On Fri, May 14, 2010 at 10:59 AM, Yonik Seeley
<[email protected]> wrote:

> So it seems like the biggest issue we might have in cutting over would
> be the field cache and sorting?  Instead of using String.compareTo we
> need one that compares as UTF-32 (or longer term, don't even create
> strings of course...)

Admittedly not having looked at all the places that do
String.compareTo on terms [I am sure there are probably more?], I
wonder if its worth the effort up front to just go completely to bytes
for Term and everything.

I worry about using a special String comparator: any code, including
stuff thats not in Lucene/Solr, that puts these things into ordered
collections for example could have bugs.

If we instead move all of this to BytesRef where the comparator just
works by default, it seems less scary.

-- 
Robert Muir
[email protected]

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to