relaton-iso 1.11.0 → 1.12.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: f7fa7c91589b331ac4748d70c790b814bd294b73c8c35c12a1e7a6a2fa54ca38
-  data.tar.gz: 64ea8882a8c4ec26278f4bea0b8af25879b7ead65cc156117dadbeab8b606400
+  metadata.gz: 4479e38048aa0dfae8bcc85f1e9de03b5fe0561048b658ec47b3df8ca64794eb
+  data.tar.gz: c297ddc7b15d8186b85fbb7d4d3f84863d7df6e20ad243f0740364262fe43807
 SHA512:
-  metadata.gz: 1c8716c7c2ddb6fb8644b528cbddccf2c900897326b1f3897946247886e21c25452f71783cc696ba44cba7a2247b11577085c71f780c91fb060f4d72748c1af4
-  data.tar.gz: 1a91a1433ddf1312edb01e15348207573ed30f5d022b376757b2fda69ad8ae4be62090e726a8baede35e1ff3cdf95dcf5146bd863c6758e5dea4ccc047309dd8
+  metadata.gz: 6fd11d9fe01bd36052cf2762df6e8f728d361fbe23410dcb0229e95436a9f7909e51e51a8be0abc9f7bd269ffd10d642cb0aa7792dba76c0c174b606e319bf58
+  data.tar.gz: 060edbed6bb5b11033911db2200a4e19e98585c4133a1b4a09eaa79b582f2cf23b39f424b84d3110429ff1800247cec4f2b6522a82bbbdc45cb093e63a339bda

data/.github/workflows/rake.yml

@@ -10,27 +10,4 @@ on:
 
 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: [ '3.0', '2.7', '2.6', '2.5' ]
-        os: [ ubuntu-latest, windows-latest, macos-latest ]
-        experimental: [ false ]
-    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
+    uses: relaton/support/.github/workflows/rake.yml@master

data/Gemfile

@@ -1,6 +1,3 @@
-Encoding.default_external = Encoding::UTF_8
-Encoding.default_internal = Encoding::UTF_8
-
 source "https://rubygems.org"
 
 git_source(:github) { |repo_name| "https://github.com/#{repo_name}" }

data/bin/thor

@@ -0,0 +1,29 @@
+#!/usr/bin/env ruby
+# frozen_string_literal: true
+
+#
+# This file was generated by Bundler.
+#
+# The application 'thor' 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("thor", "thor")

data/lib/relaton_iso/document_identifier.rb

@@ -0,0 +1,27 @@
+module RelatonIso
+  class DocumentIdentifier < RelatonBib::DocumentIdentifier
+    def id
+      id_str = @id.to_s.sub(/\sED\d+/, "")
+      if @all_parts
+        if type == "URN"
+          return "#{@id.urn}:ser"
+        else
+          return "#{id_str} (all parts)"
+        end
+      end
+      type == "URN" ? @id.urn.to_s : id_str
+    end
+
+    def remove_part
+      @id.part = nil
+    end
+
+    def remove_date
+      @id.year = nil
+    end
+
+    def all_parts
+      @all_parts = true
+    end
+  end
+end

data/lib/relaton_iso/hit.rb

@@ -4,13 +4,13 @@ module RelatonIso
   # Hit.
   class Hit < RelatonBib::Hit
     # @return [RelatonIsoBib::IsoBibliographicItem]
-    attr_writer :fetch
+    attr_writer :fetch, :pubid
 
     # Parse page.
-    # @param lang [String, NilClass]
+    # @param lang [String, nil]
     # @return [RelatonIso::IsoBibliographicItem]
     def fetch(lang = nil)
-      @fetch ||= Scrapper.parse_page @hit, lang
+      @fetch ||= Scrapper.parse_page self, lang
     end
 
     # @return [Integer]
@@ -23,5 +23,10 @@ module RelatonIso
       else 4
       end
     end
+
+    # @return [Pubid::Iso::Identifier]
+    def pubid
+      @pubid ||= Pubid::Iso::Identifier.parse_from_title(hit[:title])
+    end
   end
 end

data/lib/relaton_iso/hit_collection.rb

@@ -11,25 +11,21 @@ module RelatonIso
     # @param text [String] reference to search
     def initialize(text)
       super
-      @array = text.match?(/^ISO\sTC\s184\/SC\s?4/) ? fetch_github : fetch_iso
+      @array = text.match?(/^ISO\s(?:TC\s184\/SC\s?4|IEC\sDIR\s(?:\d|IEC|JTC))/) ? fetch_github : fetch_iso
     end
 
