UTF-8 searches

[jkacmarcik]

We're in the process of setting up a journal that will include articles (in html) that include UTF-8 character encoding. Everything works well except the searching, so I fixed it.

includes/functions.inc.php:
line 128

Code: Select all

array_push($vars, htmlspecialchars($var));

becomes

Code: Select all

array_push($vars, str_replace('&', '&', htmlspecialchars($var)));

line 151

Code: Select all

array_push($vars, htmlspecialchars($var));

becomes

Code: Select all

array_push($vars, str_replace('&', '&', htmlspecialchars($var)));

htmlspecialchars() strips out html tags, slashes and ampersands -- however the ampersands are needed in UTF-8.

So, is there any reason that the ampersands can't be left alone? There shouldn't be any [new] security issues, if someone wanted to get an ampersand past as-is they could use a doubly encode one.

-jeffrey