Skip to main content

A simple binary serialization/deserialization library for Python

Project description

(Py)Bran

(Py)Bran - Boring, plain and simple python binary serialization/deserialization library

Its called (Py)Bran because Bran was taken. It does (cereal)ization. Yes. Also because Bran (the ceareal) is boring like this library.

MIT license PyPI Version PyPI Versions PyPI Status

CI Status security: bandit PyLint Score Coverage %

Installing

pip install pybran

Using

Registering your Schemas

In order for Bran to work with complex objects, it needs to know the object Schema. Schemas can be defined in two ways:

  • Automatically with Decorators
  • Manually

Automatically with Decorators

from pybran import schema, field

@schema
class MyClass:
    test = field(1)
    other = field(MyClass2())

@schema
class MyClass2:
    test2 = field(2)

Manually

from pybran import register_class

class MyClass:
    test = 1
    other = MyClass2()

class MyClass2:
    test2 = 2
    
register_class(MyClass2, {"test2": MyClass2.test2})
register_class(MyClass, {"test": MyClass.test, "other": MyClass.other})

Registering field aliases

Bran generates an enumeration for each field so it knows which field it's deserializing when doing so. Writing an enumeration is more efficient than writing the field name itself.

If you want to override the enumeration bran generates, you can do so by specifying the alias parameter when registering a field. You can also manually register aliases if need be. Registering with annotations is done like so:

@schema
class MyClass:
    test = field(1, alias=b'\x05')

Manual registration can be done like so:

from pybran import register_class

class MyClass:
    test = 1
    
register_class(MyClass, {"test2": MyClass.test}, {"test2": b'\x05'})

Aliases and field information can be fetched from the ClassDefinition definition stored in the class_registry

from pybran import class_registry, ClassDefinition

@schema
class MyClass:
    test = field(1)

class_definition: ClassDefinition = class_registry.get(MyClass)

The ClassDefinition object contains two internal Registry objects:

  • fields: Tracks the class fields and their types
  • aliases: Aliases registered for the class fields
from pybran import class_registry, ClassDefinition

@schema
class MyClass:
    test = field(1)
    test2 = field(1, alias=b'\x05')

class_definition: ClassDefinition = class_registry.get(MyClass)

class_definition.fields.get("test") # Returns type(int)
class_definition.fields.get("test2") # Returns type(int)

class_definition.aliases.get("test") # Returns 1 (first registered enumeration)
class_definition.aliases.get("test2") # Returns b'\x05'

class_definition.aliases.get(1) # Returns "test"
class_definition.aliases.get(b'\x05') # Returns "test2"

Serializing

Registering Serializers

Bespoke Serializers can be registered with the Loader instance using the Loader.register method.

from pybran.loaders import Loader
from pybran.serializers import DefaultSerializer

loader = Loader()

# Register the type MyClass to use the serializer DefaultSerializer 
loader.register(MyClass, DefaultSerializer)

Bran automatically comes configured with a serializer registry that maps the following types:

from pybran.serializers import *

{
    bool: BoolSerializer,
    int: IntSerializer,
    float: FloatSerializer,
    str: StringSerializer,
    set: SetSerializer,
    dict: MappingSerializer,
    list: ArraySerializer,
    tuple: ArraySerializer
}

When importing the Loader, these are already preconfigured. If you wish to specify your own serializer registry to use upon instantiation, then you can override it when creating a new Loader object.

from pybran.loaders import Loader
from pybran.exceptions import BranSerializerException

serializers = {
    int: MyCustomSerializer
}

loader = Loader(serializers)

# Using MyCustomSerializer
try:
    serialized = loader.serialize(1)
except BranSerializerException as e:
    print(e)

Serializing Directly

Serializing an object is as simple as calling the Loader.serialize method

Any serialization errors will raise a BranSerializerException

from pybran.loaders import Loader
from pybran.exceptions import BranSerializerException

loader = Loader()

try:
    serialized = loader.serialize(1)
except BranSerializerException as e:
    print(e)

Deserializing Directly

Deserializing is just as simple

Any deserialization errors will raise a BranSerializerException

import io

from pybran.loaders import Loader
from pybran.exceptions import BranSerializerException

loader = Loader()

serialized = b'\x00\x00\x00\x00'
try:
    myint = loader.deserialize(io.BytesIO(serialized), int)
except BranSerializerException as e:
    print(e)

Serializing with Automatic Type Tagging

When serializing/deserializing a type, type information can be configured to be stored so the deserializer automatically knows which object to deserialize to.

This can be done by passing the tagging = True kwarg when calling serialize/deserialize

import io

from pybran.serializers import DefaultSerializer
from pybran.loaders import Loader

loader = Loader()
loader.register(MyClass, DefaultSerializer)

myobj = MyClass()

serialized = loader.serialize(myobj, tagging=True)
myobj_deserialized = loader.deserialize(io.BytesIO(serialized), tagging=True)

Reading from a file

The Loader is also capable of reading and writing to a file

Can raise a BranFileException

from pybran.loaders import Loader
from pybran.serializers import DefaultSerializer
from pybran.exceptions import BranSerializerException, BranFileException

loader = Loader()
loader.register(MyClass, DefaultSerializer)

try:
    myobject = loader.read("path/my_file", MyClass)
except (BranSerializerException, BranFileException) as e:
    print(e)

Writing to a file

Can raise a BranFileException

from pybran.loaders import Loader
from pybran.serializers import DefaultSerializer
from pybran.exceptions import BranSerializerException, BranFileException

loader = Loader()
loader.register(MyClass, DefaultSerializer)

myobject = MyClass()

try:
    loader.write("path/my_file", myobject)
except (BranSerializerException, BranFileException) as e:
    print(e)

Writing a Custom Serializer

Writing a custom serializer can be done by extending the Serializer class, and registering the class with your Loader instance.

from pybran.serializers import Serializer
from pybran.loaders import Loader

class MyCustomSerializer(Serializer):
    def serialize(self, loader, obj, **kwargs):
        # Implement custom serialization logic here

    def deserialize(self, loader, cls, data, **kwargs):
        # Implement custom deserialization logic here

loader = Loader()
loader.register(MyClass, MyCustomSerializer)

Overriding automatic Type ID generation

PyBran uses a cache to enumerate types with an identification token. This cache takes the form of a symmetric dict. This allows PyBran, when its necessary to store type information, to write the enumeration/ID of the Type, rather than the entire type information.

By default, Type IDs are generated with an Atomic counter. When PyBran registers a class, it automatically registers an enumeration for the class itself and any unseen types contained within.

pybran/decorators.py

...

type_registry = Registry(lambda k: TypeId().get_id())

...

This can be overridden by changing the default_value_generator of the type registry, like so:

import pybran

bespoke_map = {
    int: b'I',
    float: b'L'
}

def generate_key(k: type):
    return bespoke_map.get(type, pybran.TypeId().get_id())

pybran.type_registry.default_value_generator = generate_key
pybran.refresh()

Refresh MUST be called to refresh the registry caches if you update the default value generator functions. Otherwise the registry will contain both old and new autogenerated values.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pybran-0.1.0.tar.gz (14.1 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

pybran-0.1.0-py3-none-any.whl (12.4 kB view details)

Uploaded Python 3

File details

Details for the file pybran-0.1.0.tar.gz.

File metadata

  • Download URL: pybran-0.1.0.tar.gz
  • Upload date:
  • Size: 14.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.2 importlib_metadata/4.8.1 pkginfo/1.7.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.2 CPython/3.9.5

File hashes

Hashes for pybran-0.1.0.tar.gz
Algorithm Hash digest
SHA256 9773926fb71194a0e4334b76fabd454a28ff571362897537022742a497dfc331
MD5 8b84ac7ee0c81d1a4f4bcabc1f5e097c
BLAKE2b-256 3b7cb46b9dcf884198dd647c51d8c91039f236730efeef5027a2557d4579d648

See more details on using hashes here.

File details

Details for the file pybran-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: pybran-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 12.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.2 importlib_metadata/4.8.1 pkginfo/1.7.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.2 CPython/3.9.5

File hashes

Hashes for pybran-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 32e26918f2fd7945999056cf4f8c7f3c3aacac9f171028983b855b6d42299689
MD5 c8f1f9ebe77788d0e58f9f80f8faf706
BLAKE2b-256 3c534a8e00aa7291857961e24ac310371a121b05a0e72249f5daaee69ff185a0

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page