
Security News
Axios Supply Chain Attack Reaches OpenAI macOS Signing Pipeline, Forces Certificate Rotation
OpenAI rotated macOS signing certificates after a malicious Axios package reached its CI pipeline in a broader software supply chain attack.
mixbook-http-client
Advanced tools
Composable TypeScript HTTP client for browser and NodeJS apps, with built-in mocking and 'middlewares' support
Composable TypeScript HTTP client for browser and NodeJS apps, with built-in mocking and "middlewares" support.
The idea is - you have some base HTTP clients:
BrowserHttpClient),NodeHttpClient)MockHttpClient)They implement the same interface, and completely interchangeable. Which makes it very convenient to use them in isomorphic apps, and also convenient to test.
And then there're a bunch of "middleware" clients (similar to middlewares in web apps), which "enhance" those base clients. They usually wrap those base clients, and implement the same interface, so you can also use them interchangeably. We have:
RetryableHttpClient - will retry a request on condition (e.g. if the request is failing) after some delay.LoggableHttpClient - will log the info about the requests using the provided logger.MsgPackHttpClient - will add methods for making calls to APIs, that respond in msgpack format.The library is pretty small (~6kb compressed if you use everything except MsgPackHttpClient, and ~25kb if you use MsgPackHttpClient).
It's also tree-shaking friendly, so if you don't use some of those base or middleware clients, they won't be included into the bundle.
You can build your own "middlewares", they just need to implement the same IHttpClient interface, and that's it.
So, there're a bunch of various HttpClient classes, all of them conform to the same interface. You can build your custom uber http client with the features you need.
You create an instance of HttpClient, for a browser:
import {BrowserHttpClient} from "mixbook-http-client/build/browser";
const client = new BrowserHttpClient();
or for a NodeJS app:
import {NodeHttpClient} from "mixbook-http-client/build/node";
const client = new NodeHttpClient();
They all implement the same interface, IHttpClient. Then you make a call, like:
const response = await client.get("https://www.example.com/coupon/FOOBAR");
All the call methods (get, post, put, delete) return Promise<IResponse>. If you need to track the upload/download progress via onUploadProgress/onDownloadProgress, or be able to abort the request, or set the request timeout, you should use (send), it will return IHttpClientSession, it has all of that.
You can also set the default headers or the default timeout via the constructor arguments of browser or NodeJS clients. They always gonna be used with all the requests made via those clients. Like:
import {NodeHttpClient} from "mixbook-http-client/build/node";
const client = new NodeHttpClient({timeout: 1000, headers: {"content-type": "application/json"}});
There is MockHttpClient which also implements IHttpClient.
import {MockHttpClient, MockResponse} from "mixbook-http-client/build/mock";
const client = new MockHttpClient();
client.mockRequest(
{method: "GET", url: "https://example.com/coupon/FOOBAR"},
new MockResponse(200, JSON.stringify({data: "ok"}))
);
const json = (await client.get("https://example.com/coupon/FOOBAR").promise).json;
expect(json.data).to.eq("ok");
expect(client.executedRequests.filter(r => r.request.url === "https://example.com/coupon/FOOBAR").length).to.eq(1);
By default, it only matches mocked requests by URL. If you want it to take request headers/body into account, you can use constructor arguments shouldUseBody and shouldUseHeaders.
By default, if you make several requests that match the same URL (and optionally headers/body), it will return the same mocked response. If you want to test a sequence of requests (like first /foo returns {data: "foo"}, second /foo returns {data: "bar"}, you can use isRepeating: false constructor argument.
You can also make your client "retryable" by wrapping it into RetryableHttpClient. It also implements IHttpClient. It will retry the request if the retry condition is true (by default it's - 3 additional attempts for GET requests which are failed or with >=500 status code), after some delay. It's a proxy class, so you use it by wrapping another http client:
import {BrowserHttpClient} from "mixbook-http-client/build/browser";
import {RetryableHttpClient} from "mixbook-http-client/build/retryable";
const retryableClient1 = new RetryableHttpClient(new BrowserHttpClient());
/// or
const retryableClient2 = new RetryableHttpClient(new BrowserHttpClient(), {
retryCondition: (request, response, error, retryCount) =>
retryCount < 10 && (response == null || response.status >= 400),
delaysInMilliseconds: [500, 1000, 1500],
});
You can make your client "loggable" by wrapping it into LoggableHttpClient. It also implements IHttpClient. It accepts an object, that conforms to ILoggable interface (e.g., just a regular console), and gonna use that object to log information about requests and responses.
import {BrowserHttpClient} from "mixbook-http-client/build/browser";
import {LoggableHttpClient} from "mixbook-http-client/build/loggable";
const loggableClient = new LoggableHttpClient(new BrowserHttpClient(), console);
You can use MsgPackHttpClient to talk to APIs, that send MsgPack responses. It also implements IHttpClient, but in addition it adds some other methods, like getMsgPack, postMsgPack, sendMsgPack. Same as RetryableHttpClient - it's a proxy, you initialize it by passing another IHttpClient as an
argument.
The reason why it's not part of HttpClient base class (which would make some sense) - it depends on msgpack-lite, which is quite heavy. So, if you don't use MsgPackHttpClient, the compressed/gzipped size of mixbook-http-client will be around 7kb, but if you do use it - around 25kb.
So, as an example, you build a client, which has all of those features, like this:
import {NodeHttpClient} from "mixbook-http-client/build/node";
import {MsgPackHttpClient} from "mixbook-http-client/build/msgPackClient";
import {LoggableHttpClient} from "mixbook-http-client/build/loggable";
import {RetryableHttpClient} from "mixbook-http-client/build/retryable";
const client = new LoggableHttpClient(
new RetryableHttpClient(new MsgPackHttpClient(new NodeHttpClient()), {
retryCondition: (request, response, error, retryCount) =>
retryCount < 10 && (response == null || response.status >= 400),
delaysInMilliseconds: [500, 1000, 1500],
}),
console
);
FAQs
Composable TypeScript HTTP client for browser and NodeJS apps, with built-in mocking and 'middlewares' support
The npm package mixbook-http-client receives a total of 2 weekly downloads. As such, mixbook-http-client popularity was classified as not popular.
We found that mixbook-http-client demonstrated a not healthy version release cadence and project activity because the last version was released a year ago. It has 1 open source maintainer collaborating on the project.
Did you know?

Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.

Security News
OpenAI rotated macOS signing certificates after a malicious Axios package reached its CI pipeline in a broader software supply chain attack.

Security News
Open source is under attack because of how much value it creates. It has been the foundation of every major software innovation for the last three decades. This is not the time to walk away from it.

Security News
Socket CEO Feross Aboukhadijeh breaks down how North Korea hijacked Axios and what it means for the future of software supply chain security.