From 8fdff2748ffbc4ce3d36c75f2bd33182b641e895 Mon Sep 17 00:00:00 2001
From: Eugen Rochko <eugen@zeonfederated.com>
Date: Fri, 16 Aug 2019 01:24:03 +0200
Subject: Add more accurate account search (#11537)

* Add more accurate account search

When ElasticSearch is available, a more accurate search is implemented:

- Using edge n-gram index for acct and display name
- Using asciifolding and cjk width normalization on display names
- Using Gaussian decay on account activity for additional scoring (recency)
- Using followers/friends ratio for additional scoring (spamminess)
- Using followers number for additional scoring (size)

The exact match precedence only takes effect when the input conforms
to the username format and the username part of it is complete, i.e.
when the user started typing the domain part.

* Support single-letter usernames

* Fix tests

* Fix not picking up account updates

* Add weights and normalization for scores, skip zero terms queries

* Use local counts for accounts index, adjust search parameters

* Fix mistakes

* Using updated_at of accounts is inadequate for remote accounts
---
 spec/services/account_search_service_spec.rb | 122 ++++++---------------------
 1 file changed, 25 insertions(+), 97 deletions(-)

(limited to 'spec')

diff --git a/spec/services/account_search_service_spec.rb b/spec/services/account_search_service_spec.rb
index 7b071b378..5b7182586 100644
--- a/spec/services/account_search_service_spec.rb
+++ b/spec/services/account_search_service_spec.rb
@@ -1,126 +1,56 @@
 require 'rails_helper'
 
 describe AccountSearchService, type: :service do
-  describe '.call' do
-    describe 'with a query to ignore' do
+  describe '#call' do
+    context 'with a query to ignore' do
       it 'returns empty array for missing query' do
         results = subject.call('', nil, limit: 10)
 
         expect(results).to eq []
       end
-      it 'returns empty array for hashtag query' do
-        results = subject.call('#tag', nil, limit: 10)
 
-        expect(results).to eq []
-      end
       it 'returns empty array for limit zero' do
         Fabricate(:account, username: 'match')
+
         results = subject.call('match', nil, limit: 0)
 
         expect(results).to eq []
       end
     end
 
-    describe 'searching for a simple term that is not an exact match' do
+    context 'searching for a simple term that is not an exact match' do
       it 'does not return a nil entry in the array for the exact match' do
-        match = Fabricate(:account, username: 'matchingusername')
-
+        account = Fabricate(:account, username: 'matchingusername')
         results = subject.call('match', nil, limit: 5)
-        expect(results).to eq [match]
-      end
-    end
 
-    describe 'searching local and remote users' do
-      describe "when only '@'" do
-        before do
-          allow(Account).to receive(:find_local)
-          allow(Account).to receive(:search_for)
-          subject.call('@', nil, limit: 10)
-        end
-
-        it 'uses find_local with empty query to look for local accounts' do
-          expect(Account).to have_received(:find_local).with('')
-        end
-      end
-
-      describe 'when no domain' do
-        before do
-          allow(Account).to receive(:find_local)
-          allow(Account).to receive(:search_for)
-          subject.call('one', nil, limit: 10)
-        end
-
-        it 'uses find_local to look for local accounts' do
-          expect(Account).to have_received(:find_local).with('one')
-        end
-
-        it 'uses search_for to find matches' do
-          expect(Account).to have_received(:search_for).with('one', 10, 0)
-        end
-      end
-
-      describe 'when there is a domain' do
-        before do
-          allow(Account).to receive(:find_remote)
-        end
-
-        it 'uses find_remote to look for remote accounts' do
-          subject.call('two@example.com', nil, limit: 10)
-          expect(Account).to have_received(:find_remote).with('two', 'example.com')
-        end
-
-        describe 'and there is no account provided' do
-          it 'uses search_for to find matches' do
-            allow(Account).to receive(:search_for)
-            subject.call('two@example.com', nil, limit: 10, resolve: false)
-
-            expect(Account).to have_received(:search_for).with('two example.com', 10, 0)
-          end
-        end
-
-        describe 'and there is an account provided' do
-          it 'uses advanced_search_for to find matches' do
-            account = Fabricate(:account)
-            allow(Account).to receive(:advanced_search_for)
-            subject.call('two@example.com', account, limit: 10, resolve: false)
-
-            expect(Account).to have_received(:advanced_search_for).with('two example.com', account, 10, nil, 0)
-          end
-        end
+        expect(results).to eq [account]
       end
     end
 
-    describe 'with an exact match' do
-      it 'returns exact match first, and does not return duplicates' do
-        partial = Fabricate(:account, username: 'exactness')
-        exact = Fabricate(:account, username: 'exact')
-
-        results = subject.call('exact', nil, limit: 10)
-        expect(results.size).to eq 2
-        expect(results).to eq [exact, partial]
-      end
-    end
-
-    describe 'when there is a local domain' do
+    context 'when there is a local domain' do
       around do |example|
         before = Rails.configuration.x.local_domain
+
         example.run
+
         Rails.configuration.x.local_domain = before
       end
 
       it 'returns exact match first' do
         remote     = Fabricate(:account, username: 'a', domain: 'remote', display_name: 'e')
         remote_too = Fabricate(:account, username: 'b', domain: 'remote', display_name: 'e')
-        exact = Fabricate(:account, username: 'e')
+        exact      = Fabricate(:account, username: 'e')
+
         Rails.configuration.x.local_domain = 'example.com'
 
         results = subject.call('e@example.com', nil, limit: 2)
+
         expect(results.size).to eq 2
         expect(results).to eq([exact, remote]).or eq([exact, remote_too])
       end
     end
 
-    describe 'when there is a domain but no exact match' do
+    context 'when there is a domain but no exact match' do
       it 'follows the remote account when resolve is true' do
         service = double(call: nil)
         allow(ResolveAccountService).to receive(:new).and_return(service)
@@ -138,23 +68,21 @@ describe AccountSearchService, type: :service do
       end
     end
 
-    describe 'should not include suspended accounts' do
-      it 'returns the fuzzy match first, and does not return suspended exacts' do
-        partial = Fabricate(:account, username: 'exactness')
-        exact = Fabricate(:account, username: 'exact', suspended: true)
+    it 'returns the fuzzy match first, and does not return suspended exacts' do
+      partial = Fabricate(:account, username: 'exactness')
+      exact   = Fabricate(:account, username: 'exact', suspended: true)
+      results = subject.call('exact', nil, limit: 10)
 
-        results = subject.call('exact', nil, limit: 10)
-        expect(results.size).to eq 1
-        expect(results).to eq [partial]
-      end
+      expect(results.size).to eq 1
+      expect(results).to eq [partial]
+    end
 
-      it "does not return suspended remote accounts" do
-        remote = Fabricate(:account, username: 'a', domain: 'remote', display_name: 'e', suspended: true)
+    it "does not return suspended remote accounts" do
+      remote  = Fabricate(:account, username: 'a', domain: 'remote', display_name: 'e', suspended: true)
+      results = subject.call('a@example.com', nil, limit: 2)
 
-        results = subject.call('a@example.com', nil, limit: 2)
-        expect(results.size).to eq 0
-        expect(results).to eq []
-      end
+      expect(results.size).to eq 0
+      expect(results).to eq []
     end
   end
 end
-- 
cgit