What is a PHP Filter?

A PHP filter is used to validate and filter data coming from insecure sources.

To test, validate and filter user input or custom data is an important part of any web application.

The PHP filter extension is designed to make data filtering easier and quicker.

Why use a Filter?

Almost all web applications depend on external input. Usually this comes from a user or another application (like a web service). By using filters you can be sure your application gets the correct input type.

You should always filter all external data!

Input filtering is one of the most important application security issues.

What is external data?

• Input data from a form
• Cookies
• Web services data
• Server variables
• Database query results

Functions and Filters

To filter a variable, use one of the following filter functions:

• filter_var() – Filters a single variable with a specified filter
• filter_var_array() – Filter several variables with the same or different filters
• filter_input – Get one input variable and filter it
• filter_input_array – Get several input variables and filter them with the same or different filters

In the example below, we validate an integer using the filter_var() function:

The code above uses the "FILTER_VALIDATE_INT"  filter to filter the variable. Since the integer is valid, the output of the code above will be: "Integer is valid".

If we try with a variable that is not an integer (like "123abc"), the output will be: "Integer is not valid".

Validating and Sanitizing

There are two kinds of filters:

Validating filters:

• Are used to validate user input
• Strict format rules (like URL or E-Mail validating)
• Returns the expected type on success or FALSE on failure

Sanitizing filters:

• Are used to allow or disallow specified characters in a string
• No data format rules
• Always return the string

Options and Flags

Options and flags are used to add additional filtering options to the specified filters.

Different filters have different options and flags.

In the example below, we validate an integer using the filter_var() and the "min_range" and "max_range" options:

Like the code above, options must be put in an associative array with the name "options". If a flag is used it does not need to be in an array.

Since the integer is "300" it is not in the specified range, and the output of the code above will be: "Integer is not valid".

Validate Input

Let’s try validating input from a form.

The first thing we need to do is to confirm that the input data we are looking for exists.

Then we filter the input data using the filter_input() function.

In the example below, the input variable "email" is sent to the PHP page:

Example Explained

The example above has an input (email) sent to it using the "GET" method:

1. Check if an "email" input variable of the "GET" type exist
2. If the input variable exists, check if it is a valid e-mail address

Sanitize Input

Let’s try cleaning up an URL sent from a form.

First we confirm that the input data we are looking for exists.

Then we sanitize the input data using the filter_input() function.

In the example below, the input variable "url" is sent to the PHP page:

Example Explained

The example above has an input (url) sent to it using the "POST" method:

1. Check if the "url" input of the "POST" type exists
2. If the input variable exists, sanitize (take away invalid characters) and store it in the $url variable

If the input variable is a string like this "http://www.W3ååSchøøools.com/", the $url variable after the sanitizing will look like this:

Filter Multiple Inputs

A form almost always consist of more than one input field. To avoid calling the filter_var or filter_input functions over and over, we can use the filter_var_array or the filter_input_array functions.

In this example we use the filter_input_array() function to filter three GET variables. The received GET variables is a name, an age and an e-mail address:

Example Explained

The example above has three inputs (name, age and email) sent to it using the "GET" method:

1. Set an array containing the name of input variables and the filters used on the specified input variables
2. Call the filter_input_array() function with the GET input variables and the array we just set
3. Check the "age" and "email" variables in the $result variable for invalid inputs. (If any of the input variables are invalid, that input variable will be FALSE after the filter_input_array() function)

The second parameter of the filter_input_array() function can be an array or a single filter ID.

If the parameter is a single filter ID all values in the input array are filtered by the specified filter.

If the parameter is an array it must follow these rules:

• Must be an associative array containing an input variable as an array key (like the "age" input variable)
• The array value must be a filter ID or an array specifying the filter, flags and options

Using Filter Callback

It is possible to call a user defined function and use it as a filter using the FILTER_CALLBACK filter. This way, we have full control of the data filtering.

You can create your own user defined function or use an existing PHP function

The function you wish to use to filter is specified the same way as an option is specified. In an associative array with the name "options"

In the example below, we use a user created function to convert all  "_" to whitespaces:

The result from the code above should look like this:

Example Explained

The example above converts all "_" to whitespaces:

1. Create a function to replace "_" to whitespaces
2. Call the filter_var() function with the FILTER_CALLBACK filter and an array containing our function

What is an Exception

With PHP 5 came a new object oriented way of dealing with errors.

