Emojis splitted up unexpectedly (e.g. https://emojipedia.org/ninja-cat/)

[ClemensSchneider]

Hi there,

first of all, thanks a lot for this library and the efforts you put in!

I've got a scenario, where some emojis seem to be split up the wrong way.

When splitting up the following emoji-sequence:
🐱‍💻🐱‍🚀🐱‍👤

I get the following string-tokens (notice the first two matching and the ninja-cat being split into two):

Is there an easy explanation for the behavior or is there a general guideline on which emojis are supported and which aren't?

I'm on Windows 10.

Thanks!

[jasonsbarr]

See #28

[ClemensSchneider]

This seems to explain it, yes.
So is there an easy way to disallow inputting / identify non-recommended unicode emojis that are not meant to be supported cross-platform?

[shmibbles]

i'm also getting unexpected splitting: 👨🏿‍🦰 splits into [ "👨🏿‍", "🦰" ], and this emoji is officially part of unicode, check it out. Might have something to do with zero width joiners not being recognized correctly?

Edit: just did some more testing, and 👨🏿‍🦰 is split after the ZWJ (\u200d), so 👨🏿‍🦰 split looks like this:
[ "\ud83d\udc68\ud83c\udfff\u200d", "\ud83e\uddb0" ]. Maybe the red hair component (\ud83e\uddb0) and other hair components aren't recognized after the ZWJ? hope this helps

[rodrigobutta]

Hi! Same here, this emoji 👨‍🦳 gets splited into [ '👨‍', '🦳' ] that are these unicodes [ 128104, 8205 ], [ 129459 ]. Also happening with other similar emojis. I coulnd`t find a patron yet so I can fix it.

[Lemmingh]

@shmibbles

I think you're right. They were introduced in Unicode 11.0, so this depends on #24