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

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

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

revision 774 by zherczeg, Thu Dec 1 06:08:45 2011 UTC revision 795 by zherczeg, Sat Dec 10 02:20:06 2011 UTC
# Line 102  overrun before it actually does run off Line 102  overrun before it actually does run off
102  #define REQ_CASELESS   0x10000000l      /* Indicates caselessness */  #define REQ_CASELESS   0x10000000l      /* Indicates caselessness */
103  #define REQ_VARY       0x20000000l      /* Reqchar followed non-literal item */  #define REQ_VARY       0x20000000l      /* Reqchar followed non-literal item */
104    
105    /* Repeated character flags. */
106    
107    #define UTF_LENGTH     0x10000000l      /* The char contains its length. */
108    
109  /* Table for handling escaped characters in the range '0'-'z'. Positive returns  /* Table for handling escaped characters in the range '0'-'z'. Positive returns
110  are simple data values; negative values are for special things like \d and so  are simple data values; negative values are for special things like \d and so
111  on. Zero means further processing is needed (for things like \x), or the escape  on. Zero means further processing is needed (for things like \x), or the escape
# Line 470  static const char error_texts[] = Line 474  static const char error_texts[] =
474    "\\k is not followed by a braced, angle-bracketed, or quoted name\0"    "\\k is not followed by a braced, angle-bracketed, or quoted name\0"
475    /* 70 */    /* 70 */
476    "internal error: unknown opcode in find_fixedlength()\0"    "internal error: unknown opcode in find_fixedlength()\0"
477      "Not allowed UTF-8 / UTF-16 code point (>= 0xd800 && <= 0xdfff)\0"
478    ;    ;
479    
480  /* Table to identify digits and hex digits. This is used when compiling  /* Table to identify digits and hex digits. This is used when compiling
# Line 538  static const pcre_uint8 digitab[] = Line 543  static const pcre_uint8 digitab[] =
543    
544  /* This is the "abnormal" case, for EBCDIC systems not running in UTF-8 mode. */  /* This is the "abnormal" case, for EBCDIC systems not running in UTF-8 mode. */
545    
546  static const pcre_unit8 digitab[] =  static const pcre_uint8 digitab[] =
547    {    {
548    0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   0-  7  0 */    0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   0-  7  0 */
549    0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   8- 15    */    0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /*   8- 15    */
# Line 706  static int Line 711  static int
711  check_escape(const pcre_uchar **ptrptr, int *errorcodeptr, int bracount,  check_escape(const pcre_uchar **ptrptr, int *errorcodeptr, int bracount,
712    int options, BOOL isclass)    int options, BOOL isclass)
713  {  {
714  BOOL utf8 = (options & PCRE_UTF8) != 0;  /* PCRE_UTF16 has the same value as PCRE_UTF8. */
715    BOOL utf = (options & PCRE_UTF8) != 0;
716  const pcre_uchar *ptr = *ptrptr + 1;  const pcre_uchar *ptr = *ptrptr + 1;
717  int c, i;  pcre_int32 c;
718    int i;
719    
720  GETCHARINCTEST(c, ptr);           /* Get character value, increment pointer */  GETCHARINCTEST(c, ptr);           /* Get character value, increment pointer */
721  ptr--;                            /* Set pointer back to the last byte */  ptr--;                            /* Set pointer back to the last byte */
# Line 940  else Line 947  else
947      c -= CHAR_0;      c -= CHAR_0;
948      while(i++ < 2 && ptr[1] >= CHAR_0 && ptr[1] <= CHAR_7)      while(i++ < 2 && ptr[1] >= CHAR_0 && ptr[1] <= CHAR_7)
949          c = c * 8 + *(++ptr) - CHAR_0;          c = c * 8 + *(++ptr) - CHAR_0;
950      if (!utf8 && c > 0xff) *errorcodeptr = ERR51;      if (!utf && c > 0xff) *errorcodeptr = ERR51;
951      break;      break;
952    
953      /* \x is complicated. \x{ddd} is a character number which can be greater      /* \x is complicated. \x{ddd} is a character number which can be greater
954      than 0xff in utf8 mode, but only if the ddd are hex digits. If not, { is      than 0xff in utf or non-8bit mode, but only if the ddd are hex digits.
955      treated as a data character. */      If not, { is treated as a data character. */
956    
957      case CHAR_x:      case CHAR_x:
958      if ((options & PCRE_JAVASCRIPT_COMPAT) != 0)      if ((options & PCRE_JAVASCRIPT_COMPAT) != 0)
# Line 974  else Line 981  else
981      if (ptr[1] == CHAR_LEFT_CURLY_BRACKET)      if (ptr[1] == CHAR_LEFT_CURLY_BRACKET)
982        {        {
983        const pcre_uchar *pt = ptr + 2;        const pcre_uchar *pt = ptr + 2;
       int count = 0;  
984    
985        c = 0;        c = 0;
986        while (MAX_255(*pt) && (digitab[*pt] & ctype_xdigit) != 0)        while (MAX_255(*pt) && (digitab[*pt] & ctype_xdigit) != 0)
987          {          {
988          register int cc = *pt++;          register int cc = *pt++;
989          if (c == 0 && cc == CHAR_0) continue;     /* Leading zeroes */          if (c == 0 && cc == CHAR_0) continue;     /* Leading zeroes */
         count++;  
990    
991  #ifndef EBCDIC  /* ASCII/UTF-8 coding */  #ifndef EBCDIC  /* ASCII/UTF-8 coding */
992          if (cc >= CHAR_a) cc -= 32;               /* Convert to upper case */          if (cc >= CHAR_a) cc -= 32;               /* Convert to upper case */
# Line 990  else Line 995  else
995          if (cc >= CHAR_a && cc <= CHAR_z) cc += 64;  /* Convert to upper case */          if (cc >= CHAR_a && cc <= CHAR_z) cc += 64;  /* Convert to upper case */
996          c = (c << 4) + cc - ((cc >= CHAR_0)? CHAR_0 : (CHAR_A - 10));          c = (c << 4) + cc - ((cc >= CHAR_0)? CHAR_0 : (CHAR_A - 10));
997  #endif  #endif
         }  
998    
       if (*pt == CHAR_RIGHT_CURLY_BRACKET)  
         {  
999  #ifdef COMPILE_PCRE8  #ifdef COMPILE_PCRE8
1000          if (c < 0 || count > (utf8? 8:2)) *errorcodeptr = ERR34;          if (c > (utf ? 0x10ffff : 0xff)) { c = -1; break; }
1001  #else  #else
1002  #ifdef COMPILE_PCRE16  #ifdef COMPILE_PCRE16
1003          if (c < 0 || count > (utf8? 8:4)) *errorcodeptr = ERR34;          if (c > (utf ? 0x10ffff : 0xffff)) { c = -1; break; }
1004  #endif  #endif
1005  #endif  #endif
1006            }
1007    
1008          if (c < 0)
1009            {
1010            while (MAX_255(*pt) && (digitab[*pt] & ctype_xdigit) != 0) pt++;
1011            *errorcodeptr = ERR34;
1012            }
1013    
1014          if (*pt == CHAR_RIGHT_CURLY_BRACKET)
1015            {
1016            if (utf && c >= 0xd800 && c <= 0xdfff) *errorcodeptr = ERR71;
1017          ptr = pt;          ptr = pt;
1018          break;          break;
1019          }          }
# Line 1281  Arguments: Line 1294  Arguments:
1294    name         name to seek, or NULL if seeking a numbered subpattern    name         name to seek, or NULL if seeking a numbered subpattern
1295    lorn         name length, or subpattern number if name is NULL    lorn         name length, or subpattern number if name is NULL
1296    xmode        TRUE if we are in /x mode    xmode        TRUE if we are in /x mode
1297    utf8         TRUE if we are in UTF-8 mode    utf          TRUE if we are in UTF-8 / UTF-16 mode
1298    count        pointer to the current capturing subpattern number (updated)    count        pointer to the current capturing subpattern number (updated)
1299    
1300  Returns:       the number of the named subpattern, or -1 if not found  Returns:       the number of the named subpattern, or -1 if not found
# Line 1289  Returns:       the number of the named s Line 1302  Returns:       the number of the named s
1302    
1303  static int  static int
1304  find_parens_sub(pcre_uchar **ptrptr, compile_data *cd, const pcre_uchar *name, int lorn,  find_parens_sub(pcre_uchar **ptrptr, compile_data *cd, const pcre_uchar *name, int lorn,
1305    BOOL xmode, BOOL utf8, int *count)    BOOL xmode, BOOL utf, int *count)
1306  {  {
1307  pcre_uchar *ptr = *ptrptr;  pcre_uchar *ptr = *ptrptr;
1308  int start_count = *count;  int start_count = *count;
# Line 1457  for (; ptr < cd->end_pattern; ptr++) Line 1470  for (; ptr < cd->end_pattern; ptr++)
1470        {        {
1471        if (IS_NEWLINE(ptr)) { ptr += cd->nllen - 1; break; }        if (IS_NEWLINE(ptr)) { ptr += cd->nllen - 1; break; }
1472        ptr++;        ptr++;
1473  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
1474        if (utf8) while ((*ptr & 0xc0) == 0x80) ptr++;        if (utf) FORWARDCHAR(ptr);
1475  #endif  #endif
1476        }        }
1477      if (*ptr == 0) goto FAIL_EXIT;      if (*ptr == 0) goto FAIL_EXIT;
# Line 1469  for (; ptr < cd->end_pattern; ptr++) Line 1482  for (; ptr < cd->end_pattern; ptr++)
1482    
1483    if (*ptr == CHAR_LEFT_PARENTHESIS)    if (*ptr == CHAR_LEFT_PARENTHESIS)
1484      {      {
1485      int rc = find_parens_sub(&ptr, cd, name, lorn, xmode, utf8, count);      int rc = find_parens_sub(&ptr, cd, name, lorn, xmode, utf, count);
1486      if (rc > 0) return rc;      if (rc > 0) return rc;
