Pdal-parallelizer

Changing the code structure of the project and also the point clouds processing approach. Taking image_array for each point cloud instead of an entire file.

The unmanaged memory generated by a task is now suppressed when the task is finished.

- New merge_tiles option to merge all your tiles at the end of an execution. It takes into account compression, minor_version and dataformat_id specified in your tiles pipelines
- New remove_tiles option to delete all your tiles after merged

- Create _temp_ and _output_ folders if they don't exist
- Add a merge_tiles option to merge all the resulting chunks into one file
- Trigger warnings
- if the _tile_size_ value is not changed
- if the number of workers exceed the existing number of CPU avalaible
- If the _tile_size_ value exceeds the dimensions of the input cloud, it will now be adapted to the actual boundaries of the cloud.

Fix import bug. The API is now functional

New API for pdal-parallelizer :

- The process_pipelines function can now be used directly via Python

- fixing memory leaks