Answer stupid questions about MVC/MVVM engines?

Question

Angular

I am writing a project, and I understand that I am inventing bicycles.
Programmers have been riding these bikes for a long time, and I'm heavily sawing mine))
So the task:
In general, the application is an analytical presentation of data in the form of graphs, lists, trees, maps.
The work is divided into mathematical (aggregation of tabular data) and graphical (forming a grid and editing the properties of graphs, trees, tables, etc. placed in it).
The application must be able to save itself in JSON format in one line. That is, the server part does not assume the presence of a detailed description for each graph, table, etc. in the form of separate entities in the database, everything should be written in a serialized form in one cell of one row in the database.
1) Hence the difficulty with the implementation of the REST-api for the Frontend application.
Perhaps I just did not go deep and this is not a problem for these engines?
But we don’t have RESTapi, how can we turn the serialized data that came from the server into a data source (model) for the application?
2) In addition, the application is a kind of editor, I don’t really understand how it fits into the concept of these engines?
What it does: by dragging (from the component palette) layers are added to the workspace, components (graphs, tables, etc.) are added to them. Components can be moved from layer to layer, change the order of placement in the layer and size. Each component has its own widget with settings.
3) And finally, how to serialize all this and subsequently restore it to its original form?
4) Oh yes, there is also work with data and mathematics. When the chart is set up (dimensions and measures are set), it should calculate and display the values. Here you can not go deep, the only question is that it should be possible to write code inside each view (graph, table, etc.).
There is a significant limitation - ES6 is not possible to use yet.
I need to understand which engine suits me best and how to apply it? Is it really worth it?

Answer 1

For serialization, see react+redux.
In redax, the entire state of the application is stored in one tree (json object). By the simplest manipulations, you can save it and restore it again. But there is more and more to ES6: babel and webpack will have to be used.
And about 2 and 4 points - this already applies to your implementation. You can make calculations everywhere, UI can also be done anywhere (or almost any) anywhere.

Answer 2

In tasks of this level, the words "framework", "engine", etc. extremely harmful to use .
You seem to have a complete mess in the formulation of the task itself and the approach to choosing an ORM. Giving the client side the task of serialization and deserialization is not a good idea. First, figure out what the maximum number of algorithms and data structures are going to operate; then design the database and API for it, and only lastly look for the " engine ".
PS d3.js will do