multi_xml 0.2.2 → 0.3.0

data/.autotest

@@ -0,0 +1 @@
+require 'autotest/bundler'

data/.travis.yml

@@ -0,0 +1,7 @@
+rvm:
+  - 1.8.7
+  - 1.9.1
+  - 1.9.2
+  - rbx
+  - rbx-2.0
+  - ree

data/.yardopts

@@ -2,4 +2,4 @@
 --protected
 --markup markdown
 -
-LICENSE.mkd
+LICENSE.md

data/Gemfile

@@ -1,8 +1,8 @@
-source "http://rubygems.org"
+source 'http://rubygems.org'
 
 group :development, :test do
-  gem 'libxml-ruby', "~> 1.1", :require => nil
-  gem 'nokogiri', "~> 1.4", :require => nil
+  gem 'libxml-ruby', "~> 2.0", :require => nil, :platforms => :mri
+  gem 'nokogiri',    "~> 1.4", :require => nil
 end
 
 gemspec

data/{README.mkd → README.md}

@@ -4,11 +4,15 @@ A generic swappable back-end for XML parsing
 
 Installation
 ------------
-    $ [sudo] gem install multi_xml
+    gem install multi_xml
 
 Documentation
 -------------
-<http://rdoc.info/gems/multi_xml>
+[http://rdoc.info/gems/multi_xml](http://rdoc.info/gems/multi_xml)
+
+Continuous Integration
+----------------------
+[![Build Status](https://secure.travis-ci.org/sferik/multi_xml.png)](http://travis-ci.org/sferik/multi_xml)
 
 Usage Examples
 --------------
@@ -50,15 +54,15 @@ Here are some ways *you* can contribute:
 * by writing specifications
 * by writing code (**no patch is too small**: fix typos, add comments, clean up inconsistent whitespace)
 * by refactoring code
-* by resolving [issues](http://github.com/sferik/multi_xml/issues)
+* by resolving [issues](https://github.com/sferik/multi_xml/issues)
 * by reviewing patches
 
 Submitting an Issue
 -------------------
-We use the [GitHub issue tracker](http://github.com/sferik/multi_xml/issues) to track bugs and
+We use the [GitHub issue tracker](https://github.com/sferik/multi_xml/issues) to track bugs and
 features. Before submitting a bug report or feature request, check to make sure it hasn't already
 been submitted. You can indicate support for an existing issuse by voting it up. When submitting a
-bug report, please include a [Gist](http://gist.github.com/) that includes a stack trace and any
+bug report, please include a [Gist](https://gist.github.com/) that includes a stack trace and any
 details that may be necessary to reproduce the bug, including your gem version, Ruby version, and
 operating system. Ideally, a bug report should include a pull request with failing specs.
 
@@ -74,11 +78,37 @@ Submitting a Pull Request
 8. Commit and push your changes.
 9. Submit a pull request. Please do not include changes to the gemspec, version, or history file. (If you want to create your own version for some reason, please do so in a separate commit.)
 
+Supported Rubies
+----------------
+This library aims to support and is [tested
+against](http://travis-ci.org/sferik/multi_xml) the following Ruby
+implementations:
+
+* Ruby 1.8.7
+* Ruby 1.9.1
+* Ruby 1.9.2
+* [Rubinius](http://rubini.us)
+* [Ruby Enterprise Edition](http://www.rubyenterpriseedition.com/)
+
+If something doesn't work on one of these interpreters, it should be considered
+a bug.
+
+This library may inadvertently work (or seem to work) on other Ruby
+implementations, however support will only be provided for the versions listed
+above.
+
+If you would like this library to support another Ruby version, you may
+volunteer to be a maintainer. Being a maintainer entails making sure all tests
+run and pass on that implementation. When something breaks on your
+implementation, you will be personally responsible for providing patches in a
+timely fashion. If critical issues for a particular implementation exist at the
+time of a major release, support for that Ruby version may be dropped.
+
 Inspiration
 -----------
-MultiXML was inspired by [MultiJSON](http://github.com/intridea/multi_json/).
+MultiXML was inspired by [MultiJSON](https://github.com/intridea/multi_json/).
 
