#10653: PatternSynonyms should be imported/exported as part of the wildcard notation -------------------------------------+------------------------------------- Reporter: gridaphobe | Owner: mpickering Type: feature request | Status: new Priority: high | Milestone: 8.0.1 Component: Compiler | Version: 7.11 Resolution: | Keywords: | PatternSynonyms, pattern synonyms Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Revisions: -------------------------------------+------------------------------------- Comment (by gridaphobe):
A possible restriction might be that you can only add P to an export of a data type T if P's type has form ... -> T t1 ..tn.
Richard is right that I removed it because I thought there was some worry
I also prefer having this restriction. that the typechecker would have to get involved but if he says not then I think it's a good idea. As I recall, I had previously suggested that `T(..)` export '''every''' pattern with a type whose result is headed by `T`, which would have required running the typechecker just to determine which things `T(..)` refers to. This, I believe, is what was really concerning Simon. If on the other hand, the export list has to '''explicitly''' name patterns to associate with `T`, then we can do a simple validity/sanity check during typechecking, as Richard says. Thanks for forging ahead with this ticket Matthew! -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/10653#comment:33 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler