major clean up in unused libraries in requirements

app.py

@@ -26,81 +26,6 @@ try:
 except ImportError:
     print("python-dotenv not installed, skipping .env file loading")
 
-# Function to setup Tesseract
-def setup_tesseract():
-    """Setup Tesseract OCR environment."""
-    print("Setting up Tesseract OCR environment...")
-    
-    # Create tessdata directory if it doesn't exist
-    tessdata_dir = os.path.join(current_dir, "tessdata")
-    os.makedirs(tessdata_dir, exist_ok=True)
-    
-    # Set TESSDATA_PREFIX environment variable if not already set
-    if not os.environ.get('TESSDATA_PREFIX'):
-        # Check multiple possible locations
-        possible_tessdata_dirs = [
-            tessdata_dir,  # Our local tessdata directory
-            "/usr/share/tesseract-ocr/4.00/tessdata",  # Common location in Hugging Face
-            "/usr/share/tesseract-ocr/tessdata",  # Another common location
-            "/usr/local/share/tessdata",  # Standard installation location
-        ]
-        
-        # Use the first directory that exists
-        for dir_path in possible_tessdata_dirs:
-            if os.path.exists(dir_path):
-                os.environ['TESSDATA_PREFIX'] = dir_path
-                print(f"Set TESSDATA_PREFIX to {dir_path}")
-                break
-        else:
-            # If none exist, use our local directory
-            os.environ['TESSDATA_PREFIX'] = tessdata_dir
-            print(f"No existing tessdata directory found, set TESSDATA_PREFIX to {tessdata_dir}")
-    
-    # Download eng.traineddata if it doesn't exist in our local tessdata
-    eng_traineddata = os.path.join(tessdata_dir, "eng.traineddata")
-    if not os.path.exists(eng_traineddata):
-        try:
-            print("Downloading eng.traineddata...")
-            url = "https://github.com/tesseract-ocr/tessdata/raw/main/eng.traineddata"
-            urllib.request.urlretrieve(url, eng_traineddata)
-            print("Downloaded eng.traineddata")
-        except Exception as e:
-            print(f"Error downloading eng.traineddata: {e}")
-    
-    # Configure pytesseract
-    try:
-        import pytesseract
-        # Check if tesseract is in PATH
-        tesseract_cmd = shutil.which("tesseract")
-        if tesseract_cmd:
-            pytesseract.pytesseract.tesseract_cmd = tesseract_cmd
-            print(f"Set pytesseract.tesseract_cmd to {tesseract_cmd}")
-        else:
-            # Try common locations
-            common_locations = [
-                "/usr/bin/tesseract",
-                "/usr/local/bin/tesseract",
-                "/app/tesseract/tesseract"
-            ]
-            for location in common_locations:
-                if os.path.isfile(location) and os.access(location, os.X_OK):
-                    pytesseract.pytesseract.tesseract_cmd = location
-                    print(f"Set pytesseract.tesseract_cmd to {location}")
-                    break
-            else:
-                print("Warning: Could not find tesseract executable")
-    except ImportError:
-        print("pytesseract not installed")
-    
-    # Try to import tesserocr to verify it's working
-    try:
-        import tesserocr
-        print(f"tesserocr imported successfully, version: {tesserocr.tesseract_version()}")
-    except ImportError:
-        print("tesserocr not installed or not working")
-    except Exception as e:
-        print(f"Error importing tesserocr: {e}")
-
 # Load Gemini API key from environment variable
 gemini_api_key = os.getenv("GOOGLE_API_KEY")
 
@@ -131,8 +56,5 @@ except ModuleNotFoundError:
         # Try import again
         from src.main import main
 
-# Call setup function at import time
-setup_tesseract()
-
 if __name__ == "__main__":
     main()

build.sh

@@ -5,73 +5,11 @@ set -e
 
 echo "Starting build process..."
 
