Obsidian.md 👉 LaTeX

md_to_latex.py

#!/usr/bin/env python
#
# * Converts citations:
#	"[[Lastname2019a]]" 	into 	"\cite{Lastname2019a}"
#
# * Interprets code blocks with four back-ticks (````) as comments,
#   and those with three back-ticks (```) as "lstlisting" environments.
#
# * Images:
#      "![[Image Name.png]]"    into   "\figure (blablablabla)"
#   or
#      "![The caption for the image](image-name.png)"    into    "\caption{The caption for the image} etc blablabla"

import argparse
import os
import re
import subprocess
import sys

parser = argparse.ArgumentParser(description='Convert .md file to .tex')
parser.add_argument('filename', type=str,
                   help='Files to convert')
parser.add_argument('-c', '--copy', action='store_true',
                   help='Copy .tex to clipboard')

args = parser.parse_args()

if len(sys.argv) == 1:
    raise Exception('Need path to file')

with open(args.filename, 'r') as f:
    data = f.read()

# ------------------------------------------------------------------------------
# Markdown quirk replacement
# ------------------------------------------------------------------------------

# Images (note, does not support underscores in image filename)
figure_format = (r'\\begin{figure}\n'
		 r'  \\centering\n'
		 #r'  \\includegraphics[width=\\linewidth]{\1.\2}\n'
		 r'  \\includegraphics[width=\\textwidth]{\1.\2}\n'
		 r'  \\caption{FIXME: \1}\n'
		 r'  \\label{fig:\1}\n'
		 r'\\end{figure}\n')

data = re.sub(r'!\[\[(.+)\.(png|jpg)\]\]', figure_format, data)


# Images with: ![Caption text](image-path.png)
figure_format = (r'\\begin{figure}\n'
                 r'  \\centering\n'
                 #r'  \\includegraphics[width=\\linewidth]{\2.\3}\n'
                 r'  \\includegraphics[width=\\textwidth]{\2.\3}\n'
                 r'  \\caption{\1}\n'
                 r'  \\label{fig:\2}\n'
                 r'\\end{figure}\n')

data = re.sub(r'!\[(.+)]\((.+)\.(png|jpg)\)', figure_format, data)


# Citations
data = re.sub(r'\[\[([A-Z][a-zA-Z\-]+\d{4}\w?)\]\]', r'\\cite{\1}', data, flags=re.M)

# Any remaining wiki links
# Disable them:
data = re.sub(r'\[\[(.*?)\]\]', r'\\iffalse \1 \\fi', data, flags=re.DOTALL)
# Ignore the link but keep the text:
#data = re.sub(r'\[\[([^|]+)\|?(.*?)\]\]', r'\2', data, flags=re.DOTALL)  ## FIXME: needs to use \1 if there is no \2

# Comment blocks
data = re.sub(r'````(.*?)````', r'\\iffalse \1 \\fi\n', data, flags=re.DOTALL)

# Code blocks
data = re.sub(r'```(\w+)\n(.*?)```', r'\\begin{lstlisting}[language=\1]\n\2\n\\end{lstlisting}', data, flags=re.DOTALL)

# ------------------------------------------------------------------------------
# Convert to LaTeX and write to file + clipboard
# ------------------------------------------------------------------------------

out_filename = os.path.basename(sys.argv[1].replace('.md', '').replace(' ', '')) + '.tex'
print('Writing TeX file to {}'.format(out_filename))
process = subprocess.run('pandoc -f markdown -t latex -o Latex/{} << "HEREDOC"\n{}\nHEREDOC'.format(out_filename, data),
                         shell=True, check=True, stdout=subprocess.PIPE, universal_newlines=True)
# output = process.stdout

if args.copy:

    process = subprocess.run('cat Latex/{} | pbcopy'.format(out_filename),
                         shell=True, check=True, stdout=subprocess.PIPE, universal_newlines=True)
    print('Copied contents of Latex/{} to clipboard!'.format(out_filename))

To use run:

$ ./md_to_latex.py directory/My\ Notes\ File.md