Ok - at least now I learned about Luke, neat tool! :)

But I did not come any closer. But now I'm 100% sure that the field is
actually indexed. And I noted another thing as well: It is every integer
only field I have - none of them are searchable. Not in my application
and not in Luke.

I have tried using SimpleAnalyzer, KeywordAnalyzer and SatndardAnalyzer
now - none of them gives me any hits when searching for isbn.

Oh btw - I'm using 2.0...

Thomas 

-----Oprindelig meddelelse-----
Fra: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] 
Sendt: 20. september 2007 09:55
Emne: Re: Problem indexing isbn

ok about the "-", but i still think that the keyword analyzer is the
best option in case the isbn is in a separate field.

This point in the lucene faq cover the most common causes of this error.
http://wiki.apache.org/lucene-java/LuceneFAQ#head-3558e5121806fb4fce80fc
022d889484a9248b71

regards.


On 9/20/07, Michael Mitiaguin <[EMAIL PROTECTED]> wrote:
> Jokin:
> I don't  agree , StandardAnalyser doesn't break  xx-xx-xx , but  keeps

> it as  is.
> Try Luke to have a look (  rhyming unintentional :) , tab Plugin shows
that.
> )
>
> Thomas:
> If ISBN is a separate Lucene field  ( you don't concat database fields

> for searching ) , are you sure  it  is either default or you 
> explicitly use field name when searching.
>
> Regards
> Michael
> On 9/20/07, Jokin Cuadrado <[EMAIL PROTECTED]> wrote:
> >
> > standardAnalyzer will break the isbn at the "-" characters. I think 
> > the best option in this case is to use a  keyword analizer that will

> > index the field without analyzing it.
> >
> > regards
> > Jokin
> >
> > On 9/20/07, Thomas Skovsende <[EMAIL PROTECTED]> wrote:
> > > Hey guys.
> > >
> > > I have searched around abit, and I havent really found a solution 
> > > to my problem, so I'm gonna try here.
> > >
> > > I am trying to index an ISBN number using a StandardAnalyzer, but 
> > > i get absolutely no hits when I try to search on that ISBN number.

> > > Every other search I do works perfectly, so I know that everything

> > > is set up correctly.
> > >
> > > I have tried to look at the code for the StandardTokenizer and as 
> > > far as I can see, it should be able to index ISBN's.
> > >
> > > I feel pretty sure this is just another case of "stupid 
> > > developer", so if anyone have any pointers or hints, they would be
appriciated.
> > >
> > > Best regards,
> > >   Thomas Skovsende
> > >
> >
>

Reply via email to