youtube_dl/extractor/ted.py

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