{"repo": "isso-comments/isso", "pull_number": 485, "instance_id": "isso-comments__isso-485", "issue_numbers": "", "base_commit": "8e37a88d6d2b5fa3485f9aff39ff4b452ce2f578", "patch": "diff --git a/isso/utils/html.py b/isso/utils/html.py\n--- a/isso/utils/html.py\n+++ b/isso/utils/html.py\n@@ -6,61 +6,53 @@\n \n from distutils.version import LooseVersion as Version\n \n-HTML5LIB_VERSION = Version(pkg_resources.get_distribution(\"html5lib\").version)\n-HTML5LIB_SIMPLETREE = Version(\"0.95\")\n-\n-import html5lib\n-from html5lib.sanitizer import HTMLSanitizer\n-from html5lib.serializer import HTMLSerializer\n-\n+import bleach\n import misaka\n \n \n-def Sanitizer(elements, attributes):\n-\n-    class Inner(HTMLSanitizer):\n+class Sanitizer(object):\n \n+    def __init__(self, elements, attributes):\n         # attributes found in Sundown's HTML serializer [1]\n         # except for <img> tag,\n         # because images are not generated anyways.\n         #\n         # [1] https://github.com/vmg/sundown/blob/master/html/html.c\n-        allowed_elements = [\"a\", \"p\", \"hr\", \"br\", \"ol\", \"ul\", \"li\",\n+        self.elements = [\"a\", \"p\", \"hr\", \"br\", \"ol\", \"ul\", \"li\",\n                             \"pre\", \"code\", \"blockquote\",\n                             \"del\", \"ins\", \"strong\", \"em\",\n                             \"h1\", \"h2\", \"h3\", \"h4\", \"h5\", \"h6\",\n                             \"table\", \"thead\", \"tbody\", \"th\", \"td\"] + elements\n \n         # href for <a> and align for <table>\n-        allowed_attributes = [\"align\", \"href\"] + attributes\n-\n-        # remove disallowed tokens from the output\n-        def disallowed_token(self, token, token_type):\n-            return None\n+        self.attributes = [\"align\", \"href\"] + attributes\n \n-    return Inner\n \n \n-def sanitize(tokenizer, document):\n+    def sanitize(self, text):\n+        clean_html = bleach.clean(text, tags=self.elements,\n+            attributes=self.attributes, strip=True)\n \n-    parser = html5lib.HTMLParser(tokenizer=tokenizer)\n-    domtree = parser.parseFragment(document)\n+        def set_links(attrs, new=False):\n+            href_key = (None, u'href')\n \n-    if HTML5LIB_VERSION > HTML5LIB_SIMPLETREE:\n-        builder = \"etree\"\n+            if href_key not in attrs:\n+                return attrs\n+            if attrs[href_key].startswith(u'mailto:'):\n+                return attrs\n \n-        for link in domtree.findall(\".//{http://www.w3.org/1999/xhtml}a\"):\n-            if link.get('href', None):\n-                link.set(\"rel\", \"nofollow noopener\")\n+            rel_key = (None, u'rel')\n+            rel_values = [val for val in attrs.get(rel_key, u'').split(u' ') if val]\n \n-    else:\n-        builder = \"simpletree\"\n+            for value in [u'nofollow', u'noopener']:\n+                if value not in [rel_val.lower() for rel_val in rel_values]:\n+                    rel_values.append(value)\n \n-    stream = html5lib.treewalkers.getTreeWalker(builder)(domtree)\n-    serializer = HTMLSerializer(\n-        quote_attr_values=True, omit_optional_tags=False)\n+            attrs[rel_key] = u' '.join(rel_values)\n+            return attrs\n \n-    return serializer.render(stream)\n+        linker = bleach.linkifier.Linker(callbacks=[set_links])\n+        return linker.linkify(clean_html)\n \n \n def Markdown(extensions=(\"strikethrough\", \"superscript\", \"autolink\",\n@@ -100,7 +92,7 @@ def __init__(self, conf):\n             conf.getlist(\"allowed-elements\"),\n             conf.getlist(\"allowed-attributes\"))\n \n-        self._render = lambda text: sanitize(sanitizer, parser(text))\n+        self._render = lambda text: sanitizer.sanitize(parser(text))\n \n     def render(self, text):\n         return self._render(text)\ndiff --git a/setup.py b/setup.py\n--- a/setup.py\n+++ b/setup.py\n@@ -5,8 +5,8 @@\n \n from setuptools import setup, find_packages\n \n-requires = ['itsdangerous', 'Jinja2', 'misaka>=2.0,<3.0', 'html5lib<0.9999999',\n-            'werkzeug>=0.9']\n+requires = ['itsdangerous', 'Jinja2', 'misaka>=2.0,<3.0', 'html5lib',\n+            'werkzeug>=0.9', 'bleach']\n \n if sys.version_info < (2, 7):\n     raise SystemExit(\"Python 2 versions < 2.7 are not supported.\")\n", "test_patch": "diff --git a/isso/tests/test_html.py b/isso/tests/test_html.py\n--- a/isso/tests/test_html.py\n+++ b/isso/tests/test_html.py\n@@ -59,7 +59,6 @@ def test_github_flavoured_markdown(self):\n             print(\"Hello, World\")\n             </code></pre>\"\"\")\n \n-    @unittest.skipIf(html.HTML5LIB_VERSION <= html.HTML5LIB_SIMPLETREE, \"backport\")\n     def test_sanitizer(self):\n         sanitizer = html.Sanitizer(elements=[], attributes=[])\n         examples = [\n@@ -73,11 +72,10 @@ def test_sanitizer(self):\n \n         for (input, expected) in examples:\n             if isinstance(expected, list):\n-                self.assertIn(html.sanitize(sanitizer, input), expected)\n+                self.assertIn(sanitizer.sanitize(input), expected)\n             else:\n-                self.assertEqual(html.sanitize(sanitizer, input), expected)\n+                self.assertEqual(sanitizer.sanitize(input), expected)\n \n-    @unittest.skipIf(html.HTML5LIB_VERSION <= html.HTML5LIB_SIMPLETREE, \"backport\")\n     def test_sanitizer_extensions(self):\n         sanitizer = html.Sanitizer(elements=[\"img\"], attributes=[\"src\"])\n         examples = [\n@@ -85,7 +83,7 @@ def test_sanitizer_extensions(self):\n             ('<script src=\"doge.js\"></script>', '')]\n \n         for (input, expected) in examples:\n-            self.assertEqual(html.sanitize(sanitizer, input), expected)\n+            self.assertEqual(sanitizer.sanitize(input), expected)\n \n     def test_render(self):\n         conf = config.new({\n", "problem_statement": "", "hints_text": "", "created_at": "2018-10-01T02:51:13Z"}