/[pcre]/code/branches/pcre16/pcre_exec.c
ViewVC logotype

Diff of /code/branches/pcre16/pcre_exec.c

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 178 by ph10, Wed Jun 13 08:44:34 2007 UTC revision 366 by ph10, Mon Jul 14 15:45:32 2008 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) 1997-2007 University of Cambridge             Copyright (c) 1997-2008 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    #ifdef HAVE_CONFIG_H
46    #include "config.h"
47    #endif
48    
49  #define NLBLOCK md             /* Block containing newline information */  #define NLBLOCK md             /* Block containing newline information */
50  #define PSSTART start_subject  /* Field containing processed string start */  #define PSSTART start_subject  /* Field containing processed string start */
51  #define PSEND   end_subject    /* Field containing processed string end */  #define PSEND   end_subject    /* Field containing processed string end */
# Line 53  possible. There are also some static sup Line 57  possible. There are also some static sup
57  #undef min  #undef min
58  #undef max  #undef max
59    
 /* The chain of eptrblocks for tail recursions uses memory in stack workspace,  
 obtained at top level, the size of which is defined by EPTR_WORK_SIZE. */  
   
 #define EPTR_WORK_SIZE (1000)  
   
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_cbegroup       0x02  /* Could-be-empty unlimited repeat group */  #define match_cbegroup       0x02  /* Could-be-empty unlimited repeat group */
 #define match_tail_recursed  0x04  /* Tail recursive call */  
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 70  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 152  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 205  variable instead of being passed in the Line 237  variable instead of being passed in the
237  ****************************************************************************  ****************************************************************************
238  ***************************************************************************/  ***************************************************************************/
239    
240    /* Numbers for RMATCH calls. When this list is changed, the code at HEAP_RETURN
241  /* Numbers for RMATCH calls */  below must be updated in sync.  */
242    
243  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM6,  RM7,  RM8,  RM9,  RM10,  enum { RM1=1, RM2,  RM3,  RM4,  RM5,  RM6,  RM7,  RM8,  RM9,  RM10,
244         RM11,  RM12, RM13, RM14, RM15, RM16, RM17, RM18, RM19, RM20,         RM11,  RM12, RM13, RM14, RM15, RM16, RM17, RM18, RM19, RM20,
245         RM21,  RM22, RM23, RM24, RM25, RM26, RM27, RM28, RM29, RM30,         RM21,  RM22, RM23, RM24, RM25, RM26, RM27, RM28, RM29, RM30,
246         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,         RM31,  RM32, RM33, RM34, RM35, RM36, RM37, RM38, RM39, RM40,
247         RM41,  RM42, RM43, RM44, RM45, RM46, RM47 };         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. Note that the "rw" argument of RMATCH isn't  versions and production versions. Note that the "rw" argument of RMATCH isn't
# Line 384  Arguments: Line 416  Arguments:
416                   match_condassert - this is an assertion condition                   match_condassert - this is an assertion condition
417                   match_cbegroup - this is the start of an unlimited repeat                   match_cbegroup - this is the start of an unlimited repeat
418                     group that can match an empty string                     group that can match an empty string
                  match_tail_recursed - this is a tail_recursed group  
419     rdepth      the recursion depth     rdepth      the recursion depth
420    
421  Returns:       MATCH_MATCH if matched            )  these values are >= 0  Returns:       MATCH_MATCH if matched            )  these values are >= 0
# Line 586  original_ims = ims;    /* Save for reset Line 617  original_ims = ims;    /* Save for reset
617  string, the match_cbegroup flag is set. When this is the case, add the current  string, the match_cbegroup flag is set. When this is the case, add the current
618  subject pointer to the chain of such remembered pointers, to be checked when we  subject pointer to the chain of such remembered pointers, to be checked when we
619  hit the closing ket, in order to break infinite loops that match no characters.  hit the closing ket, in order to break infinite loops that match no characters.
620  When match() is called in other circumstances, don't add to the chain. If this  When match() is called in other circumstances, don't add to the chain. The
621  is a tail recursion, use a block from the workspace, as the one on the stack is  match_cbegroup flag must NOT be used with tail recursion, because the memory
622  already used. */  block that is used is on the stack, so a new one may be required for each
623    match(). */
624    
625  if ((flags & match_cbegroup) != 0)  if ((flags & match_cbegroup) != 0)
626    {    {
627    eptrblock *p;    newptrb.epb_saved_eptr = eptr;
628    if ((flags & match_tail_recursed) != 0)    newptrb.epb_prev = eptrb;
629      {    eptrb = &newptrb;
     if (md->eptrn >= EPTR_WORK_SIZE) RRETURN(PCRE_ERROR_NULLWSLIMIT);  
     p = md->eptrchain + md->eptrn++;  
     }  
   else p = &newptrb;  
   p->epb_saved_eptr = eptr;  
   p->epb_prev = eptrb;  
   eptrb = p;  
630    }    }
631    
632  /* Now start processing the opcodes. */  /* Now start processing the opcodes. */
# Line 621  for (;;) Line 646  for (;;)
646    
647    switch(op)    switch(op)
648      {      {
649        case OP_FAIL:
650        RRETURN(MATCH_NOMATCH);
651    
652        case OP_PRUNE:
653        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
654          ims, eptrb, flags, RM51);
655        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
656        RRETURN(MATCH_PRUNE);
657    
658        case OP_COMMIT:
659        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
660          ims, eptrb, flags, RM52);
661        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
662        RRETURN(MATCH_COMMIT);
663    
664        case OP_SKIP:
665        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
666          ims, eptrb, flags, RM53);
667        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
668        md->start_match_ptr = eptr;   /* Pass back current position */
669        RRETURN(MATCH_SKIP);
670    
671        case OP_THEN:
672        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
673          ims, eptrb, flags, RM54);
674        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
675        RRETURN(MATCH_THEN);
676    
677      /* Handle a capturing bracket. If there is space in the offset vector, save      /* Handle a capturing bracket. If there is space in the offset vector, save
678      the current subject position in the working slot at the top of the vector.      the current subject position in the working slot at the top of the vector.
