tri.named-struct 0.12.0

tri.named-struct supplies a class that can be used like dictionaries (or via attribute access), but with a predefined set of possible key values

tri.named_struct

tri.named_struct supplies classes that can be used like dictionaries, but with a predefined set of possible key values.

Example

from tri.named_struct import NamedStruct

class MyNamedStruct(NamedStruct):
    foo = NamedStructField()
    bar = NamedStructField()

m = MyNamedStruct(17, 42)
assert m['foo'] == 17
assert m.foo == 17
assert m == dict(foo=17, bar=42)

m.not_foo  # Will raise an AttributeError

Default values can be provided:

from tri.named_struct import NamedStruct

class MyNamedStruct(NamedStruct):
    foo = NamedStructField()
    bar = NamedStructField()
    baz = NamedStructField(default='default')

assert MyNamedStruct(17) == dict(foo=17, bar=None, baz='default')

Default values can alternatively be provided by a factory method:

from tri.named_struct import NamedStruct

class MyNamedStruct(NamedStruct):
    foo = NamedStructField(default_factory=list)

assert MyNamedStruct().foo == []

There is also a functional way to defined a NamedStruct subclass:

from tri.named_struct import named_struct

MyNamedStruct = named_struct('foo, bar')
m = MyNamedStruct(17, 42)
assert m.foo == 17
assert m.bar == 42

Running tests

You need tox installed then just make test.

License

BSD

Documentation

http://trinamedstruct.readthedocs.org.

Changelog

0.12.0 (2016-08-30)

• Fix class level override of named struct fields.

0.11.0 (2016-05-06)

• Fix deafult values shadowing values set explicitly in subclass __init__ method.

0.10.0 (2016-01-12)

• Added default_factory argument to specify default values via a callback.