GitLab

The ConDecl type in HsDecls is an uneasy compromise. For the most part,
HsSyn directly reflects the syntax written by the programmer; and that
gives just the right "pegs" on which to hang Alan's API annotations. But
ConDecl doesn't properly reflect the syntax of Haskell-98 and GADT-style
data type declarations.

To be concrete, here's a draft new data type

```lang=hs
data ConDecl name
  | ConDeclGADT
      { con_names   :: [Located name]
      , con_type    :: LHsSigType name  -- The type after the ‘::’
      , con_doc     :: Maybe LHsDocString }

  | ConDeclH98
      { con_name    :: Located name

      , con_qvars     :: Maybe (LHsQTyVars name)
        -- User-written forall (if any), and its implicit
        -- kind variables
        -- Non-Nothing needs -XExistentialQuantification

      , con_cxt       :: Maybe (LHsContext name)
        -- ^ User-written context (if any)

      , con_details   :: HsConDeclDetails name
          -- ^ Arguments

      , con_doc       :: Maybe LHsDocString
          -- ^ A possible Haddock comment.
      } deriving (Typeable)
```

Note that

    For GADTs, just keep a type. That's what the user writes.
    NB:HsType can represent records on the LHS of an arrow:

      { x:Int,y:Bool} -> T

    con_qvars and con_cxt are both Maybe because they are both
    optional (the forall and the context of an existential data type

    For ConDeclGADT the type variables of the data type do not scope
    over the con_type; whereas for ConDeclH98 they do scope over con_cxt
    and con_details.

Updates haddock submodule.

Test Plan: ./validate

Reviewers: simonpj, erikd, hvr, goldfire, austin, bgamari

Subscribers: erikd, goldfire, thomie, mpickering

Differential Revision: https://phabricator.haskell.org/D1558

GHC Trac Issues: #11028

This patch began as a modest refactoring of HsType and friends, to
clarify and tidy up exactly where quantification takes place in types.
Although initially driven by making the implementation of wildcards more
tidy (and fixing a number of bugs), I gradually got drawn into a pretty
big process, which I've been doing on and off for quite a long time.

There is one compiler performance regression as a result of all
this, in perf/compiler/T3064.  I still need to look into that.

* The principal driving change is described in Note [HsType binders]
  in HsType.  Well worth reading!

* Those data type changes drive almost everything else.  In particular
  we now statically know where

       (a) implicit quantification only (LHsSigType),
           e.g. in instance declaratios and SPECIALISE signatures

       (b) implicit quantification and wildcards (LHsSigWcType)
           can appear, e.g. in function type signatures

* As part of this change, HsForAllTy is (a) simplified (no wildcards)
  and (b) split into HsForAllTy and HsQualTy.  The two contructors
  appear when and only when the correponding user-level construct
  appears.  Again see Note [HsType binders].

  HsExplicitFlag disappears altogether.

* Other simplifications

     - ExprWithTySig no longer needs an ExprWithTySigOut variant

     - TypeSig no longer needs a PostRn name [name] field
       for wildcards

     - PatSynSig records a LHsSigType rather than the decomposed
       pieces

     - The mysterious 'GenericSig' is now 'ClassOpSig'

* Renamed LHsTyVarBndrs to LHsQTyVars

* There are some uninteresting knock-on changes in Haddock,
  because of the HsSyn changes

I also did a bunch of loosely-related changes:

* We already had type synonyms CoercionN/CoercionR for nominal and
  representational coercions.  I've added similar treatment for

      TcCoercionN/TcCoercionR

      mkWpCastN/mkWpCastN

  All just type synonyms but jolly useful.

* I record-ised ForeignImport and ForeignExport

* I improved the (poor) fix to Trac #10896, by making
  TcTyClsDecls.checkValidTyCl recover from errors, but adding a
  harmless, abstract TyCon to the envt if so.

* I did some significant refactoring in RnEnv.lookupSubBndrOcc,
  for reasons that I have (embarrassingly) now totally forgotten.
  It had to do with something to do with import and export

Updates haddock submodule.

This patch does some signficant refactoring to the treatment
of default methods in class declarations, and more generally
to the type checking of type/class decls.

