docs: add stub pages for notebooks relocated by PR #4617

[AnitaGeorge404]

PR #4617 removed application notebooks from the docs without leaving any pointer at their old paths, so those URLs now 404. Ten of the removed notebooks have a confirmed home in the NVIDIA/cuda-q-academic repository (five relocated as-is, five re-authored or expanded into teaching modules). Add an orphan RST stub at each old path linking to its new location, and add matching "moved" notices to applications.rst at the positions the original notebook cards occupied.

The remaining seven notebooks removed by PR #4617 have no identified counterpart and are left untouched.

Summary

PR #4617 removed 17 application notebooks from docs/sphinx/applications/python/ without leaving any pointer at their old paths, so those URLs now return 404s (#5386). This PR restores a useful landing page at each old path for the notebooks that have an identifiable successor, instead of a dead link.

Changes

Ten of the seventeen removed notebooks have a confirmed successor in the NVIDIA/cuda-q-academic repository. For each, this PR adds an :orphan: RST stub at the notebook's original path (so the old URL builds to a real page again, without being added to any toctree) that links to the new location, plus a matching "moved" notice in applications.rst at the position the original notebook card occupied:

Relocated as-is (same notebook, moved):

• quantum_teleportation → cuda-q-academic/qis-examples/quantum_teleportation.ipynb
• bernstein_vazirani → cuda-q-academic/qis-examples/bernstein_vazirani.ipynb
• quantum_fourier_transform → cuda-q-academic/qis-examples/quantum_fourier_transform.ipynb
• deutsch_algorithm → cuda-q-academic/qis-examples/deutsch_algorithm.ipynb
• grovers → cuda-q-academic/qis-examples/grovers.ipynb

Re-authored / expanded into a teaching module (equivalent current material, not a verbatim copy):

• adapt_vqe → cuda-q-academic/chemistry-simulations/adapt_vqe.ipynb
• qm_mm_pe → cuda-q-academic/chemistry-simulations/qmmm.ipynb
• hybrid_quantum_neural_networks → cuda-q-academic/quantum-machine-learning-and-data-analysis/01_an_introduction_to_hybrid_quantum_neural_networks.ipynb
• unitary_compilation_diffusion_models → cuda-q-academic/ai-for-quantum/01_compiling_unitaries_diffusion.ipynb
• quantum_pagerank → cuda-q-academic/quantum-machine-learning-and-data-analysis/04_quantum_pagerank.ipynb (identified during this PR's investigation; same PageRank-via-quantum-stochastic-walks topic, now expanded into a fuller QML module)

Also fixes a latent bug that these additions would otherwise have triggered: applications.rst has an inline script and filter.js that both call .split(',') on every .notebook-entry's data-tags attribute, unconditionally on page load and again on every filter click. A .notebook-entry without data-tags throws and breaks tag rendering/filtering for the whole page. Each new "moved" notice carries the same data-tags its original entry had, and a small custom.css rule collapses the now-unused image column for these text-only cards.

Validation

• Verified all ten destination paths exist in NVIDIA/cuda-q-academic (main) by cloning the repository and checking the files directly, plus a spot-check that the GitHub blob page for one of them (quantum_teleportation.ipynb) renders and not a 404.
• Parsed each new .rst file with docutils — no errors.
• Ran a scoped sphinx-build -b html -n -W (nitpicky, warnings-as-errors) against the applications.rst page tree, including the new stubs and all notebooks it links to: build succeeded with zero warnings.
• Confirmed each of the ten old paths (e.g. applications/python/quantum_teleportation.html) builds to a real page and is not referenced by any toctree (orphan behavior confirmed).
• git diff --check — no whitespace errors.
• Reviewed git status/git diff to confirm only the intended 12 files changed (10 new stubs, applications.rst, custom.css).

I was not able to run the repository's full scripts/build_docs.sh, since it builds the CUDA-Q C++/Python package from source and executes every notebook in the documentation (GPU-dependent), which is outside what this environment can do. The scoped Sphinx build above covers the actual pages this PR touches.

Scope

Of the seventeen notebooks removed by PR #4617, seven are left untouched because no defensible successor could be found:

• digitized_counterdiabatic_qaoa, cost_minimization, divisive_clustering_coresets, edge_detection, vqe_advanced — no matching content found anywhere in cuda-q-academic or the current docs.
• qaoa and adapt_qaoa have only partial thematic overlap with newer cuda-q-academic notebooks (e.g. a QAOA max-cut lab, and an Adapt-QAOA section embedded inside a larger, differently-scoped module) — not clean 1:1 replacements, so adding a redirect for them would be misleading. Left out of this PR pending maintainer input on whether either is an intended replacement.

Separately, while investigating, I noticed docs/sphinx/applications/python/uccsd_wf_ansatz.ipynb and generate_fermionic_ham.ipynb still contain markdown links to the now-dead vqe_advanced.html. That's an existing dangling-link issue in files unrelated to this PR's scope, so it isn't fixed here.

Related issue

Closes #5386

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.