Mailing-List: contact java-user-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: java-user@lucene.apache.org
Received-SPF: pass (herse.apache.org: local policy)
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
  s=s1024; d=yahoo.com;
  h=Message-ID:X-YMail-OSG:Received:X-Mailer:Date:From:Subject:To:MIME-Version:Content-Type:Content-Transfer-Encoding;
  b=bPq03+1tCabWQL22tERDfbsWjfKtVxaq2Mk9LzLR/gqdEmDFWOcZsG/BP1m0qXWXivHx3GfSm68AfCaoO6zdF3SNFrpyO/shBU8UcfUmHlbkmqNy/CxSl2U+sH76/JXigZzPgXiu6uWt9i5HIctL1btcBpC57HjTx+fTXyWddao=
  ;
Message-ID: <20070315030944.90761.qmail@web50302.mail.re2.yahoo.com>
Date: Wed, 14 Mar 2007 20:09:44 -0700 (PDT)
From: Otis Gospodnetic <otis_gospodnetic@yahoo.com>
Subject: Re: Performance between Filter and HitCollector?
To: java-user@lucene.apache.org
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

eks dev and others - have you tried using the code from LUCENE-584?  Notice=
d any performance increase when you disabled scoring?  I'd like to look at =
that patch soon and commit it if everything is in place and makes sense, so=
 I'm curious if you or anyone else already tried this patch...=0A=0AThanks,=
=0AOtis=0A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .=0ASi=
mpy -- http://www.simpy.com/  -  Tag  -  Search  -  Share=0A=0A----- Origin=
al Message ----=0AFrom: eks dev <eksdev@yahoo.co.uk>=0ATo: java-user@lucene=
.apache.org=0ASent: Wednesday, March 14, 2007 3:59:25 PM=0ASubject: Re: Per=
formance between Filter and HitCollector?=0A=0Ajust to complete this fine a=
nswer,=0Athere is also Matcher patch (https://issues.apache.org/jira/browse=
/LUCENE-584)  that could bring the best of both worlds via e.g. ConstantSco=
ringQuery or another abstraction that enables disabling Scoring (where appr=
opriate)=0A=0A----- Original Message ----=0AFrom: Chris Hostetter <hossman_=
lucene@fucit.org>=0ATo: java-user@lucene.apache.org=0ASent: Wednesday, 14 M=
arch, 2007 7:15:06 PM=0ASubject: Re: Performance between Filter and HitColl=
ector?=0A=0A=0Ait's kind of an Apples/Oranges comparison .. in the examples=
 you gave=0Abelow, one is executing an arbitrary query (which oculd be anyt=
hing) the=0Aother is doing a simple TermEnumeration.=0A=0AAsuming that Quer=
y is a TermQuery, the Filter is theoreticaly going to be=0Afaster becuase i=
t does't have to compute any Scores ... generally speaking=0Aa a Filter wil=
l alwyas be a little faster then a functionally equivilent=0AQuery for the =
purposes of building up a simple BitSet of matching=0Adocuments because teh=
 Query involves the score calcuations ... but the=0AQuery is generally more=
 usable.=0A=0AThe Query can also be more efficient in other ways, because t=
he=0AHitCollector doesn't *have* to build a BitSet, it can deal with the=0A=
results in whatever way it wants (where as a Filter allways generates a=0AB=
itSet).=0A=0ASolr goes the HitCollector route for a few reasons:=0A  1) all=
ows us to use hte DocSet abstraction which allows other=0A     performance =
benefits over straight BitSets=0A  2) allows us to have simpler code that b=
uilds DocSets and DocLists=0A     (DocLists know about scores, sorting, and=
 pagination) in a single=0A     pass when scores or sorting are requested.=
=0A=0A=0A=0A-Hoss=0A=0A=0A-------------------------------------------------=
--------------------=0ATo unsubscribe, e-mail: java-user-unsubscribe@lucene=
.apache.org=0AFor additional commands, e-mail: java-user-help@lucene.apache=
.org=0A=0A=0A=0A=0A=0A=0A        =0A_______________________________________=
____________________ =0AAll New Yahoo! Mail =96 Tired of unwanted email com=
e-ons? Let our SpamGuard protect you. http://uk.docs.yahoo.com/nowyoucan.ht=
ml=0A=0A-------------------------------------------------------------------=
--=0ATo unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org=0AFor =
additional commands, e-mail: java-user-help@lucene.apache.org=0A=0A=0A=0A

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