Implement topological equality check for IntersectionMatrix #1133
Conversation
|
@michaelkirk There are some cases that will be useful for this in the |
|
All the files in textxml/general are parsed and processed. We only skip the operations we don't know how to support which may live across files or be a subset of a particular file. This seemed like the best way to get as much test coverage as possible. I hope you don't mind, but I pushed a commit to your branch that prints a little bit more about the stats of how many test cases are being skipped and why in jts_test_runner. You can see a lot more detail (too much detail probably) if you run with Feel free to remove it if you don't want to include it in your branch. |
|
Just to be clear - the tests in TestRelateLC and TestRelateLL are already being run. You can convince yourself by watching the tests fail after you tweak the expected output. |
|
Ahhh OK. Thank you! I'll have a further poke around. |
|
re: the CI failure, I'll open a PR soon that bumps our MSRV to 1.70 (>6 months old) and updates the "latest" containers. |
| self.0[CoordPos::Inside][CoordPos::Inside] != Dimensions::Empty | ||
| && self.0[CoordPos::Inside][CoordPos::Outside] == Dimensions::Empty | ||
| && self.0[CoordPos::Outside][CoordPos::Inside] == Dimensions::Empty | ||
| && self.0[CoordPos::Outside][CoordPos::OnBoundary] == Dimensions::Empty |
There was a problem hiding this comment.
I think we need one more condition:
&& self.0[CoordPos::OnBoundary][CoordPos::Outside] == Dimensions::Empty
There was a problem hiding this comment.
Damn, thought I had them all. I looked at the other is_ methods for inspiration, but evidently got a bit confused. Which matrix entry does it correspond to?
There was a problem hiding this comment.
TBH, I firstly just intuited that any checks for an "is equal" relationship should be symmetrical - and noticed they weren't.
Then I noticed there are 5 constraints in the de-9im spec string, but you only had 4 comparisons and deduced which non-symmetrical check was missing.
But to try to answer your question directly, I think it's like this:
Spec for "is topo equal": [T*F**FFF*]
| Inside_a | OnBoundary_a | Outside_a | |
|---|---|---|---|
| Inside_b | T | * | F |
| OnBounary_b | * | * | F (missing) |
| Outside_b | F | F | * |
Corresponding Indexes into de-9im string
| Inside_a | OnBoundary_a | Outside_a | |
|---|---|---|---|
| Inside_b | 0 | 1 | 2 |
| OnBounary_b | 3 | 4 | 5 (missing) |
| Outside_b | 6 | 7 | 8 |
So we were missing a check for the 6th element (which has an index of 5).
[T*F**FFF*]
[ ^ ]
CHANGES.mdif knowledge of this change could be valuable to users.Fixes #1132