@@ -32,7 +32,7 @@ This will return an object containing information extracted from the RDF respons ## Requirements ## * [Ruby 1.8.5 or better](http://ruby-lang.org) -* [libxml-ruby](http://libxml.rubyforge.org/), [libxml2](http://xmlsoft.org/) +* [nokogiri](http://nokogiri.rubyforge.org/nokogiri/), [libxml2](http://xmlsoft.org/) * [curb](http://curb.rubyforge.org/), [libcurl](http://curl.haxx.se/) * [json](http://json.rubyforge.org/) README.markdown @@ -18,7 +18,7 @@ begin s.authors = ['Abhay Kumar'] s.files = FileList["[A-Z]*", "{bin,generators,lib,test}/**/*"] s.rubyforge_project = 'calais' - s.add_dependency 'libxml-ruby', '>= 0.5.4' + s.add_dependency 'nokogiri', '>= 1.3.3' s.add_dependency 'json', '>= 1.1.3' s.add_dependency 'curb', '>= 0.1.4' end Rakefile @@ -41,16 +41,16 @@ Gem::Specification.new do |s| s.specification_version = 2 if Gem::Version.new(Gem::RubyGemsVersion) >= Gem::Version.new('1.2.0') then - s.add_runtime_dependency(%q<libxml-ruby>, [">= 0.5.4"]) + s.add_runtime_dependency(%q<nokogiri>, [">= 1.3.3"]) s.add_runtime_dependency(%q<json>, [">= 1.1.3"]) s.add_runtime_dependency(%q<curb>, [">= 0.1.4"]) else - s.add_dependency(%q<libxml-ruby>, [">= 0.5.4"]) + s.add_dependency(%q<nokogiri>, [">= 1.3.3"]) s.add_dependency(%q<json>, [">= 1.1.3"]) s.add_dependency(%q<curb>, [">= 0.1.4"]) end else - s.add_dependency(%q<libxml-ruby>, [">= 0.5.4"]) + s.add_dependency(%q<nokogiri>, [">= 1.3.3"]) s.add_dependency(%q<json>, [">= 1.1.3"]) s.add_dependency(%q<curb>, [">= 0.1.4"]) end calais.gemspec @@ -5,7 +5,7 @@ require 'iconv' require 'set' require 'rubygems' -require 'xml/libxml' +require 'nokogiri' require 'json' require 'curb' lib/calais.rb @@ -38,36 +38,37 @@ module Calais def params_xml check_params - - params_node = XML::Node.new('c:params') + document = Nokogiri::XML::Document.new + + params_node = Nokogiri::XML::Node.new('c:params', document) params_node['xmlns:c'] = 'http://s.opencalais.com/1/pred/' params_node['xmlns:rdf'] = 'http://www.w3.org/1999/02/22-rdf-syntax-ns#' - - processing_node = XML::Node.new('c:processingDirectives') + + processing_node = Nokogiri::XML::Node.new('c:processingDirectives', document) processing_node['c:contentType'] = AVAILABLE_CONTENT_TYPES[@content_type] if @content_type processing_node['c:outputFormat'] = AVAILABLE_OUTPUT_FORMATS[@output_format] if @output_format processing_node['c:reltagBaseURL'] = @reltag_base_url.to_s if @reltag_base_url - + processing_node['c:enableMetadataType'] = @metadata_enables.join(';') unless @metadata_enables.empty? processing_node['c:discardMetadata'] = @metadata_discards.join(';') unless @metadata_discards.empty? processing_node['c:omitOutputtingOriginalText'] = 'true' if @omit_outputting_original_text - - user_node = XML::Node.new('c:userDirectives') + + user_node = Nokogiri::XML::Node.new('c:userDirectives', document) user_node['c:allowDistribution'] = @allow_distribution.to_s unless @allow_distribution.nil? user_node['c:allowSearch'] = @allow_search.to_s unless @allow_search.nil? user_node['c:externalID'] = @external_id.to_s if @external_id user_node['c:submitter'] = @submitter.to_s if @submitter - + params_node << processing_node params_node << user_node - + if @external_metadata - external_node = XML::Node.new('c:externalMetadata') + external_node = Nokogiri::XML::Node.new('c:externalMetadata', document) external_node << @external_metadata params_node << external_node end - - params_node.to_s + + params_node.to_xml(:indent => 2) end private lib/calais/client.rb @@ -47,14 +47,14 @@ module Calais class Instance attr_accessor :prefix, :exact, :suffix, :offset, :length - # Makes a new Instance object from an appropriate LibXML::XML::Node. + # Makes a new Instance object from an appropriate Nokogiri::XML::Node. def self.from_node(node) instance = self.new - instance.prefix = node.find_first("c:prefix").content - instance.exact = node.find_first("c:exact").content - instance.suffix = node.find_first("c:suffix").content - instance.offset = node.find_first("c:offset").content.to_i - instance.length = node.find_first("c:length").content.to_i + instance.prefix = node.xpath("c:prefix[1]").first.content + instance.exact = node.xpath("c:exact[1]").first.content + instance.suffix = node.xpath("c:suffix[1]").first.content + instance.offset = node.xpath("c:offset[1]").first.content.to_i + instance.length = node.xpath("c:length[1]").first.content.to_i instance end @@ -76,112 +76,112 @@ module Calais private def extract_data - doc = XML::Parser.string(@raw_response).parse + doc = Nokogiri::XML(@raw_response) - if doc.root.find("/Error").first - raise Calais::Error, doc.root.find("/Error/Exception").first.content + if doc.root.xpath("/Error[1]").first + raise Calais::Error, doc.root.xpath("/Error/Exception").first.content end - doc.root.find("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:docinfometa]}')]/..").each do |node| + doc.root.xpath("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:docinfometa]}')]/..").each do |node| @language = node['language'] @submission_date = DateTime.parse node['submissionDate'] - attributes = extract_attributes(node.find("*[contains(name(), 'c:')]")) + attributes = extract_attributes(node.xpath("*[contains(name(), 'c:')]")) @signature = attributes.delete('signature') @submitter_code = attributes.delete('submitterCode') - node.remove! + node.remove end - doc.root.find("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:docinfo]}')]/..").each do |node| + doc.root.xpath("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:docinfo]}')]/..").each do |node| @request_id = node['calaisRequestID'] - attributes = extract_attributes(node.find("*[contains(name(), 'c:')]")) + attributes = extract_attributes(node.xpath("*[contains(name(), 'c:')]")) @doc_title = attributes.delete('docTitle') - @doc_date = Date.parse attributes.delete('docDate') + @doc_date = Date.parse(attributes.delete('docDate')) - node.remove! + node.remove end - @categories = doc.root.find("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:doccat]}')]/..").map do |node| + @categories = doc.root.xpath("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:doccat]}')]/..").map do |node| category = Category.new - category.name = node.find_first("c:categoryName").content - score = node.find_first("c:score") + category.name = node.xpath("c:categoryName[1]").first.content + score = node.xpath("c:score[1]").first category.score = score.content.to_f unless score.nil? - node.remove! + node.remove category end - @relevances = doc.root.find("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:relevances]}')]/..").inject({}) do |acc, node| - subject_hash = node.find_first("c:subject")[:resource].split('/')[-1] - acc[subject_hash] = node.find_first("c:relevance").content.to_f + @relevances = doc.root.xpath("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:relevances]}')]/..").inject({}) do |acc, node| + subject_hash = node.xpath("c:subject[1]").first[:resource].split('/')[-1] + acc[subject_hash] = node.xpath("c:relevance[1]").first.content.to_f - node.remove! + node.remove acc end - @entities = doc.root.find("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:entities]}')]/..").map do |node| + @entities = doc.root.xpath("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:entities]}')]/..").map do |node| extracted_hash = node['about'].split('/')[-1] rescue nil entity = Entity.new entity.calais_hash = CalaisHash.find_or_create(extracted_hash, @hashes) entity.type = extract_type(node) - entity.attributes = extract_attributes(node.find("*[contains(name(), 'c:')]")) + entity.attributes = extract_attributes(node.xpath("*[contains(name(), 'c:')]")) entity.relevance = @relevances[extracted_hash] entity.instances = extract_instances(doc, extracted_hash) - node.remove! + node.remove entity end - @relations = doc.root.find("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:relations]}')]/..").map do |node| + @relations = doc.root.xpath("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:relations]}')]/..").map do |node| extracted_hash = node['about'].split('/')[-1] rescue nil relation = Relation.new relation.calais_hash = CalaisHash.find_or_create(extracted_hash, @hashes) relation.type = extract_type(node) - relation.attributes = extract_attributes(node.find("*[contains(name(), 