[Bug] findColumn throws NullPointerException instead of SQLException for an unknown column name

[PDGGK]

Search before asking

• I searched in the issues and found nothing similar.

Version

master (2.0.x). The affected code is also present in released 2.0.x versions.

Describe the bug and provide the minimal reproduce step

IoTDBRpcDataSet.findColumn(String) returns columnOrdinalMap.get(columnName) directly. When the column name is not present, Map<String, Integer>.get returns null, which auto-unboxes to int and throws a NullPointerException. The JDBC contract (ResultSet.findColumn) requires a SQLException when the column is not found, and the sibling findColumnNameByIndex already guards its input and throws — the guard was simply omitted for findColumn.

This surfaces at the JDBC boundary (IoTDBJDBCResultSet.findColumn) and through SessionDataSet.DataIterator.findColumn, both of which delegate to the method above. The same missing guard also exists in IoTDBJDBCDataSet.findColumn.

Minimal reproduce step:

1. Execute any query over JDBC and obtain the ResultSet.
2. Call resultSet.findColumn("a_column_name_that_is_not_in_the_result").
3. A raw NullPointerException is thrown instead of a SQLException.

What did you expect to see?

findColumn throws a SQLException (at the JDBC boundary) that names the missing column, per the ResultSet.findColumn contract.

What did you see instead?

A NullPointerException, from unboxing the null returned by the column-ordinal map lookup.

Anything else?

The fix mirrors the existing findColumnNameByIndex guard: throw a checked StatementExecutionException at the rpc layer (reusing the existing UNKNOWN_COLUMN_NAME message), converted to SQLException at the JDBC boundary. I'd like to work on this and will open a PR (which also fixes the identical, currently-unused-but-public IoTDBJDBCDataSet.findColumn).

Are you willing to submit a PR?

• I'm willing to submit a PR!