LonamiWebs/Telethon
1
1
Fork 0
mirror of https://github.com/LonamiWebs/Telethon.git synced 2024-11-22 09:26:37 +03:00
Code Issues Packages Projects Releases Wiki Activity
3,053 Commits 5 Branches 171 Tags 25 MiB
3f589b287d
Commit Graph

151 Commits

Author SHA1 Message Date
Lonami Exo
1974b663a2 Add missing type to _store_own_updates 2024-04-28 10:16:56 +02:00
Lonami Exo
7ce0b2f940 Fix invalid date type in UpdateShort 2023-09-12 17:16:55 +02:00
Lonami Exo
128b707488 Don't treat asyncio.IncompleteReadError as unhandled 
The library will behave the same, but the log severity is lowered.
 2023-08-03 19:01:10 +02:00
Lonami Exo
4ff7ac6b75 Handle CancelledError inside mtprotosender recv loop 2023-06-02 19:04:51 +02:00
Lonami Exo
be59c36ed3 Handle errors at connection level 
Closes #4042.
 2023-03-12 17:43:36 +01:00
Lonami Exo
d81eb0b2e8 Apply pts returned by some additional requests 
When a bot account sends a message, deletes it, and sends a new one,
very reliably it would detect a gap, and as a result recover the
second message it sent, processing it itself (because the hack with
`_self_outgoing` cannot possibly work when catching up).

Now certain `rpc_result` are also processed as-if they were updates
(including the ones from deleting messages), which solves this gap
issue. Not entirely sure if it's a hack or the intended way to do it
(since Telegram *does* return proper `updates` for other RPCs), but
it seems to solve this particular problem.

