dbt Classes Instance Methods is_string, is_numeric, is_number, is_float don't seem to work

[fivetran-catfritz]

Describe the bug

The same issue with the methods persists in the Databricks adapter as was reported in the Spark adapter in this issue.

dbt Classes Instance Methods is_string, is_numeric, is_number, is_float don't seem to work

For our instance, col.is_string() does not works, and for one of our customers, the col.is_number() also does not work.

Steps To Reproduce

• Take any string field output from get_columns_in_relation and print the result of col.is_string().

Expected behavior

• When col.datatype == 'string' then col.is_string() should evaluate to true.

Screenshots and log output

• I ran this print statement.

{{ print('Column: ' ~ col.name ~ ' | Data Type: ' ~ col.data_type ~ ' | Is Number: ' ~ col.is_number() ~ ' | Is String: ' ~ col.is_string()) }}

And the result shows it's not identifying correctly.

Column: name03 | Data Type: string | Is Number: False | Is String: False

System information

The output of dbt --version:

Core:
  - installed: 1.11.6
  - latest:    1.11.7 - Update available!

Plugins:
  - databricks: 1.11.6 - Up to date!

The operating system you're using:
macOS

The output of python --version:
Python 3.13.9

Additional context

Add any other context about the problem here.

[psaikaushik]

I can pick this up. The same issue was reported in dbt-spark (#658) and the root cause is the same — the Column base class methods like is_string() check against type names that don't include Spark/Databricks types.

Looking at the code, is_string() in the base Column class likely checks for types like text, character varying, etc. (PostgreSQL-style names), but Databricks returns string as the dtype. The fix is to override the string_type classmethod (or the equivalent type classification method) in the Databricks Column implementation to include string in the set of recognized string types.

Same pattern for is_number() — Databricks uses int, bigint, double, etc. which may not all be in the base class's numeric type set.

I'll look at the Column adapter class and send a PR.

[psaikaushik]

PR up: #1390

Overrides is_string(), is_number(), is_float(), is_integer(), and is_numeric() on DatabricksColumn with the full set of Spark/Databricks type names. The base Column class only knows PostgreSQL types, which is why these all returned False.