On 1/1/2013 1:03 PM, dafna wrote:
hi everyone,
I have a system written in .net that I'm using lucene .net.
I'm thinking of changing that system to be based on solr instead of lucene.
My situation is that I have a lot of code that was written in .net, that I
don't want to waste because moving to solr makes me to change code to java.
I have an idea, but I want to make sure is OK to do that:
I want to keep the .net code as it is. The only change will be the write to
the index and the read from the index.
I want to translate (by object serialize or hard coded serialize) the Query
in search side & document in index side to text, and then call the http
interface from the .net code.
Is that idea can work ? what am I losing with that solution ?
My concern is about the analyzers. Is it ok to call the .net analyzers (in
search & index), and then let solr get the data without any data handling ?
Have you seen the SolrNet project?
http://code.google.com/p/solrnet/
Moving to SolrJ and Java would probably be better, but if you'd rather
stick to .NET, you have options.
All of the analyzers, tokenizers, filters, and similar such definitions
are configured on the server side, not in the client.
Thanks,
Shawn
