Uniword provides an HTML import pipeline that converts HTML content into OOXML document parts. This enables importing web content, rich text from web editors, and HTML-formatted data into Word documents.

1. Pipeline Overview

The HTML-to-OOXML conversion follows these steps:

  1. Parse HTML — The HTML input is parsed into a DOM tree

  2. Map elements — HTML elements are mapped to OOXML equivalents

  3. Convert styles — CSS inline styles and class-based styles are converted to OOXML formatting

  4. Handle images — Base64-encoded or linked images are embedded as document parts

  5. Build document — The converted elements are assembled into the document structure

2. Element Mapping

HTML elements map to OOXML elements as follows:

HTML OOXML

<p>

<w:p> (Paragraph)

<h1> - <h6>

<w:p> with heading styles

<strong>, <b>

<w:r> with <w:b> (Bold)

<em>, <i>

<w:r> with <w:i> (Italic)

<u>

<w:r> with <w:u> (Underline)

<ul>

<w:p> with bulleted list numbering

<ol>

<w:p> with numbered list numbering

<table>

<w:tbl> (Table)

<img>

<w:drawing> (Inline image)

<a>

<w:hyperlink> (Hyperlink)

<br>

<w:br> (Break)

3. Style Conversion

CSS properties are converted to OOXML formatting properties:

  • font-size maps to <w:sz> (half-points)

  • color maps to <w:color> (hex RGB)

  • font-family maps to <w:rFonts>

  • text-align maps to <w:jc> (justification)

  • background-color maps to <w:shd> (shading)

4. Usage 

# Import HTML into a document
doc = Uniword::Document.new
importer = Uniword::HtmlImporter.new(doc)
importer.import('<h1>Title</h1><p>Paragraph text</p>')
doc.save('output.docx')

5. Limitations

Not all HTML features have OOXML equivalents:

  • CSS positioning (position: absolute) is not directly supported

  • CSS flexbox/grid layouts have no OOXML counterpart

  • JavaScript-generated content cannot be imported

  • Complex nested lists may require post-processing