Other requests such as reading history, mentions or reactions also
return an instance of this type, but the `pts_count` should be 0,
and at worst it should simply trigger a gap, which shouldn't be a
big deal.
 2023-01-14 12:31:01 +01:00
Lonami Exo
83bafa25e3 Stop using asyncio.get_event_loop() 
It is deprecated in newer Python versions.
Closes #4013.
 2023-01-11 21:02:29 +01:00
Lonami Exo
2ffac2dcdb Handle DestroyAuthKey result more gracefully 2022-10-21 15:52:00 +02:00
Lonami Exo
f902c9293a Fix log message formatting when obj was not updates 2022-10-21 15:50:37 +02:00
Lonami Exo
a7db08d020 Fix sender destroy session handling was not running 2022-10-21 15:49:18 +02:00
Lonami Exo
7d21b40401 Revert "Make sessions async" 
This reverts commit d2de0f3aca.
 2022-08-30 12:32:21 +02:00
Lonami Exo
1af6d9a873 Properly log RpcError with no parent request 
This should get rid of the unexpected BufferError traceback.
 2022-05-31 11:02:34 +02:00
Lonami Exo
3a44f56f64 Also process own updates in MessageBox 2022-05-19 16:40:32 +02:00
Lonami Exo
184984ac51 Protect against potential replay attacks 
See #3753.
 2022-05-18 12:24:28 +02:00
Lonami Exo
db09a92bc5 Make use of the new MessageBox 2022-05-13 13:17:16 +02:00
Lonami Exo
fdb0720fe9 Don't reset the auth_key upon receiving -404 2022-05-12 12:03:48 +02:00
Tulir Asokan
d2de0f3aca Make sessions async 
SQLiteSession is not updated, don't try to use it
 2022-05-12 10:36:10 +02:00
Lonami Exo
4c771bf2af Fix setting logout result was not checking for future cancellation 2021-02-27 15:14:44 +01:00
Lonami Exo
292a36f760 Handle DestroySessionRes 
Should close #1706.
 2021-02-27 15:13:53 +01:00
Lonami Exo
8f0de3d285 Fix TypeNotFoundError was not being propagated 
Closes #1697. This would cause deadlocks, as the request future
would never be resolved, so await would wait forever.
 2021-02-11 19:27:57 +01:00
Lonami Exo
ee0fc5cc29 Fix AttributeError on reconnect with no retries 
Closes #1646.
 2020-12-11 21:43:39 +01:00
Lonami Exo
becfe2ce7a Start reconnect if a second ping is sent without a pong for the first 
May help with #1564.
 2020-12-11 17:18:25 +01:00
Lonami Exo
026c992395 Don't try to reconnect when authkey is invalid 2020-10-07 10:40:34 +02:00
Lonami Exo
09f4c5c708 Only reset auth_key on error -404 
This error is "auth key not found", and the authorization key should
probably not be reset on other error codes. This might address #1457.
 2020-10-05 14:08:21 +02:00
Lonami Exo
185a93a105 Expect BufferError during automatic reconnect 
This seems to occur whe the Telegram servers are dying and logging
everyone out.
 2020-10-05 14:07:11 +02:00
Lonami Exo
10b2b60415 Fix requests were not re-enqueued if sending failed 2020-09-29 21:07:38 +02:00
yash-dk
1d6fd7898a
Consider all reconnect attempts as retrying (#1557) 
This means that a value of 0 retries will no longer try to reconnect.
 2020-09-13 09:43:01 +02:00
Lonami Exo
1c3e7dda01 Avoid explicitly passing the loop to asyncio 
This behaviour is deprecated and will be removed in future versions
of Python. Technically, it could be considered a bug (invalid usage
causing different behaviour from the expected one), and in practice
it should not break much code (because .get_event_loop() would likely
be the same event loop anyway).
 2020-07-25 18:39:35 +02:00
Lonami Exo
1e94fe25fa Log requests that trigger struct.error 
The exception hardly provides any valuable information.
This will hopefully help troubleshooting why the error happens.
 2020-02-20 13:40:08 +01:00
Lonami Exo
3c253734ac Clear old docs and fix formatting in ConnectionError messages 2020-01-07 12:20:01 +01:00
Lonami Exo
cd37478e31 Don't send pings unless the connection is made 
This will hopefully avoid batching tens of ping requests which
we don't care about their results.
 2019-12-02 18:36:20 +01:00
Lonami Exo
5dcc30dcc6 Handle connection errors during auth key generation 
This should help with spurious server-side disconnects during
auth_key generation, which happen most commonly on user DC
migrations.
 2019-10-24 13:48:29 +02:00
Lonami Exo
08b78f0c47 Unify retry loops in mtprotosender._connect 
Now the retry count is not twice its value.
 2019-10-24 13:40:09 +02:00
Lonami Exo
3039915ce9 Factor out parts from mtprotosender._connect 
This will help unifying the retry loop and reconnecting
if the server disconnects us during auth key generation
which will be done in a follow-up commit.
 2019-10-24 13:36:32 +02:00
Lonami Exo
e24dd3ad75 Prevent double-connect causing double-reads later 
Which leads to "RuntimeError: readexactly() called while another
coroutine is already waiting for incoming data" being raised,
and causing everything to break or halt.
 2019-08-11 19:05:11 +02:00
Lonami Exo
f5de2cd9a0 Fix logging bug on disconnect 2019-08-08 09:32:18 +02:00
Lonami Exo
a7a7c4add2 Pass all Updates when building events 2019-06-30 16:34:34 +02:00
Lonami Exo
aa2b3daccc Factor out setting entities to events 2019-06-30 16:34:34 +02:00
Lonami
86cdb7c1f8
Remove accidentally-committed debug print 2019-06-15 23:14:59 +02:00
Lonami Exo
40d32cee95 Assert that dispatched updates are Updates 2019-06-15 21:15:57 +02:00
Lonami Exo
e5f1b2afa3 Update logging calls to use proper formatting 2019-05-17 12:30:13 +02:00
Lonami Exo
c1be0bd2e8 Fix disconnection without a previous connection 2019-05-11 16:53:11 +02:00
Lonami Exo
9965cda968 Save pts and date in a tuple for immutability 
This way it is easy and cheap to copy the two required values
to all incoming updates in case we need to getDifference since
the previous pts/date to fetch entities.

This is still a work in progress.
 2019-04-10 21:10:34 +04:00
Lonami Exo
ad963fd23e Don't clear pending_ack on disconnect 
Upon reconnecting, we must make sure to send all `pending_ack`,
or Telegram may resend some responses (e.g. causing duplicated
updates).
 2019-03-28 12:16:15 +01:00
Lonami Exo
5554b414e1 Propagate the last error on reconnect, not always ConnectionError 2019-03-28 10:11:33 +01:00
Lonami Exo
7523869875 Except IOError and not ConnectionError 
PySocks' errors do not subclass ConnectionError so the library
was unnecessarily logging them as unexpected, when any IOError
was actually expected.
 2019-03-28 09:54:35 +01:00
Lonami Exo
f6fe580eb7 Safer auto reconnect to prevent double connect 2019-03-22 19:01:40 +01:00
Lonami Exo
436fb64289 Prevent double autoreconnect like #1112 2019-03-22 16:21:18 +01:00
Lonami Exo
04ba2e1fc7 Revert disconnect() to be async again (#1133) 
It's the only way to properly clean all background tasks,
which the library makes heavy use for in MTProto/Connection
send and receive loops.

Some parts of the code even relied on the fact that it was
asynchronous (it used to return a future so you could await
it and not be breaking).

It's automatically syncified to reduce the damage of being
a breaking change.
 2019-03-21 12:21:00 +01:00
Dmitry D. Chernov
fc2977fc0d Tiny style fixes 2019-02-07 06:55:34 +10:00