X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fdailymotion.py;h=360113f9c46840ad4e0f38a8cc9974f36d6a25c5;hb=7fad1c6328b02ba9f23d37f374a05255abfe38a6;hp=eb2322d54778673bf4d7e718d0393c80b2217e8c;hpb=953e32b2c1be077e65bba844010a5a2707af2e2b;p=youtube-dl diff --git a/youtube_dl/extractor/dailymotion.py b/youtube_dl/extractor/dailymotion.py index eb2322d54..360113f9c 100644 --- a/youtube_dl/extractor/dailymotion.py +++ b/youtube_dl/extractor/dailymotion.py @@ -1,14 +1,11 @@ import re import json import itertools -import socket from .common import InfoExtractor -from .subtitles import SubtitlesIE +from .subtitles import SubtitlesInfoExtractor from ..utils import ( - compat_http_client, - compat_urllib_error, compat_urllib_request, compat_str, get_element_by_attribute, @@ -18,42 +15,17 @@ from ..utils import ( ) -class DailyMotionSubtitlesIE(SubtitlesIE): - - def _get_available_subtitles(self, video_id): - request = compat_urllib_request.Request('https://api.dailymotion.com/video/%s/subtitles?fields=id,language,url' % video_id) - try: - sub_list = compat_urllib_request.urlopen(request).read().decode('utf-8') - except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: - self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err)) - return {} - info = json.loads(sub_list) - if (info['total'] > 0): - sub_lang_list = dict((l['language'], l['url']) for l in info['list']) - return sub_lang_list - self._downloader.report_warning(u'video doesn\'t have subtitles') - return {} - - def _get_subtitle_url(self, sub_lang, sub_name, video_id, format): - sub_lang_list = self._get_available_subtitles(video_id) - return sub_lang_list[sub_lang] - - def _request_automatic_caption(self, video_id, webpage): - self._downloader.report_warning(u'Automatic Captions not supported by dailymotion') - return {} - - -class DailymotionIE(DailyMotionSubtitlesIE): #,InfoExtractor): +class DailymotionIE(SubtitlesInfoExtractor): """Information Extractor for Dailymotion""" - _VALID_URL = r'(?i)(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/video/([^/]+)' + _VALID_URL = r'(?i)(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/(?:embed/)?video/([^/]+)' IE_NAME = u'dailymotion' _TEST = { u'url': u'http://www.dailymotion.com/video/x33vw9_tutoriel-de-youtubeur-dl-des-video_tech', u'file': u'x33vw9.mp4', u'md5': u'392c4b85a60a90dc4792da41ce3144eb', u'info_dict': { - u"uploader": u"Alex and Van .", + u"uploader": u"Amphora Alex and Van .", u"title": u"Tutoriel de Youtubeur\"DL DES VIDEO DE YOUTUBE\"" } } @@ -65,6 +37,7 @@ class DailymotionIE(DailyMotionSubtitlesIE): #,InfoExtractor): video_id = mobj.group(1).split('_')[0].split('?')[0] video_extension = 'mp4' + url = 'http://www.dailymotion.com/video/%s' % video_id # Retrieve video webpage to extract further information request = compat_urllib_request.Request(url) @@ -87,41 +60,29 @@ class DailymotionIE(DailyMotionSubtitlesIE): #,InfoExtractor): embed_url = 'http://www.dailymotion.com/embed/video/%s' % video_id embed_page = self._download_webpage(embed_url, video_id, u'Downloading embed page') - info = self._search_regex(r'var info = ({.*?}),', embed_page, 'video info') + info = self._search_regex(r'var info = ({.*?}),$', embed_page, + 'video info', flags=re.MULTILINE) info = json.loads(info) # TODO: support choosing qualities - for key in ['stream_h264_hd1080_url', 'stream_h264_hd_url', - 'stream_h264_hq_url', 'stream_h264_url', + for key in ['stream_h264_hd1080_url','stream_h264_hd_url', + 'stream_h264_hq_url','stream_h264_url', 'stream_h264_ld_url']: - if info.get(key): # key in info and info[key]: + if info.get(key):#key in info and info[key]: max_quality = key - self.to_screen(u'%s: Using %s' % (video_id, key)) + self.to_screen(u'Using %s' % key) break else: raise ExtractorError(u'Unable to extract video URL') video_url = info[max_quality] # subtitles - video_subtitles = None - video_webpage = None - - if self._downloader.params.get('writesubtitles', False) or self._downloader.params.get('allsubtitles', False): - video_subtitles = self._extract_subtitles(video_id) - elif self._downloader.params.get('writeautomaticsub', False): - video_subtitles = self._request_automatic_caption(video_id, video_webpage) - + video_subtitles = self.extract_subtitles(video_id) if self._downloader.params.get('listsubtitles', False): self._list_available_subtitles(video_id) return - if 'length_seconds' not in info: - self._downloader.report_warning(u'unable to extract video duration') - video_duration = '' - else: - video_duration = compat_urllib_parse.unquote_plus(video_info['length_seconds'][0]) - return [{ 'id': video_id, 'url': video_url, @@ -132,3 +93,46 @@ class DailymotionIE(DailyMotionSubtitlesIE): #,InfoExtractor): 'subtitles': video_subtitles, 'thumbnail': info['thumbnail_url'] }] + + def _get_available_subtitles(self, video_id): + try: + sub_list = self._download_webpage( + 'https://api.dailymotion.com/video/%s/subtitles?fields=id,language,url' % video_id, + video_id, note=False) + except ExtractorError as err: + self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err)) + return {} + info = json.loads(sub_list) + if (info['total'] > 0): + sub_lang_list = dict((l['language'], l['url']) for l in info['list']) + return sub_lang_list + self._downloader.report_warning(u'video doesn\'t have subtitles') + return {} + + +class DailymotionPlaylistIE(InfoExtractor): + _VALID_URL = r'(?:https?://)?(?:www\.)?dailymotion\.[a-z]{2,3}/playlist/(?P.+?)/' + _MORE_PAGES_INDICATOR = r'' + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + playlist_id = mobj.group('id') + video_ids = [] + + for pagenum in itertools.count(1): + webpage = self._download_webpage('https://www.dailymotion.com/playlist/%s/%s' % (playlist_id, pagenum), + playlist_id, u'Downloading page %s' % pagenum) + + playlist_el = get_element_by_attribute(u'class', u'video_list', webpage) + video_ids.extend(re.findall(r'data-id="(.+?)" data-ext-id', playlist_el)) + + if re.search(self._MORE_PAGES_INDICATOR, webpage, re.DOTALL) is None: + break + + entries = [self.url_result('http://www.dailymotion.com/video/%s' % video_id, 'Dailymotion') + for video_id in video_ids] + return {'_type': 'playlist', + 'id': playlist_id, + 'title': get_element_by_id(u'playlist_name', webpage), + 'entries': entries, + }