[eluser]Alleknalle[/eluser]
Hi all,
I'm having hidden characters in my database and I don't know how I can make my checks safe enough to delete them.
The problem is the following:
When you paste it in Notepad++ a ? appears at the beginning of the line. In Netbeans a kind of dash appears.
When I convert the encoding, with mb_convert_encoding, to UTF-8  appears.
And when I use htmlentities or htmlspecialchars
appears.
It's not a big problem for my database and also not for the output, but when I put it into my CI session, my session is destroyed.
Now I would like to delete this kind of characters, they aren't usefull, but I don't know how.
I think I can't always delete all htmlspecialchars because I'm using WYSIWYG editors too.
Can anyone help me?
To clean the input I'm using trim and xss_clean now. For this input field I'm also using strip_tags.
EDIT:
db collation is utf8_unicode_ci
CI config charset is UTF-8
When I echo my post in CI I see nothing, not even in my source.
When I echo my post in simple, not CI, script I see nothing, but in my source I can see