[fuente: http://developer.chrome.com/extensions/getstarted.html]

Extensions allow you to add functionality to Chrome without diving deeply into native code. You can create new extensions for Chrome with those core technologies that you’re already familiar with from web development: HTML, CSS, and JavaScript. If you’ve ever built a web page, you should feel right at home with extensions pretty quickly; we’ll put that to the test right now by walking through the construction of a simple extension that will give you one-click access to pictures of kittens. Kittens!

We’ll do so by implementing a UI element we call a browser action, which allows us to place a clickable icon right next to Chrome’s Omnibox for easy access. Clicking that icon will open a popup window filled with kittenish goodness, which will look something like this:

If you’d like to follow along at home (and you should!), create a shiny new directory on your computer, and pop open your favourite text editor. Let’s get going!

Something to Declare

The very first thing we’ll need to create is a manifest file named manifest.json. The manifest is nothing more than a JSON-formatted table of contents, containing properties like your extension’s name and description, its version number, and so on. At a high level, we’ll use it to declare to Chrome what the extension is going to do, and what permissions it requires in order to do those things.

In order to display kittens, we’ll want to tell Chrome that we’d like to create a browser action, and that we’d like free-reign to access kittens from a particular source on the net. A manifest file containing those instructions looks like this:

{
  "manifest_version": 2,

  "name": "One-click Kittens",
  "description": "This extension demonstrates a browser action with kittens.",
  "version": "1.0",

  "permissions": [
    "https://secure.flickr.com/"
  ],
  "browser_action": {
    "default_icon": "icon.png",
    "default_popup": "popup.html"
  }
}

Go ahead and save that data to a file named manifest.json in the directory you created, or download a copy of manifest.json from our sample repository .

What does it mean?

The attribute names are fairly self-descriptive, but let’s walk through the manifest line-by-line to make sure we’re all on the same page.

The first line, which declares that we’re using version 2 of the manifest file format, is mandatory (version 1 is old, deprecated, and generally not awesome).

The next block defines the extension’s name, description, and version. These will be used both inside of Chrome to show a user which extensions you have installed, and also on the Chrome Web Store to display your extension to potentially new users. The name should be short and snappy, and the description no longer than a sentence or so (you’ll have more room for a detailed description later).

The final block first requests permission to work with data on https://secure.flickr.com/, and declares that this extension implements a browser action, assigning it a default icon and popup in the process.

Resources

You probably noticed that manifest.json pointed at two resource files when defining the browser action: icon.png and popup.html. Both resources must exist inside the extension package, so let’s create them now:

• icon.png will be displayed next to the Omnibox, waiting for user interaction. Download a copy of icon.png from our sample repository, Download a copy of icon.png from our sample repository , and save it into the directory you’re working in. You could also create your own if you’re so inclined; it’s just a 19px-square PNG file.
• popup.html will be rendered inside the popup window that’s created in response to a user’s click on the browser action. It’s a standard HTML file, just like you’re used to from web development, giving you more or less free reign over what the popup displays. Download a copy of popup.html from our sample repository , and save it into the directory you’re working in.popup.html requires an additional JavaScript file in order to do the work of grabbing kitten images from the web and loading them into the popup. To save you some effort, just download a copy of popup.js from our sample repository , and save it into the directory you’re working in.

You should now have four files in your working directory: icon.png, manifest.json, popup.html, popup.js. The next step is to load those files into Chrome.

Load the extension

Extensions that you download from the Chrome Web Store are packaged up as .crx files, which is great for distribution, but not so great for development. Recognizing this, Chrome gives you a quick way of loading up your working directory for testing. Let’s do that now.

1. Visit chrome://extensions in your browser (or open up the Chrome menu by clicking the icon to the far right of the Omnibox:. and select Extensions under the Tools menu to get to the same place).
2. Ensure that the Developer Mode checkbox in the top right-hand corner is checked.
3. Click Load unpacked extension… to pop up a file-selection dialog.
4. Navigate to the directory in which your extension files live, and select it.

