How to Convert from JSON to CSV at The Command Line

  <div data-element="header">
    <h4 class="text-2xl font-bold pb-2">Get notified about new articles!</h4>
  </div>
  <!-- <div class="formkit-subheader" data-element="subheader" style="color: rgb(104, 104, 104); font-size: 18px;">

    <div class="formkit-field">
      <input
        class="border rounded-md p-2 mb-2 w-2/3"
        name="email_address"
        aria-label="Email Address"
        placeholder="Email Address"
        required=""
        type="email"
      />
    </div>
    <button
      data-element="submit"
      class="formkit-submit formkit-submit w-44 pt-1 pb-1"
      style="
        color: rgb(255, 255, 255);
        background-color: rgb(22, 119, 190);
        border-radius: 4px;
        font-weight: 400;
      "
    >
      <div class="formkit-spinner">
        <div></div>
        <div></div>
        <div></div>
      </div>
      <span class="">Subscribe</span>
    </button>
  </div>
  <div class="formkit-guarantee" data-element="guarantee">
    <p class="text-slate-400 text-sm">
      We won't send you spam. Unsubscribe at any time.
    </p>
  </div>
</div>
<style></style>

Convert JSON to CSV via the Command Line Using JQ

If you don’t want to install dasel or if you just love jq (brew install jq) then this solution may work well for you. Before I discovered dasel this was the main approach I used:

$ cat simple.json| jq -r '(map(keys) | add | unique) as $cols | map(. as $row | $cols | map($row[.])) as $rows | $cols, $rows[] | @csv' 
"color","id","value"
"red",1,"#f00"
"green",2,"#0f0"
"blue",3,"#00f"

A full explanation for how this works is beyond the scope of this article, but the idea is to use (map(keys) | add | unique) to get the column names and then gather the values for those columns into $row and use the @csv filter to convert the array of rows in a CSV format.

Convert JSON to CSV via the Command Line and Choose Ordering Column

The downside to the previous two approaches is that you can’t specify which columns to exclude nor their order. Both jq and dazel support a query language for customizing the output, but if you don’t want to dive into CSS selectors, the jsonv tool is a great alternative.

To convert, we will use jsonv and pipe it our JSON file. Then, we will specify the columns to include, and we redirect its output to a file.

cat simple.json | jsonv id,color,value > simple.csv

This gives us a simple CSV file:

$ cat simple.csv
1,"red","#f00"
2,"green","#0f0"
3,"blue","#00f"
4,"cyan","#0ff"
5,"magenta","#f0f"
6,"yellow","#ff0"
7,"black","#000"

jsonv handles more complex examples as well. Under the hood, it uses gnuawk (gawk). You can install it using curl:

curl -Ls https://raw.github.com/archan937/jsonv.sh/master/install.sh | bash

You can install gawk using brew (brew install gawk) or your package manager of choice.

Convert CSV to JSON at The Command Line

For converting CSV to JSON, we can use dasel again. The read (-r) and write (-w) options mean that it’s easy to convert from any of its supported file formats (JSON, YAML, TOML, XML, and CSV).

We can get our original JSON document back from CSV like this:

$ cat sample.csv
ID, color, value
1,"red","#f00"
2,"green","#0f0"
3,"blue","#00f"

$ dasel -r csv -w json < sample.csv
{
  "ID": "1",
  "color": "red",
  "value": "#f00"
}
{
  "ID": "2",
  "color": "green",
  "value": "#0f0"
}
{
  "ID": "3",
  "color": "blue",
  "value": "#00f"
}

Convert CSV to JSON Command Line with csvtojson

Another option is to grab the npm tool csvtojson (npm i --save csvtojson). If you don’t already have npm and Node.js installed then installing dasel is a simple solution. Once installed, you can convert to CSV by sending input to csvtojson over standard in:

csvtojson < sample.csv
[
{"ID":"1","color":"red","value":"#f00"},
{"ID":"2","color":"green","value":"#0f0"},
{"ID":"3","color":"blue","value":"#00f"}
]

Conclusion

You now have the knowledge and the tools you need to convert JSON to CSV and CSV to JSON. jq, dasel, csvtojson and jsonv are handy command line tools that can be used on Linux, MacOS and Windows under WSL 2.