Re: [Haskell-cafe] message passing style like in Haskell?

After some fiddling with this style, here is what I came up with for the 8 queens problem in the 99 problem set. It's quite entertaining ... ( note: it's brute force and requires a combination library ) queens2 n = n.permutations.filter all_satisfied where all_satisfied queens = queens.diff_col && queens.diff_diag diff_col queens = queens.unique.is queens diff_diag queens = n .combinations 2 .map (map (subtract 1)) .map (id &&& flip cherry_pick queens) .any same_dist.not where same_dist (row_pair, col_pair) = row_pair.foldl1 (-).abs == col_pair.foldl1 (-).abs -- generic helper cherry_pick ids xs = ids.map (xs !!) is a b = a == b unique xs = nub xs Guess this can conclude this experiment :) jinjing On Sun, Jun 22, 2008 at 1:10 AM, Ian Lynagh wrote:

On Fri, Jun 20, 2008 at 07:57:58AM +0200, Ketil Malde wrote:

...

"Albert Y. C. Lai" writes:

...

...

While we are kind of on this topic, what makes the characters ħ þ prefix operator by default, while º and most other odd ones infix?

...

alphanumeric vs non-alphanumeric

Testing this, I find that isAlpha is True also for 'º', but as the OP claims, Haskell will use it as a(n infix) symbol.

This is a bug in GHC. The characters <= '\255' were done specially, but incorrectly for many of those >= '\128'. I'll fix it, probably by just removing the specialisation for them.

Thanks Ian

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe