Hy,

First of all congratulations for the work done on Hibernate Search!
I work on industrialization of Hibernate Search and this opens up interesting prospects.

Unfortunately, a lot of Analyzer are not compatible.

Does it exist list of compatible Analyzer ?

Regarding the Road Map (Devoxx) : improved integration with Solr.
will you solve this problem ?

Sincerely.

G. S.

Hello,
What would you exactly need from "improved integration with Solr" ?

The Lucene Analyzers have changed interface since version 3.0, especially they are now missing an empty constructor so we have to update the way they are initialized.
Solr is currently having the same problem; the Solr team is working to upgrade to Lucene 3 too but - as of writing - didn't finish yet.
Which current Analyzers are incompatible? The problem is that you can create your own analyzer, we can't predict all needs to initialize them; but you could create a Search friendly constructor.
While working on the new loaders, I'd like to hear from current trouble so we can fix that.

I don't know exactly for "improved integration with Solr", I just readed it on a web site what resume Devoxx's day and Emmanuel Bernard's presentation.

about empty constructor, i see what you say, i had this problem.

I have also a problem with MultiFieldQueryParser and parse method.
the indexation is allright with for example : NGramTokenizerFactory and NGramFilterFactory
(maxGramSize = 30 and minGramSize = 2)

but when i instantiate : MultiFieldQueryParser(fields, fullTextEntityManager.getSearchFactory().getAnalyzer(XXX.class))

the parsing return bad lucene query : xxx:"t e s t te es st t e s t tes est test"

=> problem..

do you work on this type of problems ?

Thx
G.S.