add optional language attribute to title and description

front-matter · Jan 31, 2024 · 2734afe · 2734afe
1 parent d3fc72a
commit 2734afe
Show file tree

Hide file tree

Showing 14 changed files with 555 additions and 135 deletions.
diff --git a/Gemfile.lock b/Gemfile.lock
@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    commonmeta-ruby (3.11.0)
+    commonmeta-ruby (3.12.0)
       activesupport (>= 4.2.5, < 8.0)
       addressable (~> 2.8.1, < 2.8.2)
       base32-url (>= 0.7.0, < 1)
@@ -58,15 +58,16 @@ GEM
       rubocop (~> 1.0)
     concurrent-ruby (1.2.3)
     connection_pool (2.4.1)
-    crack (0.4.5)
+    crack (0.4.6)
+      bigdecimal
       rexml
     crass (1.0.6)
     csl (2.0.0)
       namae (~> 1.0)
       rexml
     csl-styles (2.0.1)
       csl (~> 2.0)
-    diff-lcs (1.5.0)
+    diff-lcs (1.5.1)
     docile (1.4.0)
     domain_name (0.6.20240107)
     drb (2.2.0)
@@ -154,7 +155,7 @@ GEM
       iniparser (>= 0.1.0)
     public_suffix (4.0.7)
     racc (1.7.3)
-    rack (3.0.8)
+    rack (3.0.9)
     rack-test (2.1.0)
       rack (>= 1.3)
     rainbow (3.1.1)

diff --git a/lib/commonmeta/author_utils.rb b/lib/commonmeta/author_utils.rb
@@ -25,8 +25,8 @@ def datacite_contributor_roles = {
       "Researcher" => "Other",
       "Sponsor" => "Other",
       "Supervisor" => "Supervision",
-      "WorkPackageLeader" => "Other"
-    }
+      "WorkPackageLeader" => "Other",
+    }    
 
     def get_one_author(author)
       # basic sanity checks
@@ -55,20 +55,20 @@ def get_one_author(author)
            parse_attributes(author.fetch("identifier", nil), first: true) ||
            parse_attributes(author.fetch("sameAs", nil), first: true)
       id = normalize_orcid(id) || normalize_ror(id) if id.present?
-      
+
       # DataCite metadata
       if id.nil? && author["nameIdentifiers"].present?
         id = Array.wrap(author.dig("nameIdentifiers")).find do |ni|
           normalize_name_identifier(ni).present?
         end
         id = normalize_name_identifier(id) if id.present?
-      # Crossref metadata
+        # Crossref metadata
       elsif id.nil? && author["ORCID"].present?
         id = author.fetch("ORCID")
         id = normalize_orcid(id)
-      # JSON Feed metadata
+        # JSON Feed metadata
       elsif id.nil? && author["url"].present?
-        id = author.fetch("url")       
+        id = author.fetch("url")
       end
 
       # parse author type, i.e. "Person", "Organization" or not specified
@@ -168,6 +168,9 @@ def is_personal_name?(name: nil)
       # check if a name has only one word, e.g. "FamousOrganization", not including commas
       return false if name.to_s.split(" ").size == 1 && name.to_s.exclude?(",")
 
+      # check if name contains words known to be used in organization names
+      return false if %w[University College Institute School Center Department Laboratory Library Museum Foundation Society Association Company Corporation Collaboration Consortium Incorporated Inc. Institut Research Science].any? { |word| name.to_s.include?(word) }
+
       # check for suffixes, e.g. "John Smith, MD"
       return true if name && %w[MD PhD].include?(name.split(", ").last)
 

diff --git a/lib/commonmeta/readers/datacite_reader.rb b/lib/commonmeta/readers/datacite_reader.rb
@@ -4,170 +4,176 @@ module Commonmeta
   module Readers
     module DataciteReader
       def get_datacite(id: nil, **options)
-        return { 'string' => nil, 'state' => 'not_found' } unless id.present?
+        return { "string" => nil, "state" => "not_found" } unless id.present?
 
         api_url = datacite_api_url(id, options)
         response = HTTP.get(api_url)
-        return { 'string' => nil, 'state' => 'not_found' } unless response.status.success?
+        return { "string" => nil, "state" => "not_found" } unless response.status.success?
 
