Connection through a proxy

[freak]

This node falls below the community's threshold of quality. You may see it by logging in.

[Corion]

The simplest script is:

# in your shell:
perl -MLWP::Simple -e "get 'http://www.example.com/'"

assuming, that you have set your environment variables up correctly, that is, HTTP_PROXY=http://www.proxy-example.com:8080/.

[TwistedGreen]

Will you need to use authentication to get through the proxy? If so, you may run into some difficulties using the usual modules. I have had to use the CPAN module Net::HTTPTunnel instead of any of the LWP::UserAgent modules to do this.

An example:

use Net::HTTPTunnel;
use HTTP::Request;

my $proxy_host = "192.168.1.1";
my $proxy_port = "83";
my $proxy_user = "username";
my $proxy_pass = "password";

my $ht = Net::HTTPTunnel->new( 'proxy-host' => $proxy_host,
                               'proxy-port' => $proxy_port,
                               'remote-host' => 'www.perlmonks.org',
                               'remote-port' => '80',
                               'proxy-user' => $proxy_user,
                               'proxy-pass' => $proxy_pass,
                               'timeout' => 20);
                                   
$ht or die "request timed out";

# generate an HTTP request
my $req = HTTP::Request->new(GET => "http://www.perlmonks.org/index.pl
+?node_id=317556");
    
print $ht $req->as_string;                    # print the request to t
+he socket
    
my $response;
while (<$ht>) {$response .= $_;}            # read response (the HTML 
+web page) from the socket

close $ht;                                    # close the socket

The page is now stored in $response, to do with as you please.

[mpolo]

The newer LWP::UserAgent modules accept an http_proxy in the form http://username:password@server:port/ . You can, for instance, set $ENV{HTTP_PROXY} to a string of that form and then use LWP::Simple, etc.

If you have a Microsoft proxy server that you are going through, you will have to indicate your domain as well, so the format will be http://domain\username:password@server:port/ (careful with quoting of the backslash).

[mce]

Hi,

What is the difference with this module, and the Basic authentication that comes with LWP?

See the credentials function in LWP::UserAgent.

[TwistedGreen]

From what I gathered when I was working on this problem, the main difference is in the level of abstraction of the modules. LWP::UserAgent will not give you a direct socket, while Net::HTTPTunnel will return a TCP socket which is tunneled transparently through the proxy you specify. You can then write what you will to the server, such as the HTTP GET request as shown in the example.

But if you just want to retrieve a page like I did, the critical difference was that Net::HTTPTunnel worked swimmingly with my proxy while LWP::UserAgent seemed unable to authenticate properly. Perhaps a more experienced monk can explain the underlying mechanics.