        them. The condition is true if any one is set. This is tediously similar
919        goto TAIL_RECURSE;        to the code above, but not close enough to try to amalgamate. */
920    
921          if (!condition && condcode == OP_NCREF)
922            {
923            int refno = offset >> 1;
924            uschar *slotA = md->name_table;
925    
926            for (i = 0; i < md->name_count; i++)
927              {
928              if (GET2(slotA, 0) == refno) break;
929              slotA += md->name_entry_size;
930              }
931    
932            /* Found a name for the number - there can be only one; duplicate names
933            for different numbers are allowed, but not vice versa. First scan down
934            for duplicates. */
935    
936            if (i < md->name_count)
937              {
938              uschar *slotB = slotA;
939              while (slotB > md->name_table)
940                {
941                slotB -= md->name_entry_size;
942                if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)
943                  {
944                  offset = GET2(slotB, 0) << 1;
945                  condition = offset < offset_top &&
946                    md->offset_vector[offset] >= 0;
947                  if (condition) break;
948                  }
949                else break;
950                }
951    
952              /* Scan up for duplicates */
953    
954              if (!condition)
955                {
956                slotB = slotA;
957                for (i++; i < md->name_count; i++)
958                  {
959                  slotB += md->name_entry_size;
960                  if (strcmp((char *)slotA + 2, (char *)slotB + 2) == 0)
961                    {
962                    offset = GET2(slotB, 0) << 1;
963                    condition = offset < offset_top &&
964                      md->offset_vector[offset] >= 0;
965                    if (condition) break;
966                    }
967                  else break;
968                  }
969                }
970              }
971            }
972    
973          /* Chose branch according to the condition */
974    
975          ecode += condition? 3 : GET(ecode, 1);
976          }
977    
978        else if (condcode == OP_DEF)     /* DEFINE - always false */
979          {
980          condition = FALSE;
981          ecode += GET(ecode, 1);
982        }        }
983    
984      /* The condition is an assertion. Call match() to evaluate it - setting      /* The condition is an assertion. Call match() to evaluate it - setting
985      the final argument TRUE causes it to stop at the end of an assertion. */      the final argument match_condassert causes it to stop at the end of an
986        assertion. */
987    
988      else      else
989        {        {
990        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL,
991            match_condassert | match_isgroup);            match_condassert, RM3);
992        if (rrc == MATCH_MATCH)        if (rrc == MATCH_MATCH)
993          {          {
994          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE+2);          condition = TRUE;
995            ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);
996          while (*ecode == OP_ALT) ecode += GET(ecode, 1);          while (*ecode == OP_ALT) ecode += GET(ecode, 1);
997          }          }
998        else if (rrc != MATCH_NOMATCH)        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
999          {          {
1000          RRETURN(rrc);         /* Need braces because of following else */          RRETURN(rrc);         /* Need braces because of following else */
1001          }          }
1002        else ecode += GET(ecode, 1);        else
1003            {
1004            condition = FALSE;
1005            ecode += codelink;
1006            }
1007          }
1008    
1009        /* We are now at the branch that is to be obeyed. As there is only one,      /* We are now at the branch that is to be obeyed. As there is only one,
1010        we can use tail recursion to avoid using another stack frame. */      we can use tail recursion to avoid using another stack frame, except when
1011        match_cbegroup is required for an unlimited repeat of a possibly empty
1012        group. If the second alternative doesn't exist, we can just plough on. */
1013    
1014        if (condition || *ecode == OP_ALT)
1015          {
1016        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1017        flags = match_isgroup;        if (op == OP_SCOND)        /* Possibly empty group */
1018        goto TAIL_RECURSE;          {
1019            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, match_cbegroup, RM49);
1020            RRETURN(rrc);
1021            }
1022          else                       /* Group must match something */
1023            {
1024            flags = 0;
1025            goto TAIL_RECURSE;
1026            }
1027        }        }
1028      /* Control never reaches here */      else                         /* Condition false & no alternative */
1029          {
1030          ecode += 1 + LINK_SIZE;
1031          }
1032        break;
1033    
1034    
1035        /* Before OP_ACCEPT there may be any number of OP_CLOSE opcodes,
1036        to close any currently open capturing brackets. */
1037    
1038        case OP_CLOSE:
1039        number = GET2(ecode, 1);
1040        offset = number << 1;
1041    
1042      /* Skip over conditional reference or large extraction number data if  #ifdef DEBUG
1043      encountered. */        printf("end bracket %d at *ACCEPT", number);
1044          printf("\n");
1045    #endif
1046    
1047      case OP_CREF:      md->capture_last = number;
1048      case OP_BRANUMBER:      if (offset >= md->offset_max) md->offset_overflow = TRUE; else
1049          {
1050          md->offset_vector[offset] =
1051            md->offset_vector[md->offset_end - number];
1052          md->offset_vector[offset+1] = eptr - md->start_subject;
1053          if (offset_top <= offset) offset_top = offset + 2;
1054          }
1055      ecode += 3;      ecode += 3;
1056      break;      break;
1057    
     /* End of the pattern. If we are in a recursion, we should restore the  
     offsets appropriately and continue from after the call. */  
1058    
1059        /* End of the pattern, either real or forced. If we are in a top-level
1060        recursion, we should restore the offsets appropriately and continue from
1061        after the call. */
1062    
1063        case OP_ACCEPT:
1064      case OP_END:      case OP_END:
1065      if (md->recursive != NULL && md->recursive->group_num == 0)      if (md->recursive != NULL && md->recursive->group_num == 0)
1066        {        {
# Line 745  for (;;) Line 1069  for (;;)
1069        md->recursive = rec->prevrec;        md->recursive = rec->prevrec;
1070        memmove(md->offset_vector, rec->offset_save,        memmove(md->offset_vector, rec->offset_save,
1071          rec->saved_max * sizeof(int));          rec->saved_max * sizeof(int));
1072        md->start_match = rec->save_start;        offset_top = rec->save_offset_top;
1073          mstart = rec->save_start;
1074        ims = original_ims;        ims = original_ims;
1075        ecode = rec->after_call;        ecode = rec->after_call;
1076        break;        break;
1077        }        }
1078    
1079      /* Otherwise, if PCRE_NOTEMPTY is set, fail if we have matched an empty      /* Otherwise, if we have matched an empty string, fail if PCRE_NOTEMPTY is
1080      string - backtracking will then try other alternatives, if any. */      set, or if PCRE_NOTEMPTY_ATSTART is set and we have matched at the start of
1081        the subject. In both cases, backtracking will then try other alternatives,
1082        if any. */
1083    
1084        if (eptr == mstart &&
1085            (md->notempty ||
1086              (md->notempty_atstart &&
1087                mstart == md->start_subject + md->start_offset)))
1088          RRETURN(MATCH_NOMATCH);
1089    
1090        /* Otherwise, we have a match. */
1091    
1092      if (md->notempty && eptr == md->start_match) RRETURN(MATCH_NOMATCH);      md->end_match_ptr = eptr;           /* Record where we ended */
1093      md->end_match_ptr = eptr;          /* Record where we ended */      md->end_offset_top = offset_top;    /* and how many extracts were taken */
1094      md->end_offset_top = offset_top;   /* and how many extracts were taken */      md->start_match_ptr = mstart;       /* and the start (\K can modify) */
1095      RRETURN(MATCH_MATCH);      RRETURN(MATCH_MATCH);
1096    
1097      /* Change option settings */      /* Change option settings */
# Line 777  for (;;) Line 1112  for (;;)
1112      case OP_ASSERTBACK:      case OP_ASSERTBACK:
1113      do      do
1114        {        {
1115        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,
1116          match_isgroup);          RM4);
1117        if (rrc == MATCH_MATCH) break;        if (rrc == MATCH_MATCH) break;
1118        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1119        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
1120        }        }
1121      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 804  for (;;) Line 1139  for (;;)
1139      case OP_ASSERTBACK_NOT:      case OP_ASSERTBACK_NOT:
1140      do      do
1141        {        {
1142        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,
1143          match_isgroup);          RM5);
1144        if (rrc == MATCH_MATCH) RRETURN(MATCH_NOMATCH);        if (rrc == MATCH_MATCH) RRETURN(MATCH_NOMATCH);
1145        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1146        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1147        }        }
1148      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 826  for (;;) Line 1161  for (;;)
1161  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
1162      if (utf8)      if (utf8)
1163        {        {
1164        c = GET(ecode,1);        i = GET(ecode, 1);
1165        for (i = 0; i < c; i++)        while (i-- > 0)
1166          {          {
1167          eptr--;          eptr--;
1168          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1169          BACKCHAR(eptr)          BACKCHAR(eptr);
1170          }          }
1171        }        }
1172      else      else
# Line 840  for (;;) Line 1175  for (;;)
1175      /* No UTF-8 support, or not in UTF-8 mode: count is byte count */      /* No UTF-8 support, or not in UTF-8 mode: count is byte count */
1176    
1177        {        {
1178        eptr -= GET(ecode,1);        eptr -= GET(ecode, 1);
1179        if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);        if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
1180        }        }
1181    
1182      /* Skip to next op code */      /* Save the earliest consulted character, then skip to next op code */
1183    
1184        if (eptr < md->start_used_ptr) md->start_used_ptr = eptr;
1185      ecode += 1 + LINK_SIZE;      ecode += 1 + LINK_SIZE;
1186      break;      break;
1187    
# Line 862  for (;;) Line 1198  for (;;)
1198        cb.offset_vector    = md->offset_vector;        cb.offset_vector    = md->offset_vector;
1199        cb.subject          = (PCRE_SPTR)md->start_subject;        cb.subject          = (PCRE_SPTR)md->start_subject;
1200        cb.subject_length   = md->end_subject - md->start_subject;        cb.subject_length   = md->end_subject - md->start_subject;
1201        cb.start_match      = md->start_match - md->start_subject;        cb.start_match      = mstart - md->start_subject;
1202        cb.current_position = eptr - md->start_subject;        cb.current_position = eptr - md->start_subject;
1203        cb.pattern_position = GET(ecode, 2);        cb.pattern_position = GET(ecode, 2);
1204        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);        cb.next_item_length = GET(ecode, 2 + LINK_SIZE);
# Line 897  for (;;) Line 1233  for (;;)
1233      case OP_RECURSE:      case OP_RECURSE:
1234        {        {
1235        callpat = md->start_code + GET(ecode, 1);        callpat = md->start_code + GET(ecode, 1);
1236        new_recursive.group_num = *callpat - OP_BRA;        new_recursive.group_num = (callpat == md->start_code)? 0 :
1237            GET2(callpat, 1 + LINK_SIZE);
       /* For extended extraction brackets (large number), we have to fish out  
       the number from a dummy opcode at the start. */  
   
       if (new_recursive.group_num > EXTRACT_BASIC_MAX)  
         new_recursive.group_num = GET2(callpat, 2+LINK_SIZE);  
1238    
1239        /* Add to "recursing stack" */        /* Add to "recursing stack" */
1240    
# Line 929  for (;;) Line 1260  for (;;)
1260    
1261        memcpy(new_recursive.offset_save, md->offset_vector,        memcpy(new_recursive.offset_save, md->offset_vector,
1262              new_recursive.saved_max * sizeof(int));              new_recursive.saved_max * sizeof(int));
1263        new_recursive.save_start = md->start_match;        new_recursive.save_start = mstart;
1264        md->start_match = eptr;        new_recursive.save_offset_top = offset_top;
1265          mstart = eptr;
1266    
1267        /* OK, now we can do the recursion. For each top-level alternative we        /* OK, now we can do the recursion. For each top-level alternative we
1268        restore the offset and recursion data. */        restore the offset and recursion data. */
1269    
1270        DPRINTF(("Recursing into group %d\n", new_recursive.group_num));        DPRINTF(("Recursing into group %d\n", new_recursive.group_num));
1271          flags = (*callpat >= OP_SBRA)? match_cbegroup : 0;
1272        do        do
1273          {          {
1274          RMATCH(rrc, eptr, callpat + 1 + LINK_SIZE, offset_top, md, ims,          RMATCH(eptr, callpat + _pcre_OP_lengths[*callpat], offset_top,
1275              eptrb, match_isgroup);            md, ims, eptrb, flags, RM6);
1276          if (rrc == MATCH_MATCH)          if (rrc == MATCH_MATCH)
1277            {            {
1278            DPRINTF(("Recursion matched\n"));            DPRINTF(("Recursion matched\n"));
# Line 948  for (;;) Line 1281  for (;;)
1281              (pcre_free)(new_recursive.offset_save);              (pcre_free)(new_recursive.offset_save);
1282            RRETURN(MATCH_MATCH);            RRETURN(MATCH_MATCH);
1283            }            }
1284          else if (rrc != MATCH_NOMATCH)          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
1285            {            {
