relaton-w3c 1.0.2 → 1.3.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 215e2c9d5f297c9aa8314eac2ed1c9fde29f9921e25ffb55024c826f02119e1d
-  data.tar.gz: df2f2918cf0e2b6f92139cd270e8e6c4b967b9545fe1373a6ae44f3e4eb00596
+  metadata.gz: d7dd4430200f1bc42ef19c5e7efd4eb0bee2167d17c80ca44533f0b08e6bdfe6
+  data.tar.gz: 62b4b98f90f7541d8099647f287e94cbe62600ef5943bcd7e315b7d8990f8424
 SHA512:
-  metadata.gz: 2445c87f6ddc4cebf8e0839d5c57261d85244574464c30f3941e17205ffc89f617343e7fa59ab51d1b3ff9971c83313cad074e5f78edf62b48d9d7e32b5230fb
-  data.tar.gz: 71a59b22bdaa10181ba60faa84fdc28fe50f22a4ba8429d1609a994130e98d6322df244eaebde827e946293b7f68b183f25e0d327053400a3ef6559f984ec974
+  metadata.gz: 5480d5eb9a41ef48c7e4a9237f793e96016c32ed3b1cd982b784f4e3e8f0ba84de136b13993ef7a5d6d2fda92540cf5be98339fd198bc93e5486e572958a5feb
+  data.tar.gz: 12de49efbd61c4417ba7d31062f63560883cbe46f5ca9405218e84cb7d815f2717ea691cfd83175aff5b072cc44f7c3284fea526991d9395e9c47789d283b593

data/.github/workflows/ubuntu.yml

@@ -35,4 +35,5 @@ jobs:
           bundle install --jobs 4 --retry 3
       - name: Run specs
         run: |
+          unset JAVA_TOOL_OPTIONS
           bundle exec rake

data/.rubocop.yml

@@ -5,6 +5,6 @@
 inherit_from:
   - https://raw.githubusercontent.com/riboseinc/oss-guides/master/ci/rubocop.yml
 AllCops:
-  TargetRubyVersion: 2.3
+  TargetRubyVersion: 2.4
 Rails:
-  Enabled: true
+  Enabled: false

data/grammars/biblio.rng

@@ -88,7 +88,7 @@
       <text/>
     </element>
   </define>
-  <define name="LocalizedString">
+  <define name="LocalizedString1">
     <optional>
       <!-- multiple languages and scripts possible: comma delimit them if so -->
       <attribute name="language"/>
@@ -98,6 +98,16 @@
     </optional>
     <text/>
   </define>
+  <define name="LocalizedString">
+    <choice>
+      <ref name="LocalizedString1"/>
+      <oneOrMore>
+        <element name="variant">
+          <ref name="LocalizedString1"/>
+        </element>
+      </oneOrMore>
+    </choice>
+  </define>
   <!--
     Unlike UML, change type to format: type is overloaded
     Would be need if plain were default value and could omit the attribute
@@ -121,7 +131,7 @@
     </optional>
     <ref name="LocalizedStringOrXsAny"/>
   </define>
-  <define name="LocalizedStringOrXsAny">
+  <define name="LocalizedStringOrXsAny1">
     <optional>
       <!-- multiple languages and scripts possible: comma delimit them if so -->
       <attribute name="language"/>
@@ -136,6 +146,16 @@
       </choice>
     </oneOrMore>
   </define>
+  <define name="LocalizedStringOrXsAny">
+    <choice>
+      <ref name="LocalizedStringOrXsAny1"/>
+      <oneOrMore>
+        <element name="variant">
+          <ref name="LocalizedStringOrXsAny1"/>
+        </element>
+      </oneOrMore>
+    </choice>
+  </define>
   <define name="contributor">
     <element name="contributor">
       <zeroOrMore>
@@ -512,7 +532,7 @@
   </define>
   <define name="LocalityType">
     <data type="string">
-      <param name="pattern">section|clause|part|paragraph|chapter|page|whole|table|annex|figure|note|list|example|volume|issue|time|locality:[a-zA-Z0-9_]+</param>
+      <param name="pattern">section|clause|part|paragraph|chapter|page|whole|table|annex|figure|note|list|example|volume|issue|time|anchor|locality:[a-zA-Z0-9_]+</param>
     </data>
   </define>
   <define name="referenceFrom">
@@ -641,9 +661,9 @@
     <optional>
       <ref name="status"/>
     </optional>
-    <optional>
+    <zeroOrMore>
       <ref name="copyright"/>
-    </optional>
+    </zeroOrMore>
     <zeroOrMore>
       <ref name="docrelation"/>
     </zeroOrMore>
@@ -1001,7 +1021,17 @@
       <optional>
         <ref name="to"/>
       </optional>
-      <ref name="owner"/>
+      <oneOrMore>
+        <ref name="owner"/>
+      </oneOrMore>
+      <optional>
+        <ref name="copyright_scope"/>
+      </optional>
+    </element>
+  </define>
+  <define name="copyright_scope">
+    <element name="scope">
+      <text/>
     </element>
   </define>
   <define name="from">

