smarty 3

Intended Audience

This article is intended for PHP programmers and HTML designers interested in applying a new technique for web development – PHP templating. Advanced knowledge of PHP programming and HTML is assumed.

Smarty3 Overview

The theoretical web development process is that: first the designer makes the interface, and breaks it down into HTML pieces for the programmer then the programmer implements the PHP business logic into the HTML.

That’s fine in theory, but in practice, from my experience, the client frequently comes with more requirements, or maybe more modifications to the design or to the business logic. When this happens , the HTML is modified (or words rebuilt ) programmer changes the code inside HTML.

The problem with this scenario is that the programmer needs to be on stand-by until the designer completes the layout and the HTML files. Another problem is that if there is a major design change then the programmer will change the code to fit in the new page. And that’s why I recommand Smarty. Smarty is a templating engine for PHP.

You can download it from http://www.smarty.net or svn checkout http://smarty-php.googlecode.com/svn/branches/Smarty3Alpha/

The installation process is very simple. Just read the documentation and follow up the instructions.

So what is Smarty ? Smarty is a set of PHP classes that compile the templates into PHP scripts. Smarty is a template engine and a very useful tool for designers and programmers.
Smarty for Designers

Designers work with HTML files. To work with Smarty, you work with template files. These files are are made up of static content but combined with Smarty mark-up tags. All the template files have a .html or .tpl extension. The Smarty template tags are enclosed within { and } delimiters or custom delimiters which can be defined or bypassed in the class.

Let’s consider the basic structure of a web page. There is a header, a middle part, and a footer. A template file that includes the header and the footer looks like this:

{include file=”header.tpl”}
<form name=”form1″>
Label1 <input type=”text” name=”text1″>
<input type=”submit” value=”submit”>
</form>
{include file=”footer.tpl”}

All the templates should reside in a single template directory. After calling a template for the first time, the compiled template will reside in templates_c.

Smarty language is very poweful. All the variables that come from PHP are identified in Smarty with {$Variable_Name} (we precede them with a $ sign). So if we have a variable in PHP that is called $MyName, then to print it in Smarty you have to write something like:

<html>
<body>
Welcome, {$MyName} <br>
</body>
</html>

The power of Smarty lies also in its flexibility. You can insert IFs and LOOPs into the template. The syntax for IF is:

{if <condition> }
html code
{else}
html code
{/if}

Let’s say you have a dynamic menu based on links. Depending on the link you click, you go to a specific page. So you get from PHP a variable $Menu with a integer value, depending on the page you are. The template looks like :

{if ($Menu == 1) }
Option 1
{else}
<a href=”option1.php”>Option 1</a>
{/if}
{if ($Menu == 2)}
Option 2
{else}
<a href=”option2.php”>Option 2</a>
{/if}

For coding a loop let’s suppose you get an array like the following from PHP :

<table>
<tr
{section name=user loop=$userID}
{if $smarty.section.user.iteration is odd}
bgcolor=#efefef
{else}
bgcolor=#ffffff
{/if}
>
<td> ID = {$userID[user]}  </td>
<td> Name = {$name[user]}     </td>
<td> Address = {$address[user]} </td>
</tr>
{sectionelse}
<tr>
<td>
There is no user.
</td>
</tr>
</section>
</table>

Iteration is an internal counter for Smarty. It helps us to know the current iteration of the section. I use this internal variable to make alternate row colors in the table by checking if current iteration value is odd or not.

An alternative for LOOPS is FOREACH which is used to loop over a single associative array.

<foreach from=$users item=current_user>
Name = {$current_user}
<foreachelse}
No user available.
</foreach>

The main difference between SECTION and FOREACH is that for SECTION you can start from a specific value, and can also set a step for the iteration, whereas for FOREACH you have to loop over all values.
Smarty for Programmers

The advantage for programmers is that they write the code in a PHP file without having to mix the instructions with HTML. Furthermore, if the designer changes the layout of a page the programmer doesn’t have to change the code to suit the new layout since the functionalities won’t change. You do your work in your files, assign to the templates all the values needed to print on the site and go out for a beer. You won’t get phone calls asking you to change a bit of code because the designer changed the layout and now a set of internal errors cropped up.

In the PHP file you need to include the Smarty class require ‘Smarty.class.php’. After that you need to instantiate the smarty with $smarty = new Smarty.

To assign a variable to the template you need to $smarty->assign(‘UserName’, ‘John Doe’). After everything is finished you call the method to display the template $smarty->display(‘index.tpl’).

A sample code looks like this (index.php) :

<?php
require ‘Smarty.class.php’;
$smarty = new Smarty;

$smarty->assign(‘Username’, ‘John Doe’);
$smarty->display(‘index.tpl’);
?>

The template (index.tpl) looks like this:

<html>
<body>
Welcome {$Username}
</body>
</html>

You can also create an array in PHP an pass it to the template:

$tmp = array ( ‘UID’=> ’10′,  &’Name’ => ‘John Doe’, ‘Address’=>’Home address’);
$smarty->assign(‘info’, $tmp);

Sample Script

This script connects to a local database and select all the products from the ‘Products’ table. Then it passes all the values to the template, which prints them on the screen.

INDEX.PHP

<?php
require ‘Smarty.class.php’;
$smarty = new Smarty;

$hostname = “localhost”;
$dbUser = “sqluser”;
$dbPass = “sqlpass”;
$dbName = “sqldb”;
// connect to the database
$conn = mysql_connect($hostname, $dbUser, $dbPass) or die(“Cannot connect to the database”);

mysql_select_db($dbName);

$sql = “SELECT product_id, info FROM products ORDER BY product_id ASC”;
// get all the products from the table
$res = mysql_query($sql);
$results = array();
$i=0;
while ($row=mysql_fetch_array($res)) {
$tmp = array(
‘product_id’ => $r['product_id'],
‘info’=> $r['info']
);
$results[$i++] = $tmp;
}
// pass the results to the template
$smarty->assign(‘results’, $results);
// load the template
$smarty->display(‘index.tpl’);
?>

INDEX.TPL

<html>
<body>
Here’s a table with the results: <br>
<table cellpadding=1 cellspacing=0 border=0 width=100%>
{section name=nr loop=$results}
<tr {if $smarty.section.nr.iteration is odd} bgcolor=”#efefef”{/if}>
<td width=15%>
<nobr><a href=&#8221;show-product.php?id={$results[nr].product_id}”>Press here</a>

<td width=29%><a href=”show.php?id={$results[nr].prodID}”
{popup inarray=$smarty.section.nr.iteration}
>{$results[nr].info}</a></td>
</tr>

{sectionelse}
<tr><td align=”center”><br><b>no product </b> <br> </td></tr>
{/section}

</table>

<br>

Here’s a select with the results: <br>
<select name=”mys”>
{section name=nr loop=$results}
<option value=”{$results[nr].product_id}”>{$results[nr].info}</option>
{/section}
</select>

</body>
</html>
Summary

Smarty is a great tool for both designers and developers. By using Smarty you can reduce the site development and maintenance times. If you are a developer you no longer need to mix PHP code with HTML code. Just take care of business logic and leave the HTML to the designer.

svn + apache2

Many of my developers work from windows including myself. Creating secure access over ssh and svn (svn+ssh://)
using Zend studio and svn (subversion plugin for zend) was the initial idea, from a quick pass over subversion, but once svnserver was up and running and I tried it, this wasn’t as simple. There are several articles on integrating subversion, ssh and windows. None of the solutions looked simple or elegant. And wide open unsecured traffic was not acceptable.

I determined best practice pointed us to subversion access through web_dav_svn -> web_dav -> apache2 (https://). This brought authentication away from centralized auth and allowed fine grained control over the access granted via htpasswd and .htaccess files. It allowed us to restrict unsecured access and redirect http to https, and encrypt using SSL. It kept us from having to work ssh onto windows for each developer, instead we could have a simple eclipse/subclipse plugin access demonstrated and from that point they were able to customize on their own.

What I wanted was:

* untar subversion-1.6.6 and subversion-deps-1.6.6 in /apps/src (they layer over each other)
* first build and install the -deps apr and apr-util into /apps/local
* then build apache2 against /usr/local/apr and /usr/local/apr-util
* install apache2
* build and install serf
* remove serf, apr, and apr-util subdirectories and source code from within subversion-1.6.6
* build 1.6.6 against apxs in apache, without Berkeley DB, without neon, and specifying /usr/local/apr, /usr/local/apr-util, and /usr/local/serf
* install and test

process

* create subdirectory /apps/src, place all tarballs in this directory
* untar subversion & subversion-deps version 1.6.6 (these tar onto each other)
* cd subversion-1.6.6/apr
* ./configure –prefix=/usr/local/subversion/apr
* make && make install
* cd ../apr-util
* ./configure –prefix=/usr/local/subversion/apr-util –with-apr=/usr/local/subversion/apr
* make && make install
* cd ../neon
* ./configure –prefix=/usr/local/subversion/neon
*
* apache2: untar httpd-2.2.14
* ./configure –prefix=/usr/local/subversion/apache –enable-dav –enable-dav-fs –enable-dav-lock –enable-expires –enable-headers –enable-info –enable-logio –enable-proxy –enable-rewrite –enable-unique-id –with-apr=/usr/local/subversion/apr –with-apr-util=/usr/local/subversion/apr-util –enable-so –enable-mods-shared=all
* make && make install

This way apache2 builds against apr and apr-util compatible with subversion 1.6.6
and then build subversion against it as well..
* ln -s /apps/apache2_2.2.14 /apps/apache2
* compile and install serf
* ./configure –-prefix=/usr/local/subversion/serf –with-apr=/usr/local/subversion/apr –with-apr-util=/usr/local/subversion/apr-util
* make && make install
* remove serf, pr and apr-util from subversion

compile subversion

./configure –prefix=/apps/svn –with-ssl –with-libs=/usr/local/ssl –without-berkeley-db –with-apxs=/apps/apache2/bin/apxs –with-openssl=/usr/local/ssl –without-neon –with-serf=/usr/local/serf –with-apr=/usr/local/apr –with-apr-util=/usr/local/apr-util

make && make install

test:

/usr/local/subversion/svn/bin/svnadmin create /usr/local/subversion/repository

chown subversion:subversion /usr/local/subversion

chown -Rv subversion.subversion /usr/local/subversion

root@dedicated[/bin] $

no core dump…

ONWARD to configure and test apache2 and subversion…
httpd.conf:

* change all references to apache2_2.2.14 to apache2 (makes the httpd.conf generic rather than subject to needing a migration after a point release upgrade…)
* change port 80 to a non-priveliged port (8080)
* check for
o LoadModule dav_module modules/mod_dav.so
o LoadModule dav_module modules/mod_dav.so
o LoadModule ssl_module modules/mod_ssl.so
* add in ServerName hostname.domain.com:8443 Some of the apache level sanity validation requires a statement of the local host.
* add in SSL stuff (this IS httpd from source – the default httpd.conf had the ssl-module load statement, but no explicit SSL configuration

#
# Note: The following must must be present to support
# starting without SSL on platforms with no /dev/random equivalent
# but a statically compiled-in mod_ssl.
#

<IfModule ssl_module>
SSLRandomSeed startup builtin
SSLRandomSeed connect builtin
</IfModule>

# =================================================
# SSL/TLS settings
# =================================================

Listen 0.0.0.0:8080
Listen 0.0.0.0:8443

SSLEngine on
#SSLOptions +StrictRequire

#<Directory />
# SSLRequireSSL
#</Directory>

SSLProtocol -all +TLSv1 +SSLv3
SSLCipherSuite HIGH:MEDIUM:!aNULL:+SHA1:+MD5:+HIGH:+MEDIUM

SSLMutex file:/apps/apache2/logs/ssl_mutex

SSLRandomSeed startup file:/dev/urandom 1024
SSLRandomSeed connect file:/dev/urandom 1024

SSLSessionCache shm:/apps/apache2/logs/ssl_cache_shm
SSLSessionCacheTimeout 600

SSLPassPhraseDialog builtin
SSLCertificateFile /apps/apache2/conf/ssl.crt/server.crt
SSLCertificateKeyFile /apps/apache2/conf/ssl.key/server.key

SSLVerifyClient none
SSLProxyEngine off

<IfModule mime.c>
AddType application/x-x509-ca-cert .crt
AddType application/x-pkcs7-crl .crl
</IfModule>

* create ssl certificate (self-signed). I installed in /apps/apache2/conf/ssl.crt and ssl.key, naming the .crt and ,key files for the server hostname and then symbolically linking them to the generic “server.crt” and “server.key”.
* restart apache2 and test https://host:8443/ – you should get the “It works!” apache test page, thus validating the SSL certificate and setup from a browser level
*
* parent directory for svn

<Location /svn>
DAV svn
SVNParentPath /apps/repos
</Location>

Preferred method to store PHP arrays (json_encode vs serialize)

I need to store a multi-dimensional associative array of data in a flat file for caching purposes. I might occasionally come across the need to convert it to JSON for use in my web app but the vast majority of the time I will be using the array directly in PHP.

Would it be more efficient to store the array as JSON or as a PHP serialized array in this text file? I’ve looked around and it seems that in the newest versions of PHP (5.3), json_decode is actually faster than unserialize.

I’m currently leaning towards storing the array as JSON as I feel its easier to read by a human if necessary, it can be used in both PHP and JavaScript with very little effort, and from what I’ve read, it might even be faster to decode (not sure about encoding, though).

Does anyone know of any pitfalls? Anyone have good benchmarks to show the performance benefits of either method?

Launch of Dazzle CMS Blog

First off i like to welcome you all to our new blog, so you get too see what been under development and how we have progressed,

as we start looking at getting more developers working beside us im sure they post some of there status updates too.

as up until now the project has been off / on for some time

but now has been under constant development for a few months now

all our old code that we start back when we first started the project was when php 5 was released and now its been slowly updated and written how php 5 code should be and now does require php 5.3 but this could change to atleast 5.2 as im thinking of implementing some wrappers for the 5.3  code

the project itself has been alot inflused by phpnuke, & nuke-evolution and the whole reason behind this is for a while i worked with phpnuke and even worked with the developer of phpnuke when phpnuke 7.7 was released

and so much i wanted to change about phpnuke but i had no control of major core changes which i really wanted todo with phpnuke for sometime

so what i did was to rewrite how phpnuke functions and make everythign better and easier to use

for example the biggest part of control of an CMS is design wanted to be able to design every aspect of the portal this means chaneg the looks and feel of the modules as well as the site

so smarty was used as our template engine so smarty is wrapped into our template engine but it goes behond templates almost every aspect is assigned into smarty from form elements , as for the form elements you can control the html type too as some elements only work under a particlar version of doctype for example xhtml / html 4.0.1 & the new html5

every major class this includes sessions, users, database, templates, filesystem etc are all plugin based you can extend any of the classes just by writing an new class and dropping it right into the plugin folder or into the modules class folder for example if its only used by the module you created

as soon we tied up the core features and frozen the feature releases we will post full complete list but we will post up an summary shorly

~Dazzle Development Team~