[utils] Fix urljoin for paths with non-http(s) schemes

jchenpku · Jan 20, 2019 · fad4ceb · fad4ceb
1 parent 6945b9e
commit fad4ceb
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 1 deletion.
diff --git a/test/test_utils.py b/test/test_utils.py
@@ -507,6 +507,8 @@ def test_urljoin(self):
         self.assertEqual(urljoin('http://foo.de/', ''), None)
         self.assertEqual(urljoin('http://foo.de/', ['foobar']), None)
         self.assertEqual(urljoin('http://foo.de/a/b/c.txt', '.././../d.txt'), 'http://foo.de/d.txt')
+        self.assertEqual(urljoin('http://foo.de/a/b/c.txt', 'rtmp://foo.de'), 'rtmp://foo.de')
+        self.assertEqual(urljoin(None, 'rtmp://foo.de'), 'rtmp://foo.de')
 
     def test_url_or_none(self):
         self.assertEqual(url_or_none(None), None)

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
@@ -1868,7 +1868,7 @@ def urljoin(base, path):
         path = path.decode('utf-8')
     if not isinstance(path, compat_str) or not path:
         return None
-    if re.match(r'^(?:https?:)?//', path):
+    if re.match(r'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path):
         return path
     if isinstance(base, bytes):
         base = base.decode('utf-8')