Publications:ML-like Inference for Classifiers

Jump to: navigation, search

Do not edit this section

Keep all hand-made modifications below

Title ML-like Inference for Classifiers
Author Cristiano Calcagno and Eugenio Moggi and Walid Taha
Year 2004
PublicationType Conference Paper
Conference ESOP'04. European Symposium on Programming
Diva url
Abstract Environment classifiers were proposed as a new approach totyping multi-stage languages. Safety was established in the simply-typedand let-polymorphic settings. While the motivation for classifiers was thefeasibility of inference, this was in fact not established. This paper startswith the observation that inference for the full classifier-based systemfails. We then identify a subset of the original system for which inferenceis possible. This subset, which uses implicit classifiers, retains significantexpressivity (e.g. it can embed the calculi of Davies and Pfenning) andeliminates the need for classifier names in terms. Implicit classifiers wereimplemented in MetaOCaml, and no changes were needed to make anexisting test suite acceptable by the new type checker.