fix(route/cjlu/yjsy): use Puppeteer to fetch CJLU Graduate School notices after anti-crawler update (#20359)

* fix(route/yjsy): use puppeteer to bypass WAF

* fix: chage waiting-util to get fully loaded page

* fix: remove 'stylesheet' from excludeResourceTypes in yjsy route

* chore: provide default title for notifications in yjsy route

* fix: add HTTP headers for Puppeteer page requests in yjsy route

* fix: replace global ua in config with a static value

* fix: change waitUntil from 'networkidle0' to 'networkidle2'

* fix: set ua with setUserAgent

* fix: ensure user agent is set after HTTP headers in Puppeteer page requests

* fix: add limit parameter to handler for controlling result set size

* fix: update resource type handling in request interception

- use default maxAge managed by env

lib/routes/cjlu/yjsy/index.ts

@@ -4,6 +4,7 @@ import { load } from 'cheerio';
 import { parseDate } from '@/utils/parse-date';
 import cache from '@/utils/cache';
 import timezone from '@/utils/timezone';
+import { getPuppeteerPage } from '@/utils/puppeteer';
 
 const host = 'https://yjsy.cjlu.edu.cn/';
 
@@ -11,18 +12,49 @@ const titleMap = new Map([
     ['yjstz', '中量大研究生院 —— 研究生通知'],
     ['jstz', '中量大研究生院 —— 教师通知'],
 ]);
+const 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,en;q=0.8,en-GB;q=0.7,en-US;q=0.6',
+    'Cache-Control': 'max-age=0',
+    Connection: 'keep-alive',
+    Referer: 'https://yjsy.cjlu.edu.cn',
+    'Sec-Fetch-Dest': 'document',
+    'Sec-Fetch-Mode': 'navigate',
+    'Sec-Fetch-Site': 'same-origin',
+    'Sec-Fetch-User': '?1',
+    'Upgrade-Insecure-Requests': '1',
+    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/141.0.0.0 Safari/537.36 Edg/141.0.0.0',
+    'sec-ch-ua': '"Microsoft Edge";v="141", "Not?A_Brand";v="8", "Chromium";v="141"',
+    'sec-ch-ua-mobile': '?0',
+    'sec-ch-ua-platform': '"Windows"',
+};
+
+const allowedResourceTypes = new Set(['document', 'script']);
 
 export const route: Route = {
     path: '/yjsy/:cate',
     categories: ['university'],
     example: '/cjlu/yjsy/yjstz',
     parameters: {
-        cate: '订阅的类型，支持 yjstz（研究生通知）和 jstz（教师通知）',
+        cate: {
+            description: '订阅的类型，支持 yjstz（研究生通知）和 jstz（教师通知）',
+            default: 'yjstz',
+            options: [
+                {
+                    label: '教师通知',
+                    value: 'jstz',
+                },
+                {
+                    label: '研究生通知',
+                    value: 'yjstz',
+                },
+            ],
+        },
     },
     features: {
         requireConfig: false,
-        requirePuppeteer: false,
-        antiCrawler: false,
+        requirePuppeteer: true,
+        antiCrawler: true,
         supportRadar: true,
         supportBT: false,
         supportPodcast: false,
@@ -50,17 +82,33 @@ export const route: Route = {
 
 async function handler(ctx) {
     const cate = ctx.req.param('cate');
+    const limit = ctx.req.query('limit') ? Number.parseInt(ctx.req.query('limit'), 10) : 10;
+    const url = `${host}index/${cate}.htm`;
 
-    const response = await ofetch(`${cate}.htm`, {
-        baseURL: `${host}/index/`,
-        responseType: 'text',
+    const { page, destory, browser } = await getPuppeteerPage(url, {
+        onBeforeLoad: async (page) => {
+            await page.setExtraHTTPHeaders(headers);
+            await page.setUserAgent(headers['User-Agent']);
+            await page.setRequestInterception(true);
+            page.on('request', (request) => {
+                allowedResourceTypes.has(request.resourceType()) ? request.continue() : request.abort();
+            });
+        },
+        gotoConfig: { waitUntil: 'networkidle2' },
     });
 
+    const cookies = await browser.cookies();
+    const cookieString = cookies.map((c) => `${c.name}=${c.value}`).join('; ');
+
+    const response = await page.content();
+    await destory();
+
     const $ = load(response);
 
     const list = $('div.grid685.right div.body ul')
         .find('li')
         .toArray()
+        .slice(0, limit)
         .map((element) => {
             const item = $(element);
 
@@ -71,7 +119,7 @@ async function handler(ctx) {
             const route = href.startsWith('../') ? href.replace(/^\.\.\//, '') : href;
 
             return {
-                title: a.attr('title') ?? titleMap.get(cate),
+                title: a.attr('title') ?? titleMap.get(cate) ?? '中量大研究生院通知',
                 pubDate: timezone(parseDate(timeStr, 'YYYY/MM/DD'), +8),
                 link: `${host}${route}`,
                 description: '',
@@ -87,6 +135,11 @@ async function handler(ctx) {
 
                 const res = await ofetch(item.link, {
                     responseType: 'text',
+                    headers: {
+                        ...headers,
+                        Cookie: cookieString,
+                        Referer: url,
+                    },
                 });
                 const $ = load(res);