Closed KellerFuchs closed 6 years ago
@kaniini I'm doing that as a prelude to trying to use python-ssh2
(as I find it much easier to make that sort of changes if I have tooling that picks up simple-ish mistakes)
I don't understand why we would use python-ssh2? this entire application is built on top of asyncio and that would be even worse performance?
@kaniini Oh, I misunderstood the lack of answer on IRC as signifying you had no opposition to trying it. python-ssh2 is still usable asynchronously, and it's 5-10× faster than asyncssh (mostly because it's a wrapper around libssh2, which is written in C). I suggested that as a way to address the performance issue mentionned in the README.
If you think it's a bad idea, I just won't do it :3
PS: Regardless of whether python-ssh2 turns out to be a better ssh library for antissh, good linting and tests might be useful (and uncovered some warts), so I'd really want your opinion/feedback.
I have no objection to this PR then.
@kaniini I think this should be good to go, if you are OK with it.
You (or your favorite CI service) can run tox
to run the checks; there are still a couple of things flagged by pylint
, but I think they are false positives (with the exception of TODO).
gently pings @kaniini
flake8
,pylint
, andpydocstrings
.flake8
.pylint
.