File Format Processing in Python

For Office files, openpyxl reads and writes Excel workbooks, while XlsxWriter only writes them but offers rich formatting and charts. python-docx and python-pptx create and edit Word documents and PowerPoint decks. For PDFs, pypdf is pure Python and handles splitting, merging, cropping and basic text extraction. PyMuPDF is much faster for rendering and extraction but is licensed under the AGPL or a commercial license. pdfminer.six gives detailed control over text layout, ReportLab generates PDFs from code, and WeasyPrint renders HTML and CSS to PDF.

To feed documents into language models, MarkItDown from Microsoft converts Office files, PDFs and more to Markdown, and Docling from IBM parses layout and tables into structured output. pypandoc wraps pandoc for conversions between many markup formats. For Markdown itself, markdown-it-py follows CommonMark closely, Python-Markdown is the long-standing extensible implementation, and Mistune is fast.

Python 3.11 added tomllib to the standard library for reading TOML (it does not write). PyYAML is the common YAML library; use safe_load for untrusted input. tablib exports tabular data to several formats, and pyelftools parses ELF binaries and DWARF debug information. When a format has both a reading and a writing library, check which side you need first: several popular packages here only read or only write, and choosing the wrong one usually means switching libraries halfway through a project.