route 人民网今日头条一览

yubinbai · Sep 9, 2021 · ad446f6 · ad446f6
1 parent d445580
commit ad446f6
Show file tree

Hide file tree

Showing 4 changed files with 63 additions and 6 deletions.
diff --git a/docs/finance.md b/docs/finance.md
@@ -161,12 +161,6 @@ pageClass: routes
 
 <Route author="yubinbai" example="/sina/finance" path="/sina/finance" />
 
-## 新浪财经
-
-### 新浪财经－国內
-
-<Route author="yubinbai" example="/sina/finance" path="/sina/finance" />
-
 ## 雪球
 
 ### 用户动态

diff --git a/docs/traditional-media.md b/docs/traditional-media.md
@@ -828,6 +828,10 @@ category 对应的关键词有
 
 <Route author="nczitzk" example="/people/cpc/24h" path="/people/cpc/24h"/>
 
+### 人民网 今日头条一览
+
+<Route author="yubinbai" example="/people/toutiao" path="/people/toutiao"/>
+
 ## 人民日报社 国际金融报
 
 ### 栏目

diff --git a/lib/router.js b/lib/router.js
@@ -1568,6 +1568,7 @@ router.get('/people/opinion/:id', require('./routes/people/opinion'));
 router.get('/people/env/:id', require('./routes/people/env'));
 router.get('/people/xjpjh/:keyword?/:year?', require('./routes/people/xjpjh'));
 router.get('/people/cpc/24h', require('./routes/people/cpc/24h'));
+router.get('/people/toutiao', lazyloadRouteHandler('./routes/people/toutiao'));
 
 // 北极星电力网
 router.get('/bjx/huanbao', require('./routes/bjx/huanbao'));

diff --git a/lib/routes/people/toutiao.js b/lib/routes/people/toutiao.js
@@ -0,0 +1,58 @@
+const cheerio = require('cheerio');
+const got = require('@/utils/got');
+const iconv = require('iconv-lite');
+const userAgent = 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/93.0.4577.63 Safari/537.36';
+const cookie = 'HMF_CI=1af542e784845025d7b6dbadb116e866b9d8964dc614f5f1b703405cb77295e7a2;'; // required, otherwise 403 forbidden
+
+module.exports = async (ctx) => {
+    const title = '人民网 今日头条一览';
+    const link = `http://www.people.com.cn/GB/59476/index.html`;
+
+    const response = await got.get(link, {
+        responseType: 'buffer',
+        headers: {
+            'User-Agent': userAgent,
+            Cookie: cookie,
+        },
+    });
+    const $ = cheerio.load(response.data);
+
+    const list = $('body > table:nth-child(2) > tbody > tr > td > table > tbody > tr:nth-child(2) > td > table > tbody > tr > td:nth-child(1) > table:nth-child(2) > tbody > tr > td > *')
+        .slice(0, 20)
+        .map(function () {
+            const link = $(this).find('a').attr('href') ?? $(this).attr('href');
+            return link;
+        })
+        .filter((e) => !!e)
+        .get();
+
+    const out = await Promise.all(
+        list.map((itemUrl) =>
+            ctx.cache.tryGet(itemUrl, async () => {
+                const response = await got.get(itemUrl, {
+                    responseType: 'buffer',
+                    headers: {
+                        'User-Agent': userAgent,
+                        Cookie: cookie,
+                    },
+                });
+                response.data = iconv.decode(response.data, 'gbk');
+                const $ = cheerio.load(response.data);
+                const title = $('body > div.main > div.layout.rm_txt.cf > div.col.col-1 > h1').html().trim();
+                const description = $('body > div.main > div.layout.rm_txt.cf > div.col.col-1').html().trim();
+                const single = {
+                    title,
+                    link: itemUrl,
+                    description,
+                };
+                return Promise.resolve(single);
+            })
+        )
+    );
+
+    ctx.state.data = {
+        title,
+        link,
+        item: out,
+    };
+};