7
0
mirror of https://github.com/ytdl-org/youtube-dl.git synced 2024-12-15 01:36:31 +00:00
youtube-dl/youtube_dl/extractor/ted.py

121 lines
4.8 KiB
Python
Raw Normal View History

2014-01-17 02:52:17 +00:00
from __future__ import unicode_literals
2013-06-23 19:55:53 +00:00
import json
import re
from .subtitles import SubtitlesInfoExtractor
2013-06-23 19:55:53 +00:00
from ..utils import (
RegexNotFoundError,
)
2014-01-17 02:52:17 +00:00
class TEDIE(SubtitlesInfoExtractor):
2013-06-23 19:55:53 +00:00
_VALID_URL=r'''http://www\.ted\.com/
(
((?P<type_playlist>playlists)/(?P<playlist_id>\d+)) # We have a playlist
|
((?P<type_talk>talks)) # We have a simple talk
)
(/lang/(.*?))? # The url may contain the language
/(?P<name>\w+) # Here goes the name and then ".html"
'''
2013-06-27 18:46:46 +00:00
_TEST = {
2014-01-17 02:52:17 +00:00
'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
'file': '102.mp4',
2014-01-17 02:54:54 +00:00
'md5': '4ea1dada91e4174b53dac2bb8ace429d',
2014-01-17 02:52:17 +00:00
'info_dict': {
"description": "md5:c6fa72e6eedbd938c9caf6b2702f5922",
"title": "Dan Dennett: The illusion of consciousness"
2013-06-27 18:46:46 +00:00
}
}
2013-06-23 19:55:53 +00:00
@classmethod
def suitable(cls, url):
"""Receives a URL and returns True if suitable for this IE."""
return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
def _real_extract(self, url):
m=re.match(self._VALID_URL, url, re.VERBOSE)
if m.group('type_talk'):
return self._talk_info(url)
2013-06-23 19:55:53 +00:00
else :
playlist_id=m.group('playlist_id')
name=m.group('name')
self.to_screen(u'Getting info of playlist %s: "%s"' % (playlist_id,name))
return [self._playlist_videos_info(url,name,playlist_id)]
2013-11-15 13:33:51 +00:00
def _playlist_videos_info(self, url, name, playlist_id):
2013-06-23 19:55:53 +00:00
'''Returns the videos of the playlist'''
2013-11-15 13:33:51 +00:00
webpage = self._download_webpage(
2014-01-17 02:52:17 +00:00
url, playlist_id, 'Downloading playlist webpage')
2013-11-15 13:33:51 +00:00
matches = re.finditer(
r'<p\s+class="talk-title[^"]*"><a\s+href="(?P<talk_url>/talks/[^"]+\.html)">[^<]*</a></p>',
webpage)
2013-06-23 19:55:53 +00:00
playlist_title = self._html_search_regex(r'div class="headline">\s*?<h1>\s*?<span>(.*?)</span>',
webpage, 'playlist title')
2013-11-15 13:33:51 +00:00
playlist_entries = [
self.url_result(u'http://www.ted.com' + m.group('talk_url'), 'TED')
for m in matches
]
return self.playlist_result(
playlist_entries, playlist_id=playlist_id, playlist_title=playlist_title)
2013-06-23 19:55:53 +00:00
def _talk_info(self, url, video_id=0):
"""Return the video for the talk in the url"""
m = re.match(self._VALID_URL, url,re.VERBOSE)
video_name = m.group('name')
webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
self.report_extraction(video_name)
# If the url includes the language we get the title translated
2013-07-19 14:13:31 +00:00
title = self._html_search_regex(r'<span .*?id="altHeadline".+?>(?P<title>.*)</span>',
2013-06-23 19:55:53 +00:00
webpage, 'title')
json_data = self._search_regex(r'<script.*?>var talkDetails = ({.*?})</script>',
webpage, 'json data')
info = json.loads(json_data)
desc = self._html_search_regex(r'<div class="talk-intro">.*?<p.*?>(.*?)</p>',
webpage, 'description', flags = re.DOTALL)
thumbnail = self._search_regex(r'</span>[\s.]*</div>[\s.]*<img src="(.*?)"',
webpage, 'thumbnail')
2013-10-04 08:32:34 +00:00
formats = [{
'ext': 'mp4',
'url': stream['file'],
'format': stream['id']
2013-11-15 13:06:38 +00:00
} for stream in info['htmlStreams']]
video_id = info['id']
# subtitles
video_subtitles = self.extract_subtitles(video_id, webpage)
if self._downloader.params.get('listsubtitles', False):
self._list_available_subtitles(video_id, webpage)
return
2013-11-15 13:06:38 +00:00
return {
'id': video_id,
2013-10-04 08:32:34 +00:00
'title': title,
'thumbnail': thumbnail,
'description': desc,
'subtitles': video_subtitles,
2013-10-04 08:32:34 +00:00
'formats': formats,
}
def _get_available_subtitles(self, video_id, webpage):
try:
options = self._search_regex(r'(?:<select name="subtitles_language_select" id="subtitles_language_select">)(.*?)(?:</select>)', webpage, 'subtitles_language_select', flags=re.DOTALL)
languages = re.findall(r'(?:<option value=")(\S+)"', options)
if languages:
sub_lang_list = {}
for l in languages:
url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
sub_lang_list[l] = url
return sub_lang_list
except RegexNotFoundError:
self._downloader.report_warning(u'video doesn\'t have subtitles')
return {}