1487      if (*ptr == 0) goto FAIL_EXIT;      if (*ptr == 0) goto FAIL_EXIT;
1488      }      }
# Line 1515  Arguments: Line 1528  Arguments:
1528    name         name to seek, or NULL if seeking a numbered subpattern    name         name to seek, or NULL if seeking a numbered subpattern
1529    lorn         name length, or subpattern number if name is NULL    lorn         name length, or subpattern number if name is NULL
1530    xmode        TRUE if we are in /x mode    xmode        TRUE if we are in /x mode
1531    utf8         TRUE if we are in UTF-8 mode    utf          TRUE if we are in UTF-8 / UTF-16 mode
1532    
1533  Returns:       the number of the found subpattern, or -1 if not found  Returns:       the number of the found subpattern, or -1 if not found
1534  */  */
1535    
1536  static int  static int
1537  find_parens(compile_data *cd, const pcre_uchar *name, int lorn, BOOL xmode,  find_parens(compile_data *cd, const pcre_uchar *name, int lorn, BOOL xmode,
1538    BOOL utf8)    BOOL utf)
1539  {  {
1540  pcre_uchar *ptr = (pcre_uchar *)cd->start_pattern;  pcre_uchar *ptr = (pcre_uchar *)cd->start_pattern;
1541  int count = 0;  int count = 0;
# Line 1535  matching closing parens. That is why we Line 1548  matching closing parens. That is why we
1548    
1549  for (;;)  for (;;)
1550    {    {
1551    rc = find_parens_sub(&ptr, cd, name, lorn, xmode, utf8, &count);    rc = find_parens_sub(&ptr, cd, name, lorn, xmode, utf, &count);
1552    if (rc > 0 || *ptr++ == 0) break;    if (rc > 0 || *ptr++ == 0) break;
1553    }    }
1554    
# Line 1618  and doing the check at the end; a flag s Line 1631  and doing the check at the end; a flag s
1631    
1632  Arguments:  Arguments:
1633    code     points to the start of the pattern (the bracket)    code     points to the start of the pattern (the bracket)
1634    utf8     TRUE in UTF-8 mode    utf      TRUE in UTF-8 / UTF-16 mode
1635    atend    TRUE if called when the pattern is complete    atend    TRUE if called when the pattern is complete
1636    cd       the "compile data" structure    cd       the "compile data" structure
1637    
# Line 1630  Returns:   the fixed length, Line 1643  Returns:   the fixed length,
1643  */  */
1644    
1645  static int  static int
1646  find_fixedlength(pcre_uchar *code, BOOL utf8, BOOL atend, compile_data *cd)  find_fixedlength(pcre_uchar *code, BOOL utf, BOOL atend, compile_data *cd)
1647  {  {
1648  int length = -1;  int length = -1;
1649    
# Line 1657  for (;;) Line 1670  for (;;)
1670      case OP_ONCE:      case OP_ONCE:
1671      case OP_ONCE_NC:      case OP_ONCE_NC:
1672      case OP_COND:      case OP_COND:
1673      d = find_fixedlength(cc + ((op == OP_CBRA)? IMM2_SIZE : 0), utf8, atend, cd);      d = find_fixedlength(cc + ((op == OP_CBRA)? IMM2_SIZE : 0), utf, atend, cd);
1674      if (d < 0) return d;      if (d < 0) return d;
1675      branchlength += d;      branchlength += d;
1676      do cc += GET(cc, 1); while (*cc == OP_ALT);      do cc += GET(cc, 1); while (*cc == OP_ALT);
# Line 1691  for (;;) Line 1704  for (;;)
1704      cs = ce = (pcre_uchar *)cd->start_code + GET(cc, 1);  /* Start subpattern */      cs = ce = (pcre_uchar *)cd->start_code + GET(cc, 1);  /* Start subpattern */
1705      do ce += GET(ce, 1); while (*ce == OP_ALT);           /* End subpattern */      do ce += GET(ce, 1); while (*ce == OP_ALT);           /* End subpattern */
1706      if (cc > cs && cc < ce) return -1;                    /* Recursion */      if (cc > cs && cc < ce) return -1;                    /* Recursion */
1707      d = find_fixedlength(cs + 2, utf8, atend, cd);      d = find_fixedlength(cs + 2, utf, atend, cd);
1708      if (d < 0) return d;      if (d < 0) return d;
1709      branchlength += d;      branchlength += d;
1710      cc += 1 + LINK_SIZE;      cc += 1 + LINK_SIZE;
# Line 1750  for (;;) Line 1763  for (;;)
1763      case OP_NOTI:      case OP_NOTI:
1764      branchlength++;      branchlength++;
1765      cc += 2;      cc += 2;
1766  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
1767      if (utf8 && cc[-1] >= 0xc0) cc += PRIV(utf8_table4)[cc[-1] & 0x3f];      if (utf && HAS_EXTRALEN(cc[-1])) cc += GET_EXTRALEN(cc[-1]);
1768  #endif  #endif
1769      break;      break;
1770    
# Line 1764  for (;;) Line 1777  for (;;)
1777      case OP_NOTEXACTI:      case OP_NOTEXACTI:
1778      branchlength += GET2(cc,1);      branchlength += GET2(cc,1);
1779      cc += 2 + IMM2_SIZE;      cc += 2 + IMM2_SIZE;
1780  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
1781      if (utf8 && cc[-1] >= 0xc0) cc += PRIV(utf8_table4)[cc[-1] & 0x3f];      if (utf && HAS_EXTRALEN(cc[-1])) cc += GET_EXTRALEN(cc[-1]);
1782  #endif  #endif
1783      break;      break;
1784    
# Line 1945  length. Line 1958  length.
1958    
1959  Arguments:  Arguments:
1960    code        points to start of expression    code        points to start of expression
1961    utf8        TRUE in UTF-8 mode    utf         TRUE in UTF-8 / UTF-16 mode
1962    number      the required bracket number or negative to find a lookbehind    number      the required bracket number or negative to find a lookbehind
1963    
1964  Returns:      pointer to the opcode for the bracket, or NULL if not found  Returns:      pointer to the opcode for the bracket, or NULL if not found
1965  */  */
1966    
1967  const pcre_uchar *  const pcre_uchar *
1968  PRIV(find_bracket)(const pcre_uchar *code, BOOL utf8, int number)  PRIV(find_bracket)(const pcre_uchar *code, BOOL utf, int number)
1969  {  {
1970  for (;;)  for (;;)
1971    {    {
# Line 2032  for (;;) Line 2045  for (;;)
2045    a multi-byte character. The length in the table is a minimum, so we have to    a multi-byte character. The length in the table is a minimum, so we have to
2046    arrange to skip the extra bytes. */    arrange to skip the extra bytes. */
2047    
2048  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2049      if (utf8) switch(c)      if (utf) switch(c)
2050        {        {
2051        case OP_CHAR:        case OP_CHAR:
2052        case OP_CHARI:        case OP_CHARI:
# Line 2063  for (;;) Line 2076  for (;;)
2076        case OP_MINQUERYI:        case OP_MINQUERYI:
2077        case OP_POSQUERY:        case OP_POSQUERY:
2078        case OP_POSQUERYI:        case OP_POSQUERYI:
2079        if (code[-1] >= 0xc0) code += PRIV(utf8_table4)[code[-1] & 0x3f];        if (HAS_EXTRALEN(code[-1])) code += GET_EXTRALEN(code[-1]);
2080        break;        break;
2081        }        }
2082  #else  #else
2083      (void)(utf8);  /* Keep compiler happy by referencing function argument */      (void)(utf);  /* Keep compiler happy by referencing function argument */
2084  #endif  #endif
2085      }      }
2086    }    }
# Line 2084  instance of OP_RECURSE. Line 2097  instance of OP_RECURSE.
2097    
2098  Arguments:  Arguments:
2099    code        points to start of expression    code        points to start of expression
2100    utf8        TRUE in UTF-8 mode    utf         TRUE in UTF-8 / UTF-16 mode
2101    
2102  Returns:      pointer to the opcode for OP_RECURSE, or NULL if not found  Returns:      pointer to the opcode for OP_RECURSE, or NULL if not found
2103  */  */
2104    
2105  static const pcre_uchar *  static const pcre_uchar *
2106  find_recurse(const pcre_uchar *code, BOOL utf8)  find_recurse(const pcre_uchar *code, BOOL utf)
2107  {  {
2108  for (;;)  for (;;)
2109    {    {
# Line 2152  for (;;) Line 2165  for (;;)
2165      by a multi-byte character. The length in the table is a minimum, so we have      by a multi-byte character. The length in the table is a minimum, so we have
2166      to arrange to skip the extra bytes. */      to arrange to skip the extra bytes. */
2167    
2168  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2169      if (utf8) switch(c)      if (utf) switch(c)
2170        {        {
2171        case OP_CHAR:        case OP_CHAR:
2172        case OP_CHARI:        case OP_CHARI:
# Line 2183  for (;;) Line 2196  for (;;)
2196        case OP_MINQUERYI:        case OP_MINQUERYI:
2197        case OP_POSQUERY:        case OP_POSQUERY:
2198        case OP_POSQUERYI:        case OP_POSQUERYI:
2199        if (code[-1] >= 0xc0) code += PRIV(utf8_table4)[code[-1] & 0x3f];        if (HAS_EXTRALEN(code[-1])) code += GET_EXTRALEN(code[-1]);
2200        break;        break;
2201        }        }
2202  #else  #else
2203      (void)(utf8);  /* Keep compiler happy by referencing function argument */      (void)(utf);  /* Keep compiler happy by referencing function argument */
2204  #endif  #endif
2205      }      }
2206    }    }
# Line 2210  bracket whose current branch will alread Line 2223  bracket whose current branch will alread
2223  Arguments:  Arguments:
2224    code        points to start of search    code        points to start of search
2225    endcode     points to where to stop    endcode     points to where to stop
2226    utf8        TRUE if in UTF8 mode    utf         TRUE if in UTF-8 / UTF-16 mode
2227    cd          contains pointers to tables etc.    cd          contains pointers to tables etc.
