relaton-itu 1.7.3 → 1.7.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 971645ffb5b0111df7d841c47f6ac1a64904cc85274218f2f6ff622dbba3b576
-  data.tar.gz: 68bc7e0bbb844896c7bf30c6d3e5747a9d6f0e68aa9384f9e761f2881f3be2bc
+  metadata.gz: 249fe896ec8a77979ca15d6a42da98ad2ac2620cfe8dc0f468cd14277c5a35b0
+  data.tar.gz: 62415ed835abc49cf00d3048b52556b3f718a4ad0dc531ec2c20572b95305210
 SHA512:
-  metadata.gz: 83a9d8ee974bfe44cad5715db0199dfcf1eec9709d2a1d778d08cdc21379a69063bb6d6f0ee818b962f825c684d680bb4dda5cbba9b0ca48b24dc8dde4876cdd
-  data.tar.gz: 96a7cf63dd9ff84a115959d7766544e8d8f5f1328fa976737875b7808152939f1168e6a1644e0eb5e4af5cd22f6077b232cf45968e6945bc1a86d2a7721f8e19
+  metadata.gz: 65a5bcf91f851cc4ec3139fad83b0c83f143b1bafefe8c9638e34072f7d82f76b77517fedf153fcdf7b63ee903fcd372c64e378edaeb231835f477f004f7e94a
+  data.tar.gz: 61cdc7df34b24f5d3f3e56b967e9e2b34337bc164b32691747853660cbb6c9e337a6eee058162f97e218da356f4028d7c107b4be4ed7dc10f99253e6985fccd9

data/.github/workflows/rake.yml

@@ -0,0 +1,46 @@
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
+name: rake
+
+on:
+  push:
+    branches: [ master, main ]
+    tags: [ v* ]
+  pull_request:
+
+jobs:
+  rake:
+    name: Test on Ruby ${{ matrix.ruby }} ${{ matrix.os }}
+    runs-on: ${{ matrix.os }}
+    continue-on-error: ${{ matrix.experimental }}
+    strategy:
+      fail-fast: false
+      matrix:
+        ruby: [ '2.7', '2.6', '2.5', '2.4' ]
+        os: [ ubuntu-latest, windows-latest, macos-latest ]
+        experimental: [ false ]
+        include:
+          - ruby: '3.0'
+            os: 'ubuntu-latest'
+            experimental: true
+          - ruby: '3.0'
+            os: 'windows-latest'
+            experimental: true
+          - ruby: '3.0'
+            os: 'macos-latest'
+            experimental: true
+    steps:
+      - uses: actions/checkout@v2
+        with:
+          submodules: true
+
+      # https://github.com/ruby-debug/debase/issues/89#issuecomment-686827382
+      - if: matrix.os == 'macos-latest' && matrix.ruby == '2.5'
+        run: echo BUNDLE_BUILD__DEBASE="--with-cflags=\"-Wno-error=implicit-function-declaration\"" >> $GITHUB_ENV
+
+      - uses: ruby/setup-ruby@v1
+        with:
+          ruby-version: ${{ matrix.ruby }}
+          bundler-cache: true
+
+      - run: bundle exec rake

data/README.adoc

@@ -97,6 +97,29 @@ RelatonItu::ItuBibliography.get "ITU-T G.989.2/Amd 1"
 ...
 ----
 
+=== Get ITU-R documents
+
+[source,ruby]
+----
+RelatonItu::ItuBibliography.get "ITU-R REC-BO.600-1"
+[relaton-itu] ("ITU-R REC-BO.600-1") fetching...
+[relaton-itu] ("ITU-R REC-BO.600-1") found R-REC-BO.600-1
+=> #<RelatonItu::ItuBibliographicItem:0x007f84e9930498
+...
+----
+
+=== Get ITU-R Radio Regulations (RR)
+
+[source,ruby]
+----
+RelatonItu::ItuBibliography.get 'ITU-R RR (2020)'
+[relaton-itu] ("ITU-R RR") fetching...
+[relaton-itu] WARNING: invalid doctype: publication
+[relaton-itu] ("ITU-R RR") found ITU-R RR
+=> #<RelatonItu::ItuBibliographicItem:0x007fb4b55eb660
+...
+----
+
 === Create bibliographic item form YAML
 [source,ruby]
 ----
@@ -104,11 +127,7 @@ hash = YAML.load_file 'spec/examples/itu_bib_item.yml'
 => {"id"=>"ITU-T L.163 (11/2018)",
 ...
 
-bib_hash = RelatonItu::HashConverter.hash_to_bib hash
-=> {:id=>"ITU-T L.163 (11/2018)",
-...
-
-RelatonItu::ItuBibliographicItem.new bib_hash
+RelatonItu::ItuBibliographicItem.from_hash hash
 => #<RelatonItu::ItuBibliographicItem:0x007fd88ac02aa0
 ...
 ----

data/bin/rspec

@@ -0,0 +1,29 @@
+#!/usr/bin/env ruby
+# frozen_string_literal: true
+
+#
+# This file was generated by Bundler.
+#
+# The application 'rspec' is installed as part of a gem, and
+# this file is here to facilitate running it.
+#
+
+require "pathname"
+ENV["BUNDLE_GEMFILE"] ||= File.expand_path("../../Gemfile",
+  Pathname.new(__FILE__).realpath)
+
+bundle_binstub = File.expand_path("../bundle", __FILE__)
+
+if File.file?(bundle_binstub)
+  if File.read(bundle_binstub, 300) =~ /This file was generated by Bundler/
+    load(bundle_binstub)
+  else
+    abort("Your `bin/bundle` was not generated by Bundler, so this binstub cannot run.
+Replace `bin/bundle` by running `bundle binstubs bundler --force`, then run this command again.")
+  end
+end
+
+require "rubygems"
+require "bundler/setup"
+
+load Gem.bin_path("rspec-core", "rspec")

data/lib/relaton_itu.rb

@@ -1,3 +1,4 @@
+require "mechanize"
 require "relaton_itu/version"
 require "relaton_itu/itu_bibliography"
 require "digest/md5"

data/lib/relaton_itu/editorial_group.rb

@@ -20,7 +20,7 @@ module RelatonItu
         warn "[relaton-itu] WARNING: invalid bureau: #{bureau}"
       end
       @bureau = bureau
-      @group = group.is_a?(Hash) ? ItuGroup.new(group) : group
+      @group = group.is_a?(Hash) ? ItuGroup.new(**group) : group
       @subgroup = subgroup.is_a?(Hash) ? ItuGroup.new(subgroup) : subgroup
       @workgroup = workgroup.is_a?(Hash) ? ItuGroup.new(workgroup) : workgroup
     end

data/lib/relaton_itu/hash_converter.rb

@@ -3,11 +3,20 @@ module RelatonItu
     class << self
       private
 
+      #
+      # Ovverides superclass's method
+      #
+      # @param item [Hash]
+      # @retirn [RelatonItu::ItuBibliographicItem]
+      def bib_item(item)
+        ItuBibliographicItem.new(**item)
+      end
+
       def editorialgroup_hash_to_bib(ret)
         eg = ret[:editorialgroup]
         return unless eg
 
-        ret[:editorialgroup] = EditorialGroup.new eg
+        ret[:editorialgroup] = EditorialGroup.new **eg
       end
 
       # @param ret [Hash]
@@ -15,7 +24,7 @@ module RelatonItu
         return unless ret[:structuredidentifier]
 
         ret[:structuredidentifier] = StructuredIdentifier.new(
-          ret[:structuredidentifier]
+          **ret[:structuredidentifier]
         )
       end
     end

data/lib/relaton_itu/hit.rb

@@ -3,10 +3,12 @@
 module RelatonItu
   # Hit.
   class Hit < RelatonBib::Hit
+    attr_writer :fetch
+
     # Parse page.
     # @return [RelatonItu::ItuBibliographicItem]
     def fetch
-      @fetch ||= Scrapper.parse_page hit, hit_collection.gi_imp
+      @fetch ||= Scrapper.parse_page self, hit_collection.gi_imp
     end
   end
 end

data/lib/relaton_itu/hit_collection.rb

@@ -12,25 +12,48 @@ module RelatonItu
     # @return [TrueClass, FalseClass]
     attr_reader :gi_imp
 
+    # @return [Mechanize]
+    attr_reader :agent
+
     # @param ref [String]
     # @param year [String]
-    def initialize(ref, year = nil)
+    def initialize(ref, year = nil) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength
       text = ref.sub /(?<=\.)Imp\s?(?=\d)/, ""
       super text, year
+      @agent = Mechanize.new
+      agent.user_agent_alias = "Mac Safari"
       @gi_imp = /\.Imp\d/.match?(ref)
-      uri = URI "#{DOMAIN}/net4/ITU-T/search/GlobalSearch/Search"
-      data = { json: params.to_json }
-      resp = Net::HTTP.post(uri, data.to_json,
-                            "Content-Type" => "application/json")
-      @array = hits JSON.parse(resp.body)
+      if ref.match? /^(ITU-T|ITU-R\sRR)/
+        url = "#{DOMAIN}/net4/ITU-T/search/GlobalSearch/Search"
+        data = { json: params.to_json }
+        resp = agent.post url, data.to_json, "Content-Type" => "application/json"
+        @array = hits JSON.parse(resp.body)
+      elsif ref.match? /^ITU-R/
+        rf = ref.sub(/^ITU-R\s/, "").upcase
+        url = "https://raw.githubusercontent.com/relaton/relaton-data-itu-r/master/data/#{rf}.yaml"
+        resp = Net::HTTP.get_response(URI(url))
+        if resp.code == "404"
+          @array = []
+          return
+        end
+
+        hash = YAML.safe_load resp.body
+        item_hash = HashConverter.hash_to_bib(hash)
+        item = ItuBibliographicItem.new **item_hash
+        hit = Hit.new({ url: url }, self)
+        hit.fetch = item
+        @array = [hit]
+      end
     end
 
