Re: [GHC] #15852: Bad axiom produced for polykinded data family

10 Nov 2018

      #15852: Bad axiom produced for polykinded data family
-------------------------------------+-------------------------------------
        Reporter:  goldfire          |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  Compiler          |              Version:  8.7
      Resolution:                    |             Keywords:  TypeFamilies,
                                     |  TypeInType
Operating System:  Unknown/Multiple  |         Architecture:
                                     |  Unknown/Multiple
 Type of failure:  None/Unknown      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by RyanGlScott):

 `IfaceTyConInfo` or `IfaceTyConSort` might be possible candidates for
 storing this information. Currently, these are defined as:

 {{{#!hs
 data IfaceTyConInfo   -- Used to guide pretty-printing
                       -- and to disambiguate D from 'D (they share a name)
   = IfaceTyConInfo { ifaceTyConIsPromoted :: IsPromoted
                    , ifaceTyConSort       :: IfaceTyConSort }

 -- | The various types of TyCons which have special, built-in syntax.
 data IfaceTyConSort = IfaceNormalTyCon          -- ^ a regular tycon

                     | IfaceTupleTyCon !Arity !TupleSort
                       -- ^ e.g. @(a, b, c)@ or @(#a, b, c#)@.

                     | IfaceSumTyCon !Arity
                       -- ^ e.g. @(a | b | c)@

                     | IfaceEqualityTyCon
                       -- ^ A heterogeneous equality TyCon
 }}}

 (The Haddocks for `IfaceTyConSort` are misleading, since //every//
 `IfaceTyCon` has one, not just those with special, built-in syntax.
 Regardless of how we fix this bug, we should change that documentation to
 reflect reality.)

 My vote would be to extend one of these data types with an extra field of
 type `IfaceTyConParent`, which would be sufficient to tell if a `TyCon`
 came from a data family instance or not. Then it would just be a matter of
 plumbing this information to `pprAxBranch` from its call sites.

-- 
Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/15852#comment:3
GHC http://www.haskell.org/ghc/
The Glasgow Haskell Compiler