Closed tkellen closed 9 years ago
@tkellen I definitely think this could go in interpret. it would allow for the tests to be co-located and it's just another method to expose, so interpret could continue being used as it is now but just expose the prepare
method also. thoughts?
My only hesitation is that we'd be shipping interpret with liftoff but expecting consumers to provide their own. That seems weird.
I was thinking we could make it the default but they could override if desired
Hrm. I think I'm gonna leave them separate. If someone wants the loading behavior but also wants to provide an entirely different interpret object, I don't think they should have to inherit all of the existing ones (as they would if this was bundled with interpret).
That's fair. Want me to review/merge this? Any idea why travis is failing?
Yeah, a review/merge would be fantastic. Haven't looked into why this is failing but it's all green on my end.
looks like i messed up the changelog, force pushing a fix for that in a sec.
ready for review
hmm, toml test is failing for me
It looks like the requires aren't being cleaned up in the tests. Let me dig in a bit
bad variable scoping, perhaps? i just force pushed a fix. sorry, i should probably be doing fixups now.
yeah, need to do fixups now. It seems you removed the afterEach
cleanup block
Shouldn't be needed, but I could be wrong. Thanks for looking.
When toml is throwing, iced-coffee-script and babel loaders are showing up in the stack trace. not sure why things aren't being cleaned up.
AHHA! Coffee-script seems to be hook Module.prototype.load
in addition to require.extensions
....ugh
Shesh. Like every one of my modules for cli tools (interpret, v8flags, flagged-respawn, rechoir) cuts into the really gross underpinnings of node.
Got almost everything sorted out but literate coffee/iced on the .md
extension isn't working with my test alterations. If I let coffee-script hijack the load
stuff works but I want to get each test running from a clean cache.
got it. you are going to hate me :stuck_out_tongue:
alright, alright, commit the code already. what abomination awaits? :p
@tkellen reassigning to you for review of all my changes. Your stuff looks good, just need a sanity check on all the test stuff.
Looking good! Will cut a release in a min.
PS: thanks!
cc @phated do you think this revised functionality should just be in interpret?