Logic and Complexity of Type Inference in Programming Languages

这是一个详细的研究逻辑和复杂性的类型 推理及其在函数和逻辑程序设计中的应用 语言 类型化lambda演算提供了一个函数式的“最简单” 这样的语言对于这项研究。 计算可以嵌入类型 这门语言的学科，以及关于 自然扩展的可类型性的判定过程可以被推导出来 语言，包括ML式的let多态，定点 操作符和二阶多态lambda演算。 其他 将分析复杂的打字形式，复杂性和统一性， 相关计算模型 中的有界性现象 数据库逻辑程序，并决定 将研究可类型性和实际计算类型。

This is a detailed study of the logic and complexity of type inference, with applications to functional and logic programming languages. The typed lambda calculus provides a functional "simplest" such language for this research. Computation can be embedded in type disciplines for this language, and complexity results concerning the decision process of typability can be derived for natural extensions to the language, including ML-style let-polymorphism, fixed point operators, and second-order polymorphic lambda calculus. Other complex forms of typing will be analyzed, complexity- and unification- related models of computation. The phenomenon of boundedness in database logic programs, and the relationship between deciding typability and actually computing types will be studied.