679      We mustn't change the current values of the data slot, because they may be      We mustn't change the current values of the data slot, because they may be
# Line 662  for (;;) Line 715  for (;;)
715          {          {
716          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,          RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md,
717            ims, eptrb, flags, RM1);            ims, eptrb, flags, RM1);
718          if (rrc != MATCH_NOMATCH) RRETURN(rrc);          if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
719          md->capture_last = save_capture_last;          md->capture_last = save_capture_last;
720          ecode += GET(ecode, 1);          ecode += GET(ecode, 1);
721          }          }
# Line 677  for (;;) Line 730  for (;;)
730        RRETURN(MATCH_NOMATCH);        RRETURN(MATCH_NOMATCH);
731        }        }
732    
733      /* Insufficient room for saving captured contents. Treat as a non-capturing      /* FALL THROUGH ... Insufficient room for saving captured contents. Treat
734      bracket. */      as a non-capturing bracket. */
735    
736        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
737        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
738    
739      DPRINTF(("insufficient capture room: treat as non-capturing\n"));      DPRINTF(("insufficient capture room: treat as non-capturing\n"));
740    
741        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
742        /* VVVVVVVVVVVVVVVVVVVVVVVVV */
743    
744      /* Non-capturing bracket. Loop for all the alternatives. When we get to the      /* Non-capturing bracket. Loop for all the alternatives. When we get to the
745      final alternative within the brackets, we would return the result of a      final alternative within the brackets, we would return the result of a
746      recursive call to match() whatever happened. We can reduce stack usage by      recursive call to match() whatever happened. We can reduce stack usage by
747      turning this into a tail recursion. */      turning this into a tail recursion, except in the case when match_cbegroup
748        is set.*/
749    
750      case OP_BRA:      case OP_BRA:
751      case OP_SBRA:      case OP_SBRA:
# Line 693  for (;;) Line 753  for (;;)
753      flags = (op >= OP_SBRA)? match_cbegroup : 0;      flags = (op >= OP_SBRA)? match_cbegroup : 0;
754      for (;;)      for (;;)
755        {        {
756        if (ecode[GET(ecode, 1)] != OP_ALT)        if (ecode[GET(ecode, 1)] != OP_ALT)   /* Final alternative */
757          {          {
758          ecode += _pcre_OP_lengths[*ecode];          if (flags == 0)    /* Not a possibly empty group */
759          flags |= match_tail_recursed;            {
760          DPRINTF(("bracket 0 tail recursion\n"));            ecode += _pcre_OP_lengths[*ecode];
761          goto TAIL_RECURSE;            DPRINTF(("bracket 0 tail recursion\n"));
762              goto TAIL_RECURSE;
763              }
764    
765            /* Possibly empty group; can't use tail recursion. */
766    
767            RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,
768              eptrb, flags, RM48);
769            RRETURN(rrc);
770          }          }
771    
772        /* For non-final alternatives, continue the loop for a NOMATCH result;        /* For non-final alternatives, continue the loop for a NOMATCH result;
# Line 706  for (;;) Line 774  for (;;)
774    
775        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,        RMATCH(eptr, ecode + _pcre_OP_lengths[*ecode], offset_top, md, ims,
776          eptrb, flags, RM2);          eptrb, flags, RM2);
777        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
778        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
779        }        }
780      /* Control never reaches here. */      /* Control never reaches here. */
# Line 754  for (;;) Line 822  for (;;)
822          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);          ecode += 1 + LINK_SIZE + GET(ecode, LINK_SIZE + 2);
823          while (*ecode == OP_ALT) ecode += GET(ecode, 1);          while (*ecode == OP_ALT) ecode += GET(ecode, 1);
824          }          }
825        else if (rrc != MATCH_NOMATCH)        else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
826          {          {
827          RRETURN(rrc);         /* Need braces because of following else */          RRETURN(rrc);         /* Need braces because of following else */
828          }          }
# Line 766  for (;;) Line 834  for (;;)
834        }        }
835    
836      /* 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,
837      we can use tail recursion to avoid using another stack frame. If the second      we can use tail recursion to avoid using another stack frame, except when
838      alternative doesn't exist, we can just plough on. */      match_cbegroup is required for an unlimited repeat of a possibly empty
839        group. If the second alternative doesn't exist, we can just plough on. */
840    
841      if (condition || *ecode == OP_ALT)      if (condition || *ecode == OP_ALT)
842        {        {
843        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
844        flags = match_tail_recursed | ((op == OP_SCOND)? match_cbegroup : 0);        if (op == OP_SCOND)        /* Possibly empty group */
845        goto TAIL_RECURSE;          {
846            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, match_cbegroup, RM49);
847            RRETURN(rrc);
848            }
849          else                       /* Group must match something */
850            {
851            flags = 0;
852            goto TAIL_RECURSE;
853            }
854        }        }
855      else      else                         /* Condition false & no 2nd alternative */
856        {        {
857        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
858        }        }
859      break;      break;
860    
861    
862      /* End of the pattern. If we are in a top-level recursion, we should      /* End of the pattern, either real or forced. If we are in a top-level
863      restore the offsets appropriately and continue from after the call. */      recursion, we should restore the offsets appropriately and continue from
864        after the call. */
865    
866        case OP_ACCEPT:
867      case OP_END:      case OP_END:
868      if (md->recursive != NULL && md->recursive->group_num == 0)      if (md->recursive != NULL && md->recursive->group_num == 0)
869        {        {
# Line 805  for (;;) Line 884  for (;;)
884      if (md->notempty && eptr == mstart) RRETURN(MATCH_NOMATCH);      if (md->notempty && eptr == mstart) RRETURN(MATCH_NOMATCH);
885      md->end_match_ptr = eptr;           /* Record where we ended */      md->end_match_ptr = eptr;           /* Record where we ended */
886      md->end_offset_top = offset_top;    /* and how many extracts were taken */      md->end_offset_top = offset_top;    /* and how many extracts were taken */
887      md->start_match_ptr = mstart;  /* and the start (\K can modify) */      md->start_match_ptr = mstart;       /* and the start (\K can modify) */
888      RRETURN(MATCH_MATCH);      RRETURN(MATCH_MATCH);
889    
890      /* Change option settings */      /* Change option settings */
# Line 829  for (;;) Line 908  for (;;)
908        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,
909          RM4);          RM4);
910        if (rrc == MATCH_MATCH) break;        if (rrc == MATCH_MATCH) break;
911        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
912        ecode += GET(ecode, 1);        ecode += GET(ecode, 1);
913        }        }
914      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 856  for (;;) Line 935  for (;;)
935        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, NULL, 0,
936          RM5);          RM5);
937        if (rrc == MATCH_MATCH) RRETURN(MATCH_NOMATCH);        if (rrc == MATCH_MATCH) RRETURN(MATCH_NOMATCH);
938        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
939        ecode += GET(ecode,1);        ecode += GET(ecode,1);
940        }        }
941      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 880  for (;;) Line 959  for (;;)
959          {          {
960          eptr--;          eptr--;
961          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);          if (eptr < md->start_subject) RRETURN(MATCH_NOMATCH);
962          BACKCHAR(eptr)          BACKCHAR(eptr);
963          }          }
964        }        }
965      else      else
# Line 993  for (;;) Line 1072  for (;;)
1072              (pcre_free)(new_recursive.offset_save);              (pcre_free)(new_recursive.offset_save);
1073            RRETURN(MATCH_MATCH);            RRETURN(MATCH_MATCH);
1074            }            }
1075          else if (rrc != MATCH_NOMATCH)          else if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN)
1076            {            {
1077            DPRINTF(("Recursion gave error %d\n", rrc));            DPRINTF(("Recursion gave error %d\n", rrc));
1078            RRETURN(rrc);            RRETURN(rrc);
# Line 1027  for (;;) Line 1106  for (;;)
1106    
1107      do      do
1108        {        {
1109        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM7);
         eptrb, 0, RM7);  
1110        if (rrc == MATCH_MATCH) break;        if (rrc == MATCH_MATCH) break;
1111        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH && rrc != MATCH_THEN) RRETURN(rrc);
1112        ecode += GET(ecode,1);        ecode += GET(ecode,1);
1113        }        }
1114      while (*ecode == OP_ALT);      while (*ecode == OP_ALT);
# Line 1073  for (;;) Line 1151  for (;;)
1151    
1152      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1153        {        {
1154        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM8);
         RM8);  
1155        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1156        ecode = prev;        ecode = prev;
