#14529: Refactor ConDecl -------------------------------------+------------------------------------- Reporter: simonpj | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1 Resolution: | Keywords: 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 simonpj): Here's what I have {{{ data ConDecl pass = ConDeclGADT { con_names :: [Located (IdP pass)] -- The next four fields describe the type after the '::' -- See Note [GADT abstract syntax] , con_forall :: Bool -- ^ True <=> explicit forall -- False => hsq_explicit is empty , con_qvars :: LHsQTyVars pass , con_mb_cxt :: Maybe (LHsContext pass) -- ^ User-written context (if any) , con_args :: HsConDeclDetails pass -- ^ Arguments; never InfixCon , con_res_ty :: LHsType pass -- ^ Result type , con_doc :: Maybe LHsDocString -- ^ A possible Haddock comment. } | ConDeclH98 { con_name :: Located (IdP pass) , con_forall :: Bool -- ^ True <=> explicit user-written forall -- e.g. data T a = forall b. MkT b (b->a) -- con_qvars = {b} -- False => hsq_implicit, hsq_explicit both empty , con_qvars :: LHsQTyVars pass -- Existentials only , con_mb_cxt :: Maybe (LHsContext pass) -- ^ User-written context (if any) , con_args :: HsConDeclDetails pass -- ^ Arguments , con_doc :: Maybe LHsDocString -- ^ A possible Haddock comment. } }}} `LHsQTYVars` seems right to me. I need those implicit binders. But not, I grant you, in the H98 case, because we never bind implicit variables there. I can change that too. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/14529#comment:2 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler