On Sat, Feb 2, 2013 at 1:08 PM, Peter Jones
I've been working on a few customizations to xmonad but as I play with Haskell I keep seeing a pattern of duplication in my code that I don't know how to resolve. Here's an example:
-- | Enables 'focusFollowsMouse' for tiled windows only. For this to -- work you need to turn off 'focusFollowsMouse' in your configuration -- and then add this function to your 'handleEventHook'. focusFollowsTiledOnly :: Event -> X All focusFollowsTiledOnly e@(CrossingEvent {ev_window = w, ev_event_type = t}) | isNormalEnter = whenX bothTiled (focus w) >> continueHooks where isNormalEnter = t == enterNotify && ev_mode e == notifyNormal bothTiled = (&&) <$> notFloating w <*> currentIsTiled currentIsTiled = currentWindow >>= maybe (return True) notFloating currentWindow = gets $ W.peek . windowset notFloating w' = gets $ not . M.member w' . W.floating . windowset continueHooks = return . mempty $ True focusFollowsTiledOnly _ = return . mempty $ True
The last two lines demonstrate the pattern I've been seeing. The only way I know how to remove this duplication is to move it out into a top-level function. Is that correct?
Hi Peter, Maybe you'll like this version below better. You can use mempty instead of 'return . mempty $ True'. The latter is a bit confusing since the True doesn't end up in value, since there's a "instance Monoid b => Monoid (a -> b)" whose mempty ignores the argument `a'. import qualified XMonad.StackSet as W import XMonad import Data.Monoid import Data.Map as M import Control.Applicative focusFollowsTiledOnly :: Event -> X All focusFollowsTiledOnly e | CrossingEvent {ev_window = w, ev_event_type = t} <- e, let isNormalEnter = t == enterNotify && ev_mode e == notifyNormal bothTiled = notFloating w <&&> currentIsTiled currentIsTiled = currentWindow >>= maybe (return True) notFloating currentWindow = gets $ W.peek . windowset notFloating w' = gets $ not . M.member w' . W.floating . windowset, isNormalEnter = whenX bothTiled (focus w) >> mempty | otherwise = mempty -- Adam