[Umbrella] FIP-17: Support Full KV Scan for Primary Key Tables

### Search before asking

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

### Motivation

Fluss KV currently supports point lookups, but RocksDB allows iterators, range, and prefix scans.

For small tables (hundreds of rows or a few thousands), that can be used to create dashboards often need to periodically read all keys.

**Goal:** Assume I have a front-end and I want to create a dashboard to show < 1-2k values.. In this scenario, I could use the fluss client directly to periodically query the whole kv store

### Solution

We can expose functionality to the api to allow a full scan (with snapshot consistency) to simplify such use cases.

-------------------
An implementation can be found [here](https://github.com/apache/fluss/pull/2809) and for easier reviews it can be broken down to the following individuals PRs: