Hi David James wrote:

Subject: Bug? Variants don't overlay when they contain strings Date: Tuesday, December 28, 1999 11:25 AM

Should the different variants in a variant record overlay in the same memory?

Previous Pascals I have used have guaranteed this, and I've got low-level code that relies on this.

The simple example below shows that this is not true for GPC - the

version

as reported by gpc -v is: Reading specs from /usr/local/lib/gcc-lib/i586-pc-linux-gnu/2.95.2/specs gpc version 19991030, based on 2.95.2 19991024 (release)

The output from the program is:

addr of i is: -1073742896 addr of s is: -1073742760 addr of ic is: -1073742668

program daj15(input,output); type trec = record case integer of 1:(i:array [1..34] of integer); 2:(s:string(80)); 3:(ic:packed array[0..79] of char); end;

var r:trec;

procedure sub(r:trec); begin writeln('addr of i is: ',integer(addr(r.i)):1); writeln('addr of s is: ',integer(addr(r.s)):1); writeln('addr of ic is: ',integer(addr(r.ic)):1); end;

as an aside, if you change the cast "integer" to "cardinal" the output will be a positive number

begin sub(r); end.

If the string s is removed from the record, then the same address is reported for i and ic.

You have more than an overlay problem. In standard pascal you expect ic[ 1 ] to align with the first character position of s and thus one character to the left is the length of s. Thus ic[0] is the length of s.

In extended pascal there is 8 bytes to the left of the first charactor position makingt ic[ 0 ] useless So you end up leaving s out of Trec but what you can do is index s like an array[ 1..CAPACITY] and use Setlength when you need to.

Two different kinds of strings with the same name - string - does make a bit of confusion. Perhaps the oldstyle strings could be renamed "short string" ?? ( didn't think changing the spelling to "stwing" would fly )

Russ russwhit@mind.net