Extract a specific field from JSON output using jq

Question

I have a JSON output as follows:

{   "example": {     "sub-example": [       {         "name": "123-345",         "tag" : 100       },       {         "name": "234-456",         "tag" : 100       },       {         "name": "4a7-a07a5",         "tag" : 100       }     ]   } } 

I want to extract the values of the three "name" fields and store it in three variables.

I tried cat json_file | jq '.["example.sub-example.name"]' to extract the value of the "name" field but that doesn't work.

Can anyone tell me how to achieve this using jq (or some other method)?

Answer 1

If you just want to extract the name fields, the command you're looking for is jq '.example."sub-example" | .[] | .name'. If you want to keep the names in an array, wrap the whole jq expression in square brackets.

Answer 2

In jq 1.3, you can use the filter to extract the values:

.example["sub-example"] | .[] | .name 

Or more compactly:

.example["sub-example"][].name 

These of course also work with later versions of jq as well.

Reading into shell variables

Rather than populating separate shell variables (which would require knowing in advance how many values there are), consider populating a shell array. For example, using a bash shell with mapfile (aka readarray):

mapfile -t ary < <(< json_file jq '.example."sub-example"[].name') 

You could alternatively use a shell while loop. Etc etc. There are many SO Qs on this topic.