relaton-iso 1.12.0 → 1.12.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 3600bb96dbe0079a7a3decb1ed1e72d4bf8b0b1b3e3dbf895ce09babea1180e8
-  data.tar.gz: d01275c03e71b7bcc0137c3b72035cd19894eb39812d70f1ddadbb1e2c0354c7
+  metadata.gz: 4479e38048aa0dfae8bcc85f1e9de03b5fe0561048b658ec47b3df8ca64794eb
+  data.tar.gz: c297ddc7b15d8186b85fbb7d4d3f84863d7df6e20ad243f0740364262fe43807
 SHA512:
-  metadata.gz: 4e6d8c80fef320926e2c3fd71be2963e8d68216ee26079640495a610af32efce41fba270929a61bb989d8ddae1bc04bcf168428782e1e9d51643ca1ed49b59c1
-  data.tar.gz: 5d26dd354ca5d1cc6eeaac3ae612343ec63aa79fa5b584720490f7f0780937e47396cabf1fb0e4b2822161a9f7baad27df7c9bd4e9789bbb7387b3e91f80fc52
+  metadata.gz: 6fd11d9fe01bd36052cf2762df6e8f728d361fbe23410dcb0229e95436a9f7909e51e51a8be0abc9f7bd269ffd10d642cb0aa7792dba76c0c174b606e319bf58
+  data.tar.gz: 060edbed6bb5b11033911db2200a4e19e98585c4133a1b4a09eaa79b582f2cf23b39f424b84d3110429ff1800247cec4f2b6522a82bbbdc45cb093e63a339bda

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

@@ -1,16 +1,15 @@
 module RelatonIso
   class DocumentIdentifier < RelatonBib::DocumentIdentifier
-    attr_accessor :all_parts
-
     def id
+      id_str = @id.to_s.sub(/\sED\d+/, "")
       if @all_parts
         if type == "URN"
           return "#{@id.urn}:ser"
-            else
-          return "#{@id} (all parts)"
+        else
+          return "#{id_str} (all parts)"
         end
       end
-      type == "URN" ? @id.urn.to_s : @id.to_s
+      type == "URN" ? @id.urn.to_s : id_str
     end
 
     def remove_part

data/lib/relaton_iso/hit.rb

@@ -7,10 +7,10 @@ module RelatonIso
     attr_writer :fetch, :pubid
 
     # Parse page.
-    # @param lang [String, NilClass]
+    # @param lang [String, nil]
     # @return [RelatonIso::IsoBibliographicItem]
-    def fetch(lang = nil, all_parts = false)
-      @fetch ||= Scrapper.parse_page @hit, lang, all_parts
+    def fetch(lang = nil)
+      @fetch ||= Scrapper.parse_page self, lang
     end
 
     # @return [Integer]
@@ -26,7 +26,7 @@ module RelatonIso
 
     # @return [Pubid::Iso::Identifier]
     def pubid
-      Pubid::Iso::Identifier.parse_from_title(hit[:title])
+      @pubid ||= Pubid::Iso::Identifier.parse_from_title(hit[:title])
     end
   end
 end

data/lib/relaton_iso/hit_collection.rb

@@ -21,7 +21,7 @@ module RelatonIso
       hit = @array.min_by { |h| h.pubid.part }
       return @array.first.fetch lang unless hit
 
-      bibitem = hit.fetch(lang, true)
+      bibitem = hit.fetch(lang)
       all_parts_item = bibitem.to_all_parts
       @array.reject { |h| h.hit[:uuid] == hit.hit[:uuid] }.each do |hi|
         isobib = RelatonIsoBib::IsoBibliographicItem.new(

data/lib/relaton_iso/iso_bibliography.rb

@@ -73,26 +73,32 @@ module RelatonIso
         end
       end
 
-      def matches_base?(query_pubid, pubid, any_types_stages: false)
+      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)
+          ((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
 
       # @param hit_collection [RelatonIso::HitCollection]
       # @param year [String]
       # @return [RelatonIso::HitCollection]
-      def filter_hits_by_year(hit_collection, year)
+      def filter_hits_by_year(hit_collection, year) # rubocop:disable Metrics/MethodLength,Metrics/AbcSize,Metrics/CyclomaticComplexity,Metrics/PerceivedComplexity
         missed_years = []
 
         # 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_years << hit.pubid.year
+            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
@@ -106,9 +112,7 @@ module RelatonIso
 
       private
 
-      # rubocop:disable Metrics/MethodLength
-
-      def fetch_ref_err(query_pubid, year)
+      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."
@@ -124,14 +128,12 @@ module RelatonIso
         nil
       end
 
-      # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity
-
       # Search for hits. If no found then trying missed stages and ISO/IEC.
       #
       # @param query_pubid [Pubid::Iso::Identifier] reference without correction
       # @param opts [Hash]
       # @return [Array<RelatonIso::Hit>]
-      def isobib_search_filter(query_pubid, 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

data/lib/relaton_iso/scrapper.rb

@@ -50,52 +50,55 @@ module RelatonIso
 
     class << self
       # Parse page.
-      # @param hit_data [Hash]
+      # @param hit [RelatonIso::Hit]
       # @param lang [String, NilClass]
       # @return [RelatonIsoBib::IsoBibliographicItem]
-      def parse_page(hit_data, lang = nil, all_parts = false) # rubocop:disable Metrics/AbcSize, Metrics/MethodLength, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity
+      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_relaton_docids(Pubid::Iso::Identifier.parse(item_ref(doc))),
-          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
 
-      # Fetch relaton docids.
-      # @param pubid [Pubid::Iso::Identifier]
-      # @param edition [String]
-      # @param langs [Array<Hash>]
-      # @param stage [Float]
+      #
+      # Create document ids.
+      #
+      # @param doc [Nokogiri::HTML::Document] document
+      # @param pubid [Pubid::Iso::Identifier] pubid
+      #
       # @return [Array<RelatonBib::DocumentIdentifier>]
-      def fetch_relaton_docids(pubid)
-        pubid.urn_stage = 60.60
+      #
+      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"),
@@ -183,22 +186,29 @@ module RelatonIso
       end
       # rubocop:enable Metrics/AbcSize, Metrics/MethodLength
 
-      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
 
@@ -228,7 +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 "/"
@@ -252,6 +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 = []
@@ -263,14 +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
@@ -285,7 +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 = "1.12.0"
+  VERSION = "1.12.1"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: relaton-iso
 version: !ruby/object:Gem::Version
-  version: 1.12.0
+  version: 1.12.1
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2022-06-25 00:00:00.000000000 Z
+date: 2022-07-05 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: byebug
@@ -244,6 +244,7 @@ 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