relaton 3.0.3 → 3.0.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: a774b326943089257e3aef6e5402fc1e2d53d4ff557e8b19f845a62d64b9a99b
-  data.tar.gz: 665fa378d2380f79a4e2952df1ff3d0f7e1de6d91ead262d75261555a4170e5c
+  metadata.gz: f6cef1665e29f9d4a6e8ab48528253675e5b901399666138306c2be69c0d1b6e
+  data.tar.gz: e4eef10cab8373db1d911a4b68342b35c1bb335c15f986d24697f286e83d9a96
 SHA512:
-  metadata.gz: bfa2af357a4e288f0ce97fb82ff78ac58b39bc8f8fc8659584ebb29a4568a50175137b8b8cd32b6710aff65b8413cb4ab0d1ea674918187edd9a74592e331a6c
-  data.tar.gz: e2ad6a9822a6ffa5f9a625bb4095260b81b3906fcff8ba0cf4fbdcea0d543f9e2b14133c7dd29dad0f165c26713b16fc0f06ebfc06d134384d4d0f280ee873b1
+  metadata.gz: 41185be7ff82e3b45667eb33380fb3eee8e548e3ea24039d8743c5ed50b2fcc1b1c252536f87a1d6b0c4821e41e17b08a45e861af1d6886a8abe9df287a800a6
+  data.tar.gz: 3235473497b828e0f5d8890a59b6548994b2ab113d665f010d24e1970c6b24eaa398b298d7f83ba3b770153c558c712059ff5147b5f39d86b61ad5e35312d930

data/lib/relaton/3gpp/data_fetcher.rb

@@ -1,3 +1,4 @@
+require "yeptris/psych"
 require "fileutils"
 require "net/http"
 require "uri"
@@ -64,7 +65,7 @@ module Relaton
       # @return [String, nil] file name
       #
       def get_file(renewal) # rubocop:disable Metrics/MethodLength, Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity
-        @current = YAML.load_file CURRENT if File.exist? CURRENT
+        @current = Yeptris::Psych.load_file CURRENT if File.exist? CURRENT
         @current ||= {}
         uri = URI(CSV_URL)
         n = 0
@@ -183,7 +184,7 @@ module Relaton
       # @return [Relaton3gpp::BibliographicItem, nil] merged bibliographic item or nil if no merge has been done
       #
       def merge_duplication(bib, file)
-        hash = YAML.load_file file
+        hash = Yeptris::Psych.load_file file
         existed = Item.from_hash hash
         changed = update_source? bib, existed
         bib1, bib2, chng = transposed_relation bib, existed

data/lib/relaton/bipm/data_outcomes_parser.rb

@@ -1,3 +1,4 @@
+require "yeptris/psych"
 module Relaton::Bipm
   class DataOutcomesParser
     SHORTTYPE = {
@@ -130,7 +131,7 @@ module Relaton::Bipm
       fr_file = en_file.sub "en", "fr"
       [en_file, fr_file].map do |file|
         if File.exist? file
-          data = YAML.safe_load(File.read(file, encoding: "UTF-8"))
+          data = Yeptris::Psych.safe_load(File.read(file, encoding: "UTF-8"))
           path = file
         end
         [path, data]

data/lib/relaton/bipm/si_brochure_parser.rb

@@ -1,3 +1,4 @@
+require "yeptris/psych"
 require "moxml"
 
 module Relaton::Bipm
@@ -60,8 +61,8 @@ module Relaton::Bipm
             warn_duplicate = false
             item2 = Item.from_yaml File.read(outfile, encoding: "UTF-8")
             fix_si_brochure_id item2
-            hash1 = YAML.safe_load item1.to_yaml
-            hash2 = YAML.safe_load item2.to_yaml
+            hash1 = Yeptris::Psych.safe_load item1.to_yaml
+            hash2 = Yeptris::Psych.safe_load item2.to_yaml
             Item.from_yaml deep_merge(hash1, hash2).to_yaml
           else
             warn_duplicate = true

data/lib/relaton/core/data_fetcher.rb

@@ -1,3 +1,4 @@
+require "yeptris/psych"
 require "yeptris"
 require "fileutils"
 
@@ -121,7 +122,7 @@ module Relaton
         kept = ledger.merge(ledger_new).select do |_, row|
           row.is_a?(Hash) && row["file"] && @written.include?(row["file"])
         end
-        File.write ledger_path, YAML.dump(kept), encoding: "UTF-8"
+        File.write ledger_path, Yeptris::Psych.dump(kept), encoding: "UTF-8"
         @ledger = kept
       end
 