 Copyright
 ---------
 Copyright (c) 2010 Erik Michaels-Ober.
-See [LICENSE](https://github.com/sferik/multi_xml/blob/master/LICENSE.mkd) for details.
+See [LICENSE](https://github.com/sferik/multi_xml/blob/master/LICENSE.md) for details.

data/Rakefile

@@ -1,3 +1,5 @@
+#!/usr/bin/env rake
+
 require 'bundler'
 Bundler::GemHelper.install_tasks
 
@@ -10,7 +12,7 @@ task :default => :spec
 namespace :doc do
   require 'yard'
   YARD::Rake::YardocTask.new do |task|
-    task.files   = ['LICENSE.mkd', 'lib/**/*.rb']
+    task.files   = ['LICENSE.md', 'lib/**/*.rb']
     task.options = [
       '--no-private',
       '--protected',

data/lib/multi_xml.rb

@@ -1,47 +1,46 @@
 require 'base64'
 require 'bigdecimal'
 require 'date'
-require 'multi_xml/core_extensions'
+require 'stringio'
 require 'time'
 require 'yaml'
 
 module MultiXml
   class ParseError < StandardError; end
 
-  class << self
-
-    REQUIREMENT_MAP = [
-      ['libxml', :libxml],
-      ['nokogiri', :nokogiri],
-      ['rexml/document', :rexml]
-    ] unless defined?(REQUIREMENT_MAP)
-
-    CONTENT_ROOT = '__content__'.freeze unless defined?(CONTENT_ROOT)
-
-    # TODO: use Time.xmlschema instead of Time.parse;
-    #       use regexp instead of Date.parse
-    unless defined?(PARSING)
-      PARSING = {
-        'symbol'       => Proc.new{|symbol| symbol.to_sym},
-        'date'         => Proc.new{|date| Date.parse(date)},
-        'datetime'     => Proc.new{|time| Time.parse(time).utc rescue DateTime.parse(time).utc},
-        'integer'      => Proc.new{|integer| integer.to_i},
-        'float'        => Proc.new{|float| float.to_f},
-        'decimal'      => Proc.new{|number| BigDecimal(number)},
-        'boolean'      => Proc.new{|boolean| !%w(0 false).include?(boolean.strip)},
-        'string'       => Proc.new{|string| string.to_s},
-        'yaml'         => Proc.new{|yaml| YAML::load(yaml) rescue yaml},
-        'base64Binary' => Proc.new{|binary| binary.unpack('m').first},
-        'binary'       => Proc.new{|binary, entity| parse_binary(binary, entity)},
-        'file'         => Proc.new{|file, entity| parse_file(file, entity)}
-      }
-
-      PARSING.update(
-        'double'   => PARSING['float'],
-        'dateTime' => PARSING['datetime']
-      )
-    end
+  REQUIREMENT_MAP = [
+    ['libxml', :libxml],
+    ['nokogiri', :nokogiri],
+    ['rexml/document', :rexml]
+  ] unless defined?(REQUIREMENT_MAP)
+
+  CONTENT_ROOT = '__content__'.freeze unless defined?(CONTENT_ROOT)
+
+  # TODO: use Time.xmlschema instead of Time.parse;
+  #       use regexp instead of Date.parse
+  unless defined?(PARSING)
+    PARSING = {
+      'symbol'       => Proc.new{|symbol| symbol.to_sym},
+      'date'         => Proc.new{|date| Date.parse(date)},
+      'datetime'     => Proc.new{|time| Time.parse(time).utc rescue DateTime.parse(time).utc},
+      'integer'      => Proc.new{|integer| integer.to_i},
+      'float'        => Proc.new{|float| float.to_f},
+      'decimal'      => Proc.new{|number| BigDecimal(number)},
+      'boolean'      => Proc.new{|boolean| !%w(0 false).include?(boolean.strip)},
+      'string'       => Proc.new{|string| string.to_s},
+      'yaml'         => Proc.new{|yaml| YAML::load(yaml) rescue yaml},
+      'base64Binary' => Proc.new{|binary| binary.unpack('m').first},
+      'binary'       => Proc.new{|binary, entity| parse_binary(binary, entity)},
+      'file'         => Proc.new{|file, entity| parse_file(file, entity)}
+    }
+
+    PARSING.update(
+      'double'   => PARSING['float'],
+      'dateTime' => PARSING['datetime']
+    )
+  end
 
+  class << self
     # Get the current parser class.
     def parser
       return @parser if @parser
@@ -85,13 +84,22 @@ module MultiXml
       end
     end
 
-    # Parse an XML string into Ruby.
+    # Parse an XML string or IO into Ruby.
     #
     # <b>Options</b>
     #
     # <tt>:symbolize_keys</tt> :: If true, will use symbols instead of strings for the keys.
     def parse(xml, options={})
-      xml.strip!
+      xml ||= ''
+
+      xml.strip! if xml.respond_to?(:strip!)
+
+      xml = StringIO.new(xml) unless xml.respond_to?(:read)
+
+      char = xml.getc
+      return {} if char.nil?
+      xml.ungetc(char)
+
       begin
         hash = typecast_xml_value(undasherize_keys(parser.parse(xml))) || {}
       rescue parser.parse_error => error
@@ -118,7 +126,7 @@ module MultiXml
     private
 
