Mailing-List: contact java-user-help@lucene.apache.org; run by ezmlm
Precedence: bulk
Reply-To: java-user@lucene.apache.org
Received-SPF: neutral (nike.apache.org: local policy)
Content-Type: text/plain; charset=iso-8859-1
Mime-Version: 1.0 (Apple Message framework v1084)
Subject: Re: Phonetic search with Lucene 3.2
From: Paul Libbrecht <paul@hoplahup.net>
In-Reply-To: 
 <CABPX7Kb-5jmW=x3MKfOw9vN+JXACqpOwCEBUxmtiJxM-_H3nuA@mail.gmail.com>
Date: Wed, 9 Nov 2011 14:20:58 +0100
Content-Transfer-Encoding: quoted-printable
Message-Id: <AD23220A-5B3E-43B5-AF67-1E496E0FFBA3@hoplahup.net>
References: 
 <CABPX7KbW62zUtE=pCJuvVG0XxzU-HJmRBTZ9XHruMGU5f2ZY1w@mail.gmail.com>
 <4A5A76AB-AA58-4A5B-9DF3-4EB76C7232F0@hoplahup.net>
 <CABPX7KbPNVHjOMPE-=R_T8a-fiQ_gEXwHGRH_Z26eKvHR_9YjA@mail.gmail.com>
 <CABPX7KZJTYGHN=3i3is6vWgcaaoBTzZ3U5U=33WxzAJrEtDqsA@mail.gmail.com>
 <A9876D89-6130-4309-BE4B-98C0809A8B85@gmail.com>
 <CABPX7KYsbf13ytBWi6CDjHXiHzSpEs=BJaOObBuhdetr3O9EYg@mail.gmail.com>
 <20F12225-21C9-479D-839B-A8D5D0F4BD9F@gmail.com>
 <CABPX7Kb6y1DKwQ1kkDsGt7Wk0mA0OtmA_j8bECZ_q+6jWXwoXQ@mail.gmail.com>
 <56C49D51-4C20-48DE-AECF-1191FC7C69C4@hoplahup.net>
 <CABPX7Kb-5jmW=x3MKfOw9vN+JXACqpOwCEBUxmtiJxM-_H3nuA@mail.gmail.com>
To: java-user@lucene.apache.org

That uses Lucene 2.9.2 indeed.

paul


Le 9 nov. 2011 =E0 11:43, Felipe Carvalho a =E9crit :

> Which version of Lucene are you using? I had tried it with Lucene 3.3 =
and
> had some problems, did you have to do any customizations?
>=20
> On Wed, Nov 9, 2011 at 8:38 AM, Paul Libbrecht <paul@hoplahup.net> =
wrote:
>=20
>> We've been using
>>       http://www.tangentum.biz/en/products/phonetix/
>> which does double-metaphone.
>> Maybe that helps.
>>=20
>> paul
>>=20
>>=20
>> Le 9 nov. 2011 =E0 11:29, Felipe Carvalho a =E9crit :
>>=20
>>> Using PerFieldAnalyzerWrapper seems to be working for what I need!
>>>=20
>>> On indexing:
>>>=20
>>>       PerFieldAnalyzerWrapper wrapper =3D new =
PerFieldAnalyzerWrapper(new
>>> StandardAnalyzer(Version.LUCENE_33));
>>>       wrapper.addAnalyzer("nome", new =
MetaphoneReplacementAnalyzer());
>>>       IndexWriterConfig indexWriterConfig =3D new
>>> IndexWriterConfig(Version.LUCENE_33, wrapper);
>>>       Directory directory =3D FSDirectory.open(new File(indexPath));
>>>       IndexWriter indexWriter =3D new IndexWriter(directory,
>>> indexWriterConfig);
>>>=20
>>> On search:
>>>=20
>>>       Directory directory =3D FSDirectory.open(new
>>> File(lastIndexDir(Calendar.getInstance())));
>>>       IndexSearcher is =3D new IndexSearcher(directory);
>>>       PerFieldAnalyzerWrapper wrapper =3D new =
PerFieldAnalyzerWrapper(new
>>> StandardAnalyzer(Version.LUCENE_33));
>>>       wrapper.addAnalyzer("name", new =
MetaphoneReplacementAnalyzer());
>>>       QueryParser parser =3D new QueryParser(Version.LUCENE_33, =
"name",
>>> wrapper);
>>>       Query query =3D parser.parse(expression);
>>>       ScoreDoc[] hits =3D is.search(query, 1000).scoreDocs;
>>>=20
>>> Does anyone know any other phonetic analyzer implementation? I'm =
using
>>> MetaphoneReplacementAnalyzer from LIA examples.
>>>=20
>>> I'm looking at lucene-contrib stuff at
>>> http://lucene.apache.org/java/3_4_0/lucene-contrib/index.html but I
>> can't
>>> seem to find other phonetic analyzers.
>>>=20
>>> Thanks!
>>>=20
>>>=20
>>> On Tue, Nov 8, 2011 at 12:19 PM, Erik Hatcher =
<erik.hatcher@gmail.com
>>> wrote:
>>>=20
>>>>=20
>>>> On Nov 8, 2011, at 05:42 , Felipe Carvalho wrote:
>>>>>> Yes, quite possible, including boosting on exact matches if you =
want.
>>>> Use
>>>>>> a BooleanQuery to wrap clauses parsed once with phonetic =
analysis, and
>>>> once
>>>>>> without, including fields at indexing time for both too of =
course.
>>>>>>=20
>>>>>=20
>>>>> Would it be possible to point to an example where this is done. =
The
>> best
>>>>> example of a BooleanQuery I've found so far is this one:
>>>>>=20
>>>>=20
>> =
http://www.avajava.com/tutorials/lessons/how-do-i-combine-queries-with-a-b=
oolean-query.html
>>>>>=20
>>>>> But I couldn't find a boolean query using different analyzers for
>>>> different
>>>>> fields of the attribute.
>>>>=20
>>>> You could use two different QueryParser instances with different
>>>> analyzers.  Or use the PerFieldAnalyzerWrapper, though you'll still
>> need to
>>>> instances in order to have a different default field for each
>> expression.
>>>> But then use the techniques you saw in that article (or in Lucene =
in
>>>> Action, since you mentioned having that) to combine Query objects =
into a
>>>> BooleanQuery.
>>>>=20
>>>>      Erik
>>>>=20
>>>>=20
>>>> =
---------------------------------------------------------------------
>>>> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
>>>> For additional commands, e-mail: java-user-help@lucene.apache.org
>>>>=20
>>>>=20
>>=20
>>=20
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: java-user-unsubscribe@lucene.apache.org
>> For additional commands, e-mail: java-user-help@lucene.apache.org
>>=20
>>=20


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