Skip to main content
Join the official 2019 Python Developers SurveyStart the survey!

Pythonic, yet C-style structs.

Project description


Pythonic, yet C-Style structs/unions.

Structs are similar to namedtuples, but they allow type assertion and are defined the same way as any class. Creating a struct is easy.

$ pip install struction

from struction import Struct, default, between, clamp

class MyStruct(Struct):
    # any types specified *must* be a class/type
    field_0 = int
    field_1 = str

    # you can also allow multiple types
    multi_type = int, str

    # and default values!
    with_default = int, default(10)

    with_multi_and_default = int, str, default(10)

    # it's also possible to set a specific range for fields
    with_range = int, float, between(1, 10)  # setting this to a value < 1 or > 10 will raise ValueError

    # or, values can be clamped to a range
    with_clamp = int, float, clamp(-5, 5)  # setting this to any value outside of range will clamp it

Once a struct is created, it’s fields can be changed, but they must match the given type or a TypeError will be raised. Using del on a field resets it to its default value.

It’s also possible to nest structs. Any structs that are nested will automatically be initialized.

class NestMe(Struct):
    field_0 = int
    field_1 = int

class Nester(Struct):
    abc = str
    nest = OnlyInt  # nested struct
>>> print(Nester())
# Nester {
#     abc = None
#     nest = OnlyInt {
#         field_0 = None
#         field_1 = None
#     }
# }

>>> print(Nester(nest=None))
# Nester {
#     abc = None
#     nest = None
# }

If you don’t want strict types, you can also use a TypecastingStruct. This will attempt to typecast the given value to the field’s type. If it can’t be typecasted, it will then raise a TypeError.

from struction import TypecastingStruct

class Test(TypecastingStruct):
    i = int
    f = float
    s = str
    all = float, int, str
>>> test = Test()
>>> test.i = 5.3
>>> test.f = 100
>>> test.s = {"a": 1, "b":2}
>>> print(test)
# Struct Test {
#     all = None
#     f = 100.0
#     i = 5
#     s = "{'a': 1, 'b': 2}"
# }
>>> # If multiple types are allowed for a field, the value will be
>>> # casted to the first type that doesn't throw an Exception
>>> test.all = ("a", 1, "b", 2, "c", 3)
>>> test.all
# '("a", 1, "b", 2, "c", 3)'

Note: Typecasting only works at runtime. The values still need to match their types at class definition.


These can be applied to any Struct class.

  • Struct.dict() : dict with struct’s fields. {name: value, …}
  • Struct.fields() : list of fields struct has.
  • str(Struct) : Multi-line representation of struct.
  • repr(Struct) : Single line representation of struct.

Project details

Download files

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

Files for Struction, version 0.3.0
Filename, size File type Python version Upload date Hashes
Filename, size Struction-0.3.0.tar.gz (5.4 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page