Re: Data instances in ghc 6.2.1?

Claus Reinke wrote:

> For context: I wanted to show people in the haskell tools thread
>
> - how easy it is these days to write a Haskell program transformation
>    for instrumenting source code with asserts, e.g. to get location info
>    for heads of empty lists :-)
>  
Yes, it is easy!

> - why these lovely new tools won't be used because even without 
>    transformations they don't reproduce the source :-(
Yes, but then, we should really work on reproducability of code!

After all, you have gathered relevant experiences in the HaRe project.
So the requirements are clear.

How much work is it to incorporate comments and layout in a systematic
manner? In GDK, we don't have any problems doing this for C-based
combinator parsing. We use a simple wrapper pattern. Whenever, we fetch
a token from the lexer, then we build a sandwich token that also comprises
the comment/layout (identation) from before the token. Ok, since we go
for an abstract syntax, we also need to catch keywords and all this.
Also, we must switch off all normalisation in the parse tree construction.
(Are there any normalisations going on in Language/Haskell? Probably, not
In, gcc, for example, there is, which is problem.)

> But as they can't use this in the current release anyway, I'll leave
> it at that.
>  

I have committed the extended Language/Haskell/Syntax.hs.
Why not demonstrate, what you said, anyway, using CVS built from head?

I would love to help with both:
- this showcase
- and improving Language/Haskell

Cheers,
Ralf