Skip to main content
Version: 1.0

Network settings

Network related settings, as well some login related settings, are defined in the API class.

// Get a File or Storage object
const file = File.fromURL(url)
const storage = new Storage(options, callback)

// The API object can be accessed using:
const api = file.api // or storage.api

The default API object used by the File class can be accessed via:

import { API } from 'megajs'
const globalApi = API.getGlobalApi()

In V0 this library used the request library, but now fetch is used and to make the library easier to use some shortcuts were added:

// The user-agent used when requesting is available via:
api.userAgent // default: megajs/{version}

// In Node the Agents used for HTTP and HTTPs are available via
api.httpAgent
api.httpsAgent

Those agents can be replaced in order if you need to use a proxy. Keep in mind that you need to use an agent that supports keep-alive.

You can replace the MEGA gateway, which is useful when testing to point to a mock server:

api.gateway // default: https://g.api.mega.co.nz/

You can also replace the fetch wrapper function used:

api.fetch = someFetchLikeFunction

But by doing so the api.userAgent, api.httpAgent and api.httpsAgent shortcuts will be ignored and will need to be reimplemented.

As v1.2.0 you can replace the fetch module function. It defaults to globalThis.fetch and, if nonexistent, it requires node-fetch. To replace it override API.fetchModule:

import { API } from 'megajs'
API.fetchModule = someFetchImplementation

If you replace API.fetchModule all code using the library will be affected, but the api.userAgent, api.httpAgent and api.httpsAgent shortcuts will NOT be ignored, as the reasoning for it is just to allow the user to provide an alternative fetch implementation for the current environment.

Notice that some future version will not import node-fetch if a global fetch doesn't exist in order to make npm install faster and as modern Node.js already include a global fetch implementation by default.

Files created with new File and File.fromURL by default use a non-logged API instance and can be quite limited. To workaround those limits accounts can be used, so downloading will use the limits from those:

// First log in
const storage = new Storage(options, callback)

// Wait for the callback then get a shared file using the logged API
const sharedFile = File.fromURL(url, {
  api: storage.api
})

// Now downloading the shared file will use the limits from that account

Finally the properties of the API class can also be provided as Storage arguments:

const storage = new Storage({
  email: 'user@example.com',
  password: 'correct horse battery example',
  userAgent: 'ExampleClient/1.0'
}, callback)

And that's the end of the tutorial. The next page just includes some advanced and unusual usages of the library.