Merge pull request #8092 from bpfoley/twitter-thumbnail

[youtube-dl] / youtube_dl / compat.py

diff --git a/youtube_dl/compat.py b/youtube_dl/compat.py
index cf10835ca55d11e0d8803c550b41a198ac039827..74702786a76ee612a7f9ec91567752e22e70eda6 100644 (file)
--- a/youtube_dl/compat.py
+++ b/youtube_dl/compat.py
@@ -77,6 +77,11 @@ try:
 except ImportError:  # Python 2
     from urllib import urlretrieve as compat_urlretrieve
 
+try:
+    from html.parser import HTMLParser as compat_HTMLParser
+except ImportError:  # Python 2
+    from HTMLParser import HTMLParser as compat_HTMLParser
+
 
 try:
     from subprocess import DEVNULL
@@ -181,20 +186,20 @@ except ImportError:  # Python < 3.4
             # parameter := attribute "=" value
             url = req.get_full_url()
 
-            scheme, data = url.split(":", 1)
-            mediatype, data = data.split(",", 1)
+            scheme, data = url.split(':', 1)
+            mediatype, data = data.split(',', 1)
 
             # even base64 encoded data URLs might be quoted so unquote in any case:
             data = compat_urllib_parse_unquote_to_bytes(data)
-            if mediatype.endswith(";base64"):
+            if mediatype.endswith(';base64'):
                 data = binascii.a2b_base64(data)
                 mediatype = mediatype[:-7]
 
             if not mediatype:
-                mediatype = "text/plain;charset=US-ASCII"
+                mediatype = 'text/plain;charset=US-ASCII'
 
             headers = email.message_from_string(
-                "Content-type: %s\nContent-length: %d\n" % (mediatype, len(data)))
+                'Content-type: %s\nContent-length: %d\n' % (mediatype, len(data)))
 
             return compat_urllib_response.addinfourl(io.BytesIO(data), headers, url)
 
@@ -216,10 +221,19 @@ except ImportError:  # Python 2.6
 if sys.version_info[0] >= 3:
     compat_etree_fromstring = xml.etree.ElementTree.fromstring
 else:
-    # on python 2.x the the attributes of a node are str objects instead of
-    # unicode
+    # python 2.x tries to encode unicode strings with ascii (see the
+    # XMLParser._fixtext method)
     etree = xml.etree.ElementTree
 
+    try:
+        _etree_iter = etree.Element.iter
+    except AttributeError:  # Python <=2.6
+        def _etree_iter(root):
+            for el in root.findall('*'):
+                yield el
+                for sub in _etree_iter(el):
+                    yield sub
+
     # on 2.6 XML doesn't have a parser argument, function copied from CPython
     # 2.7 source
     def _XML(text, parser=None):
@@ -231,11 +245,16 @@ else:
     def _element_factory(*args, **kwargs):
         el = etree.Element(*args, **kwargs)
         for k, v in el.items():
-            el.set(k, v.decode('utf-8'))
+            if isinstance(v, bytes):
+                el.set(k, v.decode('utf-8'))
         return el
 
     def compat_etree_fromstring(text):
-        return _XML(text, parser=etree.XMLParser(target=etree.TreeBuilder(element_factory=_element_factory)))
+        doc = _XML(text, parser=etree.XMLParser(target=etree.TreeBuilder(element_factory=_element_factory)))
+        for el in _etree_iter(doc):
+            if el.text is not None and isinstance(el.text, bytes):
+                el.text = el.text.decode('utf-8')
+        return doc
 
 try:
     from urllib.parse import parse_qs as compat_parse_qs
@@ -254,7 +273,7 @@ except ImportError:  # Python 2
             nv = name_value.split('=', 1)
             if len(nv) != 2:
                 if strict_parsing:
-                    raise ValueError("bad query field: %r" % (name_value,))
+                    raise ValueError('bad query field: %r' % (name_value,))
                 # Handle case of a control-name with no equal sign
                 if keep_blank_values:
                     nv.append('')
@@ -312,6 +331,9 @@ def compat_ord(c):
         return ord(c)
 
 
+compat_os_name = os._name if os.name == 'java' else os.name
+
+
 if sys.version_info >= (3, 0):
     compat_getenv = os.getenv
     compat_expanduser = os.path.expanduser
@@ -332,7 +354,7 @@ else:
     # The following are os.path.expanduser implementations from cpython 2.7.8 stdlib
     # for different platforms with correct environment variables decoding.
 
-    if os.name == 'posix':
+    if compat_os_name == 'posix':
         def compat_expanduser(path):
             """Expand ~ and ~user constructions.  If user or $HOME is unknown,
             do nothing."""
@@ -356,7 +378,7 @@ else:
                 userhome = pwent.pw_dir
             userhome = userhome.rstrip('/')
             return (userhome + path[i:]) or '/'
-    elif os.name == 'nt' or os.name == 'ce':
+    elif compat_os_name == 'nt' or compat_os_name == 'ce':
         def compat_expanduser(path):
             """Expand ~ and ~user constructs.
 
@@ -419,7 +441,7 @@ if sys.version_info < (3, 0) and sys.platform == 'win32':
 else:
     compat_getpass = getpass.getpass
 
-# Old 2.6 and 2.7 releases require kwargs to be bytes
+# Python < 2.6.5 require kwargs to be bytes
 try:
     def _testfunc(x):
         pass
@@ -452,7 +474,7 @@ if sys.version_info < (2, 7):
         if err is not None:
             raise err
         else:
-            raise socket.error("getaddrinfo returns an empty list")
+            raise socket.error('getaddrinfo returns an empty list')
 else:
     compat_socket_create_connection = socket.create_connection
 
@@ -526,6 +548,7 @@ else:
     from tokenize import generate_tokens as compat_tokenize_tokenize
 
 __all__ = [
+    'compat_HTMLParser',
     'compat_HTTPError',
     'compat_basestring',
     'compat_chr',
@@ -542,6 +565,7 @@ __all__ = [
     'compat_itertools_count',
     'compat_kwargs',
     'compat_ord',
+    'compat_os_name',
     'compat_parse_qs',
     'compat_print',
     'compat_shlex_split',