test/fixtures/articles/noroot.xml @@ -1,10 +1,10 @@ -require 'xmlsimple' +require 'xml' module WithXml - # Customize the xml serialization for ActiveRecord objects. It uses the with_xml block to + + # Customize the xml serialization for ActiveRecord objects. It uses the with_xml block to # define options declared in the ActiveRecord that allows for the customization of the to_xml and # from_xml call. - # def self.included(base) base.extend ClassMethods end @@ -14,7 +14,7 @@ module WithXml end module ClassMethods - + # Specifies the options of handling incoming (from_xml) xml data and serialization (to_xml) # # class Article < ActiveRecord::Base @@ -38,16 +38,15 @@ module WithXml module InstanceMethods - ## # Override the default from_xml and perform some mapping goodness + # +xml+ string of the xml to import def from_xml(xml) - mapper = SupportingClasses::Binder.new("<opt>#{xml}</opt>", self) - mapped_attributes = mapper.bind + mapper = SupportingClasses::Binder.new("#{xml}", self) + mapped_attributes = mapper.map self.attributes = mapped_attributes if mapped_attributes self end - - ## + # Override the default to_xml with our custom options def to_xml(options = {}, &block) with_xml = self.class.read_inheritable_attribute(:with_xml) @@ -56,7 +55,7 @@ module WithXml end super(options, &block) end - + end module SupportingClasses @@ -69,26 +68,25 @@ module WithXml attr_reader :serialize_opts def initialize(name, opts, binder_table, &block) + @matchers = [] @name = name.to_sym - @binders = binder_table[@name] = [] + @binders = binder_table[@name] = {} instance_eval(&block) if block extract_options!(opts) end - ## # Adds each 'bind' option to the binders array - def bind(name, xml_attr_opts) - @binders << SupportingClasses::XmlAttr.new(name, xml_attr_opts) + # +name+ of the attribute + def map(name, xml_attr_opts) + @binders[name] = SupportingClasses::XmlAttr.new(name, xml_attr_opts) end - ## # Adds the options used to export the xml (to_xml) def serialize(opts) @serialize_opts = opts end - ## - # Return true or false if + # Return true or false if we have a match def has_match?(name) matchers = @opts[:match] || Array.new return true if matchers.find {|n| n == name} @@ -96,10 +94,12 @@ module WithXml private + # Parse options decalared in the with_xml def extract_options!(opts) return nil if opts.nil? @opts = opts @matchers = opts[:alias] if opts[:alias] + @matchers.collect!{|x| x.to_s} end end @@ -107,10 +107,13 @@ module WithXml class XmlAttr attr_reader :name attr_reader :matchers - - def initialize(name, opts) - @name, @opts = name, opts - @matchers = opts[:to] if opts[:to] + + def initialize(name, opts = {}) + @matchers = [] + @name, @opts = name.to_s, opts + opts[:to] = [] unless opts[:to] + opts[:to].insert(0, name.to_s) + @matchers = opts[:to].flatten end end @@ -123,56 +126,79 @@ module WithXml @with_xml.matchers << activerecord.class.class_name.downcase end - ## # Creates a new hash map that matches the ActiveRecord fields from the given xml - def bind - xml_hash = Hash.from_xml(xml) - new_xml = find_hash_for(xml_hash, @with_xml.matchers) do |key, value| - break(value) + def map + xml_attrs = {} + return xml_attrs unless valid_xml?(xml) + + activerecord.attributes.each_key do |key| + if !@with_xml.binders.has_key?(key.to_sym) + @with_xml.binders[key.to_sym] = SupportingClasses::XmlAttr.new(key) + end + end + + root_xml = find_root_element(xml) + if root_xml + xml_attrs = find_attributes_for(root_xml) end - return hash_for(new_xml) + return xml_attrs end private - ## - # Searches all the keys (even nested) and returns the first one - def find_hash_for(hash, desired_keys, &block) - return false unless Hash === hash - hash.each_pair do |key, value| - if desired_keys.include?(key) || find_hash_for(value, desired_keys, &block) - yield(key, value) - return true + # Determine if the xml is well-formed or not + # +xml+ string to validate + def valid_xml?