Develop binary-search strategies for length exploration

Develop binary-search strategies for DueList’s length-exploration procedure by supplying the semantic monotonicity analysis required to determine when such strategies are applicable.

Background

DueList’s feedback loop iteratively refines length abstractions by excluding list lengths that the under-approximating solver has shown to be unsatisfiable. The paper notes that the current exploration strategy delegates the order in which lengths are considered to the underlying SMT solver, which may explore lengths in non-increasing order.

An increasing-length strategy could guarantee discovery of shortest lists in satisfiable instances, but might fail to exploit inferred length information. The authors identify binary search as an alternative and leave its development unresolved because applying it would require semantic analysis of monotonicity.