From 0a79a862a7e46ec2b944b3f0271e4f545b52a143 Mon Sep 17 00:00:00 2001
From: pandada8 <pandada8@gmail.com>
Date: Wed, 2 Oct 2024 03:34:24 +0800
Subject: [PATCH] feat(routes/ceph): add ceph blog (#16980)

* feat(routes/ceph): add ceph blog

* Apply suggestions from code review

Co-authored-by: Tony <TonyRL@users.noreply.github.com>

* fix(route/ceph): add 1hour cache for index, fix url double slash

* fix(route/ceph): remove unnecessary caching

---------
---
 lib/routes/ceph/blog.ts      | 72 ++++++++++++++++++++++++++++++++++++
 lib/routes/ceph/namespace.ts |  7 ++++
 2 files changed, 79 insertions(+)
 create mode 100644 lib/routes/ceph/blog.ts
 create mode 100644 lib/routes/ceph/namespace.ts

diff --git a/lib/routes/ceph/blog.ts b/lib/routes/ceph/blog.ts
new file mode 100644
index 000000000..7c5e1dc9a
--- /dev/null
+++ b/lib/routes/ceph/blog.ts
@@ -0,0 +1,72 @@
+import { Data, Route } from '@/types';
+import cache from '@/utils/cache';
+import got from '@/utils/got';
+import { load } from 'cheerio';
+import { parseDate } from '@/utils/parse-date';
+import { Context } from 'hono';
+
+export const route: Route = {
+    path: '/blog/:topic?',
+    categories: ['blog'],
+    example: '/ceph/blog/a11y',
+    parameters: {
+        category: 'filter blog post by category, return all posts if not specified',
+    },
+    features: {
+        requireConfig: false,
+        requirePuppeteer: false,
+        antiCrawler: false,
+        supportBT: false,
+        supportPodcast: false,
+        supportScihub: false,
+    },
+    radar: [
+        {
+            source: ['ceph.io/'],
+        },
+    ],
+    name: 'Blog',
+    maintainers: ['pandada8'],
+    handler,
+    url: 'ceph.io',
+};
+
+async function handler(ctx: Context): Promise<Data> {
+    const { category } = ctx.req.param();
+    const url = category ? `https://ceph.io/en/news/blog/category/${category}/` : 'https://ceph.io/en/news/blog/';
+    const response = await got.get(url);
+    const data = response.data;
+    const $ = load(data);
+    const list = $('#main .section li')
+        .toArray()
+        .map((e) => {
+            const element = $(e);
+            const title = element.find('a').text().trim();
+            const pubDate = parseDate(element.find('time').attr('datetime'));
+            return {
+                title,
+                link: new URL(element.find('a').attr('href'), 'https://ceph.io').href,
+                pubDate,
+            };
+        });
+
+    const result = await Promise.all(
+        list.map((item) =>
+            cache.tryGet(item.link, async () => {
+                const itemReponse = await got.get(item.link);
+                const data = itemReponse.data;
+                const item$ = load(data);
+
+                item.author = item$('#main section > div:nth-child(1) span').text().trim();
+                item.description = item$('#main section > div:nth-child(2) > div').html();
+                return item;
+            })
+        )
+    );
+
+    return {
+        title: 'Ceph Blog',
+        link: url,
+        item: result,
+    };
+}
diff --git a/lib/routes/ceph/namespace.ts b/lib/routes/ceph/namespace.ts
new file mode 100644
index 000000000..0585fb5b5
--- /dev/null
+++ b/lib/routes/ceph/namespace.ts
@@ -0,0 +1,7 @@
+import type { Namespace } from '@/types';
+
+export const namespace: Namespace = {
+    name: 'Ceph',
+    url: 'ceph.io',
+    description: 'Ceph is an open source distributed storage system designed to evolve with data.',
+};