[yt-dlp.git] / youtube_dl / extractor / generic.py

# encoding: utf-8

from __future__ import unicode_literals

import os
import re

from .common import InfoExtractor
from .youtube import YoutubeIE
from ..utils import (
    compat_urllib_error,
    compat_urllib_parse,
    compat_urllib_request,
    compat_urlparse,

    ExtractorError,
    HEADRequest,
    smuggle_url,
    unescapeHTML,
    unified_strdate,
    url_basename,
)
from .brightcove import BrightcoveIE
from .ooyala import OoyalaIE


class GenericIE(InfoExtractor):
    IE_DESC = 'Generic downloader that works on some sites'
    _VALID_URL = r'.*'
    IE_NAME = 'generic'
    _TESTS = [
        {
            'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
            'file': '13601338388002.mp4',
            'md5': '6e15c93721d7ec9e9ca3fdbf07982cfd',
            'info_dict': {
                'uploader': 'www.hodiho.fr',
                'title': 'R\u00e9gis plante sa Jeep',
            }
        },
        # embedded vimeo video
        {
            'add_ie': ['Vimeo'],
            'url': 'http://skillsmatter.com/podcast/home/move-semanticsperfect-forwarding-and-rvalue-references',
            'file': '22444065.mp4',
            'md5': '2903896e23df39722c33f015af0666e2',
            'info_dict': {
                'title': 'ACCU 2011: Move Semantics,Perfect Forwarding, and Rvalue references- Scott Meyers- 13/04/2011',
                'uploader_id': 'skillsmatter',
                'uploader': 'Skills Matter',
            }
        },
        # bandcamp page with custom domain
        {
            'add_ie': ['Bandcamp'],
            'url': 'http://bronyrock.com/track/the-pony-mash',
            'file': '3235767654.mp3',
            'info_dict': {
                '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 : l’analyse technique',
                'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
                'uploader': 'BFM BUSINESS',
            },
            'params': {
                'skip_download': True,
            },
        },
        # Direct link to a video
        {
            'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
            'file': 'trailer.mp4',
            'md5': '67d406c2bcb6af27fa886f31aa934bbe',
            'info_dict': {
                'id': 'trailer',
                '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',
            'file': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ.mp4',
            'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
            'info_dict': {
                'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
                'ext': 'mp4',
                'title': '2cc213299525360.mov',  # that's what we get
            },
        },
    ]

    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.')
        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)

    def _send_head(self, url):
        """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
            HEADRequest also on the redirected URL
            """
            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"))
                    return HEADRequest(newurl,
                                       headers=newheaders,
                                       origin_req_host=req.get_origin_req_host(),
                                       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())

        response = opener.open(HEADRequest(url))
        if response is None:
            raise ExtractorError('Invalid URL protocol')
        return response

    def _real_extract(self, url):
        parsed_url = compat_urlparse.urlparse(url)
        if not parsed_url.scheme:
            self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
            return self.url_result('http://' + url)
        video_id = os.path.splitext(url.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

        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)

        self.report_extraction(video_id)

        # 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
        video_title = self._html_search_regex(
            r'(?s)<title>(.*?)</title>', webpage, 'video title',
            default='video')

        # video uploader is domain name
        video_uploader = self._search_regex(
            r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')

        # Look for BrightCove:
        bc_url = BrightcoveIE._extract_brightcove_url(webpage)
        if bc_url is not None:
            self.to_screen('Brightcove video detected.')
            surl = smuggle_url(bc_url, {'Referer': url})
            return self.url_result(surl, 'Brightcove')

        # Look for embedded (iframe) Vimeo player
        mobj = re.search(
            r'<iframe[^>]+?src="((?:https?:)?//player.vimeo.com/video/.+?)"', webpage)
        if mobj:
            player_url = unescapeHTML(mobj.group(1))
            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=|embedSWF\(\s*)
            (["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
                (?:embed|v)/.+?)
            \1''', webpage)
        if matches:
            urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
                     for tuppl in matches]
            return self.playlist_result(
                urlrs, playlist_id=video_id, playlist_title=video_title)