2228    
2229  Returns:      TRUE if what is matched could be empty  Returns:      TRUE if what is matched could be empty
# Line 2218  Returns:      TRUE if what is matched co Line 2231  Returns:      TRUE if what is matched co
2231    
2232  static BOOL  static BOOL
2233  could_be_empty_branch(const pcre_uchar *code, const pcre_uchar *endcode,  could_be_empty_branch(const pcre_uchar *code, const pcre_uchar *endcode,
2234    BOOL utf8, compile_data *cd)    BOOL utf, compile_data *cd)
2235  {  {
2236  register int c;  register int c;
2237  for (code = first_significant_code(code + PRIV(OP_lengths)[*code], TRUE);  for (code = first_significant_code(code + PRIV(OP_lengths)[*code], TRUE);
# Line 2266  for (code = first_significant_code(code Line 2279  for (code = first_significant_code(code
2279    
2280      do      do
2281        {        {
2282        if (could_be_empty_branch(scode, endcode, utf8, cd))        if (could_be_empty_branch(scode, endcode, utf, cd))
2283          {          {
2284          empty_branch = TRUE;          empty_branch = TRUE;
2285          break;          break;
# Line 2322  for (code = first_significant_code(code Line 2335  for (code = first_significant_code(code
2335        empty_branch = FALSE;        empty_branch = FALSE;
2336        do        do
2337          {          {
2338          if (!empty_branch && could_be_empty_branch(code, endcode, utf8, cd))          if (!empty_branch && could_be_empty_branch(code, endcode, utf, cd))
2339            empty_branch = TRUE;            empty_branch = TRUE;
2340          code += GET(code, 1);          code += GET(code, 1);
2341          }          }
# Line 2344  for (code = first_significant_code(code Line 2357  for (code = first_significant_code(code
2357      actual length is stored in the compiled code, so we must update "code"      actual length is stored in the compiled code, so we must update "code"
2358      here. */      here. */
2359    
2360  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
2361      case OP_XCLASS:      case OP_XCLASS:
2362      ccode = code += GET(code, 1);      ccode = code += GET(code, 1);
2363      goto CHECK_CLASS_REPEAT;      goto CHECK_CLASS_REPEAT;
# Line 2354  for (code = first_significant_code(code Line 2367  for (code = first_significant_code(code
2367      case OP_NCLASS:      case OP_NCLASS:
2368      ccode = code + PRIV(OP_lengths)[OP_CLASS];      ccode = code + PRIV(OP_lengths)[OP_CLASS];
2369    
2370  #ifdef SUPPORT_UTF8  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
2371      CHECK_CLASS_REPEAT:      CHECK_CLASS_REPEAT:
2372  #endif  #endif
2373    
# Line 2443  for (code = first_significant_code(code Line 2456  for (code = first_significant_code(code
2456      /* In UTF-8 mode, STAR, MINSTAR, POSSTAR, QUERY, MINQUERY, POSQUERY, UPTO,      /* In UTF-8 mode, STAR, MINSTAR, POSSTAR, QUERY, MINQUERY, POSQUERY, UPTO,
2457      MINUPTO, and POSUPTO may be followed by a multibyte character */      MINUPTO, and POSUPTO may be followed by a multibyte character */
2458    
2459  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2460      case OP_STAR:      case OP_STAR:
2461      case OP_STARI:      case OP_STARI:
2462      case OP_MINSTAR:      case OP_MINSTAR:
# Line 2456  for (code = first_significant_code(code Line 2469  for (code = first_significant_code(code
2469      case OP_MINQUERYI:      case OP_MINQUERYI:
2470      case OP_POSQUERY:      case OP_POSQUERY:
2471      case OP_POSQUERYI:      case OP_POSQUERYI:
2472      if (utf8 && code[1] >= 0xc0) code += PRIV(utf8_table4)[code[1] & 0x3f];      if (utf && HAS_EXTRALEN(code[1])) code += GET_EXTRALEN(code[1]);
2473      break;      break;
2474    
2475      case OP_UPTO:      case OP_UPTO:
# Line 2465  for (code = first_significant_code(code Line 2478  for (code = first_significant_code(code
2478      case OP_MINUPTOI:      case OP_MINUPTOI:
2479      case OP_POSUPTO:      case OP_POSUPTO:
2480      case OP_POSUPTOI:      case OP_POSUPTOI:
2481      if (utf8 && code[1 + IMM2_SIZE] >= 0xc0) code += PRIV(utf8_table4)[code[1 + IMM2_SIZE] & 0x3f];      if (utf && HAS_EXTRALEN(code[1 + IMM2_SIZE])) code += GET_EXTRALEN(code[1 + IMM2_SIZE]);
2482      break;      break;
2483  #endif  #endif
2484    
# Line 2509  Arguments: Line 2522  Arguments:
2522    code        points to start of the recursion    code        points to start of the recursion
2523    endcode     points to where to stop (current RECURSE item)    endcode     points to where to stop (current RECURSE item)
2524    bcptr       points to the chain of current (unclosed) branch starts    bcptr       points to the chain of current (unclosed) branch starts
2525    utf8        TRUE if in UTF-8 mode    utf         TRUE if in UTF-8 / UTF-16 mode
2526    cd          pointers to tables etc    cd          pointers to tables etc
2527    
2528  Returns:      TRUE if what is matched could be empty  Returns:      TRUE if what is matched could be empty
# Line 2517  Returns:      TRUE if what is matched co Line 2530  Returns:      TRUE if what is matched co
2530    
2531  static BOOL  static BOOL
2532  could_be_empty(const pcre_uchar *code, const pcre_uchar *endcode,  could_be_empty(const pcre_uchar *code, const pcre_uchar *endcode,
2533    branch_chain *bcptr, BOOL utf8, compile_data *cd)    branch_chain *bcptr, BOOL utf, compile_data *cd)
2534  {  {
2535  while (bcptr != NULL && bcptr->current_branch >= code)  while (bcptr != NULL && bcptr->current_branch >= code)
2536    {    {
2537    if (!could_be_empty_branch(bcptr->current_branch, endcode, utf8, cd))    if (!could_be_empty_branch(bcptr->current_branch, endcode, utf, cd))
2538      return FALSE;      return FALSE;
2539    bcptr = bcptr->outer;    bcptr = bcptr->outer;
2540    }    }
# Line 2656  value in the reference (which is a group Line 2669  value in the reference (which is a group
2669  Arguments:  Arguments:
2670    group      points to the start of the group    group      points to the start of the group
2671    adjust     the amount by which the group is to be moved    adjust     the amount by which the group is to be moved
2672    utf8       TRUE in UTF-8 mode    utf        TRUE in UTF-8 / UTF-16 mode
2673    cd         contains pointers to tables etc.    cd         contains pointers to tables etc.
2674    save_hwm   the hwm forward reference pointer at the start of the group    save_hwm   the hwm forward reference pointer at the start of the group
2675    
# Line 2664  Returns:     nothing Line 2677  Returns:     nothing
2677  */  */
2678    
2679  static void  static void
2680  adjust_recurse(pcre_uchar *group, int adjust, BOOL utf8, compile_data *cd,  adjust_recurse(pcre_uchar *group, int adjust, BOOL utf, compile_data *cd,
2681    pcre_uchar *save_hwm)    pcre_uchar *save_hwm)
2682  {  {
2683  pcre_uchar *ptr = group;  pcre_uchar *ptr = group;
2684    
2685  while ((ptr = (pcre_uchar *)find_recurse(ptr, utf8)) != NULL)  while ((ptr = (pcre_uchar *)find_recurse(ptr, utf)) != NULL)
2686    {    {
2687    int offset;    int offset;
2688    pcre_uchar *hc;    pcre_uchar *hc;
# Line 2875  sense to automatically possessify the re Line 2888  sense to automatically possessify the re
2888    
2889  Arguments:  Arguments:
2890    previous      pointer to the repeated opcode    previous      pointer to the repeated opcode
2891    utf8          TRUE in UTF-8 mode    utf           TRUE in UTF-8 / UTF-16 mode
2892    ptr           next character in pattern    ptr           next character in pattern
2893    options       options bits    options       options bits
2894    cd            contains pointers to tables etc.    cd            contains pointers to tables etc.
# Line 2884  Returns:        TRUE if possessifying is Line 2897  Returns:        TRUE if possessifying is
2897  */  */
2898    
2899  static BOOL  static BOOL
2900  check_auto_possessive(const pcre_uchar *previous, BOOL utf8,  check_auto_possessive(const pcre_uchar *previous, BOOL utf,
2901    const pcre_uchar *ptr, int options, compile_data *cd)    const pcre_uchar *ptr, int options, compile_data *cd)
2902  {  {
2903  int c, next;  pcre_int32 c, next;
2904  int op_code = *previous++;  int op_code = *previous++;
2905    
2906  /* Skip whitespace and comments in extended mode */  /* Skip whitespace and comments in extended mode */
# Line 2904  if ((options & PCRE_EXTENDED) != 0) Line 2917  if ((options & PCRE_EXTENDED) != 0)
2917          {          {
2918          if (IS_NEWLINE(ptr)) { ptr += cd->nllen; break; }          if (IS_NEWLINE(ptr)) { ptr += cd->nllen; break; }
2919          ptr++;          ptr++;
2920  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2921          if (utf8) while ((*ptr & 0xc0) == 0x80) ptr++;          if (utf) FORWARDCHAR(ptr);
2922  #endif  #endif
2923          }          }
2924        }        }
# Line 2923  if (*ptr == CHAR_BACKSLASH) Line 2936  if (*ptr == CHAR_BACKSLASH)
2936    if (temperrorcode != 0) return FALSE;    if (temperrorcode != 0) return FALSE;
2937    ptr++;    /* Point after the escape sequence */    ptr++;    /* Point after the escape sequence */
2938    }    }
2939    else if (!MAX_255(*ptr) || (cd->ctypes[*ptr] & ctype_meta) == 0)
 else if ((cd->ctypes[*ptr] & ctype_meta) == 0)  
2940    {    {
2941  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2942    if (utf8) { GETCHARINC(next, ptr); } else    if (utf) { GETCHARINC(next, ptr); } else
2943  #endif  #endif
2944    next = *ptr++;    next = *ptr++;
2945    }    }
   
2946  else return FALSE;  else return FALSE;
2947    
2948  /* Skip whitespace and comments in extended mode */  /* Skip whitespace and comments in extended mode */
# Line 2948  if ((options & PCRE_EXTENDED) != 0) Line 2959  if ((options & PCRE_EXTENDED) != 0)
2959          {          {
2960          if (IS_NEWLINE(ptr)) { ptr += cd->nllen; break; }          if (IS_NEWLINE(ptr)) { ptr += cd->nllen; break; }
2961          ptr++;          ptr++;
2962  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2963          if (utf8) while ((*ptr & 0xc0) == 0x80) ptr++;          if (utf) FORWARDCHAR(ptr);
2964  #endif  #endif
2965          }          }
2966        }        }
# Line 2969  the next item is a character. */ Line 2980  the next item is a character. */
2980  if (next >= 0) switch(op_code)  if (next >= 0) switch(op_code)
2981    {    {
2982    case OP_CHAR:    case OP_CHAR:
2983  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2984    GETCHARTEST(c, previous);    GETCHARTEST(c, previous);
2985  #else  #else
2986    c = *previous;    c = *previous;
# Line 2981  if (next >= 0) switch(op_code) Line 2992  if (next >= 0) switch(op_code)
2992    high-valued characters. */    high-valued characters. */
2993    
2994    case OP_CHARI:    case OP_CHARI:
2995  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
2996    GETCHARTEST(c, previous);    GETCHARTEST(c, previous);
2997  #else  #else
2998    c = *previous;    c = *previous;
2999  #endif  #endif
3000    if (c == next) return FALSE;    if (c == next) return FALSE;
3001  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3002    if (utf8)    if (utf)
3003      {      {
3004      unsigned int othercase;      unsigned int othercase;
3005      if (next < 128) othercase = cd->fcc[next]; else      if (next < 128) othercase = cd->fcc[next]; else
# Line 3000  if (next >= 0) switch(op_code) Line 3011  if (next >= 0) switch(op_code)
3011      return (unsigned int)c != othercase;      return (unsigned int)c != othercase;
3012      }      }
3013    else    else
3014  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
3015    return (c != cd->fcc[next]);  /* Non-UTF-8 mode */    return (c != cd->fcc[next]);  /* Non-UTF-8 mode */
3016    
3017    /* For OP_NOT and OP_NOTI, the data is always a single-byte character. These    /* For OP_NOT and OP_NOTI, the data is always a single-byte character. These
# Line 3012  if (next >= 0) switch(op_code) Line 3023  if (next >= 0) switch(op_code)
3023    
3024    case OP_NOTI:    case OP_NOTI:
3025    if ((c = *previous) == next) return TRUE;    if ((c = *previous) == next) return TRUE;
3026  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3027    if (utf8)    if (utf)
3028      {      {
3029      unsigned int othercase;      unsigned int othercase;
3030      if (next < 128) othercase = cd->fcc[next]; else      if (next < 128) othercase = cd->fcc[next]; else
# Line 3025  if (next >= 0) switch(op_code) Line 3036  if (next >= 0) switch(op_code)
3036      return (unsigned int)c == othercase;      return (unsigned int)c == othercase;
3037      }      }
3038    else    else
3039  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF */
3040    return (c == cd->fcc[next]);  /* Non-UTF-8 mode */    return (c == cd->fcc[next]);  /* Non-UTF-8 mode */
3041    
3042    /* Note that OP_DIGIT etc. are generated only when PCRE_UCP is *not* set.    /* Note that OP_DIGIT etc. are generated only when PCRE_UCP is *not* set.
