relaton 3.0.0.pre.alpha.4 → 3.0.0.pre.alpha.5

data/lib/relaton/gb/scraper.rb

@@ -15,7 +15,7 @@ module Relaton
 
       @prefixes = nil
 
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Moxml::Document]
       # @param src [String]
       # @param hit [RelatonGb::Hit]
       # @return [Hash]
@@ -41,7 +41,7 @@ module Relaton
         [Docidentifier.new(content: docref, type: "Chinese Standard", primary: true)]
       end
 
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Moxml::Document]
       # @param docref [Strings]
       # @return [Array<Relaton::Bib::Contributor>]
       def get_contributors(doc, docref)
@@ -67,22 +67,22 @@ module Relaton
         Bib::TypedLocalizedString.new language: lang, content: content
       end
 
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Moxml::Document]
       # @return [Array<Relaton::Bib::Title>]
       def get_titles(doc)
-        tzh = doc.at("//td[contains(text(), '中文标准名称')]/b").text
+        tzh = doc.at_xpath("//td[contains(text(), '中文标准名称')]/b").text
         titles = Relaton::Bib::Title.from_string tzh, "zh", "Hans"
-        ten = doc.at("//td[contains(text(), '英文标准名称')]").text.match(/[\w\s]+/).to_s
+        ten = doc.at_xpath("//td[contains(text(), '英文标准名称')]").text.match(/[\w\s]+/).to_s
         return titles if ten.empty?
 
         titles + Relaton::Bib::Title.from_string(ten, "en", "Latn")
       end
 
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Moxml::Document]
       # @param status [String, NilClass]
       # @return [Relaton::Bib::Status]
       def get_status(doc, status = nil)
-        status ||= doc.at("//td[contains(., '标准状态')]/span")&.text&.strip
+        status ||= doc.at_xpath("//td[contains(., '标准状态')]/span")&.text&.strip
         return unless STAGES[status]
 
         stage = Bib::Status::Stage.new content: STAGES[status]
@@ -91,17 +91,17 @@ module Relaton
 
       private
 
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Moxml::Document]
       # @return [Array<String>]
       def get_ccs(doc)
-        code = doc.at("//div[contains(text(), '中国标准分类号')]/following-sibling::div").text.strip
+        code = doc.at_xpath("//div[contains(text(), '中国标准分类号')]/following-sibling::div").text.strip
         [CCS.new(code: code)]
       end
 
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Moxml::Document]
       # @return [Array<Relaton::Bib::ICS>]
       def get_ics(doc)
-        ics = doc.at("//div[contains(text(), '国际标准分类号')]/following-sibling::div"\
+        ics = doc.at_xpath("//div[contains(text(), '国际标准分类号')]/following-sibling::div"\
                     " | //dt[contains(text(), '国际标准分类号')]/following-sibling::dd")
         return [] unless ics
 
@@ -109,10 +109,10 @@ module Relaton
         [Bib::ICS.new(code: code)]
       end
 
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Moxml::Document]
       # @return [String]
       def get_scope(doc)
-        issued = doc.at("//div[contains(., '发布单位')]/following-sibling::div")
+        issued = doc.at_xpath("//div[contains(., '发布单位')]/following-sibling::div")
         case issued&.text
         when /国家标准/ then "national"
         when /^行业标准/ then "sector"
@@ -150,12 +150,12 @@ module Relaton
         (Bib::Uri.new(type: "src", content: src))
       end
 
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Moxml::Document]
       # @return [Array<Hash>]
       #   * :type [String] type of date
       #   * :on [String] date
       def get_dates(doc)
-        date = doc.at("//div[contains(text(), '发布日期')]/following-sibling::div"\
+        date = doc.at_xpath("//div[contains(text(), '发布日期')]/following-sibling::div"\
                       " | //dt[contains(text(), '发布日期')]/following-sibling::dd")
         [Bib::Date.new(type: "published", at: date.text.delete("\r\n\t\t"))]
       end
@@ -177,7 +177,7 @@ module Relaton
         Doctype.new content: "standard"
       end
 
