[Xapian-discuss] Search::Xapian add_database'd search results are odd?

[Eric Parusel]

Olly Betts wrote:
> On Tue, Dec 21, 2004 at 02:25:59PM -0800, Eric Parusel wrote:
> 
>>	From your wording, I take it the mapping is possibly due for a 
>>	change in the future?  I'll keep an eye out for that...
> 
> 
> I don't have a plan to change it, but I wouldn't rule it out totally.
> It's problematic if one database is very large, as you overrun 32 bits
> more quickly.  But it has nice features, such as merged docids being stable
> over adding documents to the underlying databases.
> 
> When this came up before, someone (might have been me) suggested we
> should have a method to return the underlying database and docid
> but I've not had a chance to look at doing that.

Hmm, ok, for me it's the docid's themselves that are important.
If the offset value is within range, I suppose I can't easily reverse 
the equation to get out the original docid value, right?

A method to return the underlying db and docid would be useful to myself 
too.  Heck, the docids are unique (set by me) so I don't even need to 
know the underlying db they're in...

Thanks for everything,
Eric