# Line 3117  switch(op_code) Line 3128  switch(op_code)
3128    {    {
3129    case OP_CHAR:    case OP_CHAR:
3130    case OP_CHARI:    case OP_CHARI:
3131  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3132    GETCHARTEST(c, previous);    GETCHARTEST(c, previous);
3133  #else  #else
3134    c = *previous;    c = *previous;
# Line 3347  pcre_uint8 classbits[32]; Line 3358  pcre_uint8 classbits[32];
3358  must not do this for other options (e.g. PCRE_EXTENDED) because they may change  must not do this for other options (e.g. PCRE_EXTENDED) because they may change
3359  dynamically as we process the pattern. */  dynamically as we process the pattern. */
3360    
3361  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3362  BOOL utf8 = (options & PCRE_UTF8) != 0;  /* PCRE_UTF16 has the same value as PCRE_UTF8. */
3363  pcre_uint8 utf8_char[6];  BOOL utf = (options & PCRE_UTF8) != 0;
3364    pcre_uchar utf_chars[6];
3365  #else  #else
3366  BOOL utf8 = FALSE;  BOOL utf = FALSE;
3367  #endif  #endif
3368    
3369  /* Helper variables for OP_XCLASS opcode (for characters > 255). */  /* Helper variables for OP_XCLASS opcode (for characters > 255). */
# Line 3414  for (;; ptr++) Line 3426  for (;; ptr++)
3426    int tempbracount;    int tempbracount;
3427    pcre_uchar mcbuffer[8];    pcre_uchar mcbuffer[8];
3428    
3429    /* Get next byte in the pattern */    /* Get next character in the pattern */
3430    
3431    c = *ptr;    c = *ptr;
3432    
# Line 3459  for (;; ptr++) Line 3471  for (;; ptr++)
3471        }        }
3472    
3473      *lengthptr += (int)(code - last_code);      *lengthptr += (int)(code - last_code);
3474      DPRINTF(("length=%d added %d c=%c\n", *lengthptr, (int)(code - last_code),      DPRINTF(("length=%d added %d c=%c (0x%x)\n", *lengthptr,
3475        c));        (int)(code - last_code), c, c));
3476    
3477      /* If "previous" is set and it is not at the start of the work space, move      /* If "previous" is set and it is not at the start of the work space, move
3478      it back to there, in order to avoid filling up the work space. Otherwise,      it back to there, in order to avoid filling up the work space. Otherwise,
# Line 3546  for (;; ptr++) Line 3558  for (;; ptr++)
3558          {          {
3559          if (IS_NEWLINE(ptr)) { ptr += cd->nllen - 1; break; }          if (IS_NEWLINE(ptr)) { ptr += cd->nllen - 1; break; }
3560          ptr++;          ptr++;
3561  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3562          if (utf8) while ((*ptr & 0xc0) == 0x80) ptr++;          if (utf) FORWARDCHAR(ptr);
3563  #endif  #endif
3564          }          }
3565        if (*ptr != 0) continue;        if (*ptr != 0) continue;
# Line 3726  for (;; ptr++) Line 3738  for (;; ptr++)
3738        {        {
3739        const pcre_uchar *oldptr;        const pcre_uchar *oldptr;
3740    
3741  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
3742        if (utf8 && c > 127)        if (utf && HAS_EXTRALEN(c))
3743          {                           /* Braces are required because the */          {                           /* Braces are required because the */
3744          GETCHARLEN(c, ptr, ptr);    /* macro generates multiple statements */          GETCHARLEN(c, ptr, ptr);    /* macro generates multiple statements */
3745          }          }
# Line 3945  for (;; ptr++) Line 3957  for (;; ptr++)
3957              SETBIT(classbits, 0x20); /* SPACE */              SETBIT(classbits, 0x20); /* SPACE */
3958              SETBIT(classbits, 0xa0); /* NSBP */              SETBIT(classbits, 0xa0); /* NSBP */
3959  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3960              if (utf8)              if (utf)
3961                {                {
3962                xclass = TRUE;                xclass = TRUE;
3963                *class_uchardata++ = XCL_SINGLE;                *class_uchardata++ = XCL_SINGLE;
3964                class_uchardata += PRIV(ord2utf8)(0x1680, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x1680, class_uchardata);
3965                *class_uchardata++ = XCL_SINGLE;                *class_uchardata++ = XCL_SINGLE;
3966                class_uchardata += PRIV(ord2utf8)(0x180e, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x180e, class_uchardata);
3967                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
3968                class_uchardata += PRIV(ord2utf8)(0x2000, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x2000, class_uchardata);
3969                class_uchardata += PRIV(ord2utf8)(0x200A, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x200A, class_uchardata);
3970                *class_uchardata++ = XCL_SINGLE;                *class_uchardata++ = XCL_SINGLE;
3971                class_uchardata += PRIV(ord2utf8)(0x202f, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x202f, class_uchardata);
3972                *class_uchardata++ = XCL_SINGLE;                *class_uchardata++ = XCL_SINGLE;
3973                class_uchardata += PRIV(ord2utf8)(0x205f, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x205f, class_uchardata);
3974                *class_uchardata++ = XCL_SINGLE;                *class_uchardata++ = XCL_SINGLE;
3975                class_uchardata += PRIV(ord2utf8)(0x3000, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x3000, class_uchardata);
3976                }                }
3977  #endif  #endif
3978              continue;              continue;
# Line 3980  for (;; ptr++) Line 3992  for (;; ptr++)
3992                }                }
3993    
3994  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
3995              if (utf8)              if (utf)
3996                {                {
3997                xclass = TRUE;                xclass = TRUE;
3998                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
3999                class_uchardata += PRIV(ord2utf8)(0x0100, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x0100, class_uchardata);
4000                class_uchardata += PRIV(ord2utf8)(0x167f, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x167f, class_uchardata);
4001                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
4002                class_uchardata += PRIV(ord2utf8)(0x1681, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x1681, class_uchardata);
4003                class_uchardata += PRIV(ord2utf8)(0x180d, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x180d, class_uchardata);
4004                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
4005                class_uchardata += PRIV(ord2utf8)(0x180f, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x180f, class_uchardata);
4006                class_uchardata += PRIV(ord2utf8)(0x1fff, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x1fff, class_uchardata);
4007                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
4008                class_uchardata += PRIV(ord2utf8)(0x200B, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x200B, class_uchardata);
4009                class_uchardata += PRIV(ord2utf8)(0x202e, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x202e, class_uchardata);
4010                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
4011                class_uchardata += PRIV(ord2utf8)(0x2030, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x2030, class_uchardata);
4012                class_uchardata += PRIV(ord2utf8)(0x205e, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x205e, class_uchardata);
4013                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
4014                class_uchardata += PRIV(ord2utf8)(0x2060, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x2060, class_uchardata);
4015                class_uchardata += PRIV(ord2utf8)(0x2fff, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x2fff, class_uchardata);
4016                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
4017                class_uchardata += PRIV(ord2utf8)(0x3001, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x3001, class_uchardata);
4018                class_uchardata += PRIV(ord2utf8)(0x7fffffff, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x10ffff, class_uchardata);
4019                }                }
4020  #endif  #endif
4021              continue;              continue;
# Line 4015  for (;; ptr++) Line 4027  for (;; ptr++)
4027              SETBIT(classbits, 0x0d); /* CR */              SETBIT(classbits, 0x0d); /* CR */
4028              SETBIT(classbits, 0x85); /* NEL */              SETBIT(classbits, 0x85); /* NEL */
4029  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
4030              if (utf8)              if (utf)
4031                {                {
4032                xclass = TRUE;                xclass = TRUE;
4033                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
4034                class_uchardata += PRIV(ord2utf8)(0x2028, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x2028, class_uchardata);
4035                class_uchardata += PRIV(ord2utf8)(0x2029, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x2029, class_uchardata);
4036                }                }
4037  #endif  #endif
4038              continue;              continue;
# Line 4043  for (;; ptr++) Line 4055  for (;; ptr++)
4055                }                }
4056    
4057  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
4058              if (utf8)              if (utf)
4059                {                {
4060                xclass = TRUE;                xclass = TRUE;
4061                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
4062                class_uchardata += PRIV(ord2utf8)(0x0100, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x0100, class_uchardata);
4063                class_uchardata += PRIV(ord2utf8)(0x2027, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x2027, class_uchardata);
4064                *class_uchardata++ = XCL_RANGE;                *class_uchardata++ = XCL_RANGE;
4065                class_uchardata += PRIV(ord2utf8)(0x2029, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x2029, class_uchardata);