-      # @param doc [Nokogiri::HTML::Document]
+      # @param doc [Moxml::Document]
       # @param ref [String]
       # @return [Relaton::Gb::GbType]
       def get_gbtype(doc, ref)

data/lib/relaton/gb/sec_scraper.rb

@@ -3,7 +3,7 @@
 
 require "net/http"
 require "json"
-require "nokogiri"
+require "moxml"
 require_relative "scraper"
 require_relative "item"
 require_relative "hit_collection"
@@ -44,7 +44,7 @@ module Relaton
         def scrape_doc(hit)
           src = "https://hbba.sacinfo.org.cn/stdDetail/#{hit.pid}"
           page_uri = URI src
-          doc = Nokogiri::HTML Net::HTTP.get(page_uri)
+          doc = Moxml.new.parse_html Net::HTTP.get(page_uri)
           ItemData.new(**scrapped_data(doc, src, hit))
         rescue SocketError, Timeout::Error, Errno::EINVAL, Errno::ECONNRESET, EOFError,
               Net::HTTPBadResponse, Net::HTTPHeaderSyntaxError, Net::ProtocolError,
@@ -54,14 +54,14 @@ module Relaton
 
         private
 
-        # @param doc [Nokogiri::HTML::Document]
+        # @param doc [Moxml::Document]
         # @return [Array<Relaton::Bib::Title>]
         def get_titles(doc)
-          tzh = doc.at("//h4").text.delete("\r\n\t")
+          tzh = doc.at_xpath("//h4").text.delete("\r\n\t")
           Bib::Title.from_string(tzh, "zh", "Hans")
         end
 
-        # @param _doc [Nokogiri::HTML::Document]
+        # @param _doc [Moxml::Document]
         # @param ref [String]
         # @return [Hash]
         #   * :type [String]
@@ -72,16 +72,16 @@ module Relaton
         #   { type: "technical", name: name }
         # end
 
-        # @param _doc [Nokogiri::HTML::Document]
+        # @param _doc [Moxml::Document]
         # @return [String]
         def get_scope(_doc)
           "sector"
         end
 
-        # @param doc [Nokogiri::HTML::Document]
+        # @param doc [Moxml::Document]
         # @return [Array<String>]
         def get_ccs(doc)
-          array(doc.at("//dt[contains(text(), '中国标准分类号')]/following-sibling::dd")).map do |cc|
+          array(doc.at_xpath("//dt[contains(text(), '中国标准分类号')]/following-sibling::dd")).map do |cc|
             text = Cnccs.fetch(cc.text.strip)&.description
             CCS.new code: cc.text, text: text
           end

data/lib/relaton/gb/t_scraper.rb

@@ -1,7 +1,7 @@
 # encoding: UTF-8
 # frozen_string_literal: true
 
-require "nokogiri"
+require "moxml"
 require_relative "scraper"
 require_relative "hit_collection"
 require_relative "hit"
@@ -23,8 +23,8 @@ module Relaton
           xpath = '//table[contains(@class, "standard_list_table")]/tr/td/a'
           t_xpath = "../preceding-sibling::td[4]"
           hits = doc.xpath(xpath).map do |h|
-            docref = h.at(t_xpath).text.gsub(/â\u0080\u0094/, "-")
-            status = h.at("../preceding-sibling::td[1]").text.delete "\r\n"
+            docref = h.at_xpath(t_xpath).text.gsub(/â\u0080\u0094/, "-")
+            status = h.at_xpath("../preceding-sibling::td[1]").text.delete "\r\n"
             pid = h[:href].sub(%r{/$}, "")
             Hit.new pid: pid, docref: docref, status: status, scraper: self
           end
@@ -55,7 +55,7 @@ module Relaton
         private
 
