I also like to point out that when you die like you're doing now, the user will get a 500 error, possibly embarrasing whoever decides to use this (user has no idea that's what it's supposed to do when it can't read) ;D. A friendly error message might be in order (see CGI::Carp).

One more thing, you store under $header = "text/html", which is not neccessary when you use &CGI::header, because that is the default, and there really is no need to keep it in a "separate" variable (you prolly just got a little carried away with the configurating ;D)

Also, you might wanna add files that begin with . to the list of stuff not ok to see, as well as the actual script that's displaying the directory structure (unless you want it to show up if its there)

And, you also ought to look into the other parameters for the header method (you might wanna specify an expiration time, like print header(-type=>'text/html', -expires => '+5m');

And, look into Ovids cgi intro course, and look into

   $CGI::DISABLE_UPLOADS = 1;# Disable uploads
   $CGI::POST_MAX        =-1;# Maximum number of bytes per post
[download]

Also, since you're going to be using CGI to generate the html, you might as well generate "valid" html, check http://validator.w3.org/ to see about errors, a good starting point is specifying '-dtd'   => "-//W3C//DTD HTML 4.0 Transitional//EN" in start_html.

I think that's plenty to ponder, but I suggest you go and check out perlsec first, cause it's the most important.

 
___crazyinsomniac_______________________________________
Disclaimer: Don't blame. It came from inside the void
perl -e "$q=$_;map({chr unpack qq;H*;,$_}split(q;;,q*H*));print;$q/$q;"

• Using form input in a pathname without sanitizing it; specifically:
• Allowing sneaky examination of the system by user input of paths starting with lots of dotdot entries.

    -- Chip Salzenberg, Free-Floating Agent of Chaos

1. There is a great potential for an endless recursive loop to be established if your program processes a symbolic link to a directory higher in the directory tree - This is a common trap that File::Find-like subroutines get caught by. To correct this, simply add a negative test for symbolic link with your -d test. ie.

   if (-d $dir && !-l _) {
   
       ...
   
   };
   [download]

2. And less importantly, the sort by_lc readdir(DIR) block in the op_dir subroutine could more easily be written as thus:

   foreach (sort { lc($a) cmp lc($b) } readdir(DIR)) {
       push (@dir, "$dir\/$_") if (m/$non_include/);
   };
   [download]

Ooohhh, Rob no beer function well without!

#!/usr/bin/perl -wT

#####################################################
#####Web Site Directory Print                   #####
#####Copyright 2001, Jonathan Clover            #####
#####                                           #####
#####Description: Program to Print out in plain #####
#####text tab-deliminated representation of a   #####
#####directory structure. Has the ability to not#####
#####include any directory, as well as allows   #####
#####for a specific web directory to be         #####
#####specified within the quiry string formated #####
#####like "?dir=/about".                        #####
#####################################################

use strict;
use CGI;

$CGI::DISABLE_UPLOADS = 1;
$CGI::POST_MAX        = 51_200;

my ($web_dir, $default_dir, $tab, @non_include);

####################
###Configurations###
####################

#Home Web Directory
$web_dir = "/www2/nati";

#Default Directory to Start in if you 
#wish it not to be the Home Web Directory
#$default_dir = "/www2/nati";

#Tab String to use for print out
$tab = "\ \ \ \ ";

#Files not to include
#Those that start with . are never included
#for obvious reasons(aka infinite recursion
#and people trying to %#&@ with the script)
#Regex accepted as values in list
@non_include = ('_',          #Front Page Hidden Folders
        'Merchant2',  #The Online Store Data Folder
        'webstats');  #The Web Statstistics Folder

###########################
###End of Configurations###
###########################

my $cur = CGI->new();
my $start;
my $non_include = '^(\.|'.join('|', @non_include).')';

###Allow for Param's from a web interface###
if ( $cur->param("dir") ){
    my ( $temp ) = ( $cur->param("dir") =~ /([\w\/]+)$/ ); 
    $start = $web_dir.$temp;
}
else { $start = defined($default_dir)? $default_dir : $web_dir; }

###Calculate the number of tabs to be
###used when printing out results
my @start = split('/', $start);
my $tabs = $#start;

###Start the program and print out as plain text
print "Content-Type: text/html\n\n",
      "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.0 Strict//EN\">\n",
      "<html>\n",
      "<head>\n",
      "\t<title>Dir Structure Print</title>\n",
      "</head>\n",
      "<body>\n";

dir_tree($start);

print "</body>\n</html>";

###Subrotines Below###
sub dir_tree {
    my ($dir) = @_;
    my @dir = split('/', $dir);
    if(-e $dir){
        print $tab x ($#dir - $tabs),
              $dir[$#dir]."<br>\n";
    }
    else {
        print $dir[$#dir]." does not exist";
    }
    if(-d $dir && !(-l $dir)){
        foreach (op_dir($dir)){
            dir_tree($_);
        }
    }
}

sub op_dir {
    my $dir = shift;
    my @dir;
    opendir(DIR, $dir) || die "Couldn't open dir: $!";
    foreach (sort {lc($a) cmp lc($b)} readdir(DIR)){
        if ($_ !~ m/$non_include/){
            push(@dir, "$dir\/$_");
        }
    }
    return @dir;
}
[download]

$_=crypt("hssq","cr");m-[funki.g.jim.bed.wax]-i;$_=eval"$`\(\"czEW\",\
+"pr\"\).$`(\"CCSBD\",\"Cl\")";s+ltO8f+ +;s=kt|g|m.|YA=\"=g;s|[ej]|\"\
+.\"|g;eval;
[download]