     # @param lang [String, NilClass]
     # @return [RelatonIsoBib::IsoBibliographicItem]
     def to_all_parts(lang = nil) # rubocop:disable Metrics/CyclomaticComplexity
       # parts = @array.reject { |h| h.hit["docPart"]&.empty? }
-      hit = @array.min_by do |h|
-        IsoBibliography.ref_components(h.hit[:title])[1].to_i
-      end
+      hit = @array.min_by { |h| h.pubid.part }
       return @array.first.fetch lang unless hit
 
-      bibitem = hit.fetch lang
+      bibitem = hit.fetch(lang)
       all_parts_item = bibitem.to_all_parts
       @array.reject { |h| h.hit[:uuid] == hit.hit[:uuid] }.each do |hi|
-        %r{^(?<fr>ISO(?:\s|/)[^-/:()]+(?:-[\w-]+)?(?::\d{4})?
-          (?:/\w+(?:\s\w+)?\s\d+(?:\d{4})?)?)}x =~ hi.hit[:title]
         isobib = RelatonIsoBib::IsoBibliographicItem.new(
-          formattedref: RelatonBib::FormattedRef.new(content: fr),
+          formattedref: RelatonBib::FormattedRef.new(content: hi.pubid.to_s),
         )
         all_parts_item.relation << RelatonBib::DocumentRelation.new(
           type: "instance", bibitem: isobib,

data/lib/relaton_iso/iso_bibliography.rb

@@ -27,175 +27,152 @@ module RelatonIso
       # @option opts [Boolean] :keep_year if undated reference should return
       #   actual reference with year
       #
-      # @return [String] Relaton XML serialisation of reference
+      # @return [RelatonIsoBib::IsoBibliographicItem] Relaton XML serialisation of reference
       def get(ref, year = nil, opts = {}) # rubocop:disable Metrics/CyclomaticComplexity,Metrics/MethodLength,Metrics/PerceivedComplexity,Metrics/AbcSize
         code = ref.gsub(/\u2013/, "-")
-        # %r{\s(?<num>\d+)(?:-(?<part>[\d-]+))?(?::(?<year1>\d{4}))?} =~ code
-        _, _part, year1, = ref_components ref
-        year ||= year1
+
+        # parse "all parts" request
         code.sub! " (all parts)", ""
         opts[:all_parts] ||= $~ && opts[:all_parts].nil?
-        # opts[:keep_year] ||= opts[:keep_year].nil?
-        # code.sub!("#{num}-#{part}", num) if opts[:all_parts] && part
-        # if %r[^ISO/IEC DIR].match? code
-        #   return RelatonIec::IecBibliography.get(code, year, opts)
-        # end
 
-        ret = isobib_get1(code, year, opts)
-        return nil if ret.nil?
+        query_pubid = Pubid::Iso::Identifier.parse(code)
+        query_pubid.year = year if year
+
+        hits = isobib_search_filter(query_pubid, opts)
 
-        if (year && opts[:keep_year].nil?) || opts[:keep_year] || opts[:all_parts]
+        # return only first one if not all_parts
+        ret = if !opts[:all_parts] || hits.size == 1
+                hits.any? && hits.first.fetch(opts[:lang])
+              else
+                hits.to_all_parts(opts[:lang])
+              end
+
+        if ret
+          warn "[relaton-iso] (\"#{query_pubid}\") found #{ret.docidentifier.first.id}"
+        else
+          return fetch_ref_err(query_pubid, query_pubid.year)
+        end
+
+        if (query_pubid.year && opts[:keep_year].nil?) || opts[:keep_year] || opts[:all_parts]
           ret
         else
           ret.to_most_recent_reference
         end
       end
 
