Use pytest+jupytext for notebook testing

[mpharrigan]

Replace the kernel-based notebook execution in execute_notebook() with jupytext-based in-memory conversion and exec(). This avoids parallel kernel startup overhead and eliminates flaky kernel failures

Changes:

• qualtran/testing.py: Rewrite execute_notebook() to use jupytext to convert notebooks to Python scripts in memory, then execute them in a temporary directory.
• dev_tools/check-notebook-tests.py: New CI script that verifies every committed .ipynb (in qualtran/ and tutorials/, excluding dev_tools/) has a corresponding @pytest.mark.notebook test calling execute_notebook(). Prints actionable error messages on failure.
• check/pytest-notebook: New check script that runs the coverage check then pytest -m notebook with MPLBACKEND=agg to prevent interactive matplotlib windows.
• .github/workflows/pr.yaml: The notebooks CI job now runs check/pytest-notebook instead of the old execute-notebooks.py script. The docs build workflow (execute-notebooks.py, notebook_execution.py) is unchanged.
• pyproject.toml: Add jupytext to test dependencies.
• Added @pytest.mark.notebook tests for 61 notebooks that were missing them, plus 3 tutorial notebooks in tutorials/quantum-programming/.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[gemini-code-assist]

Code Review

This pull request replaces the kernel-based Jupyter notebook testing with a faster in-process execution method using jupytext. It adds a verification script to ensure all committed notebooks have corresponding tests and updates numerous test files to include notebook execution checks. Feedback highlights the need for recursive path matching in the notebook discovery script and suggests improvements to the execute_notebook function to handle sys.modules isolation, working directory consistency, and IPython magics.

[mpharrigan]

xref #1745
xref #1807