unit test plugin for dbt
- column types -
- automatic derivation
- ddl upfront including drop + df.to_sql load/replace
-
create schema for tablenames with dot-notation
-
include directory parsing to find all dataset-definitions
-
script file execution (sql commands from a file) . sqlalchemy function execute_file
-
expected result execution and check wether result set is empty --> positive test result . return 0 in case of empty result set . return df with resultset otherwise
-
collect TestSuite and Test from directory Structure . TestSuite has a setup method (xml-datasets + init.dml) and a list of Tests . Test has an init.dml and expected.dml . TestResult --> list of TestSuites/Tests and a repot of green/red tests with error messages
-
build it into dbt