solr home

2009-09-25 Thread Park, Michael
I already have a handful of solr instances running .  However, I'm
trying to install solr (1.4) on a new linux server with tomcat using a
context file (same way I usually do):

 



   



 

However it throws an exception due to the following:

SEVERE: Could not start SOLR. Check solr/home property

java.lang.RuntimeException: Can't find resource 'solrconfig.xml' in
classpath or 'solr/conf/', cwd=/opt/local/solr/fedora_solr

at
org.apache.solr.core.SolrResourceLoader.openResource(SolrResourceLoader.
java:198)

at
org.apache.solr.core.SolrResourceLoader.openConfig(SolrResourceLoader.ja
va:166)

 

Any ideas why this is happening?

 

Thanks, Mike



RE: what is too large for an indexed field

2009-09-21 Thread Park, Michael
I'm using the solr.WhitespaceTokenizerFactory and the
solr.LowerCaseFilterFactory.  Is it safe to assume that a token would be
created for each word?  

I can't image anything that would be even close to 16383 chars. Is there
a way to dissect the tokens? 

Thanks, Mike

-Original Message-
From: Mark Miller [mailto:markrmil...@gmail.com] 
Sent: Monday, September 21, 2009 3:42 PM
To: solr-user@lucene.apache.org
Subject: Re: what is too large for an indexed field

Park, Michael wrote:
> I am trying to place the value of around 390,000 characters into a
> single field.  However, my search results have become inaccurate.  Is
> this too large?  I tried bumping the maxFieldLength in the
> solrconfig.xml file to 500,000 and it hasn't fixed the problem.
>
>  
>
> Thanks,
>
> Mike
>
>
>   
How large is your largest token? There is hard limit of (I think) 16383
chars.

-- 
- Mark

http://www.lucidimagination.com





RE: what is too large for an indexed field

2009-09-21 Thread Park, Michael
I get no results back on a search.  But I can see the actual word or phrase in 
the stored doc.

-Original Message-
From: ysee...@gmail.com [mailto:ysee...@gmail.com] On Behalf Of Yonik Seeley
Sent: Monday, September 21, 2009 4:18 PM
To: solr-user@lucene.apache.org
Subject: Re: what is too large for an indexed field

On Mon, Sep 21, 2009 at 3:27 PM, Park, Michael  wrote:
> I am trying to place the value of around 390,000 characters into a
> single field.  However, my search results have become inaccurate.

Do you mean that the document should score higher, or that the
document doesn't match a particular query?
If the former, keep in mind that length normalization penalizes long documents.

-Yonik
http://www.lucidimagination.com


RE: Solr + autocomplete

2007-11-12 Thread Park, Michael
Will I need to use Solr 1.3 with the EdgeNGramFilterFactory in order to
get the autosuggest feature?

-Original Message-
From: Chris Hostetter [mailto:[EMAIL PROTECTED] 
Sent: Monday, November 12, 2007 1:05 PM
To: solr-user@lucene.apache.org
Subject: RE: Solr + autocomplete


: "Error loading class 'solr.EdgeNGramFilterFactory'".  For some reason

EdgeNGramFilterFactory didn't exist when Solr 1.2 was released, but the 
EdgeNGramTokenizerFactory did.  (the javadocs that come with each
release 
list all of the various factories in that release)


-Hoss



RE: Solr + autocomplete

2007-11-12 Thread Park, Michael
Thanks Ryan,

This looks like the way to go.  However, when I set up my schema I get,
"Error loading class 'solr.EdgeNGramFilterFactory'".  For some reason
the class is not found.  I tried the stable 1.2 build and even tried the
nightly build.  I'm using "".

Any suggestions?

Thanks,
Mike

-Original Message-
From: Ryan McKinley [mailto:[EMAIL PROTECTED] 
Sent: Monday, October 15, 2007 4:44 PM
To: solr-user@lucene.apache.org
Subject: Re: Solr + autocomplete

> 
> I would imagine there is a library to set up an autocomplete search
with
> Solr.  Does anyone have any suggestions?  Scriptaculous has a
JavaScript
> autocomplete library.  However, the server must return an unordered
> list.
> 

Solr does not provide an autocomplete UI, but it can return JSON that a 
JS library can use to populate an autocomplete.

Depending on you index size/ query speed, you may be fine with a 
standard faceting prefix filter.  If the index is large, you may want to

index using the EdgeNGramFilterFactory.

Check the last comment in:
https://issues.apache.org/jira/browse/SOLR-357

ryan




RE: Solr + autocomplete

2007-10-18 Thread Park, Michael
Thx!  I remember coming across extjs a ways back.  It was very slick.
I'll give it a try.

