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

   1 # encoding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import json
   6 import base64
   7 import xml.etree.ElementTree
   8
   9 # Soompi uses the same subtitle encryption as crunchyroll
  10 from .crunchyroll import CrunchyrollIE
  11
  12
  13 class SoompiIE(CrunchyrollIE):
  14     IE_NAME = 'soompi'
  15     _VALID_URL = r'^https?://tv\.soompi\.com/en/watch/(?P<id>[0-9]+)'
  16     _TESTS = [{
  17         'url': 'http://tv.soompi.com/en/watch/23363',
  18         'info_dict': {
  19             'id': '23363',
  20             'ext': 'mp4',
  21             'title': 'Liar Game CM1',
  22             'description': '15sec'
  23         },
  24         'params': {
  25             'skip_download': True,
  26         },
  27     }]
  28
  29     def _get_episodes(self, webpage, episode_filter=None):
  30         episodes = json.loads(
  31             self._search_regex(r'\s+VIDEOS\s+= (\[.+?\]);', webpage, "episodes meta"))
  32         return [ep for ep in episodes if episode_filter is None or episode_filter(ep)]
  33
  34     def _get_subtitles(self, video_id, show_format_xml):
  35         subtitles = {}
  36         subtitle_info_nodes = show_format_xml.findall('./{default}preload/subtitles/subtitle')
  37         subtitle_nodes = show_format_xml.findall('./{default}preload/subtitle')
  38
  39         sub_langs = {}
  40         for i in subtitle_info_nodes:
  41             sub_langs[i.attrib["id"]] = i.attrib["title"]
  42
  43         for s in subtitle_nodes:
  44             lang_code = sub_langs.get(s.attrib["id"], None)
  45             if lang_code is None:
  46                 continue
  47
  48             sub_id = int(s.attrib["id"])
  49             iv = base64.b64decode(s.find("iv").text)
  50             data = base64.b64decode(s.find("data").text)
  51             subtitle = self._decrypt_subtitles(data, iv, sub_id).decode('utf-8')
  52             sub_root = xml.etree.ElementTree.fromstring(subtitle)
  53
  54             subtitles[lang_code] = [{
  55                 'ext': 'srt', 'data': self._convert_subtitles_to_srt(sub_root)
  56             }, {
  57                 'ext': 'ass', 'data': self._convert_subtitles_to_ass(sub_root)
  58             }]
  59         return subtitles
  60
  61     def _real_extract(self, url):
  62         video_id = self._match_id(url)
  63
  64         webpage = self._download_webpage(
  65             url, video_id, note="Downloading episode page",
  66             errnote="Video may not be available for your location")
  67         vid_formats = re.findall(r"\?quality=q([0-9]+)", webpage)
  68
  69         show_meta = json.loads(
  70             self._search_regex(r'\s+var show = (\{.+?\});', webpage, "show meta"))
  71         episodes = self._get_episodes(
  72             webpage, episode_filter=lambda x: x['id'] == video_id)
  73
  74         title = episodes[0]["name"]
  75         description = episodes[0]["description"]
  76         duration = int(episodes[0]["duration"])
  77         slug = show_meta["slug"]
  78
  79         formats = []
  80         show_format_xml = None
  81         for vf in vid_formats:
  82             show_format_url = "http://tv.soompi.com/en/show/%s/%s-config.xml?mode=hls&quality=q%s" \
  83                               % (slug, video_id, vf)
  84             show_format_xml = self._download_xml(
  85                 show_format_url, video_id, note="Downloading q%s show xml" % vf)
  86             avail_formats = self._extract_m3u8_formats(
  87                 show_format_xml.find('./{default}preload/stream_info/file').text,
  88                 video_id, ext="mp4", m3u8_id=vf, preference=int(vf))
  89             formats.extend(avail_formats)
  90         self._sort_formats(formats)
  91
  92         subtitles = self.extract_subtitles(video_id, show_format_xml)
  93
  94         return {
  95             'id': video_id,
  96             'title': title,
  97             'description': description,
  98             'duration': duration,
  99             'formats': formats,
 100             'subtitles': subtitles
 101         }
 102
 103
 104 class SoompiShowIE(SoompiIE):
 105     IE_NAME = 'soompi:show'
 106     _VALID_URL = r'^https?://tv\.soompi\.com/en/shows/(?P<id>[0-9a-zA-Z\-_]+)'
 107     _TESTS = [{
 108         'url': 'http://tv.soompi.com/en/shows/liar-game',
 109         'info_dict': {
 110             'id': 'liar-game',
 111             'title': 'Liar Game',
 112             'description': 'md5:52c02bce0c1a622a95823591d0589b66',
 113         },
 114         'playlist_count': 14,
 115     }]
 116
 117     def _real_extract(self, url):
 118         show_id = self._match_id(url)
 119
 120         webpage = self._download_webpage(url, show_id, note="Downloading show page")
 121         title = self._og_search_title(webpage).replace("SoompiTV | ", "")
 122         description = self._og_search_description(webpage)
 123
 124         episodes = self._get_episodes(webpage)
 125         entries = []
 126         for ep in episodes:
 127             entries.append(self.url_result(
 128                 'http://tv.soompi.com/en/watch/%s' % ep['id'], 'Soompi', ep['id']))
 129
 130         return self.playlist_result(entries, show_id, title, description)