Function read_dataset_dataframe 

pub fn read_dataset_dataframe(
    path: &Path,
) -> Result<(DataFrame, BTreeMap<String, String>), Error>

Read an EdgeFirst dataset annotation file (.arrow/.ipc/.parquet) and its file-level metadata (schema_version, labels, category_metadata).

Format is selected from the file extension: .parquet reads the Apache Parquet footer key-value metadata; anything else (including .arrow/.ipc) reads Arrow IPC custom schema metadata.

If only the metadata is needed (not the DataFrame), prefer the cheaper read_dataset_metadata, which skips decoding row data entirely.

Errors

Returns an error if the file cannot be opened or the DataFrame/metadata cannot be decoded for the selected format.