| Re: Implementing Johnson algorithm |
|
 |
|
 |
|
 |
|
 |
Group: comp.lang.haskell · Group Profile
Author: Dan DoelDan Doel Date: Sep 21, 2007 00:52
andrea wrote:
> Yes thanks, I then found the answer on a book...
> newtype instead is still a data declaration but only accepts 1
> constructor and one argument, am I right?
Yes, that's right. The fact that it is so limited allows the compiler to
optimize things in ways that wouldn't hold for general data declarations,
but as far as how it works, it's about the same as a data declaration with
one constructor and one argument.
|