/[pcre]/code/trunk/pcre_dfa_exec.c
ViewVC logotype

Diff of /code/trunk/pcre_dfa_exec.c

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

revision 345 by ph10, Mon Apr 28 15:10:02 2008 UTC revision 406 by ph10, Mon Mar 23 12:05:43 2009 UTC
# Line 3  Line 3 
3  *************************************************/  *************************************************/
4    
5  /* PCRE is a library of functions to support regular expressions whose syntax  /* PCRE is a library of functions to support regular expressions whose syntax
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 (but see
7    below for why this module is different).
8    
9                         Written by Philip Hazel                         Written by Philip Hazel
10             Copyright (c) 1997-2008 University of Cambridge             Copyright (c) 1997-2009 University of Cambridge
11    
12  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
13  Redistribution and use in source and binary forms, with or without  Redistribution and use in source and binary forms, with or without
# Line 60  applications. */ Line 61  applications. */
61  #define SP "                   "  #define SP "                   "
62    
63    
   
64  /*************************************************  /*************************************************
65  *      Code parameters and static tables         *  *      Code parameters and static tables         *
66  *************************************************/  *************************************************/
# Line 511  for (;;) Line 511  for (;;)
511      stateblock *current_state = active_states + i;      stateblock *current_state = active_states + i;
512      const uschar *code;      const uschar *code;
513      int state_offset = current_state->offset;      int state_offset = current_state->offset;
514      int count, codevalue;      int count, codevalue, rrc;
 #ifdef SUPPORT_UCP  
     int chartype, script;  
 #endif  
515    
516  #ifdef DEBUG  #ifdef DEBUG
517      printf ("%.*sProcessing state %d c=", rlevel*2-2, SP, state_offset);      printf ("%.*sProcessing state %d c=", rlevel*2-2, SP, state_offset);
# Line 760  for (;;) Line 757  for (;;)
757        if ((md->moptions & PCRE_NOTEOL) == 0)        if ((md->moptions & PCRE_NOTEOL) == 0)
758          {          {
759          if (clen == 0 ||          if (clen == 0 ||
760              (IS_NEWLINE(ptr) &&              ((md->poptions & PCRE_DOLLAR_ENDONLY) == 0 && IS_NEWLINE(ptr) &&
761                 ((ims & PCRE_MULTILINE) != 0 || ptr == end_subject - md->nllen)                 ((ims & PCRE_MULTILINE) != 0 || ptr == end_subject - md->nllen)
762              ))              ))
763            { ADD_ACTIVE(state_offset + 1, 0); }            { ADD_ACTIVE(state_offset + 1, 0); }
# Line 825  for (;;) Line 822  for (;;)
822        if (clen > 0)        if (clen > 0)
823          {          {
824          BOOL OK;          BOOL OK;
825          int category = _pcre_ucp_findprop(c, &chartype, &script);          const ucd_record * prop = GET_UCD(c);
826          switch(code[1])          switch(code[1])
827            {            {
828            case PT_ANY:            case PT_ANY:
# Line 833  for (;;) Line 830  for (;;)
830            break;            break;
831    
832            case PT_LAMP:            case PT_LAMP:
833            OK = chartype == ucp_Lu || chartype == ucp_Ll || chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;
834            break;            break;
835    
836            case PT_GC:            case PT_GC:
837            OK = category == code[2];            OK = _pcre_ucp_gentype[prop->chartype] == code[2];
838            break;            break;
839    
840            case PT_PC:            case PT_PC:
841            OK = chartype == code[2];            OK = prop->chartype == code[2];
842            break;            break;
843    
844            case PT_SC:            case PT_SC:
845            OK = script == code[2];            OK = prop->script == code[2];
846            break;            break;
847    
848            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
