/**
* Accessor for starting value.
* @return string Parsed value.
* @access public
*/
function getDefault()
{
return $this->_wrap(SimpleHtmlSaxParser::decodeHtml($this->getContent()));
}
/**
* Turns HTML into text browser visible text. Images
* are converted to their alt text and tags are supressed.
* Entities are converted to their visible representation.
* @param string $html HTML to convert.
* @return string Plain text.
* @access public
* @static
*/
function normalise($html)
{
$text = preg_replace('|<!--.*?-->|', '', $html);
$text = preg_replace('|<img.*?alt\\s*=\\s*"(.*?)".*?>|', ' \\1 ', $text);
$text = preg_replace('|<img.*?alt\\s*=\\s*\'(.*?)\'.*?>|', ' \\1 ', $text);
$text = preg_replace('|<img.*?alt\\s*=\\s*([a-zA-Z_]+).*?>|', ' \\1 ', $text);
$text = preg_replace('|<.*?>|', '', $text);
$text = SimpleHtmlSaxParser::decodeHtml($text);
$text = preg_replace('|\\s+|', ' ', $text);
return trim($text);
}
/**
* Turns HTML into text browser visible text. Images
* are converted to their alt text and tags are supressed.
* Entities are converted to their visible representation.
* @param string $html HTML to convert.
* @return string Plain text.
* @access public
*/
static function normalise($html)
{
$text = preg_replace('|<!--.*?-->|', '', $html);
$text = preg_replace('|<script[^>]*>.*?</script>|', '', $text);
$text = preg_replace('|<img[^>]*alt\\s*=\\s*"([^"]*)"[^>]*>|', ' \\1 ', $text);
$text = preg_replace('|<img[^>]*alt\\s*=\\s*\'([^\']*)\'[^>]*>|', ' \\1 ', $text);
$text = preg_replace('|<img[^>]*alt\\s*=\\s*([a-zA-Z_]+)[^>]*>|', ' \\1 ', $text);
$text = preg_replace('|<[^>]*>|', '', $text);
$text = SimpleHtmlSaxParser::decodeHtml($text);
$text = preg_replace('|\\s+|', ' ', $text);
return trim(trim($text), "�");
// TODO: The \xAO is a . Add a test for this.
}
/**
* Turns HTML into text browser visible text. Images
* are converted to their alt text and tags are supressed.
* Entities are converted to their visible representation.
* @param string $html HTML to convert.
* @return string Plain text.
* @access public
* @static
*/
function normalise($html)
{
$text = preg_replace('|<!--.*?-->|', '', $html);
$text = preg_replace('|<script[^>]*>.*?</script>|', '', $text);
$text = preg_replace('|<img[^>]*alt\\s*=\\s*"([^>]*)"[^>]*>|', ' \\1 ', $text);
$text = preg_replace('|<img[^>]*alt\\s*=\\s*\'([^>]*)\'[^>]*>|', ' \\1 ', $text);
$text = preg_replace('|<img[^>]*alt\\s*=\\s*([a-zA-Z_]+)[^>]*>|', ' \\1 ', $text);
$text = preg_replace('|<[^>]*>|', '', $text);
$text = SimpleHtmlSaxParser::decodeHtml($text);
$text = preg_replace('|\\s+|', ' ', $text);
return trim($text);
}
