Free for all links page mods, just need to make it refuse certian words and from any page except certian ones. 500points

You should build a queue so when links are submitted they must be approved by editors.  Editors can be registered users in which you provide the status because they have been around a while.  This way, each link can be inspected and insured it is valid and in the proper location.

I want to make it self controled, so i will ad these featured to help control it

I'm surprised that you're only getting 1,000 per day!  Allowing anyone and everyone to post links without being reviewed is only asking for trouble.  You could limit the submissions to allow only 1 link per domain and/or IP address, but that still won't help  by very much.  So, until you decide to moderate the submissions, you won't be able to fix the problem.

On another note, your code could use some work.  You should look into using some of the Perl modules that will make the scripts easier to write, more robust, and more secure.  Here are a few things to read up on and use.

1)enable warnings
2)use the strict pragma in EVERY Perl script you write
3)run CGI scripts in taint mode
4)use the CGI module
5)use file locks  (i.e., flock) for reading/writing the flat file DB files (i.e., database.txt)
6)reading/writing MySQL (or other) database files
7)use one of the modules for calculating and formatting the date (POSIX, Date::Time, Time::Format, etc,)

Here's how the beginning of CGI scripts are generally written.

 #!/usr/bin/perl -Tw

use strict;
use CGI;
use CGI::Carp 'fatalsToBrowser';

I would recommend that you create a procedure so that when users register you send them a link to their e-mail address that needs to be verified - this should help a little with the spam issue. Also, check to see if the spammers are using web based e-mail accounts or not and maybe ban these or put in extra steps (e.g. must be verifued by admin) for those users.

I would also look to build an administrative back end with some kind of analysis as to how many postings are being done per user - blocking the IP address won't stop serious spammers as they will probably be on broadband and the ISPs tend to dynamically change their IP address even when thry're connected. If someone is porting like crazy, then you can assume they are a spammer and ban them (or take their e-mail address and spam them back!)

I need help writting the perl...i dont want an admin section. to much work. i have way to many sites. juts want to do the aboved mentioned stuff.

You need to have the capacity to identify mis-spelled words as well, and have a "quarantine" status for those caught by the anti-spam measure that may actually be legitimate.  Maybe use soundex to identify possible issues.

Anti-spam is big business now, because it is so hard to do effectively.

Is it worth your while to invest in a commercial anti-spam engine?

What you could do, you may not want to of course, is to have it count down from 10 seconds before it adds the link therefore any auto-spamming programs or people who add spam will be discouraged from adding.
Just a suggestion.

The simplest technique I know is to display a short code. e.g. a random four digit number, and ask the visitor to enter that code. Only accept an entry if the visitor's code matches the displayed code.

Hackers can design a workaround which automatically reads the displayed code, but it wouldn't be worth their while unless your links page is almost vying with google for popularity. In that case you could add refinements like displaying the random code as an image at a randomly chosen location on the page, using random fonts.

Extending on mtgradwell's thought:

Or use a jpeg to display the image of the digits in a funky font on a patterned background.  That moves spoofing the digits into the realm of advanced image processing, which is beyond the means of the average spammer today (not guaranteeing tomorrow).

Another option could be to call the submitted and scan the returned page for certain keywords ("screenscraping"). I'm not sure how this is done with perl, since I've only used asp. A quick search on google shows that you choose from a couple of modules to implement this.

The best thing I would do, is flag by IP, IP addresses can be spoofed, but setting a limit on how many per IP is a good idea, that and e-mail address (but to be honest e-mail addresses are easy to come by)

this'd only work on if it was the same person doing this

http://BlogX.co.uk try flooding that with comments

any help on the actual programming of the perl to do this


1. Have the abilty in the script to deny certian words (sex, make money, etc, etc)
2. DENY HTML/CSS
3. Deny anypost except from certian pages
example my page
my friedns page
4. Deny more than 3 post per 24 hour period

and


it would be nice to have an email field so the user HAS to enter their email and it then saves their email and url in referrers.dat, for my records.

http://www.blahblah.com (youremail@email.com)

>> 1. Have the abilty in the script to deny certian words (sex, make money, etc, etc)

This is going to be very difficult to write from scratch.  First you need to put together the list of words (AND PHRASES?), then you'll need to take each one of them and list their “normal” misspelling, then build a list of their possible obfuscations.  Once you have that “database” assembled, then you can start working on how to handle the searching/matching.

For this part of the script, I think you'd be better off by following swift99's suggestion of investing in a commercial product.

>> 2. DENY HTML/CSS

Based on the links you currently have, you can filter most of this with 1 or 2 simple regex's.  For example, here's one that will strip out the style tags from the links:

   $FORM{'url'} =~ s/style=[^>]+//i;

If you want to strip out the form submissions, this regex should do the job:

   $FORM{'url'} =~ s!/\?[^"]+!!;

>> 3. Deny anypost except from certian pages

I'm not exactly sure what you mean.

