dlt.pipeline

pipeline

@overload
def pipeline(pipeline_name: str = None,
             pipelines_dir: str = None,
             pipeline_salt: TSecretStrValue = None,
             destination: TDestinationReferenceArg = None,
             staging: TDestinationReferenceArg = None,
             dataset_name: str = None,
             import_schema_path: str = None,
             export_schema_path: str = None,
             full_refresh: bool = None,
             dev_mode: bool = False,
             refresh: TRefreshMode = None,
             progress: TCollectorArg = _NULL_COLLECTOR,
             _impl_cls: Type[TPipeline] = Pipeline) -> TPipeline

View source on GitHub

Creates a new instance of dlt pipeline, which moves the data from the source ie. a REST API to a destination ie. database or a data lake.

Notes:

The pipeline functions allows you to pass the destination name to which the data should be loaded, the name of the dataset and several other options that govern loading of the data. The created Pipeline object lets you load the data from any source with run method or to have more granular control over the loading process with extract, normalize and load methods.

Please refer to the following doc pages

• Write your first pipeline walkthrough: https://dlthub.com/docs/walkthroughs/create-a-pipeline
• Pipeline architecture and data loading steps: https://dlthub.com/docs/reference
• List of supported destinations: https://dlthub.com/docs/dlt-ecosystem/destinations

Arguments:

• pipeline_name str, optional - A name of the pipeline that will be used to identify it in monitoring events and to restore its state and data schemas on subsequent runs. Defaults to the file name of pipeline script with dlt_ prefix added.

• pipelines_dir str, optional - A working directory in which pipeline state and temporary files will be stored. Defaults to user home directory: ~/dlt/pipelines/.

• pipeline_salt TSecretStrValue, optional - A random value used for deterministic hashing during data anonymization. Defaults to a value derived from the pipeline name. Default value should not be used for any cryptographic purposes.

• destination TDestinationReferenceArg, optional - A name of the destination to which dlt will load the data, or a destination module imported from dlt.destination. May also be provided to run method of the pipeline.

• staging TDestinationReferenceArg, optional - A name of the destination where dlt will stage the data before final loading, or a destination module imported from dlt.destination. May also be provided to run method of the pipeline.

• dataset_name str, optional - A name of the dataset to which the data will be loaded. A dataset is a logical group of tables ie. schema in relational databases or folder grouping many files. May also be provided later to the run or load methods of the Pipeline. If not provided at all then defaults to the pipeline_name

• import_schema_path str, optional - A path from which the schema yaml file will be imported on each pipeline run. Defaults to None which disables importing.

• export_schema_path str, optional - A path where the schema yaml file will be exported after every schema change. Defaults to None which disables exporting.

• full_refresh bool, optional - Deprecated use dev_mode instead.

• dev_mode bool, optional - When set to True, each instance of the pipeline with the pipeline_name starts from scratch when run and loads the data to a separate dataset. The datasets are identified by dataset_name_ + datetime suffix. Use this setting whenever you experiment with your data to be sure you start fresh on each run. Defaults to False.

• refresh TRefreshMode, optional - Fully or partially reset sources during pipeline run. When set here the refresh is applied on each run of the pipeline. To apply refresh only once you can pass it to pipeline.run or extract instead. The following refresh modes are supported:

  • drop_sources: Drop tables and source and resource state for all sources currently being processed in run or extract methods of the pipeline. (Note: schema history is erased)
  • drop_resources: Drop tables and resource state for all resources being processed. Source level state is not modified. (Note: schema history is erased)
  • drop_data: Wipe all data and resource state for all resources being processed. Schema is not modified.

• progress(TCollectorArg) - A progress monitor that shows progress bars, console or log messages with current information on sources, resources, data items etc. processed in extract, normalize and load stage. Pass a string with a collector name or configure your own by choosing from dlt.progress module. We support most of the progress libraries: try passing tqdm, enlighten or alive_progress or log to write to console/log.

• _impl_cls Type[TPipeline], optional - A class of the pipeline to use. Defaults to Pipeline.

Returns:

• TPipeline - An instance of Pipeline class or a subclass. Please check the documentation of run method for information on what to do with it.

pipeline

@overload
def pipeline() -> Pipeline

View source on GitHub

When called without any arguments, returns the recently created Pipeline instance. If not found, it creates a new instance with all the pipeline options set to defaults.

attach

@with_config(
    spec=PipelineConfiguration,
    sections=(known_sections.PIPELINES, ),
    section_arg_name="pipeline_name",
)
def attach(pipeline_name: str = None,
           pipelines_dir: str = None,
           pipeline_salt: TSecretStrValue = None,
           destination: TDestinationReferenceArg = None,
           staging: TDestinationReferenceArg = None,
           progress: TCollectorArg = _NULL_COLLECTOR,
           dataset_name: str = None,
           **injection_kwargs: Any) -> Pipeline

View source on GitHub

Attaches to the working folder of pipeline_name in pipelines_dir or in default directory. Pre-configured destination and staging factories may be provided. If not present, default factories are created from pipeline state.

