dict-validator

Python module for validating dictionaries against a schema.

I built this module for use in a Flask web service that accepts JSON data and must validate it before proceeding. I first considered using Voluptuous, which has a similar end result. However, I don't really like the method they use of defining the schema itself as a dictionary. So I have created an object-oriented method of validating dictionaries, where you define your schema as a class or set of nested classes that indicate which attributes should exist in the dictionary.

Features

• express a dictionary schema in a clear, object-oriented manner
• type-check various attributes of the dictionary
• specify whether an attribute should be required or optional
• support for list types
• support for nested schemas

Example

class InnerDict(DictValidator):
  x = Field(int)                      # x is a required attribute and must be an integer
  y = Field(str, required=False)      # y is an optional attribute, and should be a string if it exists
  z = ArrayField(str)                 # z is a required array of strings
  
class MyDict(DictValidator):
  a = StructuredDictField(InnerDict)  # a is a required dictionary attribute that conforms to the schema above


dict1 = MyDict({})
dict2 = MyDict({'a': 5})
dict3 = MyDict({'a': {}})
dict4 = MyDict({'a': {'x': 5, 'y': 'string', 'z': ['a','b','c']}})

valid1 = dict1.validate()    # invalid because the dict does not contain a
valid2 = dict2.validate()    # invalid because a is not a dict
valid3 = dict3.validate()    # invalid because a fails the InnerDict validator
valid4 = dict4.validate()    # valid

Improvements

• add validators to individual fields (range validation on integers, length validation for strings, etc.)
• add support for attribute names that are reserved keywords in python. maybe escape with a double underscore.