2016-10-02 19:39:18 +08:00
|
|
|
|
# coding: utf-8
|
2014-02-11 01:20:41 +08:00
|
|
|
|
from __future__ import unicode_literals
|
|
|
|
|
|
|
|
|
|
from .common import InfoExtractor
|
2016-12-05 00:20:14 +08:00
|
|
|
|
from ..compat import (
|
|
|
|
|
compat_str,
|
|
|
|
|
compat_urlparse,
|
|
|
|
|
)
|
2016-04-10 05:02:35 +08:00
|
|
|
|
from ..utils import (
|
|
|
|
|
int_or_none,
|
|
|
|
|
qualities,
|
|
|
|
|
unified_strdate,
|
|
|
|
|
)
|
2014-02-11 01:20:41 +08:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
class FirstTVIE(InfoExtractor):
|
2015-02-14 04:04:28 +08:00
|
|
|
|
IE_NAME = '1tv'
|
|
|
|
|
IE_DESC = 'Первый канал'
|
2016-08-22 00:55:47 +08:00
|
|
|
|
_VALID_URL = r'https?://(?:www\.)?1tv\.ru/(?:[^/]+/)+(?P<id>[^/?#]+)'
|
2014-02-11 01:20:41 +08:00
|
|
|
|
|
2015-02-14 04:04:28 +08:00
|
|
|
|
_TESTS = [{
|
2016-08-22 02:11:51 +08:00
|
|
|
|
# single format
|
2016-08-22 00:55:47 +08:00
|
|
|
|
'url': 'http://www.1tv.ru/shows/naedine-so-vsemi/vypuski/gost-lyudmila-senchina-naedine-so-vsemi-vypusk-ot-12-02-2015',
|
|
|
|
|
'md5': 'a1b6b60d530ebcf8daacf4565762bbaf',
|
2014-02-11 01:20:41 +08:00
|
|
|
|
'info_dict': {
|
2016-08-22 00:55:47 +08:00
|
|
|
|
'id': '40049',
|
2014-02-11 01:20:41 +08:00
|
|
|
|
'ext': 'mp4',
|
2016-12-05 00:20:14 +08:00
|
|
|
|
'title': 'Гость Людмила Сенчина. Наедине со всеми. Выпуск от 12.02.2015',
|
2017-01-02 20:08:07 +08:00
|
|
|
|
'thumbnail': r're:^https?://.*\.(?:jpg|JPG)$',
|
2016-04-10 05:02:35 +08:00
|
|
|
|
'upload_date': '20150212',
|
|
|
|
|
'duration': 2694,
|
2014-02-11 11:26:52 +08:00
|
|
|
|
},
|
2015-02-14 04:04:28 +08:00
|
|
|
|
}, {
|
2016-08-22 02:11:51 +08:00
|
|
|
|
# multiple formats
|
2016-08-22 00:55:47 +08:00
|
|
|
|
'url': 'http://www.1tv.ru/shows/dobroe-utro/pro-zdorove/vesennyaya-allergiya-dobroe-utro-fragment-vypuska-ot-07042016',
|
2016-08-22 02:11:51 +08:00
|
|
|
|
'info_dict': {
|
|
|
|
|
'id': '364746',
|
|
|
|
|
'ext': 'mp4',
|
2016-12-05 00:20:14 +08:00
|
|
|
|
'title': 'Весенняя аллергия. Доброе утро. Фрагмент выпуска от 07.04.2016',
|
2017-01-02 20:08:07 +08:00
|
|
|
|
'thumbnail': r're:^https?://.*\.(?:jpg|JPG)$',
|
2016-08-22 02:11:51 +08:00
|
|
|
|
'upload_date': '20160407',
|
|
|
|
|
'duration': 179,
|
|
|
|
|
'formats': 'mincount:3',
|
|
|
|
|
},
|
|
|
|
|
'params': {
|
|
|
|
|
'skip_download': True,
|
|
|
|
|
},
|
2016-12-05 00:20:14 +08:00
|
|
|
|
}, {
|
|
|
|
|
'url': 'http://www.1tv.ru/news/issue/2016-12-01/14:00',
|
|
|
|
|
'info_dict': {
|
|
|
|
|
'id': '14:00',
|