         # rubocop:disable Metrics/MethodLength
-        # @param doc [Nokogiri::HTML::Document]
+        # @param doc [Moxml::Document]
         # @param src [String]
         # @param hit [RelatonGb::Hit]
         # @return [Hash]
@@ -89,7 +89,7 @@ module Relaton
 
         def get_titles(doc)
           xpz = '//td[contains(.,"中文标题")]/following-sibling::td[1]'
-          titles = Bib::Title.from_string doc.at(xpz)
+          titles = Bib::Title.from_string doc.at_xpath(xpz)
             .text, "zh", "Hans"
           xpe = '//td[contains(.,"英文标题")]/following-sibling::td[1]'
           ten = doc.xpath(xpe).text

data/lib/relaton/gost/processor.rb

@@ -12,6 +12,7 @@ module Relaton
       def initialize
         @short = :relaton_gost
         @prefix = "GOST"
+        @pubid_identifier = :Gost # Db cache key
         # Both Latin "GOST" and Cyrillic "ГОСТ" route here. The trailing
         # \b keeps the prefix from swallowing longer tokens ("GOSTA …").
         @defaultprefix = %r{^(?:GOST|ГОСТ)\b}

data/lib/relaton/iala/processor.rb

@@ -8,6 +8,7 @@ module Relaton
       def initialize
         @short = :relaton_iala
         @prefix = "IALA"
+        @pubid_identifier = :Iala # Db cache key
         @defaultprefix = %r{^IALA\s}
         @idtype = "IALA"
       end

data/lib/relaton/iana/data_fetcher.rb

@@ -51,11 +51,11 @@ module Relaton
       end
 
       def parse(content)
-        xml = Nokogiri::XML(content)
-        registry = xml.at("/xmlns:registry")
+        xml = Moxml.parse(content)
+        registry = xml.at_xpath("/xmlns:registry", NS)
         doc = Parser.parse registry, nil, @errors
         save_doc doc
-        registry.xpath("./xmlns:registry").each { |r| save_doc Parser.parse(r, doc, @errors) }
+        registry.xpath("./xmlns:registry", NS).each { |r| save_doc Parser.parse(r, doc, @errors) }
       end
 
       #

data/lib/relaton/iana/parser.rb

@@ -1,10 +1,14 @@
 module Relaton
   module Iana
+    # The IANA registry XML uses a default namespace, and moxml (unlike raw
+    # nokogiri) does not bind the literal `xmlns:` prefix to it — every
+    # namespaced query passes this binding explicitly.
+    NS = { "xmlns" => "http://www.iana.org/assignments" }.freeze
     class Parser
       #
       # Document parser initalization
       #
-      # @param [Nokogiri::XML::Element] xml
+      # @param [Moxml::Element] xml
       #
       def initialize(xml, rootdoc, errors = {})
         @xml = xml
@@ -15,7 +19,7 @@ module Relaton
       #
       # Initialize document parser and run it
       #
-      # @param [Nokogiri::XML::Element] xml
+      # @param [Moxml::Element] xml
       #
       # @return [Relaton::Iana::ItemData, nil] bibliographic item
       #
@@ -50,7 +54,7 @@ module Relaton
       # @return [Array<Relaton::Bib::Title>] title
       #
       def parse_title
-        content = @xml.at("./xmlns:title")&.text || @xml[:id]
+        content = @xml.at_xpath("./xmlns:title", NS)&.text || @xml[:id]
         result = [Bib::Title.new(content: content, language: "en", script: "Latn")]
         @errors[:title] &&= result.empty?
         result
@@ -121,7 +125,7 @@ module Relaton
       # @return [Array<Relaton::Bib::Date>] date
       #
       def parse_date
-        d = @xml.xpath("./xmlns:created|./xmlns:published|./xmlns:updated").map do |dt|
+        d = @xml.xpath("./xmlns:created|./xmlns:published|./xmlns:updated", NS).map do |dt|
           Bib::Date.new(type: dt.name, at: dt.text)
         end
         result = d.none? && @rootdoc ? @rootdoc.date : d

