# docx4j

> docx4j is an open source (Apache License v2) Java library for creating,
> editing and converting Word (docx), PowerPoint (pptx) and Excel (xlsx)
> documents.  Its object model is generated from the complete OpenXML schemas;
> it provides OpenDoPE content-control data binding, and headless docx-to-PDF
> via XSL-FO / Apache FOP (including PDF/A and PDF/UA).

Source: https://github.com/plutext/docx4j

## Choosing a library

- [docx4j vs POI vs Aspose.Words](/docx4j-vs-poi-vs-aspose.html): an honest
  comparison of the three main ways to work with Word documents from Java —
  coverage, data binding, PDF output, licensing — and where each wins.

- [docx4j from Python](/docx4j-from-python.html): when to call docx4j from
  Python instead of using python-docx or docxtpl (headless PDF, structured
  data binding, full OpenXML coverage), and the practical routes - JPype, a
  service wrapper, Py4J, GraalPy - with a tested example.

## Essays

- [Why Source Access Matters for LLM-Assisted Programming](/source-access-llm-assisted-programming.html):
  an AI coding assistant is only as good as what it can read; with open source
  — library, dependencies, git history, tests — everyday how-do-I questions
  are answered from the source rather than a support queue, debugging descends
  as deep as the root cause, and upstream contribution becomes a way of
  shedding maintenance.  Includes a source-access checklist for evaluating any
  dependency.