Exception handling is used to change the normal flow of the code execution if a specified error (exceptional) condition occurs. This condition is called an exception.

This is what normally happens when an exception is triggered:

• The current code state is saved
• The code execution will switch to a predefined (custom) exception handler function
• Depending on the situation, the handler may then resume the execution from the saved code state, terminate the script execution or continue the script from a different location in the code

We will show different error handling methods:

• Basic use of Exceptions
• Creating a custom exception handler
• Multiple exceptions
• Re-throwing an exception
• Setting a top level exception handler

Note: Exceptions should only be used with error conditions, and should not be used to jump to another place in the code at a specified point.

Basic Use of Exceptions

When an exception is thrown, the code following it will not be executed, and PHP will try to find the matching "catch" block.

If an exception is not caught, a fatal error will be issued with an "Uncaught Exception" message.

Lets try to throw an exception without catching it:

The code above will get an error like this:

Try, throw and catch

To avoid the error from the example above, we need to create the proper code to handle an exception.

Proper exception code should include:

1. Try – A function using an exception should be in a "try" block. If the exception does not trigger, the code will continue as normal. However if the exception triggers, an exception is "thrown"
2. Throw – This is how you trigger an exception. Each "throw" must have at least one "catch"
3. Catch – A "catch" block retrieves an exception and creates an object containing the exception information

Lets try to trigger an exception with valid code:

The code above will get an error like this:

Example explained:

The code above throws an exception and catches it:

1. The checkNum() function is created. It checks if a number is greater than 1. If it is, an exception is thrown
2. The checkNum() function is called in a "try" block
3. The exception within the checkNum() function is thrown
4. The "catch" block retrives the exception and creates an object ($e) containing the exception information
5. The error message from the exception is echoed by calling $e->getMessage() from the exception object

However, one way to get around the "every throw must have a catch" rule is to set a top level exception handler to handle errors that slip through.

Creating a Custom Exception Class

Creating a custom exception handler is quite simple. We simply create a special class with functions that can be called when an exception occurs in PHP. The class must be an extension of the exception class.

The custom exception class inherits the properties from PHP’s exception class and you can add custom functions to it.

Lets create an exception class:

The new class is a copy of the old exception class with an addition of the errorMessage() function. Since it is a copy of the old class, and it inherits the properties and methods from the old class, we can use the exception class methods like getLine() and getFile() and getMessage().

Example explained:

The code above throws an exception and catches it with a custom exception class:

1. The customException() class is created as an extension of the old exception class. This way it inherits all methods and properties from the old exception class
2. The errorMessage() function is created. This function returns an error message if an e-mail address is invalid
3. The $email variable is set to a string that is not a valid e-mail address
4. The "try" block is executed and an exception is thrown since the e-mail address is invalid
5. The "catch" block catches the exception and displays the error message

Multiple Exceptions

It is possible for a script to use multiple exceptions to check for multiple conditions.

It is possible to use several if..else blocks, a switch, or nest multiple exceptions. These exceptions can use different exception classes and return different error messages:

Example explained:

The code above tests two conditions and throws an exception if any of the conditions are not met:

1. The customException() class is created as an extension of the old exception class. This way it inherits all methods and properties from the old exception class
2. The errorMessage() function is created. This function returns an error message if an e-mail address is invalid
3. The $email variable is set to a string that is a valid e-mail address, but contains the string "example"
4. The "try" block is executed and an exception is not thrown on the first condition
5. The second condition triggers an exception since the e-mail contains the string "example"
6. The "catch" block catches the exception and displays the correct error message

If there was no customException catch, only the base exception catch, the exception would be handled there

Re-throwing Exceptions

Sometimes, when an exception is thrown, you may wish to handle it differently than the standard way. It is possible to throw an exception a second time within a "catch" block.

A script should hide system errors from users. System errors may be important for the coder, but is of no interest to the user. To make things easier for the user you can re-throw the exception with a user friendly message:

Example explained:

The code above tests if the email-address contains the string "example" in it, if it does, the exception is re-thrown:

1. The customException() class is created as an extension of the old exception class. This way it inherits all methods and properties from the old exception class
2. The errorMessage() function is created. This function returns an error message if an e-mail address is invalid
3. The $email variable is set to a string that is a valid e-mail address, but contains the string "example"
4. The "try" block contains another "try" block to make it possible to re-throw the exception
5. The exception is triggered since the e-mail contains the string "example"
6. The "catch" block catches the exception and re-throws a "customException"
7. The "customException" is caught and displays an error message

If the exception is not caught in its current "try" block, it will search for a catch block on "higher levels".

