Personal tools

GHC/Stand-alone deriving declarations

From HaskellWiki

< GHC(Difference between revisions)
Jump to: navigation, search
(+cat)
m (Standalone deriving: fix broken link)
(6 intermediate revisions by 4 users not shown)
Line 1: Line 1:
 
[[Category:GHC|Stand-alone deriving declarations]]
 
[[Category:GHC|Stand-alone deriving declarations]]
Bjorn Bringert has recently implemented "stand-alone deriving" declarations, documented briefly here [http://www.haskell.org/ghc/dist/current/docs/users_guide/type-extensions.html#stand-alone-deriving]. There are a few loose ends which I summarise here:
 
 
== Syntax ==
 
   
The current syntax is
+
''This page is from an early point in the life of the stand-alone deriving mechanism. Please see the linked documentation for an up-to-date account of the present situation.''
<haskell>
 
deriving Show for T
 
</haskell>
 
There seems to be a consensus that this would be better:
 
<haskell>
 
derive instance Show T
 
</haskell>
 
so that it looks more like a regular instance declaration. Here <hask>derive</hask> is not a new keyword; it's a "special-id", distinguished by the following <hask>instance</hask> keyword. That means that <hask>derive</hask> can still be used freely as a regular varid.
 
(Perhaps <hask>derived</hask> would sound more declarative than <hask>derive</hask>.)
 
   
== Context on the declaration ==
+
== Standalone deriving ==
   
Because it looks like a regular instance declaration, it would arguably be reasonable to require the programmer to supply the context. It seems odd to say:
+
GHC supports so-called "stand-alone deriving" declarations, which are described in the [http://www.haskell.org/ghc/docs/latest/html/users_guide/deriving.html#stand-alone-deriving user manual section].
  +
  +
This page mentions points that may not be immediately obvious from the manual.
  +
  +
== Deriving data types with non-standard contexts ==
  +
  +
In Haskell 98, and GHC, you can't say this
 
<haskell>
 
<haskell>
derive instance Show (T a)
+
data T m = MkT (m Int) deriving Eq
 
</haskell>
 
</haskell>
and perhaps cleaner to say
+
because the instance declaration would have a non-standard context. It would have to look like this:
 
<haskell>
 
<haskell>
derive instance Show a => Show (T a)
+
instance Eq (m Int) => Eq (T m) where ...
 
</haskell>
 
</haskell>
(At the moment, the compiler figures out the appropriate context, but at some point that automation may run out of steam.)
+
Of course, you can write the instance manually, but then you have to write
  +
all that tiresome code for equality. Standalone deriving lets you supply the context yourself, but have GHC write the code:
  +
<haskell>
  +
data T m = MkT (m Int)
   
== Alternative Syntax ==
+
deriving instance Eq (m Int) => Eq (T m)
Alternatively the syntax could be:
 
<haskell>
 
instance Show a => Show (T a) derived
 
 
</haskell>
 
</haskell>
which has the advantage that the most important thing, namely the fact we are introducing an instance, is first, and the particular way we are introducing it (by deriving instead of explicitly by a <hask>where</hask> clause or using class defaults only) is second. (Another advantage is that the reserved id <hask>derived</hask> is now inside the <hask>instance</hask> construct not before it which is in line with the way other reserved ids are used in Haskell 98.)
+
Of course, you'll need to add the flags <hask>-XFlexibleContexts</hask> and <hask>-XUndecideableInstances</hask> to allow this instance declaration, but that's fair enough.
   
There was a suggestion to allow just
+
The same applies to data type declarations involving type functions.
<haskell>
+
instance Show a => Show (T a)
+
== Variations (not implemented) ==
</haskell>
+
where the fact that the compiler is automatically deriving the content of the instance declaration is implicit in the absence of a <hask>where</hask> clause. The rationale was that the
+
This section collects some un-implemented ideas.
<hask>where</hask> clause is already optional - if a class does not have a body then we can instead say that all instances of it are derived. However this is to misunderstand the current meaning of a missing <hask>where</hask> clause. In Haskell'98 it means that all the methods of the class are ''undefined'' for this instance (except for defaults). That is the opposite meaning to this proposal, which would say that all methods are fully defined.
 
   
== Interaction with "newtype-deriving" ==
+
=== Interaction with "newtype-deriving" ===
   
GHC's "newtype deriving mechanism" (see [http://www.haskell.org/ghc/dist/current/docs/users_guide/type-extensions.html#newtype-deriving]) should obviously work in a standalone deriving setting too. But perhaps it can be generalised a little. Currently you can only say
+
GHC's "newtype deriving mechanism" (see [http://www.haskell.org/ghc/dist/current/docs/users_guide/deriving.html#newtype-deriving]) should obviously work in a standalone deriving setting too. But perhaps it can be generalised a little. Currently you can only say
 
<haskell>
 
<haskell>
deriving C a for Foo
+
deriving instance C a Foo
 
</haskell>
 
</haskell>
 
(where Foo is the newtype), and get an instance for <hask>(C a Foo)</hask>. But what if you want and instance for <hask>C Foo a</hask>, where the new type is not the last parameter. You can't do that at the moment. However, even with the new instance-like syntax, it's not clear to me how to signal the type to be derived. Consider
 
(where Foo is the newtype), and get an instance for <hask>(C a Foo)</hask>. But what if you want and instance for <hask>C Foo a</hask>, where the new type is not the last parameter. You can't do that at the moment. However, even with the new instance-like syntax, it's not clear to me how to signal the type to be derived. Consider
Line 38: Line 35:
 
newtype Foo = F Int
 
newtype Foo = F Int
 
newtype Bar = B Bool
 
newtype Bar = B Bool
derive instance C Foo Bar
+
deriving instance C Foo Bar
 
</haskell>
 
</haskell>
 
Which of these thee instances do we want?
 
Which of these thee instances do we want?
Line 56: Line 53:
 
newtype IdSet = IdSet IntSet
 
newtype IdSet = IdSet IntSet
 
 
derive instance SetLike IntSet Int => SetLike IdSet Id
+
deriving instance SetLike IntSet Int => SetLike IdSet Id
 
</haskell>
 
</haskell>
   
== Duplicate instances ==
+
=== Duplicate instances ===
   
 
Suppose two modules, M1 and M2 both contain an identical standalone deriving declaration
 
Suppose two modules, M1 and M2 both contain an identical standalone deriving declaration
 
<haskell>
 
<haskell>
derive Show T
+
deriving Show T
 
</haskell>
 
</haskell>
 
Then, can you import M1 and M2 into another module X and use show on values of type T, or will you get an overlapping instance error? Since both instances are derived in the very same way, their code must be identical, so arguably we can choose either. (There is some duplicated code of course.)
 
Then, can you import M1 and M2 into another module X and use show on values of type T, or will you get an overlapping instance error? Since both instances are derived in the very same way, their code must be identical, so arguably we can choose either. (There is some duplicated code of course.)
Line 84: Line 81:
 
-- (they need to be derived to ensure the output can
 
-- (they need to be derived to ensure the output can
 
-- be parsed in our non-Haskell code).
 
-- be parsed in our non-Haskell code).
derive instance Show A
+
deriving instance Show A
derive instance Show B
+
deriving instance Show B
 
</haskell>
 
</haskell>
 
The writer of MC already knows that MA and MB defined instances of Show for A and B. He just wants to ensure that nobody changes either module to use a non-derived instance; if someone does try to use a non-derived instance:
 
The writer of MC already knows that MA and MB defined instances of Show for A and B. He just wants to ensure that nobody changes either module to use a non-derived instance; if someone does try to use a non-derived instance:

Revision as of 16:10, 28 October 2009


This page is from an early point in the life of the stand-alone deriving mechanism. Please see the linked documentation for an up-to-date account of the present situation.

Contents

1 Standalone deriving

GHC supports so-called "stand-alone deriving" declarations, which are described in the user manual section.

This page mentions points that may not be immediately obvious from the manual.

2 Deriving data types with non-standard contexts

In Haskell 98, and GHC, you can't say this

  data T m = MkT (m Int) deriving Eq

because the instance declaration would have a non-standard context. It would have to look like this:

  instance Eq (m Int) => Eq (T m) where ...

Of course, you can write the instance manually, but then you have to write all that tiresome code for equality. Standalone deriving lets you supply the context yourself, but have GHC write the code:

  data T m = MkT (m Int)
 
  deriving instance Eq (m Int) => Eq (T m)
Of course, you'll need to add the flags
-XFlexibleContexts
and
-XUndecideableInstances
to allow this instance declaration, but that's fair enough.

The same applies to data type declarations involving type functions.

3 Variations (not implemented)

This section collects some un-implemented ideas.

3.1 Interaction with "newtype-deriving"

GHC's "newtype deriving mechanism" (see [1]) should obviously work in a standalone deriving setting too. But perhaps it can be generalised a little. Currently you can only say

  deriving instance C a Foo
(where Foo is the newtype), and get an instance for
(C a Foo)
. But what if you want and instance for
C Foo a
, where the new type is not the last parameter. You can't do that at the moment. However, even with the new instance-like syntax, it's not clear to me how to signal the type to be derived. Consider
  newtype Foo = F Int
  newtype Bar = B Bool
  deriving instance C Foo Bar

Which of these thee instances do we want?

  instance C Foo Bool => C Foo Bar
  instance C Int Bar  => C Foo Bar
  instance C Int Bool => C Foo Bar

The obvious way to signal this is to give the instance context (just as above). This is perhaps another reason for having an explicit instance context in a standalone deriving declaration.

Incidentally, notice that the third of the alternatives in the previous bullet unwraps two newtypes simultaneously. John Meacham suggested this example:

  class SetLike m k  where 
  instance SetLike IntSet Int where
 
  newtype Id = Id Int
  newtype IdSet = IdSet IntSet
 
  deriving instance SetLike IntSet Int => SetLike IdSet Id

3.2 Duplicate instances

Suppose two modules, M1 and M2 both contain an identical standalone deriving declaration

  deriving Show T

Then, can you import M1 and M2 into another module X and use show on values of type T, or will you get an overlapping instance error? Since both instances are derived in the very same way, their code must be identical, so arguably we can choose either. (There is some duplicated code of course.)

This situation is expected to be common, as the main use of the standalone feature is to obtain derived instances that were omitted when the data type was defined.

But, that means whether or not an instance was derived is now part of the module's. Programs would be able to use this (mis)feature to perform a compile-time check and execute code differently depending on whether any given instance is derived or hand-coded:

   module MA(A) where
   data A = A deriving Show
 
   module MB(B) where
   data B = B deriving Show
 
   module MC where
   import MA
   import MB
 
   -- verify that the A and B Show instances were derived
   -- (they need to be derived to ensure the output can
   -- be parsed in our non-Haskell code).
   deriving instance Show A 
   deriving instance Show B

The writer of MC already knows that MA and MB defined instances of Show for A and B. He just wants to ensure that nobody changes either module to use a non-derived instance; if someone does try to use a non-derived instance:

   module MA(A) where
   data A = A
   instance Show A where
       show _ = "a"

then they will get an overlapping instance error in MC.

The result is that programs would be able to require, for any Class, not just that an instance of the class was defined for a type, but that a /derived/ instance was defined. Is this good?