I'm trying to get the distinct sample_id values from a PostgreSQL database table. This would be simple besides the fact that they're embedded in a JSONB field.
Here's what my database looks like when I run select * from event limit 10:
id | step | event_date | event_status | payload
----+------+------------+--------------+-----------------------------
1 | 1 | 2018-01-10 | success | {"x": "y", "sample_id": [0]}
2 | 2 | 2018-01-12 | success | {"x": "y", "sample_id": [0]}
3 | 3 | 2018-01-14 | success | {"x": "y", "sample_id": [0]}
4 | 1 | 2018-01-13 | success | {"x": "y", "sample_id": [1, 38, 63]}
5 | 2 | 2018-01-15 | success | {"x": "y", "sample_id": [1]}
6 | 3 | 2018-01-17 | pending | {"x": "y", "sample_id": [1]}
7 | 1 | 2018-01-16 | success | {"x": "y", "sample_id": [2]}
8 | 2 | 2018-01-18 | success | {"x": "y", "sample_id": [2, 55]}
9 | 3 | 2018-08-20 | success | {"x": "y", "sample_id": [2]}
10 | 1 | 2018-01-19 | success | {"x": "y", "sample_id": [3]}
For the following output I should be getting following data back: [0, 1, 2, 3, 38, 55, 63] (sorted order doesn't really matter).
I've looked into using the JSONB field functions but having no luck. Any idea how to construct such a query?