> > (for example if you need separate TFs for each document type). > > I wonder if in this precise case it wouldn't be pertinent to > have a single index with the various document types each > having each their own fields set. Isn't TF calculated field by field ?
Oh, you are right :) So i will start testing with one "mixed type" index and perhaps use IndexReaderFactory afterwards in comparison. Thanks, Kai Gülzau