@@ -349,7 +350,7 @@ module Relaton
 
       def ledger
         @ledger ||= if File.exist?(ledger_path)
-                      YAML.safe_load_file(ledger_path) || {}
+                      Yeptris::Psych.safe_load_file(ledger_path) || {}
                     else
                       {}
                     end

data/lib/relaton/core/processor.rb

@@ -134,6 +134,21 @@ module Relaton
         copy
       end
 
+      # A copy of `pubid` with each given component the parse leaves nil, for
+      # #item_cache_key. `pubid` itself when there is nothing to add.
+      #
+      # @param pubid [Pubid::Identifier]
+      # @param values [Hash{Symbol => String, nil}]
+      # @return [Pubid::Identifier]
+      def with_components(pubid, **values)
+        values = values.select { |attr, value| value && pubid.public_send(attr).nil? }
+        return pubid if values.empty?
+
+        copy = pubid.class.from_hash pubid.to_hash
+        values.each { |attr, value| copy.public_send "#{attr}=", value }
+        copy
+      end
+
       # The cache key for a query: the parsed pubid with the `year` and
       # `all_parts` options folded in. The publication date range is not part
       # of the identity, so it never enters the key.
@@ -153,6 +168,20 @@ module Relaton
         pubid
       end
 
+      # The cache key of a fetched document: its primary docidentifier as
+      # parsed (`pubid`), plus the components that docid omits by design and
+      # the item keeps elsewhere. A flavor whose printed docid drops the
+      # edition (IHO `B-12`, ECMA `ECMA-262`) overrides this, or every
+      # edition gets one key and one cache file (relaton#408). The same set
+      # as relaton-cli's `MachineIndex::REFINABLE_KEYS`.
+      #
+      # @param pubid [Pubid::Identifier] the docid's parse; not mutated
+      # @param _item [Relaton::Bib::ItemData] the fetched document
+      # @return [Pubid::Identifier]
+      def item_cache_key(pubid, _item)
+        pubid
+      end
+
       def initialize
         raise "This is an abstract class!"
       end

data/lib/relaton/db/CLAUDE.md

@@ -223,6 +223,17 @@ relaton#189 item 2 / relaton#204. Two lutaml-store `FileSystem` stores under
   both naming the same file. There are no redirect entries. A document is
   deleted with the last row that points to it; a row whose document is gone is
   a miss.
+- **The item row's key must tell editions apart (relaton#408).** `Db#item_key`
+  parses the primary docidentifier and passes it through
+  `Core::Processor#item_cache_key(pubid, item)`. A flavor whose printed docid
+  omits the edition by design overrides the hook to add it from the item:
+  IHO adds `version` from `edition` (`IHO B-12` → `IHO B-12 3.0.0`), ECMA adds
+  `edition` and the `volume` extent (`ECMA-262` → `ECMA-262 ed16`), through
+  `#with_components`, which fills only components the parse leaves nil, on a
+  copy. Without it every edition got one item key and one file, and a fetch of
+  `IHO B-12 2.0.3` overwrote the file the `IHO B-12 3.0.0` query row named. A
+  cache with such a file is dropped by the next `Relaton::VERSION` bump
+  (`grammar_hash`), so no separate cache-version bump was made.
 
 Matching: the **exact** row first (canonical `to_hash`, not pubid `==`, which
 is false after a `from_hash` round trip for some flavors, e.g. UN). Only a
@@ -233,8 +244,9 @@ still applies and it never answers with another query's year-stripped copy.
 `ISO 19115:2003 === ISO 19115-1:2003` (and the same for IEC, BSI, JIS, GB).
 `Cache#subset_of?` therefore also requires that every component the row adds
 (at any depth) is allowed: only a language for the dated fallback
-(`LANGUAGE_COMPONENTS`), a year/date/month or a language for the date-range
-`candidates` (`EDITION_COMPONENTS`). A row with no `root.number` (DOI, ISBN)
+(`LANGUAGE_COMPONENTS`), a year/date/month, an edition/version (the
+`item_cache_key` components; not a volume, which names another document) or a
+language for the date-range `candidates` (`EDITION_COMPONENTS`). A row with no `root.number` (DOI, ISBN)
 goes to one of 256 digest buckets (`<flavor>/~<hex>`).
 
