chicken-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Chicken-users] long pointer equivalent


From: Perry E. Metzger
Subject: Re: [Chicken-users] long pointer equivalent
Date: 27 Aug 2002 08:40:54 -0400
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2

"felix" <address@hidden> writes:
> Right. This should be handled transparently. So should be supply
> an `off_t' type (perhaps better named)? Is there a generic term for
> these platform-specific, large integer types?

No. The C and POSIX standards specify a number of types, which might
be of several different sizes, including things like size_t's,
off_t's, etc.  That's pretty much not something we can avoid easily.

However, what you *do* know is that the underlying types have to be
one of 8, 16, 32 or 64 bit widths (or perhaps even 128 bits -- I don't
know much about what long long's look like on some platforms.)
We also know that the types are fixed on any given
platform. Therefore, it should be possible to determine the signatures
of C functions being interfaced to at build time, and to translate
those signatures into things Chicken can deal with.

What you can't get away from, though, is the need to deal with integral
types of a variety of bit lengths, if not at the Scheme level then in
the foreign function interface.


-- 
Perry E. Metzger                address@hidden
--
"Ask not what your country can force other people to do for you..."




reply via email to

[Prev in Thread] Current Thread [Next in Thread]