4066                class_uchardata += PRIV(ord2utf8)(0x7fffffff, class_uchardata);                class_uchardata += PRIV(ord2utf)(0x10ffff, class_uchardata);
4067                }                }
4068  #endif  #endif
4069              continue;              continue;
# Line 4138  for (;; ptr++) Line 4150  for (;; ptr++)
4150            goto LONE_SINGLE_CHARACTER;            goto LONE_SINGLE_CHARACTER;
4151            }            }
4152    
4153  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4154          if (utf8)          if (utf)
4155            {                           /* Braces are required because the */            {                           /* Braces are required because the */
4156            GETCHARLEN(d, ptr, ptr);    /* macro generates multiple statements */            GETCHARLEN(d, ptr, ptr);    /* macro generates multiple statements */
4157            }            }
# Line 4188  for (;; ptr++) Line 4200  for (;; ptr++)
4200          matching for characters > 127 is available only if UCP support is          matching for characters > 127 is available only if UCP support is
4201          available. */          available. */
4202    
4203  #ifdef SUPPORT_UTF  #if defined SUPPORT_UTF && !(defined COMPILE_PCRE8)
4204          if (utf8 && (d > 255 || ((options & PCRE_CASELESS) != 0 && d > 127)))          if ((d > 255) || (utf && ((options & PCRE_CASELESS) != 0 && d > 127)))
4205  #endif  #elif defined  SUPPORT_UTF
4206  #ifndef COMPILE_PCRE8          if (utf && (d > 255 || ((options & PCRE_CASELESS) != 0 && d > 127)))
4207    #elif !(defined COMPILE_PCRE8)
4208          if (d > 255)          if (d > 255)
4209  #endif  #endif
4210  #if defined SUPPORT_UTF || defined COMPILE_PCRE16  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
4211            {            {
4212            xclass = TRUE;            xclass = TRUE;
4213    
# Line 4203  for (;; ptr++) Line 4216  for (;; ptr++)
4216            they fit with the basic range. */            they fit with the basic range. */
4217    
4218  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
4219    #ifndef COMPILE_PCRE8
4220              if (utf && (options & PCRE_CASELESS) != 0)
4221    #else
4222            if ((options & PCRE_CASELESS) != 0)            if ((options & PCRE_CASELESS) != 0)
4223    #endif
4224              {              {
4225              unsigned int occ, ocd;              unsigned int occ, ocd;
4226              unsigned int cc = c;              unsigned int cc = c;
# Line 4234  for (;; ptr++) Line 4251  for (;; ptr++)
4251                else                else
4252                  {                  {
4253                  *class_uchardata++ = XCL_RANGE;                  *class_uchardata++ = XCL_RANGE;
4254                  class_uchardata += PRIV(ord2utf8)(occ, class_uchardata);                  class_uchardata += PRIV(ord2utf)(occ, class_uchardata);
4255                  }                  }
4256                class_uchardata += PRIV(ord2utf8)(ocd, class_uchardata);                class_uchardata += PRIV(ord2utf)(ocd, class_uchardata);
4257                }                }
4258              }              }
4259  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
# Line 4246  for (;; ptr++) Line 4263  for (;; ptr++)
4263    
4264            *class_uchardata++ = XCL_RANGE;            *class_uchardata++ = XCL_RANGE;
4265  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
4266            class_uchardata += PRIV(ord2utf8)(c, class_uchardata);  #ifndef COMPILE_PCRE8
4267            class_uchardata += PRIV(ord2utf8)(d, class_uchardata);            if (utf)
4268                {
4269                class_uchardata += PRIV(ord2utf)(c, class_uchardata);
4270                class_uchardata += PRIV(ord2utf)(d, class_uchardata);
4271                }
4272              else
4273                {
4274                *class_uchardata++ = c;
4275                *class_uchardata++ = d;
4276                }
4277  #else  #else
4278              class_uchardata += PRIV(ord2utf)(c, class_uchardata);
4279              class_uchardata += PRIV(ord2utf)(d, class_uchardata);
4280    #endif
4281    #else /* SUPPORT_UTF */
4282            *class_uchardata++ = c;            *class_uchardata++ = c;
4283            *class_uchardata++ = d;            *class_uchardata++ = d;
4284  #endif  #endif /* SUPPORT_UTF */
4285    
4286            /* With UCP support, we are done. Without UCP support, there is no            /* With UCP support, we are done. Without UCP support, there is no
4287            caseless matching for UTF characters > 127; we can use the bit map            caseless matching for UTF characters > 127; we can use the bit map
# Line 4259  for (;; ptr++) Line 4289  for (;; ptr++)
4289            can still use  */            can still use  */
4290    
4291  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
4292            continue;    /* With next character in the class */  #ifndef COMPILE_PCRE8
4293  #else            if (utf)
4294  #ifdef SUPPORT_UTF  #endif
4295                continue;    /* With next character in the class */
4296    #endif  /* SUPPORT_UCP */
4297    
4298    #if defined SUPPORT_UTF && !defined(SUPPORT_UCP) && !(defined COMPILE_PCRE8)
4299              if (utf)
4300                {
4301                if ((options & PCRE_CASELESS) == 0 || c > 127) continue;
4302                /* Adjust upper limit and fall through to set up the map */
4303                d = 127;
4304                }
4305              else
4306                {
4307                if (c > 255) continue;
4308                /* Adjust upper limit and fall through to set up the map */
4309                d = 255;
4310                }
4311    #elif defined SUPPORT_UTF && !defined(SUPPORT_UCP)
4312            if ((options & PCRE_CASELESS) == 0 || c > 127) continue;            if ((options & PCRE_CASELESS) == 0 || c > 127) continue;
4313            /* Adjust upper limit and fall through to set up the map */            /* Adjust upper limit and fall through to set up the map */
4314            d = 127;            d = 127;
# Line 4269  for (;; ptr++) Line 4316  for (;; ptr++)
4316            if (c > 255) continue;            if (c > 255) continue;
4317            /* Adjust upper limit and fall through to set up the map */            /* Adjust upper limit and fall through to set up the map */
4318            d = 255;            d = 255;
4319  #endif  /* SUPPORT_UTF */  #endif  /* SUPPORT_UTF && !SUPPORT_UCP && !COMPILE_PCRE8 */
 #endif  /* SUPPORT_UCP */  
4320            }            }
4321  #endif  /* SUPPORT_UTF8 || COMPILE_PCRE16 */  #endif  /* SUPPORT_UTF || !COMPILE_PCRE8 */
4322    
4323          /* We use the bit map for 8 bit mode, or when the characters fall          /* We use the bit map for 8 bit mode, or when the characters fall
4324          partially or entirely to [0-255] ([0-127] for UCP) ranges. */          partially or entirely to [0-255] ([0-127] for UCP) ranges. */
# Line 4303  for (;; ptr++) Line 4349  for (;; ptr++)
4349    
4350        /* Handle a character that cannot go in the bit map */        /* Handle a character that cannot go in the bit map */
4351    
4352  #ifdef SUPPORT_UTF  #if defined SUPPORT_UTF && !(defined COMPILE_PCRE8)
4353        if (utf8 && (c > 255 || ((options & PCRE_CASELESS) != 0 && c > 127)))        if ((c > 255) || (utf && ((options & PCRE_CASELESS) != 0 && c > 127)))
4354  #endif  #elif defined SUPPORT_UTF
4355  #ifndef COMPILE_PCRE8        if (utf && (c > 255 || ((options & PCRE_CASELESS) != 0 && c > 127)))
4356    #elif !(defined COMPILE_PCRE8)
4357        if (c > 255)        if (c > 255)
4358  #endif  #endif
4359  #if defined SUPPORT_UTF || defined COMPILE_PCRE16  #if defined SUPPORT_UTF || !(defined COMPILE_PCRE8)
4360          {          {
4361          xclass = TRUE;          xclass = TRUE;
4362          *class_uchardata++ = XCL_SINGLE;          *class_uchardata++ = XCL_SINGLE;
4363  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
4364          class_uchardata += PRIV(ord2utf8)(c, class_uchardata);  #ifndef COMPILE_PCRE8
4365  #else          /* In non 8 bit mode, we can get here even
4366          *class_uchardata++ = c;          if we are not in UTF mode. */
4367            if (!utf)
4368              *class_uchardata++ = c;
4369            else
4370  #endif  #endif
4371              class_uchardata += PRIV(ord2utf)(c, class_uchardata);
4372    #else /* SUPPORT_UTF */
4373            *class_uchardata++ = c;
4374    #endif /* SUPPORT_UTF */
4375    
4376  #ifdef SUPPORT_UCP  #ifdef SUPPORT_UCP
4377    #ifdef COMPILE_PCRE8
4378          if ((options & PCRE_CASELESS) != 0)          if ((options & PCRE_CASELESS) != 0)
4379    #else
4380            /* In non 8 bit mode, we can get here even
4381            if we are not in UTF mode. */
4382            if (utf && (options & PCRE_CASELESS) != 0)
4383    #endif
4384            {            {
4385            unsigned int othercase;            unsigned int othercase;
4386            if ((othercase = UCD_OTHERCASE(c)) != c)            if ((othercase = UCD_OTHERCASE(c)) != c)
4387              {              {
4388              *class_uchardata++ = XCL_SINGLE;              *class_uchardata++ = XCL_SINGLE;
4389              class_uchardata += PRIV(ord2utf8)(othercase, class_uchardata);              class_uchardata += PRIV(ord2utf)(othercase, class_uchardata);
4390              }              }
4391            }            }
4392  #endif  /* SUPPORT_UCP */  #endif  /* SUPPORT_UCP */
4393    
4394          }          }
4395        else        else
4396  #endif  /* SUPPORT_UTF8 */  #endif  /* SUPPORT_UTF || COMPILE_PCRE16 */
   
4397        /* Handle a single-byte character */        /* Handle a single-byte character */
4398          {          {
4399          classbits[c/8] |= (1 << (c&7));          classbits[c/8] |= (1 << (c&7));
# Line 4346  for (;; ptr++) Line 4405  for (;; ptr++)
4405          class_charcount++;          class_charcount++;
4406          class_lastchar = c;          class_lastchar = c;
4407          }          }
4408    
4409        }        }
4410    
4411      /* Loop until ']' reached. This "while" is the end of the "do" far above.      /* Loop until ']' reached. This "while" is the end of the "do" far above.
