Mailing-List: contact java-user-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: java-user@lucene.apache.org
Received-SPF: pass (nike.apache.org: domain of serera@gmail.com designates
 74.125.78.25 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :content-type;
        b=tXXGeRHyJccXdTzDppDVnktVV0N8DavhMvAIkQOkY/8zF0bJ5aCEas3hotMUPiO5Ag
         MPxwjZ7Hb/XiBebvgc/pEOL/7t7gSLGjSATHCAu8ynsy4NnMA1Ozy+WSqFNTCC9NxsnV
         syw3TYwVO3OiY56vEQOehZ79JbvnWoOaLHW28=
MIME-Version: 1.0
In-Reply-To: <004101ca2bc1$eab54070$710bc30a@sv.us.sonicwall.com>
References: <003101ca2bb9$419a0f00$710bc30a@sv.us.sonicwall.com>
	 <786fde50909020403n3f2d72d8gc35a0d7bc6d21c62@mail.gmail.com>
	 <004101ca2bc1$eab54070$710bc30a@sv.us.sonicwall.com>
Date: Wed, 2 Sep 2009 14:51:09 +0300
Message-ID: <786fde50909020451p3ef98563s40f94a595c059087@mail.gmail.com>
Subject: Re: First result in the group
From: Shai Erera <serera@gmail.com>
To: java-user@lucene.apache.org
Content-Type: multipart/alternative; boundary=0016e6dbe74c022190047296e21e

--0016e6dbe74c022190047296e21e
Content-Type: text/plain; charset=ISO-8859-1

I see ... the solution I have in mind is not simple, but it follows the
Collector approach. Index categories as payloads of documents such that
there is one field (cats:all for example) that includes a posting list for
all documents, each has the categories it is associated w/ in its payload:
cats:all --> 0 [cat1, cat2] 1 [cat1, cat3, cat4] 4 [cat2, cat3, cat4] ...

Then when Collector.collect() is called, skip to that doc ID and read its
categories and store the ID in category maps (i.e., you'll have maps for
cat1, cat2, ... catN, each will include all doc IDs, sorted by score, which
were collected by this query).

Then you can fetch all categories whose maps/sets are not empty and display
M docs from each.

If you know in advance which categories are requested to be grouped by, for
example I want a group by on categories 1, 3, 4 and 7, you can optimize the
solution further, but I'm not sure if that's what you requested.

Also, if you can translate category Strings to integers, you can store more
efficient payloads ...

Shai

On Wed, Sep 2, 2009 at 2:38 PM, Ganesh <emailgane@yahoo.co.in> wrote:

> I have a field called category and all documents will have belong to some
> category( say some belong to X and some Y etc). The field values may change
> dynamically. I want the search results to be filterted to retrieve one
> document per category.
>
> This is similar to 'group by' feature in database.
>
> Regards
> Ganesh
>
>
> ----- Original Message -----
> From: "Shai Erera" <serera@gmail.com>
> To: <java-user@lucene.apache.org>
> Sent: Wednesday, September 02, 2009 4:33 PM
> Subject: Re: First result in the group
>
>
> > What do you mean by "first result in the group"? What is a group?
> >
> > On Wed, Sep 2, 2009 at 1:36 PM, Ganesh <emailgane@yahoo.co.in> wrote:
> >
> >> Hello all,
> >>
> >> I want to retrieve the first result in the group. How to acheive this?
> >> Currently i am parsing all the results, using a hash and avoiding
> duplicate
> >> entries.
> >>
> >> Is there any better way?
> >>
> >> Regards
> >> Ganesh
> >> Send instant messages to your online friends
> http://in.messenger.yahoo.com
> >>
> >> ---------------------------------------------------------------------
> >> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> >> For additional commands, e-mail: java-user-help@lucene.apache.org
> >>
> >>
> >
> Send instant messages to your online friends http://in.messenger.yahoo.com
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
> For additional commands, e-mail: java-user-help@lucene.apache.org
>
>

--0016e6dbe74c022190047296e21e--