I have json files falling in our blob with two fields, 1. offset(integer), 2. value(base64).

This value column is json with unicode. so they sent it as base64. Challenge is this json is very large with 100+ fields. so we cannot define the schema. We can only have some schemahints. so Autoloader is the best fit.

I tried with Autoloader with schemahints and other options. It always picks value as string and unable to parse it back to json without providing the schema. I want databricks to infer the schema. Any help is much appreciated. Thanks Advance.

Attached the sameple json!