POST /data/.../:revision-id/changes
GET /data/.../:revision-id/changes/:change-id
POST /data/.../:revision-id/deletes
(And I'm about to add a POST .../corrections)
Each change has a 'kind' associated: append/retraction/correction.
Each change gets an URI matching /data/.../:revision-id/changes/:change-id (note the 'changes' part, the URI no longer carries information about the 'kind').
This PR renames two routes to be consistent with our internal naming (:datahost.change/kind -> #{:dh/ChangeKindAppend :dh/ChangeKindRetract}):
POST /data/.../:revision-id/appends
POST /data/.../:revision-id/retractions
This would leave us with GET .../changes. The only issue is we don't allow getting the metadata of a change (only CSVs at the moment), so querying that endpoint won't tell us whether the returned CSV is an append, retraction or correction. The user is expected to get that information via another call (see here)
We have the following routes
(And I'm about to add a
POST .../corrections)/data/.../:revision-id/changes/:change-id(note the 'changes' part, the URI no longer carries information about the 'kind').This PR renames two routes to be consistent with our internal naming (
:datahost.change/kind->#{:dh/ChangeKindAppend :dh/ChangeKindRetract}):This would leave us with
GET .../changes. The only issue is we don't allow getting the metadata of a change (only CSVs at the moment), so querying that endpoint won't tell us whether the returned CSV is an append, retraction or correction. The user is expected to get that information via another call (see here)