# Line 4384  for (;; ptr++) Line 4444  for (;; ptr++)
4444    
4445  #ifdef SUPPORT_UTF  #ifdef SUPPORT_UTF
4446      if (class_charcount == 1 && !xclass &&      if (class_charcount == 1 && !xclass &&
4447        (!utf8 || !negate_class || class_lastchar < 128))        (!utf || !negate_class || class_lastchar < 128))
 #elif defined COMPILE_PCRE8  
     if (class_charcount == 1)  
4448  #else  #else
4449      if (class_charcount == 1 && !xclass)      if (class_charcount == 1)
4450  #endif  #endif
4451        {        {
4452        zeroreqchar = reqchar;        zeroreqchar = reqchar;
# Line 4407  for (;; ptr++) Line 4465  for (;; ptr++)
4465        /* For a single, positive character, get the value into mcbuffer, and        /* For a single, positive character, get the value into mcbuffer, and
4466        then we can handle this with the normal one-character code. */        then we can handle this with the normal one-character code. */
4467    
4468  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4469        if (utf8 && class_lastchar > 127)        if (utf && class_lastchar > 127)
4470          mclength = PRIV(ord2utf8)(class_lastchar, mcbuffer);          mclength = PRIV(ord2utf)(class_lastchar, mcbuffer);
4471        else        else
4472  #endif  #endif
4473          {          {
# Line 4593  for (;; ptr++) Line 4651  for (;; ptr++)
4651        {        {
4652        op_type = (*previous == OP_CHAR)? 0 : OP_STARI - OP_STAR;        op_type = (*previous == OP_CHAR)? 0 : OP_STARI - OP_STAR;
4653    
4654        /* Deal with UTF-8 characters that take up more than one byte. It's        /* Deal with UTF characters that take up more than one character. It's
4655        easier to write this out separately than try to macrify it. Use c to        easier to write this out separately than try to macrify it. Use c to
4656        hold the length of the character in bytes, plus 0x80 to flag that it's a        hold the length of the character in bytes, plus UTF_LENGTH to flag that
4657        length rather than a small character. */        it's a length rather than a small character. */
4658    
4659  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4660        if (utf8 && (code[-1] & 0x80) != 0)        if (utf && NOT_FIRSTCHAR(code[-1]))
4661          {          {
4662          pcre_uchar *lastchar = code - 1;          pcre_uchar *lastchar = code - 1;
4663          while((*lastchar & 0xc0) == 0x80) lastchar--;          BACKCHAR(lastchar);
4664          c = code - lastchar;            /* Length of UTF-8 character */          c = code - lastchar;            /* Length of UTF-8 character */
4665          memcpy(utf8_char, lastchar, c); /* Save the char */          memcpy(utf_chars, lastchar, IN_UCHARS(c)); /* Save the char */
4666          c |= 0x80;                      /* Flag c as a length */          c |= UTF_LENGTH;                /* Flag c as a length */
4667          }          }
4668        else        else
4669  #endif  #endif /* SUPPORT_UTF */
   
       /* Handle the case of a single byte - either with no UTF8 support, or  
       with UTF-8 disabled, or for a UTF-8 character < 128. */  
4670    
4671          /* Handle the case of a single charater - either with no UTF support, or
4672          with UTF disabled, or for a single character UTF character. */
4673          {          {
4674          c = code[-1];          c = code[-1];
4675          if (repeat_min > 1) reqchar = c | req_caseopt | cd->req_varyopt;          if (repeat_min > 1) reqchar = c | req_caseopt | cd->req_varyopt;
# Line 4625  for (;; ptr++) Line 4682  for (;; ptr++)
4682    
4683        if (!possessive_quantifier &&        if (!possessive_quantifier &&
4684            repeat_max < 0 &&            repeat_max < 0 &&
4685            check_auto_possessive(previous, utf8, ptr + 1, options, cd))            check_auto_possessive(previous, utf, ptr + 1, options, cd))
4686          {          {
4687          repeat_type = 0;    /* Force greedy */          repeat_type = 0;    /* Force greedy */
4688          possessive_quantifier = TRUE;          possessive_quantifier = TRUE;
# Line 4646  for (;; ptr++) Line 4703  for (;; ptr++)
4703        c = previous[1];        c = previous[1];
4704        if (!possessive_quantifier &&        if (!possessive_quantifier &&
4705            repeat_max < 0 &&            repeat_max < 0 &&
4706            check_auto_possessive(previous, utf8, ptr + 1, options, cd))            check_auto_possessive(previous, utf, ptr + 1, options, cd))
4707          {          {
4708          repeat_type = 0;    /* Force greedy */          repeat_type = 0;    /* Force greedy */
4709          possessive_quantifier = TRUE;          possessive_quantifier = TRUE;
# Line 4670  for (;; ptr++) Line 4727  for (;; ptr++)
4727    
4728        if (!possessive_quantifier &&        if (!possessive_quantifier &&
4729            repeat_max < 0 &&            repeat_max < 0 &&
4730            check_auto_possessive(previous, utf8, ptr + 1, options, cd))            check_auto_possessive(previous, utf, ptr + 1, options, cd))
4731          {          {
4732          repeat_type = 0;    /* Force greedy */          repeat_type = 0;    /* Force greedy */
4733          possessive_quantifier = TRUE;          possessive_quantifier = TRUE;
# Line 4750  for (;; ptr++) Line 4807  for (;; ptr++)
4807          we have to insert the character for the previous code. For a repeated          we have to insert the character for the previous code. For a repeated
4808          Unicode property match, there are two extra bytes that define the          Unicode property match, there are two extra bytes that define the
4809          required property. In UTF-8 mode, long characters have their length in          required property. In UTF-8 mode, long characters have their length in
4810          c, with the 0x80 bit as a flag. */          c, with the UTF_LENGTH bit as a flag. */
4811    
4812          if (repeat_max < 0)          if (repeat_max < 0)
4813            {            {
4814  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4815            if (utf8 && c >= 128)            if (utf && (c & UTF_LENGTH) != 0)
4816              {              {
4817              memcpy(code, utf8_char, c & 7);              memcpy(code, utf_chars, IN_UCHARS(c & 7));
4818              code += c & 7;              code += c & 7;
4819              }              }
4820            else            else
# Line 4779  for (;; ptr++) Line 4836  for (;; ptr++)
4836    
4837          else if (repeat_max != repeat_min)          else if (repeat_max != repeat_min)
4838            {            {
4839  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4840            if (utf8 && c >= 128)            if (utf && (c & UTF_LENGTH) != 0)
4841              {              {
4842              memcpy(code, utf8_char, c & 7);              memcpy(code, utf_chars, IN_UCHARS(c & 7));
4843              code += c & 7;              code += c & 7;
4844              }              }
4845            else            else
# Line 4809  for (;; ptr++) Line 4866  for (;; ptr++)
4866    
4867        /* The character or character type itself comes last in all cases. */        /* The character or character type itself comes last in all cases. */
4868    
4869  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
4870        if (utf8 && c >= 128)        if (utf && (c & UTF_LENGTH) != 0)
4871          {          {
4872          memcpy(code, utf8_char, c & 7);          memcpy(code, utf_chars, IN_UCHARS(c & 7));
4873          code += c & 7;          code += c & 7;
4874          }          }
4875        else        else
# Line 4836  for (;; ptr++) Line 4893  for (;; ptr++)
4893    
4894      else if (*previous == OP_CLASS ||      else if (*previous == OP_CLASS ||
4895               *previous == OP_NCLASS ||               *previous == OP_NCLASS ||
4896  #if defined SUPPORT_UTF8 || !defined COMPILE_PCRE8  #if defined SUPPORT_UTF || !defined COMPILE_PCRE8
4897               *previous == OP_XCLASS ||               *previous == OP_XCLASS ||
4898  #endif  #endif
4899               *previous == OP_REF ||               *previous == OP_REF ||
# Line 4939  for (;; ptr++) Line 4996  for (;; ptr++)
4996          if (repeat_max <= 1)    /* Covers 0, 1, and unlimited */          if (repeat_max <= 1)    /* Covers 0, 1, and unlimited */
4997            {            {
4998            *code = OP_END;            *code = OP_END;
4999            adjust_recurse(previous, 1, utf8, cd, save_hwm);            adjust_recurse(previous, 1, utf, cd, save_hwm);
5000            memmove(previous + 1, previous, IN_UCHARS(len));            memmove(previous + 1, previous, IN_UCHARS(len));
5001            code++;            code++;
5002            if (repeat_max == 0)            if (repeat_max == 0)
# Line 4963  for (;; ptr++) Line 5020  for (;; ptr++)
5020            {            {
5021            int offset;            int offset;
5022            *code = OP_END;            *code = OP_END;
5023            adjust_recurse(previous, 2 + LINK_SIZE, utf8, cd, save_hwm);            adjust_recurse(previous, 2 + LINK_SIZE, utf, cd, save_hwm);
5024            memmove(previous + 2 + LINK_SIZE, previous, IN_UCHARS(len));            memmove(previous + 2 + LINK_SIZE, previous, IN_UCHARS(len));
5025            code += 2 + LINK_SIZE;            code += 2 + LINK_SIZE;
