smg 0.1.0 → 0.2.0

data/lib/smg/http.rb

@@ -0,0 +1,66 @@
+require 'smg/http/request'
+require 'smg/http/exceptions'
+
+module SMG #:nodoc:
+  module HTTP
+
+    module Model
+
+      def site(value)
+        @site = value
+      end
+
+      def params(value)
+        @params = value
+      end
+
+      def get(path, options = {}, &block)
+        http Net::HTTP::Get, path, options, &block
+      end
+
+      def head(path, options = {}, &block)
+        http Net::HTTP::Head, path, options, &block
+      end
+
+      def delete(path, options = {}, &block)
+        http Net::HTTP::Delete, path, options, &block
+      end
+
+      def post(path, options = {}, &block)
+        http Net::HTTP::Post, path, options, &block
+      end
+
+      def put(path, options = {}, &block)
+        http Net::HTTP::Put, path, options, &block
+      end
+
+      private
+
+      def http(verb, path, options = {})
+        raise "site URI missed" unless @site
+        opts = options.dup
+        uri = uri_for(path, opts.delete(:query))
+        response = SMG::HTTP::Request.new(verb, uri, opts).perform
+        parse block_given? ? yield(response) : response.body
+      end
+
+      def uri_for(path, query = nil)
+        ret = Addressable::URI.parse(@site)
+        ret.path = path
+        qvalues = {}
+        qvalues.update(@params) if @params
+        qvalues.update(query) if query
+        ret.query_values = qvalues unless qvalues.empty?
+        ret
+      end
+
+    end
+
+    def self.append_features(base)
+      base.extend Model
+    end
+
+  end
+end
+
+# EOF

data/lib/smg/http/exceptions.rb

@@ -0,0 +1,37 @@
+module SMG #:nodoc:
+  module HTTP #:nodoc:
+
+    class ConnectionError < StandardError
+
+      def initialize(response, message = nil)
+        @response = response
+        @message = message
+      end
+
+      def to_s
+        message =  "Action failed with code: #{@response.code}."
+        message << " Message: #{@response.message}" if @response.respond_to?(:message)
+        message
+      end
+
+    end
+
+    class RedirectionError < ConnectionError
+    end
+
+    class TimeoutError < ConnectionError
+
+      def initialize(message)
+        @message = message
+      end
+
+      def to_s
+        @message
+      end
+
+    end
+
+  end
+end
+
+# EOF

