Only allow dotted quad IPv4 addresses #40257
Conversation
See JuliaLang#40255 This commit makes Class A addresses a parse error in Julia because they are probably not what the user wanted anyway.
|
The code LGTM. Mustn't the "1 failing" check be a false alarm? Shouldn't we merge this and backport to 1.7 (assuming future LTS)? I.e. a good exception to new features after feature freeze, as strictly taking one away, that we do not want, improving security. |
|
We need to decide if this is breaking or not (if breaking, it needs to wait for v2 or possibly just be dropped) |
|
In a sense this is breaking, or "technically breaking". I was going by the assumption @StefanKarpinski preapproved of this in the issue (not marked 2.0; neither this PR he self-assigned, so I was just reminding in case forgotten): "The todo here is to [..] make sure that we error on the ones that we don't support. Arguably, erroring on weird potentially ambiguous formats is better for security than correctly parsing all possible formats" This "weird" format should be even more rare than octal, that @sbromberger hadn't seen in 20+ years, so was dropped (admittedly pre-1.0): We could at least do PkgEval, to see if anything known breaks. |
|
It's definitely a minor change at least, so too late for 1.7. PkgEval should be done to see if any registered packages use this. If not, I'd be ok for 1.8. |
|
Bump. This was maybe forgotten because of "1 failing" buildbot/tester_macos64 which is strange and I can't see what's going on there. Can we do a PkgEval, since it must be a false alarm(?) and then merge this for 1.8? |
See #40255
This commit makes Class A addresses a parse error in Julia because they
are probably not what the user wanted anyway.