If the extension is valid, it’ll be loaded up and active right away! If it’s invalid, an error message will be displayed at the top of the page. Correct the error, and try again.

Fiddle with Code

Now that you’ve got your first extension up and running, let’s fiddle with things so that you have an idea what your development process might look like. As a trivial example, let’s change the data source to search for pictures of puppies instead of kittens.

Hop into popup.js, and edit line 11 from var QUERY = 'kittens'; to read var QUERY = 'puppies';, and save your changes.

If you click on your extension’s browser action again, you’ll note that your change hasn’t yet had an effect. You’ll need to let Chrome know that something has happened, either explicitly by going back to the extension page (chrome://extensions, or Tools > Extensions under the Chrome menu), and clicking Reload under your extension, or by reloading the extensions page itself (either via the reload button to the left of the Omnibox, or by hitting F5 or Ctrl-R).

Once you’ve reloaded the extension, click the browser action icon again. Puppies galore!

What next?

You now know about the manifest file’s central role in bringing things together, and you’ve mastered the basics of declaring a browser action, and rendering some kittens (or puppies!) in response to a user’s click. That’s a great start, and has hopefully gotten you interested enough to explore further. There’s a lot more out there to play around with.

• The Chrome Extension Overview backs up a bit, and fills in a lot of detail about extensions’ architecture in general, and some specific concepts you’ll want to be familiar with going forward. It’s the best next step on your journey towards extension mastery.
• No one writes perfect code on the first try, which means that you’ll need to learn about the options available for debugging your creations. Our debugging tutorial is perfect for that, and is well worth carefully reading.
• Chrome extensions have access to powerful APIs above and beyond what’s available on the open web: browser actions are just the tip of the iceburg. Our chrome.* APIs documentation will walk you through each API in turn.
• Finally, the developer’s guide has dozens of additional links to pieces of documentation you might be interested in.

Overview

Once you’ve finished this page and the Getting Started tutorial, you’ll be all set to start writing extensions.

The basics

An extension is a zipped bundle of files—HTML, CSS, JavaScript, images, and anything else you need—that adds functionality to the Google Chrome browser. Extensions are essentially web pages, and they can use all the APIs that the browser provides to web pages, from XMLHttpRequest to JSON to HTML5.

Extensions can interact with web pages or servers using content scripts or cross-origin XMLHttpRequests. Extensions can also interact programmatically with browser features such as bookmarks and tabs.

Extension UIs

Many extensions—but not packaged apps—add UI to Google Chrome in the form of browser actions or page actions. Each extension can have at most one browser action or page action. Choose a browser action when the extension is relevant to most pages. Choose a page action when the extension’s icon should appear or disappear, depending on the page.

This mail extension uses a browser action(icon in the toolbar).	This map extension uses a page action(icon in the address bar) and content script(code injected into a web page).	This news extension features a browser action that, when clicked, shows a popup.

Extensions (and packaged apps) can also present a UI in other ways, such as adding to the Chrome context menu, providing an options page, or using a content script that changes how pages look. See the Developer’s Guide for a complete list of extension features, with links to implementation details for each one.

Files

Each extension has the following files:

• A manifest file
• One or more HTML files (unless the extension is a theme)
• Optional: One or more JavaScript files
• Optional: Any other files your extension needs—for example, image files

While you’re working on your extension, you put all these files into a single folder. When you distribute your extension, the contents of the folder are packaged into a special ZIP file that has a .crx suffix. If you upload your extension using the Chrome Developer Dashboard, the .crx file is created for you. For details on distributing extensions, see Hosting.

Referring to files

You can put any file you like into an extension, but how do you use it? Usually, you can refer to the file using a relative URL, just as you would in an ordinary HTML page. Here’s an example of referring to a file named myimage.png that’s in a subfolder named images.

