Also dispatch GI.isgeometry on Type.

[evetion]

This fixes a small oversight from #290.

[rafaqz]

Was wondering what was going on

[rafaqz]

Well, theres still something odd going on with the tests.

[evetion]

Yep, it's because GDAL v1.4 😢 @visr

[rafaqz]

Pinning 1.3 works for me, that will let my other PRs get merged as well.

[visr]

What was up with GDAL.jl v1.4? If it is just the regular #158 we shouldn't pin it. I tested ArchGDAL locally before making the GDAL release, I believe it was just that.

[rafaqz]

I'm not sure its just that? There are a lot of nothing return values and a method error:

https://github.com/yeesian/ArchGDAL.jl/runs/6903399033?check_suite_focus=true#step:20:296

[evetion]

See https://github.com/yeesian/ArchGDAL.jl/runs/6903398191?check_suite_focus=true

More tests were failing than just a tolerance it seemed, not sure why. And pinning untill we make a release that works with 1.4 seems right?

[evetion]

Failures are due to OSGeo/gdal@1164ba2#diff-360c35bb4393b380a99c8d258b1a8f20dbcc9634e4ccb6e22fe7ad82593892e9L405.

So the manual flag checks implemented in #290 to fix #303 are now not needed anymore.

[visr]

Ha nice find. For all these bool integer returns I guess we need to test > 0.
At least GDAL.jl has a new testset now: JuliaGeo/GDAL.jl@4f96fac