A lightweight polyfill library for Promise-based WebExtension APIs in Chrome
Go to file
dependabot[bot] b8e374a37a chore(deps-dev): bump eslint from 3.19.0 to 6.6.0 (#209)
Bumps [eslint](https://github.com/eslint/eslint) from 3.19.0 to 6.6.0.
- [Release notes](https://github.com/eslint/eslint/releases)
- [Changelog](https://github.com/eslint/eslint/blob/master/CHANGELOG.md)
- [Commits](https://github.com/eslint/eslint/compare/v3.19.0...v6.6.0)

Signed-off-by: dependabot[bot] <support@github.com>
2019-11-21 14:24:41 +01:00
scripts test(integration): Use node for running test scripts (#167) 2019-01-28 18:49:25 +01:00
src fix: Throw error when loaded in non-extension contexts (#190) 2019-07-18 15:21:49 +02:00
test fix: Throw error when loaded in non-extension contexts (#190) 2019-07-18 15:21:49 +02:00
.editorconfig chore: Add `.editorconfig` (#188) 2019-09-08 15:38:25 +02:00
.eslintignore chore: Small tweaks to the eslint config files (follow-up for #167) 2019-01-28 18:56:40 +01:00
.eslintrc chore: Small tweaks to the eslint config files (follow-up for #167) 2019-01-28 18:56:40 +01:00
.gitattributes chore: Added .gitattributes to force eol=lf to fix #101 (#103) 2018-04-09 15:44:30 +02:00
.gitignore test: introduced a test suite for unit testing. 2016-11-07 16:33:42 +01:00
.travis.yml fix(travis): Added xvfb service to travis config (#203) 2019-09-24 21:20:42 +02:00
CODE_OF_CONDUCT.md chore: Add Mozilla Code of Conduct file 2019-04-16 11:39:12 +02:00
CONTRIBUTING.md doc: Rewrite README.md and move build and contribution details into a separate doc file (#147) 2018-07-19 15:56:49 +02:00
Gruntfile.js test(integration): Use node for running test scripts (#167) 2019-01-28 18:49:25 +01:00
LICENSE Initial commit 2016-10-06 12:02:51 -07:00
README.md docs(README): Add suggested webpack usage (#194) 2019-08-01 18:53:23 +02:00
api-metadata.json fix: Remove runtime.getBrowserInfo() (#200) 2019-08-01 19:01:38 +02:00
package.json chore(deps-dev): bump eslint from 3.19.0 to 6.6.0 (#209) 2019-11-21 14:24:41 +01:00

README.md

WebExtension browser API Polyfill Build Status npm version

This library allows extensions that use the Promise-based WebExtension/BrowserExt API being standardized by the W3 Browser Extensions group to run on Google Chrome with minimal or no changes.

This library doesn't (and it is not going to) polyfill API methods or options that are missing on Chrome but natively provided on Firefox, and so the extension has to do its own "runtime feature detection" in those cases (and then eventually polyfill the missing feature on its own or enable/disable some of the features accordingly).

Table of contents

Supported Browsers

Browser Support Level
Chrome Officially Supported (with automated tests)
Firefox Officially Supported as a NO-OP (with automated tests for comparison with the behaviors on Chrome)
Opera Unofficially Supported as a Chrome-compatible target (but not explicitly tested in automation)
Edge Not supported (See MSEdge Support section)

The polyfill is being tested explicitly (with automated tests that run on every pull request) on officially supported browsers (that are currently the last stable versions of Chrome and Firefox).

On Firefox, this library is actually acting as a NO-OP: it detects that the browser API object is already defined and it does not create any custom wrappers. Firefox is still included in the automated tests, to ensure that no wrappers are being created when running on Firefox, and for comparison with the behaviors implemented by the library on Chrome.

Installation

A new version of the library is built from this repository and released as an npm package.

The npm package is named after this repo: webextension-polyfill.

For the extension that already include a package.json file, the last released version of this library can be quickly installed using:

npm install --save-dev webextension-polyfill

Inside the dist/ directory of the npm package, there are both the minified and non-minified builds (and their related source map files):

  • node_modules/webextension-polyfill/dist/browser-polyfill.js
  • node_modules/webextension-polyfill/dist/browser-polyfill.min.js

For extensions that do not include a package.json file and/or prefer to download and add the library directly into their own code repository, all the versions released on npm are also available for direct download from unpkg.com:

and linked to the Github releases:

Basic Setup

In order to use the polyfill, it must be loaded into any context where browser APIs are accessed. The most common cases are background and content scripts, which can be specified in manifest.json (make sure to include the browser-polyfill.js script before any other scripts that use it):

{
  // ...

