Mailing-List: contact user-help@lucenenet.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@lucenenet.apache.org
Received-SPF: pass (nike.apache.org: domain of henko.sch@gmail.com designates
 74.125.82.46 as permitted sender)
MIME-Version: 1.0
From: Schnell Henrik <henko.sch@gmail.com>
Date: Fri, 10 Apr 2015 15:24:48 +0200
Message-ID: 
 <CA+9s_TrvpOkDbYRCxoQ2vkPpmsbWfLqRwXqOvsiw_k-hoimzwQ@mail.gmail.com>
Subject: Classification and WhitespaceAnalyzer problems
To: user@lucenenet.apache.org
Content-Type: multipart/alternative; boundary=f46d043c7fbc7204d505135eb1c4

--f46d043c7fbc7204d505135eb1c4
Content-Type: text/plain; charset=UTF-8

Hello,

I have noticed that you have added Lucene.Net.Classification
implementations, so I thought I would try them with a large corpus to see
how the different algorithms perform with classifying different texts. I
cloned the latest branch from github and opened the solution. I could build
it successfully, so I grabbed the dll's and included them in my project. So
far so good.

But then I noticed that I cannot instantiate an IndexWriter because it
needs an IndexWriterConfig which needs an Analyzer and I could not find any
Analyzer implementations, only the abstract Analyzer class.

Then I have noticed that the WhitespaceAnalyzer.cs is there in the
src\Lucene.Net.Core\Analysis directory but it is not included in the
Lucene.Net project that is in the solution, so it was not built into the
dll's. Ok, so I tried to include all the neccessary files for the
WhitespaceAnalyzer, but they don't build, it seems they use an older lucene
api and are not compatible anymore.

So my question is: how could I try the new Classification features? I'm not
sure I could write an own Analyzer. Are there any working implementations
currently?

Thank you,
Henrik

--f46d043c7fbc7204d505135eb1c4--