I have dataset of documents and also hierarchy or users. For simplicity of reproducible example each user belongs to a team, and each team belongs to a department. Some of documents in my dataset have different visibility rules and I want to reproduce them in the search engine. For example: • Article with list of employees, available for the whole company • Article with internal department rules, available for

department1

(i.e. for all teams in this department) • Article with internal team’s drafts, available for

team1

My first idea was to use arrays for such visibility fields and then fullfill them with corresponding values. If visibility chosen as

department1

, we will add corresponding teams too. Solution looks like this Problems with this solution are: • Some people can be outside teams or departments. For example CEO or other leaders. • We need to update index every time when new teams are created, otherwise they will get empty results (e.g. newcomer team5 won’t get even publicly available documents)

Some people can be outside teams or departments. For example CEO or other leaders.

You could add them to an internal admin “system” team of sorts, that is not exposed to your end users

We need to update index every time when new teams are created, otherwise they will get empty results (e.g. newcomer team5 won’t get even publicly available documents)

You could create another attribute called

is_public: true|false

in each document

Could you please check the latest gist? I got only

"available for all"

document, but expected

"available for department1 (so for all nested teams)"

too, because

departments

contains

department1

🤔

Sorry, my bad. Had outdated docker build

Now it’s working as it should be! Thank you for the OR, it’s lifesaving