basicserial

Contents

• Overview
• Usage
• License

Overview

Does this look familiar?

>>> import json
>>> from datetime import date
>>> MY_DATA = {'foo': 123, 'bar': date(2018, 5, 22)}
>>> json.dumps(MY_DATA)
Traceback (most recent call last):
    ...
TypeError: datetime.date(2018, 5, 22) is not JSON serializable

It's one thing when your serialization tools don't know how to handle your custom classes, but it's annoying when they don't handle the built-in and/or common data types. Thus, basicserial was born.

This package is a thin wrapper around the common serialization tools that can do the following for you when working with JSON, YAML, and TOML:

• Automatically serializes the following types to common-sense representations:

  Type	JSON	YAML	TOML
  set	array	sequence	array
  frozenset	array	sequence	array
  Decimal	number	float	float
  Fraction	string	string	string
  date	string (ISO 8601)	timestamp	string (ISO 8601)
  time	string (ISO 8601)	string (ISO 8601)	string (ISO 8601)
  datetime	string (ISO 8601)	timestamp	string (ISO 8601)
  complex	string	string	string
  OrderedDict	object	map	key/value
  defaultdict	object	map	key/value
  namedtuple	object	map	key/value
  UserDict	object	map	key/value
  UserList	array	sequence	array
  UserString	string	string	string
  UUID	string	string	string

• Can serialize Enum members appropriately based on their type.

• Can automatically deserialize dates, times, and datetimes into the native Python objects.

• Provides a simple flag for generating "pretty" strings.

Usage

To use this package, install it from PyPI (pip install basicserial). Then, make sure you install the serialization package you'd like basicserial to use:

• For YAML, it supports PyYAML and ruamel.yaml.
• For TOML, it supports toml, pytoml, qtoml, tomlkit, and tomli/tomli-w.
• For JSON, it supports Python's built-in json module, simplejson, orjson, rapidjson, ujson, hyperjson, and pysimdjson.

basicserial will automatically find a package to use, but if you want to use a specific one, you can specify its name via the pkg argument to the functions.

JSON:

>>> print(basicserial.to_json(MY_DATA))
{"foo": 123, "bar": "2018-05-22"}

>>> print(basicserial.to_json(MY_DATA, pretty=True))
{
  "foo": 123,
  "bar": "2018-05-22"
}

>>> basicserial.from_json(basicserial.to_json(MY_DATA))
{u'foo': 123, u'bar': datetime.date(2018, 5, 22)}

>>> basicserial.from_json(basicserial.to_json(MY_DATA), native_datetimes=False)
{u'foo': 123, u'bar': u'2018-05-22'}

YAML:

>>> print(basicserial.to_yaml(MY_DATA))
{bar: 2018-05-22, foo: 123}

>>> print(basicserial.to_yaml(MY_DATA, pretty=True))
bar: 2018-05-22
foo: 123

>>> basicserial.from_yaml(basicserial.to_yaml(MY_DATA))
{u'foo': 123, u'bar': datetime.date(2018, 5, 22)}

>>> basicserial.from_yaml(basicserial.to_yaml(MY_DATA), native_datetimes=False)
{'foo': 123, 'bar': u'2018-05-22'}

TOML:

>>> print(basicserial.to_toml(MY_DATA))
foo = 123
bar = "2018-05-22"

>>> print(basicserial.to_toml(MY_DATA, pretty=True))
foo = 123
bar = "2018-05-22"

>>> basicserial.from_toml(basicserial.to_toml(MY_DATA))
{u'foo': 123, u'bar': datetime.date(2018, 5, 22)}

>>> basicserial.from_toml(basicserial.to_toml(MY_DATA), native_datetimes=False)
{u'foo': 123, u'bar': u'2018-05-22'}

License

This project is released under the terms of the MIT License.