How to set C-c C-k to check kind?

[HaskellZhangSong]

I found there is a pieces of code to check the kind of a term in inf-haskell.el
(defun inferior-haskell-kind (type)
"Query the haskell process for the kind of the given expression."
(interactive

I bind the key as following
(define-key map (kbd "C-c C-k") 'inferior-haskell-kind)
and restart emacs but still cannot use it.
I am a new user of haskell-mode. Could any explain to me a little bit.

[sergv]

Assuming that map refers to haskell-mode-related keymap like haskell-mode-map, one reason why inferior-haskell-kind could not be used is that C-c C-k binding is shadowed by the same keybinding in some minor mode. To test this just temporarily rebind inferior-haskell-kind to some bizzare combination that is unlikely to be bound and try it out, or type in, while you're in haskell-mode buffer and there's some type name at point, M-x eval-expression (inferior-haskell-kind (haskell-ident-at-point)).

If it turns out to be true, i.e. inferior-haskell-kind function is working for you, you can start hunting for offending mode that binds C-c C-k as follows: in buffer with active haskell mode issue M-x eval-expression (minor-mode-key-binding (kbd "C-c C-k")) which should return list of minor modes that bind C-c C-k. If returned list is empty you still can try, in the same buffer with active haskell mode, C-h m to get list of all active modes in the buffer and search for C-c C-k there. When mode name with C-c C-k keybinding is found you can disable it like this (add-hook <MODE>-hook (lambda () (define-key <MODE>-map (kbd "C-c C-k") nil))), of course using actual mode name instead of <MODE>.

[ardumont]

To sum up, in this case, I believe it's:

(add-hook 'inferior-haskell-mode-hook
          (lambda ()
            (define-key inferior-haskell-mode-map (kbd "C-c C-k") nil)
            (define-key inferior-haskell-mode-map (kbd "C-c C-k") 'inferior-haskell-kind)))

[gracjan]

@HaskellZhangSong: did you manage to solve this problem? If not do not hessitate to write a note here and we will reopen this issue again.