BIDSHandler roadmap idea

Features that are mostly implement already:

• add functionality.

  This is mostly complete. There are some issues still with .fif files that are large and have been split still.

New feature ideas:

• query functionality.

  The ability to query the BIDS file structure for information. Still need to decide the best way to implement this in terms of the API. One possible solution is to have something like .find(**kwargs) where the kwargs can be token, condition, value, obj and maybe others?

  An example would be .find(token='task', condition='=', value='restingstate', obj='subject') to find all the subjects with a 'restingstate' task.

• rename functionality.

  To extend the functionality relating to manipulating BIDS data, the ability to rename certain parameters could be useful. This way you can fix issues that may have occured during generation, or if certain parameters need to be changed (eg. task name changes). This would automatically handle renaming all affected files and could be applied on any scale (eg. apply task = restingstate -> task = resting to an entire project).

• Tests need to be written for other MEG data formats and other modalities need to be tested (EEG?)

• Add methods to allow for easy extraction of the required information to load raw data into a mne.raw object (potentially). For most manufacturers data this is simply scan.raw_file, however for KIT data we need the mrk file, elp and hsp files, so having all these returned by a single function could be useful.

• Return text lists of the projects, subjects etc somehow (.subject_list maybe?)

• Write actual API docs which are auto-generated (and are on the gh-docs branch?)