commonmeta-ruby 3.0.10 → 3.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: dbd0cff9167f3cf4cbef16d1b2f344dfbaaaa0c5f121f385f8e2703fedd42b0d
-  data.tar.gz: 601f92cc97809057378e7dd89e36d1918a18280201b25a25ad8a13ecd1412a04
+  metadata.gz: 6489e2d1d9225e98fbf809c51986da2d9959b25dadc51ab5637a8fa43181e743
+  data.tar.gz: 0b4d31c968f52baf74848fc51290123db1645503f55840172c01aacb43ec5753
 SHA512:
-  metadata.gz: d0e2953f19c0364ad7f774fe38be92bf9df6b975c02568e980ac4cfb7a0fae590aeda39df9d3c80332701db4cc7bfd81648c987b1e96eed6cd1595baf758376d
-  data.tar.gz: 76b133fe3e101ba09c62c362d41035fd323e262575338866bacdbd2e07544b4b1e63583c304c827f9c386c22c7ea45c7f2b0c6eb304cf060b10f8e32e77e3796
+  metadata.gz: c6224cacab63d37765875e6e23fe1c70b7443e603b895866d2b192fb5d823d7d98d8613eae03e56debf9fea1a02f7a13f45a4ed5b0d2f37990368a6fe18c8182
+  data.tar.gz: c07f1ff76a51a313a96a881077b1a3f5ee918a3cf7012c5edbdd8dce24381d856994b840b76563185fc220ac3f3c75c5edf4e06c6c69524ab48b526b63c1d5d8

