[compat] Fix compat_shlex_split for non-ASCII input

Closes #9871

diff --git a/test/test_compat.py b/test/test_compat.py
index 1d7ac9f1696b8e66b1f8bd31d2f1694cc65468f6..b574249489a3ded4cf5dcd66e49c123829c2331e 100644 (file)
--- a/test/test_compat.py
+++ b/test/test_compat.py
@@ -88,6 +88,7 @@ class TestCompat(unittest.TestCase):
     def test_compat_shlex_split(self):
         self.assertEqual(compat_shlex_split('-option "one two"'), ['-option', 'one two'])
         self.assertEqual(compat_shlex_split('-option "one\ntwo" \n -flag'), ['-option', 'one\ntwo', '-flag'])
+        self.assertEqual(compat_shlex_split('-val 中文'), ['-val', '中文'])
 
     def test_compat_etree_fromstring(self):
         xml = '''

diff --git a/youtube_dl/compat.py b/youtube_dl/compat.py
index 67db1c7c6595b0f219d393da7df1363c48103f26..83b96d38fe66f063608a4669b9880da1481fda23 100644 (file)
--- a/youtube_dl/compat.py
+++ b/youtube_dl/compat.py
@@ -1,3 +1,4 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
 import binascii
@@ -2594,15 +2595,16 @@ except ImportError:  # Python < 3.3
             return "'" + s.replace("'", "'\"'\"'") + "'"
 
 
-if sys.version_info >= (2, 7, 3):
+try:
+    assert shlex.split('中文') == ['中文']
     compat_shlex_split = shlex.split
-else:
+except (AssertionError, UnicodeWarning, UnicodeEncodeError):
     # Working around shlex issue with unicode strings on some python 2
     # versions (see http://bugs.python.org/issue1548891)
     def compat_shlex_split(s, comments=False, posix=True):
         if isinstance(s, compat_str):
             s = s.encode('utf-8')
-        return shlex.split(s, comments, posix)
+        return list(map(lambda s: s.decode('utf-8'), shlex.split(s, comments, posix)))
 
 
 def compat_ord(c):