You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I agree that adding crow's feet would be great, but sadly I don't think this is possible using the current software package used to generate the ERDs :-( If we wanted to add crow's feet, we'd need to investigate alternatives or seek to build something new.
At the moment, ERDs generated by the docs are generated via the Python package jts_erd and are actually also generated from the tabular data package. See conf.py, L451–L516 for how this is done.
I've had an investigate into the jts_erd docs including the examples and cannot find a way to provide crow's feet inside the package. We'd either need to fork it, or develop something else from scratch.
Unfortunately, nothing I know of can generate ERDs directly from JSON Schema.
If the core issue is that the documentation lacks clear descriptions of where particular schemas are embedded under others in a 1-to-many fashion, we can consider using the compiled schemas and pointing a docson viewer at them. This renders beautiful views of JSON Schemas, and ODSC maintains its own fork adding features for use with specific standards. This would generate a view of the data which demonstrates the embedded nature of JSON, as well as highlight where there are arrays.
(I hope this is the right place to log this.)
This is the excellent ERD generated by ODSC software for the UK profile.
This is the ERD generated by Porism software from the tabular data package.
The ODSC one is generally better but it lacks "crows' feet" denoting the direction of one-to-many relationships. Can they be added?
The text was updated successfully, but these errors were encountered: