SEO URLs without mod_rewrite and FilesMatch

MyEgo.cz

home foto blogy mywindows.cz kontakt

SEO URLs without mod_rewrite and FilesMatch

Technologie 31.10.04

Many CMSs solve the need for Fancy URLs (URLs that are easy to follow by the user, and favoured by Search Engines as well) by using mod_rewrite, and/or FilesMatch (available only on Apache), or both.

Common solutions:

E.g., to get $itemid from URL like http://../article/this-is-a-title we need to create following .htaccess file:

<FilesMatch "^article$"> 
 ForceType application/x-httpd-php 
</FilesMatch> 

In this example we are forcing /article/ part of the URL to be processed as a PHP application.

Next, we will create a new PHP file called article (without .php extension), which will process the URL, and determine proper article number, and pass the control to our CMS.

<?php
// some internal function
include ('somecommonfncs.php');
// get article id
$itemid = getItemIdFromName($_SERVER['REQUEST_URI']);
// and process it
processIt();
?>

Seems easy, right? Well, this solution has several big problems in real-life, especially when open-source projects and code portability are concerned.

Problems:

  • FilesMatch is not available under Microsoft IIS, it works fine only with Apache, thus this solution may not run at your selected web-hosting company (if they do not use Apache),
  • Even if your webhosting company does use Apache HTTP server, you may not be able to access/create .htaccess, or it costs you more money to be able to modify it,
  • Some browsers (namely, MSIE) behave in a very strange way when downloading CSS files again for URLs modified in a runtime by FilesMatch,
  • Google Page Rank (GPR) is different for the domain, and for the article fancy URL, if somebody is linking to your article, it does not even fully add to Google Page Rank value of a domain!

Solutions:

Instead of using FilesMatch, and .htaccess file, we will use simple $_GET variables. Our final URL will look like:

http://domain.com/?article=this-is-a-title

And code will look like:

<?php
// some internal function
include ('somecommonfncs.php');
// get article id
$data=$_GET['article'];
if (isset($d)) $itemid = getItemIdFromName($data);
// and process it
processIt();
?>

Same approach can be applied for member ids, archivelists ids, category ids, and many others.

This approach does have many advantages and no short-commings:

  • This SEO-friendly solution does work with any web-server configuration, be it Apache, and/or IIS, with or without mod_rewrite,
  • Google Page Rank for your domain increases, even when linking to an individual article,
  • Browsers (MSIE) do not reload CSS files again and again,
  • You can safely use international characters in URL parameters, thus improving your Page Rank even more.

You can see example of these new fancy URLs at http://demo.blogcms.com/.

BLOG:CMS is an open-source Personal Content Management System, licensed under GNU GPL, developed by Radek Hulán.


Komentáře

  1. 1 Adam Hošek 01.11.04, 15:40:25
    FB

    Well, I've read somewhere that many Search Engines (but Google) don't really count with URLs with '?'; (the one before Query String :)). I don't know how it is true, but I think your solution is one of the best if you can't use mod_rewrite or anything.
    I've thought about it, too. Have not implemented by myself but seen on some websites already: why not use the same way as you do but: http://web/?/article/the-title/. Or other type of customized Query String like http://web/?/article:the-title. The one with slashes is more like the use of mod_rewrite. What do you think about it?

  2. 2 Ondřej Kokeš 01.11.04, 16:15:57
    FB

    Your code:
    <FilesMatch "^article$">
    ForceType application/x-httpd-php
    </FilesMatch>
    I don't know why it's like this. If you "parse" just one file, you don't have to use regexp. So my version:
    <Files article>
    ForceType application/x-httpd-php
    </Files>

  3. 3 Radek Hulán 01.11.04, 16:20:30

    [2] this examples's taken from BLOG:CMS source code, and simplified, it does make sense in the whole picture ;-)

Nový komentář

Pro přidání komentáře se musíte nejdříve  registrovat Facebookem

Banan.cz