data/lib/smg/http/request.rb

@@ -0,0 +1,126 @@
+require 'net/http'
+
+gem 'addressable', '>= 2.1.1'
+require 'addressable/uri'
+
+module SMG #:nodoc:
+  module HTTP #:nodoc:
+    class Request
+
+      attr_reader :verb, :uri, :headers, :body, :limit, :timeout, :proxy
+
+      DEFAULT_LIMIT = 5
+
+      VERBS = [Net::HTTP::Get, Net::HTTP::Post, Net::HTTP::Put, Net::HTTP::Delete, Net::HTTP::Head]
+
+      def initialize(verb, uri, options = {})
+        raise "unknown verb: #{verb}" unless VERBS.include?(verb)
+
+        @verb     = verb
+        @uri      = Addressable::URI.parse(uri)
+        @proxy    = options[:proxy] ? Addressable::URI.parse(options[:proxy]) : nil
+        @headers  = options[:headers] ? options[:headers].to_hash : {}
+        @limit    = options[:no_follow] ? 1 : options[:limit] || DEFAULT_LIMIT
+        @body     = options[:body]
+        @timeout  = options[:timeout] ? options[:timeout].to_i : nil
+      end
+
+      def perform
+        setup
+        response = http.request(@request)
+        handle_response(response)
+      rescue Timeout::Error => e
+        raise TimeoutError, e.message
+      end
+
+      protected
+
+      #--
+      # 10.3 Redirection 3xx
+      # This class of status code indicates that further action needs to be
+      # taken by the user agent in order to fulfill the request. The action
+      # required MAY be carried out by the user agent without interaction
+      # with the user if and only if the method used in the second request is
+      # GET or HEAD. A client SHOULD detect infinite redirection loops, since
+      # such loops generate network traffic for each redirection.
+      #
+      # If the (301, 302 or 307) status code is received in response to a request other
+      # than GET or HEAD, the user agent MUST NOT automatically redirect the
+      # request unless it can be confirmed by the user, since this might
+      # change the conditions under which the request was issued.
+      #
+      # 10.3.1 300 Multiple Choices
+      # If the server has a preferred choice of representation, it SHOULD
+      # include the specific URI for that representation in the Location
+      # field; user agents MAY use the Location field value for automatic
+      # redirection.
+      #
+      # 10.3.4 303 See Other
+      # The response to the request can be found under a different URI and
+      # SHOULD be retrieved using a GET method on that resource.
+      #
+      # 10.3.6 305 Use Proxy
+      # The requested resource MUST be accessed through the proxy given by
+      # the Location field. The Location field gives the URI of the proxy.
+      # The recipient is expected to repeat this single request via the
+      # proxy. 305 responses MUST only be generated by origin servers.
+      #++
+
+      def perform_redirection(response)
+        location = response["Location"]
+        raise RedirectionError.new(response, "Location field-value missed") unless location
+
+        case response
+        when Net::HTTPUseProxy; @proxy = Addressable::URI.parse(location)
+        when Net::HTTPSeeOther; @verb, @uri = Net::HTTP::Get, Addressable::URI.parse(location)
+        else
+          raise RedirectionError.new(response, "automatical redirection is NOT allowed") unless
+            @verb == Net::HTTP::Get ||
+            @verb == Net::HTTP::Head
+          @uri = Addressable::URI.parse(location)
+        end
+
+        perform
+      end
+
+      def handle_response(response)
+        case response
+        when Net::HTTPOK
+          response
+        when Net::HTTPMultipleChoice,   # 300
+             Net::HTTPMovedPermanently, # 301
+             Net::HTTPFound,            # 302
+             Net::HTTPSeeOther,         # 303
+             Net::HTTPUseProxy,         # 305
+             Net::HTTPTemporaryRedirect # 307
+
+          raise RedirectionError.new(response, "redirection level too deep") unless (@limit -= 1) > 0
+          perform_redirection(response)
+        else
+          raise ConnectionError.new(response)
+        end
+      end
+
+      def setup
+        @request = verb.new(@uri.request_uri, @headers)
+        @request.basic_auth(@uri.user, @uri.password) if @uri.user && @uri.password
+        @request.body = @body if @body
+        nil
+      end
+
+      def http
+        http = @proxy ?
+          Net::HTTP.new(@uri.host, @uri.port, @proxy.host, @proxy.port, @proxy.user, @proxy.password) :
+          Net::HTTP.new(@uri.host, @uri.port)
+
+        return http unless @timeout
+        http.open_timeout = @timeout
+        http.read_timeout = @timeout
+        http
+      end
+
+    end
+  end
+end
+
+# EOF

data/lib/smg/mapping.rb

@@ -3,6 +3,7 @@ module SMG #:nodoc:
 
     attr_reader :elements, :nested, :attributes
     attr_reader :root
+    attr_reader :parsed
 
     def initialize
       @elements   = {}
@@ -33,10 +34,19 @@ module SMG #:nodoc:
       @root = normalize_path(path)
     end
 
+    def refresh!
+      @parsed ||= []
+      @parsed.clear
+      nil
+    end
+
     private
 
     def normalize_path(path)
-      path.to_s.split("/")
+      path = path.to_s.squeeze("/")
+      path = path[0..-2] if path[-1] == ?/
+      path = path[1..-1] if path[0] == ?/
+      path.split("/")
     end
 
     def handle_path(path)

data/lib/smg/mapping/element.rb

@@ -2,46 +2,47 @@ module SMG #:nodoc:
   class Mapping #:nodoc:
 
     class Element
-      attr_reader :path, :name, :accessor, :data_class, :cast_to, :at, :context
+      attr_reader :path, :name, :accessor, :data_class, :cast_to, :at, :context, :with
 
       def initialize(path, options = {})
 