     private
 
     # @return [String]
     def group
-      @group ||= if %r{OB|Operational Bulletin}.match? text then "Publications"
-                 else "Recommendations"
+      @group ||= case text
+                 when %r{OB|Operational Bulletin}, %r{^ITU-R\sRR}
+                   "Publications"
+                 when %r{^ITU-T} then "Recommendations"
                  end
     end
 
@@ -46,7 +69,7 @@ module RelatonItu
         "ExactPhrase" => false,
         "CollectionName" => "General",
         "CollectionGroup" => group,
-        "Sector" => "t",
+        "Sector" => text.match(/(?<=^ITU-)\w/).to_s.downcase,
         "Criterias" => [{
           "Name" => "Search in",
           "Criterias" => [
@@ -93,7 +116,7 @@ module RelatonItu
         code  = h["Media"]["Name"]
         title = h["Title"]
         url   = h["Redirection"]
-        type  = group.downcase[0...-1]
+        type  = h["Collection"]["Group"].downcase[0...-1]
         Hit.new({ code: code, title: title, url: url, type: type }, self)
       end
     end

data/lib/relaton_itu/itu_bibliographic_item.rb

@@ -14,5 +14,12 @@ module RelatonItu
       end
       super
     end
+
+    # @param hash [Hash]
+    # @return [RelatonItu::ItuBibliographicItem]
+    def self.from_hash(hash)
+      item_hash = ::RelatonItu::HashConverter.hash_to_bib(hash)
+      new **item_hash
+    end
   end
 end

data/lib/relaton_itu/itu_bibliography.rb

@@ -19,20 +19,27 @@ module RelatonItu
       # @param text [String]
       # @return [RelatonItu::HitCollection]
       def search(text, year = nil)
+        # code = text.sub(/(?<=ITU-T\s\w)\.(\w+\.)(?=\d+)/, ' \1')
+        if text =~ /(ITU-T\s\w)\.(Suppl\.|Annex)\s?(\w?\d+)/
+          correct_ref = "#{$~[1]} #{$~[2]} #{$~[3]}"
+          warn "[relaton-itu] WARNING: Incorrect reference #{text}"
+          warn "[relaton-itu] the reference should be #{correct_ref}"
+        end
         HitCollection.new text, year
       rescue SocketError, Timeout::Error, Errno::EINVAL, Errno::ECONNRESET,
              EOFError, Net::HTTPBadResponse, Net::HTTPHeaderSyntaxError,
-             Net::ProtocolError, OpenSSL::SSL::SSLError
-        raise RelatonBib::RequestError, "Could not access http://www.itu.int"
+             Net::ProtocolError, URI::InvalidURIError => e
+        raise RelatonBib::RequestError, e.message
       end
 
       # @param code [String] the ISO standard Code to look up (e..g "ISO 9000")
       # @param year [String] the year the standard was published (optional)
-      # @param opts [Hash] options; restricted to :all_parts if all-parts reference is required
+      # @param opts [Hash] options; restricted to :all_parts if all-parts
+      #   reference is required
       # @return [String] Relaton XML serialisation of reference
       def get(code, year = nil, opts = {}) # rubocop:disable Metrics/CyclomaticComplexity,Metrics/MethodLength,Metrics/PerceivedComplexity
         if year.nil?
-          /^(?<code1>[^\s]+\s[^\s]+)\s\(\d{2}\/(?<year1>\d+)\)$/ =~ code
+          /^(?<code1>[^\s]+\s[^\s]+)\s\((\d{2}\/)?(?<year1>\d+)\)$/ =~ code
           unless code1.nil?
             code = code1
             year = year1
@@ -54,11 +61,13 @@ module RelatonItu
         id = year ? "#{code}:#{year}" : code
         warn "[relaton-itu] WARNING: no match found online for #{id}. "\
           "The code must be exactly like it is on the standards website."
-        warn "[relaton-itu] (There was no match for #{year}, though there were matches "\
-          "found for #{missed_years.join(', ')}.)" unless missed_years.empty?
-        if /\d-\d/ =~ code
-          warn "[relaton-itu] The provided document part may not exist, or the document "\
-            "may no longer be published in parts."
+        unless missed_years.empty?
+          warn "[relaton-itu] (There was no match for #{year}, though there "\
+            "were matches found for #{missed_years.join(', ')}.)"
+        end
+        if /\d-\d/.match? code
+          warn "[relaton-itu] The provided document part may not exist, or "\
+            "the document may no longer be published in parts."
         else
           warn "[relaton-itu] If you wanted to cite all document parts for the reference, "\
             "use \"#{code} (all parts)\".\nIf the document is not a standard, "\
@@ -69,10 +78,11 @@ module RelatonItu
 
