I'm trying to use the JQ command to filter a json object to selectively extract keys. Here's a sample object that I've placed in file x.txt:
{
"activities" : {
"-KSndgjqvmQkWVKHCpLh" : {
"create_device" : "...",
"stop_time_utc" : "2016-11-01T23:08:08Z"
},
"-KSoBSrBh6PZcjRocGD7" : {
"create_device" : "..."
},
"-KSptboGjo8g4bieUbGM" : {
"create_device" : "...",
"stop_time_utc" : "2017-01-17T23:08:08Z"
}
}
}
The following command can extract all of the activity keys:
cat x.txt | jq '.activities | keys'
[
"-KSndgjqvmQkWVKHCpLh",
"-KSoBSrBh6PZcjRocGD7",
"-KSptboGjo8g4bieUbGM"
]
I've been googling and experimenting for a few hours trying to filter the object to select only the activity entries that have a stop_time_utc value, and use something like a "select(.stop_time_utc | fromdateiso8601 > now)" to only pick activities that have expired. For example, I'd like to use filters to create an array from the sample object with only the one relevant entry:
[
"-KSndgjqvmQkWVKHCpLh"
]
Is attempting this with the keys option the wrong route? Any ideas or suggestions would be much appreciated.
The filter with_entries/1
is your friend, e.g.:
.activities | with_entries( select(.value | has("stop_time_utc") ) )
produces:
{
"-KSndgjqvmQkWVKHCpLh": {
"create_device": "...",
"stop_time_utc": "2016-11-01T23:08:08Z"
},
"-KSptboGjo8g4bieUbGM": {
"create_device": "...",
"stop_time_utc": "2017-01-17T23:08:08Z"
}
It's now easy to add additional selection criteria, extract the key names of interest, etc. For example:
.activities
| with_entries( select( (.value.stop_time_utc? | fromdateiso8601?) < now ) )
| keys