Mailing-List: contact java-user-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: java-user@lucene.apache.org
Received-SPF: neutral (athena.apache.org: local policy)
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
  s=s1024; d=yahoo.co.uk;
  h=Message-ID:X-YMail-OSG:Received:X-Mailer:References:Date:From:Subject:To:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding;
  b=QgKih3TH+HrrlISn6XwKteuCNTjvHruY/kbJf2pS25pSlGdtQ1KpcxoDnRG0aXeuWXkYYHiGIXneZeV0aVb077aK1+r3rRpWa8iQCmL61BO+32TbLYtVYeEAz1NAXml/Sm3ylz/3G36GF9WQJxL2XtuJD/oDTbwdYcXcYHXJ5aQ=;
Message-ID: <519838.55818.qm@web24704.mail.ird.yahoo.com>
References: <322510.28299.qm@web24702.mail.ird.yahoo.com>
 <86faf560907070153g220ba978ide2944325c5c9e8b@mail.gmail.com>
Date: Tue, 7 Jul 2009 09:39:54 +0000 (GMT)
From: mark harwood <markharw00d@yahoo.co.uk>
Subject: Re: Boolean retrieval
To: java-user@lucene.apache.org
In-Reply-To: <86faf560907070153g220ba978ide2944325c5c9e8b@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable


Seems a long-winded way of producing a BooleanFilter but I guess you are tr=
ying to work with user input in the form of query strings.=0A=0AThe bug in =
your code is that clause.getQuery().getString() is not producing terms that=
 are in your index - the first call to getTermsFilter passes the string  "+=
f1:aaa +f1:bbb" which is not a term in the index.=0A=0AGiven the requiremen=
t is to ignore scoring I would recommend (as someone else suggested) lookin=
g at the IndexSearch.search method that takes a HitCollector and simply acc=
umulate all results, regardless of score.=0A=0A=0A=0A=0A=0A----- Original M=
essage ----=0AFrom: Lukas Michelbacher <mmmasterluke@gmail.com>=0ATo: java-=
user@lucene.apache.org=0ASent: Tuesday, 7 July, 2009 9:53:24=0ASubject: Re:=
 Boolean retrieval=0A=0ATo test my Boolean queries, I have a small test col=
lection where each document=0Acontains one of 1024 possible combinations of=
 the strings "aaa", "bbb",=0A... "jjj".  I tried wrapping a Boolean query l=
ike this (it's based on an=0Aolder post to this list [1])=0A=0A=0Aprivate s=
tatic TermsFilter getTermsFilter(String field, String text) {=0A  TermsFilt=
er tf =3D new TermsFilter();=0A  tf.addTerm(new Term(field, text));=0A  ret=
urn tf;=0A}=0A=0AQuery q =3D new QueryParser("f1", new StandardAnalyzer()).=
parse("(aaa=0AAND bbb) OR ccc");=0AIndexSearcher searcher =3D new IndexSear=
cher(indexDir);=0ATopDocCollector collector =3D new TopDocCollector(1024);=
=0A=0ABooleanQuery bc =3D (BooleanQuery) q;=0ABooleanFilter finalFilter =3D=
 new BooleanFilter();=0ABooleanFilter boolFilt =3D new BooleanFilter();=0A=
=0A// add each clause of the original query to the filter=0Afor (BooleanCla=
use clause : bc.getClauses()) {=0A  boolFilt.add(new FilterClause(getTermsF=
ilter("f1",=0Aclause.getQuery().toString()), clause.getOccur()));=0A  Syste=
m.out.println(clause.getQuery().toString());=0A}=0A=0AfinalFilter.add(new F=
ilterClause(boolFilt, BooleanClause.Occur.MUST));=0A=0AConstantScoreQuery c=
sq =3D new ConstantScoreQuery(finalFilter);=0Asearcher.search(csq, finalFil=
ter, collector);=0A=0AScoreDoc[] hits =3D collector.topDocs().scoreDocs;=0A=
System.out.println("Found " + collector.getTotalHits() + " hits");=0A=0AThe=
 result is 0 hits (should be 640).=0A=0A[1] tinyurl.com/ml52ye=0A=0A2009/7/=
4 Mark Harwood <markharw00d@yahoo.co.uk>:=0A>=0A> Check out booleanfilter i=
n contrib/queries. It can be wrapped in a constantScoreQuery=0A>=0A>=0A>=0A=
> On 4 Jul 2009, at 17:37, Lukas Michelbacher <michells@ims.uni-stuttgart.d=
e> wrote:=0A>=0A>=0A> This is about an experiment comparing plain Boolean r=
etrieval with=0A> vector-space-based retrieval.=0A>=0A> I would like to dis=
able all of Lucene's scoring mechanisms and just=0A> run a true Boolean que=
ry that returns exactly the documents that match a=0A> query specified in B=
oolean syntax (OR, AND, NOT). No scoring or sorting=0A> required.=0A>=0A> A=
s far as I can see, this is not supported out of the box.  Which classes=0A=
> would I have to modify?=0A>=0A> Would it be enough to create a subclass o=
f Similarity and to ignore all terms but one (coord, say) and make this ter=
m return 1 if the query matches the document and 0 otherwise?=0A>=0A> Lukas=
=0A>=0A> --=0A> Lukas Michelbacher=0A> Institute for Natural Language Proce=
ssing=0A> Universit=E4t Stuttgart=0A> email: michells@ims.uni-stuttgart.de=
=0A=0A---------------------------------------------------------------------=
=0ATo unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org=0AFor ad=
ditional commands, e-mail: java-user-help@lucene.apache.org=0A=0A=0A      

---------------------------------------------------------------------
To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-user-help@lucene.apache.org