From 209383faabc13a1724b93ac172e54859c76e43a2 Mon Sep 17 00:00:00 2001
From: Vinoth Kannan <vinothkannan@vinkas.com>
Date: Fri, 2 Jun 2017 15:09:06 +0530
Subject: [PATCH] FEATURE: pull onebox images

---
 app/jobs/regular/pull_hotlinked_images.rb |  4 +-
 lib/cooked_post_processor.rb              | 21 +++++++++--
 spec/jobs/pull_hotlinked_images_spec.rb   | 46 ++++++++++++++++++++++-
 3 files changed, 64 insertions(+), 7 deletions(-)

diff --git a/app/jobs/regular/pull_hotlinked_images.rb b/app/jobs/regular/pull_hotlinked_images.rb
index c14dae92fec..316dfda9fcb 100644
--- a/app/jobs/regular/pull_hotlinked_images.rb
+++ b/app/jobs/regular/pull_hotlinked_images.rb
@@ -94,12 +94,14 @@ module Jobs
         # we never want that job to bump the topic
         options = { bypass_bump: true }
         post.revise(Discourse.system_user, changes, options)
+      elsif downloaded_urls.present?
+        post.trigger_post_process(true)
       end
     end
 
     def extract_images_from(html)
       doc = Nokogiri::HTML::fragment(html)
-      doc.css("img[src]") - doc.css(".onebox-result img") - doc.css("img.avatar")
+      doc.css("img[src]") - doc.css("img.avatar")
     end
 
     def is_valid_image_url(src)
diff --git a/lib/cooked_post_processor.rb b/lib/cooked_post_processor.rb
index 48f1d38b494..48fcfe4a733 100644
--- a/lib/cooked_post_processor.rb
+++ b/lib/cooked_post_processor.rb
@@ -51,7 +51,7 @@ class CookedPostProcessor
   end
 
   def keep_reverse_index_up_to_date
-    upload_ids = Set.new
+    upload_ids = Array.new
 
     @doc.css("a/@href", "img/@src").each do |media|
       if upload = Upload.get_from_url(media.value)
@@ -59,6 +59,8 @@ class CookedPostProcessor
       end
     end
 
+    upload_ids.concat(oneboxed_image_uploads.pluck(:id))
+
     values = upload_ids.map{ |u| "(#{@post.id},#{u})" }.join(",")
     PostUpload.transaction do
       PostUpload.delete_all(post_id: @post.id)
@@ -105,7 +107,12 @@ class CookedPostProcessor
   end
 
   def oneboxed_images
-    @doc.css(".onebox-result img, .onebox img")
+    @doc.css(".onebox-body img, .onebox img")
+  end
+
+  def oneboxed_image_uploads
+    urls = oneboxed_images.map { |img| img["src"] }
+    Upload.where(origin: urls)
   end
 
   def limit_size!(img)
@@ -301,8 +308,14 @@ class CookedPostProcessor
       Oneboxer.onebox(url, args)
     end
 
-    # make sure we grab dimensions for oneboxed images
-    oneboxed_images.each { |img| limit_size!(img) }
+    uploads = oneboxed_image_uploads.select(:url, :origin)
+    oneboxed_images.each do |img|
+      upload = uploads.detect { |u| u.origin == img["src"] }
+      next unless upload.present?
+      img["src"] = upload.url
+      # make sure we grab dimensions for oneboxed images
+      limit_size!(img)
+    end
 
     # respect nofollow admin settings
     if !@cooking_options[:omit_nofollow] && SiteSetting.add_rel_nofollow_to_user_content
diff --git a/spec/jobs/pull_hotlinked_images_spec.rb b/spec/jobs/pull_hotlinked_images_spec.rb
index ed680f9983f..220ed4e79cf 100644
--- a/spec/jobs/pull_hotlinked_images_spec.rb
+++ b/spec/jobs/pull_hotlinked_images_spec.rb
@@ -3,10 +3,12 @@ require 'jobs/regular/pull_hotlinked_images'
 
 describe Jobs::PullHotlinkedImages do
 
+  let(:image_url) { "http://wiki.mozilla.org/images/2/2e/Longcat1.png" }
+
   before do
     png = Base64.decode64("R0lGODlhAQABALMAAAAAAIAAAACAAICAAAAAgIAAgACAgMDAwICAgP8AAAD/AP//AAAA//8A/wD//wBiZCH5BAEAAA8ALAAAAAABAAEAAAQC8EUAOw==")
-    stub_request(:get, "http://wiki.mozilla.org/images/2/2e/Longcat1.png").to_return(body: png)
-    stub_request(:head, "http://wiki.mozilla.org/images/2/2e/Longcat1.png")
+    stub_request(:get, image_url).to_return(body: png)
+    stub_request(:head, image_url)
     SiteSetting.download_remote_images_to_local = true
     FastImage.expects(:size).returns([100, 100]).at_least_once
   end
@@ -29,4 +31,44 @@ describe Jobs::PullHotlinkedImages do
     expect(post.raw).to match(/^<img src='\/uploads/)
   end
 
+  describe 'onebox' do
+
+    let(:media) { "File:Brisbane_May_2013201.jpg" }
+    let(:url) { "https://commons.wikimedia.org/wiki/#{media}" }
+    let(:api_url) { "https://en.wikipedia.org/w/api.php?action=query&titles=#{media}&prop=imageinfo&iilimit=50&iiprop=timestamp|user|url&iiurlwidth=500&format=json" }
+
+    before do
+      SiteSetting.queue_jobs = true
+      stub_request(:get, url).to_return(body: '')
+      stub_request(:head, url)
+      stub_request(:get, api_url).to_return(body: "{
+        \"query\": {
+          \"pages\": {
+            \"-1\": {
+              \"title\": \"#{media}\",
+              \"imageinfo\": [{
+                \"thumburl\": \"#{image_url}\",
+                \"url\": \"#{image_url}\",
+                \"descriptionurl\": \"#{url}\"
+              }]
+            }
+          }
+        }
+      }")
+      stub_request(:head, api_url)
+    end
+
+    it 'replaces image src' do
+      post = Fabricate(:post, raw: "#{url}")
+
+      Jobs::ProcessPost.new.execute(post_id: post.id)
+      Jobs::PullHotlinkedImages.new.execute(post_id: post.id)
+      Jobs::ProcessPost.new.execute(post_id: post.id)
+      post.reload
+
+      expect(post.cooked).to match(/<img src=.*\/uploads/)
+    end
+
+  end
+
 end