|
|
sybperl-l Archive
Up Prev Next
From: "Srinivasan, Ravikumar" <ravikumar dot srinivasan at gs dot com>
Subject: RE: DBD::Sybase 1.07 install on Solaris 8 - failing t/exec.t test
#11
Date: Nov 2 2005 8:15PM
I would think this is a sybase feature :-)
Floats tend to do this from time to time.
Reason I tend use decimal instead.
-Ravi
-----Original Message-----
From: owner-sybperl-l@peppler.org [mailto:owner-sybperl-l@peppler.org] On
Behalf Of Greg Earle
Sent: Wednesday, November 02, 2005 3:03 PM
To: sybperl-l@peppler.org
Subject: Re: DBD::Sybase 1.07 install on Solaris 8 - failing t/exec.t test
#11
On Nov 2, 2005, at 11:37 AM, Michael Peppler wrote:
> On Tue, 2005-11-01 at 18:38 -0800, Greg Earle wrote:
>
>> I thought at first perhaps my Sybase server/login/password/database
>> information was incorrect, but I don't think I would've gotten this
>> far if it hadn't been.
>>
>> That part of t/exec.t is
>>
>> [...]
>> $rc = $sth->execute(undef, 25, 3.2234, "jan 3 2001", 5.4);
>> ok(defined($rc), "exec dbitest 3");
>> my @out = $sth->func('syb_output_params');
>> ok($out[0] == 5.4, "out param 1");
>>
>> Neither my Perl nor my (non-existent) Sybase knowledge are good
>> enough to tell what's going on here. Any ideas, Mike?
>
> It looks like the output parameter that should be 5.4 isn't quite that.
> I'm not sure what the reason might be, but you could edit the script
> and
> print out the @out list, like this:
>
> print "@out\n";
>
> right after the $sth->func() call to see what you get.
Oh, duh. Why didn't I think of that :-)
Looks like an arithmetic error:
[...]
ok 10 - exec dbitest 3
5.40000000000000035527136788005009
not ok 11 - out param 1
# Failed test (t/exec.t at line 89)
[...]
So it's looking for 5.4, and getting something really close to it :-)
Where do I look now? Something in the way I built Perl that's
returning these oddball long-precision results? Or in Sybase?
Thanks,
- Greg
|