         body = JSON.parse(response.body)
-        client = Array.wrap(body.fetch('included', nil)).find do |m|
-          m['type'] == 'clients'
+        client = Array.wrap(body.fetch("included", nil)).find do |m|
+          m["type"] == "clients"
         end
-        client_id = client.to_h.fetch('id', nil)
-        provider_id = Array.wrap(client.to_h.fetch('relationships', nil)).find do |m|
-          m['provider'].present?
-        end.to_h.dig('provider', 'data', 'id')
-
-        { 'string' => response.body.to_s,
-          'provider_id' => provider_id,
-          'client_id' => client_id }
+        client_id = client.to_h.fetch("id", nil)
+        provider_id = Array.wrap(client.to_h.fetch("relationships", nil)).find do |m|
+          m["provider"].present?
+        end.to_h.dig("provider", "data", "id")
+
+        { "string" => response.body.to_s,
+          "provider_id" => provider_id,
+          "client_id" => client_id }
       end
 
       def read_datacite(string: nil, **_options)
         errors = jsonlint(string)
-        return { 'errors' => errors } if errors.present?
+        return { "errors" => errors } if errors.present?
 
         read_options = ActiveSupport::HashWithIndifferentAccess.new(_options.except(:doi, :id, :url,
                                                                                     :sandbox, :validate, :ra))
 
         meta = string.present? ? JSON.parse(string) : {}
 
         # optionally strip out the message wrapper from API
-        meta = meta.dig('data', 'attributes') if meta.dig('data').present?
+        meta = meta.dig("data", "attributes") if meta.dig("data").present?
 
         meta.transform_keys!(&:underscore)
 
-        id = normalize_doi(meta.fetch('doi', nil))
+        id = normalize_doi(meta.fetch("doi", nil))
 
-        resource_type_general = meta.dig('types', 'resourceTypeGeneral')
-        resource_type = meta.dig('types', 'resourceType')
+        resource_type_general = meta.dig("types", "resourceTypeGeneral")
+        resource_type = meta.dig("types", "resourceType")
         # if resource_type is one of the new resource_type_general types introduced in schema 4.3, use it
         type = Commonmeta::Utils::DC_TO_CM_TRANSLATIONS.fetch(resource_type, nil) ||
-               Commonmeta::Utils::DC_TO_CM_TRANSLATIONS.fetch(resource_type_general, 'Other')
+               Commonmeta::Utils::DC_TO_CM_TRANSLATIONS.fetch(resource_type_general, "Other")
 
-        alternate_identifiers = Array.wrap(meta.fetch('alternate_identifiers', nil)).map do |i|
+        alternate_identifiers = Array.wrap(meta.fetch("alternate_identifiers", nil)).map do |i|
           i.transform_keys! { |k| k.camelize(:lower) }
         end
-        url = meta.fetch('url', nil)
-        titles = Array.wrap(meta.fetch('titles', nil)).map do |title|
-          title.compact
+        url = meta.fetch("url", nil)
+        titles = Array.wrap(meta.fetch("titles", nil)).map do |title|
+          { "title" => title.fetch("title", nil),
+            "type" => title.fetch("titleType", nil),
+            "language" => title.fetch("lang", nil) }.compact
         end
-        contributors = get_authors(from_datacite(meta.fetch('creators', nil)))
-        contributors += get_authors(from_datacite(meta.fetch('contributors', nil)))
-        if meta.fetch('publisher', nil).is_a?(Hash)
-          publisher = { 'name' => meta.fetch('publisher', nil).fetch('name', nil) }
-        elsif meta.fetch('publisher', nil).is_a?(String)
-          publisher = { 'name' => meta.fetch('publisher', nil) }
+        contributors = get_authors(from_datacite(meta.fetch("creators", nil)))
+        contributors += get_authors(from_datacite(meta.fetch("contributors", nil)))
+        if meta.fetch("publisher", nil).is_a?(Hash)
+          publisher = { "name" => meta.fetch("publisher", nil).fetch("name", nil) }
+        elsif meta.fetch("publisher", nil).is_a?(String)
+          publisher = { "name" => meta.fetch("publisher", nil) }
         else
           publisher = nil
         end
 
