Closed pawelwilczewski closed 10 months ago
It would be good to add some test scenarios like the one in the mentioned issue.
Stole a little bit of test code from the issue - hopefully the author a-mckinley doesn't mind
@george-zubrienko done! Let me know if anything else needs changing!
I noticed it won't work with .schema.loads()
because I never made changes to the mm.py
. I'm afraid I won't have free time to get around to doing that in the nearest future...
I'd think a fix similar to what I did could be implemented in the else
clause in line 87 in mm.py
I'd think a fix similar to what I did could be implemented in the
else
clause in line 87 in mm.py
I'll review what is here this week, then we'll open a new issue and somebody can pick it up ;)
I'd think a fix similar to what I did could be implemented in the
else
clause in line 87 in mm.pyI'll review what is here this week, then we'll open a new issue and somebody can pick it up ;)
would love to see this one go through as we have another PR that adds more support for undiscriminated unions we would like to go in after this https://github.com/pawelwilczewski/dataclasses-json/pull/1
Not too sure why the poetry installation might have failed...
Merging https://github.com/lidatong/dataclasses-json/pull/476 will fix, sorry for delays in reviewing, on my todo
No worries at all, and no pressure!
Seems like there were some issues with the refactor. I'll roll back and you can automatically merge - unless you'd like to give it another go. Cheers!
I'll take a look why and suggest a new one. The one I added was from the phone so might have missed the if
condition
Yeah I gave it an attempt but kinda just skimmed through it relying on github actions rather than properly testing
If you can fix would be great, if not I'll prepare a proper suggestion within next 2 days. I want this to be included in 0.6.0 :)
I believe this does it now
Awesome, thanks a lot for the contribution, will be released this weekend!
Happy to contribute. Thanks a lot for your overall and this PR's input!
Fixed the issue described in #239.
I'm not sure this is the best way of fixing this as the comment suggests the type should have already been decoded but this is my best attempt.
Frankly, I haven't tested it properly (works for my use cases) but it seems to me the logic is valid.
I am not very familiar with the repo, so please let me know if I'm violating some of the rules set here and I'll try to adhere.