> Now, I would like to obtain the List of all Terms (and their corresponding
> position) from each document (hits.doc(i)).
Try IndexReader.getTermFreqVector(), which will return an instance of
TermPositionVector when the corresponding field has been indexed with
storeTermVector==true.
--
You should perhaps go about implementing an automatic index backup feature of
some sort. In the case of index corruption you would at least be able to go
back to the latest backup.
--
Maik Schreiber * http://www.blizzy.de
GPG public key: http://pgp.mit.edu:11371/pks/lookup?op=get&a
at's exactly the problem you get when following the practice of using
existing exception types "when your exception does not have something new to
say". The thing is, subclassing an existing exception type actually _does_
make the new type more meaningful and specialized.
--
Original Message
Subject: Contribution: LuceneIndexAccessor
Date: Mon, 16 May 2005 18:11:23 +0200
From: Maik Schreiber <[EMAIL PROTECTED]>
Reply-To: java-dev@lucene.apache.org
To: java-dev@lucene.apache.org
Hi,
My company would like to make the following contribution to
, you can open multiple
readers and searchers while concurrently having opened a writer. Closing the
writer will block until all readers and searchers have been released by client
code (so they can be closed).
--
Maik Schreiber * http://www.blizzy.de
GPG public key: http://pgp.mit.edu:11371/p
h my own words. See for yourself if
they match :)
Aside from all that, did you really want to query for the _phrase_ "audio
cable"? Perhaps you wanted the words "audio" and "cable" to appear anywhere
in the fields, but not necessarily as a phrase.
--
Maik Schreiber
ndle this by itself but relies on outside synchronization.
You might want to have a look at LuceneIndexAccessor (please see
http://issues.apache.org/bugzilla/show_bug.cgi?id=34995)
--
Maik Schreiber * http://www.blizzy.de
GPG public key: http://pgp.mit.edu:11371/pks/lookup?op=get&searc
using HTTP authentication for XML-RPC requests.
--
Maik Schreiber * http://www.blizzy.de
GPG public key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x1F11D713
Key fingerprint: CF19 AFCE 6E3D 5443 9599 18B5 5640 1F11 D713
--
it) in the index so I
> know when e.g. the request comes to delete all documents I also which ones.
I'm not sure if you're talking about authentication for using the XML-RPC
interface here...?
--
Maik Schreiber * http://www.blizzy.de
GPG public key: http://pgp.mit.edu:11371/pks
and store this as Field.Keyword in the lucene index
> and "AND" all queries for that key too.
Okay, I see. In our system we don't filter search results by user. HTTP
authentication is only used to protect the XML-RPC interface so that not
everyone can use it at will.
--
Maik Schre
lines, and add those documents with a "text" field into a Lucene index.
--
Maik Schreiber * http://www.blizzy.de
GPG public key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x1F11D713
Key fingerprint: CF19 AFCE
p (a user is in exactly one group).
Searching in the "permissions" field adds to the score, however, so that more
restrictive documents (having fewer groups in the field) tend to get a higher
score, thus showing up more towards the top of the list. I just don't want
that, though...
Just a quick question: How do I add non-scoring fields to a query? Set
boost to 0?
Yes, just use permissions:blah^0
Cool, thanks.
--
Maik Schreiber * http://www.blizzy.de
GPG public key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x1F11D713
Key fingerprint: CF19 AFCE 6E3D
> You can also use a filter to filter your results. As far as I know
> Filter does not effect the score
Yes, but wouldn't a filter be at least a little slower in this simple case?
Perhaps I should just do a few timing tests...
--
Maik Schreiber * http://www.blizzy.de
GPG publi
>
> [...]
Nice, thanks for sharing the insigt.
--
Maik Schreiber * http://www.blizzy.de
GPG public key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x1F11D713
Key fingerprint: CF19 AFCE 6E3D 5443 9599 18B5 5640 1F11 D713
-
15 matches
Mail list logo
