Hum! It seems to be exactly what I need. Thanks! I'll look for it in the docs.
Rafael Calsaverini Data Scientist @ Catho <http://catho.com.br> cell: +55 11 7525.6222 * * *8d21881718d00d997686177be1c27360493b23ea0258f5e6534437e6* On Wed, Aug 21, 2013 at 12:08 PM, Erick Erickson <erickerick...@gmail.com>wrote: > Have you tried "nested queries"? I think you can specify the full edismax > syntax on the URL in a nested query for your second search field... > > Best, > Erick > > > On Tue, Aug 20, 2013 at 5:17 PM, Rafael Calsaverini < > rafael.calsaver...@gmail.com> wrote: > > > Hi there, > > > > > > suppose I have documents with fields that can be loosely grouped by what > > kind of information they contain. As an example, imagine the documents > > representing people and some fields are related to how they're called > > (name, surname, nickname, etc...) and some fields are related to where > they > > live (street, city, state, ...). > > > > Imagine I have an application that provide two search fields to look for > > people: who do you want to find, and where does he live. So I just get > two > > strings, and each of them might contain information related to multiple > > fields. > > > > Is there a way to do something like edismax for each group of field? > > Something like: > > > > (name^2 surname^2 nickname):(rafael calsaverini) AND (street city > > state):(rua dos bobos sao paulo SP) > > > > or whatever is the adequate syntax. > > > > The alternative would be to build a parser for one of the fields, but if > I > > want to allow for a completely free field for the "where" part, it might > > not be easy to predict what the user would put there. Is there anybody > else > > with a similar problem? > > > > Thanks. > > > > > > > > Rafael Calsaverini > > Data Scientist @ Catho <http://catho.com.br> > > cell: +55 11 7525.6222 > > > > * > > * > > *8d21881718d00d997686177be1c27360493b23ea0258f5e6534437e6* > > >
