What happens?

Selecting MAP datatype into dataframes creates a column of object datatype but then trying to reinsert from the dataframe throws a ConversionException: Conversion Error: Unimplemented type for cast (STRUCT(a INTEGER) -> MAP(VARCHAR, INTEGER))

There seems to be no option for doing bulk inserting into MAP columns due to this.

To Reproduce

import duckdb

# Establish a connection to DuckDB (in-memory)
conn = duckdb.connect(":memory:")

# Create a table with a MAP column
conn.execute("""
    CREATE TABLE my_table (
        id INTEGER,
        data MAP(VARCHAR, INTEGER)
    )
""")

# manually insert the data
conn.execute("INSERT INTO my_table (id, data) VALUES (1, MAP {'a': 1})")

# Verify the data
result = conn.execute("SELECT * FROM my_table").fetch_df()
print(result)
#    id      data
# 0   1  {'a': 1}
print(result["data"].dtype)
# object

# try to reinsert queried data
conn.execute("INSERT INTO my_table (id, data) SELECT * FROM result")
# ConversionException: Conversion Error: Unimplemented type for cast (STRUCT(a INTEGER) -> MAP(VARCHAR, INTEGER))

OS:

macOS Intel Silicon

DuckDB Version:

1.2

DuckDB Client:

Python

Hardware:

No response

Full Name:

Raj Veerappan

Affiliation:

Personal

What is the latest build you tested with? If possible, we recommend testing with the latest nightly build.

I have tested with a stable release

Did you include all relevant data sets for reproducing the issue?

Yes

Did you include all code required to reproduce the issue?

• Yes, I have

Did you include all relevant configuration (e.g., CPU architecture, Python version, Linux distribution) to reproduce the issue?

• Yes, I have