       def search_filter(code, year) # rubocop:disable Metrics/AbcSize,Metrics/CyclomaticComplexity,Metrics/MethodLength,Metrics/PerceivedComplexity
         %r{
-          ^(?<pref1>ITU)?(-(?<type1>\w))?\s?(?<code1>[^\s\/]+)
+          ^(?<pref1>ITU)?(-(?<type1>\w))?\s?(?<code1>[^\s\/]+(?:\/\w[\.\d]+)?)
+          (\s\(?(?<ver1>v\d+)\)?)?
           (\s\(((?<month1>\d{2})\/)?(?<year1>\d{4})\))?
           (\s-\s(?<buldate1>\d{2}\.\w{1,4}\.\d{4}))?
-          (\/(?<corr1>(Amd|Cor)\s?\d+))?
+          (\s(?<corr1>(Amd|Cor|Amendment|Corrigendum)\.?\s?\d+))?
           (\s\(((?<cormonth1>\d{2})\/)?(?<coryear1>\d{4})\))?
         }x =~ code
         year ||= year1
@@ -81,21 +91,27 @@ module RelatonItu
         warn "[relaton-itu] (\"#{code}\") fetching..."
         result = search(code)
         code1.sub! /(?<=\.)Imp(?=\d)/, "" if result.gi_imp
+        if corr1
+          corr1.sub!(/[\.\s]+/, " ").sub!("Amendment", "Amd")
+          corr1.sub!("Corrigendum", "Corr")
+        end
         result.select do |i|
+          next true unless i.hit[:code]
+
           %r{
             ^(?<pref2>ITU)?(-(?<type2>\w))?\s?(?<code2>[\S]+)
+            (\s\(?(?<ver2>v\d+)\)?)?
             (\s\(((?<month2>\d{2})\/)?(?<year2>\d{4})\))?
             (\s(?<corr2>(Amd|Cor)\.\s?\d+))?
             (\s\(((?<cormonth2>\d{2})\/)?(?<coryear2>\d{4})\))?
           }x =~ i.hit[:code]
           /:[^\(]+\((?<buldate2>\d{2}\.\w{1,4}\.\d{4})\)/ =~ i.hit[:title]
           corr2&.sub! /\.\s?/, " "
-          pref1 == pref2 && (!type1 || type1 == type2) && code1 == code2 &&
+          pref1 == pref2 && (!type1 || type1 == type2) && code2.include?(code1) &&
             (!year || year == year2) && (!month1 || month1 == month2) &&
             corr1 == corr2 && (!coryear1 || coryear1 == coryear2) &&
-            buldate1 == buldate2 && (!cormonth1 || cormonth1 == cormonth2)
-          # i.hit[:code] &&
-          #   i.hit[:code].match(docidrx).to_s == c
+            buldate1 == buldate2 && (!cormonth1 || cormonth1 == cormonth2) &&
+            (!ver1 || ver1 == ver2)
         end
       end
 
@@ -108,10 +124,11 @@ module RelatonItu
       def isobib_results_filter(result, year)
         missed_years = []
         result.each do |r|
-          return { ret: r.fetch } if !year
-
-          /\(\d{2}\/(?<pyear>\d{4})\)/ =~ r.hit[:code]
-          return { ret: r.fetch } if year == pyear
+          /\((\d{2}\/)?(?<pyear>\d{4})\)/ =~ r.hit[:code]
+          if !year || year == pyear
+            ret = r.fetch
+            return { ret: ret } if ret
+          end
 
           missed_years << pyear
         end

data/lib/relaton_itu/itu_group.rb

@@ -63,7 +63,7 @@ module RelatonItu
       @type = type
       @name = name
       @acronym = acronym
-      @period = period.is_a?(Hash) ? Period.new(period) : period
+      @period = period.is_a?(Hash) ? Period.new(**period) : period
     end
 
     # @param builder [Nokogiri::XML::Builder]

data/lib/relaton_itu/processor.rb

@@ -26,8 +26,7 @@ module RelatonItu
     # @param hash [Hash]
     # @return [RelatonItu::ItuBibliographicItem]
     def hash_to_bib(hash)
-      item_hash = ::RelatonItu::HashConverter.hash_to_bib(hash)
-      ::RelatonItu::ItuBibliographicItem.new item_hash
+      ::RelatonItu::ItuBibliographicItem.from_hash hash
     end
 