1286            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1287              if (new_recursive.offset_save != stacksave)
1288                (pcre_free)(new_recursive.offset_save);
1289            RRETURN(rrc);            RRETURN(rrc);
1290            }            }
1291    
# Line 982  for (;;) Line 1317  for (;;)
1317    
1318      do      do
1319        {        {
1320        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM7);
         eptrb, match_isgroup);  
1321        if (rrc == MATCH_MATCH) break;        if (rrc == MATCH_MATCH) break;
1322        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1323        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1324        }        }
1325      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 997  for (;;) Line 1331  for (;;)
1331      /* Continue as from after the assertion, updating the offsets high water      /* Continue as from after the assertion, updating the offsets high water
1332      mark, since extracts may have been taken. */      mark, since extracts may have been taken. */
1333    
1334      do ecode += GET(ecode,1); while (*ecode == OP_ALT);      do ecode += GET(ecode, 1); while (*ecode == OP_ALT);
1335    
1336      offset_top = md->end_offset_top;      offset_top = md->end_offset_top;
1337      eptr = md->end_match_ptr;      eptr = md->end_match_ptr;
# Line 1028  for (;;) Line 1362  for (;;)
1362    
1363      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1364        {        {
1365        RMATCH(rrc, eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM8);
1366        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1367        ecode = prev;        ecode = prev;
1368        flags = match_isgroup;        flags = 0;
1369        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1370        }        }
1371      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
1372        {        {
1373        RMATCH(rrc, eptr, prev, offset_top, md, ims, eptrb, match_isgroup);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, match_cbegroup, RM9);
1374        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1375        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1376        flags = 0;        flags = 0;
# Line 1051  for (;;) Line 1385  for (;;)
1385      do ecode += GET(ecode,1); while (*ecode == OP_ALT);      do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1386      break;      break;
1387    
1388      /* BRAZERO and BRAMINZERO occur just before a bracket group, indicating      /* BRAZERO, BRAMINZERO and SKIPZERO occur just before a bracket group,
1389      that it may occur zero times. It may repeat infinitely, or not at all -      indicating that it may occur zero times. It may repeat infinitely, or not
1390      i.e. it could be ()* or ()? in the pattern. Brackets with fixed upper      at all - i.e. it could be ()* or ()? or even (){0} in the pattern. Brackets
1391      repeat limits are compiled as a number of copies, with the optional ones      with fixed upper repeat limits are compiled as a number of copies, with the
1392      preceded by BRAZERO or BRAMINZERO. */      optional ones preceded by BRAZERO or BRAMINZERO. */
1393    
1394      case OP_BRAZERO:      case OP_BRAZERO:
1395        {        {
1396        next = ecode+1;        next = ecode+1;
1397        RMATCH(rrc, eptr, next, offset_top, md, ims, eptrb, match_isgroup);        RMATCH(eptr, next, offset_top, md, ims, eptrb, 0, RM10);
1398        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1399        do next += GET(next,1); while (*next == OP_ALT);        do next += GET(next,1); while (*next == OP_ALT);
1400        ecode = next + 1+LINK_SIZE;        ecode = next + 1 + LINK_SIZE;
1401        }        }
1402      break;      break;
1403    
1404      case OP_BRAMINZERO:      case OP_BRAMINZERO:
1405        {        {
1406        next = ecode+1;        next = ecode+1;
1407        do next += GET(next,1); while (*next == OP_ALT);        do next += GET(next, 1); while (*next == OP_ALT);
1408        RMATCH(rrc, eptr, next + 1+LINK_SIZE, offset_top, md, ims, eptrb,        RMATCH(eptr, next + 1+LINK_SIZE, offset_top, md, ims, eptrb, 0, RM11);
         match_isgroup);  
1409        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1410        ecode++;        ecode++;
1411        }        }
1412      break;      break;
1413    
1414      /* End of a group, repeated or non-repeating. If we are at the end of      case OP_SKIPZERO:
1415      an assertion "group", stop matching and return MATCH_MATCH, but record the        {
1416      current high water mark for use by positive assertions. Do this also        next = ecode+1;
1417      for the "once" (not-backup up) groups. */        do next += GET(next,1); while (*next == OP_ALT);
1418          ecode = next + 1 + LINK_SIZE;
1419          }
1420        break;
1421    
1422        /* End of a group, repeated or non-repeating. */
1423    
1424      case OP_KET:      case OP_KET:
1425      case OP_KETRMIN:      case OP_KETRMIN:
1426      case OP_KETRMAX:      case OP_KETRMAX:
1427      prev = ecode - GET(ecode, 1);      prev = ecode - GET(ecode, 1);
     saved_eptr = eptrb->epb_saved_eptr;  
1428    
1429      /* Back up the stack of bracket start pointers. */      /* If this was a group that remembered the subject start, in order to break
1430        infinite repeats of empty string matches, retrieve the subject start from
1431        the chain. Otherwise, set it NULL. */
1432    
1433      eptrb = eptrb->epb_prev;      if (*prev >= OP_SBRA)
1434          {
1435          saved_eptr = eptrb->epb_saved_eptr;   /* Value at start of group */
1436          eptrb = eptrb->epb_prev;              /* Backup to previous group */
1437          }
1438        else saved_eptr = NULL;
1439    
1440        /* If we are at the end of an assertion group, stop matching and return
1441        MATCH_MATCH, but record the current high water mark for use by positive
1442        assertions. Do this also for the "once" (atomic) groups. */
1443    
1444      if (*prev == OP_ASSERT || *prev == OP_ASSERT_NOT ||      if (*prev == OP_ASSERT || *prev == OP_ASSERT_NOT ||
1445          *prev == OP_ASSERTBACK || *prev == OP_ASSERTBACK_NOT ||          *prev == OP_ASSERTBACK || *prev == OP_ASSERTBACK_NOT ||
# Line 1102  for (;;) Line 1450  for (;;)
1450        RRETURN(MATCH_MATCH);        RRETURN(MATCH_MATCH);
1451        }        }
1452    
1453      /* In all other cases except a conditional group we have to check the      /* For capturing groups we have to check the group number back at the start
1454      group number back at the start and if necessary complete handling an      and if necessary complete handling an extraction by setting the offsets and
1455      extraction by setting the offsets and bumping the high water mark. */      bumping the high water mark. Note that whole-pattern recursion is coded as
1456        a recurse into group 0, so it won't be picked up here. Instead, we catch it
1457        when the OP_END is reached. Other recursion is handled here. */
1458    
1459      if (*prev != OP_COND)      if (*prev == OP_CBRA || *prev == OP_SCBRA)
1460        {        {
1461        number = *prev - OP_BRA;        number = GET2(prev, 1+LINK_SIZE);
   
       /* For extended extraction brackets (large number), we have to fish out  
       the number from a dummy opcode at the start. */  
   
       if (number > EXTRACT_BASIC_MAX) number = GET2(prev, 2+LINK_SIZE);  
1462        offset = number << 1;        offset = number << 1;
1463    
1464  #ifdef DEBUG  #ifdef DEBUG
# Line 1121  for (;;) Line 1466  for (;;)
1466        printf("\n");        printf("\n");
1467  #endif  #endif
1468    
1469        /* Test for a numbered group. This includes groups called as a result        md->capture_last = number;
1470        of recursion. Note that whole-pattern recursion is coded as a recurse        if (offset >= md->offset_max) md->offset_overflow = TRUE; else
       into group 0, so it won't be picked up here. Instead, we catch it when  
       the OP_END is reached. */  
   
       if (number > 0)  
1471          {          {
1472          md->capture_last = number;          md->offset_vector[offset] =
1473          if (offset >= md->offset_max) md->offset_overflow = TRUE; else            md->offset_vector[md->offset_end - number];
1474            {          md->offset_vector[offset+1] = eptr - md->start_subject;
1475            md->offset_vector[offset] =          if (offset_top <= offset) offset_top = offset + 2;
1476              md->offset_vector[md->offset_end - number];          }
1477            md->offset_vector[offset+1] = eptr - md->start_subject;  
1478            if (offset_top <= offset) offset_top = offset + 2;        /* Handle a recursively called group. Restore the offsets
1479            }        appropriately and continue from after the call. */
1480    
1481          /* Handle a recursively called group. Restore the offsets        if (md->recursive != NULL && md->recursive->group_num == number)
1482          appropriately and continue from after the call. */          {
1483            recursion_info *rec = md->recursive;
1484          if (md->recursive != NULL && md->recursive->group_num == number)          DPRINTF(("Recursion (%d) succeeded - continuing\n", number));
1485            {          md->recursive = rec->prevrec;
1486            recursion_info *rec = md->recursive;          mstart = rec->save_start;
1487            DPRINTF(("Recursion (%d) succeeded - continuing\n", number));          memcpy(md->offset_vector, rec->offset_save,
1488            md->recursive = rec->prevrec;            rec->saved_max * sizeof(int));
1489            md->start_match = rec->save_start;          offset_top = rec->save_offset_top;
1490            memcpy(md->offset_vector, rec->offset_save,          ecode = rec->after_call;
1491              rec->saved_max * sizeof(int));          ims = original_ims;
1492            ecode = rec->after_call;          break;
           ims = original_ims;  
           break;  
           }  
1493          }          }
1494        }        }
1495    
1496      /* Reset the value of the ims flags, in case they got changed during      /* For both capturing and non-capturing groups, reset the value of the ims
1497      the group. */      flags, in case they got changed during the group. */
1498    
1499      ims = original_ims;      ims = original_ims;
1500      DPRINTF(("ims reset to %02lx\n", ims));      DPRINTF(("ims reset to %02lx\n", ims));
# Line 1175  for (;;) Line 1513  for (;;)
1513    
1514      /* The repeating kets try the rest of the pattern or restart from the      /* The repeating kets try the rest of the pattern or restart from the
1515      preceding bracket, in the appropriate order. In the second case, we can use      preceding bracket, in the appropriate order. In the second case, we can use
1516      tail recursion to avoid using another stack frame. */      tail recursion to avoid using another stack frame, unless we have an
1517        unlimited repeat of a group that can match an empty string. */
1518    
1519        flags = (*prev >= OP_SBRA)? match_cbegroup : 0;
1520    
1521      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1522        {        {
1523        RMATCH(rrc, eptr, ecode + 1+LINK_SIZE, offset_top, md, ims, eptrb, 0);        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM12);
1524        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1525          if (flags != 0)    /* Could match an empty string */
1526            {
1527            RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM50);
1528            RRETURN(rrc);
1529            }
1530        ecode = prev;        ecode = prev;
       flags = match_isgroup;  
1531        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1532        }        }
1533      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
1534        {        {
1535        RMATCH(rrc, eptr, prev, offset_top, md, ims, eptrb, match_isgroup);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM13);
1536        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1537        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1538        flags = 0;        flags = 0;
# Line 1202  for (;;) Line 1547  for (;;)
1547      if ((ims & PCRE_MULTILINE) != 0)      if ((ims & PCRE_MULTILINE) != 0)
1548        {        {
1549        if (eptr != md->start_subject &&        if (eptr != md->start_subject &&
1550            (eptr == md->end_subject ||            (eptr == md->end_subject || !WAS_NEWLINE(eptr)))
            eptr < md->start_subject + md->nllen ||  
            !IS_NEWLINE(eptr - md->nllen)))  
1551          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
1552        ecode++;        ecode++;
1553        break;        break;
# Line 1225  for (;;) Line 1568  for (;;)
1568      ecode++;      ecode++;
1569      break;      break;
1570    
1571        /* Reset the start of match point */
1572    
1573        case OP_SET_SOM:
1574        mstart = eptr;
1575        ecode++;
1576        break;
1577    
1578      /* Assert before internal newline if multiline, or before a terminating      /* Assert before internal newline if multiline, or before a terminating
1579      newline unless endonly is set, else end of subject unless noteol is set. */      newline unless endonly is set, else end of subject unless noteol is set. */
1580    
# Line 1244  for (;;) Line 1594  for (;;)
1594        if (!md->endonly)        if (!md->endonly)
1595          {          {
1596          if (eptr != md->end_subject &&          if (eptr != md->end_subject &&
1597              (eptr != md->end_subject - md->nllen || !IS_NEWLINE(eptr)))              (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))
1598            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1599          ecode++;          ecode++;
1600          break;          break;
# Line 1263  for (;;) Line 1613  for (;;)
1613    
1614      case OP_EODN:      case OP_EODN:
1615      if (eptr != md->end_subject &&      if (eptr != md->end_subject &&
1616          (eptr != md->end_subject - md->nllen || !IS_NEWLINE(eptr)))          (!IS_NEWLINE(eptr) || eptr != md->end_subject - md->nllen))
1617        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
1618      ecode++;      ecode++;
1619      break;      break;
# Line 1276  for (;;) Line 1626  for (;;)
1626    
1627        /* Find out if the previous and current characters are "word" characters.        /* Find out if the previous and current characters are "word" characters.
1628        It takes a bit more work in UTF-8 mode. Characters > 255 are assumed to        It takes a bit more work in UTF-8 mode. Characters > 255 are assumed to
1629        be "non-word" characters. */        be "non-word" characters. Remember the earliest consulted character for
1630          partial matching. */
1631    
1632  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
1633        if (utf8)        if (utf8)
1634          {          {
1635          if (eptr == md->start_subject) prev_is_word = FALSE; else          if (eptr == md->start_subject) prev_is_word = FALSE; else
1636            {            {
1637            const uschar *lastptr = eptr - 1;            USPTR lastptr = eptr - 1;
1638            while((*lastptr & 0xc0) == 0x80) lastptr--;            while((*lastptr & 0xc0) == 0x80) lastptr--;
1639              if (lastptr < md->start_used_ptr) md->start_used_ptr = lastptr;
1640            GETCHAR(c, lastptr);            GETCHAR(c, lastptr);
1641            prev_is_word = c < 256 && (md->ctypes[c] & ctype_word) != 0;            prev_is_word = c < 256 && (md->ctypes[c] & ctype_word) != 0;
1642            }            }
1643          if (eptr >= md->end_subject) cur_is_word = FALSE; else          if (eptr >= md->end_subject)
1644              {
1645              SCHECK_PARTIAL();
1646              cur_is_word = FALSE;
1647              }
1648            else
1649            {            {
1650            GETCHAR(c, eptr);            GETCHAR(c, eptr);
1651            cur_is_word = c < 256 && (md->ctypes[c] & ctype_word) != 0;            cur_is_word = c < 256 && (md->ctypes[c] & ctype_word) != 0;
# Line 1297  for (;;) Line 1654  for (;;)
1654        else        else
1655  #endif  #endif
1656    
1657        /* More streamlined when not in UTF-8 mode */        /* Not in UTF-8 mode */
1658    
1659          {          {
1660          prev_is_word = (eptr != md->start_subject) &&          if (eptr == md->start_subject) prev_is_word = FALSE; else
1661            ((md->ctypes[eptr[-1]] & ctype_word) != 0);            {
1662          cur_is_word = (eptr < md->end_subject) &&            if (eptr <= md->start_used_ptr) md->start_used_ptr = eptr - 1;
1663            ((md->ctypes[*eptr] & ctype_word) != 0);            prev_is_word = ((md->ctypes[eptr[-1]] & ctype_word) != 0);
1664              }
1665            if (eptr >= md->end_subject)
1666              {
1667              SCHECK_PARTIAL();
1668              cur_is_word = FALSE;
1669              }
1670            else cur_is_word = ((md->ctypes[*eptr] & ctype_word) != 0);
1671          }          }
1672    
1673        /* Now see if the situation is what we want */        /* Now see if the situation is what we want */
# Line 1317  for (;;) Line 1681  for (;;)
1681      /* Match a single character type; inline for speed */      /* Match a single character type; inline for speed */
1682    
1683      case OP_ANY:      case OP_ANY:
1684      if ((ims & PCRE_DOTALL) == 0)      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
1685        /* Fall through */
1686    
1687        case OP_ALLANY:
1688        if (eptr++ >= md->end_subject)
1689        {        {
1690        if (eptr <= md->end_subject - md->nllen && IS_NEWLINE(eptr))        SCHECK_PARTIAL();
1691          RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
1692        }        }
1693      if (eptr++ >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (utf8) while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
     if (utf8)  
       while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  
1694      ecode++;      ecode++;
1695      break;      break;
1696    
# Line 1332  for (;;) Line 1698  for (;;)
1698      any byte, even newline, independent of the setting of PCRE_DOTALL. */      any byte, even newline, independent of the setting of PCRE_DOTALL. */
1699    
1700      case OP_ANYBYTE:      case OP_ANYBYTE:
1701      if (eptr++ >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr++ >= md->end_subject)
1702          {
1703          SCHECK_PARTIAL();
1704          RRETURN(MATCH_NOMATCH);
1705          }
1706      ecode++;      ecode++;
1707      break;      break;
1708    
1709      case OP_NOT_DIGIT:      case OP_NOT_DIGIT:
1710      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1711          {
1712          SCHECK_PARTIAL();
1713          RRETURN(MATCH_NOMATCH);
1714          }
1715      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1716      if (      if (
1717  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1350  for (;;) Line 1724  for (;;)
1724      break;      break;
1725    
1726      case OP_DIGIT:      case OP_DIGIT:
1727      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1728          {
1729          SCHECK_PARTIAL();
1730          RRETURN(MATCH_NOMATCH);
1731          }
1732      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1733      if (      if (
1734  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1363  for (;;) Line 1741  for (;;)
1741      break;      break;
1742    
1743      case OP_NOT_WHITESPACE:      case OP_NOT_WHITESPACE:
1744      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1745          {
1746          SCHECK_PARTIAL();
1747          RRETURN(MATCH_NOMATCH);
1748          }
1749      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1750      if (      if (
1751  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1376  for (;;) Line 1758  for (;;)
1758      break;      break;
1759    
1760      case OP_WHITESPACE:      case OP_WHITESPACE:
1761      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1762          {
1763          SCHECK_PARTIAL();
1764          RRETURN(MATCH_NOMATCH);
1765          }
1766      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1767      if (      if (
1768  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1389  for (;;) Line 1775  for (;;)
1775      break;      break;
1776    
1777      case OP_NOT_WORDCHAR:      case OP_NOT_WORDCHAR:
1778      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1779          {
1780          SCHECK_PARTIAL();
1781          RRETURN(MATCH_NOMATCH);
1782          }
1783      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1784      if (      if (
1785  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1402  for (;;) Line 1792  for (;;)
1792      break;      break;
1793    
1794      case OP_WORDCHAR:      case OP_WORDCHAR:
1795      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
1796          {
1797          SCHECK_PARTIAL();
1798          RRETURN(MATCH_NOMATCH);
1799          }
1800      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1801      if (      if (
1802  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1414  for (;;) Line 1808  for (;;)
1808      ecode++;      ecode++;
1809      break;      break;
1810    
1811  #ifdef SUPPORT_UCP      case OP_ANYNL:
1812      /* Check the next character by Unicode property. We will get here only      if (eptr >= md->end_subject)
1813      if the support is in the binary; otherwise a compile-time error occurs. */        {
1814          SCHECK_PARTIAL();
1815      case OP_PROP:        RRETURN(MATCH_NOMATCH);
1816      case OP_NOTPROP:        }
     if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);  
1817      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1818        switch(c)
1819        {        {
1820        int chartype, script;        default: RRETURN(MATCH_NOMATCH);
1821        int category = _pcre_ucp_findprop(c, &chartype, &script);        case 0x000d:
1822          if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
1823          break;
1824    
1825        switch(ecode[1])        case 0x000a:
1826          {        break;
         case PT_ANY:  
         if (op == OP_NOTPROP) RRETURN(MATCH_NOMATCH);  
         break;  
1827    
1828          case PT_LAMP:        case 0x000b:
1829          if ((chartype == ucp_Lu ||        case 0x000c:
1830               chartype == ucp_Ll ||        case 0x0085:
1831               chartype == ucp_Lt) == (op == OP_NOTPROP))        case 0x2028:
1832            RRETURN(MATCH_NOMATCH);        case 0x2029:
1833          if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
1834          break;
1835          }
1836        ecode++;
1837        break;
1838    
1839        case OP_NOT_HSPACE:
1840        if (eptr >= md->end_subject)
1841          {
1842          SCHECK_PARTIAL();
1843          RRETURN(MATCH_NOMATCH);
1844          }
1845        GETCHARINCTEST(c, eptr);
1846        switch(c)
1847          {
1848          default: break;
1849          case 0x09:      /* HT */
1850          case 0x20:      /* SPACE */
1851          case 0xa0:      /* NBSP */
1852          case 0x1680:    /* OGHAM SPACE MARK */
1853          case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
1854          case 0x2000:    /* EN QUAD */
1855          case 0x2001:    /* EM QUAD */
1856          case 0x2002:    /* EN SPACE */
1857          case 0x2003:    /* EM SPACE */
1858          case 0x2004:    /* THREE-PER-EM SPACE */
1859          case 0x2005:    /* FOUR-PER-EM SPACE */
1860          case 0x2006:    /* SIX-PER-EM SPACE */
1861          case 0x2007:    /* FIGURE SPACE */
1862          case 0x2008:    /* PUNCTUATION SPACE */
1863          case 0x2009:    /* THIN SPACE */
1864          case 0x200A:    /* HAIR SPACE */
1865          case 0x202f:    /* NARROW NO-BREAK SPACE */
1866          case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
1867          case 0x3000:    /* IDEOGRAPHIC SPACE */
1868          RRETURN(MATCH_NOMATCH);
1869          }
1870        ecode++;
1871        break;
1872    
1873        case OP_HSPACE:
1874        if (eptr >= md->end_subject)
1875          {
1876          SCHECK_PARTIAL();
1877          RRETURN(MATCH_NOMATCH);
1878          }
1879        GETCHARINCTEST(c, eptr);
1880        switch(c)
1881          {
1882          default: RRETURN(MATCH_NOMATCH);
1883          case 0x09:      /* HT */
1884          case 0x20:      /* SPACE */
1885          case 0xa0:      /* NBSP */
1886          case 0x1680:    /* OGHAM SPACE MARK */
1887          case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
1888          case 0x2000:    /* EN QUAD */
1889          case 0x2001:    /* EM QUAD */
1890          case 0x2002:    /* EN SPACE */
1891          case 0x2003:    /* EM SPACE */
1892          case 0x2004:    /* THREE-PER-EM SPACE */
1893          case 0x2005:    /* FOUR-PER-EM SPACE */
1894          case 0x2006:    /* SIX-PER-EM SPACE */
1895          case 0x2007:    /* FIGURE SPACE */
1896          case 0x2008:    /* PUNCTUATION SPACE */
1897          case 0x2009:    /* THIN SPACE */
1898          case 0x200A:    /* HAIR SPACE */
1899          case 0x202f:    /* NARROW NO-BREAK SPACE */
1900          case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
1901          case 0x3000:    /* IDEOGRAPHIC SPACE */
1902          break;
1903          }
1904        ecode++;
1905        break;
1906    
1907        case OP_NOT_VSPACE:
1908        if (eptr >= md->end_subject)
1909          {
1910          SCHECK_PARTIAL();
1911          RRETURN(MATCH_NOMATCH);
1912          }
1913        GETCHARINCTEST(c, eptr);
1914        switch(c)
1915          {
1916          default: break;
1917          case 0x0a:      /* LF */
1918          case 0x0b:      /* VT */
1919          case 0x0c:      /* FF */
1920          case 0x0d:      /* CR */
1921          case 0x85:      /* NEL */
1922          case 0x2028:    /* LINE SEPARATOR */
1923          case 0x2029:    /* PARAGRAPH SEPARATOR */
1924          RRETURN(MATCH_NOMATCH);
1925          }
1926        ecode++;
1927        break;
1928    
1929        case OP_VSPACE:
1930        if (eptr >= md->end_subject)
1931          {
1932          SCHECK_PARTIAL();
1933          RRETURN(MATCH_NOMATCH);
1934          }
1935        GETCHARINCTEST(c, eptr);
1936        switch(c)
1937          {
1938          default: RRETURN(MATCH_NOMATCH);
1939          case 0x0a:      /* LF */
1940          case 0x0b:      /* VT */
1941          case 0x0c:      /* FF */
1942          case 0x0d:      /* CR */
1943          case 0x85:      /* NEL */
1944          case 0x2028:    /* LINE SEPARATOR */
1945          case 0x2029:    /* PARAGRAPH SEPARATOR */
1946          break;
1947          }
1948        ecode++;
1949        break;
1950    
1951    #ifdef SUPPORT_UCP
1952        /* Check the next character by Unicode property. We will get here only
1953        if the support is in the binary; otherwise a compile-time error occurs. */
1954    
1955        case OP_PROP:
1956        case OP_NOTPROP:
1957        if (eptr >= md->end_subject)
1958          {
1959          SCHECK_PARTIAL();
1960          RRETURN(MATCH_NOMATCH);
1961          }
1962        GETCHARINCTEST(c, eptr);
1963          {
1964          const ucd_record *prop = GET_UCD(c);
1965    
1966          switch(ecode[1])
1967            {
1968            case PT_ANY:
1969            if (op == OP_NOTPROP) RRETURN(MATCH_NOMATCH);
1970            break;
1971    
1972            case PT_LAMP:
1973            if ((prop->chartype == ucp_Lu ||
1974                 prop->chartype == ucp_Ll ||
1975                 prop->chartype == ucp_Lt) == (op == OP_NOTPROP))
1976              RRETURN(MATCH_NOMATCH);
1977           break;           break;
1978    
1979          case PT_GC:          case PT_GC:
1980          if ((ecode[2] != category) == (op == OP_PROP))          if ((ecode[2] != _pcre_ucp_gentype[prop->chartype]) == (op == OP_PROP))
1981            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1982          break;          break;
1983    
1984          case PT_PC:          case PT_PC:
1985          if ((ecode[2] != chartype) == (op == OP_PROP))          if ((ecode[2] != prop->chartype) == (op == OP_PROP))
1986            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1987          break;          break;
1988    
1989          case PT_SC:          case PT_SC:
1990          if ((ecode[2] != script) == (op == OP_PROP))          if ((ecode[2] != prop->script) == (op == OP_PROP))
1991            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1992          break;          break;
1993    
1994          default:          default:
1995          RRETURN(PCRE_ERROR_INTERNAL);          RRETURN(PCRE_ERROR_INTERNAL);
         break;  
1996          }          }
1997    
1998        ecode += 3;        ecode += 3;
# Line 1467  for (;;) Line 2003  for (;;)
2003      is in the binary; otherwise a compile-time error occurs. */      is in the binary; otherwise a compile-time error occurs. */
2004    
2005      case OP_EXTUNI:      case OP_EXTUNI:
2006      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
2007          {
2008          SCHECK_PARTIAL();
2009          RRETURN(MATCH_NOMATCH);
2010          }
2011      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2012        {        {
2013        int chartype, script;        int category = UCD_CATEGORY(c);
       int category = _pcre_ucp_findprop(c, &chartype, &script);  
2014        if (category == ucp_M) RRETURN(MATCH_NOMATCH);        if (category == ucp_M) RRETURN(MATCH_NOMATCH);
2015        while (eptr < md->end_subject)        while (eptr < md->end_subject)
2016          {          {
# Line 1480  for (;;) Line 2019  for (;;)
2019            {            {
2020            GETCHARLEN(c, eptr, len);            GETCHARLEN(c, eptr, len);
2021            }            }
2022          category = _pcre_ucp_findprop(c, &chartype, &script);          category = UCD_CATEGORY(c);
2023          if (category != ucp_M) break;          if (category != ucp_M) break;
2024          eptr += len;          eptr += len;
2025          }          }
# Line 1501  for (;;) Line 2040  for (;;)
2040      case OP_REF:      case OP_REF:
2041        {        {
2042        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
2043        ecode += 3;                                 /* Advance past item */        ecode += 3;
2044    
2045          /* If the reference is unset, there are two possibilities:
2046    
2047          (a) In the default, Perl-compatible state, set the length to be longer
2048          than the amount of subject left; this ensures that every attempt at a
2049          match fails. We can't just fail here, because of the possibility of
2050          quantifiers with zero minima.
2051    
2052        /* If the reference is unset, set the length to be longer than the amount        (b) If the JavaScript compatibility flag is set, set the length to zero
2053        of subject left; this ensures that every attempt at a match fails. We        so that the back reference matches an empty string.
2054        can't just fail here, because of the possibility of quantifiers with zero  
2055        minima. */        Otherwise, set the length to the length of what was matched by the
2056          referenced subpattern. */
2057        length = (offset >= offset_top || md->offset_vector[offset] < 0)?  
2058          md->end_subject - eptr + 1 :        if (offset >= offset_top || md->offset_vector[offset] < 0)
2059          md->offset_vector[offset+1] - md->offset_vector[offset];          length = (md->jscript_compat)? 0 : md->end_subject - eptr + 1;
2060          else
2061            length = md->offset_vector[offset+1] - md->offset_vector[offset];
2062    
2063        /* Set up for repetition, or handle the non-repeated case */        /* Set up for repetition, or handle the non-repeated case */
2064    
# Line 1539  for (;;) Line 2087  for (;;)
2087          break;          break;
2088    
2089          default:               /* No repeat follows */          default:               /* No repeat follows */
2090          if (!match_ref(offset, eptr, length, md, ims)) RRETURN(MATCH_NOMATCH);          if (!match_ref(offset, eptr, length, md, ims))
2091              {
2092              CHECK_PARTIAL();
2093              RRETURN(MATCH_NOMATCH);
2094              }
2095          eptr += length;          eptr += length;
2096          continue;              /* With the main loop */          continue;              /* With the main loop */
2097          }          }
# Line 1555  for (;;) Line 2107  for (;;)
2107    
2108        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
2109          {          {
2110          if (!match_ref(offset, eptr, length, md, ims)) RRETURN(MATCH_NOMATCH);          if (!match_ref(offset, eptr, length, md, ims))
2111              {
2112              CHECK_PARTIAL();
2113              RRETURN(MATCH_NOMATCH);
2114              }
2115          eptr += length;          eptr += length;
2116          }          }
2117    
# Line 1570  for (;;) Line 2126  for (;;)
2126          {          {
2127          for (fi = min;; fi++)          for (fi = min;; fi++)
2128            {            {
2129            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM14);
2130            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2131            if (fi >= max || !match_ref(offset, eptr, length, md, ims))            if (fi >= max) RRETURN(MATCH_NOMATCH);
2132              if (!match_ref(offset, eptr, length, md, ims))
2133                {
2134                CHECK_PARTIAL();
2135              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2136                }
2137            eptr += length;            eptr += length;
2138            }            }
2139          /* Control never gets here */          /* Control never gets here */
# Line 1586  for (;;) Line 2146  for (;;)
2146          pp = eptr;          pp = eptr;
2147          for (i = min; i < max; i++)          for (i = min; i < max; i++)
2148            {            {
2149            if (!match_ref(offset, eptr, length, md, ims)) break;            if (!match_ref(offset, eptr, length, md, ims))
2150                {
2151                CHECK_PARTIAL();
2152                break;
2153                }
2154            eptr += length;            eptr += length;
2155            }            }
2156          while (eptr >= pp)          while (eptr >= pp)
2157            {            {
2158            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM15);
2159            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2160            eptr -= length;            eptr -= length;
2161            }            }
# Line 1600  for (;;) Line 2164  for (;;)
2164        }        }
2165      /* Control never gets here */      /* Control never gets here */
2166    
   
   
2167      /* Match a bit-mapped character class, possibly repeatedly. This op code is      /* Match a bit-mapped character class, possibly repeatedly. This op code is
2168      used when all the characters in the class have values in the range 0-255,      used when all the characters in the class have values in the range 0-255,
2169      and either the matching is caseful, or the characters are in the range      and either the matching is caseful, or the characters are in the range
# Line 1656  for (;;) Line 2218  for (;;)
2218          {          {
2219          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2220            {            {
2221            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
2222                {
2223                SCHECK_PARTIAL();
2224                RRETURN(MATCH_NOMATCH);
2225                }
2226            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
2227            if (c > 255)            if (c > 255)
2228              {              {
# Line 1674  for (;;) Line 2240  for (;;)
2240          {          {
2241          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2242            {            {
2243            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
2244                {
2245                SCHECK_PARTIAL();
2246                RRETURN(MATCH_NOMATCH);
2247                }
2248            c = *eptr++;            c = *eptr++;
2249            if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);            if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2250            }            }
# Line 1696  for (;;) Line 2266  for (;;)
2266            {            {
2267            for (fi = min;; fi++)            for (fi = min;; fi++)
2268              {              {
2269              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM16);
2270              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2271              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2272                if (eptr >= md->end_subject)
2273                  {
2274                  SCHECK_PARTIAL();
2275                  RRETURN(MATCH_NOMATCH);
2276                  }
2277              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
2278              if (c > 255)              if (c > 255)
2279                {                {
# Line 1716  for (;;) Line 2291  for (;;)
2291            {            {
2292            for (fi = min;; fi++)            for (fi = min;; fi++)
2293              {              {
2294              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM17);
2295              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2296              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2297                if (eptr >= md->end_subject)
2298                  {
2299                  SCHECK_PARTIAL();
2300                  RRETURN(MATCH_NOMATCH);
2301                  }
2302              c = *eptr++;              c = *eptr++;
2303              if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);              if ((data[c/8] & (1 << (c&7))) == 0) RRETURN(MATCH_NOMATCH);
2304              }              }
# Line 1739  for (;;) Line 2319  for (;;)
2319            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2320              {              {
2321              int len = 1;              int len = 1;
2322              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject)
2323                  {
2324                  SCHECK_PARTIAL();
2325                  break;
2326                  }
2327              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
2328              if (c > 255)              if (c > 255)
2329                {                {
# Line 1753  for (;;) Line 2337  for (;;)
2337              }              }
2338            for (;;)            for (;;)
2339              {              {
2340              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM18);
2341              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2342              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
2343              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 1765  for (;;) Line 2349  for (;;)
2349            {            {
2350            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2351              {              {
2352              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject)
2353                  {
2354                  SCHECK_PARTIAL();
2355                  break;
2356                  }
2357              c = *eptr;              c = *eptr;
2358              if ((data[c/8] & (1 << (c&7))) == 0) break;              if ((data[c/8] & (1 << (c&7))) == 0) break;
2359              eptr++;              eptr++;
2360              }              }
2361            while (eptr >= pp)            while (eptr >= pp)
2362              {              {
2363              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM19);
2364              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2365              eptr--;              eptr--;
2366              }              }
# Line 1785  for (;;) Line 2373  for (;;)
2373    
2374    
2375      /* Match an extended character class. This opcode is encountered only      /* Match an extended character class. This opcode is encountered only
2376      in UTF-8 mode, because that's the only time it is compiled. */      when UTF-8 mode mode is supported. Nevertheless, we may not be in UTF-8
2377        mode, because Unicode properties are supported in non-UTF-8 mode. */
2378    
2379  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
2380      case OP_XCLASS:      case OP_XCLASS:
# Line 1826  for (;;) Line 2415  for (;;)
2415    
2416        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
2417          {          {
2418          if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);          if (eptr >= md->end_subject)
2419          GETCHARINC(c, eptr);            {
2420              SCHECK_PARTIAL();
2421              RRETURN(MATCH_NOMATCH);
2422              }
2423            GETCHARINCTEST(c, eptr);
2424          if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);          if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);
2425          }          }
2426    
# Line 1843  for (;;) Line 2436  for (;;)
2436          {          {
2437          for (fi = min;; fi++)          for (fi = min;; fi++)
2438            {            {
2439            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM20);
2440            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2441            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
2442            GETCHARINC(c, eptr);            if (eptr >= md->end_subject)
2443                {
2444                SCHECK_PARTIAL();
2445                RRETURN(MATCH_NOMATCH);
2446                }
2447              GETCHARINCTEST(c, eptr);
2448            if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);            if (!_pcre_xclass(c, data)) RRETURN(MATCH_NOMATCH);
2449            }            }
2450          /* Control never gets here */          /* Control never gets here */
# Line 1860  for (;;) Line 2458  for (;;)
2458          for (i = min; i < max; i++)          for (i = min; i < max; i++)
2459            {            {
2460            int len = 1;            int len = 1;
2461            if (eptr >= md->end_subject) break;            if (eptr >= md->end_subject)
2462            GETCHARLEN(c, eptr, len);              {
2463                SCHECK_PARTIAL();
2464                break;
2465                }
2466              GETCHARLENTEST(c, eptr, len);
2467            if (!_pcre_xclass(c, data)) break;            if (!_pcre_xclass(c, data)) break;
2468            eptr += len;            eptr += len;
2469            }            }
2470          for(;;)          for(;;)
2471            {            {
2472            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM21);
2473            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2474            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
2475            BACKCHAR(eptr)            if (utf8) BACKCHAR(eptr);
2476            }            }
2477          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2478          }          }
# Line 1888  for (;;) Line 2490  for (;;)
2490        length = 1;        length = 1;
2491        ecode++;        ecode++;
2492        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
2493        if (length > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);        if (length > md->end_subject - eptr)
2494            {
2495            CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
2496            RRETURN(MATCH_NOMATCH);
2497            }
2498        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);        while (length-- > 0) if (*ecode++ != *eptr++) RRETURN(MATCH_NOMATCH);
2499        }        }
2500      else      else
# Line 1896  for (;;) Line 2502  for (;;)
2502    
2503      /* Non-UTF-8 mode */      /* Non-UTF-8 mode */
2504        {        {
2505        if (md->end_subject - eptr < 1) RRETURN(MATCH_NOMATCH);        if (md->end_subject - eptr < 1)
2506            {
2507            SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */
2508            RRETURN(MATCH_NOMATCH);
2509            }
2510        if (ecode[1] != *eptr++) RRETURN(MATCH_NOMATCH);        if (ecode[1] != *eptr++) RRETURN(MATCH_NOMATCH);
2511        ecode += 2;        ecode += 2;
2512        }        }
# Line 1912  for (;;) Line 2522  for (;;)
2522        ecode++;        ecode++;
2523        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
2524    
2525        if (length > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);        if (length > md->end_subject - eptr)
2526            {
2527            CHECK_PARTIAL();             /* Not SCHECK_PARTIAL() */
2528            RRETURN(MATCH_NOMATCH);
2529            }
2530    
2531        /* If the pattern character's value is < 128, we have only one byte, and        /* If the pattern character's value is < 128, we have only one byte, and
2532        can use the fast lookup table. */        can use the fast lookup table. */
# Line 1926  for (;;) Line 2540  for (;;)
2540    
2541        else        else
2542          {          {
2543          int dc;          unsigned int dc;
2544          GETCHARINC(dc, eptr);          GETCHARINC(dc, eptr);
2545          ecode += length;          ecode += length;
2546    
# Line 1936  for (;;) Line 2550  for (;;)
2550          if (fc != dc)          if (fc != dc)
2551            {            {
2552  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2553            if (dc != _pcre_ucp_othercase(fc))            if (dc != UCD_OTHERCASE(fc))
2554  #endif  #endif
2555              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2556            }            }
# Line 1947  for (;;) Line 2561  for (;;)
2561    
2562      /* Non-UTF-8 mode */      /* Non-UTF-8 mode */
2563        {        {
2564        if (md->end_subject - eptr < 1) RRETURN(MATCH_NOMATCH);        if (md->end_subject - eptr < 1)
2565            {
2566            SCHECK_PARTIAL();            /* This one can use SCHECK_PARTIAL() */
2567            RRETURN(MATCH_NOMATCH);
2568            }
2569        if (md->lcc[ecode[1]] != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);        if (md->lcc[ecode[1]] != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);
2570        ecode += 2;        ecode += 2;
2571        }        }
2572      break;      break;
2573    
2574      /* Match a single character repeatedly; different opcodes share code. */      /* Match a single character repeatedly. */
2575    
2576      case OP_EXACT:      case OP_EXACT:
2577      min = max = GET2(ecode, 1);      min = max = GET2(ecode, 1);
2578      ecode += 3;      ecode += 3;
2579      goto REPEATCHAR;      goto REPEATCHAR;
2580    
2581        case OP_POSUPTO:
2582        possessive = TRUE;
2583        /* Fall through */
2584    
2585      case OP_UPTO:      case OP_UPTO:
2586      case OP_MINUPTO:      case OP_MINUPTO:
2587      min = 0;      min = 0;
# Line 1968  for (;;) Line 2590  for (;;)
2590      ecode += 3;      ecode += 3;
2591      goto REPEATCHAR;      goto REPEATCHAR;
2592    
2593        case OP_POSSTAR:
2594        possessive = TRUE;
2595        min = 0;
2596        max = INT_MAX;
2597        ecode++;
2598        goto REPEATCHAR;
2599    
2600        case OP_POSPLUS:
2601        possessive = TRUE;
2602        min = 1;
2603        max = INT_MAX;
2604        ecode++;
2605        goto REPEATCHAR;
2606    
2607        case OP_POSQUERY:
2608        possessive = TRUE;
2609        min = 0;
2610        max = 1;
2611        ecode++;
2612        goto REPEATCHAR;
2613    
2614      case OP_STAR:      case OP_STAR:
2615      case OP_MINSTAR:      case OP_MINSTAR:
2616      case OP_PLUS:      case OP_PLUS:
# Line 1976  for (;;) Line 2619  for (;;)
2619      case OP_MINQUERY:      case OP_MINQUERY:
2620      c = *ecode++ - OP_STAR;      c = *ecode++ - OP_STAR;
2621      minimize = (c & 1) != 0;      minimize = (c & 1) != 0;
2622    
2623      min = rep_min[c];                 /* Pick up values from tables; */      min = rep_min[c];                 /* Pick up values from tables; */
2624      max = rep_max[c];                 /* zero for max => infinity */      max = rep_max[c];                 /* zero for max => infinity */
2625      if (max == 0) max = INT_MAX;      if (max == 0) max = INT_MAX;
2626    
2627      /* Common code for all repeated single-character matches. We can give      /* Common code for all repeated single-character matches. */
     up quickly if there are fewer than the minimum number of characters left in  
     the subject. */  
2628    
2629      REPEATCHAR:      REPEATCHAR:
2630  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
# Line 1991  for (;;) Line 2633  for (;;)
2633        length = 1;        length = 1;
2634        charptr = ecode;        charptr = ecode;
2635        GETCHARLEN(fc, ecode, length);        GETCHARLEN(fc, ecode, length);
       if (min * length > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);  
2636        ecode += length;        ecode += length;
2637    
2638        /* Handle multibyte character matching specially here. There is        /* Handle multibyte character matching specially here. There is
# Line 1999  for (;;) Line 2640  for (;;)
2640    
2641        if (length > 1)        if (length > 1)
2642          {          {
         int oclength = 0;  
         uschar occhars[8];  
   
2643  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2644          int othercase;          unsigned int othercase;
2645          if ((ims & PCRE_CASELESS) != 0 &&          if ((ims & PCRE_CASELESS) != 0 &&
2646              (othercase = _pcre_ucp_othercase(fc)) >= 0 &&              (othercase = UCD_OTHERCASE(fc)) != fc)
              othercase >= 0)  
2647            oclength = _pcre_ord2utf8(othercase, occhars);            oclength = _pcre_ord2utf8(othercase, occhars);
2648            else oclength = 0;
2649  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2650    
2651          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2652            {            {
2653            if (memcmp(eptr, charptr, length) == 0) eptr += length;            if (eptr <= md->end_subject - length &&
2654            /* Need braces because of following else */              memcmp(eptr, charptr, length) == 0) eptr += length;
2655            else if (oclength == 0) { RRETURN(MATCH_NOMATCH); }  #ifdef SUPPORT_UCP
2656              else if (oclength > 0 &&
2657                       eptr <= md->end_subject - oclength &&
2658                       memcmp(eptr, occhars, oclength) == 0) eptr += oclength;
2659    #endif  /* SUPPORT_UCP */
2660            else            else
2661              {              {
2662              if (memcmp(eptr, occhars, oclength) != 0) RRETURN(MATCH_NOMATCH);              CHECK_PARTIAL();
2663              eptr += oclength;              RRETURN(MATCH_NOMATCH);
2664              }              }
2665            }            }
2666    
# Line 2028  for (;;) Line 2670  for (;;)
2670            {            {
2671            for (fi = min;; fi++)            for (fi = min;; fi++)
2672              {              {
2673              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM22);
2674              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2675              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
2676              if (memcmp(eptr, charptr, length) == 0) eptr += length;              if (eptr <= md->end_subject - length &&
2677              /* Need braces because of following else */                memcmp(eptr, charptr, length) == 0) eptr += length;
2678              else if (oclength == 0) { RRETURN(MATCH_NOMATCH); }  #ifdef SUPPORT_UCP
2679                else if (oclength > 0 &&
2680                         eptr <= md->end_subject - oclength &&
2681                         memcmp(eptr, occhars, oclength) == 0) eptr += oclength;
2682    #endif  /* SUPPORT_UCP */
2683              else              else
2684                {                {
2685                if (memcmp(eptr, occhars, oclength) != 0) RRETURN(MATCH_NOMATCH);                CHECK_PARTIAL();
2686                eptr += oclength;                RRETURN(MATCH_NOMATCH);
2687                }                }
2688              }              }
2689            /* Control never gets here */            /* Control never gets here */
2690            }            }
2691          else  
2692            else  /* Maximize */
2693            {            {
2694            pp = eptr;            pp = eptr;
2695            for (i = min; i < max; i++)            for (i = min; i < max; i++)
2696              {              {
2697              if (eptr > md->end_subject - length) break;              if (eptr <= md->end_subject - length &&
2698              if (memcmp(eptr, charptr, length) == 0) eptr += length;                  memcmp(eptr, charptr, length) == 0) eptr += length;
2699              else if (oclength == 0) break;  #ifdef SUPPORT_UCP
2700                else if (oclength > 0 &&
2701                         eptr <= md->end_subject - oclength &&
2702                         memcmp(eptr, occhars, oclength) == 0) eptr += oclength;
2703    #endif  /* SUPPORT_UCP */
2704              else              else
2705                {                {
2706                if (memcmp(eptr, occhars, oclength) != 0) break;                CHECK_PARTIAL();
2707                eptr += oclength;                break;
2708                }                }
2709              }              }
2710            while (eptr >= pp)  
2711             {            if (possessive) continue;
2712             RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);  
2713             if (rrc != MATCH_NOMATCH) RRETURN(rrc);            for(;;)
2714             eptr -= length;              {
2715             }              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM23);
2716            RRETURN(MATCH_NOMATCH);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2717                if (eptr == pp) { RRETURN(MATCH_NOMATCH); }
2718    #ifdef SUPPORT_UCP
2719                eptr--;
2720                BACKCHAR(eptr);
2721    #else   /* without SUPPORT_UCP */
2722                eptr -= length;
2723    #endif  /* SUPPORT_UCP */
2724                }
2725            }            }
2726          /* Control never gets here */          /* Control never gets here */
2727          }          }
# Line 2075  for (;;) Line 2734  for (;;)
2734  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF8 */
2735    
2736      /* When not in UTF-8 mode, load a single-byte character. */      /* When not in UTF-8 mode, load a single-byte character. */
2737        {  
2738        if (min > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);      fc = *ecode++;
       fc = *ecode++;  
       }  
2739    
2740      /* The value of fc at this point is always less than 256, though we may or      /* The value of fc at this point is always less than 256, though we may or
2741      may not be in UTF-8 mode. The code is duplicated for the caseless and      may not be in UTF-8 mode. The code is duplicated for the caseless and
# Line 2096  for (;;) Line 2753  for (;;)
2753        {        {
2754        fc = md->lcc[fc];        fc = md->lcc[fc];
2755        for (i = 1; i <= min; i++)        for (i = 1; i <= min; i++)
2756            {
2757            if (eptr >= md->end_subject)
2758              {
2759              SCHECK_PARTIAL();
2760              RRETURN(MATCH_NOMATCH);
2761              }
2762          if (fc != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);          if (fc != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);
2763            }
2764        if (min == max) continue;        if (min == max) continue;
2765        if (minimize)        if (minimize)
2766          {          {
2767          for (fi = min;; fi++)          for (fi = min;; fi++)
2768            {            {
2769            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM24);
2770            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2771            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max) RRETURN(MATCH_NOMATCH);
2772                fc != md->lcc[*eptr++])            if (eptr >= md->end_subject)
2773                {
2774                SCHECK_PARTIAL();
2775              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2776                }
2777              if (fc != md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);
2778            }            }
2779          /* Control never gets here */          /* Control never gets here */
2780          }          }
2781        else        else  /* Maximize */
2782          {          {
2783          pp = eptr;          pp = eptr;
2784          for (i = min; i < max; i++)          for (i = min; i < max; i++)
2785            {            {
2786            if (eptr >= md->end_subject || fc != md->lcc[*eptr]) break;            if (eptr >= md->end_subject)
2787                {
2788                SCHECK_PARTIAL();
2789                break;
2790                }
2791              if (fc != md->lcc[*eptr]) break;
2792            eptr++;            eptr++;
2793            }            }
2794    
2795            if (possessive) continue;
2796    
2797          while (eptr >= pp)          while (eptr >= pp)
2798            {            {
2799            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM25);
2800            eptr--;            eptr--;
2801            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2802            }            }
# Line 2133  for (;;) Line 2809  for (;;)
2809    
2810      else      else
2811        {        {
2812        for (i = 1; i <= min; i++) if (fc != *eptr++) RRETURN(MATCH_NOMATCH);        for (i = 1; i <= min; i++)
2813            {
2814            if (eptr >= md->end_subject)
2815              {
2816              SCHECK_PARTIAL();
2817              RRETURN(MATCH_NOMATCH);
2818              }
2819            if (fc != *eptr++) RRETURN(MATCH_NOMATCH);
2820            }
2821    
2822        if (min == max) continue;        if (min == max) continue;
2823    
2824        if (minimize)        if (minimize)
2825          {          {
2826          for (fi = min;; fi++)          for (fi = min;; fi++)
2827            {            {
2828            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM26);
2829            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2830            if (fi >= max || eptr >= md->end_subject || fc != *eptr++)            if (fi >= max) RRETURN(MATCH_NOMATCH);
2831              if (eptr >= md->end_subject)
2832                {
2833                SCHECK_PARTIAL();
2834              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2835                }
2836              if (fc != *eptr++) RRETURN(MATCH_NOMATCH);
2837            }            }
2838          /* Control never gets here */          /* Control never gets here */
2839          }          }
2840        else        else  /* Maximize */
2841          {          {
2842          pp = eptr;          pp = eptr;
2843          for (i = min; i < max; i++)          for (i = min; i < max; i++)
2844            {            {
2845            if (eptr >= md->end_subject || fc != *eptr) break;            if (eptr >= md->end_subject)
2846                {
2847                SCHECK_PARTIAL();
2848                break;
2849                }
2850              if (fc != *eptr) break;
2851            eptr++;            eptr++;
2852            }            }
2853            if (possessive) continue;
2854    
2855          while (eptr >= pp)          while (eptr >= pp)
2856            {            {
2857            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM27);
2858            eptr--;            eptr--;
2859            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2860            }            }
# Line 2169  for (;;) Line 2867  for (;;)
2867      checking can be multibyte. */      checking can be multibyte. */
2868    
2869      case OP_NOT:      case OP_NOT:
2870      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject)
2871          {
2872          SCHECK_PARTIAL();
2873          RRETURN(MATCH_NOMATCH);
2874          }
2875      ecode++;      ecode++;
2876      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
2877      if ((ims & PCRE_CASELESS) != 0)      if ((ims & PCRE_CASELESS) != 0)
# Line 2206  for (;;) Line 2908  for (;;)
2908      ecode += 3;      ecode += 3;
2909      goto REPEATNOTCHAR;      goto REPEATNOTCHAR;
2910    
2911        case OP_NOTPOSSTAR:
2912        possessive = TRUE;
2913        min = 0;
2914        max = INT_MAX;
2915        ecode++;
2916        goto REPEATNOTCHAR;
2917    
2918        case OP_NOTPOSPLUS:
2919        possessive = TRUE;
2920        min = 1;
2921        max = INT_MAX;
2922        ecode++;
2923        goto REPEATNOTCHAR;
2924    
2925        case OP_NOTPOSQUERY:
2926        possessive = TRUE;
2927        min = 0;
2928        max = 1;
2929        ecode++;
2930        goto REPEATNOTCHAR;
