Make account search blazing fast and rank followers/followees higher in the results

6 files changed, 50 insertions, 15 deletions

diff --git a/Gemfile b/Gemfile
index c97f80bde..56a8a6594 100644
--- a/Gemfile
+++ b/Gemfile
@@ -47,7 +47,6 @@ gem 'rack-attack'
 gem 'rack-cors', require: 'rack/cors'
 gem 'sidekiq'
 gem 'rails-settings-cached'
-gem 'pg_search'
 gem 'simple-navigation'
 gem 'statsd-instrument'
 gem 'ruby-oembed', require: 'oembed'
diff --git a/Gemfile.lock b/Gemfile.lock
index 8ab50773f..c8a5991d0 100644
--- a/Gemfile.lock
+++ b/Gemfile.lock
@@ -254,10 +254,6 @@ GEM
     parser (2.3.1.2)
       ast (~> 2.2)
     pg (0.18.4)
-    pg_search (1.0.6)
-      activerecord (>= 3.1)
-      activesupport (>= 3.1)
-      arel
     pghero (1.6.2)
       activerecord
     powerpack (0.1.1)
@@ -491,7 +487,6 @@ DEPENDENCIES
   paperclip (~> 5.1)
   paperclip-av-transcoder
   pg
-  pg_search
   pghero
   pry-rails
   puma
diff --git a/app/models/account.rb b/app/models/account.rb
index 1eb886ee3..978dc2d71 100644
--- a/app/models/account.rb
+++ b/app/models/account.rb
@@ -2,7 +2,6 @@
 
 class Account < ApplicationRecord
   include Targetable
-  include PgSearch
 
   MENTION_RE = /(?:^|[^\/\w])@([a-z0-9_]+(?:@[a-z0-9\.\-]+[a-z0-9]+)?)/i
   IMAGE_MIME_TYPES = ['image/jpeg', 'image/png', 'image/gif'].freeze
@@ -56,9 +55,6 @@ class Account < ApplicationRecord
   # PuSH subscriptions
   has_many :subscriptions, dependent: :destroy
 
-  pg_search_scope :search_for, against: { display_name: 'A', username: 'B', domain: 'C' },
-                               using: { tsearch: { prefix: true } }
-
   scope :remote, -> { where.not(domain: nil) }
   scope :local, -> { where(domain: nil) }
   scope :without_followers, -> { where('(select count(f.id) from follows as f where f.target_account_id = accounts.id) = 0') }
@@ -212,6 +208,42 @@ SQL
       Account.find_by_sql([sql, account.id, account.id, limit])
     end
 
+    def search_for(terms, limit = 10)
+      textsearch  = '(setweight(to_tsvector(\'simple\', accounts.display_name), \'A\') || setweight(to_tsvector(\'simple\', accounts.username), \'B\') || setweight(to_tsvector(\'simple\', coalesce(accounts.domain, \'\')), \'C\'))'
+      query       = 'to_tsquery(\'simple\', \'\'\' \' || ? || \' \'\'\' || \':*\')'
+
+      sql = <<SQL
+        SELECT
+          accounts.*,
+          ts_rank_cd(#{textsearch}, #{query}, 32) AS rank
+        FROM accounts
+        WHERE #{query} @@ #{textsearch}
+        ORDER BY rank DESC
+        LIMIT ?
+SQL
+
+      Account.find_by_sql([sql, terms, terms, limit])
+    end
+
+    def advanced_search_for(terms, account, limit = 10)
+      textsearch  = '(setweight(to_tsvector(\'simple\', accounts.display_name), \'A\') || setweight(to_tsvector(\'simple\', accounts.username), \'B\') || setweight(to_tsvector(\'simple\', coalesce(accounts.domain, \'\')), \'C\'))'
+      query       = 'to_tsquery(\'simple\', \'\'\' \' || ? || \' \'\'\' || \':*\')'
+
+      sql = <<SQL
+        SELECT
+          accounts.*,
+          (count(f.id) + 1) * ts_rank_cd(#{textsearch}, #{query}, 32) AS rank
+        FROM accounts
+        LEFT OUTER JOIN follows AS f ON (accounts.id = f.account_id AND f.target_account_id = ?) OR (accounts.id = f.target_account_id AND f.account_id = ?)
+        WHERE #{query} @@ #{textsearch}
+        GROUP BY accounts.id
+        ORDER BY rank DESC
+        LIMIT ?
+SQL
+
+      Account.find_by_sql([sql, terms, account.id, account.id, terms, limit])
+    end
+
     def following_map(target_account_ids, account_id)
       follow_mapping(Follow.where(target_account_id: target_account_ids, account_id: account_id), :target_account_id)
     end
diff --git a/app/services/search_service.rb b/app/services/search_service.rb
index 6f740e149..19fc16973 100644
--- a/app/services/search_service.rb
+++ b/app/services/search_service.rb
@@ -1,7 +1,7 @@
 # frozen_string_literal: true
 
