X-Git-Url: http://git.cielonegro.org/gitweb.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fcondenast.py;h=e8f2b5a07591410c16fe6fe096678a12006abe48;hb=a4690b3244a42a833146c406e622c96045b23df5;hp=f336a3c620a04e8bb643309b4812725e8f50e1d1;hpb=e1fb24569065e93b0406a0f19a3ab75611a8c968;p=youtube-dl.git diff --git a/youtube_dl/extractor/condenast.py b/youtube_dl/extractor/condenast.py index f336a3c62..e8f2b5a07 100644 --- a/youtube_dl/extractor/condenast.py +++ b/youtube_dl/extractor/condenast.py @@ -1,15 +1,18 @@ # coding: utf-8 +from __future__ import unicode_literals import re -import json from .common import InfoExtractor -from ..utils import ( - compat_urllib_parse, - orderedSet, +from ..compat import ( + compat_urllib_parse_urlencode, compat_urllib_parse_urlparse, compat_urlparse, ) +from ..utils import ( + orderedSet, + remove_end, +) class CondeNastIE(InfoExtractor): @@ -20,30 +23,56 @@ class CondeNastIE(InfoExtractor): # The keys are the supported sites and the values are the name to be shown # to the user and in the extractor description. - _SITES = {'wired': u'WIRED', - 'gq': u'GQ', - 'vogue': u'Vogue', - 'glamour': u'Glamour', - 'wmagazine': u'W Magazine', - 'vanityfair': u'Vanity Fair', - } - - _VALID_URL = r'http://(video|www).(?P%s).com/(?Pwatch|series|video)/(?P.+)' % '|'.join(_SITES.keys()) - IE_DESC = u'Condé Nast media group: %s' % ', '.join(sorted(_SITES.values())) - - _TEST = { - u'url': u'http://video.wired.com/watch/3d-printed-speakers-lit-with-led', - u'file': u'5171b343c2b4c00dd0c1ccb3.mp4', - u'md5': u'1921f713ed48aabd715691f774c451f7', - u'info_dict': { - u'title': u'3D Printed Speakers Lit With LED', - u'description': u'Check out these beautiful 3D printed LED speakers. You can\'t actually buy them, but LumiGeek is working on a board that will let you make you\'re own.', - } + _SITES = { + 'allure': 'Allure', + 'architecturaldigest': 'Architectural Digest', + 'arstechnica': 'Ars Technica', + 'bonappetit': 'Bon Appétit', + 'brides': 'Brides', + 'cnevids': 'Condé Nast', + 'cntraveler': 'Condé Nast Traveler', + 'details': 'Details', + 'epicurious': 'Epicurious', + 'glamour': 'Glamour', + 'golfdigest': 'Golf Digest', + 'gq': 'GQ', + 'newyorker': 'The New Yorker', + 'self': 'SELF', + 'teenvogue': 'Teen Vogue', + 'vanityfair': 'Vanity Fair', + 'vogue': 'Vogue', + 'wired': 'WIRED', + 'wmagazine': 'W Magazine', } + _VALID_URL = r'https?://(?:video|www|player)\.(?P%s)\.com/(?Pwatch|series|video|embed(?:js)?)/(?P[^/?#]+)' % '|'.join(_SITES.keys()) + IE_DESC = 'Condé Nast media group: %s' % ', '.join(sorted(_SITES.values())) + + EMBED_URL = r'(?:https?:)?//player\.(?P%s)\.com/(?Pembed(?:js)?)/.+?' % '|'.join(_SITES.keys()) + + _TESTS = [{ + 'url': 'http://video.wired.com/watch/3d-printed-speakers-lit-with-led', + 'md5': '1921f713ed48aabd715691f774c451f7', + 'info_dict': { + 'id': '5171b343c2b4c00dd0c1ccb3', + 'ext': 'mp4', + 'title': '3D Printed Speakers Lit With LED', + 'description': 'Check out these beautiful 3D printed LED speakers. You can\'t actually buy them, but LumiGeek is working on a board that will let you make you\'re own.', + } + }, { + # JS embed + 'url': 'http://player.cnevids.com/embedjs/55f9cf8b61646d1acf00000c/5511d76261646d5566020000.js', + 'md5': 'f1a6f9cafb7083bab74a710f65d08999', + 'info_dict': { + 'id': '55f9cf8b61646d1acf00000c', + 'ext': 'mp4', + 'title': '3D printed TSA Travel Sentry keys really do open TSA locks', + } + }] + def _extract_series(self, url, webpage): title = self._html_search_regex(r'
.*?

