Is it possible to get a matching percentage from a full text query. I.e.

A boolean query made up of termqueries for a keyword (joined with a SHOULD) and I need a projection that returns (keyword matches in index/keywords in boolean query). If all keywords are found, it will return 1, if 2 out of 5 keywords were found, it would return 0.4, etc.

Is this currently possible?

Hi, you can have something similar using projections:


But this score is not only about the number of matched keywords, look for scoring in Lucene's documentation: it's a pretty complex science and fully customizable.
You might consider the score got from Lucene a better value than 1..5, or you normalize it someway, or you implement your own scoring formula as it can be overriden.
See @Similarity to implement this.