Create an extension that customizes the DevTools UI

아티클
03/28/2023

This tutorial shows how to build your own DevTools extension from scratch. It's a good way to experience everything that's needed for typical development. At the end of the tutorial, you'll have a working extension that creates its own panel and interacts with the DevTools APIs.

First review the prerequisites. Then download and run the final extension from the Solution section to see what you'll be building in this tutorial. Otherwise, start at Step 1 - Create a Basic DevTools Extension to build the extension from scratch.

Prerequisites

Microsoft Edge Browser to run the extension.
A code editor such as Visual Studio Code to follow along with the tutorial.

Solution

In this section, you download and run the final state of the extension that the rest of this tutorial will teach you to build. Later, you'll start from scratch and learn to write the code to create your own extension.

Get the final extension code by downloading this zip file and extracting its content to a folder on your computer.
Open Microsoft Edge and go to edge://extensions/.
Turn on the Developer mode toggle.
Click Load unpacked and navigate to the folder where you extracted the zip file. In this folder, select Demos-main > devtools-extension > sample 4, and then click Select Folder.

The edge://extensions page in Microsoft Edge, showing the Developer mode and Load unpacked buttons

Your DevTools extension should now be loaded. Open a new tab in Microsoft Edge, and then open DevTools by pressing F12.

The extension creates a Sample Panel tab in the DevTools window:

Microsoft Edge, showing DevTools on the side, with the extension's Sample Panel selected

Step 1 - Create a basic DevTools extension

A basic DevTools extension consists of two files, as shown in Step 1 code:

A manifest file.

{
    "name": "DevTools Sample Extension",
    "description": "A Basic DevTools Extension",
    "manifest_version": 3,
    "version": "1.0",
    "devtools_page": "devtools.html"
}

Property	Description
name	The name of the extension that will appear under `edge://extensions/`.
description	The description of the extension that will be displayed under the name of the extension.
version	The version of the extension that will appear next to the name of the extension.
manifest_version	Determines the set of features that the extension will be using, such as service workers or network request modification. The current version is version `3`. To learn more about this version and the differences with version `2`, see Overview and timelines for migrating to Manifest V3.
devtools_page	The path to an HTML file that will be run every time the DevTools UI is opened. Though the page is not rendered in DevTools, it will be used to load the necessary JavaScript files for the extension.

An html file to match the devtools_page field in the manifest file.

<!DOCTYPE html>
<html>
  <head>
    <meta charset="UTF-8" />
  </head>
  <body>
    A Basic DevTools Extension.
  </body>
</html>

To load and test your DevTools extension in Microsoft Edge, use the Developer mode:

In Microsoft Edge, go to edge://extensions/.
Turn on the Developer mode toggle.
Click Load unpacked, navigate to the folder where you wrote the code for your extension, and click Select folder.

Your DevTools extension is now loaded. Open a new tab and then open DevTools (F12).

Your extension is loaded, but isn't displayed in DevTools yet, because you haven't created a panel for the extension.

Step 2 - Add a DevTools panel

In this step, you'll create a new panel in DevTools. You can find the code for this step in Step 2 code, or write it yourself by following the instructions below.

A panel is a tab in the main toolbar in DevTools, similar to the Elements, Console, and Sources tools shown below:

DevTools panels and sidebars

In this step, we will create a basic DevTools extension with a sample panel.

Create a devtools.js file with the following code:
```
chrome.devtools.panels.create("Sample Panel", "icon.png", "panel.html", panel => {
    // code invoked on panel creation
});
```
The create method requires four parameters: title, iconpath, pagePath, and a callback function. Note that while the panel's icon is a required parameter, it does not get displayed in Microsoft Edge DevTools currently.

To learn more about the chrome.devtools.panels extension API, see the chrome.devtools.panels API reference.
Load the file from the HTML page specified in the devtools_page manifest field earlier (the devtools.html file), by adding a script element in the source code.
```
<!DOCTYPE html>
<html>
  <head>
    <meta charset="UTF-8" />
  </head>
  <body>
    <script src="devtools.js"></script>
  </body>
</html>
```
Create the panel.html file the you referenced in the previous chrome.devtools.panels.create method call. This webpage will contain the user interface of the panel your extension is adding to DevTools.
```
<!DOCTYPE html>
<html>
  <head>
    <meta charset="UTF-8" />
  </head>
  <body>
    <h1>A Basic DevTools Extension with Panel</h1>
  </body>
</html>
```

To test your changes in Microsoft Edge, reload your extension from the edge://extensions/ page:

Navigate (or go back) to the edge://extensions/ page.
Find the unpacked extension entry you loaded in step 1.
Click Reload.

