Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/ted.py

   1 from __future__ import unicode_literals
   2
   3 import json
   4 import re
   5
   6 from .subtitles import SubtitlesInfoExtractor
   7
   8 from ..utils import (
   9     RegexNotFoundError,
  10 )
  11
  12
  13 class TEDIE(SubtitlesInfoExtractor):
  14     _VALID_URL=r'''http://www\.ted\.com/
  15                    (
  16                         ((?P<type_playlist>playlists)/(?P<playlist_id>\d+)) # We have a playlist
  17                         |
  18                         ((?P<type_talk>talks)) # We have a simple talk
  19                    )
  20                    (/lang/(.*?))? # The url may contain the language
  21                    /(?P<name>\w+) # Here goes the name and then ".html"
  22                    '''
  23     _TEST = {
  24         'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
  25         'file': '102.mp4',
  26         'md5': '4ea1dada91e4174b53dac2bb8ace429d',
  27         'info_dict': {
  28             "description": "md5:c6fa72e6eedbd938c9caf6b2702f5922",
  29             "title": "Dan Dennett: The illusion of consciousness"
  30         }
  31     }
  32
  33     @classmethod
  34     def suitable(cls, url):
  35         """Receives a URL and returns True if suitable for this IE."""
  36         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
  37
  38     def _real_extract(self, url):
  39         m=re.match(self._VALID_URL, url, re.VERBOSE)
  40         if m.group('type_talk'):
  41             return self._talk_info(url)
  42         else :
  43             playlist_id=m.group('playlist_id')
  44             name=m.group('name')
  45             self.to_screen(u'Getting info of playlist %s: "%s"' % (playlist_id,name))
  46             return [self._playlist_videos_info(url,name,playlist_id)]
  47
  48
  49     def _playlist_videos_info(self, url, name, playlist_id):
  50         '''Returns the videos of the playlist'''
  51
  52         webpage = self._download_webpage(
  53             url, playlist_id, 'Downloading playlist webpage')
  54         matches = re.finditer(
  55             r'<p\s+class="talk-title[^"]*"><a\s+href="(?P<talk_url>/talks/[^"]+\.html)">[^<]*</a></p>',
  56             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             self.url_result(u'http://www.ted.com' + m.group('talk_url'), 'TED')
  63             for m in matches
  64         ]
  65         return self.playlist_result(
  66             playlist_entries, playlist_id=playlist_id, playlist_title=playlist_title)
  67
  68     def _talk_info(self, url, video_id=0):
  69         """Return the video for the talk in the url"""
  70         m = re.match(self._VALID_URL, url,re.VERBOSE)
  71         video_name = m.group('name')
  72         webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
  73         self.report_extraction(video_name)
  74         # If the url includes the language we get the title translated
  75         title = self._html_search_regex(r'<span .*?id="altHeadline".+?>(?P<title>.*)</span>',
  76                                         webpage, 'title')
  77         json_data = self._search_regex(r'<script.*?>var talkDetails = ({.*?})</script>',
  78                                     webpage, 'json data')
  79         info = json.loads(json_data)
  80         desc = self._html_search_regex(r'<div class="talk-intro">.*?<p.*?>(.*?)</p>',
  81                                        webpage, 'description', flags = re.DOTALL)
  82
  83         thumbnail = self._search_regex(r'</span>[\s.]*</div>[\s.]*<img src="(.*?)"',
  84                                        webpage, 'thumbnail')
  85         formats = [{
  86             'ext': 'mp4',
  87             'url': stream['file'],
  88             'format': stream['id']
  89         } for stream in info['htmlStreams']]
  90
  91         video_id = info['id']
  92
  93         # subtitles
  94         video_subtitles = self.extract_subtitles(video_id, webpage)
  95         if self._downloader.params.get('listsubtitles', False):
  96             self._list_available_subtitles(video_id, webpage)
  97             return
  98
  99         return {
 100             'id': video_id,
 101             'title': title,
 102             'thumbnail': thumbnail,
 103             'description': desc,
 104             'subtitles': video_subtitles,
 105             'formats': formats,
 106         }
 107
 108     def _get_available_subtitles(self, video_id, webpage):
 109         try:
 110             options = self._search_regex(r'(?:<select name="subtitles_language_select" id="subtitles_language_select">)(.*?)(?:</select>)', webpage, 'subtitles_language_select', flags=re.DOTALL)
 111             languages = re.findall(r'(?:<option value=")(\S+)"', options)
 112             if languages:
 113                 sub_lang_list = {}
 114                 for l in languages:
 115                     url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
 116                     sub_lang_list[l] = url
 117                 return sub_lang_list
 118         except RegexNotFoundError:
 119             self._downloader.report_warning(u'video doesn\'t have subtitles')
 120         return {}