Skip to content

PDF Output

NormaDocs can produce a PDF in addition to the DOCX. The DOCX is always written; the PDF is opt-in via the --format flag.

Usage

# DOCX only (default)
normadocs paper.md

# DOCX + PDF
normadocs paper.md --format all

# PDF only
normadocs paper.md --format pdf

The PDF appears in the same output directory as the DOCX, with the same prefix and a .pdf extension.

Engines

NormaDocs supports two PDF engines. Pick one and install the corresponding system or Python dependency.

LibreOffice is the recommended path. It faithfully renders the DOCX the formatter just produced, so the PDF matches the DOCX exactly.

Install:

# Debian / Ubuntu
sudo apt install libreoffice-writer

# macOS
brew install --cask libreoffice

# Windows: download from libreoffice.org

Verify:

soffice --version

WeasyPrint (Python-only alternative)

WeasyPrint is a pure-Python PDF engine. It does not need LibreOffice, but the rendered PDF may differ slightly from the DOCX because the conversion path is different (DOCX → HTML → PDF instead of DOCX → PDF via headless Word).

Install:

pip install 'normadocs[pdf]'

Choosing an engine

Engine Pros Cons
LibreOffice Exact DOCX-to-PDF match. Requires ~300 MB system install.
WeasyPrint Python-only, no system deps. PDF may differ from DOCX.

If both are installed, NormaDocs uses LibreOffice. The choice is automatic based on which binary is on PATH.

Common failure modes

  • soffice: command not found — LibreOffice is not installed or not on PATH. Install it (see above) or use WeasyPrint.
  • PDF generation hangs — usually a LibreOffice profile lock. Kill any stale soffice processes and retry.
  • PDF is empty — the DOCX conversion succeeded but the DOCX itself is empty. Check the DOCX first.
  • ImportError: weasyprint — install via pip install 'normadocs[pdf]' and re-run.

See Troubleshooting for more.

Disabling PDF

If you only need DOCX (the default), pass --format docx explicitly, or omit --format entirely. PDF generation is opt-in.