Suck'em ratings

cli.js

const { GoogleResults, BingResults, DDGResults } = require('./sniffratings');

function parseArguments(args) {
	const argsMap = {};
	let currentKey = null;

	for (let i = 2; i < args.length; i++) {
		const arg = args[i];

		if (arg.startsWith('-')) {
			currentKey = arg.slice(1);
			argsMap[currentKey] = true;
		} else if (currentKey) {
			argsMap[currentKey] = arg;
			currentKey = null;
		}
	}

	return argsMap;
}

function handleRatingSearch(company, sp) {
	company = company.trim();

	if (!company) return Promise.reject('empty_content')

	let fetcher = { getReviews: () => Promise.reject("invalid_search_provider") }
	if (sp == 'google') {
		fetcher = GoogleResults.init(company)
	} else if (sp == 'ddg') {
		fetcher = DDGResults.init(company)
	}

	return fetcher.getReviews();
}

async function main() {
	const argsMap = parseArguments(process.argv);

	if (!argsMap.company) {
		console.error('Error: Company name is required. Use -company option.');
		return;
	}

	const company = argsMap.company;
	const sp = argsMap.search || 'ddg';

	try {
		const ratings = await handleRatingSearch(company, sp);
		console.log(`Ratings for ${company}:`, ratings);
	} catch (error) {
		console.error('Error retrieving ratings:', error);
	}
}

main()

grissmonkey.js

// ==UserScript==
// @name     RatingSniffer
// @version  1.1
// @match        https://www.linkedin.com/jobs/*
// @run-at      document-idle
// ==/UserScript==

(function(_window) {
	const MAX_LOOP = 10000;

	function $(el) {
		return document.querySelector(el)
	}

	var SP = 'ddg'; //google

	function waitTill(fn, comp, cb) {
		let loopCount = 0;

		let interval = setInterval(() => {
			let [values, ok] = comp(fn());

			console.log("running", ok, values);

			if (ok) {
				console.log("result received")
				clearInterval(interval)
				cb(null, values);
				return
			}

			if (loopCount == MAX_LOOP) {
				clearInterval(interval);
				cb("max loop", values);

				return
			}


			loopCount += 1;

		}, 2000)
	}

	function getNewRater() {
		let ratingEl = document.createElement("a");

		ratingEl.href = "#";
		ratingEl.id = "berater_el"
		ratingEl.innerText = "GetRated"
		ratingEl.className = "jobs-save-button artdeco-button artdeco-button--secondary artdeco-button--3"
		ratingEl.style = "margin-left: 10px;";

		ratingEl.onclick = () => {
			let companyName = $(".job-details-jobs-unified-top-card__company-name a").innerText.trim();

			getCompanyRatings(companyName, SP).then(res => alert(res)).catch(err => alert(err))
		}

		return ratingEl;
	}


	function debounce(func, delay) {
		let debounceTimer;
		delay = delay || 2000

		return function() {
			const context = this;
			const args = arguments;
			clearTimeout(debounceTimer);
			debounceTimer = setTimeout(() => func.apply(context, args), delay);
		};
	};

	const mutator = (elq) => (mutationsList, observer) => {
		let hasMutated = false;

		//TODO: add a debouncer

		for (const mutation of mutationsList) {
			if (mutation.type === 'childList' || mutation.type === 'attributes') {
				// console.log("lala");

				hasMutated = true
				break;
			}
		}

		console.log("mutated?", hasMutated);

		if (!hasMutated) return;

		let element = $(elq);
		let html = element.innerText;
		//console.log("html", !html.includes("GetRated"));

		if (hasMutated && !html.includes("GetRated")) {
			let ratingEl = getNewRater();
			console.log("appending to ", elq);


			element.appendChild(ratingEl);
		}
	}

	let watcherQs = [
		".jobs-details__main-content--single-pane .mt5 div:first-child",
		".job-details-jobs-unified-top-card__container--two-pane .mt5 div:first-child"
	]

	//fireup
	// jobs-details__main-content jobs-details__main-content--single-pane
	// job-details-jobs-unified-top-card__container--two-pane"
	//
	waitTill(() => watcherQs.slice(0, 1),
		(values) => [values, values && values.length > 0],
		(err, elements) => {
			if (err != null) {
				return err
			}

			// console.log("creating rater", elements);

			elements.forEach((element, i) => {
				if (!element) return;

				// first run
				let ratingEl = getNewRater();
				element.appendChild(ratingEl);

				const observer = new MutationObserver(debounce(mutator(watcherQs[i])));
				const config = { childList: true, attributes: true };

				observer.observe(element, config);
			});
		});


	function getCompanyRatings(company, sp) {
		sp ||= 'ddg'

		const url = `http://localhost:3000/rating?company=${company}&sp=${sp}`;

		return fetch(url).then(resp => resp.json()).then(data => {
			if (!data.success) return Promise.reject('failed_response');

			console.log(data);

			let texts = data.result.ratings.map(r => `${r.provider}: ${r.rating}`)
			texts.push(`data_breached: ${data.result.breached}`);
			return texts.join("\n");
		});
	}


	//   _window.addEventListener('load', fireUp);
	_window.getRatings = getCompanyRatings;

})(window);

install.sh

#!/usr/bin/env bash

curl 'https://raw.githubusercontent.com/ikouchiha47/devtools/master/berater/install.sh' | bash -s -- -y

server.js

const http = require('http');
const url = require('url');

const { GoogleResults, DDGResults } = require('./sniffratings');

function handleRatingSearch(company, sp) {
	company = company.trim();

	if (!company) return Promise.reject('empty_content')

	let fetcher = { getReviews: () => Promise.reject("invalid_search_provider") }
	if (sp == 'google') {
		fetcher = GoogleResults.init(company)
	} else if (sp == 'ddg') {
		fetcher = DDGResults.init(company)
	}

	return fetcher.getReviews();
	// Refine google search or google search twice parallely for each

