Concrete Type Inference: Delivering Object-Oriented Applications

نویسندگان

  • Ole Agesen
  • David M. Ungar
چکیده

types describe properties or invariants of objects. For example, Emerald’s [13] interface types list a set of operations that objects must support and the types of arguments and results of the operations. Abstract types characterize the externally observable behavior of objects and need not distinguish between different implementations of the same behavior. Returning to our example, both ListStack and ArrayStack instances may have an interface type like Stack_IF = [push:elmType→void; pop:void→elmType]†. Stack_IF specifies that objects have a push and a pop operation, but leaves unspecified how these operations are implemented. Any object that defines a push and a pop operation, regardless of how these are implemented, has the abstract type Stack_IF. Smalltalk and Self, while having no linguistic notion of type, nevertheless employ interface types: dynamic type checking in these languages entails verifying that objects “understand” (i.e., define a method for) the messages that they receive. In fact, interfaces are so important that they are introduced on the very first page of Goldberg and Robson’s Smalltalk description [46]. Many statically-typed object-oriented languages, including Simula, Beta, C++, and Eiffel use class types‡. The property that a class type expresses is: “being an instance of the given class or one of its subclasses.” Thus, referring again to the classes in Figure 4, the class type Stack includes instances of the classes Stack, ListStack, and ArrayStack, whereas the class type ArrayStack includes ArrayStack instances only. In fact, for a given program (under the closed-world view discussed in Section 2.1.4), a class type can always be expanded to a concrete type, e.g.: class type Stack = {Stack, ListStack, ArrayStack} class type ArrayStack = {ArrayStack}. We place class types between the concrete types and abstract types in Figure 3. First, class types are not fully abstract: they partially reveal the implementation of objects. For example, if class Stack defines an instance variable size, it can be asserted that any object of class type Stack has this instance variable††. Beta’s “inner” mechanism for method combination [77] likewise implies that methods defined in class Stack extend to any object of class type Stack. Second, class types are not fully concrete; knowing a class type for an object only implies that the specified † This example is not in Emerald syntax. ‡ Other statically-typed languages, including Strongtalk [19] and TOOPL [20], use variations of interface types. †† However, a few languages permit subclasses to override instance variables with get/set methods. Stack ListStack ArrayStack Sequence

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Incremental Inference of Concrete Types

Concrete type information is invaluable for program optimization. The determination of concrete types is, in general, a ow sensitive global data ow problem. As a result, its solution is hampered by the very program structures for whose optimization its results are most critical: dynamic dispatch (as in object-oriented programs) and rst class functions (including function pointers). Constraint b...

متن کامل

Modular Concrete Type-inference for Statically Typed Object-oriented Programming Languages ?

The problem of concrete type-inference for statically typed object-oriented programming languages (e.g., Java, C ++) determines at each program point, those objects to which a reference may refer or a pointer may point during execution. We present a new technique called analysis-using-abstract-values which performs modular and demand-driven concrete type-inference of a robust subset of Java wit...

متن کامل

Complexity of Concrete Type-Inference in the Presence of Exceptions

Concrete type-inference for statically typed object-oriented programming languages (e.g., Java, C ++) determines at each program point, those objects to which a reference may refer or a pointer may point during execution. A precise compile-time solution for this problem requires a ow-sensitive analysis. Our new complexity results for concrete type-inference distinguish the diiculty of the intra...

متن کامل

Declarative Semantics in Object-Oriented Software Development - A Taxonomy and Survey

One of the modern paradigms to develop an application is object oriented analysis and design. In this paradigm, there are several objects and each object plays some specific roles in applications. In an application, we must distinguish between procedural semantics and declarative semantics for their implementation in a specific programming language. For the procedural semantics, we can write a ...

متن کامل

Dynamic vs. Static Optimization Techniques for Object-Oriented Languages

Object-oriented programs can be optimized either dynamically, i.e., based on run-time information, or statically, i.e., based on program analysis alone. Two promising optimization techniques for object-oriented languages are type feedback (dynamic) and concrete type inference (static). We directly compare the two techniques, evaluating their effectiveness on a suite of 23 SELF programs while ke...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1995