Based on some of my suggestions, I've reworked portions of your script that you may want to review.  Each of the subs uses a different method for outputting the html; however, you should use use only 1 of those methods.  My purpose was to show you the various options.

#!/usr/local/bin/perl

use strict;
use POSIX qw(strftime);
use CGI qw(:all);
use CGI::Carp 'fatalsToBrowser';


# Define Variables

my $filename = 'addalink.htm';
my $linksurl = 'http://www.ccds.addr.com/wrc/links/addalink.htm';
my $linkscgi = 'links.pl';
my $linkstitle = 'Free for all Links Page';
my $database = 'database.txt';

my $q = new CGI;
my %FORM = $q->Vars;


# Done
##############################################################################


&no_url if ($FORM{'url'} eq 'http://' || $FORM{'url'} !~ /^(f|ht)tp:\/\/\w+\.\w+/);
&no_title if (!($FORM{'title'}));

# strip out the style tags
$FORM{'url'} =~ s/style=[^>]+//i;

# strip out the query string
$FORM{'url'} =~ s!/\?[^"]+!!;


# Enter our tags and sections into an associative array

my %sections = (
            busi => 'Business',
            comp => 'Computers',
            educ => 'Education',
            ente => 'Entertainment',
            gove => 'Government',
            pers => 'Personal',
            misc => 'Miscellaneous',
            );


# Suck previous link file into an array

open FILE,$filename or die $!;
my @lines = <FILE>;
close(FILE);

my $i=1;
foreach my $line (@lines) {    
    if ($line =~ /\<li\>\<a href\=\"([^\"]+)\">([^<]+)<\/a>/) {
        if ($FORM{'url'} eq $1) {
            &repeat_url;
        }
        $i++;
    }
}

# Open Link File to Output
open FILE,">$filename" or die $!;

foreach my $line (@lines) { # For every line in our data

   if ($line =~ /<!--time-->/) {
      my $date = strftime("%A, %B %d, %Y at %H:%M:%S %Z", localtime(time));
      print FILE "<!--time--><b>Last link was added $date</b><hr>\n";
   }
   elsif ($line =~ /<!--number-->/) {
      print FILE "<!--number--><b>There are <i>$i</i> links on this ";
      print FILE "page.</b><br>\n";
   }
   else {
       print FILE $line;
   }

   foreach my $tag ( keys %sections) { # For every tag
      if ( ($FORM{'section'} eq $sections{$tag}) &&
         ($line =~ /<!--$tag-->/) ) {

         print FILE "<li><a href=\"$FORM{'url'}\">$FORM{'title'}</a>\n";
      }
   }
}

close (FILE);

# Return Link File
#print "Location: $linksurl\n\n";

if ($database ne '') {
    open (DATABASE,">>$database" or die $!);
    print DATABASE "$FORM{'url'}\n";
    close(DATABASE);
}


sub no_url {  #  output HTML using a here document
   print <<HTML;
      Content-type: text/html\n\n
      <html><head><title>ERROR: No URL</title></head>
      <body bgcolor=#FFFFFF text=#000000><center>
      <h1>No URL</h1></center>
      You forgot to enter a url you wanted added to the Free for all link page.
      Another possible problem was that your link was invalid.<p>
      <form method=POST action="$linkscgi">
      <input type=hidden name="title" value="$FORM{'title'}">
      <input type=hidden name="section""
      value="$FORM{'section'}">
      URL: <input type=text name="url" size=50><p>
      <input type=submit> * <input type=reset>
      <hr>
      <a href="$linksurl">$linkstitle</a>
      </form></body></html>
HTML

   exit;
}

sub no_title {  #  output HTML using the CGI.pm object oriented syntax
   print (
      $q->header,
      $q->start_html(-title=>'ERROR: No Title', -bgcolor=>'#FFFFFF', -text=>'#000000'),
      $q->br."\n",
      $q->h1({-align=>center}, 'No Title') . "\n",
      $q->p('You forgot to enter a title you wanted added to the Free for all link page.',
            $q->br."\n",
            'Another possible problem is that you title contained illegal characters.')."\n",
      $q->start_form(-method=>'POST', -action=>$linkscgi),
      $q->hidden(-name=>'url', -value=>$FORM{'url'})."\n",
      $q->hidden(-name=>'title', -value=>$FORM{'section'})."\n".'TITLE: ',
      $q->textfield(-name=>'title', -size=>50)."\n",
      $q->submit, ' * ', $q->reset,
      $q->hr."\n",
      $q->a({-href=>$linksurl}, $linkstitle)."\n",
      $q->end_form,
      $q->end_html,
   );

   exit;
}

sub repeat_url {  # output using the CGI.pm standard syntax
   print (
      header,
      start_html(-title=>'ERROR: Repeat URL', -bgcolor=>'#FFFFFF', -text=>'#000000')."\n",
      h1({-align=>center},  'Repeat URL')."\n",
      p("Sorry, this URL is already in the Free For All Link Page\n",
        "You cannot add this URL to it again.  Sorry.\n"),
      a({-href=>$linksurl}, $linkstitle),
      end_html,
    );

   exit;
}

You may be interested in using the URI::URL module for parsing the URL.

