Closed roboguy13 closed 8 years ago
Okay, the differences in those tests look like normal variation between how GHC 7.10.2 and HEAD prints certain things. The detailed differences can be seen here. I also notice that I don't see the stray [0m
s in the output of this newer test, so that might have just been a fluke.
I went ahead and added a separate tests/golden-ghc-7.11
directory to this branch with the HEAD version of the output. This directory is conditionally selected by the test suite (by CPP) if the GHC version is > 7.10.2. Otherwise, it uses the original golden files.
There were more changes in HEAD than I was expecting, but I think I have this more or less right. I would like to find a way for the CPP conditionals to look nicer though.
It looks like the tests don't pass on HEAD because HEAD seems to
show
certain things differently (I# 1
becomesI# 1#
). The thing that worries me the most is that occasionally I'll see an escape code printed literally in the HEAD version (in particular,[0m
).