Update scraper.js

scraper.js

@@ -15,44 +15,46 @@ async function getPinterestVideos(query) {
         const searchUrl = `https://www.pinterest.com/search/pins/?q=${encodeURIComponent(query)}&rs=typed`;
         await page.goto(searchUrl, { waitUntil: 'networkidle2', timeout: 30000 });
 
-        // Scroll deeply to get ~100 pins
-        for (let i = 0; i < 5; i++) {
-            await page.evaluate(() => window.scrollBy(0, 2000));
-            await new Promise(r => setTimeout(r, 1000));
-        }
+        // --- SCROLLING LOGIC TO FIND MORE PINS ---
+        await page.evaluate(async () => {
+            for (let i = 0; i < 3; i++) { // Scroll 3 times to load ~60 pins
+                window.scrollBy(0, window.innerHeight * 2);
+                await new Promise(r => setTimeout(r, 1500));
+            }
+        });
 
         const pinUrls = await page.evaluate(() => {
-            const anchors = Array.from(document.querySelectorAll('a[href*="/pin/"]'));
-            return [...new Set(anchors.map(a => a.href))];
+            return Array.from(document.querySelectorAll('a[href*="/pin/"]'))
+                .map(a => a.href).filter(href => href.includes('/pin/'));
         });
 
-        const limit = 80; 
-        const selectedPins = pinUrls.slice(0, limit);
-        console.error(`Scraped ${selectedPins.length} pins. Extracting video data...`);
-
-        // Process pins in small batches to avoid crashing the Hugging Face CPU
-        for (let i = 0; i < selectedPins.length; i += 5) {
-            const batch = selectedPins.slice(i, i + 5);
-            await Promise.all(batch.map(async (url) => {
-                const p = await browser.newPage();
-                try {
-                    await p.goto(url, { waitUntil: 'domcontentloaded', timeout: 8000 });
-                    const html = await p.content();
-                    const match = html.match(/https:\/\/v1\.pinimg\.com\/videos\/mc\/[^\s"']+/);
-                    if (match) {
-                        let raw = match[0].replace(/\\u002F/g, '/').split('"')[0].split("'")[0];
-                        const hMatch = raw.match(/([a-f0-9]{32})/);
-                        if (hMatch) {
-                            const h = hMatch[1];
-                            const hd = `https://v1.pinimg.com/videos/mc/720p/${h.substring(0,2)}/${h.substring(2,4)}/${h.substring(4,6)}/${h}.mp4`;
-                            process.stdout.write(hd + "\n");
-                        }
+        // Get up to 60 unique pins to check
+        const uniquePins = [...new Set(pinUrls)].slice(0, 60);
+        console.error(`Found ${uniquePins.length} potential pins. Checking for video files...`);
+
+        for (const pinUrl of uniquePins) {
+            const pPage = await browser.newPage();
+            try {
+                // We go to each pin and try to find the video source
+                await pPage.goto(pinUrl, { waitUntil: 'domcontentloaded', timeout: 10000 });
+                const content = await pPage.content();
+                const videoRegex = /https:\/\/v1\.pinimg\.com\/videos\/mc\/[^\s"']+/g;
+                const matches = content.match(videoRegex);
+
+                if (matches) {
+                    let rawUrl = matches[0].replace(/\\u002F/g, '/').replace(/[奖励"']/g, '');
+                    const hashMatch = rawUrl.match(/([a-f0-9]{32})/);
+                    if (hashMatch) {
+                        const h = hashMatch[1];
+                        // Output the 720p URL for Python to check
+                        process.stdout.write(`https://v1.pinimg.com/videos/mc/720p/${h.substring(0,2)}/${h.substring(2,4)}/${h.substring(4,6)}/${h}.mp4\n`);
                     }
-                } catch (e) {} finally { await p.close(); }
-            }));
+                }
+            } catch (e) {} 
+            await pPage.close();
         }
-    } catch (err) {
-        console.error(err);
+    } catch (error) {
+        console.error("Scraper Error: " + error.message);
     } finally {
         if (browser) await browser.close();
     }