relaton-nist 1.19.0 → 1.19.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: cd31209270fb62d1b7dabbddf7f379eb7b33a21832f3e290dd9ed49c94d715c4
-  data.tar.gz: 8f15ad43f7d1edfd0a9135a6b750bf1414c3fd24216fd30c730586b9cd342f5c
+  metadata.gz: 84f25c99f7dfa62a09294c042a3774332fdb018841028aa7ef0dde5b07c9788c
+  data.tar.gz: 618fca8f81fe125d4c265c02d1cf4a65c8188b960f7dd91332e91b4ec22bb05d
 SHA512:
-  metadata.gz: ea6daf8959d56b98ab2e02e63315c2ecf7bd483824ec2ad0832fc9239e32d57e556942d86edbaa4c88655904d2e04e9a05c50cb4bbf38486773ab3cb0986c851
-  data.tar.gz: 2ca580e13556eb72455cc61f9a783d30894675d3f98630dd2de589a3a2a894380bc96a26453eb345e727eec37a7ecb8535b35aa011b34cab9580b858fc0e70ae
+  metadata.gz: 2e4c3cfdec6fbe121b8b669e7beca74df804bbda725c358fc0be0bccc8853d12644f79937edf9eebd06fb06a0a91b92b37d61ad9267784ed6a3f476d2f3ea8ac
+  data.tar.gz: 517f957beaf97d42b0cf0d8fc9749ef9af2f99f2db0510a994be4e4ac56003f2d6c5ad2fa46d6f574b0868c520e79797911dbbe5d6a1516382f6625f486b6e17

data/lib/relaton_nist/data_fetcher.rb

@@ -1,10 +1,12 @@
 # frozen_string_literal: true
 
 require "yaml"
+require "loc_mods"
+require_relative "mods_parser"
 
 module RelatonNist
   class DataFetcher
-    URL = "https://raw.githubusercontent.com/usnistgov/NIST-Tech-Pubs/nist-pages/xml/allrecords.xml"
+    URL = "https://github.com/usnistgov/NIST-Tech-Pubs/releases/download/May2024/allrecords-MODS.xml"
 
     def initialize(output, format)
       @output = output
@@ -68,16 +70,18 @@ module RelatonNist
     end
 
     def fetch_tech_pubs
-      docs = Nokogiri::XML OpenURI.open_uri URL
-      docs.xpath(
-        "/body/query/doi_record/report-paper/report-paper_metadata",
-      ).each { |doc| write_file TechPubsParser.parse(doc, series) }
+      docs = LocMods::Collection.from_xml OpenURI.open_uri(URL)
+      # docs.xpath(
+      #   "/body/query/doi_record/report-paper/report-paper_metadata",
+      # )
+      docs.mods.each { |doc| write_file ModsParser.new(doc, series).parse }
     end
 
     def add_static_files
       Dir["./static/*.yaml"].each do |file|
         hash = YAML.load_file file
-        write_file RelatonNist::NistBibliographicItem.from_hash(hash)
+        bib = RelatonNist::NistBibliographicItem.from_hash(hash)
+        index.add_or_update bib.docidentifier[0].id, file
       end
     end
 