     # Returns hash of XML grammar

data/lib/relaton_itu/scrapper.rb

@@ -24,18 +24,18 @@ module RelatonItu
     }.freeze
 
     class << self
-      # rubocop:disable Metrics/AbcSize, Metrics/MethodLength
-
       # Parse page.
-      # @param hit_data [Hash]
+      # @param hit [RelatonItu::Hit]
       # @return [Hash]
-      def parse_page(hit_data, imp = false)
-        url, doc = get_page hit_data[:url]
+      def parse_page(hit, imp = false) # rubocop:disable Metrics/AbcSize, Metrics/MethodLength
+        doc = get_page hit
+        return unless doc.code == "200"
+
         if imp
           a = doc.at "//span[contains(@id, 'tab_ig_uc_rec')]/a"
           return unless a
 
-          url, doc = get_page URI.join(url, a[:href]).to_s
+          doc = get_page hit, a[:href].to_s
         end
 
         # Fetch edition.
@@ -44,67 +44,65 @@ module RelatonItu
         ItuBibliographicItem.new(
           fetched: Date.today.to_s,
           type: "standard",
-          docid: fetch_docid(doc, hit_data[:title]),
+          docid: fetch_docid(doc, hit.hit[:title]),
           edition: edition,
           language: ["en"],
           script: ["Latn"],
           title: fetch_titles(doc),
-          doctype: hit_data[:type],
+          doctype: hit.hit[:type],
           docstatus: fetch_status(doc),
           ics: [], # fetch_ics(doc),
           date: fetch_dates(doc),
-          contributor: fetch_contributors(hit_data[:code]),
-          editorialgroup: fetch_workgroup(hit_data[:code], doc),
-          abstract: fetch_abstract(doc),
-          copyright: fetch_copyright(hit_data[:code], doc),
-          link: fetch_link(doc, url),
+          contributor: fetch_contributors(hit.hit[:code]),
+          editorialgroup: fetch_workgroup(hit.hit[:code], doc),
+          abstract: fetch_abstract(doc, hit),
+          copyright: fetch_copyright(hit.hit[:code], doc),
+          link: fetch_link(doc),
           relation: fetch_relations(doc),
           place: ["Geneva"]
         )
       end
-      # rubocop:enable Metrics/AbcSize
 
       private
 
       # Fetch abstracts.
-      # @param doc [Nokigiri::HTML::Document]
-      # @return [Array<Array>]
-      def fetch_abstract(doc)
-        abstract_url = doc.at('//table/tr/td/span[contains(@id, "lbl_dms")]/div')
-        return [] unless abstract_url
-
-        url = abstract_url[:onclick].match(/https?[^']+/).to_s
-        d = Nokogiri::HTML Net::HTTP.get(URI(url)).encode(undef: :replace, replace: "")
-        abstract_content = d.css("p.MsoNormal").text.gsub(/\r\n/, "")
-          .squeeze(" ").gsub(/\u00a0/, "")
+      # @param doc [Mechanize::Page]
+      # @param hit [RelatonItu::Hit]
+      # @return [Array<Hash>]
+      def fetch_abstract(doc, hit) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength
+        abstract_url = doc.at '//table/tr/td/span[contains(@id, "lbl_dms")]/div'
+        content = if abstract_url
+                    url = abstract_url[:onclick].match(/https?[^']+/).to_s
+                    rsp = hit.hit_collection.agent.get url
+                    d = Nokogiri::HTML rsp.body.encode(undef: :replace, replace: "")
+                    d.css("p.MsoNormal").text.gsub(/\r\n/, "").squeeze(" ").gsub(/\u00a0/, "")
+                  elsif a = doc.at('//table/tr/td/span[contains(@class, "observation")]/text()')
+                    a.text.strip
+                  end
+        return [] unless content
 
         [{
-          content: abstract_content,
+          content: content,
           language: "en",
           script: "Latn",
         }]
       end
 
       # Get page.
-      # @param path [String] page's path
+      # @param hit [RelatonItu::Hit]
+      # @param url [String, nil]
       # @return [Array<String, Nokogiri::HTML::Document>]
-      def get_page(url)
-        uri = URI url
-        resp = Net::HTTP.get_response(uri)
-        until resp.code == "200"
-          uri = URI resp["location"] if resp.code.match? /^30/
-          resp = Net::HTTP.get_response(uri)
-        end
-        [uri.to_s, Nokogiri::HTML(resp.body)]
+      def get_page(hit, url = nil)
+        uri = url || hit.hit[:url]
+        hit.hit_collection.agent.get uri
       rescue SocketError, Timeout::Error, Errno::EINVAL, Errno::ECONNRESET,
              EOFError, Net::HTTPBadResponse, Net::HTTPHeaderSyntaxError,
              Net::ProtocolError, OpenSSL::SSL::SSLError
-        raise RelatonBib::RequestError, "Could not access #{url}"
+        raise RelatonBib::RequestError, "Could not access #{uri}"
       end
-      # rubocop:enable Metrics/MethodLength
 
       # Fetch docid.
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Mechanize::Page]
       # @param title [String]
       # @return [Hash]
       def fetch_docid(doc, title)
@@ -117,16 +115,18 @@ module RelatonItu
         docids
       end
 
-      def createdocid(text)
+      # @param text [String]
+      # @return [RelatonBib::DocumentIdentifier]
+      def createdocid(text) # rubocop:disable Metrics/MethodLength
         %r{
           ^(?<code>((ITU-\w|ISO\/IEC)\s)?[^\(:]+)
-          (\(((?<month>\d{2})\/)?(?<year>\d{4})\))?
+          (\(((?<_month>\d{2})\/)?(?<_year>\d{4})\))?
           (:[^\(]+\((?<buldate>\d{2}\.\w{1,4}\.\d{4})\))?
           (\s(?<corr>(Amd|Cor)\.\s?\d+))?
-          # (\s\(((?<cormonth>\d{2})\/)?(?<coryear>\d{4})\))?
+          # (\s\(((?<_cormonth>\d{2})\/)?(?<_coryear>\d{4})\))?
         }x =~ text.squeeze(" ")
         corr&.sub! /\.\s?/, " "
-        id = [code.sub(/[[:space:]]$/, ""), corr].compact.join "/"
+        id = [code.sub(/[[:space:]]$/, ""), corr].compact.join " "
         id += " - #{buldate}" if buldate
         type = id.match(%r{^\w+}).to_s
         type = "ITU" if type == "G"
@@ -134,7 +134,7 @@ module RelatonItu
       end
 
       # Fetch status.
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Mechanize::Page]
       # @return [RelatonBib::DocumentStatus, NilClass]
       def fetch_status(doc)
         s = doc.at("//table/tr/td/span[contains(@id, 'Label7')]",
@@ -147,7 +147,7 @@ module RelatonItu
 
       # Fetch workgroup.
       # @param code [String]
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Mechanize::Page]
       # @return [RelatonItu::EditorialGroup, NilClass]
       def fetch_workgroup(code, doc)
         wg = doc.at('//table/tr/td/span[contains(@id, "Label8")]/a')
@@ -155,8 +155,7 @@ module RelatonItu
 
         group = wg && itugroup(wg.text)
         EditorialGroup.new(
-          bureau: code.match(/(?<=-)./).to_s,
-          group: group
+          bureau: code.match(/(?<=-)./).to_s, group: group
         )
       end
 
@@ -176,24 +175,24 @@ module RelatonItu
         ItuGroup.new name: name, type: type, acronym: acronym
       end
 
-      # rubocop:disable Metrics/MethodLength
-
       # Fetch relations.
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Mechanize::Page]
       # @return [Array<Hash>]
       def fetch_relations(doc)
-        doc.xpath('//div[contains(@id, "tab_sup")]//table/tr[position()>2]').map do |r|
+        doc.xpath('//div[contains(@id, "tab_sup")]//table/tr[position()>2]')
+          .map do |r|
           ref = r.at('./td/span[contains(@id, "title_e")]/nobr/a')
-          fref = RelatonBib::FormattedRef.new(content: ref.text, language: "en", script: "Latn")
-          bibitem = ItuBibliographicItem.new(formattedref: fref, type: "standard")
+          fref = RelatonBib::FormattedRef.new(content: ref.text, language: "en",
+                                              script: "Latn")
+          bibitem = ItuBibliographicItem.new(formattedref: fref,
+                                             type: "standard")
           { type: "complements", bibitem: bibitem }
         end
       end
-      # rubocop:enable Metrics/MethodLength
 
       # Fetch titles.
-      # @param doc [Nokogiri::HTML::Document]
-      # @return [Array<Hash>]
+      # @param doc [Mechanize::Page]
+      # @return [RelatonBib::TypedTitleStringCollection]
       def fetch_titles(doc)
         t = doc.at("//td[@class='title']|//div/table[1]/tr[4]/td/strong")
         return [] unless t
@@ -202,27 +201,29 @@ module RelatonItu
       end
 
