processing data from client-side storage for the page rendering

Question

What is the best way to process data from client-side storage for the page rendering? To make the situation clear: we intend to have flexible abstract architecture - things can be changed on backend and it should cause changes on frontend part.

So, e.g., on frontend I get deeply nested JSON with form basic structure - avaliable services, attributes, etc. I want to parse it as some plain array, use this array for page rendering (using appropriate widgets for different types of services), and store in some client storage for future, and when user visits the page again - use the previosly cached data from client for page rendering (if no new services added on server). The technologies I intend to use are angularJS, localForage, and angular-localforage lib. So, I have following questions:

1. What is the best way to proceed data - I plan to get a whole array (concerning specific page) from client storage in one transaction, put it into Angular app and use from the app then for page rendering. Is it appropriate and effective? Or maybe I should take date data in small transactions, service by service? The same question with cahing data to client storage.
2. If I have some big transaction, do I have any risk of making other things go slow, or this risk is neutralized in wrapping or storage architecture?
3. I couldn't find actual information about migration in case user changes the storage type - from localStorage to indexeddb for example. The issue is open, so it seems that I'll need to check the type of storage every time, and if changed - recache the data from server again. Is it the only possible way?
4. Are there any issues/drawbacks with selected architecture/technologies? Can you advice to change or add someting?
5. What is appropriate for searching pieces of data in simple key-value structure provided by localForage? E.g., if I need elements with parent_id=42 - should I get whole data needed to app, organize it, and perform queries in app? Or do some tricky storage structure? Or there is some better way?

Answer 1

point no 5 You can use IDBKeyRange.bound. There are various parameters for it.