The user can right now either add comments via the DAEphys setvariable or enter them directly in the comment notebook. The serialization for the setvariable is
which makes it machine readable iff the comment does not have EOLs.
When the user enters text into the notebook directly, we don't enforce anything and all bets are off.
We do store the comment in the NWB in /general/user_comment and in the LBN
Ideas:
Allow to define an input comment format with strings/variable and names. This would be done in JSON and from that we would build up a panel. Maybe one can leverage some ideas from https://www.sciencedirect.com/science/article/pii/S2352711018300505 without getting totally lost in details. Plain text entries are still supported of course. Even with EOLs.
With that input panel we can then store the notebook contents machine readable in JSON. Or in JSONL, see https://jsonlines.org, if we want to have entries for each sweep. We already use JSONL for the logfiles.
timjarsky
commented
1 year ago
The user can right now either add comments via the DAEphys setvariable or enter them directly in the comment notebook. The serialization for the setvariable is
https://github.com/AllenInstitute/MIES/blob/10f9a784e984475e0f6d733ba315d49cdd571168/Packages/MIES/MIES_DAEphys.ipf#L4297
which makes it machine readable iff the comment does not have EOLs. When the user enters text into the notebook directly, we don't enforce anything and all bets are off.
We do store the comment in the NWB in /general/user_comment and in the LBN
Ideas: