[5.x]: zero width non-joiner stripped out rich text fields causing unwanted ligatures

[low]

What happened?

Description

It seems rich text fields like CKEditor strip out some invisible but necessary characters in some languages like Persian/Farsi.

Steps to reproduce

1. In a rich text field, paste this text: می‌شویم
2. Hit save. Now the text shows as میشویم

Expected behavior

I would expect the text to remain as می‌شویم

Actual behavior

The zero with non-joiner in the middle of those characters is stripped out, causing unwanted ligatures.

Craft CMS version

5 and 4

PHP version

8.1+

Operating system and version

No response

Database type and version

MySQL 8.1

Image driver and version

No response

Installed plugins and versions

[linear]

CMS-1899

PT-3068

[i-just]

Hi, thanks for reporting! I raised a PR for this.

[anita-chouhan]

Hello @i-just, we are also experiencing the same issue with Kannada text.

Input

ಫೆಲೋಶಿಪ್‌ಗಾಗಿ

Output

ಫೆಲೋಶಿಪ್ಗಾಗಿ

Craft CMS version

5.9.12

Plugin version

4.11.1