 Expiry (`Cache#valid_row?`): an undated row, and **every** `not_found` row,

data/lib/relaton/db/cache.rb

@@ -36,8 +36,11 @@ module Relaton
       # reads an omitted `part` as "any part" (`ISO 19115:2003 ===
       # ISO 19115-1:2003`), which names another document.
       LANGUAGE_COMPONENTS = %w[language languages].freeze
-      # What a row may add to a key and still be one of its editions.
-      EDITION_COMPONENTS = (%w[year date month] + LANGUAGE_COMPONENTS).freeze
+      # What a row may add to a key and still be one of its editions. An
+      # `edition` or a `version` comes from Core::Processor#item_cache_key
+      # (IHO, ECMA; relaton#408). A `volume` names another document.
+      EDITION_COMPONENTS =
+        (%w[year date month edition version] + LANGUAGE_COMPONENTS).freeze
 
       # @return [String]
       attr_reader :dir

data/lib/relaton/db.rb

@@ -382,13 +382,17 @@ module Relaton
       [query, query && processor.cache_key(code, year, opts, query)]
     end
 
-    # The key of a fetched document, from its primary identifier. The
-    # identifier is data, so an unparseable one gives no key.
+    # The key of a fetched document, from its primary identifier and what
+    # that identifier omits (Core::Processor#item_cache_key, relaton#408).
+    # The identifier is data, so an unparseable one gives no key.
     def item_key(bib, stdclass)
       docid = bib.docidentifier.detect(&:primary) || bib.docidentifier.first
       return unless docid&.content
 
-      cache_key docid.content, nil, {}, stdclass
+      key = cache_key docid.content, nil, {}, stdclass
+      return key unless key.is_a? ::Pubid::Identifier
+
+      @registry[stdclass].item_cache_key key, bib
     rescue ::Pubid::Errors::Error
       nil
     end

data/lib/relaton/ecma/processor.rb

@@ -21,6 +21,14 @@ module Relaton
         Bibliography.get(code, date, opts)
       end
 
+      # An ECMA record prints `ECMA-262` with no ` ed<N>`/` vol<N>`, and keeps
+      # them in `edition` and in a `volume` extent (relaton#408).
+      def item_cache_key(pubid, item)
+        volume = item.extent.flat_map(&:locality)
+          .detect { |l| l.type == "volume" }&.reference_from
+        with_components pubid, edition: item.edition&.content, volume: volume
+      end
+
       #
       # Fetch all the documents from a source
       #

data/lib/relaton/equivalence.rb

@@ -1,4 +1,4 @@
-require "yaml"
+require "yeptris/psych"
 
 module Relaton
   # Cross-SDO equivalence as data: chain patterns and explicit equivalences
@@ -23,7 +23,7 @@ module Relaton
                 else
                   [path]
                 end
-        files.each { |f| tables << YAML.safe_load_file(f, aliases: true) }
+        files.each { |f| tables << Yeptris::Psych.safe_load_file(f, aliases: true) }
         tables
       end
 

data/lib/relaton/gb/scraper.rb

@@ -1,7 +1,7 @@
+require "yeptris/psych"
 # encoding: UTF-8
 # frozen_string_literal: true
 
-require "yaml"
 require "gb_agencies"
 
 module Relaton
@@ -132,7 +132,7 @@ module Relaton
       # @param pref [String]
       # @return [Hash{String=>String}]
       def prefix(pref)
-        @prefixes ||= YAML.load_file File.join(__dir__, "yaml/prefixes.yaml")
+        @prefixes ||= Yeptris::Psych.load_file File.join(__dir__, "yaml/prefixes.yaml")
         @prefixes[pref]
       end
 

data/lib/relaton/ieee/data_fetcher.rb

@@ -1,3 +1,4 @@
+require "yeptris/psych"
 require "etc"
 require "zip"
 require_relative "../ieee"
@@ -77,7 +78,7 @@ module Relaton
         # @param file [String]
         # @return [String, nil]
         def primary_docid(file)
-          yaml = YAML.safe_load(File.read(file, encoding: "UTF-8"),
+          yaml = Yeptris::Psych.safe_load(File.read(file, encoding: "UTF-8"),
                                 permitted_classes: [Symbol, Date, Time])
           id = yaml && yaml["docidentifier"]&.find do |i|
             i["type"] == "IEEE" && i["primary"] == true && i["trademark"].nil?

data/lib/relaton/iho/processor.rb

@@ -25,6 +25,12 @@ module Relaton
         Bibliography.get(code, date, opts)
       end
 
