[deploy] Add unique index on articles user_id, title and body_markdown (#10284)

* Add unique index on articles user_id, title and body_markdown

* Add test

.rubocop_todo.yml

@@ -11,10 +11,3 @@ require:
 # one by one as the offenses are removed from the code base.
 # Note that changes in the inspected code, or installation of new
 # versions of RuboCop, may require this file to be generated again.
-
-# Offense count: 2
-# Configuration parameters: Include.
-# Include: app/models/**/*.rb
-Rails/UniqueValidationWithoutIndex:
-  Exclude:
-    - 'app/models/article.rb'

db/migrate/20200911140318_add_unique_index_to_articles_user_id_title_body_markdown.rb

@@ -0,0 +1,26 @@
+class AddUniqueIndexToArticlesUserIdTitleBodyMarkdown < ActiveRecord::Migration[6.0]
+  disable_ddl_transaction!
+
+  def up
+    # to avoid the error "Values larger than 1/3 of a buffer page cannot be indexed."
+    # due to the fact that large text column cannot be indexed by btree indexes
+    # we are going to need to build an index on the hash of the `body_markdown` column.
+    # We also cannot use a `HASH` index as it's not supported by unique columns.
+    # I don't recommend using GiN or GiST as well
+    # See <https://www.postgresql.org/message-id/AANLkTin3p6VS1Z=TtqUV-5cG4TZpTjUMfuPNzWJFgnr5@mail.gmail.com>,
+    # <https://www.postgresql.org/docs/11/pgcrypto.html#id-1.11.7.34.5> and
+    # <https://www.postgresql.org/docs/11/indexes-types.html>
+    # NOTE: using SQL as I couldn't find a way to have Rails generate it correctly
+    ActiveRecord::Base.connection.execute(
+      <<~SQL
+        CREATE UNIQUE INDEX CONCURRENTLY IF NOT EXISTS "index_articles_on_user_id_and_title_and_digest_body_markdown"
+        ON "articles"
+        USING btree ("user_id", "title", digest("body_markdown", 'sha512'::text));
+      SQL
+    )
+  end
+
+  def down
+    remove_index :articles, name: :index_articles_on_user_id_and_title_and_digest_body_markdown, algorithm: :concurrently
+  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: 2020_09_10_205316) do
+ActiveRecord::Schema.define(version: 2020_09_11_140318) do
 
   # These are extensions that must be enabled in order to support this database
   enable_extension "citext"
@@ -153,6 +153,7 @@ ActiveRecord::Schema.define(version: 2020_09_10_205316) do
     t.string "video_source_url"
     t.string "video_state"
     t.string "video_thumbnail_url"
+    t.index "user_id, title, digest(body_markdown, 'sha512'::text)", name: "index_articles_on_user_id_and_title_and_digest_body_markdown", unique: true
     t.index ["boost_states"], name: "index_articles_on_boost_states", using: :gin
     t.index ["canonical_url"], name: "index_articles_on_canonical_url", unique: true
     t.index ["comment_score"], name: "index_articles_on_comment_score"

spec/models/article_spec.rb

@@ -43,6 +43,15 @@ RSpec.describe Article, type: :model do
 
     it { is_expected.not_to allow_value("foo").for(:main_image_background_hex_color) }
 
+    describe "#body_markdown" do
+      it "is unique scoped for user_id and title" do
+        art2 = build(:article, body_markdown: article.body_markdown, user: article.user, title: article.title)
+
+        expect(art2).not_to be_valid
+        expect(art2.errors.full_messages.to_sentence).to match("markdown has already been taken")
+      end
+    end
+
     describe "#after_commit" do
       it "on update enqueues job to index article to elasticsearch" do
         article.save