        # Look for embedded Dailymotion player
        matches = re.findall(
            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
        if matches:
            urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Dailymotion')
                     for tuppl in matches]
            return self.playlist_result(
                urlrs, playlist_id=video_id, playlist_title=video_title)

        # 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 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)=([^"&]+)', webpage)
        if mobj is not None:
            return OoyalaIE._build_url_result(mobj.group(1))

        # Look for Aparat videos
        mobj = re.search(r'<iframe src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
        if mobj is not None:
            return self.url_result(mobj.group(1), 'Aparat')

        # Look for MPORA videos
        mobj = re.search(r'<iframe .*?src="(http://mpora\.com/videos/[^"]+)"', webpage)
        if mobj is not None:
            return self.url_result(mobj.group(1), 'Mpora')

        # Look for embedded Novamov player
        mobj = re.search(
            r'<iframe[^>]+?src=(["\'])(?P<url>http://(?:(?:embed|www)\.)?novamov\.com/embed\.php.+?)\1', webpage)
        if mobj is not None:
            return self.url_result(mobj.group('url'), 'Novamov')

        # Look for embedded Facebook player
        mobj = re.search(
            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')

        # Start with something easy: JW Player in SWFObject
        mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
        if mobj is None:
            # Look for gorilla-vid style embedding
            mobj = re.search(r'(?s)jw_plugins.*?file:\s*["\'](.*?)["\']', webpage)
        if mobj is None:
            # Broaden the search a little bit
            mobj = re.search(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
        if mobj is None:
            # Broaden the search a little bit: JWPlayer JS loader
            mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
        if mobj is None:
            # Try to find twitter cards info
            mobj = re.search(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
        if mobj is None:
            # 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.search(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
            # 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:
                mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
        if mobj is None:
            # HTML5 video
            mobj = re.search(r'<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage, flags=re.DOTALL)
        if mobj is None:
            raise ExtractorError('Unsupported URL: %s' % url)

        # It's possible that one of the regexes
        # matched, but returned an empty group:
        if mobj.group(1) is None:
            raise ExtractorError('Did not find a valid video URL at %s' % url)

        video_url = mobj.group(1)
        video_url = compat_urlparse.urljoin(url, video_url)
        video_id = compat_urllib_parse.unquote(os.path.basename(video_url))

        # Sometimes, jwplayer extraction will result in a YouTube URL
        if YoutubeIE.suitable(video_url):
            return self.url_result(video_url, 'Youtube')

