You have a JSON shaped problem, and now you want to write a short query for getting something out of it.
1{
2 "a": [
3 {
4 "x": 100
5 },
6 {
7 "x": 200
8 }
9 ]
10}
RFC 6901 introduces JSON Pointer:
tldr: / separated key names, arrays are 0 indexed.
No special syntax for all array elements etc...
# 200
/a/1/x
Not very flexible, but it is used in JSON Patch: RFC 6902
JSON Path is basically XPath for JSON.
tldr: . separated, .. recursive descent, [::] array access, ?() script filter (eg on child attributes)
# 100, 200
$..x
$.a[:].x
# [{ "x": 100 }]
$.a[?(@.x == 100)]
tldr: . separated, no recursive descent(?), [::] array access, | pipe and use functions.
# 100, 200
a[*].x
# { "x": 100 }
a[?x == 100]
jq
is a more general purpose utility that can do advanced queries
and reshape data as necessary.
tldr: . separated, .. recursive descent, [] array access, | pipe and use functions for more.
# 100, 200
..x
.a[].x
# { "x": 100 }
.a[] | select(.x == 100)
cel (spec) is a minimal language for evaluating single expressions. With macros it becomes viable to query nested data structures.
# 100, 200
o.a.map(n, n.x)
# { "x": 100 }
o.a.filter(n, n.x == 100)