From 5625bd0617d6c661c66049b420556ae442102ffc Mon Sep 17 00:00:00 2001 From: remitamine Date: Mon, 21 Dec 2015 21:06:10 +0100 Subject: [PATCH] [br] add support for br-klassik.de and improve extraction - extend _VALID_URL to match both br.de and br-klassik.de - extract all formats(hls,hds and rtmp) - use xpath_element and xpath_text for xml info extraction --- youtube_dl/extractor/br.py | 127 +++++++++++++++++++++++-------------- 1 file changed, 78 insertions(+), 49 deletions(-) diff --git a/youtube_dl/extractor/br.py b/youtube_dl/extractor/br.py index 66e394e10..e66854538 100644 --- a/youtube_dl/extractor/br.py +++ b/youtube_dl/extractor/br.py @@ -1,18 +1,21 @@ # coding: utf-8 from __future__ import unicode_literals +import re + from .common import InfoExtractor from ..utils import ( ExtractorError, int_or_none, parse_duration, + xpath_element, + xpath_text, ) class BRIE(InfoExtractor): IE_DESC = 'Bayerischer Rundfunk Mediathek' - _VALID_URL = r'https?://(?:www\.)?br\.de/(?:[a-z0-9\-_]+/)+(?P[a-z0-9\-_]+)\.html' - _BASE_URL = 'http://www.br.de' + _VALID_URL = r'(?Phttps?://(?:www\.)?br(?:-klassik)?\.de)/(?:[a-z0-9\-_]+/)+(?P[a-z0-9\-_]+)\.html' _TESTS = [ { @@ -22,7 +25,7 @@ class BRIE(InfoExtractor): 'id': '48f656ef-287e-486f-be86-459122db22cc', 'ext': 'mp4', 'title': 'Die böse Überraschung', - 'description': 'Betriebliche Altersvorsorge: Die böse Überraschung', + 'description': 'md5:ce9ac81b466ce775b8018f6801b48ac9', 'duration': 180, 'uploader': 'Reinhard Weber', 'upload_date': '20150422', @@ -30,23 +33,23 @@ class BRIE(InfoExtractor): }, { 'url': 'http://www.br.de/nachrichten/oberbayern/inhalt/muenchner-polizeipraesident-schreiber-gestorben-100.html', - 'md5': 'a44396d73ab6a68a69a568fae10705bb', + 'md5': 'af3a3a4aa43ff0ce6a89504c67f427ef', 'info_dict': { 'id': 'a4b83e34-123d-4b81-9f4e-c0d3121a4e05', - 'ext': 'mp4', + 'ext': 'flv', 'title': 'Manfred Schreiber ist tot', - 'description': 'Abendschau kompakt: Manfred Schreiber ist tot', + 'description': 'md5:b454d867f2a9fc524ebe88c3f5092d97', 'duration': 26, } }, { - 'url': 'http://www.br.de/radio/br-klassik/sendungen/allegro/premiere-urauffuehrung-the-land-2015-dance-festival-muenchen-100.html', + 'url': 'https://www.br-klassik.de/audio/peeping-tom-premierenkritik-dance-festival-muenchen-100.html', 'md5': '8b5b27c0b090f3b35eac4ab3f7a73d3d', 'info_dict': { 'id': '74c603c9-26d3-48bb-b85b-079aeed66e0b', 'ext': 'aac', 'title': 'Kurzweilig und sehr bewegend', - 'description': '"The Land" von Peeping Tom: Kurzweilig und sehr bewegend', + 'description': 'md5:0351996e3283d64adeb38ede91fac54e', 'duration': 296, } }, @@ -57,7 +60,7 @@ class BRIE(InfoExtractor): 'id': '6ba73750-d405-45d3-861d-1ce8c524e059', 'ext': 'mp4', 'title': 'Umweltbewusster Häuslebauer', - 'description': 'Uwe Erdelt: Umweltbewusster Häuslebauer', + 'description': 'md5:d52dae9792d00226348c1dbb13c9bae2', 'duration': 116, } }, @@ -68,7 +71,7 @@ class BRIE(InfoExtractor): 'id': 'd982c9ce-8648-4753-b358-98abb8aec43d', 'ext': 'mp4', 'title': 'Folge 1 - Metaphysik', - 'description': 'Kant für Anfänger: Folge 1 - Metaphysik', + 'description': 'md5:bb659990e9e59905c3d41e369db1fbe3', 'duration': 893, 'uploader': 'Eva Maria Steimle', 'upload_date': '20140117', @@ -77,28 +80,31 @@ class BRIE(InfoExtractor): ] def _real_extract(self, url): - display_id = self._match_id(url) + base_url, display_id = re.search(self._VALID_URL, url).groups() page = self._download_webpage(url, display_id) xml_url = self._search_regex( r"return BRavFramework\.register\(BRavFramework\('avPlayer_(?:[a-f0-9-]{36})'\)\.setup\({dataURL:'(/(?:[a-z0-9\-]+/)+[a-z0-9/~_.-]+)'}\)\);", page, 'XMLURL') - xml = self._download_xml(self._BASE_URL + xml_url, None) + xml = self._download_xml(base_url + xml_url, display_id) medias = [] for xml_media in xml.findall('video') + xml.findall('audio'): + media_id = xml_media.get('externalId') media = { - 'id': xml_media.get('externalId'), - 'title': xml_media.find('title').text, - 'duration': parse_duration(xml_media.find('duration').text), - 'formats': self._extract_formats(xml_media.find('assets')), - 'thumbnails': self._extract_thumbnails(xml_media.find('teaserImage/variants')), - 'description': ' '.join(xml_media.find('shareTitle').text.splitlines()), - 'webpage_url': xml_media.find('permalink').text + 'id': media_id, + 'title': xpath_text(xml_media, 'title', 'title', True), + 'duration': parse_duration(xpath_text(xml_media, 'duration')), + 'formats': self._extract_formats(xpath_element( + xml_media, 'assets'), media_id), + 'thumbnails': self._extract_thumbnails(xpath_element( + xml_media, 'teaserImage/variants'), base_url), + 'description': xpath_text(xml_media, 'desc'), + 'webpage_url': xpath_text(xml_media, 'permalink'), + 'uploader': xpath_text(xml_media, 'author'), } - if xml_media.find('author').text: - media['uploader'] = xml_media.find('author').text - if xml_media.find('broadcastDate').text: - media['upload_date'] = ''.join(reversed(xml_media.find('broadcastDate').text.split('.'))) + broadcast_date = xpath_text(xml_media, 'broadcastDate') + if broadcast_date: + media['upload_date'] = ''.join(reversed(broadcast_date.split('.'))) medias.append(media) if len(medias) > 1: @@ -109,35 +115,58 @@ class BRIE(InfoExtractor): raise ExtractorError('No media entries found') return medias[0] - def _extract_formats(self, assets): - - def text_or_none(asset, tag): - elem = asset.find(tag) - return None if elem is None else elem.text - - formats = [{ - 'url': text_or_none(asset, 'downloadUrl'), - 'ext': text_or_none(asset, 'mediaType'), - 'format_id': asset.get('type'), - 'width': int_or_none(text_or_none(asset, 'frameWidth')), - 'height': int_or_none(text_or_none(asset, 'frameHeight')), - 'tbr': int_or_none(text_or_none(asset, 'bitrateVideo')), - 'abr': int_or_none(text_or_none(asset, 'bitrateAudio')), - 'vcodec': text_or_none(asset, 'codecVideo'), - 'acodec': text_or_none(asset, 'codecAudio'), - 'container': text_or_none(asset, 'mediaType'), - 'filesize': int_or_none(text_or_none(asset, 'size')), - } for asset in assets.findall('asset') - if asset.find('downloadUrl') is not None] - + def _extract_formats(self, assets, media_id): + formats = [] + for asset in assets.findall('asset'): + format_url = xpath_text(asset, ['downloadUrl', 'url']) + asset_type = asset.get('type') + if asset_type == 'HDS': + f4m_formats = self._extract_f4m_formats( + format_url + '?hdcore=3.2.0', media_id, f4m_id='hds', fatal=False) + if f4m_formats: + formats.extend(f4m_formats) + elif asset_type == 'HLS': + m3u8_formats = self._extract_m3u8_formats( + format_url, media_id, 'mp4', 'm3u8_native', m3u8_id='hds', fatal=False) + if m3u8_formats: + formats.extend(m3u8_formats) + else: + format_info = { + 'ext': xpath_text(asset, 'mediaType'), + 'width': int_or_none(xpath_text(asset, 'frameWidth')), + 'height': int_or_none(xpath_text(asset, 'frameHeight')), + 'tbr': int_or_none(xpath_text(asset, 'bitrateVideo')), + 'abr': int_or_none(xpath_text(asset, 'bitrateAudio')), + 'vcodec': xpath_text(asset, 'codecVideo'), + 'acodec': xpath_text(asset, 'codecAudio'), + 'container': xpath_text(asset, 'mediaType'), + 'filesize': int_or_none(xpath_text(asset, 'size')), + } + format_url = self._proto_relative_url(format_url) + if format_url: + http_format_info = format_info.copy() + http_format_info.update({ + 'url': format_url, + 'format_id': 'http-%s' % asset_type, + }) + formats.append(http_format_info) + server_prefix = xpath_text(asset, 'serverPrefix') + if server_prefix: + rtmp_format_info = format_info.copy() + rtmp_format_info.update({ + 'url': server_prefix, + 'play_path': xpath_text(asset, 'fileName'), + 'format_id': 'rtmp-%s' % asset_type, + }) + formats.append(rtmp_format_info) self._sort_formats(formats) return formats - def _extract_thumbnails(self, variants): + def _extract_thumbnails(self, variants, base_url): thumbnails = [{ - 'url': self._BASE_URL + variant.find('url').text, - 'width': int_or_none(variant.find('width').text), - 'height': int_or_none(variant.find('height').text), - } for variant in variants.findall('variant')] + 'url': base_url + xpath_text(variant, 'url'), + 'width': int_or_none(xpath_text(variant, 'width')), + 'height': int_or_none(xpath_text(variant, 'height')), + } for variant in variants.findall('variant') if xpath_text(variant, 'url')] thumbnails.sort(key=lambda x: x['width'] * x['height'], reverse=True) return thumbnails -- 2.22.2