I was made admin on an OpenNeuro dataset and asked to update the dataset (first time doing this) to fix a validation error. I set up the CLI, downloaded the dataset, edited one file to fix the validation error, and re-uploaded. Precisely:
uvx openneuro-py@latest download --dataset=ds003104 # download
deno run jsr:@openneuro/cli login # login
nano ds003104/sub-01/meg/sub-01_coordsystem.json # edit the file to fix the validation error
deno run jsr:@openneuro/cli upload --dataset ds003104 ds003104/ # re-upload
Then, when in the “draft” view, dozens of files that I didn’t touch now show hash mismatches. (Meanwhile, I realized it is possible through the web UI to manually update a single file — what I would have done in the first place, if I’d known it was an option). I was hoping to reset the draft state to the state of the latest snapshot, so I could do the single-file-manual-update.
Questions:
- is there a way to reset the staging area / draft state via the CLI? If not, is there some other way to do it?
- any idea why I saw all those hash mismatches?
NB: After walking away for an hour and coming back, all those hash mismatches are gone and I was able to update the dataset 
. Still would like to know the answers to those questions though, if anyone has insight.