Closed urschrei closed 5 months ago
We now additionally check that neither of the input geometry dimensions are empty for Crosses
and Overlaps
There is what looks to be a straight forward typo in
is_overlaps
, but other than that LGTM.One take it or leave it (truly) style thing.
Thanks for seeing this through!
Many thanks for the reviews. Our long national nightmare is over.
[x] I added an entry to
CHANGES.md
if knowledge of this change could be valuable to users.I want to do a little more checking with regard to the logic we use to determine whether the input dimensions differ (which is a requirement of theCrosses
predicate), and conversely checking whether they're the same (a requirement of theOverlaps
predicate). I'm currently doing this by interrogating the IM for the following two relations:a-interior / b-exterior
anda-exterior / b-interior
. I assert that if they're not equal the input geometry dimensions differ, and vice versa, but I'm not yet 100 % certain that this is true for all valid input geometries.Addressed in https://github.com/georust/geo/pull/1136/commits/af8a50cf1da7ba014b3778751e16c95b1995bfae for all non multi-geometries.