ES: index snippet content only up to 1 MB

To avoid indexing timeout.
1 MB is chosen because it's repo file index limit,
and snippet will be migrated from database to repo.

doc/integration/elasticsearch.md

@@ -333,6 +333,10 @@ curl --request PUT localhost:9200/gitlab-production/_settings --data '{
 
 Enable Elasticsearch search in **Admin > Settings > Integrations**. That's it. Enjoy it!
 
+### Index limit
+
+Currently for repository and snippet files, GitLab would only index up to 1 MB of content, in order to avoid indexing timeout.
+
 ## GitLab Elasticsearch Rake Tasks
 
 There are several rake tasks available to you via the command line:

ee/changelogs/unreleased/12111-snippet-index.yml

+---
+title: "Elasticsearch: index snippet content only up to 1 MB"
+merge_request: 15215
+author:
+type: changed

ee/lib/elastic/latest/snippet_instance_proxy.rb

@@ -3,6 +3,8 @@
 module Elastic
   module Latest
     class SnippetInstanceProxy < ApplicationInstanceProxy
+      MAX_INDEX_SIZE = 1.megabyte
+
       def as_indexed_json(options = {})
         # We don't use as_json(only: ...) because it calls all virtual and serialized attributes
         # https://gitlab.com/gitlab-org/gitlab-ee/issues/349
@@ -22,6 +24,10 @@ module Elastic
           data[attr.to_s] = safely_read_attribute_for_elasticsearch(attr)
         end
 
+        if data['content'].bytesize > MAX_INDEX_SIZE
+          data['content'] = data['content'].mb_chars.limit(MAX_INDEX_SIZE).to_s # rubocop: disable CodeReuse/ActiveRecord
+        end
+
         # ES6 is now single-type per index, so we implement our own typing
         data['type'] = es_type
 

ee/spec/lib/elastic/latest/snippet_instance_proxy_spec.rb

+# frozen_string_literal: true
+
+require 'spec_helper'
+
+describe Elastic::Latest::SnippetInstanceProxy do
+  let(:snippet) { create(:personal_snippet) }
+
+  subject { described_class.new(snippet) }
+
+  context '#as_indexed_json' do
+    it 'serializes snippet as hash' do
+      expect(subject.as_indexed_json.with_indifferent_access).to include(
+        id: snippet.id,
+        title: snippet.title,
+        file_name: snippet.file_name,
+        content: snippet.content,
+        created_at: snippet.created_at,
+        updated_at: snippet.updated_at,
+        project_id: snippet.project_id,
+        author_id: snippet.author_id,
+        visibility_level: snippet.visibility_level
+      )
+    end
+  end
+end

ee/spec/lib/gitlab/elastic/snippet_search_results_spec.rb

@@ -61,4 +61,14 @@ describe Gitlab::Elastic::SnippetSearchResults, :elastic do
       expect(results.snippet_blobs_count).to eq(1)
     end
   end
+
+  context 'when content is too long' do
+    let(:content) { "abc" + (" " * Elastic::Latest::SnippetInstanceProxy::MAX_INDEX_SIZE) + "xyz" }
+    let(:snippet) { create(:personal_snippet, :public, content: content) }
+
+    it 'indexes up to a limit' do
+      expect(described_class.new(nil, 'abc').snippet_blobs_count).to eq(1)
+      expect(described_class.new(nil, 'xyz').snippet_blobs_count).to eq(0)
+    end
+  end
 end