pandoc-select-code

Pandoc filter to extract only the code blocks. You might use this, for example, to extract sample code from a tutorial.

I originally developed this filter because I wanted something like literate Haskell, but using Markdown. I wanted to be able to write a Markdown file with Haskell code blocks, and then be able to compile and execute it, or use Pandoc to produce a PDF. Cabal doesn't support that, so one workaround is to pre-process the file using Pandoc with this filter and the --write=plain option to produce a source file that can be compiled.

I also use this when I'm updating a tutorial I wrote some time ago; I'll extract the code examples and verify that they still work with the latest software.

Installation

This package is available from Hackage, or as a Nix flake.

From Hackage

To install from Hackage, use cabal install. The package name is pandoc-select-code.

As a Nix flake

Note: Flakes must be enabled in your Nix or NixOS installation.

One way to use the Nix flake is to create a shell.nix with pandoc and this package, like so:

with (import <nixpkgs> {});
let
  pandoc-include-plus = (builtins.getFlake git+https://codeberg.org/mhwombat/pandoc-include-plus).packages.${builtins.currentSystem}.default;
in
mkShell {
  buildInputs = [
    pandoc
    pandoc-select-code
    # add any other software you want to use in the shell.
  ];
}

Enter the shell using nix-shell. Now you can use the commands below.

Transforming to other formats

You can process your Markdown document either using pandoc directly, or through Hakyll.

With Pandoc

Use this filter by adding --filter=pandoc-select-code to your pandoc command. For example:

pandoc --filter=pandoc-linear-table myfile.md --output=myfile.pdf

With Hakyll

Use this filter as a transform in Hakyll. For example, you could modify site.hs, adding

import Text.Pandoc.Filters.SelectCode (transform)

and changing

pandocCompiler

to

pandocCompilerWithTransform defaultHakyllReaderOptions defaultHakyllWriterOptions transform