GeezerVison

find_shortest_path_bfs indexed self.graph[value] directly while expanding the BFS queue. Any node that is reachable in the graph but is not present as a dict key — a leaf node with no outgoing edges, which is normal for an adjacency-list graph — raised a KeyError instead of being treated as having no neighbors.

The other traversal methods on this class (find_path_dfs, find_all_paths_dfs, find_shortest_path_dfs) already handle this correctly via self.graph.get(start, []). This brings find_shortest_path_bfs in line with them.

Repro before the fix:

graph = {"A": ["B"], "B": ["C", "D"]}  # C and D are leaf nodes, no key in graph
GraphSearch(graph).find_shortest_path_bfs("A", "Z")  # KeyError: 'D'

Added a regression test (tests/test_graph_search.py) that fails before this change and passes after.

Ran the full suite locally with the new test passing. Also ran codespell, flake8, isort, black --check, and mypy on the changed files with no issues.