Writing v3 table metadata

[rambleraptor]

Part of #1818

This might be early. This enables writing v3 metadata

Rationale for this change

We'll need this at some point for v3 support.

Are these changes tested?

Unit test included.

Are there any user-facing changes?

• Adds support for writing v3 table metadata.

[glesperance]

Hey @rambleraptor, similar to my comment on your deletion_vector_write PR, I've added Spark interop tests for the V3 metadata work: glesperance/iceberg-python@f57ea40c

• test_spark_reads_pyiceberg_v3_metadata (pyiceberg → Spark): Verifies Spark can read V3 table metadata written by pyiceberg
• test_pyiceberg_reads_spark_v3_table (Spark → pyiceberg): Verifies pyiceberg can read V3 tables created by Spark

Happy to open a PR against your branch, or feel free to cherry-pick if easier.