Function column_definition

python/pathway/internals/schema.py:756–827

Creates column definition Args: primary_key: should column be a part of a primary key. default_value: default value replacing blank entries. The default value of the column must be specified explicitly, otherwise there will be no default value. dt

(
    *,
    primary_key: bool = False,
    default_value: Any | None = _no_default_value_marker,
    dtype: Any | None = None,
    name: str | None = None,
    append_only: bool | None = None,
    description: str | None = None,
    example: Any = None,
    source_component: Literal["key", "payload"] = "payload",
    _serialized_default_value: Any | None = None,
    _serialized_example: Any | None = None,
)

def column_definition(
    *,
    primary_key: bool = False,
    default_value: Any | None = _no_default_value_marker,
    dtype: Any | None = None,
    name: str | None = None,
    append_only: bool | None = None,
    description: str | None = None,
    example: Any = None,
    source_component: Literal["key", "payload"] = "payload",
    _serialized_default_value: Any | None = None,
    _serialized_example: Any | None = None,
) -> Any:  # Return any so that mypy does not complain
    """Creates column definition

    Args:
        primary_key: should column be a part of a primary key.
        default_value: default value replacing blank entries. The default value of the
            column must be specified explicitly,
            otherwise there will be no default value.
        dtype: data type. When used in schema class,
            will be deduced from the type annotation.
        name: name of a column. When used in schema class,
            will be deduced from the attribute name.
        append_only: whether column is append-only. if unspecified, defaults to False
            or to value specified at the schema definition level
        description: human-readable description for the column. Used by HTTP input
            connector in automated OpenAPI schema generation.
        example: example of the column value. Used by HTTP input connector in automated
            OpenAPI schema generation.
        source_component: the part of the input message from which the value should be
            parsed when JSON format is used. In Kafka, this can be used to specify that
            certain fields must be parsed from the message key.

    Returns:
        Column definition.

    Example:

    >>> import pathway as pw
    >>> class NewSchema(pw.Schema):
    ...   key: int = pw.column_definition(primary_key=True)
    ...   timestamp: str = pw.column_definition(name="@timestamp")
    ...   data: str
    >>> NewSchema
    <pathway.Schema types={'key': <class 'int'>, '@timestamp': <class 'str'>, 'data': <class 'str'>}, \
id_type=pathway.engine.Pointer[int]>
    """

    if _serialized_default_value is not None:
        if not isinstance(default_value, _Undefined):
            raise ValueError(
                "Maximum one of {'default_value', '_serialized_default_value'} must be specified"
            )
        default_value = api.deserialize(base64.b64decode(_serialized_default_value))
    if _serialized_example is not None:
        if example is not None:
            raise ValueError(