-        @name       = (options[:as] || options[:at] || path.last).to_sym
+        @name       = (options[:as] || options[:at] || path.last.gsub(":","_")).to_sym
         @path       = path
         @collection = !!options[:collection]
+        @with       = options[:with] ? normalize_conditions(options[:with]) : nil
         @accessor   = @collection ? :"append_to_#{@name}" : :"#{@name}="
         @data_class = nil
         @cast_to    = nil
         @context    = nil
 
         if options.key?(:context)
-          if Array === options[:context]
-            @context = options[:context].compact
-            @context.uniq!
-            @context = nil if @context.empty?
-          else
-            raise ArgumentError, ":context should be an Array"
-          end
+          raise ArgumentError, "+options[:context]+ should be an Array of Symbols" unless
+            Array === options[:context] &&
+            options[:context].all?{ |c| Symbol === c }
+
+          @context = options[:context].compact
+          @context.uniq!
+          @context = nil if @context.empty?
         end
 
         if options.key?(:class)
           klass = options[:class]
           if SMG::Model === klass
             @data_class = klass
-          elsif TypeCasts.key?(klass)
+          elsif SMG::Mapping::TypeCasts.key?(klass)
             @cast_to = klass
           else
-            raise ArgumentError, ":class should be an SMG::Model or a valid typecast"
+            raise ArgumentError, "+options[:class]+ should be an SMG::Model or a valid typecast"
           end
         end
 
-        #ignore :at on nested collections of resources
+        #ignore options[:at] on nested collections of resources
         @at = options.key?(:at) && !@data_class ? options[:at].to_s : nil
 
       end
 
       def cast(value)
-        @cast_to ? ::SMG::Mapping::TypeCasts[@cast_to, value] : value
+        @cast_to ? SMG::Mapping::TypeCasts[@cast_to, value] : value
       rescue
         raise ArgumentError, "#{value.inspect} is not a valid source for #{@cast_to.inspect}"
       end
@@ -54,6 +55,21 @@ module SMG #:nodoc:
         @context.nil? || @context.include?(context)
       end
 
+      def with?(attrh)
+        @with.nil? || @with.all? { |k,v| attrh[k] == v }
+      end
+
+      private
+
+      def normalize_conditions(conditions)
+        ret = {}
+        conditions.each do |k,v|
+          v = v.to_s unless v.nil?
+          ret[k.to_s] = v
+        end
+        ret
+      end
+
     end
 
   end

data/lib/smg/mapping/typecasts.rb

@@ -24,7 +24,8 @@ module SMG #:nodoc:
       self.typecasts = {}
       self.typecasts[ :string   ] = lambda{ |v| v.to_s }
       self.typecasts[ :integer  ] = lambda{ |v| v.to_i }
-      self.typecasts[ :boolean  ] = lambda{ |v| v.to_s.strip != 'false' }
+      self.typecasts[ :float    ] = lambda{ |v| v.to_f }
+      self.typecasts[ :boolean  ] = lambda{ |v| v.nil? ? nil : (v.to_s.strip != 'false') }
       self.typecasts[ :symbol   ] = lambda{ |v| v.to_sym }
       self.typecasts[ :datetime ] = lambda{ |v| Time.parse(v).utc }
       self.typecasts[ :date     ] = lambda{ |v| Date.parse(v) }

data/lib/smg/model.rb

@@ -44,10 +44,16 @@ module SMG #:nodoc:
     end
 
     def parse(data, context = nil)
-      resource = new
-      doc = SMG::Document.new(resource,context)
+      doc = SMG::Document.new(resource = new,context)
       ::Nokogiri::XML::SAX::Parser.new(doc).parse(data)
-      resource.instance_variable_set(:@_parsed, true)
+      resource.parsed!
+      resource
+    end
+
+    def parse_file(fname, context = nil)
+      doc = SMG::Document.new(resource = new,context)
+      ::Nokogiri::XML::SAX::Parser.new(doc).parse_file(fname)
+      resource.parsed!
       resource
     end
 

