Vijay Kiran

Add support for checks that are based on the relative percentage difference eg ```yaml checks for CUSTOMERS: - change percent between avg last 7 and count < 10 % ```

Depends on https://github.com/sodadata/soda-core/issues/1243

See docs/soda_checks_yaml.md. L69 find the TODO Consider if we should push it to the user to define the right variables and avoid clashes between the variable names when comparing? So...

SodaCL validation

2

Implement an API to validate a scan. The goal is to log as much errors as possible without opening any warehouse connection.

with this checks: ```yaml checks for dim_customer: - invalid_count(marital_status) = 0: valid values: - W ``` Scan reports no errors: ```shell ➜ soda scan -c configuration.yml -d adventureworks checks1.yml Soda...

Schema check type names matching

1

Relevant checks.yml ``` checks for actor: - count > 0 - column types: actor_id: integer first_name: varchar last_update: timestamp ``` ``` DEBUG | column types [FAILED] DEBUG | column_type_mismatch[last_update] expected(timestamp)...

Add CLI create

1

Add CLI command for validating the connection and building the first checks. Eg `soda connect` will check all the configured datasources. ``` soda connect datasource redshift_xyz [OK] datasource postgres_dev [OK]...

All data sources have reserved keywords, if these are used for table/column names we have to quote them. Can we just quote everything to get around the issue?

Similar support for Hive as in Soda SQL https://docs.soda.io/soda/warehouse_types.html#apache-hive-experimental