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    commonmeta-ruby (3.0.10)
+    commonmeta-ruby (3.2.1)
       activesupport (>= 4.2.5, < 8.0)
       addressable (~> 2.8.1, < 2.8.2)
       base32-url (>= 0.5.0, < 1)
@@ -13,7 +13,7 @@ PATH
       gender_detector (~> 2.0)
       http (~> 5.1, >= 5.1.1)
       json-ld-preloaded (~> 3.2, >= 3.2.2)
-      json_schemer (~> 0.2.23)
+      json_schemer (~> 1.0.1)
       jsonlint (~> 0.3.0)
       loofah (~> 2.19)
       namae (~> 1.0)
@@ -25,7 +25,7 @@ PATH
 GEM
   remote: https://rubygems.org/
   specs:
-    activesupport (7.0.4.3)
+    activesupport (7.0.5)
       concurrent-ruby (~> 1.0, >= 1.0.2)
       i18n (>= 1.6, < 2)
       minitest (>= 5.1)
@@ -57,14 +57,12 @@ GEM
     docile (1.4.0)
     domain_name (0.5.20190701)
       unf (>= 0.0.5, < 1.0.0)
-    ebnf (2.3.2)
+    ebnf (2.3.3)
       htmlentities (~> 4.3)
       rdf (~> 3.2)
       scanf (~> 1.0)
       sxp (~> 1.2)
       unicode-types (~> 1.8)
-    ecma-re-validator (0.4.0)
-      regexp_parser (~> 2.2)
     edtf (3.1.1)
       activesupport (>= 3.0, < 8.0)
     ffi (1.15.5)
@@ -86,25 +84,24 @@ GEM
     http-cookie (1.0.5)
       domain_name (~> 0.5)
     http-form_data (2.3.0)
-    i18n (1.12.0)
+    i18n (1.14.0)
       concurrent-ruby (~> 1.0)
     json (2.6.3)
-    json-canonicalization (0.3.1)
-    json-ld (3.2.3)
+    json-canonicalization (0.3.2)
+    json-ld (3.2.5)
       htmlentities (~> 4.3)
-      json-canonicalization (~> 0.3)
+      json-canonicalization (~> 0.3, >= 0.3.2)
       link_header (~> 0.0, >= 0.0.8)
       multi_json (~> 1.15)
-      rack (~> 2.2)
-      rdf (~> 3.2, >= 3.2.9)
+      rack (>= 2.2, < 4)
+      rdf (~> 3.2, >= 3.2.10)
     json-ld-preloaded (3.2.2)
       json-ld (~> 3.2)
       rdf (~> 3.2)
-    json_schemer (0.2.24)
-      ecma-re-validator (~> 0.3)
+    json_schemer (1.0.1)
       hana (~> 1.3)
       regexp_parser (~> 2.0)
-      uri_template (~> 0.7)
+      simpleidn (~> 0.2)
     jsonlint (0.3.0)
       oj (~> 3)
       optimist (~> 3)
@@ -113,19 +110,19 @@ GEM
     llhttp-ffi (0.4.0)
       ffi-compiler (~> 1.0)
       rake (~> 13.0)
-    loofah (2.20.0)
+    loofah (2.21.3)
       crass (~> 1.0.2)
-      nokogiri (>= 1.5.9)
+      nokogiri (>= 1.12.0)
     matrix (0.4.2)
     minitest (5.18.0)
     multi_json (1.15.0)
     namae (1.1.1)
-    nokogiri (1.14.3-arm64-darwin)
+    nokogiri (1.15.2-arm64-darwin)
       racc (~> 1.4)
-    oj (3.14.3)
+    oj (3.15.0)
     optimist (3.0.1)
-    parallel (1.22.1)
-    parser (3.2.2.0)
+    parallel (1.23.0)
+    parser (3.2.2.1)
       ast (~> 2.4.1)
     postrank-uri (1.1)
       addressable (>= 2.4.0)
@@ -133,12 +130,12 @@ GEM
       public_suffix (>= 4.0.0, < 5)
     public_suffix (4.0.7)
     racc (1.6.2)
-    rack (2.2.6.4)
+    rack (3.0.7)
     rack-test (2.1.0)
       rack (>= 1.3)
     rainbow (3.1.1)
     rake (13.0.6)
-    rdf (3.2.9)
+    rdf (3.2.10)
       link_header (~> 0.0, >= 0.0.8)
     rdf-aggregate-repo (3.2.1)
       rdf (~> 3.2)
@@ -158,20 +155,20 @@ GEM
     rdf-turtle (3.2.1)
       ebnf (~> 2.3)
       rdf (~> 3.2)
-    rdf-vocab (3.2.5)
+    rdf-vocab (3.2.6)
       rdf (~> 3.2, >= 3.2.4)
     rdf-xsd (3.2.1)
       rdf (~> 3.2)
       rexml (~> 3.2)
-    regexp_parser (2.7.0)
+    regexp_parser (2.8.0)
     rexml (3.2.5)
     rspec (3.12.0)
       rspec-core (~> 3.12.0)
       rspec-expectations (~> 3.12.0)
       rspec-mocks (~> 3.12.0)
-    rspec-core (3.12.1)
+    rspec-core (3.12.2)
       rspec-support (~> 3.12.0)
-    rspec-expectations (3.12.2)
+    rspec-expectations (3.12.3)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.12.0)
     rspec-mocks (3.12.5)
@@ -181,7 +178,7 @@ GEM
     rspec-xsd (0.1.0)
       nokogiri (~> 1.6)
       rspec (~> 3)
-    rubocop (1.50.1)
+    rubocop (1.52.0)
       json (~> 2.3)
       parallel (~> 1.10)
       parser (>= 3.2.0.0)
@@ -191,18 +188,21 @@ GEM
       rubocop-ast (>= 1.28.0, < 2.0)
       ruby-progressbar (~> 1.7)
       unicode-display_width (>= 2.4.0, < 3.0)
-    rubocop-ast (1.28.0)
+    rubocop-ast (1.29.0)
       parser (>= 3.2.1.0)
-    rubocop-capybara (2.17.1)
+    rubocop-capybara (2.18.0)
       rubocop (~> 1.41)
-    rubocop-performance (1.17.1)
+    rubocop-factory_bot (2.23.1)
+      rubocop (~> 1.33)
+    rubocop-performance (1.18.0)
       rubocop (>= 1.7.0, < 2.0)
       rubocop-ast (>= 0.4.0)
     rubocop-rake (0.6.0)
       rubocop (~> 1.0)
-    rubocop-rspec (2.19.0)
+    rubocop-rspec (2.22.0)
       rubocop (~> 1.33)
       rubocop-capybara (~> 2.17)
+      rubocop-factory_bot (~> 2.22)
     ruby-progressbar (1.13.0)
     scanf (1.0.0)
     simplecov (0.22.0)
@@ -211,11 +211,13 @@ GEM
       simplecov_json_formatter (~> 0.1)
     simplecov-html (0.12.3)
     simplecov_json_formatter (0.1.4)
-    sxp (1.2.3)
+    simpleidn (0.2.1)
+      unf (~> 0.1.4)
+    sxp (1.2.4)
       matrix (~> 0.4)
       rdf (~> 3.2)
-    temple (0.10.0)
-    thor (1.2.1)
+    temple (0.10.2)
+    thor (1.2.2)
     tilt (2.1.0)
     tzinfo (2.0.6)
       concurrent-ruby (~> 1.0)
@@ -224,7 +226,6 @@ GEM
     unf_ext (0.0.8.2)
     unicode-display_width (2.4.2)
     unicode-types (1.8.0)
-    uri_template (0.7.0)
     vcr (6.1.0)
     webmock (3.18.1)
       addressable (>= 2.8.0)

data/bin/commonmeta

@@ -2,7 +2,7 @@
 
 require File.expand_path("../../lib/commonmeta", __FILE__)
 
