from __future__ import unicode_literals import re from .common import InfoExtractor class MDRIE(InfoExtractor): _VALID_URL = r'^(?Phttps?://(?:www\.)?mdr\.de)/(?:.*)/(?Pvideo|audio)(?P[^/_]+)(?:_|\.html)' # No tests, MDR regularily deletes its videos _TEST = { 'url': 'http://www.mdr.de/fakt/video189002.html', 'only_matching': True, } def _real_extract(self, url): m = re.match(self._VALID_URL, url) video_id = m.group('video_id') domain = m.group('domain') # determine title and media streams from webpage html = self._download_webpage(url, video_id) title = self._html_search_regex(r'(.*?)', html, 'title') xmlurl = self._search_regex( r'dataURL:\'(/(?:.+)/(?:video|audio)[0-9]+-avCustom.xml)', html, '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': '%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': '%s-%d' % (media_type, vbr), }) formats.append(format) self._sort_formats(formats) return { 'id': video_id, 'title': title, 'formats': formats, }