Processing newline-delimited JSON (or NDJSON) into a Snowflake table
This recipe walks you through the process of loading NDJSON data and transforming it into a relational form to then be loaded into a table. NDJSON is a JSON format in which each row is valid JSON in itself and therefore can be processed independently of the complete document. For more details about NDJSON, please see http://ndjson.org/.
Getting ready
Since our objective is to demonstrate the processing of NDJSON, we will not describe the process of creating stages and external data from cloud storage as it has already been covered in other recipes. For simplicity, we have made available a sample NDJSON file in a public cloud bucket that we will read and process. The NDJSON sample file can be found at https://github.com/PacktPublishing/Snowflake-Cookbook/blob/master/Chapter03/r6/ndjson_sample.json if you would like to download it and use it in your own cloud storage buckets.
How to do it…
To demonstrate...