🚀 go-pugleaf

In article <1992Dec16.164821.19079@walter.bellcore.com> dnew@thumper.bellcore.com (Darren New) writes:

			.................

>> 	use_double(c,d)

>> where, for example, c and d would be 32-bit objects and double is 64-bit.

>This is a trivial modification, using a union with two 32-bit objects
>overlapping one 64-bit object.

>However, C cannot choose different functions based on that their value
>is returned as.  What should it do if you say

>process(use_double(c,d))

>and you haven't declared what types "process" uses returns.  It's a tiny
>amount of work to use different functions depending on what you want to
>pass and what you want back, compared to actually getting it to work.

My idea about "use" is much simpler.  The expression use_double(...), as
I intend it, means that the bit pattern, or concatenated bit pattern, 
described in the parentheses is to be now used as a double number in
any application.  Therefore, in   process(use_double(...)), process just
assumes that what is presented to it is a double, just as the cast to
double declares that the result is double.
-- 
Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907-1399
Phone: (317)494-6054
hrubin@snap.stat.purdue.edu (Internet, bitnet)  
{purdue,pur-ee}!snap.stat!hrubin(UUCP)

hrubin@pop.stat.purdue.edu (Herman Rubin) writes:

>The expression use_double(...), as
>I intend it, means that the bit pattern, or concatenated bit pattern, 
>described in the parentheses is to be now used as a double number in
>any application.  Therefore, in   process(use_double(...)), process just
>assumes that what is presented to it is a double, just as the cast to
>double declares that the result is double.

I don't know why I bother, but here it is anyway...

In Gnu C:
	#define use_double(var) ({ \
		union { typeof(var) old; double new; } tmp; \
		tmp.old = var; \
		tmp.new \
	})

In C++:
	template<class T>
	double use_double(T var) {
		union { T old; double new; } tmp;
		tmp.old = var;
		return tmp.new
	}

Satisfied, Mr. Rubin?

-- 
Fergus Henderson             fjh@munta.cs.mu.OZ.AU      
This .signature virus is a self-referential statement that is true - but 
you will only be able to consistently believe it if you copy it to your own
.signature file!