What I'm doing is a step beyond that. I'm using Error (which implements a try-catch type system) and have the following code:
my $dbh = try {
DBI->connect(
$conn_string,
$props{USER},
$props{PWD},
{
PrintError => 0,
RaiseError => 1,
},
)
} otherwise {
throw Error::Database -text => (
$DBI::errstr
|| $@
|| "Unknown Error: '$conn_string'"
);
};
$dbh->{HandleError} = sub { throw Error::Database -text => $DBI::er
+rstr };
return $dbh;
I've defined Error::Database as such:
package Error::Database;
######################################################################
+##########
use 5.6.0;
use strict;
use warnings;
######################################################################
+##########
# This is a baseclass I use for extending the standard Error class in
+ways that don't matter here.
use Error::Base;
our @ISA = qw(Error::Base);
######################################################################
+##########
sub new
{
my $class = shift;
my $self = $class->SUPER::new(@_) || return undef;
my $txt = $self->text;
if ($txt =~ /text=([^=]+)./)
{
$self->{-text} = $1;
}
elsif ($txt =~ /Message String: (.*?)$/)
{
$self->{-text} = $1;
}
return $self;
}
1;
The reason for the regexes is I wanted to strip off meaningless guck to me and only giv eme the good stuff.
------ We are the carpenters and bricklayers of the Information Age. Don't go borrowing trouble. For programmers, this means Worry only about what you need to implement. Please remember that I'm crufty and crochety. All opinions are purely mine and all code is untested, unless otherwise specified. |