[GHC] #13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

[GHC] #13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase

GHC - devs mailing list
#13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase
-------------------------------------+-------------------------------------
           Reporter:  dfeuer         |             Owner:  (none)
               Type:  bug            |            Status:  new
           Priority:  normal         |         Milestone:  8.4.1
          Component:  Compiler       |           Version:  8.2.1-rc1
           Keywords:                 |  Operating System:  Unknown/Multiple
  PatternSynonyms,                   |
  PatternMatchWarnings               |
       Architecture:                 |   Type of failure:  Incorrect
  Unknown/Multiple                   |  error/warning at compile-time
          Test Case:                 |        Blocked By:
           Blocking:                 |   Related Tickets:
Differential Rev(s):                 |         Wiki Page:
-------------------------------------+-------------------------------------
 In #8779, mpickering added a `COMPLETE` pragma to allow exhaustiveness
 checking for matches involving pattern synonyms. Unfortunately, there is
 still one piece missing: the interaction with `EmptyCase`. Suppose I write

 {{{#!hs
 newtype Fin (n :: Nat) = Fin Natural  -- constructor not exported

 pattern FZ :: () => forall m. n ~ 'S m => Fin n
 pattern FS :: () => n ~ 'S m => Fin m -> Fin n
 {-# COMPLETE FZ, FS #-}
 }}}

 I would very much like to be able to write

 {{{#!hs
 finZAbsurd :: Fin 'Z -> Void
 finZAbsurd x = case x of
 }}}

 but this gives me a pattern coverage warning! That's certainly not what we
 want. I believe what we actually want is this:

   When checking empty case, check that ''at least one'' complete pattern
 set is impossible.

 In this case, it would see two complete pattern sets: the built-in `{Fin}`
 and the user-decreed `{FZ, FS}`. Since neither `FZ` nor `FS` have matching
 types, we should conclude that the empty case is fine.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13717>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler

_______________________________________________
ghc-tickets mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-tickets
Reply | Threaded
Open this post in threaded view
|

Re: [GHC] #13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase

GHC - devs mailing list
#13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase
-------------------------------------+-------------------------------------
        Reporter:  dfeuer            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.4.1
       Component:  Compiler          |              Version:  8.2.1-rc1
      Resolution:                    |             Keywords:
                                     |  PatternSynonyms,
                                     |  PatternMatchWarnings
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Incorrect         |  Unknown/Multiple
  error/warning at compile-time      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------
Description changed by dfeuer:

@@ -8,1 +8,1 @@
- pattern FZ :: () => forall m. n ~ 'S m => Fin n
+ pattern FZ :: () => n ~ 'S m => Fin n

New description:

 In #8779, mpickering added a `COMPLETE` pragma to allow exhaustiveness
 checking for matches involving pattern synonyms. Unfortunately, there is
 still one piece missing: the interaction with `EmptyCase`. Suppose I write

 {{{#!hs
 newtype Fin (n :: Nat) = Fin Natural  -- constructor not exported

 pattern FZ :: () => n ~ 'S m => Fin n
 pattern FS :: () => n ~ 'S m => Fin m -> Fin n
 {-# COMPLETE FZ, FS #-}
 }}}

 I would very much like to be able to write

 {{{#!hs
 finZAbsurd :: Fin 'Z -> Void
 finZAbsurd x = case x of
 }}}

 but this gives me a pattern coverage warning! That's certainly not what we
 want. I believe what we actually want is this:

   When checking empty case, check that ''at least one'' complete pattern
 set is impossible.

 In this case, it would see two complete pattern sets: the built-in `{Fin}`
 and the user-decreed `{FZ, FS}`. Since neither `FZ` nor `FS` have matching
 types, we should conclude that the empty case is fine.

--

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13717#comment:1>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler

_______________________________________________
ghc-tickets mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-tickets
Reply | Threaded
Open this post in threaded view
|

Re: [GHC] #13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase

GHC - devs mailing list
In reply to this post by GHC - devs mailing list
#13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase
-------------------------------------+-------------------------------------
        Reporter:  dfeuer            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.4.1
       Component:  Compiler          |              Version:  8.2.1-rc1
      Resolution:                    |             Keywords:
                                     |  PatternSynonyms,
                                     |  PatternMatchWarnings
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Incorrect         |  Unknown/Multiple
  error/warning at compile-time      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------
Changes (by Iceland_jack):

 * cc: Iceland_jack (added)


--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13717#comment:2>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler

_______________________________________________
ghc-tickets mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-tickets
Reply | Threaded
Open this post in threaded view
|

Re: [GHC] #13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase

GHC - devs mailing list
In reply to this post by GHC - devs mailing list
#13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase
-------------------------------------+-------------------------------------
        Reporter:  dfeuer            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.4.1
       Component:  Compiler          |              Version:  8.2.1-rc1
      Resolution:                    |             Keywords:
                                     |  PatternSynonyms,
                                     |  PatternMatchWarnings
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Incorrect         |  Unknown/Multiple
  error/warning at compile-time      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by RyanGlScott):

 I tried to reproduce this error, but I couldn't. I don't know what
 implementations for `FZ` and `FS` you had in mind, so I tried whipping up
 my own example from scratch:

 {{{#!hs
 {-# LANGUAGE DataKinds #-}
 {-# LANGUAGE EmptyCase #-}
 {-# LANGUAGE GADTs #-}
 {-# LANGUAGE PatternSynonyms #-}
 {-# OPTIONS_GHC -Wall #-}
 module T (T(TInt, TBool), tCharAbsurd1) where

 import Data.Void

 data T a where
   MkTInt  :: T Int
   MkTBool :: T Bool

 pattern TInt :: () => n ~ Int => T n
 pattern TInt = MkTInt

 pattern TBool :: () => n ~ Bool => T n
 pattern TBool = MkTBool
 {-# COMPLETE TInt, TBool #-}

 tCharAbsurd1 :: T Char -> Void
 tCharAbsurd1 x = case x of {}
 }}}

 This, however, compiles without any warnings. I even tried using `T` in
 another module:

 {{{#!hs
 {-# LANGUAGE EmptyCase #-}
 {-# OPTIONS_GHC -Wall #-}
 module Bug where

 import Data.Void
 import T

 tCharAbsurd2 :: T Char -> Void
 tCharAbsurd2 x = case x of {}
 }}}

 This too compiles without warning:

 {{{
 $ ghc/inplace/bin/ghc-stage2 --interactive Bug.hs
 GHCi, version 8.3.20170516: http://www.haskell.org/ghc/  :? for help
 Loaded GHCi configuration from /home/rgscott/.ghci
 [1 of 2] Compiling T                ( T.hs, interpreted )
 [2 of 2] Compiling Bug              ( Bug.hs, interpreted )
 Ok, modules loaded: Bug, T.
 }}}

 How did you get this supposed pattern coverage warning?

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13717#comment:3>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler

_______________________________________________
ghc-tickets mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-tickets
Reply | Threaded
Open this post in threaded view
|

Re: [GHC] #13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase

GHC - devs mailing list
In reply to this post by GHC - devs mailing list
#13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase
-------------------------------------+-------------------------------------
        Reporter:  dfeuer            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.4.1
       Component:  Compiler          |              Version:  8.2.1-rc1
      Resolution:                    |             Keywords:
                                     |  PatternSynonyms,
                                     |  PatternMatchWarnings
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Incorrect         |  Unknown/Multiple
  error/warning at compile-time      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by dfeuer):

 Here's a full reproduction:

 {{{#!hs
 {-# LANGUAGE DataKinds #-}
 {-# LANGUAGE PatternSynonyms #-}
 {-# LANGUAGE ScopedTypeVariables #-}
 {-# LANGUAGE GADTs #-}
 {-# LANGUAGE KindSignatures #-}
 {-# LANGUAGE ViewPatterns #-}
 {-# LANGUAGE StandaloneDeriving #-}
 {-# LANGUAGE Trustworthy #-}

 module Fin (Nat (..), Fin (FZ, FS), FinView (..), viewFin) where
 import Numeric.Natural
 import Unsafe.Coerce

 data Nat = Z | S Nat

 -- Fin *must* be exported abstractly (or placed in an Unsafe
 -- module) to maintain type safety.
 newtype Fin (n :: Nat) = Fin Natural
 deriving instance Show (Fin n)

 data FinView n where
   VZ :: FinView ('S n)
   VS :: !(Fin n) -> FinView ('S n)

 viewFin :: Fin n -> FinView n
 viewFin (Fin 0) = unsafeCoerce VZ
 viewFin (Fin n) = unsafeCoerce (VS (Fin (n - 1)))

 pattern FZ :: () => n ~ 'S m => Fin n
 pattern FZ <- (viewFin -> VZ) where
   FZ = Fin 0

 pattern FS :: () => n ~ 'S m => Fin m -> Fin n
 pattern FS m <- (viewFin -> VS m) where
   FS (Fin m) = Fin (1 + m)

 {-# COMPLETE FZ, FS #-}
 }}}

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13717#comment:4>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler

_______________________________________________
ghc-tickets mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-tickets
Reply | Threaded
Open this post in threaded view
|

Re: [GHC] #13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase

GHC - devs mailing list
In reply to this post by GHC - devs mailing list
#13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase
-------------------------------------+-------------------------------------
        Reporter:  dfeuer            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.4.1
       Component:  Compiler          |              Version:  8.2.1-rc1
      Resolution:                    |             Keywords:
                                     |  PatternSynonyms,
                                     |  PatternMatchWarnings
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Incorrect         |  Unknown/Multiple
  error/warning at compile-time      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by dfeuer):

 For the record, I can implement `finZAbsurd` in a messier way that the
 pattern checker accepts:

 {{{#!hs
 finZAbsurd :: Fin 'Z -> a
 finZAbsurd = finZAbsurd' Refl
   where
     finZAbsurd' :: n :~: 'Z -> Fin n -> a
     finZAbsurd' pf FZ = case pf of
     finZAbsurd' pf (FS _) = case pf of
 }}}

 but that's a lot uglier.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13717#comment:5>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler

_______________________________________________
ghc-tickets mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-tickets
Reply | Threaded
Open this post in threaded view
|

Re: [GHC] #13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase

GHC - devs mailing list
In reply to this post by GHC - devs mailing list
#13717: Pattern synonym exhaustiveness checks don't play well with EmptyCase
-------------------------------------+-------------------------------------
        Reporter:  dfeuer            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.4.1
       Component:  Compiler          |              Version:  8.2.1-rc1
      Resolution:                    |             Keywords:
                                     |  PatternSynonyms,
                                     |  PatternMatchWarnings
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Incorrect         |  Unknown/Multiple
  error/warning at compile-time      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by RyanGlScott):

 dfeuer, thanks. Now I see why my example didn't trigger any warnings: the
 definition of the `T` type itself (i.e., the fact that it was a GADT) was
 what ruled out the possibility of matching on any of its constructors in
 `tCharAbsurd`. On the other hand, `Fin` does not have this property, and
 you're relying on the types of the conlikes in its `COMPLETE` set to rule
 out the possibility of matching anything.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13717#comment:6>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler

_______________________________________________
ghc-tickets mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-tickets