json-schema
json-schema copied to clipboard
Bugfix: Problem with items being an array of schema
Hey guys @justinrainbow @bighappyface @shmax @martin-helmich,
I think this test should be working, just like the 20 or the 21, it is quite similar but still produces a wrong result.
How could we resolve this case ? We need your help on this 😢
My colleague @alex-pex found the bug and it fits perfectly with the other test cases We are waiting for the merge :smiley:
@erayd Hi, I changed the number of the test case
Are you talking about this one ? #417
@erayd did you make any progress solving this issue?
@erayd we still have the problem and no update on this issue, can you look into it?
@alex-pex I've looked into some of it, but not all yet - I'm woefully short of free time to work on this library. It'll happen at some point, but I can't promise when - my paid work takes priority.
@shmax I don't suppose you have a spare moment to investigate this? You reviewed all the original stuff for the defaults feature, so hopefully will know what to look at...
Sure, I'll look into it tomorrow...
@erayd I spent the day working on this, but wasn't very productive (I'm currently living out of a suitcase, and the laptop I brought with me isn't set up for development, so I burnt up all my time setting up my environment). I haven't gotten as far as understanding how the recursive default stuff works, but I read through your comments on #417 and if nothing else was able to confirm that this proposed fix doesn't seem to help with that issue.
I'm really crushed for time. If you're worried about #417, maybe we can see if @wlalele would be willing to add a test for it to this PR, and see if he can account for it? At this point he certainly knows more about $path->seFromDefault() than I do and I'm unlikely to be able to catch up before next weekend.
Same issue: https://github.com/justinrainbow/json-schema/issues/632
The changes seems good for this PR. However I feel the same sentiment as @erayd about #417 / #632 I think this is going to be another topic on the milestone for this library. So I'm not inclined to merge this PR ATM. Edit: See #727 for current priorities (applying defaults to objects in arrays is currently not our top priority)
On a side note the actions for this PR seems to be lost due to time passing by. Perhaps a git commit --amend --no-edit && git push --force-with-lease can resolve that.