Server Side Nested Navigation Script

Abstract

The purpose of the script is to recursively parse the z-nav.inc files published to the production server to generate nested navigation structure reflecting the navigation.

An inital navigation file is loaded and parsed for <li> nodes.

Given the /z-nav.inc starting point:

/z-nav.inc contents
<li><a href="/index.php">Home</a></li>
<li><a href="/section_a/index.php">Section A</a></li>
<li><a href="/about.php">About</a></li>

The script will simply render out the first link, then see the link to a sub-section on the second link and attempt to load the navigation file for that section.

/section_a/z-nav.inc contents
<li><a href="/section_a/freshmen_college.php">Why college?</a></li>
<li><a href="/section_a/freshmen_ready.php">Get ready</a></li>

The script will render out the links in the section_a navigation file and then proceed to the third link of the inital navigation file. This happens recursively resulting in the following output:

<ul>
    <li><a href="/index.php">Home</a></li>
    <li>
        <a href="/section_a/index.php">Section A</a>
        <ul>
            <li><a href="/section_a/freshmen_college.php">Why college?</a></li>
            <li><a href="/section_a/freshmen_ready.php">Get ready</a></li>
        </ul>
    </li>
    <li><a href="/about.php">About</a></li>
</ul>

Reference

The entry point of the script expects a query string parameter $nav to define the path of the navigation starting point. If no path is given /z-nav.inc is the default.

parse_str($_SERVER['QUERY_STRING']);
$nav = (isset($nav)) ? $nav : "/z-nav.inc";

Change this value to reflect your navigation file names.

The path is also converted to the servers local file system path.

$server_file_path = $_SERVER["DOCUMENT_ROOT"] . $nav;

You can set how many levels deep the script should process by setting the $nest_limit.

$nest_limit = -1;

Set to -1 for limitless nesting.

Then the script initializes the variable $html as the opening root <ul> node.

$html = "<ul>";

process_nav_file()

The inital call to process_nav_file is then executed given the path to the first z-nav.inc file

process_nav_file($server_file_path);

This function maintains the current nesting level upon entry and exit of it's scope:

$nest_level++;
...
$nest_level--;

If the given navigation file exists a regex pattern is defined to identify <li> nodes within it.

$regex = '/(<li[^>]*?>[\s]*?<a[^>]*?href="([^"]+?)">[\s\S]*?<\/a>[\s]*?<\/li>)/';

Using the regex callback function preg_replace_callback, each matched <li> node will be passed as the argument to a call to match_li. In other words, match_li will be called once for each <li> node in the navigation file.

if(preg_match($regex, $file))
    preg_replace_callback($regex,"match_li",$file);

match_li()

A gloabl variable $html is mainted during the recursion containing the nested navigation HTML.

Another regex pattern is defined to determine if the current <li> node contains a link to a sub-section, that being a link ending in the default page name index.html|default.aspx or a slash /.

$regex = '/<li[^>]*?>[\s]*?<a[^>]*?href=\"([^"]*?\/)(index\.php)?\"[^>]*?>([\s\S]*?)<\/a>[\s]*?<\/li>/';

you will change the portion (index\.php) according to your implementation.

The current <li> node passed into the function is then tested with a regex to determine if it's a section link.

if(preg_match($regex, $li))
    preg_replace_callback($regex,"match_section_li",$li);
else
    $html .= $li;

If it is <li><a href="/section_a/index.php">Section A</a></li> the match_section_li function is called with the current <li> as the paramater, if it's not it is simply rendered out as the non section link that it is: <li><a href="/about.php">About</a></li>

match_section_li()

Global variables are made available to the function scope.

global $html;
global $current_directory;
global $nest_level;
global $nest_limit;
global $processed_section;

The regex defined in match_li uses 3 capture groups so that portions of the section link can be copied into local variables.

$li = $matches[0];
$href = $matches[1];
$page = $matches[2];
$title = $matches[3];

The href value of the section link is translated into a server file system path.

$file_path = $_SERVER["DOCUMENT_ROOT"] . $href . "z-nav.inc";

And we start off by allowing this section link to cause a recursive callback to the process_nav_file function.

$process_nav = true;

Many conditions will cause the section link to simply be copied as is without the z-nav.inc file in the section this <li> links to being opened.

//check if file exists
if(file_exists($file_path)){
    $file = file_get_contents($file_path);

    //make sure the file contains at least one <li>
    //so an empty <ul> is not generated
    $regex = '/(<li[^>]*?>[\s]*?<a[^>]*?>[\s\S]*?<\/a>[\s]*?<\/li>)/';

    if(!preg_match($regex, $file))
        $process_nav = false;
}
else
    $process_nav = false;  //no _nav in section

//check if this is the current sections index page link
if(!strcmp($href, $current_directory))
    $process_nav = false;

//check if class=ou-no-subnav within the li's node
if(strpos($li, "class=\"ou-no-subnav\"") !== false)
    $process_nav = false;

if($nest_level == $nest_limit)
    $process_nav = false;

//check if we've already processed this section before (cross-link)
if(in_array($href, $processed_section))
    $process_nav = false;

If all the conditions fail we can process the section's z-nav.inc file.

Keep track of processed z-nav.inc files so that they are only processed once.

if($href != '')
    $processed_section[] = $href;

The current link is rendered 'manually' instead of simply being copied from the z-nav.inc file.

$html .= "<li><a href=\"{$href}{$page}\">{$title}</a><ul class=\"level-{$nest_level}\">";

It will contain the nested <ul>

Next the call to process_nav_file is executed moving this documentation to the process_nav_file section.

Last Updated: 8/7/24