latex-unicode-math-mode

An Emacs minor mode for entering Unicode math symbols in LaTeX-mode,
with a sty file to make pdflatex Unicode-aware. This minor mode
automatically replaces inputs like \phi with ϕ and \in with ∈.
These replacements happen inside of math environments or everywhere,
depending on the configuration.

If you use pdflatex, you can add \usepackage{unicode-math-mode} to
your tex file in order to make pdflatex aware of the Unicode
characters. Use M-x latex-unicode-save-sty-file to put this file
somewhere where pdflatex will find it.

If you update this package, you may need to do this again to get the
latest sty file.

You can customize the keybindings and symbols with M-x customize-group latex-unicode-math.

Installation

From MELPA

The recommended way of installing is from MELPA.
Call M-x list-packages and look for latex-unicode-math-mode.
After installing, add this to your ~/.emacs file to enable the mode
automatically for all LaTeX files:

(require 'latex-unicode-math-mode)
;; Enable latex-unicode-math-mode automatically for all LaTeX files.
;; This converts LaTeX to Unicode inside math environments.
(add-hook 'LaTeX-mode-hook 'latex-unicode-math-mode)
;; Enable latex-unicode-mode automatically for all LaTeX files.
;; This converts LaTeX to Unicode everwhere, not only in math
;; environments.
;;(add-hook 'LaTeX-mode-hook 'latex-unicode-mode)

Note that latex-unicode-math-mode and latex-unicode-mode are
mutually exclusive. Enabling one will automatically disable the
other.

While latex-unicode-math-mode really makes sense only with AUCTeX,
latex-unicode-mode also works fine in other major modes, for example
in org-mode.

Tested with pdflatex from TeX Live 2016 and GNU Emacs 25.1.1.

Example

Say you type this:

\pi(v_1,\ldots,v_n) \in A \cup B

As you type, latex-unicode-math-mode turns it into this:

π(v_1,…,v_n) ∈ A ∪ B

There will be no change in the compiled pdf (don’t forget to add
\usepackage{unicode-math-mode} in the preamble).

If you later decide that you do do not like Unicode symbols after all,
you can easily revert everything back to LaTeX macros, see the next
section.

Additional Features

M-x latex-unicode-convert-region converts all LaTeX macros in the
active region to their Unicode equivalents.

M-x latex-unicode-convert-buffer is the same, but applies to the
whole buffer.

M-x latex-unicode-invert-region and M-x latex-unicode-invert-buffer revert all Unicode symbols in the current
region/buffer back to their LaTeX equivalents.

Comparison to the TeX input method

Emacs comes with an input method called TeX, enabled via M-x set-input-method TeX (see
https://github.com/emacs-mirror/emacs/blob/master/lisp/leim/quail/latin-ltx.el).
This input method is very similar to latex-unicode-math-mode. Some
pros and contras are as follows.

latex-unicode-math-mode:

• Pro:
  • Only active in math environments.
  • Has a .sty file to make pdflatex understand the Unicode symbols.
  • Has shorthand notation for common symbols such as != for \neq
    and ==​> for \Longrightarrow.
  • Can convert/invert regions and buffers.
  • Customizable (customization group latex-unicode-math).
• Contra:
  • Very new and likely to change. Expect rough edges and bugs.

The TeX input method:

• Pro:
  • Covers subscripts and superscripts.
  • Has auto-completion (thanks to quail).
• Contra:
  • All or nothing: Does not automatically deactivate itself outside
    of math environments.
  • Does not work with pdflatex: I do not know how to handle Unicode
    superscripts in pdflatex combined with the ' (prime) character,
    which is active in math mode.

Bugs

Due to the implementation, this minor mode will probably break your
input method. If you write in English or in another European
language, chances are you are not using one.