collections/backtracking_utils library

Generic depth-first backtracking solver with pruning and result limits (roadmap #486).

Backtracking explores a tree of partial states: from a state it generates candidate choices, applys each to descend, prunes any branch whose partial state fails isValid before recursing, and records a state that passes isComplete. This is the engine behind constraint puzzles (N-Queens, Sudoku), subset/permutation search, and exact-cover problems — anything where "try, check, undo, try the next" beats enumerating the whole space.

The solver is immutable and stateless between runs: every callback is pure over the State/Choice types you supply, and a fresh State is produced by apply rather than mutating in place, so the same solver can be reused.

Classes

BacktrackingSolver<State, Choice>

A reusable backtracking search over states of type State reached by applying choices of type Choice.