Declare defaulted move ops noexcept (SonarQube BLOCKER cpp:S5018)

[rgsl888prabhu]

SonarQube rule cpp:S5018 flags = defaulted move constructors and move-assignment operators that arent declared noexcept. This PR adds noexcept to the seven flagged sites on main, clearing 7 of 8 BLOCKER findings.

All affected classes already move trivially-noexcept members (std::vector, std::unique_ptr, rmm::device_uvector, raw pointers, scalars), so the specifier compiles cleanly. None of these four types are currently stored by value in an STL container (mip_node_t is always behind unique_ptr or raw pointer; the other three dont appear in containers), so on the current code this is a runtime no-op — it just satisfies the lint rule.

[coderabbitai]

📝 Walkthrough

Walkthrough

This PR standardizes move special member declarations by adding explicit noexcept to move constructors and move assignment operators in four types while keeping them defaulted.

Changes

Move Special Members Exception Specification

Layer / File(s)	Summary
Add noexcept to move special members cpp/include/cuopt/linear_programming/optimization_problem.hpp, cpp/include/cuopt/linear_programming/pdlp/pdlp_warm_start_data.hpp, cpp/src/branch_and_bound/deterministic_workers.hpp, cpp/src/branch_and_bound/mip_node.hpp	Move constructors and move assignment operators in optimization_problem_t, pdlp_warm_start_data_t, deterministic_diving_worker_t, and mip_node_t are updated from defaulted-but-not-noexcept to explicitly noexcept = default.

---

🎯 1 (Trivial) | ⏱️ ~3 minutes

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 10.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately summarizes the main change: adding noexcept to defaulted move operations to resolve SonarQube blocker cpp:S5018 across multiple files.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description check	✅ Passed	The pull request description is clearly related to the changeset, explaining the specific SonarQube rule being addressed and the rationale for adding noexcept to defaulted move operations.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/sonar-blocker-noexcept-moves

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[mlubin]

so the specifier is a no-op at runtime — it just lets the rule pass and, where it matters, lets STL containers use move-not-copy when these types are stored

Sorry which one is it? Does it change the behavior of STL containers when these types or stored or is it a no-op?

[rgsl888prabhu]

Sorry, that was contradictory wording on my part. Quick clarification:

In general, noexcept on a move ctor IS observable: std::vector (and friends) reallocate via std::move_if_noexcept, so adding noexcept lets reallocations move elements instead of copy them. So it's not strictly a runtime no-op — but only when the type is stored by value in such a container.

For these four types specifically, I checked the codebase:

• optimization_problem_t, pdlp_warm_start_data_t, deterministic_diving_worker_t — never stored by value in any STL container.
• mip_node_t — only ever held as std::unique_ptr<mip_node_t> or raw mip_node_t* in containers (std::vector, std::deque, std::list); the pointer types' moves are noexcept regardless of mip_node_t's declaration.

So in practice on the current code, the change really is a runtime no-op; it only satisfies the lint rule and would start mattering if a future change put one of these in a container by value. I've also dropped the misleading paragraph from the PR description.