lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mark Miller <>
Subject Re: Mixing Case and Case-Insensitive Searching
Date Sat, 12 May 2007 19:58:26 GMT

> I'd love to see a formal syntax like this officially enter the Lucene
> standard query language someday.
I doubt that this is something that is ever going to really happen. 
There are a couple of approaches to the problem and there are other 
similar  problems (like  allowing stemmed and unstemmed searches)  and I 
think you are going to be stuck with roll your own for this type of 
thing. The problem with it being in the syntax is that it requires you 
to use certain analyzer's and indexing schemes...something not 
guaranteed...and if you where to add a case sensitive approach to the 
core of Lucene it would slow down the common case for a somewhat rare 
case. I think the brilliance of Lucenes approach is that it provides 
building blocks for your own search solution. I wouldn't consider it out 
of the normal to write your own parser or modify Lucenes parser...I 
think a lot of people do. Your solution is a good one, and I think its 
the right approach with Lucene...there are a lot of cool things you can 
do with different analyzers and token positions...but adding more 
support than what is there will tie down the sweet open endedness. Maybe 
a sandbox package with all of the necessary analyzers and the modified 
query parser would be more likely...even still it seems a lot of stuff 
does not even make the sandbox. These guys keep quite intent on keeping 
Lucene lean and mean and I am quite thankful for it.

As far as solr, a lot of the code can just be stolen or you might check 
out embedded solr and I think there is something else similar to 
embedded solr that might not require a web server...

- Mark

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

View raw message