@expertbridge-ui/connection

Connection modules for Expertbridge:

• ExpertbridgeClient requests and authentication
• (future) i18n locales and translation

ExpertbridgeClient

The ExpertbridgeClient handles all client-side requests to the Expertbridge backend. It can be configured for use within the Expertbridge application, or used to issue CORS requests in other applications. At a high-level it supports:

• CSRF token authentication
  • a token may be passed at configuration time, else the client will handle fetching and passing the token in all subsequent requests.
  • queues requests in the case that another request is made before the token is received.
  • it checks for a token before every request, and will fail if no token was received or if it has expired. In either case the user should be directed to re-authenticate.
• supports GET and POST requests (no PUT or DELETE)
• timeouts
• query aborts through the AbortController API
• conditional GET requests using If-None-Match and ETag headers

Example usage

// appSetup.js
import { ExpertbridgeClient } from `@expertbridge-ui/connection`;
// or import ExpertbridgeClient from `@expertbridge-ui/connection/lib|esm/ExpertbridgeClient`;

ExpertbridgeClient.configure(...clientConfig);
ExpertbridgeClient.init(); // CSRF auth, can also chain `.configure().init();

// anotherFile.js
import { ExpertbridgeClient } from `@expertbridge-ui/connection`;

ExpertbridgeClient.post(...requestConfig)
  .then(({ request, json }) => ...)
  .catch((error) => ...);

API

Client Configuration

The following flags can be passed in the client config call ExpertbridgeClient.configure(...clientConfig);

• protocol = 'http:'
• host
• headers
• credentials = 'same-origin' (set to include for non-Expertbridge apps)
• mode = 'same-origin' (set to cors for non-Expertbridge apps)
• timeout
• csrfToken you can configure the client with a CSRF token at configuration time, else the client will attempt to fetch this before any other requests are issued

Per-request Configuration

The following flags can be passed on a per-request call ExpertbridgeClient.get/post(...requestConfig);

• url or endpoint
• headers
• body
• timeout
• signal (for aborting, from const { signal } = (new AbortController()))
• for POST requests
  • postPayload (key values are added to a new FormData())
  • stringify whether to call JSON.stringify on postPayload values

Request aborting

Per-request aborting is implemented through the AbortController API:

import { ExpertbridgeClient } from '@expertbridge-ui/connection';
import AbortController from 'abortcontroller-polyfill';

const controller = new AbortController();
const { signal } = controller;

ExpertbridgeClient.get({ ..., signal }).then(...).catch(...);

if (IWantToCancelForSomeReason) {
  signal.abort(); // Promise is rejected, request `catch` is invoked
}

Development

@data-ui/build-config is used to manage the build configuration for this package including babel builds, jest testing, eslint, and prettier.