ruby-changes:3333
From: ko1@a...
Date: 1 Jan 2008 14:44:10 +0900
Subject: [ruby-changes:3333] matz - Ruby:r14826 (trunk): * lib/rexml: 1.9 patch from Sam Ruby mentioned in his blog:
matz 2008-01-01 14:43:50 +0900 (Tue, 01 Jan 2008)
New Revision: 14826
Modified files:
trunk/ChangeLog
trunk/lib/rexml/doctype.rb
trunk/lib/rexml/element.rb
trunk/lib/rexml/encoding.rb
trunk/lib/rexml/entity.rb
trunk/lib/rexml/formatters/pretty.rb
trunk/lib/rexml/functions.rb
trunk/lib/rexml/parsers/baseparser.rb
trunk/lib/rexml/parsers/sax2parser.rb
trunk/lib/rexml/parsers/treeparser.rb
trunk/lib/rexml/parsers/xpathparser.rb
trunk/lib/rexml/source.rb
trunk/lib/rexml/syncenumerator.rb
trunk/lib/rexml/text.rb
trunk/lib/rexml/validation/validation.rb
trunk/lib/rexml/xpath_parser.rb
Log:
* lib/rexml: 1.9 patch from Sam Ruby mentioned in his blog:
<http://intertwingly.net/blog/2007/12/31/Porting-REXML-to-Ruby-1-9>
[ruby-core:14639]
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/parsers/treeparser.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/formatters/pretty.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/text.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/validation/validation.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/parsers/xpathparser.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/ChangeLog?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/syncenumerator.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/entity.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/parsers/sax2parser.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/xpath_parser.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/doctype.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/parsers/baseparser.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/element.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/source.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/functions.rb?r1=14826&r2=14825&diff_format=u
http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/rexml/encoding.rb?r1=14826&r2=14825&diff_format=u
Index: ChangeLog
===================================================================
--- ChangeLog (revision 14825)
+++ ChangeLog (revision 14826)
@@ -1,3 +1,9 @@
+Tue Jan 1 14:41:56 2008 Yukihiro Matsumoto <matz@r...>
+
+ * lib/rexml: 1.9 patch from Sam Ruby mentioned in his blog:
+ <http://intertwingly.net/blog/2007/12/31/Porting-REXML-to-Ruby-1-9>
+ [ruby-core:14639]
+
Tue Jan 1 14:15:04 2008 Yukihiro Matsumoto <matz@r...>
* string.c (rb_str_substr): offset movement bug. a patch from
Index: lib/rexml/parsers/xpathparser.rb
===================================================================
--- lib/rexml/parsers/xpathparser.rb (revision 14825)
+++ lib/rexml/parsers/xpathparser.rb (revision 14826)
@@ -332,12 +332,12 @@
predicates << expr[1..-2] if expr
end
#puts "PREDICATES = #{predicates.inspect}"
- predicates.each{ |expr|
- #puts "ORING #{expr}"
+ predicates.each{ |pred|
+ #puts "ORING #{pred}"
preds = []
parsed << :predicate
parsed << preds
- OrExpr(expr, preds)
+ OrExpr(pred, preds)
}
#puts "PREDICATES = #{predicates.inspect}"
path
Index: lib/rexml/parsers/baseparser.rb
===================================================================
--- lib/rexml/parsers/baseparser.rb (revision 14825)
+++ lib/rexml/parsers/baseparser.rb (revision 14826)
@@ -242,6 +242,11 @@
@document_status = :after_doctype
@source.read if @source.buffer.size<2
md = @source.match(/\s*/um, true)
+ if @source.encoding == "UTF-8"
+ if @source.buffer.respond_to? :force_encoding
+ @source.buffer.force_encoding(Encoding::UTF_8)
+ end
+ end
end
end
if @document_status == :in_doctype
Index: lib/rexml/parsers/sax2parser.rb
===================================================================
--- lib/rexml/parsers/sax2parser.rb (revision 14825)
+++ lib/rexml/parsers/sax2parser.rb (revision 14826)
@@ -149,17 +149,26 @@
procs = get_procs( :end_prefix_mapping, event[1] )
listeners = get_listeners( :end_prefix_mapping, event[1] )
if procs or listeners
- namespace_mapping.each do |prefix, uri|
+ namespace_mapping.each do |ns_prefix, ns_uri|
# notify observers of namespaces
- procs.each { |ob| ob.call( prefix ) } if procs
- listeners.each { |ob| ob.end_prefix_mapping(prefix) } if listeners
+ procs.each { |ob| ob.call( ns_prefix ) } if procs
+ listeners.each { |ob| ob.end_prefix_mapping(ns_prefix) } if listeners
end
end
when :text
#normalized = @parser.normalize( event[1] )
#handle( :characters, normalized )
copy = event[1].clone
- @entities.each { |key, value| copy = copy.gsub("&#{key};", value) }
+
+ esub = proc { |match|
+ if @entities.has_key?($1)
+ @entities[$1].gsub(Text::REFERENCE, &esub)
+ else
+ match
+ end
+ }
+
+ copy.gsub!( Text::REFERENCE, &esub )
copy.gsub!( Text::NUMERICENTITY ) {|m|
m=$1
m = "0#{m}" if m[0] == ?x
Index: lib/rexml/parsers/treeparser.rb
===================================================================
--- lib/rexml/parsers/treeparser.rb (revision 14825)
+++ lib/rexml/parsers/treeparser.rb (revision 14826)
@@ -30,7 +30,10 @@
return
when :start_element
tag_stack.push(event[1])
- el = @build_context = @build_context.add_element( event[1], event[2] )
+ el = @build_context = @build_context.add_element( event[1] )
+ event[2].each do |key, value|
+ el.attributes[key]=Attribute.new(key,value,self)
+ end
when :end_element
tag_stack.pop
@build_context = @build_context.parent
Index: lib/rexml/validation/validation.rb
===================================================================
--- lib/rexml/validation/validation.rb (revision 14825)
+++ lib/rexml/validation/validation.rb (revision 14826)
@@ -33,8 +33,8 @@
sattr = [:start_attribute, nil]
eattr = [:end_attribute]
text = [:text, nil]
- k,v = event[2].find { |k,v|
- sattr[1] = k
+ k,v = event[2].find { |key,value|
+ sattr[1] = key
#puts "Looking for #{sattr.inspect}"
m = @current.next( sattr )
#puts "Got #{m.inspect}"
@@ -47,7 +47,7 @@
@current = m
else
#puts "Didn't get end"
- text[1] = v
+ text[1] = value
#puts "Looking for #{text.inspect}"
m = m.next( text )
#puts "Got #{m.inspect}"
Index: lib/rexml/element.rb
===================================================================
--- lib/rexml/element.rb (revision 14825)
+++ lib/rexml/element.rb (revision 14826)
@@ -296,7 +296,7 @@
raise "First argument must be either an element name, or an Element object" if element.nil?
el = @elements.add(element)
attrs.each do |key, value|
- el.attributes[key]=Attribute.new(key,value,self)
+ el.attributes[key]=value
end if attrs.kind_of? Hash
el
end
@@ -552,7 +552,11 @@
def attribute( name, namespace=nil )
prefix = nil
- prefix = namespaces.index(namespace) if namespace
+ if namespaces.respond_to? :key
+ prefix = namespaces.key(namespace) if namespace
+ else
+ prefix = namespaces.index(namespace) if namespace
+ end
prefix = nil if prefix == 'xmlns'
attributes.get_attribute( "#{prefix ? prefix + ':' : ''}#{name}" )
end
@@ -704,7 +708,6 @@
# A private helper method
def each_with_something( test, max=0, name=nil )
num = 0
- child=nil
@elements.each( name ){ |child|
yield child if test.call(child) and num += 1
return if max>0 and num == max
@@ -754,7 +757,6 @@
raise "index (#{index}) must be >= 1" if index < 1
name = literalize(name) if name
num = 0
- child = nil
@element.find { |child|
child.kind_of? Element and
(name.nil? ? true : child.has_name?( name )) and
@@ -1217,7 +1219,8 @@
def get_attribute_ns(namespace, name)
each_attribute() { |attribute|
if name == attribute.name &&
- namespace == attribute.namespace()
