Embed metadata about when/how parquets were generated

[mjcollin]

The parquets should have metadata about when they were generated, data source, citation, perhaps code version, perhaps machine, runtime, etc. so they can stand by themselves instead of relying on things like file name to indicate these things.

Also, an assigned UUID for a GUID too. (Handle?)

[mjcollin]

Looks like not yet, but there is an open ticket to be able to write parquet footer metadata but it has seen no progress in years:

https://stackoverflow.com/questions/42433111/how-to-add-extra-metadata-when-writing-to-parquet-files-using-spark

https://issues.apache.org/jira/browse/SPARK-10803

Possibly write a .metadata full of JSON when writing data sets but how would that interact with the long term plan of naming paths //