        # here's a fun little line of code for you:
        video_id = os.path.splitext(video_id)[0]

        return {
            'id': video_id,
            'url': video_url,
            'uploader': video_uploader,
            'title': video_title,
        }
Commit	Line	Data
cfe50f04 JMF	1	# encoding: utf-8
cfe50f04 JMF	2
79649588 PH	3	from __future__ import unicode_literals
79649588 PH	4
9b122384 PH	5	import os
	6	import re
	7
	8	from .common import InfoExtractor
fc9713a1	9	from .youtube import YoutubeIE
9b122384 PH	10	from ..utils import (
	11	compat_urllib_error,
	12	compat_urllib_parse,
	13	compat_urllib_request,
a5caba1e	14	compat_urlparse,
9b122384 PH	15
9b122384 PH	16	ExtractorError,
aa94a6d3	17	HEADRequest,
9d4660ca PH	18	smuggle_url,
9d4660ca PH	19	unescapeHTML,
42393ce2 PH	20	unified_strdate,
42393ce2 PH	21	url_basename,
9b122384	22	)
cfe50f04	23	from .brightcove import BrightcoveIE
c0d0b01f	24	from .ooyala import OoyalaIE
9b122384	25
0838239e	26
9b122384	27	class GenericIE(InfoExtractor):
79649588	28	IE_DESC = 'Generic downloader that works on some sites'
9b122384	29	_VALID_URL = r'.*'
79649588	30	IE_NAME = 'generic'
cfe50f04 JMF	31	_TESTS = [
cfe50f04 JMF	32	{
79649588 PH	33	'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
	34	'file': '13601338388002.mp4',
	35	'md5': '6e15c93721d7ec9e9ca3fdbf07982cfd',
	36	'info_dict': {
	37	'uploader': 'www.hodiho.fr',
	38	'title': 'R\u00e9gis plante sa Jeep',
cfe50f04 JMF	39	}
cfe50f04 JMF	40	},
9d4660ca PH	41	# embedded vimeo video
9d4660ca PH	42	{
79649588 PH	43	'add_ie': ['Vimeo'],
	44	'url': 'http://skillsmatter.com/podcast/home/move-semanticsperfect-forwarding-and-rvalue-references',
	45	'file': '22444065.mp4',
	46	'md5': '2903896e23df39722c33f015af0666e2',
	47	'info_dict': {
	48	'title': 'ACCU 2011: Move Semantics,Perfect Forwarding, and Rvalue references- Scott Meyers- 13/04/2011',
	49	'uploader_id': 'skillsmatter',
	50	'uploader': 'Skills Matter',
9d4660ca	51	}
c19f7764 JMF	52	},
	53	# bandcamp page with custom domain
	54	{
79649588 PH	55	'add_ie': ['Bandcamp'],
	56	'url': 'http://bronyrock.com/track/the-pony-mash',
	57	'file': '3235767654.mp3',
	58	'info_dict': {
	59	'title': 'The Pony Mash',
	60	'uploader': 'M_Pallante',
c19f7764	61	},
79649588	62	'skip': 'There is a limit of 200 free downloads / month for the test song',
c19f7764	63	},
eeb165e6	64	# embedded brightcove video
dd5bcdc4 JMF	65	# it also tests brightcove videos that need to set the 'Referer' in the
dd5bcdc4 JMF	66	# http requests
eeb165e6	67	{
79649588 PH	68	'add_ie': ['Brightcove'],
	69	'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
	70	'info_dict': {
	71	'id': '2765128793001',
	72	'ext': 'mp4',
	73	'title': 'Le cours de bourse : l’analyse technique',
	74	'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
	75	'uploader': 'BFM BUSINESS',
eeb165e6	76	},
79649588 PH	77	'params': {
79649588 PH	78	'skip_download': True,
eeb165e6 JMF	79	},
eeb165e6 JMF	80	},
42393ce2 PH	81	# Direct link to a video
42393ce2 PH	82	{
79649588 PH	83	'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
	84	'file': 'trailer.mp4',
	85	'md5': '67d406c2bcb6af27fa886f31aa934bbe',
	86	'info_dict': {
	87	'id': 'trailer',
	88	'title': 'trailer',
	89	'upload_date': '20100513',
42393ce2	90	}
c0d0b01f JMF	91	},
	92	# ooyala video
	93	{
79649588	94	'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
3486df38	95	'file': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ.mp4',
79649588 PH	96	'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
	97	'info_dict': {
	98	'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
	99	'ext': 'mp4',
3486df38	100	'title': '2cc213299525360.mov', # that's what we get
c0d0b01f JMF	101	},
c0d0b01f JMF	102	},
cfe50f04	103	]
9b122384 PH	104
	105	def report_download_webpage(self, video_id):
