Hi,
I've read in Lucene In Action about how Nutch collates the results from several IndexSearcher machines. Part of Lucene's scoring is idf. How does an individual IndexSearcher machine know a term's idf if it doesn't have access to the whole index? Thanks.
