youtube-dl/youtube_dl/extractor/generic.py

888 lines
36 KiB
Python
Raw Normal View History

# encoding: utf-8
from __future__ import unicode_literals
2013-06-23 18:31:45 +00:00
import os
import re
from .common import InfoExtractor
from .youtube import YoutubeIE
2013-06-23 18:31:45 +00:00
from ..utils import (
compat_urllib_error,
compat_urllib_parse,
compat_urllib_request,
2013-08-28 10:47:27 +00:00
compat_urlparse,
compat_xml_parse_error,
2013-06-23 18:31:45 +00:00
ExtractorError,
float_or_none,
2013-12-20 16:05:28 +00:00
HEADRequest,
orderedSet,
parse_xml,
smuggle_url,
unescapeHTML,
unified_strdate,
url_basename,
2013-06-23 18:31:45 +00:00
)
from .brightcove import BrightcoveIE
from .ooyala import OoyalaIE
from .rutv import RUTVIE
from .smotri import SmotriIE
2013-06-23 18:31:45 +00:00
2013-06-23 18:31:45 +00:00
class GenericIE(InfoExtractor):
IE_DESC = 'Generic downloader that works on some sites'
2013-06-23 18:31:45 +00:00
_VALID_URL = r'.*'
IE_NAME = 'generic'
_TESTS = [
{
'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
2014-04-14 11:51:46 +00:00
'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
'info_dict': {
2014-04-14 11:51:46 +00:00
'id': '13601338388002',
'ext': 'mp4',
'uploader': 'www.hodiho.fr',
'title': 'R\u00e9gis plante sa Jeep',
}
},
# bandcamp page with custom domain
{
'add_ie': ['Bandcamp'],
'url': 'http://bronyrock.com/track/the-pony-mash',
'info_dict': {
2014-04-14 11:56:29 +00:00
'id': '3235767654',
'ext': 'mp3',
'title': 'The Pony Mash',
'uploader': 'M_Pallante',
},
'skip': 'There is a limit of 200 free downloads / month for the test song',
},
# embedded brightcove video
# it also tests brightcove videos that need to set the 'Referer' in the
# http requests
{
'add_ie': ['Brightcove'],
'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
'info_dict': {
'id': '2765128793001',
'ext': 'mp4',
'title': 'Le cours de bourse : lanalyse technique',
'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
'uploader': 'BFM BUSINESS',
},
'params': {
'skip_download': True,
},
},
2014-01-28 02:35:32 +00:00
{
# https://github.com/rg3/youtube-dl/issues/2253
'url': 'http://bcove.me/i6nfkrc3',
'md5': '0ba9446db037002366bab3b3eb30c88c',
'info_dict': {
2014-04-14 11:56:29 +00:00
'id': '3101154703001',
'ext': 'mp4',
2014-01-28 02:35:32 +00:00
'title': 'Still no power',
'uploader': 'thestar.com',
'description': 'Mississauga resident David Farmer is still out of power as a result of the ice storm a month ago. To keep the house warm, Farmer cuts wood from his property for a wood burning stove downstairs.',
},
'add_ie': ['Brightcove'],
},
{
'url': 'http://www.championat.com/video/football/v/87/87499.html',
'md5': 'fb973ecf6e4a78a67453647444222983',
'info_dict': {
'id': '3414141473001',
'ext': 'mp4',
'title': 'Видео. Удаление Дзагоева (ЦСКА)',
'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
'uploader': 'Championat',
},
},
# Direct link to a video
{
'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
'md5': '67d406c2bcb6af27fa886f31aa934bbe',
'info_dict': {
'id': 'trailer',
'ext': 'mp4',
'title': 'trailer',
'upload_date': '20100513',
}
},
# ooyala video
{
'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
'info_dict': {
'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
'ext': 'mp4',
2014-01-21 00:40:34 +00:00
'title': '2cc213299525360.mov', # that's what we get
},
},
# google redirect
{
'url': 'http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CCUQtwIwAA&url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DcmQHVoWB5FY&ei=F-sNU-LLCaXk4QT52ICQBQ&usg=AFQjCNEw4hL29zgOohLXvpJ-Bdh2bils1Q&bvm=bv.61965928,d.bGE',
'info_dict': {
'id': 'cmQHVoWB5FY',
'ext': 'mp4',
'upload_date': '20130224',
'uploader_id': 'TheVerge',
'description': 'Chris Ziegler takes a look at the Alcatel OneTouch Fire and the ZTE Open; two of the first Firefox OS handsets to be officially announced.',
'uploader': 'The Verge',
'title': 'First Firefox OS phones side-by-side',
},
'params': {
'skip_download': False,
}
},
2014-02-24 00:15:51 +00:00
# embed.ly video
{
'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
'info_dict': {
'id': '9ODmcdjQcHQ',
'ext': 'mp4',
'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
'upload_date': '20140225',
'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
'uploader': 'Tested',
'uploader_id': 'testedcom',
2014-02-24 00:15:51 +00:00
},
# No need to test YoutubeIE here
'params': {
'skip_download': True,
},
},
