[https://www.loom.com/share/682e25f621d14f57be042772f6e3bf2f](https://www.loom.com/share/682e25f621d14f57be042

Question

[https://www.loom.com/share/682e25f621d14f57be042772f6e3bf2f](https://www.loom.com/share/682e25f621d14f57be042772f6e3bf2f "smartCard-inline")

shyam · Accepted Answer

[https://www.loom.com/share/682e25f621d14f57be042772f6e3bf2f](https://www.loom.com/share/682e25f621d14f57be042772f6e3bf2f "smartCard-inline") The contentUiRenderer is running twice in dev. This is normal, the issue is the 2nd time it runs it ignores it’s first match and goes to the 2nd match. If a link like [https://twitter.com/s/dog](https://twitter.com/s/dog "‌") is saved it will match the url first then on rerender match the /s as a space and run that logic. Since the last render is what saves it then delivers a broken link for twitter. Suggested Fix: - Convert regex’s to match on , <@>, and - Make the front end mark these before sending to server - Have server capture these, validate their accuracy then add the key=x from the uuid in the post for security - Rename Badwords function in server to contentFilter and have it find links, /s, \@‌username and then process the same way cusswords do to include the id security check

useContentRenderer update

Activity