-        container = meta.fetch('container', nil)
-        funding_references = meta.fetch('funding_references', nil)
+        container = meta.fetch("container", nil)
+        funding_references = meta.fetch("funding_references", nil)
 
         date = {}
-        date['created'] =
-          get_iso8601_date(meta.dig('created')) || get_date(meta.dig('dates'), 'Created')
-        date['published'] =
-          get_iso8601_date(meta.dig('published')) || get_date(meta.dig('dates'),
-                                                              'Issued') || get_iso8601_date(meta.dig('publication_year'))
-        date['registered'] = get_iso8601_date(meta.dig('registered'))
-        date['updated'] =
-          get_iso8601_date(meta.dig('updated')) || get_date(meta.dig('dates'), 'Updated')
-
-        descriptions = Array.wrap(meta.fetch('descriptions', nil)).map do |description|
-          description.compact
+        date["created"] =
+          get_iso8601_date(meta.dig("created")) || get_date(meta.dig("dates"), "Created")
+        date["published"] =
+          get_iso8601_date(meta.dig("published")) || get_date(meta.dig("dates"),
+                                                              "Issued") || get_iso8601_date(meta.dig("publication_year"))
+        date["registered"] = get_iso8601_date(meta.dig("registered"))
+        date["updated"] =
+          get_iso8601_date(meta.dig("updated")) || get_date(meta.dig("dates"), "Updated")
+
+        descriptions = Array.wrap(meta.fetch("descriptions", nil)).map do |description|
+          description_type = description.fetch("descriptionType", nil)
+          description_type = "Other" unless %w[Abstract Methods TechnicalInfo].include?(description_type)
+          { "description" => description.fetch("description", nil),
+            "type" => description_type,
+            "language" => description.fetch("lang", nil) }.compact
         end
-        license = Array.wrap(meta.fetch('rights_list', nil)).find do |r|
-          r['rightsUri'].present?
+        license = Array.wrap(meta.fetch("rights_list", nil)).find do |r|
+          r["rightsUri"].present?
         end
-        license = hsh_to_spdx('rightsURI' => license['rightsUri']) if license.present?
-        version = meta.fetch('version', nil)
-        subjects = meta.fetch('subjects', nil)
-        language = meta.fetch('language', nil)
-        geo_locations = meta.fetch('geo_locations', nil)
-        references = (Array.wrap(meta.fetch('related_identifiers',
-                                            nil)) + Array.wrap(meta.fetch('related_items',
+        license = hsh_to_spdx("rightsURI" => license["rightsUri"]) if license.present?
+        version = meta.fetch("version", nil)
+        subjects = meta.fetch("subjects", nil)
+        language = meta.fetch("language", nil)
+        geo_locations = meta.fetch("geo_locations", nil)
+        references = (Array.wrap(meta.fetch("related_identifiers",
+                                            nil)) + Array.wrap(meta.fetch("related_items",
                                                                           nil))).select do |r|
-                       %w[References Cites IsSupplementedBy].include?(r['relationType'])
-                     end.map do |reference|
+          %w[References Cites IsSupplementedBy].include?(r["relationType"])
+        end.map do |reference|
           get_datacite_reference(reference)
         end
-        files = Array.wrap(meta.fetch("content_url", nil)).map { |file| { "url" => file } }    
-        formats = meta.fetch('formats', nil)
-        sizes = meta.fetch('sizes', nil)
-        schema_version = meta.fetch('schema_version', nil) || 'http://datacite.org/schema/kernel-4'
-        state = id.present? || read_options.present? ? 'findable' : 'not_found'
-
-        { 'id' => id,
-          'type' => type,
-          'additional_type' => resource_type == type ? nil : resource_type,
-          'url' => url,
-          'titles' => titles,
-          'contributors' => contributors,
-          'container' => container,
-          'publisher' => publisher,
-          'provider' => 'DataCite',
-          'alternate_identifiers' => alternate_identifiers.presence,
-          'references' => references,
-          'funding_references' => funding_references,
-          'files' => files.presence,
-          'date' => date.compact,
-          'descriptions' => descriptions,
-          'license' => license,
-          'version' => version,
-          'subjects' => subjects,
-          'language' => language,
-          'geo_locations' => geo_locations,
-          'formats' => formats,
-          'sizes' => sizes,
-          'state' => state }.compact # .merge(read_options)
+        files = Array.wrap(meta.fetch("content_url", nil)).map { |file| { "url" => file } }
+        formats = meta.fetch("formats", nil)
+        sizes = meta.fetch("sizes", nil)
+        schema_version = meta.fetch("schema_version", nil) || "http://datacite.org/schema/kernel-4"
+        state = id.present? || read_options.present? ? "findable" : "not_found"
+
+        { "id" => id,
+          "type" => type,
+          "additional_type" => resource_type == type ? nil : resource_type,
+          "url" => url,
+          "titles" => titles,
+          "contributors" => contributors,
+          "container" => container,
+          "publisher" => publisher,
+          "provider" => "DataCite",
+          "alternate_identifiers" => alternate_identifiers.presence,
+          "references" => references,
+          "funding_references" => funding_references,
+          "files" => files.presence,
+          "date" => date.compact,
+          "descriptions" => descriptions,
+          "license" => license,
+          "version" => version,
+          "subjects" => subjects,
+          "language" => language,
+          "geo_locations" => geo_locations,
+          "formats" => formats,
+          "sizes" => sizes,
+          "state" => state }.compact # .merge(read_options)
       end
 
