Merge pull request #95 from beyondwords-io/allow-a-wildcard-in-the-ta…

…gs-option Add a way to bypass the options[:tags] whitelist
cantino · Nov 11, 2023 · cc1b0d2 · cc1b0d2
2 parents 599ed39 + aeec549
commit cc1b0d2
Show file tree

Hide file tree

Showing 3 changed files with 25 additions and 1 deletion.
diff --git a/lib/readability.rb b/lib/readability.rb
@@ -426,6 +426,8 @@ def sanitize(node, candidates, options = {})
 
       # We'll sanitize all elements using a whitelist
       base_whitelist = @options[:tags] || %w[div p]
+      all_whitelisted = base_whitelist.include?("*")
+
       # We'll add whitespace instead of block elements,
       # so a<br>b will have a nice space between them
       base_replace_with_whitespace = %w[br hr h1 h2 h3 h4 h5 h6 dl dd ol li ul address blockquote center]
@@ -438,7 +440,7 @@ def sanitize(node, candidates, options = {})
 
       ([node] + node.css("*")).each do |el|
         # If element is in whitelist, delete all its attributes
-        if whitelist[el.node_name]
+        if all_whitelisted || whitelist[el.node_name]
           el.attributes.each { |a, x| el.delete(a) unless @options[:attributes] && @options[:attributes].include?(a.to_s) }
 
           # Otherwise, replace the element with its contents

diff --git a/spec/fixtures/nested_images.html b/spec/fixtures/nested_images.html
@@ -0,0 +1,11 @@
+<html>
+  <body>
+    <article>
+      <section>
+        <figure>
+          <img src="http://example.com/image.jpeg" />
+        </figure>
+      </section>
+    </article>
+  </body>
+</html>
diff --git a/spec/readability_spec.rb b/spec/readability_spec.rb
@@ -56,6 +56,7 @@
       @nytimes  = File.read(File.dirname(__FILE__) + "/fixtures/nytimes.html")
       @thesun   = File.read(File.dirname(__FILE__) + "/fixtures/thesun.html")
       @ch       = File.read(File.dirname(__FILE__) + "/fixtures/codinghorror.html")
+      @nested   = File.read(File.dirname(__FILE__) + "/fixtures/nested_images.html")
 
       FakeWeb::Registry.instance.clean_registry
 
@@ -103,6 +104,16 @@
       ])
     end
 
+    it "should be able to preserve deeply nested image tags in the article's content by whitelisting all tags" do
+      @doc = Readability::Document.new(@nested, attributes: ["src"])
+      expect(@doc.images).to be_empty
+
+      @doc = Readability::Document.new(@nested, attributes: ["src"], tags: ["figure", "image"])
+      expect(@doc.images).to be_empty
+
+      @doc = Readability::Document.new(@nested, attributes: ["src"], tags: ["*"])
+      expect(@doc.content).to include('<img src="http://example.com/image.jpeg" />')
+    end
 
     it "should not try to download local images" do
       @doc = Readability::Document.new(<<-HTML)