测试

routes/index.js

@@ -21,7 +21,7 @@ router.post('/rsCookie', async (req, res) => {
         let cookie = req.body['cookieBase64'];
         let userAgent = req.body['userAgentBase64'];
         console.log(`${uuid};接收到 ${areaName} 请求：${url}`)
-        fs.writeFileSync(`./back/${uuid}.url`, `${url}\n ${userAgent}\n ${cookie} \n`);
+        // fs.writeFileSync(`./back/${uuid}.url`, `${url}\n ${userAgent}\n ${cookie} \n`);
         if (url == null || url === '') {
             return res.status(500).send('error url')
         }
@@ -105,22 +105,29 @@ async function handle(url, htmlStr, jsText, cookie, userAgent, uuid, loadHtmlJs)
     if (cookie != null) {
         let cookieList = CookieStr2List(cookie);
         console.log(`${uuid};cookie 加载长度--->`, cookieList, baseUrl)
-        fs.writeFileSync(`./back/${uuid}.cookie`, cookie)
+        // fs.writeFileSync(`./back/${uuid}.cookie`, cookie)
         for (let i = 0; i < cookieList.length; i++) {
             cookieJar.setCookieSync(cookieList[i], baseUrl);
         }
     }
+    // 设置cookie的回掉
+    const superSetCookie = cookieJar.setCookie;
+    cookieJar.setCookie = function (cookie, currentUrl, options, callback) {
+        console.log(`${uuid};正在设置 Cookie:`, cookie, currentUrl);
+        // store.set(uuid, cookie)
+        return superSetCookie.call(this, cookie, currentUrl, options, callback);
+    };
     // 加载dom
     let dom = await jsDom(htmlStr);
     // console.log('html 加载长度--->', dom.serialize())
     console.log(`${uuid};html 加载长度--->`, dom.serialize().length)
-    fs.writeFileSync(`./back/${uuid}.html`, dom.serialize())
+    // fs.writeFileSync(`./back/${uuid}.html`, dom.serialize())
 
     window = dom.window
-    // js执行成功后的瑞树会跳转页面 会触发onbeforeunload钩子
+    // js执行成功后会跳转页面 会触发onbeforeunload钩子
     window.onbeforeunload = async (url) => {
         const cookies = cookieJar.getCookieStringSync(baseUrl);
-        // console.debug(`${url} 生成cookie：`, cookies);
+        console.debug(`${url} 生成cookie：`, cookies);
         store.set(uuid, cookies)
         // window.close();
     }
@@ -134,33 +141,24 @@ async function handle(url, htmlStr, jsText, cookie, userAgent, uuid, loadHtmlJs)
         js = jsText;
     }
     console.log(`${uuid};js 加载长度--->`, js.length)
-    fs.writeFileSync(`./back/${uuid}.js`, js)
+    // fs.writeFileSync(`./back/${uuid}.js`, js)
 
     // 执行 js
     let script = new Script(js);
     let internalVMContext = dom.getInternalVMContext();
     script.runInContext(internalVMContext);
     // 等待 onbeforeunload 钩子触发后的回掉
-    let timeout = -1;
+    // let val = await store.waitGetAndDelete(uuid, 100, 10)
-    let reTry = -1;
-    if (process.env.NODE_ENV === 'prod') {
-        timeout = 100
-        reTry = 10
-    } else {
-        timeout = 100
-        reTry = 10
-    }
-    let val = await store.waitGetAndDelete(uuid, timeout, reTry)
 
     // bug 关闭后 部分参数是需要永久保存在内存中的 比如 cookie 实例 下次调用会报错
     // internalVMContext.close()
     // window.close()
     // dom = null
 
-    if (val != null) {
+    // if (val != null) {
-        return val;
+    //     return val;
-    }
+    // }
-    console.log(`${uuid} 执行超时`)
+    // console.log(`${uuid} 执行超时`)
     return cookieJar.getCookieStringSync(baseUrl);
     // throw new Error(`${uuid}执行超时`)
 }

test/rs6/yunnan.js

@@ -4,32 +4,49 @@ const {jsdomFromText, browser, jsdomFromUrl} = require("sdenv");
 const {Script} = require("vm");
 let fs = require('fs');
 const express = require("express");
-const cheerio = require('cheerio');
+// const cheerio = require('cheerio');
 
 
-const app = express();
+// const app = express();
 
 var baseUrl = "https://app.yunnan.chinatax.gov.cn"
 let url = `https://app.yunnan.chinatax.gov.cn/xxmh/html/dfts/index_frame.html`;
 