Detailed Documentation:
http://search.cpan.org/~gaas/libwww-perl-5.10/lib/URI/URL.pm

Lastest Version:
http://search.cpan.org/~gaas/URI-1.31/URI/URL.pm

Here's a test script I ran that parsed the URL from 2 of your links.

use URI::URL;

$url1 = new URI::URL 'http://www.juiceboosted.com/index.php?requestId=2&Id=ix1pq';
$url2 = new URI::URL 'http://www.proofofprofit.com/mikl4545/?Jul28?Jul28?Jul28?Jul28';

foreach $url ($url1, $url2) {
   print 'scheme: ' . $url->scheme . "\n";
   print 'netloc: ' . $url->netloc . "\n";
   print 'path: ' . $url->path . "\n";
   print 'query: ' . $url->query . "\n";
   print $/ x 2;
}

-- output --
scheme: http
netloc: www.juiceboosted.com
path: /index.php
query: requestId=2&Id=ix1pq


scheme: http
netloc: www.proofofprofit.com
path: /mikl4545/
query: Jul28?Jul28?Jul28?Jul28

You might be wondering why I'm suggesting using these modules.  Well, these modules (and others that I haven't mentioned) will make it easier to parse, compare, and save or reject the links that are being submitted.  They will also make your scripts easier to write/read/maintain.

FishMonger's got the goods here ... follow up on his excellent work.

swift99,  thank you for the compliment. :)


Caiapfas ,

I have another suggestion.  Currently, you're reading-in and rewriting the entire 'addalink.htm' file.  It would be more effiecient to use a (relational) database such as MySQL to store the urls and other related info and have the 'links.pl' script dynamicly create/output the html links page.  If you don't want to use MySQL (or some other relational database), you can use various types of flat files as if they were a relational database.  I found this info in one of Tintin's posts in another question.

http://www.experts-exchange.com/Programming/Programming_Languages/Perl/Q_21076389.html

http://search.cpan.org/~jzucker/DBD-AnyData-0.08/AnyData.pm

This may not be as good as MySQL, but it still would be more effiecient that what you're currently using.

I recommend Firebird over MySql because Firebird has true two-phase commit in all implementations, but you have to pick and choose MySql implementations to get two-phase commit and the reliability that goes with it.

FishMonger,


Almost there.


1. Have the abilty in the script to deny certian words (sex, make money, etc, etc)
For this one I can make the list of words, no SWEAT..but can you make the script check badwords.txt before accepting the post. I can handle the rest.

2. DENY HTML/CSS/Javascript
Will it deny javascript/html?

3. Deny anypost except from certian pages
example my page
my friedns page
They have programs out there that just has the url to my perl script and can add links to the page from that program.
or

all you need to steal is this code...

<form method=POST action="http://www.ccds.addr.com/cgi-bin/addalink/links.pl" >

Title: <input type=text name="title" size=30><br>
URL: <input type=text name="url" size=55><br>
Section to be placed in: <select name="section">

<option> Business
<option> Computers
<option> Education
<option> Entertainment
<option> Government
<option> Personal
<option selected> Miscellaneous

</select>

<input type=submit value="Add"> * <input type=reset>

</form>


and you can add to my page...So i want to deny post/accepting post only from certian pages laided out in the perl script

example :

# Referrers

$referrers = "http://www.blablah.com";
$referrers = "http://www.blablah22.com";


4. Deny more than 3 post per 24 hour period from same IP

does it do this?


it would be nice to have an email field so the user HAS to enter their email and it then saves their email and url in referrers.dat, for my records.

http://www.blahblah.com (youremail@email.com)

I'll need you to post some lines from your badwords.txt file, so I can work out how to parse it and do the pattern matching against it.  Do you need to run these tests against the url itself, or the link discription, or both?


>> <form method=POST action="http://www.ccds.addr.com/cgi-bin/addalink/links.pl" >

It's not a good idea to use the full address, it would be much better to use relative paths.  It would be even better to have it refer to itsef (using a variable) and have the links.pl script build the page dynamiclly.

Fish,

If possible can we make the badword.txt non-case sentive...meaning if i have the word free and the user type freE it still gets denied..

link to file
http://www.ticketstogo.com/aadownload/badwords.txt


and how about a simple list of referrers? to deny post excetp from certian urls.
and does it deny javascript/html?

There are several methods for checking for the bad words.  Here's one method.

open BAD, "badword.txt" or die $!;
@badwords = <BAD>;
close BAD;

$badwords = join('|', @badwords);

&badurl if ($FORM{'title'} =~ /$badwords/i or $FORM{'url'} =~ /$badwords/i);

sub badurl {

   insert code for displayng the rejection page due to inapropriate words

}

I'm tied up on other projects at the moment, but as soon as I can, I'll respond to your other requests.

One small point I forgot; you need to remove the \n from each element of the array, prior to the join statement.

chomp @badwords;
$badwords = join('|', @badwords);

FishMonger,

I will have to wait for you.. but i'm getting spammed so much on that page.
please helllllllllllpppppppppppppppppppppp

I'm still tied up, so I still can't do any scripting, but I need to know if you have the ability to install a couple modules.  If not, then we'll have to do things the hard way.  My recomendation is to throw away your current script and static html page and use a more robust cgi script, but I don't have the time to write the complete script for you.  My plan is to work within your current approach and show you a few module options that should help resolve some, but not all, of the issues.

what do you mean install modules? how would i do this, but files in the same direcotry?

also, I think these few measures added will cut down on 90% of my spam issues...
its all i want/need.

Caiapfas,

Unfortunately, I still haven't had the time to work on you script needs.  My work projects and entertaining visiting relatives has kept me pretty busy.  I have a couple of ideas that will help resolve some, if not, most of your concerns, unfortunately I may not be able to work on the code until the weekend.

>> what do you mean install modules? how would i do this

Modules are additional (external) functions and subroutines that you import into your scripts, which will make yours scripts more robust, without having you to “reinvent the wheel” by writing complicated functions.  For example, the cg i module that I showed you is a standard module distributed with Perl that, among other things, simplifies the importing/processing of form submissions.  If you are hosting the site(s) on your own server, installing Perl modules can be as easy as this command:

ppm install module_name

or

perl -MCPAN -e install module_name

However, if the site is hosted by your ISP (or someone else), it can/will be more complicated, but is still do-able.

Caiapfas,

I haven't heard weather or not you'll be able to install any modules. If you will not be able to install any modules, then it will not be beneficial to either of us if I rewrite most or all of your script to use those modules. So instead, I'll make a few suggestions that will help clear up some of the issues.

Your 'url' field requires the person to specify the protocol, however, your page doesn't mention that requirement. So, you might want to add a selection field (just to the left of the url field) listing the accepted protocols with http:// being the default.

Currently, you have 3 separate "rejection" subroutines, of which, 2 are nearly identical. Following along with that approach, one of my prior posts suggested adding another sub for rejecting the "bad words". Instead, It would be better to have 1 rejection subroutine and pass a variable to it that holds the rejection message(s). As you perform each of the tests, i.e., 'no title', 'bad words', 'no url', etc., you can build up the rejection message and then call the rejection sub, if needed.

When testing the URL, you might want to check its length and reject it if it’s greater than X chars.  Currently, some of them are quite long.  You can also do a simple regex and reject if certain keywords like style and script are found.

>> Deny any post except from certain pages
There are a couple of CGI Environmental Variables that might help with this, namely: HTTP_REFERER and REMOTE_ADDR.  For example, as a starting point, you could do something similar to this:

if($ENV{HTTP_REFERER} !~ m!http://(www\.)?$mydomain/!) {
  $reject .= “your rejection message”;
}

If you add a required email field (and test for its validity), the combination of the $ENV{HTTP_ADDR} and the email address can be logged in a referrers.dat file.

Now, everything you need, can be done without the use of any additional modules, however, the modules will make it much easier to design, adds readability and maintainability as well as making it more bullet proof.  And, if you switch to using a relational database instead of the flat files, you'll add speed, and flexibility.

ok, i perfer not to install any modules.
I only want meager security measures i mentioned for this will cut 95% of the spam.

For the referer denyer...how can i add more than one.

also..could you kindly add the script bits you outlayed on this page in the complete code for me. I tryed many errors =}

and the only thing we havent written code for it seems is the referrers.dat file. I would like the script to add the referrer and email address to this file. and deny adding anymore links ffor the same refferer

and then the very last

only 3 submits in 48 hours per ip, can be logged in the refferers.dat file

Without the benefit of the modules and the database, some of these tasks are going to be difficult and/or make your site run slow due to the extra reading/writing of the flat text files.

My schedule is really full right now, but as soon as I can, I'll work on the script.  However, I can't promise to provide you a complete and fully debugged script that does everything you want.  Afterall you need to remember that people on this site are here providing free tech support, and would prefer to assist you in writing the script, instead of writing it for you.

