### Feature Request / Improvement Currently, `large_string` data type is converted to `string` ([link](https://github.com/apache/iceberg-python/blob/main/pyiceberg/io/pyarrow.py#L879-L880)) This breaks the parquet writer when we're writing an Arrow table with a `large_string` column See https://github.com/pola-rs/polars/issues/9795