Namespace DocWright.Formats.Docx

Classes

DocxReader

Reads Office Open XML WordprocessingML (.docx) into the mutable DocWright document model. The main document part is consumed in a single streaming pass (one paragraph or table row materialized at a time); styles, numbering, theme and settings are parsed from their small parts; everything DocWright does not model is preserved — unknown XML as islands on the nearest node, unknown package parts opaquely — so nothing is silently dropped.

DocxWriter

Writes the DocWright document model back to an Office Open XML WordprocessingML package.

This is the second consumer of the DOM, parallel to the style resolver and independent of it: the writer never touches resolved or layout types, and reads only what the reader put in the model. It writes what it reads — no field re-evaluation, no renumbering, no normalization — because fidelity to the source is worth more than tidy output.

Preserved content is re-emitted rather than reconstructed. Unknown-XML islands go back out verbatim at their schema position, and opaque package parts are streamed through byte for byte with the relationships that reach them.