universalCaptchaSolver

This powerful function allows you to solve any captcha without having to identify it. To do this, Grimport creates a split-screen session between a specific section of your PC screen and the device of a human agent solving the captcha. The advantage is that mouse movements and keyboard typing are faithfully reproduced as if the captcha solver were actually behind the screen and the fingerprint is not altered (the solver does not pass through a proxy, there is no DSN leak and the fingerprint is not altered by a browser with different dimensions, for example).
This function requires the captcha to be loaded in Firefox Developper and the Grimport Bridge extension to be up to date. To do this, use the firefox function.
The function is synchronised between crawlers so that it cannot be executed twice simultaneously.

Installation

Advice

Example

setCloudCode(111111)

firefox("browse", "https://captcha.com/demos/features/captcha-demo.aspx")

firefox("waitLoaded")



universalCaptchaSolver("#demoForm fieldset", { ->

	if(firefox("javascript", """ document.querySelector(".correct") ? document.querySelector(".correct").innerHTML : '' """)) return true 

	else return false

})

setCloudCode(111111)

firefox("setHeaders", [

	/.*/ : [ "Accept-Language": "en-US,en;q=0.5" ] //we ensure that all pages are in English

])



firefox("browse", "https://nopecha.com/demo/hcaptcha#easy")

firefox("waitLoaded")



positionIframe = firefox("positionInScreen", "iframe") //capture the initial position

screenZone = [ //extension of the area

	"x": get(positionIframe, "x"),

	"y": get(positionIframe, "y") - 50,

	"width": get(positionIframe, "width") + 500,

	"height": get(positionIframe, "height") + 650,

	"from": "screenInFirefox",

]



universalCaptchaSolver(

	screenZone, 

	{ ->

		if(firefox("javascript", """ document.querySelector(".response") ? document.querySelector(".response").innerHTML : '' """)) return true 

		else return false

	}, 

	"Task: click on 'I am a human' and solve the captcha to make it disappear and validate!", //solving the captcha in two stages using a puzzle

	true //for test, false in production,

	{-> 	//refresh the page to reset everything

		firefox("browse", "https://nopecha.com/demo/hcaptcha#easy")

		firefox("waitLoaded")

	},

	3600, //one hour

	40

)

screenZone =  firefox("javascript", """
 captchaFrame = document.querySelector('#captcha__frame'); //the Datadome catpcha box



// Get the details of the captcha__frame

 captchaFrameRect = captchaFrame.getBoundingClientRect();



// Creating a table with coordinates

 coordinates = {

    x: captchaFrameRect.left + window.scrollX,

    y: captchaFrameRect.top + window.scrollY,

    width: captchaFrameRect.width,

    height: captchaFrameRect.height,

    from: 'pageInFirefox'

};



//return the table (it will be returned as an object directly usable by screenZone)

coordinates""", 4) // index 4 was found by iterative testing. It represents the index of the frame in which Javascript is executed. We start with index 1 and gradually increase until we find a frame that works. There may be frames within frames and there may be many of them.

	

//we solve the captcha

universalCaptchaSolver(screenZone,{-->

	iframeDatadome = regex(/(?si)<iframe [^<>]*src="([^<>"]+captcha[^<>"]+)"/, firefox("sourceCode"), "outerHTML") //it's not about frame 4. regex will run on the main frame. It simply detects the iframe containing the other iframes which themselves contain frame 4. This method is sufficient because when the captcha is solved, the target site no longer has any iframes containing the word "captcha".

	if(iframeDatadome) return false

	return true

})

identification = {->
	console("identification")

	setCloudCode(11111111111)
	//cookie page
	firefox("browse", "https://www.website.com/")
	firefox("waitLoaded")
	await(10*1000) //ajax functions load the cookie in a complex way, so we wait to make sure everything is in place

	//universalCaptchaSolver
	positionCaptcha = firefox("positionInScreen", "#captcha__frame", "*") //note the last argument "*", which is used to search for the element in iframes
	if(positionCaptcha)
	{	
		positionCaptcha.put("from", "screenInFirefox") //the capture mode is added to the table because it is not in firefox/positionInScreen
		universalCaptchaSolver(positionCaptcha,{->
			if(contains(firefox("javascript","el = document.querySelector('.captcha__human__container'); if(!el) ''; else el.outerHTML", "*"), "blocked")) return true  //in this case, the site blocks completely without offering a captcha, so we exit universalCaptchaSolver and deal with this case later with an IP change
			else if(firefox("javascript","""el = document.querySelector("iframe[src*='geo.captcha-delivery.com']"); if(!el) ''; else el.outerHTML""", "*")) return false //the iframe is present in the main frame, which means that the captcha is still there
			else return true //the captcha is solved
		}, "Solve the captcha (usually a puzzle piece to be moved to the right place). Refresh as necessary (wait 5 secondes after do it).", false) //the instructions are important, guide the captcha solver carefully and in detail. At the slightest oddity, it will move on to another captcha.
		
	}

	//use of the cookie in grimport
	setCookieValue(firefox("getAllCookies", ".website.com"), "https://www.website.com/")
	setCookieValue(firefox("getAllCookies", ".www.website.com"), "https://www.website.com/")
	setCookieValue(firefox("getAllCookies", "www.website.com"), "https://www.website.com/")

	//check for blocking
	if(contains(firefox("javascript","el = document.querySelector('.captcha__human__container'); if(!el) ''; else el.outerHTML", "*"), "blocked"))
	{
		console("Total blocking - you have to change your ip and cookie")
		return false //failed identification
	}
	return true //successful identification
}

nbHTTPerrors = 0 //HTTP error counter
actionHttpError({httpErrorIdentifier,errorLink,codeHTTP-> //actionHttpError is triggered every time an HTTP error occurs somewhere in Grimport
	if(equals(codeHTTP, 403)) //reverse engineering of requests has established that in the event of a 403 error you need to identify yourself again
	{
		nbHTTPerrors++ //each time an error occurs, the counter is incremented

		successIdentification = false
		if(nbHTTPerrors > 0 && nbHTTPerrors<=4)
		{
			successIdentification = identification()
		}
		if(nbHTTPerrors > 4 || !successIdentification) //if there are many requests for identification or if there is a problem with identification, the IP is changed
		{
			nbHTTPerrors=0
			proxyChange(["method":"nordvpn_by_command","zone":"europe",
			"action_between_change":{->
				firefox("clearAllCookiesAllDomains") //Deleting cookies on Firefox
				clearAllCookies() //Deleting cookies in Grimport Crawler
				identification() //we identify the web page test and the proxy change
			})
		}
})

See also

Parameters

{jsonResponse -> ... }

{jsonResponse -> ... }