searchanddisplace-docker/docker/ingest/Dockerfile

FROM rcarjan/nginx-php:7.4

LABEL maintainer="Radu Liviu Carjan"

## Add required files
RUN mkdir /var/www/dewarp
ADD dewarp /var/www/dewarp

## Install libreoffice
RUN apt-add-repository -y ppa:libreoffice/ppa && \
    apt-get install -y \
    libreoffice \
    software-properties-common

# Install python & popple PDF convertor
RUN add-apt-repository -y ppa:deadsnakes/ppa && \
    apt-get install -y \
    supervisor \
    python3.8 \
    python3.8-dev \
    python3.8-distutils \
    libpoppler-cpp-dev \
    poppler-utils

## Install Tesseract OCR, Pandoc & other dependencies
RUN add-apt-repository -y ppa:alex-p/tesseract-ocr-devel && \
    apt-get install -y \
    tesseract-ocr \
    unpaper \
    unoconv \
    pandoc

## Configure correct python version, install PIP
RUN rm /usr/bin/python3 && \
    ln -s /usr/bin/python3.8 /usr/bin/python3 && \
    apt-get install -y python-is-python3 && \
    curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py && \
    python get-pip.py && \
    rm -rf get-pip.py && \
    pip install --upgrade pip

## Install PIP packages
RUN pip install \
    pdftotext \
    supervisor \
    opencv-python
    
WORKDIR /var/www/dewarp
RUN pip install -r requirements.txt

RUN mkdir /var/log/queue

WORKDIR /var/www/ingest