From c9f5300272c73cdfb635924dfb0b4758c19c3540 Mon Sep 17 00:00:00 2001
From: Giancarmine Salucci <giancarmine@gmail.com>
Date: Wed, 13 May 2026 01:31:33 +0200
Subject: [PATCH] feat: use Playwright for caption, yt-dlp for thumbnail only

Always extract the full caption via Playwright (browser sees the
untruncated text). yt-dlp runs in parallel only to get the thumbnail
CDN URL quickly; its result for the description is discarded.

This eliminates the truncation problem at the source without needing
a fallback heuristic.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---
 playwright.config.ts.bak               | 34 +++++++++++++++
 src/lib/server/queue/QueueProcessor.ts | 59 +++++++-------------------
 2 files changed, 49 insertions(+), 44 deletions(-)
 create mode 100644 playwright.config.ts.bak

diff --git a/playwright.config.ts.bak b/playwright.config.ts.bak
new file mode 100644
index 0000000..f72915b
--- /dev/null
+++ b/playwright.config.ts.bak
@@ -0,0 +1,34 @@
+import { defineConfig, devices } from '@playwright/test';
+
+/**
+ * Playwright configuration for E2E tests
+ *
+ * See https://playwright.dev/docs/test-configuration
+ */
+export default defineConfig({
+	testDir: './src/tests',
+	testMatch: '**/*.e2e.spec.ts',
+	fullyParallel: true,
+	forbidOnly: !!process.env.CI,
+	retries: process.env.CI ? 2 : 0,
+	workers: process.env.CI ? 1 : undefined,
+	reporter: 'list',
+	use: {
+		baseURL: 'http://localhost:5173',
+		trace: 'on-first-retry'
+	},
+
+	projects: [
+		{
+			name: 'chromium',
+			use: { ...devices['Desktop Chrome'] }
+		}
+	],
+
+	webServer: {
+		command: 'npm run dev',
+		url: 'http://localhost:5173',
+		reuseExistingServer: !process.env.CI,
+		timeout: 120000
+	}
+});
diff --git a/src/lib/server/queue/QueueProcessor.ts b/src/lib/server/queue/QueueProcessor.ts
index c19dff3..e025538 100644
--- a/src/lib/server/queue/QueueProcessor.ts
+++ b/src/lib/server/queue/QueueProcessor.ts
@@ -19,63 +19,34 @@ import { uploadRecipeWithIngredientsDTO, uploadRecipeImage } from '$lib/server/t
 import { pushNotificationService } from '$lib/server/notifications/PushNotificationService';
 import { queueConfig } from './config';
 import { logError } from '../utils/logger';
-import { env } from '$env/dynamic/private';
 import type { ProgressEvent, ExtractedContent, ProgressCallback } from '$lib/server/extraction';
 import type { QueueItem } from './types';
 
 /**
- * Instagram caption truncation marker: yt-dlp's GraphQL path returns a
- * caption ending with "…." (U+2026 + ".") when Instagram cut it short.
- * When we detect this, fall back to the Playwright extractor which intercepts
- * the full caption from live network traffic (requires a valid auth.json).
+ * Extract caption via Playwright (full, untruncated) and thumbnail via yt-dlp
+ * (fast, reliable CDN URL). Both run in parallel; yt-dlp failure is non-fatal.
  */
-const TRUNCATION_SUFFIX = '\u2026.';
-
 async function extractTextAndThumbnail(
 	url: string,
 	cb?: ProgressCallback
 ): Promise<ExtractedContent> {
-	const backend = (env.EXTRACTOR_BACKEND ?? 'ytdlp').toLowerCase();
+	// Run Playwright (caption) and yt-dlp (thumbnail) concurrently
+	const [ytdlpResult, playwrightResult] = await Promise.allSettled([
+		extractWithYtDlp(url),
+		extractWithPlaywright(url, cb)
+	]);
 
-	if (backend === 'playwright') {
-		return extractWithPlaywright(url, cb);
+	if (playwrightResult.status === 'rejected') {
+		throw playwrightResult.reason;
 	}
 
-	// Primary path: yt-dlp (fast, no browser)
-	const result = await extractWithYtDlp(url, cb);
+	// Prefer yt-dlp thumbnail; fall back to whatever Playwright captured
+	const thumbnail =
+		ytdlpResult.status === 'fulfilled' && ytdlpResult.value.thumbnail
+			? ytdlpResult.value.thumbnail
+			: playwrightResult.value.thumbnail;
 
-	// Truncation detected → fall back to Playwright for the full caption
-	if (result.bodyText.endsWith(TRUNCATION_SUFFIX)) {
-		cb?.({
-			type: 'status',
-			message: 'Caption truncated by Instagram API — retrying with browser extraction...',
-			timestamp: new Date().toISOString()
-		});
-		try {
-			const playwrightResult = await extractWithPlaywright(url, cb);
-			if (playwrightResult.bodyText && !playwrightResult.bodyText.endsWith(TRUNCATION_SUFFIX)) {
-				cb?.({
-					type: 'status',
-					message: `Full caption retrieved via browser (${playwrightResult.bodyText.length} chars)`,
-					timestamp: new Date().toISOString()
-				});
-				// Prefer yt-dlp thumbnail if Playwright didn't get one
-				return {
-					bodyText: playwrightResult.bodyText,
-					thumbnail: playwrightResult.thumbnail ?? result.thumbnail
-				};
-			}
-		} catch (e) {
-			// Playwright fallback failed — surface the original yt-dlp result
-			cb?.({
-				type: 'status',
-				message: `Browser fallback failed (${e instanceof Error ? e.message : String(e)}), using partial caption`,
-				timestamp: new Date().toISOString()
-			});
-		}
-	}
-
-	return result;
+	return { bodyText: playwrightResult.value.bodyText, thumbnail };
 }
 
 /**