feat: argparse for validate.py; pretty JSON output in schema generator

[SoundMatt]

Closes #101

Changes

• validate.py: Replace bare sys.argv indexing with argparse. Running without arguments now prints a proper usage message instead of throwing IndexError. The --quiet flag is preserved as a named argument.

• ifex_to_json_schema.py: Capture all print() output via io.StringIO / redirect_stdout, then parse and re-emit with json.dumps(indent=2). The schema is now pretty-printed without needing an external pipe step.

• pretty_print_json.py: Deleted. Its sole purpose was piping the schema generator output through json.dumps; that is now done internally.

Test

# Usage message instead of IndexError
python -m ifex.output_filters.schema.validate

# Pretty-printed schema directly
python -m ifex.output_filters.schema.ifex_to_json_schema | head -20

[gunnar-mb]

Good fix, but CI/CD must be updated since these files use pretty_print_json.py‎:

• .github/workflows/create-new-release.yml
• .github/workflows/buildcheck.yml

Also since JSON publication is part of release flow, test first.

[SoundMatt]

Thanks for the review! Pushed a fix for the CI gap — create-new-release.yml and buildcheck.yml were both still calling pretty_print_json.py (now deleted), piping ifex_to_json_schema.py's output through it. Since ifex_to_json_schema.py pretty-prints its own output now, both workflows just write straight to ifex-core-idl-schema.json in one step.

Re: testing the release flow — happy to do a dry run if there's a way to trigger create-new-release.yml without cutting an actual release, otherwise let me know what you'd like to see before merging.