Hi Stone

GSearch supports the Lucene query syntax, see

http://lucene.apache.org/java/2_9_2/queryparsersyntax.html

It says: "Lucene supports fielded data. When performing a search you can either 
specify a field, or use the default field." GSearch supports the default 
field(s) through the property "fgsindex.defaultQueryFields" in the 
index.properties file.

Further, the syntax page says: "Lucene supports single and multiple character 
wildcard searches within single terms (not within phrase queries)." So there 
cannot be wildcards in field names.

I hope this clarifies.

Best
Gert

On 28/04/2010, at 23.30, Stone Xiang wrote:

Hi Steve,

Thanks for the information. I played with gSearch a little bit and looks like 
it has some limitation.
For example, I can issue the Lucene search expression

dsm.rss:"podcast"

to search the "rss" data stream for "podcast" since each data stream will be 
indexed as a Lucene field.
But what if I want to search across all the data streams without explicitly 
listing all the data streams?

dsm.ds1:"podcast"
dsm.ds2:"podcast"
...

is really awkward. Does gSearch accept a simple

dsm.*:"podcast"

expression to search ALL data streams?

Thanks.


Stone

________________________________
From: Steve Bayliss 
<[email protected]<mailto:[email protected]>>
To: Stone Xiang <[email protected]<mailto:[email protected]>>; 
[email protected]<mailto:[email protected]>
Sent: Fri, April 23, 2010 4:35:15 AM
Subject: RE: [Fedora-commons-developers] Full-text search for Fedora 
DataStreams Against Resource Index?

Hi Stone

Theoretically you can do this - but you would need to add the content in 
RELS-EXT or RELS-INT, for example

<info:fedora/myns:somepid/DS1> <hasDatastreamContent> "literal content of 
datastream"

However it would probably be more appropriate to use search/indexing tools 
designed for this - for example Fedora GSearch/SOLR.

Regards
Steve
-----Original Message-----
From: Stone Xiang [mailto:[email protected]]
Sent: 21 April 2010 19:19
To: 
[email protected]<mailto:[email protected]>
Subject: [Fedora-commons-developers] Full-text search for Fedora DataStreams 
Against Resource Index?

Hi, can I run full-text search of Fedora data streams against the resource 
index?
I know one can write SparQL expression to search Dublin Core fields against 
Mulgara
data store since all the Dublin Core fields can be stored in Mulgara as 
triples. But what
about data stream content? If I want to find out all the digital objects whose 
title,
description or document content of any of their data streams contains the word 
"painting",
will resource index (or Mulgara) help?

Thanks.


<ATT00001..c><ATT00002..c>

------------------------------------------------------------------------------
_______________________________________________
Fedora-commons-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/fedora-commons-developers

Reply via email to