- def _real_extract(self, url):
- mobj = re.match(self._VALID_URL, url)
- uploader_id = mobj.group('username')
-
- entries = []
- page_count = 0
- media_url = 'http://instagram.com/%s/media' % uploader_id
- while True:
- page = self._download_json(
- media_url, uploader_id,
- note='Downloading page %d ' % (page_count + 1),
- )
- page_count += 1
-
- for it in page['items']:
- if it.get('type') != 'video':
+ _gis_tmpl = None
+
+ def _entries(self, data):
+ def get_count(suffix):
+ return int_or_none(try_get(
+ node, lambda x: x['edge_media_' + suffix]['count']))
+
+ uploader_id = data['entry_data']['ProfilePage'][0]['graphql']['user']['id']
+ csrf_token = data['config']['csrf_token']
+ rhx_gis = data.get('rhx_gis') or '3c7ca9dcefcf966d11dacf1f151335e8'
+
+ self._set_cookie('instagram.com', 'ig_pr', '1')
+
+ cursor = ''
+ for page_num in itertools.count(1):
+ variables = json.dumps({
+ 'id': uploader_id,
+ 'first': 12,
+ 'after': cursor,
+ })
+
+ if self._gis_tmpl:
+ gis_tmpls = [self._gis_tmpl]
+ else:
+ gis_tmpls = [
+ '%s' % rhx_gis,
+ '',
+ '%s:%s' % (rhx_gis, csrf_token),
+ '%s:%s:%s' % (rhx_gis, csrf_token, std_headers['User-Agent']),
+ ]
+
+ for gis_tmpl in gis_tmpls:
+ try:
+ media = self._download_json(
+ 'https://www.instagram.com/graphql/query/', uploader_id,
+ 'Downloading JSON page %d' % page_num, headers={
+ 'X-Requested-With': 'XMLHttpRequest',
+ 'X-Instagram-GIS': hashlib.md5(
+ ('%s:%s' % (gis_tmpl, variables)).encode('utf-8')).hexdigest(),
+ }, query={
+ 'query_hash': '42323d64886122307be10013ad2dcc44',
+ 'variables': variables,
+ })['data']['user']['edge_owner_to_timeline_media']
+ self._gis_tmpl = gis_tmpl
+ break
+ except ExtractorError as e:
+ if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
+ if gis_tmpl != gis_tmpls[-1]:
+ continue
+ raise
+
+ edges = media.get('edges')
+ if not edges or not isinstance(edges, list):
+ break
+
+ for edge in edges:
+ node = edge.get('node')
+ if not node or not isinstance(node, dict):