-if (ARGV & %w(--version -v help --help encode decode encode_id decode_id)).empty?
+if (ARGV & %w(--version -v help --help encode decode encode_id decode_id json_feed)).empty?
   Commonmeta::CLI.start(ARGV.dup.unshift("convert"))
 else
   Commonmeta::CLI.start

data/commonmeta.gemspec

@@ -30,7 +30,7 @@ Gem::Specification.new do |s|
   s.add_dependency 'http', '~> 5.1', '>= 5.1.1'
   s.add_dependency 'json-ld-preloaded', '~> 3.2', '>= 3.2.2'
   s.add_dependency 'jsonlint', '~> 0.3.0'
-  s.add_dependency 'json_schemer', '~> 0.2.23'
+  s.add_dependency 'json_schemer', '~> 1.0.1'
   s.add_dependency 'loofah', '~> 2.19'
   s.add_dependency 'namae', '~> 1.0'
   s.add_dependency 'postrank-uri', '~> 1.1'

data/lib/commonmeta/cli.rb

@@ -9,6 +9,7 @@ module Commonmeta
   class CLI < Thor
     include Commonmeta::DoiUtils
     include Commonmeta::Utils
+    include Commonmeta::Readers::JsonFeedReader
 
     def self.exit_on_failure?
       true
@@ -79,6 +80,12 @@ module Commonmeta
       puts decode_container_id(id)
     end
 
+    desc '', 'feed'
+
+    def json_feed(id)
+      puts get_json_feed(id)
+    end
+
     default_task :convert
   end
 end

data/lib/commonmeta/metadata_utils.rb

@@ -13,6 +13,7 @@ require_relative 'readers/codemeta_reader'
 require_relative 'readers/crossref_reader'
 require_relative 'readers/crossref_xml_reader'
 require_relative 'readers/datacite_reader'
+require_relative 'readers/json_feed_reader'
 require_relative 'readers/npm_reader'
 require_relative 'readers/ris_reader'
 require_relative 'readers/schema_org_reader'
@@ -46,6 +47,7 @@ module Commonmeta
     include Commonmeta::Readers::CrossrefXmlReader
     include Commonmeta::Readers::CslReader
     include Commonmeta::Readers::DataciteReader
+    include Commonmeta::Readers::JsonFeedReader
     include Commonmeta::Readers::NpmReader
     include Commonmeta::Readers::RisReader
     include Commonmeta::Readers::SchemaOrgReader

data/lib/commonmeta/readers/json_feed_reader.rb

@@ -0,0 +1,90 @@
+# frozen_string_literal: true
+
+module Commonmeta
+  module Readers
+    module JsonFeedReader
+      def get_json_feed_item(id: nil, **_options)
+        return { "string" => nil, "state" => "not_found" } unless id.present?
+
+        url = normalize_id(id)
+        response = HTTP.get(url)
+        return { "string" => nil, "state" => "not_found" } unless response.status.success?
+  
+        { "string" => response.body.to_s }
+      end
+
+      def read_json_feed_item(string: nil, **options)
+        read_options = ActiveSupport::HashWithIndifferentAccess.new(options.except(:doi, :id, :url,
+                                                                                   :sandbox, :validate, :ra))
+
+        meta = string.present? ? JSON.parse(string) : {}
+
+        id = options[:doi] ? normalize_doi(options[:doi]) : normalize_id(meta.fetch("id", nil))
+        url = normalize_url(meta.fetch("url", nil))
+        type = "Article"
+        creators = if meta.fetch("authors", nil).present?
+            get_authors(from_json_feed(Array.wrap(meta.fetch("authors"))))
+          else
+            [{ "type" => "Organization", "name" => ":(unav)" }]
+          end
+        titles = [{ "title" => meta.fetch("title", nil) }]
+        publisher = { "name" => meta.dig("blog", "title") }
+
+        date = {}
+        date["published"] = get_iso8601_date(meta.dig("date_published")) if meta.dig("date_published").present?
+        date["updated"] = get_iso8601_date(meta.dig("date_modified")) if meta.dig("date_modified").present?
+
+        license = if meta.dig("blog", "license").present?
+            hsh_to_spdx("rightsURI" => meta.dig("blog", "license"))
+          end
+        home_page_url = normalize_url(meta.dig("blog", "home_page_url"))
+        container = if meta.dig("blog", "title").present?
+            { "type" => "Periodical",
+              "title" => meta.dig("blog", "title"),
+              "identifier" => home_page_url,
+              "identifierType" => "URL" }
+          end
+
+        descriptions = if meta.fetch("summary", nil).present?
+            [{ "description" => sanitize(meta.fetch("summary", nil)),
+               "descriptionType" => "Abstract" }]
+          else
+            []
+          end
+        language = meta.fetch("language", nil) || meta.fetch("blog", "language", nil)
+        state = id.present? || read_options.present? ? "findable" : "not_found"
+        subjects = Array.wrap(meta.fetch("tags", nil)).reduce([]) do |sum, subject|
+          sum += name_to_fos(subject)
+
+          sum
+        end
+
+        { "id" => id,
+          "type" => type,
+          "url" => url,
+          "titles" => titles,
+          "creators" => creators,
+          "publisher" => publisher,
+          "container" => container,
+          "date" => date,
+          "language" => language,
+          "descriptions" => descriptions,
+          "license" => license,
+          "subjects" => subjects.presence,
+          "state" => state }.compact.merge(read_options)
+      end
+
+      def get_json_feed(id)
+        # get JSON Feed items not registered as DOIs
+        return { "string" => nil, "state" => "not_found" } unless id.present?
+
+        url = json_feed_url(id)
+        response = HTTP.get(url)
+        return { "string" => nil, "state" => "not_found" } unless response.status.success?
+
+        blog = JSON.parse(response.body.to_s)
+        blog["items"].select { |item| !validate_doi(item["id"]) }.map { |item| item["short_id"] }.join(",").presence
+      end
+    end
+  end
+end