(xml) + begin + XML::Document.string(xml) + rescue Exception + return false + # Return nil if an exception is thrown + end + end + + # Based on the aliases, find the root element of the xml document + # +xml+ string that contains the root elements + def find_root_element(xml) + doc = XML::Document.string(xml) + @with_xml.matchers.each do |root| + if new_xml = find_element(doc, root) + return new_xml end end return false end - - ## - # Update the xml hash with the correct keys and values - def hash_for(xml) - return false unless Hash === xml - xml.each do |element_name, element_value| - field = field_for(element_name) - xml.delete(element_name) - if activerecord.has_attribute?(field) - xml[field] = element_value.strip unless element_value.nil? + + # Given, the list of :to options, find the first ActiveRecord field that matches the given key + def find_attributes_for(element) + attrs = {} + @with_xml.binders.each do |key, binder| + binder.matchers.each do |matcher| + if node = find_element(element, matcher) + attrs[binder.name] = find_text(node) + end end end + return attrs + end + + # Find the first element matching the xpath (ignores case) + # +doc+ to be searched + def find_element(doc, xpath = '/') + xml = (doc.find_first("//#{xpath}") || + doc.find_first("//#{xpath.downcase}") || + doc.find_first( "//#{xpath.upcase}")) return xml end - ## - # Given, the list of :to options, find the first ActiveRecord field that matches the given key - def field_for(element) - return element unless String === element - atrribute_name = element.downcase # ignore case by default - @with_xml.binders.each do |binder| - if binder.name.to_s == atrribute_name || binder.matchers.include?(atrribute_name) - return binder.name - end + # Find and format each element text + def find_text(doc) + return if doc.nil? + text = "" + doc.children.each do |child| + text << "#{child.content.strip} " unless child.content.strip.empty? end - return atrribute_name + text.strip end end lib/with_xml.rb @@ -1,7 +1,9 @@ class Article < ActiveRecord::Base + with_xml :feeds, :alias => ["magazine", "blog", "newspaper"] do - bind :heading, :to => ["subject", "title"] - bind :body, :to => ["content", "message", "post"] + map :heading, :to => ["subject", "title"] + map :body, :to => ["content[@type='copy']", "message", "post"] serialize :except => [ :name ], :skip_instruct => true end -end \ No newline at end of file + attr_accessor :sources +end test/fixtures/article.rb @@ -1,4 +1,8 @@ -<article> +<ARTICLE xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema"> <heading>Page Six</heading> <body>Content for article</body> -</article> \ No newline at end of file + <sources> + <source>Source 1</source> + <source>Source 2</source> + </sources> +</ARTICLE> \ No newline at end of file test/fixtures/articles/article.xml @@ -1,4 +1,9 @@ <magazine> <title>The Washington Post</title> - <content>This is the content</content> + <nested> + <content type="copy">This is the content</content> + <content type="copy2">This is NOT the content</content> + <content type="copy3">This is NOT the content</content> + <content type="copy4">This is NOT the content</content> + </nested> </magazine> \ No newline at end of file test/fixtures/articles/magazine.xml @@ -1,5 +1,5 @@ ActiveRecord::Schema.define(:version => 1) do create_table :articles do |t| - t.string :name, :heading, :body, :author + t.string :name, :heading, :body, :author, :sources end end \ No newline at end of file test/schema.rb @@ -3,31 +3,39 @@ require File.expand_path(File.dirname(__FILE__) + "/test_helper") class WithXmlTest < Test::Unit::TestCase fixtures :articles include TagMatchers - + + def test_field_override + article = articles(:magazine) # bind :heading, :to => ["subject", "title"] + assert_equal "New Yorker", article.heading + magazine = load_xml_article("magazine.xml") #title = "The Washington Post" + article.from_xml(magazine) + assert_equal "The Washington Post", article.heading + end + # test options settings - + def test_name_setter fixture = Article.with_xml(:test, {:alias => ["DAILYNEWS"]} ) do - bind :heading, :to => ["subject", "title"] + map :heading, :to => ["subject", "title"] end