Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!iuvax!purdue!mentor.cc.purdue.edu!pur-ee!pur-phy!murphy
From: murphy@pur-phy (William J. Murphy)
Newsgroups: comp.lang.pascal
Subject: Re: complex numbers
Message-ID: <2361@pur-phy>
Date: 14 Jul 89 18:13:52 GMT
References: <2751@ohstpy.mps.ohio-state.edu> <2360@pur-phy>
Reply-To: murphy@newton.physics.purdue.edu.UUCP (William J. Murphy)
Organization: Purdue Univ. Physics Dept., W. Lafayette, IN
Lines: 44

In article <2360@pur-phy> hal@newton.physics.purdue.edu.UUCP (Hal Chambers) writes:
>In article <2751@ohstpy.mps.ohio-state.edu> heff@ohstpy.mps.ohio-state.edu writes:
>>. . . to write equations in source code such as
>>                     a =: b + c;
>>                     a := (b+c)/(d+e);
>>as can be done in fortran.  For that matter, are there any languages beside
>>fortran which allow easy computation of complex numbers?
>
>Although Pascal (and Modula-2, etc.) allows the definition of a Complex
>type,  there is no way for the programmer to define the operators
>other than as explicit procedure calls (functions can't return a
>structured type).
>
>So the examples above would appear as something like:
>    >   a =: b + c;
>		cadd(a, b,c);
>
>    >   a := (b+c)/(d+e);
>		cadd(c1, b,c);
>		cadd(c2, d,e);
>		cdiv(a, c1,c2);
>Ugly!!!!
This is one approach, another which works and is not quite so ugly is
to create a memory pool for temporary calculation and use pointers
into that pool.  This way you can write
    TempPtr := cdiv( cadd(^b, ^c), cadd(^d, ^e) );
    a.x     := TempPtr^.x;
    a.y     := TempPtr^.y;

Don't flame the syntax, I haven't done this for a month at least.
The point is that you can use pointers to the variable rather
than passing a bunch of values and then you must be certain to 
reassign the values in the TempPtr into the original variable.

The pool is treated as a circular buffer which has an arbitrary
size (read big enough to fit your longest expression) I use a buffer
of 150 double precision complex numbers. 

If you are interested in a copy of the Turbo-Pascal source that
I have, send e-mail to me. I feel that it is too long to post on the
net.

Bill Murphy
murphy@newton.physics.purdue.edu