This repo shows an example for how to create a glossary using the quarto extension nutshell.

Create a glossary.md file for your project with terms as second level headings and their explanation as a paragraph below

# Glossary

## Quarto

Quarto is an open-source scientific and technical publishing system built on Pandoc

The prerender script glossary.R looks as follows

files <- Sys.getenv("QUARTO_PROJECT_INPUT_FILES")
glossary <- readLines("glossary.md")
terms <- glossary[stringr::str_detect(glossary, "^## ")] |> stringr::str_remove_all("## ")

for (f in files){
    if (f != "glossary.md"){
      co <- readLines(f)
      if (any(stringr::str_detect(co, paste0(terms, collapse = "|")))) {
          print(f)
          co <- stringi::stri_replace_all_regex(co, paste0("(?<![:#])", terms),
              paste0("[:", terms, "](/glossary.html#", terms, ")"), vectorize_all = FALSE)
          writeLines(co, f)
      }
    }
}

The script reads all files that are to be rendered and replaces every term from the glossary with a nutshell link. For our example, all instances of Quarto turn into [:Quarto](/glossary.html#Quarto). It thus changes all input files.

Add the script to _quarto.yml like this

project:
  type: book
  pre-render: glossary.R

The rendered demo book included in this repo with a glossary can be found on quarto.pub

The pre-render R script is not perfect and still struggles with some cases. Also, I am not sure if a pre-render script is the best solution for this. Probably a lua filter would make more sense, but my lua skills are still limited.