from frictionless import Resource, Step, steps, transform

# General


def test_transform():
    target = transform(
        "data/transform.csv",
        steps=[
            steps.table_normalize(),
            steps.table_melt(field_name="id"),
        ],
    )
    assert isinstance(target, Resource)
    assert target.schema.to_descriptor() == {
        "fields": [
            {"name": "id", "type": "integer"},
            {"name": "variable", "type": "string"},
            {"name": "value", "type": "any"},
        ]
    }
    assert target.read_rows() == [
        {"id": 1, "variable": "name", "value": "germany"},
        {"id": 1, "variable": "population", "value": 83},
        {"id": 2, "variable": "name", "value": "france"},
        {"id": 2, "variable": "population", "value": 66},
        {"id": 3, "variable": "name", "value": "spain"},
        {"id": 3, "variable": "population", "value": 47},
    ]


def test_transform_custom_step():
    # Create step
    class custom(Step):
        def transform_resource(self, resource: Resource):
            current = resource.to_copy()

            # Data
            def data():
                with current:
                    for row in current.row_stream:  # type: ignore
                        row["id"] = row["id"] * row["id"]
                        yield row

            # Meta
            resource.data = data

    # Transform resource
    target = transform("data/transform.csv", steps=[custom()])
    assert isinstance(target, Resource)
    assert target.schema.to_descriptor() == {
        "fields": [
            {"type": "integer", "name": "id"},
            {"type": "string", "name": "name"},
            {"type": "integer", "name": "population"},
        ]
    }
    assert target.read_rows() == [
        {"id": 1, "name": "germany", "population": 83},
        {"id": 4, "name": "france", "population": 66},
        {"id": 9, "name": "spain", "population": 47},
    ]