Find a way to automate testing

[RyanSchw]

Given the depth of code #7 has as part of feature adding, I think it would be good to revisit the idea of creating some test code for core functionality. I have no clue how this could be done, but I'm curious to see if @AdamBusch or @jacklecroy have any interest or ideas on how to create some test cases.

[RyanSchw]

Maybe something like this could be useful?

[jacklecroy]

Yeah we could try using pytest or something. We would have to manually write tests for everything, though.

[RyanSchw]

Sorry, I can better explain. Writing the tests should be a manual process anyways, so I don't mean automatically create tests. I mean that we'll manually write the tests and then use TravisCI (which we use for the linter already) to execute those tests.

The point of difficulty will be testing something like a serial input (so we'll probably have to find mock libraries for everything). Once we figure out how to mock everything, we could use pytest or the native unittest module to test those assertions.

[jacklecroy]

OK, is there anything right now you think should have testing?

[RyanSchw]

Ideally all of it, lol. Would be good to start off creating tests for the things you know already work (actual physical tests) so that you have a point of reference. Essentially we should be able to verify the system/code will work without having to go outside and run the code on the RPi.

…

________________________________ From: Jack LeCroy <notifications@github.com> Sent: Wednesday, January 30, 2019 6:19 PM To: PurdueElectricRacing/Wireless-Telemetry-Server Cc: Ryan Joseph Schwartz; Author Subject: Re: [PurdueElectricRacing/Wireless-Telemetry-Server] Find a way to automate testing (#11) OK, is there anything right now you think should have testing? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<#11 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/Aeu3M-VXKdXpu8rYxvGg9E7fmMYVY_9Oks5vIiiXgaJpZM4aa9M5>.