add cached word count to attributes, backfill task

app/jobs/cache_attribute_word_count_job.rb

@@ -0,0 +1,30 @@
+class CacheAttributeWordCountJob < ApplicationJob
+  queue_as :cache
+
+  def perform(*args)
+    attribute_id = args.shift
+    attribute    = Attribute.find_by(id: attribute_id)
+
+    return if attribute.nil?
+    return if attribute.value.nil? || attribute.value.blank?
+
+    word_count = WordCountAnalyzer::Counter.new(
+      ellipsis:          'no_special_treatment',
+      hyperlink:         'count_as_one',
+      contraction:       'count_as_one',
+      hyphenated_word:   'count_as_one',
+      date:              'no_special_treatment',
+      number:            'count',
+      numbered_list:     'ignore',
+      xhtml:             'remove',
+      forward_slash:     'count_as_multiple_except_dates',
+      backslash:         'count_as_one',
+      dotted_line:       'ignore',
+      dashed_line:       'ignore',
+      underscore:        'ignore',
+      stray_punctuation: 'ignore'
+    ).count(attribute.value)
+
+    attribute.update!(word_count_cache: word_count)
+  end
+end

app/models/page_data/attribute.rb

@@ -19,6 +19,10 @@ class Attribute < ApplicationRecord
     end
   end
 
+  after_commit do
+    CacheAttributeWordCountJob.perform_later(self.id) if saved_changes.key?('value')
+  end
+
   after_save do
     entity.touch
   end

db/migrate/20211007215520_add_word_count_cache_to_attributes.rb

@@ -0,0 +1,5 @@
+class AddWordCountCacheToAttributes < ActiveRecord::Migration[6.0]
+  def change
+    add_column :attributes, :word_count_cache, :integer
+  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: 2021_09_15_030031) do
+ActiveRecord::Schema.define(version: 2021_10_07_215520) do
 
   create_table "active_storage_attachments", force: :cascade do |t|
     t.string "name", null: false
@@ -165,6 +165,7 @@
     t.datetime "created_at"
     t.datetime "updated_at"
     t.datetime "deleted_at"
+    t.integer "word_count_cache"
     t.index ["attribute_field_id", "deleted_at", "entity_id", "entity_type"], name: "attributes_afi_deleted_at_entity_id_entity_type"
     t.index ["attribute_field_id", "deleted_at"], name: "index_attributes_on_attribute_field_id_and_deleted_at"
     t.index ["attribute_field_id", "user_id", "entity_type", "entity_id", "deleted_at"], name: "attributes_afi_ui_et_ei_da"

lib/tasks/backfill.rake

@@ -1,4 +1,28 @@
 namespace :backfill do
+  desc "Backfill cached word counts on all attributes"
+  task attribute_word_count_caches: :environment do
+    Attribute.where(cached_word_count: nil).find_each do |attribute|
+      word_count = WordCountAnalyzer::Counter.new(
+        ellipsis:          'no_special_treatment',
+        hyperlink:         'count_as_one',
+        contraction:       'count_as_one',
+        hyphenated_word:   'count_as_one',
+        date:              'no_special_treatment',
+        number:            'count',
+        numbered_list:     'ignore',
+        xhtml:             'remove',
+        forward_slash:     'count_as_multiple_except_dates',
+        backslash:         'count_as_one',
+        dotted_line:       'ignore',
+        dashed_line:       'ignore',
+        underscore:        'ignore',
+        stray_punctuation: 'ignore'
+      ).count(attribute.value)
+
+      attribute.update_column(:word_count_cache, word_count)
+    end
+  end
+
   desc "Backfill cached word counts on all documents"
   task document_word_count_caches: :environment do
     Document.where(cached_word_count: nil).where.not(body: [nil, ""]).find_each(batch_size: 500) do |document|