| modifyLayout | 
| :: LayoutClass l a |  |  | => m a | the layout modifier
 |  | -> Workspace WorkspaceId (l a) a | current workspace
 |  | -> Rectangle | screen rectangle
 |  | -> X ([(a, Rectangle)], Maybe (l a)) |  |  modifyLayout allows you to intercept a call to runLayout
   before it is called on the underlying layout, in order to
   perform some effect in the X monad, and/or modify some of
   the parameters before passing them on to the runLayout
   method of the underlying layout.
 The default implementation of modifyLayout simply calls
   runLayout on the underlying layout.
  |  
  | 
 | 
| handleMess :: m a -> SomeMessage -> X (Maybe (m a)) | 
handleMess allows you to spy on messages to the underlying
   layout, in order to have an effect in the X monad, or alter
   the layout modifier state in some way (by returning Just
   nm, where nm is a new modifier).  In all cases, the
   underlying layout will also receive the message as usual,
   after the message has been processed by handleMess.
 If you wish to possibly modify a message before it reaches
   the underlying layout, you should use
   handleMessOrMaybeModifyIt instead.  If you do not need to
   modify messages or have access to the X monad, you should use
   pureMess instead.
 The default implementation of handleMess calls unhook
   when receiving a Hide or ReleaseResources method (after
   which it returns Nothing), and otherwise passes the message
   on to pureMess.
  | 
 | 
| handleMessOrMaybeModifyIt :: m a -> SomeMessage -> X (Maybe (Either (m a) SomeMessage)) | 
handleMessOrMaybeModifyIt allows you to intercept messages
   sent to the underlying layout, in order to have an effect in
   the X monad, alter the layout modifier state, or produce a
   modified message to be passed on to the underlying layout.
 The default implementation of handleMessOrMaybeModifyIt
   simply passes on the message to handleMess.
  | 
 | 
| pureMess :: m a -> SomeMessage -> Maybe (m a) | 
pureMess allows you to spy on messages sent to the
   underlying layout, in order to possibly change the layout
   modifier state.
 The default implementation of pureMess ignores messages
   sent to it, and returns Nothing (causing the layout
   modifier to remain unchanged).
  | 
 | 
| redoLayout | 
| :: m a | the layout modifier
 |  | -> Rectangle | screen rectangle
 |  | -> Stack a | current window stack
 |  | -> [(a, Rectangle)] | (window,rectangle) pairs returned
 by the underlying layout
 |  | -> X ([(a, Rectangle)], Maybe (m a)) |  |  redoLayout allows you to intercept a call to runLayout on
   workspaces with at least one window, after it is called on
   the underlying layout, in order to perform some effect in the
   X monad, possibly return a new layout modifier, and/or
   modify the results of runLayout before returning them.
 If you don't need access to the X monad, use pureModifier
   instead.  Also, if the behavior you need can be cleanly
   separated into an effect in the X monad, followed by a pure
   transformation of the results of runLayout, you should
   consider implementing hook and pureModifier instead of
   redoLayout.
 If you also need to perform some action when runLayout is
   called on an empty workspace, see emptyLayoutMod.
 The default implementation of redoLayout calls hook and
   then pureModifier.
  |  
  | 
 | 
| pureModifier | 
| :: m a | the layout modifier
 |  | -> Rectangle | screen rectangle
 |  | -> Stack a | current window stack
 |  | -> [(a, Rectangle)] | (window, rectangle) pairs returned
 by the underlying layout
 |  | -> ([(a, Rectangle)], Maybe (m a)) |  |  pureModifier allows you to intercept a call to runLayout
   after it is called on the underlying layout, in order to
   modify the list of window/rectangle pairings it has returned,
   and/or return a new layout modifier.
 The default implementation of pureModifier returns the
   window rectangles unmodified.
  |  
  | 
 | 
| emptyLayoutMod :: m a -> Rectangle -> [(a, Rectangle)] -> X ([(a, Rectangle)], Maybe (m a)) | 
emptyLayoutMod allows you to intercept a call to
   runLayout on an empty workspace, after it is called on
   the underlying layout, in order to perform some effect in the
   X monad, possibly return a new layout modifier, and/or
   modify the results of runLayout before returning them.
 If you don't need access to the X monad, then tough luck.
   There isn't a pure version of emptyLayoutMod.
 The default implementation of emptyLayoutMod ignores its
   arguments and returns an empty list of window/rectangle
   pairings.
 NOTE: emptyLayoutMod will likely be combined with
   redoLayout soon!
  | 
 | 
| hook :: m a -> X () | 
hook is called by the default implementation of
   redoLayout, and as such represents an X action which is to
   be run each time runLayout is called on the underlying
   layout, after runLayout has completed.  Of course, if you
   override redoLayout, then hook will not be called unless
   you explicitly call it.
 The default implementation of hook is return () (i.e., it
   has no effect).
  | 
 | 
| unhook :: m a -> X () | 
unhook is called by the default implementation of
   handleMess upon receiving a Hide or a ReleaseResources
   message.
 The default implementation, of course, does nothing.
  | 
 | 
| modifierDescription :: m a -> String | 
| modifierDescription is used to give a String description to
   this layout modifier.  It is the empty string by default; you
   should only override this if it is important that the
   presence of the layout modifier be displayed in text
   representations of the layout (for example, in the status bar
   of a XMonad.Hooks.DynamicLog user).
 | 
 | 
| modifyDescription :: LayoutClass l a => m a -> l a -> String | 
| modifyDescription gives a String description for the entire
   layout (modifier + underlying layout).  By default, it is
   derived from the concatenation of the modifierDescription
   with the description of the underlying layout, with a
   "smart space" in between (the space is not included if the
   modifierDescription is empty).
 |