/[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 216 by ph10, Wed Aug 15 14:35:57 2007 UTC revision 218 by ph10, Thu Aug 16 10:13:23 2007 UTC
# Line 1106  for (;;) Line 1106  for (;;)
1106    {    {
1107    int d;    int d;
1108    register int op = *cc;    register int op = *cc;
   
1109    switch (op)    switch (op)
1110      {      {
1111      case OP_CBRA:      case OP_CBRA:
# Line 1195  for (;;) Line 1194  for (;;)
1194    
1195      case OP_TYPEEXACT:      case OP_TYPEEXACT:
1196      branchlength += GET2(cc,1);      branchlength += GET2(cc,1);
1197        if (cc[3] == OP_PROP || cc[3] == OP_NOTPROP) cc += 2;
1198      cc += 4;      cc += 4;
1199      break;      break;
1200    
# Line 1303  for (;;) Line 1303  for (;;)
1303      code += _pcre_OP_lengths[c];      code += _pcre_OP_lengths[c];
1304      }      }
1305    
1306    /* In UTF-8 mode, opcodes that are followed by a character may be followed by    /* Otherwise, we can get the item's length from the table, except that for
1307    a multi-byte character. The length in the table is a minimum, so we have to    repeated character types, we have to test for \p and \P, which have an extra
1308    arrange to skip the extra bytes. */    two bytes of parameters. */
1309    
1310    else    else
1311      {      {
1312        switch(c)
1313          {
1314          case OP_TYPESTAR:
1315          case OP_TYPEMINSTAR:
1316          case OP_TYPEPLUS:
1317          case OP_TYPEMINPLUS:
1318          case OP_TYPEQUERY:
1319          case OP_TYPEMINQUERY:
1320          case OP_TYPEUPTO:
1321          case OP_TYPEMINUPTO:
1322          case OP_TYPEEXACT:
1323          case OP_TYPEPOSSTAR:
1324          case OP_TYPEPOSPLUS:
1325          case OP_TYPEPOSQUERY:
1326          case OP_TYPEPOSUPTO:
1327          if (code[1] == OP_PROP || code[1] == OP_NOTPROP) code += 2;
1328          break;
1329          }
1330    
1331        /* Add in the fixed length from the table */
1332    
1333      code += _pcre_OP_lengths[c];      code += _pcre_OP_lengths[c];
1334    
1335      /* In UTF-8 mode, opcodes that are followed by a character may be followed by
1336      a multi-byte character. The length in the table is a minimum, so we have to
1337      arrange to skip the extra bytes. */
1338    
1339  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
1340      if (utf8) switch(c)      if (utf8) switch(c)
1341        {        {
# Line 1360  for (;;) Line 1386  for (;;)
1386    register int c = *code;    register int c = *code;
1387    if (c == OP_END) return NULL;    if (c == OP_END) return NULL;
1388    if (c == OP_RECURSE) return code;    if (c == OP_RECURSE) return code;
1389    
1390    /* XCLASS is used for classes that cannot be represented just by a bit    /* XCLASS is used for classes that cannot be represented just by a bit
1391    map. This includes negated single high-valued characters. The length in    map. This includes negated single high-valued characters. The length in
1392    the table is zero; the actual length is stored in the compiled code. */    the table is zero; the actual length is stored in the compiled code. */
1393    
1394    if (c == OP_XCLASS) code += GET(code, 1);    if (c == OP_XCLASS) code += GET(code, 1);
1395    
1396    /* Otherwise, we get the item's length from the table. In UTF-8 mode, opcodes    /* Otherwise, we can get the item's length from the table, except that for
1397    that are followed by a character may be followed by a multi-byte character.    repeated character types, we have to test for \p and \P, which have an extra
1398    The length in the table is a minimum, so we have to arrange to skip the extra    two bytes of parameters. */
   bytes. */  
1399    
1400    else    else
1401      {      {
1402        switch(c)
1403          {
1404          case OP_TYPESTAR:
1405          case OP_TYPEMINSTAR:
1406          case OP_TYPEPLUS:
1407          case OP_TYPEMINPLUS:
1408          case OP_TYPEQUERY:
1409          case OP_TYPEMINQUERY:
1410          case OP_TYPEUPTO:
1411          case OP_TYPEMINUPTO:
1412          case OP_TYPEEXACT:
1413          case OP_TYPEPOSSTAR:
1414          case OP_TYPEPOSPLUS:
1415          case OP_TYPEPOSQUERY:
1416          case OP_TYPEPOSUPTO:
1417          if (code[1] == OP_PROP || code[1] == OP_NOTPROP) code += 2;
1418          break;
1419          }
1420    
1421        /* Add in the fixed length from the table */
1422    
1423      code += _pcre_OP_lengths[c];      code += _pcre_OP_lengths[c];
1424    
1425        /* In UTF-8 mode, opcodes that are followed by a character may be followed
1426        by a multi-byte character. The length in the table is a minimum, so we have
1427        to arrange to skip the extra bytes. */
1428    
1429  #ifdef SUPPORT_UTF8  #ifdef SUPPORT_UTF8
1430      if (utf8) switch(c)      if (utf8) switch(c)
1431        {        {
# Line 2671  for (;; ptr++) Line 2722  for (;; ptr++)
2722            else inescq = TRUE;            else inescq = TRUE;
2723            continue;            continue;
2724            }            }
2725            else if (-c == ESC_E) continue;  /* Ignore orphan \E */
2726    
2727          if (c < 0)          if (c < 0)
2728            {            {
# Line 4745  for (;; ptr++) Line 4797  for (;; ptr++)
4797        goto FAILED;        goto FAILED;
4798        }        }
4799    
4800      /* In the pre-compile phase, update the length by the length of the nested      /* In the pre-compile phase, update the length by the length of the group,
4801      group, less the brackets at either end. Then reduce the compiled code to      less the brackets at either end. Then reduce the compiled code to just a
4802      just the brackets so that it doesn't use much memory if it is duplicated by      set of non-capturing brackets so that it doesn't use much memory if it is
4803      a quantifier. */      duplicated by a quantifier.*/
4804    
4805      if (lengthptr != NULL)      if (lengthptr != NULL)
4806        {        {
# Line 4758  for (;; ptr++) Line 4810  for (;; ptr++)
4810          goto FAILED;          goto FAILED;
4811          }          }
4812        *lengthptr += length_prevgroup - 2 - 2*LINK_SIZE;        *lengthptr += length_prevgroup - 2 - 2*LINK_SIZE;
4813        code++;        *code++ = OP_BRA;
4814        PUTINC(code, 0, 1 + LINK_SIZE);        PUTINC(code, 0, 1 + LINK_SIZE);
4815        *code++ = OP_KET;        *code++ = OP_KET;
4816        PUTINC(code, 0, 1 + LINK_SIZE);        PUTINC(code, 0, 1 + LINK_SIZE);
4817          break;    /* No need to waste time with special character handling */
4818        }        }
4819    
4820      /* Otherwise update the main code pointer to the end of the group. */      /* Otherwise update the main code pointer to the end of the group. */
4821    
4822      else code = tempcode;      code = tempcode;
4823    
4824      /* For a DEFINE group, required and first character settings are not      /* For a DEFINE group, required and first character settings are not
4825      relevant. */      relevant. */

Legend:
Removed from v.216  
changed lines
  Added in v.218

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12