data/lib/relaton/iana/processor.rb

@@ -8,6 +8,7 @@ module Relaton
       def initialize # rubocop:disable Lint/MissingSuper
         @short = :relaton_iana
         @prefix = "IANA"
+        @pubid_identifier = :Iana # Db cache key
         @defaultprefix = %r{^IANA\s}
         @idtype = "IANA"
         @datasets = %w[iana-registries]
@@ -35,6 +36,14 @@ module Relaton
         DataFetcher.fetch(source, **opts)
       end
 
+      # `get` reads a reference pubid cannot parse as a miss, so it gets no key
+      # here either: it is not cached.
+      def cache_pubid(ref)
+        super
+      rescue StandardError
+        nil
+      end
+
       # @param xml [String]
       # @return [Relaton::Iana::ItemData]
       def from_xml(xml)

data/lib/relaton/iec/data_parser.rb

@@ -289,38 +289,53 @@ module Relaton
       #
       # @return [Array<Relaton::Bib::Relation>] relation
       #
-      def relation # rubocop:disable Metrics/MethodLength
+      def relation
+        result = fetch_relations
+        @errors[:relation] &&= result.empty?
+        result
+      end
+
+      #
+      # Fetch relations from the webstore. Retry on network errors only.
+      # A non-success response (the legacy endpoint now redirects to a 404
+      # HTML page) or a malformed body gives no relations.
+      #
+      # @return [Array<Relaton::Bib::Relation>] relations
+      #
+      def fetch_relations # rubocop:disable Metrics/MethodLength
         try = 0
-        result = begin
+        begin
           uri = URI "#{DOMAIN}/webstore/webstore.nsf/AjaxRequestXML?Openagent&url=#{urn_id}"
           resp = Net::HTTP.get_response uri
-          doc = Nokogiri::XML resp.body
-          create_relations doc
         rescue StandardError => e
           try += 1
           try < 3 ? retry : raise(e)
         end
-        @errors[:relation] &&= result.empty?
-        result
+        return [] if resp.is_a?(Net::HTTPResponse) && !resp.is_a?(Net::HTTPSuccess)
+
+        create_relations Moxml.parse(resp.body)
+      rescue Moxml::ParseError => e
+        Util.warn "Failed to parse relations for `#{urn_id}`: #{e.message.lines.first&.strip}"
+        []
       end
 
       #
       # Create relations.
       #
-      # @param [Nokogiri::XML::Document] doc XML document
+      # @param [Moxml::Document] doc XML document
       #
       # @return [Array<Relaton::Bib::Relation>] relations
       #
       def create_relations(doc) # rubocop:disable Metrics/MethodLength
         doc.xpath('//ROW[STATUS[.!="PREPARING" and .!="PUBLISHED"]]')
           .map do |r|
-          r_type = r.at("STATUS").text.downcase
+          r_type = r.at_xpath("STATUS").text.downcase
           type = case r_type
                 when "revised", "replaced" then "updates"
                 when "withdrawn" then "obsoletes"
                 else r_type
                 end
-          ref = r.at("FULL_NAME").text
+          ref = r.at_xpath("FULL_NAME").text
           docid = Docidentifier.new(content: ref, type: "IEC", primary: true)
           bibitem = ItemData.new(formattedref: Bib::Formattedref.new(content: ref), docidentifier: [docid])
           Relation.new type: type, bibitem: bibitem

data/lib/relaton/iec/processor.rb

@@ -34,6 +34,12 @@ module Relaton
         DataFetcher.fetch(source, **opts)
       end
 
