You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
So in the softDelete function the params.provider is inherited from the original method (remove), but it should be an internal call like it was in previous versions (feathers-plus) of softDelete, this was done by setting the provider inside of softDelete to undefined. This is better for permissions etc, cause for example people are allowed to delete (and so call the remove hook) but aren't allowed to patch.
Now the question is how will this be changed, are we adding that back by default like something:
I think the removeParams is more flexible in my opinion but it is also a bit double cause we have a deletedQuery, what is technically inside the params.query.
Or maybe we should add an removeProvider function, any thoughts?
Depending whats the outcome of this is I can make a PR for either implementation.
The text was updated successfully, but these errors were encountered:
feathers-hooks-common/lib/services/soft-delete.js
Line 37 in f91393d
So in the
softDelete
function theparams.provider
is inherited from the original method (remove), but it should be an internal call like it was inprevious versions(feathers-plus) ofsoftDelete
, this was done by setting the provider inside ofsoftDelete
toundefined
. This is better for permissions etc, cause for example people are allowed to delete (and so call the remove hook) but aren't allowed to patch.Now the question is how will this be changed, are we adding that back by default like something:
Or are we going to add it with a property option like:
And merge it just like how we do with the query.
I think the
removeParams
is more flexible in my opinion but it is also a bit double cause we have adeletedQuery
, what is technically inside theparams.query
.Or maybe we should add an
removeProvider
function, any thoughts?Depending whats the outcome of this is I can make a PR for either implementation.
The text was updated successfully, but these errors were encountered: