Using Sockets in PHP: getting articles from newsgroups

PHP can open a socket port on a remote or local host. This article is a small example of using sockets: Connect to a Usenet newsgroup server, talk to servers, and download articles from newsgroups.


Open a socket in PHP
Use Fsockopen () to open a socket. This function can be used in both PHP3 and PHP4. The function declaration is this:


int Fsockopen (string hostname, int port _
[, int errno [, String errstr [, double timeout]]]


This function will open a TCP connection to port ports connected to the host hostname. Hostname can be a valid domain name, or an IP address. For UDP connections, you must specify the protocol: Udp://hostname. For UNIX domains, the host name uses the path to the socket, in which case port ports must be set to 0. The optional timeout parameter is used to set the time, in seconds, to wait for a socket to open.


For more information on fsockopen (), please refer to: http://www.php.net/manual/function.fsockopen.php


Network News Transfer Protocol
Access to a newsgroup server is required through a protocol called NNTP (Network News Transfer Protocol). This agreement has detailed details in the rfc977 and can be obtained in http://www.w3.org/Protocols/rfc977/rfc977.html. This document describes how to connect to the NNTP server, how to talk to the server, and the different commands to complete these tasks.


Connection
Connecting to an NNTP server requires knowing its host name (or IP address) and the port it listens on. To avoid a connection attempt failure causing the program to hang, you should use the timeout parameter.
<?php
$cfgServer = "Your.news.host";
$cfgPort = 119;
$cfgTimeOut = 10;

Open a socket
if (! $cfgTimeOut)
Without timeout
$usenet _handle = Fsockopen ($cfgServer, $cfgPort);
Else
With timeout
$usenet _handle = Fsockopen ($cfgServer, $cfgPort, & $errno, & $errstr, $cfgTimeOut);

if (! $usenet _handle) {
echo "Connection failed."
Exit ();
}
else {
echo "Connected."
$tmp = fgets ($usenet _handle, 1024);
}

?>
Talking to the server

Now that we have connected to the server, we can talk to the server through the socket we opened earlier. Let's say we get the last 10 articles from a newsgroup. RFC977 points out that the first step is to select the correct newsgroup with the group command:
GROUP GGG
Parameter GGG is the name of the newsgroup you want to select (for example, "Net.news"), which is required. Lists of available newsgroups can be obtained by using the List command. When the command to select a newsgroup succeeds, returns the article number of the first and last articles in the group, and the number of articles in the group.


Here is an example:
chrome:~$ Telnet My.news.host 119
Trying aa.bb.cc.dd ...
Connected to My.news.host.
Escape character is ' ^] '.
My.news.host internetnews NNRP Server INN 2.2.2 13-dec-1999 ready (posting OK).
GROUP Alt.test
211 232 222996 223235 Alt.test
Quit
205.

After receiving the command GROUP alt.test, the server returns "211 232 222996 223235 alt.test". 211 is the return code defined in the RFC, indicating that the command was executed successfully. The return information also points out that there are now 232 articles, the earliest article is numbered 222996, and the latest article is numbered 223235. We see that 222996+232 is not equal to 223235. The missing 7 articles were deleted from the server for some reason, possibly because it was canceled by its rightful author (which is possible, and it is easy to do), or deleted because it is an irrigation article.


To be aware, some servers may require authentication before they select newsgroups, depending on whether they are a public or private server. It is also possible that the server allows anyone to read the article, but it requires authentication to publish the article.

<?php

$cfgUser = "xxxxxx";
$CFGPASSWD = "yyyyyy";
$cfgNewsGroup = "alt.php";

Identification required on private server
if ($cfgUser) {
Fputs ($usenet _handle, "AUTHINFO USER". $cfgUser. " n ");
$tmp = fgets ($usenet _handle, 1024);
Fputs ($usenet _handle, "AUTHINFO Pass". $cfgPasswd. " n ");
$tmp = fgets ($usenet _handle, 1024);

Check error

if ($tmp!= "281 okrn") {
echo "502 Authentication Errorn";
Exit ();
}
}

Select newsgroup

Fput ($usenet _handle, "GROUP". $cfgNewsGroup. " n ");
$tmp = fgets ($usenet _handle, 1024);

if ($tmp = = "Authentication required for Commandrn") {
Echo $tmp;
Exit ();
}

$info = Split ("", $tmp);
$first = $info [2];
$last = $info [3];

printf ("A:%sn", $first);
printf ("Last:%lastn", $last);

?>