From a08dfd27a86b8458dfab52cdf5ee656394ff3800 Mon Sep 17 00:00:00 2001 From: Philipp Hagemeister Date: Sun, 23 Jun 2013 20:48:32 +0200 Subject: Move MyVideo into its own file --- youtube_dl/InfoExtractors.py | 150 +------------------------------------------ 1 file changed, 1 insertion(+), 149 deletions(-) (limited to 'youtube_dl/InfoExtractors.py') diff --git a/youtube_dl/InfoExtractors.py b/youtube_dl/InfoExtractors.py index a3cc2e6f5..5c6152427 100755 --- a/youtube_dl/InfoExtractors.py +++ b/youtube_dl/InfoExtractors.py @@ -25,6 +25,7 @@ from .extractor.dailymotion import DailymotionIE from .extractor.gametrailers import GametrailersIE from .extractor.generic import GenericIE from .extractor.metacafe import MetacafeIE +from .extractor.myvideo import MyVideoIE from .extractor.statigram import StatigramIE from .extractor.photobucket import PhotobucketIE from .extractor.vimeo import VimeoIE @@ -194,155 +195,6 @@ class FacebookIE(InfoExtractor): -class MyVideoIE(InfoExtractor): - """Information Extractor for myvideo.de.""" - - _VALID_URL = r'(?:http://)?(?:www\.)?myvideo\.de/watch/([0-9]+)/([^?/]+).*' - IE_NAME = u'myvideo' - - # Original Code from: https://github.com/dersphere/plugin.video.myvideo_de.git - # Released into the Public Domain by Tristan Fischer on 2013-05-19 - # https://github.com/rg3/youtube-dl/pull/842 - def __rc4crypt(self,data, key): - x = 0 - box = list(range(256)) - for i in list(range(256)): - x = (x + box[i] + compat_ord(key[i % len(key)])) % 256 - box[i], box[x] = box[x], box[i] - x = 0 - y = 0 - out = '' - for char in data: - x = (x + 1) % 256 - y = (y + box[x]) % 256 - box[x], box[y] = box[y], box[x] - out += chr(compat_ord(char) ^ box[(box[x] + box[y]) % 256]) - return out - - def __md5(self,s): - return hashlib.md5(s).hexdigest().encode() - - def _real_extract(self,url): - mobj = re.match(self._VALID_URL, url) - if mobj is None: - raise ExtractorError(u'invalid URL: %s' % url) - - video_id = mobj.group(1) - - GK = ( - b'WXpnME1EZGhNRGhpTTJNM01XVmhOREU0WldNNVpHTTJOakpt' - b'TW1FMU5tVTBNR05pWkRaa05XRXhNVFJoWVRVd1ptSXhaVEV3' - b'TnpsbA0KTVRkbU1tSTRNdz09' - ) - - # Get video webpage - webpage_url = 'http://www.myvideo.de/watch/%s' % video_id - webpage = self._download_webpage(webpage_url, video_id) - - mobj = re.search('source src=\'(.+?)[.]([^.]+)\'', webpage) - if mobj is not None: - self.report_extraction(video_id) - video_url = mobj.group(1) + '.flv' - - video_title = self._html_search_regex('([^<]+)', - webpage, u'title') - - video_ext = self._search_regex('[.](.+?)$', video_url, u'extension') - - return [{ - 'id': video_id, - 'url': video_url, - 'uploader': None, - 'upload_date': None, - 'title': video_title, - 'ext': u'flv', - }] - - # try encxml - mobj = re.search('var flashvars={(.+?)}', webpage) - if mobj is None: - raise ExtractorError(u'Unable to extract video') - - params = {} - encxml = '' - sec = mobj.group(1) - for (a, b) in re.findall('(.+?):\'(.+?)\',?', sec): - if not a == '_encxml': - params[a] = b - else: - encxml = compat_urllib_parse.unquote(b) - if not params.get('domain'): - params['domain'] = 'www.myvideo.de' - xmldata_url = '%s?%s' % (encxml, compat_urllib_parse.urlencode(params)) - if 'flash_playertype=MTV' in xmldata_url: - self._downloader.report_warning(u'avoiding MTV player') - xmldata_url = ( - 'http://www.myvideo.de/dynamic/get_player_video_xml.php' - '?flash_playertype=D&ID=%s&_countlimit=4&autorun=yes' - ) % video_id - - # get enc data - enc_data = self._download_webpage(xmldata_url, video_id).split('=')[1] - enc_data_b = binascii.unhexlify(enc_data) - sk = self.__md5( - base64.b64decode(base64.b64decode(GK)) + - self.__md5( - str(video_id).encode('utf-8') - ) - ) - dec_data = self.__rc4crypt(enc_data_b, sk) - - # extracting infos - self.report_extraction(video_id) - - video_url = None - mobj = re.search('connectionurl=\'(.*?)\'', dec_data) - if mobj: - video_url = compat_urllib_parse.unquote(mobj.group(1)) - if 'myvideo2flash' in video_url: - self._downloader.report_warning(u'forcing RTMPT ...') - video_url = video_url.replace('rtmpe://', 'rtmpt://') - - if not video_url: - # extract non rtmp videos - mobj = re.search('path=\'(http.*?)\' source=\'(.*?)\'', dec_data) - if mobj is None: - raise ExtractorError(u'unable to extract url') - video_url = compat_urllib_parse.unquote(mobj.group(1)) + compat_urllib_parse.unquote(mobj.group(2)) - - video_file = self._search_regex('source=\'(.*?)\'', dec_data, u'video file') - video_file = compat_urllib_parse.unquote(video_file) - - if not video_file.endswith('f4m'): - ppath, prefix = video_file.split('.') - video_playpath = '%s:%s' % (prefix, ppath) - video_hls_playlist = '' - else: - video_playpath = '' - video_hls_playlist = ( - video_filepath + video_file - ).replace('.f4m', '.m3u8') - - video_swfobj = self._search_regex('swfobject.embedSWF\(\'(.+?)\'', webpage, u'swfobj') - video_swfobj = compat_urllib_parse.unquote(video_swfobj) - - video_title = self._html_search_regex("(.*?)", - webpage, u'title') - - return [{ - 'id': video_id, - 'url': video_url, - 'tc_url': video_url, - 'uploader': None, - 'upload_date': None, - 'title': video_title, - 'ext': u'flv', - 'play_path': video_playpath, - 'video_file': video_file, - 'video_hls_playlist': video_hls_playlist, - 'player_url': video_swfobj, - }] - class ComedyCentralIE(InfoExtractor): """Information extractor for The Daily Show and Colbert Report """ -- cgit v1.2.3