RE: [sv-ec] multiple non-static array dimensions

Dave writes --

     "There are a number of bugs in the BNF that prevent constructing a
dynamic array that is a member of another array, or even a member of a
structure. Same problem for constructing classes. I think this was just
an oversight."

Dave, Would you please file a Mantis item with representative examples
you believe should be legal, but the current BNF disallows?

Thanks,

-- Brad 

-----Original Message-----
From: owner-sv-ec@eda.org [mailto:owner-sv-ec@eda.org] On Behalf Of
Rich, Dave
Sent: Tuesday, September 18, 2007 5:20 PM
To: Steven Sharp; sv-ec@eda-stds.org; michael.burns@freescale.com
Subject: RE: [sv-ec] multiple non-static array dimensions

There are several mantis items already open on this.

The general consensus has been that an array is a single dimensional
type whose element type many be any type, including another array
(except that a packed array may not contain unpacked types). A
dimensional type is indexing method that could be a fixed-sized,
dynamic, associative or queue type. So a multi-dimensional array is just
a special case of an array whose element type is a similar array
dimensional type.


There are a number of bugs in the BNF that prevent constructing a
dynamic array that is a member of another array, or even a member of a
structure. Same problem for constructing classes. I think this was just
an oversight.


Dave


> -----Original Message-----
> From: owner-sv-ec@server.eda.org [mailto:owner-sv-ec@server.eda.org]
On
> Behalf Of Steven Sharp
> Sent: Tuesday, September 18, 2007 4:38 PM
> To: sv-ec@server.eda-stds.org; michael.burns@freescale.com
> Subject: Re: [sv-ec] multiple non-static array dimensions
> 
> 
> >Are multiple dynamic array dimensions allowed?
> >
> >   int var1[][];
> 
> Yes.
> 
> 
> >I haven't been able to find anything definitive in 1800-2005 or in
draft
> 3a,
> >though 3a does define a dynamic array as "_any_ dimension of an
unpacked
> array
> >whose size can be set or changed at run time"; to me, this statement
> means that
> >a dynamic array does not have to be the only unpacked dimension, but
it
> does
> not
> >necessarily imply that multiple dynamic unpacked dimensions are
allowed.
> 
> The LRM was rather unclear about this.  I misinterpreted it initially 
> myself.
> 
> 
> >If this is allowed, I would expect that the outer array would need to
be
> >"new"ed, then each array-in-array would need to be "new"ed
separately,
> yes?
> 
> Yes.  The outer array is essentially an array of handles to dynamic 
> arrays.
> 
> 
> > And,
> >also that arbitrary combinations of queues and associative arrays
would
> also be
> >allowed as unpacked dimensions.
> 
> Yes.
> 
> 
> Steven Sharp
> sharp@cadence.com
> 
> 
> --
> This message has been scanned for viruses and dangerous content by 
> MailScanner, and is believed to be clean.


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



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