+ namespace == attribute.namespace() &&
+ ( !namespace.empty? || !attribute.fully_expanded_name.index(':') )
return attribute
end
}
Index: lib/rexml/source.rb
===================================================================
--- lib/rexml/source.rb (revision 14825)
+++ lib/rexml/source.rb (revision 14826)
@@ -59,6 +59,9 @@
@to_utf = true
else
@to_utf = false
+ if @buffer.respond_to? :force_encoding
+ @buffer.force_encoding Encoding::UTF_8
+ end
end
end
@@ -147,13 +150,13 @@
str = @source.read( 2 )
if encoding
self.encoding = encoding
- elsif 0xfe == str[0] && 0xff == str[1]
+ elsif str[0,2] == "\xfe\xff"
@line_break = "\000>"
- elsif 0xff == str[0] && 0xfe == str[1]
+ elsif str[0,2] == "\xff\xfe"
@line_break = ">\000"
- elsif 0xef == str[0] && 0xbb == str[1]
+ elsif str[0,2] == "\xef\xbb"
str += @source.read(1)
- str = '' if (0xbf == str[2])
+ str = '' if (str[2,1] == "\xBF")
@line_break = ">"
else
@line_break = ">"
@@ -193,6 +196,9 @@
str = @source.readline(@line_break)
str = decode(str) if @to_utf and str
@buffer << str
+ if not @to_utf and @buffer.respond_to? :force_encoding
+ @buffer.force_encoding Encoding::UTF_8
+ end
rescue Exception, NameError
@source = nil
end
Index: lib/rexml/doctype.rb
===================================================================
--- lib/rexml/doctype.rb (revision 14825)
+++ lib/rexml/doctype.rb (revision 14826)
@@ -117,7 +117,6 @@
unless @children.empty?
next_indent = indent + 1
output << ' ['
- child = nil # speed
@children.each { |child|
output << "\n"
f.write( child, output )
Index: lib/rexml/functions.rb
===================================================================
--- lib/rexml/functions.rb (revision 14825)
+++ lib/rexml/functions.rb (revision 14826)
@@ -256,9 +256,15 @@
end
}
- string(string).unpack('U*').collect { |c|
- if map.has_key? c then map[c] else c end
- }.compact.pack('U*')
+ if ''.respond_to? :chars
+ string(string).chars.collect { |c|
+ if map.has_key? c then map[c] else c end
+ }.compact.join
+ else
+ string(string).unpack('U*').collect { |c|
+ if map.has_key? c then map[c] else c end
+ }.compact.pack('U*')
+ end
end
# UNTESTED
Index: lib/rexml/entity.rb
===================================================================
--- lib/rexml/entity.rb (revision 14825)
+++ lib/rexml/entity.rb (revision 14826)
@@ -139,7 +139,7 @@
if @parent
matches.each do |entity_reference|
entity_value = @parent.entity( entity_reference[0] )
- rv.gsub!( /%#{entity_reference};/um, entity_value )
+ rv.gsub!( /%#{entity_reference.join};/um, entity_value )
end
end
return rv
Index: lib/rexml/syncenumerator.rb
===================================================================
--- lib/rexml/syncenumerator.rb (revision 14825)
+++ lib/rexml/syncenumerator.rb (revision 14826)
@@ -6,8 +6,7 @@
# Enumerable objects.
def initialize(*enums)
@gens = enums
- @biggest = @gens[0]
- @gens.each {|x| @biggest = x if x.size > @biggest.size }
+ @length = @gens.collect {|x| x.size }.max
end
# Returns the number of enumerated Enumerable objects, i.e. the size
@@ -24,8 +23,8 @@
# Enumerates rows of the Enumerable objects.
def each
- @biggest.zip( *@gens ) {|a|
- yield(*a[1..-1])
+ @length.times {|i|
+ yield @gens.collect {|x| x[i]}
}
self
end
Index: lib/rexml/text.rb
===================================================================
--- lib/rexml/text.rb (revision 14825)
+++ lib/rexml/text.rb (revision 14826)
@@ -308,37 +308,24 @@
# Unescapes all possible entities
def Text::unnormalize( string, doctype=nil, filter=nil, illegal=nil )
- rv = string.clone
- rv.gsub!( /\r\n?/, "\n" )
- matches = rv.scan( REFERENCE )
- return rv if matches.size == 0
- rv.gsub!( NUMERICENTITY ) {|m|
- m=$1
- m = "0#{m}" if m[0] == ?x
- [Integer(m)].pack('U*')
- }
- matches.collect!{|x|x[0]}.compact!
