Update Dockerfile to copy PDF file first and adjust PYTHONPATH; modify initialize_rag to support new PDF path structure

Dockerfile

@@ -7,15 +7,18 @@ ENV PATH="/home/user/.local/bin:$PATH"
 
 WORKDIR /app
 
-# Copy only the midterm project files
+# Copy requirements and install dependencies
 COPY --chown=user requirements.txt requirements.txt
 RUN pip install --no-cache-dir --upgrade -r requirements.txt
 
-# Copy the app directory
+# Copy the PDF file first
+COPY --chown=user Grid_Code.pdf Grid_Code.pdf
+
+# Copy the rest of the application
 COPY --chown=user . .
 
-# Set environment variable for Python path
-ENV PYTHONPATH=/app
+# Set environment variables for Python path
+ENV PYTHONPATH="${PYTHONPATH}:/app:/app/src"
 
 # Run streamlit on port 7860 for Hugging Face Spaces
 CMD ["streamlit", "run", "app.py", "--server.port", "7860", "--server.address", "0.0.0.0"]

app.py

@@ -112,19 +112,16 @@ def initialize_rag():
 
     # Try multiple possible paths for the PDF
     possible_paths = [
-        "app/data/raw/grid_code.pdf",  # Local path
-        "/app/app/data/raw/grid_code.pdf",  # Docker path
-        Path(__file__).parent
-        / "app"
-        / "data"
-        / "raw"
-        / "grid_code.pdf",  # Absolute path
+        "Grid_Code.pdf",  # Base directory (local and Docker)
+        "/app/Grid_Code.pdf",  # Docker container path
+        Path(__file__).parent / "Grid_Code.pdf",  # Absolute path
     ]
 
     data_path = None
     for path in possible_paths:
         if isinstance(path, str):
             path = Path(path)
+        logger.info(f"Checking path: {path}")
         if path.exists():
             data_path = str(path)
             logger.info(f"Found PDF at: {data_path}")