bolognese 0.2.2 → 0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 3a6b175047cc9fe1cae1eba94b41164dd0c9e406
-  data.tar.gz: f78c9e066599d1a2b6db5c66cc1a19e8024dd73b
+  metadata.gz: d4728c17ca6677056bae7bb34fd334b2a018bbf5
+  data.tar.gz: 69525494936ec8e6e4fe396f9b490b9fe19f1aa8
 SHA512:
-  metadata.gz: fede61f6811145761b605513ae01345508e8f60a7e2d807d22ac17fd80096326d7ab4af1a8b3ca117c1856adb2f31012a23ed3392023d8f9213b424d32c2e459
-  data.tar.gz: 4ddc0ca66cda0fbb1ea3d10021d97b1f532f733dd221c39509d1994a4c51da09f27da515fdb0536d086d425d3d0075a00fb6f498e9813802e19b0bcde70cbb87
+  metadata.gz: 85aa5235f96786e7f005ee588abb8d9f6bff889102d4b8d7a5b2ec91badcfd6b1f84cf2f0fdf33813768b396f98774a1ed4cd570089255628f105c2effc862d9
+  data.tar.gz: 6db1eb5f1ab14ec751468e4192a0d09a05a12dbf79193a80c0ceb30e9deb41c641d76d9331eba1775422308baa94f531abbae8d92cb2b0908c8f9c84ebec06ca

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    Bolognese (0.2)
+    bolognese (0.3)
       activesupport (~> 4.2, >= 4.2.5)
       builder (~> 3.2, >= 3.2.2)
       dotenv (~> 2.1, >= 2.1.1)
@@ -103,7 +103,7 @@ PLATFORMS
   ruby
 
 DEPENDENCIES
-  Bolognese!
+  bolognese!
   bundler (~> 1.0)
   codeclimate-test-reporter (~> 1.0, >= 1.0.0)
   rack-test (~> 0)

data/lib/bolognese/cli.rb

@@ -17,13 +17,15 @@ module Bolognese
     end
 
     desc "read pid", "read metadata for PID"
-    method_option :as, :default => "schema_org"
+    method_option :as, default: "schema_org"
     def read(pid)
       provider = Metadata.new(pid).provider
 
       case
       when provider == "crossref" && options[:as] == "crossref"
         puts Crossref.new(pid).raw
+      when provider == "crossref" && options[:as] == "datacite"
+        puts Crossref.new(pid).as_datacite
       when provider == "crossref"
         puts Crossref.new(pid).as_schema_org
       when provider == "datacite" && options[:as] == "datacite"

data/lib/bolognese/crossref.rb

@@ -1,13 +1,15 @@
 require_relative 'doi_utils'
+require_relative 'datacite_utils'
 require_relative 'utils'
 
 module Bolognese
   class Crossref < Metadata
     include Bolognese::DoiUtils
     include Bolognese::Utils
+    include Bolognese::DataciteUtils
 
     # CrossRef types from https://api.crossref.org/types
-    CROSSREF_TYPE_TRANSLATIONS = {
+    CR_TO_SO_TRANSLATIONS = {
       "Proceedings" => nil,
       "ReferenceBook" => nil,
       "JournalIssue" => nil,
@@ -55,6 +57,10 @@ module Bolognese
       metadata.present?
     end
 
+    def doi
+      doi_from_url(id)
+    end
+
     def journal_metadata
       metadata.dig("crossref", "journal", "journal_metadata").presence || {}
     end
@@ -82,7 +88,7 @@ module Bolognese
     end
 
     def type
-      CROSSREF_TYPE_TRANSLATIONS[additional_type] || "CreativeWork"
+      CR_TO_SO_TRANSLATIONS[additional_type] || "CreativeWork"
     end
 
     def name
@@ -110,6 +116,10 @@ module Bolognese
       end
     end
 
+    def keywords
+
+    end
+
     def author
       person = bibliographic_metadata.dig("contributors", "person_name")
       Array(person).select { |a| a["contributor_role"] == "author" }.map do |a|
@@ -130,6 +140,14 @@ module Bolognese
       end.presence
     end
 
+    def version
+
+    end
+
+    def date_created
+
+    end
+
     def date_published
       pub_date = bibliographic_metadata.fetch("publication_date", nil) ||
         bibliographic_metadata.fetch("acceptance_date", nil)
@@ -173,6 +191,10 @@ module Bolognese
        end.presence
     end
 