data/lib/relaton_nist/mods_parser.rb

@@ -0,0 +1,209 @@
+module RelatonNist
+  class ModsParser
+    RELATION_TYPES = {
+      "otherVersion" => "editionOf",
+      "preceding" => "updates",
+      "succeeding" => "updatedBy",
+    }.freeze
+
+    ATTRS = %i[type docid title link abstract date doctype contributor relation place series].freeze
+
+    def initialize(doc, series)
+      @doc = doc
+      @series = series
+    end
+
+    # @return [RelatonNist::NistBibliographicItem]
+    def parse
+      args = ATTRS.each_with_object({}) do |attr, hash|
+        hash[attr] = send("parse_#{attr}")
+      end
+      NistBibliographicItem.new(**args)
+    end
+
+    def parse_type
+      "standard"
+    end
+
+    # @return [Array<RelatonBib::DocumentIdentifier>]
+    def parse_docid
+      [
+        { type: "NIST", id: pub_id, primary: true },
+        { type: "DOI", id: doi },
+      ].map { |id| RelatonBib::DocumentIdentifier.new(**id) }
+    end
+
+    # @return [String]
+    def pub_id
+      get_id_from_str doi
+    end
+
+    def get_id_from_str(str)
+      str.match(/\/((?:NBS|NIST).+)/)[1].gsub(".", " ")
+    end
+
+    # @return [String]
+    def doi
+      url = @doc.location.reduce(nil) { |m, l| m || l.url.detect { |u| u.usage == "primary display" } }
+      id = url.content.match(/10\.6028\/.+/)[0]
+      case id
+      when "10.6028/NBS.CIRC.sup" then "10.6028/NBS.CIRC.24e7sup"
+      when "10.6028/NBS.CIRC.supJun1925-Jun1926" then "10.6028/NBS.CIRC.24e7sup2"
+      when "10.6028/NBS.CIRC.supJun1925-Jun1927" then "10.6028/NBS.CIRC.24e7sup3"
+      when "10.6028/NBS.CIRC.24supJuly1922" then "10.6028/NBS.CIRC.24e6sup"
+      when "10.6028/NBS.CIRC.24supJan1924" then "10.6028/NBS.CIRC.24e6sup2"
+      else id
+      end
+    end
+
+    # @return [Array<RelatonBib::TypedTitleString>]
+    def parse_title
+      title = @doc.title_info.reduce([]) do |a, ti|
+        next a if ti.type == "alternative"
+
+        a += ti.title.map { |t| create_title(t, "title-main", ti.non_sort[0]) }
+        a + ti.sub_title.map { |t| create_title(t, "title-part") }
+      end
+      if title.size > 1
+        content = title.map { |t| t.title.content }.join(" - ")
+        title << create_title(content, "main")
+      elsif title.size == 1
+        title[0].instance_variable_set :@type, "main"
+      end
+      title
+    end
+
+    def create_title(title, type, non_sort = nil)
+      content = title.gsub("\n", " ").squeeze(" ").strip
+      content = "#{non_sort.content}#{content}".squeeze(" ") if non_sort
+      RelatonBib::TypedTitleString.new content: content, type: type, language: "en", script: "Latn"
+    end
+
+    def parse_link
+      @doc.location.map do |location|
+        url = location.url.first
+        type = url.usage == "primary display" ? "doi" : "src"
+        RelatonBib::TypedUri.new content: url.content, type: type
+      end
+    end
+
+    def parse_abstract
+      @doc.abstract.map do |a|
+        content = a.content.gsub("\n", " ").squeeze(" ").strip
+        RelatonBib::FormattedString.new content: content, language: "en", script: "Latn"
+      end
+    end
+
+    def parse_date
+      date = @doc.origin_info[0].date_issued.map do |di|
+        create_date(di, "issued")
+      # end + @doc.record_info[0].record_creation_date.map do |rcd|
+      #   create_date(rcd, "created")
+      # end + @doc.record_info[0].record_change_date.map do |rcd|
+      #   create_date(rcd, "updated")
+      end
+      date.compact
+    end
+
+    def create_date(date, type)
+      RelatonBib::BibliographicDate.new type: type, on: decode_date(date)
+    rescue Date::Error
+    end
+
+    def decode_date(date)
+      if date.encoding == "marc" && date.content.size == 6
+        Date.strptime(date.content, "%y%m%d").to_s
+      elsif date.encoding == "iso8601"
+        Date.strptime(date.content, "%Y%m%d").to_s
+      else date.content
+      end
+    end
+
+    def parse_doctype
+      RelatonBib::DocumentType.new(type: "standard")
+    end
+
+    def parse_contributor
+      # eaxclude primary contributors to avoid duplication
+      @doc.name.reject { |n| n.usage == "primary" }.map do |name|
+        entity, default_role = create_entity(name)
+        next unless entity
+
+        role = name.role.reduce([]) do |a, r|
+          a + r.role_term.map { |rt| { type: rt.content } }
+        end
+        role << { type: default_role } if role.empty?
+        RelatonBib::ContributionInfo.new entity: entity, role: role
+      end.compact
+    end
+
+    def create_entity(name)
+      case name.type
+      when "personal" then [create_person(name), "author"]
+      when "corporate" then [create_org(name), "publisher"]
+      end
+    end
+
+    def create_person(name)
+      # exclude typed name parts because they are not actual name parts
+      cname = name.name_part.reject(&:type).map(&:content).join(" ")
+      complatename = RelatonBib::LocalizedString.new cname, "en"
+      fname = RelatonBib::FullName.new completename: complatename
+      name_id = name.name_identifier[0]
+      identifier = RelatonBib::PersonIdentifier.new "uri", name_id.content if name_id
+      RelatonBib::Person.new name: fname, identifier: [identifier]
+    end
+
+    def create_org(name)
+      names = name.name_part.reject(&:type).map { |n| n.content.gsub("\n", " ").squeeze(" ").strip }
+      url = name.name_identifier[0]&.content
+      id = RelatonBib::OrgIdentifier.new "uri", url if url
+      RelatonBib::Organization.new name: names, identifier: [id]
+    end
+
+    def parse_relation
+      @doc.related_item.reject { |ri| ri.type == "series" }.map do |ri|
+        type = RELATION_TYPES[ri.type]
+        RelatonBib::DocumentRelation.new(type: type, bibitem: create_related_item(ri))
+      end
+    end
+
+    def create_related_item(item)
+      item_id = get_id_from_str related_item_id(item)
+      docid = RelatonBib::DocumentIdentifier.new type: "NIST", id: item_id
+      fref = RelatonBib::FormattedRef.new content: item_id
+      NistBibliographicItem.new(docid: [docid], formattedref: fref)
+    end
+
+    def related_item_id(item)
+      if item.other_type && item.other_type[0..6] == "10.6028"
+        item.other_type
+      else
+        item.name[0].name_part[0].content
+      end
+    end
+
+    def parse_place
+      @doc.origin_info.select { |p| p.event_type == "publisher"}.map do |p|
+        place = p.place[0].place_term[0].content
+        /(?<city>\w+), (?<state>\w+)/ =~ place
+        RelatonBib::Place.new city: city, region: create_region(state)
+      end
+    end
+
+    def create_region(state)
+      [RelatonBib::Place::RegionType.new(iso: state)]
+    rescue ArgumentError
+      []
+    end
+
+    def parse_series
+      @doc.related_item.select { |ri| ri.type == "series" }.map do |ri|
+        tinfo = ri.title_info[0]
+        tcontent = tinfo.title[0].strip
+        title = RelatonBib::TypedTitleString.new content: tcontent
+        RelatonBib::Series.new title: title, number: tinfo.part_number[0]
+      end
+    end
+  end
+end

