-
-
Notifications
You must be signed in to change notification settings - Fork 1.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
CVE With JSON5 v1.0.1 #2770
Comments
tsconfig-paths v3 depends on json5 v1, and v1.0.2 fixes the issue (which doesn't actually apply to eslint-plugin-import anyways, it's a false positive) so there's literally nothing we need to do here. |
@ljharb - it's easy to dismiss an issue for those reasons, but what you've done is damned us to transitive dependency scanning hell. Were you to add a Similarly, if you were to remain up to date with the latest packages, this wouldn't happen. Ignoring it on your end just means all your consumers with scanning software have to put up with the problem. You're expecting them to know that this is a false positive because the library's not really used. |
I really don't know what you mean. You have the freedom to trivially update your lockfile to the latest dependency - you can even It is both unreasonable and unnecessary to expect maintainers to keep ranged dependencies up to date. You are responsible for keeping your entire dep graph up to date within the specified ranges, not random maintainers. Separately, |
So if I submit a PR with the 1.0.2 fix in, you'll reject it, and instead explain why I shouldn't need to have a clean dependency graph? |
tsconfig-paths v3 would need that fix, since we don't directly depend on json5. You do need to have a clean dependency graph - it's just that the onus is on YOU to keep it so when updates are in-range, not on transitive maintainers. |
A few thoughts: A lot of people use this library. That's a lot of people who may face the same issue I've faced, which is trying to track down the importance/otherwise of a transitive dependency and whether it's safe to ignore/ There are a lot of known issues in this I made an experiment in PR #2771 to see if updating As far as I can tell, the unit tests pass... They may fail on the PR build. |
Unfortunately no, updating to v4 is a nonstarter; see #2447. |
Interestingly, you're on I appreciate the challenge of going for full backwards compatibility and how that stretches against dependency management. Thanks @ljharb for helping me get through the discussion to this final conclusion. |
Recommended upgrade to 1.0.2/2.2.2
https://github.com/advisories/GHSA-9c47-m6qq-7p4h , https://github.com/json5/json5/issues/199 , https://github.com/json5/json5/issues/295 , https://github.com/json5/json5/security/advisories/GHSA-9c47-m6qq-7p4h , https://nvd.nist.gov/vuln/detail/CVE-2022-46175
Evidence
This is a consequence of being on the v3.x of
tsconfig-paths
- the CVE is fixed in the later version oftsconfig-paths
which uses[email protected]
The text was updated successfully, but these errors were encountered: