Merge pull request #7320 from remitamine/adobetv

[youtube-dl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 764a89ccaccc34e9236d84b445f9f2d8342e0244..d0606b4bcd3d4706912f753441608dff721d7699 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -36,6 +36,7 @@ import zlib
 from .compat import (
     compat_basestring,
     compat_chr,
 from .compat import (
     compat_basestring,
     compat_chr,

+    compat_etree_fromstring,

     compat_html_entities,
     compat_http_client,
     compat_kwargs,
     compat_html_entities,
     compat_http_client,
     compat_kwargs,


@@ -365,13 +366,20 @@ def sanitize_path(s):
     if drive_or_unc:
         norm_path.pop(0)
     sanitized_path = [
     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)
 
 
         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 = []
 def orderedSet(iterable):
     """ Remove all duplicates from the input iterable """
     res = []


@@ -395,10 +403,14 @@ def _htmlentity_transform(entity):
             numstr = '0%s' % numstr
         else:
             base = 10
             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
 
     # Unknown entity in name, return its literal representation

-    return ('&%s;' % entity)
+    return '&%s;' % entity

 
 
 def unescapeHTML(s):
 
 
 def unescapeHTML(s):


@@ -651,6 +663,16 @@ def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
     return hc
 
 
     return hc
 
 

+def handle_youtubedl_headers(headers):
+    filtered_headers = headers
+
+    if 'Youtubedl-no-compression' in filtered_headers:
+        filtered_headers = dict((k, v) for k, v in filtered_headers.items() if k.lower() != 'accept-encoding')
+        del filtered_headers['Youtubedl-no-compression']
+
+    return filtered_headers
+
+

 class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
     """Handler for HTTP requests and responses.
 
 class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
     """Handler for HTTP requests and responses.
 


@@ -658,7 +680,7 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
     the standard headers to every HTTP request and handles gzipped and
     deflated responses from web servers. If compression is to be avoided in
     a particular request, the original request in the program code only has
     the standard headers to every HTTP request and handles gzipped and
     deflated responses from web servers. If compression is to be avoided in
     a particular request, the original request in the program code only has

-    to include the HTTP header "Youtubedl-No-Compression", which will be
+    to include the HTTP header "Youtubedl-no-compression", which will be

     removed before making the real request.
 
     Part of this code was copied from:
     removed before making the real request.
 
     Part of this code was copied from:


@@ -719,10 +741,8 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
             # The dict keys are capitalized because of this bug by urllib
             if h.capitalize() not in req.headers:
                 req.add_header(h, v)
             # The dict keys are capitalized because of this bug by urllib
             if h.capitalize() not in req.headers:
                 req.add_header(h, v)

-        if 'Youtubedl-no-compression' in req.headers:
-            if 'Accept-encoding' in req.headers:
-                del req.headers['Accept-encoding']
-            del req.headers['Youtubedl-no-compression']
+
+        req.headers = handle_youtubedl_headers(req.headers)

 
         if sys.version_info < (2, 7) and '#' in req.get_full_url():
             # Python 2.6 is brain-dead when it comes to fragments
 
         if sys.version_info < (2, 7) and '#' in req.get_full_url():
             # Python 2.6 is brain-dead when it comes to fragments


@@ -910,7 +930,8 @@ 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')
         timetuple = email.utils.parsedate_tz(date_str)
         if timetuple:
             upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')

-    return compat_str(upload_date)
+    if upload_date is not None:
+        return compat_str(upload_date)

 
 
 def determine_ext(url, default_ext='unknown_video'):
 
 
 def determine_ext(url, default_ext='unknown_video'):


@@ -919,6 +940,21 @@ def determine_ext(url, default_ext='unknown_video'):
     guess = url.partition('?')[0].rpartition('.')[2]
     if re.match(r'^[A-Za-z0-9]+$', guess):
         return guess
     guess = url.partition('?')[0].rpartition('.')[2]
     if re.match(r'^[A-Za-z0-9]+$', 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
 
     else:
         return default_ext
 


@@ -1662,30 +1698,9 @@ def urlencode_postdata(*args, **kargs):
 
 
 def encode_dict(d, encoding='utf-8'):
 
 
 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 = {
 
 
 US_RATINGS = {


@@ -1988,7 +2003,7 @@ def dfxp2srt(dfxp_data):
 
         return out
 
 
         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')
 
     out = []
     paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall(_x('.//ttaf1:p')) or dfxp.findall('.//p')