blob: 95d20eaf6373f584b40bff33ad79d0cb3f5b9931 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
|
$NetBSD: patch-src_Darcs_UI_SelectChanges.hs,v 1.1 2022/02/14 14:45:09 pho Exp $
GHC 9.2 cannot seem to unify a rank-2 type "forall wX wY. FL p wX wY -> FL
p wX wY" with just "a -> a". Attempting to do so results in a compilation
error. pho@ has no idea why. Is this a compiler bug???
Work around the issue by expanding "maybe" by hand.
--- src/Darcs/UI/SelectChanges.hs.orig 2022-02-14 13:33:09.429529658 +0000
+++ src/Darcs/UI/SelectChanges.hs
@@ -346,7 +346,9 @@ runInvertibleSelection ps psc = runReade
-- after selecting with a splitter, the results may not be canonical
canonizeAfterSplitter :: (FL p :> FL p) wA wB -> (FL p :> FL p) wA wB
canonizeAfterSplitter (x :> y) =
- let canonizeIfNeeded = maybe id canonizeSplit mspl
+ let canonizeIfNeeded = \a -> case mspl of
+ Just spl -> canonizeSplit spl a
+ Nothing -> a
in canonizeIfNeeded x :> canonizeIfNeeded y
-- retrieve the results of patch selection
|
