relaton-w3c 1.3.0 → 1.5.pre

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: '0609ad90edb48cdb1289f78c18cc532b50b5fe4b7e734dfb3e3d093985978494'
-  data.tar.gz: c61ffb78a67b314c325f06806ec63b0265a664837b3d36b1ee81eeaddc917a8c
+  metadata.gz: 1e722b7477c8308d552d6074c1ff45e0475f4b514972f84d616a979c6cad47de
+  data.tar.gz: 50e24a13d39126eeec22a7dc04c2c8ed51ded7f57d3ab8be88a5c834808f3d15
 SHA512:
-  metadata.gz: 14f21761fb55b87d2b4bbb1cc6777181c18b9cd591e63cc3ab6d2f910304cda96a84aa4dfe848ef90de0fb0baea0019b3cf017c47c971ad9e8693b33fedecb5f
-  data.tar.gz: f7ade432b4e528ab2b9bc1c0a4e29fb7b5cdac64cd574f1adcf2e698a0e6bea8fdc4b86d8b5dc7052f1c183b33699046cfb51b4a99fe280c884d101c6f5e81bb
+  metadata.gz: 4dba20d4305cd58b0542b4ac6f455a87a2fd96d44513ea34a576987df910891e0df7e64eede636f76c9d9499706bac2eaa90baa238bcc527a470bc555ed707b3
+  data.tar.gz: 39cfc5b3e6138b6d209b193cc94fa6fb6b2f56d2ab78284e0b2b950cd08707c8dd3c67d35afbf1a9cc1fa0a7a07cf598065f62a085d26948e11173f59580d40e

data/README.adoc

@@ -10,7 +10,7 @@ Add this line to your application's Gemfile:
 
 [source,ruby]
 ----
-gem 'relaton_w3c'
+gem 'relaton-w3c'
 ----
 
 And then execute:
@@ -19,14 +19,16 @@ And then execute:
 
 Or install it yourself as:
 
-    $ gem install relaton_w3c
+    $ gem install relaton-w3c
 
 == Usage
 
 === Search for a standard using keywords
 
+[source,ruby]
 ----
 require 'relaton_w3c'
+=> true
 
 hits = RelatonW3c::W3cBibliography.search("W3C JSON-LD 1.1")
 => <RelatonW3c::HitCollection:0x007f93b5e4ff48 @ref=W3C JSON-LD 1.1 @fetched=false>
@@ -38,40 +40,43 @@ item = hits[0].fetch
 
 === XML serialization
 
+[source,ruby]
 ----
 item.to_xml
-=> "<bibitem type="standard">
-        <fetched>2020-04-07</fetched>
-        <title type="main" format="text/plain" language="en" script="Latn">JSON-LD 1.1</title>
-        <title format="text/plain" language="en" script="Latn">A JSON-based Serialization for Linked Data</title>
-        <uri type="src">https://www.w3.org/TR/2020/CR-json-ld11-20200316/</uri>
-        <date type="published">
-            <on>2020</on>
-        </date>
-        ...
+=> "<bibitem id="CR-json-ld11-20200316" type="standard">
+      <fetched>2020-04-07</fetched>
+      <title type="main" format="text/plain" language="en" script="Latn">JSON-LD 1.1</title>
+      <title format="text/plain" language="en" script="Latn">A JSON-based Serialization for Linked Data</title>
+      <uri type="src">https://www.w3.org/TR/2020/CR-json-ld11-20200316/</uri>
+      <date type="published">
+          <on>2020</on>
+      </date>
+      ...
     </bibitem>"
 ----
 
 With argument `bibdata: true` it ouputs XML wrapped by `bibdata` element and adds flavour `ext` element.
 
+[source,ruby]
 ----
 item.to_xml bibdata: true
 => "<bibdata type="standard">
-        <fetched>2020-04-07</fetched>
-        <title type="main" format="text/plain" language="en" script="Latn">JSON-LD 1.1</title>
-        <title format="text/plain" language="en" script="Latn">A JSON-based Serialization for Linked Data</title>
-        <uri type="src">https://www.w3.org/TR/2020/CR-json-ld11-20200316/</uri>
-        <date type="published">
-            <on>2020</on>
-        </date>
-        ...
-        <ext>
-            <doctype>candidateRecommendation</doctype>
-        </ext>
+      <fetched>2020-04-07</fetched>
+      <title type="main" format="text/plain" language="en" script="Latn">JSON-LD 1.1</title>
+      <title format="text/plain" language="en" script="Latn">A JSON-based Serialization for Linked Data</title>
+      <uri type="src">https://www.w3.org/TR/2020/CR-json-ld11-20200316/</uri>
+      <date type="published">
+          <on>2020</on>
+      </date>
+      ...
+      <ext>
+        <doctype>candidateRecommendation</doctype>
+      </ext>
     </bibdata>"
 ----
 
 === Get document by title
+[source,ruby]
 ----
 RelatonW3c::W3cBibliography.get "W3C JSON-LD 1.1"
 [relaton-w3c] ("W3C JSON-LD 1.1") fetching...
@@ -81,6 +86,7 @@ RelatonW3c::W3cBibliography.get "W3C JSON-LD 1.1"
 ----
 
 === Get document by title and type
+[source,ruby]
 ----
 RelatonW3c::W3cBibliography.get "W3C Candidate Recommendation JSON-LD 1.1"
 [relaton-w3c] ("W3C Candidate Recommendation JSON-LD 1.1") fetching...
@@ -90,6 +96,7 @@ RelatonW3c::W3cBibliography.get "W3C Candidate Recommendation JSON-LD 1.1"
 ----
 
 === Get document by title and short type
+[source,ruby]
 ----
 RelatonW3c::W3cBibliography.get "W3C CR JSON-LD 1.1"
 [relaton-w3c] ("W3C CR JSON-LD 1.1") fetching...
@@ -99,6 +106,7 @@ RelatonW3c::W3cBibliography.get "W3C CR JSON-LD 1.1"
 ----
 
 === Get document by title, type, and date
+[source,ruby]
 ----
 RelatonW3c::W3cBibliography.get "W3C WD JSON-LD 1.1 2019-10-18"
 [relaton-w3c] ("W3C WD JSON-LD 1.1 2019-10-18") fetching...
@@ -108,6 +116,7 @@ RelatonW3c::W3cBibliography.get "W3C WD JSON-LD 1.1 2019-10-18"
 ----
 
 === Create bibliographic item from XML
+[source,ruby]
 ----
 RelatonW3c::XMLParser.from_xml File.read('spec/fixtures/cr_json_ld11.xml')
 => #<RelatonW3c::W3cBibliographicItem:0x007f9381efce98
@@ -115,6 +124,7 @@ RelatonW3c::XMLParser.from_xml File.read('spec/fixtures/cr_json_ld11.xml')
 ----
 
 === Create bibliographic item from YAML
+[source,ruby]
 ----
 hash = YAML.load_file 'spec/fixtures/cr_json_ld11.yml'
 => {"title"=>

data/lib/relaton_w3c/hit_collection.rb

@@ -45,7 +45,12 @@ 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
+      if result.empty?
+        result = data.select { |h| h["link"].split("/").last.match? /#{title}/ }
       end
       result.map { |h| Hit.new(h, self) }
     end
@@ -54,7 +59,7 @@ 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 +29,38 @@ 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" }
-        elsif hit["title"]
+          title = doc.at("//*[contains(@id, 'title')]")&.text
+          titles << { content: title, type: "main" } if title && !title.empty?
+          subtitle = doc.at(
+            "//h2[@id='subtitle']|//p[contains(@class, 'subline')]"
+          )&.text
+          titles << { content: subtitle, tipe: "subtitle" } if subtitle
+        end
+        if titles.empty? && 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 +86,27 @@ 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 +124,19 @@ 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 +161,8 @@ 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 = "1.3.0".freeze
+  VERSION = "1.5.pre".freeze
 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", ">= 1.3.0"
+  spec.add_dependency "relaton-bib", ">= 1.5.pre"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: relaton-w3c
 version: !ruby/object:Gem::Version
-  version: 1.3.0
+  version: 1.5.pre
 platform: ruby
 authors:
 - Ribose Inc.
-autorequire: 
+autorequire:
 bindir: exe
 cert_chain: []
-date: 2020-08-21 00:00:00.000000000 Z
+date: 2020-10-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: debase
@@ -114,14 +114,14 @@ dependencies:
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 1.3.0
+        version: 1.5.pre
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 1.3.0
+        version: 1.5.pre
 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
@@ -173,12 +173,12 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: 2.4.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - ">="
+  - - ">"
     - !ruby/object:Gem::Version
-      version: '0'
+      version: 1.3.1
 requirements: []
 rubygems_version: 3.0.6
-signing_key: 
+signing_key:
 specification_version: 4
 summary: 'RelatonIso: retrieve W3C Standards for bibliographic using the IsoBibliographicItem
   model'