On Tue, 24 May 2011, Christanto Leonardo wrote:
What is the minimum jar required to use the best Tika detection can offer?
My hunch is it'd be tika-core, all the tika-core dependencies,
tika-parsers, poi, and a few bits of commons, but you'd need to do some
tests...
Currently I am using this code to do detection (if this is the best way
to do detection):
Detector detector = new
ContainerAwareDetector(MimeTypes.getDefaultMimeTypes());
Tika tika = new Tika(detector);
String mimeType = tika.detect(TikaInputStream.get(in));
Jukka has done a bit of refactoring, so now I think you can use
CompositeDetector instead of ContainerAwareDetector, and it'll pick up the
container parsers dynamically for you
Nick