milkfarm-onix 0.7.7

data/lib/onix/lists/product_form_detail.rb

@@ -0,0 +1,157 @@
+# coding: utf-8
+
+module ONIX
+  module Lists
+    # Code list 78
+    PRODUCT_FORM_DETAIL = {
+      "A101" => "CD standard audio format",
+      "A102" => "SACD super audio format",
+      "A103" => "MP3 format",
+      "A104" => "WAV format",
+      "A105" => "Real Audio format",
+      "A106" => "WMA",
+      "A107" => "AAC",
+      "A108" => "Ogg/Vorbis",
+      "A109" => "Audible",
+      "A110" => "FLAC",
+      "A111" => "AIFF",
+      "A112" => "ALAC",
+      "A201" => "DAISY 2: full audio with title only (no navigation)",
+      "A202" => "DAISY 2: full audio with navigation",
+      "A203" => "DAISY 2: full audio with navigation and partial text",
+      "A204" => "DAISY 2: full audio and full text",
+      "A205" => "DAISY 2: full text and some audio",
+      "A206" => "DAISY 2: full text and no audio",
+      "A207" => "DAISY 3: full audio with title only (no navigation)",
+      "A208" => "DAISY 3: full audio with navigation",
+      "A209" => "DAISY 3: full audio with navigation and partial text",
+      "A210" => "DAISY 3: full audio and full text",
+      "A211" => "DAISY 3: full text and some audio",
+      "A212" => "DAISY 3: full text and no audio",
+      "B101" => "Mass market (rack) paperback",
+      "B102" => "Trade paperback (US)",
+      "B103" => "Digest format paperback",
+      "B104" => "A-format paperback",
+      "B105" => "B-format paperback",
+      "B106" => "Trade paperback (UK)",
+      "B107" => "Tall rack paperback (US)",
+      "B108" => "A5: Tankobon",
+      "B109" => "B5: Tankobon",
+      "B110" => "B6: Tankobon",
+      "B111" => "A6: Bunko",
+      "B112" => "B40-dori: Shinsho",
+      "B113" => "Pocket (Sweden)",
+      "B114" => "Storpocket (Sweden)",
+      "B115" => "Kartonnage (Sweden)",
+      "B116" => "Flexband (Sweden)",
+      "B201" => "Coloring / join-the-dot book",
+      "B202" => "Lift-the-flap book",
+      "B203" => "Fuzzy book",
+      "B204" => "Miniature book",
+      "B205" => "Moving picture / flicker book",
+      "B206" => "Pop-up book",
+      "B207" => "Scented / 'smelly' book",
+      "B208" => "Sound story / 'noisy' book",
+      "B209" => "Sticker book",
+      "B210" => "Touch-and-feel book",
+      "B211" => "Toy / die-cut book",
+      "B212" => "Die-cut book",
+      "B213" => "Book-as-toy",
+      "B214" => "Soft-to-touch book",
+      "B215" => "Fuzzy-felt book",
+      "B221" => "Picture book",
+      "B301" => "Loose leaf - sheets & binder",
+      "B302" => "Loose leaf - binder only",
+      "B303" => "Loose leaf - sheets only",
+      "B304" => "Sewn",
+      "B305" => "Unsewn / adhesive bound",
+      "B306" => "Library binding",
+      "B307" => "Reinforced binding",
+      "B308" => "Half bound",
+      "B309" => "Quarter bound",
+      "B310" => "Saddle-sewn",
+      "B311" => "Comb bound",
+      "B312" => "Wire-O",
+      "B313" => "Concealed wire",
+      "B401" => "Cloth over boards",
+      "B402" => "Paper over boards",
+      "B403" => "Leather, real",
+      "B404" => "Leather, imitation",
+      "B405" => "Leather, bonded",
+      "B406" => "Vellum",
+      "B407" => "Plastic",
+      "B408" => "Vinyl",
+      "B409" => "Cloth",
+      "B410" => "Imitation cloth",
+      "B411" => "Velvet",
+      "B412" => "Flexible plastic/vinyl cover",
+      "B413" => "Plastic-covered",
+      "B414" => "Vinyl-covered",
+      "B415" => "Laminated cover",
+      "B501" => "With dust jacket",
+      "B502" => "With printed dust jacket",
+      "B503" => "With translucent dust cover",
+      "B504" => "With flaps",
+      "B505" => "With thumb index",
+      "B506" => "With ribbon marker(s)",
+      "B507" => "With zip fastener",
+      "B508" => "With button snap fastener",
+      "B509" => "With leather edge lining",
+      "B601" => "Turn-around book",
+      "B602" => "Unflipped manga format",
+      "B701" => "UK Braille Grade 1",
+      "B702" => "UK Braille Grade 2",
+      "B703" => "US Braille",
+      "D101" => "Real Video format",
+      "D102" => "Quicktime format",
+      "D103" => "AVI format",
+      "D104" => "Windows Media Video format",
+      "D105" => "MPEG-4",
+      "D201" => "MS-DOS",
+      "D202" => "Windows",
+      "D203" => "Macintosh",
+      "D204" => "UNIX / LINUX",
+      "D205" => "Other operating system(s)",
+      "D206" => "Palm OS",
+      "D207" => "Windows Mobile",
+      "D301" => "Microsoft XBox",
+      "D302" => "Nintendo Gameboy Color",
+      "D303" => "Nintendo Gameboy Advanced",
+      "D304" => "Nintendo Gameboy",
+      "D305" => "Nintendo Gamecube",
+      "D306" => "Nintendo 64",
+      "D307" => "Sega Dreamcast",
+      "D308" => "Sega Genesis/Megadrive",
+      "D309" => "Sega Saturn",
+      "D310" => "Sony PlayStation 1",
+      "D311" => "Sony PlayStation 2",
+      "D312" => "Nintendo Dual Screen",
+      "D313" => "Sony PlayStation 3",
+      "D314" => "Xbox 360",
+      "D315" => "Nintendo Wii",
+      "D316" => "Sony PlayStation Portable (PSP)",
+      "L101" => "Laminated",
+      "P101" => "Desk calendar",
+      "P102" => "Mini calendar",
+      "P103" => "Engagement calendar",
+      "P104" => "Day by day calendar",
+      "P105" => "Poster calendar",
+      "P106" => "Wall calendar",
+      "P107" => "Perpetual calendar",
+      "P108" => "Advent calendar",
+      "P109" => "Bookmark calendar",
+      "P110" => "Student calendar",
+      "P111" => "Project calendar",
+      "P112" => "Almanac calendar",
+      "P113" => "Other calendar",
+      "P114" => "Other calendar or organiser product",
+      "P201" => "Hardback (stationery)",
+      "P202" => "Paperback / softback (stationery)",
+      "P203" => "Spiral bound (stationery)",
+      "P204" => "Leather / fine binding (stationery)",
+      "V201" => "PAL",
+      "V202" => "NTSC",
+      "V203" => "SECAM",
+    }
+  end
+end

