Characters getting mangled during subset

[tobys-at-canva]

I am attempting to subset a font and finding that the resulting characters are completely mangled.

The font I am subsetting is Noto Sans CJK SC Bold from https://www.google.com/get/noto/help/cjk/.

My exact code is this:

var Font = require('fonteditor-core').Font;
var fs = require('fs');
var buffer = fs.readFileSync('fonts/NotoSansCJKsc-Bold.otf');

// read font data
var font = Font.create(buffer, {
  type: 'otf', 
  subset: [0, 0x5b99, 0x6d2a], 
  hinting: false, 
  compound2simple: false, 
  inflate: null, 
  combinePath: false,
});

var buffer = font.write({
  type: 'woff', 
  hinting: true,
  deflate: null,
});

fs.writeFileSync('build/output.woff', buffer);

And while I can perform other operations without issue (eg, convert the full file to ttf), this code results in mangled characters in the output file.

Any ideas on what I may be able to try to get these subsets to look correct?

[Qix-]

I'm seeing this as well. Will try a manual approach as a workaround.

[zapalagrzegorz]

Also experiencing problems with font Poppins after subsetting. Not an issue on v.1.0.5

[gongpeione]

same here

[varlinux]

subset is a Decimal array:

var font = Font.create(buffer, {
  type: 'otf', 
  // subset: [0, 0x5b99, 0x6d2a], // false
  subset: [0, 23449, 27946], // true, 0x5b99 -> 23449, 0x6d2a => 27946
  hinting: false, 
  compound2simple: false, 
  inflate: null, 
  combinePath: false,
});