-      def ref_components(ref)
-        %r{
-          ^(?<code>ISO(?:\s|/)[^-/:()]+\d+)
-          (?:-(?<part>[\w-]+))?
-          (?::(?<year>\d{4}))?
-          (?:/(?<corr>\w+(?:\s\w+)?\s\d+)(?:(?<coryear>\d{4}))?)?
-        }x =~ ref
-        [code&.strip, part, year, corr, coryear]
+      # @param query_pubid [Pubid::Iso::Identifier]
+      # @param pubid [Pubid::Iso::Identifier]
+      # @param all_parts [Boolean] match with any parts when true
+      # @return [Boolean]
+      def matches_parts?(query_pubid, pubid, all_parts: false)
+        if all_parts
+          # match only with documents with part number
+          !pubid.part.nil?
+        else
+          query_pubid.part == pubid.part
+        end
       end
 
-      private
+      def matches_base?(query_pubid, pubid, any_types_stages: false) # rubocop:disable Metrics/AbcSize,Metrics/CyclomaticComplexity,Metrics?PerceivedComplexity
+        query_pubid.publisher == pubid.publisher &&
+          query_pubid.number == pubid.number &&
+          query_pubid.copublisher == pubid.copublisher &&
+          ((any_types_stages && query_pubid.stage.nil?) || query_pubid.stage == pubid.stage) &&
+          ((any_types_stages && query_pubid.type.nil?) || query_pubid.type == pubid.type)
+      end
 
-      # rubocop:disable Metrics/MethodLength
+      # @param hit_collection [RelatonIso::HitCollection]
+      # @param year [String]
+      # @return [RelatonIso::HitCollection]
+      def filter_hits_by_year(hit_collection, year) # rubocop:disable Metrics/MethodLength,Metrics/AbcSize,Metrics/CyclomaticComplexity,Metrics/PerceivedComplexity
+        missed_years = []
 
-      def fetch_ref_err(code, year, missed_years)
-        id = year ? "#{code}:#{year}" : code
-        warn "[relaton-iso] WARNING: no match found online for #{id}. "\
-             "The code must be exactly like it is on the standards website."
-        unless missed_years.empty?
+        # filter by year
+        hits = hit_collection.select do |hit|
+          if hit.pubid.year == year
+            true
+          elsif hit.pubid.year.nil? && hit.hit[:year].to_s == year
+            hit.pubid.year = year
+            true
+          else
+            missed_year = hit.pubid.year || hit.hit[:year].to_s
+            if missed_year && !missed_year.empty? && !missed_years.include?(missed_year)
+              missed_years << missed_year
+            end
+            false
+          end
+        end
+
+        if hits.empty? && !missed_years.empty?
           warn "[relaton-iso] (There was no match for #{year}, though there "\
                "were matches found for #{missed_years.join(', ')}.)"
         end
-        if /\d-\d/.match? code
+        hits
+      end
+
+      private
+
+      def fetch_ref_err(query_pubid, year) # rubocop:disable Metrics/MethodLength
+        id = year ? "#{query_pubid}:#{year}" : query_pubid
+        warn "[relaton-iso] WARNING: no match found online for #{id}. "\
+             "The code must be exactly like it is on the standards website."
+        if /\d-\d/.match? query_pubid.to_s
           warn "[relaton-iso] The provided document part may not exist, "\
                "or the document may no longer be published in parts."
         else
           warn "[relaton-iso] If you wanted to cite all document parts for "\
-               "the reference, use \"#{code} (all parts)\".\nIf the document "\
+               "the reference, use \"#{query_pubid} (all parts)\".\nIf the document "\
                "is not a standard, use its document type abbreviation "\
                "(TS, TR, PAS, Guide)."
         end
         nil
       end
 
-      # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity
-
       # Search for hits. If no found then trying missed stages and ISO/IEC.
       #
-      # @param code [String] reference without correction
+      # @param query_pubid [Pubid::Iso::Identifier] reference without correction
       # @param opts [Hash]
       # @return [Array<RelatonIso::Hit>]
-      def isobib_search_filter(code, opts)
-        ref = remove_part code, opts[:all_parts]
-        warn "[relaton-iso] (\"#{code}\") fetching..."
-        result = search(ref)
-        res = search_code result, code, opts
+      def isobib_search_filter(query_pubid, opts) # rubocop:disable Metrics/MethodLength,Metrics/AbcSize
+        query_pubid.part = nil if opts[:all_parts]
+        warn "[relaton-iso] (\"#{query_pubid}\") fetching..."
+        # fetch hits collection
+        hit_collection = search(query_pubid.to_s(with_date: false))
+        # filter only matching hits
+        res = filter_hits hit_collection, query_pubid,
+                          all_parts: opts[:all_parts]
         return res unless res.empty?
 