 class SearchService < BaseService
-  def call(query, limit, resolve = false)
+  def call(query, limit, resolve = false, account = nil)
     return if query.blank? || query.start_with?('#')
 
     username, domain = query.gsub(/\A@/, '').split('@')
@@ -9,13 +9,12 @@ class SearchService < BaseService
 
     if domain.nil?
       exact_match = Account.find_local(username)
-      results     = Account.search_for(username)
+      results     = account.nil? ? Account.search_for(username, limit) : Account.advanced_search_for(username, account, limit)
     else
       exact_match = Account.find_remote(username, domain)
-      results     = Account.search_for("#{username} #{domain}")
+      results     = account.nil? ? Account.search_for("#{username} #{domain}", limit) : Account.advanced_search_for("#{username} #{domain}", account, limit)
     end
 
-    results = results.limit(limit).to_a
     results = [exact_match] + results.reject { |a| a.id == exact_match.id } if exact_match
 
     if resolve && !exact_match && !domain.nil?
diff --git a/db/migrate/20170317193015_add_search_index_to_accounts.rb b/db/migrate/20170317193015_add_search_index_to_accounts.rb
new file mode 100644
index 000000000..14e174147
--- /dev/null
+++ b/db/migrate/20170317193015_add_search_index_to_accounts.rb
@@ -0,0 +1,9 @@
+class AddSearchIndexToAccounts < ActiveRecord::Migration[5.0]
+  def up
+    execute 'CREATE INDEX search_index ON accounts USING gin((setweight(to_tsvector(\'simple\', accounts.display_name), \'A\') || setweight(to_tsvector(\'simple\', accounts.username), \'B\') || setweight(to_tsvector(\'simple\', coalesce(accounts.domain, \'\')), \'C\')));'
+  end
+
+  def down
+    remove_index :accounts, name: :search_index
+  end
+end
diff --git a/db/schema.rb b/db/schema.rb
index 4ec85ef2b..c571fa3a8 100644
--- a/db/schema.rb
+++ b/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: 20170304202101) do
+ActiveRecord::Schema.define(version: 20170317193015) do
 
   # These are extensions that must be enabled in order to support this database
   enable_extension "plpgsql"
@@ -43,6 +43,7 @@ ActiveRecord::Schema.define(version: 20170304202101) do
     t.boolean  "silenced",                default: false, null: false
     t.boolean  "suspended",               default: false, null: false
     t.boolean  "locked",                  default: false, null: false
+    t.index "(((setweight(to_tsvector('simple'::regconfig, (display_name)::text), 'A'::\"char\") || setweight(to_tsvector('simple'::regconfig, (username)::text), 'B'::\"char\")) || setweight(to_tsvector('simple'::regconfig, (COALESCE(domain, ''::character varying))::text), 'C'::\"char\")))", name: "search_index", using: :gin
     t.index ["username", "domain"], name: "index_accounts_on_username_and_domain", unique: true, using: :btree
   end