|
|
|
|
'title': 'Выпуск новостей в 14:00 1 декабря 2016 года. Новости. Первый канал',
|
|
|
|
|
'description': 'md5:2e921b948f8c1ff93901da78ebdb1dfd',
|
|
|
|
|
},
|
|
|
|
|
'playlist_count': 13,
|
|
|
|
|
}, {
|
|
|
|
|
'url': 'http://www.1tv.ru/shows/tochvtoch-supersezon/vystupleniya/evgeniy-dyatlov-vladimir-vysockiy-koni-priveredlivye-toch-v-toch-supersezon-fragment-vypuska-ot-06-11-2016',
|
|
|
|
|
'only_matching': True,
|
2015-02-14 04:04:28 +08:00
|
|
|
|
}]
|
2014-02-11 01:20:41 +08:00
|
|
|
|
|
|
|
|
|
def _real_extract(self, url):
|
2016-08-22 00:55:47 +08:00
|
|
|
|
display_id = self._match_id(url)
|
2014-02-11 01:20:41 +08:00
|
|
|
|
|
2016-08-22 00:55:47 +08:00
|
|
|
|
webpage = self._download_webpage(url, display_id)
|
|
|
|
|
playlist_url = compat_urlparse.urljoin(url, self._search_regex(
|
2016-12-05 00:20:14 +08:00
|
|
|
|
r'data-playlist-url=(["\'])(?P<url>(?:(?!\1).)+)\1',
|
|
|
|
|
webpage, 'playlist url', group='url'))
|
|
|
|
|
|
|
|
|
|
parsed_url = compat_urlparse.urlparse(playlist_url)
|
|
|
|
|
qs = compat_urlparse.parse_qs(parsed_url.query)
|
|
|
|
|
item_ids = qs.get('videos_ids[]') or qs.get('news_ids[]')
|
|
|
|
|
|
|
|
|
|
items = self._download_json(playlist_url, display_id)
|
|
|
|
|
|
|
|
|
|
if item_ids:
|
|
|
|
|
items = [
|
|
|
|
|
item for item in items
|
|
|
|
|
if item.get('uid') and compat_str(item['uid']) in item_ids]
|
|
|
|
|
else:
|
|
|
|
|
items = [items[0]]
|
|
|
|
|
|
|
|
|
|
entries = []
|
|
|
|
|
QUALITIES = ('ld', 'sd', 'hd', )
|
|
|
|
|
|
|
|
|
|
for item in items:
|
|
|
|
|
title = item['title']
|
|
|
|
|
quality = qualities(QUALITIES)
|
|
|
|
|
formats = []
|
2017-01-20 01:47:04 +08:00
|
|
|
|
path = None
|
2016-12-05 00:20:14 +08:00
|
|
|
|
for f in item.get('mbr', []):
|
|
|
|
|
src = f.get('src')
|
|
|
|
|
if not src or not isinstance(src, compat_str):
|
|
|
|
|
continue
|
|
|
|
|
tbr = int_or_none(self._search_regex(
|
|
|
|
|
r'_(\d{3,})\.mp4', src, 'tbr', default=None))
|
2017-01-20 01:47:04 +08:00
|
|
|
|
if not path:
|
|
|
|
|
path = self._search_regex(
|
|
|
|
|
r'//[^/]+/(.+?)_\d+\.mp4', src,
|
|
|
|
|
'm3u8 path', default=None)
|
2016-12-05 00:20:14 +08:00
|
|
|
|
formats.append({
|
|
|
|
|
'url': src,
|
|
|
|
|
'format_id': f.get('name'),
|
|
|
|
|
'tbr': tbr,
|
2017-01-20 01:47:04 +08:00
|
|
|
|
'source_preference': quality(f.get('name')),
|
2017-06-01 23:19:52 +08:00
|
|
|
|
# quality metadata of http formats may be incorrect
|
|
|
|
|
'preference': -1,
|
2016-12-05 00:20:14 +08:00
|
|
|
|
})
