commonmeta-ruby 3.3.8 → 3.3.9

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e717e42ba9e8264cfc3ebb4bcb774c58b7f003afad4ffea7da385cd66e84eb27
-  data.tar.gz: bee1fe28c91acb00e3b5afae8b4305d6036dd4924f554b44d49145ac388fcf69
+  metadata.gz: 39731365147aea3875a44db36625a0cd825826f7f0f94755601634b2716fcef9
+  data.tar.gz: 7591e80c87796e20933352f49079437a9faac148e47889009a5ee3298471ee4c
 SHA512:
-  metadata.gz: 1cc77541969f79992f5e8b6f0677939e64397ffbedfbb75c40d2d534c2a64ad2f8a29b86df1570d76cdd0245f2f80a01e422052c065ca5e45085b85fb9dda7ec
-  data.tar.gz: 1389f97cf9c919eda5a384a247b7cddb142b0bfb50ce4c7bb77d83f8c1da6812a20c3277a1a0860d42b4952c63c8862d29c6061e0c11cf2a1436ea0424cd3fc9
+  metadata.gz: c9c3593d73cadfd1697163801d0505a74c536a93e2a40c2bcaac5e12e7c6e703970a4a207f9282331fe065257ec26856db4e214342926019334e64229b920059
+  data.tar.gz: 69a053002420f6aa808b15c6a77344fa02e7a7c26e208a00b7be68d489540356beb47481e8f7565e102905233b51830a1dbd591a4a6435d5bc4c646a3f632d01

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    commonmeta-ruby (3.3.8)
+    commonmeta-ruby (3.3.9)
       activesupport (>= 4.2.5, < 8.0)
       addressable (~> 2.8.1, < 2.8.2)
       base32-url (>= 0.7.0, < 1)
@@ -27,7 +27,7 @@ PATH
 GEM
   remote: https://rubygems.org/
   specs:
-    activesupport (7.0.5)
+    activesupport (7.0.5.1)
       concurrent-ruby (~> 1.0, >= 1.0.2)
       i18n (>= 1.6, < 2)
       minitest (>= 5.1)
@@ -113,6 +113,7 @@ GEM
       oj (~> 3)
       optimist (~> 3)
     jwt (2.7.1)
+    language_server-protocol (3.17.0.3)
     latex-decode (0.4.0)
     link_header (0.0.8)
     llhttp-ffi (0.4.0)
@@ -186,12 +187,13 @@ GEM
     rspec-mocks (3.12.5)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.12.0)
-    rspec-support (3.12.0)
+    rspec-support (3.12.1)
     rspec-xsd (0.1.0)
       nokogiri (~> 1.6)
       rspec (~> 3)
-    rubocop (1.52.1)
+    rubocop (1.53.1)
       json (~> 2.3)
+      language_server-protocol (>= 3.17.0)
       parallel (~> 1.10)
       parser (>= 3.2.2.3)
       rainbow (>= 2.2.2, < 4.0)
@@ -245,7 +247,7 @@ GEM
     unicode-display_width (2.4.2)
     unicode-types (1.8.0)
     uuidtools (2.2.0)
-    vcr (6.1.0)
+    vcr (6.2.0)
     webmock (3.18.1)
       addressable (>= 2.8.0)
       crack (>= 0.3.2)

data/lib/commonmeta/author_utils.rb

@@ -63,6 +63,8 @@ module Commonmeta
         type = 'Person'
       elsif type.blank? && is_personal_name?(name: name) && name.to_s.exclude?(';')
         type = 'Person'
+      elsif type.blank? && name.present? && !is_personal_name?(name: name)
+        type = 'Organization'
       end
 
       # parse author contributor role
@@ -127,8 +129,8 @@ module Commonmeta
     def is_personal_name?(name: nil)
       return true if name_exists?(name.to_s.split.first) || name_exists?(name.to_s.split(', ').last)
 
-      # check if a name has only one word, e.g. "FamousOrganization"
-      return false if name.to_s.split(' ').size == 1
+      # 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 for suffixes, e.g. "John Smith, MD"
       return true if %w[MD PhD].include? name.split(', ').last

data/lib/commonmeta/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Commonmeta
-  VERSION = '3.3.8'
+  VERSION = '3.3.9'
 end

data/spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/ghost_post_with_institutional_author.yml

