Merge pull request #162 from MITLibraries/tco-110-journals

Split Detector::Journals model into two, extending BulkChecker into the detection model

app/models/detector/journal.rb

@@ -1,23 +1,16 @@
 # frozen_string_literal: true
 
-# == Schema Information
-#
-# Table name: detector_journals
-#
-#  id              :integer          not null, primary key
-#  name            :string
-#  additional_info :json
-#  created_at      :datetime         not null
-#  updated_at      :datetime         not null
-#
 class Detector
-  # Detector::Journal stores information about academic journals loaded from external sources to allow us to check our
-  # incoming Terms against these information
-  class Journal < ApplicationRecord
-    before_save :downcase_fields!
+  # Detector::Journal handles the comparison between incoming Term records and our known list of academic journals
+  # (which are managed by the separate Journal model).
+  class Journal
+    attr_reader :detections
 
-    def self.table_name_prefix
-      'detector_'
+    # shared singleton methods
+    extend Detector::BulkChecker
+
+    def initialize(phrase)
+      @detections = Detector::Journal.full_term_match(phrase)
     end
 
     # Identify journals in which the incoming phrase matches a Journal.name exactly
@@ -30,9 +23,9 @@ def self.table_name_prefix
     #
     # @param phrase [String]. A string representation of a search term (not an actual Term object!)
     #
-    # @return [Set of Detector::Journal] A set of ActiveRecord Detector::Journal relations.
+    # @return [Set of Journal] A set of ActiveRecord Journal records.
     def self.full_term_match(phrase)
-      Journal.where(name: phrase.downcase)
+      ::Journal.where(name: phrase.downcase)
     end
 
     # Identify journals in which the incoming phrase contains one or more Journal names
@@ -41,12 +34,12 @@ def self.full_term_match(phrase)
     #
     # @param phrase [String]. A string representation of a search term (not an actual Term object!)
     #
-    # @return [Set of Detector::Journal] A set of ActiveRecord Detector::Journal relations.
+    # @return [Set of Journal] A set of ActiveRecord Journal records.
     def self.partial_term_match(phrase)
-      Journal.all.select { |journal| phrase.downcase.include?(journal.name) }
+      ::Journal.all.select { |journal| phrase.downcase.include?(journal.name) }
     end
 
-    # Look up any matching Detector::Journal records, building on the full_term_match method. If a match is found, a
+    # Look up any matching Journal records, building on the full_term_match method. If a match is found, a
     # Detection record is created to indicate this success.
     #
     # @note This does not care whether multiple matching journals are detected. If _any_ match is found, a Detection
@@ -65,13 +58,5 @@ def self.record(term)
 
       nil
     end
-
-    private
-
-    # Downcasing all names before saving allows for more efficient matching by ensuring our index is lowercase.
-    # If we find we need the non-lowercase Journal name in the future, we could store that as `additional_info` json
-    def downcase_fields!
-      name.downcase!
-    end
   end
 end

app/models/journal.rb

@@ -0,0 +1,28 @@
+# frozen_string_literal: true
+
+# == Schema Information
+#
+# Table name: journals
+#
+#  id              :integer          not null, primary key
+#  name            :string
+#  additional_info :json
+#  created_at      :datetime         not null
+#  updated_at      :datetime         not null
+#
+
+# Journal is the list of academic journals which are known to TACOS. This list of records is referred to by the
+# Detector::Journal model in order to determine whether a given term matches a known journal. The names of these
+# journals are stored in lowercase, which matches how the Detector::Journal processes incoming terms, in order to
+# prevent capitalization differences resulting in a false negative.
+class Journal < ApplicationRecord
+  before_save :downcase_fields!
+
+  private
+
+  # Downcasing all names before saving allows for more efficient matching by ensuring our index is lowercase.
+  # If we find we need the non-lowercase Journal name in the future, we could store that as `additional_info` json
+  def downcase_fields!
+    name.downcase!
+  end
+end

db/migrate/20241218192753_create_journals.rb

@@ -0,0 +1,11 @@
+class CreateJournals < ActiveRecord::Migration[7.1]
+  def change
+    create_table :journals do |t|
+      t.string :name
+      t.json :additional_info
+
+      t.timestamps
+    end
+    add_index :journals, :name
+  end
+end

db/migrate/20241219144452_drop_detector_journals.rb

@@ -0,0 +1,15 @@
+class DropDetectorJournals < ActiveRecord::Migration[7.1]
+  def up
+    drop_table :detector_journals
+  end
+
+  def down
+    create_table :detector_journals do |t|
+      t.string :name
+      t.json :additional_info
+
+      t.timestamps
+    end
+    add_index :detector_journals, :name
+  end
+end

docs/reference/classes.md

@@ -30,6 +30,8 @@ classDiagram
 
   Detector "1" --> "0..*" DetectorCategory
 
+  DetectorJournal -- Journal : references
+
   Confirmation --> Term
   Confirmation --> Category
   User --> Confirmation : provides many
@@ -110,6 +112,11 @@ classDiagram
     DetectorSuggestedResource: record()
     DetectorSuggestedResource: update_fingerprint()
 
