HiNice to have again new update of BaseX.
In past I had some performance problems, when I had to process one document, enriching the elements by data from another database.
I have a traffic-feed document, containing elements loc-id, referencing to detailed location description.
Detailed location description is located in another document, which we call PredefinedLocationSet. In some cases we have about 11000 locations or even more, total size of open PredefinedLocationSet document is about 8 MB. Traffic feed can in some cases be of similar nature (the same number of "records" and the same size).
My task was to return (not to modify) traffic feed with detailed location description embedded.
All my xqueries, trying to accomplish something like that, seemed to me rather slow, regardless of my attempts to find any sort of optimization.
I wonder, if map:map approach is something, what has a chance to be faster. Any technical background, hints or ideas are welcome.
My idea is to start my xquery by building up map from PredefinedLocationSet (using loc-id as the key) and then run through traffic-feed to get loc-id, find related location details and return enriched "record".
With best regards
Jan Vlcinsky