<img src="images/myimage.png">

As you might notice while you use the Google Chrome debugger, every file in an extension is also accessible by an absolute URL like this:

chrome-extension://<extensionID>/<pathToFile>

In that URL, the <extensionID> is a unique identifier that the extension system generates for each extension. You can see the IDs for all your loaded extensions by going to the URL chrome://extensions. The <pathToFile> is the location of the file under the extension’s top folder; it’s the same as the relative URL.

While you’re working on an extension (before it’s packaged), the extension ID can change. Specifically, the ID of an unpacked extension will change if you load the extension from a different directory; the ID will change again when you package the extension. If your extension’s code needs to specify the full path to a file within the extension, you can use the @@extension_id predefined message to avoid hardcoding the ID during development.

When you package an extension (typically, by uploading it with the dashboard), the extension gets a permanent ID, which remains the same even after you update the extension. Once the extension ID is permanent, you can change all occurrences of @@extension_id to use the real ID.

The manifest file

The manifest file, called manifest.json, gives information about the extension, such as the most important files and the capabilities that the extension might use. Here’s a typical manifest file for a browser action that uses information from google.com:

{
  "name": "My Extension",
  "version": "2.1",
  "description": "Gets information from Google.",
  "icons": { "128": "icon_128.png" },
  "background": {
    "persistent": false,
    "scripts": ["bg.js"]
  },
  "permissions": ["http://*.google.com/", "https://*.google.com/"],
  "browser_action": {
    "default_title": "",
    "default_icon": "icon_19.png",
    "default_popup": "popup.html"
  }
}

For details, see Manifest Files.

Architecture

Many extensions have a background page, an invisible page that holds the main logic of the extension. An extension can also contain other pages that present the extension’s UI. If an extension needs to interact with web pages that the user loads (as opposed to pages that are included in the extension), then the extension must use a content script.

The background page

The following figure shows a browser that has at least two extensions installed: a browser action (yellow icon) and a page action (blue icon). Both the browser action and the page action have background pages. This figure shows the browser action’s background page, which is defined by background.html and has JavaScript code that controls the behavior of the browser action in both windows.

There are two types of background pages: persistent background pages, and event pages. Persistent background pages, as the name suggests, are always open. Event pages are opened and closed as needed. Unless you absolutely need your background page to run all the time, prefer to use an event page.

See Event Pages and Background Pages for more details.

UI pages

Extensions can contain ordinary HTML pages that display the extension’s UI. For example, a browser action can have a popup, which is implemented by an HTML file. Any extension can have an options page, which lets users customize how the extension works. Another type of special page is the override page. And finally, you can use tabs.create or window.open() to display any other HTML files that are in the extension.

The HTML pages inside an extension have complete access to each other’s DOMs, and they can invoke functions on each other.

The following figure shows the architecture of a browser action’s popup. The popup’s contents are a web page defined by an HTML file (popup.html). This extension also happens to have a background page (background.html). The popup doesn’t need to duplicate code that’s in the background page because the popup can invoke functions on the background page.

See Browser Actions, Options, Override Pages, and the Communication between pages section for more details.

Content scripts

If your extension needs to interact with web pages, then it needs a content script. A content script is some JavaScript that executes in the context of a page that’s been loaded into the browser. Think of a content script as part of that loaded page, not as part of the extension it was packaged with (its parent extension).

Content scripts can read details of the web pages the browser visits, and they can make changes to the pages. In the following figure, the content script can read and modify the DOM for the displayed web page. It cannot, however, modify the DOM of its parent extension’s background page.

Content scripts aren’t completely cut off from their parent extensions. A content script can exchange messages with its parent extension, as the arrows in the following figure show. For example, a content script might send a message whenever it finds an RSS feed in a browser page. Or a background page might send a message asking a content script to change the appearance of its browser page.

For more information, see Content Scripts.

Using the chrome.* APIs

In addition to having access to all the APIs that web pages and apps can use, extensions can also use Chrome-only APIs (often called chrome.* APIs) that allow tight integration with the browser. For example, any extension or web app can use the standard window.open() method to open a URL. But if you want to specify which window that URL should be displayed in, your extension can use the Chrome-only tabs.create method instead.

