eegdash.local_bids module#

Local BIDS discovery helpers.

These utilities support offline workflows (no DB/S3) by discovering BIDS recordings on the filesystem and returning EEGDash v2 records.

eegdash.local_bids.discover_local_bids_records(dataset_root: str | Path, filters: dict[str, Any]) → list[dict[str, Any]][source]

Discover local BIDS recordings and build EEGDash v2 records.

Parameters:

  • dataset_root (str | Path) – Local dataset directory (e.g., /path/to/ds005509).

  • filters (dict) – Filters dict. Must include 'dataset' and may include BIDS entities like 'subject', 'session', 'task', 'run', plus 'modality' (default: 'eeg').

Returns:

A list of v2 records, one for each matched recording file.

Return type:

list[dict[str, Any]]

Notes

Matching is performed via mne_bids.find_matching_paths() using datatypes/suffixes derived from the 'modality' filter. The returned records use storage.backend='local' and point storage.base at dataset_root.