Hello Lucene developers and users. I'm currently researching
I'd like to extend an information that Explanation class provides. This
class currently provides only score computation for document or query.
Particularly I'd like to include the following information to it:
- concrete document fields, that contain searched tokens
- their number in the document founded
- document length
- some other custom information
It would be great if this information, describing matching process would be
As I can see the good place for this information is Explanation class
(explain parameter for Elasticsearch). This class is final currently and
its constructor is private. So it is not possible to extend it. I think it
would be right to make it extendable at least because Similarity mechanism
I imaging that it would be very comfortable to call the query to
Elasticsearch with "explain" parameter and get additional information
there, that would be filled by custom Similarity algorithm or plugin.
If "explain" is not appropriate place for it - could you advise me a better
way to extend/describe matching/querying documents by concrete query?