2008/11/8 Johannes Waldmann
Looking at this funny new feature http://haskell.org/ghc/docs/6.10.1/html/users_guide/syntax-extns.html#genera... I have just one question - why doesn't this work with the do-notation?
I avoid list comprehensions because I feel that "return" belongs at the end, not in front.
If I recall correctly, "putting the SQL-select where it belongs" is a slogan used by Hijlsberg to justify the LINQ syntax for C#, and of course he is right.
Now ghc copies LINQ (syntactically), but stops halfway?
Hi Johannes, There is no technical reason the syntax could not be extended to do notation - see the discussion by Michael Adams on the http://haskell.org/haskellwiki/Simonpj/Talk:ListComp page for a taste of how that would work (note that his translation is however not totally correct, IIRC). The only reason that I didn't actually implement this feature is that neither I nor SPJ could think of a use case for this syntax outside the list monad. I don't think we considered the possibility you might use do notation for the list monad, as it's not an idiom that seems to occur often. If you can come up with such a use case I could probably find the time to implement the extra translation steps! On reflection, it does seem a bit like an annoying irregularity to the implementation. Cheers, Max