	// Get rating from google search and also try to get rating from website. Combine to average.
	// 
	// GoogleResults.init('Turing').getReviews().
	// 	then(results => results.filter(r => r.status == 'fulfilled')).
	// 	then(results => Promise.allSettled(results.map((r, i) => write(`scraped-${i}.html`, r.value)))).
	// 	catch(err => console.error(err));

}

const requestListener = function(req, res) {
	res.setHeader("Content-Type", "application/json");

	const parsedUrl = url.parse(req.url, true);
	const pathname = parsedUrl.pathname;
	const company = parsedUrl.query.company;
	const sp = parsedUrl.query.sp || 'ddg';

	if (pathname == "/rating") {
		return handleRatingSearch(company, sp).then(result => {
			res.writeHead(200);
			console.log(result);
			res.end(JSON.stringify({ success: true, result }));
		}).catch(err => {
			res.writeHead(422);
			res.end(`{"error": ${err}, "success": false}`)
		})
	}

	res.end(`{"error": "invalidpath"}`)
};


const server = http.createServer(requestListener);
const port = process.env.PORT || 3000;
server.listen(port, () => {
	console.log(`Server is running on :${port}`);
});

sniffratings.js

const fs = require('node:fs/promises');

const GoogleURL = `https://google.com/search?q`
const DDGURL = `https://html.duckduckgo.com/html?q`
const BingURL = `https://www.bing.com/search?q`
const Providers = ["glassdoor.com", "ambitionbox.com"]

const write = (fileName, data) => {
	log.on("writing to file " + fileName);
	return fs.writeFile(`${fileName}`, data, { flag: 'w' })
}

let agents = [
	"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.3",
	"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/125.0.0.0 Safari/537.3",
	"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.3",
	"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36 Edg/126.0.0.",
	"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/125.0.0.0 Safari/537.36 Edg/125.0.0.",
	"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/125.0.0.0 Safari/537.36 OPR/111.0.0.",
]

const ua = {
	"User-Agent": agents[Math.floor(Math.random() * 10) % 6]
}

const log = {
	on: console.log.bind(console),
	off: () => { },
}

const searchProviders = (company, searchprovider, url, qbuilder) => {
	const resolver = (query, provider) => {
		const searchURL = `${url}=${query.replaceAll(" ", "+")}`;
		let scope = { text: '', provider: '' };

		log.on(searchURL);

		return fetch(searchURL, { headers: { ...ua } }).then(resp => resp.text()).then(t => {
			if (t.includes('captcha')) {
				return Promise.reject(`provider(${provider}) for your search result bitched out`)
			}

			scope.text = t; scope.provider = provider;
			const filename = `tmp/${searchprovider}_${company.toLowerCase()}_${provider.split('.').shift()}.html`;
			return write(filename, t);
		}).then(() => scope).catch(err => { log.on(err); throw err });
	}

	return Promise.allSettled(
		Providers.map(provider => resolver(qbuilder(company, provider), provider))
	).then(results => (
		results.
			filter(result => result.status === 'fulfilled').
			map(result => result.value)
	))
}

const findRatingsInSearch = (searchResult, provider, rx) => {
	// const results = {};
	while ((m = rx.exec(searchResult)) !== null) {
		if (m.index === rx.lastIndex) {
			regex.lastIndex++;
		}

		if (m && m.groups && m.groups.rating) {
			// results[provider] = m.groups.rating
			return m.groups.rating
		}
	}

	return null;
}


const GoogleResults = {
	_company: '',
	_baseURL: GoogleURL,
	_buildQ: (company, provider) => `site://${provider} ${company} reviews`,

	init: function(company) { this._company = company; return this; },
	rx: /<a href=([^>]+)\/?>/gm,
	frx: /\/url\?q=([^&]+)/,
	rrx: /Rating[\s\S]*?(?<rating>(\d\.\d))/gm,

	getReviews: function() {
		return searchProviders(this._company, 'google', this._baseURL, this._buildQ).then(results => {
			return results.map(result => ({
				rating: findRatingsInSearch(result.text, result.provider, this.rrx),
				provider: result.provider,
			}));
		});
	},
	_aggregateReviews: function() {
		// let searchRes = search(this._company, GoogleURL, buildQuery(this._company));
		// return searchRes.
		// then(data => cleanHTML(data, GoogleResults.rx)).
		// then(urls => findURLs(urls, this._company, GoogleResults.frx)).
		return Promise.reject("not_implemented")
	},
}

const BingResults = {
	...GoogleResults,
	_baseURL: BingURL,

	getReviews: function() {
		return searchProviders(this._company, 'bing', this._baseURL, this._buildQ).then(results => {
			return results.map(result => ({
				rating: '',
				provider: result.provider,
			}));
		});
	}
}

const DDGResults = {
	...GoogleResults,
	_baseURL: DDGURL,
	_buildQ: (company, provider) => `site:${provider} ${company} ratings`,

	rrx: /<a class="result__snippet"[^>]+>[\s\S]*?(?<rating>(\d\.\d))/m,
	getReviews: function() {
		return searchProviders(this._company, 'ddg', this._baseURL, this._buildQ).then(results => {
			return results.map(result => ({
				rating: findRatingsInSearch(result.text, result.provider, this.rrx),
				provider: result.provider,
			}));
		});
	}
}

module.exports = {
	GoogleResults: GoogleResults,
	BingResults: BingResults,
	DDGResults: DDGResults,
}