	106	"""Report webpage download."""
	107	if not self._downloader.params.get('test', False):
79649588	108	self._downloader.report_warning('Falling back on generic information extractor.')
9b122384 PH	109	super(GenericIE, self).report_download_webpage(video_id)
	110
	111	def report_following_redirect(self, new_url):
	112	"""Report information extraction."""
79649588	113	self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
9b122384	114
42393ce2	115	def _send_head(self, url):
9b122384	116	"""Check if it is a redirect, like url shorteners, in case return the new url."""
9b122384 PH	117
	118	class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
	119	"""
	120	Subclass the HTTPRedirectHandler to make it use our
aa94a6d3	121	HEADRequest also on the redirected URL
9b122384 PH	122	"""
	123	def redirect_request(self, req, fp, code, msg, headers, newurl):
	124	if code in (301, 302, 303, 307):
	125	newurl = newurl.replace(' ', '%20')
	126	newheaders = dict((k,v) for k,v in req.headers.items()
	127	if k.lower() not in ("content-length", "content-type"))
aa94a6d3	128	return HEADRequest(newurl,
9b122384 PH	129	headers=newheaders,
	130	origin_req_host=req.get_origin_req_host(),
	131	unverifiable=True)
	132	else:
	133	raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
	134
	135	class HTTPMethodFallback(compat_urllib_request.BaseHandler):
	136	"""
	137	Fallback to GET if HEAD is not allowed (405 HTTP error)
	138	"""
	139	def http_error_405(self, req, fp, code, msg, headers):
	140	fp.read()
	141	fp.close()
	142
	143	newheaders = dict((k,v) for k,v in req.headers.items()
	144	if k.lower() not in ("content-length", "content-type"))
	145	return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
	146	headers=newheaders,
	147	origin_req_host=req.get_origin_req_host(),
	148	unverifiable=True))
	149
	150	# Build our opener
	151	opener = compat_urllib_request.OpenerDirector()
	152	for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
	153	HTTPMethodFallback, HEADRedirectHandler,
	154	compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
	155	opener.add_handler(handler())
	156
aa94a6d3	157	response = opener.open(HEADRequest(url))
9b122384	158	if response is None:
79649588	159	raise ExtractorError('Invalid URL protocol')
42393ce2	160	return response
9b122384 PH	161
9b122384 PH	162	def _real_extract(self, url):
a7130543 JMF	163	parsed_url = compat_urlparse.urlparse(url)
	164	if not parsed_url.scheme:
	165	self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
	166	return self.url_result('http://' + url)
42393ce2	167	video_id = os.path.splitext(url.split('/')[-1])[0]
a7130543	168
79649588	169	self.to_screen('%s: Requesting header' % video_id)
c1d1facd	170
30934689	171	try:
42393ce2 PH	172	response = self._send_head(url)
	173
	174	# Check for redirect
	175	new_url = response.geturl()
	176	if url != new_url:
	177	self.report_following_redirect(new_url)
cecaaf3f	178	return self.url_result(new_url)
42393ce2 PH	179
	180	# Check for direct link to a video
	181	content_type = response.headers.get('Content-Type', '')
3e785145	182	m = re.match(r'^(?P<type>audio\|video\|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
42393ce2 PH	183	if m:
	184	upload_date = response.headers.get('Last-Modified')
	185	if upload_date:
	186	upload_date = unified_strdate(upload_date)
42393ce2 PH	187	return {
	188	'id': video_id,
	189	'title': os.path.splitext(url_basename(url))[0],
	190	'formats': [{
	191	'format_id': m.group('format_id'),
	192	'url': url,
79649588	193	'vcodec': 'none' if m.group('type') == 'audio' else None
42393ce2 PH	194	}],
	195	'upload_date': upload_date,
	196	}
	197
30934689 PH	198	except compat_urllib_error.HTTPError:
	199	# This may be a stupid server that doesn't like HEAD, our UA, or so
	200	pass
9b122384	201
9b122384 PH	202	try:
	203	webpage = self._download_webpage(url, video_id)
