A subroutine example is below, msg me if you cant work out which bits might be useful.

P.S I have a suspicion setting the umask to 000 might not be a good thing but this sub is renaming the files...

sub storeFiles {
    my $self = shift;
    my $cgi = shift;
    my $mode = shift; # going to be insert or update

    my $fileerror = '0';
    my @filecols = ('DOC_1', 'DOC_2', 'IMG_1', 'IMG_2', 'IMG_3', 'IMG_
+4');
    my $umm = umask(000); 

    foreach my $doc (@filecols) {
        if ($cgi->param($doc)) {
            # if insert this is an add, else a replace action for upda
+tes

            my $longfile = $cgi->param($doc);
            my $tempfile = $cgi->tmpFileName($longfile);
            my $size = -s $tempfile;
             
            if (($size < 1) || ($size > 2000000)) {
                $fileerror = '1';
                last;
            }

            $longfile =~ s/ //g;                            # strip sp
+aces
            $longfile =~ s/^(.*[\/\\:])*([^\/\\:]+)$/$2/;   # remove p
+ath      

            if ($longfile !~ /\./) {
                $fileerror = '2';
                last;
            }

            $longfile =~ s/\.(?=.*\.)//g;                   # substitu
+te all but last dot (courtesy of L.Rosler)

            my ($f, $ext) = split /\./, $longfile;

            if (!$self->_isValidFile($doc, $ext)) {
                $fileerror = '3';
                last;
            }

            my $newfile = $DATA_DIR . $self->{'PAGE'}[2] . "/" . $self
+->{'DOCID'}[2] . $doc . "." . $ext;
            eval {
                copy ($tempfile, $newfile) or die "cant copy temp *$te
+mpfile* to new *$newfile* $! \n longfile is $longfile";
            };
            if ($@) {
                $fileerror = '4';
                last;
            } else {
                $self->{$doc}[2] = $self->{'DOCID'}[2] . $doc . "." . 
+$ext;              
            }
        } elsif ($cgi->param("exists$doc")) {
            $self->{$doc}[2] = $cgi->param("exists$doc");
        } else {
            $self->{$doc}[2] = '';
        }

        if (($cgi->param("chk$doc")) && (!$cgi->param($doc))) {
            # unlink file if update checkbox is checked        
            my $page = $self->{'PAGE'}[2];
            if (($page == 22) || ($page == 26) || ($page == 27)) {
                my $dir = "../data/$page";
                opendir (DIRH, $dir) or Util->showError("Can't Open de
+signated directory $dir $!");
                my @files = grep /^$self->{'DOCID'}[2]$doc/, readdir D
+IRH;
                @files = map ("$dir/$_", @files);
                closedir DIRH;
                unlink @files or Util->showError("oops cant unlink you
+r chosen files @files $!");
            }
            $self->{$doc}[2] = ''; 
        }
    }
    umask($umm);
    return $fileerror;
}
[download]

my $file = $query->param('upload_file_name');
while (<$file>) {
   print "Got a line: $_";
}
[download]

You may wish to examine the 'FILE UPLOAD' section of the CGI.pm documentation. It goes into considerable detail about all of this, as well as the tmpFileName method another poster mentioned. It even has examples.

<FORM ACTION="/path/somescript.cgi" METHOD="POST" ENCTYPE="multipart/f
+orm-data">
[download]

my $file = $q->param( 'file' ) or someErrorRoutine( $q, "No file recei
+ved" );
my $fh   = $q->upload( $file );
[download]

Cheers,
Ovid

Join the Perlmonks Setiathome Group or just go the the link and check out our stats.

That's easy when you use CGI.pm.

use CGI qw(:standard);
print
  hr,
  start_multipart_form,
  filefield('fieldname'),
  submit,
  end_form,
  hr;
[download]

-- Randal L. Schwartz, Perl hacker

AgentM Systems or Nasca Enterprises is not responsible for the comments made by AgentM- anywhere.