fix: use utf-8 encoding for subprocess to handle non-ascii chars

fix: use utf-8 encoding for subprocess to handle non-ascii chars #31