Write to Delta

SparkDeltaDestination

Bases: DestinationInterface

The Spark Delta Destination is used to write data to a Delta table.

Examples

#Delta Destination for Streaming Queries

from rtdip_sdk.pipelines.destinations import SparkDeltaDestination

delta_destination = SparkDeltaDestination(
    data=df,
    options={
        "checkpointLocation": "/{CHECKPOINT-LOCATION}/"
    },
    destination="DELTA-TABLE-PATH",
    mode="append",
    trigger="10 seconds",
    query_name="DeltaDestination",
    query_wait_interval=None
)

delta_destination.write_stream()

#Delta Destination for Batch Queries

from rtdip_sdk.pipelines.destinations import SparkDeltaDestination

delta_destination = SparkDeltaDestination(
    data=df,
    options={
        "overwriteSchema": True
    },
    destination="DELTA-TABLE-PATH",
    mode="append",
    trigger="10 seconds",
    query_name="DeltaDestination",
    query_wait_interval=None
)

delta_destination.write_batch()

Parameters:

Name	Type	Description	Default
data	DataFrame	Dataframe to be written to Delta	required
options	dict	Options that can be specified for a Delta Table write operation (See Attributes table below). Further information on the options is available for batch and streaming.	required
destination	str	Either the name of the Hive Metastore or Unity Catalog Delta Table or the path to the Delta table	required
mode	optional str	Method of writing to Delta Table - append/overwrite (batch), append/update/complete (stream). Default is append	'append'
trigger	optional str	Frequency of the write operation. Specify "availableNow" to execute a trigger once, otherwise specify a time period such as "30 seconds", "5 minutes". Set to "0 seconds" if you do not want to use a trigger. (stream) Default is 10 seconds	'10 seconds'
query_name	optional str	Unique name for the query in associated SparkSession. (stream) Default is DeltaDestination	'DeltaDestination'
query_wait_interval	optional int	If set, waits for the streaming query to complete before returning. (stream) Default is None	None

Attributes:

Name	Type	Description
checkpointLocation	str	Path to checkpoint files. (Streaming)
txnAppId	str	A unique string that you can pass on each DataFrame write. (Batch & Streaming)
txnVersion	str	A monotonically increasing number that acts as transaction version. (Batch & Streaming)
maxRecordsPerFile	int str	Specify the maximum number of records to write to a single file for a Delta Lake table. (Batch)
replaceWhere	str	Condition(s) for overwriting. (Batch)
partitionOverwriteMode	str	When set to dynamic, overwrites all existing data in each logical partition for which the write will commit new data. Default is static. (Batch)
overwriteSchema	bool str	If True, overwrites the schema as well as the table data. (Batch)

Source code in src/sdk/python/rtdip_sdk/pipelines/destinations/spark/delta.py

class SparkDeltaDestination(DestinationInterface):
    """
    The Spark Delta Destination is used to write data to a Delta table.

    Examples
    --------
    ```python
    #Delta Destination for Streaming Queries

    from rtdip_sdk.pipelines.destinations import SparkDeltaDestination

    delta_destination = SparkDeltaDestination(
        data=df,
        options={
            "checkpointLocation": "/{CHECKPOINT-LOCATION}/"
        },
        destination="DELTA-TABLE-PATH",
        mode="append",
        trigger="10 seconds",
        query_name="DeltaDestination",
        query_wait_interval=None
    )

    delta_destination.write_stream()
    ```
    ```python
    #Delta Destination for Batch Queries

    from rtdip_sdk.pipelines.destinations import SparkDeltaDestination

    delta_destination = SparkDeltaDestination(
        data=df,
        options={
            "overwriteSchema": True
        },
        destination="DELTA-TABLE-PATH",
        mode="append",
        trigger="10 seconds",
        query_name="DeltaDestination",
        query_wait_interval=None
    )

    delta_destination.write_batch()
    ```

    Parameters:
        data (DataFrame): Dataframe to be written to Delta
        options (dict): Options that can be specified for a Delta Table write operation (See Attributes table below). Further information on the options is available for [batch](https://docs.delta.io/latest/delta-batch.html#write-to-a-table){ target="_blank" } and [streaming](https://docs.delta.io/latest/delta-streaming.html#delta-table-as-a-sink){ target="_blank" }.
        destination (str): Either the name of the Hive Metastore or Unity Catalog Delta Table **or** the path to the Delta table
        mode (optional str): Method of writing to Delta Table - append/overwrite (batch), append/update/complete (stream). Default is append
        trigger (optional str): Frequency of the write operation. Specify "availableNow" to execute a trigger once, otherwise specify a time period such as "30 seconds", "5 minutes". Set to "0 seconds" if you do not want to use a trigger. (stream) Default is 10 seconds
        query_name (optional str): Unique name for the query in associated SparkSession. (stream) Default is DeltaDestination
        query_wait_interval (optional int): If set, waits for the streaming query to complete before returning. (stream) Default is None

