From f22b5a6b960bca57990ede16d0e06603d1b9ab0d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Sergey=20M=E2=80=A4?= Date: Sat, 26 Dec 2020 22:37:41 +0700 Subject: [PATCH] [instagram] Improve extraction (closes #22880) --- youtube_dl/extractor/instagram.py | 128 +++++++++++++++--------------- 1 file changed, 65 insertions(+), 63 deletions(-) diff --git a/youtube_dl/extractor/instagram.py b/youtube_dl/extractor/instagram.py index 0755896de..3f5648c0c 100644 --- a/youtube_dl/extractor/instagram.py +++ b/youtube_dl/extractor/instagram.py @@ -122,9 +122,9 @@ class InstagramIE(InfoExtractor): webpage = self._download_webpage(url, video_id) - (video_url, description, thumbnail, timestamp, uploader, + (media, video_url, description, thumbnail, timestamp, uploader, uploader_id, like_count, comment_count, comments, height, - width) = [None] * 11 + width) = [None] * 12 shared_data = self._parse_json( self._search_regex( @@ -137,69 +137,71 @@ class InstagramIE(InfoExtractor): (lambda x: x['entry_data']['PostPage'][0]['graphql']['shortcode_media'], lambda x: x['entry_data']['PostPage'][0]['media']), dict) - if not media: - additional_data = self._parse_json( - self._search_regex(r'window\.__additionalDataLoaded\(\'[^\']+\',\s*({.+?})\);', - webpage, 'additional data', default='{}'), - video_id, fatal=False) - if additional_data: - media = try_get( - additional_data, - lambda x: x['graphql']['shortcode_media'], - dict) - if media: - video_url = media.get('video_url') - height = int_or_none(media.get('dimensions', {}).get('height')) - width = int_or_none(media.get('dimensions', {}).get('width')) - description = try_get( - media, lambda x: x['edge_media_to_caption']['edges'][0]['node']['text'], - compat_str) or media.get('caption') - thumbnail = media.get('display_src') - timestamp = int_or_none(media.get('taken_at_timestamp') or media.get('date')) - uploader = media.get('owner', {}).get('full_name') - uploader_id = media.get('owner', {}).get('username') + # _sharedData.entry_data.PostPage is empty when authenticated (see + # https://github.com/ytdl-org/youtube-dl/pull/22880) + if not media: + additional_data = self._parse_json( + self._search_regex( + r'window\.__additionalDataLoaded\s*\(\s*[^,]+,\s*({.+?})\s*\)\s*;', + webpage, 'additional data', default='{}'), + video_id, fatal=False) + if additional_data: + media = try_get( + additional_data, lambda x: x['graphql']['shortcode_media'], + dict) + if media: + video_url = media.get('video_url') + height = int_or_none(media.get('dimensions', {}).get('height')) + width = int_or_none(media.get('dimensions', {}).get('width')) + description = try_get( + media, lambda x: x['edge_media_to_caption']['edges'][0]['node']['text'], + compat_str) or media.get('caption') + thumbnail = media.get('display_src') + timestamp = int_or_none(media.get('taken_at_timestamp') or media.get('date')) + uploader = media.get('owner', {}).get('full_name') + uploader_id = media.get('owner', {}).get('username') - def get_count(key, kind): - return int_or_none(try_get( - media, (lambda x: x['edge_media_%s' % key]['count'], - lambda x: x['%ss' % kind]['count']))) - like_count = get_count('preview_like', 'like') - comment_count = get_count('to_comment', 'comment') + def get_count(key, kind): + return int_or_none(try_get( + media, (lambda x: x['edge_media_%s' % key]['count'], + lambda x: x['%ss' % kind]['count']))) + like_count = get_count('preview_like', 'like') + comment_count = get_count('to_comment', 'comment') - comments = [{ - 'author': comment.get('user', {}).get('username'), - 'author_id': comment.get('user', {}).get('id'), - 'id': comment.get('id'), - 'text': comment.get('text'), - 'timestamp': int_or_none(comment.get('created_at')), - } for comment in media.get( - 'comments', {}).get('nodes', []) if comment.get('text')] - if not video_url: - edges = try_get( - media, lambda x: x['edge_sidecar_to_children']['edges'], - list) or [] - if edges: - entries = [] - for edge_num, edge in enumerate(edges, start=1): - node = try_get(edge, lambda x: x['node'], dict) - if not node: - continue - node_video_url = url_or_none(node.get('video_url')) - if not node_video_url: - continue - entries.append({ - 'id': node.get('shortcode') or node['id'], - 'title': 'Video %d' % edge_num, - 'url': node_video_url, - 'thumbnail': node.get('display_url'), - 'width': int_or_none(try_get(node, lambda x: x['dimensions']['width'])), - 'height': int_or_none(try_get(node, lambda x: x['dimensions']['height'])), - 'view_count': int_or_none(node.get('video_view_count')), - }) - return self.playlist_result( - entries, video_id, - 'Post by %s' % uploader_id if uploader_id else None, - description) + comments = [{ + 'author': comment.get('user', {}).get('username'), + 'author_id': comment.get('user', {}).get('id'), + 'id': comment.get('id'), + 'text': comment.get('text'), + 'timestamp': int_or_none(comment.get('created_at')), + } for comment in media.get( + 'comments', {}).get('nodes', []) if comment.get('text')] + if not video_url: + edges = try_get( + media, lambda x: x['edge_sidecar_to_children']['edges'], + list) or [] + if edges: + entries = [] + for edge_num, edge in enumerate(edges, start=1): + node = try_get(edge, lambda x: x['node'], dict) + if not node: + continue + node_video_url = url_or_none(node.get('video_url')) + if not node_video_url: + continue + entries.append({ + 'id': node.get('shortcode') or node['id'], + 'title': 'Video %d' % edge_num, + 'url': node_video_url, + 'thumbnail': node.get('display_url'), + 'width': int_or_none(try_get(node, lambda x: x['dimensions']['width'])), + 'height': int_or_none(try_get(node, lambda x: x['dimensions']['height'])), + 'view_count': int_or_none(node.get('video_view_count')), + }) + return self.playlist_result( + entries, video_id, + 'Post by %s' % uploader_id if uploader_id else None, + description) if not video_url: video_url = self._og_search_video_url(webpage, secure=False)