We should provide the ability for the TP to generate a schema file for offline use from a live data store. This would give the best of both worlds - type safety from a known good schema but simplify the whole CI/CD process by making it easy to create a schema.
There should be the ability to generate schema for at least blobs and tables, as well as supply some "depth" parameters e.g. folder depth for blobs, and number of rows for type inference from tables.
We should provide the ability for the TP to generate a schema file for offline use from a live data store. This would give the best of both worlds - type safety from a known good schema but simplify the whole CI/CD process by making it easy to create a schema.
There should be the ability to generate schema for at least blobs and tables, as well as supply some "depth" parameters e.g. folder depth for blobs, and number of rows for type inference from tables.