_ Git - youtube-dl/blob - youtube_dl/extractor/instagram.py

   1 from __future__ import unicode_literals
   2
   3 import itertools
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..compat import compat_str
   8 from ..utils import (
   9     get_element_by_attribute,
  10     int_or_none,
  11     lowercase_escape,
  12     try_get,
  13 )
  14
  15
  16 class InstagramIE(InfoExtractor):
  17     _VALID_URL = r'(?P<url>https?://(?:www\.)?instagram\.com/p/(?P<id>[^/?#&]+))'
  18     _TESTS = [{
  19         'url': 'https://instagram.com/p/aye83DjauH/?foo=bar#abc',
  20         'md5': '0d2da106a9d2631273e192b372806516',
  21         'info_dict': {
  22             'id': 'aye83DjauH',
  23             'ext': 'mp4',
  24             'title': 'Video by naomipq',
  25             'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
  26             'thumbnail': r're:^https?://.*\.jpg',
  27             'timestamp': 1371748545,
  28             'upload_date': '20130620',
  29             'uploader_id': 'naomipq',
  30             'uploader': 'Naomi Leonor Phan-Quang',
  31             'like_count': int,
  32             'comment_count': int,
  33             'comments': list,
  34         },
  35     }, {
  36         # missing description
  37         'url': 'https://www.instagram.com/p/BA-pQFBG8HZ/?taken-by=britneyspears',
  38         'info_dict': {
  39             'id': 'BA-pQFBG8HZ',
  40             'ext': 'mp4',
  41             'title': 'Video by britneyspears',
  42             'thumbnail': r're:^https?://.*\.jpg',
  43             'timestamp': 1453760977,
  44             'upload_date': '20160125',
  45             'uploader_id': 'britneyspears',
  46             'uploader': 'Britney Spears',
  47             'like_count': int,
  48             'comment_count': int,
  49             'comments': list,
  50         },
  51         'params': {
  52             'skip_download': True,
  53         },
  54     }, {
  55         # multi video post
  56         'url': 'https://www.instagram.com/p/BQ0eAlwhDrw/',
  57         'playlist': [{
  58             'info_dict': {
  59                 'id': 'BQ0dSaohpPW',
  60                 'ext': 'mp4',
  61                 'title': 'Video 1',
  62             },
  63         }, {
  64             'info_dict': {
  65                 'id': 'BQ0dTpOhuHT',
  66                 'ext': 'mp4',
  67                 'title': 'Video 2',
  68             },
  69         }, {
  70             'info_dict': {
  71                 'id': 'BQ0dT7RBFeF',
  72                 'ext': 'mp4',
  73                 'title': 'Video 3',
  74             },
  75         }],
  76         'info_dict': {
  77             'id': 'BQ0eAlwhDrw',
  78             'title': 'Post by instagram',
  79             'description': 'md5:0f9203fc6a2ce4d228da5754bcf54957',
  80         },
  81     }, {
  82         'url': 'https://instagram.com/p/-Cmh1cukG2/',
  83         'only_matching': True,
  84     }, {
  85         'url': 'http://instagram.com/p/9o6LshA7zy/embed/',
  86         'only_matching': True,
  87     }]
  88
  89     @staticmethod
  90     def _extract_embed_url(webpage):
  91         mobj = re.search(
  92             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?instagram\.com/p/[^/]+/embed.*?)\1',
  93             webpage)
  94         if mobj:
  95             return mobj.group('url')
  96
  97         blockquote_el = get_element_by_attribute(
  98             'class', 'instagram-media', webpage)
  99         if blockquote_el is None:
 100             return
 101
 102         mobj = re.search(
 103             r'<a[^>]+href=([\'"])(?P<link>[^\'"]+)\1', blockquote_el)
 104         if mobj:
 105             return mobj.group('link')
 106
 107     def _real_extract(self, url):
 108         mobj = re.match(self._VALID_URL, url)
 109         video_id = mobj.group('id')
 110         url = mobj.group('url')
 111
 112         webpage = self._download_webpage(url, video_id)
 113
 114         (video_url, description, thumbnail, timestamp, uploader,
 115          uploader_id, like_count, comment_count, comments, height,
 116          width) = [None] * 11
 117
 118         shared_data = self._parse_json(
 119             self._search_regex(
 120                 r'window\._sharedData\s*=\s*({.+?});',
 121                 webpage, 'shared data', default='{}'),
 122             video_id, fatal=False)
 123         if shared_data:
 124             media = try_get(
 125                 shared_data,
 126                 (lambda x: x['entry_data']['PostPage'][0]['graphql']['shortcode_media'],
 127                  lambda x: x['entry_data']['PostPage'][0]['media']),
 128                 dict)
 129             if media:
 130                 video_url = media.get('video_url')
 131                 height = int_or_none(media.get('dimensions', {}).get('height'))
 132                 width = int_or_none(media.get('dimensions', {}).get('width'))
 133                 description = media.get('caption')
 134                 thumbnail = media.get('display_src')
 135                 timestamp = int_or_none(media.get('date'))
 136                 uploader = media.get('owner', {}).get('full_name')
 137                 uploader_id = media.get('owner', {}).get('username')
 138                 like_count = int_or_none(media.get('likes', {}).get('count'))
 139                 comment_count = int_or_none(media.get('comments', {}).get('count'))
 140                 comments = [{
 141                     'author': comment.get('user', {}).get('username'),
 142                     'author_id': comment.get('user', {}).get('id'),
 143                     'id': comment.get('id'),
 144                     'text': comment.get('text'),
 145                     'timestamp': int_or_none(comment.get('created_at')),
 146                 } for comment in media.get(
 147                     'comments', {}).get('nodes', []) if comment.get('text')]
 148                 if not video_url:
 149                     edges = try_get(
 150                         media, lambda x: x['edge_sidecar_to_children']['edges'],
