CKEditor issue with some special unicode characters

[Jan P.]

Dear ProcessWire community,
I recently found an issue with unicode characters used in CKEditor / Input fields. Most likely it seems to be a CKEditor configuration issue. Maybe someone has a hint to solve this.

First of all my setup:
ProcessWire 3.0.184 (latest master release) running on PHP 7.4.28
Input field Textarea / CKEditor with Markup HTML / use ACF (yes / no – doesn’t matter) / use HTML Purifier no

When I put a random Chinese character into this field (decimal 27721) so Ampersand Sharp 27721 Semicolon – it works perfectly out of box. This unicode character will be displayed and rendered.

Now my problem:
When I put one of those new fancy smiley unicode characters into this field (decimal 128513) the whole source code breaks after this point but only after pressing the save button. It is totally weird behaviour. The significant difference is the high decimal number.

I tried a few configuration suggestions from the CKEditor Reference Page for example:
config.entities = true
config.entities_processNumerical = true
config.entities_additional and exactly allowing decimal numbers

Nothing works. Does anybody have a good idea or is it a known CKEditor bug?

 

[zoeck]

Check out this thread ? 

Do you have your database charset set to utf8mb4?