+    def related_identifiers
+
+    end
+
     def provider
       { "@type" => "Organization",
         "name" => "Crossref" }

data/lib/bolognese/datacite.rb

@@ -1,12 +1,14 @@
 require_relative 'doi_utils'
+require_relative 'datacite_utils'
 require_relative 'utils'
 
 module Bolognese
   class Datacite < Metadata
     include Bolognese::DoiUtils
+    include Bolognese::DataciteUtils
     include Bolognese::Utils
 
-    DATACITE_TYPE_TRANSLATIONS = {
+    DC_TO_SO_TRANSLATIONS = {
       "Audiovisual" => "VideoObject",
       "Collection" => "Collection",
       "Dataset" => "Dataset",
@@ -42,9 +44,13 @@ module Bolognese
       metadata.present?
     end
 
+    def doi
+      doi_from_url(id)
+    end
+
     def type
       k = metadata.dig("resourceType", "resourceTypeGeneral")
-      DATACITE_TYPE_TRANSLATIONS[k.to_s.dasherize] || "CreativeWork"
+      DC_TO_SO_TRANSLATIONS[k.to_s.dasherize] || "CreativeWork"
     end
 
     def additional_type

data/lib/bolognese/datacite_utils.rb

@@ -0,0 +1,211 @@
+module Bolognese
+  module DataciteUtils
+    #   @resource_type = metadata.fetch("resource_type", nil)
+    #   @rights_list = metadata.fetch("rights_list", nil)
+    #   @descriptions = metadata.fetch("descriptions", nil)
+    #   @contributors = metadata.fetch("contributors", nil)
+    #   @alternate_identifier = metadata.fetch("alternate_identifier", nil)
+    #   @media = metadata.fetch("media", nil)
+
+    SO_TO_DC_TRANSLATIONS = {
+      "VideoObject" => "Audiovisual",
+      "Collection" => "Collection",
+      "Dataset" => "Dataset",
+      "Event" => "Event",
+      "ImageObject" => "Image",
+      "Service" => "Service",
+      "SoftwareSourceCode" => "Software",
+      "AudioObject" => "Sound",
+      "ScholarlyArticle" => "Text",
+      "CreativeWork" => "Other"
+    }
+
+    LICENSE_NAMES = {
+      "http://creativecommons.org/publicdomain/zero/1.0/" => "Public Domain (CC0 1.0)",
+      "https://creativecommons.org/licenses/by/4.0/" => "Creative Commons Attribution 4.0 (CC-BY 4.0)",
+      "https://creativecommons.org/licenses/by-nc/4.0/" => "Creative Commons Attribution Noncommercial 4.0 (CC-BY-NC 4.0)",
+      "https://creativecommons.org/licenses/by-sa/4.0/" => "Creative Commons Attribution Share Alike 4.0 (CC-BY-SA 4.0)",
+      "https://creativecommons.org/licenses/by-nc-nd/4.0/" => "Creative Commons Attribution Noncommercial No Derivaties 4.0 (CC-BY-NC-ND 4.0)"
+    }
+
+    SCHEMA = File.expand_path("../../../resources/kernel-4.0/metadata.xsd", __FILE__)
+
+    def schema
+      Nokogiri::XML::Schema(open(SCHEMA))
+    end
+
+    def datacite_xml
+      Nokogiri::XML::Builder.new(:encoding => 'UTF-8') do |xml|
+        xml.resource(root_attributes) do
+          insert_work(xml)
+        end
+      end.to_xml
+    end
+
+    def as_datacite
+      if schema.validate(Nokogiri::XML(datacite_xml))
+        datacite_xml
+      end
+    end
+
+    def validation_errors
+      @validation_errors ||= schema.validate(Nokogiri::XML(datacite_xml))
+    end
+
+    def insert_work(xml)
+      insert_identifier(xml)
+      insert_creators(xml)
+      insert_titles(xml)
+      insert_publisher(xml)
+      insert_publication_year(xml)
+      insert_resource_type(xml)
+      insert_alternate_identifiers(xml)
+      insert_subjects(xml)
+      insert_contributors(xml)
+      insert_dates(xml)
+      insert_related_identifiers(xml)
+      insert_version(xml)
+      insert_rights_list(xml)
+      insert_descriptions(xml)
+    end
+
+    def insert_identifier(xml)
+      xml.identifier(doi, 'identifierType' => "DOI")
+    end
+
+    def insert_creators(xml)
+      xml.creators do
+        Array(author).each do |creator|
+          xml.creator do
+            insert_person(xml, creator, "creator")