# Line 994  for (;;) Line 991  for (;;)
991        if (clen > 0)        if (clen > 0)
992          {          {
993          BOOL OK;          BOOL OK;
994          int category = _pcre_ucp_findprop(c, &chartype, &script);          const ucd_record * prop = GET_UCD(c);
995          switch(code[2])          switch(code[2])
996            {            {
997            case PT_ANY:            case PT_ANY:
# Line 1002  for (;;) Line 999  for (;;)
999            break;            break;
1000    
1001            case PT_LAMP:            case PT_LAMP:
1002            OK = chartype == ucp_Lu || chartype == ucp_Ll || chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;
1003            break;            break;
1004    
1005            case PT_GC:            case PT_GC:
1006            OK = category == code[3];            OK = _pcre_ucp_gentype[prop->chartype] == code[3];
1007            break;            break;
1008    
1009            case PT_PC:            case PT_PC:
1010            OK = chartype == code[3];            OK = prop->chartype == code[3];
1011            break;            break;
1012    
1013            case PT_SC:            case PT_SC:
1014            OK = script == code[3];            OK = prop->script == code[3];
1015            break;            break;
1016    
1017            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
# Line 1043  for (;;) Line 1040  for (;;)
1040        case OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS:        case OP_EXTUNI_EXTRA + OP_TYPEPOSPLUS:
1041        count = current_state->count;  /* Already matched */        count = current_state->count;  /* Already matched */
1042        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }        if (count > 0) { ADD_ACTIVE(state_offset + 2, 0); }
1043        if (clen > 0 && _pcre_ucp_findprop(c, &chartype, &script) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1044          {          {
1045          const uschar *nptr = ptr + clen;          const uschar *nptr = ptr + clen;
1046          int ncount = 0;          int ncount = 0;
# Line 1057  for (;;) Line 1054  for (;;)
1054            int nd;            int nd;
1055            int ndlen = 1;            int ndlen = 1;
1056            GETCHARLEN(nd, nptr, ndlen);            GETCHARLEN(nd, nptr, ndlen);
1057            if (_pcre_ucp_findprop(nd, &chartype, &script) != ucp_M) break;            if (UCD_CATEGORY(nd) != ucp_M) break;
1058            ncount++;            ncount++;
1059            nptr += ndlen;            nptr += ndlen;
1060            }            }
# Line 1216  for (;;) Line 1213  for (;;)
1213        if (clen > 0)        if (clen > 0)
1214          {          {
1215          BOOL OK;          BOOL OK;
1216          int category = _pcre_ucp_findprop(c, &chartype, &script);          const ucd_record * prop = GET_UCD(c);
1217          switch(code[2])          switch(code[2])
1218            {            {
1219            case PT_ANY:            case PT_ANY:
# Line 1224  for (;;) Line 1221  for (;;)
1221            break;            break;
1222    
1223            case PT_LAMP:            case PT_LAMP:
1224            OK = chartype == ucp_Lu || chartype == ucp_Ll || chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;
1225            break;            break;
1226    
1227            case PT_GC:            case PT_GC:
1228            OK = category == code[3];            OK = _pcre_ucp_gentype[prop->chartype] == code[3];
1229            break;            break;
1230    
1231            case PT_PC:            case PT_PC:
1232            OK = chartype == code[3];            OK = prop->chartype == code[3];
1233            break;            break;
1234    
1235            case PT_SC:            case PT_SC:
1236            OK = script == code[3];            OK = prop->script == code[3];
1237            break;            break;
1238    
1239            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
# Line 1274  for (;;) Line 1271  for (;;)
1271        QS2:        QS2:
1272    
1273        ADD_ACTIVE(state_offset + 2, 0);        ADD_ACTIVE(state_offset + 2, 0);
