On 08/30/2007 04:32 AM, Beginner wrote:
Hi,
I want all the output plus any error messages to got to a log file. I
used the BEGIN block to direct STDERR into the file:
BEGIN {
open(STDERR, ">>/usr/local/myreports/report.log") || die "Can't
write to file: $!\n";
}
use strict;
use warnings;
...
### Start some logging ###
my $log;
my $logfile = "$dist_dir/report.log";
open($log,">>$logfile") || die "Can't write to $logfile: $!\n";
print $log "$0 called at ", &tm," with pid $$\n";
I am not massively happy with this method. I can't seem to use a
scalar that I can share with the rest of the script for the log file
and if I do a perl -c myscript.pl, all the messages go to the log
file.
Q1) Can I use a scalar instead of having to give the full file path
in the BEGIN block?
Yes,
our $logfile;
INIT {
$logfile = '/usr/local/myreports/report.log';
open (STDERR, '>>', $logfile) || die "Can't write to file: $!\n";
}
I think that code should also allow "perl -c myscript.pl" to send output
to the console.
I haven't tried it but I suspect is would be fall
within the scope of the BEGIN block and as far as I know, you cannot
declare anything before a BEGIN block. Failing that, can anyone
suggest a different way of doing this.
Q2) You sometimes see scripts accept the verbose/debug command line
argument. How can you implement a system so that you can ask for more
messages? Once you've set $debug = 1; do you have to pepper your
script with
print $log "some message\n" if $debug == 1;
TIA,
Dp.
I've seen it done that way before.
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
http://learn.perl.org/