       def format_contributor(contributor)
-        type = contributor.fetch('nameType', nil)
-
-        { 'name' => type == 'Person' ? nil : contributor.fetch('name', nil),
-          'type' => type,
-          'givenName' => contributor.fetch('givenName', nil),
-          'familyName' => contributor.fetch('familyName', nil),
-          'nameIdentifiers' => contributor.fetch('nameIdentifiers', nil).presence,
-          'affiliations' => contributor.fetch('affiliations', nil).presence,
-          'contributorType' => contributor.fetch('contributorType', nil) }.compact
+        type = contributor.fetch("nameType", nil)
+
+        { "name" => type == "Person" ? nil : contributor.fetch("name", nil),
+          "type" => type,
+          "givenName" => contributor.fetch("givenName", nil),
+          "familyName" => contributor.fetch("familyName", nil),
+          "nameIdentifiers" => contributor.fetch("nameIdentifiers", nil).presence,
+          "affiliations" => contributor.fetch("affiliations", nil).presence,
+          "contributorType" => contributor.fetch("contributorType", nil) }.compact
       end
 
       def get_datacite_reference(reference)
         return nil unless reference.present? || !reference.is_a?(Hash)
 
-        key = reference['relatedIdentifier']
+        key = reference["relatedIdentifier"]
         doi = nil
         url = nil
 
-        case reference['relatedIdentifierType']
-        when 'DOI'
-          doi = normalize_doi(reference['relatedIdentifier'])
-        when 'URL'
-          url = reference['relatedIdentifier']
+        case reference["relatedIdentifierType"]
+        when "DOI"
+          doi = normalize_doi(reference["relatedIdentifier"])
+        when "URL"
+          url = reference["relatedIdentifier"]
         else
-          url = reference['relatedIdentifier']
+          url = reference["relatedIdentifier"]
         end
 
         {
-          'key' => key,
-          'doi' => doi,
-          'url' => url,
-          'contributor' => reference.dig('author'),
-          'title' => reference.dig('article-title'),
-          'publisher' => reference.dig('publisher'),
-          'publicationYear' => reference.dig('year'),
-          'volume' => reference.dig('volume'),
-          'issue' => reference.dig('issue'),
-          'firstPage' => reference.dig('first-page'),
-          'lastPage' => reference.dig('last-page'),
-          'containerTitle' => reference.dig('journal-title'),
-          'edition' => nil,
-          'unstructured' => doi.nil? ? reference.dig('unstructured') : nil
+          "key" => key,
+          "doi" => doi,
+          "url" => url,
+          "contributor" => reference.dig("author"),
+          "title" => reference.dig("article-title"),
+          "publisher" => reference.dig("publisher"),
+          "publicationYear" => reference.dig("year"),
+          "volume" => reference.dig("volume"),
+          "issue" => reference.dig("issue"),
+          "firstPage" => reference.dig("first-page"),
+          "lastPage" => reference.dig("last-page"),
+          "containerTitle" => reference.dig("journal-title"),
+          "edition" => nil,
+          "unstructured" => doi.nil? ? reference.dig("unstructured") : nil,
         }.compact
       end
     end

diff --git a/lib/commonmeta/schema_utils.rb b/lib/commonmeta/schema_utils.rb
@@ -5,7 +5,7 @@
 
 module Commonmeta
   module SchemaUtils
-    COMMONMETA = File.read(File.expand_path("../../resources/commonmeta_v0.10.6.json",
+    COMMONMETA = File.read(File.expand_path("../../resources/commonmeta_v0.10.7.json",
                                             __dir__))
 
     def json_schema_errors

diff --git a/lib/commonmeta/version.rb b/lib/commonmeta/version.rb
@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Commonmeta
-  VERSION = '3.11.0'
+  VERSION = '3.12.0'
 end
diff --git a/lib/commonmeta/writers/commonmeta_writer.rb b/lib/commonmeta/writers/commonmeta_writer.rb
@@ -8,4 +8,4 @@ def commonmeta
       end
     end
   end
-end
+end
-Original file line number
+Diff line change
@@ Expand Up / @@ -8,4 +8,4 @@ def commonmeta @@
           end
         end
       end
-    end
+    end