1157        flags = match_tail_recursed;        flags = 0;
1158        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1159        }        }
1160      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
# Line 1085  for (;;) Line 1162  for (;;)
1162        RMATCH(eptr, prev, offset_top, md, ims, eptrb, match_cbegroup, RM9);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, match_cbegroup, RM9);
1163        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1164        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1165        flags = match_tail_recursed;        flags = 0;
1166        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1167        }        }
1168      /* Control never gets here */      /* Control never gets here */
# Line 1097  for (;;) Line 1174  for (;;)
1174      do ecode += GET(ecode,1); while (*ecode == OP_ALT);      do ecode += GET(ecode,1); while (*ecode == OP_ALT);
1175      break;      break;
1176    
1177      /* BRAZERO and BRAMINZERO occur just before a bracket group, indicating      /* BRAZERO, BRAMINZERO and SKIPZERO occur just before a bracket group,
1178      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
1179      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
1180      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
1181      preceded by BRAZERO or BRAMINZERO. */      optional ones preceded by BRAZERO or BRAMINZERO. */
1182    
1183      case OP_BRAZERO:      case OP_BRAZERO:
1184        {        {
# Line 1123  for (;;) Line 1200  for (;;)
1200        }        }
1201      break;      break;
1202    
1203        case OP_SKIPZERO:
1204          {
1205          next = ecode+1;
1206          do next += GET(next,1); while (*next == OP_ALT);
1207          ecode = next + 1 + LINK_SIZE;
1208          }
1209        break;
1210    
1211      /* End of a group, repeated or non-repeating. */      /* End of a group, repeated or non-repeating. */
1212    
1213      case OP_KET:      case OP_KET:
# Line 1216  for (;;) Line 1301  for (;;)
1301    
1302      /* 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
1303      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
1304      tail recursion to avoid using another stack frame. */      tail recursion to avoid using another stack frame, unless we have an
1305        unlimited repeat of a group that can match an empty string. */
1306    
1307      flags = (*prev >= OP_SBRA)? match_cbegroup : 0;      flags = (*prev >= OP_SBRA)? match_cbegroup : 0;
1308    
1309      if (*ecode == OP_KETRMIN)      if (*ecode == OP_KETRMIN)
1310        {        {
1311        RMATCH(eptr, ecode + 1+LINK_SIZE, offset_top, md, ims, eptrb, 0,        RMATCH(eptr, ecode + 1 + LINK_SIZE, offset_top, md, ims, eptrb, 0, RM12);
         RM12);  
1312        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1313          if (flags != 0)    /* Could match an empty string */
1314            {
1315            RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM50);
1316            RRETURN(rrc);
1317            }
1318        ecode = prev;        ecode = prev;
       flags |= match_tail_recursed;  
1319        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1320        }        }
1321      else  /* OP_KETRMAX */      else  /* OP_KETRMAX */
# Line 1234  for (;;) Line 1323  for (;;)
1323        RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM13);        RMATCH(eptr, prev, offset_top, md, ims, eptrb, flags, RM13);
1324        if (rrc != MATCH_NOMATCH) RRETURN(rrc);        if (rrc != MATCH_NOMATCH) RRETURN(rrc);
1325        ecode += 1 + LINK_SIZE;        ecode += 1 + LINK_SIZE;
1326        flags = match_tail_recursed;        flags = 0;
1327        goto TAIL_RECURSE;        goto TAIL_RECURSE;
1328        }        }
1329      /* Control never gets here */      /* Control never gets here */
# Line 1366  for (;;) Line 1455  for (;;)
1455      /* Match a single character type; inline for speed */      /* Match a single character type; inline for speed */
1456    
1457      case OP_ANY:      case OP_ANY:
1458      if ((ims & PCRE_DOTALL) == 0)      if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
1459        {      /* Fall through */
1460        if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);  
1461        }      case OP_ALLANY:
1462      if (eptr++ >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr++ >= md->end_subject) RRETURN(MATCH_NOMATCH);
1463      if (utf8)      if (utf8) while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
       while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  
1464      ecode++;      ecode++;
1465      break;      break;
1466    
# Line 1471  for (;;) Line 1559  for (;;)
1559        case 0x000d:        case 0x000d:
1560        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;        if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
1561        break;        break;
1562    
1563        case 0x000a:        case 0x000a:
1564          break;
1565    
1566        case 0x000b:        case 0x000b:
1567        case 0x000c:        case 0x000c:
1568        case 0x0085:        case 0x0085:
1569        case 0x2028:        case 0x2028:
1570        case 0x2029:        case 0x2029:
1571          if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
1572        break;        break;
1573        }        }
1574      ecode++;      ecode++;
# Line 1587  for (;;) Line 1679  for (;;)
1679      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1680      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1681        {        {
1682        int chartype, script;        const ucd_record * prop = GET_UCD(c);
       int category = _pcre_ucp_findprop(c, &chartype, &script);  
1683    
1684        switch(ecode[1])        switch(ecode[1])
1685          {          {
# Line 1597  for (;;) Line 1688  for (;;)
1688          break;          break;
1689    
1690          case PT_LAMP:          case PT_LAMP:
1691          if ((chartype == ucp_Lu ||          if ((prop->chartype == ucp_Lu ||
1692               chartype == ucp_Ll ||               prop->chartype == ucp_Ll ||
1693               chartype == ucp_Lt) == (op == OP_NOTPROP))               prop->chartype == ucp_Lt) == (op == OP_NOTPROP))
1694            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1695           break;           break;
1696    
1697          case PT_GC:          case PT_GC:
1698          if ((ecode[2] != category) == (op == OP_PROP))          if ((ecode[2] != _pcre_ucp_gentype[prop->chartype]) == (op == OP_PROP))
1699            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1700          break;          break;
1701    
1702          case PT_PC:          case PT_PC:
1703          if ((ecode[2] != chartype) == (op == OP_PROP))          if ((ecode[2] != prop->chartype) == (op == OP_PROP))
1704            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1705          break;          break;
1706    
1707          case PT_SC:          case PT_SC:
1708          if ((ecode[2] != script) == (op == OP_PROP))          if ((ecode[2] != prop->script) == (op == OP_PROP))
1709            RRETURN(MATCH_NOMATCH);            RRETURN(MATCH_NOMATCH);
1710          break;          break;
1711    
# Line 1633  for (;;) Line 1724  for (;;)
1724      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);      if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
1725      GETCHARINCTEST(c, eptr);      GETCHARINCTEST(c, eptr);
1726        {        {
1727        int chartype, script;        int category = UCD_CATEGORY(c);
       int category = _pcre_ucp_findprop(c, &chartype, &script);  
1728        if (category == ucp_M) RRETURN(MATCH_NOMATCH);        if (category == ucp_M) RRETURN(MATCH_NOMATCH);
1729        while (eptr < md->end_subject)        while (eptr < md->end_subject)
1730          {          {
# Line 1643  for (;;) Line 1733  for (;;)
1733            {            {
1734            GETCHARLEN(c, eptr, len);            GETCHARLEN(c, eptr, len);
1735            }            }
1736          category = _pcre_ucp_findprop(c, &chartype, &script);          category = UCD_CATEGORY(c);
1737          if (category != ucp_M) break;          if (category != ucp_M) break;
1738          eptr += len;          eptr += len;
1739          }          }
# Line 1664  for (;;) Line 1754  for (;;)
1754      case OP_REF:      case OP_REF:
1755        {        {
1756        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */        offset = GET2(ecode, 1) << 1;               /* Doubled ref number */
1757        ecode += 3;                                 /* Advance past item */        ecode += 3;
1758    
1759          /* If the reference is unset, there are two possibilities:
1760    
1761          (a) In the default, Perl-compatible state, set the length to be longer
1762          than the amount of subject left; this ensures that every attempt at a
1763          match fails. We can't just fail here, because of the possibility of
1764          quantifiers with zero minima.
1765    
1766          (b) If the JavaScript compatibility flag is set, set the length to zero
1767          so that the back reference matches an empty string.
1768    
1769          Otherwise, set the length to the length of what was matched by the
1770          referenced subpattern. */
1771    
1772        /* If the reference is unset, set the length to be longer than the amount        if (offset >= offset_top || md->offset_vector[offset] < 0)
1773        of subject left; this ensures that every attempt at a match fails. We          length = (md->jscript_compat)? 0 : md->end_subject - eptr + 1;
1774        can't just fail here, because of the possibility of quantifiers with zero        else
1775        minima. */          length = md->offset_vector[offset+1] - md->offset_vector[offset];
   
       length = (offset >= offset_top || md->offset_vector[offset] < 0)?  
