Perhaps this pattern is what you need.

$server_name =~ m{
    \A                   # start of string
    .{1,3}               # 1-3 characters
    \.                   # a literal period
    product1help\.com    # literal 'product1help.com'
}xms;
[download]

If you want to know what the prefix matched actually was, use this:

$server_name =~ m{
    \A                   # start of string
    (                    # start capture 1
        .{1,3}           # 1-3 characters
    )                    # end capture 1
    \.                   # a literal period
    product1help\.com    # literal 'product1help.com'
}xms;
my $prefix = $1;
[download]

I don't understand why this is not working, if my url is sef.mlxhelp.com it fails, if its www.mlxhelp.com it works.. what gives can anyone clarify why this fails? why does it even look at the prefix part?

 
 sub transform
  {
      if (!$Session->{'isAuthenticated'}) 
      { 
      if ($Request->ServerVariables("SERVER_NAME")->item()=~/(\mlxhelp
+*)\.com/) 
         { 
         $Response->write("this is what I got $1");
          $trns=$Server->MapPath("Transforms/pLogin.xsl");
          return $trns;
          }
          if ($Request->ServerVariables("SERVER_NAME")->item()=~/(\tem
+pohelp*)\.com/) 
          { 
          $Response->write("this is what I got $1"); 
          $trns=$Server->MapPath("Transforms/tpLogin.xsl");
          return $trns;
          }
       }
[download]

In a regular expression, "\t" is a tab character, and I'm not sure what "\m" is, if anything.

use Test::More 'tests' => 2;

ok( 'sef.mlxhelp.com' =~ /(\.mlxhelp)\.com/,
    q{'sef.mlxhelp.com' matches} );
ok( 'www.mlxhelp.com' =~ /(\.mlxhelp)\.com/,
    q{'www.mlxhelp.com' matches} );
[download]

Note that these will always put ".mlxhelp" in $1 when they match because the capturing parentheses are around that literal string.