ExpiresFilter is a Java Servlet API port of Apache mod_expires to add 'Expires' and 'Cache-Control: max-age=' headers to HTTP response according to its 'Content-Type'.

Following documentation is inspired by mod_expires

Summary

This filter controls the setting of the Expires HTTP header and the max-age directive of the Cache-Control HTTP header in server responses. The expiration date can set to be relative to either the time the source file was last modified, or to the time of the client access.

These HTTP headers are an instruction to the client about the document's validity and persistence. If cached, the document may be fetched from the cache rather than from the source until this time has passed. After that, the cache copy is considered "expired" and invalid, and a new copy must be obtained from the source.

To modify Cache-Control directives other than max-age (see RFC 2616 section 14.9), you can use other servlet filters or Apache Httpd mod_headers module.

Filter Configuration

Basic configuration to add 'Expires' and 'Cache-Control: max-age=' headers to images, css and javascript

 <web-app ...>
    ...
    <filter>
       <filter-name>ExpiresFilter</filter-name>
       <filter-class>org.apache.catalina.filters.ExpiresFilter</filter-class>
       <init-param>
          <param-name>ExpiresByType image</param-name>
          <param-value>access plus 10 minutes</param-value>
       </init-param>
       <init-param>
          <param-name>ExpiresByType text/css</param-name>
          <param-value>access plus 10 minutes</param-value>
       </init-param>
       <init-param>
          <param-name>ExpiresByType application/javascript</param-name>
          <param-value>access plus 10 minutes</param-value>
       </init-param>
    </filter>
    ...
    <filter-mapping>
       <filter-name>ExpiresFilter</filter-name>
       <url-pattern>/*</url-pattern>
       <dispatcher>REQUEST</dispatcher>
    </filter-mapping>
    ...
 </web-app>
 
 

Configuration Parameters

ExpiresByType <content-type>

This directive defines the value of the Expires header and the max-age directive of the Cache-Control header generated for documents of the specified type (e.g., text/html). The second argument sets the number of seconds that will be added to a base time to construct the expiration date. The Cache-Control: max-age is calculated by subtracting the request time from the expiration date and expressing the result in seconds.

The base time is either the last modification time of the file, or the time of the client's access to the document. Which should be used is specified by the <code> field; M means that the file's last modification time should be used as the base time, and A means the client's access time should be used. The duration is expressed in seconds. A2592000 stands for access plus 30 days in alternate syntax.

The difference in effect is subtle. If M (modification in alternate syntax) is used, all current copies of the document in all caches will expire at the same time, which can be good for something like a weekly notice that's always found at the same URL. If A ( access or now in alternate syntax) is used, the date of expiration is different for each client; this can be good for image files that don't change very often, particularly for a set of related documents that all refer to the same images (i.e., the images will be accessed repeatedly within a relatively short timespan).

Example:

 <init-param>
    <param-name>ExpiresByType text/html</param-name>
    <param-value>access plus 1 month 15 days 2 hours</param-value>
 </init-param>

 <init-param>
    <!-- 2592000 seconds = 30 days -->
    <param-name>ExpiresByType image/gif</param-name>
    <param-value>A2592000</param-value>
 </init-param>
 
 

Note that this directive only has effect if ExpiresActive On has been specified. It overrides, for the specified MIME type only, any expiration date set by the ExpiresDefault directive.

You can also specify the expiration time calculation using an alternate syntax, described earlier in this document.

ExpiresExcludedResponseStatusCodes

This directive defines the http response status codes for which the ExpiresFilter will not generate expiration headers. By default, the 304 status code ("Not modified") is skipped. The value is a comma separated list of http status codes.

This directive is useful to ease usage of ExpiresDefault directive. Indeed, the behavior of 304 Not modified (which does specify a Content-Type header) combined with Expires and Cache-Control:max-age= headers can be unnecessarily tricky to understand.

Configuration sample :

 <init-param>
    <param-name>ExpiresExcludedResponseStatusCodes</param-name>
    <param-value>302, 500, 503</param-value>
 </init-param>
 
 

ExpiresDefault

This directive sets the default algorithm for calculating the expiration time for all documents in the affected realm. It can be overridden on a type-by-type basis by the ExpiresByType directive. See the description of that directive for details about the syntax of the argument, and the "alternate syntax" description as well.

