add utf8proc so we can detect unicode char width

pkg/utf8proc/main.zig

@@ -0,0 +1,8 @@
+pub const c = @import("c.zig");
+
+/// Given a codepoint, return a character width analogous to `wcwidth(codepoint)`,
+/// except that a width of 0 is returned for non-printable codepoints
+/// instead of -1 as in `wcwidth`.
+pub fn charwidth(codepoint: u21) u8 {
+    return @intCast(u8, c.utf8proc_charwidth(@intCast(i32, codepoint)));
+}