yt-dlp/yt_dlp/extractor/rtp.py

99 lines
3.3 KiB
Python
Raw Normal View History

import base64
import json
import re
import urllib.parse
from .common import InfoExtractor
from ..utils import js_to_json
2014-12-12 19:22:24 +01:00
class RTPIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?rtp\.pt/play/p(?P<program_id>[0-9]+)/(?P<id>[^/?#]+)/?'
_TESTS = [{
2014-12-12 19:22:24 +01:00
'url': 'http://www.rtp.pt/play/p405/e174042/paixoes-cruzadas',
'md5': 'e736ce0c665e459ddb818546220b4ef8',
2014-12-12 19:22:24 +01:00
'info_dict': {
2014-12-21 15:28:40 +01:00
'id': 'e174042',
2014-12-12 19:22:24 +01:00
'ext': 'mp3',
'title': 'Paixões Cruzadas',
'description': 'As paixões musicais de António Cartaxo e António Macedo',
'thumbnail': r're:^https?://.*\.jpg',
2014-12-12 19:22:24 +01:00
},
}, {
'url': 'http://www.rtp.pt/play/p831/a-quimica-das-coisas',
'only_matching': True,
}]
2014-12-12 19:22:24 +01:00
_RX_OBFUSCATION = re.compile(r'''(?xs)
atob\s*\(\s*decodeURIComponent\s*\(\s*
(\[[0-9A-Za-z%,'"]*\])
\s*\.\s*join\(\s*(?:""|'')\s*\)\s*\)\s*\)
''')
def __unobfuscate(self, data, *, video_id):
if data.startswith('{'):
data = self._RX_OBFUSCATION.sub(
lambda m: json.dumps(
base64.b64decode(urllib.parse.unquote(
''.join(self._parse_json(m.group(1), video_id)),
)).decode('iso-8859-1')),
data)
return js_to_json(data)
2014-12-12 19:22:24 +01:00
def _real_extract(self, url):
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
title = self._html_search_meta(
'twitter:title', webpage, display_name='title', fatal=True)
f, config = self._search_regex(
r'''(?sx)
var\s+f\s*=\s*(?P<f>".*?"|{[^;]+?});\s*
var\s+player1\s+=\s+new\s+RTPPlayer\s*\((?P<config>{(?:(?!\*/).)+?})\);(?!\s*\*/)
''', webpage,
'player config', group=('f', 'config'))
f = self._parse_json(
f, video_id,
lambda data: self.__unobfuscate(data, video_id=video_id))
config = self._parse_json(
config, video_id,
lambda data: self.__unobfuscate(data, video_id=video_id))
formats = []
if isinstance(f, dict):
f_hls = f.get('hls')
if f_hls is not None:
formats.extend(self._extract_m3u8_formats(
f_hls, video_id, 'mp4', 'm3u8_native', m3u8_id='hls'))
f_dash = f.get('dash')
if f_dash is not None:
formats.extend(self._extract_mpd_formats(f_dash, video_id, mpd_id='dash'))
2019-05-28 05:58:12 +02:00
else:
formats.append({
'format_id': 'f',
'url': f,
'vcodec': 'none' if config.get('mediaType') == 'audio' else None,
})
subtitles = {}
vtt = config.get('vtt')
if vtt is not None:
for lcode, lname, url in vtt:
subtitles.setdefault(lcode, []).append({
'name': lname,
'url': url,
})
2014-12-12 19:22:24 +01:00
return {
'id': video_id,
'title': title,
'formats': formats,
2019-05-28 05:58:12 +02:00
'description': self._html_search_meta(['description', 'twitter:description'], webpage),
'thumbnail': config.get('poster') or self._og_search_thumbnail(webpage),
'subtitles': subtitles,
2014-12-12 19:22:24 +01:00
}