Alternate Syntax

The ExpiresDefault and ExpiresByType directives can also be defined in a more readable syntax of the form:

 <init-param>
    <param-name>ExpiresDefault</param-name>
    <param-value><base> [plus] (<num> <type>)*</param-value>
 </init-param>

 <init-param>
    <param-name>ExpiresByType type/encoding</param-name>
    <param-value><base> [plus] (<num> <type>)*</param-value>
 </init-param>
 
 

where <base> is one of:

• access
• now (equivalent to 'access')
• modification

The plus keyword is optional. <num> should be an integer value (acceptable to Integer.parseInt()), and <type> is one of:

• years
• months
• weeks
• days
• hours
• minutes
• seconds

For example, any of the following directives can be used to make documents expire 1 month after being accessed, by default:

 <init-param>
    <param-name>ExpiresDefault</param-name>
    <param-value>access plus 1 month</param-value>
 </init-param>

 <init-param>
    <param-name>ExpiresDefault</param-name>
    <param-value>access plus 4 weeks</param-value>
 </init-param>

 <init-param>
    <param-name>ExpiresDefault</param-name>
    <param-value>access plus 30 days</param-value>
 </init-param>
 
 

The expiry time can be fine-tuned by adding several ' <num> <type>' clauses:

 <init-param>
    <param-name>ExpiresByType text/html</param-name>
    <param-value>access plus 1 month 15 days 2 hours</param-value>
 </init-param>

 <init-param>
    <param-name>ExpiresByType image/gif</param-name>
    <param-value>modification plus 5 hours 3 minutes</param-value>
 </init-param>
 
 

Note that if you use a modification date based setting, the Expires header will not be added to content that does not come from a file on disk. This is due to the fact that there is no modification time for such content.

Expiration headers generation eligibility

A response is eligible to be enriched by ExpiresFilter if :

1. no expiration header is defined (Expires header or the max-age directive of the Cache-Control header),
2. the response status code is not excluded by the directive ExpiresExcludedResponseStatusCodes,
3. the Content-Type of the response matches one of the types defined the in ExpiresByType directives or the ExpiresDefault directive is defined.

Note :

• If Cache-Control header contains other directives than max-age, they are concatenated with the max-age directive that is added by the ExpiresFilter.

Expiration configuration selection

The expiration configuration if elected according to the following algorithm:

1. ExpiresByType matching the exact content-type returned by HttpServletResponse.getContentType() possibly including the charset (e.g. 'text/xml;charset=UTF-8'),
2. ExpiresByType matching the content-type without the charset if HttpServletResponse.getContentType() contains a charset (e.g. ' text/xml;charset=UTF-8' -> 'text/xml'),
3. ExpiresByType matching the major type (e.g. substring before '/') of HttpServletResponse.getContentType() (e.g. 'text/xml;charset=UTF-8' -> 'text '),
4. ExpiresDefault

Implementation Details

When to write the expiration headers ?

The ExpiresFilter traps the 'on before write response body' event to decide whether it should generate expiration headers or not.

To trap the 'before write response body' event, the ExpiresFilter wraps the http servlet response's writer and outputStream to intercept calls to the methods write(), print(), close() and flush(). For empty response body (e.g. empty files), the write(), print(), close() and flush() methods are not called; to handle this case, the ExpiresFilter, at the end of its doFilter() method, manually triggers the onBeforeWriteResponseBody() method.

Configuration syntax

The ExpiresFilter supports the same configuration syntax as Apache Httpd mod_expires.

A challenge has been to choose the name of the <param-name> associated with ExpiresByType in the <filter> declaration. Indeed, Several ExpiresByType directives can be declared when web.xml syntax does not allow to declare several <init-param> with the same name.

The workaround has been to declare the content type in the <param-name> rather than in the <param-value>.

Designed for extension : the open/close principle

The ExpiresFilter has been designed for extension following the open/close principle.

Key methods to override for extension are :

• isEligibleToExpirationHeaderGeneration(HttpServletRequest, XHttpServletResponse)
• getExpirationDate(XHttpServletResponse)

Troubleshooting

To troubleshoot, enable logging on the org.apache.catalina.filters.ExpiresFilter.

Extract of logging.properties

Sample of initialization log message :

Sample of per-request log message where ExpiresFilter adds an expiration date

Sample of per-request log message where ExpiresFilter does not add an expiration date