There is no common DOM to manipulate or generate documents, this library is mainly for conversion. However, the Docx package provides some helper methods on top of the Open XML SDK that may be extended in the future.
If your main purpose is creating documents from scratch you can consider the following libraries: OfficeIMO, OpenXML-Office, ClosedXML, ShapeCrawler, QuestPDF.

Supported features

• Binary formats: almost all doc/xls/ppt features were supported by the original project, but exceptions occurred when using .NET (rather than .NET Framework) or loading specific documents/encodings. Most errors should be fixed now but more work is needed to make the library reliable; if you find other bugs, you are welcome to open an issue (please attach a sample file if the issue only occurs for specific documents).
• DOCX to RTF:
  • Font formatting, paragraphs, tables and lists
    • Not all properties are supported, e.g. advanced positioning and conditional formatting for tables or some list types are not recognized.
  • Images:
    • JPEG, PNG and EMF are supported.
    • Only inline images are supported (wrap layouts are not yet implemented).
  • Hyperlinks and bookmarks
  • Page setup: size, orientation, margins, borders, background color
  • Header and footer
  • Drop caps
  • Fields (partial) and page numbers
  • TODO: footnotes/endnotes, math formulas, drawings, OLE objects and more
• DOCX to Markdown:
  • Text and basic formatting
    • Bold, italic, underline, strikethrough, superscript, subscript
    • Any highlight color is converted to <mark>
  • Inline images
    • ImagesOutputFolder needs to be set to an existing directory, otherwise images are skipped. An absolute URI is used by default; to produce a relative URI set ImagesBaseUriOverride to any not-null folder path (empty string or "." means same folder as the Markdown file, "../images" means images subfolder in the parent folder).
    • Some image types are not recognized (e.g. WordPad embeds images in a different way compared to MS Word and other word processors).
    • Images should be in JPEG, PNG or GIF format to be supported by browsers; BMP is partially supported but not recommended. There is currently no automatic image conversion implemented.
    • Crop and effects are not supported.
  • Tables (values only)
  • External hyperlinks
  • Page breaks are converted to horizontal lines
  • TODO: H1-H6 headers (Word styles), bookmarks (internal hyperlinks), lists, math formulas, charts
• Markdown to DOCX:
  • Basic Markdown features
  • External hyperlinks
  • Bookmarks for internal hyperlinks to headings (GitHub-like auto-identifiers)
  • Images
    • The converter attempts to read local images and download online images (http/https URLs only). If this behavior is not desired, set SkipImages to true.
    • Images specified as absolute URLs are processed by default. For relative URLs ImagesBaseUri needs to be set to an absolute local directory path or http(s) URL, which will be combined with the image URL at runtime, such as: C:\Data + ./images/image1.jpg (all kind of URIs should be recognized).
    • WEBP and AVIF images are ignored as they are not supported in DOCX documents; base64 is also ignored as it is rarely used and not supported by many Markdown processors.
  • Tables (experimental)
  • TODO: other internal hyperlinks types, HTML tags (<u>, <sup>, <sub>, <mark>, ...), math and other extensions

Roadmap

• Support more elements and attributes, and fix issues on edge cases
• Reverse RTF to DOCX conversion
• Documentation: for now you can refer to the sample app. When ready, any documentation will be available in the project Wiki.

Credits

Dependencies:

• Open XML SDK
• Markdig - for DocSharp.Markdown only

Forked:

• b2xtranslator
• markdig.docx

Others:

• Html2OpenXml for images header decoding and unit conversions.

License

DocSharp is licensed under MIT license and can be used for both open source and commercial projects.
If you find the library useful, adding a star is highly appreciated.