ZFC-Digital · 854771076 · Nov 5, 2024 · Nov 5, 2024 · Nov 5, 2024 · Nov 5, 2024
diff --git a/README.md b/README.md
@@ -1,5 +1,9 @@
 # CF Clearance Scraper
 
+### 请我喝杯咖啡
+ETH 0x72691a36ed1fac3b197fb42612dc15a8958bf9f2
+
+
 This library was created for testing and training purposes to retrieve the page source of websites, create Cloudflare Turnstile tokens and create Cloudflare WAF sessions.
 
 Cloudflare protection not only checks cookies in the request. It also checks variables in the header. For this reason, it is recommended to use it with the sample code in this readme file.

diff --git a/install.sh b/install.sh
@@ -0,0 +1,14 @@
+curl -L https://github.com/854771076/cf-clearance-scraper/archive/refs/heads/master.zip -o cf-clearance-scraper-main.zip
+unzip cf-clearance-scraper-main.zip
+cd cf-clearance-scraper-main
+
+# 构建镜像
+docker build -t captcha_cracker .
+
+# 运行容器
+docker run -d --restart unless-stopped -p 3000:3000 \
+    -e PORT=3000 \
+    -e browserLimit=20 \
+    -e timeOut=60000 \
+    -e authToken=authToken123456 \
+    captcha_cracker
diff --git a/package-lock.json b/package-lock.json
diff --git a/src/data/axios.min.js b/src/data/axios.min.js
diff --git a/src/data/fakePage.html b/src/data/fakePage.html
@@ -5,6 +5,8 @@
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
     <title></title>
+    <input type="hidden" id="cf-response" name="cf-response">
+    <input type="hidden" id="cf-status" name="cf-status" value="0">
 </head>
 
 <body>
@@ -14,15 +16,18 @@
         window.onloadTurnstileCallback = function () {
             turnstile.render('.turnstile', {
                 sitekey: '<site-key>',
+
                 callback: function (token) {
-                    var c = document.createElement('input');
-                    c.type = 'hidden';
-                    c.name = 'cf-response';
+                    var c=document.querySelector('#cf-response');
+                    var status=document.querySelector('#cf-status');
                     c.value = token;
-                    document.body.appendChild(c);
+                    status.value="1"
+
+
                 },
             });
         };
+
 
     </script>
 </body>

diff --git a/src/data/reCAPTCHAV3.html b/src/data/reCAPTCHAV3.html
@@ -0,0 +1,34 @@
+<!DOCTYPE html>
+<html lang="en">
+
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title></title>
+</head>
+
+<body>
+    <script src="https://www.google.com/recaptcha/api.js?render=<reCAPTCHA_site_key>" defer></script>
+    <script>
+        window.onload= function () {
+            console.log('begin')
+            grecaptcha.ready(function () {
+                console.log('in')
+                grecaptcha.execute('<reCAPTCHA_site_key>', { action: 'submit' }).then(function (token) {
+                    console.log("token:",token)
+                    let input = document.createElement('input');
+                    input.id = 'my_token';
+                    input.name = 'my_token';
+                    input.value = token;  // Ensure `token` is defined earlier in your code
+
+                    // Append the input to the body (or another container)
+                    document.body.appendChild(input);
+
+                    document.appendChild(input)
+                });
+            });
+        };
+    </script>
+</body>
+
+</html>
diff --git a/src/data/reCAPTCHAV3_enterprise.html b/src/data/reCAPTCHAV3_enterprise.html
@@ -0,0 +1,34 @@
+<!DOCTYPE html>
+<html lang="en">
+
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title></title>
+</head>
+
+<body>
+    <script src="https://www.google.com/recaptcha/enterprise.js?render=<reCAPTCHA_site_key>" defer></script>
+    <script>
+        window.onload= function () {
+            console.log('begin')
+            grecaptcha.enterprise.ready(function () {
+                console.log('in')
+                grecaptcha.enterprise.execute('<reCAPTCHA_site_key>', { action: 'submit' }).then(function (token) {
+                    console.log("token:",token)
+                    let input = document.createElement('input');
+                    input.id = 'my_token';
+                    input.name = 'my_token';
+                    input.value = token;  // Ensure `token` is defined earlier in your code
+
+                    // Append the input to the body (or another container)
+                    document.body.appendChild(input);
+
+                    document.appendChild(input)
+                });
+            });
+        };
+    </script>
+</body>
+
+</html>
diff --git a/src/endpoints/getSource.js b/src/endpoints/getSource.js
@@ -1,4 +1,4 @@
-function getSource({ url, proxy }) {
+function getSource({ url, proxy,callback,axios }) {
     return new Promise(async (resolve, reject) => {
 
