FROM python:3.12-slim

# Install system dependencies for building llama-cpp-python
RUN apt-get update && apt-get install -y \
    build-essential \
    python3-dev \
    cmake \
    && rm -rf /var/lib/apt/lists/*

# Set working directory
WORKDIR /app

# Install required Python libraries
RUN pip install --no-cache-dir \
    fastapi \
    uvicorn \
    pydantic \
    huggingface_hub==0.35.3 \
    pandas==2.2.2 \
    tiktoken==0.12.0 \
    pymupdf==1.26.5 \
    langchain==0.3.27 \
    langchain-community==0.3.31 \
    chromadb==1.1.1 \
    sentence-transformers==5.1.1 \
    llama-cpp-python==0.2.28 \
    'numpy<2.1.0'

# Copy the application logic and FastAPI server
COPY app_logic.py main.py ./

# Copy the persistent vector database directory
COPY chroma_db/ ./chroma_db/

# Create the directory structure for the model to match main.py's MODEL_PATH
RUN mkdir -p /root/.cache/huggingface/hub/models--TheBloke--Mistral-7B-Instruct-v0.1-GGUF/snapshots/731a9fc8f06f5f5e2db8a0cf9d256197eb6e05d1/

# Copy the quantized model file to the snapshot directory
COPY mistral-7b-instruct-v0.1.Q4_K_M.gguf /root/.cache/huggingface/hub/models--TheBloke--Mistral-7B-Instruct-v0.1-GGUF/snapshots/731a9fc8f06f5f5e2db8a0cf9d256197eb6e05d1/

# Expose port 8000 for FastAPI
EXPOSE 8000

# Command to start the FastAPI application
CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "8000"]