data/lib/smg/resource.rb

@@ -6,7 +6,11 @@ module SMG #:nodoc:
     end
 
     def parsed?
-      @_parsed
+      @_parsed ||= false
+    end
+
+    def parsed!
+      @_parsed = true
     end
 
   end

data/spec/collect_spec.rb

@@ -0,0 +1,254 @@
+# encoding: utf-8
+
+require File.expand_path File.join(File.dirname(__FILE__), 'spec_helper')
+
+describe SMG::Model, ".collect" do
+
+  include Spec::Matchers::HaveInstanceMethodMixin
+
+  before :each do
+    @klass = Class.new { include SMG::Resource }
+    @data = data = File.read(FIXTURES_DIR + 'discogs/Genosha+Recordings.xml')
+  end
+
+  it "defines appropriate reader and writer" do
+    @klass.root 'resp/label'
+    @klass.collect 'releases/release/catno', :as => :catalogue_numbers
+    @klass.should have_instance_method 'catalogue_numbers'
+    @klass.should have_instance_method 'append_to_catalogue_numbers'
+  end
+
+  it "never overrides readers" do
+    @klass.root 'resp/label'
+    @klass.collect 'urls/url', :as => :urls
+    @klass.class_eval <<-CODE
+    def urls
+      unless @urls.nil?
+        @urls.map{|u| u.empty? ? nil : u.strip}.compact.join(', ')
+      end
+    end
+    CODE
+    label = @klass.parse(@data)
+    label.urls.should == 'http://www.genosharecordings.com/, http://www.myspace.com/genosharecordings'
+  end
+
+  it "never overrides writers" do
+    @klass.root 'resp/label'
+    @klass.collect 'urls/url', :as => :urls
+    @klass.class_eval <<-CODE
+    def append_to_urls(value)
+      unless value.nil? || value.empty?
+        @urls ||= []
+        @urls << URI.parse(value)
+      end
+    end
+    CODE
+    label = @klass.parse(@data)
+    label.urls.should be_an_instance_of ::Array
+    label.urls.should == [
+      URI.parse('http://www.genosharecordings.com/'),
+      URI.parse('http://www.myspace.com/genosharecordings')
+      ]
+  end
+
+end
+
+describe SMG::Model, ".collect", "without :class option" do
+
+  before :each do
+    @klass = Class.new { include SMG::Resource }
+    @data = data = File.read(FIXTURES_DIR + 'discogs/Genosha+Recordings.xml')
+  end
+
+  it "collects texts" do
+    @klass.root 'resp/label'
+    @klass.collect 'releases/release/catno', :as => :catalogue_numbers
+    label = @klass.parse(@data)
+    label.catalogue_numbers.should be_an_instance_of ::Array
+    label.catalogue_numbers.should == ["GEN 001", "GEN 001", "GEN 002", "GEN 003", "GEN 006", "GEN 008", "GEN 008", "GEN 009", "GEN 012", "GEN 013", "GEN 015", "GEN004", "GEN005", "GEN006½", "GEN007", "GEN009", "GEN010", "GEN011", "GEN012", "GEN014", "GEN016", "GEN017", "GENCD01", "GENOSHA 018"]
+  end
+
+  it "collects attributes" do
+    @klass.root 'resp/label'
+    @klass.collect 'releases/release', :at => :id     , :as => :release_ids
+    @klass.collect 'releases/release', :at => :status , :as => :release_statuses
+    label = @klass.parse(@data)
+    label.release_ids.should be_an_instance_of ::Array
+    label.release_statuses.should be_an_instance_of ::Array
+    label.release_ids.should == ["183713", "1099277", "183735", "225253", "354681", "1079143", "448035", "1083336", "1079145", "814757", "964449", "254166", "341855", "387611", "396345", "448709", "529057", "662611", "683859", "915651", "1021944", "1494949", "354683", "1825580"] 
+    label.release_statuses.should == ["Accepted"]*24
+  end
+
+  it "is able to build multiple datasets" do
+      custom_xml = <<-XML
+<releases>
+  <release id="2259548" status="Accepted" type="TrackAppearance">Signal Flow Podcast 03</release>
+  <release id="2283715" status="Accepted" type="TrackAppearance">United Hardcore Forces</release>