Highlights:

* When the class has a generic-default method, such as
     class C a where
       op :: a -> a -> Bool
       default op :: Ord a => a -> a -> a
  the ClassOpItem records the type of the generic-default,
  in this case the type (Ord a => a -> a -> a)

* I killed off Class.DefMeth in favour of the very-similar
  BasicTypes.DefMethSpec.  However it turned out to be better
  to use a Maybe, thus
      Maybe (DefMethSpec Type)
  with Nothing meaning "no default method".

* In TcTyClsDecls.tcTyClGroup, we used to accumulate a [TyThing],
  but I found a way to make it much simpler, accumulating only
  a [TyCon].  Much less wrapping and unwrapping.

* On the way I also fixed Trac #10896 in a better way. Instead
  of killing off all ambiguity checks whenever there are any type
  errors (the fix in commit 8e8b9ed9), I instead recover in
  TcTyClsDecls.checkValidTyCl.

There was a lot of associated simplification all round

After the changes, the three functions used to print type families
were identical, so they are refactored into one.

Original RHSs of data instance declarations are recreated and
printed in user error messages.

RHSs containing representation TyCons are printed in the
Coercion Axioms section in a typechecker dump.

Add vbar to the list of SDocs exported by Outputable.
Replace all text "|" docs with it.

Fixes #10839

Reviewers: goldfire, jstolarek, austin, bgamari

Reviewed By: jstolarek

Subscribers: jstolarek, thomie

Differential Revision: https://phabricator.haskell.org/D1441

GHC Trac Issues: #10839

We no longer parse old-style GADT syntax but there was some left-over
code for emitting deprecation warnings.

Updates haddock submodule.

Test Plan: ./validate

Reviewers: austin, bgamari

Reviewed By: bgamari

Subscribers: thomie

Differential Revision: https://phabricator.haskell.org/D1460

This implements DuplicateRecordFields, the first part of the
OverloadedRecordFields extension, as described at
https://ghc.haskell.org/trac/ghc/wiki/Records/OverloadedRecordFields/DuplicateRecordFields

This includes fairly wide-ranging changes in order to allow multiple
records within the same module to use the same field names.  Note that
it does *not* allow record selector functions to be used if they are
ambiguous, and it does not have any form of type-based disambiguation
for selectors (but it does for updates). Subsequent parts will make
overloading selectors possible using orthogonal extensions, as
described on the wiki pages.  This part touches quite a lot of the
codebase, and requires changes to several GHC API datatypes in order
to distinguish between field labels (which may be overloaded) and
selector function names (which are always unique).

The Haddock submodule has been adapted to compile with the GHC API
changes, but it will need further work to properly support modules
that use the DuplicateRecordFields extension.

Test Plan: New tests added in testsuite/tests/overloadedrecflds; these
will be extended once the other parts are implemented.

Reviewers: goldfire, bgamari, simonpj, austin

Subscribers: sjcjoosten, haggholm, mpickering, bgamari, tibbe, thomie,
goldfire

Differential Revision: https://phabricator.haskell.org/D761

This fixes #10810 by cleaning up pretty-printing of constructor
declarations. This change also removes a (in my opinion) deeply
bogus orphan instance OutputableBndr [Located name], making
HsDecls now a non-orphan module. Yay all around.

Test case: th/T10810

This is part of #10811. It removes the "family" keyword from
associated type family declarations, and it adds the "type"
keyword to associated type family defaults.

For details see #6018, Phab:D202 and the wiki page:

https://ghc.haskell.org/trac/ghc/wiki/InjectiveTypeFamilies

This patch also wires-in Maybe data type and updates haddock submodule.

Test Plan: ./validate

Reviewers: simonpj, goldfire, austin, bgamari

Subscribers: mpickering, bgamari, alanz, thomie, goldfire, simonmar,
             carter

Differential Revision: https://phabricator.haskell.org/D202

GHC Trac Issues: #6018

Accept next docstrings (`-- | Docstring`) on GADT constructors.

I have confirmed that this adds no shift/reduce conflicts.

Test Plan: haddockA034

Reviewers: austin, simonpj, simonmar

Reviewed By: simonmar

Subscribers: Fuuzetsu, simonmar, thomie, mpickering, edsko

Differential Revision: https://phabricator.haskell.org/D1086

Summary:
The strings used in a WARNING pragma are captured via

    strings :: { Located ([AddAnn],[Located FastString]) }
        : STRING { sL1 $1 ([],[L (gl $1) (getSTRING $1)]) }
    ..

The STRING token has a method getSTRINGs that returns the original
source text for a string.