# funnyordie embed
{
'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
'md5': '7cf780be104d40fea7bae52eed4a470e',
'info_dict': {
'id': '18e820ec3f',
'ext': 'mp4',
'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
},
},
# RUTV embed
{
'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
'info_dict': {
'id': '776940',
'ext': 'mp4',
'title': 'Охотское море стало целиком российским',
'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
},
'params': {
# m3u8 download
'skip_download': True,
},
2014-03-20 15:33:23 +00:00
},
# Embedded TED video
{
'url': 'http://en.support.wordpress.com/videos/ted-talks/',
'md5': 'deeeabcc1085eb2ba205474e7235a3d5',
'info_dict': {
'id': '981',
'ext': 'mp4',
'title': 'My web playroom',
'uploader': 'Ze Frank',
'description': 'md5:ddb2a40ecd6b6a147e400e535874947b',
}
},
# Embeded Ustream video
{
'url': 'http://www.american.edu/spa/pti/nsa-privacy-janus-2014.cfm',
'md5': '27b99cdb639c9b12a79bca876a073417',
'info_dict': {
2014-04-04 16:56:29 +00:00
'id': '45734260',
'ext': 'flv',
'uploader': 'AU SPA: The NSA and Privacy',
'title': 'NSA and Privacy Forum Debate featuring General Hayden and Barton Gellman'
}
},
# nowvideo embed hidden behind percent encoding
{
'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
'md5': '2baf4ddd70f697d94b1c18cf796d5107',
'info_dict': {
'id': '06e53103ca9aa',
'ext': 'flv',
'title': 'Macross Episode 001 Watch Macross Episode 001 onl',
'description': 'No description',
},
},
# arte embed
{
'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
'md5': '7653032cbb25bf6c80d80f217055fa43',
'info_dict': {
'id': '048195-004_PLUS7-F',
'ext': 'flv',
'title': 'X:enius',
'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
'upload_date': '20140320',
},
'params': {
'skip_download': 'Requires rtmpdump'
}
},
# smotri embed
{
'url': 'http://rbctv.rbc.ru/archive/news/562949990879132.shtml',
'md5': 'ec40048448e9284c9a1de77bb188108b',
'info_dict': {
'id': 'v27008541fad',
'ext': 'mp4',
'title': 'Крым и Севастополь вошли в состав России',
'description': 'md5:fae01b61f68984c7bd2fa741e11c3175',
'duration': 900,
'upload_date': '20140318',
'uploader': 'rbctv_2012_4',
'uploader_id': 'rbctv_2012_4',
},
},
# Condé Nast embed
{
'url': 'http://www.wired.com/2014/04/honda-asimo/',
'md5': 'ba0dfe966fa007657bd1443ee672db0f',
'info_dict': {
'id': '53501be369702d3275860000',
'ext': 'mp4',
'title': 'Hondas New Asimo Robot Is More Human Than Ever',
}
},
# Dailymotion embed
{
'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
'md5': '441aeeb82eb72c422c7f14ec533999cd',
'info_dict': {
'id': 'k2mm4bCdJ6CQ2i7c8o2',
'ext': 'mp4',
'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
'uploader': 'Spi0n',
},
'add_ie': ['Dailymotion'],
},
# YouTube embed
{
'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
'info_dict': {
'id': 'FXRb4ykk4S0',
'ext': 'mp4',
'title': 'The NBL Auction 2014',
'uploader': 'BADMINTON England',
'uploader_id': 'BADMINTONEvents',
'upload_date': '20140603',
'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
},
'add_ie': ['Youtube'],
'params': {
'skip_download': True,
}
},
# MTVSercices embed
{
'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
'md5': '35727f82f58c76d996fc188f9755b0d5',
'info_dict': {
'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
'ext': 'mp4',
'title': 'Review',
'description': 'Mario\'s life in the fast lane has never looked so good.',
},
},
2014-05-21 09:55:37 +00:00
# YouTube embed via <data-embed-url="">
{
'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
'info_dict': {
'id': 'jpSGZsgga_I',
2014-05-21 09:55:37 +00:00
'ext': 'mp4',
'title': 'Asphalt 8: Airborne - Launch Trailer',
'uploader': 'Gameloft',
'uploader_id': 'gameloft',
'upload_date': '20130821',
'description': 'md5:87bd95f13d8be3e7da87a5f2c443106a',
},
'params': {
'skip_download': True,
2014-05-21 09:55:37 +00:00
}
},
# Camtasia studio
{
'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
'playlist': [{
'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
'info_dict': {
'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
'ext': 'flv',
'duration': 2235.90,
}
}, {
'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
'info_dict': {
'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
'ext': 'flv',
'duration': 2235.93,
}
}],
'info_dict': {
'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
}
}
]
2013-06-23 18:31:45 +00:00
def report_download_webpage(self, video_id):
"""Report webpage download."""