If no local pipeline state is found, dlt will attempt to restore the pipeline from the provided destination and dataset.

run

def run(data: Any,
        *,
        destination: TDestinationReferenceArg = None,
        staging: TDestinationReferenceArg = None,
        dataset_name: str = None,
        table_name: str = None,
        write_disposition: TWriteDispositionConfig = None,
        columns: Sequence[TColumnSchema] = None,
        schema: Schema = None,
        loader_file_format: TLoaderFileFormat = None,
        table_format: TTableFormat = None,
        schema_contract: TSchemaContract = None,
        refresh: TRefreshMode = None) -> LoadInfo

View source on GitHub

Loads the data in data argument into the destination specified in destination and dataset specified in dataset_name.

Notes:

This method will extract the data from the data argument, infer the schema, normalize the data into a load package (ie. jsonl or PARQUET files representing tables) and then load such packages into the destination.

The data may be supplied in several forms:

• a list or Iterable of any JSON-serializable objects ie. dlt.run([1, 2, 3], table_name="numbers")
• any Iterator or a function that yield (Generator) ie. dlt.run(range(1, 10), table_name="range")
• a function or a list of functions decorated with @dlt.resource ie. dlt.run([chess_players(title="GM"), chess_games()])
• a function or a list of functions decorated with @dlt.source.

Please note that dlt deals with bytes, datetime, decimal and uuid objects so you are free to load binary data or documents containing dates.

Execution: The run method will first use sync_destination method to synchronize pipeline state and schemas with the destination. You can disable this behavior with restore_from_destination configuration option. Next it will make sure that data from the previous is fully processed. If not, run method normalizes and loads pending data items. Only then the new data from data argument is extracted, normalized and loaded.

Arguments:

• data Any - Data to be loaded to destination

• destination TDestinationReferenceArg, optional - A name of the destination to which dlt will load the data, or a destination module imported from dlt.destination. If not provided, the value passed to dlt.pipeline will be used.

• staging TDestinationReferenceArg, optional - A name of the destination where dlt will stage the data before final loading, or a destination module imported from dlt.destination. May also be provided to run method of the pipeline.

• dataset_name str, optional - A name of the dataset to which the data will be loaded. A dataset is a logical group of tables ie. schema in relational databases or folder grouping many files. If not provided, the value passed to dlt.pipeline will be used. If not provided at all then defaults to the pipeline_name

• table_name str, optional - The name of the table to which the data should be loaded within the dataset. This argument is required for a data that is a list/Iterable or Iterator without __name__ attribute. The behavior of this argument depends on the type of the data:

  • generator functions: the function name is used as table name, table_name overrides this default
  • @dlt.resource: resource contains the full table schema and that includes the table name. table_name will override this property. Use with care!
  • @dlt.source: source contains several resources each with a table schema. table_name will override all table names within the source and load the data into single table.

• write_disposition TWriteDispositionConfig, optional - Controls how to write data to a table. Accepts a shorthand string literal or configuration dictionary. Allowed shorthand string literals: append will always add new data at the end of the table. replace will replace existing data with new data. skip will prevent data from loading. "merge" will deduplicate and merge data based on "primary_key" and "merge_key" hints. Defaults to "append". Write behaviour can be further customized through a configuration dictionary. For example, to obtain an SCD2 table provide write_disposition={"disposition": "merge", "strategy": "scd2"}. Please note that in case of dlt.resource the table schema value will be overwritten and in case of dlt.source, the values in all resources will be overwritten.

• columns Sequence[TColumnSchema], optional - A list of column schemas. Typed dictionary describing column names, data types, write disposition and performance hints that gives you full control over the created table schema.

• schema Schema, optional - An explicit Schema object in which all table schemas will be grouped. By default dlt takes the schema from the source (if passed in data argument) or creates a default one itself.

• loader_file_format TLoaderFileFormat, optional - The file format the loader will use to create the load package. Not all file_formats are compatible with all destinations. Defaults to the preferred file format of the selected destination.

• table_format TTableFormat, optional - Can be "delta" or "iceberg". The table format used by the destination to store tables. Currently you can select table format on filesystem and Athena destinations.

• schema_contract TSchemaContract, optional - On override for the schema contract settings, this will replace the schema contract settings for all tables in the schema. Defaults to None.

• refresh TRefreshMode, optional - Fully or partially reset sources before loading new data in this run. The following refresh modes are supported:

  • drop_sources: Drop tables and source and resource state for all sources currently being processed in run or extract methods of the pipeline. (Note: schema history is erased)
  • drop_resources: Drop tables and resource state for all resources being processed. Source level state is not modified. (Note: schema history is erased)
  • drop_data: Wipe all data and resource state for all resources being processed. Schema is not modified.

Raises:

• PipelineStepFailed - when a problem happened during extract, normalize or load steps.

Returns:

• LoadInfo - Information on loaded data including the list of package ids and failed job statuses. Please not that dlt will not raise if a single job terminally fails. Such information is provided via LoadInfo.