Since FAST computes a document signature for every document it indexes, is it possible to query FAST for all the documents having at least one duplicate in the index?
If yes, what would be the FQL query for that?
Duplication is handled query time, so there is no property which will allow you to count or filter on this. But I agree getting a list of all duplicates would be nice to clean up data.
If you query with trimming you will get a count per item letting you know the duplicates (fcount I think..but not 100% sure), but you would need to iterate all results to get this.