OpyAPI - JsonSchema and OpenAPI tools for python

JsonSchema and OpenAPI tools for python.

Features

• Validation builder for json schema
• Custom string formatters support
• Full JsonSchema draft-7 support
• Support for JsonReferences, $anchor
• Dataclass generation from OpenAPI schema
• Great performance

Installation

With pip,

pip install opyapi

or through poetry

poetry add opyapi

Usage

It is recommended to get familiar with json-schema if you haven't yet. Understanding Json Schema is a great place to start and learn Json Schema's basics

Simple usage

Library can be used to validate data, against defined json-schema like so:

from opyapi import validate

assert validate(
    {"name": "Test", "age":12}, 
    {
        "type": "object",
        "properties": {
            "name": {
                "type": "string",
            },
            "age": {
              "type": "integer",  
            }
        }
    }
)

Using references

Validating documents against complex schemas is also possible, please consider the following example with json references:

from opyapi import validate

schema = {
    "type": "object",
    "properties": {
        "name": {
            "type": "string",
            "minLength": 2,
        },
        "email": {
            "$ref": "#/$defs/email"
        }
    },
    "$defs": {
        "email": {
            "type": "string",
            "format": "email"
        }
    }
}

validate({"name": "Bob", "email": "bob@test.com"}, schema)

Custom string formats support

from opyapi import StringFormat, validate
import re

def my_format_validator(value: str) -> str:
    if re.match("^my-", value):
        return value
    raise ValueError(f"Could not validate {value}")

StringFormat["my-format"] = my_format_validator

validate("my-test", {"type": "string", "format": "my-format"})  # passes
validate("test", {"type": "string", "format": "my-format"})  # fails

In the above example opyapi.StringFormat is used to register new custom format, which is recognised during validation.

Re-using validators

There are scenarios where same validator should be used multiple times, in these scenarios to improve performance it is better to use build_validator_for which returns a validator function for the passed schema:

from opyapi import build_validator_for

my_validator = build_validator_for({
    "type": "object",
    "properties": {
        "name": {
            "type": "string",
            "minLength": 2,
        },
        "email": {
            "type": "string",
            "format": "email",
        }
    },
})

assert my_validator({"name": "Bob", "email": "bob@email.com"})

Catching errors

opyapi provides versatile error package, which is really simple to use. You are only required to define try/except block within you can catch one of the errors defined in the package depending on your scenario.

from opyapi import build_validator_for
from opyapi.errors import ValidationError

my_validator = build_validator_for({
    "type": "object",
    "properties": {
        "name": {
            "type": "string",
            "minLength": 2,
        },
        "email": {
            "type": "string",
            "format": "email",
        }
    },
})

try:
    my_validator({"name": "Bob", "email": "invalid"})
except ValidationError as e:
    print(f"Validation failed from the reason: {e}")

Errors structure

The following graph represents exception used in opyapi. All exceptions are part of opyapi.errors module.

ValueError
    ┗━ ValidationError
            ┣━ TypeValidationError
            ┣━ EnumValidationError
            ┣━ FormatValidationError
            ┣━ ItemsValidationError
            ┃           ┣━ UniqueItemsValidationError
            ┃           ┣━ AdditionalItemsValidationError
            ┃           ┣━ MinimumItemsValidationError
            ┃           ┗━ MaximumItemsValidationError
            ┣━ MultipleOfValidationError
            ┣━ ComparisonValidationError
            ┃           ┣━ EqualityValidationError
            ┃           ┗━ RangeValidationError
            ┃                       ┣━ MinimumValidationError
            ┃                       ┣━ MaximumValidationError
            ┃                       ┣━ ExclusiveMinimumValidationError
            ┃                       ┗━ ExclusiveMaximumValidationError
            ┗━ ObjectValidationError
                        ┣━ PropertyValidationError
                        ┃           ┣━ RequiredPropertyValidationError
                        ┃           ┣━ PropertyValueValidationError
                        ┃           ┣━ PropertyNameValidationError
                        ┃           ┗━ AdditionalPropertiesValidationError
                        ┣━ ObjectSizeValidationError
                        ┃           ┣━ MinimumPropertiesValidationError
                        ┃           ┗━ MaximumPropertiesValidationError
                        ┗━ DependencyValidationError

opyapi.errors.ValidationError

Generic validation error.

opyapi.errors.TypeValidationError

Raised when passed type does not conform type defined in the schema.

Can be triggered by following example schema:

{
  "type": "integer"
}

opyapi.errors.EnumValidationError

Raised when value does not conform enum definition.

Can be triggered by following example schema:

{
  "enum": ["integer", "string", "boolean"]
}

opyapi.errors.FormatValidationError

Raised when value does not conform defined format in string schema.

Can be triggered by following example schema:

{
  "type": "string",
  "format": "date-time"
}

opyapi.errors.ItemsValidationError

Generic Exception raised when validation of an array fails for some reason.

opyapi.errors.UniqueItemsValidationError

Raised when items in an array are expected to be unique but passed value does not conform the requirement. This exception extends generic opyapi.errors.ItemsValidationError exception.

Can be triggered by following example schema:

{
  "type": "array",
  "uniqueItems": true
}

opyapi.errors.AdditionalItemsValidationError

Raised when additional items in an array either does not conform the schema or are not expected. This exception extends generic opyapi.errors.ItemsValidationError exception.

Can be triggered by the following schema:

{
  "type": "array",
  "items": [
    {"type": "string"},
    {"type":  "integer"}        
  ],
  "additionalItems": false
}

opyapi.errors.MinimumItemsValidationError, opyapi.errors.MaximumItemsValidationError

Raised when number of items in an array does not conform maximum or minimum items specified in schema. This exception extends generic opyapi.errors.ItemsValidationError exception.

Can be triggered by following example schema:

{
  "type": "array",
  "minimumItems": 2,
  "MaximumItems": 10
}

opyapi.errors.MultipleOfValidationError

Raised when validated number is not multiplication of passed value.

Can be triggered by following example schema:

{
  "type": "numerical",
  "multipleOf": 2
}

opyapi.errors.ComparisonValidationError

Raised when comparison operation fails. This is a generic exception used by other comparison errors.

opyapi.errors.EqualityValidationError

Raised when validated value is not the same as defined expected value. This exception extends opyapi.errors.ComparisonValidationError.

Can be triggered by following example schema:

{
  "const": "test"
}

opyapi.errors.RangeValidationError

Raised when (exclusive) minimum, (exclusive) maximum comparisons fail. This exception extends opyapi.errors.ComparisonValidationError.

opyapi.errors.MinimumValidationError, opyapi.errors.MaximumValidationError, opyapi.errors.ExclusiveMinimumValidationError, opyapi.errors.ExclusiveMaximumValidationError

Raised when passed numerical (or integer) value is not within expected range defined in schema.

Can be triggered by the following example schema:

{
  "type": "numerical",
  "minimum": 2
}

opyapi.errors.ObjectValidationError

Generic exception raised when validation of an object fails for some reason.

opyapi.errors.PropertyValidationError

Generic exception raised when validation of object's property fails for some reason. This exception extends opyapi.errors.ObjectValidationError exception.

opyapi.errors.RequiredPropertyValidationError

Raised when required property is not present in passed object. This exception extends opyapi.errors.PropertyValidationError exception.

Can be triggered by the following example schema:

{
  "type": "object",
  "properties": {
    "name" : {"type": "string"}
  },
  "required": ["name"]
}

opyapi.errors.PropertyValueValidationError

Raised when property contains invalid value. This exception extends opyapi.errors.PropertyValidationError exception.

Can be triggered by the following example schema:

{
  "type": "object",
  "properties": {
    "name" : {"type": "string"}
  }
}

opyapi.errors.PropertyNameValidationError

Raised when property's names does not conform defined schema. This exception extends opyapi.errors.PropertyValidationError exception.

Can be triggered by the following example schema:

{
  "type": "object",
  "properties": {
    "name" : {"type": "string"}
  },
  "propertyNames": {
    "pattern": "^x-" 
  }
}

opyapi.errors.AdditionalPropertiesValidationError

Raised when additional properties are not allowed or set schema is not followed. This exception extends opyapi.errors.PropertyValidationError exception.

Can be triggered by the following example schema:

{
  "type": "object",
  "properties": {
    "name" : {"type": "string"}
  },
  "additionalProperties": false
}

opyapi.errors.ObjectSizeValidationError

Generic exception raised when number of properties does not conform defined schema. This exception extends opyapi.errors.ObjectValidationError exception.

opyapi.errors.MinimumPropertiesValidationError, opyapi.errors.MaximumPropertiesValidationError

Generic exception raised when number of properties does not conform defined schema. These exceptions extend opyapi.errors.ObjectSizeValidationErrorexception.

Can be triggered by the following example schema:

{
  "type": "object",
  "properties": {
    "name" : {"type": "string"}
  },
  "minProperties": 2,
  "maxProperties": 10
}

opyapi.errors.DependencyValidationError

Raised when dependent properties are defined in the schema but not provided in a passed object. This exception extends opyapi.errors.ObjectValidationError exception.

Can be triggered by the following example schema:

{
  "type": "object",
  "properties": {
    "name" : {"type": "string"}
  },
  "dependentRequired": {
    "name": ["first_name", "last_name"]
  }
}

API

validate(obj: typing.Any, schema: typing.Union[dict, opyapi.JsonSchema]): typing.Any

Validates passed object obj, and if valid returns the object, otherwise raises a ValueError exception.

from opyapi import validate

assert validate(
    {"name": "Test", "age":12}, 
    {
        "type": "object",
        "properties": {
            "name": {
                "type": "string",
            },
            "age": {
              "type": "integer",  
            }
        }
    }
)

build_validator_for(schema: typing.Union[dict, JsonSchema]) -> Callable

Creates validator function for passed json schema and returns it as a result.

from opyapi import build_validator_for

validator =  build_validator_for({
    "type": "object",
    "properties": {
        "name": {
            "type": "string",
        },
        "age": {
          "type": "integer",  
        }
    }
})