if not self._downloader.params.get('test', False):
self._downloader.report_warning('Falling back on generic information extractor.')
2013-06-23 18:31:45 +00:00
super(GenericIE, self).report_download_webpage(video_id)
def report_following_redirect(self, new_url):
"""Report information extraction."""
self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
2013-06-23 18:31:45 +00:00
def _send_head(self, url):
2013-06-23 18:31:45 +00:00
"""Check if it is a redirect, like url shorteners, in case return the new url."""
class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
"""
Subclass the HTTPRedirectHandler to make it use our
2013-12-20 16:05:28 +00:00
HEADRequest also on the redirected URL
2013-06-23 18:31:45 +00:00
"""
def redirect_request(self, req, fp, code, msg, headers, newurl):
if code in (301, 302, 303, 307):
newurl = newurl.replace(' ', '%20')
newheaders = dict((k,v) for k,v in req.headers.items()
if k.lower() not in ("content-length", "content-type"))
try:
# This function was deprecated in python 3.3 and removed in 3.4
origin_req_host = req.get_origin_req_host()
except AttributeError:
origin_req_host = req.origin_req_host
2013-12-20 16:05:28 +00:00
return HEADRequest(newurl,
2013-06-23 18:31:45 +00:00
headers=newheaders,
origin_req_host=origin_req_host,
2013-06-23 18:31:45 +00:00
unverifiable=True)
else:
raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
class HTTPMethodFallback(compat_urllib_request.BaseHandler):
"""
Fallback to GET if HEAD is not allowed (405 HTTP error)
"""
def http_error_405(self, req, fp, code, msg, headers):
fp.read()
fp.close()
newheaders = dict((k,v) for k,v in req.headers.items()
if k.lower() not in ("content-length", "content-type"))
return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
headers=newheaders,
origin_req_host=req.get_origin_req_host(),
unverifiable=True))
# Build our opener
opener = compat_urllib_request.OpenerDirector()
for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
HTTPMethodFallback, HEADRedirectHandler,
compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
opener.add_handler(handler())
2013-12-20 16:05:28 +00:00
response = opener.open(HEADRequest(url))
2013-06-23 18:31:45 +00:00
if response is None:
raise ExtractorError('Invalid URL protocol')
return response
2013-06-23 18:31:45 +00:00
def _extract_rss(self, url, video_id, doc):
playlist_title = doc.find('./channel/title').text
playlist_desc_el = doc.find('./channel/description')
playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
entries = [{
'_type': 'url',
'url': e.find('link').text,
'title': e.find('title').text,
} for e in doc.findall('./channel/item')]
return {
'_type': 'playlist',
'id': url,
'title': playlist_title,
'description': playlist_desc,
'entries': entries,
}
def _extract_camtasia(self, url, video_id, webpage):
""" Returns None if no camtasia video can be found. """
camtasia_cfg = self._search_regex(
r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
webpage, 'camtasia configuration file', default=None)
if camtasia_cfg is None:
return None
title = self._html_search_meta('DC.title', webpage, fatal=True)
camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
camtasia_cfg = self._download_xml(
camtasia_url, video_id,
note='Downloading camtasia configuration',
errnote='Failed to download camtasia configuration')
fileset_node = camtasia_cfg.find('./playlist/array/fileset')
entries = []
for n in fileset_node.getchildren():
url_n = n.find('./uri')
if url_n is None:
continue
entries.append({
'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
'title': '%s - %s' % (title, n.tag),
'url': compat_urlparse.urljoin(url, url_n.text),
'duration': float_or_none(n.find('./duration').text),
})
return {
'_type': 'playlist',
'entries': entries,
'title': title,
}
2013-06-23 18:31:45 +00:00
def _real_extract(self, url):
if url.startswith('//'):
return {
