If you have a map of nested maps, because I have recently struggled myself with a query, I'd like to show you how you can update each of the nested maps in a consistent fashion, using multiple approaches. Your input, a single map, may resemble the following:

[ShaneMellwig]

If you have a map of nested maps, because I have recently struggled myself with a query, I'd like to show you how you can update each of the nested maps in a consistent fashion, using multiple approaches. Your input, a single map, may resemble the following:

apple:
  kind: fruit
  sugar:
    min: 0.01
    max: 0.37
  edible: true
potato:
  kind: vegetable
  starch:
    min: 0.15
    max: 0.46
  edible: true
acorn:
  kind: nut

Say you want to add one or several properties to each of the maps (keyed apple, potato and acorn), the shortest expression is probably something like:

Approach 1

with_entries(.value += { "planet": "Earth", "season": "fall" })

With the above, the map expressed with { "planet": "Earth", "season": "fall" } is merged with every nested map (bound to . inside with_entries):

apple:
  kind: fruit
  sugar:
    min: 0.01
    max: 0.37
  edible: true
  planet: Earth
  season: fall
potato:
  kind: vegetable
  starch:
    min: 0.15
    max: 0.46
  edible: true
  planet: Earth
  season: fall
acorn:
  kind: nut
  planet: Earth
  season: fall

Approach 1.1

There's at least one variation on the above that produces equivalent result:

with_entries(.value.planet = "Earth" | .value.season = "fall")

Approach 1.2

The above, can be further shortened to:

with_entries(with(.value; .planet = "Earth" | .season = "fall"))

The way I understand it, with_entries is implemented in a manner where it's equivalent to:

[ to_entries.[] | with(.value; .planet = "Earth" | .season = "fall") ] | from_entries

Approach 2

with(.[]; .planet = "Earth" | .season = "fall")

Remember the importance of operator precedence rules, assuming you only wanted to add one of the properties, planet, the below snippet will not yield the expected result:

.[] | .planet = "Earth"

On the other hand, the following amendment on the above does work:

(.[] | .planet) = "Earth"

I hope all this is useful to someone.

Originally posted by @amn in #2233

[ShaneMellwig]

If you have a map of nested maps, because I have recently struggled myself with a query, I'd like to show you how you can update each of the nested maps in a consistent fashion, using multiple approaches. Your input, a single map, may resemble the following:

apple:
kind: fruit
sugar:
min: 0.01
max: 0.37
edible: true
potato:
kind: vegetable
starch:
min: 0.15
max: 0.46
edible: true
acorn:
kind: nut
Say you want to add one or several properties to each of the maps (keyed apple, potato and acorn), the shortest expression is probably something like:

Approach 1

with_entries(.value += { "planet": "Earth", "season": "fall" })

With the above, the map expressed with { "planet": "Earth", "season": "fall" } is merged with every nested map (bound to . inside with_entries):

apple:
kind: fruit
sugar:
min: 0.01
max: 0.37
edible: true
planet: Earth
season: fall
potato:
kind: vegetable
starch:
min: 0.15
max: 0.46
edible: true
planet: Earth
season: fall
acorn:
kind: nut
planet: Earth
season: fall

Approach 1.1

There's at least one variation on the above that produces equivalent result:

with_entries(.value.planet = "Earth" | .value.season = "fall")

Approach 1.2

The above, can be further shortened to:

with_entries(with(.value; .planet = "Earth" | .season = "fall"))

The way I understand it, with_entries is implemented in a manner where it's equivalent to:

[ to_entries.[] | with(.value; .planet = "Earth" | .season = "fall") ] | from_entries

Approach 2

with(.[]; .planet = "Earth" | .season = "fall")

Remember the importance of operator precedence rules, assuming you only wanted to add one of the properties, planet, the below snippet will not yield the expected result:

.[] | .planet = "Earth"

On the other hand, the following amendment on the above does work:

(.[] | .planet) = "Earth"

I hope all this is useful to someone.

Originally posted by @amn in #2233
Reply