diff options
Diffstat (limited to 'compiler/Language/Haskell/Syntax/Decls.hs')
| -rw-r--r-- | compiler/Language/Haskell/Syntax/Decls.hs | 85 |
1 files changed, 54 insertions, 31 deletions
diff --git a/compiler/Language/Haskell/Syntax/Decls.hs b/compiler/Language/Haskell/Syntax/Decls.hs index baeef95b17..3ed3e5a4b0 100644 --- a/compiler/Language/Haskell/Syntax/Decls.hs +++ b/compiler/Language/Haskell/Syntax/Decls.hs @@ -73,7 +73,9 @@ module Language.Haskell.Syntax.Decls ( CImportSpec(..), -- ** Data-constructor declarations ConDecl(..), LConDecl, - HsConDeclH98Details, HsConDeclGADTDetails(..), + HsConDeclH98Details, + ConGadtSigBody(..), PrefixConGadtSigBody(..), + anonPrefixConGadtSigArgs, prefixConGadtSigRes, -- ** Document comments DocDecl(..), LDocDecl, docDeclDoc, -- ** Deprecations @@ -1062,8 +1064,7 @@ data ConDecl pass -- implicit. The 'XRec' is used to anchor exact print -- annotations, AnnForall and AnnDot. , con_mb_cxt :: Maybe (LHsContext pass) -- ^ User-written context (if any) - , con_g_args :: HsConDeclGADTDetails pass -- ^ Arguments; never infix - , con_res_ty :: LHsType pass -- ^ Result type + , con_body :: ConGadtSigBody pass -- ^ The argument and result types , con_doc :: Maybe (LHsDoc pass) -- ^ A possible Haddock -- comment. @@ -1108,17 +1109,17 @@ There are two broad ways to classify GADT constructors: data T a where K :: forall a. Ord a => [a] -> ... -> T a -This distinction is recorded in the `con_args :: HsConDetails pass`, which -tracks if we're dealing with a RecCon or PrefixCon. It is easy to distinguish -the two in the AST since record GADT constructors use HsRecTy. This distinction -is made in GHC.Parser.PostProcess.mkGadtDecl. +This distinction is recorded in the `con_body :: ConGadtSigBody pass` field, +which tracks if we're dealing with a RecConGADT or PrefixConGADT. It is easy to +distinguish the two in the AST since record GADT constructors use HsRecTy. This +distinction is made in GHC.Parser.PostProcess.mkGadtDecl. It is worth elaborating a bit more on the process of splitting the argument types of a GADT constructor, since there are some non-obvious details involved. While splitting the argument types of a record GADT constructor is easy (they are stored in an HsRecTy), splitting the arguments of a prefix GADT constructor is trickier. The basic idea is that we must split along the outermost function -arrows ((->) and (%1 ->)) in the type, which GHC.Hs.Type.splitHsFunType +arrows ((->) and (%1 ->)) in the type, which GHC.Hs.Type.splitLHsPrefixGadtSigBody accomplishes. But what about type operators? Consider: C :: a :*: b -> a :*: b -> a :+: b @@ -1182,24 +1183,18 @@ or contexts in two parts: 1. GHC, in the process of splitting apart a GADT's type, extracts out the leading `forall` and context (if they are provided). To accomplish this splitting, the renamer uses the - GHC.Hs.Type.splitLHsGADTPrefixTy function, which is careful not to remove + GHC.Hs.Type.splitLHsGadtTy function, which is careful not to remove parentheses surrounding the leading `forall` or context (as these parentheses can be syntactically significant). If the third result returned - by splitLHsGADTPrefixTy contains any `forall`s or contexts, then they must - be nested, so they will be rejected. + by splitLHsGadtTy contains any `forall`s or contexts, then they must + be nested, so they will be rejected later in the renamer. Note that this step applies to both prefix and record GADTs alike, as they - both have syntax which permits `forall`s and contexts. The difference is - where this step happens: - - * For prefix GADTs, this happens in the renamer (in rnConDecl), as we cannot - split until after the type