Skip to content


Repository files navigation


Simple LaTeX parser providing latex-to-unicode and unicode-to-latex conversion

Python: ≥ 3.4 or ≥ 2.7. The library is designed to be as backwards-compatible as reasonably possible and is able to run on old python verisons should it be necessary. (Use the script directly if you have python<3.7, poetry doesn't seem to work with old python versions.)

NEW (4/2023): PYLATEXENC 3.0alpha is in pre-release on PyPI. See new features and major changes. The documentation is still incomplete, and the new APIs are still subject to changes. The code is meant to be as backwards compatible as is reasonably possible. Feel free to try it out & submit feedback!

Unicode Text to LaTeX code

The pylatexenc.latexencode module provides a function unicode_to_latex() which converts a unicode string into LaTeX text and escape sequences. It should recognize accented characters and most math symbols. A couple of switches allow you to alter how this function behaves.

You can also run latexencode in command-line to convert plain unicode text (from the standard input or from files given on the command line) into LaTeX code, written on to the standard output.

A third party plug-in for Vim vim-latexencode by @Konfekt provides a corresponding command to operate on a given range.

Parsing LaTeX code & converting to plain text (unicode)

The pylatexenc.latexwalker module provides a series of routines that parse the LaTeX structure of given LaTeX code and returns a logical structure of objects, which can then be used to produce output in another format such as plain text. This is not a replacement for a full (La)TeX engine, rather, this module provides a way to parse a chunk of LaTeX code as mark-up code.

The pylatexenc.latex2text module builds up on top of pylatexenc.latexwalker and provides functions to convert given LaTeX code to plain text with unicode characters.

You can also run latex2text in command-line to convert LaTeX input (either from the standard input, or from files given on the command line) into plain text written on the standard output.


Full documentation is available at

To build the documentation manually, run:

> poetry install --with=builddoc
> cd doc/
doc> poetry run make html


See LICENSE.txt (MIT License).

NOTE: See copyright notice and license information for file tools/unicode.xml provided in tools/unicode.xml.LICENSE. (The file tools/unicode.xml was downloaded from as linked from

Javascript Library

Some core parts of this library can be transcribed to JavaScript. This feature is used (and was developed for) my Flexible Latex-like Markup project. See the js-transcrypt/ folder and its README file.