Up Prev Next
From: Peter Sinnott <psinnottie at aol dot com>
Subject: Re: bcp via Sybase::DBlib - seg faults and error catching
Date: Sep 12 2002 1:37PM
On Thu, Sep 12, 2002 at 03:02:43AM -0700, firstname.lastname@example.org wrote:
> > With 4 the sendrow goes ok and the first 3 elements get inserted. I can't
> > see any indication anything has gone wrong but when I call bcp_done I
> > get a seg fault.
> That's a bug - probably an off-by-one error in the C code.
For the case of sendrow with an array reference the problem seems
to be that the return from av_len is being used as the size of the array
in question but according to perlguts it's the highest index in the
array. I was kinda expecting the length myself until I checked the pod.
> As an aside - I am still on vacation in Europe, so don't expect any
> actual fixes for another couple of weeks.
I was planning on using a per row validation function to catch some
stuff that the database couldn't know about but as it turns out it
doesn't matter since we won't be using bcp at all now.
Hope you are having a nice vacation and that you are getting a little
more sun than we are getting here in sunny Ireland.
Our challenge is to conveniently provide access to progressive sources
in order that we may collaboratively foster inexpensive meta-services for
100% customer satisfaction