+          end
+        end
+      end
+    end
+
+    def insert_contributors(xml)
+      return xml unless editor.present?
+
+      xml.contributors do
+        Array(editor).each do |contributor|
+          xml.contributor("contributorType" => contributor[:contributor_type]) do
+            insert_person(xml, contributor, "contributor")
+          end
+        end
+      end
+    end
+
+    def insert_person(xml, person, type)
+      person_name = [person["familyName"], person["givenName"], person["name"]].compact.join(", ")
+
+      xml.send(:'creatorName', person_name) if type == "creator"
+      xml.send(:'contributorName', person_name) if type == "contributor"
+      xml.send(:'givenName', person[:given_name]) if person["givenName"].present?
+      xml.send(:'familyName', person[:family_name]) if person["familyName"].present?
+      xml.nameIdentifier(person["@id"], 'schemeURI' => 'http://orcid.org/', 'nameIdentifierScheme' => 'ORCID') if person["@id"].present?
+    end
+
+    def insert_titles(xml)
+      xml.titles do
+        insert_title(xml)
+      end
+    end
+
+    def insert_title(xml)
+      xml.title(name)
+    end
+
+    def insert_publisher(xml)
+      xml.publisher(is_part_of["name"])
+    end
+
+    def insert_publication_year(xml)
+      xml.publicationYear(date_published[0..3])
+    end
+
+    def insert_resource_type(xml)
+      return xml unless type.present?
+
+      xml.resourceType(additional_type, 'resourceTypeGeneral' => SO_TO_DC_TRANSLATIONS[type])
+    end
+
+    def insert_alternate_identifiers(xml)
+      return xml unless alternate_name.present?
+
+      xml.alternateIdentifiers do
+        xml.alternateIdentifier(alternate_name, 'alternateIdentifierType' => "Local accession number")
+      end
+    end
+
+    def insert_dates(xml)
+      xml.dates do
+        insert_date(xml, date_created, 'Created') if date_created.present?
+        insert_date(xml, date_published, 'Issued') if date_published.present?
+        insert_date(xml, date_modified, 'Updated') if date_modified.present?
+      end
+    end
+
+    def insert_date(xml, date, date_type)
+      xml.date(date, 'dateType' => date_type)
+    end
+
+    def insert_subjects(xml)
+      return xml unless keywords.present?
+
+      xml.subjects do
+        keywords.split(", ").each do |subject|
+          xml.subject(subject)
+        end
+      end
+    end
+
+    def insert_version(xml)
+      return xml unless version.present?
+
+      xml.version(version)
+    end
+
+    def insert_related_identifiers(xml)
+      return xml unless related_identifiers.present?
+
+      xml.relatedIdentifiers do
+        related_identifiers.each do |related_identifier|
+          xml.relatedIdentifier(related_identifier[:value], 'relatedIdentifierType' => related_identifier[:related_identifier_type], 'relationType' => related_identifier[:relation_type])
+        end
+      end
+    end
+
+    def insert_rights_list(xml)
+      return xml unless license.present?
+
+      xml.rightsList do
+        xml.rights(LICENSE_NAMES[license], 'rightsURI' => license)
+      end
+    end
+
+    def insert_descriptions(xml)
+      return xml unless description.present?
+
+      xml.descriptions do
+        xml.description(description)
+      end
+    end
+
+    def without_control(s)
+      r = ''
+      s.each_codepoint do |c|
+        if c >= 32
+          r << c
+        end
+      end
+      r
+    end
+
+    def root_attributes
+      { :'xmlns:xsi' => 'http://www.w3.org/2001/XMLSchema-instance',
+        :'xsi:schemaLocation' => 'http://datacite.org/schema/kernel-4 http://schema.datacite.org/meta/kernel-4/metadata.xsd',
+        :'xmlns' => 'http://datacite.org/schema/kernel-4' }
+    end
+  end
+end
+