         md->end_subject - eptr + 1 :  
         md->offset_vector[offset+1] - md->offset_vector[offset];  
1776    
1777        /* Set up for repetition, or handle the non-repeated case */        /* Set up for repetition, or handle the non-repeated case */
1778    
# Line 2033  for (;;) Line 2132  for (;;)
2132            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM21);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM21);
2133            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2134            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
2135            BACKCHAR(eptr)            if (utf8) BACKCHAR(eptr);
2136            }            }
2137          RRETURN(MATCH_NOMATCH);          RRETURN(MATCH_NOMATCH);
2138          }          }
# Line 2099  for (;;) Line 2198  for (;;)
2198          if (fc != dc)          if (fc != dc)
2199            {            {
2200  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2201            if (dc != _pcre_ucp_othercase(fc))            if (dc != UCD_OTHERCASE(fc))
2202  #endif  #endif
2203              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2204            }            }
# Line 2190  for (;;) Line 2289  for (;;)
2289  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2290          unsigned int othercase;          unsigned int othercase;
2291          if ((ims & PCRE_CASELESS) != 0 &&          if ((ims & PCRE_CASELESS) != 0 &&
2292              (othercase = _pcre_ucp_othercase(fc)) != NOTACHAR)              (othercase = UCD_OTHERCASE(fc)) != fc)
2293            oclength = _pcre_ord2utf8(othercase, occhars);            oclength = _pcre_ord2utf8(othercase, occhars);
2294          else oclength = 0;          else oclength = 0;
2295  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
# Line 2510  for (;;) Line 2609  for (;;)
2609              {              {
2610              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM28);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM28);
2611              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2612                if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2613              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
2614              if (d < 256) d = md->lcc[d];              if (d < 256) d = md->lcc[d];
2615              if (fi >= max || eptr >= md->end_subject || fc == d)              if (fc == d) RRETURN(MATCH_NOMATCH);
2616                RRETURN(MATCH_NOMATCH);  
2617              }              }
2618            }            }
2619          else          else
# Line 2619  for (;;) Line 2719  for (;;)
2719              {              {
2720              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM32);              RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM32);
2721              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
2722                if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2723              GETCHARINC(d, eptr);              GETCHARINC(d, eptr);
2724              if (fi >= max || eptr >= md->end_subject || fc == d)              if (fc == d) RRETURN(MATCH_NOMATCH);
               RRETURN(MATCH_NOMATCH);  
2725              }              }
2726            }            }
2727          else          else
# Line 2786  for (;;) Line 2886  for (;;)
2886            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2887              {              {
2888              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2889              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2890              }              }
2891            break;            break;
2892    
# Line 2794  for (;;) Line 2894  for (;;)
2894            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2895              {              {
2896              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2897              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2898              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
2899              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
2900                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
2901                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
# Line 2807  for (;;) Line 2907  for (;;)
2907            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2908              {              {
2909              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2910              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2911              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
2912              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
2913                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2914              }              }
# Line 2818  for (;;) Line 2918  for (;;)
2918            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2919              {              {
2920              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2921              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2922              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
2923              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
2924                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2925              }              }
# Line 2829  for (;;) Line 2929  for (;;)
2929            for (i = 1; i <= min; i++)            for (i = 1; i <= min; i++)
2930              {              {
2931              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2932              GETCHARINC(c, eptr);              GETCHARINCTEST(c, eptr);
2933              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_script = UCD_SCRIPT(c);
2934              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
2935                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
2936              }              }
# Line 2849  for (;;) Line 2949  for (;;)
2949          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2950            {            {
2951            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
2952            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
2953            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);
2954            while (eptr < md->end_subject)            while (eptr < md->end_subject)
2955              {              {
# Line 2858  for (;;) Line 2958  for (;;)
2958                {                {
2959                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
2960                }                }
2961              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
2962              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
2963              eptr += len;              eptr += len;
2964              }              }
# Line 2876  for (;;) Line 2976  for (;;)
2976          case OP_ANY:          case OP_ANY:
2977          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
2978            {            {
2979            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject || IS_NEWLINE(eptr))
                ((ims & PCRE_DOTALL) == 0 && IS_NEWLINE(eptr)))  
2980              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
2981            eptr++;            eptr++;
2982            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
2983            }            }
2984          break;          break;
2985    
2986            case OP_ALLANY:
2987            for (i = 1; i <= min; i++)
2988              {
2989              if (eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
2990              eptr++;
2991              while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
2992              }
2993            break;
2994    
2995          case OP_ANYBYTE:          case OP_ANYBYTE:
2996          eptr += min;          eptr += min;
2997          break;          break;
# Line 2899  for (;;) Line 3007  for (;;)
3007              case 0x000d:              case 0x000d:
3008              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3009              break;              break;
3010    
3011              case 0x000a:              case 0x000a:
3012                break;
3013    
3014              case 0x000b:              case 0x000b:
3015              case 0x000c:              case 0x000c:
3016              case 0x0085:              case 0x0085:
3017              case 0x2028:              case 0x2028:
3018              case 0x2029:              case 0x2029:
3019                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3020              break;              break;
3021              }              }
3022            }            }
# Line 2941  for (;;) Line 3053  for (;;)
3053              }              }
3054            }            }
3055          break;          break;
3056    
3057          case OP_HSPACE:          case OP_HSPACE:
3058          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3059            {            {
# Line 2973  for (;;) Line 3085  for (;;)
3085              }              }
3086            }            }
3087          break;          break;
3088    
3089          case OP_NOT_VSPACE:          case OP_NOT_VSPACE:
3090          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3091            {            {
# Line 2993  for (;;) Line 3105  for (;;)
3105              }              }
3106            }            }
3107          break;          break;
3108    
3109          case OP_VSPACE:          case OP_VSPACE:
3110          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3111            {            {
# Line 3009  for (;;) Line 3121  for (;;)
3121              case 0x85:      /* NEL */              case 0x85:      /* NEL */
3122              case 0x2028:    /* LINE SEPARATOR */              case 0x2028:    /* LINE SEPARATOR */
3123              case 0x2029:    /* PARAGRAPH SEPARATOR */              case 0x2029:    /* PARAGRAPH SEPARATOR */
3124              break;              break;
3125              }              }
3126            }            }
3127          break;          break;
# Line 3038  for (;;) Line 3150  for (;;)
3150          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3151            {            {
3152            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject ||
3153               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_space) != 0))               (*eptr < 128 && (md->ctypes[*eptr] & ctype_space) != 0))
3154              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3155            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3156            }            }
3157          break;          break;
3158    
# Line 3058  for (;;) Line 3170  for (;;)
3170          for (i = 1; i <= min; i++)          for (i = 1; i <= min; i++)
3171            {            {
3172            if (eptr >= md->end_subject ||            if (eptr >= md->end_subject ||
3173               (*eptr < 128 && (md->ctypes[*eptr++] & ctype_word) != 0))               (*eptr < 128 && (md->ctypes[*eptr] & ctype_word) != 0))
3174              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3175            while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;            while (++eptr < md->end_subject && (*eptr & 0xc0) == 0x80);
3176            }            }
3177          break;          break;
3178    
# Line 3088  for (;;) Line 3200  for (;;)
3200        switch(ctype)        switch(ctype)
3201          {          {
3202          case OP_ANY:          case OP_ANY:
