WPILib Test Harness Revival

[panicnot42]

There exists a test harness on FIRST forge, however, it hasn't been updated since '09. Has any team considered reviving/rewriting the project? It would seem - from an inital glance - that one would have to reimplement ChipObject with dummy/logging functions

[William Kunkel]

I've been thinking about maybe trying to revive it over the summer, once I had more free time. But I haven't really looked extensively into the existing code.

[panicnot42]

I was thinking more of a ground-up rewrite. I'd like to have it modify less stock WPI code so updates can be made quickly. If I can get enough interest on this thread I'd say it'd be worth starting.

[virtuald]

Originally Posted by panicnot42

I was thinking more of a ground-up rewrite. I'd like to have it modify less stock WPI code so updates can be made quickly. If I can get enough interest on this thread I'd say it'd be worth starting.

As the original author of the test harness, it'd be good to see someone take it on and make it better. We don't use C++ anymore (we use python now), so I have no interest in maintaining it anymore.

Implementing ChipObject, while doable, seems like a rather annoying task. It would be nice if the library could be easily stubbed out so you didn't have to modify a lot of code to write a test harness, but chances are you'll have to modify more than you would like.

One idea I had was to write something to automatically parse the C++ files and generate the test harness files from that. Then anytime there's an update, you just regenerate it. Probably wouldn't be that terrible to do, you might be able to repurpose something like SWIG or SIP to do it.