[Feature Request]: Add read() support to Java ClickHouseIO (currently write-only)

[nitinware]

What would you like to happen?

Motivation

The Java ClickHouseIO connector at sdks/java/io/clickhouse/ currently supports only writes. Reading from ClickHouse in a Beam pipeline requires falling back to JdbcIO with the ClickHouse JDBC driver, which loses ClickHouse-specific schema fidelity (LowCardinality, Array(T), Nullable(T), Nested, tuples) and native columnar transport.

Proposal

Add ClickHouseIO.read() and ClickHouseIO.readAll() PTransforms symmetric to the existing write path:

• Native Row output using the ClickHouse client's native format
• Query/projection via user-supplied SELECT template
• Splittable reader for parallel reads across partitions / shards,
  modeled on JdbcIO's splitting pattern
• Reuse the existing connection configuration builder from the writer

Benefits

• Closes an obvious read/write asymmetry — every other major DB IO in
  Beam (JDBC, Cassandra, Mongo, HBase, Spanner, BigQuery, Snowflake)
  supports both directions
• ClickHouse is increasingly common as an analytics source; users
  currently work around this with JdbcIO at the cost of type fidelity

Testing plan

• Unit tests mirroring the existing writer test file
• Integration test using the module's existing testcontainers-based
  ClickHouse setup (ClickHouseContainer is already wired in build.gradle)
• No new dependencies required

Issue Priority

Priority: 2 (default / most feature requests should be filed as P2)

Issue Components

• Component: Python SDK
• Component: Java SDK
• Component: Go SDK
• Component: Typescript SDK
• Component: IO connector
• Component: Beam YAML
• Component: Beam examples
• Component: Beam playground
• Component: Beam katas
• Component: Website
• Component: Infrastructure
• Component: Spark Runner
• Component: Flink Runner
• Component: Prism Runner
• Component: Twister2 Runner
• Component: Hazelcast Jet Runner
• Component: Google Cloud Dataflow Runner

[nitinware]

.take-issue