A couple of years ago, I realised that actually quite a few people were using my PySLHA library and plotter, and that I should write it up for them to cite, that being the tail-wags-dog way that the academic world rolls. So I knocked something together.
While writing this, using a LaTeX class file of my own devising, I decided I wanted to render my Python code examples better than the venerable listings package can do. And I found minted, a clever LaTeX package which automatically runs Pygments via the LaTeX chell escape mechanism. Problem is, the arXiv doesn't allow -shell-escape running of LaTeX; I had to beg a favour to get my original version of the paper uploaded.
Now I'm coming up to a major new release of PySLHA, it seems worth updating that arXiv note, and maybe even trying to get it "properly" published for the usual ineffable reasons. And another minted special request isn't going to wash. But I still like its output. So I just figured out what it was doing, and fiddled together a teeny bash script that provides the same code snippets statically. I don't think this exists elsewhere, but it's not worth a proper code release, so here's the while thing in case someone finds it useful:
#! /usr/bin/env bash ## Write a .sty file defining the commands used in each Verbatim code block (bit hacky) echo "" | pygmentize -l python -f latex -P full=True | head -n -10 | grep -E -v "documentclass|inputenc" > pygtex.sty ## Make a Verbatim code block for each input code file, transforming foo.ext to foo-ext.tex for inname in $@; do outname=$(echo "$inname" | sed -e 's/[\ \.]/-/g').tex pygmentize -f latex -P verboptions='frame=leftline,framesep=1.5ex,framerule=0.8pt,fontsize=\smaller' $inname > $outname done
I called it pygtex; you can call it whatever you like. It can be called like
pygtex *snippet.py (if you've made code snippets with that name pattern)
and will write out a pygtex.sty file, and a .tex file for each snippet. Then
include them in your doc like this: