Re: [sv-ec] Mantis 958: array query functions

In reviewing 958 for the vote, I'm realizing that it perpetuates the 
fallacy that there is a unique size to a dynamically sized dimension. 
Any of the functions that take a dimension expression cannot in general 
return a meaningful result when referring to a dynamically sized 
dimension whose dimension expression is not 1:

bit [1:0][] a;
a[0] = new[4];
a[1] = new[37];

$right(a,2) = ???
$size(a,2) = ???

Granted, this is not the problem that 958 is addressing, but it's pretty 
close. Is this the sort of thing that can be filed as an errata and 
fixed after our deadline?

Mike

Jonathan Bromley wrote:
> Shalom,
>
>   
>> Re 7.5.1:
>>
>> "The number of elements in the array. Must be a non-negative integral
>> expression. The type of this operand is longint. It shall be 
>> an error if
>> the value of this operand is negative. If this operand is zero, the
>> array shall become empty so that its size
>> will be reported as zero by its .size method (see 7.5.2)."
>>
>> The second sentence, "Must be a non-negative integral expression," was
>> deleted by Mantis 1427.
>>
>> The end, "so that its size will be reported as zero by its 
>> .size method (see 7.5.2)," is not necessary.
>>     
>
> Agreed; I'll fix both.  Thanks.
>   


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.