Schema

0.7.8 · active · verified Sun Mar 29

Schema is a Python library for validating data structures, such as those obtained from config files, forms, external services, or command-line parsing, often converted from JSON/YAML to Python data types. It is actively maintained with moderate release frequency, typically addressing bug fixes and minor features, with occasional larger updates.

Warnings

gotcha Do not confuse `schema` (keleshev/schema) with `jsonschema`. This library provides a Pythonic way to define and validate data structures, and can generate JSON schema from its definitions. `jsonschema` is a separate library that strictly implements the JSON Schema specification.
Fix: Ensure you are importing from the correct library based on whether you need a Pythonic schema definition or adherence to the JSON Schema specification.
gotcha Since version 0.6.2, `schema.SchemaError` has specific subclasses like `SchemaWrongKey` and `SchemaMissingKeyError`. While `except SchemaError` will still catch these, code that relies on precise error type checking for key-related issues (e.g., `isinstance(e, SchemaWrongKey)`) may need to be updated if it was written before these subclasses were introduced and expected only the base `SchemaError` class for specific key problems.
Fix: When catching `SchemaError`, consider if more granular handling is needed for `SchemaWrongKey`, `SchemaMissingKeyError`, or other specific subclasses. Generally, catching the base `SchemaError` is sufficient for overall validation failure.
deprecated The `schema` library was tested against older Python versions (e.g., Python 2.6, 2.7, 3.2-3.9). These older Python versions are no longer officially supported by the Python Software Foundation and may have security vulnerabilities or compatibility issues with modern libraries.
Fix: Always use `schema` with a currently supported Python version (e.g., Python 3.8+). Update your Python environment if necessary.
gotcha An issue related to `Optional` fields was reverted in version 0.6.4 ("Revert the optional error commit"). This implies a temporary instability or unexpected behavior regarding `Optional` field validation in an intermediate version prior to 0.6.4. Users upgrading from a version that contained this problematic change might have encountered validation issues.
Fix: Ensure you are using version 0.6.4 or later if you experienced unexpected behavior with `Optional` fields in older installations. Always test `Optional` field behavior thoroughly after any major version upgrade.

Install

pip install schema Install latest version

Imports

Schema
```
from schema import Schema
```
And
```
from schema import And
```
Or
```
from schema import Or
```
Use
```
from schema import Use
```
Optional
```
from schema import Optional
```
Regex
```
from schema import Regex
```
SchemaError
```
from schema import SchemaError
```

Quickstart

This quickstart defines a schema for a list of personal information entries, including validation for name length, age range (with type conversion), and an optional gender field with specific allowed values. It then attempts to validate both valid and invalid data, demonstrating how `Schema.validate()` works and how `SchemaError` is raised for invalid input.

from schema import Schema, And, Use, Optional, SchemaError

schema = Schema(
    [
        {
            "name": And(str, len),
            "age": And(Use(int), lambda n: 18 <= n <= 99),
            Optional("gender"): And(str, Use(str.lower), lambda s: s in ("male", "female", "other")),
        }
    ]
)

data = [
    {"name": "Sue", "age": "28", "gender": "Female"},
    {"name": "Sam", "age": "42"},
    {"name": "Sacha", "age": "20", "gender": "Other"},
]

try:
    validated_data = schema.validate(data)
    print("Validation successful!")
    print(validated_data)
except SchemaError as e:
    print(f"Validation failed: {e}")

# Example of invalid data
invalid_data = [
    {"name": "Alice", "age": "15"} # Age too young
]

try:
    schema.validate(invalid_data)
except SchemaError as e:
    print(f"Validation failed for invalid data: {e}")

view raw JSON →