Update app.py

app.py

@@ -1,146 +1,147 @@
-import gradio as gr  
-from playwright.sync_api import sync_playwright, Error as PlaywrightError  
-import subprocess  
-import os  
-import time  
-  
-# --- Helper to ensure Playwright browser is installed ---  
-# This function attempts to install the browser if not found.  
-# It's best-effort for standard Gradio spaces; a Dockerfile is more robust.  
-def install_playwright_browser_if_needed():  
-    try:  
-        # Check if browser is callable by Playwright  
-        with sync_playwright() as p:  
-            try:  
-                browser = p.chromium.launch(headless=False)  
-                browser.close()  
-                print("Playwright Chromium browser is available.")  
-                return True  
-            except PlaywrightError:  
-                print("Playwright Chromium browser not found or not executable. Attempting installation.")  
-  
-        # If not found, try to install.  
-        # The `packages.txt` should have installed most system dependencies.  
-        # This command installs the browser itself into Playwright's managed location.  
-        print("Attempting to install Playwright Chromium browser executable...")  
-        try:  
-            # Using subprocess to run the playwright install command  
-            subprocess.run(["playwright", "install", "chromium"], check=True, capture_output=True, text=True)  
-            print("Playwright Chromium executable installed successfully.")  
-            # Verify again  
-            with sync_playwright() as p:  
-                browser = p.chromium.launch(headless=False)  
-                browser.close()  
-            print("Playwright Chromium successfully verified after installation.")  
-            return True  
-        except subprocess.CalledProcessError as e:  
-            print(f"Playwright install chromium failed. STDERR: {e.stderr} STDOUT: {e.stdout}")  
-            return False  
-        except FileNotFoundError:  
-            print("Playwright command not found. Ensure 'playwright' is in requirements.txt and installed.")  
-            return False  
-    except Exception as e:  
-        print(f"An error occurred during Playwright browser setup: {e}")  
-        return False  
-  
-# Run browser installation check when the app starts.  
-# Logs will appear in the Hugging Face Space "Logs" tab.  
-print("Initializing Space: Checking/Installing Playwright browser...")  
-BROWSER_READY = install_playwright_browser_if_needed()  
-if BROWSER_READY:  
-    print("Browser is ready.")  
-else:  
-    print("WARNING: Browser installation failed or could not be verified. Screenshot functionality may not work.")  
-  
-# --- Screenshot function ---  
-def take_web_screenshot(url: str):  
-    if not BROWSER_READY:  
-        return None, "Error: Playwright browser (Chromium) is not properly installed or configured. Cannot take screenshot."  
-  
-    if not url:  
-        return None, "Please enter a website URL."  
-      
-    # Prepend https:// if no scheme is present, as Playwright requires it.  
-    if not (url.startswith("http://") or url.startswith("https://")):  
-        url = "https://" + url  
-  
-    screenshot_path = None  # Initialize here  
-  
-    try:  
-        with sync_playwright() as p:  
-            browser = p.chromium.launch(headless=False)  
-            context = browser.new_context(  
-                user_agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/98.0.4758.102 Safari/537.36"  
-            )  
-            page = context.new_page()  
-              
-            # Set a common viewport size  
-            page.set_viewport_size({"width": 1280, "height": 720})  
-              
-            print(f"Navigating to URL: {url}")  
-            # Increased timeout and wait_until 'domcontentloaded' or 'load' can be more reliable  
-            page.goto(url, timeout=60000, wait_until="domcontentloaded") # 60 seconds timeout  
-              
-            # Give some time for lazy-loaded elements if necessary, though 'networkidle' can be slow/unreliable  
-            # page.wait_for_timeout(2000) # Optional: 2 seconds, adjust as needed  
-  
-            # Create a unique filename for the screenshot  
-            os.makedirs("screenshots", exist_ok=True) # Ensure screenshots directory exists  
-            timestamp = time.strftime("%Y%m%d-%H%M%S")  
-            # Simple sanitization for filename  
-            sanitized_url_part = "".join(c if c.isalnum() else "_" for c in url.split("://")[-1])[:50]  
-            screenshot_path = f"screenshots/screenshot_{sanitized_url_part}_{timestamp}.png"  
-              
-            print(f"Taking screenshot and saving to: {screenshot_path}")  
-            page.screenshot(path=screenshot_path, full_page=False) # Captures the viewport  
-              
-            browser.close()  
-            print("Screenshot successful.")  
