Snowflake querying items of list of jsons - sql

Question

I'm looking to query all items inside an array of jsons, similar to what Elasticsearch is doing with it's nested data type.
For example, every row in my table looks like this, where a class has a column named students with a list of students:
{
"class": "history"
"students": [
{"first_name": "joe", last_name: "doe", "age": 16},
{"first_name": "tony", last_name: "helen", "age": 10},
{"first_name": "erica", last_name: "kran", "age": 12},
]
}
{
"class": "math"
"students": [
{"first_name": "joe", last_name: "no", "age": 16},
{"first_name": "yo", last_name: "wha", "age": 18},
{"first_name": "dan", last_name: "test", "age": 12},
]
}
I want to make a query that can query inside each items in the list, for example:
Get all classes, where there is a student named joe who is over the age of 14, and there is a student named erica under the age of 14.
This query should only return the "history" class, since it's the only class that has a student with these properties.
EDIT: editing my example data to illustrate the problem:
Looking for classes that have a student named joe over 14 AND a student named erica under 14.
Only the first class should match, since it satisfies both conditions, the second class does not because it doesnt have erica, although there is a joe.

Answer 1

The equivalent with pure SQL and a flatten would be:
with data as (
select parse_json('{
"class": "history",
"students": [
{"first_name": "joe", last_name: "doe", "age": 16},
{"first_name": "tony", last_name: "helen", "age": 10},
{"first_name": "erica", last_name: "kran", "age": 17},
]
}') x
union all select parse_json('{
"class": "math",
"students": [
{"first_name": "joe", last_name: "no", "age": 12},
{"first_name": "yo", last_name: "wha", "age": 18},
{"first_name": "dan", last_name: "test", "age": 12},
]
}')
)
select x:class::string
from data, table(flatten(x:students)) y
where (y.value:first_name='joe' and y.value:age>=14)
or (y.value:first_name='erica' and y.value:age<=14)
;
"history"