sources: factor out Source::URL::Newgrounds.

Also fix it so that the image URL is set as the source for Newgrounds
posts, not the page URL. It's possible to generate the page URL from the
image URL (except for images after the first in multi-image posts).

* Page: https://www.newgrounds.com/art/view/natthelich/weaver
* Image: https://art.ngfiles.com/images/1520000/1520217_natthelich_weaver.jpg?f1606365031

test/unit/sources/newgrounds_test.rb

@@ -1,7 +1,7 @@
 require 'test_helper'
 
 module Sources
-  class NewGroundsTest < ActiveSupport::TestCase
+  class NewgroundsTest < ActiveSupport::TestCase
     context "The source for a newgrounds picture" do
       setup do
         @url = "https://www.newgrounds.com/art/view/hcnone/sephiroth"
@@ -32,7 +32,7 @@ module Sources
 
       should "get the canonical url" do
         assert_equal(@url, @image_1.canonical_url)
-        assert_equal(@url, @image_2.canonical_url)
+        assert_equal(@image_url, @image_2.canonical_url)
       end
 
       should "download an image" do
@@ -62,6 +62,18 @@ module Sources
       end
     end
 
+    context "A multi-image Newgrounds post" do
+      should "get all the images" do
+        source = Sources::Strategies.find("https://www.newgrounds.com/art/view/natthelich/weaver")
+        image_urls = [
+          "https://art.ngfiles.com/images/1520000/1520217_natthelich_weaver.jpg?f1606365031",
+          "https://art.ngfiles.com/comments/199000/iu_199826_7115981.jpg",
+        ]
+
+        assert_equal(image_urls, source.image_urls)
+      end
+    end
+
     context "A deleted or not existing picture" do
       setup do
         @fake_1 = Sources::Strategies.find("https://www.newgrounds.com/art/view/ThisUser/DoesNotExist")