The following options are supported for Zend_Filter_Alnum:

• allowwhitespace: If this option is set then whitespace characters are allowed. Otherwise they are supressed. Per default whitespaces are not allowed.

See the following example for the default behaviour of this filter.

The above example returns 'Thisismycontent123'. As you see all whitespaces and also the brackets are filtered.

Note: Zend_Filter_Alnum works on almost all languages. But actually there are three exceptions: Chinese, Japanese and Korean. Within these languages the english alphabet is use instead of the characters from these languages. The language itself is detected by using Zend_Locale.

Zend_Filter_Alnum can also allow whitespaces. This can be usefull when you want to strip special chars from a text. See the following example:

The above example returns 'This is my content 123'. As you see only the brackets are filtered whereas the whitespaces are not touched.

To change allowWhiteSpace afterwards you can use setAllowWhiteSpace() and getAllowWhiteSpace().

The following options are supported for Zend_Filter_Alpha:

• allowwhitespace: If this option is set then whitespace characters are allowed. Otherwise they are suppressed. By default whitespace characters are not allowed.

A basic example of usage is below:

The above example returns 'Thisismycontent'. Notice that the whitespace characters and brackets are removed.

Note: Zend_Filter_Alpha works on most languages; however, there are three exceptions: Chinese, Japanese and Korean. With these languages the english alphabet is used. The language is detected through the use of Zend_Locale.

Zend_Filter_Alpha can also allow whitespace characters. This can be useful when you want to strip special characters from a string. See the following example:

The above example returns 'This is my content '. Notice that the parenthesis, colon, and numbers have all been removed while the whitespace characters remain.

To change allowWhiteSpace after instantiation the method setAllowWhiteSpace() may be used.

To query the current value of allowWhiteSpace the method getAllowWhiteSpace() may be used.

There are no additional options for Zend_Filter_BaseName.

A basic example of usage is below:

This will return 'filename'.

This will return 'filename.txt'.

The following options are supported for Zend_Filter_Boolean:

• casting: When this option is set to TRUE then any given input will be casted to boolean. This option defaults to TRUE.

• locale: This option sets the locale which will be used to detect localized input.

• type: The type option sets the boolean type which should be used. Read the following for details.

By default, this filter works by casting the input to a BOOLEAN value; in other words, it operates in a similar fashion to calling (boolean) $value.

This means that without providing any configuration, Zend_Filter_Boolean accepts all input types and returns a BOOLEAN just as you would get by type casting to BOOLEAN.

Sometimes casting with (boolean) will not suffice. Zend_Filter_Boolean allows you to configure specific types to convert, as well as which to omit.

The following types can be handled:

• boolean: Returns a boolean value as is.

• integer: Converts an integer 0 value to FALSE.

• float: Converts a float 0.0 value to FALSE.

• string: Converts an empty string '' to FALSE.

• zero: Converts a string containing the single character zero ('0') to FALSE.

• empty_array: Converts an empty array to FALSE.

• null: Converts a NULL value to FALSE.

• php: Converts values according to PHP when casting them to BOOLEAN.

• false_string: Converts a string containing the word "false" to a boolean FALSE.

• yes: Converts a localized string which contains the word "no" to FALSE.

• all: Converts all above types to BOOLEAN.

All other given values will return TRUE by default.

There are several ways to select which of the above types are filtered. You can give one or multiple types and add them, you can give an array, you can use constants, or you can give a textual string. See the following examples:

You can also give an instance of Zend_Config to set the desired types. To set types after instantiation, use the setType() method.

As mentioned previously, Zend_Filter_Boolean can also recognise localized "yes" and "no" strings. This means that you can ask your customer in a form for "yes" or "no" within his native language and Zend_Filter_Boolean will convert the response to the appropriate boolean value.

To set the desired locale, you can either use the locale option, or the method setLocale().

Sometimes it is necessary to recognise only TRUE or FALSE and return all other values without changes. Zend_Filter_Boolean allows you to do this by setting the casting option to FALSE.

