How to pass paremeter in my .sql file in PostGresOperator airflow?

Question

I have sql file and I want to pass parameters to that sql file using PostGresOperator.

"""select * from table_{} where id > ID """.format(mytable,myID)

My postGresOperator

mport_redshift_table = PostgresOperator(
            task_id='copy_data_from_redshift_{}'.format(country),
            postgres_conn_id='postgres_default',
            sql="""
                   select * from table_{} where id > {}
                """.format(mytable,myID)

How can I do the same and pass my parameters in my .sql file and still use .format(mytable,myID) ?

so that I can pass them into my referenced .sql file.

Answer 1

As explained in the How-to Guide for PostgresOperator, you can place your SQL in a file within a subfolder in the dag directory:

-- dags/sql/birth_date.sql
SELECT * FROM pet WHERE birth_date BETWEEN SYMMETRIC {{ params.begin_date }} AND {{ params.end_date }};

Use params to pass in the key/value pairs that would be rendered within the SQL in your file:

get_birth_date = PostgresOperator(
    task_id="get_birth_date",
    postgres_conn_id="postgres_default",
    sql="sql/birth_date.sql",
    params={"begin_date": "2020-01-01", "end_date": "2020-12-31"},
)

Edit

If you want to do it inline, without using the file, just use any string interpolation mechanism:

sql="""select * from table_{} where id > {} """.format(mytable,myID)

or

sql=f"""select * from table_{table_name} where id > {myID} """

or if you want to use jinja, taking advantage of any of the default vairables, such as a param provided when the DAG was triggered you could do it like this:

sql=f"""select * from table_{{ params.param1 }} where id > {myID} """