initial production version

src/routes/api/scrape-images/+server.ts

@@ -0,0 +1,75 @@
+import { json } from '@sveltejs/kit';
+import type { RequestHandler } from './$types';
+
+export const POST: RequestHandler = async ({ request }) => {
+	const { url } = await request.json();
+
+	if (!url) {
+		return json({ error: 'URL is required' }, { status: 400 });
+	}
+
+	try {
+		const response = await fetch(url, {
+			headers: {
+				'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36'
+			}
+		});
+
+		if (!response.ok) {
+			return json({ error: 'Failed to fetch URL' }, { status: 400 });
+		}
+
+		const html = await response.text();
+		const baseUrl = new URL(url);
+		const origin = baseUrl.origin;
+
+		const imageUrls: string[] = [];
+		const imgRegex = /<img[^>]+src="([^">]+)"/g;
+		const ogImageRegex = /<meta[^>]+property="og:image"[^>]+content="([^">]+)"/g;
+		const twitterImageRegex = /<meta[^>]+name="twitter:image"[^>]+content="([^">]+)"/g;
+
+		function toAbsoluteUrl(imgUrl: string): string {
+			if (imgUrl.startsWith('http')) {
+				return imgUrl;
+			}
+			if (imgUrl.startsWith('//')) {
+				return `https:${imgUrl}`;
+			}
+			if (imgUrl.startsWith('/')) {
+				return `${origin}${imgUrl}`;
+			}
+			return `${origin}/${imgUrl}`;
+		}
+
+		let match;
+
+		while ((match = ogImageRegex.exec(html)) !== null) {
+			imageUrls.push(toAbsoluteUrl(match[1]));
+		}
+
+		while ((match = twitterImageRegex.exec(html)) !== null) {
+			imageUrls.push(toAbsoluteUrl(match[1]));
+		}
+
+		while ((match = imgRegex.exec(html)) !== null) {
+			const imgUrl = match[1];
+			const fullUrl = toAbsoluteUrl(imgUrl);
+			if (!imageUrls.includes(fullUrl)) {
+				imageUrls.push(fullUrl);
+			}
+		}
+
+		const filteredImages = imageUrls.filter(
+			(url) =>
+				!url.includes('logo') &&
+				!url.includes('icon') &&
+				!url.includes('sprite') &&
+				!url.endsWith('.svg') &&
+				url.length < 500
+		);
+
+		return json({ images: filteredImages.slice(0, 20) });
+	} catch (error) {
+		return json({ error: 'Failed to scrape images' }, { status: 500 });
+	}
+};