data/lib/onix/market_representation.rb

@@ -0,0 +1,19 @@
+# coding: utf-8
+
+module ONIX
+  class MarketRepresentation
+    include ROXML
+    include ONIX::Common
+
+    xml_name "MarketRepresentation"
+
+    xml_accessor :agent_name, :from => "AgentName"
+    xml_accessor :agent_role, :from => "AgentRole", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :market_country, :from => "MarketCountry"
+    xml_accessor :market_territory, :from => "MarketTerritory"
+    xml_accessor :market_country_excluded, :from => "MarketCountryExcluded"
+    xml_accessor :market_restriction_detail, :from => "MarketRestrictionDetail"
+    xml_accessor :market_publishing_status, :from => "MarketPublishingStatus", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+  end
+end
+

data/lib/onix/measure.rb

@@ -0,0 +1,14 @@
+# coding: utf-8
+
+module ONIX
+  class Measure
+    include ROXML
+    include ONIX::Common
+
+    xml_name "Measure"
+
+    xml_accessor :measure_type_code, :from => "MeasureTypeCode", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :measurement,       :from => "Measurement", :as => BigDecimal
+    xml_accessor :measure_unit_code, :from => "MeasureUnitCode"
+  end
+end

data/lib/onix/media_file.rb

@@ -0,0 +1,16 @@
+# coding: utf-8
+
+module ONIX
+  class MediaFile
+    include ROXML
+    include ONIX::Common
+
+    xml_name "MediaFile"
+
+    xml_accessor :media_file_type_code, :from => "MediaFileTypeCode", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :media_file_format_code, :from => "MediaFileFormatCode", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :image_resolution, :from => "ImageResolution"
+    xml_accessor :media_file_link_type_code, :from => "MediaFileLinkTypeCode", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :media_file_link, :from => "MediaFileLink"
+  end
+end

