Fix wrong signature call for ConversationHandler.TIMEOUT handlers (issue #1652)

[tobiaswicker]

Fix for issue #1652

[tobiaswicker]

Forced push was necessary to trigger the checks to run again, because one failed for unrelated reasons.

[Eldinnie]

`Thanks for making this fix.
Can we explore adding a test to make sure this (keeps) work(ing/s) as intended?

[tobiaswicker]

@Eldinnie PTB does have tests for the ConversationHandler and there is even a test_conversation_handler_timeout_state() test case that specifically tests the ConversationHandler.TIMEOUT state. I think the issue you are talking about, is that those tests don't use_context, right?

So would you like to see the existing tests migrated to use context or rather have separate tests? I'd opt for the latter for backward compatibility for now.

[Poolitzer]

@tobiaswicker Switching all the tests to contexts is actually an issue, #1509

[tobiaswicker]

@Poolitzer OK, thanks!

@Eldinnie does the mentioned issue satisfy your request for a test or have you had a test in mind that wouldn't be covered by migrating any of the existing tests?

[Poolitzer]

@tobiaswicker I am very sure that solving that issue will satisfy his testing needs

[Bibo-Joshi]

@Poolitzer Note, that #1509 is currently assigned only to the v13,0 milestone, which will probably not be closed too soon.

[Poolitzer]

@Bibo-Joshi I mean, we are talking about a test for a callback here. I dont think it will break soon enough that we will have a problem with adding this test later, do you?

[Bibo-Joshi]

@Poolitzer I guess not. Just wanted to mention it ;)

[Poolitzer]

@Bibo-Joshi Fair enough

[Eldinnie]

@tobiaswicker IMO the new PR should be paired with a test. One that would fail without this fix, and will pass with this fix.

[tsnoam]

@tobiaswicker thankyou for your contribution