`evil-repeat-type' returns nil making `evil-repeat-record-fail' about evil HOT 4 CLOSED

Original comment by Michael Markert (Bitbucket: cofi, GitHub: cofi):

@input-methods\
Didn't notice. But it broke: Typing ue with german-postfix got me üü (double result). But b464897 seems to fix it.

@legacy\
Well then: Great that it broke :)

from evil.

Original comment by Frank Fischer (Bitbucket: lyro, GitHub: lyro):

Thanks. The function `evil-repeat-record-command' should never be called with a non-function repeat-type ... so my fix is to remove this function completely (it was legacy anyway) ;)

But I realized that repeation with input-methods did not work (anymore?). It should work in b464897 again.

from evil.

Original comment by Michael Markert (Bitbucket: cofi, GitHub: cofi):

What I meant wrt to input method: Typing any char triggered an error. Should've made this clear :(

The change seems to fix it, I don't run into the issue anymore.

For completeness sake: I changed `evil-repeat-record-command' this way:

 (defun evil-repeat-record-command (&optional repeat-type)
   "Calls the post-repeat-information of the current command."
   ;; finish repeation of current command
   (let ((repeat-type (or repeat-type
                          (evil-repeat-type this-command t))))
-    (funcall repeat-type 'post)))
+    (if repeat-type
+      (funcall repeat-type 'post))))

But since evil-repeat-type' can return nil and repeat-type' is optional you should consider adding such a check.

from evil.

Original comment by Frank Fischer (Bitbucket: lyro, GitHub: lyro):

I'm sorry but I find this bug report really confusing and difficult to understand. I hope I fixed the bug in 8fac166. If not, please give an example of what is going wrong. Of course, you are encouraged to give hints where the bug may be, but without an example it is very hard to figure out what fails at all.

from evil.