[Enhancement] Fix GEO Functions Logic

[zclllyybb]

Search before asking

• I had searched in the issues and found no similar issues.

Description

Currently, we support a large number of geographic information functions (st_xxx), but some functions have not been correctly implemented with three-dimensional semantics, leading to incorrect or inconsistent calculation results.
This is the initial research: https://selectdb.feishu.cn/wiki/HCJYwb8HVi6k07kKseJcWTDdnbg

Solution

You need to review the current behavior of all geographic information functions and support the correct semantic for two-dimensional/ three-dimensional calculations.

Are you willing to submit PR?

• Yes I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct

[benvdh]

@zclllyybb Is it perhaps an idea too look at GDAL for these implementations? It's licensed under a MIT like License, and has APIs for both Java and C++, and GDAL is already used in many geo-related applications...

It could perhaps speed up the implementation of the geo functions too, and ensure implementations are correct...

[zclllyybb]

@zclllyybb Is it perhaps an idea too look at GDAL for these implementations? It's licensed under the Apache License, and has APIs for both Java and C++, and GDAL is already used in many geo-related applications...

It could perhaps speed up the implementation of the geo functions too, and ensure implementations are correct...

actually we already have some basic libraries like S2 to do the same thing. considering should keep same behaviour with PostGIS, I think we'd better continue finishing our fix base on it.