From fd95105ed4293098d15dec4bd48d6c761d32ab83 Mon Sep 17 00:00:00 2001 From: Remita Amine Date: Sat, 13 Jul 2019 12:47:02 +0100 Subject: [PATCH] [livejournal] Add new extractor(closes #21526) --- youtube_dl/extractor/extractors.py | 1 + youtube_dl/extractor/livejournal.py | 42 +++++++++++++++++++++++++++++ 2 files changed, 43 insertions(+) create mode 100644 youtube_dl/extractor/livejournal.py diff --git a/youtube_dl/extractor/extractors.py b/youtube_dl/extractor/extractors.py index e88ad34a85..75a53f54bd 100644 --- a/youtube_dl/extractor/extractors.py +++ b/youtube_dl/extractor/extractors.py @@ -579,6 +579,7 @@ ) from .linuxacademy import LinuxAcademyIE from .litv import LiTVIE +from .livejournal import LiveJournalIE from .liveleak import ( LiveLeakIE, LiveLeakEmbedIE, diff --git a/youtube_dl/extractor/livejournal.py b/youtube_dl/extractor/livejournal.py new file mode 100644 index 0000000000..3a9f4553f4 --- /dev/null +++ b/youtube_dl/extractor/livejournal.py @@ -0,0 +1,42 @@ +# coding: utf-8 +from __future__ import unicode_literals + +from .common import InfoExtractor +from ..compat import compat_str +from ..utils import int_or_none + + +class LiveJournalIE(InfoExtractor): + _VALID_URL = r'https?://(?:[^.]+\.)?livejournal\.com/video/album/\d+.+?\bid=(?P\d+)' + _TEST = { + 'url': 'https://andrei-bt.livejournal.com/video/album/407/?mode=view&id=51272', + 'md5': 'adaf018388572ced8a6f301ace49d4b2', + 'info_dict': { + 'id': '1263729', + 'ext': 'mp4', + 'title': 'Истребители против БПЛА', + 'upload_date': '20190624', + 'timestamp': 1561406715, + } + } + + def _real_extract(self, url): + video_id = self._match_id(url) + webpage = self._download_webpage(url, video_id) + record = self._parse_json(self._search_regex( + r'Site\.page\s*=\s*({.+?});', webpage, + 'page data'), video_id)['video']['record'] + storage_id = compat_str(record['storageid']) + title = record.get('name') + if title: + # remove filename extension(.mp4, .mov, etc...) + title = title.rsplit('.', 1)[0] + return { + '_type': 'url_transparent', + 'id': video_id, + 'title': title, + 'thumbnail': record.get('thumbnail'), + 'timestamp': int_or_none(record.get('timecreate')), + 'url': 'eagleplatform:vc.videos.livejournal.com:' + storage_id, + 'ie_key': 'EaglePlatform', + }