(.+?)

', - webpage, u'series title', flags=re.DOTALL) + webpage, 'series title', flags=re.DOTALL) url_object = compat_urllib_parse_urlparse(url) base_url = '%s://%s' % (url_object.scheme, url_object.netloc) m_paths = re.finditer(r'

.*?(.+?)

', - r'
(.+?)
', - ], - webpage, u'description', - fatal=False, flags=re.DOTALL) + def _extract_video(self, webpage, url_type): + if url_type != 'embed': + description = self._html_search_regex( + [ + r'
(.+?)
', + r'
(.+?)
', + ], + webpage, 'description', fatal=False, flags=re.DOTALL) + else: + description = None params = self._search_regex(r'var params = {(.+?)}[;,]', webpage, - u'player params', flags=re.DOTALL) - video_id = self._search_regex(r'videoId: [\'"](.+?)[\'"]', params, u'video id') - player_id = self._search_regex(r'playerId: [\'"](.+?)[\'"]', params, u'player id') - target = self._search_regex(r'target: [\'"](.+?)[\'"]', params, u'target') - data = compat_urllib_parse.urlencode({'videoId': video_id, + 'player params', flags=re.DOTALL) + video_id = self._search_regex(r'videoId: [\'"](.+?)[\'"]', params, 'video id') + player_id = self._search_regex(r'playerId: [\'"](.+?)[\'"]', params, 'player id') + target = self._search_regex(r'target: [\'"](.+?)[\'"]', params, 'target') + data = compat_urllib_parse_urlencode({'videoId': video_id, 'playerId': player_id, 'target': target, }) base_info_url = self._search_regex(r'url = [\'"](.+?)[\'"][,;]', - webpage, u'base info url', + webpage, 'base info url', default='http://player.cnevids.com/player/loader.js?') info_url = base_info_url + data info_page = self._download_webpage(info_url, video_id, - u'Downloading video info') - video_info = self._search_regex(r'var video = ({.+?});', info_page, u'video info') - video_info = json.loads(video_info) - - def _formats_sort_key(f): - type_ord = 1 if f['type'] == 'video/mp4' else 0 - quality_ord = 1 if f['quality'] == 'high' else 0 - return (quality_ord, type_ord) - best_format = sorted(video_info['sources'][0], key=_formats_sort_key)[-1] - - return {'id': video_id, - 'url': best_format['src'], - 'ext': best_format['type'].split('/')[-1], - 'title': video_info['title'], - 'thumbnail': video_info['poster_frame'], - 'description': description, - } + 'Downloading video info') + video_info = self._search_regex(r'var\s+video\s*=\s*({.+?});', info_page, 'video info') + video_info = self._parse_json(video_info, video_id) + + formats = [{ + 'format_id': '%s-%s' % (fdata['type'].split('/')[-1], fdata['quality']), + 'url': fdata['src'], + 'ext': fdata['type'].split('/')[-1], + 'quality': 1 if fdata['quality'] == 'high' else 0, + } for fdata in video_info['sources'][0]] + self._sort_formats(formats) + + return { + 'id': video_id, + 'formats': formats, + 'title': video_info['title'], + 'thumbnail': video_info['poster_frame'], + 'description': description, + } def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) site = mobj.group('site') url_type = mobj.group('type') - id = mobj.group('id') + item_id = mobj.group('id') + + # Convert JS embed to regular embed + if url_type == 'embedjs': + parsed_url = compat_urlparse.urlparse(url) + url = compat_urlparse.urlunparse(parsed_url._replace( + path=remove_end(parsed_url.path, '.js').replace('/embedjs/', '/embed/'))) + url_type = 'embed' - self.to_screen(u'Extracting from %s with the Condé Nast extractor' % self._SITES[site]) - webpage = self._download_webpage(url, id) + self.to_screen('Extracting from %s with the Condé Nast extractor' % self._SITES[site]) + webpage = self._download_webpage(url, item_id) if url_type == 'series': return self._extract_series(url, webpage) else: - return self._extract_video(webpage) + return self._extract_video(webpage, url_type)