Mailing-List: contact users-help@jackrabbit.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@jackrabbit.apache.org
Received-SPF: pass (athena.apache.org: domain of gaeremyncks@gmail.com
 designates 209.85.220.215 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=content-type:mime-version:subject:from:in-reply-to:date
         :content-transfer-encoding:message-id:references:to:x-mailer;
        b=Mv9f6yc8bTPZEkHYwffKhyAoMaFqNKBrW3ulLooaaJ8dMiclQTU/OavXpJRjbz4sOY
         WgpDOXtizslclCCHK48qN1ir9oIdThKmiaGFFf9daR7IYO6F450iiMqcUNiL/3FTfXFX
         DehXX46/cEt8sgEv1uX1p6aGBWjUuRHFic/M8=
Content-Type: text/plain; charset=us-ascii; format=flowed; delsp=yes
Mime-Version: 1.0 (Apple Message framework v1076)
Subject: Re: Excerpts on search queries.
From: Simon Gaeremynck <gaeremyncks@gmail.com>
In-Reply-To: <c3ac3bad0911270803k23cb54a6y4afb9e53e45ec1e5@mail.gmail.com>
Date: Fri, 27 Nov 2009 16:45:07 +0000
Content-Transfer-Encoding: 7bit
Message-Id: <85E27883-0EB8-4E1F-8F27-3758A6824222@gmail.com>
References: <4FF12BCA-F1E0-46D7-9332-F424561FA0F8@gmail.com>
 <c3ac3bad0911270803k23cb54a6y4afb9e53e45ec1e5@mail.gmail.com>
To: users@jackrabbit.apache.org

Thanks for the quick response!

I am now able to retrieve excerpts in one query for both content and  
properties.
According to http://issues.apache.org/jira/browse/JCR-920 it looks  
like it should work for properties (and it sort of does!)

The only thing I have to figure out is how to only show the  
highlighted bit and the content around it for that specific property.
This will probably need a custom excerpt provider.

Thanks for the help!

Simon


On 27 Nov 2009, at 16:03, Alexander Klimetschek wrote:

> On Sat, Nov 28, 2009 at 00:00, Simon Gaeremynck  
> <gaeremyncks@gmail.com> wrote:
>> I am trying to do a basic search query on the entire repository  
>> that returns
>> any node that has a  sling:resourceType property and that has a
>> property/content that matches a search string.
>> If there is a match I need to get all the properties of that node  
>> and an
>> excerpt of the matched string.
>>
>> I have the following xpath query who returns the correct nodes:
>> //*[@sling:resourceType and (jcr:contains(jcr:content,'*test*') or
>> jcr:contains(.,'*test*'))]
>>
>> This query gives me the parent of the jcr:content node and all the
>> properties on it, which is exactly what I want.
>>
>> Now, I thought I could get the excerpts out of the results by  
>> appending
>> /rep:excerpt(.) like so:
>> //*[@sling:resourceType and (jcr:contains(jcr:content,'*test*') or
>> jcr:contains(.,'*test*'))]/rep:excerpt(.)
>>
>> But this doesn't give my a usable excerpt.
>
> I guess the excerpt is only useful when done on the jcr:content node.
> So I think this query should do the trick:
>
> //*[@sling:resourceType and (jcr:contains(jcr:content,'test') or
> jcr:contains(.,'test'))]/rep:excerpt()
>
> and then you'd call the rep:excerpt() with the jcr:content as  
> parameter:
>
> Row row = ...
> Value excerpt = row.getValue("rep:excerpt(jcr:content)");
>
> In this case you could also switch between the excerpt from a
> jcr:content subnode or the node itself, depending on what exists.
>
> Note that I also changed "*test*" to "test" as that should give the
> same results. jcr:contains does a full text search, not an exact
> comparison, hence word stemming etc. all take place.
>
>> Another problem I'm facing with excerpts is when I try to get an  
>> excerpt out
>> of a property/jcr:content with no (or almost none) content around the
>> highlighted word/phrase.
>> It seems to fill up the space by taking other property values of  
>> the node.
>
> You can have a custom excerpt provider:
> http://wiki.apache.org/jackrabbit/ExcerptProvider
>
> Regards,
> Alex
>
> -- 
> Alexander Klimetschek
> alexander.klimetschek@day.com