In this case Zend_Filter_Boolean will work as described in the following table, which shows which values return TRUE or FALSE. All other given values are returned without change when casting is set to FALSE

Usage without casting

Type	True	False
Zend_Filter_Boolean::BOOLEAN	TRUE	FALSE
Zend_Filter_Boolean::INTEGER	0	1
Zend_Filter_Boolean::FLOAT	0.0	1.0
Zend_Filter_Boolean::STRING	""
Zend_Filter_Boolean::ZERO	"0"	"1"
Zend_Filter_Boolean::EMPTY_ARRAY	array()
Zend_Filter_Boolean::NULL	NULL
Zend_Filter_Boolean::FALSE_STRING	"false" (case independently)	"true" (case independently)
Zend_Filter_Boolean::YES	localized "yes" (case independently)	localized "no" (case independently)

The following example shows the behaviour when changing the casting option:

The following options are supported for Zend_Filter_Callback:

• callback: This sets the callback which should be used.

• options: This property sets the options which are used when the callback is processed.

The usage of this filter is quite simple. Let's expect we want to create a filter which reverses a string.

As you can see it's really simple to use a callback to define a own filter. It is also possible to use a method, which is defined within a class, by giving an array as callback.

To get the actual set callback use getCallback() and to set another callback use setCallback().

Note: Possible exceptions
You should note that defining a callback method which can not be called will raise an exception.

It is also possible to define default parameters, which are given to the called method as array when the filter is executed. This array will be concatenated with the value which will be filtered.

When you would call the above method definition manually it would look like this:

The following options are supported for Zend_Filter_Compress and Zend_Filter_Decompress:

• adapter: The compression adapter which should be used. It defaults to Gz.

• options: Additional options which are given to the adapter at initiation. Each adapter supports it's own options.

The following compression formats are supported by their own adapter:

• Bz2

• Gz

• Lzf

• Rar

• Tar

• Zip

Each compression format has different capabilities as described below. All compression filters may be used in approximately the same ways, and differ primarily in the options available and the type of compression they offer (both algorithmically as well as string vs. file vs. directory)

To create a compression filter you need to select the compression format you want to use. The following description takes the Bz2 adapter. Details for all other adapters are described after this section.

The two filters are basically identical, in that they utilize the same backends. Zend_Filter_Compress should be used when you wish to compress items, and Zend_Filter_Decompress should be used when you wish to decompress items.

For instance, if we want to compress a string, we have to initiate Zend_Filter_Compress and indicate the desired adapter.

To use a different adapter, you simply specify it to the constructor.

You may also provide an array of options or Zend_Config object. If you do, provide minimally the key "adapter", and then either the key "options" or "adapterOptions" (which should be an array of options to provide to the adapter on instantiation).

Note: Default compression Adapter
When no compression adapter is given, then the Gz adapter will be used.

Almost the same usage is we want to decompress a string. We just have to use the decompression filter in this case.

To get the compressed string, we have to give the original string. The filtered value is the compressed version of the original string.

Decompression works the same way.

Note: Note on string compression
Not all adapters support string compression. Compression formats like Rar can only handle files and directories. For details, consult the section for the adapter you wish to use.

Creating an archive file works almost the same as compressing a string. However, in this case we need an additional parameter which holds the name of the archive we want to create.

In the above example the uncompressed string is compressed, and is then written into the given archive file.

Note: Existing archives will be overwritten
The content of any existing file will be overwritten when the given filename of the archive already exists.

When you want to compress a file, then you must give the name of the file with its path.

You may also specify a directory instead of a filename. In this case the whole directory with all its files and subdirectories will be compressed into the archive.

Note: Do not compress large or base directories
You should never compress large or base directories like a complete partition. Compressing a complete partition is a very time consuming task which can lead to massive problems on your server when there is not enough space or your script takes too much time.

Decompressing an archive file works almost like compressing it. You must specify either the archive parameter, or give the filename of the archive when you decompress the file.

Some adapters support decompressing the archive into another subdirectory. In this case you can set the target parameter.

Note: Directories to extract to must exist
When you want to decompress an archive into a directory, then that directory must exist.

The Bz2 Adapter can compress and decompress:

• Strings

• Files

• Directories

This adapter makes use of PHP's Bz2 extension.

To customize compression, this adapter supports the following options:

• Archive: This parameter sets the archive file which should be used or created.

• Blocksize: This parameter sets the blocksize to use. It can be from '0' to '9'. The default value is '4'.

All options can be set at instantiation or by using a related method. For example, the related methods for 'Blocksize' are getBlocksize() and setBlocksize(). You can also use the setOptions() method which accepts all options as array.

The Gz Adapter can compress and decompress:

• Strings

• Files

• Directories

This adapter makes use of PHP's Zlib extension.