       # Fetch dates
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Mechanize::Page]
       # @return [Array<Hash>]
-      def fetch_dates(doc) # rubocop:disable Metrics/CyclomaticComplexity
+      def fetch_dates(doc) # rubocop:disable Metrics/CyclomaticComplexity,Metrics/PerceivedComplexity
         dates = []
         date = doc.at("//table/tr/td/span[contains(@id, 'Label5')]",
                       "//p[contains(.,'Approved in')]")
         pdate = date&.text&.match(/\d{4}-\d{2}-\d{2}/).to_s || ob_date(doc)
         if pdate && !pdate&.empty?
           dates << { type: "published", on: pdate }
+        elsif pdate = ob_date(doc)
+          dates << { type: "published", on: pdate }
         end
         dates
       end
 
       # Scrape Operational Bulletin date.
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Mechanize::Page]
       # @return [String]
       def ob_date(doc)
         pdate = doc.at('//table/tbody/tr/td[contains(text(), "Year:")]')
         return unless pdate
 
-        roman_to_arabic pdate.text.match(%r{(?<=Year: )\d{2}.\w+.\d{4}}).to_s
+        roman_to_arabic pdate.text.match(%r{(?<=Year: )(\d{2}.\w+.)?\d{4}}).to_s
       end
 
       # Convert roman month number in string date to arabic number
@@ -230,12 +231,15 @@ module RelatonItu
       # @return [String]
       def roman_to_arabic(date)
         %r{(?<rmonth>[IVX]+)} =~ date
-        month = ROMAN_MONTHS.index(rmonth) + 1
-        Date.parse(date.sub(%r{[IVX]+}, month.to_s)).to_s
+        if ROMAN_MONTHS.index(rmonth)
+          month = ROMAN_MONTHS.index(rmonth) + 1
+          Date.parse(date.sub(%r{[IVX]+}, month.to_s)).to_s
+        else date
+        end
       end
 
       # Fetch contributors
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Mechanize::Page]
       # @return [Array<Hash>]
       def fetch_contributors(code)
         return [] unless code
@@ -251,11 +255,10 @@ module RelatonItu
       end
 
       # Fetch links.
-      # @param doc [Nokogiri::HTML::Document]
-      # @param url [String]
+      # @param doc [Mechanize::Page]
       # @return [Array<Hash>]
-      def fetch_link(doc, url)
-        links = [{ type: "src", content: url }]
+      def fetch_link(doc)
+        links = [{ type: "src", content: doc.uri.to_s }]
         obp_elm = doc.at(
           '//a[@title="Persistent link to download the PDF file"]',
           "//font[contains(.,'PDF')]/../.."
@@ -266,6 +269,8 @@ module RelatonItu
         links
       end
 
+      # @param type [String]
+      # @param elm [Nokogiri::XML::Element]
       def typed_link(type, elm)
         {
           type: type,
@@ -275,7 +280,7 @@ module RelatonItu
 
       # Fetch copyright.
       # @param code [String]
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Mechanize::Page]
       # @return [Array<Hash>]
       def fetch_copyright(code, doc)
         abbreviation = code.match(/^[^-]+/).to_s

data/lib/relaton_itu/version.rb

@@ -1,3 +1,3 @@
 module RelatonItu
-  VERSION = "1.7.3".freeze
+  VERSION = "1.7.8".freeze
 end

data/lib/relaton_itu/xml_parser.rb

@@ -8,7 +8,7 @@ module RelatonItu
       # @param item_hash [Hash]
       # @return [RelatonItu::ItuBibliographicItem]
       def bib_item(item_hash)
-        ItuBibliographicItem.new item_hash
+        ItuBibliographicItem.new **item_hash
       end
 
       # @param ext [Nokogiri::XML::Element]

data/relaton-itu.gemspec

@@ -26,16 +26,17 @@ Gem::Specification.new do |spec|
   spec.require_paths = ["lib"]
   spec.required_ruby_version = Gem::Requirement.new(">= 2.4.0")
 
-  spec.add_development_dependency "debase"
+  # spec.add_development_dependency "debase"
   spec.add_development_dependency "equivalent-xml", "~> 0.6"
   spec.add_development_dependency "pry-byebug"
-  spec.add_development_dependency "rake", "~> 10.0"
+  spec.add_development_dependency "rake", "~> 13.0"
   spec.add_development_dependency "rspec", "~> 3.0"
-  spec.add_development_dependency "ruby-debug-ide"
+  # spec.add_development_dependency "ruby-debug-ide"
   spec.add_development_dependency "ruby-jing"
   spec.add_development_dependency "simplecov"
   spec.add_development_dependency "vcr", "~> 5.0.0"
   spec.add_development_dependency "webmock"
 