data/lib/onix/normaliser.rb

@@ -0,0 +1,187 @@
+# coding: utf-8
+
+require 'tempfile'
+require 'fileutils'
+
+module ONIX
+
+  # A standalone class that can be used to normalise ONIX files
+  # into a standardised form. If you're accepting ONIX files from a wide range
+  # of suppliers, you're guarunteed to get all sorts of dialects.
+  #
+  # This will create a new file that:
+  #
+  # - is UTF-8 encoded
+  # - uses reference tags, not short
+  # - has no named entities (ndash, etc) other than & < and >
+  #
+  # Usage:
+  #
+  #   ONIX::Normaliser.process("oldfile.xml", "newfile.xml")
+  #   
+  # Dependencies:
+  #
+  # At this stage the class depends on several external apps, all commonly available
+  # on *nix systems: xsltproc, isutf8, iconv and sed
+  #
+  class Normaliser
+
+    class << self
+
+      # normalise oldfile and save it as newfile. oldfile
+      # will be left untouched
+      #
+      def process(oldfile, newfile)
+        self.new(oldfile, newfile).run
+      end
+    end
+
+    def initialize(oldfile, newfile)
+      raise ArgumentError, "#{oldfile} does not exist" unless File.file?(oldfile)
+      raise ArgumentError, "#{newfile} already exists" if File.file?(newfile)
+      raise "xsltproc app not found" unless app_available?("xsltproc")
+      raise "isutf8 app not found"   unless app_available?("isutf8")
+      raise "iconv app not found"    unless app_available?("iconv")
+      raise "sed app not found"      unless app_available?("sed")
+      raise "tr app not found"       unless app_available?("tr")
+
+      @oldfile = oldfile
+      @newfile = newfile
+      @curfile = next_tempfile
+      FileUtils.cp(@oldfile, @curfile)
+      @head    = File.open(@oldfile, "r") { |f| f.read(1024) }
+    end
+
+    def run
+      # remove short tags
+      if @head.include?("ONIXmessage")
+        dest = next_tempfile
+        to_reference_tags(@curfile, dest)
+        @curfile = dest
+      end
+
+      # convert to utf8
+      dest = next_tempfile
+      to_utf8(@curfile, dest)
+      @curfile = dest
+
+      # remove control chars
+      dest = next_tempfile
+      remove_control_chars(@curfile, dest)
+      @curfile = dest
+
+      # remove entities
+      dest = next_tempfile
+      replace_named_entities(@curfile, dest)
+      @curfile = dest
+
+      FileUtils.cp(@curfile, @newfile)
+    end
+
+    #private
+
+    # check the specified app is available on the system
+    #
+    def app_available?(app)
+      `which #{app}`.strip == "" ? false : true
+    end
+
+    # generate a temp filename
+    #
+    def next_tempfile
+      p = nil
+      Tempfile.open("onix") do |tf|
+        tf.close
+        p = tf.path
+      end
+      p
+    end
+
+    # uses an XSLT stylesheet provided by edituer to convert
+    # a file from short tags to long tags.
+    #
+    # more detail here:
+    #   http://www.editeur.org/files/ONIX%203/ONIX%20tagname%20converter%20v2.htm
+    #
+    def to_reference_tags(src, dest)
+      inpath = File.expand_path(src)
+      outpath = File.expand_path(dest)
+      xsltpath = File.dirname(__FILE__) + "/../../support/switch-onix-2.1-short-to-reference.xsl"
+      `xsltproc -o #{outpath} #{xsltpath} #{inpath}`
+    end
+
+    # ensure the file is valid utf8, then make sure it's declared as such.
+    #
+    # The following behaviour is  expected:
+    #
+    #   file is valid utf8, is marked correctly
+    #     - copied untouched
+    #   file is valid utf8, is marked incorrectly or has no marked encoding
+    #     - copied and encoding mark fixed or added
+    #   file is no utf8, encoding is marked
+    #     - file is converted to utf8 and enecoding mark is updated
+    #   file is not utf8, encoding is not marked
+    #     - file is copied untouched
+    #
+    def to_utf8(src, dest)
+      inpath = File.expand_path(src)
+      outpath = File.expand_path(dest)
+
+      m, src_enc = *@head.match(/encoding=.([a-zA-Z0-9\-]+)./i)
+
+      # ensure the file is actually utf8
+      if `isutf8 #{inpath}`.strip == ""
+        if src_enc.to_s.downcase == "utf-8"
+          FileUtils.cp(inpath, outpath)
+        else
+          FileUtils.cp(inpath, outpath)
+          `sed -i 's/<?xml.*?>/<?xml version=\"1.0\" encoding=\"UTF-8\"?>/g' #{outpath}`
+        end
+      elsif src_enc
+        `iconv --from-code=#{src_enc} --to-code=UTF-8 #{inpath} > #{outpath}`
+        `sed -i 's/#{src_enc}/UTF-8/' #{outpath}`
+      else
+          FileUtils.cp(inpath, outpath)
+      end
+    end
+
+    # XML files shouldn't contain low ASCII control chars. Strip them.
+    #
+    def remove_control_chars(src, dest)
+      inpath = File.expand_path(src)
+      outpath = File.expand_path(dest)
+      `cat #{inpath} | tr -d "\\000-\\010\\013\\014\\016-\\037" > #{outpath}`
+    end
+
+    # replace all named entities in the specified file with
+    # numeric entities.
+    #
+    def replace_named_entities(src, dest)
+      inpath = File.expand_path(src)
+      outpath = File.expand_path(dest)
+
+      cmd = "sed " + entity_map.map do |named, numeric|
+        "-e 's/\\&#{named};/\\&#{numeric};/g'"
+      end.join(" ") + " #{inpath} > #{outpath}"
+      #raise cmd
+      `#{cmd}`
+    end
+
+    # return a named entity to numeric entity mapping, build by extracting
+    # data from the ONIX DTD
+    #
+    def entity_map
+      return @map if @map
+
+      path = File.dirname(__FILE__) + "/../../support/entities.txt"
+      @map = {}
+      File.read(path).split.each do |line|
+        elements = line.split(":")
+        @map[elements.first] = elements.last
+      end
+      @map
+    end
+
+  end
+
+end