5026            *previous++ = OP_BRAZERO + repeat_type;            *previous++ = OP_BRAZERO + repeat_type;
# Line 5165  for (;; ptr++) Line 5222  for (;; ptr++)
5222              pcre_uchar *scode = bracode;              pcre_uchar *scode = bracode;
5223              do              do
5224                {                {
5225                if (could_be_empty_branch(scode, ketcode, utf8, cd))                if (could_be_empty_branch(scode, ketcode, utf, cd))
5226                  {                  {
5227                  *bracode += OP_SBRA - OP_BRA;                  *bracode += OP_SBRA - OP_BRA;
5228                  break;                  break;
# Line 5188  for (;; ptr++) Line 5245  for (;; ptr++)
5245                {                {
5246                int nlen = (int)(code - bracode);                int nlen = (int)(code - bracode);
5247                *code = OP_END;                *code = OP_END;
5248                adjust_recurse(bracode, 1 + LINK_SIZE, utf8, cd, save_hwm);                adjust_recurse(bracode, 1 + LINK_SIZE, utf, cd, save_hwm);
5249                memmove(bracode + 1 + LINK_SIZE, bracode, IN_UCHARS(nlen));                memmove(bracode + 1 + LINK_SIZE, bracode, IN_UCHARS(nlen));
5250                code += 1 + LINK_SIZE;                code += 1 + LINK_SIZE;
5251                nlen += 1 + LINK_SIZE;                nlen += 1 + LINK_SIZE;
# Line 5265  for (;; ptr++) Line 5322  for (;; ptr++)
5322        else if (*tempcode == OP_EXACT || *tempcode == OP_NOTEXACT)        else if (*tempcode == OP_EXACT || *tempcode == OP_NOTEXACT)
5323          {          {
5324          tempcode += PRIV(OP_lengths)[*tempcode];          tempcode += PRIV(OP_lengths)[*tempcode];
5325  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
5326          if (utf8 && tempcode[-1] >= 0xc0)          if (utf && HAS_EXTRALEN(tempcode[-1]))
5327            tempcode += PRIV(utf8_table4)[tempcode[-1] & 0x3f];            tempcode += GET_EXTRALEN(tempcode[-1]);
5328  #endif  #endif
5329          }          }
5330    
# Line 5304  for (;; ptr++) Line 5361  for (;; ptr++)
5361    
5362          default:          default:
5363          *code = OP_END;          *code = OP_END;
5364          adjust_recurse(tempcode, 1 + LINK_SIZE, utf8, cd, save_hwm);          adjust_recurse(tempcode, 1 + LINK_SIZE, utf, cd, save_hwm);
5365          memmove(tempcode + 1 + LINK_SIZE, tempcode, IN_UCHARS(len));          memmove(tempcode + 1 + LINK_SIZE, tempcode, IN_UCHARS(len));
5366          code += 1 + LINK_SIZE;          code += 1 + LINK_SIZE;
5367          len += 1 + LINK_SIZE;          len += 1 + LINK_SIZE;
# Line 5613  for (;; ptr++) Line 5670  for (;; ptr++)
5670          /* Search the pattern for a forward reference */          /* Search the pattern for a forward reference */
5671    
5672          else if ((i = find_parens(cd, name, namelen,          else if ((i = find_parens(cd, name, namelen,
5673                          (options & PCRE_EXTENDED) != 0, utf8)) > 0)                          (options & PCRE_EXTENDED) != 0, utf)) > 0)
5674            {            {
5675            PUT2(code, 2+LINK_SIZE, i);            PUT2(code, 2+LINK_SIZE, i);
5676            code[1+LINK_SIZE]++;            code[1+LINK_SIZE]++;
# Line 5803  for (;; ptr++) Line 5860  for (;; ptr++)
5860                *errorcodeptr = ERR49;                *errorcodeptr = ERR49;
5861                goto FAILED;                goto FAILED;
5862                }                }
5863              if (namelen + 3 > cd->name_entry_size)              if (namelen + IMM2_SIZE + 1 > cd->name_entry_size)
5864                {                {
5865                cd->name_entry_size = namelen + 3;                cd->name_entry_size = namelen + IMM2_SIZE + 1;
5866                if (namelen > MAX_NAME_SIZE)                if (namelen > MAX_NAME_SIZE)
5867                  {                  {
5868                  *errorcodeptr = ERR48;                  *errorcodeptr = ERR48;
# Line 5834  for (;; ptr++) Line 5891  for (;; ptr++)
5891    
5892              for (i = 0; i < cd->names_found; i++)              for (i = 0; i < cd->names_found; i++)
5893                {                {
5894                int crc = memcmp(name, slot+2, namelen);                int crc = memcmp(name, slot+IMM2_SIZE, IN_UCHARS(namelen));
5895                if (crc == 0)                if (crc == 0)
5896                  {                  {
5897                  if (slot[2+namelen] == 0)                  if (slot[IMM2_SIZE+namelen] == 0)
5898                    {                    {
5899                    if (GET2(slot, 0) != cd->bracount + 1 &&                    if (GET2(slot, 0) != cd->bracount + 1 &&
5900                        (options & PCRE_DUPNAMES) == 0)                        (options & PCRE_DUPNAMES) == 0)
# Line 5889  for (;; ptr++) Line 5946  for (;; ptr++)
5946                }                }
5947    
5948              PUT2(slot, 0, cd->bracount + 1);              PUT2(slot, 0, cd->bracount + 1);
5949              memcpy(slot + 2, name, IN_UCHARS(namelen));              memcpy(slot + IMM2_SIZE, name, IN_UCHARS(namelen));
5950              slot[2 + namelen] = 0;              slot[IMM2_SIZE + namelen] = 0;
5951              }              }
5952            }            }
5953    
# Line 5958  for (;; ptr++) Line 6015  for (;; ptr++)
6015            temp = cd->end_pattern;            temp = cd->end_pattern;
6016            cd->end_pattern = ptr;            cd->end_pattern = ptr;
6017            recno = find_parens(cd, name, namelen,            recno = find_parens(cd, name, namelen,
6018              (options & PCRE_EXTENDED) != 0, utf8);              (options & PCRE_EXTENDED) != 0, utf);
6019            cd->end_pattern = temp;            cd->end_pattern = temp;
6020            if (recno < 0) recno = 0;    /* Forward ref; set dummy number */            if (recno < 0) recno = 0;    /* Forward ref; set dummy number */
6021            }            }
# Line 5974  for (;; ptr++) Line 6031  for (;; ptr++)
6031            for (i = 0; i < cd->names_found; i++)            for (i = 0; i < cd->names_found; i++)
6032              {              {
6033              if (STRNCMP_UC_UC(name, slot+IMM2_SIZE, namelen) == 0 &&              if (STRNCMP_UC_UC(name, slot+IMM2_SIZE, namelen) == 0 &&
6034                  slot[2+namelen] == 0)                  slot[IMM2_SIZE+namelen] == 0)
6035                break;                break;
6036              slot += cd->name_entry_size;              slot += cd->name_entry_size;
6037              }              }
# Line 5985  for (;; ptr++) Line 6042  for (;; ptr++)
6042              }              }
6043            else if ((recno =                /* Forward back reference */            else if ((recno =                /* Forward back reference */
6044                      find_parens(cd, name, namelen,                      find_parens(cd, name, namelen,
6045                        (options & PCRE_EXTENDED) != 0, utf8)) <= 0)                        (options & PCRE_EXTENDED) != 0, utf)) <= 0)
6046              {              {
6047              *errorcodeptr = ERR15;              *errorcodeptr = ERR15;
6048              goto FAILED;              goto FAILED;
# Line 6089  for (;; ptr++) Line 6146  for (;; ptr++)
6146              {              {
6147              *code = OP_END;              *code = OP_END;
6148              if (recno != 0)              if (recno != 0)
6149                called = PRIV(find_bracket)(cd->start_code, utf8, recno);                called = PRIV(find_bracket)(cd->start_code, utf, recno);
6150    
6151              /* Forward reference */              /* Forward reference */
6152    
6153              if (called == NULL)              if (called == NULL)
6154                {                {
6155                if (find_parens(cd, NULL, recno,                if (find_parens(cd, NULL, recno,
6156                      (options & PCRE_EXTENDED) != 0, utf8) < 0)                      (options & PCRE_EXTENDED) != 0, utf) < 0)
6157                  {                  {
6158                  *errorcodeptr = ERR15;                  *errorcodeptr = ERR15;
6159                  goto FAILED;                  goto FAILED;
# Line 6120  for (;; ptr++) Line 6177  for (;; ptr++)
6177              conditional subpatterns will be picked up then. */              conditional subpatterns will be picked up then. */
6178    
6179              else if (GET(called, 1) == 0 && cond_depth <= 0 &&              else if (GET(called, 1) == 0 && cond_depth <= 0 &&
6180                       could_be_empty(called, code, bcptr, utf8, cd))                       could_be_empty(called, code, bcptr, utf, cd))
6181                {                {
6182                *errorcodeptr = ERR40;                *errorcodeptr = ERR40;
6183                goto FAILED;                goto FAILED;
# Line 6618  for (;; ptr++) Line 6675  for (;; ptr++)
6675    
6676            {            {
6677            previous = (-c > ESC_b && -c < ESC_Z)? code : NULL;            previous = (-c > ESC_b && -c < ESC_Z)? code : NULL;
6678            *code++ = (!utf8 && c == -ESC_C)? OP_ALLANY : -c;            *code++ = (!utf && c == -ESC_C)? OP_ALLANY : -c;
6679            }            }
6680          }          }
6681        continue;        continue;
# Line 6628  for (;; ptr++) Line 6685  for (;; ptr++)
6685      a value > 127. We set its representation in the length/buffer, and then      a value > 127. We set its representation in the length/buffer, and then
