It’s not possible to flatten this dataset into 1 line, because columns have non-unique values. For instance, these columns have more than 1 value:
- metrics{*}
- name
- columnType
- dataType
After you received the JSON response, you need to derive a few tables, and in each derived table parse only a subset (sub-object) of the received JSON by specifying the root path of that sub-object. The number of derived tables depends on the JSON structure.
Once you flatten sub-JSON’s, you can merge/append all of them together in one table as needed.
Here is an example how you can flatten data when every column has only 1 unique value:
flatten.morph (6.1 KB)