+      # `IEV` is not a document identifier (`get` answers it with the IEV
+      # vocabulary), so it gets no key: it is not cached.
+      def cache_pubid(ref)
+        ref.strip.casecmp?("IEV") ? nil : super
+      end
+
       # @param xml [String]
       # @return [Relaton::Iec::ItemData]
       def from_xml(xml)

data/lib/relaton/iec.rb

@@ -1,6 +1,6 @@
 require "digest/md5"
 require "net/http"
-require "nokogiri"
+require "moxml"
 require "pubid"
 require "zip"
 require "relaton/index"

data/lib/relaton/ieee/data_fetcher.rb

@@ -147,7 +147,7 @@ module Relaton
       # IdamsParser#parse_relation, so mutates crossrefs under a mutex.
       #
       # @param [String] docnumber of main document
-      # @param [Nokogiri::XML::Element] amsid relation data
+      # @param [Moxml::Element] amsid relation data
       #
       def add_crossref(docnumber, amsid)
         return if RELATION_TYPES[amsid.type] == false

data/lib/relaton/ieee/processor.rb

@@ -36,6 +36,14 @@ module Relaton
         DataFetcher.fetch(source, **opts)
       end
 
+      # `get` reads a reference pubid cannot parse as a miss, so it gets no key
+      # here either: it is not cached.
+      def cache_pubid(ref)
+        super
+      rescue StandardError
+        nil
+      end
+
       # @param xml [String]
       # @return [Relaton::Ieee::ItemData]
       def from_xml(xml)

data/lib/relaton/ietf/data_fetcher.rb

@@ -332,7 +332,7 @@ module Relaton
       #
       # Get RFC index
       #
-      # @return [Nokogiri::XML::Document] RFC index
+      # @return [Moxml::Document] RFC index
       #
       def rfc_index
         uri = URI "https://www.rfc-editor.org/rfc-index.xml"

data/lib/relaton/ietf/processor.rb

@@ -6,6 +6,7 @@ module Relaton
       def initialize # rubocop:disable Lint/MissingSuper
         @short = :relaton_ietf
         @prefix = "IETF"
+        @pubid_identifier = :Ietf # Db cache key
         @defaultprefix = /^((IETF|RFC|BCP|FYI|STD)\s|I-D[.\s])/
         @idtype = "IETF"
         @datasets = %w[ietf-rfcsubseries ietf-internet-drafts ietf-rfc-entries]

data/lib/relaton/ietf/rfc/entry.rb

@@ -195,8 +195,7 @@ module Relaton
             formattedref: build_formattedref,
             date: build_subseries_date(rfc_index),
             relation: build_relations(rfc_index, wg_names: wg_names),
-            series: build_series,
-            ext: Ext.new(doctype: Doctype.new(content: "rfc"), stream: stream, flavor: "ietf"),
+            ext: Ext.new(doctype: Doctype.new(content: "rfc"), stream: canonical_stream, flavor: "ietf"),
           )
         end
 
@@ -259,13 +258,6 @@ module Relaton
           ItemData.new(formattedref: Bib::Formattedref.new(content: ref), docidentifier: [docid])
         end
 
-        def build_series
-          return [] unless stream
-
-          t = Bib::Title.new(content: stream)
-          [Bib::Series.new(type: "stream", title: [t])]
-        end
-
         # --- RFC entry builders ---
 
         def build_rfc_docid
@@ -388,7 +380,7 @@ module Relaton
         def build_rfc_series
           series = build_rfc_is_also_series
           series << Bib::Series.new(title: [Bib::Title.new(content: "RFC")], number: shortnum)
-          series + build_rfc_stream_series
+          series
         end
 
         def build_rfc_is_also_series
@@ -401,27 +393,31 @@ module Relaton
           end
         end
 
