From 4b2305e5a04751f4387bbfe93f35c4d835a3c74f Mon Sep 17 00:00:00 2001
From: Jens Kraemer <jk@jkraemer.net>
Date: Thu, 15 Jun 2017 11:45:24 +0800
Subject: [PATCH] moves text extraction to an ActiveJob Job.

- without further configuration, this will run inline as before
- users may setup DelayedJob or any other ActiveJob backend to benefit from
  background text extraction.
---
 app/jobs/extract_fulltext_job.rb       | 19 +++++++++++++++++++
 app/models/attachment.rb               | 22 +++++++++++-----------
 test/jobs/extract_fulltext_job_test.rb | 23 +++++++++++++++++++++++
 3 files changed, 53 insertions(+), 11 deletions(-)
 create mode 100644 app/jobs/extract_fulltext_job.rb
 create mode 100644 test/jobs/extract_fulltext_job_test.rb

diff --git a/app/jobs/extract_fulltext_job.rb b/app/jobs/extract_fulltext_job.rb
new file mode 100644
index 0000000..0a9fd24
--- /dev/null
+++ b/app/jobs/extract_fulltext_job.rb
@@ -0,0 +1,19 @@
+class ExtractFulltextJob < ActiveJob::Base
+  queue_as :text_extraction
+
+  def perform(attachment_id)
+    if att = find_attachment(attachment_id) and
+      att.readable? and
+      text = Redmine::TextExtractor.new(att).text
+
+      att.update_column :fulltext, text
+    end
+  end
+
+  private
+
+  def find_attachment(id)
+    Attachment.find_by_id id
+  end
+
+end
diff --git a/app/models/attachment.rb b/app/models/attachment.rb
index 0ee8637..ca9cbc1 100644
--- a/app/models/attachment.rb
+++ b/app/models/attachment.rb
@@ -415,15 +415,6 @@ class Attachment < ActiveRecord::Base
     digest.size < 64 ? "MD5" : "SHA256" if digest.present?
   end
 
-  def extract_fulltext
-    if Redmine::Configuration['enable_fulltext_search'] and
-      readable? and
-      text = Redmine::TextExtractor.new(self).text
-
-      update_column :fulltext, text
-    end
-  end
-
   private
 
   def reuse_existing_file_if_possible
@@ -482,10 +473,19 @@ class Attachment < ActiveRecord::Base
     time.strftime("%Y/%m")
   end
 
+  def extract_fulltext
+    if Redmine::Configuration['enable_fulltext_search']
+      ExtractFulltextJob.perform_later(id)
+    end
+  end
+
+  # attempts to extract the fulltext of any attachments that do not have a
+  # fulltext set currently. This runs inline, does not enqueue any background
+  # jobs.
   def self.extract_fulltext
     if Redmine::Configuration['enable_fulltext_search']
-      Attachment.where(fulltext: nil).find_in_batches do |group|
-        group.each{|a| a.extract_fulltext}
+      Attachment.where(fulltext: nil).pluck(:id).each do |id|
+        ExtractFulltextJob.perform_now id
       end
     else
       logger.info "fulltext search is disabled, check configuration.yml"
diff --git a/test/jobs/extract_fulltext_job_test.rb b/test/jobs/extract_fulltext_job_test.rb
new file mode 100644
index 0000000..ed4b666
--- /dev/null
+++ b/test/jobs/extract_fulltext_job_test.rb
@@ -0,0 +1,23 @@
+require 'test_helper'
+
+class ExtractFulltextJobTest < ActiveJob::TestCase
+
+  def test_should_extract_fulltext
+    att = nil
+    Redmine::Configuration.with 'enable_fulltext_search' => false do
+      att = Attachment.create(
+        :container => Issue.find(1),
+        :file => uploaded_test_file("testfile.txt", "text/plain"),
+        :author => User.find(1),
+        :content_type => 'text/plain')
+    end
+    att.reload
+    assert_nil att.fulltext
+
+    ExtractFulltextJob.perform_now(att.id)
+
+    att.reload
+    assert att.fulltext.include?("this is a text file for upload tests\r\nwith multiple lines")
+  end
+
+end
-- 
2.1.4