Re: RFC: general sequences

On Mon, May 23, 2005 at 04:52:04PM +0200, Josef Svenningsson wrote:

On 5/23/05, Tomasz Zielonka wrote:

...

Another thing: O(1) reverse would be very nice if it could be implemented with no overhead for other operations ;-)

That should be trivial to implement. Just add a bit in the Seq data type which indicates whether the sequence should be treated backwards or forwards. The reverse operation then just toggles that bit. Or am I missing something?

True -- it would mean a bit more code, though.

Also, I have some comments about the complexities in the documentation. First of all I think it is useful to say what exactly indices like n and i refer to even if it might not be that hard to figure out. Secondly it seems that the complexity of the indexing functions is wrong. Shouldn't they be O(log(min(i,n-i)))?

Ouch! thanks.