Normalise EmptyCase types using the constraint solver

Summary:
Certain `EmptyCase` expressions were mistakently producing
warnings since their types did not have as many type families reduced
as they could have. The most direct way to fix this is to normalise
these types initially using the constraint solver to solve for any
local equalities that may be in scope.

Test Plan: make test TEST=T14813

Reviewers: simonpj, bgamari, goldfire

Reviewed By: simonpj

Subscribers: rwbarton, carter

GHC Trac Issues: #14813

Differential Revision: https://phabricator.haskell.org/D5094

1 files changed, 7 insertions, 0 deletions

diff --git a/compiler/utils/MonadUtils.hs b/compiler/utils/MonadUtils.hs
index 39a76e1cf2..e86bc49708 100644
--- a/compiler/utils/MonadUtils.hs
+++ b/compiler/utils/MonadUtils.hs
@@ -21,6 +21,7 @@ module MonadUtils
         , foldlM, foldlM_, foldrM
         , maybeMapM
         , whenM, unlessM
+        , filterOutM
         ) where
 
 -------------------------------------------------------------------------------
@@ -31,6 +32,7 @@ import GhcPrelude
 
 import Maybes
 
+import Control.Applicative
 import Control.Monad
 import Control.Monad.Fix
 import Control.Monad.IO.Class
@@ -199,3 +201,8 @@ whenM mb thing = do { b <- mb
 unlessM :: Monad m => m Bool -> m () -> m ()
 unlessM condM acc = do { cond <- condM
                        ; unless cond acc }
+
+-- | Like 'filterM', only it reverses the sense of the test.
+filterOutM :: (Applicative m) => (a -> m Bool) -> [a] -> m [a]
+filterOutM p =
+  foldr (\ x -> liftA2 (\ flg -> if flg then id else (x:)) (p x)) (pure [])