Hi,
--- Brian Goetz <[EMAIL PROTECTED]> wrote:
> > I am getting a problem when using an analyzer that
> > uses the PorterStemFilter with queries that have
> terms
> > in them. I am using Lucene 1.0, with the small
> porter
> > stemmer fix.
>
> [ Tal; This question needs to go into the FAQ ]
Definitely.
>
> You must use the same analyzer to index the
> documents as you do to
> process the queries. Otherwise, the terms won't
> match. So if you didn't
> use the stemmer when building the index base, you
> can't use it for the
> queries, and if you did use it, you have to use it
> for the queries.
>
Thanks for the reply, but I am using the same analyzer
for the index, as I am for the searching. So, I don't
think that is my problem.
Let me try and explain the problem more clearly...
I have this query
contents:product
That is, I want to search the field "contents" for the
term 'product'.
The porter stemmer filter parses this and produces
these tokens
content
product
But since the field was stored as "contents", that is
Field.Text("contents","product") and the Field class
does not analyze the key of a Field, it does not find
a match.
So, I have resorted to running the keys of my fields
through the analyzer before passing them to the Field
method.
This seems to work fine.
I was just trying to work out if perhaps the analyzer
is meant to be clever enough to identify field
prefixes and not parse/tokenize them?
Thanks,
Chris
=====
Need somewhere to Live in London - http://freeflats.com
__________________________________________________
Do You Yahoo!?
Get personalized email addresses from Yahoo! Mail - only $35
a year! http://personal.mail.yahoo.com/
_______________________________________________
Lucene-users mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/lucene-users
