Add read-only support for Hadoop Catalogs

[peasee]

Is your feature request related to a problem or challenge?

iceberg-rust should implement a HadoopCatalog to read file-based catalogs, through existing FileIO stores like file://, s3://, s3a://, etc.

I see there has been a PR with a community contribution that has been effectively declined, due to concerns over deprecating file-based catalogs: #1313

However, my opinion is that there are surely many users with large Hadoop Catalogs who could benefit from this crate implementing a Hadoop Catalog. As the user from the PR mentions, StaticTable can be ineffective for large catalogs.

Describe the solution you'd like

My proposal is that instead of implementing a Hadoop Catalog with full read-write functionality, the catalog implements read-only functionality. This may favor better with the community, by avoiding creating more file-based Hadoop Catalogs - while still providing users with these catalogs a method to read their catalogs (and potentially migrate to other catalogs).

As a disclaimer, I am already working on an implementation for this solution. I would love to contribute it back to the crate, so I am creating this issue early for discussion. I will be primarily focusing on file://, s3:// and/or s3a:// FileIO connections.

Willingness to contribute

I can contribute to this feature independently