French version here.

ScreenshotHTML

ScreenshotHTML allow you to take screenshots of HTML code, using Puppeteer (a Chromium-automated-based browser).

⚠️ ScreenshotHTML use headless mode by default, so no navigation window will be visible.

Installation

NodeJS 14 (or latest) is required.

# Using npm
npm i screenshothtml

# Or using yarn
yarn add screenshothtml

# Or using pnpm
pnpm i screenshothtml

This library does not support browsers, you can use html2canvas and canvas2image to get a similar result.

Options

var image = await screenshot({ <the options> })

Option	Type	Example	Description
html	string	<h1>Hello world!</h1>	HTML code to capture
maxTimeout	int	15000	Maximum time to wait for the page to load (in ms)
viewportWidth	int	1600	Viewport width
viewportHeight	int	900	Viewport height
autoCloseAfter	int	100000	Edit the time (in ms) required before the browser close after the screenshot
puppeteerArguments	object	{headless: false}	Add additional arguments passed to Puppeteer while creating the browser
waitForDomLoaded	boolean	false	Wait before the DOM finish to load before taking a screenshot, more info here.

APIs

const screenshot = require('screenshothtml')

Main :

screenshot()

Do a screenshot of the HTML code.

list of options.

Close the browser :

await screenshot.close()

Manually close the browser.

Use the browser :

screenshot.browser

Return the browser instance, as it is used by the library.

Last screenshot :

screenshot.browserLastUse

Return the date corresponding to the last screenshot.

Puppeteer arguments :

screenshot.puppeteerArgs

Return the arguments passed to Chromium while creating the browser, do not confuse with the option puppeteerArguments that is different.

ScreenshotHTML version :

screenshot.version

Return the used version of the library.

Examples

You can find examples in the example folder of this repo.

Speed

When the environment variable SCREENSHOTHTML_SHOW_STATISTICS is set, statistics about the speed are displayed in the terminal.

Some results on the examples in this repo, tested on my computer.

With the option waitForDomLoaded defined with false:

• browserOpen: ~330ms
• pageOpen: ~115ms
• gotoPage: ~990ms
• screenshot: ~310ms
• closePage: ~3ms

With the option waitForDomLoaded defined avec true :

• browserOpen: ~220ms
• pageOpen: ~120ms
• gotoPage: ~20ms
• screenshot: ~480ms
• closePage: ~4ms

Load a URL instead of code

Instead of inserting HTML code, you can load a URL by replacing the code by url:<the url of the site>. This also works with localhost.

Browser close

After a screenshot, the browser is automatically closed after 60 seconds, you can change this time with the option autoCloseAfter (negative number cancel the auto close, and 0 close without waiting).

The browser can also be closed manually with the close() function.

waitForDomLoaded

The option waitForDomLoaded (boolean) can be used to define if the browser should wait for the DOM to be loaded before taking a screenshot. If the value is false (or not defined), the browser will use networkidle0, and when it is activated (true), the screenshots are faster, but some elements and scripts may be not load, which can be useful if the page only contains a simple HTML code.

License

MIT © Johan