         if (!url) return reject('Missing url parameter')
@@ -34,13 +34,37 @@ function getSource({ url, proxy }) {
             });
             page.on('response', async (res) => {
                 try {
-                    if ([200, 302].includes(res.status()) && [url, url + '/'].includes(res.url())) {
+                    // if ([200, 302].includes(res.status()) && [url, url + '/'].includes(res.url())) {
+                    if (true) {   
                         await page.waitForNavigation({ waitUntil: 'load', timeout: 5000 }).catch(() => { });
+                        if (axios){
+                            //读取src\data\axios.min.js
+                            const axios = await fs.readFileSync(path.join(__dirname, '../src/data/axios.min.js'), 'utf8')
+                            //在页面中执行axios.min.js
+                            await page.evaluate(axios)
+                        }
                         const html = await page.content();
-                        await context.close()
-                        isResolved = true
-                        clearInterval(cl)
-                        resolve(html)
+                        if(callback){
+                            //在页面中执行callback()
+                            await page.evaluate(eval(callback))
+                            //等待id为status的元素加载
+                            await page.waitForSelector('#status', { timeout: 60000 })
+                            //获取id为status的元素的文本
+                            const text = await page.$eval('#status', el => el.innerText)
+                            //返回执行结果
+                            await context.close()
+                            isResolved = true
+                            clearInterval(cl)
+                            resolve(text)
+                        }else{
+                            await context.close()
+                            isResolved = true
+                            clearInterval(cl)
+                            //控制台执行callback
+
+                            resolve(html)
+                        }
+
                     }
                 } catch (e) { }
             })

