jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ard Schrijvers <a.schrijv...@onehippo.com>
Subject Re: jcr:like property analyzer
Date Thu, 18 Jun 2009 22:18:59 GMT
Hello Neman,

first of all I recommend you to use a french analyzer. Secondly,
though, wildcard queries in combination with analyzing in highly
complex. I would always recommend the 'contains' and not 'like'
useage. But you would still have the wildcard issue in combination
with "é", "è" and "ê". The lucene guys can tell you all about this,

Regards Ard

On Thu, Jun 18, 2009 at 10:53 PM, Ian Boston<ieb@tfd.co.uk> wrote:
> You *should* be able to replace the analyzer with a Latin1 Anayser like the
> one that uses the ISOLantin1Filter [1] in repository.xml according to [2]
>
> I have a vague memory that back at lucene 1.9.1 there were a bunch of bugs
> in teh ISOLatin1Filter that caused problems with some european languages. I
> submitted a patch, which I think made it in.
>
> Unfortunately I cant remember where our version of the ISOLatin1Filter is,
> it was several years ago.
>
> HTH
> Ian
>
> [1]
> http://www.crosswire.org/jsword/java2html/org/crosswire/jsword/index/lucene/analysis/SimpleLuceneAnalyzer.java.html
> [2] http://wiki.apache.org/jackrabbit/Search
>
> On 18 Jun 2009, at 17:58, neman ould sidahmed wrote:
>
>> Thank you for your response, If I try  "select * from item where title
>> like 'Le bon march%' " I get one result.
>> The question is how can I index and search the sentense "Le bon marché de
>> Samedi" and get a result with the queries :
>> "select * from item where title like 'Le bon marche%' ""select * from item
>> where title like 'Le bon marché%' ""select * from item where title like 'Le
>> bon marchê%' ""select * from item where title like 'Le bon marchë%' "
>> Explication : I index the sentense like "Le bon marche de Samedi" and in
>> search, The analyzer convert the sentense to "Le bon marche de Samedi"
>> without accent.
>> Regards,
>> nould.
>> --- En date de : Jeu 18.6.09, Alexander Klimetschek <aklimets@day.com> a
>> écrit :
>>
>> De: Alexander Klimetschek <aklimets@day.com>
>> Objet: Re: jcr:like property analyzer
>> À: users@jackrabbit.apache.org
>> Date: Jeudi 18 Juin 2009, 18h37
>>
>> On Thu, Jun 18, 2009 at 5:23 PM, neman ould sidahmed<nouldsid@yahoo.fr>
>> wrote:
>>>
>>> Hello everyone, I would like know, how I can index and search a property
>>> that contains frensh character like "é", "è" and "ê".
>>> For example : I index the property "tilte" with the value "Le bon marché
>>> de Samedi". but If i use the query : "select * from item where title like
>>> 'Le bon marche%' ", i get no result, the true question is how can use an
>>> analyzer for property "title" and with function like.
>>
>> Did you try "select * from item where title like 'Le bon march%' " ?
>>
>> Regards,
>> Alex
>>
>> --
>> Alexander Klimetschek
>> alexander.klimetschek@day.com
>>
>>
>>
>
>

Mime
View raw message