Optimizing MySQL and a Zend_Db_Profiler example

published on October 13, 2008.
Heads-up! You're reading an old post and the information in it is quite probably outdated.

Last night I came across on a post on TechFounder, about using Zend_Db_Profiler and a good example with real data on optimizing MySQL queries. For “geeks” who SQL speak fluently this will probably be no new stuff, but for great number of web developers (me included!) this will probably come in handy.


Starting with Zend Framework

published on October 07, 2008.
Heads-up! You're reading an old post and the information in it is quite probably outdated.

Zend Framework is a big & heavy object-oriented framework for PHP. I started working with ZF a couple of months ago, I liked it’s documention (it’s very well documented) and decided to stick with this framework. Here is the latest version of the framework — at the time of writing v1.6.1.

It supports the MVC pattern, which helps separating business logic from viewing logic. It supports a great number of API’s, such as Delicious API, Flickr API, Yahoo API, Akismet API and many more.

The advantages of using a framework is that it is enforcing the developer to write code using a coding standard, it is well documented and well supported, and it is a lot easier to work in a 2+ person team using a framework. If you are a one—man team, someday you may want to add more developers to your projects; the process of their settling in will be very comfortable if you are using a framework.

Choose yourself a framework that best suits your needs, or write your own (be sure to make good documentation, also!). To be honest, I wasn’t looking at other frameworks, just ZF, but I knew right away that it is good for me. Prior to this post I did a little research on other frameworks, and I’m still sure that I made the right choice by choosing ZF.

You can read a bit more about ZF in general on the overview page.

How does it work?

Before anything, we should take a look how does the ZF work, when used in the MVC manner. ZF has a thingy, called Front Controller. When a user is accessing a web page, the Front Controller is called: it’s determining what should be done with the input and which further objects should be instantiated and methods called, and in what order.

E.g., if one makes a page request like: http://example.com/news/last/, first, the Front Controller is called. The Front Controller sets up the environment, loads up some files and classes, etc., then it calls a controller called “News” and an action called “Last” which is to be found inside the “News” controller. If it fails to find the “News” controller or the “Last” action, than it can show the user some error page, or to print out the error itself, depending how it is set up. If everything is OK, then it shows the user the content…

This explanation is very basic, as I intend to dedicate one big post to the Front Controller itself, going deep into details…

Some terms explained

Bootstrap file: all page requests are routed through this file, the Front Controller object is created here.

A module is a part of an application which has it’s own controllers, actions, view scripts, models, configuration files. For example, a page can have a default module and a blog module, where each module has its own Index Controller, Administrator Controller, and have its own unique controllers, like a Comments Controller for the blog module.

A controller is a class which has its own actions and can have its own functions. It controls the data received from the user or from the database, and decides what to do with it. The controller is responsible for one set of things, e.g. a News Controller would list latest news, list news from a particular source, show the archive, etc.

An action is a function inside a controller, which is responsible for doing some action, e.g. action for showing news.

A model receives data from the Controller, and sends data to the Controller. Database related stuff — selecting, inserting, updating, deleting — should be only in the model. Filtering data that is to be inserted into the database should be done in the Controller, not in the model.

A view script is responsible to show the data received from the Controller to the user.

A view helper script is to help to do some automating in the view scripts, like formatting dates, generating form elements, etc.

Just for the record, in further examples, “Dummy” will be referring to a module, “Foo” will be referring to a controller inside the “Dummy” module and “Bar” will be referring to an action inside the “Foo” controller.

Basic file structure

Here’s an example of a file structure for a ZF based application — after the # sign are comments:

|  |--Zend/ # Zend core
|--application/ # Core of our application
|   |--default/ # The Default module
|       |--config/ # Some configuration files
|          |--config.ini
|       |--controllers/ # Controllers go here
|          |--IndexController.php
|          |--FooController.php
|       |--models/ # Models...
|          |--ModelName.php
|       |--views/ # View related stuff...
|          |--helpers/
|          |--scripts/
|             |--index/ # View files for the Index Controller
|               |--index.phtml # For the default index action
|             |--foo/ # View files for the Foo Controller
|               |--index.phtml # For the default index action
|               |--bar.phtml # For a bar action in the Foo Controller
|             |--layout.phtml # For layout
|   |--dummy/ # A Dummy module...
|       |--config/
|          |--config.ini
|       |--controllers/
|          |--IndexController.php
|          |--FooController.php
|       |--models/
|       |--views/
|          |--helpers/
|          |--scripts/
|             |--index/
|               |--index.phtml
|             |--foo/
|               |--index.phtml
|               |--bar.phtml

With this file structure, http://example.com/ should point to the public folder; this way, the application or the library can not be accessed through the browser, which improves security of the application.