data/lib/commonmeta/utils.rb

@@ -468,6 +468,8 @@ module Commonmeta
         "cff"
       elsif %r{\A(http|https):/(/)?github\.com/(.+)\z}.match?(id)
         "cff"
+      elsif %r{\A(http|https):/(/)?rogue-scholar\.org/api/posts/(.+)\z}.match?(id)
+        "json_feed_item"
       else
         "schema_org"
       end
@@ -875,6 +877,12 @@ module Commonmeta
       }
     end
 
+    def from_json_feed(element)
+      mapping = { "url" => "id" }
+
+      map_hash_keys(element: element, mapping: mapping)
+    end
+
     def from_csl(element)
       Array.wrap(element).map do |a|
         if a["literal"].present?
@@ -1068,6 +1076,8 @@ module Commonmeta
 
       return iso8601_time.split(".").first + "Z" if iso8601_time.to_s.include? "."
 
+      return iso8601_time.split("+").first + "Z" if iso8601_time.to_s.include? "+"
+
       iso8601_time
     end
 
@@ -1084,6 +1094,15 @@ module Commonmeta
       dd.fetch("date", nil)
     end
 
+    def get_link(links, link_type)
+      ll = Array.wrap(links).find { |d| d["rel"] == link_type } || {}
+      ll.fetch("href", nil)
+    end
+
+    def rogue_scholar_api_url(id, _options = {})
+      "https://rogue-scholar.org/api/posts/#{id}"
+    end
+
     # convert commonmeta dates to DataCite format
     def get_dates_from_date(date)
       return nil if date.nil?
@@ -1361,5 +1380,9 @@ module Commonmeta
     def decode_container_id(id)
       Base32::URL.decode(id)
     end
+
+    def json_feed_url(id)
+      "https://rogue-scholar.org/api/blogs/#{id}"
+    end
   end
 end

data/lib/commonmeta/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Commonmeta
-  VERSION = '3.0.10'
+  VERSION = '3.2.1'
 end

data/spec/author_utils_spec.rb

@@ -40,6 +40,16 @@ describe Commonmeta::Metadata, vcr: true do
       expect(subject.is_personal_name?(name: author['name'])).to be true
     end
 
+    it 'has unknown given name' do
+      author = { 'name' => 'Rintze Zelle' }
+      expect(subject.is_personal_name?(name: author['name'])).to be false
+    end
+
+    it 'has middle initial' do
+      author = { 'name' => 'Martin H. Fenner' }
+      expect(subject.is_personal_name?(name: author['name'])).to be true
+    end
+
     it 'has no info' do
       author = { 'name' => 'M Fenner' }
       expect(subject.is_personal_name?(name: author['name'])).to be false

data/spec/cli_spec.rb

@@ -340,4 +340,12 @@ describe Commonmeta::CLI do
       expect { subject.decode_id input }.to output(/18397685862\n/).to_stdout
     end
   end
+
+  describe "json_feed", vcr: true do
+    let(:input) { "prmb582" }
+
+    it "blog post" do
+      expect { subject.json_feed input }.to output(/1jdkoe52,4qe9ypg1,1jdkwod5,56glp9d9,1xdnk0d0,56glmmd9,1xdnoje0,31epl0dn,q0dqm6go\n/).to_stdout
+    end
+  end
 end