Create tutorials for the docs

[efaulhaber]

This has been on my to-do list for a long time. PointNeighbors.jl should be easy to use, but just from the docstrings, it is very difficult to figure out how it works.

This PR

• adds several tutorials to demonstrate usage,
• adds GPU tests that (so far only) test the GPU tutorial on GPUs in CI,
• throws an error if the search_radius is an Integer. This error can happen easily and was previously throwing a cryptic "trying to convert float to int" exception. I don't want to add a hard dispatch to AbstractFloat because theoretically, this might work with more special custom data types. Just not with integers.
• makes ParallelUpdate the default strategy for the FullGridCellList with DynamicVectorOfVectors backend. This is a change that I wanted to do for a long time now. ParallelUpdate is faster in most applications, especially on GPUs. Furthermore, it supports initialization on GPUs, whereas the previous default required updating the NHS on the CPU before adapting it to the GPU.

Find the rendered docs here: https://trixi-framework.github.io/PointNeighbors.jl/previews/PR153/

Closes #124.

[Copilot]

Pull request overview

This PR improves the usability of PointNeighbors.jl by adding a new Tutorials section to the documentation (generated from Literate.jl sources) and by adding a constructor guard against integer search_radius types that would lead to incorrect/failed distance computations.

Changes:

• Add four new worked tutorials (basic usage, n-body, periodicity, GPU usage) and a Tutorials index page in the docs.
• Integrate Literate.jl generation into docs/make.jl and add required docs build dependencies.
• Add an Integer-typed search_radius validation guard to multiple neighborhood search constructors.

Reviewed changes

Copilot reviewed 10 out of 10 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
src/nhs_trivial.jl	Reject integer search_radius in TrivialNeighborhoodSearch constructor to avoid integer distance conversions.
src/nhs_precomputed.jl	Reject integer search_radius in PrecomputedNeighborhoodSearch constructor for the same reason.
src/nhs_grid.jl	Clarify n_points docstring wording; reject integer search_radius in GridNeighborhoodSearch constructor.
docs/src/tutorial.md	Adds Tutorials landing page with links to each tutorial.
docs/Project.toml	Adds Literate, plus tutorial-imported deps (Adapt, KernelAbstractions) to the docs environment.
docs/make.jl	Runs Literate conversion into docs/src/tutorials and adds Tutorials to the Documenter page structure.
docs/literate/src/tut_basic_usage.jl	New basic fixed-radius neighbor search tutorial source.
docs/literate/src/tut_n_body.jl	New n-body cutoff-radius example tutorial source.
docs/literate/src/tut_periodicity.jl	New periodic domain tutorial source using PeriodicBox.
docs/literate/src/tut_gpu_usage.jl	New GPU usage tutorial source (KernelAbstractions/Adapt-based workflow).

---

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

[codecov]

Codecov Report

❌ Patch coverage is 57.14286% with 3 lines in your changes missing coverage. Please review.
✅ Project coverage is 86.00%. Comparing base (11a445c) to head (829990e).

Files with missing lines	Patch %	Lines
src/nhs_grid.jl	50.00%	1 Missing ⚠️
src/nhs_precomputed.jl	50.00%	1 Missing ⚠️
src/nhs_trivial.jl	50.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #153      +/-   ##
==========================================
- Coverage   86.06%   86.00%   -0.06%     
==========================================
  Files          15       15              
  Lines         739      743       +4     
==========================================
+ Hits          636      639       +3     
- Misses        103      104       +1     

Flag	Coverage Δ
unit	86.00% <57.14%> (-0.06%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[efaulhaber]

/run-gpu-tests