X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fndr.py;h=2447c812e021e73991082aefab4bd98e6dd000a1;hb=HEAD;hp=79a13958b05e25a1c9e586168bb3a10742fbe01f;hpb=98c2c0febc7e686278ceaadc73eb40fee1b46752;p=youtube-dl diff --git a/youtube_dl/extractor/ndr.py b/youtube_dl/extractor/ndr.py index 79a13958b..2447c812e 100644 --- a/youtube_dl/extractor/ndr.py +++ b/youtube_dl/extractor/ndr.py @@ -1,130 +1,402 @@ -# encoding: utf-8 +# coding: utf-8 from __future__ import unicode_literals import re from .common import InfoExtractor from ..utils import ( - ExtractorError, + determine_ext, int_or_none, + merge_dicts, + parse_iso8601, qualities, - parse_duration, + try_get, + urljoin, ) class NDRBaseIE(InfoExtractor): def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) - video_id = mobj.group('id') + display_id = next(group for group in mobj.groups() if group) + webpage = self._download_webpage(url, display_id) + return self._extract_embed(webpage, display_id) - page = self._download_webpage(url, video_id, 'Downloading page') - title = self._og_search_title(page).strip() - description = self._og_search_description(page) - if description: - description = description.strip() +class NDRIE(NDRBaseIE): + IE_NAME = 'ndr' + IE_DESC = 'NDR.de - Norddeutscher Rundfunk' + _VALID_URL = r'https?://(?:www\.)?ndr\.de/(?:[^/]+/)*(?P[^/?#]+),[\da-z]+\.html' + _TESTS = [{ + # httpVideo, same content id + 'url': 'http://www.ndr.de/fernsehen/Party-Poette-und-Parade,hafengeburtstag988.html', + 'md5': '6515bc255dc5c5f8c85bbc38e035a659', + 'info_dict': { + 'id': 'hafengeburtstag988', + 'display_id': 'Party-Poette-und-Parade', + 'ext': 'mp4', + 'title': 'Party, Pötte und Parade', + 'description': 'md5:ad14f9d2f91d3040b6930c697e5f6b4c', + 'uploader': 'ndrtv', + 'timestamp': 1431108900, + 'upload_date': '20150510', + 'duration': 3498, + }, + 'params': { + 'skip_download': True, + }, + }, { + # httpVideo, different content id + 'url': 'http://www.ndr.de/sport/fussball/40-Osnabrueck-spielt-sich-in-einen-Rausch,osna270.html', + 'md5': '1043ff203eab307f0c51702ec49e9a71', + 'info_dict': { + 'id': 'osna272', + 'display_id': '40-Osnabrueck-spielt-sich-in-einen-Rausch', + 'ext': 'mp4', + 'title': 'Osnabrück - Wehen Wiesbaden: Die Highlights', + 'description': 'md5:32e9b800b3d2d4008103752682d5dc01', + 'uploader': 'ndrtv', + 'timestamp': 1442059200, + 'upload_date': '20150912', + 'duration': 510, + }, + 'params': { + 'skip_download': True, + }, + }, { + # httpAudio, same content id + 'url': 'http://www.ndr.de/info/La-Valette-entgeht-der-Hinrichtung,audio51535.html', + 'md5': 'bb3cd38e24fbcc866d13b50ca59307b8', + 'info_dict': { + 'id': 'audio51535', + 'display_id': 'La-Valette-entgeht-der-Hinrichtung', + 'ext': 'mp3', + 'title': 'La Valette entgeht der Hinrichtung', + 'description': 'md5:22f9541913a40fe50091d5cdd7c9f536', + 'uploader': 'ndrinfo', + 'timestamp': 1290626100, + 'upload_date': '20140729', + 'duration': 884, + }, + 'params': { + 'skip_download': True, + }, + }, { + 'url': 'https://www.ndr.de/Fettes-Brot-Ferris-MC-und-Thees-Uhlmann-live-on-stage,festivalsommer116.html', + 'only_matching': True, + }] - duration = int_or_none(self._html_search_regex(r'duration: (\d+),\n', page, 'duration', default=None)) - if not duration: - duration = parse_duration(self._html_search_regex( - r'(\d+:\d+)', - page, 'duration', default=None)) + def _extract_embed(self, webpage, display_id): + embed_url = self._html_search_meta( + 'embedURL', webpage, 'embed URL', + default=None) or self._search_regex( + r'\bembedUrl["\']\s*:\s*(["\'])(?P(?:(?!\1).)+)\1', webpage, + 'embed URL', group='url') + description = self._search_regex( + r']+itemprop="description">([^<]+)

', + webpage, 'description', default=None) or self._og_search_description(webpage) + timestamp = parse_iso8601( + self._search_regex( + r']+itemprop="(?:datePublished|uploadDate)"[^>]+content="([^"]+)"', + webpage, 'upload date', default=None)) + info = self._search_json_ld(webpage, display_id, default={}) + return merge_dicts({ + '_type': 'url_transparent', + 'url': embed_url, + 'display_id': display_id, + 'description': description, + 'timestamp': timestamp, + }, info) - formats = [] - mp3_url = re.search(r'''\{src:'(?P