[Haskell-cafe] Some thoughts on Type-Directed Name Resolution

[Gábor Lehel]

On Mon, Feb 20, 2012 at 4:41 AM, AntC <anthony_clayden at clear.net.nz> wrote:
> Folks, I've put my 'Record in Haskell' proposal on the wiki
> http://hackage.haskell.org/trac/ghc/wiki/Records  as suggestion 5 Declared
> Overloaded Record Fields.
>
> Thanks to the voiciferousness on this thread, dot notation is completely
> optional.
>
> Feedback welcome.

Thanks for writing it up, I admit I was having trouble following
across the various email threads.

Surprisingly, your ideas are very similar to my own. I'm not sure if
this is a good thing or a bad sign, but naturally I'm in favor.

I was wondering whether it wouldn't make sense to have some syntax
within the record itself, instead of at the top level, to declare,
"I'm definitely declaring a new record field", versus "I'm definitely
re-using an existing record field", versus "If this record field
already exists I'm re-using it, otherwise I'm declaring it". (It
doesn't necessarily make sense to have all three - the second one
might be cumbersome, or the third one might be error-prone - but they
seem like the options.)

The existing, unadorned record syntax would mean "I'm definitely
declaring a new record field", because that's what it already means.
Simply leaving off the type annotation to indicate otherwise sadly
wouldn't work because, as you mention, that means that it's the same
type as the next field.

So something like:

data Rec1 = Rec1 { field1 :: Int, field2 :: Char } -- declare field1
:: Int and field2 :: Char fields

data Rec2 = Rec2 { import field1, field3 :: String } -- reuse field1
:: Int, declare field3 :: String

data Rec3 = Rec3 { field3 :: String } -- declare field3 :: String, but
error: already declared

Hopefully someone can think of better syntax than my "import field1" above.

Regarding the polymorphic update / higher-rank fields issues, I'm not
competent to address them in earnest, but: isn't this primarily an
ImpredicativeTypes issue? If GHC had full support for
ImpredicativeTypes (whatever that means), would it work?

~g