Hi Toby, Well, the requirement we have is just around text indexing, although I see your point. I'm no expert on this stuff, so I'll get out of my depth pretty quickly (like now!), but how does the lucene stuff in org.apache.lucene.search.highlight do this? Could we allow the client code direct access to these classes?
Miro -- it highly depends on what is index. if it's just simple string properties, an occurrrence (pos+len) indication would surely be helpful for highlighting the results. but what about indexed office documents, pdfs etc? i think it would be of little help for the display-logic to know the positions (bytes, words, pos ?) of that documents. regards, toby On 5/9/06, Miro Walker <[EMAIL PROTECTED]> wrote:
Just a quick question for the list. At the moment, the search capability within JCR doesn't directly support highlighting of matched terms. Does anyone have any ideas about how this functionality (which is available in Lucene) could be exposed through JCR/Jackrabbit? Postprocessing search results in the client app is both inefficient and inaccurate, and ideally we'd want to be able to expose the Lucene functionality through the JCR API. Any ideas? Miro
