nub

[Richard Braakman]

On Thu, Nov 28, 2002 at 10:21:53PM +0000, Alistair Bayley wrote:
> Wouldn't this have been better called "unique"? (analogous to the Unix
> program "uniq"). I was looking for a "unique" in the GHC Data.List library a
> few days ago, and didn't see one, so I wrote my own (not realising that was
> what "nub" did).

No, Unix uniq makes only a single pass.

uniq = map head . group

(A much tidier definition than the one I gave a few days ago :-)

By contrast, "nub" removes duplicate elements from the list no matter
where they are.

I think both functions are useful.  If I understand it right, uniq can
evaluate its argument list lazily, while nub cannot.  There's no real
need to put uniq in the standard library, though.

Hmm, but with that said, I don't think I disagree with you.  Renaming
"nub" to "unique" makes it clear that it is similar, but not identical
to what Unix "uniq" does.

Richard Braakman