 151                         list) or []
 152                     if edges:
 153                         entries = []
 154                         for edge_num, edge in enumerate(edges, start=1):
 155                             node = try_get(edge, lambda x: x['node'], dict)
 156                             if not node:
 157                                 continue
 158                             node_video_url = try_get(node, lambda x: x['video_url'], compat_str)
 159                             if not node_video_url:
 160                                 continue
 161                             entries.append({
 162                                 'id': node.get('shortcode') or node['id'],
 163                                 'title': 'Video %d' % edge_num,
 164                                 'url': node_video_url,
 165                                 'thumbnail': node.get('display_url'),
 166                                 'width': int_or_none(try_get(node, lambda x: x['dimensions']['width'])),
 167                                 'height': int_or_none(try_get(node, lambda x: x['dimensions']['height'])),
 168                                 'view_count': int_or_none(node.get('video_view_count')),
 169                             })
 170                         return self.playlist_result(
 171                             entries, video_id,
 172                             'Post by %s' % uploader_id if uploader_id else None,
 173                             description)
 174
 175         if not video_url:
 176             video_url = self._og_search_video_url(webpage, secure=False)
 177
 178         formats = [{
 179             'url': video_url,
 180             'width': width,
 181             'height': height,
 182         }]
 183
 184         if not uploader_id:
 185             uploader_id = self._search_regex(
 186                 r'"owner"\s*:\s*{\s*"username"\s*:\s*"(.+?)"',
 187                 webpage, 'uploader id', fatal=False)
 188
 189         if not description:
 190             description = self._search_regex(
 191                 r'"caption"\s*:\s*"(.+?)"', webpage, 'description', default=None)
 192             if description is not None:
 193                 description = lowercase_escape(description)
 194
 195         if not thumbnail:
 196             thumbnail = self._og_search_thumbnail(webpage)
 197
 198         return {
 199             'id': video_id,
 200             'formats': formats,
 201             'ext': 'mp4',
 202             'title': 'Video by %s' % uploader_id,
 203             'description': description,
 204             'thumbnail': thumbnail,
 205             'timestamp': timestamp,
 206             'uploader_id': uploader_id,
 207             'uploader': uploader,
 208             'like_count': like_count,
 209             'comment_count': comment_count,
 210             'comments': comments,
 211         }
 212
 213
 214 class InstagramUserIE(InfoExtractor):
 215     _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<id>[^/]{2,})/?(?:$|[?#])'
 216     IE_DESC = 'Instagram user profile'
 217     IE_NAME = 'instagram:user'
 218     _TEST = {
 219         'url': 'https://instagram.com/porsche',
 220         'info_dict': {
 221             'id': 'porsche',
 222             'title': 'porsche',
 223         },
 224         'playlist_count': 5,
 225         'params': {
 226             'extract_flat': True,
 227             'skip_download': True,
 228             'playlistend': 5,
 229         }
 230     }
 231
 232     def _entries(self, uploader_id):
 233         query = {
 234             '__a': 1,
 235         }
 236
 237         def get_count(kind):
 238             return int_or_none(try_get(
 239                 node, lambda x: x['%ss' % kind]['count']))
 240
 241         for page_num in itertools.count(1):
 242             page = self._download_json(
 243                 'https://instagram.com/%s/' % uploader_id, uploader_id,
 244                 note='Downloading page %d' % page_num,
 245                 fatal=False, query=query)
 246             if not page:
 247                 break
 248
 249             nodes = try_get(page, lambda x: x['user']['media']['nodes'], list)
 250             if not nodes:
 251                 break
 252
 253             max_id = None
 254
 255             for node in nodes:
 256                 node_id = node.get('id')
 257                 if node_id:
 258                     max_id = node_id
 259
 260                 if node.get('__typename') != 'GraphVideo' and node.get('is_video') is not True:
 261                     continue
 262                 video_id = node.get('code')
 263                 if not video_id:
 264                     continue
 265
 266                 info = self.url_result(
 267                     'https://instagram.com/p/%s/' % video_id,
 268                     ie=InstagramIE.ie_key(), video_id=video_id)
 269
 270                 description = try_get(
 271                     node, [lambda x: x['caption'], lambda x: x['text']['id']],
 272                     compat_str)
 273                 thumbnail = node.get('thumbnail_src') or node.get('display_src')
 274                 timestamp = int_or_none(node.get('date'))
 275
 276                 comment_count = get_count('comment')
 277                 like_count = get_count('like')
 278                 view_count = int_or_none(node.get('video_views'))
 279
 280                 info.update({
 281                     'description': description,
 282                     'thumbnail': thumbnail,
 283                     'timestamp': timestamp,
 284                     'comment_count': comment_count,
 285                     'like_count': like_count,
 286                     'view_count': view_count,
 287                 })
 288
 289                 yield info
 290
 291             if not max_id:
 292                 break
 293
 294             query['max_id'] = max_id
 295
 296     def _real_extract(self, url):
 297         uploader_id = self._match_id(url)
 298         return self.playlist_result(
 299             self._entries(uploader_id), uploader_id, uploader_id)