[Feature] Support Semi-Structured Data Columns in Unit Test dict fixture data

[ronco]

Is this your first time submitting a feature request?

• I have read the expectations for open source contributors
• I have searched the existing issues, and I could not find an existing issue for this feature
• I am requesting a straightforward extension of existing dbt functionality, rather than a Big Idea better suited to a discussion

Describe the feature

When writing unit tests that include semi-structured data columns (SUPER in Redshift for example), the fixture data can only be staged via SQL. This feature calls for understanding structured data in dict fixtures.

An example fixture definition would be something like this:

- input: ref('table_with_super')
  format: dict
  rows:
    - { structured_column: {foo: bar}, int_column: 123 }

Describe alternatives you've considered

This functionality can be achieved using SQL fixtures like so:
SELECT json_parse('{"foo":"bar"}')::SUPER as structured_column
This is quite cumbersome as you have to specify all significant columns in your sql.

Who will this benefit?

Anyone making use of semi-structured data columns.

Are you interested in contributing this feature?

I am open to contributing if this is of interest.

Anything else?

No response