As input I have:
[
{
"backup": [
{
"timestamp": { "start": 1642144383, "stop": 1642144386 },
"info": { "size": 1200934840},
"type": "full"
},
{
"timestamp": {"start": 1642144388, "stop": 1642144392 },
"info": { "size": 1168586300
},
"type": "incr"
}
],
"name": "dbname1"
},
{
"backup": [
{
"timestamp": { "start": 1642144383, "stop": 1642144386 },
"info": { "size": 1200934840},
"type": "full"
},
{
"timestamp": {"start": 1642144388, "stop": 1642144392 },
"info": { "size": 1168586300
},
"type": "incr"
}
],
"name": "dbname2"
}
]
using
jq '.[]
| [ .backup[] + {name} ]
| max_by(.timestamp.stop)
'
(thanks @pmf) I can re-order this to
{
"timestamp": {
"start": 1642144388,
"stop": 1642144392
},
"info": {
"size": 1168586300
},
"type": "incr",
"name": "dbname1"
}
{
"timestamp": {
"start": 1642144388,
"stop": 1642144392
},
"info": {
"size": 1168586300
},
"type": "incr",
"name": "dbname2"
}
Selected is the dict containing the max timestamp and the name added to it, being the last created backup of a database. There are multiple databases possible. How can I form the output to a cleanly formatted array? I was hoping for
[
{
"timestamp": {
"start": 1642144388,
"stop": 1642144392
},
"info": {
"size": 1168586300
},
"type": "incr",
"name": "dbname1"
},
{
"timestamp": {
"start": 1642144388,
"stop": 1642144392
},
"info": {
"size": 1168586300
},
"type": "incr",
"name": "dbname2"
}
]
And yes, I can add this using sed but I feel jq should be able to do this. So the question is how can should this be written?
Instead of .[] | …
use map(…)
to retain the array.
jq 'map([.backup[] + {name}] | max_by(.timestamp.stop))'
[
{
"timestamp": {
"start": 1642144388,
"stop": 1642144392
},
"info": {
"size": 1168586300
},
"type": "incr",
"name": "dbname1"
},
{
"timestamp": {
"start": 1642144388,
"stop": 1642144392
},
"info": {
"size": 1168586300
},
"type": "incr",
"name": "dbname2"
}
]