youtube_dl/extractor/pinterest.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import re
   6
   7 from .common import InfoExtractor
   8 from ..compat import compat_str
   9 from ..utils import (
  10     determine_ext,
  11     float_or_none,
  12     int_or_none,
  13     try_get,
  14     unified_timestamp,
  15     url_or_none,
  16 )
  17
  18
  19 class PinterestBaseIE(InfoExtractor):
  20     _VALID_URL_BASE = r'https?://(?:[^/]+\.)?pinterest\.(?:com|fr|de|ch|jp|cl|ca|it|co\.uk|nz|ru|com\.au|at|pt|co\.kr|es|com\.mx|dk|ph|th|com\.uy|co|nl|info|kr|ie|vn|com\.vn|ec|mx|in|pe|co\.at|hu|co\.in|co\.nz|id|com\.ec|com\.py|tw|be|uk|com\.bo|com\.pe)'
  21
  22     def _call_api(self, resource, video_id, options):
  23         return self._download_json(
  24             'https://www.pinterest.com/resource/%sResource/get/' % resource,
  25             video_id, 'Download %s JSON metadata' % resource, query={
  26                 'data': json.dumps({'options': options})
  27             })['resource_response']
  28
  29     def _extract_video(self, data, extract_formats=True):
  30         video_id = data['id']
  31
  32         title = (data.get('title') or data.get('grid_title') or video_id).strip()
  33
  34         urls = []
  35         formats = []
  36         duration = None
  37         if extract_formats:
  38             for format_id, format_dict in data['videos']['video_list'].items():
  39                 if not isinstance(format_dict, dict):
  40                     continue
  41                 format_url = url_or_none(format_dict.get('url'))
  42                 if not format_url or format_url in urls:
  43                     continue
  44                 urls.append(format_url)
  45                 duration = float_or_none(format_dict.get('duration'), scale=1000)
  46                 ext = determine_ext(format_url)
  47                 if 'hls' in format_id.lower() or ext == 'm3u8':
  48                     formats.extend(self._extract_m3u8_formats(
  49                         format_url, video_id, 'mp4', entry_protocol='m3u8_native',
  50                         m3u8_id=format_id, fatal=False))
  51                 else:
  52                     formats.append({
  53                         'url': format_url,
  54                         'format_id': format_id,
  55                         'width': int_or_none(format_dict.get('width')),
  56                         'height': int_or_none(format_dict.get('height')),
  57                         'duration': duration,
  58                     })
  59             self._sort_formats(
  60                 formats, field_preference=('height', 'width', 'tbr', 'format_id'))
  61
  62         description = data.get('description') or data.get('description_html') or data.get('seo_description')
  63         timestamp = unified_timestamp(data.get('created_at'))
  64
  65         def _u(field):
  66             return try_get(data, lambda x: x['closeup_attribution'][field], compat_str)
  67
  68         uploader = _u('full_name')
  69         uploader_id = _u('id')
  70
  71         repost_count = int_or_none(data.get('repin_count'))
  72         comment_count = int_or_none(data.get('comment_count'))
  73         categories = try_get(data, lambda x: x['pin_join']['visual_annotation'], list)
  74         tags = data.get('hashtags')
  75
  76         thumbnails = []
  77         images = data.get('images')
  78         if isinstance(images, dict):
  79             for thumbnail_id, thumbnail in images.items():
  80                 if not isinstance(thumbnail, dict):
  81                     continue
  82                 thumbnail_url = url_or_none(thumbnail.get('url'))
  83                 if not thumbnail_url:
  84                     continue
  85                 thumbnails.append({
  86                     'url': thumbnail_url,
  87                     'width': int_or_none(thumbnail.get('width')),
  88                     'height': int_or_none(thumbnail.get('height')),
  89                 })
  90
  91         return {
  92             'id': video_id,
  93             'title': title,
  94             'description': description,
  95             'duration': duration,
  96             'timestamp': timestamp,
  97             'thumbnails': thumbnails,
