feat(route): add EFE Noticias route (#22165)

* feat(route): add EFE Noticias route

Add route for EFE (Agencia EFE) Spanish news agency with 10 categories:
mundo, espana, economia, cultura, ciencia-y-tecnologia, deportes,
salud, medio-ambiente, educacion, euro-efe.

Features:
- Article images via enclosure and inline in description
- Ad filtering (auto-banner, srr-main, promotional logos)
- Clean img attributes for RSS reader compatibility
- Configurable item limit via query parameter

* fix(route/efe): address review comments

- Remove `limit` from `parameters` (query param, not path param)
- Scope article link selector to `.elementor-loop-container` to avoid anti-pattern #1
- Remove unnecessary `.first()` on `.elementor-widget-theme-post-content`
- Remove `.srr-main` from cleanup (not inside content area), keep `decoding`/`loading` on images

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: address review comments for EFE route

- Fix parameters to structured object with only path params
- Scope selector to .elementor-post article links only
- Remove unnecessary .first() on title
- Stop stripping image attributes (decoding, loading, width, etc.)

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: revert overly specific selector that broke all routes

.elementor-post class doesn't exist on EFE site, use attribute
selector to pre-filter article URLs instead.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: fetch articles sequentially to avoid 429 rate limiting

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: suppress no-await-in-loop lint for sequential rate-limited fetches

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: use specific selector for article title links only

Select only .elementor-widget-theme-post-title a inside .e-loop-item
to avoid matching image links, category links, or duplicate entries.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: use single quotes for static string (oxlint stylistic/quotes)

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix: use Promise.all and remove Set (selector matches once per article)

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* refactor: use Array.map instead of push for link collection

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

* fix(efe): use pMap to limit concurrency and avoid 429 rate limiting

---------

Co-authored-by: mlkgrnt <mlkgrnt@users.noreply.github.com>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>

Author: mlkgrnt

lib/routes/efe/index.ts

@@ -0,0 +1,97 @@
+import { load } from 'cheerio';
+import pMap from 'p-map';
+
+import type { Route } from '@/types';
+import cache from '@/utils/cache';
+import ofetch from '@/utils/ofetch';
+import { parseDate } from '@/utils/parse-date';
+
+const rootUrl = 'https://efe.com';
+
+const categories: Record<string, string> = {
+    mundo: 'Mundo',
+    espana: 'España',
+    economia: 'Economía',
+    cultura: 'Cultura',
+    'ciencia-y-tecnologia': 'Ciencia y Tecnología',
+    deportes: 'Deportes',
+    salud: 'Salud',
+    'medio-ambiente': 'Medio Ambiente',
+    educacion: 'Educación',
+    'euro-efe': 'EuroEFE',
+};
+
+export const route: Route = {
+    path: '/:category?',
+    name: 'Category',
+    maintainers: ['mlkgrnt'],
+    example: '/efe/mundo',
+    parameters: {
+        category: {
+            description: 'Category slug, see table below. Defaults to mundo.',
+            default: 'mundo',
+        },
+    },
+    handler,
+    categories: ['new-media'],
+    features: {
+        requireConfig: false,
+        requirePuppeteer: false,
+        antiCrawler: false,
+        supportBT: false,
+        supportPodcast: false,
+        supportScihub: false,
+    },
+    radar: [
+        {
+            source: ['efe.com/:category'],
+            target: '/:category',
+        },
+    ],
+};
+
+async function handler(ctx) {
+    const category = ctx.req.param('category') || 'mundo';
+    const limit = ctx.req.query('limit') ? Number.parseInt(ctx.req.query('limit'), 10) : 20;
+    const pageUrl = `${rootUrl}/${category}/`;
+
+    const response = await ofetch(pageUrl);
+    const $ = load(response);
+
+    const links = $('.e-loop-item .elementor-widget-theme-post-title a[href]')
+        .toArray()
+        .map((el) => $(el).attr('href'))
+        .filter((href): href is string => !!href && href.startsWith(`${rootUrl}/${category}/`) && /\/\d{4}-\d{2}-\d{2}\//.test(href));
+
+    const items = await pMap(
+        links.slice(0, limit),
+        (link) =>
+            cache.tryGet(link, async () => {
+                const detail = await ofetch(link);
+                const $detail = load(detail);
+
+                const title = $detail('title').text();
+                const dateMatch = detail.match(/"datePublished":\s*"([^"]+)"/);
+                const pubDate = dateMatch ? parseDate(dateMatch[1]) : undefined;
+
+                const image = $detail('meta[property="og:image"]').attr('content');
+                const content = $detail('.elementor-widget-theme-post-content');
+                content.find('.auto-banner').remove();
+                const description = (image ? `<figure><img src="${image}"></figure>` : '') + (content.html() || '');
+
+                return {
+                    title,
+                    link,
+                    pubDate,
+                    description,
+                };
+            }),
+        { concurrency: 2 }
+    );
+
+    return {
+        title: `EFE Noticias - ${categories[category] || category}`,
+        link: pageUrl,
+        item: items,
+    };
+}

lib/routes/efe/namespace.ts

@@ -0,0 +1,7 @@
+import type { Namespace } from '@/types';
+
+export const namespace: Namespace = {
+    name: 'EFE Noticias',
+    url: 'efe.com',
+    lang: 'es',
+};