data/lib/relaton_w3c/hash_converter.rb

@@ -1,4 +1,11 @@
 module RelatonW3c
   class HashConverter < RelatonBib::HashConverter
+    class << self
+      # @param item_hash [Hash]
+      # @return [RelatonW3c::W3cBibliographicItem]
+      def bib_item(item_hash)
+        W3cBibliographicItem.new item_hash
+      end
+    end
   end
 end

data/lib/relaton_w3c/hit_collection.rb

@@ -45,7 +45,9 @@ module RelatonW3c
       /(?<title>.+)\s(?<date>\d{4}-\d{2}-\d{2})$/ =~ title_date
       title ||= title_date
       result = data.select do |hit|
-        hit["title"] == title && type_date_filter(hit, type, date)
+        (hit["title"].casecmp?(title) ||
+          hit["link"].split("/").last.match?(/#{title}/)) &&
+          type_date_filter(hit, type, date)
       end
       result.map { |h| Hit.new(h, self) }
     end
@@ -54,7 +56,7 @@ module RelatonW3c
     # @param type [String]
     # @param date [String]
     # @return [TrueClass, FalseClass]
-    def type_date_filter(hit, type, date)
+    def type_date_filter(hit, type, date) # rubocop:disable Metrics/AbcSize
       if type && hit["type"] != short_type(type) || date && hit["date"] != date
         history = get_history hit, type, date
         return false unless history.any?

data/lib/relaton_w3c/scrapper.rb

@@ -13,11 +13,12 @@ module RelatonW3c
 
       # @param hit [Hash]
       # @return [RelatonW3c::W3cBibliographicItem]
-      def parse_page(hit)
+      def parse_page(hit) # rubocop:disable Metrics/AbcSize, Metrics/MethodLength
         resp = Net::HTTP.get_response URI.parse(hit["link"])
         doc = resp.code == "200" ? Nokogiri::HTML(resp.body) : nil
         W3cBibliographicItem.new(
           type: "standard",
+          docid: fetch_docid(hit),
           fetched: Date.today.to_s,
           language: ["en"],
           script: ["Latn"],
@@ -28,28 +29,37 @@ module RelatonW3c
           doctype: fetch_doctype(hit, doc),
           contributor: fetch_contributor(hit, doc),
           relation: fetch_relation(doc),
-          keyword: hit["keyword"],
+          keyword: hit["keyword"]
         )
       end
 
       private
 
+      # @param hit [Hash]
+      # @return [Array<RelatonBib::DocumentIdentifier>]
+      def fetch_docid(hit)
+        id = hit["link"].split("/").last
+        [RelatonBib::DocumentIdentifier.new(id: id, type: "W3C")]
+      end
+
       # @param hit [Hash]
       # @param doc [Nokogiri::HTML::Document]
       # @return [Array<RelatonBib::TypedTitleString>]
-      def fetch_title(hit, doc)
+      def fetch_title(hit, doc) # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/MethodLength, Metrics/PerceivedComplexity
         titles = []
         if doc
-          title = doc.at("//h1[@id='title']").text
-          titles << { content: title, type: "main" }
-          subtitle = doc.at("//h2[@id='subtitle']").text
-          titles << { content: subtitle, tipe: "subtitle" }
+          title = doc.at("//h1[contains(@id, 'title')]")&.text
+          titles << { content: title, type: "main" } if title
+          subtitle = doc.at(
+            "//h2[@id='subtitle']|//p[contains(@class, 'subline')]"
+          )&.text
+          titles << { content: subtitle, tipe: "subtitle" } if subtitle
         elsif hit["title"]
           titles << { content: hit["title"], type: "main" }
         end
         titles.map do |t|
           title = RelatonBib::FormattedString.new(
-            content: t[:content], language: "en", script: "Latn",
+            content: t[:content], language: "en", script: "Latn"
           )
           RelatonBib::TypedTitleString.new(type: t[:type], title: title)
         end
@@ -75,10 +85,27 @@ module RelatonW3c
       # @param doc [Nokogiri::HTML::Document, NilClass]
       # @return [Array<RelatonBib::BibliographicDate>]
       def fetch_date(hit, doc)
-        on = hit["datepub"] || doc && doc.at("//h2/time[@datetime]")[:datetime]
+        on = hit["datepub"] || doc&.at("//h2/time[@datetime]")&.attr(:datetime)
+        on ||= fetch_date1(doc) || fetch_date2(doc)
         [RelatonBib::BibliographicDate.new(type: "published", on: on)] if on
       end
 
+      # @param doc [Nokogiri::HTML::Document, NilClass]
+      # @return [String]
+      def fetch_date1(doc)
+        d = doc&.at("//h2[@property='dc:issued']")&.attr(:content)
+        d&.match(/\d{4}-\d{2}-\d{2}/)&.to_s
+      end
+
+      # @param doc [Nokogiri::HTML::Document, NilClass]
+      # @return [String]
+      def fetch_date2(doc)
+        d = doc&.at("//h2[contains(@id, 'w3c-recommendation')]")
+        return unless d
+
+        Date.parse(d.attr(:id.match(/\d{2}-\w+-\d{4}/).to_s)).to_s
+      end
+
       # @param hit [Hash]
       # @param doc [Nokogiri::HTML::Document, NilClass]
       # @return [String]
@@ -96,17 +123,19 @@ module RelatonW3c
       # @param hit [Hash]
       # @param doc [Nokogiri::HTML::Document, NilClass]
       # @return [Array<RelatonBib::ContributionInfo>]
-      def fetch_contributor(hit, doc)
+      def fetch_contributor(hit, doc) # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/MethodLength, Metrics/PerceivedComplexity
         if doc
-          editors = find_contribs(doc, "Editors").map do |ed|
-            parse_contrib ed, "editor"
+          editors = find_contribs(doc, "Editors").reduce([]) do |mem, ed|
+            c = parse_contrib ed, "editor"
+            mem << c if c
+            mem
           end
-          contribs = find_contribs(doc, "Authors").reduce(editors) do |mem, athr|
-            ed = mem.detect { |e| e[:id] && e[:id] == athr["data-editor-id"] }
+          contribs = find_contribs(doc, "Authors").reduce(editors) do |mem, ath|
+            ed = mem.detect { |e| e[:id] && e[:id] == ath["data-editor-id"] }
             if ed
               ed[:role] << { type: "author" }
             else
-              mem << parse_contrib(athr, "author")
+              mem << parse_contrib(ath, "author")
             end
             mem
           end
@@ -131,6 +160,8 @@ module RelatonW3c
       # @return [Hash]
       def parse_contrib(element, type)
         p = element.at("a")
+        return unless p
+
         contrib = {
           name: p.text,
           url: p[:href],