3203          if ((ims & PCRE_DOTALL) == 0)          for (i = 1; i <= min; i++)
3204            {            {
3205            for (i = 1; i <= min; i++)            if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);
3206              {            eptr++;
             if (IS_NEWLINE(eptr)) RRETURN(MATCH_NOMATCH);  
             eptr++;  
             }  
3207            }            }
3208          else eptr += min;          break;
3209    
3210            case OP_ALLANY:
3211            eptr += min;
3212          break;          break;
3213    
3214          case OP_ANYBYTE:          case OP_ANYBYTE:
# Line 3117  for (;;) Line 3229  for (;;)
3229              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;              if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3230              break;              break;
3231              case 0x000a:              case 0x000a:
3232                break;
3233    
3234              case 0x000b:              case 0x000b:
3235              case 0x000c:              case 0x000c:
3236              case 0x0085:              case 0x0085:
3237                if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3238              break;              break;
3239              }              }
3240            }            }
# Line 3150  for (;;) Line 3265  for (;;)
3265              case 0x09:      /* HT */              case 0x09:      /* HT */
3266              case 0x20:      /* SPACE */              case 0x20:      /* SPACE */
3267              case 0xa0:      /* NBSP */              case 0xa0:      /* NBSP */
3268              break;              break;
3269              }              }
3270            }            }
3271          break;          break;
# Line 3184  for (;;) Line 3299  for (;;)
3299              case 0x0c:      /* FF */              case 0x0c:      /* FF */
3300              case 0x0d:      /* CR */              case 0x0d:      /* CR */
3301              case 0x85:      /* NEL */              case 0x85:      /* NEL */
3302              break;              break;
3303              }              }
3304            }            }
3305          break;          break;
# Line 3259  for (;;) Line 3374  for (;;)
3374              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3375              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3376              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3377              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3378              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
3379                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
3380                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
# Line 3274  for (;;) Line 3389  for (;;)
3389              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3390              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3391              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3392              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3393              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
3394                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3395              }              }
# Line 3287  for (;;) Line 3402  for (;;)
3402              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3403              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3404              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3405              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3406              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
3407                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3408              }              }
# Line 3300  for (;;) Line 3415  for (;;)
3415              if (rrc != MATCH_NOMATCH) RRETURN(rrc);              if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3416              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);              if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3417              GETCHARINC(c, eptr);              GETCHARINC(c, eptr);
3418              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_script = UCD_SCRIPT(c);
3419              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
3420                RRETURN(MATCH_NOMATCH);                RRETURN(MATCH_NOMATCH);
3421              }              }
# Line 3322  for (;;) Line 3437  for (;;)
3437            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3438            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);            if (fi >= max || eptr >= md->end_subject) RRETURN(MATCH_NOMATCH);
3439            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3440            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
3441            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);            if (prop_category == ucp_M) RRETURN(MATCH_NOMATCH);
3442            while (eptr < md->end_subject)            while (eptr < md->end_subject)
3443              {              {
# Line 3331  for (;;) Line 3446  for (;;)
3446                {                {
3447                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
3448                }                }
3449              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3450              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
3451              eptr += len;              eptr += len;
3452              }              }
# Line 3350  for (;;) Line 3465  for (;;)
3465            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM42);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM42);
3466            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3467            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
3468                 (ctype == OP_ANY && (ims & PCRE_DOTALL) == 0 &&                 (ctype == OP_ANY && IS_NEWLINE(eptr)))
                 IS_NEWLINE(eptr)))  
3469              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3470    
3471            GETCHARINC(c, eptr);            GETCHARINC(c, eptr);
3472            switch(ctype)            switch(ctype)
3473              {              {
3474              case OP_ANY:        /* This is the DOTALL case */              case OP_ANY:        /* This is the non-NL case */
3475              break;              case OP_ALLANY:
   
3476              case OP_ANYBYTE:              case OP_ANYBYTE:
3477              break;              break;
3478    
# Line 3371  for (;;) Line 3484  for (;;)
3484                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3485                break;                break;
3486                case 0x000a:                case 0x000a:
3487                  break;
3488    
3489                case 0x000b:                case 0x000b:
3490                case 0x000c:                case 0x000c:
3491                case 0x0085:                case 0x0085:
3492                case 0x2028:                case 0x2028:
3493                case 0x2029:                case 0x2029:
3494                  if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3495                break;                break;
3496                }                }
3497              break;              break;
# Line 3508  for (;;) Line 3624  for (;;)
3624            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM43);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM43);
3625            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3626            if (fi >= max || eptr >= md->end_subject ||            if (fi >= max || eptr >= md->end_subject ||
3627                 ((ims & PCRE_DOTALL) == 0 && IS_NEWLINE(eptr)))                 (ctype == OP_ANY && IS_NEWLINE(eptr)))
3628              RRETURN(MATCH_NOMATCH);              RRETURN(MATCH_NOMATCH);
3629    
3630            c = *eptr++;            c = *eptr++;
3631            switch(ctype)            switch(ctype)
3632              {              {
3633              case OP_ANY:   /* This is the DOTALL case */              case OP_ANY:     /* This is the non-NL case */
3634              break;              case OP_ALLANY:
   
3635              case OP_ANYBYTE:              case OP_ANYBYTE:
3636              break;              break;
3637    
# Line 3527  for (;;) Line 3642  for (;;)
3642                case 0x000d:                case 0x000d:
3643                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;                if (eptr < md->end_subject && *eptr == 0x0a) eptr++;
3644                break;                break;
3645    
3646                case 0x000a:                case 0x000a:
3647                  break;
3648    
3649                case 0x000b:                case 0x000b:
3650                case 0x000c:                case 0x000c:
3651                case 0x0085:                case 0x0085:
3652                  if (md->bsr_anycrlf) RRETURN(MATCH_NOMATCH);
3653                break;                break;
3654                }                }
3655              break;              break;
# Line 3645  for (;;) Line 3764  for (;;)
3764              int len = 1;              int len = 1;
3765              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
3766              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
3767              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3768              if ((prop_chartype == ucp_Lu ||              if ((prop_chartype == ucp_Lu ||
3769                   prop_chartype == ucp_Ll ||                   prop_chartype == ucp_Ll ||
3770                   prop_chartype == ucp_Lt) == prop_fail_result)                   prop_chartype == ucp_Lt) == prop_fail_result)
# Line 3660  for (;;) Line 3779  for (;;)
3779              int len = 1;              int len = 1;
3780              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
3781              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
3782              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3783              if ((prop_category == prop_value) == prop_fail_result)              if ((prop_category == prop_value) == prop_fail_result)
3784                break;                break;
3785              eptr+= len;              eptr+= len;
# Line 3673  for (;;) Line 3792  for (;;)
3792              int len = 1;              int len = 1;
3793              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
3794              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
3795              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_chartype = UCD_CHARTYPE(c);
3796              if ((prop_chartype == prop_value) == prop_fail_result)              if ((prop_chartype == prop_value) == prop_fail_result)
3797                break;                break;
3798              eptr+= len;              eptr+= len;
# Line 3686  for (;;) Line 3805  for (;;)
3805              int len = 1;              int len = 1;
3806              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
3807              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
3808              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_script = UCD_SCRIPT(c);
3809              if ((prop_script == prop_value) == prop_fail_result)              if ((prop_script == prop_value) == prop_fail_result)
3810                break;                break;
3811              eptr+= len;              eptr+= len;
