Webknossos

- Added option `make_relative: bool` to `wkcuber.api.dataset.Dataset.add_symlink_layer` to make the symlink relative. [360](https://github.com/scalableminds/webknossos-cuber/pull/360)

- Added `add_copy_layer()` to `wkcuber.api.dataset.Dataset` to copy the layer of a different dataset. [345](https://github.com/scalableminds/webknossos-cuber/pull/345)
- Added `View.read_bbox()` which takes only a single bounding box as parameter (instead of an offset and size). [347](https://github.com/scalableminds/webknossos-cuber/pull/347)

Breaking Changes in Config & CLI
- Some breaking changes in the dataset API: [339](https://github.com/scalableminds/webknossos-cuber/pull/339)
- The interfaces of the methods `Layer.add_mag` and `Layer.get_or_add_mag` have changed: the parameter `block_type` is now replaced with `compress`.
- Previously `Layer.mags` was of type `Dict[str, MagView]`. This was now changed to `Dict[Mag, MagView]`.
- Renamed `LayerTypes` to `LayerCategories`.

Added
- Added multiple small features: [339](https://github.com/scalableminds/webknossos-cuber/pull/339)
- Names of datasets can now be passed optionally when creating a dataset.
- The `Layer` does now expose the `largest_segment_id`.
- Add methods to get category specific layers for a given dataset.

Breaking Changes in Config & CLI
- Refactoring of the dataset API: [331](https://github.com/scalableminds/webknossos-cuber/pull/331)
- Tiff-support is now dropped (`TiffDataset` and `TiledTiffDataset` are now removed (alongside their corresponding `Layer` and `MagDataset` classes))
- Module names are now lowercase (previously: `wkcuber.api.Dataset`, `wkcuber.api.Layer`, `wkcuber.api.View`, `wkcuber.api.properties.DatasetProperties`, `wkcuber.api.properties.LayerProperties`, `wkcuber.api.properties.ResolutionProperties`)
- Some classes are renamed (`WKDataset` -> `Dataset`, `WKMagDataset` -> `MagView`)
- The "Layer types" (previously `Layer.COLOR` and `Layer.SEGMENTATION`) are now moved into their own class `wkcuber.api.layer.LayerTypes`.
- `View` (in particular `get_view()`) is refactored to be safer (this is also a breaking change).
- The attribute `path_to_mag_dataset` was renamed to `path_to_mag_view`
- Changes for `View.get_view()` (these changes also apply for `MagView.get_view()` (previously `MagDataset.get_view()`)):
- The parameter `relative_offset` was renamed to `offset`.
- The parameter `is_bounded` was dropped (`View`s are now always bounded).
- The order of the parameters `size` and `offset` was changed, so that `offset` is now the first parameter.
- The shorthand `wkcuber.api.dataset.Dataset.get_view()` was removed.
- The flag `--write_tiff` of `convert_nifti` was removed.

Fixed
- Use an os independent path separator for regexes. [334](https://github.com/scalableminds/webknossos-cuber/pull/334)

- Added distribution flags to wkcuber in order to limit the parallel jobs or use a cluster scheduler.
- Added a function for converting BoundingBox to BoundingBoxNamedTuple.
- Fixed a bug for writing compressed data. This previously caused an error when downsampling certain datasets.