Add convenience function to write certain EEG metadata

[hoechenberger]

Describe the problem

Working with EEG data, I find myself writing the following to ensure more complete metadata:

        bp_written = mne_bids.write_raw_bids(...)

        # Update EEG metadata
        bp_eeg_json = bp_written.update(extension=".json")
        updated_eeg_metdata = {
            "EEGReference": "Oz",
            "EEGGround": "Fpz",
            "CapManufacturer": "EasyCap",
            "ManufacturersModelName": "actiCHamp"
        }
        mne_bids.update_sidecar_json(bids_path=bp_eeg_json, entries=updated_eeg_metdata)

Since this information (reference, ground electrode, amplifier and cap manufacturer) should ideally always be written for EEG, it would be great to make this easier for users.

Describe your solution

Add a new method BIDSPath.update_eeg_metadata() to conveniently add this information:

def update_eeg_metadata(
    *,
    amplifier_model: str,
    cap_manufacturer: str,
    reference: str,
    ground: str,
) -> None:
    ...

Describe possible alternatives

Add this functional to write_raw_bids(), but I don't think we'll want to have so many new parameters.

Additional context

This is related to #713

[agramfort]

I find that it overfits the EEG use case and it makes us maintain a mapping between our parameter names and the BIDS standard.

my 2c

[hoechenberger]

Any other ideas on how we could make the process easier for users then? I think we're making it too hard for users to produce datasets with "complete" metadata

[sappelhoff]

I follow the same workflow as you @hoechenberger.

I think that if we don't want to add a convenience function to the code base (in order to avoid overfitting for EEG), perhaps we can add an example on how to do it.