-        # try stages
-        case code
-        when %r{^\w+/[^/]+\s\d+} # code like ISO/IEC 123, ISO/IEC/IEE 123
-          res = try_stages(result, opts) do |st|
-            code.sub(%r{^(?<pref>[^\s]+\s)}) { "#{$~[:pref]}#{st} " }
-          end
-          return res unless res.empty?
-        when %r{^\w+\s\d+} # code like ISO 123
-          res = try_stages(result, opts) do |st|
-            code.sub(%r{^(?<pref>\w+)}) { "#{$~[:pref]}/#{st}" }
-          end
-          return res unless res.empty?
-        end
+        # lookup for documents with stages when no match without stage
+        res = filter_hits hit_collection, query_pubid,
+                          all_parts: opts[:all_parts], any_types_stages: true
+        return res unless res.empty?
 
-        if %r{^ISO\s}.match? code # try ISO/IEC if ISO not found
+        # TODO: do this at pubid-iso
+        if query_pubid.publisher == "ISO" && query_pubid.copublisher.nil? # try ISO/IEC if ISO not found
           warn "[relaton-iso] Attempting ISO/IEC retrieval"
-          c = code.sub "ISO", "ISO/IEC"
-          res = search_code result, c, opts
+          query_pubid.copublisher = "IEC"
+          res = filter_hits hit_collection, query_pubid, all_parts: opts[:all_parts]
         end
         res
       end
-      # rubocop:enable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/MethodLength, Metrics/PerceivedComplexity
 
-      def remove_part(ref, all_parts)
-        return ref unless all_parts
-
-        ref.sub %r{(\S+\s\d+)[\d-]+}, '\1'
-      end
-
-      # @param result [RelatonIso::HitCollection]
-      # @param opts [Hash]
+      # @param hits [RelatonIso::HitCollection]
+      # @param query_pubid [Pubid::Iso::Identifier]
+      # @param all_parts [Boolean]
+      # @param any_stages [Boolean]
       # @return [RelatonIso::HitCollection]
-      def try_stages(result, opts)
-        res = nil
-        %w[NP WD CD DIS FDIS PRF IS AWI TR].each do |st| # try stages
-          c = yield st
-          res = search_code result, c, opts
-          return res unless res.empty?
-        end
-        res
-      end
-
-      # @param result [RelatonIso::HitCollection]
-      # @param code [String]
-      # @param opts [Hash]
-      # @return [RelatonIso::HitCollection]
-      def search_code(result, code, opts) # rubocop:disable Metrics/AbcSize,Metrics/CyclomaticComplexity,Metrics/PerceivedComplexity
-        code1, part1, _, corr1, coryear1 = ref_components code
-        result.select do |i|
-          code2, part2, _, corr2, coryear2 = ref_components i.hit[:title]
-          code1 == code2 && ((opts[:all_parts] && part2) || (!opts[:all_parts] && part1 == part2)) &&
-            corr1 == corr2 && (!coryear1 || coryear1 == coryear2)
-        end
-      end
-
-      # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/MethodLength, Metrics/PerceivedComplexity
-
-      # Sort through the results from RelatonIso, fetching them three at a time,
-      # and return the first result that matches the code, matches the year
-      # (if provided), and which # has a title (amendments do not).
-      # Only expects the first page of results to be populated.
-      # Does not match corrigenda etc (e.g. ISO 3166-1:2006/Cor 1:2007)
-      # If no match, returns any years which caused mismatch, for error
-      # reporting
-      def isobib_results_filter(result, year, opts)
-        missed_years = []
-        hits = result.reduce!([]) do |hts, h|
-          if !year || (%r{:(?<iyear>\d{4})(?!.*:\d{4})} =~ h.hit[:title] && iyear == year)
-            hts << h
-          else
-            missed_years << iyear
-            hts
-          end
+      def filter_hits(hit_collection, query_pubid, all_parts: false, any_types_stages: false) # rubocop:disable Metrics/AbcSize,Metrics/CyclomaticComplexity,Metrics/PerceivedComplexity
+        # filter out
+        result = hit_collection.select do |i|
+          hit_pubid = i.pubid
+          matches_base?(query_pubid, hit_pubid, any_types_stages: any_types_stages) &&
+            matches_parts?(query_pubid, hit_pubid, all_parts: all_parts) &&
+            query_pubid.corrigendum == hit_pubid.corrigendum &&
+            query_pubid.amendment == hit_pubid.amendment
         end
-        return { years: missed_years } unless hits.any?
 