A warning of the form

    {-# WARNING Logic
              , mkSolver
              , mkSimpleSolver
              , mkSolverForLogic
              , solverSetParams
              , solverPush
              , solverPop
              , solverReset
              , solverGetNumScopes
              , solverAssertCnstr
              , solverAssertAndTrack
              , solverCheck
              , solverCheckAndGetModel
              , solverGetReasonUnknown
              "New Z3 API support is still incomplete and fragile: \
              \you may experience segmentation faults!"
      #-}

returns the concatenated warning string rather than the original source.

This patch now deals with all remaining instances of getSTRING to bring
in a SourceText for each.

This updates the haddock submodule as well, for the AST change.

Test Plan: ./validate

Reviewers: hvr, austin, goldfire

Reviewed By: austin

Subscribers: bgamari, thomie, mpickering

Differential Revision: https://phabricator.haskell.org/D907

GHC Trac Issues: #10313

Fixes #9840 and #10306, and includes an alternative resolution to #8028.
This permits empty closed type families, and documents them in the user
guide. It updates the Haddock submodule to support the API change.

Test Plan: Added `indexed-types/should_compile/T9840` and updated
`indexed-types/should_fail/ClosedFam4` and `th/T8028`.

Reviewers: austin, simonpj, goldfire

Reviewed By: goldfire

Subscribers: bgamari, jstolarek, thomie, goldfire

Differential Revision: https://phabricator.haskell.org/D841

GHC Trac Issues: #9840, #10306

The production for opt_kind_sig is

  opt_kind_sig :: { Located (Maybe (LHsKind RdrName)) }
          :                             { noLoc Nothing }
          | '::' kind                   {% ajl (sLL $1 $> (Just $2)) AnnDcolon (gl $1) }

The outer Location is used only to get the full span for the enclosing
declration, and is then stripped. The inner LHsKind then has a SrcSpan that does
not include the '::'

Extend the SrcSpan on $2 to include $1

Reviewed By: austin

Differential Revision: https://phabricator.haskell.org/D813

GHC Trac Issues: #10209

As Trac #10047 points out, a quasi-quotation [n|...blah...|] is supposed
to behave exactly like $(n "...blah...").  But it doesn't!  This was outright
wrong: quasiquotes were being run even inside brackets.

Now that TH supports both typed and untyped splices, a quasi-quote is properly
regarded as a particular syntax for an untyped splice. But apart from that
they should be treated the same.  So this patch refactors the handling of
quasiquotes to do just that.

The changes touch quite a lot of files, but mostly in a routine way.
The biggest changes by far are in RnSplice, and more minor changes in
TcSplice.  These are the places where there was real work to be done.
Everything else is routine knock-on changes.

* No more QuasiQuote forms in declarations, expressions, types, etc.
  So we get rid of these data constructors
    * HsBinds.QuasiQuoteD
    * HsExpr.HsSpliceE
    * HsPat.QuasiQuotePat
    * HsType.HsQuasiQuoteTy

* We get rid of the HsQuasiQuote type altogether

* Instead, we augment the HsExpr.HsSplice type to have three
  consructors, for the three types of splice:
    * HsTypedSplice
    * HsUntypedSplice
    * HsQuasiQuote
  There are some related changes in the data types in HsExpr near HsSplice.
  Specifically: PendingRnSplice, PendingTcSplice, UntypedSpliceFlavour.

* In Hooks, we combine rnQuasiQuoteHook and rnRnSpliceHook into one.
  A smaller, clearer interface.

* We have to update the Haddock submodule, to accommodate the hsSyn changes

Summary: It looks like during .lhs -> .hs switch the comments were not updated. So doing exactly that.

Reviewers: austin, jstolarek, hvr, goldfire

Reviewed By: austin, jstolarek

Subscribers: thomie, goldfire

Differential Revision: https://phabricator.haskell.org/D621

GHC Trac Issues: #9986

Summary:
Add a reference note to each AnnKeywordId haddock comment so GHC
developers will have an idea why they are there.

Add a new test to ghc-api/annotations to serve as a template for other
GHC developers when they need to update the parser. It provides output
which checks that each SrcSpan that an annotation is attached to
actually appears in the `ParsedSource`, and lists the individual
annotations. The idea is that a developer writes a version of this
which parses a sample file using whatever syntax is changed in
Parser.y, and can then check that all the annotations come through.

