fix: prevent undefined behavior in subgraph::global_to_local(edge_descriptor) when edge not in subgraph

[Becheler]

Fixes bug #124

Before submitting

• This PR targets the develop branch.
• I searched for an existing PR or issue covering the same change.
• My contribution is licensed under the Boost Software License 1.0.

Type of change

• Bug fix
• New feature or API addition
• Refactor (no behavior change)
• Documentation
• Build, CI, or tooling
• Other (specify below)

Does this PR introduce a breaking change?

• Yes (describe migration impact below)
• No

What this PR does

• Commit 1: document safe utilities (find_vertex and find_edge) and lock in invariants without touching header.
• Commit 2: Add the sad-path test exercising global_to_local on an edge not in the subgraph. Red on CI (memory corruption / garbage instead of a benign default).
• Commit 3: Fix subgraph.hpp (global_to_local delegates to find_edge + assert) turning CI error into a loud meaningfull assertion failure that mirrors find_vertex
• Commit 4 adapts the test to check find_edge first, CI green
• I rebased on develop so I lost the per-commit actual CI green/red checks :'(

Motivation

In #124 it was shown that subgraph::global_to_local(edge_descriptor) is only defined when the input edge is actually in the subgraph. For edges not in the subgraph, the behavior is undefined (it dereferences end()). But its signature, name, and documentation all present it as being well-defined irrespective of the input.

Testing

Added a new subgraph test file. There are currently 5 subgraph_XXX.cpp files that should be tightened in a single test file, but it's outside the scope of this first PR.

Note: Although not a breaking change, someone will notice the change if in a debug build their code that previously invoked the UB and happened to get away with it (garbage returned but never actually used): this will now abort at the assertion instead of running away silently.

Checklist

• Existing tests pass (b2 in the test/ directory).
• New behavior is covered by a test, or this is a docs / build / refactor change.
• Documentation was updated if user-facing behavior changed.
• No new compiler warnings on the platforms I built against.

[github-actions]

Compiler-warning counts vs develop (auto-generated).
PR run 29330400514 vs develop run 29329400156 (31187f2195).

Job	Baseline	After	Delta
macos (clang, 14)	4	4	0
macos (clang, 17)	4	4	0
macos (clang, 20)	4	4	0
ubuntu (clang-19, 14)	4	4	0
ubuntu (clang-19, 17)	4	4	0
ubuntu (clang-19, 20)	4	4	0
ubuntu (clang-19, 23)	4	4	0
ubuntu (gcc-14, 14)	10	10	0
ubuntu (gcc-14, 17)	10	10	0
ubuntu (gcc-14, 20)	10	10	0
ubuntu (gcc-14, 23)	10	10	0
windows_msvc_14_3 (msvc-14.3)	971	971	0

[cppalliance-bot]

An automated preview of the documentation is available at https://524.graph.prtest3.cppalliance.org/index.html

If more commits are pushed to the pull request, the docs will rebuild at the same URL.

2026-07-14 11:55:25 UTC

[Becheler]

Current CI error relies on caller ability to identify undefined behavior:

subgraph_global_local.cpp(123): test 'sg.global_to_local(e01) == edge_t()' failed in function 'void test_absent_edge_returns_default() [Directedness = boost::directedS]'
1 error detected.

[Becheler]

Error is now precise and loud in debug mode:

subgraph_global_local: ../../../boost/graph/subgraph.hpp:237: edge_descriptor boost::subgraph<boost::adjacency_list<boost::vecS, boost::vecS, boost::directedS, boost::property<boost::vertex_color_t, int>, boost::property<boost::edge_index_t, unsigned long, boost::property<boost::edge_weight_t, int>>>>::global_to_local(edge_descriptor) const [Graph = boost::adjacency_list<boost::vecS, boost::vecS, boost::directedS, boost::property<boost::vertex_color_t, int>, boost::property<boost::edge_index_t, unsigned long, boost::property<boost::edge_weight_t, int>>>]: 
Assertion `(in_subgraph)&&("global_to_local: edge is not in this subgraph. " 
"Use find_edge() to check membership first.")' failed.
Aborted (core dumped)

[github-actions]

Boost dependency footprint vs develop (auto-generated).
PR run 29330400599 vs develop run 29329400139 (31187f2195).

Header-inclusion weights (graph files pulling each direct dependency in):

No header-inclusion-weight changes.

Transitive Boost modules: 70 → 70 (0)

[jeremy-murphy]

Congratulations on understanding this class, it's bewildering!