|
2017-01-20 01:47:04 +08:00
|
|
|
|
# m3u8 URL format is reverse engineered from [1] (search for
|
|
|
|
|
# master.m3u8). dashEdges (that is currently balancer-vod.1tv.ru)
|
|
|
|
|
# is taken from [2].
|
|
|
|
|
# 1. http://static.1tv.ru/player/eump1tv-current/eump-1tv.all.min.js?rnd=9097422834:formatted
|
|
|
|
|
# 2. http://static.1tv.ru/player/eump1tv-config/config-main.js?rnd=9097422834
|
|
|
|
|
if not path and len(formats) == 1:
|
|
|
|
|
path = self._search_regex(
|
|
|
|
|
r'//[^/]+/(.+?$)', formats[0]['url'],
|
|
|
|
|
'm3u8 path', default=None)
|
|
|
|
|
if path:
|
|
|
|
|
if len(formats) == 1:
|
|
|
|
|
m3u8_path = ','
|
|
|
|
|
else:
|
|
|
|
|
tbrs = [compat_str(t) for t in sorted(f['tbr'] for f in formats)]
|
|
|
|
|
m3u8_path = '_,%s,%s' % (','.join(tbrs), '.mp4')
|
|
|
|
|
formats.extend(self._extract_m3u8_formats(
|
|
|
|
|
'http://balancer-vod.1tv.ru/%s%s.urlset/master.m3u8'
|
|
|
|
|
% (path, m3u8_path),
|
|
|
|
|
display_id, 'mp4',
|
|
|
|
|
entry_protocol='m3u8_native', m3u8_id='hls', fatal=False))
|
2016-12-05 00:20:14 +08:00
|
|
|
|
self._sort_formats(formats)
|
|
|
|
|
|
|
|
|
|
thumbnail = item.get('poster') or self._og_search_thumbnail(webpage)
|
|
|
|
|
duration = int_or_none(item.get('duration') or self._html_search_meta(
|
|
|
|
|
'video:duration', webpage, 'video duration', fatal=False))
|
|
|
|
|
upload_date = unified_strdate(self._html_search_meta(
|
|
|
|
|
'ya:ovs:upload_date', webpage, 'upload date', default=None))
|
2014-02-11 01:20:41 +08:00
|
|
|
|
|
2016-12-05 00:20:14 +08:00
|
|
|
|
entries.append({
|
2016-12-06 00:28:57 +08:00
|
|
|
|
'id': compat_str(item.get('id') or item['uid']),
|
2016-12-05 00:20:14 +08:00
|
|
|
|
'thumbnail': thumbnail,
|
|
|
|
|
'title': title,
|
|
|
|
|
'upload_date': upload_date,
|
|
|
|
|
'duration': int_or_none(duration),
|
|
|
|
|
'formats': formats
|
2016-08-22 00:55:47 +08:00
|
|
|
|
})
|
2016-04-10 05:02:35 +08:00
|
|
|
|
|
2016-08-22 00:55:47 +08:00
|
|
|
|
title = self._html_search_regex(
|
|
|
|
|
(r'<div class="tv_translation">\s*<h1><a href="[^"]+">([^<]*)</a>',
|
|
|
|
|
r"'title'\s*:\s*'([^']+)'"),
|
2016-12-05 00:20:14 +08:00
|
|
|
|
webpage, 'title', default=None) or self._og_search_title(
|
|
|
|
|
webpage, default=None)
|
2016-08-22 00:55:47 +08:00
|
|
|
|
description = self._html_search_regex(
|
|
|
|
|
r'<div class="descr">\s*<div> </div>\s*<p>([^<]*)</p></div>',
|
|
|
|
|
webpage, 'description', default=None) or self._html_search_meta(
|
2016-12-05 00:20:14 +08:00
|
|
|
|
'description', webpage, 'description', default=None)
|
2014-02-11 01:20:41 +08:00
|
|
|
|
|
2016-12-05 00:20:14 +08:00
|
|
|
|
return self.playlist_result(entries, display_id, title, description)
|