X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/848723ea972c09f28787db91d8c06e98a274ab89..87046f03007c9b8a95c6297423e5660f795b186d:/youtube_dl/extractor/funnyordie.py diff --git a/youtube_dl/extractor/funnyordie.py b/youtube_dl/extractor/funnyordie.py index 7c40e67..f85e7de 100644 --- a/youtube_dl/extractor/funnyordie.py +++ b/youtube_dl/extractor/funnyordie.py @@ -3,22 +3,48 @@ from __future__ import unicode_literals import re from .common import InfoExtractor +from ..utils import ( + ExtractorError, + float_or_none, + int_or_none, + unified_timestamp, +) class FunnyOrDieIE(InfoExtractor): - _VALID_URL = r'^(?:https?://)?(?:www\.)?funnyordie\.com/videos/(?P[0-9a-f]+)/.*$' - _TEST = { + _VALID_URL = r'https?://(?:www\.)?funnyordie\.com/(?Pembed|articles|videos)/(?P[0-9a-f]+)(?:$|[?#/])' + _TESTS = [{ 'url': 'http://www.funnyordie.com/videos/0732f586d7/heart-shaped-box-literal-video-version', - 'file': '0732f586d7.mp4', - 'md5': 'f647e9e90064b53b6e046e75d0241fbd', + 'md5': 'bcd81e0c4f26189ee09be362ad6e6ba9', 'info_dict': { - 'description': ('Lyrics changed to match the video. Spoken cameo ' - 'by Obscurus Lupa (from ThatGuyWithTheGlasses.com). Based on a ' - 'concept by Dustin McLean (DustFilms.com). Performed, edited, ' - 'and written by David A. Scott.'), + 'id': '0732f586d7', + 'ext': 'mp4', 'title': 'Heart-Shaped Box: Literal Video Version', + 'description': 'md5:ea09a01bc9a1c46d9ab696c01747c338', + 'thumbnail': r're:^http:.*\.jpg$', + 'uploader': 'DASjr', + 'timestamp': 1317904928, + 'upload_date': '20111006', + 'duration': 318.3, + }, + }, { + 'url': 'http://www.funnyordie.com/embed/e402820827', + 'info_dict': { + 'id': 'e402820827', + 'ext': 'mp4', + 'title': 'Please Use This Song (Jon Lajoie)', + 'description': 'Please use this to sell something. www.jonlajoie.com', + 'thumbnail': r're:^http:.*\.jpg$', + 'timestamp': 1398988800, + 'upload_date': '20140502', + }, + 'params': { + 'skip_download': True, }, - } + }, { + 'url': 'http://www.funnyordie.com/articles/ebf5e34fc8/10-hours-of-walking-in-nyc-as-a-man', + 'only_matching': True, + }] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -26,14 +52,111 @@ class FunnyOrDieIE(InfoExtractor): video_id = mobj.group('id') webpage = self._download_webpage(url, video_id) - video_url = self._search_regex( - [r'type="video/mp4" src="(.*?)"', r'src="([^>]*?)" type=\'video/mp4\''], - webpage, 'video URL', flags=re.DOTALL) + links = re.findall(r']+src=(["\'])(?P.+?/master\.m3u8[^"\']*)\1', + webpage, 'm3u8 url', group='url') + + formats = [] + + m3u8_formats = self._extract_m3u8_formats( + m3u8_url, video_id, 'mp4', 'm3u8_native', + m3u8_id='hls', fatal=False) + source_formats = list(filter( + lambda f: f.get('vcodec') != 'none', m3u8_formats)) + + bitrates = [int(bitrate) for bitrate in re.findall(r'[,/]v(\d+)(?=[,/])', m3u8_url)] + bitrates.sort() + + if source_formats: + self._sort_formats(source_formats) + + for bitrate, f in zip(bitrates, source_formats or [{}] * len(bitrates)): + for path, ext in links: + ff = f.copy() + if ff: + if ext != 'mp4': + ff = dict( + [(k, v) for k, v in ff.items() + if k in ('height', 'width', 'format_id')]) + ff.update({ + 'format_id': ff['format_id'].replace('hls', ext), + 'ext': ext, + 'protocol': 'http', + }) + else: + ff.update({ + 'format_id': '%s-%d' % (ext, bitrate), + 'vbr': bitrate, + }) + ff['url'] = self._proto_relative_url( + '%s%d.%s' % (path, bitrate, ext)) + formats.append(ff) + self._check_formats(formats, video_id) + + formats.extend(m3u8_formats) + self._sort_formats( + formats, field_preference=('height', 'width', 'tbr', 'format_id')) + + subtitles = {} + for src, src_lang in re.findall(r']+\bclass=["\']channel-preview-name[^>]+>(.+?)