Python Marshmallow set different fields as 'required' for each schema variant of the same schema

Question

I have the following schema:

class MySchema(Schema):
    id = fields.Str()
    name = fields.Str()
    value = fields.Str()
    description = fields.Str()

I want to validate some input data with Flask for different types of requests.

For GET requests, I want the id field to be required (id = fields.Str(required=True)).

For POST requests, I do not need the id to be present, but the name, value and description fields to have required=True. I know how to exclude the id when validating the input data: schema = MySchema(exclude=["id"]).

For PUT requests, I need id and at least one of name, value or description.

I could just have 3 different schemas but that I would avoid that as much as possible.

I was thinking something along the lines of schema = MySchema(require=["id"]), where I can dynamically set the required fields when I load the schema.

Answer 1

I found a nice solution to this issue, using the context property of the schema and the @validates_schema decorator.

schema = MySchema()
schema.context["method"] = "POST"

...

schema = MySchema()
schema.context["method"] = "PUT"

In MySchema:

from marshmallow import Schema, validates_schema, ValidationError, fields

class MySchema(Schema):
    id = fields.Str()
    name = fields.Str()
    value = fields.Str()
    description = fields.Str()

    @validates_schema
    def validate_schema(self, data, **kwargs):
        if "method" in self.context:
            required_fields = []
            missing_fields = []
            if self.context["method"] == "POST":
                required_fields = ["name", "value", "description"]
            
            if self.context["method"] == "PUT":
                required_fields = ["id"]
                
            for field in required_fields:
                if field not in data:
                    missing_fields.append(field)
            
            if len(missing_fields) >= 1:
                raise ValidationError({missing_field:["Missing data for required field."] for missing_field in missing_fields})

This approach can be used to also pass the list of required fields as a key value pair in the context dictionary, if you only know at runtime which fields are to be required.