1274        if (clen > 0 && _pcre_ucp_findprop(c, &chartype, &script) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1275          {          {
1276          const uschar *nptr = ptr + clen;          const uschar *nptr = ptr + clen;
1277          int ncount = 0;          int ncount = 0;
# Line 1289  for (;;) Line 1286  for (;;)
1286            int nd;            int nd;
1287            int ndlen = 1;            int ndlen = 1;
1288            GETCHARLEN(nd, nptr, ndlen);            GETCHARLEN(nd, nptr, ndlen);
1289            if (_pcre_ucp_findprop(nd, &chartype, &script) != ucp_M) break;            if (UCD_CATEGORY(nd) != ucp_M) break;
1290            ncount++;            ncount++;
1291            nptr += ndlen;            nptr += ndlen;
1292            }            }
# Line 1463  for (;;) Line 1460  for (;;)
1460        if (clen > 0)        if (clen > 0)
1461          {          {
1462          BOOL OK;          BOOL OK;
1463          int category = _pcre_ucp_findprop(c, &chartype, &script);          const ucd_record * prop = GET_UCD(c);
1464          switch(code[4])          switch(code[4])
1465            {            {
1466            case PT_ANY:            case PT_ANY:
# Line 1471  for (;;) Line 1468  for (;;)
1468            break;            break;
1469    
1470            case PT_LAMP:            case PT_LAMP:
1471            OK = chartype == ucp_Lu || chartype == ucp_Ll || chartype == ucp_Lt;            OK = prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || prop->chartype == ucp_Lt;
1472            break;            break;
1473    
1474            case PT_GC:            case PT_GC:
1475            OK = category == code[5];            OK = _pcre_ucp_gentype[prop->chartype] == code[5];
1476            break;            break;
1477    
1478            case PT_PC:            case PT_PC:
1479            OK = chartype == code[5];            OK = prop->chartype == code[5];
1480            break;            break;
1481    
1482            case PT_SC:            case PT_SC:
1483            OK = script == code[5];            OK = prop->script == code[5];
1484            break;            break;
1485    
1486            /* Should never occur, but keep compilers from grumbling. */            /* Should never occur, but keep compilers from grumbling. */
# Line 1516  for (;;) Line 1513  for (;;)
1513        if (codevalue != OP_EXTUNI_EXTRA + OP_TYPEEXACT)        if (codevalue != OP_EXTUNI_EXTRA + OP_TYPEEXACT)
1514          { ADD_ACTIVE(state_offset + 4, 0); }          { ADD_ACTIVE(state_offset + 4, 0); }
1515        count = current_state->count;  /* Number already matched */        count = current_state->count;  /* Number already matched */
1516        if (clen > 0 && _pcre_ucp_findprop(c, &chartype, &script) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1517          {          {
1518          const uschar *nptr = ptr + clen;          const uschar *nptr = ptr + clen;
1519          int ncount = 0;          int ncount = 0;
# Line 1530  for (;;) Line 1527  for (;;)
1527            int nd;            int nd;
1528            int ndlen = 1;            int ndlen = 1;
1529            GETCHARLEN(nd, nptr, ndlen);            GETCHARLEN(nd, nptr, ndlen);
1530            if (_pcre_ucp_findprop(nd, &chartype, &script) != ucp_M) break;            if (UCD_CATEGORY(nd) != ucp_M) break;
1531            ncount++;            ncount++;
1532            nptr += ndlen;            nptr += ndlen;
1533            }            }
# Line 1710  for (;;) Line 1707  for (;;)
1707            other case of the character. */            other case of the character. */
1708    
1709  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
1710            othercase = _pcre_ucp_othercase(c);            othercase = UCD_OTHERCASE(c);
1711  #else  #else
1712            othercase = NOTACHAR;            othercase = NOTACHAR;
1713  #endif  #endif
# Line 1735  for (;;) Line 1732  for (;;)
1732        to wait for them to pass before continuing. */        to wait for them to pass before continuing. */
1733    
1734        case OP_EXTUNI:        case OP_EXTUNI:
