utf8_ch_len should count the number of items written in uft8_ch, however it starts unitialized and could lead to issues when ch >= 0x200000