Add news api with NPR and Al Jazeera test sources

package.json

@@ -13,6 +13,7 @@
   },
   "dependencies": {
     "@unocss/preset-web-fonts": "^0.61.0",
+    "cheerio": "^1.0.0-rc.12",
     "hono": "^4.4.6",
     "nodemon": "^3.1.0",
     "unocss": "^0.61.0",

public/styles/uno.css

@@ -43,12 +43,10 @@
 .bg-gray-300{--un-bg-opacity:1;background-color:rgb(209 213 219 / var(--un-bg-opacity));}
 .bg-indigo-200{--un-bg-opacity:1;background-color:rgb(199 210 254 / var(--un-bg-opacity));}
 .bg-transparent{background-color:transparent;}
-.bg-white,
-[bg-white=""]{--un-bg-opacity:1;background-color:rgb(255 255 255 / var(--un-bg-opacity));}
+.bg-white{--un-bg-opacity:1;background-color:rgb(255 255 255 / var(--un-bg-opacity));}
 .dark .dark\:bg-blue-600{--un-bg-opacity:1;background-color:rgb(37 99 235 / var(--un-bg-opacity));}
 .dark .dark\:bg-gray-700{--un-bg-opacity:1;background-color:rgb(55 65 81 / var(--un-bg-opacity));}
-.dark .dark\:bg-gray-900,
-.dark [dark\:bg-gray-900=""]{--un-bg-opacity:1;background-color:rgb(17 24 39 / var(--un-bg-opacity));}
+.dark .dark\:bg-gray-900{--un-bg-opacity:1;background-color:rgb(17 24 39 / var(--un-bg-opacity));}
 .dark .dark\:bg-indigo-600{--un-bg-opacity:1;background-color:rgb(79 70 229 / var(--un-bg-opacity));}
 .hover\:bg-blue-400:hover{--un-bg-opacity:1;background-color:rgb(96 165 250 / var(--un-bg-opacity));}
 .p-0{padding:0;}
@@ -66,8 +64,7 @@
 .text-5xl{font-size:3rem;line-height:1;}
 .text-base{font-size:1rem;line-height:1.5rem;}
 .text-sm{font-size:0.875rem;line-height:1.25rem;}
-.dark .dark\:text-white,
-.dark [dark\:text-white=""]{--un-text-opacity:1;color:rgb(255 255 255 / var(--un-text-opacity));}
+.dark .dark\:text-white{--un-text-opacity:1;color:rgb(255 255 255 / var(--un-text-opacity));}
 .text-black{--un-text-opacity:1;color:rgb(0 0 0 / var(--un-text-opacity));}
 .text-neutral-500{--un-text-opacity:1;color:rgb(115 115 115 / var(--un-text-opacity));}
 .text-slate-900{--un-text-opacity:1;color:rgb(15 23 42 / var(--un-text-opacity));}

src/util/api.ts

@@ -0,0 +1,69 @@
+import { load } from "cheerio";
+import { z } from "zod";
+
+const articleSchema = z.object({
+  title: z.string().min(5),
+  link: z.string().url(),
+  source: z.string(),
+});
+
+type NewsSource = {
+  name: string;
+  url: (page: number) => string;
+  listSelector: string;
+  baseUrl?: string;
+};
+
+const newsSources: NewsSource[] = [
+  {
+    name: "NPR",
+    url: (page: number) => `http://text.npr.org?page=${page}`,
+    listSelector: "ul > li > a",
+    baseUrl: "http://text.npr.org",
+  },
+  {
+    name: "Al Jazeera",
+    url: (page: number) => `https://www.aljazeera.com/us-canada?page=${page}`,
+    listSelector: "article .gc__content a",
+    baseUrl: "https://www.aljazeera.com",
+  },
+];
+
+const isValidArticle = (article: { title: string; link: string }) => {
+  try {
+    articleSchema.parse(article);
+    return true;
+  } catch (e) {
+    return false;
+  }
+};
+
+const fetchArticlesFromSource = async (
+  source: NewsSource,
+  page: number = 1,
+) => {
+  const response = await fetch(source.url(page));
+  const text = await response.text();
+
+  const $ = load(text);
+  const articles: { title: string; link: string; source: string }[] = [];
+
+  $(source.listSelector).each((_, element) => {
+    const title = $(element).text().trim();
+    const link = source.baseUrl
+      ? `${source.baseUrl}${$(element).attr("href")}`
+      : $(element).attr("href");
+
+    if (title && link) {
+      articles.push({
+        title,
+        link,
+        source: source.name,
+      });
+    }
+  });
+
+  return articles.filter(isValidArticle);
+};
+
+export { fetchArticlesFromSource, isValidArticle, newsSources };

test/api.test.ts

@@ -0,0 +1,28 @@
+import { describe, expect, it } from "bun:test";
+import { fetchArticlesFromSource, newsSources } from "../src/util/api";
+
+describe("Article Fetching Functions", () => {
+  it("Should fetch and parse NPR articles", async () => {
+    const articles = await fetchArticlesFromSource(newsSources[0], 1);
+    expect(articles.length).toBeGreaterThanOrEqual(10);
+    articles.forEach((article) => {
+      expect(article).toMatchObject({
+        title: expect.any(String),
+        link: expect.any(String),
+        source: "NPR",
+      });
+    });
+  });
+
+  it("Should fetch and parse Al Jazeera articles", async () => {
+    const articles = await fetchArticlesFromSource(newsSources[1], 1);
+    expect(articles.length).toBeGreaterThanOrEqual(10);
+    articles.forEach((article) => {
+      expect(article).toMatchObject({
+        title: expect.any(String),
+        link: expect.any(String),
+        source: "Al Jazeera",
+      });
+    });
+  });
+});

tsconfig.json

@@ -1,11 +1,11 @@
 {
-  "compilerOptions": { 
+  "compilerOptions": {
     "jsx": "react-jsx",
     "jsxImportSource": "hono/jsx",
     "allowImportingTsExtensions": true,
     "noEmit": true,
-    "types": [
-      "bun-types"
-    ]
+    "types": ["bun-types"],
+    "esModuleInterop": true,
+    "moduleResolution": "node"
   }
 }