* lib/rexml/parsers/baseparser.rb (LETTER, DIGIT):

always use POSIX charclass.

* lib/rexml/parsers/baseparser.rb (NAMECHAR):
  remove duplicated range.

* lib/rexml/xmltokens.rb (NCNAME_STR, NAMECHAR): ditto.

* lib/rexml/parsers/xpathparser.rb (PathExpr): ditto.

* lib/rexml/text.rb (REXML::Text#initialize):
  initialize @parent.

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@27228 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

ChangeLog

@@ -1,3 +1,18 @@
+Tue Apr  6 05:59:12 2010  NARUSE, Yui  <naruse@ruby-lang.org>
+
+	* lib/rexml/parsers/baseparser.rb (LETTER, DIGIT):
+	  always use POSIX charclass.
+
+	* lib/rexml/parsers/baseparser.rb (NAMECHAR):
+	  remove duplicated range.
+
+	* lib/rexml/xmltokens.rb (NCNAME_STR, NAMECHAR): ditto.
+
+	* lib/rexml/parsers/xpathparser.rb (PathExpr): ditto. 
+
+	* lib/rexml/text.rb (REXML::Text#initialize):
+	  initialize @parent.
+
 Mon Apr  5 19:54:58 2010  NARUSE, Yui  <naruse@ruby-lang.org>
 
 	* re.c (make_regexp): use onig_new_with_source to keep

lib/rexml/parsers/baseparser.rb

@@ -25,15 +25,8 @@ module REXML
     #
     # Nat Price gave me some good ideas for the API.
     class BaseParser
-      if String.method_defined? :encode
-        # Oniguruma / POSIX [understands unicode]
-        LETTER = '[[:alpha:]]'
-        DIGIT = '[[:digit:]]'
-      else
-        # Ruby < 1.9 [doesn't understand unicode]
-        LETTER = 'a-zA-Z'
-        DIGIT = '\d'
-      end
+      LETTER = '[:alpha:]'
+      DIGIT = '[:digit:]'
 
       COMBININGCHAR = '' # TODO
       EXTENDER = ''      # TODO
@@ -42,7 +35,7 @@ module REXML
       NAME_STR= "(?:(#{NCNAME_STR}):)?(#{NCNAME_STR})"
       UNAME_STR= "(?:#{NCNAME_STR}:)?#{NCNAME_STR}"
 
-      NAMECHAR = '[\-\w\d\.:]'
+      NAMECHAR = '[\-\w:]'
       NAME = "([\\w:]#{NAMECHAR}*)"
       NMTOKEN = "(?:#{NAMECHAR})+"
       NMTOKENS = "#{NMTOKEN}(\\s+#{NMTOKEN})*"

lib/rexml/parsers/xpathparser.rb

@@ -551,7 +551,7 @@ module REXML
           end
         end
         #puts "BEFORE WITH '#{rest}'"
-        rest = LocationPath(rest, n) if rest =~ /\A[\/\.\@\[\w_*]/
+        rest = LocationPath(rest, n) if rest =~ /\A[\/\.\@\[\w*]/
         parsed.concat(n)
         return rest
       end

lib/rexml/text.rb

@@ -93,12 +93,11 @@ module REXML
       entity_filter=nil, illegal=NEEDS_A_SECOND_CHECK )
 
       @raw = false
+      @parent = nil
 
       if parent
         super( parent )
         @raw = parent.raw
-      else
-        @parent = nil
       end
 
       @raw = raw unless raw.nil?

lib/rexml/xmltokens.rb

@@ -2,10 +2,10 @@ module REXML
   # Defines a number of tokens used for parsing XML.  Not for general
   # consumption.
   module XMLTokens
-    NCNAME_STR= '[\w:][\-\w\d.]*'
+    NCNAME_STR= '[\w:][\-\w.]*'
     NAME_STR= "(?:#{NCNAME_STR}:)?#{NCNAME_STR}"
 
-    NAMECHAR = '[\-\w\d\.:]'
+    NAMECHAR = '[\-\w\.:]'
     NAME = "([\\w:]#{NAMECHAR}*)"
     NMTOKEN = "(?:#{NAMECHAR})+"
     NMTOKENS = "#{NMTOKEN}(\\s+#{NMTOKEN})*"