1735        if (clen > 0 && _pcre_ucp_findprop(c, &chartype, &script) != ucp_M)        if (clen > 0 && UCD_CATEGORY(c) != ucp_M)
1736          {          {
1737          const uschar *nptr = ptr + clen;          const uschar *nptr = ptr + clen;
1738          int ncount = 0;          int ncount = 0;
# Line 1743  for (;;) Line 1740  for (;;)
1740            {            {
1741            int nclen = 1;            int nclen = 1;
1742            GETCHARLEN(c, nptr, nclen);            GETCHARLEN(c, nptr, nclen);
1743            if (_pcre_ucp_findprop(c, &chartype, &script) != ucp_M) break;            if (UCD_CATEGORY(c) != ucp_M) break;
1744            ncount++;            ncount++;
1745            nptr += nclen;            nptr += nclen;
1746            }            }
# Line 1911  for (;;) Line 1908  for (;;)
1908            if (utf8 && d >= 128)            if (utf8 && d >= 128)
1909              {              {
1910  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
1911              otherd = _pcre_ucp_othercase(d);              otherd = UCD_OTHERCASE(d);
1912  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
1913              }              }
1914            else            else
# Line 1949  for (;;) Line 1946  for (;;)
1946            if (utf8 && d >= 128)            if (utf8 && d >= 128)
1947              {              {
1948  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
1949              otherd = _pcre_ucp_othercase(d);              otherd = UCD_OTHERCASE(d);
1950  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
1951              }              }
1952            else            else
# Line 1985  for (;;) Line 1982  for (;;)
1982            if (utf8 && d >= 128)            if (utf8 && d >= 128)
1983              {              {
1984  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
1985              otherd = _pcre_ucp_othercase(d);              otherd = UCD_OTHERCASE(d);
1986  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
1987              }              }
1988            else            else
# Line 2017  for (;;) Line 2014  for (;;)
2014            if (utf8 && d >= 128)            if (utf8 && d >= 128)
2015              {              {
2016  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2017              otherd = _pcre_ucp_othercase(d);              otherd = UCD_OTHERCASE(d);
2018  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2019              }              }
2020            else            else
# Line 2052  for (;;) Line 2049  for (;;)
2049            if (utf8 && d >= 128)            if (utf8 && d >= 128)
2050              {              {
2051  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
2052              otherd = _pcre_ucp_othercase(d);              otherd = UCD_OTHERCASE(d);
2053  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
2054              }              }
2055            else            else
# Line 2203  for (;;) Line 2200  for (;;)
2200          {          {
2201          int local_offsets[1000];          int local_offsets[1000];
2202          int local_workspace[1000];          int local_workspace[1000];
2203          int condcode = code[LINK_SIZE+1];          int codelink = GET(code, 1);
2204            int condcode;
2205    
2206            /* Because of the way auto-callout works during compile, a callout item
2207            is inserted between OP_COND and an assertion condition. This does not
2208            happen for the other conditions. */
2209    
2210            if (code[LINK_SIZE+1] == OP_CALLOUT)
2211              {
2212              rrc = 0;
2213              if (pcre_callout != NULL)
2214                {
2215                pcre_callout_block cb;
2216                cb.version          = 1;   /* Version 1 of the callout block */
2217                cb.callout_number   = code[LINK_SIZE+2];
2218                cb.offset_vector    = offsets;
2219                cb.subject          = (PCRE_SPTR)start_subject;
2220                cb.subject_length   = end_subject - start_subject;
2221                cb.start_match      = current_subject - start_subject;
2222                cb.current_position = ptr - start_subject;