     # TODO: Add support for other encodings
-    def self.parse_binary(binary, entity) #:nodoc:
+    def parse_binary(binary, entity) #:nodoc:
       case entity['encoding']
       when 'base64'
         Base64.decode64(binary)
@@ -127,7 +135,7 @@ module MultiXml
       end
     end
 
-    def self.parse_file(file, entity)
+    def parse_file(file, entity)
       f = StringIO.new(Base64.decode64(file))
       f.extend(FileLike)
       f.original_filename = entity['name']
@@ -172,13 +180,14 @@ module MultiXml
       case value
       when Hash
         if value['type'] == 'array'
-          _, entries = Array.wrap(value.detect{|key, value| key != 'type'})
-          if entries.blank? || (value.is_a?(Hash) && c = value[CONTENT_ROOT] && c.blank?)
+          _, entries = value.detect {|key, _| key != 'type'}
+
+          if entries.nil? || (entries.is_a?(String) && entries.strip.empty?)
             []
           else
             case entries
             when Array
-              entries.map{|value| typecast_xml_value(value)}
+              entries.map {|entry| typecast_xml_value(entry)}
             when Hash
               [typecast_xml_value(entries)]
             else
@@ -195,7 +204,7 @@ module MultiXml
         elsif value['type'] == 'string' && value['nil'] != 'true'
           ''
         # blank or nil parsed values are represented by nil
-        elsif value.blank? || value['nil'] == 'true'
+        elsif value.empty? || value['nil'] == 'true'
           nil
         # If the type is the only element which makes it then
         # this still makes the value nil, except if type is
@@ -203,8 +212,8 @@ module MultiXml
         elsif value['type'] && value.size == 1 && !value['type'].is_a?(Hash)
           nil
         else
-          xml_value = value.inject({}) do |hash, (key, value)|
-            hash[key] = typecast_xml_value(value)
+          xml_value = value.inject({}) do |hash, (k, v)|
+            hash[k] = typecast_xml_value(v)
             hash
           end
 

data/lib/multi_xml/parsers/libxml.rb

@@ -1,79 +1,30 @@
 require 'libxml' unless defined?(LibXML)
+require 'multi_xml/parsers/libxml2_parser'
 
 module MultiXml
   module Parsers
     module Libxml #:nodoc:
+      include Libxml2Parser
+
       extend self
-      def parse_error; ::LibXML::XML::Error; end
 
-      # Parse an XML Document string or IO into a simple hash using LibXML.
-      # xml::
-      #   XML Document string or IO to parse
-      def parse(xml)
-        if !xml.respond_to?(:read)
-          xml = StringIO.new(xml || '')
-        end
+      def parse_error() ::LibXML::XML::Error end
 
-        char = xml.getc
-        if char.nil?
-          {}
-        else
-          xml.ungetc(char)
-          LibXML::XML::Parser.io(xml).parse.to_hash
-        end
+      def parse(xml)
+        node_to_hash(LibXML::XML::Parser.io(xml).parse.root)
       end
-    end
-  end
-end
 
-module LibXML #:nodoc:
-  module Conversions #:nodoc:
-    module Document #:nodoc:
-      def to_hash
-        root.to_hash
+      def each_child(node, &block)
+        node.each_child &block
       end
-    end
 
-    module Node #:nodoc:
-      CONTENT_ROOT = '__content__'.freeze unless defined?(CONTENT_ROOT)
-
-      # Convert XML document to hash
-      #
-      # hash::
-      #   Hash to merge the converted element into.
-      def to_hash(hash={})
-        node_hash = {}
-
-        # Insert node hash into parent hash correctly.
-        case hash[name]
-          when Array then hash[name] << node_hash
-          when Hash  then hash[name] = [hash[name], node_hash]
-          when nil   then hash[name] = node_hash
-        end
-
-        # Handle child elements
-        each_child do |c|
-          if c.element?
-            c.to_hash(node_hash)
-          elsif c.text? || c.cdata?
-            node_hash[CONTENT_ROOT] ||= ''
-            node_hash[CONTENT_ROOT] << c.content
-          end
-        end
-
-        # Remove content node if it is blank
-        if node_hash.length > 1 && node_hash[CONTENT_ROOT].blank?
-          node_hash.delete(CONTENT_ROOT)
-        end
-
-        # Handle attributes
-        each_attr { |a| node_hash[a.name] = a.value }
+      def each_attr(node, &block)
+        node.each_attr &block
+      end
 
-        hash
+      def node_name(node)
+        node.name
       end
     end
   end
 end
-
-LibXML::XML::Document.send(:include, LibXML::Conversions::Document)
-LibXML::XML::Node.send(:include, LibXML::Conversions::Node)

data/lib/multi_xml/parsers/libxml2_parser.rb

@@ -0,0 +1,66 @@
+module MultiXml
+  module Parsers
+    module Libxml2Parser #:nodoc:
+      # Convert XML document to hash
+      #
+      # node::
+      #   The XML node object to convert to a hash.
+      #
+      # hash::
+      #   Hash to merge the converted element into.
+      def node_to_hash(node, hash={})
+        node_hash = {MultiXml::CONTENT_ROOT => ''}
+
+        name = node_name(node)
+
+        # Insert node hash into parent hash correctly.
+        case hash[name]
+          when Array then hash[name] << node_hash
+          when Hash  then hash[name] = [hash[name], node_hash]
+          when nil   then hash[name] = node_hash
+        end
+
+        # Handle child elements
+        each_child(node) do |c|
+          if c.element?
+            node_to_hash(c, node_hash)
+          elsif c.text? || c.cdata?
+            node_hash[MultiXml::CONTENT_ROOT] << c.content
+          end
+        end
+
+        # Remove content node if it is empty
+        if node_hash[MultiXml::CONTENT_ROOT].strip.empty?
+          node_hash.delete(MultiXml::CONTENT_ROOT)
+        end
+
+        # Handle attributes
+        each_attr(node) {|a| node_hash[node_name(a)] = a.value }
+
+        hash
+      end
+
+      # Parse an XML Document IO into a simple hash.
+      # xml::
+      #   XML Document IO to parse
+      def parse(xml)
+        raise NotImplementedError, "inheritor should define #{__method__}"
+      end
+
+      # :stopdoc:
+      private
+
+      def each_child(*args)
+        raise NotImplementedError, "inheritor should define #{__method__}"
+      end
+
+      def each_attr(*args)
+        raise NotImplementedError, "inheritor should define #{__method__}"
+      end
+
+      def node_name(*args)
+        raise NotImplementedError, "inheritor should define #{__method__}"
+      end
+    end
+  end
+end

data/lib/multi_xml/parsers/nokogiri.rb

@@ -1,79 +1,32 @@
 require 'nokogiri' unless defined?(Nokogiri)
+require 'multi_xml/parsers/libxml2_parser'
 
 module MultiXml
   module Parsers
     module Nokogiri #:nodoc:
+      include Libxml2Parser
+
       extend self
-      def parse_error; ::Nokogiri::XML::SyntaxError; end
 
-      # Parse an XML Document string or IO into a simple hash using Nokogiri.
-      # xml::
-      #   XML Document string or IO to parse
-      def parse(xml)
-        if !xml.respond_to?(:read)
-          xml = StringIO.new(xml || '')
-        end
+      def parse_error() ::Nokogiri::XML::SyntaxError end
 
-        char = xml.getc
-        if char.nil?
-          {}
-        else
-          xml.ungetc(char)
-          doc = ::Nokogiri::XML(xml)
-          raise doc.errors.first if doc.errors.length > 0
-          doc.to_hash
-        end
+      def parse(xml)
+        doc = ::Nokogiri::XML(xml)
+        raise doc.errors.first if doc.errors.length > 0
+        node_to_hash(doc.root)
       end
 
-      module Conversions #:nodoc:
-        module Document #:nodoc:
-          def to_hash
-            root.to_hash
-          end
-        end
-
-        module Node #:nodoc:
-          CONTENT_ROOT = '__content__'.freeze unless defined?(CONTENT_ROOT)
-
-          # Convert XML document to hash
-          #
-          # hash::
-          #   Hash to merge the converted element into.
-          def to_hash(hash={})
-            node_hash = {}
-
-            # Insert node hash into parent hash correctly.
-            case hash[name]
-              when Array then hash[name] << node_hash
-              when Hash  then hash[name] = [hash[name], node_hash]
-              when nil   then hash[name] = node_hash
-            end
-
-            # Handle child elements
-            children.each do |c|
-              if c.element?
-                c.to_hash(node_hash)
-              elsif c.text? || c.cdata?
-                node_hash[CONTENT_ROOT] ||= ''
-                node_hash[CONTENT_ROOT] << c.content
-              end
-            end
-
-            # Remove content node if it is blank and there are child tags
-            if node_hash.length > 1 && node_hash[CONTENT_ROOT].blank?
-              node_hash.delete(CONTENT_ROOT)
-            end
-
-            # Handle attributes
-            attribute_nodes.each { |a| node_hash[a.node_name] = a.value }
+      def each_child(node, &block)
+        node.children.each &block
+      end
 
-            hash
-          end
-        end
+      def each_attr(node, &block)
+        node.attribute_nodes.each &block
       end
 
-      ::Nokogiri::XML::Document.send(:include, Conversions::Document)
-      ::Nokogiri::XML::Node.send(:include, Conversions::Node)
+      def node_name(node)
+        node.node_name
+      end
     end
   end
 end