+      # An IHO record prints `B-12` and keeps its edition in `edition`, so
+      # the edition goes on the key as the version (relaton#408).
+      def item_cache_key(pubid, item)
+        with_components pubid, version: item.edition&.content
+      end
+
       # @param xml [String]
       # @return [Relaton::Iho::ItemData]
       def from_xml(xml)

data/lib/relaton/index/file_io.rb

@@ -1,3 +1,4 @@
+require "yeptris/psych"
 module Relaton
   module Index
     #
@@ -219,12 +220,12 @@ module Relaton
       end
 
       def load_index(yaml, save = false)
-        index = YAML.safe_load(yaml, permitted_classes: [Symbol])
+        index = Yeptris::Psych.safe_load(yaml, permitted_classes: [Symbol])
         save index if save
         return deserialize_pubid(index) if check_format(index)
 
         report_invalid_index(save, "Wrong structure of")
-      rescue Psych::SyntaxError
+      rescue Yeptris::Psych::SyntaxError
         report_invalid_index(save, "YAML parsing error when reading")
       rescue InvalidIndexError
         report_invalid_index(save, "Wrong structure of")
@@ -430,7 +431,7 @@ module Relaton
         end
         yaml = yaml.read unless yaml.is_a?(String)
         Util.info "Downloaded index from `#{url}` for sqlite materialization", progname
-        raw = YAML.safe_load(yaml, permitted_classes: [Symbol])
+        raw = Yeptris::Psych.safe_load(yaml, permitted_classes: [Symbol])
         return nil unless check_format(raw)
 
         Enumerator.new do |y|
@@ -439,7 +440,7 @@ module Relaton
             y << [number, number, r[:id], r[:file]]
           end
         end
-      rescue Psych::SyntaxError, SocketError, OpenURI::HTTPError, Errno::ECONNRESET,
+      rescue Yeptris::Psych::SyntaxError, SocketError, OpenURI::HTTPError, Errno::ECONNRESET,
              OpenSSL::SSL::SSLError => e
         Util.info "SQLite index build failed (#{e.message}); falling back", progname
         nil
@@ -512,8 +513,8 @@ module Relaton
         return unless yaml
 
         begin
-          raw = YAML.safe_load(yaml, permitted_classes: [Symbol])
-        rescue Psych::SyntaxError
+          raw = Yeptris::Psych.safe_load(yaml, permitted_classes: [Symbol])
+        rescue Yeptris::Psych::SyntaxError
           warn_local_index_error("YAML parsing error when reading")
           delete_sidecar
           return [[], []]
@@ -613,7 +614,7 @@ module Relaton
           yaml = zip.entries.first.get_input_stream.read
         end
         Util.info "Downloaded index from `#{url}`", progname
-        raw = YAML.safe_load(yaml, permitted_classes: [Symbol])
+        raw = Yeptris::Psych.safe_load(yaml, permitted_classes: [Symbol])
         if check_format(raw)
           save raw
           raw = nil # release the parsed copy; read_raw_file loads the sidecar's
@@ -622,7 +623,7 @@ module Relaton
           warn_remote_index_error "Wrong structure of"
           [[], []]
         end
-      rescue Psych::SyntaxError
+      rescue Yeptris::Psych::SyntaxError
         warn_remote_index_error "YAML parsing error when reading"
         [[], []]
       end

data/lib/relaton/index/shard_source.rb

@@ -1,3 +1,4 @@
+require "yeptris/psych"
 require "yeptris"
 require "json"
 require "net/http"
@@ -152,11 +153,11 @@ module Relaton
           yaml = unzip(body)
           raise Error, "empty archive at `#{url(path)}`" unless yaml
 
-          index = YAML.safe_load(yaml, permitted_classes: [Symbol])
+          index = Yeptris::Psych.safe_load(yaml, permitted_classes: [Symbol])
           raise Error, "wrong structure of `#{url(path)}`" unless @file_io.check_basic_format(index)
 
           @file_io.deserialize_pubid(index)
-        rescue FileIO::InvalidIndexError, Psych::SyntaxError, Zip::Error, Zlib::Error => e
+        rescue FileIO::InvalidIndexError, Yeptris::Psych::SyntaxError, Zip::Error, Zlib::Error => e
           raise Error, "cannot read `#{url(path)}`: #{e.message}"
         end
       end

