snowflake.snowpark.Session.call

Session.call(sproc_name: str, *args: Any, statement_params: Optional[Dict[str, Any]] = None, block: bool = True, log_on_exception: bool = False, return_dataframe: Optional[bool] = None) → Union[Any, AsyncJob][source]

Calls a stored procedure by name.

Parameters:

  • sproc_name – The name of stored procedure in Snowflake.

  • args – Arguments should be basic Python types.

  • statement_params – Dictionary of statement level parameters to be set while executing this action.

  • block – Whether to block until the result is available. When it is False, this function executes the stored procedure asynchronously and returns an AsyncJob.

  • log_on_exception – Log warnings if they arise when trying to determine if the stored procedure as a table return type.

  • return_dataframe – When set to True, the return value of this function is a DataFrame object. This is useful when the given stored procedure’s return type is a table.

Returns:

The stored procedure result (scalar value or DataFrame) When block=False: An AsyncJob object for retrieving results later

Return type:

When block=True

Example:

>>> import snowflake.snowpark
>>> from snowflake.snowpark.functions import sproc
>>>
>>> session.add_packages('snowflake-snowpark-python')
>>>
>>> @sproc(name="my_copy_sp", replace=True)
... def my_copy(session: snowflake.snowpark.Session, from_table: str, to_table: str, count: int) -> str:
...     session.table(from_table).limit(count).write.save_as_table(to_table)
...     return "SUCCESS"
>>> _ = session.sql("create or replace table test_from(test_str varchar) as select randstr(20, random()) from table(generator(rowCount => 100))").collect()
>>> _ = session.sql("drop table if exists test_to").collect()
>>> session.call("my_copy_sp", "test_from", "test_to", 10)
'SUCCESS'
>>> session.table("test_to").count()
10

Example:

>>> from snowflake.snowpark.dataframe import DataFrame
>>>
>>> @sproc(name="my_table_sp", replace=True)
... def my_table(session: snowflake.snowpark.Session, x: int, y: int, col1: str, col2: str) -> DataFrame:
...     return session.sql(f"select {x} as {col1}, {y} as {col2}")
>>> session.call("my_table_sp", 1, 2, "a", "b").show()
-------------
|"A"  |"B"  |
-------------
|1    |2    |
-------------