From 276c9897720fe087924aef4ac80cf528e621b832 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Sergey=20M=E2=80=A4?= Date: Sun, 16 Aug 2015 03:07:05 +0600 Subject: [PATCH] [playtvak] Improve and simplify --- youtube_dl/extractor/playtvak.py | 145 ++++++++++++++++++++----------- 1 file changed, 92 insertions(+), 53 deletions(-) diff --git a/youtube_dl/extractor/playtvak.py b/youtube_dl/extractor/playtvak.py index 6dff6650c2..4e5034dc61 100644 --- a/youtube_dl/extractor/playtvak.py +++ b/youtube_dl/extractor/playtvak.py @@ -1,23 +1,22 @@ # coding: utf-8 from __future__ import unicode_literals -import re - from .common import InfoExtractor from ..compat import ( compat_urlparse, compat_urllib_parse, ) -from ..utils import ExtractorError - - -def _extract_json(code): - return re.sub( - r'(?s)^VideoPlayer.data\("", ({.*})\);?\s*?(?://[^\n]*)*$', r'\1', code) +from ..utils import ( + ExtractorError, + int_or_none, + parse_iso8601, + qualities, +) class PlaytvakIE(InfoExtractor): - _VALID_URL = r'https?://.*?(playtvak|idnes|lidovky|metro)\.cz/.*\?c=(?P[A-Z][0-9]{6}_[0-9]{6}_.*)' + IE_DESC = 'Playtvak.cz, iDNES.cz and Lidovky.cz' + _VALID_URL = r'https?://(?:.+?\.)?(?:playtvak|idnes|lidovky|metro)\.cz/.*\?(?:c|idvideo)=(?P[^&]+)' _TESTS = [{ 'url': 'http://www.playtvak.cz/vyzente-vosy-a-srsne-ze-zahrady-dn5-/hodinovy-manzel.aspx?c=A150730_150323_hodinovy-manzel_kuko', 'md5': '4525ae312c324b4be2f4603cc78ceb4a', @@ -25,8 +24,12 @@ class PlaytvakIE(InfoExtractor): 'id': 'A150730_150323_hodinovy-manzel_kuko', 'ext': 'mp4', 'title': 'Vyžeňte vosy a sršně ze zahrady', - 'thumbnail': 'http://oidnes.cz/15/074/mobil/KUK5cea00_010hodmanel58154.jpg', - 'description': 'Málo co kazí atmosféru venkovního posezení tak jako neustálé bzučení kolem hlavy. Vyzkoušejte náš lapač a odpuzovač vos a sršňů.', + 'description': 'md5:f93d398691044d303bc4a3de62f3e976', + 'thumbnail': 're:(?i)^https?://.*\.(?:jpg|png)$', + 'duration': 279, + 'timestamp': 1438732860, + 'upload_date': '20150805', + 'is_live': False, } }, { # live video test 'url': 'http://slowtv.playtvak.cz/planespotting-0pr-/planespotting.aspx?c=A150624_164934_planespotting_cat', @@ -34,8 +37,8 @@ class PlaytvakIE(InfoExtractor): 'id': 'A150624_164934_planespotting_cat', 'ext': 'flv', 'title': 're:^Přímý přenos iDNES.cz [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$', - 'thumbnail': 'http://data.idnes.cz/soubory/servisni-play-porady/89A150630_ACEK_026_VIDEOPLAYER-STREA.PNG', 'description': 'Sledujte provoz na ranveji Letiště Václava Havla v Praze', + 'thumbnail': 're:(?i)^https?://.*\.(?:jpg|png)$', 'is_live': True, }, 'params': { @@ -48,8 +51,12 @@ class PlaytvakIE(InfoExtractor): 'id': 'A150809_104116_domaci_pku', 'ext': 'mp4', 'title': 'Zavřeli jsme mraženou pizzu do auta. Upekla se', - 'thumbnail': 'http://i.idnes.cz/15/081/vidw/SHA5d1786_pizzaauto.jpg', - 'description': 'Na sociálních sítích se objevila výzva, aby lidé, kteří v horkých letních dnech uvidí v zaparkovaném autě zavřeného psa, neváhali rozbít okénko. Zastánci tohoto postoje argumentují zdravím zvířete, které v dusnu může zkolabovat. Policie doporučuje nejprve volat tísňovou linku.', + 'description': 'md5:01e73f02329e2e5760bd5eed4d42e3c2', + 'thumbnail': 're:(?i)^https?://.*\.(?:jpg|png)$', + 'duration': 39, + 'timestamp': 1438969140, + 'upload_date': '20150807', + 'is_live': False, } }, { # lidovky.cz 'url': 'http://www.lidovky.cz/dalsi-demonstrace-v-praze-o-migraci-duq-/video.aspx?c=A150808_214044_ln-video_ELE', @@ -58,70 +65,102 @@ class PlaytvakIE(InfoExtractor): 'id': 'A150808_214044_ln-video_ELE', 'ext': 'mp4', 'title': 'Táhni! Demonstrace proti imigrantům budila emoce', - 'thumbnail': 'http://i.idnes.cz/15/081/vidw/PID5d1d52_vandas3.jpg', - 'description': 'Desítky lidí se sešly v Praze na protest proti imigrantům. Současně probíhala i demonstrace na jejich podporu. Na Staroměstském náměstí vystoupil i předseda dělnické strany Tomáš Vandas a kontroverzní slovenský politik Marian Kotleba. Dalšího slovenského nacionalistu Mariána Magáta odvedla policie.', + 'description': 'md5:97c81d589a9491fbfa323c9fa3cca72c', + 'thumbnail': 're:(?i)^https?://.*\.(?:jpg|png)$', + 'timestamp': 1439052180, + 'upload_date': '20150808', + 'is_live': False, } + }, { + 'url': 'http://www.playtvak.cz/embed.aspx?idvideo=V150729_141549_play-porad_kuko', + 'only_matching': True, }] def _real_extract(self, url): video_id = self._match_id(url) + webpage = self._download_webpage(url, video_id) - infourl = self._html_search_regex(r'Misc.videoFLV\({ data: "([^"]+)"', webpage, 'xmlinfourl') - parsedurl = compat_urlparse.urlparse(infourl) - qs = compat_urlparse.parse_qs(parsedurl.query) - if 'reklama' in qs: # Don't ask for ads - qs['reklama'] = ['0'] - qs['type'] = ['js'] # Ask for JS-based info file - newquery = compat_urllib_parse.urlencode(qs, True) - infourl = compat_urlparse.urlunparse(parsedurl[:4] + (newquery, '')) - jsoninfo = self._download_json(infourl, video_id, transform_source=_extract_json) + + info_url = self._html_search_regex( + r'Misc\.videoFLV\(\s*{\s*data\s*:\s*"([^"]+)"', webpage, 'info url') + + parsed_url = compat_urlparse.urlparse(info_url) + + qs = compat_urlparse.parse_qs(parsed_url.query) + qs.update({ + 'reklama': ['0'], + 'type': ['js'], + }) + + info_url = compat_urlparse.urlunparse( + parsed_url._replace(query = compat_urllib_parse.urlencode(qs, True))) + + json_info = self._download_json( + info_url, video_id, + transform_source=lambda s: s[s.index('{'):s.rindex('}') + 1]) item = None - for i in jsoninfo['items']: - if i['type'] == 'video' or i['type'] == 'stream': + for i in json_info['items']: + if i.get('type') == 'video' or i.get('type') == 'stream': item = i break - if item is None: + if not item: raise ExtractorError('No suitable stream found') - title = item['title'] - thumbnail = item['image'] - is_live = item['type'] == 'stream' - if is_live: - title = self._live_title(title) + + quality = qualities(['low', 'middle', 'high']) formats = [] for fmt in item['video']: - format_entry = {'url': fmt['file'], - 'format_id': ("%s_%s" % (fmt['format'], fmt['quality'])), - } - if fmt['quality'] == 'middle': - format_entry['quality'] = -2 - elif fmt['quality'] == 'low': - format_entry['quality'] = -3 + video_url = fmt.get('file') + if not video_url: + continue - if fmt['format'] == 'mp4': - format_entry['ext'] = 'mp4' - elif fmt['format'] == 'webm': - format_entry['ext'] = 'webm' - elif fmt['format'] == 'apple': - format_entry['ext'] = 'mp4' - format_entry['protocol'] = 'm3u8' + format_ = fmt['format'] + format_id = '%s_%s' % (format_, fmt['quality']) + preference = None + + if format_ in ['mp4', 'webm']: + ext = format_ + elif format_ == 'rtmp': + ext = 'flv' + elif format_ == 'apple': + ext = 'mp4' # Some streams have mp3 audio which does not play # well with ffmpeg filter aac_adtstoasc - format_entry['preference'] = -1 - elif fmt['format'] == 'rtmp': - format_entry['ext'] = 'flv' + preference = -1 + elif format_ == 'adobe': # f4m manifest fails with 404 in 80% of requests + continue else: # Other formats not supported yet continue - formats.append(format_entry) - + formats.append({ + 'url': video_url, + 'ext': ext, + 'format_id': format_id, + 'quality': quality(fmt.get('quality')), + 'preference': preference, + }) self._sort_formats(formats) + + title = item['title'] + is_live = item['type'] == 'stream' + if is_live: + title = self._live_title(title) + timestamp = None + duration = None + if not is_live: + duration = int_or_none(item.get('length')) + timestamp = item.get('published') + if timestamp: + timestamp = parse_iso8601(timestamp[:-5]) + return { 'id': video_id, 'title': title, - 'thumbnail': thumbnail, 'description': self._og_search_description(webpage), + 'thumbnail': item.get('image'), + 'duration': duration, + 'timestamp': timestamp, 'is_live': is_live, 'formats': formats, }