Linear effects, exceptions, and resource safety: a Curry-Howard correspondence for destructors
Abstract
We analyse the problem of combining linearity, effects, and exceptions, in abstract models of programming languages, as the issue of providing some kind of strength for a monad $T(- \oplus E)$ in a linear setting. We consider in particular for $T$ the allocation monad, which we introduce to model and study resource-safety properties. We apply these results to a series of two linear effectful calculi for which we establish their resource-safety properties. The first calculus is a linear call-by-push-value language with two allocation effects $\mathit{new}$ and $\mathit{delete}$. The resource-safety properties follow from the linear (and even ordered) character of the typing rules. We then explain how to integrate exceptions on top of linearity and effects by adjoining default destruction actions to types, as inspired by C++/Rust destructors. We see destructors as objects $\delta : A\rightarrow TI$ in the slice category over $TI$. This construction gives rise to a second calculus, an affine ordered call-by-push-value language with exceptions and destructors, in which the weakening rule performs a side-effect. As in C++/Rust, a ``move'' operation is necessary to allow random-order release of resources, as opposed to last-in-first-out order. Moving resources is modelled as an exchange rule that performs a side-effect.