Conversation
|
Is it possible to apply this change to v1.4.1 of d3-color as well? For those who still need to support ES5 it would be helpful to have the vulnerability fixed on the ES5 compatible version of d3-color as well. I am happy to submit a PR off of the v1.4.1 tag if bumping that lesser version to v1.4.2 is something you are willing to do. Please let me know? |
|
Would love to see a backport if possible; we would also appreciate if we could get a backport to the v2 tag as well because that's the highest major version supported by I'm happy to help with these if there's anything I can do, including getting the GitHub advisories updated. |
We would also like the dependencies of d3-color updated for v2 for d3-interpolate and d3-scale to support ES5 compatibility. |
@G-Rath and others who have 👍. I'm new to contributing to open source so would appreciate your help if possible. Since I don't have PR access on this repo, I forked it to |
|
All, I have logged an issue with d3 for backporting this fix to v2.x at #108. |
Fixes #97. Supersedes #89 and #99. The problem was that this expression is fundamentally ambiguous:
Since both the dot and the digits preceding the dot are optional, there’s a combinatorial explosion of possible valid matches. If we instead combine it into an optional group and make the dot required for that group, the explosion is avoided:
Demo: https://observablehq.com/d/4b1d645fe3da1226