JQ command for JSON Processing

Installation on Ubuntu

• apt-get install jq

Some Filter

• jq ‘.’ file.json: print whole json file as . is root

• echo ‘{“fruit”:{“name”:”apple”,”color”:”green”,”price”:1.20}}’ | jq ‘.fruit’: accessing fruit key

• echo ‘{“fruit”:{“name”:”apple”,”color”:”green”,”price”:1.20}}’ | jq ‘.fruit.name’: accessing name of fruit

• jq ‘.fruit.color,.fruit.price’ fruit.json: accessing multiple keys

• echo ‘{ “with space”: “hello” }’ | jq ‘.”with space”‘: key with space

• echo ‘[“x”,”y”,”z”]’ | jq ‘.[]’: iterate over an array

• jq ‘.[] | .name’ fruits.json or jq ‘.[].name’ fruits.json: iterate over an array and access name key of each object

• jq ‘.[1].price’ fruits.json: access 1 object in the array

• echo ‘[1,2,3,4,5,6,7,8,9,10]’ | jq ‘.[6:9]’: slicing to return subarray of an array, first index is inclusive, last index is exclusive which means from item 6-8

• echo ‘[1,2,3,4,5,6,7,8,9,10]’ | jq ‘.[:6]’: this is equal to [0:6]

• jq ‘.[-2:]’: start from 2 item from the end until the end of the array

• jq ‘.fruit | keys’ fruit.json: get keys of objects

• jq ‘.fruit | length‘ fruit.json: return length or number of properties in an object

• jq ‘.fruit.name | length‘ fruit.json: get length of string

• jq ‘map(has(“name”))‘ fruit.json: check if there’s name property

• jq ‘map(.price+2)‘ fruit.json: do operation to an elements

• jq ‘[.[].price] | min‘ fruits.json or jq ‘[.[].price] | max‘: get min|max value

• jq ‘.[] | select(.price>0.5)‘ fruits.json: select item with criteria

• jq ‘.[] | select(.color==”yellow”)‘ fruits.json

• jq ‘.[] | select(.color==”yellow” and .price>=0.5)‘ fruits.json

• jq ‘.[] | select(.name|test(“^a.”)) | .price’ fruits.json: Using regular expression to check condition: get price of all item whose name start with letter a

• jq ‘map(.color) | unique‘: find unique colors, map function create another array and pass it to unique function

• jq ‘del(.fruit.name)‘ fruits.json: delete a key and value

• jq ‘.query.pages | [.[] | map(.) | .[] | {page_title: .title, page_description: .extract}]‘ wikipedia.json: transforming from a large json data to a customized one which just get 2 properties which are title and description

• jq -r '.results[].series[] | .columns, .values[] | @csv' bgp.txt: convert json to csv

• jq -r '.results[].series[] | .columns, .values[] | @tsv' bgp.txt: convert json to tsv (tab)

• jq ".result[] | {id,name}": list result, filter id and name fields

  • jq '.result[] | "(.id) (.name)"': similarly, but output in plain text

• Check if value in a list

# file autoconnect.json
{
    "auto": 1,
    "networks": ["NETA", "NETB"]
}

# file test.sh
net="NETA"
file="autoconnect.json"
if jq -e --arg net "$net" '.networks | index($net)' $file > /dev/null; then
    echo "$net exists in networks!"
else
    echo "$net does not exist in networks."
fi

# Explanation
- index($net) → Checks if $net exists in the "networks" array
- -e → Makes jq return success (0) or failure (1)
- Redirecting output (> /dev/null) → Suppresses unnecessary output

Reference

• https://www.baeldung.com/linux/jq-command-json

• https://stedolan.github.io/jq/manual/#Basicfilters