[utils] Process bytestrings in urljoin (closes #12369)

[youtube-dl.git] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 807183f4a72a767b9b30b4878c15934f300def76..d293c74982f2a9f5797f101a7658536c7a800d00 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -473,7 +473,8 @@ def timeconvert(timestr):
 def sanitize_filename(s, restricted=False, is_id=False):
     """Sanitizes a string so it could be used as part of a filename.
     If restricted is set, use a stricter subset of allowed characters.
-    Set is_id if this is not an arbitrary string, but an ID that should be kept if possible
+    Set is_id if this is not an arbitrary string, but an ID that should be kept
+    if possible.
     """
     def replace_insane(char):
         if restricted and char in ACCENT_CHARS:
@@ -1747,11 +1748,16 @@ def base_url(url):
 
 
 def urljoin(base, path):
+    if isinstance(path, bytes):
+        path = path.decode('utf-8')
     if not isinstance(path, compat_str) or not path:
         return None
     if re.match(r'^(?:https?:)?//', path):
         return path
-    if not isinstance(base, compat_str) or not re.match(r'^(?:https?:)?//', base):
+    if isinstance(base, bytes):
+        base = base.decode('utf-8')
+    if not isinstance(base, compat_str) or not re.match(
+            r'^(?:https?:)?//', base):
         return None
     return compat_urlparse.urljoin(base, path)