JSON merge arrays and UNIQUE or DISTINCT - mysql

Question

In my MySQL database I have a table features with this structure:
+-------+---------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-------+---------+------+-----+---------+----------------+
| id | int(11) | NO | PRI | NULL | auto_increment |
| val | json | NO | | NULL | |
+-------+---------+------+-----+---------+----------------+
When I select everything from it, like so, I get the following:
mysql> select * from features;
+----+----------------------------------+
| id | val |
+----+----------------------------------+
| 1 | ["apple", "banana", "orange"] |
| 2 | ["apple", "orange", "pineapple"] |
| 3 | ["orange", "banana"] |
| 4 | [] |
+----+----------------------------------+
The value in the val column should always be an array of strings. This array can have any length (>= 0).
The question is:
How can I select all those array values in a single result set, not repeated? So that I get this result and use it in PHP:
+------------+
| arr_values |
+------------+
| apple |
| banana |
| orange |
| pineapple |
+------------+
The only constraint to solve this is that it should be compatible with MySQL v5.7.

Answer 1

If maximal amount of elements per JSON value is limited then (an example for not more than 10 elements)
SELECT DISTINCT JSON_EXTRACT(features.val, CONCAT('$[', numbers.num, ']')) arr_values
FROM features, ( SELECT 0 num UNION ALL
SELECT 1 UNION ALL
SELECT 2 UNION ALL
SELECT 3 UNION ALL
SELECT 4 UNION ALL
SELECT 5 UNION ALL
SELECT 6 UNION ALL
SELECT 7 UNION ALL
SELECT 8 UNION ALL
SELECT 9 ) numbers
HAVING arr_values IS NOT NULL;
If really the max array size is limited nevertheless (for example, 1000000) then it is possible to generate the dynamic table with proper amount of number. But stored procedure with iterational parsing and temporary table is more safe solution.
UPDATE.
Non-limited solution (stored procedure).
CREATE PROCEDURE get_unique ()
BEGIN
CREATE TEMPORARY TABLE temp (val JSON);
INSERT INTO temp
SELECT val
FROM features;
CREATE TEMPORARY TABLE tmp (val JSON);
cycle: LOOP
INSERT IGNORE INTO tmp
SELECT JSON_EXTRACT(val, '$[0]')
FROM temp;
DELETE
FROM temp
WHERE JSON_EXTRACT(val, '$[1]') IS NULL;
UPDATE temp
SET val = JSON_REMOVE(val, '$[0]');
IF 0 = (SELECT COUNT(*)
FROM temp) THEN
LEAVE cycle;
END IF;
END LOOP;
DROP TEMPORARY TABLE temp;
SELECT DISTINCT *
FROM tmp
WHERE val IS NOT NULL;
DROP TEMPORARY TABLE tmp;
END
fiddle