I seem to recall that the Highlighter in Solr is pluggable, so you may
want to work at that level instead of the client side. Otherwise, you
likely would have to implement your own TermVectorMapper and add that
to the TermVectorComponent capability which then feeds your client.
For an example of using TermVectorMapper, but not solving exactly your
problem (but close), see http://www.lucidimagination.com/blog/2009/05/26/accessing-words-around-a-positional-match-in-lucene/
but note that is at the Lucene level.
On Jul 13, 2009, at 2:37 PM, Walter Ravenek wrote:
Hi all,
When I'm using the TermVectorComponent I receive term vectors with
all tokens in the documents that meet my search criteria. I would be
interested in getting the offsets for just those terms in the
documents that meet the search citeria. My documents are about 200 K
and are in XML. If I have just the offsets for the hits, I can
easily implement my own highligting on the client side.
Does anyone know how to go about doing this?
--------------------------
Grant Ingersoll
http://www.lucidimagination.com/
Search the Lucene ecosystem (Lucene/Solr/Nutch/Mahout/Tika/Droids)
using Solr/Lucene:
http://www.lucidimagination.com/search
