Andy Armstrong wrote:
I've written a full text indexer that builds a Lucene index. That all seems to work fine. Now the hard part: what should the API be for querying the index? I can't find a spec for XPath full text queries although there is this document:
http://www.w3.org/TR/2003/WD-xquery-full-text-requirements-20030502/
which sets out a few requirements. Any thoughts?
Currently Xindice has two query engines: <resolver class="org.apache.xindice.core.query.XPathQueryResolver"/> <resolver class="org.apache.xindice.core.xupdate.XUpdateQueryResolver"/>
Which are made available to the client via QueryService: XPathQueryServiceImpl XUpdateQueryServiceImpl
Would it make sense to add FullTextQueryServiceImpl, or LuceneQueryServiceImpl, and use Lucene's query syntax to access lucene indexer?
Vadim
