Hi Chetan, 2014-11-06 10:23 GMT+01:00 Chetan Mehrotra <[email protected]>:
> Hi Tommaso, > > On Thu, Nov 6, 2014 at 2:31 PM, Tommaso Teofili > <[email protected]> wrote: > > I think A and C would be very tricky to solve in OSGi, I > > have dealt myself with this topic before 1.0 and there was no clean > > solution (see also [1]), also because Lucene uses ServiceLoader for some > of > > its configuration points [2][3]. > > From my limited reading of code I think Lucene uses ServiceLoader only > for Codecs. Analyzer can be constructed independently and > programatically used directly while creating Field and on query side > while creating Query terms. So that should work if we are able to get > access to Analyer instances via service registry > yes, that's right, sorry I should've expanded this a bit: I personally think we should expose all/most of Lucene configuration points, including Codecs, so I'd like to take an approach that works for all the configuration bits. Here's an example: since Lucene uses a per-field postings format one could decide to choose a specific format that better addresses its use case so that, for instance, bloom filter postings format works really well with fields having (almost) unique terms so one could choose that for a Lucene property index on jcr:uuid. Regards, Tommaso > > Chetan Mehrotra >