Asynchronous vs. synchronous methods

Most methods in the chrome.* APIs are asynchronous: they return immediately, without waiting for the operation to finish. If you need to know the outcome of that operation, then you pass a callback function into the method. That callback is executed later (potentially much later), sometime after the method returns. Here’s an example of the signature for an asynchronous method:

chrome.tabs.create(object createProperties, function callback)

Other chrome.* methods are synchronous. Synchronous methods never have a callback because they don’t return until they’ve completed all their work. Often, synchronous methods have a return type. Consider the runtime.getURL method:

string chrome.runtime.getURL()

This method has no callback and a return type of string because it synchronously returns the URL and performs no other, asynchronous work.

Example: Using a callback

Say you want to navigate the user’s currently selected tab to a new URL. To do this, you need to get the current tab’s ID (usingtabs.query) and then make that tab go to the new URL (using tabs.update).

If query() were synchronous, you might write code like this:

   //THIS CODE DOESN'T WORK
1: var tab = chrome.tabs.query({'active': true}); //WRONG!!!
2: chrome.tabs.update(tab.id, {url:newUrl});
3: someOtherFunction();

That approach fails because query() is asynchronous. It returns without waiting for its work to complete, and it doesn’t even return a value (although some asynchronous methods do). You can tell that query() is asynchronous by the callback parameter in its signature:

chrome.tabs.query(object queryInfo, function callback)

To fix the preceding code, you must use that callback parameter. The following code shows how to define a callback function that gets the results from query() (as a parameter named tab) and calls update().

   //THIS CODE WORKS
1: chrome.tabs.query({'active': true}, function(tabs) {
2:   chrome.tabs.update(tabs[0].id, {url: newUrl});
3: });
4: someOtherFunction();

In this example, the lines are executed in the following order: 1, 4, 2. The callback function specified to query() is called (and line 2 executed) only after information about the currently selected tab is available, which is sometime after query() returns. Althoughupdate() is asynchronous, this example doesn’t use its callback parameter, since we don’t do anything about the results of the update.

More details

For more information, see the chrome.* API docs and watch this video:

Communication between pages

The HTML pages within an extension often need to communicate. Because all of an extension’s pages execute in same process on the same thread, the pages can make direct function calls to each other.

To find pages in the extension, use chrome.extension methods such as getViews() and getBackgroundPage(). Once a page has a reference to other pages within the extension, the first page can invoke functions on the other pages, and it can manipulate their DOMs.

Saving data and incognito mode

Extensions can save data using the HTML5 web storage API (such as localStorage) or by making server requests that result in saving data. Whenever you want to save something, first consider whether it’s from a window that’s in incognito mode. By default, extensions don’t run in incognito windows. You need to consider what a user expects from your extension when the browser is incognito.

Incognito mode promises that the window will leave no tracks. When dealing with data from incognito windows, do your best to honor this promise. For example, if your extension normally saves browsing history to the cloud, don’t save history from incognito windows. On the other hand, you can store your extension’s settings from any window, incognito or not.

Rule of thumb: If a piece of data might show where a user has been on the web or what the user has done, don’t store it if it’s from an incognito window.

To detect whether a window is in incognito mode, check the incognito property of the relevant tabs.Tab or windows.Window object. For example:

function saveTabData(tab, data) {
  if (tab.incognito) {
    chrome.runtime.getBackgroundPage(function(bgPage) {
      bgPage[tab.url] = data;      // Persist data ONLY in memory
    });
  } else {
    localStorage[tab.url] = data;  // OK to store data
  }
}

Now what?

Now that you’ve been introduced to extensions, you should be ready to write your own. Here are some ideas for where to go next:

• Tutorial: Getting Started
• Tutorial: Debugging
• Developer’s Guide
• Samples
• Videos, such as Extension Message Passing