Installation

$ pip install cerberedis

Example

A basic example:

from cerberus import Validator, TypeDefinition
from redis import Redis
from cerberedis import CerbeRedis

schema = {
    'name': {'type': 'string', 'required': True},
    'email': {'type': 'string', 'required': True},
}
data = {'name': 'Bob', 'email': 'bob@example.com'}

# connect to our redis server
r = Redis()
db = CerbeRedis(r)

# save the schema under the type name 'User'
# with a document id of 1
model_name, id = 'User', 1
db.save(model_name, schema, id, document)

# print the key the model is stored under

print(db.key(model_name, id))

# reload the document
loaded_document = db.load(model_name, schema, id)

A full featured example, including custom types:

from ipaddress import ip_address, IPv4Address, IPv6Address
from datetime import date, datetime
from cerberus import Validator, TypeDefinition
from redis import Redis
# alternatively use redis_mock (https://github.com/adamlwgriffiths/redis-mock)
# from redis_mock import Redis
from cerberedis import CerbeRedis

# Add custom types to both Cerberus
# https://docs.python-cerberus.org/en/stable/customize.html
Validator.types_mapping.update({
    'ipaddress': TypeDefinition('ipaddress', (IPv4Address, IPv6Address), ()),
    'ipv4address': TypeDefinition('ipv4address', (IPv4Address,), ()),
    'ipv6address': TypeDefinition('ipv6address', (IPv6Address,), ()),
})
# Provide CerbeRedis with information on how to handle these new Cerberus types
CerbeRedis.rules.update({
    # dictionary of: <cerberus type name>: [to redis function, from redis function]
    'ipaddress': [lambda x: str(x), lambda x: ip_address(x.decode('utf-8'))],
    'ipv4address': [lambda x: str(x), lambda x: IPv4Address(x.decode('utf-8'))],
    'ipv6address': [lambda x: str(x), lambda x: IPv6Address(x.decode('utf-8'))],
})

# An example Cerberus schema that uses all the built-in Cerberus types
# and the custom ones we defined above
schema = {
    'boolean': {'type': 'boolean', 'required': True},
    'binary': {'type': 'binary', 'required': True},
    'date': {'type': 'date', 'required': True},
    'datetime': {'type': 'datetime', 'required': True},
    'float': {'type': 'float', 'required': True},
    'integer': {'type': 'integer', 'required': True},
    'number': {'type': 'number', 'required': True},
    'string': {'type': 'string', 'required': True},
    'dict': {'type': 'dict', 'schema': {
        'dict_a': {'type': 'string', 'required': True},
        'dict_b': {'type': 'integer', 'required': True},
    }},
    'list': {'type': 'list', 'schema': {'type': 'integer'}},
    'set': {'type': 'set', 'schema': {'type': 'string'}},
    'ipv4address': {'type': 'ipv4address', 'required': True},
}

data = {
    'boolean': True,
    'binary': b'123',
    'date': date.today(),
    'datetime': datetime.now(),
    'float': 1.23,
    'integer': 456,
    'number': 789.0,
    'string': 'abcdefg',
    'dict': {
        'dict_a': 'dict_a_value',
        'dict_b': 9999,
    },
    'list': [1,2,3,4,5],
    'set': {'a', 'b', 'c'},
    'ipv4address': IPv4Address('127.0.0.1'),
}

# Use cerberus to validate and normalise the data
validator = Validator(schema)
document = validator.normalized(data)
if not document:
    raise ValueError(str(validator.errors))

# connect to our redis server
r = Redis()
db = CerbeRedis(r)
model_name, id = 'TestModel', 1
db.save(model_name, schema, id, document)

# reload the document
loaded_document = db.load(model_name, schema, id)

# verify the data is laid out how we expect
assert document == loaded_document

Supported Types

All existing Cerberus Types:

• boolean

• binary

• date

• datetime

• float

• integer

• number

• string

• dict

• list

• set

Limitations

• Containers cannot be nested. Ie. lists and sets cannot contain lists, sets, or dicts.

Future Work

• Use a Redis Pipeline for the load function

1.0.5

• Return None rather than empty containers. Permits us to know if the value actually exists.

1.0.4

• Fix image link

1.0.3

• Fix handling of None values for non-required fields

1.0.2

• Expose key function

1.0.1

• Version bump due to PyPi version pedantry

1.0.0

• Initial release