- - - - -
1.
Compress the files:

$ find ./data -type f -name '*.ZDL' | parallel 'name=$(echo {} | cut -d"/" -f3); ./lzss -e {} -o ./out/$name.out'

Output CSV of original, compressed files:

$ find ./data -type f -name '*.ZDL' | parallel 'name=$(echo {} | cut -d"/" -f3); ls -l {} ./out/$name.out' | awk -F" " '{print $9,$5}' | xargs -n4 | awk -F" " '{print $1 "," $2 "," $3 "," $4 "," $4*100/$2}' > result.csv

Building the Python library:
----------------------------

$ python setup.py build
$ sudo python setup.py install
$ python
>>> import lzss
>>> lzss.decode("out/6322676_4F7DDD79.ZDL.out")

- - - - -
2. to compress:   "lzss -e <input filename> -o <output filename>"
   to decompress: "lzss -d <input filename> -o <output filename>"

- - - - -
3. Note that firmware has their own version of this code checked into their own repository.
   If any changes are made to encoding here then a sample source and encoded version should
   be sent to firmware to make sure that they can still decode.
   Likewise if they change their encode then they should send an encoded file for verification.

4. The file "fw-lzss.c" is the (current) firmware/windows version of lzss.c.
   The file "fw-lzss.c.lzs" is that file LZS encoded by that version.  Our version of lzss
   should be capable of decoding it, resulting in the "fw-lzss.c" file.