Add JSONSchema parser

[jlowin]

This PR adds a new utility module jsonschema.py that converts JSON Schema definitions into Python types. This enables dynamic type creation and validation from JSON Schema specifications.

Key Features

• Converts JSON Schema into Python types
• Supports complex schema features including:
  • Basic types (string, number, integer, boolean, null)
  • Complex types (arrays, objects)
  • Format constraints (date-time, email, uri)
  • Numeric and string constraints (min/max, patterns, etc.)
  • Object properties with intelligent default handling
  • References and recursive schemas
  • Enums and constants
  • Union types

Example Usage

from marvin.utilities.jsonschema import jsonschema_to_type

schema = {
    "type": "object",
    "properties": {
        "name": {"type": "string", "minLength": 1},
        "age": {"type": "integer", "minimum": 0},
        "email": {"type": "string", "format": "email"}
    },
    "required": ["name", "age"]
}

Person = jsonschema_to_type(schema)
person = Person(name="John", age=30)

Implementation Details

• Uses modern Python type hints
• Implements intelligent caching to handle recursive schemas
• Provides robust default value handling with proper inheritance
• Sanitizes field names for Python compatibility
• Maintains original field names through aliases

Testing

The module includes extensive test coverage across multiple test classes:

• Basic Types: Validates string, number, integer, boolean, and null types
• Constraints: Tests string patterns, numeric bounds, array limits
• Complex Types:
  • Arrays with type validation and uniqueness
  • Objects with required fields and nested structures
  • Format types (datetime, email, URI)
• Advanced Features:
  • Circular references and recursive schemas
  • Property name normalization
  • Constant values and enums
  • Schema caching and hashing
  • Default value inheritance and merging