data/lib/bolognese/version.rb

@@ -1,3 +1,3 @@
 module Bolognese
-  VERSION = "0.2.2"
+  VERSION = "0.3"
 end

data/resources/kernel-4.0/include/datacite-contributorType-v4.xsd

@@ -0,0 +1,35 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!-- Version 1.0 - Created 2011-01-13 - FZ, TIB, Germany
+   2013-05 v3.0: Addition of ID to simpleType element, added values "ResearchGroup" & "Other"
+   2014-08-20 v3.1: Addition of value "DataCurator"
+   2015-05-14 v4.0 dropped value "Funder", use new "funderReference" -->
+<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema" xmlns="http://datacite.org/schema/kernel-4" targetNamespace="http://datacite.org/schema/kernel-4" elementFormDefault="qualified">
+  <xs:simpleType name="contributorType" id="contributorType">
+    <xs:annotation>
+      <xs:documentation>The type of contributor of the resource.</xs:documentation>
+    </xs:annotation>
+    <xs:restriction base="xs:string">
+      <xs:enumeration value="ContactPerson"/>
+      <xs:enumeration value="DataCollector"/>
+      <xs:enumeration value="DataCurator"/>
+      <xs:enumeration value="DataManager"/>
+      <xs:enumeration value="Distributor"/>
+      <xs:enumeration value="Editor"/>
+      <xs:enumeration value="HostingInstitution"/>
+      <xs:enumeration value="Other"/>
+      <xs:enumeration value="Producer"/>
+      <xs:enumeration value="ProjectLeader"/>
+      <xs:enumeration value="ProjectManager"/>
+      <xs:enumeration value="ProjectMember"/>
+      <xs:enumeration value="RegistrationAgency"/>
+      <xs:enumeration value="RegistrationAuthority"/>
+      <xs:enumeration value="RelatedPerson"/>
+      <xs:enumeration value="ResearchGroup"/>
+      <xs:enumeration value="RightsHolder"/>
+      <xs:enumeration value="Researcher"/>
+      <xs:enumeration value="Sponsor"/>
+      <xs:enumeration value="Supervisor"/>
+      <xs:enumeration value="WorkPackageLeader"/>
+    </xs:restriction>
+  </xs:simpleType>
+</xs:schema>

data/resources/kernel-4.0/include/datacite-dateType-v4.xsd

@@ -0,0 +1,21 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!-- Version 1.0 - Created 2011-01-13 - FZ, TIB, Germany
+     2013-05 v3.0: Addition of ID to simpleType element; addition of value "Collected"; deleted "StartDate" & "EndDate"-->
+<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema" xmlns="http://datacite.org/schema/kernel-4" targetNamespace="http://datacite.org/schema/kernel-4" elementFormDefault="qualified">
+  <xs:simpleType name="dateType" id="dateType">
+    <xs:annotation>
+      <xs:documentation>The type of date. Use RKMS‐ISO8601 standard for depicting date ranges.To indicate the end of an embargo period, use Available. To indicate the start of an embargo period, use Submitted or Accepted, as appropriate.</xs:documentation>
+    </xs:annotation>
+    <xs:restriction base="xs:string">
+      <xs:enumeration value="Accepted"/>
+      <xs:enumeration value="Available"/>
+      <xs:enumeration value="Collected"/>
+      <xs:enumeration value="Copyrighted"/>
+      <xs:enumeration value="Created"/>
+      <xs:enumeration value="Issued"/>
+      <xs:enumeration value="Submitted"/>
+      <xs:enumeration value="Updated"/>
+      <xs:enumeration value="Valid"/>
+    </xs:restriction>
+  </xs:simpleType>
+</xs:schema>

data/resources/kernel-4.0/include/datacite-descriptionType-v4.xsd

@@ -0,0 +1,19 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!-- Version 1.0 - Created 2011-01-13 - FZ, TIB, Germany
+     2013-05 v3.0: Addition of ID to simpleType element, addition of value "Methods"
+     2015-02-12 v4.0: Addition of value "TechnicalInfo"-->
+<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema" xmlns="http://datacite.org/schema/kernel-4" targetNamespace="http://datacite.org/schema/kernel-4" elementFormDefault="qualified">
+  <xs:simpleType name="descriptionType" id="descriptionType">
+    <xs:annotation>
+      <xs:documentation>The type of the description.</xs:documentation>
+    </xs:annotation>
+    <xs:restriction base="xs:string">
+      <xs:enumeration value="Abstract"/>
+      <xs:enumeration value="Methods"/>
+      <xs:enumeration value="SeriesInformation"/>
+      <xs:enumeration value="TableOfContents"/>
+      <xs:enumeration value="TechnicalInfo"/>
+      <xs:enumeration value="Other"/>
+    </xs:restriction>
+  </xs:simpleType>
+</xs:schema>