2009/7/19 Max Bolingbroke
Dear Cafe,
For fun, I spent a few hours yesterday implement support for this syntax in GHC, originally propsed by Koen Claessen:
[k, "=", v, " " | (k, v) <- [("foo", "1"), ("bar", "2")] ["foo", "=", "1", " ", "bar", "=", "2", " "]
Given that this can easily be simulated via:
[ x | (k, v) <- [("foo", "1"), ("bar", "2")], x <- [k, "=", v, " "]] ["foo","=","1"," ","bar","=","2"," "]
I believe that the added syntax (which every complete tool operating on Haskell code would have to support) is not worth its price.
This is a generalisation of list comprehensions that allows several items to be concatenated onto the result list at once, by having several comma-separated items before the pipe. One situation where I have found this to be useful is when you are writing Haskell programs that call lots of external programs and you need to set the flags based on some tests, like so:
rawSystem "myProgram" $ ["foo" | fooing_enabled] ++ ["bar1", "bar2" | baring_enabled]
I have submitted a ticket to GHC HQ with the patch (http://hackage.haskell.org/trac/ghc/ticket/3380#comment:5), but as it is just a small convenience it most likely won't make it in unless there is more demand for the feature. So, now is the time to speak up in favour of (or indeed, against) the added syntax!
All the best, Max
P.S. I also implemented tuple sections (http://hackage.haskell.org/trac/ghc/ticket/3377#comment:3) which are a lot more useful:
They are indeed, and here I would be inclined to consider the added cost worth it. -- Push the envelope. Watch it bend.