OpenEuphoria

Re: A question about certain language features

• Posted by Kat <gertie at PELL.NET> Feb 16, 2002
• 590 views

On 16 Feb 2002, at 16:52, rforno at tutopia.com wrote:

> 
> Well, I think it is worse to encounter a variable containing an arbitrary
> default value instead of one you forgot to assign.
> In the second case, you´ll get an execution error just at the point. In the
> first one, you probably will get a difficult to trace error many sentences
> after. 

This would be fine, if the interpreter caught it at compile time, and not after
a
heavily edited file was open and about to be nuked because of the error. If 
the unassigned var was default s={} and i=0,, hmm, a=? 0? ' '? ''? 

Ok, i got it !, make the last include file scan the var list, and assign
whatever
you want to them! Oh wait, no var list,, well, drat. Hey Rob, giving a varlist 
function would bypass this pre-assignment request..

s = varlist() -- nested sequence
?s[1] -- first var declared
 {"junk",'a',12}

procedure AssignDefaultVals()
for each 'a' in s[2] do
  if equal(s[each][1],SpecialCaseValue) 
   then s[each][3] = 1
   else s[each][3] = MyAtomDefaultValue
  end if
end for
end procedure

Or somesuch. Passing the ref would help here, so a new copy isn't made of 
a 200meg database of sequences. Better yet, make it a class? I dunno, it 
was a thought. 

Kat


----- Original Message ----- From: "David Cuny" <dcuny at LANSET.COM> To:
> "EUforum" <EUforum at topica.com> Sent: Friday, February 15, 2002 5:32 AM
> Subject:
> Re: A question about certain language features
> 
> 
> Robert wrote:
> 
> > In situations where it really doesn't matter how
> > something is written, I think there are advantages to
> > reducing the number of choices.
> 
> But in this case, it's not a matter of cosmetics. It actually *does* matter
> how
> (and where) something is written. By splitting the declaration from the
> assignment:
> 
>    integer foo
>    ...
>    foo = 123 -- default value
> 
> you increase the chances that the code will encounter an uninitialized
> variable. And when that happens, the Euphoria interpreter stops, and
> anything
> the user was doing is permanantly lost.
> 
> That's been my experience with Euphoria.
> 
> Isn't the end goal to provide users with robust applications?
> 
> -- David Cuny
> 
> 
> 
>