Here's another recomendation I can make; use an IDE program such as Komodo to help you write and debug your (Perl and PHP) scripts.
http://activestate.com/Products/Komodo/

If you feel that you need someone to write the complete scripts (form script and the processing script) and have them done in a hurry, I suggest closing this question, and post a new question in the Perl section.  Also, be sure to make it clear that you don't want to use any modules that are not in the base distribution.  That way other Perl experts will help and probably come up with solutions that I've missed.

ok

Where do i put this in the script?
-----------------------

open BAD, "badword.txt" or die $!;
@badwords = <BAD>;
close BAD;

chomp @badwords;
$badwords = join('|', @badwords);

&badurl if ($FORM{'title'} =~ /$badwords/i or $FORM{'url'} =~ /$badwords/i);

sub badurl {

   insert code for displayng the rejection page due to inapropriate words <<<<what goes here text????

}
-----------------------------
where do i insert this

if($ENV{HTTP_REFERER} !~ m!http://(www\.)?$mydomain/!) {
  $reject .= “your rejection message”;
}

and how can i make it allow more than one?
--------------------------


just these 2 measures will save me 1000000000's of submits

also where do i put this??


  $FORM{'url'} =~ s!/\?[^"]+!!;

Why not use the "ban list" regexp in that other script I did for you.

# Load ban list
my @banlist;
if (-s $banlistfile) {
     open (TXTFILE,"$banlistfile");
     flock(TXTFILE,LOCK_SH);
     eval {
          @banlist = <TXTFILE>;
     };
     flock(TXTFILE,LOCK_UN);
     close(TXTFILE);
     die("Could not load ban file: $@") if ($@);
}

----
and
----

# Returns 1 for banned URL
sub IsBanned {
     my($cUrl) = @_;

     # Scan through the ban list and construct regex's
     my $cExp;
     foreach (@banlist) {
          s/^\s*(.*?)\s*$/$1/;
          $cExp = $_;
          # Escape all non-word characters
          $cExp =~ s/(\W)/\\$1/g;
          # Treat * as ungreedy wildcard
          $cExp =~ s/\\\*/'(.*?)'/ge;
          # Enforce whole string comparison, use * to relax
          $cExp = '(?i)^' . $cExp . '$';

          # Fail if the URL matches pattern
          return 1 if ($cUrl =~ $cExp);
     }
     return 0;
}

----

These implement a "ban list" where you can use wildcards, like this:

Ban EVERYTHING:
*
Ban any URL containing sex anywhere in the link:
*sex*

Ban any URL ENDING with bad. URL containing bad somewhere else is ok:
*bad

Ban testing.mydomain.com, both http: and https:
http*://testing.mydomain.com/*

Ban any URL beginning with www and ending with biz, with cgi-bin somewhere in the path:
http://www.*.biz/*cgi-bin*

Remember, the beginning of the pattern and the end of the pattern must match the exact beginning and ending of the URL being tested, except when the pattern begins or ends with "*".

All patterns are case insensitive.

adg,

GREAT !! could you help me add these few measures to stop the abuse of the system?

This still needs addtional work, but it meets (most of) your needs.

#!/usr/local/bin/perl -w

use strict;
use POSIX qw(strftime);
use CGI qw(:standard);
use CGI::Carp 'fatalsToBrowser';

# Define Global Variables

my $filename = 'addalink.htm';
my $linksurl = 'http://www.ccds.addr.com/wrc/links/addalink.htm';
my $linkscgi = 'links.pl';
my $linkstitle = 'Free for all Links Page';
my $database = 'database.txt';
my (@bandwords, $bandwords, $msg);
my $url = param{'url'};
my $title = param{'title'};
my %sections = (
           busi => 'Business',
           comp => 'Computers',
           educ => 'Education',
           ente => 'Entertainment',
           gove => 'Government',
           pers => 'Personal',
           misc => 'Miscellaneous',
           );

# Done
########################################################################

if($ENV{HTTP_REFERER} !~ m!http://www.ccds.addr.com/!) {

#   The submission is NOT comming from your domain, so
#   the error handling on this could be either displaying another error page
#   or a redirect to your link submission form

}

open BAD, "bandword.txt" or die $!;
@bandwords = <BAD>;
close BAD;

chomp @bandwords;
$bandwords = join('|', @bandwords);

if (!$title || $title =~ /$bandwords/i) {
   $msg = 'You forgot to enter a title or '.
          'the title contains banned words';
   reject_submission($msg, 'title');
}

if ($url) {
   if ($url !~ /^(f|ht)tp:\/\/\w+\.\w+/) {
      $msg = 'You forgot to specify the protocol, e.g., http://';
      reject_submission($msg, 'url');
   }

   if ($url =~ /style=[^>]+/i || $url =~ /\?[^"]+/) {
      $msg = 'Your url has style tags or is a form submission, ' .
             'which are not allowed.';
      reject_submission($msg, 'url');
   }

   if ($url =~ /$bandwords/i) {
      $msg = 'Your title contains banned words';
      reject_submission($msg, 'url');
   }
}
else {
   $msg = 'You forgot to provide the url for the link';
      reject_submission($msg, 'url');
}

# Suck previous link file into an array

open FILE,$filename or die $!;
my @lines = <FILE>;
close(FILE);

my $i=1;
foreach my $line (@lines) {
   if ($line =~ /\<li\>\<a href\=\"([^\"]+)\">([^<]+)<\/a>/) {
       if ($url eq $1) {
           &repeat_url;
       }
       $i++;
   }
}

# Open Link File to Output
open FILE,">$filename" or die $!;

foreach my $line (@lines) { # For every line in our data

  if ($line =~ /<!--time-->/) {
     my $date = strftime("%A, %B %d, %Y at %H:%M:%S %Z", localtime(time));
     print FILE "<!--time--><b>Last link was added $date</b><hr>\n";
  }
  elsif ($line =~ /<!--number-->/) {
     print FILE "<!--number--><b>There are <i>$i</i> links on this ";
     print FILE "page.</b><br>\n";
  }
  else {
      print FILE $line;
  }

  foreach my $tag ( keys %sections) { # For every tag
      if ( (param{'section'} eq $sections{$tag}) &&
         ($line =~ /<!--$tag-->/) ) {

        print FILE "<li><a href=\"$url\">$title</a>\n";
      }
  }
}

close (FILE);

# Return Link File
#print "Location: $linksurl\n\n";

if ($database ne '') {
   open DATABASE,">>$database" or die $!;
   print DATABASE "$url\n";
   close(DATABASE);
}


sub reject_submission {
   my $msg = shift;
   my $section = shift;

   print <<HTML;
     Content-type: text/html\n\n
     <html><head><title>Link Rejection</title></head>
     <body bgcolor=#FFFFFF text=#000000><center>
     <h1>Link Submission Rejected</h1></center>
     $msg<p>
     <form method=POST action="$linkscgi">
     <input type=hidden name="$section" value="$section">
     <input type=hidden name="section" value="$section">
     $section: <input type=text name="$section" size=50><p>
     <input type=submit> * <input type=reset>
     <hr>
     <a href="$linksurl">$linkstitle</a>
     </form></body></html>
HTML

   exit;
}

sub repeat_url {  # output using the CGI.pm standard syntax
  print (
     header,
     start_html(-title=>'ERROR: Repeat URL', -bgcolor=>'#FFFFFF', -text=>'#000000')."\n",
     h1({-align=>center},  'Repeat URL')."\n",
     p("Sorry, this URL is already in the Free For All Link Page\n",
       "You cannot add this URL to it again.  Sorry.\n"),
     a({-href=>$linksurl}, $linkstitle),
     end_html,
   );

  exit;
}

i get error


Software error:
Bareword "center" not allowed while "strict subs" in use at /usr163/home/c/c/ccds/public_html/cgi-bin/addalink/links.pl line 152.
Execution of /usr163/home/c/c/ccds/public_html/cgi-bin/addalink/links.pl aborted due to compilation errors.

you can test it at the above link..

p.s how can i log the urls / email addresses?

As you can see, I didn't fully debug it.  The error message is informing use that I forgot to put the quotes around that word.  I also noticed a couple other errors of mine that need to be fixed.

change these lines:
   my $url = param{'url'};
   my $title = param{'title'};

   h1({-align=>center},  'Repeat URL')."\n",

to this:
   my $url = param('url');
   my $title = param('title');

   h1({-align=>'center'},  'Repeat URL')."\n",

Also, take off the openning and closing (  ) on the print statement in the repeat_url subroutine.


>> how can i log the urls / email addresses?

The first step, of course, is to add the email field to your addalink.htm page.  Then in the links.pl, you'd access it with: param('email').  You'll then need to verify that they entered a properly formatted email address, often via a regex.  However IMO, unless you take the next step and verify that it actually is a valid address, the logging of it would be pointless.  There is a CPAN module that will easily test the format and do a DNS MX lookup to verify that it is a valid address,  but you'll need to change your mind about the use/installation of additional modules.

The host name and/or IP address of the visitor can be retreived from 2 of the CGI environment variables i.e., REMOTE_HOST and REMOTE_ADDR.

Once you have the seperate pieces of info, you have several choices on how to handle the logging.  The best method is to use a relational database, but if you want to use a file, I'd suggest using a tied hash.  A tied hash is "linked" or "tied" directly to a file, so as you make changes to the hash, you are actually making changes to the file.  This will allow you to store the info as a "record".  Here's a little more info on using a tied hash.

http://www.perl.com/pub/a/2001/09/04/tiedhash.html
http://www-106.ibm.com/developerworks/linux/library/l-cptied.html

I need to jump onto a couple of other personal projects, but as soon as I can I'll work up some code examples.

ok, now all sudmissions are being denied due to contaioning no title or bad words which i know it does not have. any help?

Could you upload exactly what you have (but change the extension from .pl to .txt) any post a link to it here?