To customize the compression this adapter supports the following options:

• Archive: This parameter sets the archive file which should be used or created.

• Level: This compression level to use. It can be from '0' to '9'. The default value is '9'.

• Mode: There are two supported modes. 'compress' and 'deflate'. The default value is 'compress'.

All options can be set at initiation or by using a related method. For example, the related methods for 'Level' are getLevel() and setLevel(). You can also use the setOptions() method which accepts all options as array.

The Lzf Adapter can compress and decompress:

• Strings

Note: Lzf supports only strings
The Lzf adapter can not handle files and directories.

This adapter makes use of PHP's Lzf extension.

There are no options available to customize this adapter.

The Rar Adapter can compress and decompress:

• Files

• Directories

Note: Rar does not support strings
The Rar Adapter can not handle strings.

This adapter makes use of PHP's Rar extension.

Note: Rar compression not supported
Due to restrictions with the Rar compression format, there is no compression available for free. When you want to compress files into a new Rar archive, you must provide a callback to the adapter that can invoke a Rar compression program.

To customize the compression this adapter supports the following options:

• Archive: This parameter sets the archive file which should be used or created.

• Callback: A callback which provides compression support to this adapter.

• Password: The password which has to be used for decompression.

• Target: The target where the decompressed files will be written to.

All options can be set at instantiation or by using a related method. For example, the related methods for 'Target' are getTarget() and setTarget(). You can also use the setOptions() method which accepts all options as array.

The Tar Adapter can compress and decompress:

• Files

• Directories

Note: Tar does not support strings
The Tar Adapter can not handle strings.

This adapter makes use of PEAR's Archive_Tar component.

To customize the compression this adapter supports the following options:

• Archive: This parameter sets the archive file which should be used or created.

• Mode: A mode to use for compression. Supported are either 'NULL' which means no compression at all, 'Gz' which makes use of PHP's Zlib extension and 'Bz2' which makes use of PHP's Bz2 extension. The default value is 'NULL'.

• Target: The target where the decompressed files will be written to.

All options can be set at instantiation or by using a related method. For example, the related methods for 'Target' are getTarget() and setTarget(). You can also use the setOptions() method which accepts all options as array.

Note: Directory usage
When compressing directories with Tar then the complete file path is used. This means that created Tar files will not only have the subdirectory but the complete path for the compressed file.

The Zip Adapter can compress and decompress:

• Strings

• Files

• Directories

Note: Zip does not support string decompression
The Zip Adapter can not handle decompression to a string; decompression will always be written to a file.

This adapter makes use of PHP's Zip extension.

To customize the compression this adapter supports the following options:

• Archive: This parameter sets the archive file which should be used or created.

• Target: The target where the decompressed files will be written to.

All options can be set at instantiation or by using a related method. For example, the related methods for 'Target' are getTarget() and setTarget(). You can also use the setOptions() method which accepts all options as array.

There are no additional options for Zend_Filter_Digits.

A basic example of usage is below:

This returns "2009".

This returns "5".

There are no additional options for Zend_Filter_Dir.

A basic example of usage is below:

This returns "/etc".

This returns "C:/Temp".

The following options are supported for Zend_Filter_Encrypt and Zend_Filter_Decrypt:

• adapter: This sets the encryption adapter which should be used

• algorithm: Only MCrypt. The algorithm which has to be used. It should be one of the algorithm ciphers which can be found under » PHP's mcrypt ciphers. If not set it defaults to blowfish.

• algorithm_directory: Only MCrypt. The directory where the algorithm can be found. If not set it defaults to the path set within the mcrypt extension.

• compression: If the encrypted value should be compressed. Default is no compression.

• envelope: Only OpenSSL. The encrypted envelope key from the user who encrypted the content. You can either provide the path and filename of the key file, or just the content of the key file itself. When the package option has been set, then you can omit this parameter.

• key: Only MCrypt. The encryption key with which the input will be encrypted. You need the same key for decryption.

• mode: Only MCrypt. The encryption mode which has to be used. It should be one of the modes which can be found under » PHP's mcrypt modes. If not set it defaults to 'cbc'.

• mode_directory: Only MCrypt. The directory where the mode can be found. If not set it defaults to the path set within the Mcrypt extension.

• package: Only OpenSSL. If the envelope key should be packed with the encrypted value. Default is FALSE.

• private: Only OpenSSL. Your private key which will be used for encrypting the content. Also the private key can be either a filename with path of the key file, or just the content of the key file itself.

• public: Only OpenSSL. The public key of the user whom you want to provide the encrpted content. You can give multiple public keys by using an array. You can eigther provide the path and filename of the key file, or just the content of the key file itself.

• salt: Only MCrypt. If the key should be used as salt value. The key used for encryption will then itself also be encrypted. Default is FALSE.

• vector: Only MCrypt. The initialization vector which shall be used. If not set it will be a random vector.

As these two encryption methodologies work completely different, also the usage of the adapters differ. You have to select the adapter you want to use when initiating the filter.

To set another adapter you can also use setAdapter(), and the getAdapter() method to receive the actual set adapter.

Note: When you do not supply the adapter option or do not use setAdapter(), then the Mcrypt adapter will be used per default.

When you have installed the Mcrypt extension you can use the Mcrypt adapter. If you provide a string instead of an array of options, this string will be used as key.

You can get and set the encryption values also afterwards with the getEncryption() and setEncryption() methods.

Note: Note that you will get an exception if the mcrypt extension is not available in your environment.

Note: You should also note that all settings which be checked when you create the instance or when you call setEncryption(). If mcrypt detects problem with your settings an exception will be thrown.

You can get or set the encryption vector by calling getVector() and setVector(). A given string will be truncated or padded to the needed vector size of the used algorithm.

Note: Note that when you are not using an own vector, you must get the vector and store it. Otherwise you will not be able to decode the encoded string.

For decrypting content which was previously encrypted with Mcrypt you need to have the options with which the encryption has been called.

There is one eminent difference for you. When you did not provide a vector at encryption you need to get it after you encrypted the content by using the getVector() method on the encryption filter. Without the correct vector you will not be able to decrypt the content.

As soon as you have provided all options decryption is as simple as encryption.

Note: Note that you will get an exception if the mcrypt extension is not available in your environment.

Note: You should also note that all settings which be checked when you create the instance or when you call setEncryption(). If mcrypt detects problem with your settings an exception will be thrown.

When you have installed the OpenSSL extension you can use the OpenSSL adapter. You can get or set the public keys also afterwards with the getPublicKey() and setPublicKey() methods. The private key can also be get and set with the related getPrivateKey() and setPrivateKey() methods.

Note: Note that the OpenSSL adapter will not work when you do not provide valid keys.

When you want to encode also the keys, then you have to provide a passphrase with the setPassphrase() method. When you want to decode content which was encoded with a passphrase you will not only need the public key, but also the passphrase to decode the encrypted key.

At last, when you use OpenSSL you need to give the receiver the encrypted content, the passphrase when have provided one, and the envelope keys for decryption.

This means for you, that you have to get the envelope keys after the encryption with the getEnvelopeKey() method.

So our complete example for encrypting content with OpenSSL look like this.

Decryption with OpenSSL is as simple as encryption. But you need to have all data from the person who encrypted the content. See the following example:

Note: Note that the OpenSSL adapter will not work when you do not provide valid keys.

Optionally it could be necessary to provide the passphrase for decrypting the keys themself by using the setPassphrase() method.

At last, decode the content. Our complete example for decrypting the previously encrypted content looks like this.

The following options are supported for Zend_Filter_HtmlEntities:

• quotestyle: Equivalent to the PHP htmlentities native function parameter quote_style. This allows you to define what will be done with 'single' and "double" quotes. The following constants are accepted: ENT_COMPAT, ENT_QUOTES ENT_NOQUOTES with the default being ENT_COMPAT.

• charset: Equivalent to the PHP htmlentities native function parameter charset. This defines the character set to be used in filtering. Unlike the PHP native function the default is 'UTF-8'. See "http://php.net/htmlentities" for a list of supported character sets.

  Note: This option can also be set via the $options parameter as a Zend_Config object or array. The option key will be accepted as either charset or encoding.
  

• doublequote: Equivalent to the PHP htmlentities native function parameter double_encode. If set to false existing html entities will not be encoded. The default is to convert everything (true).

  Note: This option must be set via the $options parameter or the setDoubleEncode() method.
  

See the following example for the default behaviour of this filter.

Zend_Filter_HtmlEntities allows changing the quote style used. This can be useful when you want to leave double, single, or both types of quotes un-filtered. See the following example:

The above example returns A 'single' and "double". Notice that 'single' as well as "double" quotes are filtered.

