1
mirror of https://github.com/yt-dlp/yt-dlp synced 2024-11-26 11:31:07 +01:00

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

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

View File

@ -10,6 +10,7 @@ from ..compat import (
compat_urllib_parse_unquote,
compat_urllib_parse_urlparse,
)
from ..networking.exceptions import HTTPError
from ..utils import (
ExtractorError,
dict_get,
@ -1317,41 +1318,51 @@ class TwitterIE(TwitterBaseIE):
}
}
def _extract_status(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',
def _call_syndication_api(self, twid):
self.report_warning(
'Not all metadata or media is available via syndication endpoint', twid, only_once=True)
status = self._download_json(
'https://cdn.syndication.twimg.com/tweet-result', twid, 'Downloading syndication JSON',
headers={'User-Agent': 'Googlebot'}, query={
'id': twid,
# TODO: token = ((Number(twid) / 1e15) * Math.PI).toString(36).replace(/(0+|\.)/g, '')
'token': ''.join(random.choices('123456789abcdefghijklmnopqrstuvwxyz', k=10)),
})
if not status:
raise ExtractorError('Syndication endpoint returned empty JSON response')
# Transform the result so its structure matches that of legacy/graphql
media = []
for detail in traverse_obj(status, ((None, 'quoted_tweet'), 'mediaDetails', ..., {dict})):
detail['id_str'] = traverse_obj(detail, (
'video_info', 'variants', ..., 'url', {self._MEDIA_ID_RE.search}, 1), get_all=False) or twid
media.append(detail)
status['extended_entities'] = {'media': media}
elif self._selected_api == 'syndication':
self.report_warning(
'Not all metadata or media is available via syndication endpoint', twid, only_once=True)
status = self._download_json(
'https://cdn.syndication.twimg.com/tweet-result', twid, 'Downloading syndication JSON',
headers={'User-Agent': 'Googlebot'}, query={
'id': twid,
# TODO: token = ((Number(twid) / 1e15) * Math.PI).toString(36).replace(/(0+|\.)/g, '')
'token': ''.join(random.choices('123456789abcdefghijklmnopqrstuvwxyz', k=10)),
return status
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',
})
if not status:
raise ExtractorError('Syndication endpoint returned empty JSON response')
# Transform the result so its structure matches that of legacy/graphql
media = []
for detail in traverse_obj(status, ((None, 'quoted_tweet'), 'mediaDetails', ..., {dict})):
detail['id_str'] = traverse_obj(detail, (
'video_info', 'variants', ..., 'url', {self._MEDIA_ID_RE.search}, 1), get_all=False) or twid
media.append(detail)
status['extended_entities'] = {'media': media}
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)
else:
raise ExtractorError(f'"{self._selected_api}" is not a valid API selection', expected=True)
if self._selected_api == 'syndication':
status = self._call_syndication_api(twid)
return traverse_obj(status, 'retweeted_status', None, expected_type=dict) or {}