I don't understand the use-case. How are you using this? This module is meant to measure the visual width.

We always break line if the line fit the printWidth, need to measure the visual width to support CJK.

Does it matter, though?

• Is there really any source code that contains valid ANSI escapes (that is, not using any JavaScript escapes such as \u001b)?
• If there is, the lines containing them will be slightly off (or you might not notice it at all). Not the end of the world.

From a correctness standpoint then, yes, we shouldn't strip ANSI in Prettier's case, but from a practical standpoint I don't think it's an issue?

I think https://github.com/sindresorhus/string-length is what you actually want to use. This module is for text printed in the terminal.

CJK characters should be measured as 2-char width (visual width):

• stringLength("中"); //=> 1
• stringWidth("中"); //=> 2

So I think string-width should be what we want.

We do need the feature for text printed but not in the terminal so that we (maybe) need the ability to disable the stripAnsi, but as @lydell mentioned above, it seems not a big issue since nobody writes ANSI escapes without JavaScript escapes such as \u001b.

So I'm going to close it now unless someone open an issue about this in prettier.

@sindresorhus thanks for your quick response!

CJK characters should be measured as 2-char width (visual width) - @ikatyang

Why? Doesn't that mean that a markdown paragraph written using almost entirely CJK characters would break at 40 characters instead of 80? That seems surprising to me.

Do you want to treat it this way because those characters often print as slightly-wider-than-1-character due to them not being in the monospace font that the rest of the text is rendered in?

CJK characters are 2-char width in monospace font.