FROM crpi-vofi3w62lkohhxsp.cn-shanghai.personal.cr.aliyuncs.com/opendatalab-mineru/corex:4.4.0_torch2.7.1_vllm0.11.2_py3.10

ENV DEBIAN_FRONTEND=noninteractive
ENV MINERU_MODEL_SOURCE=local
ENV MINERU_MODEL_DIR=/models
ENV MINERU_TOOLS_CONFIG_JSON=/root/mineru.json

RUN apt-get update && apt-get install -y \
        fonts-noto-core \
        fonts-noto-cjk \
        fontconfig \
        libgl1-mesa-glx \
    && fc-cache -fv \
    && apt-get clean && rm -rf /var/lib/apt/lists/*

RUN python3 -m pip install -U pip -i https://mirrors.aliyun.com/pypi/simple && \
    python3 -m pip install 'mineru[core]>=2.7.4' --no-deps -i https://mirrors.aliyun.com/pypi/simple && \
    python3 -m pip install \
        "numpy==1.26.4" \
        "opencv-python==4.11.0.86" \
        "pydantic>=2.12,<3" \
        "modelscope" \
        "magic-pdf" \
        boto3 \
        "pdfminer.six>=20251230" \
        pypdfium2 \
        pypdf \
        reportlab \
        pdftext \
        huggingface-hub \
        json-repair \
        fast-langdetect \
        scikit-image \
        openai \
        beautifulsoup4 \
        magika \
        mineru-vl-utils \
        qwen-vl-utils \
        matplotlib \
        ultralytics \
        doclayout_yolo==0.0.4 \
        dill \
        PyYAML \
        ftfy \
        shapely \
        pyclipper \
        omegaconf \
        onnxruntime \
        python-pptx \
        python-docx \
        openpyxl \
        xlrd \
        xlwt \
        pylatexenc \
        mammoth \
        -i https://mirrors.aliyun.com/pypi/simple && \
    python3 -m pip cache purge

WORKDIR /app
COPY . .
COPY docker/entrypoint.mineru-api.sh /usr/local/bin/entrypoint.mineru-api.sh
RUN chmod +x /usr/local/bin/entrypoint.mineru-api.sh

ENTRYPOINT ["/usr/local/bin/entrypoint.mineru-api.sh"]
CMD ["mineru-api", "--host", "0.0.0.0", "--port", "8000"]