RE: [sv-ec] Mantis item 978 has been updated - ready for an email vote

Comments:

1. The new proposal changes the 3rd paragraph to 5.15 to:

"The optional 'with' clause accepts an expression enclosed in
parentheses. This expression should not include any
side effects; if it does, the results are unpredictable. In contrast,
the 'with' clause used by the randomize method (see 13.6)
accepts a set of constraints enclosed in braces."

The sentence "This expression should not include any
side effects; if it does, the results are unpredictable." is not very
standards-like. It would be better perhaps to write it as 
"If this expression includes any side effects, the results may be
unpredictable." Yes, I know this sentence was just moved from 5.15.1.


2. After inserting this new 2nd sentence, the phrase beginning the 3rd
sentence, "In contrast," no longer makes sense, as it refers to the 1st
sentence, not to the immediately preceding sentence as "in contrast"
should. Actually, I don't think the 3rd sentence is needed at all.


3. Can an array manipulation method have more than one argument? The
text implies not. Why then does Syntax 5-3 use list_of_arguments instead
of a single argument?


4. Why are array locator methods not allowed on packed arrays?


5. What does the unique() method do? Does it mean that it returns only
elements whose value appears only once? E.g, if the array elements are
'{0,1,2,1} then it would return 0 and 2.

Or does it mean that element values which appear multiple times are
returned only once? E.g, the previous example would return 0, 1, and 2.

Until now, I assumed the latter, but I suddenly saw that the description
is ambiguous.


6. Assuming the latter, the description of unique_index() should specify
that it is unspecified which of multiple indexes with the same element
value is returned.


7. Why are array reduction methods not allowed on packed arrays?

Thanks,
Shalom

> -----Original Message-----
> From: owner-sv-ec@server.eda-stds.org [mailto:owner-sv-ec@server.eda-
> stds.org] On Behalf Of Neil Korpusik
> Sent: Tuesday, July 11, 2006 11:01 PM
> To: SV_EC List
> Subject: [sv-ec] Mantis item 978 has been updated - ready for an email
> vote
> 
> Mantis item 978 has been updated, based on the feedback from the July
> 10th
> meeting, as well as a bugnote filed by Shalom. There were a few more
> syntax
> errors that were corrected. I also moved the paragraph from 5.15.1 to
> 5.15
> which discusses the optional array iterator argument since it applies
to
> all of the array manipulation methods.
> 
> I believe that mantis 978 is now ready for an email vote.
> 
> 
> After looking over this section in more detail I have the following
two
> questions:
> 
> 1. 5.15.2 Array ordering methods
> 
>    reverse() may be used on packed or unpacked arrays.
>    sort() and rsort() may only be used on unpacked arrays.
>    The LRM doesn't mention if shuffle() may be used on packed or
> unpacked
>    arrays.
> 
>    Should shuffle() only be allowed on unpacked arrays?
> 
> 2. 5.15 Array manipulation methods
> 
>    Is it an error to specify the name of an iterator when there is no
>    corresponding 'with' clause? Syntax box 5-3 implies that it is
legal.
> 
> 
> There are 4 other mantis items that deal with this same section of the
> LRM.
>    1038  5.15.2 Array ordering methods - only 1-D arrays?
>    1437  Prototype syntax for array locator methods in 5.15.1 is not
> legal
>    1490  Array ordering methods on associative arrays
>    1517  Array reduction methods used in constraints
> 
> 
> 
> --
> ---------------------------------------------------------------------
> Neil Korpusik                                     Tel: 408-720-4852
> Senior Staff Engineer                             Fax: 408-720-4850
> Frontend Technologies - ASICs & Processors (FTAP)
> Sun Microsystems
> email: neil.korpusik@sun.com
> ---------------------------------------------------------------------