Feature Request: run EXPERIMENTAL expectation (from great_expectations_experimental library) from Airflow?

[kujaska]

Hi!
Is it possible to run
1) EXPERIMENTAL expectation (from great_expectations_experimental library) from Airflow?

example: expect_queried_column_values_to_exist_in_second_table_column

Simple import to DAG does not help:

from great_expectations_experimental.expectations.expect_queried_column_values_to_exist_in_second_table_column import ExpectQueriedColumnValuesToExistInSecondTableColumn

• after DAG run getting this text in DataDocs instead of the expectation result:

expect_queried_column_values_to_exist_in_second_table_column(**{'batch_id': '0120cd462e58ed32be35bc92c0ae', 'template_dict': {'condition': '1=1', 'first_table_column': 'PROV_ID', 'second_table_column': 'PROV_ID', 'second_table_full_name': 'LINC'}}) (edited)

2) a custom expectation from great_expectations/plugins/expectations folder?
could it be run from Airflow? how?
https://docs.greatexpectations.io/docs/guides/expectations/creating_custom_expectations/how_to_use_custom_expectations/ (edited)

[kujaska]

Same needed to run a custom expectation from PLUGINS folder.

It could be done with additional IMPORT with pure GE but there is no way to do it with airflow-provider-great-expectations

Meanwhile it looks like this works:

in your DAG:

def ge_run_func():
    import great_expectations as gx
    data_context = gx.get_context(context_root_dir="your_root_dir")
    #import your custom expectation plugin from plugins folder here
    from expectations.expect_column_values_to_be_alphabetical import ExpectColumnValuesToBeAlphabetical
    result: CheckpointResult = data_context.run_checkpoint(
        checkpoint_name="your_checkpoint_name"
        )
    return (result["success"])

• and run ge_run_func in DAG with PythonOperator