"""Report information extraction."""
self.to_screen(u'%s: Extracting information' % id_or_name)
+ def report_download_webpage(self, video_id):
+ """Report webpage download."""
+ self.to_screen(u'%s: Downloading webpage' % video_id)
+
def report_age_confirmation(self):
"""Report attempt to confirm age."""
self.to_screen(u'Confirming age')
if video_description:
video_description = clean_html(video_description)
else:
- video_description = u''
+ fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage)
+ if fd_mobj:
+ video_description = unescapeHTML(fd_mobj.group(1))
+ else:
+ video_description = u''
# subtitles
video_subtitles = None
"""Report disclaimer retrieval."""
self.to_screen(u'Retrieving disclaimer')
- def report_download_webpage(self, video_id):
- """Report webpage download."""
- self.to_screen(u'%s: Downloading webpage' % video_id)
-
def _real_initialize(self):
# Retrieve disclaimer
request = compat_urllib_request.Request(self._DISCLAIMER)
def __init__(self, downloader=None):
InfoExtractor.__init__(self, downloader)
- def report_download_webpage(self, video_id):
- """Report webpage download."""
- self.to_screen(u'%s: Downloading webpage' % video_id)
-
def _real_extract(self, url):
# Extract id from URL
mobj = re.match(self._VALID_URL, url)
def __init__(self, downloader=None):
InfoExtractor.__init__(self, downloader)
- def report_download_webpage(self, video_id):
- """Report webpage download."""
- self.to_screen(u'%s: Downloading webpage' % video_id)
-
def _real_extract(self, url, new_video=True):
# Extract ID from URL
mobj = re.match(self._VALID_URL, url)
def __init__(self, downloader=None):
InfoExtractor.__init__(self, downloader)
- def report_download_webpage(self, video_id):
- """Report webpage download."""
- self.to_screen(u'%s: Downloading webpage' % video_id)
-
def _real_extract(self, url, new_video=True):
# Extract ID from URL
mobj = re.match(self._VALID_URL, url)
def __init__(self, downloader=None):
InfoExtractor.__init__(self, downloader)
- def report_download_webpage(self, video_id):
- """Report webpage download."""
- self.to_screen(u'%s: Downloading webpage' % video_id)
-
def fetch_webpage(self, url):
request = compat_urllib_request.Request(url)
try:
_VALID_URL = r'(?:http://)?(?:\w+\.)?depositfiles\.com/(?:../(?#locale))?files/(.+)'
- def report_download_webpage(self, file_id):
- """Report webpage download."""
- self.to_screen(u'%s: Downloading webpage' % file_id)
-
def _real_extract(self, url):
file_id = url.split('/')[-1]
# Rebuild url in english locale
self.report_extraction(url)
# Extract video URL
- mobj = re.search(r"jsclassref='([^']*)'", webpage)
+ mobj = re.search(r"jsclassref ?= ?'([^']*)'", webpage)
if mobj is None:
self._downloader.report_error(u'unable to extract video url')
return
_VALID_URL = r'^(?:https?://)?openclassroom.stanford.edu(?P<path>/?|(/MainFolder/(?:HomePage|CoursePage|VideoPage)\.php([?]course=(?P<course>[^&]+)(&video=(?P<video>[^&]+))?(&.*)?)?))$'
IE_NAME = u'stanfordoc'
- def report_download_webpage(self, objid):
- """Report information extraction."""
- self.to_screen(u'%s: Downloading webpage' % objid)
-
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
if mobj is None:
class YoukuIE(InfoExtractor):
_VALID_URL = r'(?:http://)?v\.youku\.com/v_show/id_(?P<ID>[A-Za-z0-9]+)\.html'
- def report_download_webpage(self, file_id):
- """Report webpage download."""
- self.to_screen(u'%s: Downloading webpage' % file_id)
-
def _gen_sid(self):
nowTime = int(time.time() * 1000)
random1 = random.randint(1000,1998)
VIDEO_TITLE_RE = r'<title>(.*?)\s+-\s+XNXX.COM'
VIDEO_THUMB_RE = r'url_bigthumb=(.*?)&'
- def report_webpage(self, video_id):
- """Report information extraction"""
- self.to_screen(u'%s: Downloading webpage' % video_id)
-
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
if mobj is None:
return
video_id = mobj.group(1)
- self.report_webpage(video_id)
+ self.report_download_webpage(video_id)
# Get webpage content
try: