\begin{code}
module FunDeps (
Equation, pprEquation,
- oclose, grow, improveOne,
+ oclose, improveOne,
checkInstCoverage, checkFunDeps,
pprFundeps
) where
import Name
import Var
import Class
-import TcGadt
import TcType
+import Unify
import InstEnv
import VarSet
import VarEnv
import Outputable
import Util
+import FastString
+
+import Data.List ( nubBy )
import Data.Maybe ( isJust )
\end{code}
%* *
%************************************************************************
+ oclose(vs,C) The result of extending the set of tyvars vs
+ using the functional dependencies from C
+
+ grow(vs,C) The result of extend the set of tyvars vs
+ using all conceivable links from C.
+
+ E.g. vs = {a}, C = {H [a] b, K (b,Int) c, Eq e}
+ Then grow(vs,C) = {a,b,c}
+
+ Note that grow(vs,C) `superset` grow(vs,simplify(C))
+ That is, simplfication can only shrink the result of grow.
+
+Notice that
+ oclose is conservative v `elem` oclose(vs,C)
+ one way: => v is definitely fixed by vs
+
+ grow is conservative if v might be fixed by vs
+ the other way: => v `elem` grow(vs,C)
+
+----------------------------------------------------------
(oclose preds tvs) closes the set of type variables tvs,
wrt functional dependencies in preds. The result is a superset
of the argument set. For example, if we have
oclose [C (x,y) z, C (x,p) q] {x,y} = {x,y,z}
because if we know x and y then that fixes z.
-Using oclose
-~~~~~~~~~~~~
-oclose is used
-
-a) When determining ambiguity. The type
- forall a,b. C a b => a
-is not ambiguous (given the above class decl for C) because
-a determines b.
-
-b) When generalising a type T. Usually we take FV(T) \ FV(Env),
-but in fact we need
- FV(T) \ (FV(Env)+)
-where the '+' is the oclosure operation. Notice that we do not
-take FV(T)+. This puzzled me for a bit. Consider
-
- f = E
-
-and suppose e have that E :: C a b => a, and suppose that b is
-free in the environment. Then we quantify over 'a' only, giving
-the type forall a. C a b => a. Since a->b but we don't have b->a,
-we might have instance decls like
- instance C Bool Int where ...
- instance C Char Int where ...
-so knowing that b=Int doesn't fix 'a'; so we quantify over it.
-
- ---------------
- A WORRY: ToDo!
- ---------------
-If we have class C a b => D a b where ....
- class D a b | a -> b where ...
-and the preds are [C (x,y) z], then we want to see the fd in D,
-even though it is not explicit in C, giving [({x,y},{z})]
-
-Similarly for instance decls? E.g. Suppose we have
- instance C a b => Eq (T a b) where ...
-and we infer a type t with constraints Eq (T a b) for a particular
-expression, and suppose that 'a' is free in the environment.
-We could generalise to
- forall b. Eq (T a b) => t
-but if we reduced the constraint, to C a b, we'd see that 'a' determines
-b, so that a better type might be
- t (with free constraint C a b)
-Perhaps it doesn't matter, because we'll still force b to be a
-particular type at the call sites. Generalising over too many
-variables (provided we don't shadow anything by quantifying over a
-variable that is actually free in the envt) may postpone errors; it
-won't hide them altogether.
-
+oclose is used (only) when generalising a type T; see extensive
+notes in TcSimplify.
+
+Note [Important subtlety in oclose]
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+Consider (oclose (C Int t) {}), where class C a b | a->b
+Then, since a->b, 't' is fully determined by Int, and the
+uniform thing is to return {t}.
+
+However, consider
+ class D a b c | b->c
+ f x = e -- 'e' generates constraint (D s Int t)
+ -- \x.e has type s->s
+Then, if (oclose (D s Int t) {}) = {t}, we'll make the function
+monomorphic in 't', thus
+ f :: forall s. D s Int t => s -> s
+
+But if this function is never called, 't' will never be instantiated;
+the functional dependencies that fix 't' may well be instance decls in
+some importing module. But the top-level defaulting of unconstrained
+type variables will fix t=GHC.Prim.Any, and that's simply a bug.
+
+Conclusion: oclose only returns a type variable as "fixed" if it
+depends on at least one type variable in the input fixed_tvs.
+
+Remember, it's always sound for oclose to return a smaller set.
+An interesting example is tcfail093, where we get this inferred type:
+ class C a b | a->b
+ dup :: forall h. (Call (IO Int) h) => () -> Int -> h
+This is perhaps a bit silly, because 'h' is fixed by the (IO Int);
+previously GHC rejected this saying 'no instance for Call (IO Int) h'.
+But it's right on the borderline. If there was an extra, otherwise
+uninvolved type variable, like 's' in the type of 'f' above, then
+we must accept the function. So, for now anyway, we accept 'dup' too.
\begin{code}
oclose :: [PredType] -> TyVarSet -> TyVarSet
oclose preds fixed_tvs
- | null tv_fds = fixed_tvs -- Fast escape hatch for common case
- | otherwise = loop fixed_tvs
+ | null tv_fds = fixed_tvs -- Fast escape hatch for common case
+ | isEmptyVarSet fixed_tvs = emptyVarSet -- Note [Important subtlety in oclose]
+ | otherwise = loop fixed_tvs
where
loop fixed_tvs
| new_fixed_tvs `subVarSet` fixed_tvs = fixed_tvs
where
new_fixed_tvs = foldl extend fixed_tvs tv_fds
- extend fixed_tvs (ls,rs) | ls `subVarSet` fixed_tvs = fixed_tvs `unionVarSet` rs
- | otherwise = fixed_tvs
+ extend fixed_tvs (ls,rs)
+ | not (isEmptyVarSet ls) -- Note [Important subtlety in oclose]
+ , ls `subVarSet` fixed_tvs = fixed_tvs `unionVarSet` rs
+ | otherwise = fixed_tvs
tv_fds :: [(TyVarSet,TyVarSet)]
-- In our example, tv_fds will be [ ({x,y}, {z}), ({x,p},{q}) ]
]
\end{code}
-Note [Growing the tau-tvs using constraints]
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-(grow preds tvs) is the result of extend the set of tyvars tvs
- using all conceivable links from pred
-
-E.g. tvs = {a}, preds = {H [a] b, K (b,Int) c, Eq e}
-Then grow precs tvs = {a,b,c}
-
-All the type variables from an implicit parameter are added, whether or
-not they are mentioned in tvs; see Note [Implicit parameters and ambiguity]
-in TcSimplify.
-
-See also Note [Ambiguity] in TcSimplify
-
-\begin{code}
-grow :: [PredType] -> TyVarSet -> TyVarSet
-grow preds fixed_tvs
- | null preds = real_fixed_tvs
- | otherwise = loop real_fixed_tvs
- where
- -- Add the implicit parameters;
- -- see Note [Implicit parameters and ambiguity] in TcSimplify
- real_fixed_tvs = foldr unionVarSet fixed_tvs ip_tvs
-
- loop fixed_tvs
- | new_fixed_tvs `subVarSet` fixed_tvs = fixed_tvs
- | otherwise = loop new_fixed_tvs
- where
- new_fixed_tvs = foldl extend fixed_tvs non_ip_tvs
-
- extend fixed_tvs pred_tvs
- | fixed_tvs `intersectsVarSet` pred_tvs = fixed_tvs `unionVarSet` pred_tvs
- | otherwise = fixed_tvs
-
- (ip_tvs, non_ip_tvs) = partitionWith get_ip preds
- get_ip (IParam _ ty) = Left (tyVarsOfType ty)
- get_ip other = Right (tyVarsOfPred other)
-\end{code}
%************************************************************************
%* *
-- We usually act on an equation by instantiating the quantified type varaibles
-- to fresh type variables, and then calling the standard unifier.
+pprEquation :: Equation -> SDoc
pprEquation (qtvs, pairs)
- = vcat [ptext SLIT("forall") <+> braces (pprWithCommas ppr (varSetElems qtvs)),
- nest 2 (vcat [ ppr t1 <+> ptext SLIT(":=:") <+> ppr t2 | (t1,t2) <- pairs])]
+ = vcat [ptext (sLit "forall") <+> braces (pprWithCommas ppr (varSetElems qtvs)),
+ nest 2 (vcat [ ppr t1 <+> ptext (sLit "~") <+> ppr t2 | (t1,t2) <- pairs])]
\end{code}
Given a bunch of predicates that must hold, such as
-- combined (for error messages)
-- Just do improvement triggered by a single, distinguised predicate
-improveOne inst_env pred@(IParam ip ty, _) preds
+improveOne _inst_env pred@(IParam ip ty, _) preds
= [ ((emptyVarSet, [(ty,ty2)]), pred, p2)
| p2@(IParam ip2 ty2, _) <- preds
, ip==ip2
, not (instanceCantMatch inst_tcs trimmed_tcs)
, eqn <- checkClsFD qtvs fd cls_tvs tys_inst tys
, let p_inst = (mkClassPred cls tys_inst,
- ptext SLIT("arising from the instance declaration at")
- <+> ppr (getSrcLoc ispec))
+ sep [ ptext (sLit "arising from the dependency") <+> quotes (pprFunDep fd)
+ , ptext (sLit "in the instance declaration at")
+ <+> ppr (getSrcLoc ispec)])
]
-improveOne inst_env eq_pred preds
+improveOne _ _ _
= []
-- tys2 = [Maybe t1, t2]
--
-- We can instantiate x to t1, and then we want to force
--- (Tree x) [t1/x] :=: t2
+-- (Tree x) [t1/x] ~ t2
--
-- This function is also used when matching two Insts (rather than an Inst
-- against an instance decl. In that case, qtvs is empty, and we are doing
-> TyVarSet -> [Type] -- Proposed new instance type
-> [Instance]
badFunDeps cls_insts clas ins_tv_set ins_tys
- = [ ispec | fd <- fds, -- fds is often empty
+ = nubBy eq_inst $
+ [ ispec | fd <- fds, -- fds is often empty, so do this first!
let trimmed_tcs = trimRoughMatchTcs clas_tvs fd rough_tcs,
ispec@(Instance { is_tcs = inst_tcs, is_tvs = tvs,
is_tys = tys }) <- cls_insts,
where
(clas_tvs, fds) = classTvsFds clas
rough_tcs = roughMatchTcs ins_tys
+ eq_inst i1 i2 = instanceDFunId i1 == instanceDFunId i2
+ -- An single instance may appear twice in the un-nubbed conflict list
+ -- because it may conflict with more than one fundep. E.g.
+ -- class C a b c | a -> b, a -> c
+ -- instance C Int Bool Bool
+ -- instance C Int Char Char
+ -- The second instance conflicts with the first by *both* fundeps
trimRoughMatchTcs :: [TyVar] -> FunDep TyVar -> [Maybe Name] -> [Maybe Name]
-- Computing rough_tcs for a particular fundep
--- class C a b c | a -> b where ...
+-- class C a b c | a -> b where ...
-- For each instance .... => C ta tb tc
--- we want to match only on the types ta, tc; so our
+-- we want to match only on the type ta; so our
-- rough-match thing must similarly be filtered.
--- Hence, we Nothing-ise the tb type right here
-trimRoughMatchTcs clas_tvs (_,rtvs) mb_tcs
+-- Hence, we Nothing-ise the tb and tc types right here
+trimRoughMatchTcs clas_tvs (ltvs, _) mb_tcs
= zipWith select clas_tvs mb_tcs
where
- select clas_tv mb_tc | clas_tv `elem` rtvs = Nothing
- | otherwise = mb_tc
+ select clas_tv mb_tc | clas_tv `elem` ltvs = mb_tc
+ | otherwise = Nothing
\end{code}
projects / ghc-hetmet.git / blobdiff