Dataset does not load if s3 prefix starts with a digit

[chinmaymore]

Context:
I am publishing OpenLineage events from AWS EMR spark application to Marquez.

Problem:
If a dataset s3 prefix path starts with a digit, the dataset fails to load in Marquez.

[boring-cyborg]

Thanks for opening your first issue in the Marquez project! Please be sure to follow the issue template!