+  spec.add_dependency "mechanize"
   spec.add_dependency "relaton-bib", "~> 1.7.0"
 end

metadata

@@ -1,29 +1,15 @@
 --- !ruby/object:Gem::Specification
 name: relaton-itu
 version: !ruby/object:Gem::Version
-  version: 1.7.3
+  version: 1.7.8
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2021-01-04 00:00:00.000000000 Z
+date: 2021-04-22 00:00:00.000000000 Z
 dependencies:
-- !ruby/object:Gem::Dependency
-  name: debase
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
 - !ruby/object:Gem::Dependency
   name: equivalent-xml
   requirement: !ruby/object:Gem::Requirement
@@ -58,14 +44,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: '13.0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '10.0'
+        version: '13.0'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
@@ -80,20 +66,6 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '3.0'
-- !ruby/object:Gem::Dependency
-  name: ruby-debug-ide
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
 - !ruby/object:Gem::Dependency
   name: ruby-jing
   requirement: !ruby/object:Gem::Requirement
@@ -150,6 +122,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: mechanize
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: relaton-bib
   requirement: !ruby/object:Gem::Requirement
@@ -172,9 +158,7 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- ".github/workflows/macos.yml"
-- ".github/workflows/ubuntu.yml"
-- ".github/workflows/windows.yml"
+- ".github/workflows/rake.yml"
 - ".gitignore"
 - ".rspec"
 - ".rubocop.yml"
@@ -183,6 +167,7 @@ files:
 - README.adoc
 - Rakefile
 - bin/console
+- bin/rspec
 - bin/setup
 - grammars/basicdoc.rng
 - grammars/biblio.rng

data/.github/workflows/macos.yml

@@ -1,34 +0,0 @@
-# Auto-generated by Cimas: Do not edit it manually!
-# See https://github.com/metanorma/cimas
-name: macos
-
-on:
-  push:
-    branches: [ master ]
-  pull_request:
-    branches: [ '**' ]
-
-jobs:
-  test-macos:
-    name: Test on Ruby ${{ matrix.ruby }} macOS
-    runs-on: macos-latest
-    strategy:
-      fail-fast: false
-      matrix:
-        ruby: [ '2.6', '2.5', '2.4' ]
-    steps:
-      - uses: actions/checkout@master
-      - name: Use Ruby
-        uses: actions/setup-ruby@v1
-        with:
-          ruby-version: ${{ matrix.ruby }}
-          architecture: 'x64'
-      - name: Update gems
-        run: |
-          sudo gem install bundler --force
-          ruby -v | grep 2.5 && bundle config set build.debase --with-cflags="-Wno-error=implicit-function-declaration"
-          ruby -v | grep 2.5 && bundle config set build.ruby-debug-ide --with-cflags="-Wno-error=implicit-function-declaration"
-          bundle install --jobs 4 --retry 3
-      - name: Run specs
-        run: |
-          bundle exec rake

data/.github/workflows/ubuntu.yml

@@ -1,33 +0,0 @@
-# Auto-generated by Cimas: Do not edit it manually!
-# See https://github.com/metanorma/cimas
-name: ubuntu
-
-on:
-  push:
-    branches: [ master ]
-  pull_request:
-    branches: [ '**' ]
-
-jobs:
-  test-linux:
-    name: Test on Ruby ${{ matrix.ruby }} Ubuntu
-    runs-on: ubuntu-latest
-    strategy:
-      fail-fast: false
-      matrix:
-        ruby: [ '2.6', '2.5', '2.4' ]
-    steps:
-      - uses: actions/checkout@master
-      - name: Use Ruby
-        uses: actions/setup-ruby@v1
-        with:
-          ruby-version: ${{ matrix.ruby }}
-          architecture: 'x64'
-      - name: Update gems
-        run: |
-          gem install bundler
-          bundle install --jobs 4 --retry 3
-      - name: Run specs
-        run: |
-          unset JAVA_TOOL_OPTIONS
-          bundle exec rake

data/.github/workflows/windows.yml

@@ -1,35 +0,0 @@
-# Auto-generated by Cimas: Do not edit it manually!
-# See https://github.com/metanorma/cimas
-name: windows
-
-on:
-  push:
-    branches: [ master ]
-  pull_request:
-    branches: [ '**' ]
-
-jobs:
-  test-windows:
-    name: Test on Ruby ${{ matrix.ruby }} Windows
-    runs-on: windows-latest
-    strategy:
-      fail-fast: false
-      matrix:
-        ruby: [ '2.6', '2.5', '2.4' ]
-    steps:
-      - uses: actions/checkout@master
-      - name: Use Ruby
-        uses: actions/setup-ruby@v1
-        with:
-          ruby-version: ${{ matrix.ruby }}
-          architecture: 'x64'
-      - name: Update gems
-        shell: pwsh
-        run: |
-          gem install bundler
-          bundle config --local path vendor/bundle
-          bundle update
-          bundle install --jobs 4 --retry 3
-      - name: Run specs
-        run: |
-          bundle exec rake