'_type': 'url',
'url': self.http_scheme() + url,
}
parsed_url = compat_urlparse.urlparse(url)
if not parsed_url.scheme:
default_search = self._downloader.params.get('default_search')
if default_search is None:
default_search = 'fixup_error'
if default_search in ('auto', 'auto_warning', 'fixup_error'):
if '/' in url:
self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
return self.url_result('http://' + url)
elif default_search != 'fixup_error':
if default_search == 'auto_warning':
if re.match(r'^(?:url|URL)$', url):
raise ExtractorError(
'Invalid URL: %r . Call youtube-dl like this: youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc" ' % url,
expected=True)
else:
self._downloader.report_warning(
'Falling back to youtube search for %s . Set --default-search "auto" to suppress this warning.' % url)
return self.url_result('ytsearch:' + url)
if default_search in ('error', 'fixup_error'):
raise ExtractorError(
('%r is not a valid URL. '
'Set --default-search "ytsearch" (or run youtube-dl "ytsearch:%s" ) to search YouTube'
) % (url, url), expected=True)
else:
assert ':' in default_search
return self.url_result(default_search + url)
2014-03-05 13:02:14 +00:00
video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
self.to_screen('%s: Requesting header' % video_id)
try:
response = self._send_head(url)
# Check for redirect
new_url = response.geturl()
if url != new_url:
self.report_following_redirect(new_url)
return self.url_result(new_url)
# Check for direct link to a video
content_type = response.headers.get('Content-Type', '')
m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
if m:
upload_date = response.headers.get('Last-Modified')
if upload_date:
upload_date = unified_strdate(upload_date)
return {
'id': video_id,
'title': os.path.splitext(url_basename(url))[0],
'formats': [{
'format_id': m.group('format_id'),
'url': url,
'vcodec': 'none' if m.group('type') == 'audio' else None
}],
'upload_date': upload_date,
}
except compat_urllib_error.HTTPError:
# This may be a stupid server that doesn't like HEAD, our UA, or so
pass
2013-06-23 18:31:45 +00:00
try:
webpage = self._download_webpage(url, video_id)
except ValueError:
# since this is the last-resort InfoExtractor, if
# this error is thrown, it'll be thrown here
raise ExtractorError('Failed to download URL: %s' % url)
2013-06-23 18:31:45 +00:00
self.report_extraction(video_id)
# Is it an RSS feed?
try:
doc = parse_xml(webpage)
if doc.tag == 'rss':
return self._extract_rss(url, video_id, doc)
except compat_xml_parse_error:
pass
# Is it a Camtasia project?
camtasia_res = self._extract_camtasia(url, video_id, webpage)
if camtasia_res is not None:
return camtasia_res
# Sometimes embedded video player is hidden behind percent encoding
# (e.g. https://github.com/rg3/youtube-dl/issues/2448)
# Unescaping the whole page allows to handle those cases in a generic way
2014-02-24 16:44:31 +00:00
webpage = compat_urllib_parse.unquote(webpage)
# it's tempting to parse this further, but you would
# have to take into account all the variations like
# Video Title - Site Name
# Site Name | Video Title
# Video Title - Tagline | Site Name
# and so on and so forth; it's just not practical
2013-12-06 08:15:04 +00:00
video_title = self._html_search_regex(
r'(?s)<title>(.*?)</title>', webpage, 'video title',
default='video')
2013-12-06 08:15:04 +00:00
# video uploader is domain name
video_uploader = self._search_regex(
r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
# Helper method
def _playlist_from_matches(matches, getter, ie=None):
urlrs = orderedSet(self.url_result(getter(m), ie) for m in matches)
return self.playlist_result(
urlrs, playlist_id=video_id, playlist_title=video_title)
2013-08-26 19:29:31 +00:00
# Look for BrightCove:
bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
if bc_urls:
self.to_screen('Brightcove video detected.')