-        if !opts[:all_parts] || hits.size == 1
-          return { ret: hits.first.fetch(opts[:lang]) }
-        end
-
-        { ret: hits.to_all_parts(opts[:lang]) }
-      end
-      # rubocop:enable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/MethodLength, Metrics/PerceivedComplexity
-
-      # @param code [String]
-      # @param year [String, NilClass]
-      # @param opts [Hash]
-      def isobib_get1(code, year, opts)
-        # return iev(code) if /^IEC 60050-/.match code
-        result = isobib_search_filter(code, opts) || return
-        ret = isobib_results_filter(result, year, opts)
-        if ret[:ret]
-          warn "[relaton-iso] (\"#{code}\") found #{ret[:ret].docidentifier.first.id}"
-          ret[:ret]
-        else
-          fetch_ref_err(code, year, ret[:years])
-        end
+        query_pubid.year ? filter_hits_by_year(result, query_pubid.year) : result
       end
     end
   end

data/lib/relaton_iso/scrapper.rb

@@ -50,43 +50,61 @@ module RelatonIso
 
     class << self
       # Parse page.
-      # @param hit_data [Hash]
+      # @param hit [RelatonIso::Hit]
       # @param lang [String, NilClass]
-      # @return [Hash]
-      def parse_page(hit_data, lang = nil) # rubocop:disable Metrics/AbcSize, Metrics/MethodLength, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity
+      # @return [RelatonIsoBib::IsoBibliographicItem]
+      def parse_page(hit, lang = nil) # rubocop:disable Metrics/AbcSize, Metrics/MethodLength, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity
         # path = "/contents/data/standard#{hit_data['splitPath']}/"\
         # "#{hit_data['csnumber']}.html"
-        doc, url = get_page "#{hit_data[:path].sub '/sites/isoorg', ''}.html"
+
+        doc, url = get_page "#{hit.hit[:path].sub '/sites/isoorg', ''}.html"
 
         # Fetch edition.
         edition = doc&.xpath("//strong[contains(text(), 'Edition')]/..")
           &.children&.last&.text&.match(/\d+/)&.to_s
+        hit.pubid.edition = edition if edition
 
         titles, abstract, langs = fetch_titles_abstract(doc, lang)
 
         RelatonIsoBib::IsoBibliographicItem.new(
           fetched: Date.today.to_s,
-          docid: fetch_docid(doc, edition, langs),
-          docnumber: fetch_docnumber(doc),
+          docid: fetch_relaton_docids(doc, hit.pubid),
+          docnumber: fetch_docnumber(hit.pubid),
           edition: edition,
           language: langs.map { |l| l[:lang] },
           script: langs.map { |l| script(l[:lang]) }.uniq,
           title: titles,
-          doctype: fetch_type(hit_data[:title]),
+          doctype: fetch_type(hit.hit[:title]),
           docstatus: fetch_status(doc),
           ics: fetch_ics(doc),
-          date: fetch_dates(doc, hit_data[:title]),
-          contributor: fetch_contributors(hit_data[:title]),
+          date: fetch_dates(doc, hit.hit[:title]),
+          contributor: fetch_contributors(hit.hit[:title]),
           editorialgroup: fetch_workgroup(doc),
           abstract: abstract,
           copyright: fetch_copyright(doc),
           link: fetch_link(doc, url),
           relation: fetch_relations(doc),
           place: ["Geneva"],
-          structuredidentifier: fetch_structuredidentifier(doc),
+          structuredidentifier: fetch_structuredidentifier(hit.pubid),
         )
       end
 
