Using the character index instead of the byte index

To replace the text properly, the character index will be needed, which in case of multibyte Unicode characters will differ from the byte index.