Strange Regex behavior

2CODE

Happy new year to all of you guys, Today I've a very strange REGEX behavior

Here's my regex code

return (preg_match('/(?:[^\x00-\x7F])/', $str) !== 1)

However it works fine outside the classes "As a normal function" but when I make it inside a class is always give me true if the text was UTF-8 or not.

Any ideas why it's always give me true inside classes and outside classes give me the correct value ?

bradgrafelman

How do you define $str when it isn't working? Can you show us a more complete example of the code you're using?

EDIT: Also, do you have display_errors set to On and error_reporting set to E_ALL? If $str didn't exist in the scope you're attempting to use it, then PHP will emit a warning and instead use NULL (I think?), and NULL certainly doesn't match your regexp pattern thus you'll always get boolean TRUE.

2CODE

Sure I have $str, Here's the full function to see what I mean

class utf8
{

public function Check($str)
{
    return (preg_match('/(?:[^\x00-\x7F])/', $str) !== 1);
}

public function seotitle ()
{

    if ($this->Check("dasdasdads")
    {
          echo 'utf8';
    } 

}

}

If you tried to use var_dump it will give you always boolen true if it was a correct UTF8 or not.

Here's a UTF-8 word to test "تجربة"

Weedpacket

2CODE;10993589 wrote:
Here's a UTF-8 word to test "تجربة"

Unfortunately, it's not; this page's character encoding is ISO-8859-1, not UTF-8 (which is probably why the quoted copy-and-paste is borked instead of rendering as تجرب&#1577😉; the Arabic characters in this page are specified as numeric entities (&#1578; etc.) since they don't exist in the ISO-8859-1 encoding.

Weedpacket

Your Check() method will return only true for plain 7-bit ASCII. Although all valid ASCII strings are valid UTF-8, there are plenty of byte sequences that are valid UTF-8 strings without being valid ASCII.

If you want to write a function to test whether a string is valid UTF-8, see [man]utf8_encode[/man] for a description of the encoding.