@@ -0,0 +1,56 @@
+---
+http_interactions:
+- request:
+    method: get
+    uri: https://rogue-scholar.org/api/posts/2b3cdd27-5123-4167-9482-3c074392e2d2
+    body:
+      encoding: UTF-8
+      string: ''
+    headers:
+      Connection:
+      - close
+      Host:
+      - rogue-scholar.org
+      User-Agent:
+      - http.rb/5.1.1
+  response:
+    status:
+      code: 200
+      message: OK
+    headers:
+      Age:
+      - '0'
+      Cache-Control:
+      - public, max-age=0, must-revalidate
+      Content-Length:
+      - '1509'
+      Content-Type:
+      - application/json; charset=utf-8
+      Date:
+      - Wed, 28 Jun 2023 07:30:24 GMT
+      Etag:
+      - '"8258swzkwh15x"'
+      Server:
+      - Vercel
+      Strict-Transport-Security:
+      - max-age=63072000
+      X-Matched-Path:
+      - "/api/posts/[slug]"
+      X-Vercel-Cache:
+      - MISS
+      X-Vercel-Id:
+      - fra1::iad1::wc777-1687937422070-9bb853322252
+      Connection:
+      - close
+    body:
+      encoding: UTF-8
+      string: '{"id":"63cef642602205003d6f50fb","uuid":"2b3cdd27-5123-4167-9482-3c074392e2d2","url":"https://blog.oa.works/nature-features-oa-reports-work-putting-oa-policy-into-practice/","title":"Nature
+        features OA.Report''s work putting OA policy into practice!","summary":"After
+        a couple of years of working to support institutions implementing their OA
+        policies, we''re so proud to see OA.Report, our development partner, and one
+        of our collaborators featured in Nature. We appreciate Nature bringing attention
+        to the progress being made in more effectively implementing OA policies.","date_published":"2023-01-24T11:11:47Z","date_modified":null,"date_indexed":"2023-06-20T06:34:19.365+00:00","authors":[{"url":null,"name":"OA.Works"}],"image":"https://blog.oa.works/content/images/2023/01/nature-website-v2.png","tags":["OA.Report"],"language":"en","references":[],"blog_id":"h7bpg11","blog":{"id":"h7bpg11","title":"OA.Works
+        Blog","description":"OA.Works Blog","language":"en","favicon":"https://blog.oa.works/favicon.png","feed_url":"https://blog.oa.works/rss/","home_page_url":"https://blog.oa.works/","user_id":"8498eaf6-8c58-4b58-bc15-27eda292b1aa","created_at":"2023-05-31T14:11:14+00:00","indexed_at":"2023-01-04","feed_format":"application/rss+xml","license":"https://creativecommons.org/licenses/by/4.0/legalcode","generator":"Ghost
+        5.53","category":"Engineering and Technology","prefix":null,"modified_at":"2023-03-23T12:24:58+00:00","version":"https://jsonfeed.org/version/1.1","backlog":true,"authors":null}}'
+  recorded_at: Wed, 28 Jun 2023 07:30:24 GMT
+recorded_with: VCR 6.1.0

data/spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/wordpress_post_with_tracking_code_on_url.yml

@@ -0,0 +1,60 @@
+---
+http_interactions:
+- request:
+    method: get
+    uri: https://rogue-scholar.org/api/posts/5d95d90d-ff59-4c8b-b7f8-44e6b45fd593
+    body:
+      encoding: UTF-8
+      string: ''
+    headers:
+      Connection:
+      - close
+      Host:
+      - rogue-scholar.org
+      User-Agent:
+      - http.rb/5.1.1
+  response:
+    status:
+      code: 200
+      message: OK
+    headers:
+      Age:
+      - '0'
+      Cache-Control:
+      - public, max-age=0, must-revalidate
+      Content-Length:
+      - '1844'
+      Content-Type:
+      - application/json; charset=utf-8
+      Date:
+      - Wed, 28 Jun 2023 07:39:01 GMT
+      Etag:
+      - '"6emrve7vie1ew"'
+      Server:
+      - Vercel
+      Strict-Transport-Security:
+      - max-age=63072000
+      X-Matched-Path:
+      - "/api/posts/[slug]"
+      X-Vercel-Cache:
+      - MISS
+      X-Vercel-Id:
+      - fra1::iad1::q9g2k-1687937937528-5f071f224fc7
+      Connection:
+      - close
+    body:
+      encoding: UTF-8
+      string: '{"id":"https://doi.org/10.59350/tpa8t-j6292","uuid":"5d95d90d-ff59-4c8b-b7f8-44e6b45fd593","url":"https://www.samuelmoore.org/2023/04/20/how-to-cultivate-good-closures-scaling-small-and-the-limits-of-openness","title":"How
+        to cultivate good closures: ‘scaling small’ and the limits of openness","summary":"Text
+        of a talk given to the COPIM end-of-project conference: “Scaling Small: Community-Owned
+        Futures for Open Access Books”, April 20th 2023 Open access publishing has
+        always had a difficult relationship with smoothness and scale. Openness implies
+        seamlessness, limitlessness or structureless-ness – or the idea that the removal
+        of price and permission barriers is what’s needed to allow research to reach
+        its full potential. The drive for seamlessness is on display in much of the
+        push for...","date_published":"2023-04-20T15:37:20Z","date_modified":"2023-06-19T11:01:51Z","date_indexed":"2023-06-20T06:34:19+00:00","authors":[{"url":"https://orcid.org/0000-0002-0504-2897","name":"Samuel
+        Moore"}],"image":null,"tags":["Open access"],"language":"en","references":null,"blog_id":"gr1by89","blog":{"id":"gr1by89","title":"Samuel
+        Moore","description":"publishing, technology, commons","language":"en","favicon":"https://www.samuelmoore.org/wp-content/uploads/2019/02/cropped-cropped-cropped-roAorange-icon-2-32x32.png","feed_url":"https://www.samuelmoore.org/feed/atom/","home_page_url":"https://www.samuelmoore.org/","user_id":"8498eaf6-8c58-4b58-bc15-27eda292b1aa","created_at":"2023-05-31T14:25:06+00:00","indexed_at":"2023-01-05","feed_format":"application/atom+xml","license":"https://creativecommons.org/licenses/by/4.0/legalcode","generator":"WordPress
+        6.2.2","category":"Social Sciences","prefix":null,"modified_at":"2023-04-20T15:37:20+00:00","version":"https://jsonfeed.org/version/1.1","backlog":true,"authors":null}}'
+  recorded_at: Wed, 28 Jun 2023 07:39:01 GMT
+recorded_with: VCR 6.1.0

