Details
-
Improvement
-
Status: Closed
-
Major
-
Resolution: Fixed
-
None
-
None
-
None
-
New
Description
When indexing lines or polygons into a shape field, we generally index multiple points for the same document. When querying, we typically visited a subset of this points. In many cases when visiting points from the same document, we already know the relationship and therefore computing the relationship between that point and the query shape is not necessary.
When using dense visitors(eg. visitors backed by a FixedBitSet), we can check if the relationship is already known and therefore skip that point.
Finally for intersects relationship we normally use sparse visitors but I wonder in the case where the number of points >> number of docs, we should use a dense visitor so we can skip points from documents where we know it intersects.
Attachments
Issue Links
- links to
