You are viewing a plain text version of this content. The canonical link for it is here.
Posted to java-user@lucene.apache.org by Waleed Tayea <Wa...@bibalex.org> on 2006/03/06 11:50:53 UTC
Exact Search
Dear All.
How can I perform an exact search on an index constructed with a
morphological analyzer.
Thanks in advance
Waleed,
Re: Multi Search vs reader?
Posted by Yonik Seeley <ys...@gmail.com>.
On 3/21/06, Brian <ok...@yahoo.com> wrote:
> I had assumed since I had multiple indexes,
> that meant I need to perform a MultiSearch. However
> several people have mentioned MultiReader. What's the
> difference, and what's the benefit of one over the
> other??
A MultiSearcher can search over Searchables, which can be remote, etc.
A MultiReader is for local indicies. When you open an IndexReader or
IndexSearcher on a single index with multiple segments, MultiReader is
actually used under the covers. You can also instantiate a
MultiReader with multiple IndexReaders.
An IndexSearcher with a MultiReader (when you can use it) should be
faster and more efficient than a MultiSearcher across multiple
IndexSearchers.
-Yonik
http://incubator.apache.org/solr Solr, The Open Source Lucene Search Server
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
Multi Search vs reader?
Posted by Brian <ok...@yahoo.com>.
I've been doing some reading and I'm a bit confused,
Could you help me clear this up.
I have roughly 12 indexes available at any 1 time for
a search. I had assumed since I had multiple indexes,
that meant I need to perform a MultiSearch. However
several people have mentioned MultiReader. What's the
difference, and what's the benefit of one over the
other??
Thanks, Brian
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
Re: MultiSearch
Posted by Otis Gospodnetic <ot...@yahoo.com>.
The Javadoc should have all the info.
If not - Lucene in Action - http://www.lucenebook.com/search?query=multisearcher
If not - Lucene in Action's free code that includes code with MultiSearcher, as you can see from snippets at the above URL.
Otis
----- Original Message ----
From: Brian <ok...@yahoo.com>
To: java-user@lucene.apache.org
Sent: Wednesday, March 15, 2006 1:39:28 PM
Subject: MultiSearch
Hello Everyone,
I currently have an IndexSearch working Great!
What I want to do now, is move to a multi Index
search. What's the best way to go about it? Is it a
simple process? Any thought's would be appreciated.
Thanks, B
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
MultiSearch
Posted by Brian <ok...@yahoo.com>.
Hello Everyone,
I currently have an IndexSearch working Great!
What I want to do now, is move to a multi Index
search. What's the best way to go about it? Is it a
simple process? Any thought's would be appreciated.
Thanks, B
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
Re: File Name Search
Posted by Erik Hatcher <er...@ehatchersolutions.com>.
If and how you tokenize is entirely dependent on how the queries need
to work. Lucene index design really is driven from querying needs
backwards.
Erik
On Mar 6, 2006, at 10:00 AM, Brian wrote:
> Cool,
> Basically I have soming similar to:
> name_division.date_order_code
>
> So I'm guessing I need to tokenize.
>
> Thanks, B
>
> --- Erik Hatcher <er...@ehatchersolutions.com> wrote:
>
>> On Mar 6, 2006, at 8:07 AM, Brian wrote:
>>> Quick Question,
>>> Is it possible to create an index & search
>> based
>>> on file names?
>>
>> Of course. One option is to simply make filename
>> another field.
>> The question is, should it be an exact match on
>> filename for
>> querying? Or should the filename get tokenized such
>> that a user
>> could query on "filename" and find "filename.txt",
>> and so on?
>>
>> Erik
>>
>>
>>
> ---------------------------------------------------------------------
>> To unsubscribe, e-mail:
>> java-user-unsubscribe@lucene.apache.org
>> For additional commands, e-mail:
>> java-user-help@lucene.apache.org
>>
>>
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail: java-user-help@lucene.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
Re: File Name Search
Posted by Brian <ok...@yahoo.com>.
Cool,
Basically I have soming similar to:
name_division.date_order_code
So I'm guessing I need to tokenize.
Thanks, B
--- Erik Hatcher <er...@ehatchersolutions.com> wrote:
> On Mar 6, 2006, at 8:07 AM, Brian wrote:
> > Quick Question,
> > Is it possible to create an index & search
> based
> > on file names?
>
> Of course. One option is to simply make filename
> another field.
> The question is, should it be an exact match on
> filename for
> querying? Or should the filename get tokenized such
> that a user
> could query on "filename" and find "filename.txt",
> and so on?