data/spec/readers/json_feed_reader_spec.rb

@@ -174,6 +174,54 @@ describe Commonmeta::Metadata, vcr: true do
       expect(subject.references.length).to eq(3)
       expect(subject.references.first).to eq("key" => "ref1", "url" => "https://sauroposeidon.files.wordpress.com/2010/04/foster-and-wedel-2014-haplocanthosaurus-from-snowmass-colorado.pdf")
     end
+    
+    it "wordpress post with tracking code on url" do
+      input = "https://rogue-scholar.org/api/posts/5d95d90d-ff59-4c8b-b7f8-44e6b45fd593"
+      subject = described_class.new(input: input)
+      puts subject.errors
+      # expect(subject.valid?).to be true
+      expect(subject.id).to eq("https://doi.org/10.59350/tpa8t-j6292")
+      expect(subject.url).to eq("https://www.samuelmoore.org/2023/04/20/how-to-cultivate-good-closures-scaling-small-and-the-limits-of-openness")
+      expect(subject.type).to eq("Article")
+      expect(subject.creators.length).to eq(1)
+      expect(subject.creators.first).to eq("familyName"=>"Moore", "givenName"=>"Samuel", "id"=>"https://orcid.org/0000-0002-0504-2897", "type"=>"Person")
+      expect(subject.titles).to eq([{ "title" => "How to cultivate good closures: ‘scaling small’ and the limits of openness" }])
+      expect(subject.license).to eq("id" => "CC-BY-4.0",
+                                    "url" => "https://creativecommons.org/licenses/by/4.0/legalcode")
+      expect(subject.date).to eq("published"=>"2023-04-20", "updated"=>"2023-06-19")
+      expect(subject.descriptions.first["description"]).to start_with("Text of a talk given to the COPIM end-of-project conference")
+      expect(subject.publisher).to eq("name" => "Samuel Moore")
+      expect(subject.subjects).to eq([{ "subject" => "Social sciences" },
+                                      { "schemeUri" => "http://www.oecd.org/science/inno/38235147.pdf",
+                                        "subject" => "FOS: Social sciences",
+                                        "subjectScheme" => "Fields of Science and Technology (FOS)" }])
+      expect(subject.language).to eq("en")
+      expect(subject.container).to eq("identifier" => "https://www.samuelmoore.org/", "identifierType" => "URL", "title" => "Samuel Moore", "type" => "Periodical")
+      expect(subject.references).to be_nil
+    end
+
+    it "ghost post with institutional author" do
+      input = "https://rogue-scholar.org/api/posts/2b3cdd27-5123-4167-9482-3c074392e2d2"
+      subject = described_class.new(input: input)
+      expect(subject.valid?).to be true
+      expect(subject.id).to eq("https://blog.oa.works/nature-features-oa-reports-work-putting-oa-policy-into-practice")
+      expect(subject.url).to eq("https://blog.oa.works/nature-features-oa-reports-work-putting-oa-policy-into-practice")
+      expect(subject.type).to eq("Article")
+      expect(subject.creators.length).to eq(1)
+      expect(subject.creators.first).to eq("name"=>"OA.Works", "type"=>"Organization")
+      expect(subject.titles).to eq([{ "title" => "Nature features OA.Report's work putting OA policy into practice!" }])
+      expect(subject.license).to eq("id" => "CC-BY-4.0",
+                                    "url" => "https://creativecommons.org/licenses/by/4.0/legalcode")
+      expect(subject.date).to eq("published" => "2023-01-24")
+      expect(subject.descriptions.first["description"]).to start_with("After a couple of years of working to support institutions implementing their OA policies")
+      expect(subject.publisher).to eq("name" => "OA.Works Blog")
+      expect(subject.subjects).to eq([{ "subject" => "Engineering and technology" },
+                                      { "schemeUri" => "http://www.oecd.org/science/inno/38235147.pdf",
+                                        "subject" => "FOS: Engineering and technology",
+                                        "subjectScheme" => "Fields of Science and Technology (FOS)" }])
+      expect(subject.language).to eq("en")
+      expect(subject.container).to eq("identifier" => "https://blog.oa.works/", "identifierType" => "URL", "title" => "OA.Works Blog", "type" => "Periodical")
+    end
 
