Commit 283a3465 authored by Ryan Scott's avatar Ryan Scott

Prevent Template Haskell splices from throwing a spurious TypeInType error

Summary:
There was a rather annoying corner case where splicing poly-kinded
Template Haskell declarations could trigger an error muttering about
`TypeInType` not being enabled, whereas the equivalent non-TH code would
compile without issue. This was causing by overzealous validity check in the
renamer, wherein failed to distinguish between two different `Exact` names
with the same `OccName`. As a result, it mistakenly believed some type
variables were being used as both type and kind variables simultaneously! Ack.

This avoids the issue by simply disabling the aforementioned validity check
for Exact names. Fixes #12503.

Test Plan: ./validate

Reviewers: austin, bgamari, goldfire

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D3022
parent 76244ec2
......@@ -1692,18 +1692,20 @@ extract_tv t_or_k ltv@(L _ tv) acc
| isRdrTyVar tv = case acc of
FKTV kvs k_set tvs t_set all
| isTypeLevel t_or_k
-> do { when (occ `elemOccSet` k_set) $
-> do { when (not_exact && occ `elemOccSet` k_set) $
mixedVarsErr ltv
; return (FKTV kvs k_set (ltv : tvs) (t_set `extendOccSet` occ)
(ltv : all)) }
| otherwise
-> do { when (occ `elemOccSet` t_set) $
-> do { when (not_exact && occ `elemOccSet` t_set) $
mixedVarsErr ltv
; return (FKTV (ltv : kvs) (k_set `extendOccSet` occ) tvs t_set
(ltv : all)) }
| otherwise = return acc
where
occ = rdrNameOcc tv
-- See Note [TypeInType validity checking and Template Haskell]
not_exact = not $ isExact tv
mixedVarsErr :: Located RdrName -> RnM ()
mixedVarsErr (L loc tv)
......@@ -1716,3 +1718,37 @@ mixedVarsErr (L loc tv)
-- just used in this module; seemed convenient here
nubL :: Eq a => [Located a] -> [Located a]
nubL = nubBy eqLocated
{-
Note [TypeInType validity checking and Template Haskell]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
extract_tv enforces an invariant that no variable can be used as both a kind
and a type unless -XTypeInType is enabled. It does so by accumulating two sets
of variables' OccNames (one for type variables and one for kind variables) that
it has seen before. If a new type variable's OccName appears in the kind set,
then it errors, and similarly for kind variables and the type set.
This relies on the assumption that any two variables with the same OccName
are the same. While this is always true of user-written code, it is not always
true in the presence of Template Haskell! GHC Trac #12503 demonstrates a
scenario where two different Exact TH-generated names can have the same
OccName. As a result, if one of these Exact names is for a type variable
and the other Exact name is for a kind variable, then extracting them both
can lead to a spurious error in extract_tv.
To avoid such a scenario, we simply don't check the invariant in extract_tv
when the name is Exact. This allows Template Haskell users to write code that
uses -XPolyKinds without needing to enable -XTypeInType.
This is a somewhat arbitrary design choice, as adding this special case causes
this code to be accepted when spliced in via Template Haskell:
data T1 k e
class C1 b
instance C1 (T1 k (e :: k))
Even if -XTypeInType is _not enabled. But accepting too many programs without
the prerequisite GHC extensions is better than the alternative, where some
programs would not be accepted unless enabling an extension which has nothing
to do with the code itself.
-}
{-# LANGUAGE KindSignatures #-}
{-# LANGUAGE PolyKinds #-}
{-# LANGUAGE TemplateHaskell #-}
{-# LANGUAGE TypeFamilies #-}
module T12503 where
import Language.Haskell.TH
data T1 k
class C1 a
$(do TyConI (DataD [] tName [ KindedTV kName kKind] _ _ _)
<- reify ''T1
d <- instanceD (cxt [])
(conT ''C1 `appT`
(conT tName `appT` sigT (varT kName) kKind))
[]
return [d])
data family T2 (a :: b)
data instance T2 b
class C2 a
$(do FamilyI (DataFamilyD tName _ _) [DataInstD [] _ [tyVar] _ _ _]
<- reify ''T2
d <- instanceD (cxt [])
(conT ''C2 `appT` (conT tName `appT` return tyVar))
[]
return [d])
......@@ -363,6 +363,7 @@ test('T12478_2', omit_ways(['ghci']), compile_and_run, ['-v0'])
test('T12478_3', omit_ways(['ghci']), compile, ['-v0'])
test('T12478_4', omit_ways(['ghci']), compile_fail, ['-v0'])
test('T12478_5', omit_ways(['ghci']), compile, ['-v0'])
test('T12503', normal, compile, ['-v0'])
test('T12513', omit_ways(['ghci']), compile_fail, ['-v0'])
test('T12530', normal, compile, ['-v0 -ddump-splices -dsuppress-uniques'])
test('T12646', normal, compile, ['-v0'])
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment