Using MultipartWriter.append_json breaks in 3.9.4 with AssertionError "assert CONTENT_DISPOSITION in payload.headers" about aiohttp HOT 8 CLOSED

Ah, wait, you're getting the payload after the append and then setting it, but the check is happening during the append. Maybe we can move the check later...

from aiohttp.

Got this issue here: https://github.com/aiogram/aiogram/actions/runs/8664683869/job/23761669236?pr=1459

Example of usage with previous version:
https://github.com/aiogram/aiogram/blob/b49939aaff2b1e37520a207b01ea124412641d01/aiogram/webhook/aiohttp_server.py#L155-L163

from aiohttp.

Through these convenience methods, we should probably add Content-Disposition automatically. But, I guess the one question is how to name the parts? Shouldn't there be an expected name by the receiving service, which the user should be using?

from aiohttp.

Note that these are just asserts, so you can likely use python -O as a workaround at the moment.

from aiohttp.

Note that these are just asserts, so you can likely use python -O as a workaround at the moment.

But this update breaks all existing tests

What's wrong with this usage?
https://github.com/aiogram/aiogram/blob/b49939aaff2b1e37520a207b01ea124412641d01/aiogram/webhook/aiohttp_server.py#L155-L163

What should I edit to make 3.9.4 works?
It seems like there's a dead end here with new assertions

from aiohttp.

What's wrong with this usage?

Wait, that code does set content-disposition. Why are we getting the error there too?
I was just looking at the code in the start of this issue, which doesn't set content-disposition.

Think you can create a test quickly in a PR here?

from aiohttp.

What's wrong with this usage?

Wait, that code does set content-disposition. Why are we getting the error there too? I was just looking at the code in the start of this issue, which doesn't set content-disposition.

Think you can create a test quickly in a PR here?

Sure, wait a minute

from aiohttp.

#8332

from aiohttp.