# Line 3702  for (;;) Line 3821  for (;;)
3821            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM44);            RMATCH(eptr, ecode, offset_top, md, ims, eptrb, 0, RM44);
3822            if (rrc != MATCH_NOMATCH) RRETURN(rrc);            if (rrc != MATCH_NOMATCH) RRETURN(rrc);
3823            if (eptr-- == pp) break;        /* Stop if tried at original pos */            if (eptr-- == pp) break;        /* Stop if tried at original pos */
3824            BACKCHAR(eptr);            if (utf8) BACKCHAR(eptr);
3825            }            }
3826          }          }
3827    
# Line 3715  for (;;) Line 3834  for (;;)
3834            {            {
3835            if (eptr >= md->end_subject) break;            if (eptr >= md->end_subject) break;
3836            GETCHARINCTEST(c, eptr);            GETCHARINCTEST(c, eptr);
3837            prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);            prop_category = UCD_CATEGORY(c);
3838            if (prop_category == ucp_M) break;            if (prop_category == ucp_M) break;
3839            while (eptr < md->end_subject)            while (eptr < md->end_subject)
3840              {              {
# Line 3724  for (;;) Line 3843  for (;;)
3843                {                {
3844                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
3845                }                }
3846              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3847              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
3848              eptr += len;              eptr += len;
3849              }              }
# Line 3741  for (;;) Line 3860  for (;;)
3860            for (;;)                        /* Move back over one extended */            for (;;)                        /* Move back over one extended */
3861              {              {
3862              int len = 1;              int len = 1;
             BACKCHAR(eptr);  
3863              if (!utf8) c = *eptr; else              if (!utf8) c = *eptr; else
3864                {                {
3865                  BACKCHAR(eptr);
3866                GETCHARLEN(c, eptr, len);                GETCHARLEN(c, eptr, len);
3867                }                }
3868              prop_category = _pcre_ucp_findprop(c, &prop_chartype, &prop_script);              prop_category = UCD_CATEGORY(c);
3869              if (prop_category != ucp_M) break;              if (prop_category != ucp_M) break;
3870              eptr--;              eptr--;
3871              }              }
# Line 3764  for (;;) Line 3883  for (;;)
3883          switch(ctype)          switch(ctype)
3884            {            {
3885            case OP_ANY:            case OP_ANY:
   
           /* Special code is required for UTF8, but when the maximum is  
           unlimited we don't need it, so we repeat the non-UTF8 code. This is  
           probably worth it, because .* is quite a common idiom. */  
   
3886            if (max < INT_MAX)            if (max < INT_MAX)
3887              {              {
3888              if ((ims & PCRE_DOTALL) == 0)              for (i = min; i < max; i++)
               {  
               for (i = min; i < max; i++)  
                 {  
                 if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;  
                 eptr++;  
                 while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  
                 }  
               }  
             else  
3889                {                {
3890                for (i = min; i < max; i++)                if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
3891                  {                eptr++;
3892                  if (eptr >= md->end_subject) break;                while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
                 eptr++;  
                 while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;  
                 }  
3893                }                }
3894              }              }
3895    
# Line 3795  for (;;) Line 3897  for (;;)
3897    
3898            else            else
3899              {              {
3900              if ((ims & PCRE_DOTALL) == 0)              for (i = min; i < max; i++)
3901                {                {
3902                for (i = min; i < max; i++)                if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
3903                  {                eptr++;
3904                  if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;                while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
                 eptr++;  
                 }  
               break;  
3905                }                }
3906              else              }
3907              break;
3908    
3909              case OP_ALLANY:
3910              if (max < INT_MAX)
3911                {
3912                for (i = min; i < max; i++)
3913                {                {
3914                c = max - min;                if (eptr >= md->end_subject) break;
3915                if (c > (unsigned int)(md->end_subject - eptr))                eptr++;
3916                  c = md->end_subject - eptr;                while (eptr < md->end_subject && (*eptr & 0xc0) == 0x80) eptr++;
               eptr += c;  
3917                }                }
3918              }              }
3919              else eptr = md->end_subject;   /* Unlimited UTF-8 repeat */
3920            break;            break;
3921    
3922            /* The byte case is the same as non-UTF8 */            /* The byte case is the same as non-UTF8 */
# Line 3836  for (;;) Line 3941  for (;;)
3941                }                }
3942              else              else
3943                {                {
3944                if (c != 0x000a && c != 0x000b && c != 0x000c &&                if (c != 0x000a &&
3945                    c != 0x0085 && c != 0x2028 && c != 0x2029)                    (md->bsr_anycrlf ||
3946                       (c != 0x000b && c != 0x000c &&
3947                        c != 0x0085 && c != 0x2028 && c != 0x2029)))
3948                  break;                  break;
3949                eptr += len;                eptr += len;
3950                }                }
# Line 3845  for (;;) Line 3952  for (;;)
3952            break;            break;
3953    
3954            case OP_NOT_HSPACE:            case OP_NOT_HSPACE:
3955            case OP_HSPACE:            case OP_HSPACE:
3956            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3957              {              {
3958              BOOL gotspace;              BOOL gotspace;
3959              int len = 1;              int len = 1;
3960              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
3961              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
3962              switch(c)              switch(c)
3963                {                {
3964                default: gotspace = FALSE; break;                default: gotspace = FALSE; break;
3965                case 0x09:      /* HT */                case 0x09:      /* HT */
3966                case 0x20:      /* SPACE */                case 0x20:      /* SPACE */
3967                case 0xa0:      /* NBSP */                case 0xa0:      /* NBSP */
# Line 3875  for (;;) Line 3982  for (;;)
3982                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */                case 0x205f:    /* MEDIUM MATHEMATICAL SPACE */
3983                case 0x3000:    /* IDEOGRAPHIC SPACE */                case 0x3000:    /* IDEOGRAPHIC SPACE */
3984                gotspace = TRUE;                gotspace = TRUE;
3985                break;                break;
3986                }                }
3987              if (gotspace == (ctype == OP_NOT_HSPACE)) break;              if (gotspace == (ctype == OP_NOT_HSPACE)) break;
3988              eptr += len;              eptr += len;
# Line 3883  for (;;) Line 3990  for (;;)
3990            break;            break;
3991    
3992            case OP_NOT_VSPACE:            case OP_NOT_VSPACE:
3993            case OP_VSPACE:            case OP_VSPACE:
3994            for (i = min; i < max; i++)            for (i = min; i < max; i++)
3995              {              {
3996              BOOL gotspace;              BOOL gotspace;
3997              int len = 1;              int len = 1;
3998              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
3999              GETCHARLEN(c, eptr, len);              GETCHARLEN(c, eptr, len);
4000              switch(c)              switch(c)
4001                {                {
4002                default: gotspace = FALSE; break;                default: gotspace = FALSE; break;
4003                case 0x0a:      /* LF */                case 0x0a:      /* LF */
4004                case 0x0b:      /* VT */                case 0x0b:      /* VT */
4005                case 0x0c:      /* FF */                case 0x0c:      /* FF */
# Line 3903  for (;;) Line 4010  for (;;)
4010                gotspace = TRUE;                gotspace = TRUE;
4011                break;                break;
4012                }                }
4013              if (gotspace == (ctype == OP_NOT_VSPACE)) break;              if (gotspace == (ctype == OP_NOT_VSPACE)) break;
4014              eptr += len;              eptr += len;
4015              }              }
4016            break;            break;
# Line 3990  for (;;) Line 4097  for (;;)
4097            }            }
4098          }          }
4099        else        else
4100  #endif  #endif  /* SUPPORT_UTF8 */
4101    
4102        /* Not UTF-8 mode */        /* Not UTF-8 mode */
4103          {          {
4104          switch(ctype)          switch(ctype)
4105            {            {
4106            case OP_ANY:            case OP_ANY:
4107            if ((ims & PCRE_DOTALL) == 0)            for (i = min; i < max; i++)
4108              {              {
4109              for (i = min; i < max; i++)              if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;
4110                {              eptr++;
               if (eptr >= md->end_subject || IS_NEWLINE(eptr)) break;  
               eptr++;  
               }  
             break;  
4111              }              }
4112            /* For DOTALL case, fall through and treat as \C */            break;
4113    
4114              case OP_ALLANY:
4115            case OP_ANYBYTE:            case OP_ANYBYTE:
4116            c = max - min;            c = max - min;
4117            if (c > (unsigned int)(md->end_subject - eptr))            if (c > (unsigned int)(md->end_subject - eptr))
# Line 4027  for (;;) Line 4131  for (;;)
4131                }                }
4132              else              else
4133                {                {
4134                if (c != 0x000a && c != 0x000b && c != 0x000c && c != 0x0085)                if (c != 0x000a &&
4135                      (md->bsr_anycrlf ||
4136                        (c != 0x000b && c != 0x000c && c != 0x0085)))
4137                  break;                  break;
4138                eptr++;                eptr++;
4139                }                }
# Line 4040  for (;;) Line 4146  for (;;)
4146              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
4147              c = *eptr;              c = *eptr;
4148              if (c == 0x09 || c == 0x20 || c == 0xa0) break;              if (c == 0x09 || c == 0x20 || c == 0xa0) break;
4149              eptr++;              eptr++;
4150              }              }
