Re: Nested interval tree encoding

From: Eric DeCosta <>
Date: Thu, 3 Jul 2008 14:38:22 -0400
Message-ID: <g4j6au$4fp$>

> Other nodes like,, and are correctly
> decoded; so this looking like a special case involving the first child
> node.
> Is there any way to detect and handle this without looking-up the parent
> node? Would using reversed continued fractions avoid this?

One possible answer, which is inelegant, but wroks is to re-compute the node as the materialzed path is extracted and perform a check at the end and adjust the 'tail' of the path as needed.

I'm hoping for something less brute-force...

-Eric Received on Thu Jul 03 2008 - 20:38:22 CEST

Original text of this message