_ Git - youtube-dl/blob - youtube_dl/extractor/ora.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 from .common import InfoExtractor
   6 from ..compat import compat_urlparse
   7 from ..utils import (
   8     get_element_by_attribute,
   9     js_to_json,
  10     qualities,
  11     unescapeHTML,
  12 )
  13
  14
  15 class OraTVIE(InfoExtractor):
  16     _VALID_URL = r'https?://(?:www\.)?(ora\.tv|unsafespeech\.com)/([^/]+/)*(?P<id>[^/\?#]+)'
  17     _TEST = {
  18         'url': 'https://www.ora.tv/larrykingnow/2015/12/16/vine-youtube-stars-zach-king-king-bach-on-their-viral-videos-0_36jupg6090pq',
  19         'md5': 'fa33717591c631ec93b04b0e330df786',
  20         'info_dict': {
  21             'id': '50178',
  22             'ext': 'mp4',
  23             'title': 'Vine & YouTube Stars Zach King & King Bach On Their Viral Videos!',
  24             'description': 'md5:ebbc5b1424dd5dba7be7538148287ac1',
  25         }
  26     }
  27
  28     def _real_extract(self, url):
  29         display_id = self._match_id(url)
  30         webpage = self._download_webpage(url, display_id)
  31
  32         ora_meta = self._parse_json(self._search_regex(
  33             r'(?s);\s*ora_meta = ({.*?});</script>', webpage, 'ora_meta'), display_id,
  34             transform_source=lambda data: js_to_json(re.sub('":(document|\().*?(:false|\(\)),', '":null,', data)))
  35
  36         video_data = ora_meta.get('video', ora_meta.get('current'))
  37         m3u8_url = video_data['hls_stream']
  38
  39         if m3u8_url:
  40             formats = self._extract_m3u8_formats(
  41                 m3u8_url, display_id, 'mp4', 'm3u8_native',
  42                 m3u8_id='hls', fatal=False)
  43             # similar to GameSpotIE
  44             m3u8_path = compat_urlparse.urlparse(m3u8_url).path
  45             QUALITIES_RE = r'((,[a-z]+\d+)+,?)'
  46             available_qualities = self._search_regex(
  47                 QUALITIES_RE, m3u8_path, 'qualities').strip(',').split(',')
  48             http_path = m3u8_path[1:].split('/', 1)[1]
  49             http_template = re.sub(QUALITIES_RE, r'%s', http_path)
  50             http_template = http_template.replace('.csmil/master.m3u8', '')
  51             http_template = compat_urlparse.urljoin(
  52                 'http://videocdn-pmd.ora.tv/', http_template)
  53             preference = qualities(
  54                 ['mobile400', 'basic400', 'basic600', 'sd900', 'sd1200', 'sd1500', 'hd720', 'hd1080'])
  55             for q in available_qualities:
  56                 formats.append({
  57                     'url': http_template % q,
  58                     'format_id': q,
  59                     'preference': preference(q),
  60                 })
  61             self._sort_formats(formats)
  62         else:
  63             return self.url_result(self._search_regex(
  64                 r'"youtube_id"\s*:\s*"([^"]+)', webpage, 'youtube id'), 'Youtube')
  65
  66         return {
  67             'id': video_data.get('id', display_id),
  68             'display_id': display_id,
  69             'title': unescapeHTML(self._og_search_title(webpage)),
  70             'description': get_element_by_attribute(
  71                 'class', 'video_txt_decription', webpage),
  72             'thumbnail': self._proto_relative_url(video_data.get('thumb')),
  73             'formats': formats,
  74         }