jse - JSON Editor

quickly edit json files from the command line

jse is pragmatic and terse. It lets you edit json fast, without needing to care about quotes, types, exact indexes, or any of the stuff that makes json a pain.

Usage

$ jse FILE COMMAND QUERY VALUE

TLDR Version

edit an existing key: e or edit

$ jse config.json edit app.version 0.3.3

add a new element: a or add

$ jse todo.json add list.shopping {task:eggs,done:false}

delete a value: d or delete

$ jse problems.json delete problems[99]

full examples with json files below

Installing

pip3 install jse

Running from Source

Requirements:

• Python 3.7+
• Click

Steps:

1. clone the repository
2. install click pip3 install click
3. make the run script executable chmod +x run.py
4. place jse on the path ln -s /path/to/run.py ~/.local/bin/jse

Using poetry is recommended if you plan to contribute

$ pip3 install poetry
$ poetry install
$ poetry shell

Examples

Assume this json file is in the current directory

# example.json
{
    "users": [
        {"name": "alice", "age": 21, "admin": false},
        {"name": "bob", "age": 57, "admin": true},
        {"name": "charlie", "age": 37, "admin": false}
    ]
}

To delete the user alice using jse, all we need to do is specify d or delete mode and the path to her user object

$ jse example.json d users[0]

We can use both index or dot notation.

$ jse example.json d users.0   #users.first or users.^ also work

# example.json
{
    "users": [
        {"name": "bob", "age": 57, "admin": true},
        {"name": "charlie", "age": 37, "admin": false}
    ]
}

Now lets make charlie an admin. To edit an existing field, use the edit command with e or edit. Edit takes a key to change and its new value.

$ jse example.json e users.1.admin true

# example.json
{
    "users": [
        {"name": "bob", "age": 57, "admin": true},
        {"name": "charlie", "age": 37, "admin": true}
    ]
}

jse is smart enough to infer datatypes from the command line. it can also accept complex nested objects and arrays in a terse, quote-free format. Lets add a new nested field to the file with add or a

$ jse example.json a highscore [{score:32.5,user:bob,metadata:{ip:192.168.1.102,client:firefox}}]

{
  "users": [
    { "name": "bob", "age": 57, "admin": true },
    { "name": "charlie", "age": 37, "admin": true }
  ],
  "highscore": [
    {
      "score": 32.5,
      "user": "bob",
      "metadata": {
        "ip": "192.168.1.102",
        "client": "firefox"
      }
    }
  ]
}

jse also understands lists, so you can add new elements to one without needing an explicit index. It will infer we are trying to append from add instead of changing the list itself to an object (edit)

$ jse example.json a highscore {score:52,user:charlie}

{
  "users": [
    { "name": "bob", "age": 57, "admin": true },
    { "name": "charlie", "age": 37, "admin": true }
  ],
  "highscore": [
    {
      "score": "32.5",
      "user": "bob",
      "metadata": {
        "ip": "192.168.1.102",
        "client": "firefox"
      }
    },
    {
      "score": 52.0,
      "user": "charlie"
    }
  ]
}

first and last (or ^ and $) can also be used as a list index for any operation

$ jse example.json a users.first {name:jon,age:22,admin:false}

{
  "users": [
    { "name": "jon", "age": 22, "admin": false },
    { "name": "bob", "age": 57, "admin": true },
    { "name": "charlie", "age": 37, "admin": true }
  ],
  "highscore": [
    {
      "score": "32.5",
      "user": "bob",
      "metadata": {
        "ip": "192.168.1.102",
        "client": "firefox"
      }
    },
    {
      "score": 52.0,
      "user": "charlie"
    }
  ]
}

jse's error messages are informative, because no one wants a generic KeyError

$ jse example.json a users.0.name "not bob"
'name' already has a value. Use the edit command to modify it

$ jse example.json d users[2]
There is no element with index 2. The largest index is 1

You can also delete mulitple keys in one command `` $ jse example.json d users.0.age users.1.age users.2.age

```json
{
    "users": [
        {
            "name": "jon",
            "admin": false,
        },
        {
            "name": "bob",
            "admin": true,
        },
        {
            "name": "charlie",
            "admin": true
        }
    ]
    ...
}

No need to specify every element, just use the * or all operator

$ jse example.json d users.*.age # or users.all.age