Jq

jq is a lightweight and flexible command-line JSON processor. jq is like sed for JSON data - you can use it to slice and filter and map and transform structured data with the same ease that sed, awk, grep, and friends let you play with text.

Example usage

$ cat azones.json

{
    "availabilityZoneInfo": [
        {
            "hosts": {
                "node-1.example.com": {
                    "nova-compute": {
                        "active": true,
                        "available": true
                    }
                },
                "node-2.example.com": {
                    "nova-compute": {
                        "active": true,
                        "available": true
                    }
                }
            },
            "zoneName": "az1",
            "zoneState": {
                "available": true
            }
        },
        {
            "hosts": {
                "node-3.example.com": {
                    "nova-compute": {
                        "active": true,
                        "available": true
                    }
                },
                "node-4.example.com": {
                    "nova-compute": {
                        "active": true,
                        "available": true
                    }
                }
            },
            "zoneName": "az2",
            "zoneState": {
                "available": true
            }
        }
    ]
}

• Capture just the availability zone names:

$ cat azones.json | jq '[.availabilityZoneInfo[] | .zoneName]'

[
  "az1",
  "az2"
]

Or, for compact instead of pretty-printed output:

$ cat azones3.json | jq -c '[.availabilityZoneInfo[] | .zoneName]'
["az1","az2"]

• Capture just the hostname (e.g., "node-1.example.com") key for availability zone "az1":

$ cat azones.json | jq '[.availabilityZoneInfo[] | select(.zoneName == "az1") | {hosts: .hosts|keys}]'

[
  {
    "hosts": [
      "node-1.example.com",
      "node-2.example.com"
    ]
  }
]

Or, for a more script-friendly output:

$ cat azones.json | jq -cM '[.availabilityZoneInfo[] | select(.zoneName == "az1") | {hosts: .hosts|keys}]' | sed -e 's/["}\[]//g;s/\]//g;s/{hosts://g;s/,/ /g'
#~OR~
$ foo=($(cat azones3.json | jq -cM '[.availabilityZoneInfo[] | select(.zoneName == "az1") | {hosts: .hosts|keys}]' | sed -e 's/["}\[]//g;s/\]//g;s/{hosts://g;s/,/ /g'))
$ echo ${foo[0]} #=> node-1.example.com

• Get just the raw values:

$ echo '{ "packet_loss": [ {"ips": "10.0.0.10 10.0.0.11 10.0.0.12", "node-17": "3/3" }] }' | jq -r '[.packet_loss[] | .ips] | .[]'
10.0.0.10 10.0.0.11 10.0.0.12

Practical example

Here is how to print out all the OpenStack compute nodes in my example environment:

#!/bin/bash
# AUTHOR: Christoph Champ <christoph.champ@gmail.com>
# Requires jq 1.5+
JQ=$(which jq)

OS_AUTH_URL=http://1.2.3.4:5000/v2.0/
OS_TENANT_NAME=admin
OS_USERNAME=admin
OS_PASSWORD=admin

INFO=$(curl -sXPOST "${OS_AUTH_URL}/tokens" \
        -H "Content-Type: application/json" \
        -d "{\"auth\":{\"tenantName\":\"$OS_TENANT_NAME\",\"passwordCredentials\":\
        {\"username\":\"$OS_USERNAME\",\"password\":\"$OS_PASSWORD\"}}}" | \
        ${JQ} -crM '[.access.token.id + "," + (.access.serviceCatalog[] | select(.name == "nova") | .endpoints[].publicURL)] | .[]')

TOKEN=${INFO%%,*}
NOVA_ENDPOINT=${INFO#*,}

IGNORE_ZONES="internal|nova"

raw=$(curl -s -H "X-Auth-Token: ${TOKEN}" "${NOVA_ENDPOINT}/os-availability-zone/detail" | \
    ${JQ} -crM '[.availabilityZoneInfo[].zoneName] | .[]' | \
    grep -vE "(${IGNORE_ZONES})" | tr '\n' ',')

IFS=',' read -r -a zones <<< "${raw%,}"

for zone in "${zones[@]}"; do
    raw=($(curl -s -H "X-Auth-Token: ${TOKEN}" "${NOVA_ENDPOINT}/os-availability-zone/detail" | \
        ${JQ} --arg zone "$zone" '[.availabilityZoneInfo[] | select(.zoneName==$zone) | .hosts|keys] | .[]' | \
        tr -d '[]",' | sed '/^$/d' | tr '\n' ',' | tr -d ' '))

    IFS=',' read -r -a nodes <<< "${raw%,}"
    for node in "${nodes[@]}"; do
        echo "node: $zone $node"
    done
done

Running the above script produces the following output:

node: az1 node-1.example.com
node: az1 node-2.example.com
node: az2 node-3.example.com
node: az2 node-4.example.com

Append to JSON

• Example of how to append key/values to an already existing JSON structure:

$ cat foo.json 
{ "name": "bob", "age": 30 }

$ cat foo.json | jq 'to_entries'
[
  {
    "key": "name",
    "value": "bob"
  },
  {
    "key": "age",
    "value": 30
  }
]

$ cat foo.json | BEARERTOKEN="Bearer abc123" jq 'to_entries | . + [{"key":"routes","value":[{"path":"api/v1","url":"http://example.com","headers":[{"name":"Authorization","content":env.BEARERTOKEN}]}]}] | from_entries'
{
  "name": "bob",
  "age": 30,
  "routes": [
    {
      "path": "api/v1",
      "url": "http://example.com",
      "headers": [
        {
          "name": "Authorization",
          "content": "Bearer abc123"
        }
      ]
    }
  ]
}

Update a specific nested value in a JSON file

$ export NEW_URL="https://172.x.x.x:6443"
$ jq --arg new_url "${NEW_URL}" '(.resources[] | select(.type == "rke_cluster") | .instances[].attributes.api_server_url) |= $new_url' foo.json
{
  "resources": [
    {
      "module": "module.rancher",
      "type": "rke_cluster",
      "instances": [
        {
          "attributes": {
            "api_server_url": "https://172.x.x.x:6443",
            "foo": "bar"
          }
        }
      ]
    }
  ]
}

Miscellaneous

• Get the randomly generated Rancher admin password, as created by the rancher2 Terraform provider:

$ jq -crM '.resources[] | select(.provider == "module.rancher.provider.rancher2.bootstrap") | {instances: .instances[]|.attributes.current_password} | .[]' terraform.tfstate

External links

• Official website
• Bash For Look Over JSON Array Using Jq
• Jsonnet — an extension of JSON