Revise the "Advanced: Access the Data Catalog in code" page into a "Use cases for the code API".

[stichbury]

Parent task #2818

Description

As I'm working through #2888 I'm getting a few comments about the code API section and how we talk about it right now. I was myself somewhat dubious when I first edited it, as it wasn't very clear what we were recommending except "don't use this unless you know what you're doing" which isn't that helpful.

Feedback from @noklam (which is very helpful):

Can we add a link to Kedro Jupyter Notebook? I think this is to demonstrate Kedro component can be used standalone. However if people need to access catalog programmatically, it's more likely they need "Hooks" for changes or they need to use the catalog for interactive analysis.

Constructing DataCatalog from scratch is rare.

Additionally, this is more like Standalone-catalog starter? should we have a separate page since I don't think this is an use case. The use case should be

• Accessing catalog during a Kedro run
• Accessing catalog in notebooks for analysis.

---

I think I'm getting the sense that we need to work through this page and structure it better to cover the use cases so that we add useful content about the standalone-datacatalog starter (for users unfamiliar with it, those wanting to use a portion of Kedro features) and any alternative reasons for using it with Hooks, as Nok suggests.

Again, this is out of the scope of work to restructure the data section, although it's a high priority piece of work because we want the standalone data catalog use case to be more prominent.

For now, I'm separating it from #2488/#2888 but will get to this soon.