Support for Arrow Flight SQL Protocol and ADBC

### Search before asking

- [x] I searched in the [issues](https://github.com/alibaba/fluss/issues) and found nothing similar.

### Motivation

Fluss is a real-time streaming storage system designed for analytical workloads, offering sub-second latency for both reads and writes. To further enhance its interoperability and performance in modern data ecosystems, it is essential to integrate Fluss with widely adopted protocols like **Arrow Flight SQL** and **ADBC (Arrow Database Connectivity)**.

- **Arrow Flight SQL** is an RPC framework that enables efficient transfer of Apache Arrow data over the network, providing a high-performance alternative to traditional JDBC/ODBC for SQL-based interactions
- **ADBC** is a standardized API for connecting Arrow-native clients with databases, engines, and storage systems. It abstracts over different database implementations, making it easier for users to interact with Fluss using Arrow-compatible tools.

By supporting these protocols, Fluss enables seamless integration with analytics tools, BI platforms, and Arrow-native systems like **DuckDB**. Since Fluss uses the Apache Arrow format as its underlying log format, it can achieve end-to-end zero-copy data transfer and processing.

### Solution

A design proposal should be submitted and discussed. For reference, see the Doris proposal for supporting Flight SQL: [Apache Doris Issue #25514 ](https://github.com/apache/doris/issues/25514).

## Reference

- Arrow Flight SQL Documentation: [https://arrow.apache.org/docs/format/FlightSql.html](https://arrow.apache.org/docs/format/FlightSql.html)
- ADBC Overview: [https://arrow.apache.org/adbc/ ](https://arrow.apache.org/adbc/)
- Arrow Flight SQL and ADBC Introduction: https://voltrondata.com/blog/simplifying-database-connectivity-with-arrow-flight-sql-and-adbc

### Willingness to contribute

- [ ] I'm willing to submit a PR!

No repository files, tests, or entry points are identified. Start by reading the Arrow Flight SQL and ADBC references and the linked Apache Doris proposal, then inspect Fluss's existing client and protocol boundaries. Done means producing a design proposal for discussion, not an implementation.

Written by the indexing model from the issue text.