2223                cb.pattern_position = GET(code, LINK_SIZE + 3);
2224                cb.next_item_length = GET(code, 3 + 2*LINK_SIZE);
2225                cb.capture_top      = 1;
2226                cb.capture_last     = -1;
2227                cb.callout_data     = md->callout_data;
2228                if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc;   /* Abandon */
2229                }
2230              if (rrc > 0) break;                      /* Fail this thread */
2231              code += _pcre_OP_lengths[OP_CALLOUT];    /* Skip callout data */
2232              }
2233    
2234            condcode = code[LINK_SIZE+1];
2235    
2236          /* Back reference conditions are not supported */          /* Back reference conditions are not supported */
2237    
# Line 2212  for (;;) Line 2240  for (;;)
2240          /* The DEFINE condition is always false */          /* The DEFINE condition is always false */
2241    
2242          if (condcode == OP_DEF)          if (condcode == OP_DEF)
2243            {            { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
           ADD_ACTIVE(state_offset + GET(code, 1) + LINK_SIZE + 1, 0);  
           }  
2244    
2245          /* The only supported version of OP_RREF is for the value RREF_ANY,          /* The only supported version of OP_RREF is for the value RREF_ANY,
2246          which means "test if in any recursion". We can't test for specifically          which means "test if in any recursion". We can't test for specifically
# Line 2224  for (;;) Line 2250  for (;;)
2250            {            {
2251            int value = GET2(code, LINK_SIZE+2);            int value = GET2(code, LINK_SIZE+2);
2252            if (value != RREF_ANY) return PCRE_ERROR_DFA_UCOND;            if (value != RREF_ANY) return PCRE_ERROR_DFA_UCOND;
2253            if (recursing > 0) { ADD_ACTIVE(state_offset + LINK_SIZE + 4, 0); }            if (recursing > 0)
2254              else { ADD_ACTIVE(state_offset + GET(code, 1) + LINK_SIZE + 1, 0); }              { ADD_ACTIVE(state_offset + LINK_SIZE + 4, 0); }
2255              else { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
2256            }            }
2257    
2258          /* Otherwise, the condition is an assertion */          /* Otherwise, the condition is an assertion */
# Line 2255  for (;;) Line 2282  for (;;)
2282                  (condcode == OP_ASSERT || condcode == OP_ASSERTBACK))                  (condcode == OP_ASSERT || condcode == OP_ASSERTBACK))
2283              { ADD_ACTIVE(endasscode + LINK_SIZE + 1 - start_code, 0); }              { ADD_ACTIVE(endasscode + LINK_SIZE + 1 - start_code, 0); }
2284            else            else
2285              { ADD_ACTIVE(state_offset + GET(code, 1) + LINK_SIZE + 1, 0); }              { ADD_ACTIVE(state_offset + codelink + LINK_SIZE + 1, 0); }
2286            }            }
2287          }          }
2288        break;        break;
# Line 2407  for (;;) Line 2434  for (;;)
2434        /* Handle callouts */        /* Handle callouts */
2435    
2436        case OP_CALLOUT:        case OP_CALLOUT:
2437          rrc = 0;
2438        if (pcre_callout != NULL)        if (pcre_callout != NULL)
2439          {          {
         int rrc;  
2440          pcre_callout_block cb;          pcre_callout_block cb;
2441          cb.version          = 1;   /* Version 1 of the callout block */          cb.version          = 1;   /* Version 1 of the callout block */
2442          cb.callout_number   = code[1];          cb.callout_number   = code[1];
# Line 2424  for (;;) Line 2451  for (;;)
2451          cb.capture_last     = -1;          cb.capture_last     = -1;
2452          cb.callout_data     = md->callout_data;          cb.callout_data     = md->callout_data;
2453          if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc;   /* Abandon */          if ((rrc = (*pcre_callout)(&cb)) < 0) return rrc;   /* Abandon */
         if (rrc == 0) { ADD_ACTIVE(state_offset + 2 + 2*LINK_SIZE, 0); }  
2454          }          }
2455          if (rrc == 0)
2456            { ADD_ACTIVE(state_offset + _pcre_OP_lengths[OP_CALLOUT], 0); }
