_ 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
  47     def _playlist_videos_info(self, url, name, playlist_id):
  48         '''Returns the videos of the playlist'''
  49
  50         webpage = self._download_webpage(
  51             url, playlist_id, u'Downloading playlist webpage')
  52         matches = re.finditer(
  53             r'<p\s+class="talk-title[^"]*"><a\s+href="(?P<talk_url>/talks/[^"]+\.html)">[^<]*</a></p>',
  54             webpage)
  55
  56         playlist_title = self._html_search_regex(r'div class="headline">\s*?<h1>\s*?<span>(.*?)</span>',
  57                                                  webpage, 'playlist title')
  58
  59         playlist_entries = [
  60             self.url_result(u'http://www.ted.com' + m.group('talk_url'), 'TED')
  61             for m in matches
  62         ]
  63         return self.playlist_result(
  64             playlist_entries, playlist_id=playlist_id, playlist_title=playlist_title)
  65
  66     def _talk_info(self, url, video_id=0):
  67         """Return the video for the talk in the url"""
  68         m = re.match(self._VALID_URL, url,re.VERBOSE)
  69         video_name = m.group('name')
  70         webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
  71         self.report_extraction(video_name)
  72         # If the url includes the language we get the title translated
  73         title = self._html_search_regex(r'<span .*?id="altHeadline".+?>(?P<title>.*)</span>',
  74                                         webpage, 'title')
  75         json_data = self._search_regex(r'<script.*?>var talkDetails = ({.*?})</script>',
  76                                     webpage, 'json data')
  77         info = json.loads(json_data)
  78         desc = self._html_search_regex(r'<div class="talk-intro">.*?<p.*?>(.*?)</p>',
  79                                        webpage, 'description', flags = re.DOTALL)
  80
  81         thumbnail = self._search_regex(r'</span>[\s.]*</div>[\s.]*<img src="(.*?)"',
  82                                        webpage, 'thumbnail')
  83         formats = [{
  84             'ext': 'mp4',
  85             'url': stream['file'],
  86             'format': stream['id']
  87         } for stream in info['htmlStreams']]
  88
  89         video_id = info['id']
  90
  91         # subtitles
  92         video_subtitles = self.extract_subtitles(video_id, webpage)
  93         if self._downloader.params.get('listsubtitles', False):
  94             self._list_available_subtitles(video_id, webpage)
  95             return
  96
  97         return {
  98             'id': video_id,
  99             'title': title,
 100             'thumbnail': thumbnail,
 101             'description': desc,
 102             'subtitles': video_subtitles,
 103             'formats': formats,
 104         }
 105
 106     def _get_available_subtitles(self, video_id, webpage):
 107         try:
 108             options = self._search_regex(r'(?:<select name="subtitles_language_select" id="subtitles_language_select">)(.*?)(?:</select>)', webpage, 'subtitles_language_select', flags=re.DOTALL)
 109             languages = re.findall(r'(?:<option value=")(\S+)"', options)
 110             if languages:
 111                 sub_lang_list = {}
 112                 for l in languages:
 113                     url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
 114                     sub_lang_list[l] = url
 115                 return sub_lang_list
 116         except RegexNotFoundError as err:
 117             self._downloader.report_warning(u'video doesn\'t have subtitles')
 118         return {}