-            return screenshot_path, f"Screenshot of {url} captured successfully!"  
-  
-    except PlaywrightError as e:  
-        error_message_detail = str(e)  
-        print(f"Playwright error: {error_message_detail}")  
-        if "net::ERR_NAME_NOT_RESOLVED" in error_message_detail:  
-            status_message = f"Error: The URL '{url}' could not be resolved. Please check the domain name."  
-        elif "Timeout" in error_message_detail:  
-            status_message = f"Error: Timeout while loading '{url}'. The page might be too slow, offline, or protected."  
-        elif "SSL_ERROR" in error_message_detail or "ssl_error" in error_message_detail.lower():  
-             status_message = f"Error: SSL certificate issue with '{url}'. The site might be insecure or misconfigured."  
-        else:  
-            status_message = f"Error: Could not take screenshot of '{url}'. Details: {error_message_detail[:200]}..."  
-        # If a screenshot was partially created or an old one exists, don't return it on error  
-        return None, status_message  
-    except Exception as e:  
-        print(f"An unexpected error occurred: {e}")  
-        return None, f"An unexpected error occurred: {str(e)[:200]}..."  
-  
-# --- Gradio Interface ---  
-iface = gr.Interface(  
-    fn=take_web_screenshot,  
-    inputs=gr.Textbox(  
-        label="Website URL",  
-        placeholder="e.g., https://www.example.com or example.com"  
-    ),  
-    outputs=[  
-        gr.Image(type="filepath", label="Website Screenshot"),  
-        gr.Textbox(label="Status")  
-    ],  
-    title="Website Screenshot Taker 📸",  
-    description=(  
-        "Enter a full website URL to capture a screenshot of its current view. "  
-        "Prefix with http:// or https:// for best results."  
-    ),  
-    examples=[  
-        ["https://gradio.app"],  
-        ["huggingface.co/spaces"],  
-        ["en.wikipedia.org/wiki/Python_(programming_language)"]  
-    ],  
-    allow_flagging="never",  
-    css=".gradio-container {max-width: 960px !important; margin: auto !important;}" # Optional: center and limit width  
-)  
-  
-# --- Main launch ---  
-if __name__ == "__main__":  
-    print("Starting Gradio application...")  
-    iface.launch()  
+import gradio as gr
+from playwright.sync_api import sync_playwright, Error as PlaywrightError
+import subprocess
+import os
+import time
+
+# --- Helper to ensure Playwright browser is installed ---
+# This function attempts to install the browser if not found.
+# It's best-effort for standard Gradio spaces; a Dockerfile is more robust.
+def install_playwright_browser_if_needed():
+    try:
+        # Check if browser is callable by Playwright
+        with sync_playwright() as p:
+            try:
+                browser = p.chromium.launch(headless=True)
+                browser.close()
+                print("Playwright Chromium browser is available.")
+                return True
+            except PlaywrightError:
+                print("Playwright Chromium browser not found or not executable. Attempting installation.")
+
+        # If not found, try to install.
+        # The `packages.txt` should have installed most system dependencies.
+        # This command installs the browser itself into Playwright's managed location.
+        print("Attempting to install Playwright Chromium browser executable...")
+        try:
+            # Using subprocess to run the playwright install command
+            subprocess.run(["playwright", "install", "chromium"], check=True, capture_output=True, text=True)
+            print("Playwright Chromium executable installed successfully.")
+            # Verify again
+            with sync_playwright() as p:
+                browser = p.chromium.launch(headless=True)
+                browser.close()
+            print("Playwright Chromium successfully verified after installation.")
+            return True
+        except subprocess.CalledProcessError as e:
+            print(f"Playwright install chromium failed. STDERR: {e.stderr} STDOUT: {e.stdout}")
+            return False
+        except FileNotFoundError:
+            print("Playwright command not found. Ensure 'playwright' is in requirements.txt and installed.")
+            return False
+    except Exception as e:
+        print(f"An error occurred during Playwright browser setup: {e}")
+        return False
+
+# Run browser installation check when the app starts.
+# Logs will appear in the Hugging Face Space "Logs" tab.
+print("Initializing Space: Checking/Installing Playwright browser...")
+BROWSER_READY = install_playwright_browser_if_needed()
+if BROWSER_READY:
+    print("Browser is ready.")
