add check for invalid geometries before st_filter

[SarahAlidoost]

What type of PR is this? (check all applicable)

• Refactor
• Feature
• Bug Fix
• Optimization
• Documentation Update

Description

Related Issues

• See this comment
• Closes Use of the S2 back-end results in invalid geometries when rertrieving buildings and water surfaces #115

Added/updated tests?

We encourage you to keep the code coverage percentage at 75% and above.

• Yes
• No, and this is why: please replace this line with details on why tests
  have not been included
• I need help with writing tests

Added entry in changelog?

For user-facing changes, add a line describing the changes in NEWS.md

• Yes

[fnattino]

@SarahAlidoost just a heads up that the Windows R-CMD-check is broken because of #126, so we currently expect it to fail

[fnattino]

Thanks a lot @SarahAlidoost for this! As I tried to explain in #115, I think that even working with GEOS vs s2 would not help as invalid polygons would remain as such.
So I think we really have to use sf::st_make_valid().

Maybe this PR could still be useful as it is good to receive the warning if some of the retrieved geometries are invalid. I would however drop the sf::st_use_s2() and re-add instead a call to sf::st_make_valid()..

[SarahAlidoost]

Thanks a lot @SarahAlidoost for this! As I tried to explain in #115, I think that even working with GEOS vs s2 would not help as invalid polygons would remain as such. So I think we really have to use sf::st_make_valid().

Maybe this PR could still be useful as it is good to receive the warning if some of the retrieved geometries are invalid. I would however drop the sf::st_use_s2() and re-add instead a call to sf::st_make_valid()..

thanks for checking this. I applied your suggestion by introducing a util function.

[fnattino]

Sorry for the late reply - this is good to go, thanks @SarahAlidoost!