Fed up with pressing TABs hundreds times a day in Haskell programming with the offside rule, I wrote a small elisp:

    http://bitbucket.org/camlspotter/offside-trap/src

This elisp understand nothing about Haskell. It just traps each indentation change of one line, and extends it to its “sticky” lines. Sticky lines are:

- It is below of the cursor line or another sticky line 
- Its indentation is as deep as or deeper than the cursor line. 

The rule is simple but works nicely:

-------------------------------------------------------------------------
  hoge = case hoge of
                   []Foo -> hogehoge             <= The cursor ([]) is here             
                     Bar -> let x = hogehoge  in <= Sticky
                            poo poo              <= Sticky
                     _ -> error "bang"           <= Sticky
  huga = 2  <= Not sticky
-------------------------------------------------------------------------
 
Pressing ONLY ONE TAB, I got the following. Puff!:

-------------------------------------------------------------------------
  hoge = case hoge of
         []Foo -> hogehoge
           Bar -> let x = hogehoge  in <= Sticky
                  poo poo              <= Sticky
           _ -> error "bang"           <= Sticky
  huga = 2  <= Not sticky
-------------------------------------------------------------------------

Of course this “offside trap” only works against codes with correct indentations. But it definitely improved my Haskell coding experience.
 
Jun