    Attributes:
        checkpointLocation (str): Path to checkpoint files. (Streaming)
        txnAppId (str): A unique string that you can pass on each DataFrame write. (Batch & Streaming)
        txnVersion (str): A monotonically increasing number that acts as transaction version. (Batch & Streaming)
        maxRecordsPerFile (int str): Specify the maximum number of records to write to a single file for a Delta Lake table. (Batch)
        replaceWhere (str): Condition(s) for overwriting. (Batch)
        partitionOverwriteMode (str): When set to dynamic, overwrites all existing data in each logical partition for which the write will commit new data. Default is static. (Batch)
        overwriteSchema (bool str): If True, overwrites the schema as well as the table data. (Batch)
    """

    data: DataFrame
    options: dict
    destination: str
    mode: str
    trigger: str
    query_name: str
    query_wait_interval: int

    def __init__(
        self,
        data: DataFrame,
        options: dict,
        destination: str,
        mode: str = "append",
        trigger: str = "10 seconds",
        query_name: str = "DeltaDestination",
        query_wait_interval: int = None,
    ) -> None:
        self.data = data
        self.options = options
        self.destination = destination
        self.mode = mode
        self.trigger = trigger
        self.query_name = query_name
        self.query_wait_interval = query_wait_interval

    @staticmethod
    def system_type():
        """
        Attributes:
            SystemType (Environment): Requires PYSPARK
        """
        return SystemType.PYSPARK

    @staticmethod
    def libraries():
        libraries = Libraries()
        libraries.add_maven_library(get_default_package("spark_delta_core"))
        return libraries

    @staticmethod
    def settings() -> dict:
        return {
            "spark.sql.extensions": "io.delta.sql.DeltaSparkSessionExtension",
            "spark.sql.catalog.spark_catalog": "org.apache.spark.sql.delta.catalog.DeltaCatalog",
        }

    def pre_write_validation(self):
        return True

    def post_write_validation(self):
        return True

    def write_batch(self):
        """
        Writes batch data to Delta. Most of the options provided by the Apache Spark DataFrame write API are supported for performing batch writes on tables.
        """
        try:
            if "/" in self.destination:
                return (
                    self.data.write.format("delta")
                    .mode(self.mode)
                    .options(**self.options)
                    .save(self.destination)
                )
            else:
                return (
                    self.data.write.format("delta")
                    .mode(self.mode)
                    .options(**self.options)
                    .saveAsTable(self.destination)
                )

        except Py4JJavaError as e:
            logging.exception(e.errmsg)
            raise e
        except Exception as e:
            logging.exception(str(e))
            raise e

    def write_stream(self):
        """
        Writes streaming data to Delta. Exactly-once processing is guaranteed
        """
        TRIGGER_OPTION = (
            {"availableNow": True}
            if self.trigger == "availableNow"
            else {"processingTime": self.trigger}
        )
        try:
            if "/" in self.destination:
                query = (
                    self.data.writeStream.trigger(**TRIGGER_OPTION)
                    .format("delta")
                    .queryName(self.query_name)
                    .outputMode(self.mode)
                    .options(**self.options)
                    .start(self.destination)
                )
            else:
                query = (
                    self.data.writeStream.trigger(**TRIGGER_OPTION)
                    .format("delta")
                    .queryName(self.query_name)
                    .outputMode(self.mode)
                    .options(**self.options)
                    .toTable(self.destination)
                )

            if self.query_wait_interval:
                while query.isActive:
                    if query.lastProgress:
                        logging.info(query.lastProgress)
                    time.sleep(self.query_wait_interval)

        except Py4JJavaError as e:
            logging.exception(e.errmsg)
            raise e
        except Exception as e:
            logging.exception(str(e))
            raise e

system_type() staticmethod

Attributes:

Name	Type	Description
SystemType	Environment	Requires PYSPARK

Source code in src/sdk/python/rtdip_sdk/pipelines/destinations/spark/delta.py

@staticmethod
def system_type():
    """
    Attributes:
        SystemType (Environment): Requires PYSPARK
    """
    return SystemType.PYSPARK

write_batch()

Writes batch data to Delta. Most of the options provided by the Apache Spark DataFrame write API are supported for performing batch writes on tables.

Source code in src/sdk/python/rtdip_sdk/pipelines/destinations/spark/delta.py

def write_batch(self):
    """
    Writes batch data to Delta. Most of the options provided by the Apache Spark DataFrame write API are supported for performing batch writes on tables.
    """
    try:
        if "/" in self.destination:
            return (
                self.data.write.format("delta")
                .mode(self.mode)
                .options(**self.options)
                .save(self.destination)
            )
        else:
            return (
                self.data.write.format("delta")
                .mode(self.mode)
                .options(**self.options)
                .saveAsTable(self.destination)
            )

    except Py4JJavaError as e:
        logging.exception(e.errmsg)
        raise e
    except Exception as e:
        logging.exception(str(e))
        raise e

write_stream()

Writes streaming data to Delta. Exactly-once processing is guaranteed

Source code in src/sdk/python/rtdip_sdk/pipelines/destinations/spark/delta.py

def write_stream(self):
    """
    Writes streaming data to Delta. Exactly-once processing is guaranteed
    """
    TRIGGER_OPTION = (
        {"availableNow": True}
        if self.trigger == "availableNow"
        else {"processingTime": self.trigger}
    )
    try:
        if "/" in self.destination:
            query = (
                self.data.writeStream.trigger(**TRIGGER_OPTION)
                .format("delta")
                .queryName(self.query_name)
                .outputMode(self.mode)
                .options(**self.options)
                .start(self.destination)
            )
        else:
            query = (
                self.data.writeStream.trigger(**TRIGGER_OPTION)
                .format("delta")
                .queryName(self.query_name)
                .outputMode(self.mode)
                .options(**self.options)
                .toTable(self.destination)
            )

        if self.query_wait_interval:
            while query.isActive:
                if query.lastProgress:
                    logging.info(query.lastProgress)
                time.sleep(self.query_wait_interval)

    except Py4JJavaError as e:
        logging.exception(e.errmsg)
        raise e
    except Exception as e:
        logging.exception(str(e))
        raise e