It has been discussed few times - you need to implement own Similarity, which will write number of tokens as a norm during indexing, and then in query time you can check the norm value per document. You can also do it on a more straightforward way: preprocess docs to derive a number_or_colors field, eg. via UpdateProcessor and filter for this field as usual.
On Wed, Jan 16, 2013 at 10:18 PM, Nguyen, Vincent (CDC/OD/OADS) (CTR) < v...@cdc.gov> wrote: > Hi, > > How do I find documents that have more than one value in a field? > > Example: > > <doc> > <arr name="color"> > <str>blue</str> > <str>red</str> > </arr> > </doc> > > Vincent Vu Nguyen > > > -- Sincerely yours Mikhail Khludnev Principal Engineer, Grid Dynamics <http://www.griddynamics.com> <mkhlud...@griddynamics.com>
