You are currently viewing the new Beta Editor Docs.
Switch to the Classic Editor Docs
You are currently viewing the Classic Editor Docs.
Switch to the Beta Editor Docs

Running an app offline

Anvil apps can function offline. In other words, an app’s client-side code will continue to work if the device loses its internet connection.

When an app is offline, client-side code (Forms and Modules) will continue to work. However, any code that tries to communicate with the server will fail with an anvil.server.AppOfflineError. This includes accessing the Users Service, working with Data Tables or any use of anvil.server.call.

This section explores some offline strategies to handle code that communicates with the server.

Installing Anvil Apps

Anvil apps are progressive web apps (PWAs) and can be installed almost like native apps.

Specific instructions for installing an Anvil app depends on the platform. On Chrome desktop, the option is available next to the address bar. In Chrome for Android or Safari for iPhone, there is an option to ‘Add to home screen’.

Installing an Anvil app isn’t necessary for an app to work offline. If you visit an Anvil app in a browser and lose connection, the app will continue to run as described above.

Is the app online?

There are two ways to know if an app is online. The first is to check using anvil.server.is_app_online(). If this returns False, subsequent communications to the server will fail.

if anvil.server.is_app_online():
    user = anvil.users.login_with_form()
else:
    # we can't use the Users Service since we're offline
    alert('It looks like you're offline')

Alternatively, wrapping code that communicates with the server in a try/except block will allow you to catch and handle an anvil.server.AppOfflineError gracefully.

try:
    anvil.server.call('save_data', data)
except anvil.server.AppOfflineError:
    ...
    # save the data in the browser

anvil.server.AppOfflineError is raised anytime an offline app attempts to communicate with the server. In the above example, we use the try/except strategy. First, we try and save data using anvil.server.call. If we find the app is offline, the code inside the except block runs. We would then need to consider strategies for storing (caching) data in the browser inside the except block. Later, when the app comes back online, we can make the server call using the cached data.

Working with data offline

If your app needs data and you want it to run offline, you’ll have to store (cache) that data in the browser to interact with it. Likewise, if a user makes changes while offline, you’ll need to store those changes until you can save them on the server.

A simple storage mechanism for browsers is the localStorage object. localStorage is a dictionary-like javascript object that persists between browser sessions and works offline. It can be used to store JSON data and can be accessed using anvil.js.

from anvil.js.window import localStorage
import json


if anvil.server.is_app_online():
    anvil.server.call('save_data', data)
else:
    localStorage.setItem('unsaved_data', json.dumps(data))

A wrapper for localStorage can be found in anvil_extras (available as a third-party dependency), which provides a cleaner python API for the same object.

from anvil_extras.storage import local_storage

if anvil.server.is_app_online():
    anvil.server.call('save_data', data)
else:
    local_storage['unsaved_data'] = data

Note that the persistence of localStorage depends on the browser. After several days of inactivity, a browser may clear the localStorage object.

localStorage is best suited for JSONable data. For more advanced offline data storage consider the browser’s IndexedDB API. IndexedDB is a low level client-side database that is suitable for storing simple data types such as strings and floats, as well as more complex data types such as bytes and lists. Working with IndexedDB can be tricky and it may be useful to find a Javascript library that wraps IndexedDB in a more user friendly API.

Example app

In the following Todo app, we explore some strategies for working with data offline. Launching the app offline is possible since it caches user data from the previous sessions. If the app goes offline during a session, attempts to save data on the server are also cached.