entries = [{
'_type': 'url',
'url': smuggle_url(bc_url, {'Referer': url}),
'ie_key': 'Brightcove'
} for bc_url in bc_urls]
return {
'_type': 'playlist',
'title': video_title,
'id': video_id,
'entries': entries,
}
# Look for embedded (iframe) Vimeo player
mobj = re.search(
r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
if mobj:
player_url = unescapeHTML(mobj.group('url'))
surl = smuggle_url(player_url, {'Referer': url})
return self.url_result(surl, 'Vimeo')
# Look for embedded (swf embed) Vimeo player
mobj = re.search(
r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
if mobj:
return self.url_result(mobj.group(1), 'Vimeo')
# Look for embedded YouTube player
matches = re.findall(r'''(?x)
(?:
<iframe[^>]+?src=|
data-video-url=|
<embed[^>]+?src=|
embedSWF\(?:\s*
)
(["\'])
(?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
(?:embed|v)/.+?)
\1''', webpage)
if matches:
return _playlist_from_matches(
matches, lambda m: unescapeHTML(m[1]), ie='Youtube')
# Look for embedded Dailymotion player
matches = re.findall(
2013-12-06 08:15:04 +00:00
r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
if matches:
return _playlist_from_matches(
matches, lambda m: unescapeHTML(m[1]))
2013-12-06 08:15:04 +00:00
# Look for embedded Wistia player
match = re.search(
r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
if match:
return {
'_type': 'url_transparent',
'url': unescapeHTML(match.group('url')),
'ie_key': 'Wistia',
'uploader': video_uploader,
'title': video_title,
'id': video_id,
}
# Look for embedded blip.tv player
mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
if mobj:
return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage)
if mobj:
return self.url_result(mobj.group(1), 'BlipTV')
# Look for embedded condenast player
matches = re.findall(
r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
webpage)
if matches:
return {
'_type': 'playlist',
'entries': [{
'_type': 'url',
'ie_key': 'CondeNast',
'url': ma,
} for ma in matches],
'title': video_title,
'id': video_id,
}
# Look for Bandcamp pages with custom domain
mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
if mobj is not None:
burl = unescapeHTML(mobj.group(1))
# Don't set the extractor because it can be a track url or an album
return self.url_result(burl)
# Look for embedded Vevo player
mobj = re.search(
r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
if mobj is not None:
return self.url_result(mobj.group('url'))
# Look for Ooyala videos
mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
if mobj is not None:
return OoyalaIE._build_url_result(mobj.group('ec'))
2013-12-20 16:05:28 +00:00
# Look for Aparat videos
mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
2013-12-20 16:05:28 +00:00
if mobj is not None:
return self.url_result(mobj.group(1), 'Aparat')
2014-01-07 07:07:46 +00:00
# Look for MPORA videos
mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
2014-01-07 07:07:46 +00:00
if mobj is not None:
return self.url_result(mobj.group(1), 'Mpora')
2014-01-08 01:11:46 +00:00
# Look for embedded NovaMov-based player
2014-01-08 01:07:11 +00:00
mobj = re.search(
r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
(?P<url>http://(?:(?:embed|www)\.)?