'c:')]")) + relation.attributes = extract_attributes(node.xpath("*[contains(name(), 'c:')]")) relation.instances = extract_instances(doc, extracted_hash) - node.remove! + node.remove relation end - @geographies = doc.root.find("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:geographies]}')]/..").map do |node| - attributes = extract_attributes(node.find("*[contains(name(), 'c:')]")) + @geographies = doc.root.xpath("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:geographies]}')]/..").map do |node| + attributes = extract_attributes(node.xpath("*[contains(name(), 'c:')]")) geography = Geography.new geography.name = attributes.delete('name') geography.calais_hash = attributes.delete('subject') geography.attributes = attributes - node.remove! + node.remove geography end - doc.root.find("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:defaultlangid]}')]/..").each { |node| node.remove! } - doc.root.find("./*").each { |node| node.remove! } + doc.root.xpath("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:defaultlangid]}')]/..").each { |node| node.remove } + doc.root.xpath("./*").each { |node| node.remove } return end def extract_instances(doc, hash) - doc.root.find("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:instances]}')]/..").select do |instance_node| - instance_node.find_first("c:subject")[:resource].split("/")[-1] == hash + doc.root.xpath("rdf:Description/rdf:type[contains(@rdf:resource, '#{MATCHERS[:instances]}')]/..").select do |instance_node| + instance_node.xpath("c:subject[1]").first[:resource].split("/")[-1] == hash end.map do |instance_node| instance = Instance.from_node(instance_node) - instance_node.remove! + instance_node.remove instance end end def extract_type(node) - node.find("*[name()='rdf:type']")[0]['resource'].split('/')[-1] + node.xpath("*[name()='rdf:type']")[0]['resource'].split('/')[-1] rescue nil end lib/calais/response.rb @@ -32,7 +32,7 @@ end describe Calais::Client, :params_xml do it 'returns an xml encoded string' do client = Calais::Client.new(:content => SAMPLE_DOCUMENT, :license_id => LICENSE_ID) - client.params_xml.should == %[<c:params xmlns:c="http://s.opencalais.com/1/pred/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">\n <c:processingDirectives/>\n <c:userDirectives/>\n</c:params>] + client.params_xml.should == %[<c:params xmlns:c=\"http://s.opencalais.com/1/pred/\" xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\"><c:processingDirectives/><c:userDirectives/></c:params>] client.content_type = :xml client.output_format = :json @@ -45,7 +45,7 @@ describe Calais::Client, :params_xml do client.external_id = Digest::SHA1.hexdigest(client.content) client.submitter = 'calais.rb' - client.params_xml.should == %[<c:params xmlns:c="http://s.opencalais.com/1/pred/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">\n <c:processingDirectives c:contentType="text/xml" c:outputFormat="application/json" c:reltagBaseURL="http://opencalais.com" c:enableMetadataType="GenericRelations" c:discardMetadata="er/Company;er/Geo"/>\n <c:userDirectives c:allowDistribution="true" c:allowSearch="true" c:externalID="1a008b91e7d21962e132bc1d6cb252532116a606" c:submitter="calais.rb"/>\n</c:params>] + client.params_xml.should == %[<c:params xmlns:c="http://s.opencalais.com/1/pred/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"><c:processingDirectives c:contentType="text/xml" c:outputFormat="application/json" c:reltagBaseURL="http://opencalais.com" c:enableMetadataType="GenericRelations" c:discardMetadata="er/Company;er/Geo"/><c:userDirectives c:allowDistribution="true" c:allowSearch="true" c:externalID="1a008b91e7d21962e132bc1d6cb252532116a606" c:submitter="calais.rb"/></c:params>] end end spec/calais/client_spec.rb ac8d10314b190aa5a1a365bf5d97475e09ad0fbe Jeremy Ashkenas jashkenas@gmail.com http://github.com/abhay/calais/commit/0810d3ad00b0aa5b85e34df9e0e6d858d7e30b81 0810d3ad00b0aa5b85e34df9e0e6d858d7e30b81 2009-09-18T11:44:36-07:00 2009-09-18T11:44:36-07:00 swapping out libxml-ruby for Nokogiri, just because libxml-ruby was segfaulting (on this machine) 8b50d07f1f6943e28704fdc9446f682a77bfbb6c Jeremy Ashkenas jashkenas@gmail.com