4151            break;            break;
4152    
# Line 4050  for (;;) Line 4156  for (;;)
4156              if (eptr >= md->end_subject) break;              if (eptr >= md->end_subject) break;
4157              c = *eptr;              c = *eptr;
4158              if (c != 0x09 && c != 0x20 && c != 0xa0) break;              if (c != 0x09 && c != 0x20 && c != 0xa0) break;
4159              eptr++;              eptr++;
4160              }              }
4161            break;            break;
4162    
# Line 4061  for (;;) Line 4167  for (;;)
4167              c = *eptr;              c = *eptr;
4168              if (c == 0x0a || c == 0x0b || c == 0x0c || c == 0x0d || c == 0x85)              if (c == 0x0a || c == 0x0b || c == 0x0c || c == 0x0d || c == 0x85)
4169                break;                break;
4170              eptr++;              eptr++;
4171              }              }
4172            break;            break;
4173    
# Line 4177  HEAP_RETURN: Line 4283  HEAP_RETURN:
4283  switch (frame->Xwhere)  switch (frame->Xwhere)
4284    {    {
4285    LBL( 1) LBL( 2) LBL( 3) LBL( 4) LBL( 5) LBL( 6) LBL( 7) LBL( 8)    LBL( 1) LBL( 2) LBL( 3) LBL( 4) LBL( 5) LBL( 6) LBL( 7) LBL( 8)
4286    LBL( 9) LBL(10) LBL(11) LBL(12) LBL(13) LBL(14) LBL(15) LBL(16)    LBL( 9) LBL(10) LBL(11) LBL(12) LBL(13) LBL(14) LBL(15) LBL(17)
4287    LBL(17) LBL(18) LBL(19) LBL(20) LBL(21) LBL(22) LBL(23) LBL(24)    LBL(19) LBL(24) LBL(25) LBL(26) LBL(27) LBL(29) LBL(31) LBL(33)
4288    LBL(25) LBL(26) LBL(27) LBL(28) LBL(29) LBL(30) LBL(31) LBL(32)    LBL(35) LBL(43) LBL(47) LBL(48) LBL(49) LBL(50) LBL(51) LBL(52)
4289    LBL(33) LBL(34) LBL(35) LBL(36) LBL(37) LBL(38) LBL(39) LBL(40)    LBL(53) LBL(54)
4290    LBL(41) LBL(42) LBL(43) LBL(44) LBL(45) LBL(46) LBL(47)  #ifdef SUPPORT_UTF8
4291      LBL(16) LBL(18) LBL(20) LBL(21) LBL(22) LBL(23) LBL(28) LBL(30)
4292      LBL(32) LBL(34) LBL(42) LBL(46)
4293    #ifdef SUPPORT_UCP
4294      LBL(36) LBL(37) LBL(38) LBL(39) LBL(40) LBL(41) LBL(44) LBL(45)
4295    #endif  /* SUPPORT_UCP */
4296    #endif  /* SUPPORT_UTF8 */
4297    default:    default:
4298    DPRINTF(("jump error in pcre match: label %d non-existent\n", frame->Xwhere));    DPRINTF(("jump error in pcre match: label %d non-existent\n", frame->Xwhere));
4299    return PCRE_ERROR_INTERNAL;    return PCRE_ERROR_INTERNAL;
# Line 4273  Returns:          > 0 => success; value Line 4385  Returns:          > 0 => success; value
4385                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
4386  */  */
4387    
4388  PCRE_EXP_DEFN int  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
4389  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_exec(const pcre *argument_re, const pcre_extra *extra_data,
4390    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,    PCRE_SPTR subject, int length, int start_offset, int options, int *offsets,
4391    int offsetcount)    int offsetcount)
# Line 4298  const uschar *start_bits = NULL; Line 4410  const uschar *start_bits = NULL;
4410  USPTR start_match = (USPTR)subject + start_offset;  USPTR start_match = (USPTR)subject + start_offset;
4411  USPTR end_subject;  USPTR end_subject;
4412  USPTR req_byte_ptr = start_match - 1;  USPTR req_byte_ptr = start_match - 1;
 eptrblock eptrchain[EPTR_WORK_SIZE];  
4413    
4414  pcre_study_data internal_study;  pcre_study_data internal_study;
4415  const pcre_study_data *study;  const pcre_study_data *study;
# Line 4361  if (re->magic_number != MAGIC_NUMBER) Line 4472  if (re->magic_number != MAGIC_NUMBER)
4472  /* Set up other data */  /* Set up other data */
4473    
4474  anchored = ((re->options | options) & PCRE_ANCHORED) != 0;  anchored = ((re->options | options) & PCRE_ANCHORED) != 0;
4475  startline = (re->options & PCRE_STARTLINE) != 0;  startline = (re->flags & PCRE_STARTLINE) != 0;
4476  firstline = (re->options & PCRE_FIRSTLINE) != 0;  firstline = (re->options & PCRE_FIRSTLINE) != 0;
4477    
4478  /* The code starts after the real_pcre block and the capture name table. */  /* The code starts after the real_pcre block and the capture name table. */
# Line 4376  end_subject = md->end_subject; Line 4487  end_subject = md->end_subject;
4487    
4488  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;  md->endonly = (re->options & PCRE_DOLLAR_ENDONLY) != 0;
4489  utf8 = md->utf8 = (re->options & PCRE_UTF8) != 0;  utf8 = md->utf8 = (re->options & PCRE_UTF8) != 0;
4490    md->jscript_compat = (re->options & PCRE_JAVASCRIPT_COMPAT) != 0;
4491    
4492  md->notbol = (options & PCRE_NOTBOL) != 0;  md->notbol = (options & PCRE_NOTBOL) != 0;
4493  md->noteol = (options & PCRE_NOTEOL) != 0;  md->noteol = (options & PCRE_NOTEOL) != 0;
# Line 4384  md->partial = (options & PCRE_PARTIAL) ! Line 4496  md->partial = (options & PCRE_PARTIAL) !
4496  md->hitend = FALSE;  md->hitend = FALSE;
4497    
4498  md->recursive = NULL;                   /* No recursion at top level */  md->recursive = NULL;                   /* No recursion at top level */
 md->eptrchain = eptrchain;              /* Make workspace generally available */  
4499    
4500  md->lcc = tables + lcc_offset;  md->lcc = tables + lcc_offset;
4501  md->ctypes = tables + ctypes_offset;  md->ctypes = tables + ctypes_offset;
4502    
4503    /* Handle different \R options. */
4504    
4505    switch (options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE))
4506      {
4507      case 0:
4508      if ((re->options & (PCRE_BSR_ANYCRLF|PCRE_BSR_UNICODE)) != 0)
4509        md->bsr_anycrlf = (re->options & PCRE_BSR_ANYCRLF) != 0;
4510      else
4511    #ifdef BSR_ANYCRLF
4512      md->bsr_anycrlf = TRUE;
4513    #else
4514      md->bsr_anycrlf = FALSE;
4515    #endif
4516      break;
4517    
4518      case PCRE_BSR_ANYCRLF:
4519      md->bsr_anycrlf = TRUE;
4520      break;
4521    
4522      case PCRE_BSR_UNICODE:
4523      md->bsr_anycrlf = FALSE;
4524      break;
4525    
4526      default: return PCRE_ERROR_BADNEWLINE;
4527      }
4528    
4529  /* Handle different types of newline. The three bits give eight cases. If  /* Handle different types of newline. The three bits give eight cases. If
4530  nothing is set at run time, whatever was used at compile time applies. */  nothing is set at run time, whatever was used at compile time applies. */
4531    
4532  switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options : (pcre_uint32)options) &  switch ((((options & PCRE_NEWLINE_BITS) == 0)? re->options :
4533         PCRE_NEWLINE_BITS)          (pcre_uint32)options) & PCRE_NEWLINE_BITS)
4534    {    {
4535    case 0: newline = NEWLINE; break;   /* Compile-time default */    case 0: newline = NEWLINE; break;   /* Compile-time default */
4536    case PCRE_NEWLINE_CR: newline = '\r'; break;    case PCRE_NEWLINE_CR: newline = '\r'; break;
# Line 4432  else Line 4569  else
4569  /* Partial matching is supported only for a restricted set of regexes at the  /* Partial matching is supported only for a restricted set of regexes at the
4570  moment. */  moment. */
4571    
4572  if (md->partial && (re->options & PCRE_NOPARTIAL) != 0)  if (md->partial && (re->flags & PCRE_NOPARTIAL) != 0)
4573    return PCRE_ERROR_BADPARTIAL;    return PCRE_ERROR_BADPARTIAL;
4574    
4575  /* Check a UTF-8 string if required. Unfortunately there's no way of passing  /* Check a UTF-8 string if required. Unfortunately there's no way of passing
# Line 4509  studied, there may be a bitmap of possib Line 4646  studied, there may be a bitmap of possib
4646    
4647  if (!anchored)  if (!anchored)
4648    {    {
4649    if ((re->options & PCRE_FIRSTSET) != 0)    if ((re->flags & PCRE_FIRSTSET) != 0)
4650      {      {
4651      first_byte = re->first_byte & 255;      first_byte = re->first_byte & 255;
4652      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)      if ((first_byte_caseless = ((re->first_byte & REQ_CASELESS) != 0)) == TRUE)
# Line 4524  if (!anchored) Line 4661  if (!anchored)
4661  /* For anchored or unanchored matches, there may be a "last known required  /* For anchored or unanchored matches, there may be a "last known required
4662  character" set. */  character" set. */
4663    
4664  if ((re->options & PCRE_REQCHSET) != 0)  if ((re->flags & PCRE_REQCHSET) != 0)
4665    {    {
4666    req_byte = re->req_byte & 255;    req_byte = re->req_byte & 255;
4667    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;    req_byte_caseless = (re->req_byte & REQ_CASELESS) != 0;
# Line 4540  the loop runs just once. */ Line 4677  the loop runs just once. */
4677  for(;;)  for(;;)
4678    {    {
4679    USPTR save_end_subject = end_subject;    USPTR save_end_subject = end_subject;
4680      USPTR new_start_match;
4681    
4682    /* Reset the maximum number of extractions we might see. */    /* Reset the maximum number of extractions we might see. */
4683    
# Line 4559  for(;;) Line 4697  for(;;)
4697    if (firstline)    if (firstline)
4698      {      {
4699      USPTR t = start_match;      USPTR t = start_match;
4700    #ifdef SUPPORT_UTF8
4701        if (utf8)
4702          {
4703          while (t < md->end_subject && !IS_NEWLINE(t))
4704            {
4705            t++;
4706            while (t < end_subject && (*t & 0xc0) == 0x80) t++;
4707            }
4708          }
4709        else
4710    #endif
4711      while (t < md->end_subject && !IS_NEWLINE(t)) t++;      while (t < md->end_subject && !IS_NEWLINE(t)) t++;
4712      end_subject = t;      end_subject = t;
4713      }      }
4714    
4715    /* Now test for a unique first byte */    /* Now advance to a unique first byte if there is one. */
4716    
4717    if (first_byte >= 0)    if (first_byte >= 0)
4718      {      {
4719      if (first_byte_caseless)      if (first_byte_caseless)
4720        while (start_match < end_subject &&        while (start_match < end_subject && md->lcc[*start_match] != first_byte)
              md->lcc[*start_match] != first_byte)  
4721          start_match++;          start_match++;
4722      else      else
4723        while (start_match < end_subject && *start_match != first_byte)        while (start_match < end_subject && *start_match != first_byte)
4724          start_match++;          start_match++;
4725      }      }
4726    
4727    /* Or to just after a linebreak for a multiline match if possible */    /* Or to just after a linebreak for a multiline match */
4728    
4729    else if (startline)    else if (startline)
4730      {      {
4731      if (start_match > md->start_subject + start_offset)      if (start_match > md->start_subject + start_offset)
4732        {        {
4733        while (start_match <= end_subject && !WAS_NEWLINE(start_match))  #ifdef SUPPORT_UTF8
4734          if (utf8)
4735            {
4736            while (start_match < end_subject && !WAS_NEWLINE(start_match))
4737              {
4738              start_match++;
4739              while(start_match < end_subject && (*start_match & 0xc0) == 0x80)
4740                start_match++;
4741              }
4742            }
4743          else
4744    #endif
4745          while (start_match < end_subject && !WAS_NEWLINE(start_match))
4746          start_match++;          start_match++;
4747    
4748        /* If we have just passed a CR and the newline option is ANY or ANYCRLF,        /* If we have just passed a CR and the newline option is ANY or ANYCRLF,
4749        and we are now at a LF, advance the match position by one more character.        and we are now at a LF, advance the match position by one more character.