data/lib/onix/other_text.rb

@@ -0,0 +1,17 @@
+# coding: utf-8
+
+module ONIX
+  class OtherText
+    include ROXML
+    include ONIX::Common
+
+    xml_name "OtherText"
+
+    xml_accessor :text_type_code, :from => "TextTypeCode", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :text_format,    :from => "TextFormat"
+    xml_accessor :text,           :from => "Text"
+    xml_accessor :text_link_type, :from => "TextLinkType", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :text_link,      :from => "TextLink"
+    xml_accessor :text_author,    :from => "TextAuthor"
+  end
+end

data/lib/onix/price.rb

@@ -0,0 +1,21 @@
+# coding: utf-8
+
+module ONIX
+  class Price
+    include ROXML
+    include ONIX::Common
+
+    xml_name "Price"
+
+    xml_accessor :price_type_code, :from => "PriceTypeCode", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :price_type_qualifier, :from => "PriceTypeQualifier", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :price_type_description, :from => "PriceTypeDescription"
+    xml_accessor :price_per, :from => "PricePer", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :minimum_order_qty, :from => "MinimumOrderQuantity", :as => Fixnum
+    xml_accessor :class_of_trade, :from => "ClassOfTrade"
+    xml_accessor :bic_discount_group_code, :from => "BICDiscountGroupCode"
+    xml_accessor :price_status, :from => "PriceStatus", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :price_amount, :from => "PriceAmount", :as => BigDecimal, :to_xml => ONIX::Formatters.decimal
+    xml_accessor :currency_code, :from => "CurrencyCode"
+  end
+end

