From acbd0ff5df5ff9d69e6707ea4fa3e3b4f9cc6528 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Sergey=20M=E2=80=A4?= Date: Fri, 29 Jun 2018 00:35:05 +0700 Subject: [PATCH] [dctptv] Restore extraction based on REST API (closes #16850) --- youtube_dl/extractor/dctp.py | 82 ++++++++++++++++++++++++------------ 1 file changed, 55 insertions(+), 27 deletions(-) diff --git a/youtube_dl/extractor/dctp.py b/youtube_dl/extractor/dctp.py index 3a6d0560e..dc0c41b8a 100644 --- a/youtube_dl/extractor/dctp.py +++ b/youtube_dl/extractor/dctp.py @@ -5,13 +5,15 @@ from ..compat import compat_str from ..utils import ( float_or_none, - unified_strdate, + int_or_none, + unified_timestamp, ) class DctpTvIE(InfoExtractor): _VALID_URL = r'https?://(?:www\.)?dctp\.tv/(?:#/)?filme/(?P[^/?#&]+)' - _TEST = { + _TESTS = [{ + # 4x3 'url': 'http://www.dctp.tv/filme/videoinstallation-fuer-eine-kaufhausfassade/', 'info_dict': { 'id': '95eaa4f33dad413aa17b4ee613cccc6c', @@ -19,31 +21,49 @@ class DctpTvIE(InfoExtractor): 'ext': 'flv', 'title': 'Videoinstallation für eine Kaufhausfassade', 'description': 'Kurzfilm', - 'upload_date': '20110407', 'thumbnail': r're:^https?://.*\.jpg$', 'duration': 71.24, + 'timestamp': 1302172322, + 'upload_date': '20110407', }, 'params': { # rtmp download 'skip_download': True, }, - } + }, { + # 16x9 + 'url': 'http://www.dctp.tv/filme/sind-youtuber-die-besseren-lehrer/', + 'only_matching': True, + }] + + _BASE_URL = 'http://dctp-ivms2-restapi.s3.amazonaws.com' def _real_extract(self, url): display_id = self._match_id(url) - webpage = self._download_webpage(url, display_id) + version = self._download_json( + '%s/version.json' % self._BASE_URL, display_id, + 'Downloading version JSON') - video_id = self._html_search_meta( - 'DC.identifier', webpage, 'video id', - default=None) or self._search_regex( - r'id=["\']uuid[^>]+>([^<]+)<', webpage, 'video id') + restapi_base = '%s/%s/restapi' % ( + self._BASE_URL, version['version_name']) - title = self._og_search_title(webpage) + info = self._download_json( + '%s/slugs/%s.json' % (restapi_base, display_id), display_id, + 'Downloading video info JSON') + + media = self._download_json( + '%s/media/%s.json' % (restapi_base, compat_str(info['object_id'])), + display_id, 'Downloading media JSON') + + uuid = media['uuid'] + title = media['title'] + ratio = '16x9' if media.get('is_wide') else '4x3' + play_path = 'mp4:%s_dctp_0500_%s.m4v' % (uuid, ratio) servers = self._download_json( 'http://www.dctp.tv/streaming_servers/', display_id, - note='Downloading server list', fatal=False) + note='Downloading server list JSON', fatal=False) if servers: endpoint = next( @@ -60,27 +80,35 @@ def _real_extract(self, url): formats = [{ 'url': endpoint, 'app': app, - 'play_path': 'mp4:%s_dctp_0500_4x3.m4v' % video_id, + 'play_path': play_path, 'page_url': url, - 'player_url': 'http://svm-prod-dctptv-static.s3.amazonaws.com/dctptv-relaunch2012-109.swf', + 'player_url': 'http://svm-prod-dctptv-static.s3.amazonaws.com/dctptv-relaunch2012-110.swf', 'ext': 'flv', }] - description = self._html_search_meta('DC.description', webpage) - upload_date = unified_strdate( - self._html_search_meta('DC.date.created', webpage)) - thumbnail = self._og_search_thumbnail(webpage) - duration = float_or_none(self._search_regex( - r'id=["\']duration_in_ms[^+]>(\d+)', webpage, 'duration', - default=None), scale=1000) + thumbnails = [] + images = media.get('images') + if isinstance(images, list): + for image in images: + if not isinstance(image, dict): + continue + image_url = image.get('url') + if not image_url or not isinstance(image_url, compat_str): + continue + thumbnails.append({ + 'url': image_url, + 'width': int_or_none(image.get('width')), + 'height': int_or_none(image.get('height')), + }) return { - 'id': video_id, - 'title': title, - 'formats': formats, + 'id': uuid, 'display_id': display_id, - 'description': description, - 'upload_date': upload_date, - 'thumbnail': thumbnail, - 'duration': duration, + 'title': title, + 'alt_title': media.get('subtitle'), + 'description': media.get('description') or media.get('teaser'), + 'timestamp': unified_timestamp(media.get('created')), + 'duration': float_or_none(media.get('duration_in_ms'), scale=1000), + 'thumbnails': thumbnails, + 'formats': formats, }