dlt.destinations.impl.bigquery.bigquery_adapter

PartitionTransformation Objects

class PartitionTransformation()

View source on GitHub

template

Template string of the transformation including column name placeholder. E.g. RANGE_BUCKET({column_name}, GENERATE_ARRAY(0, 1000000, 10000))

column_name

Column name to apply the transformation to

bigquery_partition Objects

class bigquery_partition()

View source on GitHub

Helper class to generate BigQuery partition transformations.

range_bucket

@staticmethod
def range_bucket(column_name: str,
                 start: int,
                 end: int,
                 interval: int = 1) -> PartitionTransformation

View source on GitHub

Partition by an integer column with the specified range, where:

Arguments:

• column_name - The column to partition by
• start - The start of range partitioning (inclusive)
• end - The end of the range partitioning (exclusive)
• interval - The width of each range within the partition (default: 1)

Returns:

A PartitionTransformation object for integer range partitioning

bigquery_adapter

def bigquery_adapter(
        data: Any,
        partition: Union[TColumnNames, PartitionTransformation] = None,
        cluster: TColumnNames = None,
        round_half_away_from_zero: TColumnNames = None,
        round_half_even: TColumnNames = None,
        table_description: Optional[str] = None,
        table_expiration_datetime: Optional[str] = None,
        insert_api: Optional[Literal["streaming", "default"]] = None,
        autodetect_schema: Optional[bool] = None,
        partition_expiration_days: Optional[int] = None) -> DltResource

View source on GitHub

Prepares data for loading into BigQuery.

This function takes data, which can be raw or already wrapped in a DltResource object, and prepares it for BigQuery by optionally specifying partitioning, clustering, table description and table expiration settings.

Arguments:

• data Any - The data to be transformed. This can be raw data or an instance of DltResource. If raw data is provided, the function will wrap it into a DltResource object.
• partition Union[TColumnNames, PartitionTransformation], optional - The column to partition the BigQuery table by. This can be a string representing a single column name for simple partitioning, or a PartitionTransformation object for advanced partitioning. Use the bigquery_partition helper class to create transformation objects.
• cluster TColumnNames, optional - A column name or list of column names to cluster the BigQuery table by.
• round_half_away_from_zero TColumnNames, optional - Determines how values in the column are rounded when written to the table. This mode rounds halfway cases away from zero. The columns specified must be mutually exclusive from round_half_even. See https://cloud.google.com/bigquery/docs/schemas#rounding_mode for more information.
• round_half_even TColumnNames, optional - Determines how values in the column are rounded when written to the table. This mode rounds halfway cases towards the nearest even digit. The columns specified must be mutually exclusive from round_half_away_from_zero. See https://cloud.google.com/bigquery/docs/schemas#rounding_mode for more information.
• table_description str, optional - A description for the BigQuery table.
• table_expiration_datetime str, optional - String representing the datetime when the BigQuery table expires. This is always interpreted as UTC, BigQuery's default.
• insert_api Optional[Literal["streaming", "default"]] - The API to use for inserting data into BigQuery. If "default" is chosen, the original SQL query mechanism is used. If "streaming" is chosen, the streaming API (https://cloud.google.com/bigquery/docs/streaming-data-into-bigquery) is used.
• NOTE - due to BigQuery features, streaming insert is only available for append write_disposition.
• autodetect_schema bool, optional - If set to True, BigQuery schema autodetection will be used to create data tables. This allows to create structured types from nested data.
• partition_expiration_days int, optional - For date/time based partitions it tells when partition is expired and removed. Partitions are expired based on a partitioned column value. (https://cloud.google.com/bigquery/docs/managing-partitioned-tables#partition-expiration)

Returns:

A DltResource object that is ready to be loaded into BigQuery.

Raises:

• ValueError - If any hint is invalid or none are specified.

Examples:

    data = [{"name": "Marcel", "description": "Raccoon Engineer", "date_hired": 1700784000}]
    bigquery_adapter(data, partition="date_hired", table_expiration_datetime="2024-01-30", table_description="Employee Data")

[DltResource with hints applied]

should_autodetect_schema

def should_autodetect_schema(table: PreparedTableSchema) -> bool

View source on GitHub

Tells if schema should be auto detected for a given prepared table