[BUG] set_cover Python: fix UB in vector transforms and swapped GuidedTabuSearch bindings

[jg-codes]

What

Fixes three bugs in the pybind11 wrapper (ortools/set_cover/python/set_cover.cc), and adds regression tests for each.

Disclosure: This PR was drafted with AI assistance (Claude). I reviewed and tested all changes.

Bug 1 & 2: undefined behavior in std::transform on empty vectors

VectorIntToVectorSubsetIndex and the all_subsets property both call std::transform writing to subs.begin() / subsets.begin() on a default-constructed (empty, capacity 0) vector. The output iterator is invalid from the start — this is undefined behavior per the C++ standard.

Impact: Every Python call that passes a focus list hits VectorIntToVectorSubsetIndex: next_solution(focus) on all heuristic classes, compute_coverage_in_focus, both clear_random_subsets overloads, and both clear_most_covered_elements overloads. Accessing model.all_subsets from Python triggers the second instance. In practice, the write corrupts heap metadata. I observed crashes on macOS (heap-buffer-overflow) when calling next_solution with a focus list.

Alternatives considered:

• subs.resize(n) + write to begin() — correct, but default-constructs every element before overwriting. For SubsetIndex the cost is negligible, but reserve + back_inserter avoids even that.
• Ranged constructor std::vector<SubsetIndex>(ints.begin(), ints.end()) — not viable because the element types differ (BaseInt → SubsetIndex), so a transform is needed.

Fix: reserve(n) + std::back_inserter(). This is the same pattern already used in the columns and rows properties in the same file.

Bug 3: swapped getter/setter for GuidedTabuSearch::lagrangian_factor

Lines 558–560 bind the Python name "set_lagrangian_factor" to the C++ method GetLagrangianFactor and vice versa. Calling gts.set_lagrangian_factor(0.5) from Python returns a value instead of setting one; calling gts.get_lagrangian_factor() requires an argument and mutates state.

The neighboring bindings (epsilon, penalty_factor, tabu_list_size) are wired correctly — only the lagrangian pair is transposed.

Fix: swap the two Python method name strings.

Tests added

Test	Exercises
test_all_subsets_property	all_subsets property (Bug 2)
test_focus_based_next_solution	next_solution(focus) path through VectorIntToVectorSubsetIndex (Bug 1)
test_guided_tabu_search_lagrangian_factor	set/get_lagrangian_factor round-trip (Bug 3)

Checklist

• Read CONTRIBUTING.md and PR template — targeting main as instructed
• CLA signed
• Minimal diff — only the bug fixes and their tests
• No formatting or refactoring changes
• AI-assisted: disclosed above

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.