-Original Message-
From: Bharani [mailto:[EMAIL PROTECTED] 
Sent: Thursday, October 18, 2007 5:59 AM
To: solr-user@lucene.apache.org
Subject: RE: Solr + autocomplete


You should take a look at http:\\www.extjs.com. The combo box has got an
autocomplete fultionality. Infact it even has paging built into it. I
just
did a demo using Solr for autocomplete and i got a very good responsive
GUI.
I have got about 100,000 documents with 26 fields each and get a
response <
1s

Hope that helps
-Bharani

Park, Michael wrote:
> 
> Thanks!  That's a good suggestion too.  I'll look into that.
> 
> Actually, I was hoping someone had used a reliable JS library that
> accepted JSON. 
> 
> -Original Message-
> From: Ryan McKinley [mailto:[EMAIL PROTECTED] 
> Sent: Monday, October 15, 2007 4:44 PM
> To: solr-user@lucene.apache.org
> Subject: Re: Solr + autocomplete
> 
>> 
>> I would imagine there is a library to set up an autocomplete search
> with
>> Solr.  Does anyone have any suggestions?  Scriptaculous has a
> JavaScript
>> autocomplete library.  However, the server must return an unordered
>> list.
>> 
> 
> Solr does not provide an autocomplete UI, but it can return JSON that
a 
> JS library can use to populate an autocomplete.
> 
> Depending on you index size/ query speed, you may be fine with a 
> standard faceting prefix filter.  If the index is large, you may want
to
> 
> index using the EdgeNGramFilterFactory.
> 
> Check the last comment in:
> https://issues.apache.org/jira/browse/SOLR-357
> 
> ryan
> 
> 
> 
> 

-- 
View this message in context:
http://www.nabble.com/Solr-%2B-autocomplete-tf4630140.html#a13271445
Sent from the Solr - User mailing list archive at Nabble.com.



RE: Solr + autocomplete

2007-10-15 Thread Park, Michael
Thanks!  That's a good suggestion too.  I'll look into that.

Actually, I was hoping someone had used a reliable JS library that
accepted JSON. 

-Original Message-
From: Ryan McKinley [mailto:[EMAIL PROTECTED] 
Sent: Monday, October 15, 2007 4:44 PM
To: solr-user@lucene.apache.org
Subject: Re: Solr + autocomplete

> 
> I would imagine there is a library to set up an autocomplete search
with
> Solr.  Does anyone have any suggestions?  Scriptaculous has a
JavaScript
> autocomplete library.  However, the server must return an unordered
> list.
> 

Solr does not provide an autocomplete UI, but it can return JSON that a 
JS library can use to populate an autocomplete.

Depending on you index size/ query speed, you may be fine with a 
standard faceting prefix filter.  If the index is large, you may want to

index using the EdgeNGramFilterFactory.

Check the last comment in:
https://issues.apache.org/jira/browse/SOLR-357

ryan




RE: tomcat context fragment

2007-06-06 Thread Park, Michael
Hi Chris,

No.  I set up a separate file, same as the wiki.  

It's either a tomcat version issue or a difference between how tomcat on
my Win laptop is configured vs. the configuration on our tomcat Unix
machine. 

I intend to run multiple instances of solr in production and wanted to
use the context fragments.

I have 3 test instances of solr running now (with 3 context files) and
found that whatever you set the path attribute to becomes the name of
the deployed web app (it doesn't have to match the name of the context
file, but cleaner to keep the names the same).

Here is what I found on the Apache site about this:
"The context path of this web application, which is matched
against
the beginning of each request URI to select the appropriate web
application for processing. All of the context paths within a
particular Host must be unique. If you specify a context path of an
empty string (""), you are defining the default web application for
this Host, which will process all requests not assigned to other
Contexts."

~Mike

-Original Message-
From: Chris Hostetter [mailto:[EMAIL PROTECTED] 
Sent: Wednesday, June 06, 2007 2:53 PM
To: solr-user@lucene.apache.org
Subject: RE: tomcat context fragment

: I've found the problem.
:
: The Context attribute path needed to be set:
:
: 

RE: tomcat context fragment

2007-06-06 Thread Park, Michael
I've found the problem. 

The Context attribute path needed to be set:


   



-Original Message-
From: Park, Michael [mailto:[EMAIL PROTECTED] 
Sent: Tuesday, June 05, 2007 5:28 PM
To: solr-user@lucene.apache.org
Subject: tomcat context fragment

Hello All,

 

I've been working with solr on Tomcat 5.5/Windows and had success
setting my solr home using the context fragment.  However, I cannot get
it to work on Tomcat 5.028/Unix.  I've read and re-read the Apache
Tomcat documentation and cannot find a solution.  Has anyone run into
this issue?  Is there some Tomcat setting that is preventing this from
working?

 

Thanks,

Mike



tomcat context fragment

2007-06-05 Thread Park, Michael
Hello All,

 

I've been working with solr on Tomcat 5.5/Windows and had success
setting my solr home using the context fragment.  However, I cannot get
it to work on Tomcat 5.028/Unix.  I've read and re-read the Apache
Tomcat documentation and cannot find a solution.  Has anyone run into
this issue?  Is there some Tomcat setting that is preventing this from
working?

 

Thanks,

Mike