Laurențiu Nicola

I'm not sure what went wrong. Shapefiles with multiple geometries are supported and merged into a `MULTIPOLYGON` geometry which is stored in the database.

I remember a `bindgen` "feature" causing comments to be included or not in the generated bindings based on the location of the GDAL headers.

FWIW, `valgrind` never finishes for me, but LSAN and ASAN don't complain about anything. Tested using: ``` TRYBUILD=overwrite RUSTFLAGS="-Z sanitizer=leak -C linker=clang" RUSTDOCFLAGS="-Z sanitizer=leak -C linker=clang" cargo +nightly test --target...

> Why is trybuild necessary, since it isn't part of https://github.com/georust/gdal/pull/167/files as far as I see? To check that a layer's features borrow the layer, see the single test in...

I tried, but I couldn't reproduce it with a simplified version of the code: ```rust use gdal::vector::FieldValue; use gdal::{Dataset, DatasetOptions, GdalOpenFlags}; use std::path::Path; fn main() { let dso = DatasetOptions...

No, it only works on full tiles.

I'm not sure I can do that, since the data doesn't belong to us. I can try to ask around, but these days we're pretty much swamped.

If you have some shapefiles, I can take a look. Basically, the site can be a single polygon, the attributes (fields) don't matter. For the in-situ data you need a...

MAJA support will be implemented in the next version.

Sorry, I can't say the exact date, but it should be in the following weeks or so.