DuckDB Python read_avro fails on empty Avro container #276

#!/usr/bin/env python3
"""
Repro: DuckDB Python read_avro fails on empty Avro container.

Requires:
  - fastavro
  - duckdb python package installed
"""
import tempfile
from pathlib import Path

import duckdb


def main():
    tmpdir = Path(tempfile.mkdtemp(prefix="duckdb_avro_empty_"))
    try:
        empty_avro = tmpdir / "empty.avro"
        one_avro = tmpdir / "one.avro"

        schema = {
            "type": "record",
            "name": "TestRecord",
            "fields": [
                {"name": "id", "type": "string"}
            ]
        }
        try:
            import fastavro  # type: ignore
        except Exception as e:
            raise RuntimeError("fastavro is required for this repro script") from e

        # Non-empty file via fastavro
        with open(one_avro, "wb") as out_f:
            fastavro.writer(out_f, schema, [{"id": "1"}])

        # Empty container via fastavro
        with open(empty_avro, "wb") as out_f:
            fastavro.writer(out_f, schema, [])

        print(f"DuckDB version: {duckdb.__version__}")
        con = duckdb.connect()
        try:
            try:
                row = con.execute(
                    "SELECT extension_version, installed, installed_from, install_path "
                    "FROM duckdb_extensions() WHERE extension_name = 'avro'"
                ).fetchone()
                if row:
                    ext_version, installed, installed_from, install_path = row
                    print(
                        "DuckDB avro extension: "
                        f"version={ext_version}, installed={installed}, "
                        f"from={installed_from}, path={install_path}"
                    )
                else:
                    print("DuckDB avro extension: not listed")
            except Exception as e:
                print(f"Could not read DuckDB avro extension version: {e}")
            print("Reading one.avro...")
            con.execute("SELECT count(*) FROM read_avro(?)", [str(one_avro)]).fetchall()
            print("OK: one.avro")
        except Exception as e:
            print("FAIL: one.avro:", e)

        try:
            print("Reading empty.avro...")
            con.execute("SELECT count(*) FROM read_avro(?)", [str(empty_avro)]).fetchall()
            print("OK: empty.avro")
        except Exception as e:
            print("FAIL: empty.avro:", e)
        finally:
            con.close()
    finally:
        print(f"Artifacts in: {tmpdir}")


if __name__ == "__main__":
    main()