PHP, the PHP Hypertext Processor, is an open source server-side scripting language for Web servers, which provides a real alternative to ASP, Cold-Fusion, Mod_Perl or JSP if your aim is to provide dynamic Web pages. Dynamic Web pages are pages which interact with the user, so that each user visiting the page sees customized information - which may vary each time and which may be based on a form they've just filled in, or on information extracted from a database or some other external source. Typical applications include e-commerce, online newspapers, visitors' books, ticketing systems, project management, and other groupware projects. The traditional way to produce this type of dynamic page is via CGI scripts, but these are separate programs which must be executed as a new process for each page hit, so they scale badly and rapidly become memory and processor hogs as server load increases.
PHP solves this problem by becoming a part of the Web server, essentially extending the functionality of the server itself, so that the server can do the processing without having to spawn extra processes. It's not alone in doing this, but unlike most other scripting languages for Web page development PHP also offers excellent connectivity to most of the databases in use today. Perhaps the greatest advantage of PHP, when compared to other scripting languages such as ASP or ColdFusion, is that it is open source and cross-platform. PHP's natural home is on Linux servers running Apache server software, but it runs equally well on any other Unix or Windows platform, and can be used with other Web servers.
PHP started life as a Perl program written by Rasmus Lerdorf to track visitors to his online resume. It was then rewritten in C and was extended to include support for database access. From these simple beginnings the open source community has expanded and developed PHP into a powerful server-side scripting language.
What sets PHP apart from its origins in Perl is the ability to embed PHP code inside HTML, allowing you to mix HTML and PHP in one source file rather than having to crank out the HTML from within Perl. For example, a simple PHP script to show today's date is shown in Figure 1 below. If you're familiar with HTML then the top and bottom of this code will look very familiar - it's just standard HTML.
PHP code is embedded within the HTML by bracketing it with "". The code consists of one line, a call to the PHP function "date", which returns a date in the format we've specified as a string, which is then output as part of the HTML using the "echo" function. This ability to embed PHP inside HTML means that you can continue to write your Web pages as you do now, and then embed PHP into them to provide the dynamic elements.
This is perhaps a good point at which to highlight what is meant by a server-side language. The PHP code is executed on the server, not the browser. The source of the page the browser receives back from the server is shown in Figure 2.
You can see that the PHP has been processed by the server, so no special software is needed on the client to view pages with PHP elements - the intelligence lies at the server end. This means you can use PHP with any browser - even simple, text-only ones like Lynx and w3m, or browsers in PDAs or set-top boxes. So long as the browser understands HTML it can display pages written in PHP.
In view of where it started, it's not surprising that PHP shares syntactical ideas with Perl and C, and if you've programmed in either of these or you've done any Unix shell programming, then PHP should be easy to pick up. PHP uses loosely typed variables which don't need to be declared before first use, and evaluates them according to context, so you can add a numeric string to a numeric to give a numeric, and then manipulate the result as a string. The PHP online manual gives a nice example of how a variable's type can change due to context. This is shown in Figure 3.
One of the joys of PHP is that some variables are already set for you in any PHP script. For example, the variable $HTTP_USER_AGENT will contain a string identifying the browser which requested the page. So, for example, if you're using Netscape 4.7 under NT4 to request the page, $HTTP_USER_AGENT will contain the string "Mozilla/4.7 [en] (Win NT; I)". So if you need to generate HTML which exploits browser-specific features it's easy to do. Similarly, any cookies you have set will automatically appear as variables in your script without you having to write any code to extract them. As well as simple variables PHP also lets you create arrays which act both like hash tables (associative arrays) and indexed arrays (vectors), and these arrays can be multi-dimensional.
PHP has a full range of comparative, logical and bitwise operators, auto increment and decrement, and assignment operators. It has all the control structures you might expect from a procedural language, including if/else, switch/case, for, while (test before) and do (test after). You can write your own functions or use the extensive range of functions provided as standard. Arguments can be parsed by value or reference, and functions can return a value - or multiple values if you use an array as the return variable. Variables are local to a function unless declared global.
PHP also supports include files, both static and dynamic, allowing common code such as custom function libraries and HTML fragments to be embedded in your code. This is all very good news from the developer's viewpoint, allowing code re-use wherever possible. It also means you can separate the PHP code from the HTML if you employ different people for Web design and Web programming. PHP3 includes support for object oriented programming, but there is a problem with polymorphism: once a parent's function is overridden by a child it can no longer be used. This is resolved in PHP4, which provides expanded functionality and new features for object oriented programming, and for building classes and objects.
One of the most powerful features of PHP is the ease with which you can handle forms. If you make your PHP script the action of the form then all the fields in the form appear as variables in your script automatically, ready for you to use. To take a simple example, suppose you have a page containing a form such as the one at the top of Figure 4. When the user clicks on the submit button the page action.php3 will be invoked to process that form. In action.php3 you might have the line of code at the bottom of Figure 4. As you can see, the input fields "age" and "name" have automatically become variables of the script action .php3, with names "$name" and "$age" ready for you to do what you like with them (in this case echoing them as part of the line of HTML).
The real strength of PHP lies in the extensive range of built-in functions. Those available include functions to:
Creating powerful Web applications inevitably means storing and manipulating a lot of data, so the ability to access databases is vital. PHP includes support, via function calls, for a wide variety of popular databases, including Adabase D, dBase, Unix dbm, filePro, Interbase, Informix, Microsoft SQL Server, mSQL, MySQL, Oracle, Sybase, PostgreSQL, and Solid. In addition, any database which supports ODBC, such as Microsoft Access, can be used. If you're trying to create an application which isn't database-specific then there are a number of open source abstraction layers available, including Metabase and phpDB, which will let you do this painlessly.
The home of PHP on the Web is www.php.net, and Zend Technologies have their own site at www.zend.com. These two sites between them contain links to an ever increasing list of other sites devoted to PHP. PHP is open source, so you can simply download PHP from many one of the many mirror sites. PHP also comes as a package in some Linux distributions.
If the target audience for your PHP-generated pages is the Web rather than your company intranet, then a growing number of ISPs can provide you with Web hosting on servers which support PHP, and an SQL database back-end, usually MySQL.
PHP4 is significantly faster than PHP3. This improvement in performance is particularly noticeable with larger and more complex scripts, and is the result of the PHPengine having been totally rewritten by Andi Gutmans and Zeev Suraski. The new Zend engine uses a much more efficient "compile-then-execute" method, instead of the "execute-while-parsing" model used by PHP3. The authors claim the resulting engine is up to fifty times faster than the PHP3 engine. Features in PHP4 include:
PHP4 also provides COM/DCOM support (on Windows only), allowing you to seamlessly access COM objects and instantiate them. PHP4 also has a number of additional optional functions which,with the right libraries, extend PHP to allow you to process credit cards and other financial transactions using Verisign Payment Services, and to handle Internet payments.
The figures in this article are poorly rendered in Internet Explorer owing to its odd rendering of white-space in CSS. If you need to see them clearly look at the source or get a proper browser!
[Back to Paul's homepage]