+  class Journal
+    Journal: +Integer id
+    Journal: +String name
+    Journal: +JSON additional_info
+
   class Confirmation
     Confirmation: +Integer id
     Confirmation: +Integer user_id
@@ -134,6 +141,7 @@ classDiagram
     class DetectorLcsh["Detector::Lcsh"]
     class DetectorStandardIdentifier["Detector::StandardIdentifiers"]
     class DetectorSuggestedResource["Detector::SuggestedResource"]
+    class Journal
   }
 
   namespace UserActivity {
@@ -153,6 +161,7 @@ classDiagram
   style DetectorLcsh fill:#000,stroke:#fc8d62,color:#fc8d62
   style DetectorStandardIdentifier fill:#000,stroke:#fc8d62,color:#fc8d62
   style DetectorSuggestedResource fill:#000,stroke:#fc8d62,color:#fc8d62
+  style Journal fill:#000,stroke:#fc8d62,color:#fc8d62
 
   style Categorization fill:#000,stroke:#8da0cb,color:#8da0cb,stroke-dasharray: 3 5;
   style Detection fill:#000,stroke:#8da0cb,color:#8da0cb,stroke-dasharray: 3 5;

lib/tasks/journals.rake

@@ -93,18 +93,18 @@ namespace :journals do
     end
 
     # Delete all journals. We do this to simplify the loader process to avoid consideration of updates/deletes.
-    Detector::Journal.delete_all
+    Journal.delete_all
 
     # not ideal, we should consider streaming the file rather than loading it fully into memory
     json = JSON.parse(data)
 
     json['core'].each do |journal|
-      Detector::Journal.create(name: journal['title'],
-                               additional_info: { issns: journal['issns'],
-                                                  publisher: journal['publisher'],
-                                                  alternate_titles: journal['alternate_titles'],
-                                                  type: journal['type'],
-                                                  abbreviated_title: journal['abbreviated_title'] })
+      Journal.create(name: journal['title'],
+                     additional_info: { issns: journal['issns'],
+                                        publisher: journal['publisher'],
+                                        alternate_titles: journal['alternate_titles'],
+                                        type: journal['type'],
+                                        abbreviated_title: journal['abbreviated_title'] })
     end
   end
 end

test/fixtures/detector/journals.yml → test/fixtures/journals.yml

@@ -1,6 +1,6 @@
 # == Schema Information
 #
-# Table name: detector_journals
+# Table name: journals
 #
 #  id              :integer          not null, primary key
 #  name            :string

test/models/detector/bulk_checker_test.rb

@@ -11,7 +11,9 @@ class CitationTest < ActiveSupport::TestCase
     end
 
     test 'journal_bulk_checker' do
-      skip 'Detector::Journal does not yet support bulk_checker'
+      bulk = Detector::Journal.check_all_matches(output: true)
+
+      assert_equal(1, bulk.count)
     end
 
     test 'lcsh_bulk_checker' do

test/models/detector/journal_test.rb

@@ -1,29 +1,19 @@
 # frozen_string_literal: true
 
-# == Schema Information
-#
-# Table name: detector_journals
-#
-#  id              :integer          not null, primary key
-#  name            :string
-#  additional_info :json
-#  created_at      :datetime         not null
-#  updated_at      :datetime         not null
-#
 require 'test_helper'
 
 class Detector
   class JournalTest < ActiveSupport::TestCase
     test 'exact term match on journal name' do
-      expected = detector_journals('the_new_england_journal_of_medicine')
+      expected = journals('the_new_england_journal_of_medicine')
       actual = Detector::Journal.full_term_match('the new england journal of medicine')
 
       assert_equal 1, actual.count
       assert_equal(expected, actual.first)
     end
 
     test 'mixed case exact term match on journal name' do
-      expected = detector_journals('the_new_england_journal_of_medicine')
+      expected = journals('the_new_england_journal_of_medicine')
       actual = Detector::Journal.full_term_match('The New England Journal of Medicine')
 
       assert_equal 1, actual.count
@@ -48,15 +38,6 @@ class JournalTest < ActiveSupport::TestCase
       assert_equal 2, actual.count
     end
 
-    test 'mixed titles are downcased when saved' do
-      mixed_case = 'ThIs Is A tItLe'
-      actual = Detector::Journal.create(name: mixed_case)
-      actual.reload
-
-      assert_not_equal(mixed_case, actual.name)
-      assert_equal(mixed_case.downcase, actual.name)
-    end
-
     test 'record does relevant work' do
       detection_count = Detection.count
       t = terms('journal_nature_medicine')

test/models/journal_test.rb

@@ -0,0 +1,24 @@
+# frozen_string_literal: true
+
+# == Schema Information
+#
+# Table name: journals
+#
+#  id              :integer          not null, primary key
+#  name            :string
+#  additional_info :json
+#  created_at      :datetime         not null
+#  updated_at      :datetime         not null
+#
+require 'test_helper'
+
+class JournalTest < ActiveSupport::TestCase
+  test 'mixed titles are downcased when saved' do
+    mixed_case = 'ThIs Is A tItLe'
+    actual = Journal.create(name: mixed_case)
+    actual.reload
+
+    assert_not_equal(mixed_case, actual.name)
+    assert_equal(mixed_case.downcase, actual.name)
+  end
+end