BUG: `object_catalog`.`length` may be the length of the zipfile and not the extracted DCM file.

[dmoore247]

Environment
AWS / Azure / GCP

Cluster configuration JSON:

Fully reproducible code snippet
select length, local_path from hls_radiology.tcia.object_catalog and compare with original zip files and dcm files.

** Problem Description **
Dashboards, size calculators rely upon object_catalog.length to represent the individual file. However I found that value represents the size of the original zip file (in the case where the automation unzips and then catalogs the table)

[dmoore247]

Need to update blog https://www.databricks.com/blog/7x-faster-medical-image-ingestion-python-data-source-api where it describes storage size savings.

[erinaldidb]

Quick note to fix this, we have api that extract cloud file metadata without opening it, this can be a good approach to extract the file size without reading it entirely. If it's a zip then returning the size metadata from the zip while decompressing it is the way to go.