Contexts

This page covers all about the purpose and working of Contexts.

Introduction

Context objects are classes meant to hold useful contextual data about some event. These classes are passed by library in user side functions and methods.

Schema Context

SchemaContext is accessed by Schema.context attribute. This class stores information about a schema and its state.

Field Contexts

When a field is being processed, some contextual information is required to be passed to user side function and methods. This is done by field contexts.

There are two field contexts:

• LoadContext when a field is being loaded (deserialized)

• DumpContext when a field is being dumped (serialized)

LoadContext is passed to methods and functions involved in deserialization of fields such as fields.Field.value_load() and validators while DumpContext is passed to fields.Field.value_dump().

Error Context

ErrorContext holds contextual information about a validation error. The instance of this class is passed to fields.Field.format_error().

Context States

All contexts provided by the library have a state attribute which is a dictionary. This attribute exists to allow users to easily propagate or store important state information in the context and access it later at some point.

A common use case of state is when loading a field. The fields.Field.value_load() method can store some data in state which validators can access.

Another use case is preserving the raw value of some field to access it later during the serialization of field. Example of this use case is given below:

class SumValues(fields.Field[list[int], int]):
    def value_load(self, value: list[int], ctx: oblate.LoadContext) -> int:
        if not isinstance(value, list):
            raise ValueError('Value for this field must be a list of integers')

        result = 0
        for idx, num in enumerate(value):
            if not isinstance(num, int):
                raise ValueError(f'Non-integer value at index {idx}')

            result += num

        ctx.schema.context.state['marks_raw_value'] = value
        return result

    def value_dump(self, value: int, ctx: oblate.DumpContext) -> list[int]:
        return ctx.schema.context.state['marks_raw_value']

class Student(oblate.Schema):
    name = fields.String()
    test_score = SumValues()

std = Student({'name': 'John', 'test_score': [10, 9, 5, 6]})
print(std.test_score)  # 30

print(std.dump()['test_score'])  # [10, 9, 5, 6]