Word and PowerPoint to PDF, and PDF to images. In bulk, headless, with one pipe.
3.8K
Documents in. PDFs and images out.
bsantanna/compute-document-utils turns stacks of Word and PowerPoint files into PDFs, and PDFs into page images — in bulk, headless, with one pipe. Powered by LibreOffice and Poppler.
Convert every .docx and .pptx in a folder to PDF:
find . -name '*.docx' -o -name '*.pptx' | \
docker run -i --rm -v "$PWD:/work" -w /work bsantanna/compute-document-utils convert_to_pdf
Turn every PDF into one JPEG per page:
find . -name '*.pdf' | \
docker run -i --rm -v "$PWD:/work" -w /work bsantanna/compute-document-utils extract_images
Output lands next to each source file: report.docx becomes report.pdf, and report.pdf becomes report-page-1.jpg, report-page-2.jpg, and so on.
| Command | Reads from stdin | Produces |
|---|---|---|
convert_to_pdf | Paths to .docx or .pptx files | A PDF beside each file |
extract_images | Paths to .pdf files | One 72 DPI JPEG per page beside each file |
Unsupported or missing files are skipped and reported, so one bad path never stops the batch.
| Conversion | LibreOffice Writer and Impress, headless |
| Rendering | Poppler (pdftoppm, pdftotext and friends) |
| Also included | Python 3.12, Java 25 and the tools from bsantanna/java-python-dev |
| Tag | Meaning |
|---|---|
latest | The newest release. |
YYYY.MM.XX | A fixed release, e.g. 2026.11.01. Pin this in pipelines. |
Multi-architecture: linux/amd64 and linux/arm64.
Content type
Image
Digest
sha256:79ae1845a…
Size
586.5 MB
Last updated
4 days ago
docker pull bsantanna/compute-document-utils