4750        */        */
# Line 4597  for(;;) Line 4757  for(;;)
4757        }        }
4758      }      }
4759    
4760    /* Or to a non-unique first char after study */    /* Or to a non-unique first byte after study */
4761    
4762    else if (start_bits != NULL)    else if (start_bits != NULL)
4763      {      {
4764      while (start_match < end_subject)      while (start_match < end_subject)
4765        {        {
4766        register unsigned int c = *start_match;        register unsigned int c = *start_match;
4767        if ((start_bits[c/8] & (1 << (c&7))) == 0) start_match++; else break;        if ((start_bits[c/8] & (1 << (c&7))) == 0) start_match++;
4768            else break;
4769        }        }
4770      }      }
4771    
# Line 4679  for(;;) Line 4840  for(;;)
4840      }      }
4841    
4842    /* OK, we can now run the match. */    /* OK, we can now run the match. */
4843    
4844    md->start_match_ptr = start_match;      /* Insurance */    md->start_match_ptr = start_match;
4845    md->match_call_count = 0;    md->match_call_count = 0;
4846    md->eptrn = 0;                          /* Next free eptrchain slot */    rc = match(start_match, md->start_code, start_match, 2, md, ims, NULL, 0, 0);
   rc = match(start_match, md->start_code, start_match, 2, md,  
     ims, NULL, 0, 0);  
4847    
4848    /* Any return other than MATCH_NOMATCH breaks the loop. */    switch(rc)
4849        {
4850        /* NOMATCH and PRUNE advance by one character. THEN at this level acts
4851        exactly like PRUNE. */
4852    
4853    if (rc != MATCH_NOMATCH) break;      case MATCH_NOMATCH:
4854        case MATCH_PRUNE:
4855        case MATCH_THEN:
4856        new_start_match = start_match + 1;
4857    #ifdef SUPPORT_UTF8
4858        if (utf8)
4859          while(new_start_match < end_subject && (*new_start_match & 0xc0) == 0x80)
4860            new_start_match++;
4861    #endif
4862        break;
4863    
4864        /* SKIP passes back the next starting point explicitly. */
4865    
4866        case MATCH_SKIP:
4867        new_start_match = md->start_match_ptr;
4868        break;
4869    
4870        /* COMMIT disables the bumpalong, but otherwise behaves as NOMATCH. */
4871    
4872        case MATCH_COMMIT:
4873        rc = MATCH_NOMATCH;
4874        goto ENDLOOP;
4875    
4876        /* Any other return is some kind of error. */
4877    
4878        default:
4879        goto ENDLOOP;
4880        }
4881    
4882      /* Control reaches here for the various types of "no match at this point"
4883      result. Reset the code to MATCH_NOMATCH for subsequent checking. */
4884    
4885      rc = MATCH_NOMATCH;
4886    
4887    /* If PCRE_FIRSTLINE is set, the match must happen before or at the first    /* If PCRE_FIRSTLINE is set, the match must happen before or at the first
4888    newline in the subject (though it may continue over the newline). Therefore,    newline in the subject (though it may continue over the newline). Therefore,
# Line 4696  for(;;) Line 4890  for(;;)
4890    
4891    if (firstline && IS_NEWLINE(start_match)) break;    if (firstline && IS_NEWLINE(start_match)) break;
4892    
4893    /* Advance the match position by one character. */    /* Advance to new matching position */
4894    
4895    start_match++;    start_match = new_start_match;
 #ifdef SUPPORT_UTF8  
   if (utf8)  
     while(start_match < end_subject && (*start_match & 0xc0) == 0x80)  
       start_match++;  
 #endif  
4896    
4897    /* Break the loop if the pattern is anchored or if we have passed the end of    /* Break the loop if the pattern is anchored or if we have passed the end of
4898    the subject. */    the subject. */
4899    
4900    if (anchored || start_match > end_subject) break;    if (anchored || start_match > end_subject) break;
4901    
4902    /* If we have just passed a CR and the newline option is CRLF or ANY or    /* If we have just passed a CR and we are now at a LF, and the pattern does
4903    ANYCRLF, and we are now at a LF, advance the match position by one more    not contain any explicit matches for \r or \n, and the newline option is CRLF
4904    character. */    or ANY or ANYCRLF, advance the match position by one more character. */
4905    
4906    if (start_match[-1] == '\r' &&    if (start_match[-1] == '\r' &&
4907         (md->nltype == NLTYPE_ANY ||        start_match < end_subject &&
4908          md->nltype == NLTYPE_ANYCRLF ||        *start_match == '\n' &&
4909          md->nllen == 2) &&        (re->flags & PCRE_HASCRORLF) == 0 &&
4910         start_match < end_subject &&          (md->nltype == NLTYPE_ANY ||
4911         *start_match == '\n')           md->nltype == NLTYPE_ANYCRLF ||
4912             md->nllen == 2))
4913      start_match++;      start_match++;
4914    
4915    }   /* End of for(;;) "bumpalong" loop */    }   /* End of for(;;) "bumpalong" loop */
# Line 4729  for(;;) Line 4919  for(;;)
4919  /* We reach here when rc is not MATCH_NOMATCH, or if one of the stopping  /* We reach here when rc is not MATCH_NOMATCH, or if one of the stopping
4920  conditions is true:  conditions is true:
4921    
4922  (1) The pattern is anchored;  (1) The pattern is anchored or the match was failed by (*COMMIT);
4923    
4924  (2) We are past the end of the subject;  (2) We are past the end of the subject;
4925    
# Line 4744  processing, copy those that we can. In t Line 4934  processing, copy those that we can. In t
4934  certain parts of the pattern were not used, even though there are more  certain parts of the pattern were not used, even though there are more
4935  capturing parentheses than vector slots. */  capturing parentheses than vector slots. */
4936    
4937    ENDLOOP:
4938    
4939  if (rc == MATCH_MATCH)  if (rc == MATCH_MATCH)
4940    {    {
4941    if (using_temporary_offsets)    if (using_temporary_offsets)

Legend:
Removed from v.178  
changed lines
  Added in v.366

  ViewVC Help
Powered by ViewVC 1.1.5