[udemy] Fix subtitles extraction (closes #13812)

This commit is contained in:
Sergey M․ 2017-08-04 23:44:07 +07:00
parent 11a6793f80
commit 57a38a38c3
No known key found for this signature in database
GPG key ID: 2C393E0F18A9236D

View file

@ -15,6 +15,7 @@ from ..utils import (
ExtractorError, ExtractorError,
float_or_none, float_or_none,
int_or_none, int_or_none,
js_to_json,
sanitized_Request, sanitized_Request,
unescapeHTML, unescapeHTML,
urlencode_postdata, urlencode_postdata,
@ -268,6 +269,25 @@ class UdemyIE(InfoExtractor):
f = add_output_format_meta(f, format_id) f = add_output_format_meta(f, format_id)
formats.append(f) formats.append(f)
def extract_subtitles(track_list):
if not isinstance(track_list, list):
return
for track in track_list:
if not isinstance(track, dict):
continue
if track.get('kind') != 'captions':
continue
src = track.get('src')
if not src or not isinstance(src, compat_str):
continue
lang = track.get('language') or track.get(
'srclang') or track.get('label')
sub_dict = automatic_captions if track.get(
'autogenerated') is True else subtitles
sub_dict.setdefault(lang, []).append({
'url': src,
})
download_urls = asset.get('download_urls') download_urls = asset.get('download_urls')
if isinstance(download_urls, dict): if isinstance(download_urls, dict):
extract_formats(download_urls.get('Video')) extract_formats(download_urls.get('Video'))
@ -315,23 +335,16 @@ class UdemyIE(InfoExtractor):
extract_formats(data.get('sources')) extract_formats(data.get('sources'))
if not duration: if not duration:
duration = int_or_none(data.get('duration')) duration = int_or_none(data.get('duration'))
tracks = data.get('tracks') extract_subtitles(data.get('tracks'))
if isinstance(tracks, list):
for track in tracks: if not subtitles and not automatic_captions:
if not isinstance(track, dict): text_tracks = self._parse_json(
continue self._search_regex(
if track.get('kind') != 'captions': r'text-tracks=(["\'])(?P<data>\[.+?\])\1', view_html,
continue 'text tracks', default='{}', group='data'), video_id,
src = track.get('src') transform_source=lambda s: js_to_json(unescapeHTML(s)),
if not src or not isinstance(src, compat_str): fatal=False)
continue extract_subtitles(text_tracks)
lang = track.get('language') or track.get(
'srclang') or track.get('label')
sub_dict = automatic_captions if track.get(
'autogenerated') is True else subtitles
sub_dict.setdefault(lang, []).append({
'url': src,
})
self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id')) self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))