Same name and namespace in other branches
  1. 4.7.x developer/examples/filter_example.module \filter_example_filter()
  2. 5.x developer/examples/filter_example.module \filter_example_filter()

Implementation of hook_filter().

The bulk of filtering work is done here. This hook is quite complicated, so we'll discuss each operation it defines.

File

developer/examples/filter_example.module, line 63
This is an example outlining how a module can be used to define a filter to be run on user-submitted content before it is output to the browser.

Code

function filter_example_filter($op, $delta = 0, $format = -1, $text = '') {

  // The "list" operation provides the module an opportunity to declare both how
  // many filters it defines and a human-readable name for each filter. Note that
  // the returned name should be passed through t() for translation.
  if ($op == 'list') {
    return array(
      0 => t('Substitute "foo"'),
      1 => t('Current time'),
    );
  }

  // All operations besides "list" provide a $delta argument so we know which
  // filter they refer to. We'll switch on that argument now so that we can
  // discuss each filter in turn.
  switch ($delta) {

    // First we define the simple string substitution filter.
    case 0:
      switch ($op) {

        // This description is shown in the administrative interface, unlike the
        // filter tips which are shown in the content editing interface.
        case 'description':
          return t('Substitutes a custom string for the string "foo" in the text.');

        // We don't need the "prepare" operation for this filter, but it's required
        // to at least return the input text as-is.
        case 'prepare':
          return $text;

        // The actual filtering is performed here. The supplied text should be
        // returned, once any necessary substitutions have taken place.
        case 'process':
          return str_replace('foo', variable_get('filter_example_foo_' . $format, 'bar'), $text);

        // Since we allow the administrator to define the string that gets
        // substituted when "foo" is encountered, we need to provide an interface
        // for this customization. Note that the value of $format needs to be
        // provided as part of the form name, so that different customization can
        // be done for this filter in each of the different input formats that may
        // use it.
        case 'settings':
          return form_textfield(t('Substitution string'), 'filter_example_foo_' . $format, variable_get('filter_example_foo_' . $format, 'bar'), 30, 255, t('The string to substitute for "foo" everywhere in the text.'));
      }
      break;

    // Next is our "time tag" filter.
    case 1:
      switch ($op) {

        // This description is shown in the administrative interface, unlike the
        // filter tips which are shown in the content editing interface.
        case 'description':
          return t('Inserts the current time in the place of <time /> tags.');

        // Since this filter will return a different result on each page load, we
        // need to return TRUE for "no cache" to ensure that the filter is run
        // every time the text is requested.
        case 'no cache':
          return TRUE;

        // This filter is a little trickier to implement than the previous one.
        // Since the input involves special HTML characters (< and >) we have to
        // run the filter before HTML is escaped/stripped by other filters. But
        // we want to use HTML in our result as well, and so if we run this filter
        // first our replacement string could be escaped or stripped. The solution
        // is to use the "prepare" operation to escape the special characters, and
        // to later replace our escaped version in the "process" step.
        //
        // We'll use the bytes 0xFE and 0xFF to replace < and > here. These bytes
        // are not valid in UTF-8 data and thus unlikely to cause problems.
        case 'prepare':
          return preg_replace('!<time ?/>!', '\\xFEtime /\\xFF', $text);

        // Now, in the "process" step, we'll search for our escaped time tags and
        // to the real filtering.
        case 'process':
          return str_replace('\\xFEtime /\\xFF', '<em>' . format_date(time()) . '</em>', $text);
      }
      break;
  }
}