topic Lucene Search in Alfresco Archive

Lucene Search

luisg — Tue, 29 Dec 2009 13:19:15 GMT

Hi there…I'm developing a GWT application that uses the Alfresco Repository.In my application I do some lucene searches in the Alfresco repository, but I start notice that something is wrong. So, I started to use the node browser to test the querys I use, and I found what is going on. The thing is,

Re: Lucene Search

luisg — Tue, 29 Dec 2009 14:14:25 GMT

Hi again…
I have some news here…

If I do the query:

PATH:"/app:company_home/cm:FolderA/cm:FolderB/cm:FolderC//*" AND TYPE:"{http://www.model.pt/model/content/1.0}myType" AND @cm\:name:moredataModel.xml*

I get the right result. But for the file teste_psql.sql I had tried

PATH:"/app:company_home/cm:FolderA/cm:FolderB/cm:FolderC//*" AND TYPE:"{http://www.model.pt/model/content/1.0}myType" AND @cm\:name:teste_psql.sql*

and

PATH:"/app:company_home/cm:FolderA/cm:FolderB/cm:FolderC//*" AND TYPE:"{http://www.model.pt/model/content/1.0}myType" AND @cm\:name:*teste_psql.sql*

even

PATH:"/app:company_home/cm:FolderA/cm:FolderB/cm:FolderC//*" AND TYPE:"{http://www.model.pt/model/content/1.0}myType" AND @cm\:name:teste*

and none of them worked.

thanks

Re: Lucene Search

lucille_arkenst — Tue, 29 Dec 2009 16:11:23 GMT

Luis,
I share the same frustration with Lucene Search. I want to look into it a bit more. First, can you post the link that explains to write the code this way?
Thanks,
Lucille

Re: Lucene Search

luisg — Tue, 29 Dec 2009 17:22:34 GMT

Hi…

I solved my problem, searching the documents by its id, and not for its name.

But I think this is something that should be clarified.
In my code I used the code (in java):


String query = "PATH:\""+path+"//*\" AND " +
                    "TYPE:\"{http://www.mymodel.pt/model/content/1.0}document\" AND " +
                    "@cm\\:name:\""+documentName+"\"";

Query q = new Query(Constants.QUERY_LANG_LUCENE, query);
QueryResult queryResult = getRepositoryService().query(STORE, q, false);
org.alfresco.webservice.types.ResultSet resultSet = queryResult.getResultSet();
‍‍‍‍‍‍‍‍‍

But when the program will fail in the line:


Reference reference = new Reference(STORE, rows[0].getNode().getId(), null);
‍‍‍

Giving me a nullPointerException, cause the query don't give me any results.

But, as I told you before, if I execute the query:

PATH:""+path+"//*" AND TYPE:"{http://www.mymodel.pt/model/content/1.0}document"

The document is in the list, but I don't get the document passing its name.

Luis Gonçalves

Re: Lucene Search

alex54 — Mon, 11 Jan 2010 09:52:49 GMT

Is your indexation state OK ? (I mean the synchronisation between data and indexes)
because it look like a out of synch problem…

Re: Lucene Search

maqsood — Thu, 21 Jan 2010 10:50:37 GMT

Hi luisg,

Did you get any solution for your problem? Iam also facing similar kind of problem.

Re: Lucene Search

msvoren — Sun, 24 Jan 2010 15:01:52 GMT

Try to reindex your whole reposiory.
Include following line in your repository properties file:

index.recovery.mode=FULL

and restart alfresco.

Lucene starts to behave in strange manner if indexes are not synced.
About that, can anyone answer why this happens and what one can do to prevent this happening?
This is very unpleasant if it happens on production servers..

Re: Lucene Search

maqsood — Wed, 27 Jan 2010 12:33:25 GMT

Hi msvoren,

Thanks for your help. but it didn't seem to be work out for me…
Any other solution?