[core] Decode environment variables with filesystem encoding (Fixes #3854, Fixes...

[youtube-dl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 9f49507c1256a707a139bbbba407a2b17d3a87ad..afe32ae0582824c0d02b965fb008c83f5c1cc044 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -203,6 +203,48 @@ def compat_ord(c):
     if type(c) is int: return c
     else: return ord(c)
 
+
+# Environment variables should be decoded with filesystem encoding
+# otherwise this results in issues like #3854 #2918 #3217
+if sys.version_info >= (3, 0):
+    compat_getenv = os.getenv
+    compat_expanduser = os.path.expanduser
+else:
+    def compat_getenv(key, default=None):
+        env = os.getenv(key, default)
+        if env:
+            env = env.decode(get_filesystem_encoding())
+        return env
+
+    def compat_expanduser(path):
+        """Expand ~ and ~user constructs.
+
+        If user or $HOME is unknown, do nothing."""
+        if path[:1] != '~':
+            return path
+        i, n = 1, len(path)
+        while i < n and path[i] not in '/\\':
+            i += 1
+
+        if 'HOME' in os.environ:
+            userhome = compat_getenv('HOME')
+        elif 'USERPROFILE' in os.environ:
+            userhome = compat_getenv('USERPROFILE')
+        elif not 'HOMEPATH' in os.environ:
+            return path
+        else:
+            try:
+                drive = compat_getenv('HOMEDRIVE')
+            except KeyError:
+                drive = ''
+            userhome = os.path.join(drive, compat_getenv('HOMEPATH'))
+
+        if i != 1:  # ~user
+            userhome = os.path.join(os.path.dirname(userhome), path[1:i])
+
+        return userhome + path[i:]
+
+
 # This is not clearly defined otherwise
 compiled_regex_type = type(re.compile(''))
 
@@ -799,6 +841,12 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
                 del req.headers['User-agent']
             req.headers['User-agent'] = req.headers['Youtubedl-user-agent']
             del req.headers['Youtubedl-user-agent']
+
+        if sys.version_info < (2, 7) and '#' in req.get_full_url():
+            # Python 2.6 is brain-dead when it comes to fragments
+            req._Request__original = req._Request__original.partition('#')[0]
+            req._Request__r_type = req._Request__r_type.partition('#')[0]
+
         return req
 
     def http_response(self, req, resp):
@@ -884,6 +932,7 @@ def unified_strdate(date_str):
         '%d/%m/%Y',
         '%d/%m/%y',
         '%Y/%m/%d %H:%M:%S',
+        '%d/%m/%Y %H:%M:%S',
         '%Y-%m-%d %H:%M:%S',
         '%d.%m.%Y %H:%M',
         '%d.%m.%Y %H.%M',
@@ -1197,11 +1246,14 @@ class locked_file(object):
         return self.f.read(*args)
 
 
+def get_filesystem_encoding():
+    encoding = sys.getfilesystemencoding()
+    return encoding if encoding is not None else 'utf-8'
+
+
 def shell_quote(args):
     quoted_args = []
-    encoding = sys.getfilesystemencoding()
-    if encoding is None:
-        encoding = 'utf-8'
+    encoding = get_filesystem_encoding()
     for a in args:
         if isinstance(a, bytes):
             # We may get a filename encoded with 'encodeFilename'
@@ -1251,7 +1303,7 @@ def format_bytes(bytes):
 
 
 def get_term_width():
-    columns = os.environ.get('COLUMNS', None)
+    columns = compat_getenv('COLUMNS', None)
     if columns:
         return int(columns)