Databricks Storage - robinrodricks/FluentStorage GitHub Wiki
In order to use Databricks on AWS/Azure you need to reference first.
Notebooks can be exported by calling to OpenReadAsync
or it's related methods like ReadTextAsync
. By default notebooks are exported as source code, however you can override that by appending a format specifier at the end of the path.
For instance, in order to export in Jupyter format, you would write the following:
string path = "/workspace/Shared/mynotebook"
string jupyterData = await _storage.ReadTextAsync(path + "#jupyter");
The supported formats are:
- source
- html
- jupyter
- dbc