X-Git-Url: http://git.cielonegro.org/gitweb.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=d7b737e21639679fe665d22e6ca4089f5c791618;hb=67dda51722f1ce12b956782d43047b3fff390115;hp=558c9c7d5a21c646a11720221fd1648c137ac68e;hpb=52c3a6e49d2cbc1932992d816d28bbed629daadc;p=youtube-dl.git

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 558c9c7d5..d7b737e21 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -36,6 +36,7 @@ import zlib
 from .compat import (
     compat_basestring,
     compat_chr,
+    compat_etree_fromstring,
     compat_html_entities,
     compat_http_client,
     compat_kwargs,
@@ -178,10 +179,19 @@ def xpath_with_ns(path, ns_map):
 
 
 def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
-    if sys.version_info < (2, 7):  # Crazy 2.6
-        xpath = xpath.encode('ascii')
+    def _find_xpath(xpath):
+        if sys.version_info < (2, 7):  # Crazy 2.6
+            xpath = xpath.encode('ascii')
+        return node.find(xpath)
+
+    if isinstance(xpath, (str, compat_str)):
+        n = _find_xpath(xpath)
+    else:
+        for xp in xpath:
+            n = _find_xpath(xp)
+            if n is not None:
+                break
 
-    n = node.find(xpath)
     if n is None:
         if default is not NO_DEFAULT:
             return default
@@ -356,13 +366,20 @@ def sanitize_path(s):
     if drive_or_unc:
         norm_path.pop(0)
     sanitized_path = [
-        path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|\.$)', '#', path_part)
+        path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|[\s.]$)', '#', path_part)
         for path_part in norm_path]
     if drive_or_unc:
         sanitized_path.insert(0, drive_or_unc + os.path.sep)
     return os.path.join(*sanitized_path)
 
 
+# Prepend protocol-less URLs with `http:` scheme in order to mitigate the number of
+# unwanted failures due to missing protocol
+def sanitized_Request(url, *args, **kwargs):
+    return compat_urllib_request.Request(
+        'http:%s' % url if url.startswith('//') else url, *args, **kwargs)
+
+
 def orderedSet(iterable):
     """ Remove all duplicates from the input iterable """
     res = []
@@ -386,10 +403,14 @@ def _htmlentity_transform(entity):
             numstr = '0%s' % numstr
         else:
             base = 10
-        return compat_chr(int(numstr, base))
+        # See https://github.com/rg3/youtube-dl/issues/7518
+        try:
+            return compat_chr(int(numstr, base))
+        except ValueError:
+            pass
 
     # Unknown entity in name, return its literal representation
-    return ('&%s;' % entity)
+    return '&%s;' % entity
 
 
 def unescapeHTML(s):
@@ -901,7 +922,8 @@ def unified_strdate(date_str, day_first=True):
         timetuple = email.utils.parsedate_tz(date_str)
         if timetuple:
             upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
-    return upload_date
+    if upload_date is not None:
+        return compat_str(upload_date)
 
 
 def determine_ext(url, default_ext='unknown_video'):
@@ -910,6 +932,21 @@ def determine_ext(url, default_ext='unknown_video'):
     guess = url.partition('?')[0].rpartition('.')[2]
     if re.match(r'^[A-Za-z0-9]+$', guess):
         return guess
+    elif guess.rstrip('/') in (
+            'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
+            'flv', 'f4v', 'f4a', 'f4b',
+            'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
+            'mkv', 'mka', 'mk3d',
+            'avi', 'divx',
+            'mov',
+            'asf', 'wmv', 'wma',
+            '3gp', '3g2',
+            'mp3',
+            'flac',
+            'ape',
+            'wav',
+            'f4f', 'f4m', 'm3u8', 'smil'):
+        return guess.rstrip('/')
     else:
         return default_ext
 
@@ -1653,30 +1690,9 @@ def urlencode_postdata(*args, **kargs):
 
 
 def encode_dict(d, encoding='utf-8'):
-    return dict((k.encode(encoding), v.encode(encoding)) for k, v in d.items())
-
-
-try:
-    etree_iter = xml.etree.ElementTree.Element.iter
-except AttributeError:  # Python <=2.6
-    etree_iter = lambda n: n.findall('.//*')
-
-
-def parse_xml(s):
-    class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
-        def doctype(self, name, pubid, system):
-            pass  # Ignore doctypes
-
-    parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
-    kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
-    tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
-    # Fix up XML parser in Python 2.x
-    if sys.version_info < (3, 0):
-        for n in etree_iter(tree):
-            if n.text is not None:
-                if not isinstance(n.text, compat_str):
-                    n.text = n.text.decode('utf-8')
-    return tree
+    def encode(v):
+        return v.encode(encoding) if isinstance(v, compat_basestring) else v
+    return dict((encode(k), encode(v)) for k, v in d.items())
 
 
 US_RATINGS = {
@@ -1979,7 +1995,7 @@ def dfxp2srt(dfxp_data):
 
         return out
 
-    dfxp = xml.etree.ElementTree.fromstring(dfxp_data.encode('utf-8'))
+    dfxp = compat_etree_fromstring(dfxp_data.encode('utf-8'))
     out = []
     paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall('.//p')