Update Dockerfile and requirements for better installation of magic-pdf

Dockerfile

@@ -1,120 +1,54 @@
-FROM nvidia/cuda:12.1.0-base-ubuntu22.04
-
-# Set environment variables
-ENV DEBIAN_FRONTEND=noninteractive
-ENV PYTHONUNBUFFERED=1
+FROM python:3.10-slim
 
 # Install system dependencies
-RUN apt-get update && apt-get install -y \
-    python3 \
-    python3-pip \
-    python3-venv \
-    python3-dev \
-    wget \
-    git \
+RUN apt-get update && \
+    apt-get install -y --no-install-recommends \
     build-essential \
+    curl \
     libgl1-mesa-glx \
     libglib2.0-0 \
-    imagemagick \
-    ghostscript \
-    poppler-utils \
-    libmagickwand-dev \
-    fonts-freefont-ttf \
-    ffmpeg \
-    libsm6 \
-    libxext6 \
-    libxrender-dev \
-    pkg-config \
-    libcairo2-dev \
-    sudo \
+    wget \
+    git \
     && rm -rf /var/lib/apt/lists/*
 
-# Configure ImageMagick policy to allow PDF conversion (needed for sample PDF creation)
-RUN if [ -f "/etc/ImageMagick-6/policy.xml" ]; then \
-        sed -i 's/rights="none" pattern="PDF"/rights="read|write" pattern="PDF"/g' /etc/ImageMagick-6/policy.xml; \
-    fi
-
-# Create a virtual environment
-RUN python3 -m venv /opt/mineru_venv
-ENV PATH="/opt/mineru_venv/bin:$PATH"
-
-# Upgrade pip in the virtual environment
-RUN pip install --upgrade pip
-
-# Clone the MinerU repository
-RUN git clone https://github.com/opendatalab/MinerU.git /tmp/MinerU
-
-# Install required packages
-RUN pip install --no-cache-dir torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu
-
-# Install MinerU with all features
-WORKDIR /tmp/MinerU
-RUN pip install --no-cache-dir -e ".[full]"
-
-# Install additional dependencies for the web application
-RUN pip install --no-cache-dir flask==2.3.3 flask-cors==4.0.0 werkzeug==2.3.7
-
-# Create a non-root user for Hugging Face Spaces
-RUN useradd -m -u 1000 user && \
-    echo "user ALL=(ALL) NOPASSWD:ALL" > /etc/sudoers.d/user && \
-    chmod 0440 /etc/sudoers.d/user
-
-# Create directories with proper permissions
-RUN mkdir -p /tmp/pdf_uploads /tmp/pdf_output && \
-    chmod -R 777 /tmp/pdf_uploads && \
-    chmod -R 777 /tmp/pdf_output && \
-    mkdir -p /tmp/samples && \
-    chmod -R 777 /tmp/samples
-
-# Create models directory structure with proper permissions
-RUN mkdir -p /tmp/models/MFD/YOLO && \
-    mkdir -p /tmp/models/MFR/unimernet && \
-    mkdir -p /tmp/models/table/rapid && \
-    mkdir -p /tmp/models/layout/doclayout && \
-    chmod -R 777 /tmp/models
+# Set working directory
+WORKDIR /app
 
-# Download model weights
-WORKDIR /tmp/models
+# Set environment variables
+ENV PYTHONUNBUFFERED=1 \
+    PYTHONDONTWRITEBYTECODE=1 \
+    UPLOAD_FOLDER=/tmp/pdf_uploads \
+    OUTPUT_FOLDER=/tmp/pdf_output
+
+# Create necessary directories and set permissions
+RUN mkdir -p /tmp/pdf_uploads /tmp/pdf_output /tmp/models \
+    && chmod -R 777 /tmp/pdf_uploads /tmp/pdf_output /tmp/models
+
+# Copy requirements and app files
+COPY requirements.txt /app/
+COPY *.py /app/
+COPY entrypoint.sh /app/
+RUN chmod +x /app/entrypoint.sh
 