+  <release id="2283652" status="Accepted" type="TrackAppearance">Warp Madness</release>
+  <release id="1775742" status="Accepted" type="UnofficialRelease">Stalker 2.9 Level 3 Compilation</release>
+</release>
+XML
+
+    @klass.collect 'releases/release', :at => :id, :as => :ids
+    @klass.collect 'releases/release', :at => :type, :as => :types
+    @klass.collect 'releases/release', :as => :titles
+
+    collection = @klass.parse(custom_xml)
+    collection.ids.should     == ["2259548", "2283715", "2283652", "1775742"]
+    collection.types.should   == ["TrackAppearance", "TrackAppearance", "TrackAppearance", "UnofficialRelease"]
+    collection.titles.should  == ["Signal Flow Podcast 03", "United Hardcore Forces", "Warp Madness", "Stalker 2.9 Level 3 Compilation"]
+  end
+
+  it "collects nothing when there's no matching elements" do
+    @klass.root 'resp/label'
+    @klass.collect 'releases/release', :at => :bogus, :as => :release_ids
+    label = @klass.parse(@data)
+    label.release_ids.should be_an_instance_of ::Array
+    label.release_ids.should be_empty
+  end
+
+end
+
+describe SMG::Model, ".collect", "when :class option represents SMG::Resource" do
+
+  before :each do
+    @klass = Class.new { include SMG::Resource }
+    @data = data = File.read(FIXTURES_DIR + 'discogs/Genosha+Recordings.xml')
+  end
+
+  before :all do
+    @release_class = Class.new { include SMG::Resource }
+    @release_class.extract 'release'       , :at => :id      , :as => :discogs_id
+    @release_class.extract 'release'       , :at => :status
+    @release_class.extract 'release/title'
+
+    @image_class = Class.new { include SMG::Resource }
+    @image_class.extract 'image', :at => :uri
+    @image_class.extract 'image', :at => :uri150, :as => :preview
+    @image_class.extract 'image', :at => :width
+    @image_class.extract 'image', :at => :height
+  end
+
+  it "collects resources" do
+    @klass.root 'resp/label'
+    @klass.collect 'releases/release', :as => :releases, :class => @release_class
+    label = @klass.parse(@data)
+    label.releases.should be_an_instance_of ::Array
+    label.releases.size.should == 24
+    label.releases[8].title.should == "No, We Don't Want You To Clap Your Fucking Hands"
+    label.releases[8].discogs_id.should == "1079145"
+    label.releases[8].status.should == "Accepted"
+  end
+
+  it "is able to build multiple collections" do
+    @klass.root 'resp/label'
+    @klass.collect 'releases/release' , :as => :releases    , :class => @release_class
+    @klass.collect 'images/image'     , :as => :images      , :class => @image_class
+    @klass.collect 'releases/release' , :as => :release_ids , :at => :id
+    label = @klass.parse(@data)
+
+    label.releases.should be_an_instance_of ::Array
+    label.releases.size.should == 24
+    label.releases[8].should be_an_instance_of @release_class
+    label.releases[8].title.should == "No, We Don't Want You To Clap Your Fucking Hands"
+    label.releases[8].discogs_id.should == "1079145"
+
+    label.images.should be_an_instance_of ::Array
+    label.images.size.should == 2
+    label.images[1].should be_an_instance_of @image_class
+    label.images[1].width.should == '600'
+    label.images[1].height.should == '159'
+    label.images[1].uri.should == 'http://www.discogs.com/image/L-16366-1165574398.jpeg'
+    label.images[1].preview.should == 'http://www.discogs.com/image/L-150-16366-1165574398.jpeg'
+
+    label.release_ids.should == ["183713", "1099277", "183735", "225253", "354681", "1079143", "448035", "1083336", "1079145", "814757", "964449", "254166", "341855", "387611", "396345", "448709", "529057", "662611", "683859", "915651", "1021944", "1494949", "354683", "1825580"] 
