
9 May
2009
9 May
'09
5:08 a.m.
Hi Jason,
Hi Neil, A bit off-topic, but your post reminded me: Does HLint currently help the user find space leaks? For example, does it recommend strict folds instead of lazy folds? I looked at the FAQ but this was not listed. I don't really know how feasible this is.
It spots when you use replicateM or mapM, but don't use the result, and should have actually used the _ versions. It doesn't spot foldl s that should be foldl' s - in general you only know that if you have the strictness information, which HLint doesn't. You could add the rule foldl ==> foldl', but it's not in general correct, so it won't be added to the default rule set. Thanks Neil