home *** CD-ROM | disk | FTP | other *** search
- Path: keats.ugrad.cs.ubc.ca!not-for-mail
- From: c2a192@ugrad.cs.ubc.ca (Kazimir Kylheku)
- Newsgroups: comp.lang.c
- Subject: Re: Formal type system and denotational semantics for C?
- Date: 5 Mar 1996 07:27:56 -0800
- Organization: Computer Science, University of B.C., Vancouver, B.C., Canada
- Message-ID: <4hhmhsINNqcj@keats.ugrad.cs.ubc.ca>
- References: <313C52B1.41C67EA6@cs.wisc.edu>
- NNTP-Posting-Host: keats.ugrad.cs.ubc.ca
-
- In article <313C52B1.41C67EA6@cs.wisc.edu>,
- Michael Siff <siff@cs.wisc.edu> wrote:
- >Is anyone aware of any work done for either a formal type system for C
- >or denotational semantics for C? I am particularly interested in the
- >type question. I am looking for type rules of the form:
- >
- >A |- e : t
- >-----------
- >A |- &e : t ptr
- >
- >meaning: if with assumptions A (i.e. declarations and initial
- >environment) expression e has type t, then &e has type pointer to t.
- >
- >Any information concerning the existence of such rules would be much
- >appreciated.
-
- There is the ANSI/ISO standard, but which doesn't go into this level of
- formalism. It's quite a human-readable document. :) It does state _some_ things
- formally using grammar notation to define the syntax of the language, but you
- are asking for all kinds of semantic notions to be codified also.
- --
-
-