Ap 427 - GALC: modify filters and database options (#16)

* Add migrations for AP-427

* Change variable names

* Update variable name

* Add comments and update some functions

* Fix the special items

* Push the schema.rb with latest migration timestamp

* Drop term After 1999

* Update some methods

* Add comments

* Add comments, improve performace etc.

* Add comments

* Rename migration

Author: yzhoubk

db/migrate/20250930215317_add_two_terms.rb

@@ -0,0 +1,13 @@
+# Migration to add "Digital" and "Drawing" terms to the "Medium" facet.
+class AddTwoTerms < ActiveRecord::Migration[7.0]
+  def up
+    facet = Facet.find_by!(name: "Medium")
+    Term.create!(facet: facet, value: "Digital")
+    Term.create!(facet: facet, value: "Drawing")
+  end
+
+  def down
+    facet = Facet.find_by!(name: "Medium")
+    Term.where(facet: facet, value: ["Digital", "Drawing"]).delete_all
+  end
+end

db/migrate/20250930220140_add_three_ordered_terms.rb

@@ -0,0 +1,36 @@
+# Migration summary:
+# After running `up`, the "Decade" facet will have terms:
+#   - "2000-2009" (ord: 11)
+#   - "2010-2019" (ord: 12)
+#   - "After 2020" (ord: 13)
+#   - "No Date" (ord: 14)
+#   - "After 1999" (ord: 15, to be removed after updating items_terms)
+# After running `down`, the "Decade" facet will have:
+#   - "No Date" (ord: 12)
+#   - "After 1999" (ord: 11)
+#   - The three new terms will be deleted.
+
+class AddThreeOrderedTerms < ActiveRecord::Migration[7.0]
+  def up
+    facet = Facet.find_by!(name: "Decade")
+    term  = Term.find_by!(facet: facet, value: "After 1999")
+    # NOTE: Ensure all related items_terms records are updated before removing "After 1999" to prevent orphaned references.
+    term.update!(ord: 15)
+    term  = Term.find_by!(facet: facet, value: "No Date")
+    term.update!(ord: 14)
+     
+    Term.create!(facet: facet, value: "2000-2009", ord: 11)
+    Term.create!(facet: facet, value: "2010-2019", ord: 12)
+    Term.create!(facet: facet, value: "After 2020",ord: 13)
+  end
+
+  def down
+    facet = Facet.find_by!(name: "Decade")
+    Term.where(facet: facet, value: ["2000-2009", "2010-2019", "After 2020"]).delete_all
+    term  = Term.find_by!(facet: facet, value: "After 1999")
+    term.update!(ord: 11)
+    term  = Term.find_by!(facet: facet, value: "No Date")
+    term.update!(ord: 12)
+  end
+end
+

db/migrate/20250930220504_update_term_item_relation.rb

@@ -0,0 +1,74 @@
+# Migration to reassign items with the term "After 1999" to a newly categorized term.
+class UpdateTermItemRelation < ActiveRecord::Migration[7.0]
+  def up
+    facet = Facet.find_by!(name: 'Decade')
+    t_after_2019 = Term.find_by!(facet: facet, value: 'After 1999')
+    
+    new_terms = {
+      (2000..2009) => Term.find_by!(facet: facet, value: '2000-2009'),
+      (2010..2019) => Term.find_by!(facet: facet, value: '2010-2019'),
+      (2020..)     => Term.find_by!(facet: facet, value: 'After 2020')
+    }
+
+    invalid_items = []
+
+    t_after_2019.items.find_each do |item|
+      year = integer_year(item.date)
+      unless year
+        invalid_items << { id: item.id, date: item.date }
+        next
+      end
+
+      target_term = new_terms.find { |range, _| range.cover?(year) }&.last
+      next unless target_term
+        
+      re_assign_term_to_item(item, t_after_2019, target_term)  
+    end
+
+    if invalid_items.any?
+      Rails.logger.warn("Skipped #{invalid_items.size} items due to invalid or missing dates. Example: #{invalid_items.first(5).map { |i| "ID #{i[:id]}: '#{i[:date]}'" }.join(', ')}")
+    end
+  end
+
+  def down
+    facet = Facet.find_by!(name: 'Decade')
+    target_term = Term.find_by!(facet: facet, value: 'After 1999')
+    revert_term_assignment(facet,  '2000-2009', target_term)
+    revert_term_assignment(facet, '2010-2019', target_term)
+    revert_term_assignment(facet, 'After 2020', target_term)
+  end
+
+  private
+
+  def integer_year(val)
+    return nil if val.nil?
+
+    # Try direct integer conversion
+    year = nil
+    begin
+      year = Integer(val)
+    rescue ArgumentError, TypeError
+      # Try to extract year from common date formats (e.g., "YYYY-MM-DD", "YYYY/MM/DD")
+      if val.is_a?(String)
+        if val =~ /\A(\d{4})[-\/]/
+          year = $1.to_i
+        end
+      end
+    end
+    return year
+  end
+
+  # Reassigns an item from the origin term to the target term, and removes the origin term from the item.
+  def re_assign_term_to_item(item, origin_term, target_term)
+    item.terms << target_term
+    item.terms.delete(origin_term)
+  end
+
+  # Rolls back term assignment from origin_term_value to target_term for all items.
+  def revert_term_assignment(facet, origin_term_value, target_term)
+    origin_term = Term.find_by!(facet: facet, value: origin_term_value)
+    origin_term.items.find_each do |item|
+      re_assign_term_to_item(item, origin_term, target_term)
+    end
+  end
+end

