The PhP code is in <code>~www/php</code>.
<p>The source code resides in a svn repository on
<code>repo.eecs.berkeley.edu</code> in the <code>gsrc</code> project:</p>


<p>We do our work as the <code>www</code> user on
<code>moog</code>,
so there is no need for individual logins on <code>source</code>.

<p>That said, if you wanted to check out the php source
for some reason, use:</p>
<pre>
svn co https://repo.eecs.berkeley.edu/svn/projects/gsrc/php/
</pre>

<h2>How PhP Works on the sites</h2>
<ol>
<li>A http request comes in and
<code>/usr/local/apache/conf/httpd.conf</code> is 
consulted.  This file maps the different website
urls to different locations.   This file also
makes Apache aware of PhP files.
<li>We use <i>virtual hosts</i> to run multiple
websites on one server. <code>/usr/local/apache/conf/extra/httpd-vhosts.conf</code> contains sections for each 
host. 
<br>BTW - SSL connections are handled by
<code>/usr/local/apache/conf/extra/httpd-ssl.conf</code>.
SSL is used by the login page.
<br/>For each virtual host, the <code>httpd-vhosts.conf</code> file contains
<pre>
 Include "/usr/local/apache/conf/rewrite.conf"
</pre>
<li>The <code>/usr/local/apache/conf/rewrite.conf</code>
file handles rewrites for thinks like bugzilla
and the wiki.
The last line is:
<pre>
RewriteRule ^(.*)$ /home/www/php/get.php [L,NS]
</pre>
This line says that if the request has not been handled
elsewhere, read <code>/home/www/php/get.php</code>
<br/>The rewrite log is very large, but it can be found
at <code>/usr/local/apache/conf/rewrite_log</code>
If you need to view that file, you might want to
use <code>tail</code> and <code>grep</code>
<pre>
tail -10000 /usr/local/apache/logs/rewrite_log | grep events/11/symposium2011_notice.pdf
</pre>
The above command might return:
<pre>
128.32.48.155 - - [06/Oct/2011:11:52:59 --0700] [www.e3s-center.org/sid#3d7070][rid#7e4558/subreq] (2) init rewrite engine with requested uri /e3scenter//events/11/symposium2011_notice.pdf
128.32.48.155 - - [06/Oct/2011:11:52:59 --0700] [www.e3s-center.org/sid#3d7070][rid#7e4558/subreq] (1) pass through /e3scenter//events/11/symposium2011_notice.pdf
128.32.48.155 - - [06/Oct/2011:11:52:59 --0700] [www.e3s-center.org/sid#3d7070][rid#5fd0b8/initial] (2) init rewrite engine with requested uri /events/11/symposium2011_notice.pdf
128.32.48.155 - - [06/Oct/2011:11:52:59 --0700] [www.e3s-center.org/sid#3d7070][rid#5fd0b8/initial] (2) rewrite '/events/11/symposium2011_notice.pdf' -> '/home/www/php/get.php'
</pre>
Note that the above rewrites to <code>/home/www/php/get.php</code>, which is where most files end up.

<li><code>/home/www/php/get.php</code> looks at the
server portion of the URL and then includes the 
corresponding <code>get.php</code> file.
<br>The code looks like:
<pre>
if ($sites[$_SERVER['SERVER_NAME']]) {
    include "${_SERVER['SERVER_NAME']}/get.php";
} else {
    html_unknown($_SERVER['SCRIPT_URL']);
}
</pre>

If the URL is <code>http://chess.eecs.berkeley.edu</code>,
then
<code>$_SERVER['SERVER_NAME']</code> will be set to
<code>chess.eecs.berkeley.edu</code>, so 
<code>/home/www/php/chess.eecs.berkeley.edu/get.php</code>
will be read.
<br>Each of the separate websites has corresponding
<code>get.php</code> file.

<li>The layout for each website is controlled by php
files in the <code>layout/html.inc.php</code>
file, for example, <code>/home/www/php/chess.eecs.berkeley.edu/layout/html.inc.php</code>.

<li>The <code>/home/www/php</code> directory has two
major subdirectories:
<br><code>/home/www/php/include</code> - common includes,
such as arrays that define strings
<br><code>/home/www/php/subpages</code> - used for 
subpages such as the admin pages.

</ul>
</ol>

<h2>To Upgrade PHP</h2>
<a href="/dopsysadmin/faq/55.html">See this question</a>