- if matches.size > 0
- if doctype
- matches.each do |entity_reference|
- unless filter and filter.include?(entity_reference)
- entity_value = doctype.entity( entity_reference )
- re = /&#{entity_reference};/
- rv.gsub!( re, entity_value ) if entity_value
- end
+ string.gsub( /\r\n?/, "\n" ).gsub( REFERENCE ) { |ref|
+ if ref[1] == ?#
+ if ref[2] == ?x
+ [ref[3...-1].to_i(16)].pack('U*')
+ else
+ [ref[2...-1].to_i].pack('U*')
end
+ elsif ref == '&'
+ '&'
+ elsif filter and filter.include?( ref[1...-1] )
+ ref
+ elsif doctype
+ doctype.entity( ref[1...-1] ) or ref
else
- matches.each do |entity_reference|
- unless filter and filter.include?(entity_reference)
- entity_value = DocType::DEFAULT_ENTITIES[ entity_reference ]
- re = /&#{entity_reference};/
- rv.gsub!( re, entity_value.value ) if entity_value
- end
- end
+ entity_value = DocType::DEFAULT_ENTITIES[ ref[1...-1] ]
+ entity_value ? entity_value.value : ref
end
- rv.gsub!( /&/, '&' )
- end
- rv
+ }
end
end
end
Index: lib/rexml/formatters/pretty.rb
===================================================================
--- lib/rexml/formatters/pretty.rb (revision 14825)
+++ lib/rexml/formatters/pretty.rb (revision 14826)
@@ -31,6 +31,7 @@
@level = 0
@ie_hack = ie_hack
@width = 80
+ @compact = false
end
protected
Index: lib/rexml/encoding.rb
===================================================================
--- lib/rexml/encoding.rb (revision 14825)
+++ lib/rexml/encoding.rb (revision 14826)
@@ -56,14 +56,14 @@
def check_encoding str
# We have to recognize UTF-16, LSB UTF-16, and UTF-8
- if str[0] == 0xfe && str[1] == 0xff
+ if str[0,2] == "\xfe\xff"
str[0,2] = ""
return UTF_16
- elsif str[0] == 0xff && str[1] == 0xfe
+ elsif str[0,2] == "\xff\xfe"
str[0,2] = ""
return UNILE
end
- str =~ /^\s*<\?xml\s+version\s*=\s*(['"]).*?\1\s+encoding\s*=\s*(["'])(.*?)\2/um
+ str =~ /^\s*<\?xml\s+version\s*=\s*(['"]).*?\1\s+encoding\s*=\s*(["'])(.*?)\2/m
return $3.upcase if $3
return UTF_8
end
Index: lib/rexml/xpath_parser.rb
===================================================================
--- lib/rexml/xpath_parser.rb (revision 14825)
+++ lib/rexml/xpath_parser.rb (revision 14826)
@@ -222,7 +222,7 @@
when :child
new_nodeset = []
nt = nil
- for node in nodeset
+ nodeset.each do |node|
nt = node.node_type
new_nodeset += node.children if nt == :element or nt == :document
end
@@ -266,7 +266,7 @@
when :ancestor
new_nodeset = []
- for node in nodeset
+ nodeset.each do |node|
while node.parent
node = node.parent
new_nodeset << node unless new_nodeset.include? node
@@ -277,7 +277,7 @@
when :ancestor_or_self
new_nodeset = []
- for node in nodeset
+ nodeset.each do |node|
if node.node_type == :element
new_nodeset << node
while ( node.parent )
@@ -341,7 +341,7 @@
when :descendant
results = []
nt = nil
- for node in nodeset
+ nodeset.each do |node|
nt = node.node_type
results += expr( path_stack.dclone.unshift( :descendant_or_self ),
node.children ) if nt == :element or nt == :document
@@ -376,7 +376,7 @@
when :preceding
new_nodeset = []
- for node in nodeset
+ nodeset.each do |node|
new_nodeset += preceding( node )
end
#puts "NEW NODESET => #{new_nodeset.inspect}"
@@ -385,7 +385,7 @@
when :following
new_nodeset = []
- for node in nodeset
+ nodeset.each do |node|
new_nodeset += following( node )
end
nodeset = new_nodeset
@@ -395,7 +395,7 @@
#puts "In :namespace"
new_nodeset = []
prefix = path_stack.shift
- for node in nodeset
+ nodeset.each do |node|
if (node.node_type == :element or node.node_type == :attribute)
if @namespaces
namespaces = @namespaces
--
ML: ruby-changes@q...
Info: http://www.atdot.net/~ko1/quickml