Another day, another failure to make FitLibrary do what I’m asking it to do. I’ve been trying off and on for about a month now to implement double comparisons with tolerances, and every time I think I’ve made progress it turns out to be illusory.
Yesterday I implemented a custom Parser subclass and registered it for double.class in my spreadsheet combination fixture. No luck. Based on some internet searching I tried to use the DebugCapability page, but that doesn’t actually exist (anymore?), so today I tried to use the run method in the fitlibrary.debug.DebugPage class, which seems like it should do the trick to at least let me figure out why my parser isn’t being called. That gave me a NoClassDefFound/ClassNotFoundException stack trace; I tried both a web URL and a local file path for the directory parameter, but neither worked. Of course, even finding the DebugPage class was a bit of a chore, since it’s not really documented anywhere.
I think normally this is where I’d throw my hands up and say F THIS NOISE, but the 62 green cells in my serialization test page are making me happy, so I’ve kept going. I downloaded git and got myself a GitHub account, and I guess tomorrow I’ll try building FitLibrary and its associated version of FitNesse (which is nowhere near the nightly, of course). I’m guessing that somewhere in the world there’s a better acceptance test framework for all of my needs. I am also guessing that tomorrow will be the make or break point for my attempt to use this toolset.