# Importing Shopify data into Google BigQuery

Lately I’ve been using the command line more and more at work. Normally when I work with API’s for data that needs to be imported to BigQuery, I’ll use Google Apps Script. I thought that working with the Shopify API would be a fun way of experimenting with doing this through the command line.

Calling the API is as simple as doing a curl statement with your parameters included. When I was working through this solution to begin with, I knew that creating a json file that was formatted in BigQuery’s new line delimited format might be an issue. Luckily, there’s an open source solution called JQ on Github that you can use. JQ is like sed for JSON data and makes parsing JSON super easy. With a simple -c statement, you can easily create a new line delimited json file. From there you can traverse through your code to find the items you need, using a command like '.results[]'.

From there you can set up a cronjob that calls the api every minute and gets the most recent updated file. For me since BigQuery is append only, I just called the API with an updated_at parameter and append everything to the dataset.

Also, the Shopify feed is 483 different nested and repeated key:value pairs in the JSON feed. That means you’ll end up with 483 columns in your table. I took me a while to hand code the entire JSON format for the data, but I would hate for someone to have to do that again, so I’m sharing it with you.

The example code below is designed to run every minute. If the file will be written with the current iso8601 timestamp such as 2016-11-11T09:30:00Z.json. Each time the script runs, if the file size is greater than zero, it sends a copy to Google Cloud Storage and then uploads the results to Google BigQuery. If the file size is equal to zero, then the script will stop.

If you want to use the code, you can find the code here for importing shopify data into Google BigQuery.

If you want to use this code yourself, you need to make sure that you have the Google Cloud SDK on your local system or run your code on a small Google Compute Engine Instance.