Also escape meta descriptions

page_content_extractor/html.py

@@ -143,7 +143,8 @@ def get_meta_description(self):
             for desc in descs:
                 content = desc.get('content', '')
                 if len(content) > len(self._meta_desc):
-                    self._meta_desc = content
+                    # Reason to escape https://github.com/berthubert/trifecta/issues/38
+                    self._meta_desc = escape(content)
         return self._meta_desc
 
     def get_meta_image(self):

test/test_html_parser.py

@@ -247,6 +247,15 @@ def test_longer_meta_description(self):
         parser = HtmlContentExtractor(html_doc)
         self.assertEqual(parser.get_meta_description(), "aaaa")
 
+    def test_need_escape_unsafe_meta_description(self):
+        content = "The upload handler checks that the content type starts with "image/", but this check includes the image/svg+xml content type, so the following image is accepted: <?xml version="1.0" encoding="UTF-8" standalone="no"?> <svg xmlns="http://w..."
+        html_doc = f"""
+        <meta name="description" content="{content}">
+        """
+        parser = HtmlContentExtractor(html_doc)
+        # &quot; == &#34;
+        self.assertEqual(parser.get_meta_description(), content.replace('&quot;', '&#34;'))
+
     def test_get_all_meta_images(self):
         src = 'https://opengraph.githubassets.com/740568cb37e42d5beb5c65378e1f66a0a72e5cb1650c8a45df4466e9472825a2/tikv/agatedb'
         html_doc = f"""