	204	except ValueError:
	205	# since this is the last-resort InfoExtractor, if
	206	# this error is thrown, it'll be thrown here
79649588	207	raise ExtractorError('Failed to download URL: %s' % url)
9b122384 PH	208
9b122384 PH	209	self.report_extraction(video_id)
887c6acd PH	210
	211	# it's tempting to parse this further, but you would
	212	# have to take into account all the variations like
	213	# Video Title - Site Name
	214	# Site Name \| Video Title
	215	# Video Title - Tagline \| Site Name
	216	# and so on and so forth; it's just not practical
ef4fd848	217	video_title = self._html_search_regex(
79649588 PH	218	r'(?s)<title>(.*?)</title>', webpage, 'video title',
79649588 PH	219	default='video')
ef4fd848 PH	220
	221	# video uploader is domain name
	222	video_uploader = self._search_regex(
79649588	223	r'^(?:https?://)?([^/])/.', url, 'video uploader')
887c6acd	224
627a91a9	225	# Look for BrightCove:
eeb165e6 JMF	226	bc_url = BrightcoveIE._extract_brightcove_url(webpage)
eeb165e6 JMF	227	if bc_url is not None:
79649588	228	self.to_screen('Brightcove video detected.')
79f82953 PH	229	surl = smuggle_url(bc_url, {'Referer': url})
79f82953 PH	230	return self.url_result(surl, 'Brightcove')
cfe50f04	231
7115ca84	232	# Look for embedded (iframe) Vimeo player
9d4660ca	233	mobj = re.search(
3fabeaa1	234	r'<iframe[^>]+?src="((?:https?:)?//player.vimeo.com/video/.+?)"', webpage)
9d4660ca PH	235	if mobj:
	236	player_url = unescapeHTML(mobj.group(1))
	237	surl = smuggle_url(player_url, {'Referer': url})
	238	return self.url_result(surl, 'Vimeo')
	239
7115ca84 PH	240	# Look for embedded (swf embed) Vimeo player
	241	mobj = re.search(
	242	r'<embed[^>]+?src="(https?://(?:www\.)?vimeo.com/moogaloop.swf.+?)"', webpage)
	243	if mobj:
	244	return self.url_result(mobj.group(1), 'Vimeo')
	245
53c1d3ef	246	# Look for embedded YouTube player
1f9da904 PH	247	matches = re.findall(r'''(?x)
	248	(?:<iframe[^>]+?src=\|embedSWF\(\s*)
	249	(["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
	250	(?:embed\|v)/.+?)
	251	\1''', webpage)
887c6acd PH	252	if matches:
	253	urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
	254	for tuppl in matches]
	255	return self.playlist_result(
	256	urlrs, playlist_id=video_id, playlist_title=video_title)
53c1d3ef	257
355e4fd0 PH	258	# Look for embedded Dailymotion player
355e4fd0 PH	259	matches = re.findall(
ef4fd848	260	r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
355e4fd0 PH	261	if matches:
	262	urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Dailymotion')
	263	for tuppl in matches]
	264	return self.playlist_result(
	265	urlrs, playlist_id=video_id, playlist_title=video_title)
	266
ef4fd848 PH	267	# Look for embedded Wistia player
	268	match = re.search(
	269	r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
	270	if match:
	271	return {
	272	'_type': 'url_transparent',
	273	'url': unescapeHTML(match.group('url')),
	274	'ie_key': 'Wistia',
	275	'uploader': video_uploader,
	276	'title': video_title,
	277	'id': video_id,
	278	}
	279
ee3e63e4	280	# Look for embedded blip.tv player
19dab5e6	281	mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
ee3e63e4	282	if mobj:
19dab5e6	283	return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
19dab5e6	284	mobj = re.search(r'<(?:iframe\|embed\|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/\|api\.swf#)[a-zA-Z0-9]+)', webpage)
ee3e63e4	285	if mobj:
19dab5e6	286	return self.url_result(mobj.group(1), 'BlipTV')
ee3e63e4	287
c19f7764 JMF	288	# Look for Bandcamp pages with custom domain
	289	mobj = re.search(r'<meta property="og:url"[^>]?content="(.?bandcamp\.com.*?)"', webpage)
	290	if mobj is not None:
	291	burl = unescapeHTML(mobj.group(1))
09804265 JMF	292	# Don't set the extractor because it can be a track url or an album
09804265 JMF	293	return self.url_result(burl)
