Decoding text using str_replace, e.g. 'ü' => 'ü' does not work :-(

Hello,

the data in my database is encoded in a wrong way. Now I am trying to replace characters during the runtime, but I see no effect. This is the code I am using:

			$decode_ary = array(
					'ü' => 'ü',
					'ä' => 'ä',
					'ß' => 'ß',
					'ö' => 'ö',
					'Ãœ' => 'Ü',
					'Ä' => 'Ä',
					'é' => 'é',
					'ã' => 'ã',
					'ü' => 'ü',
					'ä' => 'ä',
					'ß' => 'ß',
					'ö' => 'ö',
					'Ãœ' => 'Ü',
					'Ä' => 'Ä',
					'é' => 'é',
					'ã' => 'ã',
					'©' => '©',
					'a' => 'A',
				);
            
            $topic_title = str_ireplace(array_keys($decode_ary), array_values($decode_ary), $topic_title);

Code (markup):

Example:
Baseballschläger --> BAseballschläger

But the code returns:
BAseballschläger

...The a is replaced by A (it's just a check if the replace-function works properly), but the actual char combination is not replaced. Why?

 

str_replace only works with one-byte characters. Your database may contain multi-byte sequences representing those characters.

Without knowing your database connection encoding and display encoding it's hard to say. As a starting point, you may want to try outputting the strings as fetched from the database byte-by-byte to see what's really in there.

 

He could perhaps use preg_replace for that instead.

 

Very trange thing happens...
I discovered that using utf8_decode almost all letters are converted correctly, except for ß. Instead of Groß the result is Gro�?. Why is there such inconsistency?

utf8_decode($topic_title)

 

Maybe the ß character was stored in a different encoding and your browser is being overly helpful by displaying it anyway. Look at the actual bytes being stored and your mystery will be solved.