Nick D Montpetit wrote:
>
> Hi Everyone!
Hello,
> I'm fairly new to Perl, and completely new to submitting to the list, so
> please be easy on me. :-)
I'll try.
> The purpose of the code I wrote (listed below) is to go through the current
> directory and all of its subdirectories and report the filename, size and
> age of the x largest files, where x depends on the argument supplied on the
> command line and the number of files in the directory. I'm sure there's an
> easier way to do this with a UNIX utility (or with Perl), but this program
> has been a good learning experience for me.
>
> The code below runs without any syntax errors, but File::Find (which I love
> and use frequently) or the -e file test doesn't give the results I expect.
> Specifically, my (explicit) checks show that the test in line 11 does not
> always evaluate to TRUE for values of $_ corresponding to legitimate (i.e.,
> existing and size > 0) files, which means I'm missing files that should be
> in my final output. If I change line 11for debugging purposes to simply
>
> if ($_){
>
> all files (including the previously missed ones) are printed out (also
> along with the directories now) as expected, but for the files that would
> not have passed the (-e $_) test, the values assigned for filesize and age
> in line 12 are (tested to be) undefined. If this is any help, the files
> that don't pass the (-e $_) of test of line 11 are (perhaps coincidentally)
> the biggest files (~55 GB) in the directory.
>
> I've researched the File::Find documentation and checked the FAQ's with no
> luck, so I'm hoping someone out there can help me.
>
> Also, is there a function that returns a minimum value in an array? It's
> not my biggest concern right now, but it would make line 20 simpler.
>
> 1: #!/usr/local/bin/perl -w
> 2:
> 3: use Cwd;
> 4: use File::Find;
> 5: use FileHandle;
> 6:
> 7: ($filecount = shift) || ($filecount = 20);
> 8:
> 9: sub wanted{
> 10: my $flag = 0;
> 11: if (-e $_){
> 12: push(@filelist, [$File::Find::name, -s $_, -M $_]);
> 13: }
> 14: }
> 15:
> 16: $dir = cwd();
> 17:
> 18: find(\&wanted,$dir);
> 19:
> 20: foreach $fileref ((sort {$b->[1] <=> $a->[1]} @filelist)[0..(sort( {$a
> <=>$b} $filecount,scalar(@filelist)-1))[0]]){
> 21: write STDOUT;
> 22: }
> 23:
> 24: STDOUT -> format_name("STDOUT_BOT");
> 25: write STDOUT;
> 26:
> 27: format STDOUT_TOP=
> 28: FILENAME
> FILE SIZE (BYTES) AGE (DAYS)
> 29:
>
>------------------------------------------------------------------------------------------------------------------
> 30: .
> 31: format STDOUT=
> 32:
> @<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
>
> >>>>>>>>>>>>>>>> @>>>
> 33: $fileref->[0], $fileref->[1], sprintf("%4.0f", ($fileref->[2]))
> 34: .
> 35: format STDOUT_BOT=
> 36:
>
>-----------------------------------------------------------------------------------------------------------------
> 37: .
> .
Here is something that I think does what you want:
#!/usr/local/bin/perl -w
use strict;
use Cwd;
use File::Find;
use FileHandle;
my $filecount = shift || 20;
my @filelist;
our $fileref;
# This stores _at most_ $filecount entries in @filelist so sorting is
fast
# and you don't have to store _all_ file names. Also int(-M $_) since
you
# are only reporting the age in days. Files are stored with the largest
# file at the top of the array @filelist (largest, second largest, third
# largest, ... $filecount largest)
sub wanted {
@filelist = sort { $b->[1] <=> $a->[1] } ( @filelist, [
$File::Find::name, -s $_, int(-M $_) ] );
pop @filelist if @filelist > $filecount;
}
my $dir = cwd();
find( \&wanted, $dir );
for $fileref ( @filelist ) {
write STDOUT;
}
STDOUT->format_name("STDOUT_BOT");
write STDOUT;
# the rest of the program is unchanged from original
John
--
use Perl;
program
fulfillment
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
