Hey all,
I'm somewhat new to Lucene. Meaning I used it some time ago for a parser we
wrote to tokenize a document into word grams.
the approach I took was simple as follows:
1. extended the lucene Analyzer
2. In the tokenStream method use ShingleMatrixFilter. Passed in the
standard tokenizer,
In the second case I guess that searching might be faster by the odd
fraction of a millisecond, but any affect will likely be dwarfed by
most of the stuff on the Wiki page mentioned before.
--
Ian.
On Thu, Feb 17, 2011 at 12:56 PM, suman.holani wrote:
> Hi ,
>
> Thanks its useful.
>
> One thing
Hi ,
Thanks its useful.
One thing , if I reduce the number of fields by setting them INDEX.NO will
it affect searching speed.But yes ,I will be storing those fields in index.
Let say there are 5 fields A,B,C,D,E
---with first index having all fields config as STORE.YES and
INDEX.ANALYZED
---a
http://lucene.apache.org/java/3_0_3/fileformats.html will tell you all
you need to know about what is stored where and how.
In general, the speed of searching i.e. finding matching docs will not
be affected by the number of stored fields but retrieving data from
lots of stored fields will certainl
Try different boost values with larger gaps between low and high. If
that doesn't help, post a tiny but complete self-contained example
that demonstrates the problem. And you should always say what version
of lucene you are using.
--
Ian.
On Thu, Feb 17, 2011 at 7:56 AM, Akos Tajti wrote:
>
Hello,
I am little confused on the stored and index part of lucene
How it actually stores the indexed field and stored field
Is it that for every field indexed , all the store fields added .I mean do
we create diff indexes for every indexed field ,replicating the stored field
in each of th