2457        break;        break;
2458    
2459    
# Line 2508  Returns: > 0 => number of match Line 2536  Returns: > 0 => number of match
2536                   < -1 => some kind of unexpected problem                   < -1 => some kind of unexpected problem
2537  */  */
2538    
2539  PCRE_EXP_DEFN int  PCRE_EXP_DEFN int PCRE_CALL_CONVENTION
2540  pcre_dfa_exec(const pcre *argument_re, const pcre_extra *extra_data,  pcre_dfa_exec(const pcre *argument_re, const pcre_extra *extra_data,
2541    const char *subject, int length, int start_offset, int options, int *offsets,    const char *subject, int length, int start_offset, int options, int *offsets,
2542    int offsetcount, int *workspace, int wscount)    int offsetcount, int *workspace, int wscount)
# Line 2617  switch ((((options & PCRE_NEWLINE_BITS) Line 2645  switch ((((options & PCRE_NEWLINE_BITS)
2645           PCRE_NEWLINE_BITS)           PCRE_NEWLINE_BITS)
2646    {    {
2647    case 0: newline = NEWLINE; break;   /* Compile-time default */    case 0: newline = NEWLINE; break;   /* Compile-time default */
2648    case PCRE_NEWLINE_CR: newline = '\r'; break;    case PCRE_NEWLINE_CR: newline = CHAR_CR; break;
2649    case PCRE_NEWLINE_LF: newline = '\n'; break;    case PCRE_NEWLINE_LF: newline = CHAR_NL; break;
2650    case PCRE_NEWLINE_CR+    case PCRE_NEWLINE_CR+
2651         PCRE_NEWLINE_LF: newline = ('\r' << 8) | '\n'; break;         PCRE_NEWLINE_LF: newline = (CHAR_CR << 8) | CHAR_NL; break;
2652    case PCRE_NEWLINE_ANY: newline = -1; break;    case PCRE_NEWLINE_ANY: newline = -1; break;
2653    case PCRE_NEWLINE_ANYCRLF: newline = -2; break;    case PCRE_NEWLINE_ANYCRLF: newline = -2; break;
2654    default: return PCRE_ERROR_BADNEWLINE;    default: return PCRE_ERROR_BADNEWLINE;
# Line 2716  if ((re->flags & PCRE_REQCHSET) != 0) Line 2744  if ((re->flags & PCRE_REQCHSET) != 0)
2744    }    }
2745    
2746  /* Call the main matching function, looping for a non-anchored regex after a  /* Call the main matching function, looping for a non-anchored regex after a
2747  failed match. Unless restarting, optimize by moving to the first match  failed match. If not restarting, perform certain optimizations at the start of
2748  character if possible, when not anchored. Then unless wanting a partial match,  a match. */
 check for a required later character. */  
2749    
2750  for (;;)  for (;;)
2751    {    {
# Line 2728  for (;;) Line 2755  for (;;)
2755      {      {
2756      const uschar *save_end_subject = end_subject;      const uschar *save_end_subject = end_subject;
2757    
2758      /* Advance to a unique first char if possible. If firstline is TRUE, the      /* If firstline is TRUE, the start of the match is constrained to the first
2759      start of the match is constrained to the first line of a multiline string.      line of a multiline string. Implement this by temporarily adjusting
2760      Implement this by temporarily adjusting end_subject so that we stop      end_subject so that we stop scanning at a newline. If the match fails at
2761      scanning at a newline. If the match fails at the newline, later code breaks      the newline, later code breaks this loop. */
     this loop. */  
2762    
2763      if (firstline)      if (firstline)
2764        {        {
2765        const uschar *t = current_subject;        USPTR t = current_subject;
2766    #ifdef SUPPORT_UTF8
2767          if (utf8)
2768            {
2769            while (t < md->end_subject && !IS_NEWLINE(t))
2770              {
2771              t++;
2772              while (t < end_subject && (*t & 0xc0) == 0x80) t++;
2773              }
2774            }
2775          else
2776    #endif
2777        while (t < md->end_subject && !IS_NEWLINE(t)) t++;        while (t < md->end_subject && !IS_NEWLINE(t)) t++;
2778        end_subject = t;        end_subject = t;
2779        }        }
2780    
2781      if (first_byte >= 0)      /* There are some optimizations that avoid running the match if a known
2782        starting point is not found, or if a known later character is not present.