  "background": {
    "scripts": [
      "browser-polyfill.js",
      "background.js"
    ]
  },

  "content_scripts": [{
    // ...
    "js": [
      "browser-polyfill.js",
      "content.js"
    ]
  }]
}

For HTML documents, such as browserAction popups, or tab pages, it must be included more explicitly:

<!DOCTYPE html>
<html>
  <head>
    <script type="application/javascript" src="browser-polyfill.js"></script>
    <script type="application/javascript" src="popup.js"></script>
  </head>
  <!-- ... -->
</html>

And for dynamically-injected content scripts loaded by tabs.executeScript, it must be injected by a separate executeScript call, unless it has already been loaded via a content_scripts declaration in manifest.json:

browser.tabs.executeScript({file: "browser-polyfill.js"});
browser.tabs.executeScript({file: "content.js"}).then(result => {
  // ...
});

Basic Setup with module bundlers

This library is built as a UMD module (Universal Module Definition), and so it can also be used with module bundlers (and explicitly tested on both webpack and browserify) or AMD module loaders.

src/background.js:

var browser = require("webextension-polyfill");

browser.runtime.onMessage.addListener(async (msg, sender) => {
  console.log("BG page received message", msg, "from", sender);
  console.log("Stored data", await browser.storage.local.get());
});

browser.browserAction.onClicked.addListener(() => {
  browser.tabs.executeScript({file: "content.js"});
});

src/content.js:

var browser = require("webextension-polyfill");

browser.storage.local.set({
  [window.location.hostname]: document.title,
}).then(() => {
  browser.runtime.sendMessage(`Saved document title for ${window.location.hostname}`);
});

By using require("webextension-polyfill"), the module bundler will use the non-minified version of this library, and the extension is supposed to minify the entire generated bundles as part of its own build steps.

If the extension doesn't minify its own sources, it is still possible to explicitly ask the module bundler to use the minified version of this library, e.g.:

var browser = require("webextension-polyfill/dist/browser-polyfill.min");

...

Usage with webpack without bundling

The previous section explains how to bundle webextension-polyfill in each script. An alternative method is to include a single copy of the library in your extension, and load the library as shown in Basic Setup. You will need to install copy-webpack-plugin:

npm install --save-dev copy-webpack-plugin

In webpack.config.js, import the plugin and configure it this way. It will copy the minified file into your output folder, wherever your other webpack files are generated.

const CopyWebpackPlugin = require('copy-webpack-plugin');

module.exports = {
  /* Your regular webpack config, probably including something like this:
  output: {
    path: path.join(__dirname, 'distribution'),
    filename: '[name].js'
  },
  */
  plugins: [
    new CopyWebpackPlugin([
      {
        from: 'node_modules/webextension-polyfill/dist/browser-polyfill.js'
      }
    ])
  ]
}

And then include the file in each context, using the manifest.json just like in Basic Setup.

Using the Promise-based APIs

The Promise-based APIs in the browser namespace work, for the most part, very similarly to the callback-based APIs in Chrome's chrome namespace. The major differences are:

  • Rather than receiving a callback argument, every async function returns a Promise object, which resolves or rejects when the operation completes.

  • Rather than checking the chrome.runtime.lastError property from every callback, code which needs to explicitly deal with errors registers a separate Promise rejection handler.

  • Rather than receiving a sendResponse callback to send a response, onMessage listeners simply return a Promise whose resolution value is used as a reply.

  • Rather than nesting callbacks when a sequence of operations depend on each other, Promise chaining is generally used instead.

  • The resulting Promises can be also used with async and await, rather than dealt with directly.