Depends on D538

Test Plan: ./validate

Reviewers: simonpj, hvr, austin

Reviewed By: austin

Subscribers: thomie, jstolarek

Differential Revision: https://phabricator.haskell.org/D620

Summary:
HsTyLit now has SourceText

Update documentation of HsSyn to reflect which annotations are attached to which element.

Ensure that the parser always keeps HsSCC and HsTickPragma values, to
be ignored in the desugar phase if not needed

Bringing in SourceText for pragmas

Add Location in NPlusKPat

Add Location in FunDep

Make RecCon payload Located

Explicitly add AnnVal to RdrName where it is compound

Add Location in IPBind

Add Location to name in IEThingAbs

Add Maybe (Located id,Bool) to Match to track fun_id,infix
  This includes converting Match into a record and adding a note about why
  the fun_id needs to be replicated in the Match.

Add Location in KindedTyVar

Sort out semi-colons for parsing

  - import statements
  - stmts
  - decls
  - decls_cls
  - decls_inst

This updates the haddock submodule.

Test Plan: ./validate

Reviewers: hvr, austin, goldfire, simonpj

Reviewed By: simonpj

Subscribers: thomie, carter

Differential Revision: https://phabricator.haskell.org/D538

Signed-off-by: Austin Seipp <austin@well-typed.com>

We were wrongly simply dropping the bang, in tidy_bang_pat.

Summary:
The final design and discussion is captured at
https://ghc.haskell.org/trac/ghc/wiki/GhcAstAnnotations

This is a proof of concept implementation of a completely
separate annotation structure, populated in the parser,and tied to the
AST by means of a virtual "node-key" comprising the surrounding
SrcSpan and a value derived from the specific constructor used for the
node.

The key parts of the design are the following.

== The Annotations ==

In `hsSyn/ApiAnnotation.hs`

```lang=haskell
type ApiAnns = (Map.Map ApiAnnKey SrcSpan, Map.Map SrcSpan [Located Token])

type ApiAnnKey = (SrcSpan,AnnKeywordId)

-- ---------------------------------------------------------------------

-- | Retrieve an annotation based on the @SrcSpan@ of the annotated AST
-- element, and the known type of the annotation.
getAnnotation :: ApiAnns -> SrcSpan -> AnnKeywordId -> Maybe SrcSpan
getAnnotation (anns,_) span ann = Map.lookup (span,ann) anns

-- |Retrieve the comments allocated to the current @SrcSpan@
getAnnotationComments :: ApiAnns -> SrcSpan -> [Located Token]
getAnnotationComments (_,anns) span =
  case Map.lookup span anns of
    Just cs -> cs
    Nothing -> []

-- | Note: in general the names of these are taken from the
-- corresponding token, unless otherwise noted
data AnnKeywordId
         = AnnAs
         | AnnBang
         | AnnClass
         | AnnClose -- ^ } or ] or ) or #) etc
         | AnnComma
         | AnnDarrow
         | AnnData
         | AnnDcolon
         ....
```

== Capturing in the lexer/parser ==

The annotations are captured in the lexer / parser by extending PState to include a field

In `parser/Lexer.x`

```lang=haskell
data PState = PState {
        ....
        annotations :: [(ApiAnnKey,SrcSpan)]
        -- Annotations giving the locations of 'noise' tokens in the
        -- source, so that users of the GHC API can do source to
        -- source conversions.
     }
```

The lexer exposes a helper function to add an annotation

```lang=haskell
addAnnotation :: SrcSpan -> Ann -> SrcSpan -> P ()
addAnnotation l a v = P $ \s -> POk s {
  annotations = ((AK l a), v) : annotations s
  } ()

```

The parser also has some helper functions of the form

```lang=haskell
type MaybeAnn = Maybe (SrcSpan -> P ())

gl = getLoc
gj x = Just (gl x)

ams :: Located a -> [MaybeAnn] -> P (Located a)
ams a@(L l _) bs = (mapM_ (\a -> a l) $ catMaybes bs) >> return a
```