-# Download the YOLO model for formula detection (using curl with progress and retry)
-RUN apt-get update && apt-get install -y curl && \
-    curl -L --retry 5 --retry-delay 5 -o /tmp/models/MFD/YOLO/yolo_v8_ft.pt https://huggingface.co/opendatalab/MinerU/resolve/main/models/mfd/yolo_v8_mfd.pt && \
-    curl -L --retry 5 --retry-delay 5 -o /tmp/models/MFD/YOLO/yolo_v8_mfd.pt https://huggingface.co/opendatalab/MinerU/resolve/main/models/mfd/yolo_v8_mfd.pt && \
-    curl -L --retry 5 --retry-delay 5 -o /tmp/models/MFR/unimernet/unimernet_small.pth https://huggingface.co/opendatalab/MinerU/resolve/main/models/mfr/unimernet_small.pth && \
-    curl -L --retry 5 --retry-delay 5 -o /tmp/models/table/rapid/rapid_table.pt https://huggingface.co/opendatalab/MinerU/resolve/main/models/table/rapid_table.pt && \
-    curl -L --retry 5 --retry-delay 5 -o /tmp/models/table/rapid/slanet_plus.pt https://huggingface.co/opendatalab/MinerU/resolve/main/models/table/slanet_plus.pt && \
-    curl -L --retry 5 --retry-delay 5 -o /tmp/models/layout/doclayout/doclayout_yolo.pt https://huggingface.co/opendatalab/MinerU/resolve/main/models/layout/doclayout_yolo.pt && \
-    ls -la /tmp/models/MFD/YOLO/ && \
-    ls -la /tmp/models/MFR/unimernet/ && \
-    ls -la /tmp/models/table/rapid/ && \
-    ls -la /tmp/models/layout/doclayout/
+# Create a non-root user and change ownership
+RUN useradd -m -u 1000 appuser && \
+    chown -R appuser:appuser /app /tmp/pdf_uploads /tmp/pdf_output /tmp/models
 
-# Verify file sizes (files should not be empty)
-RUN find /tmp/models -type f -size 0 -exec echo "Warning: Empty file {}" \;
+# Install Python dependencies
+RUN pip install --no-cache-dir -r requirements.txt && \
+    pip install --no-cache-dir pymupdf
 
-# Create app directory
-RUN mkdir -p /app/samples && chown -R user:user /app
+# Switch to non-root user
+USER appuser
 
-# Copy the application files
-WORKDIR /app
-COPY . /app/
+# Install MinerU and PyMuPDF (fallback)
+RUN pip install --no-cache-dir --user minerupdf
 
-# Fix permissions for the user
-RUN chown -R user:user /app && \
-    mkdir -p /home/user/.config/magic_pdf && \
-    chown -R user:user /home/user/.config && \
-    chown -R user:user /tmp/pdf_uploads && \
-    chown -R user:user /tmp/pdf_output && \
-    chown -R user:user /tmp/models
+# Test the installation
+RUN ~/.local/bin/magic-pdf --version || echo "magic-pdf not working, will fallback to PyMuPDF"
 
-# Expose the port
+# Expose port
 EXPOSE 7860
 
-# Set up entrypoint
-RUN chmod +x /app/entrypoint.sh
-
-# Switch to non-root user for running the app
-USER user
-
-# Start the application
-CMD ["/app/entrypoint.sh"] 
+# Set the entrypoint
+ENTRYPOINT ["/app/entrypoint.sh"] 

requirements.txt

@@ -1,10 +1,8 @@
 flask==2.3.3
-werkzeug==2.3.7
 flask-cors==4.0.0
-requests>=2.31.0
-pillow>=9.4.0
-numpy>=1.24.0
-wget>=3.2
-magic-pdf[full]>=1.3.0
-uuid>=1.30
-python-magic>=0.4.27 
+werkzeug==2.3.7
+Pillow>=9.0.0
+numpy>=1.20.0
+requests>=2.25.0
+pymupdf>=1.20.0
+minerupdf