-# Install system dependencies for tesseract
-echo "Installing Tesseract and dependencies..."
+# Install system dependencies
+echo "Installing system dependencies..."
 apt-get update && apt-get install -y \
-    tesseract-ocr \
-    tesseract-ocr-eng \
-    libtesseract-dev \
-    libleptonica-dev \
-    pkg-config \
-    wget
-
-# Create tessdata directory
-TESSDATA_DIR="/usr/share/tesseract-ocr/4.00/tessdata"
-mkdir -p "$TESSDATA_DIR"
-
-# Download traineddata files directly from the official repository
-echo "Downloading Tesseract traineddata files..."
-wget -O "$TESSDATA_DIR/eng.traineddata" "https://github.com/tesseract-ocr/tessdata/raw/main/eng.traineddata"
-wget -O "$TESSDATA_DIR/osd.traineddata" "https://github.com/tesseract-ocr/tessdata/raw/main/osd.traineddata"
-
-# Set and verify TESSDATA_PREFIX
-export TESSDATA_PREFIX="$TESSDATA_DIR"
-echo "TESSDATA_PREFIX=${TESSDATA_PREFIX}" >> /etc/environment
-
-# Verify tesseract installation and data files
-echo "Verifying Tesseract installation..."
-if ! command -v tesseract &> /dev/null; then
-    echo "Tesseract installation failed!"
-    exit 1
-fi
-echo "Tesseract version: $(tesseract --version)"
-
-# Verify traineddata files
-echo "Verifying traineddata files..."
-if [ ! -f "$TESSDATA_DIR/eng.traineddata" ]; then
-    echo "eng.traineddata is missing!"
-    exit 1
-fi
-if [ ! -f "$TESSDATA_DIR/osd.traineddata" ]; then
-    echo "osd.traineddata is missing!"
-    exit 1
-fi
-
-echo "Traineddata files in $TESSDATA_DIR:"
-ls -l "$TESSDATA_DIR"
-
-# Test Tesseract functionality
-echo "Testing Tesseract functionality..."
-echo "Hello World" > test.png
-if ! tesseract test.png stdout; then
-    echo "Tesseract test failed!"
-    exit 1
-fi
-rm test.png
-
-# Clean and install tesserocr from source
-echo "Installing tesserocr from source..."
-pip uninstall -y tesserocr || true
-CPPFLAGS="-I/usr/include/tesseract" LDFLAGS="-L/usr/lib/x86_64-linux-gnu/" pip install --no-binary :all: tesserocr
-
-# Verify tesserocr installation
-echo "Verifying tesserocr installation..."
-python3 -c "
-import tesserocr
-print(f'tesserocr version: {tesserocr.__version__}')
-print(f'Available languages: {tesserocr.get_languages()}')
-print(f'TESSDATA_PREFIX: {tesserocr.get_languages()[1]}')
-"
+    wget \
+    pkg-config
 
 # Install Google Gemini API client
 echo "Installing Google Gemini API client..."

requirements.txt

@@ -1,35 +1,24 @@
 # Core dependencies
 gradio==5.14.0
-grpcio-status==1.70.0
 markdown==3.7
-multiprocess==0.70.16
-pipdeptree==2.25.0
 Pillow>=9.0.0,<11.0.0
 numpy>=1.21.0
 
-# PDF processing
-pdf2image>=1.16.0
-
-# OCR dependencies (for GOT-OCR)
-pytesseract==0.3.13
-tesseract==0.1.3
-tesserocr>=2.5.0; platform_system != "Windows"  # Only install on non-Windows systems
+# Image processing
 opencv-python-headless>=4.5.0  # Headless version for server environments
 
 # Utility dependencies
-dill==0.3.8  # Downgraded to be compatible with datasets
 python-dotenv>=1.0.0
 pydantic==2.7.1
 
 # Gemini API client
 google-genai>=0.1.0
 
-# GOT-OCR dependencies
-torch>=2.0.1
-torchvision>=0.15.2
+# GOT-OCR dependencies (as specified in documentation)
+torch==2.0.1
+torchvision==0.15.2
 transformers==4.37.2  # Pin to a specific version that works with safetensors 0.4.3
-tiktoken>=0.6.0
-verovio>=4.3.1
-accelerate>=0.28.0
-safetensors==0.4.3  # Updated to meet minimum version required by accelerate
-packaging>=21.0  # For version comparison
+tiktoken==0.6.0
+verovio==4.3.1
+accelerate==0.28.0
+safetensors==0.4.3  # Updated to meet minimum version required by accelerate

setup.sh

@@ -3,18 +3,11 @@
 # Exit on error
 set -e
 
-echo "Setting up Tesseract OCR environment..."
-
-# Install required packages if not already installed
-if ! command -v tesseract &> /dev/null; then
-    echo "Tesseract not found, attempting to install..."
-    apt-get update -y || echo "Failed to update apt, continuing anyway"
-    apt-get install -y tesseract-ocr tesseract-ocr-eng libtesseract-dev libleptonica-dev || echo "Failed to install tesseract via apt, continuing anyway"
-fi
+echo "Setting up environment..."
 
 # Install Python dependencies
 echo "Installing Python dependencies..."
-pip install -q -U pytesseract pillow opencv-python-headless pdf2image
+pip install -q -U pillow opencv-python-headless
 pip install -q -U google-genai
 echo "Python dependencies installed successfully"
 
@@ -23,54 +16,4 @@ echo "Installing GOT-OCR dependencies..."
 pip install -q -U torch==2.0.1 torchvision==0.15.2 transformers==4.37.2 tiktoken==0.6.0 verovio==4.3.1 accelerate==0.28.0 safetensors==0.4.3
 echo "GOT-OCR dependencies installed successfully"
 
-# Install tesserocr with pip
-echo "Installing tesserocr..."
-pip install -q -U tesserocr || echo "Failed to install tesserocr with pip, trying with specific compiler flags..."
-
-# If tesserocr installation failed, try with specific compiler flags
-if ! python -c "import tesserocr" &> /dev/null; then
-    echo "Trying alternative tesserocr installation..."
-    CPPFLAGS="-I/usr/local/include -I/usr/include" LDFLAGS="-L/usr/local/lib -L/usr/lib" pip install -q -U tesserocr || echo "Failed to install tesserocr with compiler flags, continuing anyway"
-fi
-
-# Create tessdata directory if it doesn't exist
-mkdir -p tessdata
-
-# Set TESSDATA_PREFIX environment variable
-export TESSDATA_PREFIX="$(pwd)/tessdata"
-echo "TESSDATA_PREFIX set to: $TESSDATA_PREFIX"
-
-# Download eng.traineddata if it doesn't exist
-if [ ! -f "tessdata/eng.traineddata" ]; then
-  echo "Downloading eng.traineddata..."
-  wget -O tessdata/eng.traineddata https://github.com/tesseract-ocr/tessdata/raw/main/eng.traineddata || \
-  curl -o tessdata/eng.traineddata https://github.com/tesseract-ocr/tessdata/raw/main/eng.traineddata
-  echo "Downloaded eng.traineddata"
-else
-  echo "eng.traineddata already exists"
-fi
-
-# Try to copy to system locations (may fail in restricted environments)
-for tessdata_dir in "/usr/share/tesseract-ocr/4.00/tessdata" "/usr/share/tesseract-ocr/tessdata" "/usr/local/share/tessdata"; do
-  if [ -d "$tessdata_dir" ]; then
-    echo "Copying eng.traineddata to $tessdata_dir..."
-    cp -f tessdata/eng.traineddata "$tessdata_dir/" 2>/dev/null || echo "Failed to copy to $tessdata_dir, continuing anyway"
-  fi
-done
-
-# Verify Tesseract installation
-echo "Verifying Tesseract installation..."
-tesseract --version || echo "Tesseract not found in PATH, but may still be available to Python"
-
-# Test tesserocr if installed
-echo "Testing tesserocr..."
-python -c "import tesserocr; print(f'tesserocr version: {tesserocr.tesseract_version()}')" || echo "tesserocr not working, but may still be able to use pytesseract"
-
-# Test pytesseract
-echo "Testing pytesseract..."
-python -c "import pytesseract; print(f'pytesseract path: {pytesseract.tesseract_cmd}')" || echo "pytesseract not working"
-
-echo "Setup completed"
-
-# Add TESSDATA_PREFIX to .env file for persistence
-echo "TESSDATA_PREFIX=$(pwd)/tessdata" >> .env 
+echo "Setup completed"