Correctly sanitize object replacement character by treating as whitespace

[alexvcasillas]

Hey there, I'm Alex from Tiptap 👋🏻

I wanted to upstream this issue to you as it's something that's affecting linkifyjs directly. It seems that linkifyjs does not handle properly the sanitization of a string when it comes to the find method that looks for link matches within a string. Here's an example of this character: \uFFFC this is an object replacement character and it looks like an empty string to the naked eye but if you pass something like find('https://google.com this) (where the space between the .com and this is the given character \uFFFC) the method will return you that the link is https://google.com this which some might say it's correct as it's a valid unicode character but people will be very confused by this as we've been by the reports we were having in our editor repo.

Just wanted to upstream this to you to see if it's something that you'd consider adding to the find method or to help anyone that might be having this issue so they know what could be about.

This was the issue in which we ended up discovering this ueberdosis/tiptap#5679 just in case you're curious what this could potentially cause :)

Hope you have a great day 🙇🏻

[nfrasser]

Hi Alex, thanks for reporting! This should be a doable fix, linkify can interpret \uFFFC as whitespace and therefore exclude anything that comes after it from the link. Will update here when the fix is available.

[alexvcasillas]

Thank you @nfrasser 🙇🏻

[nfrasser]

@alexvcasillas fix is available in the latest v4.2.0 release.