Fury Rendering Server

Fury is a Node.js wrapper for the Puppeteer screenshot and pdf function.

You can POST to /screenshot or /pdf to create either a screenshot or generate a pdf from a given URL. All parameters have to be sent in the POST body and they have to be in JSON syntax.

Unless renderEventName isn't set, Fury will wait for the screenshot or pdf creation until no network request is made for at least 500 milliseconds.

All given parameters will be forwarded to the target page via the FURY_PARAMS object in the window object of the browser.

Environment Settings

There are several environment variables which Fury is using:

• DEBUG: If DEBUG=1, Fury will open the browser not in headless mode and open the development tools in each browser.
• CONCURRENT_LIMIT: Defines the amount of max. concurrent rendering processes. Default is one concurrent rendering.
• TIMEOUT: Defines the timeout for all requests in milliseconds. Default: 30000 (30 seconds)

Request

Take a screenshot

/screenshot

Available Parameters

• url (Required): The URL to the target page for the screenshot
• width: Width in pixel of the resulting image. Default: 1920.
• height: Height in pixel of the resulting image. Default: 1080.
• cookies: Cookies which will be set before taking the screenshot. You can find a detailed explanation below.
• renderEventName: The RenderFunctionName. You can find a detailed explanation below.

Example

curl --location --request POST 'http://localhost:3000/screenshot' \
--header 'Content-Type: application/json' \
--data-raw '{
    "url": "https://google.com",
    "width": 800,
    "height": 800,
    "timeout": 60000,
    "cookies": {
        "cookie1": "value1"
    }
}'

This request will set the cookie named cookie with value1 and take a screenshot from https://google.com with dimensions of 800px x 800px. The POST will fail if the page doesn't respond after 60 seconds.

---

Create a PDF

/pdf Creates a PDF of the given page by simulating a printout.

Available Parameters

• url (Required): The URL to the target page for the screenshot
• width: Width of the page. If format is defined, this will be ignored. This parameter will be forwarded to Puppeteer, so you can find a detailed explanation here.
• height: Height of the page. If format is defined, this will be ignored. This parameter will be forwarded to Puppeteer, so you can find a detailed explanation here.
• format: Format of the page. This parameter will be forwarded to Puppeteer, so you can find a detailed explanation here.
• cookies: Cookies which will be set before taking the screenshot. You can find a detailed explanation below.
• renderEventName: The RenderFunctionName. You can find a detailed explanation below.
• printBackground: Include background media in the pdf ? This parameter will be forwarded to Puppeteer, so you can find a detailed explanation here.
• landscape: Is the page landscape orientated ? Only works, if format is provided. This parameter will be forwarded to Puppeteer, so you can find a detailed explanation here.
• margin: Margin for the printout. This parameter will be forwarded to Puppeteer, so you can find a detailed explanation here.

Example

curl --location --request POST 'http://localhost:3000/pdf' \
--header 'Content-Type: application/json' \
--data-raw '{
    "url": "https://google.com",
    "format": "A4",
    "printBackground":  true,
    "landscape": true,
    "cookies": {
        "cookie1": "value1"
    }
}

This request will create a pdf by simulating a printout of https://google.com. The page size in the pdf is A4 in landscape mode. Additionally, Fury will set the cookie cookie1 with the value value.

---

Defining Required Cookies

Both methods supports the cookies parameter to define some required cookies to process the page. This can be used for example to take a screenshot of a page which requires a login cookie.

The syntax is pretty straight forward. You define the cookies in JSON syntax, where the key represents the cookie name and the value is the value of the cookie. Syntax:

{
 "cookiename1": "cookievalue1",
 "cookiename2": "cookievalue2"
} 

RenderEventName

If the target page have to do some work before the page is ready for the screenshot, you can define with the renderEventName parameter a function name for which Fury will wait. The function name will be exposed to the global window object of the browser. After calling the function from the target page, the screenshot or the pdf will be created unless the timeout expires before.

Page errors

A page can provide an error message which will be considered after page navigation. To provide an error message, just create a <div class="furyError"> on the page and fill it with the error message if you don't want to print the page.

If this div exists and has a content, the content of the div will be returned as error.

Response

After processing the target page, Fury will send a response in JSON syntax:

{
  "data": "",
  "error": ""
}

data: The image or pdf encoded via base64 or null, if an error occured.

error: If an error occured, this will contain the error message and call stack of the error. Otherwise null.

MessagePack

If the HTTP request has the string msgpack in the HTTP Accept: header, Fury will return the same response as a MessagePack map. Because MessagePack allows binary data, this is more efficient because the data doesn't have to be encoded to base64 and it allows to return larger files.