Those who aren't subscribed to solr-dev may be interested to know that the lucene highlighter has been integrated into Solr, for both the standard request handler, and the dismax handler.
See the highlight, highlightFields, and maxSnippets params documented here: http://wiki.apache.org/solr/StandardRequestHandler When highlighting is enabled, a new "highlighting" section is added to responses. Highlighted terms are marked up like so: <em>term</em> Below is a sample for query terms ipod, USB In the real response, the markup tags <em></em> are XML escaped so each highlight snippet is a single string. <lst name="highlighting"> <lst name="IW-02"> <arr name="features"> <str>car power adapter for <em>iPod</em>, white</str> </arr> <arr name="name"> <str><em>iPod</em> & <em>iPod</em> Mini <em>USB</em> 2.0 Cable</str> </arr> </lst> <lst name="MA147LL/A"> <arr name="features"> <str>playback, Upgradeable firmware, <em>USB</em> 2.0 compatibility, Playback speed control, Rechargeable capability</str> </arr> <arr name="name"> <str>Apple 60 GB <em>iPod</em> with Video Playback Black</str> </arr> </lst> </lst> The Lucene version in Solr was just updated to fix a highlighting bug with overlapping tokens, so you may want to wait for the 7/16 nightly solr build if you use WordDelimiterFilter. -Yonik
