feat: geizhals scraper with cheerio + tests

src/lib/scrapers/geizhals.ts

@@ -0,0 +1,48 @@
+import * as cheerio from 'cheerio'
+import type { PriceScraper, ScrapeResult } from './types'
+
+const UA = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0 Safari/537.36'
+
+export const geizhalsScraper: PriceScraper = {
+  shop: 'geizhals',
+  async scrape(url: string): Promise<ScrapeResult> {
+    try {
+      const res = await fetch(url, {
+        headers: { 'User-Agent': UA, 'Accept-Language': 'de-DE,de;q=0.9' },
+        signal: AbortSignal.timeout(20_000),
+      })
+      if (!res.ok) {
+        return { price: null, currency: 'EUR', availability: 'unknown', error: `HTTP ${res.status}` }
+      }
+      const $ = cheerio.load(await res.text())
+
+      const priceTexts = [
+        $('.gh_price').first().text(),
+        $('span.gh_price strong').first().text(),
+        $('[itemprop="price"]').attr('content'),
+        $('meta[itemprop="price"]').attr('content'),
+      ].filter(Boolean) as string[]
+
+      const price = parsePrice(priceTexts[0] ?? '')
+      const name = ($('h1[itemprop="name"]').text() || $('h1').first().text() || '').trim()
+      const imageUrl = $('img.product-gallery__image').first().attr('src')
+        || $('meta[property="og:image"]').attr('content')
+        || undefined
+
+      if (price === null) {
+        return { price: null, currency: 'EUR', availability: 'unknown', name, imageUrl, error: 'price-selector-missed' }
+      }
+
+      return { price, currency: 'EUR', availability: 'in_stock', name, imageUrl }
+    } catch (err) {
+      return { price: null, currency: 'EUR', availability: 'unknown', error: (err as Error).message }
+    }
+  },
+}
+
+function parsePrice(text: string): number | null {
+  const cleaned = text.replace(/[^\d.,]/g, '').replace(/\.(?=\d{3}(\D|$))/g, '').replace(',', '.')
+  if (!cleaned) return null
+  const n = parseFloat(cleaned)
+  return Number.isFinite(n) && n > 0 ? n : null
+}

tests/scrapers/geizhals.test.ts

@@ -0,0 +1,32 @@
+import { describe, it, expect, vi, beforeEach } from 'vitest'
+import { readFileSync } from 'node:fs'
+import { join } from 'node:path'
+import { geizhalsScraper } from '@/lib/scrapers/geizhals'
+
+const fixture = readFileSync(join(__dirname, '../fixtures/geizhals-gpu.html'), 'utf-8')
+
+beforeEach(() => {
+  global.fetch = vi.fn().mockResolvedValue({
+    ok: true,
+    status: 200,
+    text: async () => fixture,
+  }) as unknown as typeof fetch
+})
+
+describe('geizhalsScraper', () => {
+  it('extracts price and name', async () => {
+    const r = await geizhalsScraper.scrape('https://geizhals.de/test')
+    expect(r.price).toBeGreaterThan(0)
+    expect(r.currency).toBe('EUR')
+    expect(r.name).toBeTruthy()
+  })
+
+  it('returns error on HTTP failure', async () => {
+    global.fetch = vi.fn().mockResolvedValue({
+      ok: false, status: 503, text: async () => '',
+    }) as unknown as typeof fetch
+    const r = await geizhalsScraper.scrape('https://geizhals.de/test')
+    expect(r.price).toBeNull()
+    expect(r.error).toMatch(/HTTP 503/)
+  })
+})