2931    
2932        case OP_NOTPOSUPTO:
2933        possessive = TRUE;
2934        min = 0;
2935        max = GET2(ecode, 1);
2936        ecode += 3;
2937        goto REPEATNOTCHAR;
2938    
2939      case OP_NOTSTAR:      case OP_NOTSTAR:
2940      case OP_NOTMINSTAR:      case OP_NOTMINSTAR:
2941      case OP_NOTPLUS:      case OP_NOTPLUS:
# Line 2218  for (;;) Line 2948  for (;;)
2948      max = rep_max[c];                 /* zero for max => infinity */      max = rep_max[c];                 /* zero for max => infinity */
2949      if (max == 0) max = INT_MAX;      if (max == 0) max = INT_MAX;
2950    
2951      /* Common code for all repeated single-byte matches. We can give up quickly      /* Common code for all repeated single-byte matches. */
     if there are fewer than the minimum number of bytes left in the  
     subject. */  
2952    
2953      REPEATNOTCHAR:      REPEATNOTCHAR:
     if (min > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);  
2954      fc = *ecode++;      fc = *ecode++;
2955    
2956      /* The code is duplicated for the caseless and caseful cases, for speed,      /* The code is duplicated for the caseless and caseful cases, for speed,
# Line 2245  for (;;) Line 2972  for (;;)
2972        /* UTF-8 mode */        /* UTF-8 mode */
2973        if (utf8)        if (utf8)
2974          {          {
2975          register int d;          register unsigned int d;
2976          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2977            {            {
2978              if (eptr >= md->end_subject)
2979                {
2980                SCHECK_PARTIAL();
2981                RRETURN(MATCH_NOMATCH);
2982                }
2983            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
2984            if (d < 256) d = md->lcc[d];            if (d < 256) d = md->lcc[d];
2985            if (fc == d) RRETURN(MATCH_NOMATCH);            if (fc == d) RRETURN(MATCH_NOMATCH);
# Line 2259  for (;;) Line 2991  for (;;)
2991        /* Not UTF-8 mode */        /* Not UTF-8 mode */
2992          {          {
2993          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2994              {
2995              if (eptr >= md->end_subject)
2996                {
2997                SCHECK_PARTIAL();
2998                RRETURN(MATCH_NOMATCH);
2999                }
3000            if (fc == md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);            if (fc == md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);
3001              }
3002          }          }
3003    
3004        if (min == max) continue;        if (min == max) continue;
# Line 2270  for (;;) Line 3009  for (;;)
3009          /* UTF-8 mode */          /* UTF-8 mode */
3010          if (utf8)          if (utf8)
3011            {            {
3012            register int d;            register unsigned int d;
3013            for (fi = min;; fi++)            for (fi = min;; fi++)
3014              {              {
3015              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM28);
3016              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3017                if (fi >= max) RRETURN(MATCH_NOMATCH);
3018                if (eptr >= md->end_subject)
3019                  {
3020                  SCHECK_PARTIAL();
3021                  RRETURN(MATCH_NOMATCH);
3022                  }
3023              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
3024              if (d < 256) d = md->lcc[d];              if (d < 256) d = md->lcc[d];
3025              if (fi >= max || eptr >= md->end_subject || fc == d)              if (fc == d) RRETURN(MATCH_NOMATCH);
               RRETURN(MATCH_NOMATCH);  
3026              }              }
3027            }            }
3028          else          else
# Line 2287  for (;;) Line 3031  for (;;)
3031            {            {
3032            for (fi = min;; fi++)            for (fi = min;; fi++)
3033              {              {
3034              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM29);
3035              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3036              if (fi >= max || eptr >= md->end_subject || fc == md->lcc[*eptr++])              if (fi >= max) RRETURN(MATCH_NOMATCH);
3037                if (eptr >= md->end_subject)
3038                  {
3039                  SCHECK_PARTIAL();
3040                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3041                  }
3042                if (fc == md->lcc[*eptr++]) RRETURN(MATCH_NOMATCH);
3043              }              }
3044            }            }
3045          /* Control never gets here */          /* Control never gets here */
# Line 2306  for (;;) Line 3055  for (;;)
3055          /* UTF-8 mode */          /* UTF-8 mode */
3056          if (utf8)          if (utf8)
3057            {            {
3058            register int d;            register unsigned int d;
3059            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3060              {              {
3061              int len = 1;              int len = 1;
3062              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject)
3063                  {
3064                  SCHECK_PARTIAL();
3065                  break;
3066                  }
3067              GETCHARLEN(d, eptr, len);              GETCHARLEN(d, eptr, len);
3068              if (d < 256) d = md->lcc[d];              if (d < 256) d = md->lcc[d];
3069              if (fc == d) break;              if (fc == d) break;
3070              eptr += len;              eptr += len;
3071              }              }
3072            for(;;)          if (possessive) continue;
3073            for(;;)
3074              {              {
3075              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM30);
3076              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3077              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
3078              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 2330  for (;;) Line 3084  for (;;)
3084            {            {
3085            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3086              {              {
3087              if (eptr >= md->end_subject || fc == md->lcc[*eptr]) break;              if (eptr >= md->end_subject)
3088                  {
3089                  SCHECK_PARTIAL();
3090                  break;
3091                  }
3092                if (fc == md->lcc[*eptr]) break;
3093              eptr++;              eptr++;
3094              }              }
3095              if (possessive) continue;
3096            while (eptr >= pp)            while (eptr >= pp)
3097              {              {
3098              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM31);
3099              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3100              eptr--;              eptr--;
3101              }              }
# Line 2354  for (;;) Line 3114  for (;;)
3114        /* UTF-8 mode */        /* UTF-8 mode */
3115        if (utf8)        if (utf8)
3116          {          {
3117          register int d;          register unsigned int d;
3118          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3119            {            {
3120              if (eptr >= md->end_subject)
3121                {
3122                SCHECK_PARTIAL();
3123                RRETURN(MATCH_NOMATCH);
3124                }
3125            GETCHARINC(d, eptr);            GETCHARINC(d, eptr);
3126            if (fc == d) RRETURN(MATCH_NOMATCH);            if (fc == d) RRETURN(MATCH_NOMATCH);
3127            }            }
# Line 2366  for (;;) Line 3131  for (;;)
3131        /* Not UTF-8 mode */        /* Not UTF-8 mode */
3132          {          {
3133          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3134              {
3135              if (eptr >= md->end_subject)
3136                {
3137                SCHECK_PARTIAL();
3138                RRETURN(MATCH_NOMATCH);
3139                }
3140            if (fc == *eptr++) RRETURN(MATCH_NOMATCH);            if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
3141              }
3142          }          }
3143    
3144        if (min == max) continue;        if (min == max) continue;
# Line 2377  for (;;) Line 3149  for (;;)
3149          /* UTF-8 mode */          /* UTF-8 mode */
3150          if (utf8)          if (utf8)
3151            {            {
3152            register int d;            register unsigned int d;
3153            for (fi = min;; fi++)            for (fi = min;; fi++)
3154              {              {
3155              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM32);
3156              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3157              GETCHARINC(d, eptr);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3158              if (fi >= max || eptr >= md->end_subject || fc == d)              if (eptr >= md->end_subject)
3159                  {
3160                  SCHECK_PARTIAL();
3161                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3162                  }
3163                GETCHARINC(d, eptr);
3164                if (fc == d) RRETURN(MATCH_NOMATCH);
3165              }              }
3166            }            }
3167          else          else
# Line 2393  for (;;) Line 3170  for (;;)
3170            {            {
3171            for (fi = min;; fi++)            for (fi = min;; fi++)
3172              {              {
3173              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM33);
3174              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3175              if (fi >= max || eptr >= md->end_subject || fc == *eptr++)              if (fi >= max) RRETURN(MATCH_NOMATCH);
3176                if (eptr >= md->end_subject)
3177                  {
3178                  SCHECK_PARTIAL();
3179                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3180                  }
3181                if (fc == *eptr++) RRETURN(MATCH_NOMATCH);
3182              }              }
3183            }            }
3184          /* Control never gets here */          /* Control never gets here */
# Line 2412  for (;;) Line 3194  for (;;)
3194          /* UTF-8 mode */          /* UTF-8 mode */
3195          if (utf8)          if (utf8)
3196            {            {
3197            register int d;            register unsigned int d;
3198            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3199              {              {
3200              int len = 1;              int len = 1;
3201              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject)
3202              GETCHARLEN(d, eptr, len);                {
3203              if (fc == d) break;                SCHECK_PARTIAL();
3204                  break;
3205                  }
3206                GETCHARLEN(d, eptr, len);
3207                if (fc == d) break;
3208              eptr += len;              eptr += len;
3209              }              }
3210              if (possessive) continue;
3211            for(;;)            for(;;)
3212              {              {
3213              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM34);
3214              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3215              if (eptr-- == pp) break;        /* Stop if tried at original pos */              if (eptr-- == pp) break;        /* Stop if tried at original pos */
3216              BACKCHAR(eptr);              BACKCHAR(eptr);
# Line 2435  for (;;) Line 3222  for (;;)
3222            {            {
3223            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3224              {              {
3225              if (eptr >= md->end_subject || fc == *eptr) break;              if (eptr >= md->end_subject)
3226                  {
3227                  SCHECK_PARTIAL();
3228                  break;
3229                  }
3230                if (fc == *eptr) break;
3231              eptr++;              eptr++;
3232              }              }
3233              if (possessive) continue;
3234            while (eptr >= pp)            while (eptr >= pp)
3235              {              {
3236              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM35);
3237              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3238              eptr--;              eptr--;
3239              }              }
# Line 2469  for (;;) Line 3262  for (;;)
3262      ecode += 3;      ecode += 3;
3263      goto REPEATTYPE;      goto REPEATTYPE;
3264    
3265        case OP_TYPEPOSSTAR:
3266        possessive = TRUE;
3267        min = 0;
3268        max = INT_MAX;
3269        ecode++;
3270        goto REPEATTYPE;
3271    
3272        case OP_TYPEPOSPLUS:
3273        possessive = TRUE;
3274        min = 1;
3275        max = INT_MAX;
3276        ecode++;
3277        goto REPEATTYPE;
3278    
3279        case OP_TYPEPOSQUERY:
3280        possessive = TRUE;
3281        min = 0;
3282        max = 1;
3283        ecode++;
3284        goto REPEATTYPE;
3285    
3286        case OP_TYPEPOSUPTO:
3287        possessive = TRUE;
3288        min = 0;
3289        max = GET2(ecode, 1);
3290        ecode += 3;
3291        goto REPEATTYPE;
3292    
3293      case OP_TYPESTAR:      case OP_TYPESTAR:
3294      case OP_TYPEMINSTAR:      case OP_TYPEMINSTAR:
3295      case OP_TYPEPLUS:      case OP_TYPEPLUS:
# Line 2500  for (;;) Line 3321  for (;;)
3321    
3322      /* First, ensure the minimum number of matches are present. Use inline      /* First, ensure the minimum number of matches are present. Use inline
3323      code for maximizing the speed, and do the type test once at the start      code for maximizing the speed, and do the type test once at the start
3324      (i.e. keep it out of the loop). Also we can test that there are at least      (i.e. keep it out of the loop). Separate the UTF-8 code completely as that
     the minimum number of bytes before we start. This isn't as effective in  
     UTF-8 mode, but it does no harm. Separate the UTF-8 code completely as that  
3325      is tidier. Also separate the UCP code, which can be the same for both UTF-8      is tidier. Also separate the UCP code, which can be the same for both UTF-8
3326      and single-bytes. */      and single-bytes. */
3327    
     if (min > md->end_subject - eptr) RRETURN(MATCH_NOMATCH);  
3328      if (min > 0)      if (min > 0)
3329        {        {
3330  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
# Line 2518  for (;;) Line 3336  for (;;)
3336            if (prop_fail_result) RRETURN(MATCH_NOMATCH);            if (prop_fail_result) RRETURN(MATCH_NOMATCH);
3337            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
3338              {              {
3339              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject)
3340              GETCHARINC(c, eptr);                {
3341                  SCHECK_PARTIAL();
3342                  RRETURN(MATCH_NOMATCH);
3343                  }
3344                GETCHARINCTEST(c, eptr);
3345              }              }
3346            break;            break;
3347    
3348            case PT_LAMP:            case PT_LAMP:
3349            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
3350              {              {
3351              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject)
3352              GETCHARINC(c, eptr);                {
3353              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);                SCHECK_PARTIAL();
3354                  RRETURN(MATCH_NOMATCH);
3355                  }
3356                GETCHARINCTEST(c, eptr);
3357                prop_chartype = UCD_CHARTYPE(c);
3358              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
3359                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
3360                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
# Line 2539  for (;;) Line 3365  for (;;)
3365            case PT_GC:            case PT_GC:
3366            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
3367              {              {
3368              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject)
3369              GETCHARINC(c, eptr);                {
3370              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);                SCHECK_PARTIAL();
3371                  RRETURN(MATCH_NOMATCH);
3372                  }
3373                GETCHARINCTEST(c, eptr);
3374                prop_category = UCD_CATEGORY(c);
3375              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
3376                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3377              }              }
# Line 2550  for (;;) Line 3380  for (;;)
3380            case PT_PC:            case PT_PC:
3381            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
3382              {              {
3383              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject)
3384              GETCHARINC(c, eptr);                {
3385              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);                SCHECK_PARTIAL();
3386                  RRETURN(MATCH_NOMATCH);
3387                  }
3388                GETCHARINCTEST(c, eptr);
3389                prop_chartype = UCD_CHARTYPE(c);
3390              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
3391                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3392              }              }
# Line 2561  for (;;) Line 3395  for (;;)
3395            case PT_SC:            case PT_SC:
3396            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
3397              {              {
3398              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject)
3399              GETCHARINC(c, eptr);                {
3400              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);                SCHECK_PARTIAL();
3401                  RRETURN(MATCH_NOMATCH);
3402                  }
3403                GETCHARINCTEST(c, eptr);
3404                prop_script = UCD_SCRIPT(c);
3405              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
3406                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3407              }              }
# Line 2571  for (;;) Line 3409  for (;;)
3409    
3410            default:            default:
3411            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
           break;  
3412            }            }
3413          }          }
3414    
# Line 2582  for (;;) Line 3419  for (;;)
3419          {          {
3420          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3421            {            {
3422              if (eptr >= md->end_subject)
3423                {
3424                SCHECK_PARTIAL();
3425                RRETURN(MATCH_NOMATCH);
3426                }
3427            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3428            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
3429            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);
3430            while (eptr < md->end_subject)            while (eptr < md->end_subject)
3431              {              {
3432              int len = 1;              int len = 1;
3433              if (!utf8) c = *eptr; else              if (!utf8) c = *eptr;
3434                {                else { GETCHARLEN(c, eptr, len); }
3435                GETCHARLEN(c, eptr, len);              prop_category = UCD_CATEGORY(c);
               }  
             prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);  
3436              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
3437              eptr += len;              eptr += len;
3438              }              }
# Line 2610  for (;;) Line 3450  for (;;)
3450          case OP_ANY:          case OP_ANY:
3451          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3452            {            {
3453            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject)
3454                 ((ims & PCRE_DOTALL) == 0 &&              {
3455                   eptr <= md->end_subject - md->nllen &&              SCHECK_PARTIAL();
                  IS_NEWLINE(eptr)))  
3456              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3457                }
3458              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
3459              eptr++;
3460              while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
3461              }
3462            break;
3463    
3464            case OP_ALLANY:
3465            for (i = 1; i <= min; i++)
3466              {
3467              if (eptr >= md->end_subject)
3468                {
3469                SCHECK_PARTIAL();
3470                RRETURN(MATCH_NOMATCH);
3471                }
3472            eptr++;            eptr++;
3473            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
3474            }            }
3475          break;          break;
3476    
3477          case OP_ANYBYTE:          case OP_ANYBYTE:
3478            if (eptr > md->end_subject - min) RRETURN(MATCH_NOMATCH);
3479          eptr += min;          eptr += min;
3480          break;          break;
3481    
3482            case OP_ANYNL:
3483            for (i = 1; i <= min; i++)
3484              {
3485              if (eptr >= md->end_subject)
3486                {
3487                SCHECK_PARTIAL();
3488                RRETURN(MATCH_NOMATCH);
3489                }
3490              GETCHARINC(c, eptr);
3491              switch(c)
3492                {
3493                default: RRETURN(MATCH_NOMATCH);
3494                case 0x000d:
3495                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3496                break;
3497    
3498                case 0x000a:
3499                break;
3500    
3501                case 0x000b:
3502                case 0x000c:
3503                case 0x0085:
3504                case 0x2028:
3505                case 0x2029:
3506                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3507                break;
3508                }
3509              }
3510            break;
3511    
3512            case OP_NOT_HSPACE:
3513            for (i = 1; i <= min; i++)
3514              {
3515              if (eptr >= md->end_subject)
3516                {
3517                SCHECK_PARTIAL();
3518                RRETURN(MATCH_NOMATCH);
3519                }
3520              GETCHARINC(c, eptr);
3521              switch(c)
3522                {
3523                default: break;
3524                case 0x09:      /* HT */
3525                case 0x20:      /* SPACE */
3526                case 0xa0:      /* NBSP */
3527                case 0x1680:    /* OGHAM SPACE MARK */
3528                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3529                case 0x2000:    /* EN QUAD */
3530                case 0x2001:    /* EM QUAD */
3531                case 0x2002:    /* EN SPACE */
3532                case 0x2003:    /* EM SPACE */
3533                case 0x2004:    /* THREE-PER-EM SPACE */
3534                case 0x2005:    /* FOUR-PER-EM SPACE */
3535                case 0x2006:    /* SIX-PER-EM SPACE */
3536                case 0x2007:    /* FIGURE SPACE */
3537                case 0x2008:    /* PUNCTUATION SPACE */
3538                case 0x2009:    /* THIN SPACE */
3539                case 0x200A:    /* HAIR SPACE */
3540                case 0x202f:    /* NARROW NO-BREAK SPACE */
3541                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3542                case 0x3000:    /* IDEOGRAPHIC SPACE */
3543                RRETURN(MATCH_NOMATCH);
3544                }
3545              }
3546            break;
3547    
3548            case OP_HSPACE:
3549            for (i = 1; i <= min; i++)
3550              {
3551              if (eptr >= md->end_subject)
3552                {
3553                SCHECK_PARTIAL();
3554                RRETURN(MATCH_NOMATCH);
3555                }
3556              GETCHARINC(c, eptr);
3557              switch(c)
3558                {
3559                default: RRETURN(MATCH_NOMATCH);
3560                case 0x09:      /* HT */
3561                case 0x20:      /* SPACE */
3562                case 0xa0:      /* NBSP */
3563                case 0x1680:    /* OGHAM SPACE MARK */
3564                case 0x180e:    /* MONGOLIAN VOWEL SEPARATOR */
3565                case 0x2000:    /* EN QUAD */
3566                case 0x2001:    /* EM QUAD */
3567                case 0x2002:    /* EN SPACE */
3568                case 0x2003:    /* EM SPACE */
3569                case 0x2004:    /* THREE-PER-EM SPACE */
3570                case 0x2005:    /* FOUR-PER-EM SPACE */
3571                case 0x2006:    /* SIX-PER-EM SPACE */
3572                case 0x2007:    /* FIGURE SPACE */
3573                case 0x2008:    /* PUNCTUATION SPACE */
3574                case 0x2009:    /* THIN SPACE */
3575                case 0x200A:    /* HAIR SPACE */
3576                case 0x202f:    /* NARROW NO-BREAK SPACE */
3577                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3578                case 0x3000:    /* IDEOGRAPHIC SPACE */
3579                break;
3580                }
3581              }
3582            break;
3583    
3584            case OP_NOT_VSPACE:
3585            for (i = 1; i <= min; i++)
3586              {
3587              if (eptr >= md->end_subject)
3588                {
3589                SCHECK_PARTIAL();
3590                RRETURN(MATCH_NOMATCH);
3591                }
3592              GETCHARINC(c, eptr);
3593              switch(c)
3594                {
3595                default: break;
3596                case 0x0a:      /* LF */
3597                case 0x0b:      /* VT */
3598                case 0x0c:      /* FF */
3599                case 0x0d:      /* CR */
3600                case 0x85:      /* NEL */
3601                case 0x2028:    /* LINE SEPARATOR */
3602                case 0x2029:    /* PARAGRAPH SEPARATOR */
3603                RRETURN(MATCH_NOMATCH);
3604                }
3605              }
3606            break;
3607    
3608            case OP_VSPACE:
3609            for (i = 1; i <= min; i++)
3610              {
3611              if (eptr >= md->end_subject)
3612                {
3613                SCHECK_PARTIAL();
3614                RRETURN(MATCH_NOMATCH);
3615                }
3616              GETCHARINC(c, eptr);
3617              switch(c)
3618                {
3619                default: RRETURN(MATCH_NOMATCH);
3620                case 0x0a:      /* LF */
3621                case 0x0b:      /* VT */
3622                case 0x0c:      /* FF */
3623                case 0x0d:      /* CR */
3624                case 0x85:      /* NEL */
3625                case 0x2028:    /* LINE SEPARATOR */
3626                case 0x2029:    /* PARAGRAPH SEPARATOR */
3627                break;
3628                }
3629              }
3630            break;
3631    
3632          case OP_NOT_DIGIT:          case OP_NOT_DIGIT:
3633          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3634            {            {
3635            if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (eptr >= md->end_subject)
3636                {
3637                SCHECK_PARTIAL();
3638                RRETURN(MATCH_NOMATCH);
3639                }
3640            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
3641            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)            if (c < 128 && (md->ctypes[c] & ctype_digit) != 0)
3642              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
# Line 2637  for (;;) Line 3646  for (;;)
3646          case OP_DIGIT:          case OP_DIGIT:
3647          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3648            {            {
3649            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject)
3650               *eptr >= 128 || (md->ctypes[*eptr++] & ctype_digit) == 0)              {
3651                SCHECK_PARTIAL();
3652                RRETURN(MATCH_NOMATCH);
3653                }
3654              if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_digit) == 0)
3655              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3656            /* 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 */
3657            }            }
# Line 2647  for (;;) Line 3660  for (;;)
3660          case OP_NOT_WHITESPACE:          case OP_NOT_WHITESPACE:
3661          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3662            {            {
3663            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject)
3664               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_space) != 0))              {
3665                SCHECK_PARTIAL();
3666              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3667            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;              }
3668              if (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0)
3669                RRETURN(MATCH_NOMATCH);
3670              while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3671            }            }
3672          break;          break;
3673    
3674          case OP_WHITESPACE:          case OP_WHITESPACE:
3675          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3676            {            {
3677            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject)
3678               *eptr >= 128 || (md->ctypes[*eptr++] & ctype_space) == 0)              {
3679                SCHECK_PARTIAL();
3680                RRETURN(MATCH_NOMATCH);
3681                }
3682              if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_space) == 0)
3683              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3684            /* 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 */
3685            }            }
# Line 2668  for (;;) Line 3689  for (;;)
3689          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3690            {            {
3691            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject ||
3692               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_word) != 0))               (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0))
