Inspired by Sahin Buyrukbilen's question here:
I made a simple read/write codec that stores all postings data into a
single text file (_X.pst), looking like this:
The codec is fully funtional – all Lucene & Solr tests pass with
-Dtests.codec=SimpleText – but, its performance is obviously poor.
However, it should be useful for debugging, transparency,
understanding just what Lucene stores in its index, etc. And it's a
quick way to gain some understanding on how a codec works...