Set a Top Level Exception Handler

The set_exception_handler() function sets a user-defined function to handle all uncaught exceptions.

The output of the code above should be something like this:

In the code above there was no "catch" block. Instead, the top level exception handler triggered. This function should be used to catch uncaught exceptions.

Rules for exceptions

• Code may be surrounded in a try block, to help catch potential exceptions
• Each try block or "throw" must have at least one corresponding catch block
• Multiple catch blocks can be used to catch different classes of exceptions
• Exceptions can be thrown (or re-thrown) in a catch block within a try block

A simple rule: If you throw something, you have to catch it.

PHP Error Handling

When creating scripts and web applications, error handling is an important part. If your code lacks error checking code, your program may look very unprofessional and you may be open to security risks.

This tutorial contains some of the most common error checking methods in PHP.

We will show different error handling methods:

• Simple "die()" statements
• Custom errors and error triggers
• Error reporting

Basic Error Handling: Using the die() function

The first example shows a simple script that opens a text file:

If the file does not exist you might get an error like this:

To avoid that the user gets an error message like the one above, we test if the file exist before we try to access it:

Now if the file does not exist you get an error like this:

The code above is more efficient than the earlier code, because it uses a simple error handling mechanism to stop the script after the error.

However, simply stopping the script is not always the right way to go. Let’s take a look at alternative PHP functions for handling errors.

Creating a Custom Error Handler

Creating a custom error handler is quite simple. We simply create a special function that can be called when an error occurs in PHP.

This function must be able to handle a minimum of two parameters (error level and error message) but can accept up to five parameters (optionally: file, line-number, and the error context):

Syntax

Error Report levels

These error report levels are the different types of error the user-defined error handler can be used for:

Now lets create a function to handle errors:

The code above is a simple error handling function. When it is triggered, it gets the error level and an error message. It then outputs the error level and message and terminates the script.

Now that we have created an error handling function we need to decide when it should be triggered.

Set Error Handler

The default error handler for PHP is the built in error handler. We are going to make the function above the default error handler for the duration of the script.

It is possible to change the error handler to apply for only some errors, that way the script can handle different errors in different ways. However, in this example we are going to use our custom error handler for all errors:

Since we want our custom function to handle all errors, the set_error_handler() only needed one parameter, a second parameter could be added to specify an error level.

Example

Testing the error handler by trying to output variable that does not exist:

The output of the code above should be something like this:

Trigger an Error

In a script where users can input data it is useful to trigger errors when an illegal input occurs. In PHP, this is done by the trigger_error() function.

Example

In this example an error occurs if the "test" variable is bigger than "1":

The output of the code above should be something like this:

An error can be triggered anywhere you wish in a script, and by adding a second parameter, you can specify what error level is triggered.

Possible error types:

• E_USER_ERROR – Fatal user-generated run-time error. Errors that can not be recovered from. Execution of the script is halted
• E_USER_WARNING – Non-fatal user-generated run-time warning. Execution of the script is not halted
• E_USER_NOTICE – Default. User-generated run-time notice. The script found something that might be an error, but could also happen when running a script normally

Example

In this example an E_USER_WARNING occurs if the "test" variable is bigger than "1". If an E_USER_WARNING occurs we will use our custom error handler and end the script:

The output of the code above should be something like this:

Now that we have learned to create our own errors and how to trigger them, lets take a look at error logging.

Error Logging

By default, PHP sends an error log to the servers logging system or a file, depending on how the error_log configuration is set in the php.ini file. By using the error_log() function you can send error logs to a specified file or a remote destination.

Sending errors messages to yourself by e-mail can be a good way of getting notified of specific errors.

Send an Error Message by E-Mail

In the example below we will send an e-mail with an error message and end the script, if a specific error occurs:

The output of the code above should be something like this:

And the mail received from the code above looks like this:

This should not be used with all errors. Regular errors should be logged on the server using the default PHP logging system.

PHP E-mail Injections

First, look at the PHP code from the previous chapter:

The problem with the code above is that unauthorized users can insert data into the mail headers via the input form.

What happens if the user adds the following text to the email input field in the form?

The mail() function puts the text above into the mail headers as usual, and now the header has an extra Cc:, Bcc:, and To: field. When the user clicks the submit button, the e-mail will be sent to all of the addresses above!

PHP Stopping E-mail Injections

The best way to stop e-mail injections is to validate the input.

The code below is the same as in the previous chapter, but now we have added an input validator that checks the email field in the form:

In the code above we use PHP filters to validate input:

• The FILTER_SANITIZE_EMAIL filter removes all illegal e-mail characters from a string
• The FILTER_VALIDATE_EMAIL filter validates value as an e-mail address

The PHP mail() Function

The PHP mail() function is used to send emails from inside a script.

Syntax

Note: For the mail functions to be available, PHP requires an installed and working email system. The program to be used is defined by the configuration settings in the php.ini file.

PHP Simple E-Mail

The simplest way to send an email with PHP is to send a text email.

In the example below we first declare the variables ($to, $subject, $message, $from, $headers), then we use the variables in the mail() function to send an e-mail:

PHP Mail Form

With PHP, you can create a feedback-form on your website. The example below sends a text message to a specified e-mail address:

This is how the example above works:

• First, check if the email input field is filled out
• If it is not set (like when the page is first visited); output the HTML form
• If it is set (after the form is filled out); send the email from the form
• When submit is pressed after the form is filled out, the page reloads, sees that the email input is set, and sends the email

Note: This is the simplest way to send e-mail, but it is not secure. In the next chapter of this tutorial you can read more about vulnerabilities in e-mail scripts, and how to validate user input to make it more secure.

PHP Session Variables

When you are working with an application, you open it, do some changes and then you close it. This is much like a Session. The computer knows who you are. It knows when you start the application and when you end. But on the internet there is one problem: the web server does not know who you are and what you do because the HTTP address doesn’t maintain state.

A PHP session solves this problem by allowing you to store user information on the server for later use (i.e. username, shopping items, etc). However, session information is temporary and will be deleted after the user has left the website. If you need a permanent storage you may want to store the data in a database.

Sessions work by creating a unique id (UID) for each visitor and store variables based on this UID. The UID is either stored in a cookie or is propagated in the URL.

Starting a PHP Session

Before you can store user information in your PHP session, you must first start up the session.

Note: The session_start() function must appear BEFORE the <html> tag:

The code above will register the user’s session with the server, allow you to start saving user information, and assign a UID for that user’s session.

Storing a Session Variable

The correct way to store and retrieve session variables is to use the PHP $_SESSION variable:

Output:

In the example below, we create a simple page-views counter. The isset() function checks if the "views" variable has already been set. If "views" has been set, we can increment our counter. If "views" doesn’t exist, we create a "views" variable, and set it to 1:

Destroying a Session

If you wish to delete some session data, you can use the unset() or the session_destroy() function.

The unset() function is used to free the specified session variable:

You can also completely destroy the session by calling the session_destroy() function:

Note: session_destroy() will reset your session and you will lose all your stored session data.

What is a Cookie?

A cookie is often used to identify a user. A cookie is a small file that the server embeds on the user’s computer. Each time the same computer requests a page with a browser, it will send the cookie too. With PHP, you can both create and retrieve cookie values.

How to Create a Cookie?

The setcookie() function is used to set a cookie.

Note: The setcookie() function must appear BEFORE the <html> tag.

Syntax

Example 1

In the example below, we will create a cookie named "user" and assign the value "Alex Porter" to it. We also specify that the cookie should expire after one hour:

Note: The value of the cookie is automatically URLencoded when sending the cookie, and automatically decoded when received (to prevent URLencoding, use setrawcookie() instead).

Example 2

You can also set the expiration time of the cookie in another way. It may be easier than using seconds.

In the example above the expiration time is set to a month (60 sec * 60 min * 24 hours * 30 days).

How to Retrieve a Cookie Value?

The PHP $_COOKIE variable is used to retrieve a cookie value.

In the example below, we retrieve the value of the cookie named "user" and display it on a page:

In the following example we use the isset() function to find out if a cookie has been set:

How to Delete a Cookie?

When deleting a cookie you should assure that the expiration date is in the past.

Delete example:

What if a Browser Does NOT Support Cookies?

If your application deals with browsers that do not support cookies, you will have to use other methods to pass information from one page to another in your application. One method is to pass the data through forms (forms and user input are described earlier in this tutorial).

The form below passes the user input to "welcome.php" when the user clicks on the "Submit" button:

Retrieve the values in the "welcome.php" file like this:

Create an Upload-File Form

To allow users to upload files from a form can be very useful.

Look at the following HTML form for uploading files:

Notice the following about the HTML form above:

• The enctype attribute of the <form> tag specifies which content-type to use when submitting the form. "multipart/form-data" is used when a form requires binary data, like the contents of a file, to be uploaded
• The type="file" attribute of the <input> tag specifies that the input should be processed as a file. For example, when viewed in a browser, there will be a browse-button next to the input field

Note: Allowing users to upload files is a big security risk. Only permit trusted users to perform file uploads.

Create The Upload Script

The "upload_file.php" file contains the code for uploading a file:

By using the global PHP $_FILES array you can upload files from a client computer to the remote server.

The first parameter is the form’s input name and the second index can be either "name", "type", "size", "tmp_name" or "error". Like this:

• $_FILES["file"]["name"] – the name of the uploaded file
• $_FILES["file"]["type"] – the type of the uploaded file
• $_FILES["file"]["size"] – the size in bytes of the uploaded file
• $_FILES["file"]["tmp_name"] – the name of the temporary copy of the file stored on the server
• $_FILES["file"]["error"] – the error code resulting from the file upload

This is a very simple way of uploading files. For security reasons, you should add restrictions on what the user is allowed to upload.

Restrictions on Upload

In this script we add some restrictions to the file upload. The user may only upload .gif or .jpeg files and the file size must be under 20 kb:

Note: For IE to recognize jpg files the type must be pjpeg, for FireFox it must be jpeg.

Saving the Uploaded File

The examples above create a temporary copy of the uploaded files in the PHP temp folder on the server.

The temporary copied files disappears when the script ends. To store the uploaded file we need to copy it to a different location:

The script above checks if the file already exists, if it does not, it copies the file to the specified folder.

Note: This example saves the file to a new folder called "upload"

Opening a File

The fopen() function is used to open files in PHP.

The first parameter of this function contains the name of the file to be opened and the second parameter specifies in which mode the file should be opened:

The file may be opened in one of the following modes:

Note: If the fopen() function is unable to open the specified file, it returns 0 (false).

Example

The following example generates a message if the fopen() function is unable to open the specified file:

Closing a File

The fclose() function is used to close an open file:

Check End-of-file

The feof() function checks if the "end-of-file" (EOF) has been reached.

The feof() function is useful for looping through data of unknown length.

Note: You cannot read from files opened in w, a, and x mode!

Reading a File Line by Line

The fgets() function is used to read a single line from a file.

Note: After a call to this function the file pointer has moved to the next line.

Example

The example below reads a file line by line, until the end of file is reached:

Reading a File Character by Character

The fgetc() function is used to read a single character from a file.

Note: After a call to this function the file pointer moves to the next character.

Example

The example below reads a file character by character, until the end of file is reached:

You can insert the content of one PHP file into another PHP file before the server executes it, with the include() or require() function.

The two functions are identical in every way, except how they handle errors:

• include() generates a warning, but the script will continue execution
• require() generates a fatal error, and the script will stop

These two functions are used to create functions, headers, footers, or elements that will be reused on multiple pages.

Server side includes saves a lot of work. This means that you can create a standard header, footer, or menu file for all your web pages. When the header needs to be updated, you can only update the include file, or when you add a new page to your site, you can simply change the menu file (instead of updating the links on all your web pages).

PHP include() Function

The include() function takes all the content in a specified file and includes it in the current file.

If an error occurs, the include() function generates a warning, but the script will continue execution.

Example 1

Assume that you have a standard header file, called "header.php". To include the header file in a page, use the include() function:

Example 2

Assume we have a standard menu file, called "menu.php", that should be used on all pages:

All pages in the Web site should include this menu file. Here is how it can be done:

If you look at the source code of the page above (in a browser), it will look like this:

PHP require() Function

The require() function is identical to include(), except that it handles errors differently.

If an error occurs, the include() function generates a warning, but the script will continue execution. The require() generates a fatal error, and the script will stop.

Error Example include() Function

Error message:

Notice that the echo statement is executed! This is because a Warning does not stop the script execution.

Error Example require() Function

Now, let’s run the same example with the require() function.

Error message:

The echo statement is not executed, because the script execution stopped after the fatal error.

It is recommended to use the require() function instead of include(), because scripts should not continue after an error.