Correct field projection not working

[Fokko Driesprong]

In Iceberg the tables are projected using field-IDs. Even if the column is renamed (and Iceberg is lazy, so existing data will not be rewritten), it should correct map the name.

Example:

from pyiceberg.catalog.sql import SqlCatalog

catalog = SqlCatalog("test_sql_catalog", uri="sqlite:///:memory:", warehouse=f"/{warehouse}")
catalog.create_namespace("default")

import pyarrow as pa

pa_schema = pa.schema([
    pa.field('animal', pa.string(), nullable=False),
    pa.field('n_legs', pa.int32(), nullable=False),
])

# Empty table
tbl = catalog.create_table("default.test_id_projection", schema=pa_schema)

# Write some data
tbl.append(
    df=pa.Table.from_arrays([pa.array(["Flamingo", "Horse"]), pa.array([2, 4])], schema=pa_schema),
)

# Rename the column
with tbl.update_schema() as upd:
    upd.rename_column("n_legs", "number_of_legs")

# Write more data
tbl.append(
    df=pa.Table.from_arrays(
        [pa.array(["Brittle stars", "Centipede"]), pa.array([5, 100])],
        schema=pa.schema([
            pa.field('animal', pa.string(), nullable=False),
            pa.field('number_of_legs', pa.int32(), nullable=False),
        ]),
    ),
)

location = tbl.metadata_location

import duckdb

duckdb.sql('INSTALL iceberg; LOAD iceberg;')
result = duckdb.sql(
    f"""
SELECT *
FROM iceberg_scan('{location}')
"""
).fetchall()

assert result == [
    ('Flamingo', 2),
    ('Horse', 4),
    ('Brittle stars', 5),
    ('Centipede', 100),
]

This fails with:

>       ).fetchall()
E       duckdb.duckdb.IOException: IO Error: Failed to read file "//private/var/folders/22/yb9h2zd55ql37h4_50xkmg7r0000gn/T/pytest-of-fokkodriesprong/pytest-8/test_sql0/default.db/test_id_projection/data/00000-0-d24c398f-f93f-420c-82d4-2a66b72c580e.parquet": schema mismatch in glob: column "number_of_legs" was read from the original file "//private/var/folders/22/yb9h2zd55ql37h4_50xkmg7r0000gn/T/pytest-of-fokkodriesprong/pytest-8/test_sql0/default.db/test_id_projection/data/00000-0-f2b7f973-f7df-4cd6-aa7a-78176b3c3d7f.parquet", but could not be found in file "//private/var/folders/22/yb9h2zd55ql37h4_50xkmg7r0000gn/T/pytest-of-fokkodriesprong/pytest-8/test_sql0/default.db/test_id_projection/data/00000-0-d24c398f-f93f-420c-82d4-2a66b72c580e.parquet".
E       Candidate names: animal, n_legs
E       If you are trying to read files with different schemas, try setting union_by_name=True