operator fixities have been resolved. - * For record GADTs, this happens in the parser (in mkGadtDecl). -2. If the GADT type is prefix, the renamer (in the ConDeclGADTPrefixPs case of - rnConDecl) will then check for nested `forall`s/contexts in the body of a - prefix GADT type, after it has determined what all of the argument types are. - This step is necessary to catch examples like MkT4 above, where the nested - quantification occurs after a visible argument type. + both have syntax which permits `forall`s and contexts. +2. The renamer (in GHC.Rename.Module.rnGADTResultTy) will then check for nested + `forall`s/contexts in the body of a GADT constructor type. This step is + necessary to catch examples like MkT4 above, where the nested quantification + occurs after a visible argument type. -} -- | The arguments in a Haskell98-style data constructor. @@ -1208,15 +1203,43 @@ type HsConDeclH98Details pass -- The Void argument to HsConDetails here is a reflection of the fact that -- type applications are not allowed in data constructor declarations. --- | The arguments in a GADT constructor. Unlike Haskell98-style constructors, --- GADT constructors cannot be declared with infix syntax. As a result, we do --- not use 'HsConDetails' here, as 'InfixCon' would be an unrepresentable --- state. (There is a notion of infix GADT constructors for the purposes of --- derived Show instances—see Note [Infix GADT constructors] in --- GHC.Tc.TyCl—but that is an orthogonal concern.) -data HsConDeclGADTDetails pass - = PrefixConGADT [HsScaled pass (LBangType pass)] - | RecConGADT (XRec pass [LConDeclField pass]) (LHsUniToken "->" "→" pass) +-- | The argument and result types in a GADT constructor. +-- See @Note [GADT abstract syntax]@. +-- +-- Unlike Haskell98-style constructors, GADT constructors cannot be declared +-- with infix syntax. As a result, we do not use 'HsConDetails' here, as +-- 'InfixCon' would be an unrepresentable state. (There is a notion of infix +-- GADT constructors for the purposes of derived 'Show' instances—see +-- @Note [Infix GADT constructors]@ in "GHC.Tc.TyCl"—but that is an +-- orthogonal concern.) +data ConGadtSigBody pass + = PrefixConGADT (PrefixConGadtSigBody pass) + | RecConGADT (XRec pass [LConDeclField pass]) (LHsUniToken "->" "→" pass) (LHsType pass) + +-- | The argument and result types in a prefix GADT constructor. This closely +-- resembles the structure of 'HsType', but with data constructor–specific +-- tweaks. See @Note [GADT abstract syntax]@. +data PrefixConGadtSigBody pass + -- | The result type. + = PCGSRes (LHsType pass) + -- | An argument followed by a function arrow (e.g., @MkT :: !Int -> ...@). + -- This is much like 'HsFunTy', except that 'PCGSAnonArg' uses an + -- 'LBangType' instead of an 'LHsType' to represent the argument. + | PCGSAnonArg (HsScaled pass (LBangType pass)) + (PrefixConGadtSigBody pass) + +-- | Retrieve the visible, non-dependent arguments in a prefix GADT +-- constructor type. Note that this takes O(/n/) time, where /n/ is the number +-- of arguments. +anonPrefixConGadtSigArgs :: PrefixConGadtSigBody pass -> [HsScaled pass (LBangType pass)] +anonPrefixConGadtSigArgs (PCGSAnonArg arg body) = arg : anonPrefixConGadtSigArgs body +anonPrefixConGadtSigArgs PCGSRes{} = [] + +-- | Retrieve the result type in a prefix GADT constructor type. Note that this +-- takes O(/n/) time, where /n/ is the number of arguments. +prefixConGadtSigRes :: PrefixConGadtSigBody pass -> LHsType pass +prefixConGadtSigRes (PCGSAnonArg _ body) = prefixConGadtSigRes body +prefixConGadtSigRes (PCGSRes res_ty) = res_ty instance Outputable NewOrData where ppr NewType = text "newtype" |