This allows annotations to be captured in the parser by means of

```
ctypedoc :: { LHsType RdrName }
        : 'forall' tv_bndrs '.' ctypedoc {% hintExplicitForall (getLoc $1) >>
                                            ams (LL $ mkExplicitHsForAllTy $2 (noLoc []) $4)
                                                [mj AnnForall $1,mj AnnDot $3] }
        | context '=>' ctypedoc         {% ams (LL $ mkQualifiedHsForAllTy   $1 $3)
                                               [mj AnnDarrow $2] }
        | ipvar '::' type               {% ams (LL (HsIParamTy (unLoc $1) $3))
                                               [mj AnnDcolon $2] }
        | typedoc                       { $1 }
```

== Parse result ==

```lang-haskell
data HsParsedModule = HsParsedModule {
    hpm_module    :: Located (HsModule RdrName),
    hpm_src_files :: [FilePath],
       -- ^ extra source files (e.g. from #includes).  The lexer collects
       -- these from '# <file> <line>' pragmas, which the C preprocessor
       -- leaves behind.  These files and their timestamps are stored in
       -- the .hi file, so that we can force recompilation if any of
       -- them change (#3589)
    hpm_annotations :: ApiAnns
  }

-- | The result of successful parsing.
data ParsedModule =
  ParsedModule { pm_mod_summary   :: ModSummary
               , pm_parsed_source :: ParsedSource
               , pm_extra_src_files :: [FilePath]
               , pm_annotations :: ApiAnns }
```

This diff depends on D426

Test Plan: sh ./validate

Reviewers: austin, simonpj, Mikolaj

Reviewed By: simonpj, Mikolaj

Subscribers: Mikolaj, goldfire, thomie, carter

Differential Revision: https://phabricator.haskell.org/D438

GHC Trac Issues: #9628

Summary:
AST changes to prepare for API annotations

Add locations to parts of the AST so that API annotations can
then be added.

The outline of the whole process is captured here
https://ghc.haskell.org/trac/ghc/wiki/GhcAstAnnotations

This change updates the haddock submodule.

Test Plan: sh ./validate

Reviewers: austin, simonpj, Mikolaj

Reviewed By: simonpj, Mikolaj

Subscribers: thomie, goldfire, carter

Differential Revision: https://phabricator.haskell.org/D426

GHC Trac Issues: #9628

Summary: It seems to be dead anyway. Also update Haddock submodule.

Test Plan: validate

Reviewers: austin

Reviewed By: austin

Subscribers: thomie, goldfire, carter, simonmar

Differential Revision: https://phabricator.haskell.org/D357

Summary:
This warning (enabled by default) reports places where a context
implicitly binds a type variable, for example

type T a = {-forall m.-} Monad m => a -> m a

Also update Haddock submodule.

Test Plan: validate

Reviewers: hvr, goldfire, simonpj, austin

Reviewed By: austin

Subscribers: simonmar, ezyang, carter

Differential Revision: https://phabricator.haskell.org/D211

GHC Trac Issues: #4426

This exposes *only* the type-classes w/o any of their methods.
This is the very first step for implementing BPP (see #9586), which
already requires breaking up several import-cycles leading back to `Prelude`.

Ideally, importing `Prelude` should be avoided in most `base` modules,
as `Prelude` does not define any entities, but rather re-exports
existing ones.

Test Plan: validate passes

Reviewers: ekmett, austin

Reviewed By: ekmett, austin

Subscribers: simonmar, ezyang, carter

Differential Revision: https://phabricator.haskell.org/D209

GHC Trac Issues: #9586

Summary:
This is a first step toward allowing generic traversals of the AST without 'landmines', by removing the `panic`s located throughout `placeHolderType`, `placeHolderKind` & co.

See more on the discussion at https://www.mail-archive.com/ghc-devs@haskell.org/msg05564.html

(This also makes a corresponding update to the `haddock` submodule.)

Test Plan: `sh validate` and new tests pass.

Reviewers: austin, simonpj, goldfire

Reviewed By: austin, simonpj, goldfire

Subscribers: edsko, Fuuzetsu, thomasw, holzensp, goldfire, simonmar, relrod, ezyang, carter

Projects: #ghc

Differential Revision: https://phabricator.haskell.org/D157

This commit also removes 'KindCheckingStrategy' and related gubbins,
instead including the notion of a CUSK into HsDecls.

This also removes the short-lived NO_OVERLAP pragama, and renames
OVERLAP to OVERLAPS.

An instance may be annotated with one of 4 pragams, to control its
interaction with other overlapping instances:

  * OVERLAPPABLE:
    this instance is ignored if a more specific candidate exists

  * OVERLAPPING:
    this instance is preferred over more general candidates

  * OVERLAPS:
    both OVERLAPPING and OVERLAPPABLE (i.e., the previous GHC behavior).
    When compiling with -XOverlappingInstances, all instance are OVERLAPS.

  * INCOHERENT:
    same as before (see manual for details).
    When compiling with -XIncoherentInstances, all instances are INCOHERENT.

In looking at Trac #9063 I decided to re-design the default
instances for associated type synonyms.  Previously it was all
jolly complicated, to support generality that no one wanted, and
was arguably undesirable.

Specifically

* The default instance for an associated type can have only
  type variables on the LHS.  (Not type patterns.)

* There can be at most one default instances declaration for
  each associated type.

To achieve this I had to do a surprisingly large amount of refactoring
of HsSyn, specifically to parameterise HsDecls.TyFamEqn over the type
of the LHS patterns.

That change in HsDecls has a (trivial) knock-on effect in Haddock, so
this commit does a submodule update too.

The net result is good though.  The code is simpler; the language
specification is simpler.  Happy days.

Trac #9263 and #9264 are thereby fixed as well.

Programmers may provide a pragma immediately after the `instance` keyword
to control the overlap/incoherence behavior for individual instances.
For example:

    instance {-# OVERLAP #-} C a where ...

I chose this notation, rather than the other two outlined in the ticket
for these reasons:

   1. Having the pragma after the type looks odd, I think.
   2. Having the pragma after there `where` does not work for
       stand-alone derived instances

I have implemented 3 pragams:

   1. NO_OVERLAP
   2. OVERLAP
   3. INCOHERENT

These correspond directly to the internal modes currently supported by
GHC.  If a pragma is specified, it will be used no matter what flags are
turned on.   For example, putting `NO_OVERLAP` on an instance will mark
it as non-overlapping, even if `OVERLAPPIN_INSTANCES` is turned on for the
module.

When printing Haskell source, and UnicodeSyntax is enabled, use the
unicode sytax characters (#8959).

The handling of typed and untyped brackets was extremely convoluted,
partly because of the evolutionary history.  I've tidied it all up.

See Note [How brackets and nested splices are handled] in TcSplice
for the full story

Main changes:

 * Untyped brackets: after the renamer, HsRnBracketOut carries
   PendingRnSplices for splices in untyped brackets.  In the
   typechecker, these pending splices are typechecked quite
   straigtforwardly, with no ps_var nonsense.

 * Typed brackets: after the renamer typed brackest still look
   like HsBracket. The type checker does the ps_var thing.

 * In TcRnTypes.ThStage, the Brack constructor, we distinguish
   the renaming from typehecking pending-stuff.  Much more
   perspicuous!

 * The "typed" flag is in HsSpliceE, not in HsSplice, because
   only expressions can be typed.  Patterns, types, declarations
   cannot.

There is further improvement to be done to make the handling of
declaration splices more uniform.

Since declaration splices are now untyped, they can be used anywhere a
declaration is valid, including in declaration brackets.

This fixes bugs #8185, #8234, and #8246. The new syntax is explained
in the comments to #8185, appears in the "Roles" subsection of the
manual, and on the [wiki:Roles] wiki page.

This change also removes the ability for a role annotation on type
synonyms, as noted in #8234.

As documented in the users' guide, you can now write

type family Foo a where ..

in a hs-boot file to declare an abstract closed type family.

Now, all open type families have result kinds that default to
*. Top-level type families have all arguments default to *,
and associated type families have all arguments that are not
mentioned in the class header default to *. Closed type
families perform kind inference, but generalize only over
those kind variables that are parametric in their use.

This is all a little fiddly and specific, but it seems to follow
use cases. This commit also includes a large
Note [Kind-checking strategies] in TcHsType that helps keep all
of this straight.

This commit changes the syntax and story around overlapping type
family instances. Before, we had "unbranched" instances and
"branched" instances. Now, we have closed type families and
open ones.

The behavior of open families is completely unchanged. In particular,
coincident overlap of open type family instances still works, despite
emails to the contrary.

A closed type family is declared like this:
> type family F a where
>   F Int = Bool
>   F a   = Char
The equations are tried in order, from top to bottom, subject to
certain constraints, as described in the user manual. It is not
allowed to declare an instance of a closed family.

AnnProvenance now has Functor, Foldable, Traversable instances.