-async function extracted(cookieJar) {
+// async function extracted(cookieJar) {
-    const [jsdomer] = await jsdomFromUrl({
+//     const [jsdomer] = await jsdomFromUrl({
-        proxy: 'http://127.0.0.1:8080',
+//         // proxy: 'http://127.0.0.1:8080',
-        strictSSL: false,
+//         strictSSL: false,
-        userAgent: 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36',
+//         userAgent: 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36',
-    }, cookieJar);
+//     }, cookieJar);
-    const dom = await jsdomer(url);
+//     const dom = await jsdomer(url);
-
+//
-    console.log(dom.serialize())
+//     console.log(dom.serialize())
+// }
+function loadJs(window, jsText) {
+    // 加载js
+    let js = '';
+    // 加载 页面上的js
+    const allScript = window.document.querySelectorAll('script[r="m"]');
+    for (let i = 0; i < allScript.length; i++) {
+        const script = allScript[i];
+        let attr = script.textContent;
+        if (attr) {
+            js += attr
+        } else {
+            js += jsText
+        }
+        js += ";\n"
+    }
+    return js;
 }
 
 
-function loadPages(urlBase64, jsBase64, htmlBase64, cookieBase64, userAgentBase64, ts) {
+async function loadPages(urlBase64, jsBase64, htmlBase64, cookieBase64, userAgentBase64, ts) {
     const htmlText = htmlBase64;
     const jsText = jsBase64;
 
     const [jsDom, cookieJar] = jsdomFromText({
-        proxy: 'http://127.0.0.1:8080',
+        // proxy: 'http://127.0.0.1:8080',
         url: url,
         referrer: url,
         contentType: "text/html",
@@ -41,11 +58,11 @@ function loadPages(urlBase64, jsBase64, htmlBase64, cookieBase64, userAgentBase6
 
     window = dom.window
     // 加载 window.$_ts
-    let $ = cheerio.load(htmlText);
+    // let $ = cheerio.load(htmlText);
-    let nextAll = $('script[r=m]');
+    // let nextAll = $('script[r=m]');
 
-    let trendJs = $('script').first().html();
+    // let trendJs = $('script').first().html();
-    let js = trendJs + ";\n" + jsText
+    // let js = trendJs + ";\n" + jsText
     // nextAll.each((index, element) => {
     //     let attr = $(element).html();
     //     if (attr) {
@@ -59,13 +76,47 @@ function loadPages(urlBase64, jsBase64, htmlBase64, cookieBase64, userAgentBase6
     cookieJar.setCookieSync(cookieBase64, baseUrl);
 
 
-
     window.onbeforeunload = async (url) => {
 
         const cookies = cookieJar.getCookieStringSync(baseUrl);
         console.debug(`${url} 生成cookie：`, cookies);
 
 
+        // let fff = await fetch("https://app.yunnan.chinatax.gov.cn/xxmh/html/dfts/index_frame.html", {
+        //     "headers": {
+        //         "accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.7",
+        //         "accept-language": "zh-CN,zh;q=0.9",
+        //         "cache-control": "no-cache",
+        //         "pragma": "no-cache",
+        //         "sec-ch-ua": "\"Not/A)Brand\";v=\"8\", \"Chromium\";v=\"126\", \"Google Chrome\";v=\"126\"",
+        //         "sec-ch-ua-mobile": "?0",
+        //         "sec-ch-ua-platform": "\"Linux\"",
+        //         "sec-fetch-dest": "document",
+        //         "sec-fetch-mode": "navigate",
+        //         "sec-fetch-site": "same-origin",
+        //         "upgrade-insecure-requests": "1",
+        //         "cookie": cookies,
+        //         "Referer": "https://app.yunnan.chinatax.gov.cn/xxmh/html/dfts/index_frame.html",
+        //     }, "body": null, "method": "GET"
+        // });
+        //
+        // console.log(fff.status, fff.statusText)
+
+
+        window.close();
+    }
+
+    browser(dom.window, 'chrome');
+
+    let js = loadJs(window, jsText);
+    let script = new Script(js);
+    let internalVMContext = dom.getInternalVMContext();
+    // 执行 js
+    script.runInContext(internalVMContext);
+
+
+    const cookies = cookieJar.getCookieStringSync(baseUrl);
+
     let fff = await fetch("https://app.yunnan.chinatax.gov.cn/xxmh/html/dfts/index_frame.html", {
         "headers": {
             "accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.7",
@@ -85,18 +136,6 @@ function loadPages(urlBase64, jsBase64, htmlBase64, cookieBase64, userAgentBase6
     });
 
     console.log(fff.status, fff.statusText)
-
-
-        window.close();
-    }
-
-    browser(dom.window, 'chrome');
-
-    let script = new Script(js);
-    let internalVMContext = dom.getInternalVMContext();
-    // 执行 js
-    script.runInContext(internalVMContext);
-
 }
 
 
@@ -104,7 +143,7 @@ async function test() {
     let response = await fetch(url, {"method": "GET"});
     // 返回页面内容
     let htmlStr = await response.text();
-    // console.log(htmlStr)
+    console.log(htmlStr)
 
 
     // 页面返回cookie
@@ -114,7 +153,7 @@ async function test() {
 
     let staticJs = fs.readFileSync('/home/liu/桌面/MO5zzCMcub4d.b4c45da.js');
 
-    loadPages(url, staticJs.toString('utf8'), htmlStr, cookies, "")
+    await loadPages(url, staticJs.toString('utf8'), htmlStr, cookies, "")
 }