(?:novamov\.com|
nowvideo\.(?:ch|sx|eu|at|ag|co)|
videoweed\.(?:es|com)|
movshare\.(?:net|sx|ag)|
divxstage\.(?:eu|net|ch|co|at|ag))
/embed\.php.+?)\1''', webpage)
2014-01-08 01:07:11 +00:00
if mobj is not None:
return self.url_result(mobj.group('url'))
# Look for embedded Facebook player
mobj = re.search(
2014-01-27 04:47:30 +00:00
r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
if mobj is not None:
return self.url_result(mobj.group('url'), 'Facebook')
# Look for embedded VK player
mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
if mobj is not None:
return self.url_result(mobj.group('url'), 'VK')
# Look for embedded ivi player
mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
if mobj is not None:
return self.url_result(mobj.group('url'), 'Ivi')
2014-01-27 04:47:30 +00:00
# Look for embedded Huffington Post player
mobj = re.search(
r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
2014-01-27 04:47:30 +00:00
if mobj is not None:
return self.url_result(mobj.group('url'), 'HuffPost')
2014-02-24 00:15:51 +00:00
# Look for embed.ly
mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
if mobj is not None:
return self.url_result(mobj.group('url'))
mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
if mobj is not None:
return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
# Look for funnyordie embed
matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
if matches:
return _playlist_from_matches(
matches, getter=unescapeHTML, ie='FunnyOrDie')
# Look for embedded RUTV player
rutv_url = RUTVIE._extract_url(webpage)
if rutv_url:
return self.url_result(rutv_url, 'RUTV')
# Look for embedded TED player
mobj = re.search(
r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
if mobj is not None:
return self.url_result(mobj.group('url'), 'TED')
# Look for embedded Ustream videos
mobj = re.search(
r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
if mobj is not None:
return self.url_result(mobj.group('url'), 'Ustream')
# Look for embedded arte.tv player
mobj = re.search(
r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
webpage)
if mobj is not None:
return self.url_result(mobj.group('url'), 'ArteTVEmbed')
# Look for embedded smotri.com player
smotri_url = SmotriIE._extract_url(webpage)
if smotri_url:
return self.url_result(smotri_url, 'Smotri')
# Look for embeded soundcloud player
mobj = re.search(
r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
webpage)
if mobj is not None:
url = unescapeHTML(mobj.group('url'))
return self.url_result(url)
2014-06-09 21:06:25 +00:00
# Look for embedded vulture.com player
mobj = re.search(
r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
webpage)
if mobj is not None:
url = unescapeHTML(mobj.group('url'))
return self.url_result(url, ie='Vulture')
# Look for embedded mtvservices player
mobj = re.search(
r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
webpage)
if mobj is not None:
url = unescapeHTML(mobj.group('url'))
return self.url_result(url, ie='MTVServicesEmbedded')
# Look for embedded yahoo player
mobj = re.search(
r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
webpage)
if mobj is not None:
return self.url_result(mobj.group('url'), 'Yahoo')
2014-08-23 13:20:49 +00:00
# Look for embedded sbs.com.au player
mobj = re.search(
r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
webpage)
if mobj is not None:
return self.url_result(mobj.group('url'), 'SBS')
2013-06-23 18:31:45 +00:00
# Start with something easy: JW Player in SWFObject
found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
if not found:
# Look for gorilla-vid style embedding
found = re.findall(r'''(?sx)
(?:
jw_plugins|
JWPlayerOptions|
jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
)
.*?file\s*:\s*["\'](.*?)["\']''', webpage)
if not found:
2013-06-23 18:31:45 +00:00
# Broaden the search a little bit
found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
if not found:
# Broaden the findall a little bit: JWPlayer JS loader
found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
if not found:
2013-06-23 18:31:45 +00:00
# Try to find twitter cards info
found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
if not found:
2013-06-23 18:31:45 +00:00
# We look for Open Graph info:
# We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
2013-06-23 18:31:45 +00:00
# We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
if m_video_type is not None:
def check_video(vurl):
vpath = compat_urlparse.urlparse(vurl).path
return '.' in vpath and not vpath.endswith('.swf')
found = list(filter(
check_video,
re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)))
if not found:
2013-08-21 02:32:22 +00:00
# HTML5 video
found = re.findall(r'(?s)<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage)
if not found:
2014-05-16 13:32:53 +00:00
found = re.search(
r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
webpage)
if found:
new_url = found.group(1)
self.report_following_redirect(new_url)
return {
'_type': 'url',
'url': new_url,
}
if not found:
raise ExtractorError('Unsupported URL: %s' % url)
2013-06-23 18:31:45 +00:00
entries = []
for video_url in found:
video_url = compat_urlparse.urljoin(url, video_url)
video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
2013-06-23 18:31:45 +00:00
# Sometimes, jwplayer extraction will result in a YouTube URL
if YoutubeIE.suitable(video_url):
entries.append(self.url_result(video_url, 'Youtube'))
continue
2013-06-23 18:31:45 +00:00
# here's a fun little line of code for you:
video_id = os.path.splitext(video_id)[0]
entries.append({
'id': video_id,
'url': video_url,
'uploader': video_uploader,
'title': video_title,
})
if len(entries) == 1:
2014-05-01 09:28:37 +00:00
return entries[0]
else:
for num, e in enumerate(entries, start=1):
e['title'] = '%s (%d)' % (e['title'], num)
return {
'_type': 'playlist',
'entries': entries,
}
2013-06-23 18:31:45 +00:00