Description
Feature Request / Improvement
This feature request arises from the discussion in #183 and is contingent upon the implementation of write support as outlined in #41. Upon the merger of PR #41, the following challenge is observed:
When attempting to write or overwrite new data to an Iceberg table, it is crucial that the column order in the pyarrow.Table dataframe or a Parquet file (without field-ids) aligns with the existing table schema. Discrepancies in column order can lead to issues during the process of writing data to a Parquet file using the Iceberg table's schema or when appending an existing Parquet file to our table, leading to incorrect data reads in subsequent operations.
To mitigate this issue,we might want a mechanism to establish a correspondence between the column names in pyarrow.Table and the Iceberg table schema. This will enable us to correctly assign field-ids to the respective columns when writing data.