The DOCX format is not a monolithic binary blob. It is a strict application of the Office Open XML (OOXML) standard. If you change a .docx extension to .zip and extract it, you will reveal the underlying architecture.
The Anatomy of a DOCX Archive
A typical valid DOCX structure contains:
[Content_Types].xml: The manifest defining MIME types for every part in the package._rels/: The relationships directory, defining how parts link together.word/document.xml: The primary payload. The structural body of the text.word/styles.xml: The stylesheet defining paragraph and run characteristics.word/media/: The directory containing embedded raster images and vectors.
Conversion to PDF: The Bottleneck
Converting DOCX to PDF requires a full rendering engine. Because DOCX relies on system fonts and dynamic pagination (calculated at runtime based on the printer driver and active margins), an external API or headless LibreOffice instance must calculate line breaks before writing the absolute coordinates to a PDF.