Skip to content

A python library for extracting equations, text, and images from .docx files

License

Notifications You must be signed in to change notification settings

hrushikeshrv/docxlatex

Repository files navigation

docxlatex

A python library for extracting text and images from .docx files with LaTeX support

Influenced by python-docx and python-docx2txt. This project aims to expand the above libraries' functionality by allowing you to extract the equations in a .docx file and converts them to valid LaTeX, primarily for use with MathJax.

docxlatex DOES NOT convert the entire .docx file to a LaTeX source file.

docxlatex is still in beta and does not work perfectly on all equations. However, it does work with reasonable accuracy on standard symbols and equations, and works reliably when the equations in the document have been converted to linear format.

Installation

Install using pip
pip install docxlatex

Usage

Import the Document class from docxlatex

from docxlatex import Document

Create a Document object, giving it the path to the .docx file, and call the get_text() method

doc = Document("path/to/document")
text = doc.get_text()

You can also set some settings on the Document object to customize behaviour, like setting the delimiters that will go before and after an equation, and flags to extract header and footer text. Read the documentation at hrushikeshrv.github.io/docxlatex for all options.

doc.inline_delimiter = "%" # "$" by default
doc.block_delimiter = "%%" # "$$" by default

Limitation

docxlatex currently requires all the mathematical equations in the .docx document to be converted into linear format to extract successfully. Make sure all the equations in the .docx file are converted into linear form.

To convert all your equations into linear format, click on any equation, go to the Equation tab, make sure LaTeX is selected, and click on Convert → All - Linear

Please report any bugs on this project's GitHub page docxlatex/issues