Re: bareword test on ebcdic.

--- Yitzchak Scott-Thoennes <[email protected]> wrote:

> On Thu, Jul 28, 2005 at 12:35:13AM -0700, rajarshi
> das wrote:
> > Nicholas Clark wrote:
> >> If you put those 3 bytes directly between the '{'
> and '}' characters in
> >> the EBCDIC version of that 1 liner, does it also
> print 3500?
> 
> > I am unable to put those three bytes in the
> 1-liner you mentioned above, since I am unable to
> print the chars corresponding to those bytes
> (www.kostis.net/charsets/ebc1047.htm) on the command
> line. 
> 
> >> I think that the regression tests tended to do
> something like
> >> 
> >> if (ord 'A' == 65) {
> >> # Do the ASCII/UTF-8 version
> >> } else {
> >> # Assume EBCDIC
> >> }
> 
> I tried to fix the attribution above; apologies if I
> got it wrong.
> 
> I think the way you want to test this is something
> like:
> 
>   $key = "\x{0442}\x{0435}\x{0441}\x{0442}";
>   if ( $hash{$key} eq eval "\$hash{$key}" )
But, would doing something like,
$key = "\x{0442}\x{0435}\x{0441}\x{0442}";
be within the scope of a bareword test ?

Also, does eval "\$hash{$key}" as in the 'if'
condition remain within the scope of a bareword test ?


> 
> It's unclear to me whether $key needs to be
> different for EBCDIC.
\x{0442} is the unicode value for the character that
we are trying to test. So, as long as we are testing
the same character, $key needs to be the same on both
platforms.
> 
> Are you just using perl on z/OS, or are you building
> it yourself?
I am building perl on z/OS and using it. 

> If the latter, Dave Mitchell has been looking for
> someone to test
> some parser changes he made on an EBCDIC platform so
> they can be
> integrated into the 5.8.x series.
> 

Thanks,
Rajarshi.


		
____________________________________________________
Start your day with Yahoo! - make it your home page 
http://www.yahoo.com/r/hs