Is it possible to query (join) 3 or more tables?

[i4git]

I have been able to get simple 2 table(resource) inner joins to work, but is there a way to join 3 or more tables? Here is an example of of the join statement I have been using:

  "joins": [
    {
      "resource": "b",
      "condition": {
        "resource": "a",
        "property": "service_id",
        "operator": "=",
        "value": {
          "resource": "b",
          "property": "srvc_id"
        }
      }
    }
  ]

[janette]

@i4git it is possible if left joins fits your use case, currently you can not define the direction of the join. Here is an example based on datastore tables from the demo site:

POST https://demo.getdkan.org/api/1/datastore/query
content-type: application/json


{
    "resources": [
        {
          "id": "cd4dff20-432c-5fe7-94fb-513e17c33a29",
          "alias": "a"
        },
        {
          "id": "31c0fbaa-36c3-53cc-8088-b8408d344d25",
          "alias": "b"
        },
        {
          "id": "fcff6691-3332-55d5-93e4-ee1efc92fd65",
          "alias": "c"
        }
    ],
    "properties": [
        {
           "resource": "a",
           "property": "price"
        },
        {
          "resource": "b",
          "property": "year"
        },
        {
          "resource": "c",
          "property": "state"
        }
    ],
    "conditions": [
      {
        "resource": "a",
        "property": "record_number",
        "value": 1,
        "operator": ">"
      }
    ],
    "joins": [
      {
        "resource": "b",
        "condition": {
          "resource": "a",
          "property": "record_number",
          "operator": "=",
          "value": {
            "resource": "b",
            "property": "record_number"
          }
        }
      },
      {
        "resource": "c",
        "condition": {
          "resource": "a",
          "property": "record_number",
          "operator": "=",
          "value": {
            "resource": "c",
            "property": "record_number"
          }
        }
      }
    ],
    "limit": 3
}

The result is:

"results": [
    {
      "price": "34.73",
      "year": "2001",
      "state": "Alaska"
    },
    {
      "price": "34.73",
      "year": "2002",
      "state": "Arizona"
    },
    {
      "price": "34.73",
      "year": "2003",
      "state": "Arkansas"
    }
  ],