>
> Erik
>
>
>
---------------------------------------------------------------------
> To unsubscribe, e-mail:
> java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail:
> java-user-help@lucene.apache.org
>
>
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
Re: File Name Search
Posted by Erik Hatcher <er...@ehatchersolutions.com>.
On Mar 6, 2006, at 8:07 AM, Brian wrote:
> Quick Question,
> Is it possible to create an index & search based
> on file names?
Of course. One option is to simply make filename another field.
The question is, should it be an exact match on filename for
querying? Or should the filename get tokenized such that a user
could query on "filename" and find "filename.txt", and so on?
Erik
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
Re: File Name Search
Posted by Brian <ok...@yahoo.com>.
GREAT!!
I don't have any questions today, I just wanted to
make sure it was possible first. I'll be starting this
in a few days (when I get an Okie Dokie...)
Then I'm sure I'll have some questions.
Thanks for the link and the reply.
V/R
B
--- Volodymyr Bychkoviak <vb...@i-hypergrid.com>
wrote:
> Yes. It possible.
> I've developed such search for our LAN shared files.
>
> I'm using technique of rotating filenames to improve
> wildcard queries
> performance.
> details :
>
http://mail-archives.apache.org/mod_mbox/lucene-java-user/200506.mbox/%3cD3188A0C-B427-4286-9D1C-AE5E98BFBE84@ehatchersolutions.com%3e
>
> Other improvements:
> WildCardQuery is rewritten into ConstantScoreQuery.
> Also I've increased IndexWriter.termInterval from
> 128(default) to 2000.
> Before this change memory consumption was too high.
>
> Now index has about 10M docs and size about 4Gb
> (primarily because of
> storing full path to file) and runs fine with
> -Xms100M memory limit.
> It has simple web interface implemented with HTTPD
> <http://elonen.iki.fi/code/nanohttpd/> java HTTP
> server.
>
> File list is refreshed every 3 days for every
> workstation.
>
> For assessing network files I'm using JCIFS
> (jcifs.samba.org)
>
> Questions?
>
> Brian wrote:
> > Quick Question,
> > Is it possible to create an index & search
> based
> > on file names?
> >
> > Thanks,
> > B
> >
> > __________________________________________________
> > Do You Yahoo!?
> > Tired of spam? Yahoo! Mail has the best spam
> protection around
> > http://mail.yahoo.com
> >
> >
>
---------------------------------------------------------------------
> > To unsubscribe, e-mail:
> java-user-unsubscribe@lucene.apache.org
> > For additional commands, e-mail:
> java-user-help@lucene.apache.org
> >
> >
> >
>
> --
> regards,
> Volodymyr Bychkoviak
>
>
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
Re: File Name Search
Posted by Volodymyr Bychkoviak <vb...@i-hypergrid.com>.
Yes. It possible.
I've developed such search for our LAN shared files.
I'm using technique of rotating filenames to improve wildcard queries
performance.
details :
http://mail-archives.apache.org/mod_mbox/lucene-java-user/200506.mbox/%3cD3188A0C-B427-4286-9D1C-AE5E98BFBE84@ehatchersolutions.com%3e
Other improvements:
WildCardQuery is rewritten into ConstantScoreQuery.
Also I've increased IndexWriter.termInterval from 128(default) to 2000.
Before this change memory consumption was too high.
Now index has about 10M docs and size about 4Gb (primarily because of
storing full path to file) and runs fine with -Xms100M memory limit.
It has simple web interface implemented with HTTPD
<http://elonen.iki.fi/code/nanohttpd/> java HTTP server.
File list is refreshed every 3 days for every workstation.
For assessing network files I'm using JCIFS (jcifs.samba.org)
Questions?
Brian wrote:
> Quick Question,
> Is it possible to create an index & search based
> on file names?
>
> Thanks,
> B
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail: java-user-help@lucene.apache.org
>
>
>
--
regards,
Volodymyr Bychkoviak
File Name Search
Posted by Brian <ok...@yahoo.com>.
Quick Question,
Is it possible to create an index & search based
on file names?
Thanks,
B
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org
Re: Exact Search
Posted by Erik Hatcher <er...@ehatchersolutions.com>.
Could you please provide an example of some sample text, the terms
that are emitted by the analyzer, and a query you'd like to work?
Erik
On Mar 6, 2006, at 5:50 AM, Waleed Tayea wrote:
> Dear All.
>
> How can I perform an exact search on an index constructed with a
> morphological analyzer.
>
>
>
> Thanks in advance
>
> Waleed,
>
>
>
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org