lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Bob Arens <>
Subject Re: Adding Fields to Documents with UnStored Fields - crazy scheme?
Date Fri, 09 Jun 2006 18:21:15 GMT
> If the old index is optimized then you might be able to iterate  
> through
> all the docs in your old index (sorted by doc id) and for each  
> iteration
> add the corresponding doc to the new index so it has a matching doc  
> id.
> The idea being that after searching on one index you could use the doc
> id to look up the corresponding doc in the other index.

True - it would be like using Lucene as a thin relational DB.  
However, I don't want to just be able to associate the new fields  
with the old, I also want to be able to search over them.

Right now, I'm thinking of rolling a custom MultiSearcher that  
operates over two indices containing identical fileIDs, and returns  
the set intersection of fileIDs returned from their respective  
searches; i.e. a search for "contents:germany title:medicine" would  
get the fileIDs containing "germany" in the contents index, the  
fileIDs containing "medicine" in the new index, find the fileIDs in  
both sets of hits, and return them.

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message