Oracle 12c Json split - sql

Question

This is how I am getting result in Oracle 12c
Id
Start Date Range
End Date Range
1
[ "2019-01-07","2019-02-17","2019-03-17"]
[ "2019-01-14","2019-02-21","2019-03-21"]
And I want it
Id
Start Date Range
End Date Range
1
2019-01-07
2019-01-14
1
2019-02-17
2019-02-21
1
2019-03-17
2019-03-21
Earlier I had asked this question for single-column split and below is the link
How to replace special characters and then break line in oracle
But when I add another column there is the cartesian product.

Answer 1

You can use json_table to extract the strings from the JSON arrays, presumably as actual dates:
select t.id, s.n, s.start_date, e.end_date
from your_table t
cross apply json_table (
t.start_range, '$[*]'
columns
n for ordinality,
start_date date path '$'
) s
join json_table (
t.end_range, '$[*]'
columns
n for ordinality,
end_date date path '$'
) e
on e.n = s.n
The for ordinality clauses provide an index into each array, and the join then matches up the 'related' array entries.
ID | N | START_DATE | END_DATE
-: | -: | :--------- | :--------
1 | 1 | 07-JAN-19 | 14-JAN-19
1 | 2 | 17-FEB-19 | 21-FEB-19
1 | 3 | 17-MAR-19 | 21-MAR-19
If you want string rather than dates for some reason you can just change the data type in the column clauses.
db<>fiddle

Answer 2

You can also do it with regexp_substr and connect by after replacing [,] and " with empty spaces.
Schema and insert statements:
create table testtable(Id int, Start_Date_Range varchar(500), End_Date_Range varchar(500));
insert into testtable values(1 ,'[ "2019-01-07","2019-02-17","2019-03-17"]', '[ "2019-01-14","2019-02-21","2019-03-21"]');
Query:
select distinct id, trim(regexp_substr(replace(replace(replace(Start_Date_Range,'"',''),'[',''),']',''),'[^,]+', 1, level) ) Start_Date_Range,
trim(regexp_substr(replace(replace(replace(end_Date_Range,'"',''),'[',''),']',''),'[^,]+', 1, level) ) End_Date_Range,
level
from testtable
connect by regexp_substr(Start_Date_Range, '[^,]+', 1, level) is not null
order by id, level;
Output:
ID
START_DATE_RANGE
END_DATE_RANGE
LEVEL
1
2019-01-07
2019-01-14
1
1
2019-02-17
2019-02-21
2
1
2019-03-17
2019-03-21
3
db<fiddle here

Answer 3

In a comment to the OP, I pointed out that the data model is not quite right. The values in the two JSON arrays are related; such data should be encoded in a single object, not two. There should be a single array of objects, each object having two members: start date and end date.
To illustrate the data model I am suggesting, I start with a sample input table (with an additional id), I use Alex Poole's answer exactly as is to generate the table the OP asked about, but then I follow that with JSON generating functions to put the data back into JSON format to illustrate how I think the input data should look like. (The provider of the JSON string should create a JSON in this format, rather than sending two separate JSON arrays of strings representing dates).
What I do not show here is how to use a single call to JSON_TABLE to split the data from the single array of objects created at the end of my query. That is a lot simpler than the query to get data out of two separate JSON arrays.
NOTE - this is not really an answer; I wrote it as an answer because it obviously wouldn't fit in a comment.
with
t (id, start_date_range, end_date_range) as (
select 1, '["2019-01-07","2019-02-17","2019-03-17"]',
'["2019-01-14","2019-02-21","2019-03-21"]' from dual union all
select 5, '["2020-04-23","2020-06-15"]',
'["2020-04-30","2020-06-19"]' from dual
)
, shown_as_table(id, n, start_date, end_date) as (
select t.id, s.n, to_char(s.start_date, 'yyyy-mm-dd'),
to_char(e.end_date, 'yyyy-mm-dd')
from t
cross apply json_table (
t.start_date_range, '$[*]'
columns
n for ordinality,
start_date date path '$'
) s
join json_table (
t.end_date_range, '$[*]'
columns
n for ordinality,
end_date date path '$'
) e
on e.n = s.n
)
select id, json_arrayagg(
json_object('start' value start_date, 'end' value end_date)
format json
order by n
) as date_range_array
from shown_as_table
group by id
;
Output:
ID DATE_RANGE_ARRAY
-- -------------------------------------------------------------------------------------------------------------------------------
1 [{"start":"2019-01-07","end":"2019-01-14"},{"start":"2019-02-17","end":"2019-02-21"},{"start":"2019-03-17","end":"2019-03-21"}]
5 [{"start":"2020-04-23","end":"2020-04-30"},{"start":"2020-06-15","end":"2020-06-19"}]