mathml-to-omml 1.0.3

Convert MathML to OMML (Office Math Markup Language) in pure Python

math_ml2omml

A pure Python library to convert MathML strings into OMML (Office Math Markup Language).

This is a modern, patched version of the original mathml2omml library, containing key bug fixes and enhancements developed for the w2w (Word-to-Word Math Compiler) project to ensure 100% parity with Microsoft Word's professional equation layouts.

Key Enhancements in this Patched Version

• Accent and Group Character Rendering: Fixed tag mismatches on <m:groupChr> inside MUnder and MOver classes. This ensures mathematical accents (like hats, tildes, dots) and underscript/overscript groupings compile and render correctly in Microsoft Word.
• Improved Alignment and Layout Compatibility: Programmatic mapping fixes to match standard Office OpenXML (OOXML) math specifications.

Installation

You can install it directly from this package directory:

pip install .

Or for development (editable mode):

pip install -e .

Usage

import math_ml2omml

mathml = '<math><mi>x</mi><mo>+</mo><mi>y</mi></math>'
omml = math_ml2omml.convert(mathml)
print(omml)

Custom Entities

If your MathML contains custom entities, pass them as a dictionary:

import html.entities
mathml = '<math><msubsup><mo>&int;</mo><mn>1</mn><mi>x</mi></msubsup></math>'
omml = math_ml2omml.convert(mathml, html.entities.name2codepoint)

Running Tests

To verify the installation and run the suite of unit tests, execute the following command in the project root:

python -m unittest discover -s tests

Author

Patched and maintained by Robert Danaraj ([e-mail]).

Acknowledgements

Special thanks and deepest gratitude to George Kerscher (IMLS Profile), Chief Innovations Officer of the DAISY Consortium and President of the International Digital Publishing Forum. As a pioneering advocate for digital document accessibility (recognized as a White House Champion of Change), his invaluable discussion, guidance, and inspiration motivated the design, development, and deployment of this project.

We also thank the contributors of the daisy-math-a11y project for inspiring our work.

Keystroke Complexity Score (KCS)

This library includes a utility calculate_kcs() to compute the keyboard entry complexity cost of inputting mathematical equations into Microsoft Word's Equation Editor GUI. It is based on the Keystroke-Level Model (KLM) (Card, Moran, and Newell, 1980), adapted specifically for the interactive auto-build behaviors of Word's UnicodeMath input syntax:

• Raw Keystrokes ($K$): Character count of the linear notation.
• Cognitive Penalties ($C_i$): Additional weight for Shift keys (+1.0) and backslash escape sequences (+1.5).
• Auto-Build Spacebars ($C_{spacebar}$): Weight for spacebar triggers required to compile structures like fractions, scripts, roots, and matrices (+2.0).
• Fragility Penalty ($F$): Additional penalties (+10.0) for structures prone to user input errors without proper grouping.

Citations

• Keystroke-Level Model (KLM): Card, S. K., Moran, T. P., and Newell, A. (1980). The Keystroke-Level Model for User Performance Time with Interactive Systems. Communications of the ACM, 23(7), 396–410.
• UnicodeMath / Linear Math: Murray Sargent III. (2010). Unicode Sincerely, Math (Unicode Technical Note #28). Microsoft Corporation. UTN #28 Specification

License

MIT License. See LICENSE for details. Portions copyright (c) 2026 Daisy Project Contributors.