+      #
+      # Create document ids.
+      #
+      # @param doc [Nokogiri::HTML::Document] document
+      # @param pubid [Pubid::Iso::Identifier] pubid
+      #
+      # @return [Array<RelatonBib::DocumentIdentifier>]
+      #
+      def fetch_relaton_docids(doc, pubid)
+        pubid.urn_stage = stage_code(doc).to_f
+        [
+          RelatonIso::DocumentIdentifier.new(id: pubid, type: "ISO", primary: true),
+          RelatonIso::DocumentIdentifier.new(id: pubid, type: "URN"),
+        ]
+      end
+
       private
 
       # Fetch titles and abstracts.
@@ -168,60 +186,29 @@ module RelatonIso
       end
       # rubocop:enable Metrics/AbcSize, Metrics/MethodLength
 
-      # Fetch docid.
-      # @param doc [Nokogiri:HTML::Document]
-      # @param edition [String]
-      # @param langs [Array<Hash>]
-      # @return [Array<RelatonBib::DocumentIdentifier>]
-      def fetch_docid(doc, edition, langs)
-        pubid = item_ref doc
-        [
-          RelatonBib::DocumentIdentifier.new(id: pubid, type: "ISO", primary: true),
-          RelatonBib::DocumentIdentifier.new(
-            id: fetch_urn(doc, pubid, edition, langs), type: "URN",
-          ),
-        ]
-      end
-
-      # @param doc [Nokogiri:HTML::Document]
-      # @param pubid [String]
-      # @param edition [String]
-      # @param langs [Array<Hash>]
-      # @returnt [String]
-      def fetch_urn(doc, pubid, edition, langs) # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity, Metrics/MethodLength
-        orig = pubid.split.first.downcase.split("/").join "-"
-        %r{(?<=)(?<type>DATA|GUIDE|ISP|IWA|PAS|R|TR|TS|TTA)} =~ pubid
-        _, part, _year, corr, = IsoBibliography.ref_components pubid
-        urn = "urn:iso:std:#{orig}"
-        urn += ":#{type.downcase}" if type
-        urn += ":#{fetch_docnumber(doc)}"
-        urn += ":-#{part}" if part
-        urn += ":stage-#{stage_code(doc)}"
-        urn += ":ed-#{edition}" if edition
-        if corr
-          corrparts = corr.split
-          urn += ":#{corrparts[0].downcase}:#{corrparts[-1]}"
-        end
-        urn += ":#{langs.map { |l| l[:lang] }.join(',')}"
-        urn
-      end
-
-      def fetch_docnumber(doc)
-        item_ref(doc)&.match(/\d+/)&.to_s
+      #
+      # Generate docnumber.
+      #
+      # @param [Pubid::Iso] pubid
+      #
+      # @return [String] docnumber
+      #
+      def fetch_docnumber(pubid)
+        pubid.to_s.match(/\d+/)&.to_s
       end
 
-      # @param doc [Nokogiri::HTML::Document]
-      def fetch_structuredidentifier(doc) # rubocop:disable Metrics/MethodLength
-        ref = item_ref doc
-        unless ref
-          return RelatonIsoBib::StructuredIdentifier.new(
-            project_number: "?", part_number: "", prefix: nil, id: "?",
-          )
-        end
-
-        m = ref.match(/^(.*?\d+)-?((?<=-)\d+|)/)
+      #
+      # Parse structuredidentifier.
+      #
+      # @param pubid [Pubid::Iso::Identifier] pubid
+      #
+      # @return [RelatonBib::StructuredIdentifier] structured identifier
+      #
+      def fetch_structuredidentifier(pubid) # rubocop:disable Metrics/MethodLength
         RelatonIsoBib::StructuredIdentifier.new(
-          project_number: m[1], part: m[2], type: "ISO",
+          project_number: "#{pubid.publisher} #{pubid.number}",
+          part: pubid&.part&.sub(/^-/, ""),
+          type: pubid.publisher,
         )
       end
 
@@ -251,7 +238,7 @@ module RelatonIso
       # Fetch workgroup.
       # @param doc [Nokogiri::HTML::Document]
       # @return [Hash]
-      def fetch_workgroup(doc) # rubocop:disable Metrics/MethodLength
+      def fetch_workgroup(doc) # rubocop:disable Metrics/MethodLength,Metrics/AbcSize,Metrics/CyclomaticComplexity
         wg_link = doc.css("div.entry-name.entry-block a")[0]
         # wg_url = DOMAIN + wg_link['href']
         workgroup = wg_link.text.split "/"
@@ -275,6 +262,7 @@ module RelatonIso
       # @param doc [Nokogiri::HTML::Document]
       # @return [Array<Hash>]
       def fetch_relations(doc) # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity
+        types = ["Now", "Now under review"]
         doc.xpath("//ul[@class='steps']/li", "//div[@class='sub-step']").reduce([]) do |a, r|
           r_type = r.at("h4", "h5").text
           date = []
@@ -286,14 +274,13 @@ module RelatonIso
                    "updates"
                  else r_type
                  end
-          if ["Now", "Now under review"].include?(type) then a
+          if types.include?(type) then a
           else
             a + r.css("a").map do |id|
-              fref = RelatonBib::FormattedRef.new(
-                content: id.text, format: "text/plain",
-              )
+              docid = RelatonBib::DocumentIdentifier.new(type: "ISO", id: id.text, primary: true)
+              fref = RelatonBib::FormattedRef.new(content: id.text, format: "text/plain")
               bibitem = RelatonIsoBib::IsoBibliographicItem.new(
-                formattedref: fref, date: date,
+                docid: [docid], formattedref: fref, date: date,
               )
               { type: type, bibitem: bibitem }
             end
@@ -308,7 +295,7 @@ module RelatonIso
       def fetch_type(ref)
         %r{
           ^(?<prefix>ISO|IWA|IEC)
-          (?:(/IEC|/IEEE|/PRF|/NP|/DGuide)*\s|/)
+          (?:(?:/IEC|/IEEE|/PRF|/NP|/DGuide)*\s|/)
           (?<type>TS|TR|PAS|AWI|CD|FDIS|NP|DIS|WD|R|Guide|(?=\d+))
         }x =~ ref
         # return "international-standard" if type_match.nil?

data/lib/relaton_iso/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module RelatonIso
-  VERSION = "1.11.0"
+  VERSION = "1.12.1"
 end

data/lib/relaton_iso.rb

@@ -2,3 +2,5 @@
 
 require "relaton_iso/version"
 require "relaton_iso/iso_bibliography"
+require "pubid-iso"
+require "relaton_iso/document_identifier"

data/relaton_iso.gemspec

@@ -42,5 +42,6 @@ Gem::Specification.new do |spec|
 
   # spec.add_dependency "relaton-iec", "~> 1.8.0"
   spec.add_dependency "algolia"
-  spec.add_dependency "relaton-iso-bib", "~> 1.11.0"
+  spec.add_dependency "relaton-iso-bib", "~> 1.12.0"
+  spec.add_dependency "pubid-iso", "~> 0.1.7"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: relaton-iso
 version: !ruby/object:Gem::Version
-  version: 1.11.0
+  version: 1.12.1
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2022-04-10 00:00:00.000000000 Z
+date: 2022-07-05 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: byebug
@@ -184,14 +184,28 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.11.0
+        version: 1.12.0
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.11.0
+        version: 1.12.0
+- !ruby/object:Gem::Dependency
+  name: pubid-iso
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.1.7
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.1.7
 description: 'RelatonIso: retrieve ISO Standards for bibliographic use using the IsoBibliographicItem
   model'
 email:
@@ -230,7 +244,9 @@ files:
 - bin/ruby-rewrite
 - bin/safe_yaml
 - bin/setup
+- bin/thor
 - lib/relaton_iso.rb
+- lib/relaton_iso/document_identifier.rb
 - lib/relaton_iso/hit.rb
 - lib/relaton_iso/hit_collection.rb
 - lib/relaton_iso/iso_bibliography.rb