You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It looks like Github's cmarkgfm has changed the option CMARK_OPT_SAFE potentially to the inverse CMARK_OPT_UNSAFE. When running as CMARK_OPT_DEFAULT XSS issues are still prevalent.
On running safemd.render(content) the following error occurs type object 'Options' has no attribute 'CMARK_OPT_SAFE'.
The text was updated successfully, but these errors were encountered:
I should probably have added an update here, but that is true. I spoke with GitHub's maintainer of cmarkgfm and got the change (safe by default) into their upstream project. Then we talked with CommonMark maintainers and eventually we managed to get it changed in all of CommonMark. So all recently updated markdown renderers based on CommonMark or their derivates should nowadays be safe by default and not exhibit any XSS-problems.
My library still adds bleach, for additional security, in case another vulnerability in markdown appears.
But I'm not sure if it's worth to maintain a whole library for that purpose. If it's desired I'll update safemd but I've mainly kept it up for people still dependant on older versions of cmark/cmarkgfm.
It looks like Github's
cmarkgfm
has changed the optionCMARK_OPT_SAFE
potentially to the inverseCMARK_OPT_UNSAFE
. When running asCMARK_OPT_DEFAULT
XSS issues are still prevalent.On running
safemd.render(content)
the following error occurstype object 'Options' has no attribute 'CMARK_OPT_SAFE'
.The text was updated successfully, but these errors were encountered: