What You Can Unstructure and How#

Unstructuring is intended to convert high-level, structured Python data (like instances of complex classes) into simple, unstructured data (like dictionaries).

Unstructuring is simpler than structuring in that no target types are required. Simply provide an argument to Converter.unstructure() and cattrs will produce a result based on the registered unstructuring hooks. A number of default unstructuring hooks are documented here.

Primitive Types and Collections#

Primitive types (integers, floats, strings…) are simply passed through. Collections are copied. There’s relatively little value in unstructuring these types directly as they are already unstructured and third-party libraries tend to support them directly.

A useful use case for unstructuring collections is to create a deep copy of a complex or recursive collection.

>>> # A dictionary of strings to lists of tuples of floats.
>>> data = {'a': [[1.0, 2.0], [3.0, 4.0]]}

>>> copy = cattrs.unstructure(data)
>>> data == copy
True
>>> data is copy
False

Typed Dicts#

TypedDicts unstructure into dictionaries, potentially unchanged (depending on the exact field types and registered hooks).

>>> from typing import TypedDict
>>> from datetime import datetime, timezone
>>> from cattrs import Converter

>>> class MyTypedDict(TypedDict):
...    a: datetime

>>> c = Converter()
>>> c.register_unstructure_hook(datetime, lambda d: d.timestamp())

>>> c.unstructure({"a": datetime(1970, 1, 1, tzinfo=timezone.utc)}, unstructure_as=MyTypedDict)
{'a': 0.0}

Generic TypedDicts work on Python 3.11 and later, since that is the first Python version that supports them in general.

On Python 3.8, using typing_extensions.TypedDict is recommended since typing.TypedDict doesn’t support all necessary features, so certain combinations of subclassing, totality and typing.Required won’t work.

Similar to attrs classes, unstructuring can be customized using cattrs.gen.typeddicts.make_dict_unstructure_fn().

>>> from typing import TypedDict
>>> from cattrs import Converter
>>> from cattrs.gen import override
>>> from cattrs.gen.typeddicts import make_dict_unstructure_fn

>>> class MyTypedDict(TypedDict):
...     a: int
...     b: int

>>> c = Converter()
>>> c.register_unstructure_hook(
...     MyTypedDict,
...     make_dict_unstructure_fn(
...         MyTypedDict,
...         c,
...         a=override(omit=True)
...     )
... )

>>> c.unstructure({"a": 1, "b": 2}, unstructure_as=MyTypedDict)
{'b': 2}

See also

Structuring TypedDicts.

New in version 23.1.0.

`pathlib.Path`#

pathlib.Path objects are unstructured into their string value.

>>> from pathlib import Path

>>> cattrs.unstructure(Path("/root"))
'/root'

New in version 23.1.0.

Customizing Collection Unstructuring#

Important

This feature is supported for Python 3.9 and later.

Sometimes it’s useful to be able to override collection unstructuring in a generic way. A common example is using a JSON library that doesn’t support sets, but expects lists and tuples instead.

Using ordinary unstructuring hooks for this is unwieldy due to the semantics of singledispatch; in other words, you’d need to register hooks for all specific types of set you’re using (set[int], set[float], set[str]…), which is not useful.

Function-based hooks can be used instead, but come with their own set of challenges - they’re complicated to write efficiently.

The Converter supports easy customizations of collection unstructuring using its unstruct_collection_overrides parameter. For example, to unstructure all sets into lists, try the following:

>>> from collections.abc import Set
>>> converter = cattrs.Converter(unstruct_collection_overrides={Set: list})

>>> converter.unstructure({1, 2, 3})
[1, 2, 3]

Going even further, the Converter contains heuristics to support the following Python types, in order of decreasing generality:

Sequence, MutableSequence, list, deque, tuple
Set, frozenset, MutableSet, set
Mapping, MutableMapping, dict, defaultdict, OrderedDict, Counter

For example, if you override the unstructure type for Sequence, but not for MutableSequence, list or tuple, the override will also affect those types. An easy way to remember the rule:

all MutableSequence s are Sequence s, so the override will apply
all list s are MutableSequence s, so the override will apply
all tuple s are Sequence s, so the override will apply

If, however, you override only MutableSequence, fields annotated as Sequence will not be affected (since not all sequences are mutable sequences), and fields annotated as tuples will not be affected (since tuples are not mutable sequences in the first place).

Similar logic applies to the set and mapping hierarchies.

Make sure you’re using the types from collections.abc on Python 3.9+, and from typing on older Python versions.

`typing.Final`#

PEP 591 Final attribute types (Final[int]) are supported and unstructured appropriately.