Lets make a small change to the test.

if (!$title) {
   $msg = 'You forgot to enter a title.';
   reject_submission($msg, 'title');
}
elsif ($title =~ /$bandwords/i) {
   $msg = "The title contained '$&' which is a banned word";
   reject_submission($msg, 'title');
}

I'm getting the error "The title contained '' which is a banned word'


Badwords.txt - http://www.ccds.addr.com/EExchange/badwords.txt
links.pl - http://www.ccds.addr.com/EExchange/links.txt


To test - http://www.ccds.addr.com/wrc/links/addalink.htm



P.s. adg, could you help me log urls and emails like in the other script?

$bandwords = join('|', map{quotemeta}@bandwords);
elsif ($title =~ /\b($bandwords)\b/i) {

that brings this error


syntax error at /usr163/home/c/c/ccds/public_html/cgi-bin/addalink/links.pl line 45, near "elsif"
Execution of /usr163/home/c/c/ccds/public_html/cgi-bin/addalink/links.pl aborted due to compilation errors.

Line 45 should be a blank line, so did you make ozo's corrections to the appropriate lines?

With ozo’s corrections, lines 43 thru 53 should be:

chomp @bandwords;
$bandwords = join('|', map{quotemeta}@bandwords);

if (!$title) {
   $msg = 'You forgot to enter a title.';
   reject_submission($msg, 'title');
}
elsif ($title =~ /\b$bandwords\b/i) {
   $msg = "The title contained '$&' which is a banned word";
   reject_submission($msg, 'title');
}

ok, now i'm get a 500 error


Internal Server Error
The server encountered an internal error or misconfiguration and was unable to complete your request.
Please contact the server administrator, webmaster@.com and inform them of the time the error occurred, and anything you might have done that may have caused the error.

More information about this error may be available in the server error log.


Check out the script here:
links.pl - http://www.ccds.addr.com/EExchange/links.txt

It might be the first line, try:

#!/usr/bin/perl -w

Try with -w and if it doesn't work try again without the -w, to see if that's killing the script too.

that wasn't the issue. before I added the lastlines of code it worked fine. but wouldnt add the link. how can i fix it?

#!/usr/local/bin/perl -w

use strict;
use POSIX qw(strftime);
use CGI qw(:standard);
use CGI::Carp qw(fatalsToBrowser);

# Define Global Variables

my $filename = '/usr163/home/c/c/ccds/public_html/wrc/links/addalink.htm';
my $linksurl = 'http://www.ccds.addr.com/wrc/links/addalink.htm';
my $linkscgi = 'http://www.ccds.addr.com/cgi-bin/addalink/links.pl';
my $linkstitle = 'Free for all Links Page';
my $database = '/usr163/home/c/c/ccds/public_html/wrc/links/database.txt';
my (@bandwords, $bandwords, $msg);
my $url = param('url');
my $title = param('title');
my $section = param('section');
my %sections = (
           busi => 'Business',
           comp => 'Computers',
           educ => 'Education',
           ente => 'Entertainment',
           gove => 'Government',
           pers => 'Personal',
           misc => 'Miscellaneous',
           );

# Done
########################################################################

#if($ENV{HTTP_REFERER} !~ m!http://www.ccds.addr.com/!) {

#   The submission is NOT comming from your domain, so
#   the error handling on this could be either displaying another error page
#   or a redirect to your link submission form

#}

if (!$title) {
   $msg = 'You forgot to enter a title.';
   reject_submission($msg, 'title');
}

if (!$url) {
   $msg = 'You forgot to provide the url for the link';
      reject_submission($msg, 'url');
}

open BAD, "badwords.txt" or die $!;
@bandwords = <BAD>;
close BAD;

chomp @bandwords;
$bandwords = join('|', map{quotemeta}@bandwords);

if ($title =~ /(\b$bandwords\b)/i) {
   $msg = "The title contained '$1' which is a banned word";
   reject_submission($msg, 'title');
}

if ($url !~ /^(f|ht)tp:\/\/\w+\.\w+/) {
   $msg = 'You forgot to specify the protocol, e.g., http://';
   reject_submission($msg, 'url');
}

if ($url =~ /style=[^>]+/i || $url =~ /\?[^"]+/) {
   $msg = 'Your url has style tags or is a form submission, ' .
          'which are not allowed.';
   reject_submission($msg, 'url');
}

if ($url =~ /($bandwords)/i) {
   $msg = "The url contained '$1' which is a banned word";
   reject_submission($msg, 'url');
}


# Suck previous link file into an array

open FILE,$filename or die $!;
my @lines = <FILE>;
close(FILE);

my $i=1;
foreach my $line (@lines) {
   if ($line =~ m!<li><a href="([^"]+)">([^<]+)</a>!) {
       if ($url eq $1) {
           &repeat_url;
       }
       $i++;
   }
}

