JDBC metadata reports java.lang.Object for IPv4 and IPv6 column class names

[ventod]

Description

ResultSetMetaData#getColumnClassName() reports java.lang.Object for ClickHouse IPv4 and IPv6 columns.

However, ResultSet#getObject() correctly returns java.net.Inet4Address and java.net.Inet6Address, respectively.

Therefore, the metadata class name appears to be inconsistent with the actual Java type returned by the driver.

Steps to reproduce

1. Create a table containing IPv4 and IPv6 columns.
2. Insert IPv4 and IPv6 values.
3. Execute a SELECT.
4. Inspect ResultSetMetaData#getColumnClassName().
5. Inspect the runtime classes returned by ResultSet#getObject().

Error Log or Exception StackTrace

No exception is thrown. The issue is an incorrect JDBC metadata value.

Column: ipv4_col
Column type name: IPv4
JDBC type: 1111 (Types.OTHER)
Column class name: java.lang.Object

Column: ipv6_col
Column type name: IPv6
JDBC type: 1111 (Types.OTHER)
Column class name: java.lang.Object

The actual objects returned by ResultSet#getObject() are:

class java.net.Inet4Address
class java.net.Inet6Address

Expected Behaviour

ResultSetMetaData#getColumnClassName() should report the Java class actually returned by ResultSet#getObject():

IPv4 -> java.net.Inet4Address
IPv6 -> java.net.Inet6Address

Specifically:

ipv4_col:
  getColumnTypeName()  -> IPv4
  getColumnType()      -> Types.OTHER
  getColumnClassName() -> java.net.Inet4Address

ipv6_col:
  getColumnTypeName()  -> IPv6
  getColumnType()      -> Types.OTHER
  getColumnClassName() -> java.net.Inet6Address

Code Example

ResultSet rs = statement.executeQuery(
    "SELECT ipv4_col, ipv6_col FROM data_types_test_table"
);

ResultSetMetaData metadata = rs.getMetaData();

for (int i = 1; i <= metadata.getColumnCount(); i++) {
    System.out.println("Column: " + metadata.getColumnName(i));
    System.out.println("Type name: " + metadata.getColumnTypeName(i));
    System.out.println("JDBC type: " + metadata.getColumnType(i));
    System.out.println("Class name: " + metadata.getColumnClassName(i));
}

rs.next();

Object ipv4 = rs.getObject("ipv4_col");
Object ipv6 = rs.getObject("ipv6_col");

System.out.println(ipv4.getClass());
System.out.println(ipv6.getClass());

Output:

Column: ipv4_col
Type name: IPv4
JDBC type: 1111
Class name: java.lang.Object

Column: ipv6_col
Type name: IPv6
JDBC type: 1111
Class name: java.lang.Object

class java.net.Inet4Address
class java.net.Inet6Address

Configuration

Client Configuration

No non-default client configuration is required to reproduce the issue.
I just created a dumb table with 2 ip columns.

Environment

• Cloud
• Client version: 0.10.0 (jdbc driver version)
• Language version: Java 17
• OS: macOS / Windows

ClickHouse Server

• ClickHouse Server version: clickhouse/clickhouse-server:26.9.11.2-distroless (docker image)
• ClickHouse Server non-default settings, if any: None
• CREATE TABLE statements for tables involved:

CREATE TABLE data_types_test_table
(
    ipv4_col IPv4,
    ipv6_col IPv6
)
ENGINE = MergeTree
ORDER BY tuple();

• Sample data for all these tables:

INSERT INTO data_types_test_table
(
    ipv4_col,
    ipv6_col
)
VALUES
(
    '10.0.0.1',
    '2001:db8::1'
);

[chernser]

@ventod

Thank you for reporting this!