     it "upstream post with references" do
       input = "https://rogue-scholar.org/api/posts/954f8138-0ecd-4090-87c5-cef1297f1470"

data/spec/readers/ris_reader_spec.rb

@@ -50,7 +50,7 @@ describe Commonmeta::Metadata, vcr: true do
       expect(subject.state).to eq('not_found')
       expect(subject.id).to eq('https://doi.org/10.7554/elife.01567')
       expect(subject.type).to eq('Dissertation')
-      expect(subject.creators).to eq([{ 'name' => 'Toparlar,Y.' }])
+      expect(subject.creators).to eq([{"familyName"=>"Toparlar", "givenName"=>"Y.", "type"=>"Person"}])
       expect(subject.titles).to eq([{ 'title' => 'A multiscale analysis of the urban heat island effect' }])
       expect(subject.descriptions.first['description']).to start_with('Designing the climates of cities')
       expect(subject.date).to eq('created' => '2018-04-25', 'published' => '2018-04-25')

data/spec/readers/schema_org_reader_spec.rb

@@ -153,9 +153,7 @@ describe Commonmeta::Metadata, vcr: true do
       expect(subject.titles).to eq([{ 'title' => 'Summary data ankylosing spondylitis GWAS' }])
       expect(subject.container).to eq('identifier' => 'https://dataverse.harvard.edu',
                                       'identifierType' => 'URL', 'title' => 'Harvard Dataverse', 'type' => 'DataRepository')
-      expect(subject.creators).to eq([{
-                                       'name' => 'International Genetics of Ankylosing Spondylitis Consortium (IGAS)'
-                                     }])
+      expect(subject.creators).to eq([{"name"=>"International Genetics of Ankylosing Spondylitis Consortium (IGAS)", "type"=>"Organization"}])
       expect(subject.subjects).to eq([{ 'subject' => 'Medicine, health and life sciences' },
                                       { 'subject' => 'genome-wide association studies' },
                                       { 'subject' => 'Ankylosing spondylitis' }])

data/spec/writers/schema_org_writer_spec.rb

@@ -177,7 +177,7 @@ describe Commonmeta::Metadata, vcr: true do
       expect(json['@id']).to eq('https://doi.org/10.23725/8na3-9s47')
       expect(json['@type']).to eq('Dataset')
       expect(json['name']).to eq('NWD165827.recab.cram')
-      expect(json['author']).to eq('name' => 'TOPMed')
+      expect(json['author']).to eq("@type"=>"Organization", "name"=>"TOPMed")
       expect(json['includedInDataCatalog']).to be_empty
       expect(json['identifier']).to eq(
         [{ '@type' => 'PropertyValue',

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: commonmeta-ruby
 version: !ruby/object:Gem::Version
-  version: 3.3.8
+  version: 3.3.9
 platform: ruby
 authors:
 - Martin Fenner
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2023-06-21 00:00:00.000000000 Z
+date: 2023-06-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activesupport
@@ -870,6 +870,7 @@ files:
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/blogger_post.yml
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/ghost_post_with_author_name_suffix.yml
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/ghost_post_with_doi.yml
+- spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/ghost_post_with_institutional_author.yml
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/ghost_post_with_organizational_author.yml
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/ghost_post_without_doi.yml
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/jekyll_post.yml
@@ -879,6 +880,7 @@ files:
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/upstream_post_with_references.yml
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/wordpress_post.yml
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/wordpress_post_with_references.yml
+- spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_json_feed_item_metadata/wordpress_post_with_tracking_code_on_url.yml
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_one_author/affiliation_is_space.yml
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_one_author/has_familyName.yml
 - spec/fixtures/vcr_cassettes/Commonmeta_Metadata/get_one_author/has_name_in_display-order_with_ORCID.yml