From 108c231d1cabfd6429c50765ab54649d6cfcca79 Mon Sep 17 00:00:00 2001
From: Guo Xiang Tan <tgx_world@hotmail.com>
Date: Mon, 8 Apr 2019 16:51:39 +0800
Subject: [PATCH] FIX: Clean up `topic_search_data` of trashed topics.

This keeps the index and table smaller.
---
 app/jobs/scheduled/reindex_search.rb | 18 +++++++++++++++++-
 spec/jobs/reindex_search_spec.rb     | 24 +++++++++++++++++++++---
 2 files changed, 38 insertions(+), 4 deletions(-)

diff --git a/app/jobs/scheduled/reindex_search.rb b/app/jobs/scheduled/reindex_search.rb
index 5a1c784e20a..802379dff06 100644
--- a/app/jobs/scheduled/reindex_search.rb
+++ b/app/jobs/scheduled/reindex_search.rb
@@ -3,6 +3,8 @@ module Jobs
   class ReindexSearch < Jobs::Scheduled
     every 2.hours
 
+    CLEANUP_GRACE_PERIOD = 1.week.ago
+
     def execute(args)
       rebuild_problem_topics
       rebuild_problem_posts
@@ -10,6 +12,7 @@ module Jobs
       rebuild_problem_users
       rebuild_problem_tags
       clean_post_search_data
+      clean_topic_search_data
     end
 
     def rebuild_problem_categories(limit: 500)
@@ -67,7 +70,7 @@ module Jobs
         .where("p.raw = ''")
         .delete_all
 
-      DB.exec(<<~SQL, deleted_at: 1.week.ago)
+      DB.exec(<<~SQL, deleted_at: CLEANUP_GRACE_PERIOD)
         DELETE FROM post_search_data
         WHERE post_id IN (
           SELECT post_id
@@ -80,6 +83,19 @@ module Jobs
       SQL
     end
 
+    def clean_topic_search_data
+      DB.exec(<<~SQL, deleted_at: CLEANUP_GRACE_PERIOD)
+      DELETE FROM topic_search_data
+      WHERE topic_id IN (
+        SELECT topic_id
+        FROM topic_search_data
+        INNER JOIN topics ON topic_search_data.topic_id = topics.id
+        WHERE topics.deleted_at IS NOT NULL
+        AND topics.deleted_at <= :deleted_at
+      )
+      SQL
+    end
+
     def load_problem_post_ids(limit)
       params = {
         locale: SiteSetting.default_locale,
diff --git a/spec/jobs/reindex_search_spec.rb b/spec/jobs/reindex_search_spec.rb
index bc6b3f6efa5..a3edcb97293 100644
--- a/spec/jobs/reindex_search_spec.rb
+++ b/spec/jobs/reindex_search_spec.rb
@@ -83,6 +83,24 @@ describe Jobs::ReindexSearch do
   end
 
   describe '#execute' do
+    it "should clean up topic_search_data of trashed topics" do
+      topic = Fabricate(:post).topic
+      topic2 = Fabricate(:post).topic
+
+      [topic, topic2].each { |t| SearchIndexer.index(t, force: true) }
+
+      freeze_time(described_class::CLEANUP_GRACE_PERIOD) do
+        topic.trash!
+      end
+
+      expect { subject.execute({}) }.to change { TopicSearchData.count }.by(-1)
+      expect(Topic.pluck(:id)).to contain_exactly(topic2.id)
+
+      expect(TopicSearchData.pluck(:topic_id)).to contain_exactly(
+        topic2.topic_search_data.topic_id
+      )
+    end
+
     it(
       "should clean up post_search_data of posts with empty raw or posts from " \
       "trashed topics"
@@ -96,18 +114,18 @@ describe Jobs::ReindexSearch do
       post3.topic.trash!
       post4 = nil
 
-      freeze_time(1.week.ago) do
+      freeze_time(described_class::CLEANUP_GRACE_PERIOD) do
         post4 = Fabricate(:post)
         post4.topic.trash!
       end
 
       expect { subject.execute({}) }.to change { PostSearchData.count }.by(-2)
 
-      expect(Post.all.pluck(:id)).to contain_exactly(
+      expect(Post.pluck(:id)).to contain_exactly(
         post.id, post2.id, post3.id, post4.id
       )
 
-      expect(PostSearchData.all.pluck(:post_id)).to contain_exactly(
+      expect(PostSearchData.pluck(:post_id)).to contain_exactly(
         post.post_search_data.post_id, post3.post_search_data.post_id
       )
     end