c19f7764	294
f25571ff PH	295	# Look for embedded Vevo player
	296	mobj = re.search(
	297	r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
	298	if mobj is not None:
	299	return self.url_result(mobj.group('url'))
	300
c0d0b01f JMF	301	# Look for Ooyala videos
	302	mobj = re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode\|ec)=([^"&]+)', webpage)
	303	if mobj is not None:
	304	return OoyalaIE._build_url_result(mobj.group(1))
	305
aa94a6d3	306	# Look for Aparat videos
42e4fcf2	307	mobj = re.search(r'<iframe src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
aa94a6d3 PH	308	if mobj is not None:
	309	return self.url_result(mobj.group(1), 'Aparat')
	310
c93c2ab1	311	# Look for MPORA videos
42e4fcf2	312	mobj = re.search(r'<iframe .*?src="(http://mpora\.com/videos/[^"]+)"', webpage)
c93c2ab1 PH	313	if mobj is not None:
c93c2ab1 PH	314	return self.url_result(mobj.group(1), 'Mpora')
5f59ee79	315
8f89e687	316	# Look for embedded Novamov player
	317	mobj = re.search(
	318	r'<iframe[^>]+?src=(["\'])(?P<url>http://(?:(?:embed\|www)\.)?novamov\.com/embed\.php.+?)\1', webpage)
	319	if mobj is not None:
	320	return self.url_result(mobj.group('url'), 'Novamov')
c93c2ab1	321
9834872b PH	322	# Look for embedded Facebook player
	323	mobj = re.search(
	324	r'<iframe[^>]+?src=(["\'])(?P<url>https://www.facebook.com/video/embed.+?)\1', webpage)
	325	if mobj is not None:
	326	return self.url_result(mobj.group('url'), 'Facebook')
	327
9b122384 PH	328	# Start with something easy: JW Player in SWFObject
9b122384 PH	329	mobj = re.search(r'flashvars: [\'"](?:.&)?file=(http[^\'"&])', webpage)
d981cef6 PH	330	if mobj is None:
	331	# Look for gorilla-vid style embedding
	332	mobj = re.search(r'(?s)jw_plugins.?file:\s["\'](.*?)["\']', webpage)
9b122384 PH	333	if mobj is None:
	334	# Broaden the search a little bit
	335	mobj = re.search(r'[^A-Za-z0-9]?(?:file\|source)=(http[^\'"&]*)', webpage)
	336	if mobj is None:
	337	# Broaden the search a little bit: JWPlayer JS loader
2f21eb2d	338	mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
9b122384 PH	339	if mobj is None:
	340	# Try to find twitter cards info
	341	mobj = re.search(r'<meta (?:property\|name)="twitter:player:stream" (?:content\|value)="(.+?)"', webpage)
	342	if mobj is None:
	343	# We look for Open Graph info:
	344	# We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
	345	m_video_type = re.search(r'<meta.?property="og:video:type".?content="video/(.*?)"', webpage)
	346	# We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
	347	if m_video_type is not None:
	348	mobj = re.search(r'<meta.?property="og:video".?content="(.*?)"', webpage)
7fea7156 PH	349	if mobj is None:
7fea7156 PH	350	# HTML5 video
08e291b5	351	mobj = re.search(r'<video[^<](?:>.?<source.*?)? src="([^"]+)"', webpage, flags=re.DOTALL)
9b122384	352	if mobj is None:
79649588	353	raise ExtractorError('Unsupported URL: %s' % url)
9b122384 PH	354
	355	# It's possible that one of the regexes
	356	# matched, but returned an empty group:
	357	if mobj.group(1) is None:
79649588	358	raise ExtractorError('Did not find a valid video URL at %s' % url)
9b122384	359
08e291b5	360	video_url = mobj.group(1)
a5caba1e	361	video_url = compat_urlparse.urljoin(url, video_url)
08e291b5	362	video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
9b122384	363
fc9713a1 PH	364	# Sometimes, jwplayer extraction will result in a YouTube URL
	365	if YoutubeIE.suitable(video_url):
	366	return self.url_result(video_url, 'Youtube')
	367
9b122384	368	# here's a fun little line of code for you:
9b122384 PH	369	video_id = os.path.splitext(video_id)[0]
9b122384 PH	370
113577e1	371	return {
fc9713a1 PH	372	'id': video_id,
fc9713a1 PH	373	'url': video_url,
9b122384	374	'uploader': video_uploader,
fc9713a1	375	'title': video_title,
113577e1	376	}