diff --git a/youtube_dl/extractor/mdr.py b/youtube_dl/extractor/mdr.py index 8b096a78a6..366352dbe7 100644 --- a/youtube_dl/extractor/mdr.py +++ b/youtube_dl/extractor/mdr.py @@ -5,35 +5,28 @@ ExtractorError, ) + class MDRIE(InfoExtractor): _VALID_URL = r'^(?P(?:https?://)?(?:www\.)?mdr\.de)/mediathek/(?:.*)/(?Pvideo|audio)(?P[^/_]+)_.*' - _TITLE = r'

(?P[^<]+)(?P[^<]+)

' - _MEDIA_XML = r'(?P/mediathek/(.+)/(video|audio)([0-9]+)-avCustom.xml)' - _MEDIA_STREAM_VIDEO = r'.*(?P[0-9]+).*(?P[^<]+)(?P[^<]+).*(?P[^<]+)' - _MEDIA_STREAM_AUDIO = r'.*(?P[A-Z0-9]+)(?P[0-9]+).*(?P[^<]+)(?P[^<]+).*(?P[^<]+)' _TESTS = [{ u'url': u'http://www.mdr.de/mediathek/themen/nachrichten/video165624_zc-c5c7de76_zs-3795826d.html', u'file': u'165624.mp4', - u'md5': u'95165945756198b8fa2dea10f0b04614', + u'md5': u'ae785f36ecbf2f19b42edf1bc9c85815', u'info_dict': { u"title": u"MDR aktuell Eins30 09.12.2013, 22:48 Uhr" }, - #u'skip': u'Requires rtmpdump' # rtmp is optional }, { - u'url': u' http://www.mdr.de/mediathek/radio/mdr1-radio-sachsen/audio718370_zc-67b21197_zs-1b9b2483.html', - u'file': u'718370.mp4', - u'md5': u'4a5b1fbb5519fb0d929c384b6ff7cb8b', + u'url': u'http://www.mdr.de/mediathek/radio/mdr1-radio-sachsen/audio718370_zc-67b21197_zs-1b9b2483.html', + u'file': u'718370.mp3', + u'md5': u'a9d21345a234c7b45dee612f290fd8d7', u'info_dict': { u"title": u"MDR 1 RADIO SACHSEN 10.12.2013, 05:00 Uhr" }, - #u'skip': u'Requires rtmpdump' # rtmp is optional }] def _real_extract(self, url): - - # determine video id from url m = re.match(self._VALID_URL, url) video_id = m.group('video_id') domain = m.group('domain') @@ -41,38 +34,46 @@ def _real_extract(self, url): # determine title and media streams from webpage html = self._download_webpage(url, video_id) - t = re.search(self._TITLE, html) - if not t: - raise ExtractorError(u'no title found') - title = t.group('title1') + t.group('title2') - m = re.search(self._MEDIA_XML, html) - if not m: - raise ExtractorError(u'no xml found') - xmlurl = m.group('xmlurl') - xml = self._download_webpage(domain+xmlurl, video_id, 'download XML').replace('\n','').replace('\r','').replace('','\n').replace('','\n') - if(mediatype == "video"): - streams = [mo.groupdict() for mo in re.finditer(self._MEDIA_STREAM_VIDEO, xml)] - if not streams: - raise ExtractorError(u'no media found') - # choose default media type and highest quality for now - stream = max([s for s in streams if s["progressiveDownloadUrl"].startswith("http://") ], - key=lambda s: int(s["frameWidth"])) - else: - streams = [mo.groupdict() for mo in re.finditer(self._MEDIA_STREAM_AUDIO, xml)] - if not streams: - raise ExtractorError(u'no media found') - # choose default media type (MP4) and highest quality for now - stream = max([s for s in streams if s["progressiveDownloadUrl"].startswith("http://") and s["mediaType"] == "MP4" ], - key=lambda s: int(s["bitrateAudio"])) - # there's two possibilities: RTMP stream or HTTP download - info = {'id': video_id, 'title': title, 'ext': 'mp4'} - if not stream["progressiveDownloadUrl"]: - self.to_screen(u'RTMP download detected') - assert stream['flashMediaServerURL'].startswith('mp4:') - info["url"] = stream["flashMediaServerApplicationURL"] - info["play_path"] = stream['flashMediaServerURL'] - else: - assert stream["progressiveDownloadUrl"].endswith('.mp4') - info["url"] = stream["progressiveDownloadUrl"] - return [info] + title = self._html_search_regex(r'

(.*?)

', html, u'title') + xmlurl = self._search_regex( + r'(/mediathek/(?:.+)/(?:video|audio)[0-9]+-avCustom.xml)', html, u'XML URL') + + doc = self._download_xml(domain + xmlurl, video_id) + formats = [] + for a in doc.findall('./assets/asset'): + url_el = a.find('.//progressiveDownloadUrl') + if url_el is None: + continue + abr = int(a.find('bitrateAudio').text) // 1000 + media_type = a.find('mediaType').text + format = { + 'abr': abr, + 'filesize': int(a.find('fileSize').text), + 'url': url_el.text, + } + + vbr_el = a.find('bitrateVideo') + if vbr_el is None: + format.update({ + 'vcodec': 'none', + 'format_id': u'%s-%d' % (media_type, abr), + }) + else: + vbr = int(vbr_el.text) // 1000 + format.update({ + 'vbr': vbr, + 'width': int(a.find('frameWidth').text), + 'height': int(a.find('frameHeight').text), + 'format_id': u'%s-%d' % (media_type, vbr), + }) + formats.append(format) + formats.sort(key=lambda f: (f.get('vbr'), f['abr'])) + if not formats: + raise ValueError('Could not find any valid formats') + + return { + 'id': video_id, + 'title': title, + 'formats': formats, + }