data/lib/relaton_nist/nist_bibliography.rb

@@ -123,7 +123,7 @@ module RelatonNist
           next if iter && r.status.iteration != iteration
           return { ret: r } if !year
 
-          r.date.select { |d| d.type == "published" }.each do |d|
+          r.date.select { |d| d.type == "published" || d.type == "issued" }.each do |d|
             return { ret: r } if year.to_i == d.on(:year)
 
             missed_years << d.on(:year)

data/lib/relaton_nist/version.rb

@@ -1,3 +1,3 @@
 module RelatonNist
-  VERSION = "1.19.0".freeze
+  VERSION = "1.19.1".freeze
 end

data/relaton_nist.gemspec

@@ -24,7 +24,8 @@ Gem::Specification.new do |spec|
   spec.required_ruby_version = Gem::Requirement.new(">= 2.7.0")
 
   spec.add_dependency "base64"
-  spec.add_dependency "relaton-bib", "~> 1.19.0"
+  spec.add_dependency "loc_mods", "~> 0.2.0"
+  spec.add_dependency "relaton-bib", "~> 1.19.2"
   spec.add_dependency "relaton-index", "~> 0.2.0"
   spec.add_dependency "rubyzip"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: relaton-nist
 version: !ruby/object:Gem::Version
-  version: 1.19.0
+  version: 1.19.1
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2024-07-03 00:00:00.000000000 Z
+date: 2024-07-12 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: base64
@@ -24,20 +24,34 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: loc_mods
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.2.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.2.0
 - !ruby/object:Gem::Dependency
   name: relaton-bib
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.19.0
+        version: 1.19.2
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.19.0
+        version: 1.19.2
 - !ruby/object:Gem::Dependency
   name: relaton-index
   requirement: !ruby/object:Gem::Requirement
@@ -98,6 +112,7 @@ files:
 - lib/relaton_nist/hash_converter.rb
 - lib/relaton_nist/hit.rb
 - lib/relaton_nist/hit_collection.rb
+- lib/relaton_nist/mods_parser.rb
 - lib/relaton_nist/nist_bibliographic_item.rb
 - lib/relaton_nist/nist_bibliography.rb
 - lib/relaton_nist/processor.rb