diff --git a/src/endpoints/getVercel.js b/src/endpoints/getVercel.js
@@ -0,0 +1,81 @@
+async function findAcceptLanguage(page) {
+    await page.setBypassCSP(true)
+    return await page.evaluate(async () => {
+        const result = await fetch('https://httpbin.org/get')
+            .then(res => res.json())
+            .then(res => (res.headers['Accept-Language'] || res.headers['accept-language']))
+            .catch(() => null)
+        return result
+    })
+}
+
+function getVercel({ url, proxy }) {
+    return new Promise(async (resolve, reject) => {
+
+        if (!url) return reject('Missing url parameter')
+        const context = await global.browser.createBrowserContext().catch(() => null);
+        if (!context) return reject('Failed to create browser context')
+
+        let isResolved = false
+
+        const { proxyRequest } = await import('puppeteer-proxy')
+
+        var cl = setTimeout(async () => {
+            if (!isResolved) {
+                await context.close()
+                reject("Timeout Error")
+            }
+        }, (global.timeOut || 60000))
+
+        try {
+            const page = await context.newPage();
+            await page.setRequestInterception(true);
+            page.on('request', async (request) => {
+                try {
+                    if (proxy) {
+                        await proxyRequest({
+                            page,
+                            proxyUrl: `http://${proxy.username ? `${proxy.username}:${proxy.password}@` : ""}${proxy.host}:${proxy.port}`,
+                            request,
+                        });
+                    } else {
+                        request.continue()
+                    }
+                } catch (e) { }
+            });
+            page.on('response', async (res) => {
+                try {
+                        await page.waitForNavigation({ waitUntil: 'load', timeout: 5000 }).catch(() => { });
+                        await page.waitForFunction(() => {
+                            return document.getElementById('footer-text')==null 
+                        })
+                        const cookies = await page.cookies()
+                        let headers = await res.request().headers()
+                        delete headers['content-type']
+                        delete headers['accept-encoding']
+                        delete headers['accept']
+                        delete headers['content-length']
+                        headers["accept-language"] = await findAcceptLanguage(page)
+                        await context.close()
+                        isResolved = true
+                        clearInterval(cl)
+                        resolve({ cookies, headers })
+
+                } catch (e) { }
+            })
+
+
+            await page.goto(url, {
+                waitUntil: 'domcontentloaded'
+            })
+        } catch (e) {
+            if (!isResolved) {
+                await context.close()
+                clearInterval(cl)
+                reject(e.message)
+            }
+        }
+
+    })
+}
+module.exports = getVercel
diff --git a/src/endpoints/solveRecaptcha.v2.js b/src/endpoints/solveRecaptcha.v2.js
@@ -0,0 +1,98 @@
+function solveRecaptcha({ url, proxy, siteKey,action }) {
+    return new Promise(async (resolve, reject) => {
+
+        if (!url) return reject('Missing url parameter')
+        if (!siteKey) return reject('Missing siteKey parameter')
+
+        const context = await global.browser.createBrowserContext().catch(() => null);
+        if (!context) return reject('Failed to create browser context')
+
+        let isResolved = false
+
+        const { proxyRequest } = await import('puppeteer-proxy')
+        const { RequestInterceptionManager } = await import('puppeteer-intercept-and-modify-requests')
+
+
+        var cl = setTimeout(async () => {
+            if (!isResolved) {
+                await context.close()
+                reject("Timeout Error")
+            }
+        }, (global.timeOut || 60000))
+
+        try {
+            const page = await context.newPage();
+            const client = await page.target().createCDPSession()
+            const interceptManager = new RequestInterceptionManager(client)
+
+            await page.setRequestInterception(true);
+            page.on('request', async (request) => {
+                try {
+                    if ([url, url + '/'].includes(request.url())) return request.abort()
+
+                    if (proxy) {
+                        await proxyRequest({
+                            page,
+                            proxyUrl: `http://${proxy.username ? `${proxy.username}:${proxy.password}@` : ""}${proxy.host}:${proxy.port}`,
+                            request,
+                        });
+                    } else {
+                        request.continue()
+                    }
+                } catch (e) { }
+            });
+
+            await interceptManager.intercept(
+                {
+                    urlPattern: url,
+                    resourceType: 'Document',
+                    modifyResponse({ body }) {
+                        return {
+                            body: String(require('fs').readFileSync('./src/data/reCAPTCHAV3.html')).replace(/<reCAPTCHA_site_key>/g, siteKey).replace(/<reCAPTCHA_site_key>/g, action),
+                            status: 200
+                        }
+                    },
+                }
+            )
+
+            await page.goto(url, {
+                waitUntil: 'domcontentloaded'
+            })
+            await page.waitForSelector('.grecaptcha-badge', {
+                timeout: 60000
+            })
+
+            await page.waitForSelector('iframe', {
+                timeout: 60000
+            });
+            const iframes = await page.$$('iframe'); 
+            const iframe1 = await iframes[0].contentFrame();
+            await iframe1.waitForSelector('#recaptcha-token', {
+                timeout: 60000
+            });
+            await page.waitForSelector('#my_token', {
+                timeout: 60000
+            });
+            const token = await page.evaluate(() => {
+                try {
+                    return document.querySelector('#my_token').value
+                } catch (e) {
+                    return null
+                }
+            })
+            isResolved = true
+            clearInterval(cl)
+            await context.close()
+            if (!token || token.length < 10) return reject('Failed to get token')
+            return resolve({token:token})
+        } catch (e) {
+            if (!isResolved) {
+                await context.close()
+                clearInterval(cl)
+                reject(e.message)
+            }
+        }
+
+    })
+}
+module.exports = solveRecaptcha