yt_dlp/extractor/pinterest.py

   1 import json
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     determine_ext,
   6     float_or_none,
   7     int_or_none,
   8     str_or_none,
   9     strip_or_none,
  10     traverse_obj,
  11     unified_timestamp,
  12     url_or_none,
  13 )
  14
  15
  16 class PinterestBaseIE(InfoExtractor):
  17     _VALID_URL_BASE = r'''(?x)
  18         https?://(?:[^/]+\.)?pinterest\.(?:
  19             com|fr|de|ch|jp|cl|ca|it|co\.uk|nz|ru|com\.au|at|pt|co\.kr|es|com\.mx|
  20             dk|ph|th|com\.uy|co|nl|info|kr|ie|vn|com\.vn|ec|mx|in|pe|co\.at|hu|
  21             co\.in|co\.nz|id|com\.ec|com\.py|tw|be|uk|com\.bo|com\.pe)'''
  22
  23     def _call_api(self, resource, video_id, options):
  24         return self._download_json(
  25             'https://www.pinterest.com/resource/%sResource/get/' % resource,
  26             video_id, 'Download %s JSON metadata' % resource, query={
  27                 'data': json.dumps({'options': options})
  28             })['resource_response']
  29
  30     def _extract_video(self, data, extract_formats=True):
  31         video_id = data['id']
  32         thumbnails = []
  33         images = data.get('images')
  34         if isinstance(images, dict):
  35             for thumbnail_id, thumbnail in images.items():
  36                 if not isinstance(thumbnail, dict):
  37                     continue
  38                 thumbnail_url = url_or_none(thumbnail.get('url'))
  39                 if not thumbnail_url:
  40                     continue
  41                 thumbnails.append({
  42                     'url': thumbnail_url,
  43                     'width': int_or_none(thumbnail.get('width')),
  44                     'height': int_or_none(thumbnail.get('height')),
  45                 })
  46
  47         info = {
  48             'title': strip_or_none(traverse_obj(data, 'title', 'grid_title', default='')),
  49             'description': traverse_obj(data, 'seo_description', 'description'),
  50             'timestamp': unified_timestamp(data.get('created_at')),
  51             'thumbnails': thumbnails,
  52             'uploader': traverse_obj(data, ('closeup_attribution', 'full_name')),
  53             'uploader_id': str_or_none(traverse_obj(data, ('closeup_attribution', 'id'))),
  54             'repost_count': int_or_none(data.get('repin_count')),
  55             'comment_count': int_or_none(data.get('comment_count')),
  56             'categories': traverse_obj(data, ('pin_join', 'visual_annotation'), expected_type=list),
  57             'tags': traverse_obj(data, 'hashtags', expected_type=list),
  58         }
  59
  60         urls = []
  61         formats = []
  62         duration = None
  63         domain = data.get('domain', '')
  64         if domain.lower() != 'uploaded by user' and traverse_obj(data, ('embed', 'src')):
  65             if not info['title']:
  66                 info['title'] = None
  67             return {
  68                 '_type': 'url_transparent',
  69                 'url': data['embed']['src'],
  70                 **info,
  71             }
  72
  73         elif extract_formats:
  74             video_list = traverse_obj(
  75                 data, ('videos', 'video_list'),
  76                 ('story_pin_data', 'pages', ..., 'blocks', ..., 'video', 'video_list'),
  77                 expected_type=dict, get_all=False, default={})
  78             for format_id, format_dict in video_list.items():
  79                 if not isinstance(format_dict, dict):
  80                     continue
  81                 format_url = url_or_none(format_dict.get('url'))
  82                 if not format_url or format_url in urls:
  83                     continue
  84                 urls.append(format_url)
  85                 duration = float_or_none(format_dict.get('duration'), scale=1000)
  86                 ext = determine_ext(format_url)
  87                 if 'hls' in format_id.lower() or ext == 'm3u8':
  88                     formats.extend(self._extract_m3u8_formats(
  89                         format_url, video_id, 'mp4', entry_protocol='m3u8_native',
  90                         m3u8_id=format_id, fatal=False))
  91                 else:
  92                     formats.append({
  93                         'url': format_url,
  94                         'format_id': format_id,
  95                         'width': int_or_none(format_dict.get('width')),
  96                         'height': int_or_none(format_dict.get('height')),
  97                         'duration': duration,
  98                     })
  99
 100         return {
 101             'id': video_id,
 102             'formats': formats,
 103             'duration': duration,
 104             'webpage_url': f'https://www.pinterest.com/pin/{video_id}/',
 105             'extractor_key': PinterestIE.ie_key(),
 106             'extractor': PinterestIE.IE_NAME,
 107             **info,
 108         }
 109
 110
 111 class PinterestIE(PinterestBaseIE):
 112     _VALID_URL = r'%s/pin/(?P<id>\d+)' % PinterestBaseIE._VALID_URL_BASE
 113     _TESTS = [{
 114         # formats found in data['videos']
 115         'url': 'https://www.pinterest.com/pin/664281013778109217/',
 116         'md5': '6550c2af85d6d9f3fe3b88954d1577fc',
 117         'info_dict': {
 118             'id': '664281013778109217',
 119             'ext': 'mp4',
 120             'title': 'Origami',
 121             'description': 'md5:e29801cab7d741ea8c741bc50c8d00ab',
 122             'duration': 57.7,
 123             'timestamp': 1593073622,
 124             'upload_date': '20200625',
 125             'repost_count': int,
 126             'comment_count': int,
 127             'categories': list,
 128             'tags': list,
 129             'thumbnail': r're:^https?://.*\.(?:jpg|png)$',
 130         },
 131     }, {
 132         # formats found in data['story_pin_data']
 133         'url': 'https://www.pinterest.com/pin/1084663891475263837/',
 134         'md5': '069ac19919ab9e1e13fa60de46290b03',
 135         'info_dict': {
 136             'id': '1084663891475263837',
 137             'ext': 'mp4',
 138             'title': 'Gadget, Cool products, Amazon product, technology, Kitchen gadgets',
 139             'description': 'md5:d0a4b6ae996ff0c6eed83bc869598d13',
 140             'uploader': 'CoolCrazyGadgets',
 141             'uploader_id': '1084664028912989237',
 142             'upload_date': '20211003',
 143             'timestamp': 1633246654.0,
 144             'duration': 14.9,
 145             'comment_count': int,
 146             'repost_count': int,
 147             'categories': 'count:9',
 148             'tags': list,
 149             'thumbnail': r're:^https?://.*\.(?:jpg|png)$',
 150         },
 151     }, {
 152         # vimeo.com embed
 153         'url': 'https://www.pinterest.ca/pin/441282463481903715/',
 154         'info_dict': {
 155             'id': '111691128',
 156             'ext': 'mp4',
 157             'title': 'Tonite Let\'s All Make Love In London (1967)',
 158             'description': 'md5:8190f37b3926807809ec57ec21aa77b2',
 159             'uploader': 'Vimeo',
 160             'uploader_id': '473792960706651251',
 161             'upload_date': '20180120',
 162             'timestamp': 1516409040,
 163             'duration': 3404,
 164             'comment_count': int,
 165             'repost_count': int,
 166             'categories': 'count:9',
 167             'tags': [],
 168             'thumbnail': r're:^https?://.*\.(?:jpg|png)$',
 169             'uploader_url': 'https://vimeo.com/willardandrade',
 170         },
 171         'params': {
 172             'skip_download': 'm3u8',
 173         },
 174     }, {
 175         'url': 'https://co.pinterest.com/pin/824721750502199491/',
 176         'only_matching': True,
 177     }]
 178
 179     def _real_extract(self, url):
 180         video_id = self._match_id(url)
 181         data = self._call_api(
 182             'Pin', video_id, {
 183                 'field_set_key': 'unauth_react_main_pin',
 184                 'id': video_id,
 185             })['data']
 186         return self._extract_video(data)
 187
 188
 189 class PinterestCollectionIE(PinterestBaseIE):
 190     _VALID_URL = r'%s/(?P<username>[^/]+)/(?P<id>[^/?#&]+)' % PinterestBaseIE._VALID_URL_BASE
 191     _TESTS = [{
 192         'url': 'https://www.pinterest.ca/mashal0407/cool-diys/',
 193         'info_dict': {
 194             'id': '585890301462791043',
 195             'title': 'cool diys',
 196         },
 197         'playlist_count': 8,
 198     }, {
 199         'url': 'https://www.pinterest.ca/fudohub/videos/',
 200         'info_dict': {
 201             'id': '682858430939307450',
 202             'title': 'VIDEOS',
 203         },
 204         'playlist_mincount': 365,
 205         'skip': 'Test with extract_formats=False',
 206     }]
 207
 208     @classmethod
 209     def suitable(cls, url):
 210         return False if PinterestIE.suitable(url) else super(
 211             PinterestCollectionIE, cls).suitable(url)
 212
 213     def _real_extract(self, url):
 214         username, slug = self._match_valid_url(url).groups()
 215         board = self._call_api(
 216             'Board', slug, {
 217                 'slug': slug,
 218                 'username': username
 219             })['data']
 220         board_id = board['id']
 221         options = {
 222             'board_id': board_id,
 223             'page_size': 250,
 224         }
 225         bookmark = None
 226         entries = []
 227         while True:
 228             if bookmark:
 229                 options['bookmarks'] = [bookmark]
 230             board_feed = self._call_api('BoardFeed', board_id, options)
 231             for item in (board_feed.get('data') or []):
 232                 if not isinstance(item, dict) or item.get('type') != 'pin':
 233                     continue
 234                 video_id = item.get('id')
 235                 if video_id:
 236                     # Some pins may not be available anonymously via pin URL
 237                     # video = self._extract_video(item, extract_formats=False)
 238                     # video.update({
 239                     #     '_type': 'url_transparent',
 240                     #     'url': 'https://www.pinterest.com/pin/%s/' % video_id,
 241                     # })
 242                     # entries.append(video)
 243                     entries.append(self._extract_video(item))
 244             bookmark = board_feed.get('bookmark')
 245             if not bookmark:
 246                 break
 247         return self.playlist_result(
 248             entries, playlist_id=board_id, playlist_title=board.get('name'))