How to Serialize Data¶

Create models from dicts¶

from_dict constructs model instances from raw dictionaries — useful for testing and data import.

from cendry import from_dict

city = from_dict(City, {
    "name": "San Francisco",
    "state": "CA",
    "country": "USA",
    "capital": False,
    "population": 870_000,
    "regions": ["west_coast"],
})

With a document ID¶

city = from_dict(City, {...}, doc_id="SF")
print(city.id)  # "SF"

Nested Maps¶

Nested Map dicts are automatically converted:

city = from_dict(City, {
    ...,
    "mayor": {"name": "London Breed", "since": 2018},
})
assert isinstance(city.mayor, Mayor)

Missing fields¶

Raises TypeError with a clear message listing all missing fields:

from_dict(City, {"name": "SF"})
# TypeError: from_dict(City): missing required fields: state, country, ...

Convert models to dicts¶

from cendry import to_dict

data = to_dict(city)
# {"name": "San Francisco", "state": "CA", ...}

Options¶

Parameter	Default	Description
`by_alias`	`False`	Use Firestore alias keys instead of Python names
`include_id`	`False`	Include the document ID in the output

to_dict(city, include_id=True)
# {"name": "San Francisco", ..., "id": "SF"}

to_dict(city, by_alias=True)
# {"displayName": "San Francisco", ...}

Testing without Firestore¶

from_dict is the primary tool for testing — no mocks needed for model construction:

def test_city_population():
    city = from_dict(City, {
        "name": "SF",
        "state": "CA",
        "country": "USA",
        "capital": False,
        "population": 870_000,
        "regions": [],
    })
    assert city.population == 870_000

Custom types in tests

If your model uses a registered custom type (like Money), from_dict runs the handler's deserialize automatically — your tests get the same conversion as production.