What if you were to make your field a multi-valued field, and at indexing time, split up the text into sentences, putting each sentence into the solr document as one of the values for the mv field? Then I think the normal highlighting code can be used to pull the entire value (i.e. sentence) of a matching mv instance within your document? I.E. put the 'overhead' into the index step, rather than trying to do it at search time?
Bob Sandiford | Lead Software Engineer | SirsiDynix P: 800.288.8020 X6943 | bob.sandif...@sirsidynix.com www.sirsidynix.com > -----Original Message----- > From: Koji Sekiguchi [mailto:k...@r.email.ne.jp] > Sent: Monday, September 05, 2011 10:33 AM > To: solr-user@lucene.apache.org > Subject: Re: Sentence aware Highlighter > > (11/09/05 23:09), O. Klein wrote: > > Using the regex in the old highlighter I had reasonable sentence > aware > > highlighting, but speed is an issue. So I tried to get this working > with the > > VFH, but this obviously didn't work with the regex. > > > > So I am looking for ways to get the same behavior but with improved > speed > > and came across https://issues.apache.org/jira/browse/LUCENE-1824, > which at > > least would be a small improvement, but the last comment confused me, > as I > > thought FVH was going to be the new highlighter for Solr. So this > patch > > would make some sense if im not mistaken. > > > > Nonetheless has anyone managed to make something like a > > SentenceAwareFragmentsBuilder? Or have some advise on how to realise > this? > > Sorry for the long delay on the issue! > I'd like to take a look into it in this week. Hopefully, BreakIterator > may be > used, which Robert mentioned in the JIRA. > > Thank you for your patience! > > koji > -- > Check out "Query Log Visualizer" for Apache Solr > http://www.rondhuit-demo.com/loganalyzer/loganalyzer.html > http://www.rondhuit.com/en/
