Solve the bug of parsing titles with unicode (cyrillic)