sebf at informatik.uni-kiel.de
Wed Feb 3 09:07:52 EST 2010
On Feb 3, 2010, at 1:48 AM, Max Bolingbroke wrote:
> 2010/2/2 Twan van Laarhoven <twanvl at gmail.com>:
>> class Quoted a where
>> parseQuote :: String -> a
>> -- for performance reasons:
>> parseQuote' :: Ghc.PackedString -> a
> Great idea!
> Thinking about it, you can use type classes to dispose of the
> QuasiQuote record entirely. Instead, have:
> class MyLang a where
> myLang :: String -> Q a
> instance MyLang Exp where
> myLang = myLangSyntaxToGHCExprForSplice . myLangExpParser
> ... etc, MyLang instances for Pat and Type too ...
With a class-based approach only one parser that creates values of the
same type could be used in a program. It would not be possible to
embed multiple languages that create TH.Exp to be spliced into a
single program. With the current syntax, I can write [$myLang| ... |]
and [$yourLang| ... |] in the same program and use different parsers
although both create Exp values.
> And then write:
> $(myLang [|...|])
This is more verbose than the proposed [myLang| ... |]. There seem to
be different goals in the different proposals in this thread: 1.
Simplify the syntax for quasi quoting (remove $, use different
brackets), 2. make it more generally applicable (allow declarations
and/or types to be quasi quoted), and 3. simplify and generalise its
implementation (invent a single mechanism that unifies quasi quoting
and TH splicing).
Underestimating the novelty of the future is a time-honored tradition.
More information about the Glasgow-haskell-users