6686      handle it as a data character. */      handle it as a data character. */
6687    
6688  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
6689      if (utf8 && c > 127)      if (utf && c > 127)
6690        mclength = PRIV(ord2utf8)(c, mcbuffer);        mclength = PRIV(ord2utf)(c, mcbuffer);
6691      else      else
6692  #endif  #endif
6693    
# Line 6651  for (;; ptr++) Line 6708  for (;; ptr++)
6708      mclength = 1;      mclength = 1;
6709      mcbuffer[0] = c;      mcbuffer[0] = c;
6710    
6711  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
6712      if (utf8 && c >= 0xc0)      if (utf && HAS_EXTRALEN(c))
6713        {        ACROSSCHAR(TRUE, ptr[1], mcbuffer[mclength++] = *(++ptr));
       while ((ptr[1] & 0xc0) == 0x80)  
         mcbuffer[mclength++] = *(++ptr);  
       }  
6714  #endif  #endif
6715    
6716      /* At this point we have the character's bytes in mcbuffer, and the length      /* At this point we have the character's bytes in mcbuffer, and the length
# Line 7360  pcre_int32 firstchar, reqchar; Line 7414  pcre_int32 firstchar, reqchar;
7414  int newline;  int newline;
7415  int errorcode = 0;  int errorcode = 0;
7416  int skipatstart = 0;  int skipatstart = 0;
7417  BOOL utf8;  BOOL utf;
7418  size_t size;  size_t size;
7419  pcre_uchar *code;  pcre_uchar *code;
7420  const pcre_uchar *codestart;  const pcre_uchar *codestart;
# Line 7428  while (ptr[skipatstart] == CHAR_LEFT_PAR Line 7482  while (ptr[skipatstart] == CHAR_LEFT_PAR
7482    int newnl = 0;    int newnl = 0;
7483    int newbsr = 0;    int newbsr = 0;
7484    
7485    if (STRNCMP_UC_C8(ptr+skipatstart+2, STRING_UTF8_RIGHTPAR, 5) == 0)  #ifdef COMPILE_PCRE8
7486      if (STRNCMP_UC_C8(ptr+skipatstart+2, STRING_UTF_RIGHTPAR, 5) == 0)
7487      { skipatstart += 7; options |= PCRE_UTF8; continue; }      { skipatstart += 7; options |= PCRE_UTF8; continue; }
7488    #endif
7489    #ifdef COMPILE_PCRE16
7490      if (STRNCMP_UC_C8(ptr+skipatstart+2, STRING_UTF_RIGHTPAR, 6) == 0)
7491        { skipatstart += 8; options |= PCRE_UTF16; continue; }
7492    #endif
7493    else if (STRNCMP_UC_C8(ptr+skipatstart+2, STRING_UCP_RIGHTPAR, 4) == 0)    else if (STRNCMP_UC_C8(ptr+skipatstart+2, STRING_UCP_RIGHTPAR, 4) == 0)
7494      { skipatstart += 6; options |= PCRE_UCP; continue; }      { skipatstart += 6; options |= PCRE_UCP; continue; }
7495    else if (STRNCMP_UC_C8(ptr+skipatstart+2, STRING_NO_START_OPT_RIGHTPAR, 13) == 0)    else if (STRNCMP_UC_C8(ptr+skipatstart+2, STRING_NO_START_OPT_RIGHTPAR, 13) == 0)
# Line 7458  while (ptr[skipatstart] == CHAR_LEFT_PAR Line 7518  while (ptr[skipatstart] == CHAR_LEFT_PAR
7518    else break;    else break;
7519    }    }
7520    
7521  utf8 = (options & PCRE_UTF8) != 0;  /* PCRE_UTF16 has the same value as PCRE_UTF8. */
7522    utf = (options & PCRE_UTF8) != 0;
7523    
7524  /* Can't support UTF8 unless PCRE has been compiled to include the code. The  /* Can't support UTF unless PCRE has been compiled to include the code. The
7525  return of an error code from PRIV(valid_utf8)() is a new feature, introduced in  return of an error code from PRIV(valid_utf)() is a new feature, introduced in
7526  release 8.13. It is passed back from pcre_[dfa_]exec(), but at the moment is  release 8.13. It is passed back from pcre_[dfa_]exec(), but at the moment is
7527  not used here. */  not used here. */
7528    
7529  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF
7530  if (utf8 && (options & PCRE_NO_UTF8_CHECK) == 0 &&  if (utf && (options & PCRE_NO_UTF8_CHECK) == 0 &&
7531       (errorcode = PRIV(valid_utf8)((PCRE_PUCHAR)pattern, -1, erroroffset)) != 0)       (errorcode = PRIV(valid_utf)((PCRE_PUCHAR)pattern, -1, erroroffset)) != 0)
7532    {    {
7533    errorcode = ERR44;    errorcode = ERR44;
7534    goto PCRE_EARLY_ERROR_RETURN2;    goto PCRE_EARLY_ERROR_RETURN2;
7535    }    }
7536  #else  #else
7537  if (utf8)  if (utf)
7538    {    {
7539    errorcode = ERR32;    errorcode = ERR32;
7540    goto PCRE_EARLY_ERROR_RETURN;    goto PCRE_EARLY_ERROR_RETURN;
# Line 7602  externally provided function. Integer ov Line 7663  externally provided function. Integer ov
7663  because nowadays we limit the maximum value of cd->names_found and  because nowadays we limit the maximum value of cd->names_found and
7664  cd->name_entry_size. */  cd->name_entry_size. */
7665    
7666  size = sizeof(real_pcre) + (length + cd->names_found * (cd->name_entry_size + 3)) * sizeof(pcre_uchar);  size = sizeof(real_pcre) + (length + cd->names_found * cd->name_entry_size) * sizeof(pcre_uchar);
7667  re = (real_pcre *)(pcre_malloc)(size);  re = (real_pcre *)(pcre_malloc)(size);
7668    
7669  if (re == NULL)  if (re == NULL)
# Line 7662  code = (pcre_uchar *)codestart; Line 7723  code = (pcre_uchar *)codestart;
7723    &firstchar, &reqchar, NULL, cd, NULL);    &firstchar, &reqchar, NULL, cd, NULL);
7724  re->top_bracket = cd->bracount;  re->top_bracket = cd->bracount;
7725  re->top_backref = cd->top_backref;  re->top_backref = cd->top_backref;
7726  re->flags = cd->external_flags;  re->flags = cd->external_flags | PCRE_MODE;
7727    
7728  if (cd->had_accept) reqchar = REQ_NONE;   /* Must disable after (*ACCEPT) */  if (cd->had_accept) reqchar = REQ_NONE;   /* Must disable after (*ACCEPT) */
7729    
# Line 7688  while (errorcode == 0 && cd->hwm > cwork Line 7749  while (errorcode == 0 && cd->hwm > cwork
7749    cd->hwm -= LINK_SIZE;    cd->hwm -= LINK_SIZE;
7750    offset = GET(cd->hwm, 0);    offset = GET(cd->hwm, 0);
7751    recno = GET(codestart, offset);    recno = GET(codestart, offset);
7752    groupptr = PRIV(find_bracket)(codestart, utf8, recno);    groupptr = PRIV(find_bracket)(codestart, utf, recno);
7753    if (groupptr == NULL) errorcode = ERR53;    if (groupptr == NULL) errorcode = ERR53;
7754      else PUT(((pcre_uchar *)codestart), offset, (int)(groupptr - codestart));      else PUT(((pcre_uchar *)codestart), offset, (int)(groupptr - codestart));
7755    }    }
# Line 7715  if (cd->check_lookbehind) Line 7776  if (cd->check_lookbehind)
7776    of zero, but that is a pathological case, and it does no harm.) When we find    of zero, but that is a pathological case, and it does no harm.) When we find
7777    one, we temporarily terminate the branch it is in while we scan it. */    one, we temporarily terminate the branch it is in while we scan it. */
7778    
7779    for (cc = (pcre_uchar *)PRIV(find_bracket)(codestart, utf8, -1);    for (cc = (pcre_uchar *)PRIV(find_bracket)(codestart, utf, -1);
7780         cc != NULL;         cc != NULL;
7781         cc = (pcre_uchar *)PRIV(find_bracket)(cc, utf8, -1))         cc = (pcre_uchar *)PRIV(find_bracket)(cc, utf, -1))
7782      {      {
7783      if (GET(cc, 1) == 0)      if (GET(cc, 1) == 0)
7784        {        {
# Line 7781  if ((re->options & PCRE_ANCHORED) == 0) Line 7842  if ((re->options & PCRE_ANCHORED) == 0)
7842        re->first_char = firstchar & 0xffff;        re->first_char = firstchar & 0xffff;
7843  #endif  #endif
7844  #endif  #endif
7845        if ((firstchar & REQ_CASELESS) != 0 && MAX_255(re->first_char)        if ((firstchar & REQ_CASELESS) != 0)
7846          && cd->fcc[re->first_char] != re->first_char)          {
7847          re->flags |= PCRE_FCH_CASELESS;  #if defined SUPPORT_UCP && !(defined COMPILE_PCRE8)
7848            /* We ignore non-ASCII first chars in 8 bit mode. */
7849            if (utf)
7850              {
7851              if (re->first_char < 128)
7852                {
7853                if (cd->fcc[re->first_char] != re->first_char)
7854                  re->flags |= PCRE_FCH_CASELESS;
7855                }
7856              else if (UCD_OTHERCASE(re->first_char) != re->first_char)
7857                re->flags |= PCRE_FCH_CASELESS;
7858              }
7859            else
7860    #endif
7861            if (MAX_255(re->first_char)
7862                && cd->fcc[re->first_char] != re->first_char)
7863              re->flags |= PCRE_FCH_CASELESS;
7864            }
7865    
7866        re->flags |= PCRE_FIRSTSET;        re->flags |= PCRE_FIRSTSET;
7867        }        }
# Line 7806  if (reqchar >= 0 && Line 7884  if (reqchar >= 0 &&
7884    re->req_char = reqchar & 0xffff;    re->req_char = reqchar & 0xffff;
7885  #endif  #endif
7886  #endif  #endif
7887    if ((reqchar & REQ_CASELESS) != 0 && MAX_255(re->req_char)    if ((reqchar & REQ_CASELESS) != 0)
7888      && cd->fcc[re->req_char] != re->req_char)      {
7889      re->flags |= PCRE_RCH_CASELESS;  #if defined SUPPORT_UCP && !(defined COMPILE_PCRE8)
7890        /* We ignore non-ASCII first chars in 8 bit mode. */
7891        if (utf)
7892          {
7893          if (re->req_char < 128)
7894            {
7895            if (cd->fcc[re->req_char] != re->req_char)
7896              re->flags |= PCRE_RCH_CASELESS;
7897            }
7898          else if (UCD_OTHERCASE(re->req_char) != re->req_char)
7899            re->flags |= PCRE_RCH_CASELESS;
7900          }
7901        else
7902    #endif
7903        if (MAX_255(re->req_char) && cd->fcc[re->req_char] != re->req_char)
7904          re->flags |= PCRE_RCH_CASELESS;
7905        }
7906    
7907    re->flags |= PCRE_REQCHSET;    re->flags |= PCRE_REQCHSET;
7908    }    }

Legend:
Removed from v.774  
changed lines
  Added in v.795

  ViewVC Help
Powered by ViewVC 1.1.5