So, here is a first example of how to use json_extract
. First, the data is a inserted in a bit different way:
insert into user (name, phone) values("oz", json('{"cell":"+491765", "home":"+498973"}'));
Now, we can select all the users phone numbers as in normal sql:
sqlite> select user.phone from user where user.name=='oz';
{"cell":"+491765","home":"+498973"}
sqlite>
But, what if we don't care about land lines and we want only cell phones?
Enter json_extract
:
sqlite> select json_extract(user.phone, '$.cell') from user;
+491765
And this is how to use json_extract
.
Using json_set
is similar. Given that the we want to update the cell phone:
sqlite> select json_set(user.phone, '$.cell', 123) from \
user;
{"cell":123,"home":"+498973"}
You can combine those function calls in other SQL queries. Thus, you can
use SQLite with structured data and with unstructured data in the form of
JSON.
Here is how to update the user cell phone only:
sqlite> update user
...> set phone =(select json_set(user.phone, '$.cell', 721) from user)
...> where name == 'oz';
sqlite> select * from user;
oz|{"cell":721,"home":"+498973"}