Your DevTools extension should now be reloaded. Open a new tab in Microsoft Edge and then open DevTools (F12). Your DevTools extension panel should be displayed in DevTools:

Microsoft Edge, with DevTools on the side, showing the new extension panel

Step 3 - Call extension APIs from a DevTools extension

In this step, you will use extension APIs to display memory information in your DevTools panel. To do this, we will need to update the permissions in the manifest file, the panel interface, and the devtools script. The source code for this step can be found in Step 3 code, or write it yourself by following the instructions below.

Use the permissions manifest member in your manifest.json file. This member defines which permissions your extension requires from the user. Some permissions are needed to use certain extension APIs.
```
"permissions": [
  "system.memory",
]
```
The system-memory permission is required to use extension APIs we will use later in this tutorial. To learn more about the available APIs and associated permissions, see the API reference.

Add the following to the body in the panel.html file to display the data in the panel.

<div>
  Available Memory Capacity: <span id="availableMemoryCapacity"></span>
</div>
<div>
  Total Memory Capacity: <span id="totalMemoryCapacity"></span>
</div>

Update the devtools.js file with the following code.

let availableMemoryCapacity;
let totalMemoryCapacity;

chrome.devtools.panels.create("Sample Panel", "icon.png", "panel.html", panel => {
    // code invoked on panel creation
    panel.onShown.addListener((extPanelWindow) => {
        availableMemoryCapacity = extPanelWindow.document.querySelector('#availableMemoryCapacity');
        totalMemoryCapacity = extPanelWindow.document.querySelector('#totalMemoryCapacity');
    });
});

setInterval(() => {
    chrome.system.memory.getInfo((data) => {
        if (availableMemoryCapacity) {
            availableMemoryCapacity.innerHTML = data.availableCapacity;
        }
        if (totalMemoryCapacity) {
            totalMemoryCapacity.innerHTML = data.capacity;
        }
    });
}, 1000);

The above code snippet does the following:

It creates a new panel Sample Panel in DevTools.
When the panel is displayed (panel.onShown listener), the availableMemoryCapacity and totalMemoryCapacity elements are retrieved from the DOM.
Next, a timer is set to run code every second after the panel is shown.
When the timer fires, the chrome.system.memory.getInfo method is used to retrieve the available and total memory capacity of the device and these values are displayed in the corresponding DOM elements.

To test your changes in Microsoft Edge, reload your extension from the edge://extensions/ page:

Navigate (or go back) to the edge://extensions/ page.
Find the unpacked extension entry you loaded in step 1.
Click Reload.

Your DevTools extension should now be reloaded. Open a new tab and then open DevTools (F12). Your DevTools extension panel should now show the available and total memory capacities.

The new extension panel in DevTools, showing memory information

More extension APIs can be found in the API Reference page.

Step 4 - Interact with the inspected webpage

In this step of the tutorial, you will add code that interacts with the inspected webpage. That code will:

Listen to click events that happen on the webpage and log them into the DevTools Console tool.
Show the mouse click position in the DevTools extension panel.
Show a greeting alert in the inspected page when user clicks a button in the DevTools extension panel.

The source code for this step can be found in Step 4 code, or write it yourself by following the instructions below.

Unfortunately, the DevTools panel you just created does not have direct access to the inspected webpage and does not run until DevTools is opened. For this you will use a content script and a background service worker.

A content script runs in the context of the inspected webpage and, in the same way other scripts are loaded by the page, they have access to the DOM and can change it.
A background service worker is a script that the browser runs in a separate thread and has access to extension APIs.

The following diagram gives an overview of how the DevTools page, inspected page, content script, and background service worker fit together in an extension.

The anatomy of a DevTools extension

In this part of the tutorial, you will detect the user clicks on a webpage using a content script. The content script will relay this info to the devtools.js file where the data will be displayed in both the console and the DevTools extension panel.

Append the following to the manifest.json file

"content_scripts": [{
  "matches": [
    "http://*/*",
    "https://*/*"
  ],
  "run_at": "document_idle",
  "js": [
    "content_script.js"
  ]
}],
"background": {
    "service_worker": "background.js"
}

Property	Value
matches	Specifies which pages this content script will be injected into.
run_at	Indicates when the browser injects the script onto the page.
js	The javascript files to be injected.