+else:
+    print("WARNING: Browser installation failed or could not be verified. Screenshot functionality may not work.")
+
+# --- Screenshot function ---
+def take_web_screenshot(url: str):
+    if not BROWSER_READY:
+        return None, "Error: Playwright browser (Chromium) is not properly installed or configured. Cannot take screenshot."
+
+    if not url:
+        return None, "Please enter a website URL."
+    
+    # Prepend https:// if no scheme is present, as Playwright requires it.
+    if not (url.startswith("http://") or url.startswith("https://")):
+        url = "https://" + url
+
+    screenshot_path = None  # Initialize here
+
+    try:
+        with sync_playwright() as p:
+            browser = p.chromium.launch(headless=True)
+            context = browser.new_context(
+                user_agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/98.0.4758.102 Safari/537.36"
+            )
+            page = context.new_page()
+            
+            # Set a common viewport size
+            page.set_viewport_size({"width": 1280, "height": 720})
+            
+            print(f"Navigating to URL: {url}")
+            # Increased timeout and wait_until 'domcontentloaded' or 'load' can be more reliable
+            page.goto(url, timeout=60000, wait_until="domcontentloaded") # 60 seconds timeout
+            
+            # Give some time for lazy-loaded elements if necessary, though 'networkidle' can be slow/unreliable
+            # page.wait_for_timeout(2000) # Optional: 2 seconds, adjust as needed
+
+            # Create a unique filename for the screenshot
+            os.makedirs("screenshots", exist_ok=True) # Ensure screenshots directory exists
+            timestamp = time.strftime("%Y%m%d-%H%M%S")
+            # Simple sanitization for filename
+            sanitized_url_part = "".join(c if c.isalnum() else "_" for c in url.split("://")[-1])[:50]
+            screenshot_path = f"screenshots/screenshot_{sanitized_url_part}_{timestamp}.png"
+            
+            print(f"Taking screenshot and saving to: {screenshot_path}")
+            page.screenshot(path=screenshot_path, full_page=False) # Captures the viewport
+            
+            browser.close()
+            print("Screenshot successful.")
+            return screenshot_path, f"Screenshot of {url} captured successfully!"
+
+    except PlaywrightError as e:
+        error_message_detail = str(e)
+        print(f"Playwright error: {error_message_detail}")
+        if "net::ERR_NAME_NOT_RESOLVED" in error_message_detail:
+            status_message = f"Error: The URL '{url}' could not be resolved. Please check the domain name."
+        elif "Timeout" in error_message_detail:
+            status_message = f"Error: Timeout while loading '{url}'. The page might be too slow, offline, or protected."
+        elif "SSL_ERROR" in error_message_detail or "ssl_error" in error_message_detail.lower():
+             status_message = f"Error: SSL certificate issue with '{url}'. The site might be insecure or misconfigured."
+        else:
+            status_message = f"Error: Could not take screenshot of '{url}'. Details: {error_message_detail[:200]}..."
+        # If a screenshot was partially created or an old one exists, don't return it on error
+        return None, status_message
+    except Exception as e:
+        print(f"An unexpected error occurred: {e}")
+        return None, f"An unexpected error occurred: {str(e)[:200]}..."
+
+# --- Gradio Interface ---
+iface = gr.Interface(
+    fn=take_web_screenshot,
+    inputs=gr.Textbox(
+        label="Website URL",
+        placeholder="e.g., https://www.example.com or example.com"
+    ),
+    outputs=[
+        gr.Image(type="filepath", label="Website Screenshot"),
+        gr.Textbox(label="Status")
+    ],
+    title="Website Screenshot Taker 📸",
+    description=(
+        "Enter a full website URL to capture a screenshot of its current view. "
+        "The system will attempt to install Chromium if it's not already present (check Space logs for details). "
+        "Prefix with http:// or https:// for best results."
+    ),
+    examples=[
+        ["https://gradio.app"],
+        ["huggingface.co/spaces"],
+        ["en.wikipedia.org/wiki/Python_(programming_language)"]
+    ],
+    allow_flagging="never",
+    css=".gradio-container {max-width: 960px !important; margin: auto !important;}" # Optional: center and limit width
+)
+
+# --- Main launch ---
+if __name__ == "__main__":
+    print("Starting Gradio application...")
+    iface.launch()