feat: 增加财新博客全文输出 (#2090)

* feat: 增加财新博客全文输出

同时更改了原财新网的路由

* refactor: 保留旧路由，向下兼容

* fix: tryGet 包含了设置缓存，不需要再次 set

* fix: 使用 config 里的统一 ua 配置

* refactor: 统一使用 cheerio 搜索子节点的方式

* style: fix formatting

* fix: 移除无用参数

* docs: 更新路由

docs/other.md

@@ -550,3 +550,11 @@ type 为 all 时，category 参数不支持 cost 和 free
 <Route name="首页" author="qiwihui" example="/paidai" path="/paidao" />
 <Route name="论坛" author="qiwihui" example="/paidai/bbs" path="/paidao/bbs" />
 <Route name="商道" author="qiwihui" example="/paidai/news" path="/paidao/news" />
+
+## 财新博客
+
+<Route name="用户博客" author="Maecenas" example="/caixin/blog/zhangwuchang" path="/caixin/blog/:column" :paramsDesc="['博客名称，可在博客主页的 URL 找到']">
+
+通过提取文章全文，以提供比官方源更佳的阅读体验.
+
+</Route>

lib/router.js

@@ -405,6 +405,8 @@ router.get('/mihoyo/bh2/:type', require('./routes/mihoyo/bh2'));
 // 央视新闻
 router.get('/cctv/:category', require('./routes/cctv/category'));
 
+// 财新博客
+router.get('/caixin/blog/:column', require('./routes/caixin/blog'));
 // 财新
 router.get('/caixin/:column/:category', require('./routes/caixin/category'));
 

lib/routes/caixin/blog.js

@@ -0,0 +1,60 @@
+const axios = require('../../utils/axios');
+const cheerio = require('cheerio');
+const { ua } = require('../../config');
+const Parser = require('rss-parser');
+const parser = new Parser({ headers: { 'User-Agent': ua } });
+
+async function load(link, need_feed_description) {
+    const response = await axios.get(link);
+    const $ = cheerio.load(response.data);
+    const article = $('.blog_content').removeAttr('style');
+    article.find('img').removeAttr('style');
+    article
+        .find('div')
+        // Non-breaking space U+00A0, ` ` in html
+        // element.children[0].data === $(element, article).text()
+        .filter((_, element) => element.children[0].data === String.fromCharCode(160))
+        .remove();
+    const description = article.html();
+
+    const item = { description };
+
+    if (need_feed_description) {
+        const author = $('div.widget.author_detail p:nth-child(2)');
+        // workaround to split author name and author article
+        const author_name = author.find('strong');
+        author_name.text(author_name.text() + '，');
+        item.feed_description = author.text();
+    }
+    return item;
+}
+
+module.exports = async (ctx) => {
+    const { column } = ctx.params;
+    const link = `http://${column}.blog.caixin.com`;
+    const feed_url = `${link}/feed`;
+    const feed = await parser.parseURL(feed_url);
+
+    const title = `财新博客 - ${/[^»]*$/.exec(feed.title)[0]}`;
+
+    const items = await Promise.all(
+        feed.items.slice(0, 10).map(async (item, index) => {
+            const link = item.link;
+            const single = {
+                title: item.title,
+                pubDate: item.pubDate,
+                link,
+                author: item['dc:creator'],
+            };
+            const other = await ctx.cache.tryGet(link, async () => await load(link, index === 0));
+            return Promise.resolve(Object.assign({}, single, other));
+        })
+    );
+
+    ctx.state.data = {
+        title,
+        link,
+        description: items[0].feed_description,
+        item: items,
+    };
+};