CV2-6562: Slow Query

app/models/bot/smooch.rb

@@ -838,7 +838,7 @@ def self.save_text_message(message)
         # strip and remove null bytes
         claim = self.extract_claim(text).gsub(/\s+/, ' ').strip.gsub("\u0000", "\\u0000")
         extra = { quote: claim }
-        pm = ProjectMedia.joins(:media).where('trim(lower(quote)) = ?', claim.downcase).where('project_medias.team_id' => team.id).last
+        pm = ProjectMedia.joins(:media).where('medias.quote_hash' => Claim.generate_hash(claim)).where('project_medias.team_id' => team.id).last
         # Don't create a new text media if it's an unconfirmed request with just a few words
         if pm.nil? && message['archived'] == CheckArchivedFlags::FlagCodes::UNCONFIRMED && ::Bot::Alegre.get_number_of_words(claim) < self.min_number_of_words_for_tipline_long_text
           return team

app/models/claim.rb

@@ -12,15 +12,20 @@ def media_type
     'quote'
   end
 
+  def self.generate_hash(claim)
+    Digest::MD5.hexdigest(claim.to_s.strip.downcase)
+  end
+
   private
 
   def remove_null_bytes
     self.quote = self.quote.gsub("\u0000", "\\u0000") unless self.quote.nil?
   end
 
   def set_uuid
-    uuid = Claim.where('lower(quote) = ?', self.quote.to_s.strip.downcase).joins("INNER JOIN project_medias pm ON pm.media_id = medias.id").first&.id
+    hash_value = Claim.generate_hash(self.quote)
+    uuid = Claim.where(quote_hash: hash_value).joins("INNER JOIN project_medias pm ON pm.media_id = medias.id").first&.id
     uuid ||= self.id
-    self.update_column(:uuid, uuid)
+    self.update_columns(uuid: uuid, quote_hash: hash_value)
   end
 end

db/migrate/20260123042611_add_quote_hash_to_media_table.rb

@@ -0,0 +1,6 @@
+class AddQuoteHashToMediaTable < ActiveRecord::Migration[6.1]
+  def change
+    add_column :medias, :quote_hash, :string
+    add_index :medias, :quote_hash
+  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.define(version: 2025_11_26_195433) do
+ActiveRecord::Schema.define(version: 2026_01_23_042611) do
 
   # These are extensions that must be enabled in order to support this database
   enable_extension "plpgsql"
@@ -474,8 +474,10 @@
     t.integer "uuid", default: 0, null: false
     t.text "original_claim"
     t.string "original_claim_hash"
+    t.string "quote_hash"
     t.index "lower((quote)::text)", name: "index_medias_on_lower_quote", where: "((type)::text = 'Claim'::text)", using: :hash
     t.index ["original_claim_hash"], name: "index_medias_on_original_claim_hash", unique: true
+    t.index ["quote_hash"], name: "index_medias_on_quote_hash"
     t.index ["url"], name: "index_medias_on_url", unique: true
   end
 

lib/tasks/migrate/20250817055243_remove_blank_media_items.rake

@@ -2,7 +2,8 @@
 namespace :check do
   namespace :migrate do
     def get_claim_uuid(id, quote)
-      uuid = Claim.where('lower(quote) = ?', quote.to_s.strip.downcase).joins("INNER JOIN project_medias pm ON pm.media_id = medias.id").first&.id
+      hash_value = Digest::MD5.hexdigest(quote.to_s.strip.downcase)
+      uuid = Claim.where(quote_hash: hash_value).joins("INNER JOIN project_medias pm ON pm.media_id = medias.id").first&.id
       uuid ||= id
     end
     # bundle exec rails check:migrate:migrate_published_and_unpublished_items
@@ -92,19 +93,36 @@ namespace :check do
       minutes = ((Time.now.to_i - started) / 60).to_i
       puts "[#{Time.now}] Done in #{minutes} minutes."
     end
+    # rake task to set quote_hash for Claims
+    # bundle exec rails check:migrate:set_claim_quote_hash
+    task set_claim_quote_hash: :environment do
+      started = Time.now.to_i
+      last_claim_id = Rails.cache.read('check:migrate:set_claim_quote_hash') || 0
+      Claim.where('id > ?', last_claim_id)
+      .find_in_batches(batch_size: 2000) do |claims|
+        c_items = []
+        claims.each do |claim|
+          print '.'
+          claim.quote_hash = Digest::MD5.hexdigest(claim.quote.to_s.strip.downcase)
+          c_items << claim.attributes
+        end
+        Claim.upsert_all(c_items)
+        Rails.cache.write('check:migrate:set_claim_quote_hash', claims.pluck(:id).max)
+      end
+      minutes = ((Time.now.to_i - started) / 60).to_i
+      puts "[#{Time.now}] Done in #{minutes} minutes."
+    end
     # rake task to set Claim uuid
     # bundle exec rails check:migrate:set_claim_uuid
     task set_claim_uuid: :environment do
       started = Time.now.to_i
-      last_claim_id = Rails.cache.read('check:migrate:set_claim_uuid') || 0
-      Claim.where('id > ?', last_claim_id).where(uuid: 0)
+      Claim.where(uuid: 0)
       .find_in_batches(batch_size: 1000) do |claims|
         claims.each do |claim|
           print '.'
           uuid = get_claim_uuid(claim.id, claim.quote)
           claim.update_column(:uuid, uuid)
         end
-        Rails.cache.write('check:migrate:set_claim_uuid', claims.pluck(:id).max)
       end
       minutes = ((Time.now.to_i - started) / 60).to_i
       puts "[#{Time.now}] Done in #{minutes} minutes."