# Open Link File to Output
open FILE,">$filename" or die $!;

foreach my $line (@lines) { # For every line in our data

  if ($line =~ /<!--time-->/) {
     my $date = strftime("%A, %B %d, %Y at %H:%M:%S %Z", localtime(time));
     print FILE "<!--time--><b>Last link was added $date</b><hr>\n";
  }
  elsif ($line =~ /<!--number-->/) {
     print FILE "<!--number--><b>There are <i>$i</i> links on this ";
     print FILE "page.</b><br>\n";
  }
  else {
     print FILE $line;
  }

  foreach my $tag ( keys %sections) { # For every tag
     if ( $section eq $sections{$tag} &&
        ($line =~ /<!--$tag-->/) ) {
        print FILE "<li><a href=\"$url\">$title</a>\n";
      }
  }
}

close (FILE);

# Return Link File
#print "Location: $linksurl\n\n";

if ($database ne '') {
   open DATABASE,">>$database" or die $!;
   print DATABASE "$url\n";
   close(DATABASE);
}


sub reject_submission {
   my $msg = shift;
   my $section = shift;

   print <<HTML;
     Content-type: text/html\n\n
     <html><head><title>Link Rejection</title></head>
     <body bgcolor=#FFFFFF text=#000000><center>
     <h1>Link Submission Rejected</h1></center>
     $msg<p>
     <form method=POST action="$linkscgi">
     <input type=hidden name="$section" value="$section">
     <input type=hidden name="section" value="$section">
     $section: <input type=text name="$section" size=50><p>
     <input type=submit> * <input type=reset>
     <hr>
     <a href="$linksurl">$linkstitle</a>
     </form></body></html>
HTML

   exit;
}

sub repeat_url {  # output using the CGI.pm standard syntax
  print
     header,
     start_html(-title=>'ERROR: Repeat URL', -bgcolor=>'#FFFFFF', -text=>'#000000')."\n",
     h1({-align=>'center'},  'Repeat URL')."\n",
     p("Sorry, this URL is already in the Free For All Link Page\n",
       "You cannot add this URL to it again.  Sorry.\n"),
     a({-href=>$linksurl}, $linkstitle),
     end_html,
   ;

  exit;
}

There still are a number of things that can and should be done to improve that script, but it works.

it still doesnt work getting a 500 error

Sounds like it's a configuration problem, possibly due to permission settings on the script or the web server isn't configured to execute scripts from that directory, i.e., no script alias.

Try running the script from the command line.  If it works, it confirms that the script is "ok", so you'll need to look at the permissions/configuration.  For example, here are a some example outputs from my test runs.

C:\Testing>links.pl title='sex and the city' url=http://www.hbo.com/city/ section=Miscellaneous
     Content-type: text/html


     <html><head><title>Link Rejection</title></head>
     <body bgcolor=#FFFFFF text=#000000><center>
     <h1>Link Submission Rejected</h1></center>
     The title contained 'sex' which is a banned word<p>
     <form method=POST action="links.pl">
     <input type=hidden name="title" value="title">
     <input type=hidden name="section" value="title">
     title: <input type=text name="title" size=50><p>
     <input type=submit> * <input type=reset>
     <hr>
     <a href="http://www.ccds.addr.com/wrc/links/addalink.htm">Free for all Links Page</a>
     </form></body></html>

C:\Testing>links.pl url=http://www.hbo.com/city/ section=Miscellaneous
     Content-type: text/html


     <html><head><title>Link Rejection</title></head>
     <body bgcolor=#FFFFFF text=#000000><center>
     <h1>Link Submission Rejected</h1></center>
     You forgot to enter a title.<p>
     <form method=POST action="links.pl">
     <input type=hidden name="title" value="title">
     <input type=hidden name="section" value="title">
     title: <input type=text name="title" size=50><p>
     <input type=submit> * <input type=reset>
     <hr>
     <a href="http://www.ccds.addr.com/wrc/links/addalink.htm">Free for all Links Page</a>
     </form></body></html>

C:\Temp>links.pl title=www.hbo.com/city url=www.hbo.com/city/ section=Miscellaneous
     Content-type: text/html


     <html><head><title>Link Rejection</title></head>
     <body bgcolor=#FFFFFF text=#000000><center>
     <h1>Link Submission Rejected</h1></center>
     You forgot to specify the protocol, e.g., http://<p>
     <form method=POST action="links.pl">
     <input type=hidden name="url" value="url">
     <input type=hidden name="section" value="url">
     url: <input type=text name="url" size=50><p>
     <input type=submit> * <input type=reset>
     <hr>
     <a href="http://www.ccds.addr.com/wrc/links/addalink.htm">Free for all Links Page</a>
     </form></body></html>


You should also take note on the link being submitted.  I chose that link (to a popular TV program) to show that just because it has a "bad" word in the title, doesn't always mean that it needs to be rejected.  

i still cant get this script to work i have it chomd to 755 and have tried 777