Add Chrome and ChromeDriver installation to Dockerfile; enhance screenshot functionality with improved options and logging

Author: npv2k1

Dockerfile

@@ -1,6 +1,28 @@
 # Use a Python image with uv pre-installed
 FROM ghcr.io/astral-sh/uv:python3.11-bookworm
 
+# Install Chrome and dependencies
+RUN apt-get update && apt-get install -y \
+    wget \
+    gnupg \
+    curl \
+    unzip \
+    && wget -q -O - https://dl-ssl.google.com/linux/linux_signing_key.pub | apt-key add - \
+    && echo "deb [arch=amd64] http://dl.google.com/linux/chrome/deb/ stable main" >> /etc/apt/sources.list.d/google-chrome.list \
+    && apt-get update \
+    && apt-get install -y \
+    google-chrome-stable \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install ChromeDriver
+RUN CHROME_VERSION=$(google-chrome --version | awk '{print $3}' | awk -F'.' '{print $1}') \
+    && CHROMEDRIVER_VERSION=$(curl -s "https://chromedriver.storage.googleapis.com/LATEST_RELEASE_$CHROME_VERSION") \
+    && wget -q "https://chromedriver.storage.googleapis.com/$CHROMEDRIVER_VERSION/chromedriver_linux64.zip" \
+    && unzip chromedriver_linux64.zip \
+    && mv chromedriver /usr/local/bin/ \
+    && chmod +x /usr/local/bin/chromedriver \
+    && rm chromedriver_linux64.zip
+
 # Create a non-root user
 RUN useradd -m -u 1000 appuser
 
@@ -50,6 +72,4 @@ USER appuser
 
 EXPOSE 7860
 # Run the FastAPI application by default
-# Uses `fastapi dev` to enable hot-reloading when the `watch` sync occurs
-# Uses `--host 0.0.0.0` to allow access from outside the container
 CMD ["python", "main.py"]

src/modules/apps/__init__.py

@@ -4,6 +4,7 @@
 from selenium import webdriver
 from selenium.webdriver.chrome.options import Options
 from selenium.webdriver.chrome.service import Service
+from subprocess import PIPE, STDOUT
 print("Gradio app loaded.")
 
 def capture_page(url: str, output_file: str = "screenshot.png"):
@@ -14,15 +15,31 @@ def capture_page(url: str, output_file: str = "screenshot.png"):
     :param output_file: The filename to save the screenshot.
     """
     options = Options()
-    options.add_argument("--headless")  # Run in headless mode
-    options.add_argument("--window-size=1920,1080")  # Set window size
+    options.add_argument('--headless')
+    options.add_argument('--no-sandbox')  # Required in Docker
+    options.add_argument('--disable-dev-shm-usage')  # Required in Docker
+    options.add_argument('--disable-gpu')  # Required in Docker
+    options.add_argument('--disable-software-rasterizer')
+    options.add_argument('--window-size=1920,1080')
+    options.add_argument('--disable-extensions')
+    options.add_argument('--disable-infobars')
 
-    # Initialize Chrome service
-    service = Service()
-    driver = webdriver.Chrome(service=service, options=options)
+    # Set up Chrome service with explicit path to chromedriver and logging
+    service = Service(
+        executable_path='/usr/local/bin/chromedriver',
+        log_output=PIPE  # Redirect logs to pipe
+    )
+    
+    # Initialize Chrome with the service and options
+    driver = webdriver.Chrome(
+        service=service,
+        options=options
+    )
 
     try:
         driver.get(url)
+        # Add a small delay to ensure page loads completely
+        driver.implicitly_wait(5)
         driver.save_screenshot(output_file)
         print(f"Screenshot saved: {output_file}")
     finally:
@@ -44,6 +61,7 @@ def capture_and_show(url: str):
         # Return the image path
         return temp_path
     except Exception as e:
+        print(f"Error in capture_and_show: {str(e)}")  # Add detailed logging
         return f"Error capturing page: {str(e)}"
 
 def create_gradio_app():