Stirling-PDF/Dockerfile.fat

# Build the application
FROM gradle:8.14-jdk21 AS build

COPY build.gradle .
COPY settings.gradle .
COPY gradlew .
COPY gradle gradle/
COPY app/core/build.gradle core/.
COPY app/common/build.gradle common/.
COPY app/proprietary/build.gradle proprietary/.
RUN ./gradlew build -x spotlessApply -x spotlessCheck -x test -x sonarqube || return 0

# Set the working directory
WORKDIR /app

# Copy the entire project to the working directory
COPY . .

# Build the application with DISABLE_ADDITIONAL_FEATURES=false
ENV DISABLE_ADDITIONAL_FEATURES=false \
    STIRLING_PDF_DESKTOP_UI=false
RUN ./gradlew clean build -x spotlessApply -x spotlessCheck -x test -x sonarqube

# Main stage
FROM alpine:3.22.2@sha256:4b7ce07002c69e8f3d704a9c5d6fd3053be500b7f1c69fc0d80990c2ad8dd412

ADD https://download.documentfoundation.org/libreoffice/stable/25.2.6/deb/aarch64/LibreOffice_25.2.6_Linux_aarch64_deb.tar.gz /tmp/lo.tar.gz
RUN mkdir -p /opt/libreoffice && \
    tar -xzf /tmp/lo.tar.gz -C /opt/libreoffice --strip-components=1 && \
    rm -f /tmp/lo.tar.gz && \
    ln -sf /opt/libreoffice/program/soffice /usr/bin/soffice.bin && \
    chmod 755 /usr/bin/soffice.bin

# Copy necessary files
COPY scripts /scripts
COPY app/core/src/main/resources/static/fonts/*.ttf /usr/share/fonts/opentype/noto/
# first /app directory is for the build stage, second is for the final image
COPY --from=build /app/app/core/build/libs/*.jar app.jar

ARG VERSION_TAG

# Set Environment Variables
ENV DISABLE_ADDITIONAL_FEATURES=true \
    VERSION_TAG=$VERSION_TAG \
    JAVA_BASE_OPTS="-XX:+UnlockExperimentalVMOptions -XX:MaxRAMPercentage=75 -XX:InitiatingHeapOccupancyPercent=20 -XX:+G1PeriodicGCInvokesConcurrent -XX:G1PeriodicGCInterval=10000 -XX:+UseStringDeduplication -XX:G1PeriodicGCSystemLoadThreshold=70" \
    JAVA_CUSTOM_OPTS="" \
    HOME=/home/stirlingpdfuser \
    PUID=1000 \
    PGID=1000 \
    UMASK=022 \
    FAT_DOCKER=true \
    INSTALL_BOOK_AND_ADVANCED_HTML_OPS=false \
    PYTHONPATH=/opt/libreoffice/program:/opt/venv/lib/python3.12/site-packages \
    UNO_PATH=/opt/libreoffice/program \
    URE_BOOTSTRAP=file:///opt/libreoffice/program/fundamentalrc \
    PATH=$PATH:/opt/venv/bin \
    STIRLING_TEMPFILES_DIRECTORY=/tmp/stirling-pdf \
    TMPDIR=/tmp/stirling-pdf \
    TEMP=/tmp/stirling-pdf \
    TMP=/tmp/stirling-pdf

# JDK for app
RUN printf '%s\n' \
      'https://dl-3.alpinelinux.org/alpine/edge/main' \
      'https://dl-3.alpinelinux.org/alpine/edge/community' \
      'https://dl-3.alpinelinux.org/alpine/edge/testing' \
      > /etc/apk/repositories

# Bash + repositories
RUN apk add --no-cache bash && \
    ln -sf /bin/bash /bin/sh

RUN --mount=type=cache,target=/var/cache/apk \
    apk upgrade --no-cache -a

RUN --mount=type=cache,target=/var/cache/apk \
    apk add --no-cache \
    ca-certificates \
    tzdata \
    tini \
    bash \
    curl \
    shadow \
    su-exec \
    openssl \
    openssl-dev \
    openjdk21-jre

RUN --mount=type=cache,target=/var/cache/apk \
    apk add --no-cache \
    ffmpeg \
    # Doc conversion
    gcompat \
    libc6-compat \
    unpaper \
    poppler-utils \
    icu-libs \
    icu-dev \
    libstdc++

# RUN --mount=type=cache,target=/var/cache/apk \
#     apk add --no-cache \
#     libreoffice

RUN --mount=type=cache,target=/var/cache/apk \
    apk add --no-cache \
    # OCR MY PDF (unpaper for descew and other advanced featues)
    tesseract-ocr-data-eng \
    tesseract-ocr-data-chi_sim \
    tesseract-ocr-data-deu \
    tesseract-ocr-data-fra \
    tesseract-ocr-data-por

RUN --mount=type=cache,target=/var/cache/apk \
    apk add --no-cache \
    unpaper \
    font-terminus font-dejavu font-noto font-noto-cjk font-awesome font-noto-extra font-liberation font-linux-libertine font-urw-base35 \
    # CV / Python
    py3-opencv \
    python3 \
    ocrmypdf \
    py3-pip \
    py3-pillow \
    py3-pdf2image

RUN --mount=type=cache,target=/var/cache/apk \
    # Calibre (musl-native) + QtWebEngine Runtime
    apk add --no-cache calibre && \
    # Calibre fixes
    apk fix --no-cache calibre

RUN python3 -m venv /opt/venv && \
    /opt/venv/bin/pip install --no-cache-dir --upgrade pip setuptools && \
    /opt/venv/bin/pip install --no-cache-dir --upgrade unoserver weasyprint && \
    ln -s /opt/libreoffice/program/uno.py /opt/venv/lib/python3.12/site-packages/ && \
    ln -s /opt/libreoffice/program/unohelper.py /opt/venv/lib/python3.12/site-packages/ && \
    ln -s /opt/libreoffice/program /opt/venv/lib/python3.12/site-packages/LibreOffice && \
    mv /usr/share/tessdata /usr/share/tessdata-original && \
    mkdir -p $HOME /configs /logs /customFiles /pipeline/watchedFolders /pipeline/finishedFolders /tmp/stirling-pdf && \
    # Configure URW Base 35 fonts
    ln -s /usr/share/fontconfig/conf.avail/69-urw-*.conf /etc/fonts/conf.d/ && \
    fc-cache -f -v && \
    chmod +x /scripts/* && \
    # User permissions
    addgroup -S stirlingpdfgroup && adduser -S stirlingpdfuser -G stirlingpdfgroup && \
    chown -R stirlingpdfuser:stirlingpdfgroup \
        $HOME \
        /scripts \
        /usr/share/fonts/opentype/noto \
        /configs \
        /customFiles \
        /pipeline \
        /tmp/stirling-pdf \
        /app.jar && \
    # chmod 755 /usr/bin/soffice.bin && \
    ln -sf /bin/busybox /bin/sh

EXPOSE 8080/tcp
# Set user and run command
ENTRYPOINT ["tini", "--", "/scripts/init.sh"]
CMD ["sh", "-c", "java -Dfile.encoding=UTF-8 -Djava.io.tmpdir=/tmp/stirling-pdf -jar /app.jar & /opt/venv/bin/unoserver --port 2003 --interface 127.0.0.1"]