Hello everybody.
I have not a usuall question about Lucene. I complete my science work which connected with Markov productions and their accelerations. And have the last pity question in this work: I need to invent how to use my results in document manipulations (so nobody interesting in core of research, everybody want to get some unreal results in my papers of using this invention in real projects - such country, such education system unfortunatly :( ). So now I`m tring to get something where I can insert (or replace real working good code with bad and slow code with production logic) some production logic in programm and then accelerate it (to show how good it is!!!pitty once again :((( ) and this something should be connected with document or symbol manipulation with open source. Firstly I found some DMS, and find out that Lucene is open source search engine is used in some of them. I looking to its source code to find out some thing that I can use for my science report but can`t find anything still. May be somebody can help me? Or may be you give me a piece of advise what programm I can look for my needs.

A small introduce to Markov productions to show you what I`m looking for. Markov producion is:
A -> B
It means that we change all ocurences of 'A' in word will change to 'B'.So in usual java it`s work like 'replace', but replace once scans work and replace all 'A'`s but don`t llok again word if we have new occurencece of 'A', but markov production will look throught word again and again until there will not be 'A'. I give yu an example. Production is:
ab -> bc
word is: aabDFGab
is we use replace we will get: abcDFGbc, but take a look, after first replacement we have a new occurence of ab. So Markov production will start look throught word once again and at finish we get: bccDFGbc.

So any tasks in lucene where I can try to apply productions, something where there a lot of replacements?