Document GHC-10498

message-index/messages/GHC-10498/function-arguments/after/Lib.hs

@@ -0,0 +1,6 @@
+module Lib where
+
+areDifferent :: Eq a => a -> a -> Maybe (a, a)
+areDifferent x y
+  | x == y = Nothing
+areDifferent x y = Just (x, y)

message-index/messages/GHC-10498/function-arguments/before/Lib.hs

@@ -0,0 +1,5 @@
+module Lib where
+
+areDifferent :: a -> a -> Maybe (a, a)
+areDifferent x x = Nothing
+areDifferent x y = Just (x, y)

message-index/messages/GHC-10498/function-arguments/index.md

@@ -0,0 +1,25 @@
+---
+title: Duplicate function arguments
+---
+
+```
+Lib.hs:4:14: error: [GHC-10498]
+    • Conflicting definitions for ‘x’
+      Bound at: Lib.hs:4:14
+                Lib.hs:4:16
+    • In an equation for ‘areDifferent’
+  |
+4 | areDifferent x x = Nothing
+  |
+```
+
+*Advanced topic:* somewhat surprisingly, you are allowed
+to duplicate arguments of type-level functions as in
+
+```haskell
+{-# LANGUAGE DataKinds, TypeFamilies #-}
+type family IsElem x xs where
+  IsElem _ '[]       = 'False
+  IsElem x (x ': xs) = 'True
+  IsElem x (_ ': xs) = IsElem x xs
+```

message-index/messages/GHC-10498/index.md

@@ -0,0 +1,12 @@
+---
+title: Conflicting definitions
+summary: The same entity has more than one definition
+severity: error
+introduced: 9.6.1
+---
+
+GHC does not allow the same entity to have more than one
+definitions. In some contexts it can figure out which one
+supposed to shadow another: local definitions take
+priority over global definitions, etc. Shadowing is just a warning, GHC-63397. However, if the definitions are on the same
+level it becomes a fatal error.

message-index/messages/GHC-10498/type-parameters/after/Lib.hs

@@ -0,0 +1,3 @@
+module Lib where
+
+data Pair a = Pair a a

message-index/messages/GHC-10498/type-parameters/before/Lib.hs

@@ -0,0 +1,3 @@
+module Lib where
+
+data Pair a a = Pair a a

message-index/messages/GHC-10498/type-parameters/index.md

@@ -0,0 +1,13 @@
+---
+title: Duplicate datatype parameters
+---
+
+```
+Lib.hs:3:11: error: [GHC-10498]
+    Conflicting definitions for ‘a’
+    Bound at: Lib.hs:3:11
+              Lib.hs:3:13
+  |
+3 | data Pair a a = Pair a a
+  |
+```

message-index/messages/GHC-10498/typo/after/Lib.hs

@@ -0,0 +1,7 @@
+module Lib where
+
+function :: Int -> Int
+function 1 = 1
+function 2 = 2
+function 3 = 3
+function _ = 4

message-index/messages/GHC-10498/typo/before/Lib.hs

@@ -0,0 +1,7 @@
+module Lib where
+
+function :: Int -> Int
+function 1 = 1
+function 2 = 2
+functlon 3 = 3
+function _ = 4

message-index/messages/GHC-10498/typo/index.md

@@ -0,0 +1,18 @@
+---
+title: Just a typo
+---
+
+```
+Lib.hs:7:1: error: [GHC-29916]
+    Multiple declarations of ‘function’
+    Declared at: Lib.hs:4:1
+                 Lib.hs:7:1
+  |
+7 | function _ = 4
+  |
+```
+
+Because of a typo (`functlon` instead of `function`)
+GHC parses the third equation as a definition of
+`functlon` (without a type signature), while the fourth
+equation becomes the second, conflicting definition of `function`.

message-index/messages/GHC-10498/wildcards/after/Lib.hs

@@ -0,0 +1,8 @@
+{-# LANGUAGE NamedFieldPuns #-}
+
+module Lib where
+
+data Foo = Foo { x :: Int, y :: Int }
+
+foo :: Foo -> Int -> Int
+foo Foo{y} x = x + y

message-index/messages/GHC-10498/wildcards/before/Lib.hs

@@ -0,0 +1,8 @@
+{-# LANGUAGE RecordWildCards #-}
+
+module Lib where
+
+data Foo = Foo { x :: Int, y :: Int }
+
+foo :: Foo -> Int -> Int
+foo Foo{..} x = x + y

message-index/messages/GHC-10498/wildcards/index.md

@@ -0,0 +1,17 @@
+---
+title: Implicit duplication because of `RecordWildCards`
+---
+
+```
+Lib.hs:8:9: error: [GHC-10498]
+    • Conflicting definitions for ‘x’
+      Bound at: Lib.hs:8:9-10
+                Lib.hs:8:13
+    • In an equation for ‘foo’
+  |
+8 | foo Foo{..} x = x + y
+  |
+```
+
+Consider using `NamedFieldPuns` to be more explicit
+about which variables are bound where.