Samuel Colvin comments

Results 1016 comments of


                                            Samuel Colvin

Layout conflict in ModelMetaclass when mixing in two models with private variables.

We often sort issue by upvote to decide priority, I also actively ignore issues with "me too" comments. I strongly suggest therefore, that you use the built in feature to...

Decimal JSON encoder is lossy

Okay, please stop the "me too" messages. We're talking about two distinct things here: 1. what `type` we use when generating JSON schema for a model - by default this...

Decimal JSON encoder is lossy

Please don't confuse this issue, for the validation JSON Schema question, see https://github.com/pydantic/pydantic/pull/5029#discussion_r1107356818.

Decimal JSON encoder is lossy

@colesiegel your question isn't clear - are you asking to change the validation or change the generated JSON schema? If you want to change the validation logic, use a `model='before'`...

Decimal JSON encoder is lossy

I'm locking this issue as the original problem in Pydantic V1 was fixed in V2.

Schema generation for discriminated unions using literal Enum values is incompatible with OpenAPI

Will be fixed in V2, ref #4666. Pr welcome to fix in v1.10.

Memory leak after updating to Pydantic v1.9.0

Anyone using large numbers of generics, would be great if you could review (and maybe install and test) #5052 - I'd like to release this soon, but I'm mindful of...

`use_enum_values` but only on `.dict()`

This is fixed I believe by `mode='json'` on `model_dump`.

Add ability to have custom encoder for builtin types

@uriyyo thanks so much for submitting this, the problem is fixed by the new serialisation logic in v2.

aliases are implemented in pydantic-core - both load and dump aliases, also more more powerful load aliases, see tests [here](https://github.com/pydantic/pydantic-core/blob/fb69c39168e8f8df3860b7164d5cedfa64acc754/tests/validators/test_typed_dict.py#L512).