The .htaccess file

The .htaccess file’s responsibility is to route requests to existing resources (existing symlinks, non-empty files, or non-empty directories) accordingly, and all other requests to the front controller. Example:

RewriteEngine on
RewriteCond %{REQUEST_FILENAME} !-f
RewriteRule .* index.php

The bootstrap file

The biggest problem is setting up correctly the bootstrap file. Here’s an example of my bootstrap file, I use it on several projects, never had any problems :)

* This is a general bootstrap file, change it to fit your needs
* Pay attention to the paths
ini_set('display_errors',1); // set this to 0 on live version

// This is my timezone, change it to yours
// See timezones here: http://www.php.net/timezones

* We need to set some include paths
* To the library
* And to the models
* And add it to the current include path
set_include_path('.' . PATH_SEPARATOR . '../library' .
					   PATH_SEPARATOR . '../application/default/models' .
                       PATH_SEPARATOR . '../application/dummy/models' .
					   PATH_SEPARATOR . get_include_path());


* This little fella loads up a class when needed
* So we don't need to bother with including class files

* This config part is needed only when you
* store stuff for db connections in a .ini file
* I do it this way all the time, so it's a part of my general bootstrap
$config = new Zend_Config_Ini('../application/default/config/db_config.ini', 'offline');
$registry = Zend_Registry::getInstance();

// Only needed if you plan to use layouts in your app

* Get an instance of the Front Controller
* Tell him where to look for controllers
* And off we go!
$frontcontroller = Zend_Controller_Front::getInstance();
        'default'   =>  '../application/default/controllers',
        'dummy'       =>  '../application/dummy/controllers'
$frontcontroller->dispatch(); // GO!!!

This kind of bootstrap file should be enough in most cases; it is for me.

This post is starting to get out of control, so I’ll stop here for now. Next time I’ll show some basic stuff with controllers, actions, views etc. Until then be sure to get familiar with the coding standard, especially with the naming conventions.

Hope that this text isn’t too confusing. I tried to keep it simple and explain all that is needed for starting with Zend Framework.

Any thoughts on ZF, or frameworks in general? Do you use any?

Wordpress paging navigation

published on October 06, 2008.
Heads-up! You're reading an old post and the information in it is quite probably outdated.

As I’m not a big fan of Wordpress plug—ins, and I wanted to use a normal page navigation, not just the default “Previous posts” and “Next posts”, I decided to play around a bit and create my own paging navigation, or pagination.


First, I wrote on a piece of paper which links I need: first page, last page, next page, previous page and the links with the page numbers. Next, I needed to see what functions are already in Wordpress, to reuse as much as I can. After a little searching, I found that the functions for the default navigation are located in the link-template.php file, under the wp-includes folder. There are the functions for the next and previous pages, and the function that creates the URL. Furthermore, I wanted a sliding pagination (like Yahoo has on it’s search page), ‘cause it’s easy to use and looks cool.

The function

So, let’s take a look at the code. I called the function simply get_pagination; it’s quite self—describing. I put it in the link-template.php file, that way, all functions for navigation are in one place.

