yggdrasil.databricks.cli.services.table¶
table ¶
ygg databricks table — the on-cluster table data-plane CLI.
This is the dedicated subcommand a deployed serverless Auto Loader job runs as
on the cluster — the single python-wheel task built by
:meth:yggdrasil.databricks.table.table.Table.auto_loader invokes the ygg
entry point with::
ygg databricks table autoload --table <catalog.schema.table> --source <path> \
--format parquet --available-now --clean-source-retention "8 days"
autoload coerces those flags and calls
:func:yggdrasil.databricks.table.auto_loader.auto_load — the Spark Structured
Streaming + cloudFiles ingestion body — then returns its summary. The
handler stays thin; the logic lives in auto_load.
This is the data plane (run the ingestion). The control-plane command that
creates / deploys the job is ygg databricks tables autoload (see
:mod:~yggdrasil.databricks.cli.services.tables).