Freezing Bidirectional Typing (Extended Abstract)
We propose Frost, a novel approach to bidirectional type inference for first-class polymorphism. Conventional bidirectional typing usually considers unidirectional information flow between functions and arguments, typically from the former to the latter. Frost allows local type information to flow back and forth between functions and arguments via pattern inference. A freezing operator is used to control the direction of information flow when there is no best choice. The flexible information flow of Frost enables expressive and predictable inference of polymorphic instantiation and implicit type abstraction. We conjecture that Frost admits a simple, sound and complete type inference algorithm and generalises to type inference for modal types.
Thu 16 OctDisplayed time zone: Perth change
13:45 - 15:30 | |||
13:45 30mTalk | Freezing Bidirectional Typing (Extended Abstract) ML Family Workshop Wenhao Tang The University of Edinburgh, Shengyi Jiang The University of Hong Kong, Bruno C. d. S. Oliveira University of Hong Kong, Sam Lindley The University of Edinburgh | ||
14:15 30mTalk | A typed approach to ontology manipulation (experience report) ML Family Workshop Davide Camino University of Torino, Italy, Andrea Zito University of Torino, Italy, Viviana Bono University of Torino, Lorenzo Bafunno University of Torino, Italy, Lorenzo Pasini University of Torino, Italy, Emanuele Rovaretto University of Torino, Italy | ||
14:45 45mTalk | From CakeML to Proof Checking, and Back AgainInvited Talk ML Family Workshop Yong Kiam Tan Institute for Infocomm Research, A*STAR |