LazFS filesystem overview
--------

LazFS filesystem is just a "wrapper". It is indended to store data in LiDAR
format. LiDAR files are transparently compressed via laszip algorithm when
application saves LiDAR data and are transparently decompressed when application
accesses the LiDAR file.

Compression ratio is around 1:8 to 1:10 so 100GB of uncompressed LiDAR data
occupies around 10 - 15 GB of disk space.

LazFS treats all files with ".las" suffix as LiDAR files so attempting to store
file with ".las" suffix which isn't a LiDAR file ends with error.

Please note that LazFS is currently experimental filesystem and it
_shouldn't_ be used in production. Make sure you have backup of all your LiDAR
data!

LazFS requirements
--------

LazFS must run on top of the filesystem which supports extended attributes. It
was tested on XFS and Ext4.

LazFS internals
--------

LazFS is mostly stateless filesystem. All stored data can be transparently
accessed directly on the underlying filesystem. Currently it stores only one
information per LiDAR file in extended attribute - decompressed file size to
speed up *stat() calls.

When application accesses a LiDAR file, it is decompressed into /tmp/ in open()
syscall and application works directly with file in /tmp/. After it closes() it,
decompressed file is removed. When files are created, they are also created and
accessed in uncompressed form via /tmp/ and are compressed when they are closed.

Example of usage
--------

./lazfs -o suid,allow_other,default_permissions <backend_dir> <target_dir>

After this call, you can work with <target_dir> and LiDAR data will be stored in
compressed form in <backend_dir>.