/[pcre]/code/trunk/pcre_compile.c

Diff of /code/trunk/pcre_compile.c

Parent Directory | Revision Log | View Patch Patch

-revision 150 by ph10,
Tue Apr 17 08:22:40 2007 UTC
+revision 171 by ph10,
Mon Jun  4 14:28:58 2007 UTC
 Line 87 
 static const short int escapes[] = {
 ,      0,      0,      0,      0,      0,      0,      0,   /* 0 - 7 */
 ,      0,    ':',    ';',    '<',    '=',    '>',    '?',   /* 8 - ? */
     '@', -ESC_A, -ESC_B, -ESC_C, -ESC_D, -ESC_E,      0, -ESC_G,   /* @ - G */
-,      0,      0,      0,      0,      0,      0,      0,   /* H - O */
+,      0,      0, -ESC_K,      0,      0,      0,      0,   /* H - O */
  -ESC_P, -ESC_Q, -ESC_R, -ESC_S,      0,      0,      0, -ESC_W,   /* P - W */
  -ESC_X,      0, -ESC_Z,    '[',   '\\',    ']',    '^',    '_',   /* X - _ */
     '`',      7, -ESC_b,      0, -ESC_d,  ESC_e,  ESC_f,      0,   /* ` - g */
 Line 208 
 static const char *error_texts[] = {
    "malformed number or name after (?(",
    "conditional group contains more than two branches",
    "assertion expected after (?(",
-   "(?R or (?digits must be followed by )",
+   "(?R or (?[+-]digits must be followed by )",
    /* 30 */
    "unknown POSIX class name",
    "POSIX collating elements are not supported",
 Line 242 
 static const char *error_texts[] = {
    /* 55 */
    "repeating a DEFINE group is not allowed",
    "inconsistent NEWLINE options",
-   "\\g is not followed by an (optionally braced) non-zero number"
+   "\\g is not followed by a braced name or an optionally braced non-zero number",
+   "(?+ or (?- or (?(+ or (?(- must be followed by a non-zero number"
  };
-Line 452 
 else
+Line 453 
 else
      /* \g must be followed by a number, either plain or braced. If positive, it
      is an absolute backreference. If negative, it is a relative backreference.
-     This is a Perl 5.10 feature. */
+     This is a Perl 5.10 feature. Perl 5.10 also supports \g{name} as a
+     reference to a named group. This is part of Perl's movement towards a
+     unified syntax for back references. As this is synonymous with \k{name}, we
+     fudge it up by pretending it really was \k. */
      case 'g':
      if (ptr[1] == '{')
        {
+       const uschar *p;
+       for (p = ptr+2; *p != 0 && *p != '}'; p++)
+         if (*p != '-' && (digitab[*p] & ctype_digit) == 0) break;
+       if (*p != 0 && *p != '}')
+         {
+         c = -ESC_k;
+         break;
+         }
        braced = TRUE;
        ptr++;
        }
-Line 1369 
 for (code = first_significant_code(code
+Line 1381 
 for (code = first_significant_code(code
    const uschar *ccode;
    c = *code;
+   /* Groups with zero repeats can of course be empty; skip them. */
+   if (c == OP_BRAZERO || c == OP_BRAMINZERO)
+     {
+     do code += GET(code, 1); while (*code == OP_ALT);
+     c = *code;
+     continue;
+     }
+   /* For other groups, scan the branches. */
    if (c == OP_BRA || c == OP_CBRA || c == OP_ONCE)
      {
      BOOL empty_branch;
-Line 1386 
 for (code = first_significant_code(code
+Line 1409 
 for (code = first_significant_code(code
        }
      while (*code == OP_ALT);
      if (!empty_branch) return FALSE;   /* All branches are non-empty */
+     c = *code;
-     /* Move past the KET and fudge things so that the increment in the "for"
-     above has no effect. */
-     c = OP_END;
-     code += 1 + LINK_SIZE - _pcre_OP_lengths[c];
      continue;
      }
-Line 2095 
 for (;; ptr++)
+Line 2113 
 for (;; ptr++)
    int class_lastchar;
    int newoptions;
    int recno;
+   int refsign;
    int skipbytes;
    int subreqbyte;
    int subfirstbyte;
-Line 3621 
 for (;; ptr++)
+Line 3640 
 for (;; ptr++)
          code[1+LINK_SIZE] = OP_CREF;
          skipbytes = 3;
+         refsign = -1;
          /* Check for a test for recursion in a named group. */
-Line 3644 
 for (;; ptr++)
+Line 3664 
 for (;; ptr++)
            terminator = '\'';
            ptr++;
            }
-         else terminator = 0;
+         else
