Using jq to parse and display multiple fields in a json serially
JsonJqJson Problem Overview
I have this Json
{
"users": [
{
"first": "Stevie",
"last": "Wonder"
},
{
"first": "Michael",
"last": "Jackson"
}
]
}
Using jq I'd like to display first and last name serially. Like so -
Stevie Wonder
Michael Jackson
This is how far I have gotten -
jq '.users[].first, .users[].last'
But it displays
"Stevie"
"Michael"
"Wonder"
"Jackson"
Notice the following:
- The double quotes that I do not want.
- The carriage return that I do not want.
- It's jumbled up. My query displays all the first names first, and then all the last names. However, I want first-last, first-last pair.
Json Solutions
Solution 1 - Json
I recommend using String Interpolation:
jq '.users[] | "\(.first) \(.last)"'
We are piping down the result of .users[]
to generate the string ".first .last" using string interpolation. \(foo)
syntax is used for string interpolation in jq
. So, for the above example, it becomes "Stevie Wonder" (".users[].first .users[].second"
working elementwise) and "Michael Jackson".
jq reference: String interpolation
Solution 2 - Json
You can use addition to concatenate strings.
> Strings are added by being joined into a larger string.
jq '.users[] | .first + " " + .last'
The above works when both first
and last
are string. If you are extracting different datatypes(number and string), then we need to convert to equivalent types. Referring to solution on this question. For example.
jq '.users[] | .first + " " + (.number|tostring)'
Solution 3 - Json
jq '.users[]|.first,.last' | paste - -
Solution 4 - Json
In addition to what others have suggested, I think that two options are worth mentioning.
Print as CSV/TSV
$ cat file.json | jq -r '.users[] | [.first, .last] | @tsv'
Stevie Wonder
Michael Jackson
cat file.json | jq -r '.users[] | [.first, .last] | @csv'
"Stevie","Wonder"
"Michael","Jackson"
The first expression, .users[]
, unnests the objects from the outer-most array as in the code given in the question. The next expression, [.first, .last]
, creates a new array of the values for each input object, and the final expression uses the built-in functions @tsv
and @csv
to print all input arrays as tab-separated and comma-seperated values, respectively.
Print as JSON values
Similarly, it is possible to construct JSON values again, which is interesting if you just want to keep a subset of the fields:
$ cat file.json | jq -c '.users[] | {first}'
{"first":"Stevie"}
{"first":"Michael"}
Solution 5 - Json
my approach will be (your json example is not well formed.. guess thats only a sample)
jq '.Front[] | [.Name,.Out,.In,.Groups] | join("|")' front.json > output.txt
returns something like this
"new.domain.com-80|8.8.8.8|192.168.2.2:80|192.168.3.29:80 192.168.3.30:80"
"new.domain.com -443|8.8.8.8|192.168.2.2:443|192.168.3.29:443 192.168.3.30:443"
and grep the output with regular expression.
Solution 6 - Json
While both of the above answers work well if key,value are strings, I had a situation to append a string and integer (jq errors using the above expressions)
Requirement: To construct a url out below json
pradeep@seleniumframework>curl http://192.168.99.103:8500/v1/catalog/service/apache-443 | jq .[0]
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 251 100 251 0 0 155k 0 --:--:-- --:--:-- --:--:-- 245k
{
"Node": "myconsul",
"Address": "192.168.99.103",
"ServiceID": "4ce41e90ede4:compassionate_wozniak:443",
"ServiceName": "apache-443",
"ServiceTags": [],
"ServiceAddress": "",
"ServicePort": 1443,
"ServiceEnableTagOverride": false,
"CreateIndex": 45,
"ModifyIndex": 45
}
Solution:
curl http://192.168.99.103:8500/v1/catalog/service/apache-443 |
jq '.[0] | "http://" + .Address + ":" + "\(.ServicePort)"'
Solution 7 - Json
This will produce an array of names
> jq '[ .users[] | (.first + " " + .last) ]' ~/test.json
[ "Stevie Wonder", "Michael Jackson"]
Solution 8 - Json
I got pretty close to what I wanted by doing something like this
cat my.json | jq '.my.prefix[] | .primary_key + ":", (.sub.prefix[] | " - " + .sub_key)' | tr -d '"'
The output of which is close enough to yaml for me to usually import it into other tools without much problem. (I am still looking for a way to basicallt export a subset of the input json)