Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!cs.utexas.edu!usc!apple!voder!pyramid!prls!philabs!linus!alliant!merk!xylogics!transfer!crackers!m2c!umvlsi!dime!smectos!eli
From: eli@smectos.gang.umass.edu (Eli Brandt)
Newsgroups: comp.lang.pascal
Subject: returning records
Message-ID: <17556@dime.cs.umass.edu>
Date: 27 Jul 90 15:31:17 GMT
References: <950036@hpclapd.HP.COM> <1990Jul25.173213.24892@uwasa.fi> <15365@thorin.cs.unc.edu>
Sender: news@dime.cs.umass.edu
Reply-To: eli@smectos.CS.UMASS.EDU (Eli Brandt)
Organization: University of Massachusetts, Amherst
Lines: 38

In article <15365@thorin.cs.unc.edu> hardarso@arras.cs.unc.edu (Kari Hardarson) writes:
>I have two questions about TP. Is there a good reason why a function
>in TP cannot return a record? Also: is there any way to implement
    (deleted; I can't answer that one)

Well, you don't want to pass records on the stack, but there should be a way to
return pointers to them with having to explicitly reference and dereference them.
One way around this is to return a string of the proper size and then typecast it:

type
  complex = record a, b: real  end;
  cstr = string[11];  { you get element 0, too }

function cAdd(x, y: complex): cstr;
var sum: complex;
begin
  with sum do
  begin
    a := x.a+y.a;
    b := x.b+y.b;
  end;
  cAdd := cstr(sum);
end;

const
  c1: complex = (a:1; b:1);   { sorry, don't remember the format }
  c2: complex = (a:-3; b:2);

var
  c3: complex;

begin
  c3 := complex(cAdd(c1, c2));
end.


Consider this more pseudocode than actual TP code.  I'm just entering this from
memory, but I have used this trick.  I ran it through cpp first to convert cAdd to complex(cAdd) to simplify the syntax.