Add an experimental e-graph equality saturation framework for KQIR optimizer

### Search before asking

- [X] I had searched in the [issues](https://github.com/apache/kvrocks/issues) and found no similar issues.

### Motivation

As e-graphes are great tool for equality saturation in program optimizer to perform term rewriting, we can build them as a builtin framework for KQIR optimizer to generate better query plans.

### Solution

_No response_

### Are you willing to submit a PR?

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

The issue names no files, tests, or entry points. Start by locating the KQIR optimizer and reviewing how it generates query plans, then establish the e-graph framework's scope and acceptance criteria; the issue currently does not define what done looks like.

Written by the indexing model from the issue text.