Create the file content_script.js with the following content:
```
document.addEventListener("click", (event) => {
  chrome.runtime.sendMessage({
      click: true,
      xPosition: event.clientX + document.body.scrollLeft,
      yPosition: event.clientY + document.body.scrollTop
    },
    response => {
      console.log("Received response", response);
    }
  );
});
```
The above code snippet prints a message to the console when the script is injected in the page. It also adds a click event listener to the page that will send a message with mouse click position in the inspected page by using the chrome.runtime.sendMessage API.
In the panel.html file, add a sayHello button and a youClickedOn label . These two elements are used to demo the interaction between the inspected page, the DevTools panel, and the background service worker. When the user clicks the sayHello button in the DevTools extension, it will display a greeting message in the inspected window. When the user clicks anywhere in the inspected page, it will display a message to show the mouse click position in the DevTools extension panel.
```
<button id="sayHello">Say Hello to the inspected page!</button>
<h2><span id="youClickedOn"></span></h2>
```

In the devtools.js file, use the chrome.runtime.connect method to create a connection to the background service worker, and then send the inspected window tabId to the service worker by using the backgroundPageConnection.postMessage method. Finally, add event listeners to the sayHello button and youClickedOn label that's defined in the panel.html file.

let youClickedOn; 
chrome.devtools.panels.create("Sample Panel", "icon.png", "panel.html", panel => {
    // code invoked on panel creation
    panel.onShown.addListener( (extPanelWindow) => {
        let sayHello = extPanelWindow.document.querySelector('#sayHello');
        youClickedOn = extPanelWindow.document.querySelector('#youClickedOn');
        sayHello.addEventListener("click", () => {
            // show a greeting alert in the inspected page
            chrome.devtools.inspectedWindow.eval('alert("Hello from the DevTools extension");');
        });             
    });
});

chrome.runtime.onMessage.addListener((request, sender, sendResponse) => {
    // Messages from content scripts should have sender.tab set
    if (sender.tab && request.click == true) {
        console.log('I am here!');
        if (youClickedOn) {
            youClickedOn.innerHTML = `You clicked on position (${request.xPosition}, ${request.yPosition}) in the inspected page.`;
        }
        sendResponse({
            xPosition: request.xPosition,
            yPosition: request.yPosition
        });
    }
});

// Create a connection to the background service worker
const backgroundPageConnection = chrome.runtime.connect({
    name: "devtools-page"
});

// Relay the tab ID to the background service worker
backgroundPageConnection.postMessage({
    name: 'init',
    tabId: chrome.devtools.inspectedWindow.tabId
});

When the user clicks the sayHello button, the DevTools extension will run a code snippet of alert("Hello from the DevTools Extension"); in the inspected window by invoking the eval() method of the inspected window chrome.devtools.inspectedWindow.

When the user clicks anywhere in the inspected window, the DevTools extension will receive a message, from the background service worker, with request.click == true and the mouse position information.

Create the background.js file and add the following code to it:

let id = null;
const connections = {};

chrome.runtime.onConnect.addListener(devToolsConnection => {
    // Assign the listener function to a variable so we can remove it later
    let devToolsListener = (message, sender, sendResponse) => {
        if (message.name == "init") {
            id = message.tabId;
            connections[id] = devToolsConnection;
            // Send a message back to DevTools
            connections[id].postMessage("Connected!");
        }
    };

    // Listen to messages sent from the DevTools page
    devToolsConnection.onMessage.addListener(devToolsListener);

    devToolsConnection.onDisconnect.addListener(() => {
        devToolsConnection.onMessage.removeListener(devToolsListener);
    });
});

The above code snippet connects the background service worker with the DevTools page. It listens to when the DevTools page connects, saves the connection, and sends a response back to the DevTools page.

This is useful when your background service worker is collecting data or performing tasks in the background that you want to be available in your DevTools extension.

To test the new sayHello button:

Go to edge://extensions/.
Find the unpacked extension entry that you loaded in Step 1.
Click the Reload button.
Open a new browser tab, open DevTools (F12), and click the Sample Panel tab.
Click the "Say Hello to The Inspected Page!" button in your panel. You should see an alert in the inspected page, as shown below.

To learn more about content scripts, see the Content Scripts documentation.

To test the youClickedOn label:

Click OK to close the alert window opened in the previous step.
Click anywhere in the inspected page. A You clicked on position (x,y) in the inspected page message should appear in the extension panel.
Open the Console tool. A message similar to the previous step should also appear in the Console tool.

다음을 통해 공유