* A pagination function
* @param integer $range: The range of the slider, works best with even numbers
* Used WP functions:
* get_pagenum_link($i) - creates the link, e.g. http://site.com/page/4
* previous_posts_link(' &laquo; '); - returns the Previous page link
* next_posts_link(' &raquo; '); - returns the Next page link
function get_pagination($range = 4){
  // $paged - number of the current page
  global $paged, $wp_query;
  // How much pages do we have?
  if ( !$max_page ) {
    $max_page = $wp_query->max_num_pages;
  // We need the pagination only if there are more than 1 page
  if($max_page > 1){
      $paged = 1;
    // On the first page, don't put the First page link
    if($paged != 1){
      echo "<a href=" . get_pagenum_link(1) . "> First </a>";
    // To the previous page
    previous_posts_link(' &laquo; ');
    // We need the sliding effect only if there are more pages than is the sliding range
    if($max_page > $range){
      // When closer to the beginning
      if($paged < $range){
        for($i = 1; $i <= ($range + 1); $i++){
          echo "<a href='" . get_pagenum_link($i) ."'";
          if($i==$paged) echo "class='current'";
          echo ">$i</a>";
      // When closer to the end
      elseif($paged >= ($max_page - ceil(($range/2)))){
        for($i = $max_page - $range; $i <= $max_page; $i++){
          echo "<a href='" . get_pagenum_link($i) ."'";
          if($i==$paged) echo "class='current'";
          echo ">$i</a>";
      // Somewhere in the middle
      elseif($paged >= $range && $paged < ($max_page - ceil(($range/2)))){
        for($i = ($paged - ceil($range/2)); $i <= ($paged + ceil(($range/2))); $i++){
          echo "<a href='" . get_pagenum_link($i) ."'";
          if($i==$paged) echo "class='current'";
          echo ">$i</a>";
    // Less pages than the range, no sliding effect needed
      for($i = 1; $i <= $max_page; $i++){
        echo "<a href='" . get_pagenum_link($i) ."'";
        if($i==$paged) echo "class='current'";
        echo ">$i</a>";
    // Next page
    next_posts_link(' &raquo; ');
    // On the last page, don't put the Last page link
    if($paged != $max_page){
      echo " <a href=" . get_pagenum_link($max_page) . "> Last </a>";

The “range” is the range of the sliding effect, i.e. how many numbers are shown besides the current number: if the range is 4, and the current page is 5, then the numbers 3, 4, 5, 6 and 7 are visible.


It’s quite simple to use it: where the pagination is needed, just call the get_pagination() function, and it will show up. Add some CSS style to it, and your good to go.

Hope someone will find this useful :)

Smush your images!

published on September 30, 2008.
Heads-up! You're reading an old post and the information in it is quite probably outdated.

I just found a nice web site where you can “smush” your images — Smushit.com. SmushIt takes an image and removes all unnecessary information about it: when was it last edited, what image editor was used etc., but keeps the quality of the image! This is more than useful for sites where there are lots of images.

There are several ways to provide images to SmushIt:

  • Upload an image
  • Provide an URL to the image
  • Use the Firefox SmushIt add-on

The first two ways are quite obvious; provide an image and it’ll process it in a few seconds.

The Firefox add-on is pretty cool: open up a web page where are the images you want to smush, click the SmushIt add-on icon (it’ll be in the right corner of the status bar), it will take you to their site and process all the images found on your web page.

When the processing is complete, there will be a table showing details about the smushing. Also a download link will be provided, to download the smushed images in one zip file.

Now run little lurker and smush your images :)

Regular expressions with PHP

published on September 22, 2008.
Heads-up! You're reading an old post and the information in it is quite probably outdated.

I just want to write some real examples. These regexps are (and always will be, ‘cause I plan to write several posts on this topic) for the PHP’s PCRE library. Here’s a good PHP PCRE cheat sheet, it’s an excellent resource for regexps. If you know nothing about regexps, first read this Wiki page.

Regexps for anchor tags

A common case is when you have a source of some web page and you want to parse out all the links from it.
An anchor tag goes something like this:

<a href="http://example.com/" title="Some website">Website</a>

Also it can have more attributes, like class, target etc. Knowing how it’s built up, we can start writing a pattern, depending on what we want.
Here are some examples, some explanations are in the comments:

// Regexp examples for <a> tags

* Different combinations...
* $matches_comb[0] contains the whole <a> tag
* $matches_comb[1] contains what's inside the "href" attribute
* $matches_comb[2] contains what's after <a> and before </a>
* with the "s" modifier mathces <a> tags that are broken in several lines,
* ie. matches <a> tags with newlines
* without the "s" modifier, matches only <a> tags without a newline

* Match only what's inside the href attributes...

* Match only what's inside the href attirbutes,
* only when it starts with http:// and includes http://
* $mathces_href_http[0] contains some trash also, nevermind,
* $mathces_href_http[1] contains exactly what we need

* Match all Email addresses - mailto:


Play around with these patterns, see what’s for what, experiment, that’s the best way to learn regexps.

Do you have some more regexps for links? Some better ones than these here?

Happy hacking!

Tags: example, pcre, php, regex, regexp.
Categories: Development, Programming.
Robert Basic

Robert Basic

Software engineer, consultant, open source contributor.

Let's work together!

If you require outsourcing or consulting help on your projects, I'm available!

Robert Basic © 2008 — 2019
Get the feed