assert_equal :test, fixture.name end - + def test_options_setter fixture = Article.with_xml(:test, {:alias => ["DAILYNEWS"]} ) do - bind :heading, :to => ["subject", "title"] + map :heading, :to => ["subject", "title"] end assert_not_nil fixture.opts[:alias] end - + def test_matchers fixture = Article.with_xml(:test, {:alias => ["BEGIN_ARTICLE", "ARTICLE"]} ) do - bind :heading, :to => ["subject", "title"] - bind :body, :to => ["message", "content"] + map :heading, :to => ["subject", "title"] + map :body, :to => ["message", "content"] end - assert_equal fixture.matchers.size, 2 + assert_equal 2, fixture.matchers.size end - + # calling from_xml def test_field_override @@ -37,7 +45,7 @@ class WithXmlTest < Test::Unit::TestCase article.from_xml(magazine) assert_equal "The Washington Post", article.heading end - + def test_from_xml_with_new_instance article = Article.new # bind :body, :to => ["content", "message", "post"] blog = load_xml_article("blog.xml") @@ -45,14 +53,14 @@ class WithXmlTest < Test::Unit::TestCase assert_equal "The Post", article.body assert_equal "TechCrunch", article.heading end - + def test_from_xml_with_unknown_field article = Article.new newspaper = load_xml_article("newspaper.xml") # contains unkown field article.from_xml(newspaper) assert_equal "The Wall Street Journal", article.heading end - + def test_mapper_with_no_matchers article = Article.new article_xml = load_xml_article("article.xml") @@ -60,24 +68,51 @@ class WithXmlTest < Test::Unit::TestCase assert_equal "Page Six", article.heading assert_equal "Content for article", article.body end - + def test_mapper_with_nested_root - article = Article.new + article = Article.new article_xml = load_xml_article("nested_article.xml") article.from_xml(article_xml.to_s) assert_equal "Page Six", article.heading assert_equal "Content for article", article.body end + def test_mapper_multiple_values + article = Article.new + article_xml = load_xml_article("article.xml") + article.from_xml(article_xml.to_s) + assert_equal "Source 1 Source 2", article.sources + end + def test_import_with_capitalized_fields article = Article.new newspaper = load_xml_article("newspaper.xml") # contains AUTHOR in all CAPS article.from_xml(newspaper) assert_equal "Caps should be ignored by default", article.author end - + + def test_from_xml_save + article = Article.new # bind :body, :to => ["content", "message", "post"] + blog = load_xml_article("article.xml") + assert article.save! + end + + def test_from_xml_with_multiple_nodes + article = articles(:magazine) + magazine = load_xml_article("magazine.xml") #body = "This is the content" + article.from_xml(magazine) + assert_equal "This is the content", article.body + end + + def test_from_xml_with_no_root + article = Article.new + noroot = load_xml_article("noroot.xml") + article.from_xml(noroot) # should not cause errors + assert_equal nil, article.body + end + # calling to_xml - + def test_default_to_xml_behaviour article = Article.new xml = load_xml_article("article.xml") @@ -85,7 +120,7 @@ class WithXmlTest < Test::Unit::TestCase export = article.to_xml assert has_tag(export, '//article/heading', "Page Six") end - + def test_export_with_options article = Article.new xml = load_xml_article("article.xml") @@ -93,7 +128,7 @@ class WithXmlTest < Test::Unit::TestCase export = article.to_xml assert has_tag(export, '//name') == false end - + def test_export_with_instruct_skipped article = Article.new xml = load_xml_article("article.xml") @@ -110,7 +145,5 @@ class WithXmlTest < Test::Unit::TestCase assert has_tag(export, '//name') assert has_tag(export, '//author') == false end - - end test/with_xml_test.rb 084151651ab20d115bb5680441b63a770f1e8627 Matt Sears matt@mattsears.com http://github.com/mattsears/with_xml/commit/ebed07243a863910cb5586fd9f942fb1dff44068 ebed07243a863910cb5586fd9f942fb1dff44068 2009-03-28T18:10:24-07:00 2009-03-28T18:10:24-07:00 Switch xml parsers from rexml to libxml and added a few more test cases 9054bc0cd2717edd7209bc6ab5a9acf10ab02ba5 Matt Sears matt@mattsears.com