+           {
+           terminator = 0;
+           if (ptr[1] == '-' || ptr[1] == '+') refsign = *(++ptr);
+           }
          /* We now expect to read a name; any thing else is an error */
-Line 3680 
 for (;; ptr++)
+Line 3704 
 for (;; ptr++)
          if (lengthptr != NULL) break;
          /* In the real compile we do the work of looking for the actual
-         reference. */
+         reference. If the string started with "+" or "-" we require the rest to
+         be digits, in which case recno will be set. */
+         if (refsign > 0)
+           {
+           if (recno <= 0)
+             {
+             *errorcodeptr = ERR58;
+             goto FAILED;
+             }
+           if (refsign == '-')
+             {
+             recno = cd->bracount - recno + 1;
+             if (recno <= 0)
+               {
+               *errorcodeptr = ERR15;
+               goto FAILED;
+               }
+             }
+           else recno += cd->bracount;
+           PUT2(code, 2+LINK_SIZE, recno);
+           break;
+           }
+         /* Otherwise (did not start with "+" or "-"), start by looking for the
+         name. */
          slot = cd->name_table;
          for (i = 0; i < cd->names_found; i++)
            {
-Line 3999 
 for (;; ptr++)
+Line 4048 
 for (;; ptr++)
          /* ------------------------------------------------------------ */
+         case '-': case '+':
          case '0': case '1': case '2': case '3': case '4':   /* Recursion or */
          case '5': case '6': case '7': case '8': case '9':   /* subroutine */
            {
            const uschar *called;
+           if ((refsign = *ptr) == '+') ptr++;
+           else if (refsign == '-')
+             {
+             if ((digitab[ptr[1]] & ctype_digit) == 0)
+               goto OTHER_CHAR_AFTER_QUERY;
+             ptr++;
+             }
            recno = 0;
            while((digitab[*ptr] & ctype_digit) != 0)
              recno = recno * 10 + *ptr++ - '0';
            if (*ptr != ')')
              {
              *errorcodeptr = ERR29;
              goto FAILED;
              }
+           if (refsign == '-')
+             {
+             if (recno == 0)
+               {
+               *errorcodeptr = ERR58;
+               goto FAILED;
+               }
+             recno = cd->bracount - recno + 1;
+             if (recno <= 0)
+               {
+               *errorcodeptr = ERR15;
+               goto FAILED;
+               }
+             }
+           else if (refsign == '+')
+             {
+             if (recno == 0)
+               {
+               *errorcodeptr = ERR58;
+               goto FAILED;
+               }
+             recno += cd->bracount;
+             }
            /* Come here from code above that handles a named recursion */
-Line 4084 
 for (;; ptr++)
+Line 4168 
 for (;; ptr++)
          /* ------------------------------------------------------------ */
          default:              /* Other characters: check option setting */
+         OTHER_CHAR_AFTER_QUERY:
          set = unset = 0;
          optset = &set;
-Line 4396 
 for (;; ptr++)
+Line 4481 
 for (;; ptr++)
        zerofirstbyte = firstbyte;
        zeroreqbyte = reqbyte;
-       /* \k<name> or \k'name' is a back reference by name (Perl syntax) */
+       /* \k<name> or \k'name' is a back reference by name (Perl syntax).
+       We also support \k{name} (.NET syntax) */
-       if (-c == ESC_k && (ptr[1] == '<' || ptr[1] == '\''))
+       if (-c == ESC_k && (ptr[1] == '<' || ptr[1] == '\'' || ptr[1] == '{'))
          {
          is_recurse = FALSE;
-         terminator = (*(++ptr) == '<')? '>' : '\'';
+         terminator = (*(++ptr) == '<')? '>' : (*ptr == '\'')? '\'' : '}';
          goto NAMED_REF_OR_RECURSE;
          }

 Legend:



Removed from v.150
 


changed lines


 
Added in v.171
 Legend:



Removed from v.150
 


changed lines


 
Added in v.171
-Removed from v.150
+Added in v.171

	ViewVC Help
Powered by ViewVC 1.1.5