Hi, It's with regards to generating a query string for Solr. I am looking for a solution where I can create the query string like
'q=name:ipod AND cat=music AND features=cool&facet=on&facet.field=cat' I understand, we may not be able to use lucene query API directly . Is there any other library through which we can create this query string. I came across Solrj but looks like it is not as reach as lucene API. Is there a way we can use lucene API and generate the query string. Thanks in Advance, Ashwin On Mon, Jun 24, 2013 at 7:56 AM, Erik Hatcher <erik.hatc...@gmail.com>wrote: > Inline below... > > On Jun 24, 2013, at 07:16 , heikki wrote: > > > hello, > > > > I'm a long-time user of Lucene, and have some questions about SOLR. > > > > 1. Is it possible to give actual Lucene queries to SOLR, bypassing any > > SOLR-side QueryParsing ? > > No, not directly as Query objects, but Solr's default query parser is > lucene syntax based, so any Lucene QueryParser syntax query can be used in > Solr. > > Since you've got Lucene API know-how, you could create a custom > QParser(Plugin) easily if you have a different syntax/format that you'd > like to use. > > > 2. Are there differences in functionality or implementation between > Faceted > > Search in Lucene and SOLR ? > > They are two entirely different codebases, no overlap whatsoever. I'm > sure the basic results are comparable, but I don't have experience with the > Lucene faceting module so I can't say what's different. > > > 3. Is it possible to use several indices (for different languages), and > > sometimes search only in one of them, sometimes (if user requested) > search > > in all of them having combined results (and relevancy calculated as if > it is > > a single index) ? > > Yes. You can create multiple Solr "cores" and query them individually or > as if they were one distributed collection (provided the schemas are > compatible). Regarding relevancy though, there still is an open issue to > implement distributed IDF, so it's not exactly as if it was a single index. > With different languages, you'll be using different fields anyway though, > so maybe the distributed IDF isn't an issue for you. > > But you could index all the documents, possibly, into the same index and > have a field that distinguishes the language and filter (or not) to query > in any combination. > > Erik > >
