PDF to text via CLI

Extract TXT, Markdown, or HTML with OpenDataLoader, then optionally convert HTML to DOCX with soffice.

Packages to install

The engine is a Java CLI plus a JRE 11+. For DOCX you also need LibreOffice Writer. A wrapper named opendataloader-pdf is optional.

default-jre-headless

libreoffice-writer-nogui

fonts-liberation

sudo apt-get update
sudo apt-get install -y default-jre-headless libreoffice-writer-nogui fonts-liberation

How to run the command

opendataloader-pdf document.pdf -o . -f html --image-output off writes document.html. -f text writes document.txt, -f markdown writes document.md. For Word: convert the HTML with soffice --headless --convert-to docx document.html.

opendataloader-pdf document.pdf -o . -f html --image-output off

Dockerfile example

Build a minimal Ubuntu image with the required packages and run the conversion inside the container.

FROM ubuntu:24.04

RUN apt-get update && \
    apt-get install -y --no-install-recommends \
        default-jre-headless \
        libreoffice-writer-nogui \
        fonts-liberation && \
    rm -rf /var/lib/apt/lists/*

WORKDIR /work

docker build -t pdf-to-text .
docker run --rm -v "$PWD:/work" -w /work pdf-to-text \
  opendataloader-pdf document.pdf -o . -f html --image-output off

Example commands

Run these locally after installing the packages, or inside the container from the Dockerfile.

opendataloader-pdf document.pdf -o . -f html --image-output off

opendataloader-pdf document.pdf -o . -f text --image-output off

java -jar opendataloader-pdf-cli.jar document.pdf -o . -f markdown --image-output off

opendataloader-pdf document.pdf -o ./output -f html --image-output off && soffice --headless --convert-to docx --outdir ./output document.html