Why does a Polars dataframe appears garbled in the VSCode terminal?

Question

I am trying to use Polars with Rust. First thing I did was to read a tiny CSV file I created with Excel and print it. The output in the VSCode terminal appears full of strange characters:

df_astro: shape: (3, 2)
ΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓö¼ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÉ
Γöé A   Γöå B   Γöé
Γöé --- Γöå --- Γöé
Γöé i64 Γöå i64 Γöé
Γò₧ΓòÉΓòÉΓòÉΓòÉΓòÉΓò¬ΓòÉΓòÉΓòÉΓòÉΓòÉΓòí
Γöé 1   Γöå 4   Γöé
Γö£ΓòîΓòîΓòîΓòîΓòîΓö╝ΓòîΓòîΓòîΓòîΓòîΓöñ
Γöé 2   Γöå 5   Γöé
Γö£ΓòîΓòîΓòîΓòîΓòîΓö╝ΓòîΓòîΓòîΓòîΓòîΓöñ
Γöé 3   Γöå 6   Γöé
ΓööΓöÇΓöÇΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÿ

Running the same program on the Command Prompt or PowerShell the output is Ok:

df_astro: shape: (3, 2)
┌─────┬─────┐
│ A   ┆ B   │
│ --- ┆ --- │
│ i64 ┆ i64 │
╞═════╪═════╡
│ 1   ┆ 4   │
├╌╌╌╌╌┼╌╌╌╌╌┤
│ 2   ┆ 5   │
├╌╌╌╌╌┼╌╌╌╌╌┤
│ 3   ┆ 6   │
└─────┴─────┘

Here is the code:

fn read_csv_to_dataframe(file_path: &str) -> Result<DataFrame,PolarsError> {
    let df = CsvReader::from_path(file_path)?
    .infer_schema(Some(5))  // Adjust the number of lines to read for schema inference
    .has_header(true)
    .with_delimiter(b',')   // Specify the delimiter if it's not a comma
    .with_ignore_parser_errors(true) // This might help if there are parsing issues
    .finish()?;

    Ok(df)
}
let df_astro = read_csv_to_dataframe(&file_astro);
    match df_astro {
        Ok(df) => println!("df_astro: {:#?}", df),
        Err(err) => println!("Error reading astro data: {}", err),
    }

Any suggestions on how to fix this?

I checked that the VSCode terminal is using UTF-8. I changed the font. I looked at the CSV file with Notepad++ and it did not have any strange characters, in each line numbers and CRLF at the end.

None of these things helped.