Annotation inference for modular checkers
نویسندگان
چکیده
interpretation [2] is a standard framework for developing and describing program analyses. We can view an annotation assistant as an abstract interpretation, where the abstract state space is the power set lattice PG and the checker is used to compute the abstract transition relation. As usual, the choice of the abstract state space controls the conservative approximations performed by the analysis. In our approach, it is easy to tune these approximations by choosing the set of candidate annotations appropriately, provided that this set remains finite and that the annotations are understood by the checker. An interesting aspect of our approach is that the checker can use arbitrary techniques (for example, weakest preconditions in the case of ESC) for performing procedure-local analysis. If these local analysis techniques allow the checker to reason about sets of intermediate states that cannot be precisely characterized using the abstract state space PG, then an annotation assistant may yield more precise results than a conventional abstract interpretation that exclusively uses these abstract states to represent sets of concrete states. The issue of annotation-based program checkers and associated annotation inference algorithms commonly arises in the study of type systems. In contrast to annotation assistants, most type inference algorithms [10,1,9] do not reuse the type checker. In many cases, this reuse may not be possible, for example, because the type checker may not allow multiple type annotation guesses for a given variable declaration. However, for a type checker that does satisfy the properties Ch0 and Ch1 (such as rccjava [7]), the approach outlined in this paper provides a useful method for prototyping a simple, though possibly inefficient, type inference algorithm.
منابع مشابه
Scalable Modular Checking of System-Specific Properties: Myth or Reality?
Annotation-based modular checkers have the potential to perform scalable checking of system-specific properties. However, such tools have seldom been deployed on large software applications of industrial relevance. We present a case study of applying a modular checker HAVOC to check properties about the synchronization protocol of a core Microsoft Windows component with more than 300,000 lines ...
متن کاملTowards Scalable Modular Checking of User-Defined Properties
Theorem-prover based modular checkers have the potential to perform scalable and precise checking of user-defined properties by combining pathsensitive intraprocedural reasoning with user-defined procedure abstractions. However, such tools have seldom been deployed on large software applications of industrial relevance due to the annotation burden required to provide the procedure abstractions....
متن کاملIntra-module Inference
Contract-based property checkers hold the potential for precise, scalable, and incremental reasoning. However, it is difficult to apply such checkers to large program modules because they require programmers to provide detailed contracts, including an interface specification, module invariants, and internal specifications. We argue that given a suitably rich assertion language, modest effort su...
متن کاملCzeSL – an error tagged corpus of Czech as a second language
Using an error-annotated learner corpus as the basis, the goal of this paper is two-fold: (i) to evaluate the practicality of the annotation scheme by computing inter-annotator agreement on a non-trivial sample of data, and (ii) to find out whether the application of automated linguistic annotation tools (taggers, spell checkers and grammar checkers) on the learner text is viable as a substitut...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- Inf. Process. Lett.
دوره 77 شماره
صفحات -
تاریخ انتشار 2001