2783        However, there is an option that disables these, for testing and for
2784        ensuring that all callouts do actually occur. */
2785    
2786        if ((options & PCRE_NO_START_OPTIMIZE) == 0)
2787        {        {
       if (first_byte_caseless)  
         while (current_subject < end_subject &&  
                lcc[*current_subject] != first_byte)  
           current_subject++;  
       else  
         while (current_subject < end_subject && *current_subject != first_byte)  
           current_subject++;  
       }  
2788    
2789      /* Or to just after a linebreak for a multiline match if possible */        /* Advance to a known first byte. */
2790    
2791      else if (startline)        if (first_byte >= 0)
2792        {          {
2793        if (current_subject > md->start_subject + start_offset)          if (first_byte_caseless)
2794              while (current_subject < end_subject &&
2795                     lcc[*current_subject] != first_byte)
2796                current_subject++;
2797            else
2798              while (current_subject < end_subject &&
2799                     *current_subject != first_byte)
2800                current_subject++;
2801            }
2802    
2803          /* Or to just after a linebreak for a multiline match if possible */
2804    
2805          else if (startline)
2806          {          {
2807          while (current_subject <= end_subject && !WAS_NEWLINE(current_subject))          if (current_subject > md->start_subject + start_offset)
2808            current_subject++;            {
2809    #ifdef SUPPORT_UTF8
2810              if (utf8)
2811                {
2812                while (current_subject < end_subject &&
2813                       !WAS_NEWLINE(current_subject))
2814                  {
2815                  current_subject++;
2816                  while(current_subject < end_subject &&
2817                        (*current_subject & 0xc0) == 0x80)
2818                    current_subject++;
2819                  }
2820                }
2821              else
2822    #endif
2823              while (current_subject < end_subject && !WAS_NEWLINE(current_subject))
2824                current_subject++;
2825    
2826              /* If we have just passed a CR and the newline option is ANY or
2827              ANYCRLF, and we are now at a LF, advance the match position by one
2828              more character. */
2829    
2830          /* If we have just passed a CR and the newline option is ANY or            if (current_subject[-1] == CHAR_CR &&
2831          ANYCRLF, and we are now at a LF, advance the match position by one more                 (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&
2832          character. */                 current_subject < end_subject &&
2833                   *current_subject == CHAR_NL)
2834          if (current_subject[-1] == '\r' &&              current_subject++;
2835               (md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&            }
              current_subject < end_subject &&  
              *current_subject == '\n')  
           current_subject++;  
2836          }          }
       }  
2837    
2838      /* Or to a non-unique first char after study */        /* Or to a non-unique first char after study */
2839    
2840      else if (start_bits != NULL)        else if (start_bits != NULL)
       {  
       while (current_subject < end_subject)  
2841          {          {
2842          register unsigned int c = *current_subject;          while (current_subject < end_subject)
2843          if ((start_bits[c/8] & (1 << (c&7))) == 0) current_subject++;            {
2844            else break;            register unsigned int c = *current_subject;
2845              if ((start_bits[c/8] & (1 << (c&7))) == 0) current_subject++;
2846                else break;
2847              }
2848          }          }
2849        }        }
2850    
# Line 2803  for (;;) Line 2866  for (;;)
2866    showed up when somebody was matching /^C/ on a 32-megabyte string... so we    showed up when somebody was matching /^C/ on a 32-megabyte string... so we
2867    don't do this when the string is sufficiently long.    don't do this when the string is sufficiently long.
2868    
2869    ALSO: this processing is disabled when partial matching is requested.    ALSO: this processing is disabled when partial matching is requested, and can
2870    */    also be explicitly deactivated. */
2871    
2872    if (req_byte >= 0 &&    if ((options & PCRE_NO_START_OPTIMIZE) == 0 &&
2873          req_byte >= 0 &&
2874        end_subject - current_subject < REQ_BYTE_MAX &&        end_subject - current_subject < REQ_BYTE_MAX &&
2875        (options & PCRE_PARTIAL) == 0)        (options & PCRE_PARTIAL) == 0)
2876      {      {
# Line 2882  for (;;) Line 2946  for (;;)
2946    not contain any explicit matches for \r or \n, and the newline option is CRLF    not contain any explicit matches for \r or \n, and the newline option is CRLF
2947    or ANY or ANYCRLF, advance the match position by one more character. */    or ANY or ANYCRLF, advance the match position by one more character. */
2948    
2949    if (current_subject[-1] == '\r' &&    if (current_subject[-1] == CHAR_CR &&
2950        current_subject < end_subject &&        current_subject < end_subject &&
2951        *current_subject == '\n' &&        *current_subject == CHAR_NL &&
2952        (re->flags & PCRE_HASCRORLF) == 0 &&        (re->flags & PCRE_HASCRORLF) == 0 &&
2953          (md->nltype == NLTYPE_ANY ||          (md->nltype == NLTYPE_ANY ||
2954           md->nltype == NLTYPE_ANYCRLF ||           md->nltype == NLTYPE_ANYCRLF ||

Legend:
Removed from v.345  
changed lines
  Added in v.406

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12