CODEWITHSUNDEEP
markdowncoqpandoc
OrenIshShalom
OrenIshShalom

Reputation: 7142

User defined language in pandoc code-block

I'm trying to write a code-block in Coq (long story :)) inside pandoc.

$ pandoc --list-highlight-languages | grep coq | wc -l
0

Since it doesn't exist, I wonder if there is an option to write my own user-defined code-block-language?

Upvotes: 1

Views: 1172

Answers (3)

OrenIshShalom
OrenIshShalom

Reputation: 7142

For completeness reasons, I am pasting a minimal xml syntax file for an html slideshow generated from a markdown file with pandoc. It is made as a skeleton for anyone writing a dsl and wants to easily mark specific words in a code block. See here many full syntax examples

<?xml version="1.0" encoding="UTF-8"?>
<language name="coq" extensions="*.v" indenter="haskell">
  <highlighting>
  <list name="kw_major">
    <item>induction</item>
    <item>Fixpoint</item>
    <item>Theorem</item>
    <item>Qed</item>
  </list>
  <list name="kw_minor">
    <item>exists</item>
    <item>forall</item>
    <item>nat</item>
  </list>
  <contexts>
    <context name="Normal" attribute="Normal" lineEndContext="#stay">
      <keyword    attribute="KwMajor" context="#stay" String="kw_major" />
      <keyword    attribute="KwMinor" context="#stay" String="kw_minor" />
      <Int        attribute="Decimal" context="#stay" />
      <DetectChar attribute="String"  context="String" char="&quot;" />
    </context>
    <context name="String" attribute="String" lineEndContext="StringSyntaxError" >
      <DetectChar attribute="String" context="#pop"  char="&quot;"       />
      <RegExpr    attribute="String" context="#stay" String="[^&quot;]*" />
    </context>
  </contexts>
  <itemDatas>
    <itemData name="KwMajor" defStyleNum="dsDataType"/>
    <itemData name="KwMinor" defStyleNum="dsKeyword"/>
    <itemData name="Decimal" defStyleNum="dsDecVal"/>
    <itemData name="String"  defStyleNum="dsString"/>
  </itemDatas>
  </highlighting>
  <general><keywords casesensitive="1" /></general>
</language>

Upvotes: 0

tarleb
tarleb

Reputation: 22609

The answer depends a little on the output format that's being targeted. The most general solution is to define a syntax definition. Pandoc can parse XML syntax definitions for the Kate editor, see https://docs.kde.org/stable5/en/kate/katepart/highlight.html for more info on the format. If you have such an XML file, then pass it to pandoc via the --syntax-definition command line option.

If the output format is LaTeX (or PDF via LaTeX), then using pandoc with the --listing option should work, esp. when combined with the tips mentioned here: https://tex.stackexchange.com/q/434523/

Likewise, minted can be used for highlighting, see the minted.lua filter.

Last but not least, using an external syntax highlighter would be possible via a filter. Lua filters are well suited for that task, as they do not require an additional tool-chain to be installed.

Upvotes: 2

Li-yao Xia
Li-yao Xia

Reputation: 33464

The pandoc executable can be extended with custom filters: https://pandoc.org/filters.html

And for transforming Pandoc documents directly in Haskell (without going through filters), a useful thing is the traversal Text.Pandoc.Walk.walkM and other related functions in pandoc-types.

For example I've used that in hakyll-alectryon to process Coq code blocks using pandoc, in Hakyll.

Upvotes: 2

Related Questions