Can you e-mail me the scripts (in a zip file or something)? I am afraid the bugsystem doesn't handle UTF8 well.

There should be no hidden UTF-8 characters in the string: all ascii compliant including the french accented e.

I did however have an error in the code where extra "(" was added. Here is the complete and correct code you can just copy-paste from any browser.

------
<?
$str = "Le Caf?< Caf? <"; 

$str = html_entity_decode(htmlentities($str,ENT_QUOTES,"UTF-8"));
echo $str;
?>

ascii is 7 bit, and accented characters are not in the 7 bit part and thus could I see them with a different charset, please send me the script.

gzipped code sent to derick@php.net.

This is not a bug at all, you're doing something wrong here ;-)

THe first part of your code is:
<?
$str = "Le Café< Café <";

$str = htmlentities($str,ENT_QUOTES,"UTF-8");

This will try to interpret the $str as an UTF-8 string and add htmlentities for all utf-8 chars that it finds in here.
Because é< is not a valid combination as a multi-byte UTF8 character, the é is simply skipped.

I don't know why the second é isn't stripped out though. 
Anyway, your code is wrong, don't try to use an iso-8859-1 string as UTF8.

replace the é with ? in my explanation, by browser used UTF8 to post the message ;-)

Thanks a bunch for the clarification. =)