administration/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-13 20:53:06 +00:00
Code Issues Projects Releases Wiki Activity

[ie/twitter] Work around API rate-limit (#8825)

Browse Source 
Closes #8762
Authored by: bashonly
This commit is contained in:
bashonly 2023-12-24 10:41:28 -06:00 committed by GitHub
parent e7d22348e7
commit 116c268438
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 42 additions and 31 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
yt_dlp/extractor/twitter.py
View File

@ -10,6 +10,7 @@ from ..compat import (
compat_urllib_parse_unquote, compat_urllib_parse_unquote,
compat_urllib_parse_urlparse, compat_urllib_parse_urlparse,
) )
from ..networking.exceptions import HTTPError
from ..utils import ( from ..utils import (
ExtractorError, ExtractorError,
dict_get, dict_get,
@ -1317,20 +1318,7 @@ class TwitterIE(TwitterBaseIE):
} }
} }
def _extract_status(self, twid): def _call_syndication_api(self, twid):
if self.is_logged_in or self._selected_api == 'graphql':
status = self._graphql_to_legacy(self._call_graphql_api(self._GRAPHQL_ENDPOINT, twid), twid)
elif self._selected_api == 'legacy':
status = self._call_api(f'statuses/show/{twid}.json', twid, {
'cards_platform': 'Web-12',
'include_cards': 1,
'include_reply_count': 1,
'include_user_entities': 0,
'tweet_mode': 'extended',
})
elif self._selected_api == 'syndication':
self.report_warning( self.report_warning(
'Not all metadata or media is available via syndication endpoint', twid, only_once=True) 'Not all metadata or media is available via syndication endpoint', twid, only_once=True)
status = self._download_json( status = self._download_json(
@ -1350,8 +1338,31 @@ class TwitterIE(TwitterBaseIE):
media.append(detail) media.append(detail)
status['extended_entities'] = {'media': media} status['extended_entities'] = {'media': media}
else: return status
raise ExtractorError(f'"{self._selected_api}" is not a valid API selection', expected=True)
def _extract_status(self, twid):
if self._selected_api not in ('graphql', 'legacy', 'syndication'):
raise ExtractorError(f'{self._selected_api!r} is not a valid API selection', expected=True)
try:
if self.is_logged_in or self._selected_api == 'graphql':
status = self._graphql_to_legacy(self._call_graphql_api(self._GRAPHQL_ENDPOINT, twid), twid)
elif self._selected_api == 'legacy':
status = self._call_api(f'statuses/show/{twid}.json', twid, {
'cards_platform': 'Web-12',
'include_cards': 1,
'include_reply_count': 1,
'include_user_entities': 0,
'tweet_mode': 'extended',
})
except ExtractorError as e:
if not isinstance(e.cause, HTTPError) or not e.cause.status == 429:
raise
self.report_warning('Rate-limit exceeded; falling back to syndication endpoint')
status = self._call_syndication_api(twid)
if self._selected_api == 'syndication':
status = self._call_syndication_api(twid)
return traverse_obj(status, 'retweeted_status', None, expected_type=dict) or {} return traverse_obj(status, 'retweeted_status', None, expected_type=dict) or {}