Make kernel_name override update notebook kernelspec metadata

[huyhoang171106]

Summary

The current execution flow appears to pass kernel_name to the execution engine, but does not reliably persist that override into the output notebook metadata. As a result, notebook consumers (e.g., JupyterHub UI) still show the original kernelspec.display_name and/or kernelspec.name. This file should be updated so that when kernel_name is provided, the in-memory notebook object has its metadata rewritten before output is saved.

Files changed

• papermill/execute.py (modified)
• papermill/cli.py (modified)
• papermill/tests/__init__.py (modified)
• papermill/tests/test_cli.py (modified)

Testing

• Not run in this environment.

What does this PR do?

Fixes #<issue_number>

Closes #856

[Copilot]

Pull request overview

This PR addresses issue #856 by ensuring that when a kernel_name override is provided to papermill (including via the CLI), the output notebook’s metadata.kernelspec is rewritten so downstream notebook consumers display the overridden kernel.

Changes:

• Add logic to persist kernel_name into nb.metadata.kernelspec.name and display_name before writing the output notebook.
• Thread kernel_name into prepare_notebook_metadata(...) so the rewrite happens consistently (including --prepare-only).
• Add a CLI test asserting the output notebook metadata reflects the overridden kernel.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 1 comment.

File	Description
papermill/execute.py	Introduces metadata rewrite helper and passes kernel_name into metadata preparation.
papermill/cli.py	Updates --kernel help text to reflect overriding notebook metadata.
papermill/tests/init.py	Adds a test constant for a kernel override name.
papermill/tests/test_cli.py	Adds a regression test validating kernelspec metadata is updated in the output notebook.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

_override_kernel_name_in_metadata assumes nb.metadata['kernelspec'] is a mapping; if a notebook has kernelspec=None (or any non-dict), this will raise when assigning ['name']/['display_name']. Consider normalizing kernelspec to an empty dict when it's missing or not a dict-like object before writing into it.

Suggested change

	kernelspec = nb.metadata.get('kernelspec', {})
	kernelspec = nb.metadata.get('kernelspec')
	if not kernelspec:
	kernelspec = {}
	elif not isinstance(kernelspec, dict):
	# Normalize non-dict, mapping-like objects to a plain dict; fall back to empty dict otherwise.
	try:
	kernelspec = dict(kernelspec)
	except TypeError:
	kernelspec = {}