cleanup

Dockers/puppeteer-api/app/utils/browser_utils.py

@@ -1,6 +1,11 @@
 from playwright.async_api import async_playwright
-from app.config import CUSTOM_USER_AGENT
+from app.config import CUSTOM_USER_AGENT, BROWSER_INSTANCE_TIMEOUT_MINUTES
 import asyncio
+import time
+
+# Global tracking for browser instances created by this module
+page_creation_times = {}
+active_pages = set()
 
 async def wait_for_network_idle(page):
     """Wait until no network requests are in flight"""
@@ -32,6 +37,10 @@ async def safe_browser_operation(url, operation_func):
         page = await context.new_page()
         page.set_default_timeout(30000)
 
+        # Track page creation time for force cleanup
+        page_creation_times[page] = time.time()
+        active_pages.add(page)
+
         # Call the operation function that uses the page
         result = await operation_func(page)
 
@@ -46,9 +55,14 @@ async def safe_browser_operation(url, operation_func):
             "url": url
         }
     finally:
-        # Ensure page is closed properly
+        # Cleanup page tracking
         if page:
             try:
+                # Remove from tracking
+                if page in active_pages:
+                    active_pages.remove(page)
+                if page in page_creation_times:
+                    del page_creation_times[page]
                 await page.close()
             except Exception as e:
                 print(f"Error closing page: {e}")
@@ -72,4 +86,33 @@ async def safe_browser_operation(url, operation_func):
             try:
                 await playwright.stop()
             except Exception as e:
-                print(f"Error closing playwright: {e}")
+                print(f"Error closing playwright: {e}")
+
+async def force_cleanup_old_pages():
+    """Force cleanup old page instances created by this module"""
+    print(f"Checking for old page instances in browser_utils (timeout: {BROWSER_INSTANCE_TIMEOUT_MINUTES} minutes)...")
+
+    current_time = time.time()
+    timeout_seconds = BROWSER_INSTANCE_TIMEOUT_MINUTES * 60
+    cleaned_pages = 0
+
+    # Clean up old pages
+    pages_to_cleanup = []
+    for page in list(active_pages):
+        creation_time = page_creation_times.get(page, 0)
+        if current_time - creation_time > timeout_seconds:
+            pages_to_cleanup.append(page)
+            print(f"Marking page for cleanup (age: {(current_time - creation_time)/60:.1f} minutes)")
+
+    for page in pages_to_cleanup:
+        try:
+            if page in active_pages:
+                active_pages.remove(page)
+            if page in page_creation_times:
+                del page_creation_times[page]
+            await page.close()
+            cleaned_pages += 1
+        except Exception as e:
+            print(f"Error cleaning up old page: {e}")
+
+    print(f"Force cleanup completed: {cleaned_pages} pages cleaned")