-        def build_rfc_stream_series
-          return [] unless stream
-
-          t = Bib::Title.new(content: stream)
-          [Bib::Series.new(type: "stream", title: [t])]
-        end
-
         STREAM_ORGS = {
           "IETF" => ["IETF", "Internet Engineering Task Force"],
           "IRTF" => ["IRTF", "Internet Research Task Force"],
           "IAB" => ["IAB", "Internet Architecture Board"],
         }.freeze
 
+        # rfc-index.xml spells the independent stream "INDEPENDENT";
+        # Ext.stream's declared values use "Independent". Match case-insensitively
+        # against the declared values so both Ext.stream and STREAM_ORGS see the
+        # canonical spelling.
+        def canonical_stream
+          return unless stream
+
+          values = Ext.attributes[:stream].options[:values]
+          values.find { |v| v.casecmp?(stream) } || stream
+        end
+
         def build_rfc_ext
-          Ext.new(doctype: Doctype.new(content: "rfc"), stream: stream, flavor: "ietf")
+          Ext.new(doctype: Doctype.new(content: "rfc"), stream: canonical_stream, flavor: "ietf")
         end
 
         def build_committee_contributor(wg_names = {})
           return if wg_acronym.nil? || wg_acronym == "NON WORKING GROUP"
 
-          abbr, name = STREAM_ORGS[stream]
+          abbr, name = STREAM_ORGS[canonical_stream]
           org = if abbr
                   Ietf::BibXMLParser.build_org(abbr, name)
                 else

data/lib/relaton/iho/processor.rb

@@ -8,6 +8,7 @@ module Relaton
       def initialize
         @short = :relaton_iho
         @prefix = "IHO"
+        @pubid_identifier = :Iho # Db cache key
         @defaultprefix = %r{^IHO\s}
         @idtype = "IHO"
       end

data/lib/relaton/index/file_io.rb

@@ -28,11 +28,11 @@ module Relaton
       #   if nil then the fiename is used to read and write file (used to create indes in GH actions)
       # @param [Pubid::Identifier] pubid class for deserialization
       #
-      # `id_keys` is accepted for backward compatibility but no longer used: the
+      # The index format check round-trips each id through `pubid_class`;
       # index format is now validated by round-tripping a sample of ids through
       # the pubid class (see #check_serialization), which understands the pubid
       # v2 (lutaml) `_type` serialization that the old key-allowlist could not.
-      def initialize(dir, url, filename, _id_keys = nil, pubid_class = nil)
+      def initialize(dir, url: nil, filename: nil, pubid_class: nil)
         @dir = dir
         @url = url
         @filename = filename

data/lib/relaton/index/pool.rb

@@ -14,7 +14,6 @@ module Relaton
       # @param [String] type <description>
       # @param [String, nil] url external URL to index, used to fetch index for searching files
       # @param [String, nil] file output file name
-      # @param [Array<Symbol>, nil] id_keys keys to check if index is correct
       # @param [String, nil] pages_url base URL of the Pages site that serves
       #   the machine index (manifest + shards), see Relaton::Index::ShardSource
       #
@@ -24,9 +23,11 @@ module Relaton
         if @pool[type.upcase.to_sym]&.actual?(**args)
           @pool[type.upcase.to_sym]
         else
+          if args.key?(:id_keys)
+            Util.warn "id_keys is deprecated and ignored by Relaton::Index"
+          end
           @pool[type.upcase.to_sym] = Type.new(
-            type, args[:url], args[:file], args[:id_keys], args[:pubid_class],
-            pages_url: args[:pages_url]
+            type, **args.slice(:url, :file, :pubid_class, :pages_url)
           )
         end
       end

data/lib/relaton/index/shard_source.rb

@@ -47,7 +47,7 @@ module Relaton
         @pages_url = pages_url.end_with?("/") ? pages_url : "#{pages_url}/"
         # Only the deserialization helpers are used: this FileIO reads and
         # writes no file.
-        @file_io = FileIO.new(dir, nil, nil, nil, pubid_class)
+        @file_io = FileIO.new(dir, pubid_class: pubid_class)
         @mutex = Mutex.new
         @state = nil
       end

