Improve DatabricksSqlSensor, DatabricksSqlStatementsSensor and DatabricksSqlStatementsOperators with query tags

[zhaorui2022]

Description

Similar to #66839

Databricks query tags help with cost attribution and track back the origin of Databricks queries. The link issue added query tags in DatabricksSqlOperator, DatabricksCopyIntoOperator, however, there are other operators that send queries directly to Databricks and can be instrumented with query tags, that is,

• DatabricksSqlSensor: https://github.com/SkastVnT/airflow/blob/ee12bab9132e0a29766e30f5cdd75cf796a53e6f/providers/databricks/src/airflow/providers/databricks/sensors/databricks_sql.py (this should also cover DatabricksPartitionSensor: https://github.com/SkastVnT/airflow/blob/ee12bab9132e0a29766e30f5cdd75cf796a53e6f/providers/databricks/src/airflow/providers/databricks/sensors/databricks_partition.py)
• DatabricksSqlStatementSensor (uses a different API DatabricksHook.post_sql_statement to send queries): https://github.com/SkastVnT/airflow/blob/ee12bab9132e0a29766e30f5cdd75cf796a53e6f/providers/databricks/src/airflow/providers/databricks/sensors/databricks.py
• DatabricksSqlStatementOperator: https://github.com/SkastVnT/airflow/blob/ee12bab9132e0a29766e30f5cdd75cf796a53e6f/providers/databricks/src/airflow/providers/databricks/operators/databricks.py#L1103

Use case/motivation

Query tags helps to track the origin of Databricks queries, and can be used for cost attribution and etc

Related issues

#66839

Are you willing to submit a PR?

• Yes I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct

[josix]

Hi, I'd like to work on this if it's still available.

[jroachgolf84]

@josix - feel free to open a PR for this issue!

[venkatamandavilli-code]

Thanks for raising this enhancement.

Adding query tags to Databricks SQL sensors and operators would be useful from an enterprise operations and governance perspective. In larger environments, teams often need better traceability between Airflow tasks and the SQL workloads executed in Databricks.

Query tags could help with monitoring, troubleshooting, cost attribution, auditability, and identifying which DAG or task triggered a specific Databricks workload. This becomes especially valuable when multiple teams share the same Databricks workspace and Airflow platform.

It may also improve supportability during incident analysis because platform teams can more easily connect downstream query behavior back to the Airflow workflow that initiated it.