New in version 23.1.0.

See also

Structuring Final.

`typing.Annotated`#

PEP 593 typing.Annotated[type, ...] are supported and are matched using the first type present in the annotated type.

Type Aliases#

Type aliases are supported on Python 3.12+ and are unstructured according to the rules for their underlying type. Their hooks can also be overriden using Converter.register_unstructure_hook(). (Since type aliases aren’t proper classes they cannot be used with Converter.register_unstructure_hook().)

Warning

Type aliases using typing.TypeAlias aren’t supported since there is no way at runtime to distinguish them from their underlying types.

>>> from datetime import datetime, UTC

>>> type IsoDate = datetime

>>> converter = cattrs.Converter()
>>> converter.register_unstructure_hook_func(
...     lambda t: t is IsoDate,
...     lambda v: v.isoformat()
... )

>>> converter.unstructure(datetime.now(UTC), unstructure_as=IsoDate)
'2023-11-20T23:10:46.728394+00:00'

New in version 24.1.0.

See also

Structuring Type Aliases.

`typing.NewType`#

NewTypes are supported and are unstructured according to the rules for their underlying type. Their hooks can also be overriden using Converter.register_unstructure_hook().

New in version 22.2.0.

See also

Structuring NewTypes.

Note

NewTypes are not supported by the legacy BaseConverter.

attrs Classes and Dataclasses#

attrs classes and dataclasses are supported out of the box. cattrs.Converters support two unstructuring strategies:

UnstructureStrategy.AS_DICT - similar to attrs.asdict(), unstructures attrs and dataclass instances into dictionaries. This is the default.
UnstructureStrategy.AS_TUPLE - similar to attrs.astuple(), unstructures attrs and dataclass instances into tuples.

>>> @define
... class C:
...     a = field()
...     b = field()

>>> inst = C(1, 'a')

>>> converter = cattrs.Converter(unstruct_strat=cattrs.UnstructureStrategy.AS_TUPLE)

>>> converter.unstructure(inst)
(1, 'a')

Mixing and Matching Strategies#

Converters publicly expose two helper methods, Converter.unstructure_attrs_asdict() and Converter.unstructure_attrs_astuple(). These methods can be used with custom unstructuring hooks to selectively apply one strategy to instances of particular classes.

Assume two nested attrs classes, Inner and Outer; instances of Outer contain instances of Inner. Instances of Outer should be unstructured as dictionaries, and instances of Inner as tuples. Here’s how to do this.

>>> @define
... class Inner:
...     a: int

>>> @define
... class Outer:
...     i: Inner

>>> inst = Outer(i=Inner(a=1))

>>> converter = cattrs.Converter()
>>> converter.register_unstructure_hook(Inner, converter.unstructure_attrs_astuple)

>>> converter.unstructure(inst)
{'i': (1,)}

Of course, these methods can be used directly as well, without changing the converter strategy.

>>> @define
... class C:
...     a: int
...     b: str

>>> inst = C(1, 'a')

>>> converter = cattrs.Converter()

>>> converter.unstructure_attrs_astuple(inst)  # Default is AS_DICT.
(1, 'a')

Unstructuring Hook Factories#

Hook factories operate one level higher than unstructuring hooks; unstructuring hooks are functions registered to a class or predicate, and hook factories are functions (registered via a predicate) that produce unstructuring hooks.

Unstructuring hooks factories are registered using Converter.register_unstructure_hook_factory().

Here’s a small example showing how to use factory hooks to skip unstructuring init=False attributes on all attrs classes.

>>> from attrs import define, has, field, fields
>>> from cattrs import override
>>> from cattrs.gen import make_dict_unstructure_fn

>>> c = cattrs.Converter()
>>> c.register_unstructure_hook_factory(
...     has,
...     lambda cl: make_dict_unstructure_fn(
...         cl, c, **{a.name: override(omit=True) for a in fields(cl) if not a.init}
...     )
... )

>>> @define
... class E:
...    an_int: int
...    another_int: int = field(init=False)

>>> inst = E(1)
>>> inst.another_int = 5
>>> c.unstructure(inst)
{'an_int': 1}

A complex use case for hook factories is described over at Using Factory Hooks.

What You Can Unstructure and How#

Primitive Types and Collections#

Typed Dicts#

pathlib.Path#

Customizing Collection Unstructuring#

typing.Final#

typing.Annotated#

Type Aliases#

typing.NewType#

attrs Classes and Dataclasses#

Mixing and Matching Strategies#

Unstructuring Hook Factories#

`pathlib.Path`#

`typing.Final`#

`typing.Annotated`#

`typing.NewType`#