[EPIC] Add Gravitino support in Daft

### Describe the proposal

Daft is a high-performance, open-source data engine designed for AI and multimodal workloads. It provides a familiar Python DataFrame API that allows users to process various data types—including images, audio, video, and text, alongside structured data—at any scale, from a single laptop to a distributed cluster.

Now some Gravitino users expect to have Daft intetgration in Daft. For example, load catalog/schema/table that registered in Gravitino, load files from fileset directly with "gvfs://" scheme et. This Epic is to record the subtasks related to this feature. The code mainly in daft repository, but have the issues in Gravitino to track the progress and status.

### Task list

Will create subtasks later.