+  end
+
+  it "collects nothing when there's no matching elements" do
+    @klass.root 'resp/label'
+    @klass.collect 'releases/bogus', :as => :releases, :class => @release_class
+    label = @klass.parse(@data)
+    label.releases.should be_an_instance_of ::Array
+    label.releases.should be_empty
+  end
+
+end
+
+describe SMG::Model, ".collect", "when :class option represents built-in typecast" do
+
+  before :each do
+    @klass = Class.new { include SMG::Resource }
+    @data = data = File.read(FIXTURES_DIR + 'discogs/Genosha+Recordings.xml')
+  end
+
+  it "makes an attempt to perform a typecast" do
+    Class.new { include SMG::Resource }
+    @klass.root 'resp/label'
+    @klass.collect 'releases/release', :at => :id, :class => :integer, :as => :release_ids
+    label = @klass.parse(@data)
+    label.release_ids.should == [183713, 1099277, 183735, 225253, 354681, 1079143, 448035, 1083336, 1079145, 814757, 964449, 254166, 341855, 387611, 396345, 448709, 529057, 662611, 683859, 915651, 1021944, 1494949, 354683, 1825580] 
+  end
+
+  it "raises an ArgumentError if typecasting fails" do
+    Class.new { include SMG::Resource }
+    @klass.root 'resp/label'
+    @klass.collect 'releases/release', :at => :id, :class => :datetime, :as => :release_ids
+    lambda { @klass.parse(@data) }.
+    should raise_error ArgumentError, %r{"183713" is not a valid source for :datetime} 
+  end
+
+end
+
+describe SMG::Model, ".collect", "with nested collections" do
+
+  before :each do
+    @klass = Class.new { include SMG::Resource }
+    @data = data = File.read(FIXTURES_DIR + 'discogs/948224.xml')
+  end
+
+  before :all do
+    @artist_class = Class.new { include SMG::Resource }
+    @artist_class.root 'artist'
+    @artist_class.extract :name
+    @artist_class.extract :role
+
+    @track_class = Class.new { include SMG::Resource }
+    @track_class.root 'track'
+    @track_class.extract :position
+    @track_class.extract :title
+    @track_class.extract :duration
+    @track_class.collect 'extraartists/artist/name', :as => :extra_artist_names
+    @track_class.collect 'extraartists/artist', :as => :extra_artists, :class => @artist_class
+  end
+
+  it "supports nested collections" do
+    @klass.root 'resp/release'
+    @klass.collect 'tracklist/track'        , :as => :tracks, :class => @track_class
+    @klass.collect 'tracklist/track/title'  , :as => :tracklist
+    release = @klass.parse(@data)
+
+    release.tracklist.should be_an_instance_of ::Array
+    release.tracklist.size.should == 6
+    release.tracklist.should == ["Butterfly V.I.P. (VIP Edit By Ophidian)", "Butterfly V.I.P. (Interlude By Cubist Boy & Tapage)", "Butterfly V.I.P. (Original Version)", "Hammerhead V.I.P. (VIP Edit By Ophidian)", "Hammerhead V.I.P. (Interlude By Cubist Boy & Tapage)", "Hammerhead V.I.P. (Original Version)"]
+
+    release.tracks.should be_an_instance_of ::Array
+    release.tracks.size.should == 6
+
+    track = release.tracks[1]
+    track.position.should == 'A2'
+    track.title.should == 'Butterfly V.I.P. (Interlude By Cubist Boy & Tapage)'
+    track.duration.should == '1:24'
+
+    track.extra_artist_names.should be_an_instance_of ::Array
+    track.extra_artist_names.size.should == 2
+    track.extra_artist_names.should == ['Cubist Boy', 'Tapage']
+
+    track.extra_artists.should be_an_instance_of ::Array
+    track.extra_artists.size.should == 2
+    track.extra_artists[1].name.should == 'Tapage'
+    track.extra_artists[1].role.should == 'Co-producer'
+  end
+
+end
+
+# EOF