data/lib/onix/product.rb

@@ -0,0 +1,73 @@
+# coding: utf-8
+
+module ONIX
+  class Product
+    include ROXML
+    include ONIX::Common
+
+    xml_name "Product"
+    
+    xml_accessor :record_reference, :from => "RecordReference"
+    xml_accessor :notification_type, :from => "NotificationType", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :product_identifiers, :from => "ProductIdentifier", :as => [ONIX::ProductIdentifier]
+    xml_accessor :product_form, :from => "ProductForm"
+    xml_accessor :product_form_detail, :from => "ProductFormDetail"
+    xml_accessor :series, :from => "Series", :as => [ONIX::Series]
+    xml_accessor :titles, :from => "Title", :as => [ONIX::Title]
+    xml_accessor :websites, :from => "Website", :as => [ONIX::Website]
+    xml_accessor :contributors, :from => "Contributor", :as => [ONIX::Contributor]
+    xml_accessor :edition_number, :from => "EditionNumber", :as => Fixnum
+    xml_accessor :languages, :from => "Language", :as => [ONIX::Language]
+    xml_accessor :number_of_pages, :from => "NumberOfPages", :as => Fixnum
+    xml_accessor :bic_main_subject, :from => "BICMainSubject"
+    xml_accessor :subjects, :from => "Subject", :as => [ONIX::Subject]
+    xml_accessor :audience_code, :from => "AudienceCode", :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :audience_ranges, :from => "AudienceRange", :as => [ONIX::AudienceRange]
+    xml_accessor :text, :from => "OtherText", :as => [ONIX::OtherText]
+    xml_accessor :media_files, :from => "MediaFile", :as => [ONIX::MediaFile]
+    xml_accessor :imprints, :from => "Imprint", :as => [ONIX::Imprint]
+    xml_accessor :publishers, :from => "Publisher", :as => [ONIX::Publisher]
+    xml_accessor :publishing_status, :from => "PublishingStatus", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor(:publication_date, :from => "PublicationDate", :to_xml => ONIX::Formatters.yyyymmdd) do |val|
+      begin
+        Date.parse(val)
+      rescue
+        nil
+      end
+    end
+    xml_accessor :copyright_year, :from => "CopyrightYear", :as => Integer
+    xml_accessor :year_first_published, :from => "YearFirstPublished", :as => Fixnum
+    xml_accessor :sales_restrictions, :from => "SalesRestriction", :as => [ONIX::SalesRestriction]
+    xml_accessor :measurements, :from => "Measure", :as => [ONIX::Measure]
+    xml_accessor :supply_details, :from => "SupplyDetail", :as => [ONIX::SupplyDetail]
+    xml_accessor :market_representations, :from => "MarketRepresentation", :as => [ONIX::MarketRepresentation]
+
+    # some deprecated attributes. Read only
+    # - See the measures array for the current way of specifying
+    #   various measurements of the product
+    xml_reader :height,     :from => "Height", :as => BigDecimal
+    xml_reader :width,      :from => "Width", :as => BigDecimal
+    xml_reader :thickness,  :from => "Thickness", :as => BigDecimal
+    xml_reader :weight,     :from => "Weight", :as => BigDecimal
+    xml_reader :dimensions, :from => "Dimensions"
+
+    def initialize
+      self.product_identifiers = []
+      self.series = []
+      self.titles = []
+      self.contributors = []
+      self.websites = []
+      self.subjects = []
+      self.audience_ranges = []
+      self.text = []
+      self.languages = []
+      self.media_files = []
+      self.imprints = []
+      self.publishers = []
+      self.sales_restrictions = []
+      self.measurements = []
+      self.supply_details = []
+      self.market_representations = []
+    end
+  end
+end

data/lib/onix/product_identifier.rb

@@ -0,0 +1,13 @@
+# coding: utf-8
+
+module ONIX
+  class ProductIdentifier
+    include ROXML
+    include ONIX::Common
+
+    xml_name "ProductIdentifier"
+
+    xml_accessor :product_id_type, :from => "ProductIDType", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :id_value, :from => "IDValue"
+  end
+end

data/lib/onix/publisher.rb

@@ -0,0 +1,16 @@
+# coding: utf-8
+
+module ONIX
+  class Publisher
+    include ROXML
+    include ONIX::Common
+
+    xml_name "Publisher"
+
+    xml_accessor :publishing_role,      :from => "PublishingRole", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :name_code_type,       :from => "NameCodeType", :as => Fixnum, :to_xml => ONIX::Formatters.two_digit
+    xml_accessor :name_code_type_name,  :from => "NameCodeTypeName"
+    xml_accessor :name_code_type_value, :from => "NameCodeTypeValue"
+    xml_accessor :publisher_name,       :from => "PublisherName"
+  end
+end