stirling-pdf-icp-brasil/Containerfile

FROM --platform=$BUILDPLATFORM docker.io/library/debian:stable-slim as debian-base
RUN apt update \
 && apt -y install curl ca-certificates libnss3-tools unzip git-core

FROM --platform=$BUILDPLATFORM debian-base as certdata.txt
RUN curl -sSL https://gist.githubusercontent.com/antoniordo/0038eb6830e4e4dafd413b61481a812b/raw/ac3dca8eeee718cb1c3398fbaad8e99ed01b4c1f/instalar-certificados-ac-icp-brasil.sh | sed -e 's,ACcompactado.zip,ACcompactadox.zip,g'| bash - \
 && truncate -s 0 /certdata.txt \
 && bash -c 'for file in /etc/ssl/certs/*.pem; do echo "Processando arquivo ${file}" ; name=$(basename $file | xargs basename -s .pem) ; openssl x509 -outform der -in ${file} -out - | nss-addbuiltin -n ${name} -t "C,C,C" >> /certdata.txt ; done'

FROM --platform=$BUILDPLATFORM debian-base as git-fetch
ARG TAG

RUN git clone --single-branch --branch v${TAG} https://github.com/Stirling-Tools/Stirling-PDF.git /stirling-pdf-src
COPY --from=certdata.txt /certdata.txt /stirling-pdf-src/app/core/src/main/resources/certdata.txt

FROM --platform=$BUILDPLATFORM docker.io/library/gradle:8.14-jdk21-ubi AS build
COPY --from=git-fetch /stirling-pdf-src /app
WORKDIR /app

RUN DISABLE_ADDITIONAL_FEATURES=false \
    STIRLING_PDF_DESKTOP_UI=false \
    ./gradlew clean build -x spotlessApply -x spotlessCheck -x test -x sonarqube

# From now on, almost directly from https://github.com/Stirling-Tools/Stirling-PDF/blob/main/Dockerfile.fat

# Main stage
FROM --platform=$TARGETPLATFORM docker.io/library/alpine:3
ARG TAG

# Copy necessary files
COPY --from=git-fetch /stirling-pdf-src/scripts /scripts
COPY --from=git-fetch /stirling-pdf-src/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

# Set Environment Variables
ENV DISABLE_ADDITIONAL_FEATURES=true \
    VERSION_TAG=v$TAG \
    JAVA_BASE_OPTS="-XX:+UnlockExperimentalVMOptions -XX:MaxRAMPercentage=75 -XX:InitiatingHeapOccupancyPercent=20 -XX:+G1PeriodicGCInvokesConcurrent -XX:G1PeriodicGCInterval=10000 -XX:+UseStringDeduplication -XX:G1PeriodicGCSystemLoadThreshold=70" \
    JAVA_CUSTOM_OPTS="" \
    HOME=/home/stirlingpdfuser \
    PUID=1000 \
    PGID=1000 \
    UMASK=022 \
    FAT_DOCKER=true \
    INSTALL_BOOK_AND_ADVANCED_HTML_OPS=false \
    PYTHONPATH=/usr/lib/libreoffice/program:/opt/venv/lib/python3.12/site-packages \
    UNO_PATH=/usr/lib/libreoffice/program \
    URE_BOOTSTRAP=file:///usr/lib/libreoffice/program/fundamentalrc \
    PATH=$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 echo "@main https://dl-cdn.alpinelinux.org/alpine/edge/main" | tee -a /etc/apk/repositories && \
    echo "@community https://dl-cdn.alpinelinux.org/alpine/edge/community" | tee -a /etc/apk/repositories && \
    echo "@testing https://dl-cdn.alpinelinux.org/alpine/edge/testing" | tee -a /etc/apk/repositories && \
    apk upgrade --no-cache -a && \
    apk add --no-cache \
    ca-certificates \
    tzdata \
    catatonit \
    bash \
    curl \
    shadow \
    su-exec \
    openssl \
    openssl-dev \
    openjdk21-jre \
    # Doc conversion
    gcompat \
    libc6-compat \
    libreoffice \
    # pdftohtml
    poppler-utils \
    # 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 \
    unpaper \
    font-terminus font-dejavu font-noto font-noto-cjk font-awesome font-noto-extra font-liberation font-linux-libertine font-urw-base35 \
    # CV
    py3-opencv \
    python3 \
    ocrmypdf \
    py3-pip \
    py3-pillow@testing \
    py3-pdf2image@testing && \
    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 /usr/lib/libreoffice/program/uno.py /opt/venv/lib/python3.12/site-packages/ && \
    ln -s /usr/lib/libreoffice/program/unohelper.py /opt/venv/lib/python3.12/site-packages/ && \
    ln -s /usr/lib/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 && \
    chown stirlingpdfuser:stirlingpdfgroup /app.jar

EXPOSE 8080/tcp
# Set user and run command
USER stirlingpdfuser
ENTRYPOINT ["/usr/bin/catatonit", "-g", "--", "/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"]