data/lib/relaton_w3c/version.rb

@@ -1,3 +1,3 @@
 module RelatonW3c
-  VERSION = "1.0.2".freeze
+  VERSION = "1.3.2".freeze
 end

data/lib/relaton_w3c/w3c_bibliographic_item.rb

@@ -5,35 +5,12 @@ module RelatonW3c
       proposedRecommendation recommendation retired workingDraft
     ].freeze
 
-    attr_reader :doctype
-
     # @param doctype [String]
     def initialize(**args)
       if args[:doctype] && !TYPES.include?(args[:doctype])
         warn "[relaton-w3c] invalid document type: #{args[:doctype]}"
       end
-      @doctype = args.delete :doctype
       super **args
     end
-
-    # @param builder [Nokogiri::XML::Builder, NilClass]
-    # @param opts [Hash]
-    # @option opts [TrueClass, FalseClass, NilClass] bibdata
-    def to_xml(builder = nil, **opts)
-      super builder, **opts do |b|
-        if opts[:bibdata] && doctype
-          b.ext do |e|
-            e.doctype doctype if doctype
-          end
-        end
-      end
-    end
-
-    # @return [Hash]
-    def to_hash
-      hash = super
-      hash["doctype"] = doctype if doctype
-      hash
-    end
   end
 end

data/lib/relaton_w3c/xml_parser.rb

@@ -1,19 +1,6 @@
 module RelatonW3c
   class XMLParser < RelatonBib::XMLParser
     class << self
-      # @param xml [String]
-      # @return [RelatonW3c::W3cBibliographicItem, NilClass]
-      def from_xml(xml)
-        doc = Nokogiri::XML xml
-        doc.remove_namespaces!
-        item = doc.at("/bibitem|/bibdata")
-        if item
-          W3cBibliographicItem.new(item_data(item))
-        else
-          warn "[relaton-w3c] can't find bibitem or bibdata element in the XML"
-        end
-      end
-
       private
 
       # Override RelatonBib::XMLParser.item_data method.
@@ -27,6 +14,12 @@ module RelatonW3c
         data[:doctype] = ext.at("./doctype")&.text
         data
       end
+
+      # @param item_hash [Hash]
+      # @return [RelatonBib::BibliographicItem]
+      def bib_item(item_hash)
+        W3cBibliographicItem.new item_hash
+      end
     end
   end
 end

data/relaton_w3c.gemspec

@@ -39,5 +39,5 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency "vcr"
   spec.add_development_dependency "webmock"
 
-  spec.add_dependency "relaton-bib", ">= 1.0.4"
+  spec.add_dependency "relaton-bib", ">= 1.3.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: relaton-w3c
 version: !ruby/object:Gem::Version
-  version: 1.0.2
+  version: 1.3.2
 platform: ruby
 authors:
 - Ribose Inc.
-autorequire: 
+autorequire:
 bindir: exe
 cert_chain: []
-date: 2020-05-26 00:00:00.000000000 Z
+date: 2020-09-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: debase
@@ -114,14 +114,14 @@ dependencies:
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 1.0.4
+        version: 1.3.0
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 1.0.4
+        version: 1.3.0
 description: 'RelatonIso: retrieve W3C Standards for bibliographic using the IsoBibliographicItem
   model'
 email:
@@ -162,7 +162,7 @@ licenses:
 - BSD-2-Clause
 metadata:
   homepage_uri: https://github.com/relaton/relaton-wc3
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -178,7 +178,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubygems_version: 3.0.6
-signing_key: 
+signing_key:
 specification_version: 4
 summary: 'RelatonIso: retrieve W3C Standards for bibliographic using the IsoBibliographicItem
   model'