data/lib/relaton/iso/hit_collection.rb

@@ -88,13 +88,30 @@ module Relaton
       end
 
       def pubid_match?(pubid)
-        match_excludings = translate_excludings(excludings) + [:all_parts]
-        match_excludings << :edition unless pubid.typed_stage&.abbr&.include?("DIR")
+        match_excludings = base_match_excludings
+        unless pubid.typed_stage&.abbr&.include?("DIR")
+          match_excludings = match_excludings + [:edition]
+        end
         # Only the candidate is built via .create (from the index) and so may
         # carry a compound part; `ref_pubid_no_year` is always a parsed pubid,
         # already split, so it needs no normalization.
         cand = normalize_compound_part(exclude_id_attrs(pubid, *match_excludings))
-        cand == exclude_id_attrs(ref_pubid_no_year, *match_excludings)
+        cand == ref_excluded(match_excludings)
+      end
+
+      # The candidate's excludings set varies only by the DIR exception; the
+      # reference side is invariant per distinct set. Exclude the ref's
+      # (possibly deep) base chain once per set, not once per candidate —
+      # on a large vocabulary compile this chain ran per candidate per
+      # lookup and dominated the bibitem-resolution phase (relaton#423).
+      def base_match_excludings
+        @base_match_excludings ||= translate_excludings(excludings) + [:all_parts]
+      end
+
+      def ref_excluded(match_excludings)
+        @ref_excluded ||= {}
+        @ref_excluded[match_excludings] ||=
+          exclude_id_attrs(ref_pubid_no_year, *match_excludings)
       end
 
       # @TODO TEMP WORKAROUND (pubid 2.x migration): the v1-generated index

data/lib/relaton/jcgm/meetings_parser.rb

@@ -1,5 +1,5 @@
+require "yeptris/psych"
 require "fileutils"
-require "yaml"
 require "date"
 
 module Relaton::Jcgm
@@ -65,7 +65,7 @@ module Relaton::Jcgm
     end
 
     def load_yaml(file)
-      YAML.safe_load File.read(file, encoding: "UTF-8"), permitted_classes: [Date, Time]
+      Yeptris::Psych.safe_load File.read(file, encoding: "UTF-8"), permitted_classes: [Date, Time]
     end
 
     # "JCGM 17th Meeting (2012)" — ordinal via pubid's Meeting.ordinal.

data/lib/relaton/nist/cswp.rb

@@ -1,4 +1,4 @@
-require "yaml"
+require "yeptris/psych"
 
 module Relaton
   module Nist
@@ -24,7 +24,7 @@ module Relaton
       class << self
         # @return [Hash] the table as written in cswp.yaml
         def table
-          @table ||= YAML.load_file File.expand_path("cswp.yaml", __dir__)
+          @table ||= Yeptris::Psych.load_file File.expand_path("cswp.yaml", __dir__)
         end
 
         #

data/lib/relaton/nist/data_fetcher.rb

@@ -1,6 +1,7 @@
 # frozen_string_literal: true
 
-require "yaml"
+require "yeptris/psych"
+
 require "mechanize"
 require "loc_mods"
 require_relative "../nist"
@@ -183,7 +184,7 @@ module Relaton
       end
 
       def series
-        @series ||= YAML.load_file File.expand_path("series.yaml", __dir__)
+        @series ||= Yeptris::Psych.load_file File.expand_path("series.yaml", __dir__)
       end
     end
   end

data/lib/relaton/sdo/store.rb

@@ -1,5 +1,7 @@
 # frozen_string_literal: true
 
+require "yeptris/psych"
+
 module Relaton
   module Sdo
     # Singleton registry of organizations, built lazily from the index manifest
@@ -35,7 +37,7 @@ module Relaton
 
       def build_index
         yaml = Fetcher.fetch_index
-        data = yaml.nil? || yaml.empty? ? {} : YAML.safe_load(yaml)
+        data = yaml.nil? || yaml.empty? ? {} : Yeptris::Psych.safe_load(yaml)
         # A corrupt/unexpected index (non-mapping YAML) degrades to "no orgs"
         # rather than crashing every lookup with a TypeError.
         organizations = data.is_a?(Hash) ? (data["organizations"] || {}) : {}

data/lib/relaton/version.rb

@@ -1,3 +1,3 @@
 module Relaton
-  VERSION = "3.0.3".freeze
+  VERSION = "3.0.4".freeze
 end

data/lib/relaton/xsf/hit.rb

@@ -1,3 +1,4 @@
+require "yeptris/psych"
 module Relaton
   module Xsf
     class Hit < Relaton::Core::Hit
@@ -6,7 +7,7 @@ module Relaton
 
         agent = Mechanize.new
         resp = agent.get hit[:url]
-        hash = YAML.safe_load resp.body
+        hash = Yeptris::Psych.safe_load resp.body
         hash["fetched"] = Date.today.to_s
         @doc = Relaton::Xsf::Item.from_yaml hash.to_yaml
       rescue StandardError => e

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: relaton
 version: !ruby/object:Gem::Version
-  version: 3.0.3
+  version: 3.0.4
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2026-10-31 00:00:00.000000000 Z
+date: 2026-11-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: addressable
@@ -366,14 +366,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.5.174
+        version: 0.6.0
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.5.174
+        version: 0.6.0
 - !ruby/object:Gem::Dependency
   name: leptris
   requirement: !ruby/object:Gem::Requirement