3693              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3694            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3695            }            }
3696          break;          break;
3697    
3698          case OP_WORDCHAR:          case OP_WORDCHAR:
3699          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3700            {            {
3701            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject)
3702               *eptr >= 128 || (md->ctypes[*eptr++] & ctype_word) == 0)              {
3703                SCHECK_PARTIAL();
3704                RRETURN(MATCH_NOMATCH);
3705                }
3706              if (*eptr >= 128 || (md->ctypes[*eptr++] & ctype_word) == 0)
3707              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3708            /* 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 */
3709            }            }
# Line 2697  for (;;) Line 3722  for (;;)
3722        switch(ctype)        switch(ctype)
3723          {          {
3724          case OP_ANY:          case OP_ANY:
3725          if ((ims & PCRE_DOTALL) == 0)          for (i = 1; i <= min; i++)
3726            {            {
3727            for (i = 1; i <= min; i++)            if (eptr >= md->end_subject)
3728              {              {
3729              if (eptr <= md->end_subject - md->nllen && IS_NEWLINE(eptr))              SCHECK_PARTIAL();
3730                RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
             eptr++;  
3731              }              }
3732              if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
3733              eptr++;
3734            }            }
3735          else eptr += min;          break;
3736    
3737            case OP_ALLANY:
3738            if (eptr > md->end_subject - min)
3739              {
3740              SCHECK_PARTIAL();
3741              RRETURN(MATCH_NOMATCH);
3742              }
3743            eptr += min;
3744          break;          break;
3745    
3746          case OP_ANYBYTE:          case OP_ANYBYTE:
3747            if (eptr > md->end_subject - min)
3748              {
3749              SCHECK_PARTIAL();
3750              RRETURN(MATCH_NOMATCH);
3751              }
3752          eptr += min;          eptr += min;
3753          break;          break;
3754    
3755            case OP_ANYNL:
3756            for (i = 1; i <= min; i++)
3757              {
3758              if (eptr >= md->end_subject)
3759                {
3760                SCHECK_PARTIAL();
3761                RRETURN(MATCH_NOMATCH);
3762                }
3763              switch(*eptr++)
3764                {
3765                default: RRETURN(MATCH_NOMATCH);
3766                case 0x000d:
3767                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3768                break;
3769                case 0x000a:
3770                break;
3771    
3772                case 0x000b:
3773                case 0x000c:
3774                case 0x0085:
3775                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3776                break;
3777                }
3778              }
3779            break;
3780    
3781            case OP_NOT_HSPACE:
3782            for (i = 1; i <= min; i++)
3783              {
3784              if (eptr >= md->end_subject)
3785                {
3786                SCHECK_PARTIAL();
3787                RRETURN(MATCH_NOMATCH);
3788                }
3789              switch(*eptr++)
3790                {
3791                default: break;
3792                case 0x09:      /* HT */
3793                case 0x20:      /* SPACE */
3794                case 0xa0:      /* NBSP */
3795                RRETURN(MATCH_NOMATCH);
3796                }
3797              }
3798            break;
3799    
3800            case OP_HSPACE:
3801            for (i = 1; i <= min; i++)
3802              {
3803              if (eptr >= md->end_subject)
3804                {
3805                SCHECK_PARTIAL();
3806                RRETURN(MATCH_NOMATCH);
3807                }
3808              switch(*eptr++)
3809                {
3810                default: RRETURN(MATCH_NOMATCH);
3811                case 0x09:      /* HT */
3812                case 0x20:      /* SPACE */
3813                case 0xa0:      /* NBSP */
3814                break;
3815                }
3816              }
3817            break;
3818    
3819            case OP_NOT_VSPACE:
3820            for (i = 1; i <= min; i++)
3821              {
3822              if (eptr >= md->end_subject)
3823                {
3824                SCHECK_PARTIAL();
3825                RRETURN(MATCH_NOMATCH);
3826                }
3827              switch(*eptr++)
3828                {
3829                default: break;
3830                case 0x0a:      /* LF */
3831                case 0x0b:      /* VT */
3832                case 0x0c:      /* FF */
3833                case 0x0d:      /* CR */
3834                case 0x85:      /* NEL */
3835                RRETURN(MATCH_NOMATCH);
3836                }
3837              }
3838            break;
3839    
3840            case OP_VSPACE:
3841            for (i = 1; i <= min; i++)
3842              {
3843              if (eptr >= md->end_subject)
3844                {
3845                SCHECK_PARTIAL();
3846                RRETURN(MATCH_NOMATCH);
3847                }
3848              switch(*eptr++)
3849                {
3850                default: RRETURN(MATCH_NOMATCH);
3851                case 0x0a:      /* LF */
3852                case 0x0b:      /* VT */
3853                case 0x0c:      /* FF */
3854                case 0x0d:      /* CR */
3855                case 0x85:      /* NEL */
3856                break;
3857                }
3858              }
3859            break;
3860    
3861          case OP_NOT_DIGIT:          case OP_NOT_DIGIT:
3862          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3863              {
3864              if (eptr >= md->end_subject)
3865                {
3866                SCHECK_PARTIAL();
3867                RRETURN(MATCH_NOMATCH);
3868                }
3869            if ((md->ctypes[*eptr++] & ctype_digit) != 0) RRETURN(MATCH_NOMATCH);            if ((md->ctypes[*eptr++] & ctype_digit) != 0) RRETURN(MATCH_NOMATCH);
3870              }
3871          break;          break;
3872    
3873          case OP_DIGIT:          case OP_DIGIT:
3874          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3875              {
3876              if (eptr >= md->end_subject)
3877                {
3878                SCHECK_PARTIAL();
3879                RRETURN(MATCH_NOMATCH);
3880                }
3881            if ((md->ctypes[*eptr++] & ctype_digit) == 0) RRETURN(MATCH_NOMATCH);            if ((md->ctypes[*eptr++] & ctype_digit) == 0) RRETURN(MATCH_NOMATCH);
3882              }
3883          break;          break;
3884    
3885          case OP_NOT_WHITESPACE:          case OP_NOT_WHITESPACE:
3886          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3887              {
3888              if (eptr >= md->end_subject)
3889                {
3890                SCHECK_PARTIAL();
3891                RRETURN(MATCH_NOMATCH);
3892                }
3893            if ((md->ctypes[*eptr++] & ctype_space) != 0) RRETURN(MATCH_NOMATCH);            if ((md->ctypes[*eptr++] & ctype_space) != 0) RRETURN(MATCH_NOMATCH);
3894              }
3895          break;          break;
3896    
3897          case OP_WHITESPACE:          case OP_WHITESPACE:
3898          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3899              {
3900              if (eptr >= md->end_subject)
3901                {
3902                SCHECK_PARTIAL();
3903                RRETURN(MATCH_NOMATCH);
3904                }
3905            if ((md->ctypes[*eptr++] & ctype_space) == 0) RRETURN(MATCH_NOMATCH);            if ((md->ctypes[*eptr++] & ctype_space) == 0) RRETURN(MATCH_NOMATCH);
3906              }
3907          break;          break;
3908    
3909          case OP_NOT_WORDCHAR:          case OP_NOT_WORDCHAR:
3910          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3911              {
3912              if (eptr >= md->end_subject)
3913                {
3914                SCHECK_PARTIAL();
3915                RRETURN(MATCH_NOMATCH);
3916                }
3917            if ((md->ctypes[*eptr++] & ctype_word) != 0)            if ((md->ctypes[*eptr++] & ctype_word) != 0)
3918              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3919              }
3920          break;          break;
3921    
3922          case OP_WORDCHAR:          case OP_WORDCHAR:
3923          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3924              {
3925              if (eptr >= md->end_subject)
3926                {
3927                SCHECK_PARTIAL();
3928                RRETURN(MATCH_NOMATCH);
3929                }
3930            if ((md->ctypes[*eptr++] & ctype_word) == 0)            if ((md->ctypes[*eptr++] & ctype_word) == 0)
3931              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3932              }
3933          break;          break;
3934    
3935          default:          default:
# Line 2768  for (;;) Line 3955  for (;;)
3955            case PT_ANY:            case PT_ANY:
3956            for (fi = min;; fi++)            for (fi = min;; fi++)
3957              {              {
3958              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM36);
3959              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3960              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3961                if (eptr >= md->end_subject)
3962                  {
3963                  SCHECK_PARTIAL();
3964                  RRETURN(MATCH_NOMATCH);
3965                  }
3966              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3967              if (prop_fail_result) RRETURN(MATCH_NOMATCH);              if (prop_fail_result) RRETURN(MATCH_NOMATCH);
3968              }              }
3969            break;            /* Control never gets here */
3970    
3971            case PT_LAMP:            case PT_LAMP:
3972            for (fi = min;; fi++)            for (fi = min;; fi++)
3973              {              {
3974              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM37);
3975              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3976              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3977                if (eptr >= md->end_subject)
3978                  {
3979                  SCHECK_PARTIAL();
3980                  RRETURN(MATCH_NOMATCH);
3981                  }
3982              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3983              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3984              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
3985                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
3986                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
3987                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3988              }              }
3989            break;            /* Control never gets here */
3990    
3991            case PT_GC:            case PT_GC:
3992            for (fi = min;; fi++)            for (fi = min;; fi++)
3993              {              {
3994              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM38);
3995              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3996              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
3997                if (eptr >= md->end_subject)
3998                  {
3999                  SCHECK_PARTIAL();
4000                  RRETURN(MATCH_NOMATCH);
4001                  }
4002              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
4003              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
4004              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
4005                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4006              }              }
4007            break;            /* Control never gets here */
4008    
4009            case PT_PC:            case PT_PC:
4010            for (fi = min;; fi++)            for (fi = min;; fi++)
4011              {              {
4012              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM39);
4013              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4014              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4015                if (eptr >= md->end_subject)
4016                  {
4017                  SCHECK_PARTIAL();
4018                  RRETURN(MATCH_NOMATCH);
4019                  }
4020              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
4021              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
4022              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
4023                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4024              }              }
4025            break;            /* Control never gets here */
4026    
4027            case PT_SC:            case PT_SC:
4028            for (fi = min;; fi++)            for (fi = min;; fi++)
4029              {              {
4030              RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM40);
4031              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4032              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max) RRETURN(MATCH_NOMATCH);
4033                if (eptr >= md->end_subject)
4034                  {
4035                  SCHECK_PARTIAL();
4036                  RRETURN(MATCH_NOMATCH);
4037                  }
4038              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
4039              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_script = UCD_SCRIPT(c);
4040              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
4041                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
4042              }              }
4043            break;            /* Control never gets here */
4044    
4045            default:            default:
4046            RRETURN(PCRE_ERROR_INTERNAL);            RRETURN(PCRE_ERROR_INTERNAL);
           break;  
4047            }            }
4048          }          }
4049    
# Line 2843  for (;;) Line 4054  for (;;)
4054          {          {
4055          for (fi = min;; fi++)          for (fi = min;; fi++)
4056            {            {
4057            RMATCH(rrc, eptr, ecode, offset_top, md, ims, eptrb, 0);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM41);
4058            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
4059            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max) RRETURN(MATCH_NOMATCH);
4060              if (eptr >= md->end_subject)
4061                {
4062                SCHECK_PARTIAL();
4063                RRETURN(MATCH_NOMATCH);
4064                }
4065            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
4066            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
4067            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);
4068   &n