The above example returns A 'single' and "double". Notice that "double" quotes are filtered while 'single' quotes are not altered.

The above example returns A 'single' and "double". Notice that neither "double" or 'single' quotes are altered.

To change or retrieve the quotestyle after instantiation, the two methods setQuoteStyle() and getQuoteStyle() may be used respectively. setQuoteStyle() accepts one parameter $quoteStyle. The following constants are accepted: ENT_COMPAT, ENT_QUOTES, ENT_NOQUOTES

To change or retrieve the charset after instantiation, the two methods setCharSet() and getCharSet() may be used respectively. setCharSet() accepts one parameter $charSet. See "http://php.net/htmlentities" for a list of supported character sets.

To change or retrieve the doublequote option after instantiation, the two methods setDoubleQuote() and getDoubleQuote() may be used respectively. setDoubleQuote() accepts one boolean parameter $doubleQuote.

There are no additional options for Zend_Filter_Int.

A basic example of usage is below:

This will return '-4'.

The following options are supported for Zend_Filter_LocalizedToNormalized and Zend_Filter_NormalizedToLocalized:

• date_format: This sets the date format to use for normalization and to detect the localized date format

• locale: This sets the locale to use for normalization and to detect the localized format

• precision: This sets the precision to use for number conversion

Normalization allows your user to enter informations in his own language notation, and you can then store the normalized value into your database for example. Localization on the other hand allows you to display normalized informations in a localized manner to your user.

Note: Please note that normalization and localization is not equal to translation. These filters can not translate strings from one language into another like you could expect with months or names of days.

The following input types can be normalized and localized:

• integer: Integer numbers. Normalization returns the english notation

• float: Float numbers. Normalization returns the english notation

• numbers: Other numbers, like real. Normalization returns the english notation

• time: Time values. The normalized value is a named array

• date: Date values. The normalized value is a named array

Any other input will be returned as is, without changing it.

Note: You should note that normalized output is always given as string. Otherwise your environment would transfer the normalized output automatically to the notation used by the locale your environment is set to.

Any given number like integer, float or real value, can be normalized. Note, that numbers in scientific notation, can actually not be handled by this filter.

So how does this normalization work in detail for numbers:

Let's expect you have set the locale 'de' as application wide locale. Zend_Filter_LocalizedToNormalized will take the set locale and use it to detect which sort of input you gave. In our example it was a value with precision. Now the filter will return you the normalized representation for this value as string.

You can also control how your normalized number has to look like. Therefor you can give all options which are also used by Zend_Locale_Format. For details about available options take a look into this Zend_Locale chapter.

Below is a example with defined precision so you can see how options work:

Input for date and time values can also be normalized. All given date and time values will be returned as array, where each date part is given within a own key.

Let's expect you have set the locale 'de' again. Now the input is automatically detected as date, and you will get a named array in return.

Of course you can also control how your date input looks like with the date_format and the locale option.

Any given number like integer, float or real value, can be localized. Note, that numbers in scientific notation, can actually not be handled by this filter.

So how does localization work in detail for numbers:

Let's expect you have set the locale 'de' as application wide locale. Zend_Filter_NormalizedToLocalized will take the set locale and use it to detect which sort of output you want to have. In our example it was a value with precision. Now the filter will return you the localized representation for this value as string.

You can also control how your localized number has to look like. Therefor you can give all options which are also used by Zend_Locale_Format. For details about how these options are used take a look into this Zend_Locale chapter.

Below is a example with defined precision so you can see how options work:

Normalized for date and time values can also be localized. All given date and time values will be returned as string, with the format defined by the set locale.

Let's expect you have set the locale 'de' again. Now the input is automatically detected as date, and will be returned in the format defined by the locale 'de'.

Of course you can also control how your date input looks like with the date_format, and the locale option.

The following options are supported for Zend_Filter_Null:

• type: The variable type which should be supported.

Per default this filter works like PHP's empty() method; in other words, if empty() returns a boolean TRUE, then a NULL value will be returned.

This means that without providing any configuration, Zend_Filter_Null will accept all input types and return NULL in the same cases as empty().

Any other value will be returned as is, without any changes.

Sometimes it's not enough to filter based on empty(). Therefor Zend_Filter_Null allows you to configure which type will be converted and which not.

The following types can be handled:

• boolean: Converts a boolean FALSE value to NULL.

• integer: Converts an integer 0 value to NULL.

• empty_array: Converts an empty array to NULL.

• string: Converts an empty string '' to NULL.

• zero: Converts a string containing the single character zero ('0') to NULL.

• all: Converts all above types to NULL. (This is the default behavior.)

There are several ways to select which of the above types are filtered. You can give one or multiple types and add them, you can give an array, you can use constants, or you can give a textual string. See the following examples:

You can also give an instance of Zend_Config to set the wished types. To set types afterwards use setType().

The following options are supported for Zend_Filter_PregReplace:

• match: The pattern which will be searched for.

• replace: The string which is used as replacement for the matches.

To use this filter properly you must give two options:

The option match has to be given to set the pattern which will be searched for. It can be a string for a single pattern, or an array of strings for multiple pattern.

To set the pattern which will be used as replacement the option replace has to be used. It can be a string for a single pattern, or an array of strings for multiple pattern.

You can use getMatchPattern() and setMatchPattern() to set the matching pattern afterwards. To set the replacement pattern you can use getReplacement() and setReplacement().

For a more complex usage take a look into PHP's » PCRE Pattern Chapter.

The following options are supported for Zend_Filter_RealPath:

• exists: This option defaults to TRUE which checks if the given path really exists.

For any given link of pathname its absolute path will be returned. References to '/./', '/../' and extra '/' characters in the input path will be stripped. The resulting path will not have any symbolic link, '/./' or '/../' character.

Zend_Filter_RealPath will return FALSE on failure, e.g. if the file does not exist. On BSD systems Zend_Filter_RealPath doesn't fail if only the last path component doesn't exist, while other systems will return FALSE.

Sometimes it is useful to get also paths when they don't exist, f.e. when you want to get the real path for a path which you want to create. You can then either give a FALSE at initiation, or use setExists() to set it.

The following options are supported for Zend_Filter_StringToLower:

• encoding: This option can be used to set an encoding which has to be used.

This is a basic example:

Per default it will only handle characters from the actual locale of your server. Characters from other charsets would be ignored. Still, it's possible to also lowercase them when the mbstring extension is available in your environment. Simply set the wished encoding when initiating the StringToLower filter. Or use the setEncoding() method to change the encoding afterwards.

Note: Setting wrong encodings
Be aware that you will get an exception when you want to set an encoding and the mbstring extension is not available in your environment.
Also when you are trying to set an encoding which is not supported by your mbstring extension you will get an exception.

The following options are supported for Zend_Filter_StringToUpper:

• encoding: This option can be used to set an encoding which has to be used.

This is a basic example for using the StringToUpper filter:

Like the StringToLower filter, this filter handles only characters from the actual locale of your server. Using different character sets works the same as with StringToLower.

The following options are supported for Zend_Filter_StringTrim:

• charlist: List of characters to remove from the beginning and end of the string. If this is not set or is null, the default behavior will be invoked, which is to remove only whitespace from the beginning and end of the string.

A basic example of usage is below:

The above example returns 'This is (my) content:'. Notice that the whitespace characters have been removed.

The above example returns 'This is (my) content'. Notice that the whitespace characters and colon are removed. You can also provide an instance of Zend_Config or an array with a 'charlist' key. To set the desired character list after instantiation, use the setCharList() method. The getCharList() return the values set for charlist.

There are no additional options for Zend_Filter_StripNewLines.

A basic example of usage is below:

The above example returns 'This is (my) content:'. Notice that all newline characters have been removed.

The following options are supported for Zend_Filter_StripTags:

• allowAttribs: This option sets the attributes which are accepted. All other attributes are stripped from the given content

• allowTags: This option sets the tags which are accepted. All other tags will be stripped from the given content

See the following example for the default behaviour of this filter:

As result you will get the stripped content 'My content'.

When the content contains broken or partitial tags then the complete following content will be erased. See the following example:

The above will return 'This contains' with the rest being stripped.

Zend_Filter_StripTags allows stripping of all but defined tags. This can be used for example to strip all tags but links from a text.

The above example strips all tags but the link. By providing an array you can set multiple tags at once.

It is also possible to strip all but allowed attributes from a tag.

The above example strips all tags but img. Additionally from the img tag all attributes but src will be stripped. By providing an array you can set multiple attributes at once.