[EHN] EEGBIDSDataset is only use for diggestion of new datasete

[bruAristimunha]

EEGDash/eegdash/data_utils.py

Line 484 in e8b82b7

class EEGBIDSDataset:

With the advance of the EEGdash, do we still need a dedicated class to do all this digestion? I feel that we can use more mne_bids here, to avoid rewriting a lot of code. @arnodelorme and @dungscout96

[dungscout96]

I'm ok with transition to mne_bids. If the functionality used in https://github.com/sccn/EEGDash/blob/develop/eegdash/bids_eeg_metadata.py can be all replaced by mne_bids I'm good with removing the class entirely

[arnodelorme]

It is very slow, though, to scan datasets when offline (not sure if this is related). It takes about 30 seconds to import 300 datasets. With MongoDB, it is more like 2s. Could we cache the information somewhere?

[bruAristimunha]

We can benchmark this before merge @arnodelorme. Can you give access to the cluster with all the datasets?