Add Early Validation for Aggregation Function Data Types During Table Creation

[wuchong]

Search before asking

• I searched in the issues and found nothing similar.

Description

Currently, Fluss does not validate whether the column data type is compatible with a given aggregation function (e.g., MAX, MIN) at table creation time. This leads to runtime failures or unexpected behavior when unsupported types—such as ARRAY, MAP, or ROW—are used with aggregation functions that require orderable types.

To improve developer experience and system robustness, we should fail fast during DDL validation.

✅ Expected Behavior

• During CREATE TABLE (or table descriptor validation), if an aggregation function is specified on a column with an unsupported data type, Fluss should throw a clear, actionable error immediately.
• The validation should reside in:
  org.apache.fluss.server.utils.TableDescriptorValidation#validateAggregationFunctionParameters

Solution

To ensure future aggregate functions are easily extensible, we should introduce a standardized way to declare and validate the data types each function supports.

One approach is adding a method such as validateDataType() to the AggFunctionType interface, similar in spirit to the existing validateParameter() that explicitly checks whether a given column type is compatible with the aggregate function.

🧪 Test Coverage

• We can add such tests in FlussAdminITCase, like FlussAdminITCase#testCreateTableWithInvalidProperty.

Willingness to contribute

• I'm willing to submit a PR!

[Jackeyzhe]

Hi @wuchong , I'd like to work on it :)

[wuchong]

@Jackeyzhe Thank you for volunteering!

However, @Prajwal-banakar previously expressed interest in working on this in the original discussion: #2296 (comment).

Before moving forward, I’d like to check whether @Prajwal-banakar is still planning to contribute this feature. Otherwise, I will assign it to you @Jackeyzhe .

[Prajwal-banakar]

Hi @wuchong , I am still very interested in this issue, but I currently have academic exams that require my full focus. Since I won't be able to start on it immediately, please feel free to assign it to @Jackeyzhe so the progress isn't delayed. I’ll look for other tasks once my exams are over!

[wuchong]

Thank you @Prajwal-banakar , I assigned this issue to @Jackeyzhe . Feel free to take other issues when you are back. Good luck to your academic exams !