clipboard: fix up html entities being escaped #6396
Conversation
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
1 Skipped Deployment
|
| json = JSON.parse( | ||
| tldrawHtmlComment | ||
| .replace(/>/g, '>') | ||
| .replace(/</g, '<') | ||
| .replace(/&/g, '&') | ||
| ) |
There was a problem hiding this comment.
We have some issues here at the moment:
Looking at the code for actually copying this stuff, it seems like we're not actually explicitly html-encoding these things ourselves. Where does the encoding come from? If we're using a regex here instead of a more robust html escaping system, why were these three chosen? I would have expected to see the reverse of this encoding explicitly happening somewhere
There was a problem hiding this comment.
had a discussion with Alex offline. annoyingly we need to go back to the original revision of this:
https://github.com/tldraw/tldraw/pull/6344/commits
we still need to base64 encode the regular non-assets json.
ugh, this means we need a version 3...
|
btw can we sdk + dotcom hotfix this when it's ready? |
followup fix for #6344
because we don't base64 encode anymore, a knock-on effect is that any characters within this html blob gets treated as such. we need to unescape these upon paste.
hat-tip to Phil for the report.
Change type
bugfiximprovementfeatureapiotherRelease notes