tweaks: try to distinguish between keystrokes and characters

Keystrokes are single integers (aided by the flags meta_key and
func_key) but in the input stream they can be encoded as escape
sequences (a series of bytes).  Characters are values in byte
range, but in UTF-8 one character can consist of multiple bytes.

Also rename two variables, because the secondary input buffer
that they refer to contains only characters (mostly just one),
never any shortcuts; and there are too many "kbinput" already.
2 files changed