reintroduce tidy in calls from egw-mail-app when only getCleanHTML is called, as it is correcting html structure issues for us

This commit is contained in:
Klaus Leithoff 2015-10-27 09:22:12 +00:00
parent 9f4a6db299
commit f06cf7bb01
3 changed files with 60 additions and 19 deletions

View File

@ -4687,23 +4687,8 @@ class emailadmin_imapbase
if ($bodyParts[$i]['charSet']===false) $bodyParts[$i]['charSet'] = translation::detect_encoding($bodyParts[$i]['body']);
// add line breaks to $bodyParts
//error_log(__METHOD__.' ('.__LINE__.') '.' Charset:'.$bodyParts[$i]['charSet'].'->'.$bodyParts[$i]['body']);
$newBody = translation::convert($bodyParts[$i]['body'], $bodyParts[$i]['charSet']);
$newBody = translation::convert_jsonsafe($bodyParts[$i]['body'], $bodyParts[$i]['charSet']);
//error_log(__METHOD__.' ('.__LINE__.') '.' MimeType:'.$bodyParts[$i]['mimeType'].'->'.$newBody);
/*
// in a way, this tests if we are having real utf-8 (the displayCharset) by now; we should if charsets reported (or detected) are correct
if (strtoupper(self::$displayCharset) == 'UTF-8')
{
$test = json_encode($newBody);
//error_log(__METHOD__.' ('.__LINE__.') '.'#'.$test.'# ->'.strlen($newBody).' Error:'.json_last_error());
if (json_last_error() != JSON_ERROR_NONE && strlen($newBody)>0)
{
// this should not be needed, unless something fails with charset detection/ wrong charset passed
error_log(__METHOD__.' ('.__LINE__.') '.' Charset Reported:'.$bodyParts[$i]['charSet'].' Carset Detected:'.translation::detect_encoding($bodyParts[$i]['body']));
$newBody = utf8_encode($newBody);
}
}
*/
//error_log(__METHOD__.' ('.__LINE__.') '.' before purify:'.$newBody);
$mailClass->activeMimeType = 'text/plain';
if ($bodyParts[$i]['mimeType'] == 'text/html') {
$mailClass->activeMimeType = $bodyParts[$i]['mimeType'];

View File

@ -2123,6 +2123,22 @@ class mail_compose
static $nonDisplayAbleCharacters = array('[\016]','[\017]',
'[\020]','[\021]','[\022]','[\023]','[\024]','[\025]','[\026]','[\027]',
'[\030]','[\031]','[\032]','[\033]','[\034]','[\035]','[\036]','[\037]');
if (extension_loaded('tidy'))
{
$tidy = new tidy();
$cleaned = $tidy->repairString($_body, mail_bo::$tidy_config,'utf8');
// Found errors. Strip it all so there's some output
if($tidy->getStatus() == 2)
{
error_log(__METHOD__.' ('.__LINE__.') '.' ->'.$tidy->errorBuffer);
}
else
{
$_body = $cleaned;
}
}
mail_bo::getCleanHTML($_body);
return preg_replace($nonDisplayAbleCharacters, '', $_body);
}

View File

@ -2778,10 +2778,50 @@ class mail_ui
$newBody = $singleBodyPart['body'];
//TODO:$newBody = $this->highlightQuotes($newBody);
#error_log(print_r($newBody,true));
if (extension_loaded('tidy'))
{
$tidy = new tidy();
$cleaned = $tidy->repairString($newBody, mail_bo::$tidy_config,'utf8');
// Found errors. Strip it all so there's some output
if($tidy->getStatus() == 2)
{
error_log(__METHOD__.' ('.__LINE__.') '.' ->'.$tidy->errorBuffer);
}
else
{
$newBody = $cleaned;
}
if (!$preserveHTML)
{
// filter only the 'body', as we only want that part, if we throw away the html
preg_match('`(<htm.+?<body[^>]*>)(.+?)(</body>.*?</html>)`ims', $newBody, $matches=array());
if ($matches[2])
{
$hasOther = true;
$newBody = $matches[2];
}
}
}
else
{
// htmLawed filter only the 'body'
preg_match('`(<htm.+?<body[^>]*>)(.+?)(</body>.*?</html>)`ims', $newBody, $matches=array());
if ($matches[2])
{
$hasOther = true;
$newBody = $matches[2];
}
$htmLawed = new egw_htmLawed();
// the next line should not be needed, but produces better results on HTML 2 Text conversion,
// as we switched off HTMLaweds tidy functionality
$newBody = str_replace(array('&amp;amp;','<DIV><BR></DIV>',"<DIV>&nbsp;</DIV>",'<div>&nbsp;</div>'),array('&amp;','<BR>','<BR>','<BR>'),$newBody);
$newBody = $htmLawed->egw_htmLawed($newBody);
if ($hasOther && $preserveHTML) $newBody = $matches[1]. $newBody. $matches[3];
}
// do the cleanup, set for the use of purifier
$newBodyBuff = $newBody;
//$newBodyBuff = $newBody;
mail_bo::getCleanHTML($newBody);
/*
// in a way, this tests if we are having real utf-8 (the displayCharset) by now; we should if charsets reported (or detected) are correct
if (strtoupper(mail_bo::$displayCharset) == 'UTF-8')
{
@ -2796,7 +2836,7 @@ class mail_ui
mail_bo::$htmLawed_config['tidy'] = $tv;
}
}
*/
// removes stuff between http and ?http
$Protocol = '(http:\/\/|(ftp:\/\/|https:\/\/))'; // only http:// gets removed, other protocolls are shown
$newBody = preg_replace('~'.$Protocol.'[^>]*\?'.$Protocol.'~sim','$1',$newBody); // removes stuff between http:// and ?http://