[X] I have searched the existing issues, and I could not find an existing issue for this bug
Current Behavior
Unit test fixtures are parsed correctly, but they leave typing up to the target database. This will result in unhandled database failures when there is no one row Null free (this handles indirectly the all values in a column Null failure case). The database won't be able to determine which type to use and default to something like VARCHAR(1) which may not be what the source table expects -- this leads to a type failure.
This change is motivated by a redshift change but could affect other adapters some day.
Expected Behavior
Let's sort this to give fixtures the best chance to succeed or in unwinnable situations, error out with a known exception.
Is this a new bug in dbt-core?
Current Behavior
Unit test fixtures are parsed correctly, but they leave typing up to the target database. This will result in unhandled database failures when there is no one row Null free (this handles indirectly the all values in a column Null failure case). The database won't be able to determine which type to use and default to something like VARCHAR(1) which may not be what the source table expects -- this leads to a type failure.
This change is motivated by a redshift change but could affect other adapters some day.
Expected Behavior
Let's sort this to give fixtures the best chance to succeed or in unwinnable situations, error out with a known exception.
Steps To Reproduce
See reprex here
Relevant log output
No response
Environment
Which database adapter are you using with dbt?
redshift
Additional Context
See redshift PR here