db/migrate/20250930230620_update_fix_term_item_relation.rb

@@ -0,0 +1,47 @@
+# Migration to update three items with nil or invalid date/year values
+class UpdateFixTermItemRelation < ActiveRecord::Migration[7.0]
+  def up
+    facet = Facet.find_by!(name: 'Decade')
+    t_after_1999 = Term.find_by!(facet: facet, value: 'After 1999')
+    t2000_2009 = Term.find_by!(facet: facet, value: '2000-2009')
+    t2020_plus = Term.find_by!(facet: facet, value: 'After 2020')
+
+    # Items with nil date
+    update_item_term(t_after_1999.id, t2000_2009.id, "i.date is null AND i.title in ('Loading a Cannon', 'Dense Populations')")
+    # Items with date = '[2020]'
+    update_item_term(t_after_1999.id, t2020_plus.id, "date = '[2020]'")
+  end
+
+  def down
+    facet = Facet.find_by!(name: 'Decade')
+    t_after_1999 = Term.find_by!(facet: facet, value: 'After 1999')
+    t2000_2009 = Term.find_by!(facet: facet, value: '2000-2009')
+    t2020_plus = Term.find_by!(facet: facet, value: 'After 2020')
+
+    # revert items with nil date
+    update_item_term(t2000_2009.id,t_after_1999.id, "i.date is null AND i.title in ('Loading a Cannon', 'Dense Populations')")
+    # revert items with date = '[2020]'
+    update_item_term(t2020_plus.id, t_after_1999.id, "date = '[2020]'")
+  end
+
+  private
+
+  def update_item_term(old_term_id, new_term_id, condition_sql)
+
+    execute <<~SQL.squish
+      WITH target AS (
+        SELECT it.term_id, it.item_id
+        FROM items i
+        JOIN items_terms it ON i.id = it.item_id
+        JOIN terms t ON it.term_id = t.id
+        WHERE t.id = #{old_term_id}
+          AND  #{condition_sql}
+      )
+      UPDATE items_terms it
+      SET term_id = #{new_term_id}
+      FROM target
+      WHERE it.item_id = target.item_id
+      AND it.term_id = target.term_id;
+    SQL
+  end
+end

db/migrate/20251001223022_delete_term_after1999.rb

@@ -0,0 +1,11 @@
+class DeleteTermAfter1999 < ActiveRecord::Migration[7.0]
+  def up
+    facet = Facet.find_by!(name: 'Decade')
+    exec_delete("DELETE FROM terms WHERE facet_id = #{facet.id} and value = 'After 1999' and ord = 15")
+  end
+
+  def down
+    facet = Facet.find_by!(name: 'Decade')
+    Term.create!(facet: facet, value: "After 1999", ord: 15)
+  end
+end

db/schema.rb

@@ -10,7 +10,7 @@
 #
 # It's strongly recommended that you check this file into your version control system.
 
-ActiveRecord::Schema[7.0].define(version: 2022_12_08_181611) do
+ActiveRecord::Schema[7.0].define(version: 2025_10_01_223022) do
   # These are extensions that must be enabled in order to support this database
   enable_extension "plpgsql"