RethinkDB Python递归文档过滤器

Question

Trying to do a recursive filter query on RethinkDB with they Python wrapper. Having a lot of trouble getting it work.

Tried a lot of variations of the query, to no avail. Essentially, I'm trying to find the rows which do not have a document nested under label with a particular user_id.

In plain english: if the current user already labeled this example, don't return it to them again.

My non-working query:

open_tasks = rdbt \
    .order_by(index=r.desc('labels_completed')) \
    .filter(r.row['locked'] == False) \
    .filter(lambda task:
            task['labels']['user_id'] != current_user.id) \
    .limit(qty) \
    .run(conn)

My dataset

[
    {
        "id": "e54893b4-b1d0-49c5-b6aa-9aa9e7d2b73b",
        "image": "https://dl.dropboxusercontent.com/u/5822/crowdlabeler/ABLXOTODWJKTXECYZTST.jpg",
        "labels": [
            {
                "account_qty_labeled": 54,
                "account_signup_date": "Tue Aug 04 2015 10:12:25 GMT-04:00",
                "compensation": 0.01,
                "dataset_id": 144,
                "label": {
                    "$$hashKey": "object:45",
                    "answer": "Yes",
                    "selected": true
                },
                "label_duration_sec": 3,
                "labeled_at": "Wed Aug 05 2015 16:26:04 GMT-05:00",
                "sess_duration_sec": 3,
                "sess_qty_labeled": 0,
                "user_id": 1
            }
        ],
        "labels_completed": 0,
        "locked": false,
        "text": "Lorem ipsum dolor sit amet, consectetur adipisicing elit. Cupiditate adipisci vero minus laudantium reprehenderit exercitationem eius, suscipit facilis laboriosam consequuntur, eligendi quis mollitia excepturi deserunt dicta, dolorem quaerat pariatur provident sint explicabo. Magnam possimus dolorum beatae quidem excepturi quibusdam dolore reprehenderit accusantium quae ad libero, voluptatum laborum, incidunt, voluptate reiciendis."
    },
    {
        "id": "9f08869e-79fd-49c0-a184-c43d2a1c95cf",
        "image": "https://dl.dropboxusercontent.com/u/5822/crowdlabeler/ACSGHDYECQWQXDHIOBYC.jpg",
        "labels": [],
        "labels_completed": 0,
        "locked": false,
        "text": "Lorem ipsum dolor sit amet, consectetur adipisicing elit. Cupiditate adipisci vero minus laudantium reprehenderit exercitationem eius, suscipit facilis laboriosam consequuntur, eligendi quis mollitia excepturi deserunt dicta, dolorem quaerat pariatur provident sint explicabo. Magnam possimus dolorum beatae quidem excepturi quibusdam dolore reprehenderit accusantium quae ad libero, voluptatum laborum, incidunt, voluptate reiciendis."
    },
    {
        "id": "9fba0a39-4cfd-4a97-b48f-e8bf2b0d46c5",
        "image": "https://dl.dropboxusercontent.com/u/5822/crowdlabeler/ADMNIUYKUHAIOHMAFXBK.jpg",
        "labels": [],
        "labels_completed": 0,
        "locked": false,
        "text": "Lorem ipsum dolor sit amet, consectetur adipisicing elit. Cupiditate adipisci vero minus laudantium reprehenderit exercitationem eius, suscipit facilis laboriosam consequuntur, eligendi quis mollitia excepturi deserunt dicta, dolorem quaerat pariatur provident sint explicabo. Magnam possimus dolorum beatae quidem excepturi quibusdam dolore reprehenderit accusantium quae ad libero, voluptatum laborum, incidunt, voluptate reiciendis."
    }
]

Thank you for your help!

Answer 1

This was a tough one but the solution ended up being two-fold.

A slight modification to the schema when saving, then using RethinkDB's contains method.

Here is the modified query that works well.

open_tasks = rdbt \
    .order_by(index=r.desc('labels_completed')) \
    .filter(r.row['locked'] == False) \
    .filter(lambda task:
            r.not_(task['labeler_ids'].contains(current_user.id))) \
    .limit(qty) \
    .run(conn)