Folksonomy + refresh, offline, background tasks

[monsieurtanuki]

Problem

The new Folksonomy feature implemented by @gciluffo in #5768 needs to take into account several aspects of offline mode.
For instance, being able to add a folksonomy key/value with no access to internet.

Proposed solution

Using background tasks would make sense.

[monsieurtanuki]

@gciluffo Do you feel like working on it, or do you prefer to recover first from that intense PR?
TBH it'd be better if I work on it - but you could review the PR.

[monsieurtanuki]

This is what needs to be implemented:

1. Refresh 1 product - when we refresh an OxF product, we should refresh its folksonomies, and there's an API to retrieve all of them
2. Refresh several products - when we refresh a list of OxF products, we should refresh all their folksonomies. Unfortunately there's no API to retrieve all folksonomies of a list of products
3. Store server data locally - in a dedicated SQFLite table, in order to access data even offline, and that's where we'll put the "refreshed" data mentioned above
4. Store transient data locally - in a dedicated Hive table. When we add/update/delete a folksonomy, we use a background task, and until it's completed we consider a transient folksonomy on top of the server values
5. Background tasks - for add/create/delete folksonomy operations. They start with setting the transient value, then they run the API operation if possible, then they refresh the data from the server and remove the transient data