Best way to process metadata files with Python on z/OS USS?

Hi everyone,

I’m working on a small internal workflow and wanted to ask for advice from people using Rocket Open-Source Languages and Tools for z/OS.

We have a process where project metadata is exported from a workstation and then transferred to z/OS USS for validation and reporting. The files are mostly CSV and JSON, and they include fields such as project ID, file name, export date, duration, owner, and status. Some of the source files are created after media work is completed in a pc video editor, but the actual task on z/OS is only to validate and process the metadata, not the media files themselves.

I’m planning to use Python on z/OS USS to:

1. Read the uploaded CSV/JSON metadata files

2. Validate required fields

3. Normalize date and filename values

4. Write a clean output file for downstream reporting

5. Log any rejected records with a reason

My questions are:

1. Are there any z/OS USS file encoding issues I should watch for when reading CSV or JSON files created on distributed systems?

2. Is Python the recommended approach for this type of lightweight validation job, or would another tool from Rocket Open AppDev for Z be better?

3. Are there best practices for handling UTF-8 input files and writing logs safely on USS?

4. Would it be better to keep the validation script fully on USS, or trigger it from an existing batch process?

Any guidance or examples from similar metadata-processing workflows would be appreciated.

Thanks!