fix: delegate Arrow schema conversion to PyIceberg

[konstantinoscs]

What this does

This PR passes a pyarrow.Schema directly to PyIceberg when creating a table and removes Tower's custom schema converter.

PyIceberg now handles field IDs, nested fields, documentation, nullability, and type validation using its standard public API.

Why

Tower's converter duplicated PyIceberg behavior and could silently lose information, particularly timestamp and decimal precision. The native conversion is easier to maintain and rejects unsupported or lossy schemas clearly.

Tests also create tables through S3 Tables and BYO-style catalog configuration, confirming that those paths still use their existing PyIceberg catalog and never request Tower-managed credentials.

Compatibility

Some schemas that were previously accepted after a lossy conversion now raise PyIceberg's normal validation errors. The reference documentation lists the supported timestamp, time, date, decimal, and timezone behavior.

Testing

• 20 schema-focused tests pass.
• The complete stack passes 225 tests, with 10 skipped.
• Black and mypy pass.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 2153591e-1183-4321-8bf9-6ec7359abaaa

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

_{Comment @coderabbitai help to get the list of available commands.}