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

   1 import json
   2 import re
   3
   4 from .subtitles import SubtitlesInfoExtractor
   5
   6 from ..utils import (
   7     compat_str,
   8     RegexNotFoundError,
   9 )
  10
  11 class TEDIE(SubtitlesInfoExtractor):
  12     _VALID_URL=r'''http://www\.ted\.com/
  13                    (
  14                         ((?P<type_playlist>playlists)/(?P<playlist_id>\d+)) # We have a playlist
  15                         |
  16                         ((?P<type_talk>talks)) # We have a simple talk
  17                    )
  18                    (/lang/(.*?))? # The url may contain the language
  19                    /(?P<name>\w+) # Here goes the name and then ".html"
  20                    '''
  21     _TEST = {
  22         u'url': u'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
  23         u'file': u'102.mp4',
  24         u'md5': u'2d76ee1576672e0bd8f187513267adf6',
  25         u'info_dict': {
  26             u"description": u"md5:c6fa72e6eedbd938c9caf6b2702f5922",
  27             u"title": u"Dan Dennett: The illusion of consciousness"
  28         }
  29     }
  30
  31     @classmethod
  32     def suitable(cls, url):
  33         """Receives a URL and returns True if suitable for this IE."""
  34         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
  35
  36     def _real_extract(self, url):
  37         m=re.match(self._VALID_URL, url, re.VERBOSE)
  38         if m.group('type_talk'):
  39             return self._talk_info(url)
  40         else :
  41             playlist_id=m.group('playlist_id')
  42             name=m.group('name')
  43             self.to_screen(u'Getting info of playlist %s: "%s"' % (playlist_id,name))
  44             return [self._playlist_videos_info(url,name,playlist_id)]
  45
  46     def _playlist_videos_info(self,url,name,playlist_id=0):
  47         '''Returns the videos of the playlist'''
  48         video_RE=r'''
  49                      <li\ id="talk_(\d+)"([.\s]*?)data-id="(?P<video_id>\d+)"
  50                      ([.\s]*?)data-playlist_item_id="(\d+)"
  51                      ([.\s]*?)data-mediaslug="(?P<mediaSlug>.+?)"
  52                      '''
  53         video_name_RE=r'<p\ class="talk-title"><a href="(?P<talk_url>/talks/(.+).html)">(?P<fullname>.+?)</a></p>'
  54         webpage=self._download_webpage(url, playlist_id, 'Downloading playlist webpage')
  55         m_videos=re.finditer(video_RE,webpage,re.VERBOSE)
  56         m_names=re.finditer(video_name_RE,webpage)
  57
  58         playlist_title = self._html_search_regex(r'div class="headline">\s*?<h1>\s*?<span>(.*?)</span>',
  59                                                  webpage, 'playlist title')
  60
  61         playlist_entries = []
  62         for m_video, m_name in zip(m_videos,m_names):
  63             talk_url='http://www.ted.com%s' % m_name.group('talk_url')
  64             playlist_entries.append(self.url_result(talk_url, 'TED'))
  65         return self.playlist_result(playlist_entries, playlist_id = playlist_id, playlist_title = playlist_title)
  66
  67     def _talk_info(self, url, video_id=0):
  68         """Return the video for the talk in the url"""
  69         m = re.match(self._VALID_URL, url,re.VERBOSE)
  70         video_name = m.group('name')
  71         webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
  72         self.report_extraction(video_name)
  73         # If the url includes the language we get the title translated
  74         title = self._html_search_regex(r'<span .*?id="altHeadline".+?>(?P<title>.*)</span>',
  75                                         webpage, 'title')
  76         json_data = self._search_regex(r'<script.*?>var talkDetails = ({.*?})</script>',
  77                                     webpage, 'json data')
  78         info = json.loads(json_data)
  79         desc = self._html_search_regex(r'<div class="talk-intro">.*?<p.*?>(.*?)</p>',
  80                                        webpage, 'description', flags = re.DOTALL)
  81
  82         thumbnail = self._search_regex(r'</span>[\s.]*</div>[\s.]*<img src="(.*?)"',
  83                                        webpage, 'thumbnail')
  84         formats = [{
  85             'ext': 'mp4',
  86             'url': stream['file'],
  87             'format': stream['id']
  88             } for stream in info['htmlStreams']]
  89
  90         video_id = info['id']
  91
  92         # subtitles
  93         video_subtitles = self.extract_subtitles(video_id, webpage)
  94         if self._downloader.params.get('listsubtitles', False):
  95             self._list_available_subtitles(video_id, webpage)
  96             return
  97
  98         info = {
  99             'id': video_id,
 100             'title': title,
 101             'thumbnail': thumbnail,
 102             'description': desc,
 103             'subtitles': video_subtitles,
 104             'formats': formats,
 105         }
 106
 107         # TODO: Remove when #980 has been merged
 108         info.update(info['formats'][-1])
 109
 110         return info
 111
 112     def _get_available_subtitles(self, video_id, webpage):
 113         try:
 114             options = self._search_regex(r'(?:<select name="subtitles_language_select" id="subtitles_language_select">)(.*?)(?:</select>)', webpage, 'subtitles_language_select', flags=re.DOTALL)
 115             languages = re.findall(r'(?:<option value=")(\S+)"', options)
 116             if languages:
 117                 sub_lang_list = {}
 118                 for l in languages:
 119                     url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
 120                     sub_lang_list[l] = url
 121                 return sub_lang_list
 122         except RegexNotFoundError as err:
 123             self._downloader.report_warning(u'video doesn\'t have subtitles')
 124         return {}