[tv5mondeplus] Fix extractor (#739)

[yt-dlp.git] / yt_dlp / extractor / mychannels.py

# coding: utf-8
from __future__ import unicode_literals

import re

from .common import InfoExtractor


class MyChannelsIE(InfoExtractor):
    _VALID_URL = r'https?://(?:www\.)?mychannels\.com/.*(?P<id_type>video|production)_id=(?P<id>[0-9]+)'
    _TEST = {
        'url': 'https://mychannels.com/missholland/miss-holland?production_id=3416',
        'md5': 'b8993daad4262dd68d89d651c0c52c45',
        'info_dict': {
            'id': 'wUUDZZep6vQD',
            'ext': 'mp4',
            'title': 'Miss Holland joins VOTE LEAVE',
            'description': 'Miss Holland | #13 Not a potato',
            'uploader': 'Miss Holland',
        }
    }

    def _real_extract(self, url):
        id_type, url_id = re.match(self._VALID_URL, url).groups()
        webpage = self._download_webpage(url, url_id)
        video_data = self._html_search_regex(r'<div([^>]+data-%s-id="%s"[^>]+)>' % (id_type, url_id), webpage, 'video data')

        def extract_data_val(attr, fatal=False):
            return self._html_search_regex(r'data-%s\s*=\s*"([^"]+)"' % attr, video_data, attr, fatal=fatal)
        minoto_id = extract_data_val('minoto-id') or self._search_regex(r'/id/([a-zA-Z0-9]+)', extract_data_val('video-src', True), 'minoto id')

        return {
            '_type': 'url_transparent',
            'url': 'minoto:%s' % minoto_id,
            'id': url_id,
            'title': extract_data_val('title', True),
            'description': extract_data_val('description'),
            'thumbnail': extract_data_val('image'),
            'uploader': extract_data_val('channel'),
        }