From f889ac45b83cd771919fba29c14cee9d24e637b7 Mon Sep 17 00:00:00 2001 From: remitamine Date: Tue, 29 Dec 2015 14:55:36 +0100 Subject: [PATCH] [ign] split long lines --- youtube_dl/extractor/ign.py | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/youtube_dl/extractor/ign.py b/youtube_dl/extractor/ign.py index d1c1c210c..c45c68c1d 100644 --- a/youtube_dl/extractor/ign.py +++ b/youtube_dl/extractor/ign.py @@ -120,19 +120,24 @@ def _real_extract(self, url): video_id = self._find_video_id(webpage) if not video_id: - return self.url_result(self._search_regex(self._EMBED_RE, webpage, 'embed url')) + return self.url_result(self._search_regex( + self._EMBED_RE, webpage, 'embed url')) return self._get_video_info(video_id) def _get_video_info(self, video_id): - api_data = self._download_json(self._API_URL_TEMPLATE % video_id, video_id) + api_data = self._download_json( + self._API_URL_TEMPLATE % video_id, video_id) formats = [] m3u8_url = api_data['refs'].get('m3uUrl') if m3u8_url: - formats.extend(self._extract_m3u8_formats(m3u8_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)) + formats.extend(self._extract_m3u8_formats( + m3u8_url, video_id, 'mp4', 'm3u8_native', + m3u8_id='hls', fatal=False)) f4m_url = api_data['refs'].get('f4mUrl') if f4m_url: - formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False)) + formats.extend(self._extract_f4m_formats( + f4m_url, video_id, f4m_id='hds', fatal=False)) for asset in api_data['assets']: formats.append({ 'url': asset['url'],