Support for oneOf composition #132
Conversation
|
Not sure how |
Likely a new version of flake8 with updated or new rules. Thanks for fixing. |
| it = iter(iterable) | ||
| # check that one is True, then check that none is True in rest of the | ||
| # iterator | ||
| return any(it) and not any(it) |
There was a problem hiding this comment.
This is really clever. Enough so that I'd like to request that you write a set of simple test cases that demonstrate it works because it's not entirely intuitive at a glance how it works. I do however like this pattern alot.
| @@ -299,7 +300,7 @@ def generate_enum_validator(enum, **kwargs): | |||
|
|
|||
|
|
|||
| @skip_if_empty | |||
| def validate_allof_anyof(value, sub_schemas, context, method, **kwargs): | |||
| def validate_composition(value, sub_schemas, context, method, default_error_message=None, **kwargs): | |||
There was a problem hiding this comment.
I think error_message_override might be a more appropriate name than default_error_message. Alternatively I'd support removing this in place of just writing 3 small functions which each raised the appropriate error message for all/any/exactly_one_of so that this function doesn't have to expand it's signature. Your call.
|
@maroux If you you can hit the two small requests for cleanup I'll get this merged. Also it appears this now needs to be rebased. |
|
Any change of merging this? |
|
I can make the cleanup commits if @maroux don't mind |
|
@pipermerriam @maroux What do you think? |
|
Yeah go for it. Sorry I haven't had a chance to work on it in so long!
…On Thu, Nov 23, 2017 at 2:17 AM Alex R ***@***.***> wrote:
@pipermerriam <https://github.com/pipermerriam> @maroux
<https://github.com/maroux> What do you think?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#132 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/ADHZdhpoBmdNpxX65WQ1fBGnQ9sKXW6Rks5s5UZLgaJpZM4LJPSw>
.
|
|
@lyssdod please do. Ping me when this is ready for review. I'm closing this issue so you can open your own PR (which it appears you've already done) |
Fixes #131