Json file with bash

Question

I need delete space and value kubernetes from json file with use bash script:

Content json file

Input

 "spec": {
        "finalizers": [
            "kubernetes"
        ]
    },

Ouput

 "spec": {
        "finalizers": []
    },

It's generally recommended to use a JSON parsing tool for JSON data. Can you use [tag:jq]? — glenn jackman, Feb 12 '21 at 18:40

score 1 · Answer 1 · answered Feb 12 '21 at 18:43

1

Given "input.json"

{
  "spec": {
          "finalizers": [
              "kubernetes"
          ]
      },
  "other": 42
  }

Then

jq '.spec.finalizers = []' input.json

outputs

{
  "spec": {
    "finalizers": []
  },
  "other": 42
}

answered Feb 12 '21 at 18:43

glenn jackman

4,630
1
17
20

Thank you very much Gleen! It worked – spawn1978 Feb 12 '21 at 19:46

score 0 · Answer 2 · answered Feb 12 '21 at 18:34

Lets use some sed magic with regex:

echo 'Input "spec": { "finalizers": [ "kubernetes" ] },' | sed -r 's/Input (.* \[) \".*\" (\].*)/\1\2/'

Which returns:

"spec": { "finalizers": [] },

sed -r uses extended regex (-E on osx), and "()" are used for capture groups, which is what the \1 and \2 refers to, so first we capture everything up to "[", then match everything from the "]" until the end.

The downside of this is that your input doesn't look like complete json, so modifying the regex to match the proper input can be much harder.

I would also recommend you to look at jq, the defacto tool for json parsing in CLI.

Thank Fredrik for you help. – spawn1978 Feb 12 '21 at 19:51 — spawn1978, Feb 12 '21 at 19:51

Json file with bash

2 Answers2