Examples

The following code will retrieve a list of URLs patterns from the storage API, retrieve a list of tabs which match any of them, reload each of those tabs, and notify the user that is has been done:

browser.storage.local.get("urls").then(({urls}) => {
  return browser.tabs.query({url: urls});
}).then(tabs => {
  return Promise.all(
    Array.from(tabs, tab => browser.tabs.reload(tab.id))
  );
}).then(() => {
  return browser.notifications.create({
    type: "basic",
    iconUrl: "icon.png",
    title: "Tabs reloaded",
    message: "Your tabs have been reloaded",
  });
}).catch(error => {
  console.error(`An error occurred while reloading tabs: ${error.message}`);
});

Or, using an async function:

async function reloadTabs() {
  try {
    let {urls} = await browser.storage.local.get("urls");

    let tabs = await browser.tabs.query({url: urls});

    await Promise.all(
      Array.from(tabs, tab => browser.tabs.reload(tab.id))
    );

    await browser.notifications.create({
      type: "basic",
      iconUrl: "icon.png",
      title: "Tabs reloaded",
      message: "Your tabs have been reloaded",
    });
  } catch (error) {
    console.error(`An error occurred while reloading tabs: ${error.message}`);
  }
}

It's also possible to use Promises effectively using two-way messaging. Communication between a background page and a tab content script, for example, looks something like this from the background page side:

browser.tabs.sendMessage(tabId, "get-ids").then(results => {
  processResults(results);
});

And like this from the content script:

browser.runtime.onMessage.addListener(msg => {
  if (msg == "get-ids") {
    return browser.storage.local.get("idPattern").then(({idPattern}) => {
      return Array.from(document.querySelectorAll(idPattern),
                        elem => elem.textContent);
    });
  }
});

or:

browser.runtime.onMessage.addListener(async function(msg) {
  if (msg == "get-ids") {
    let {idPattern} = await browser.storage.local.get("idPattern");

    return Array.from(document.querySelectorAll(idPattern),
                      elem => elem.textContent);
  }
});

Or vice versa.

Known Limitations and Incompatibilities

This library tries to minimize the amount of "special handling" that a cross-browser extension has to do to be able to run on the supported browsers from a single codebase, but there are still cases when polyfillling the missing or incompatible behaviors or features is not possible or out of the scope of this polyfill.

This section aims to keep track of the most common issues that an extension may have.

No callback supported by the Promise-based APIs on Chrome

While some of the asynchronous API methods in Firefox (the ones that return a promise) also support the callback parameter (mostly as a side effect of the backward compatibility with the callback-based APIs available on Chrome), the Promise-based APIs provided by this library do not support the callback parameter (See "#102 Cannot call browser.storage.local.get with callback").

No promise returned on Chrome for some API methods

This library takes its knowledge of the APIs to wrap and their signatures from a metadata JSON file: api-metadata.json.

If an API method is not yet included in this "API metadata" file, it will not be recognized. Promises are not supported for unrecognized APIs, and callbacks have to be used for them.

Chrome-only APIs have no promise version, because extensions that use such APIs would not be compatible with Firefox.

File an issue in this repository for API methods that support callbacks in Chrome and Firefox but are currently missing from the "API metadata" file.

Issues that happen only when running on Firefox

When an extension that uses this library doesn't behave as expected on Firefox, it is almost never an issue in this polyfill, but an issue with the native implementation in Firefox.

"Firefox only" issues should be reported upstream on Bugzilla:

API methods or options that are only available when running in Firefox

This library does not provide any polyfill for API methods and options that are only available on Firefox, and they are actually considered out of the scope of this library.

tabs.executeScript

On Firefox browser.tabs.executeScript returns a promise which resolves to the result of the content script code that has been executed, which can be an immediate value or a Promise.

On Chrome, the browser.tabs.executeScript API method as polyfilled by this library also returns a promise which resolves to the result of the content script code, but only immediate values are supported. If the content script code result is a Promise, the promise returned by browser.tabs.executeScript will be resolved to undefined.

MSEdge support

With the Microsoft announcement related to the MSEdge future there aren't many compelling reasons to officially include in this polyfill any workarounds specific to the current MSEdge versions (and then maintaining them over the time), and so MSEdge is currently unsupported.

Once the first Chrome-based MSEdge version is going to be released, we will be able to verify if any actual changes to this polyfill are needed to officially support the Chrome-based MSEdge versions.

In the meantime, for extension developers that still have to work on MSEdge extensions before it is switched to the next Chromium-based version, the MSEdge --ms-preload manifest key and the Microsoft Edge Extension Toolkit's Chrome API bridge can be used to be able to load the webextension-polyfill without any MSEdge specific changes.

The following Github repository provides some additional detail about this strategy and a minimal test extension that shows how to put it together:

Contributing to this project

Read the contributing section for additional information about how to build the library from this repository and how to contribute and test changes.