data/lib/relaton/index/type.rb

@@ -10,21 +10,18 @@ module Relaton
       # @param [String, Symbol] type type of index (ISO, IEC, etc.)
       # @param [String, nil] url external URL to index, used to fetch index for searching files
       # @param [String, nil] file output file name
-      # @param [Array<Symbol>] id_keys keys of identifier to be used for sorting index
-      #   format of index file is checked if id_keys all is provided at least in one of the IDs
       # @param [Pubid::Identifier, nil] pubid class for deserialization
       # @param [String, nil] pages_url base URL of the Pages site that serves
       #   the machine index. With it the type reads the index from there, in
       #   memory (see ShardSource), and a parsed query fetches one shard only.
       #
-      def initialize(type, url = nil, file = nil, id_keys = nil, pubid_class = nil, # rubocop:disable Metrics/ParameterLists
-                     pages_url: nil)
+      def initialize(type, url: nil, file: nil, pubid_class: nil, pages_url: nil)
         @file = file
         @dir = type.to_s.downcase
         @pubid_class = pubid_class
         @pages_url = pages_url
         filename = file || Index.config.filename
-        @file_io = FileIO.new @dir, url, filename, id_keys, pubid_class
+        @file_io = FileIO.new @dir, url: url, filename: filename, pubid_class: pubid_class
         @source = new_source
       end
 

data/lib/relaton/isbn/open_library.rb

@@ -8,23 +8,27 @@ module Relaton
 
       ENDPOINT = "http://openlibrary.org/api/volumes/brief/isbn/".freeze
 
-      def get(ref, _date = nil, _opts = {}) # rubocop:disable Metrics/MethodLength
-        Util.info "Fetching from OpenLibrary ...", key: ref
-
-        isbn = Isbn.new(ref).parse
+      # @param ref [String, Pubid::Isbn::Identifier] an ISBN-10 or ISBN-13
+      # @return [Relaton::Bib::ItemData, nil]
+      def get(ref, _date = nil, _opts = {}) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength
+        Util.info "Fetching from OpenLibrary ...", key: ref.to_s
+
+        # A parsed pubid gives its digits (`raw`); `Isbn#parse` validates them
+        # and converts an ISBN-10 to ISBN-13, as for a String.
+        isbn = Isbn.new(ref.is_a?(String) ? ref : ref.raw).parse
         unless isbn
-          Util.info "Incorrect ISBN.", key: ref
+          Util.info "Incorrect ISBN.", key: ref.to_s
           return
         end
 
         resp = request_api isbn
         unless resp
-          Util.info "Not found.", key: ref
+          Util.info "Not found.", key: ref.to_s
           return
         end
 
         bib = Parser.parse resp
-        Util.info "Found: `#{bib.docidentifier.first.content}`", key: ref
+        Util.info "Found: `#{bib.docidentifier.first.content}`", key: ref.to_s
         bib
       end
 

data/lib/relaton/isbn/processor.rb

@@ -8,6 +8,7 @@ module Relaton
       def initialize # rubocop:disable Lint/MissingSuper
         @short = :relaton_isbn
         @prefix = "ISBN"
+        @pubid_identifier = :Isbn # Db cache key
         @defaultprefix = /^ISBN\s/
         @idtype = "ISBN"
         @datasets = %w[]
@@ -22,6 +23,15 @@ module Relaton
         ::Relaton::Isbn::OpenLibrary.get(code, date, opts)
       end
 
+      # `OpenLibrary.get` reads an incorrect ISBN as a miss, so a reference
+      # pubid cannot parse (it checks the check digit too) gets no key here
+      # either: it is not cached.
+      def cache_pubid(ref)
+        super
+      rescue StandardError
+        nil
+      end
+
       #
       # @param xml [String]
       # @return [Relaton::Bib::Bibitem]