[GHC] #7730: :info and polykinds

[GHC]

#7730: :info and polykinds
--------------------------------------------+------------------------------
        Reporter:  monoidal                 |            Owner:  archblob
            Type:  bug                      |           Status:  new
        Priority:  normal                   |        Milestone:  7.10.1
       Component:  Compiler (Type checker)  |          Version:  7.6.2
      Resolution:                           |         Keywords:
Operating System:  Unknown/Multiple         |     Architecture:
 Type of failure:  None/Unknown             |  Unknown/Multiple
       Test Case:                           |       Difficulty:  Unknown
        Blocking:                           |       Blocked By:
                                            |  Related Tickets:  #8776
--------------------------------------------+------------------------------

Comment (by simonpj):

 Simon is right about `ifConOcc`. Good point.

 Rather than pass in the `Name` of the type constructor, you could simply
 pass in the `Module` in which this `IfaceDecl` lives.  Then you can easily
 make a `Name` for all of these things.  It's  bit of a nuisance, but
 perhaps better than making them all `IfaceExtNames`.  Perhaps pass in a
 `Maybe Module`, where `Nothing` means "just print the `OccName`"?

 >  @simonmar suggestion would be easier because we already pass a name to
 {{{pprIfaceDecl}}} in order to print things in context and elide
 everything not searched for

 Not exactly. We pass in a `ShowSub` (or that's what it used to be called
 in `PprTyThing`.  So yes, `ShowSub` could be augmented with that `Maybe
 Module` and then it is already being plumbed to all the right places.

 The TH route is certainly a plausible alternative; a good thought.  As I
 say above, I'm not wedded to doing all this via `IfaceSyn`.  And better
 printing for TH stuff would be good; perhaps even better than better
 printing for Iface stuff (which only hard-core people will look at).

 The place to look is `TcSplice.reifyThing` which is the moral equivalent
 of `MkIface.tyThingToIfaceDecl`.  You'll see that `reifyThing` is monadic,
 but the only reason for that is that it can fail, if TH syntax isn't rich
 enough to deal with the result.  Arguably it'd be good to remove the
 possibility of failure, and thereby force ourselves to enhance TH syntax
 to cope.

 I notice also that `reifyThing` is not doing the requisite tidying (which
 `tyThingToIfaceDecl` is careful to do) so I think there are lurking bugs
 there.

 But (and I have not got time to think this through right now), TH syntax
 models source syntax, and so doesn't have explicit kinds, whereas Iface
 syntax does....

 Probably worth completing the Iface route for now.

 Simon

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7730#comment:41>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler