使用 jq 将 json 转换为 csv

Question

I'm trying to come up with the correct jq syntax to convert json to csv.

Desired results:

<email>,<id>,<name>
e.g. 
user1@whatever.nevermind.no,0,general
user2@whatever.nevermind.no,0,general
user1@whatever.nevermind.no,1,local
...

• note that also need to ignore objects with empty "agent_priorities"

Input

[
  {
    "id": 0,
    "name": "General",
    "agent_priorities": {
      "user1@whatever.nevermind.no": "normal",
      "user2@whatever.nevermind.no": "normal"
    }
  },
  {
    "id": 1,
    "name": "local",
    "agent_priorities": {
      "user1@whatever.nevermind.no": "normal"
    }
  },
  {
    "id": 2,
    "name": "Engineering",
  }
]

Answer 1

Store the id and name in variables, then iterate over the keys of agent_priorities :

jq -r '.[] 
       | .id as $id 
       | .name as $name 
       | .agent_priorities 
       | keys 
       | .[] 
       | [., $id, $name ] 
       | @csv
      ' file.json

Answer 2

The following variant of the accepted answer checks for the existence of the "agent_priorities" key as per the requirements, and uses keys_unsorted to preserve the order of the keys:

jq -r '
       .[] 
       | select(has("agent_priorities"))
       | .id as $id 
       | .name as $name 
       | .agent_priorities 
       | keys_unsorted[]
       | [., $id, $name ] 
       | @csv
      ' file.json