  98             'uploader': uploader,
  99             'uploader_id': uploader_id,
 100             'repost_count': repost_count,
 101             'comment_count': comment_count,
 102             'categories': categories,
 103             'tags': tags,
 104             'formats': formats,
 105             'extractor_key': PinterestIE.ie_key(),
 106         }
 107
 108
 109 class PinterestIE(PinterestBaseIE):
 110     _VALID_URL = r'%s/pin/(?P<id>\d+)' % PinterestBaseIE._VALID_URL_BASE
 111     _TESTS = [{
 112         'url': 'https://www.pinterest.com/pin/664281013778109217/',
 113         'md5': '6550c2af85d6d9f3fe3b88954d1577fc',
 114         'info_dict': {
 115             'id': '664281013778109217',
 116             'ext': 'mp4',
 117             'title': 'Origami',
 118             'description': 'md5:b9d90ddf7848e897882de9e73344f7dd',
 119             'duration': 57.7,
 120             'timestamp': 1593073622,
 121             'upload_date': '20200625',
 122             'uploader': 'Love origami -I am Dafei',
 123             'uploader_id': '586523688879454212',
 124             'repost_count': 50,
 125             'comment_count': 0,
 126             'categories': list,
 127             'tags': list,
 128         },
 129     }, {
 130         'url': 'https://co.pinterest.com/pin/824721750502199491/',
 131         'only_matching': True,
 132     }]
 133
 134     def _real_extract(self, url):
 135         video_id = self._match_id(url)
 136         data = self._call_api(
 137             'Pin', video_id, {
 138                 'field_set_key': 'unauth_react_main_pin',
 139                 'id': video_id,
 140             })['data']
 141         return self._extract_video(data)
 142
 143
 144 class PinterestCollectionIE(PinterestBaseIE):
 145     _VALID_URL = r'%s/(?P<username>[^/]+)/(?P<id>[^/?#&]+)' % PinterestBaseIE._VALID_URL_BASE
 146     _TESTS = [{
 147         'url': 'https://www.pinterest.ca/mashal0407/cool-diys/',
 148         'info_dict': {
 149             'id': '585890301462791043',
 150             'title': 'cool diys',
 151         },
 152         'playlist_count': 8,
 153     }, {
 154         'url': 'https://www.pinterest.ca/fudohub/videos/',
 155         'info_dict': {
 156             'id': '682858430939307450',
 157             'title': 'VIDEOS',
 158         },
 159         'playlist_mincount': 365,
 160         'skip': 'Test with extract_formats=False',
 161     }]
 162
 163     @classmethod
 164     def suitable(cls, url):
 165         return False if PinterestIE.suitable(url) else super(
 166             PinterestCollectionIE, cls).suitable(url)
 167
 168     def _real_extract(self, url):
 169         username, slug = re.match(self._VALID_URL, url).groups()
 170         board = self._call_api(
 171             'Board', slug, {
 172                 'slug': slug,
 173                 'username': username
 174             })['data']
 175         board_id = board['id']
 176         options = {
 177             'board_id': board_id,
 178             'page_size': 250,
 179         }
 180         bookmark = None
 181         entries = []
 182         while True:
 183             if bookmark:
 184                 options['bookmarks'] = [bookmark]
 185             board_feed = self._call_api('BoardFeed', board_id, options)
 186             for item in (board_feed.get('data') or []):
 187                 if not isinstance(item, dict) or item.get('type') != 'pin':
 188                     continue
 189                 video_id = item.get('id')
 190                 if video_id:
 191                     # Some pins may not be available anonymously via pin URL
 192                     # video = self._extract_video(item, extract_formats=False)
 193                     # video.update({
 194                     #     '_type': 'url_transparent',
 195                     #     'url': 'https://www.pinterest.com/pin/%s/' % video_id,
 196                     # })
 197                     # entries.append(video)
 198                     entries.append(self._extract_video(item))
 199             bookmark = board_feed.get('bookmark')
 200             if not bookmark:
 201                 break
 202         return self.playlist_result(
 203             entries, playlist_id=board_id, playlist_title=board.get('name'))