Encoders

Encoders encode dict[str, Any] into a str

JsonEncoder

This is very similar encoder to the one from standard library - json.encoder.JSONEncoder. This one has support for the datetime.datetime, datetime.date and enum.Enum types
For datetime and date it uses the isoformat() method. For enum the name property is used

LogfmtEncoder

This encoder encodes into a key=value format

Encoding rules can be found in examples/encoder.py and tests/test_encoder_logfmt.py files. It also has support for datetime, date and Enum types

Basic example

from logdog.encoder import LogfmtEncoder

encoder = LogfmtEncoder()

obj = {
    "string": "string",
    "spaces": "with spaces",
    "number": 1.123,
    "list": ["a", "b", "c"],
    "dict": {
        "t": True,
        "f": False,
        "n": None,
    },
    "true": "true",
}

encoder.encode(obj)
# string=string spaces="with spaces" number=1.123 list=[a b c] dict={t=true f=false n=null} true="true"

Custom types support

To add support for encoding custom types, for example MyType class, you can extend existing encoders by overriding their default method

from typing import Any

from logdog.encoder import LogfmtEncoder

class MyLogfmtEncoder(LogfmtEncoder):
    def default(self, obj: Any) -> Any:
        if isinstance(obj, MyType):
            return f"{obj.some_prop}:{obj.other_prop}"
        return super().default(obj)

Writing your own encoder

You can also create your own encoder

from typing import Any

from logdog.encoder import BaseEncoder

class MyEncoder(BaseEncoder):
    def encode(self, obj: dict[str, Any]) -> str:
        return str(obj)