Technical Documents from Markdown

Table of Contents

• Technical Documents from Markdown
  • Introduction
  • Structure
  • Frontmatter
  • Requirements
    • Debian/Ubuntu Based Systems
  • ToDo

Introduction

This document system for generating PDF and HTML files from Markdown text is based on Pandoc. Inspiration for the templates comes from a LaTeX package created back in 2001, called technics.sty.

Addiva Elektronik have released this project as Open Source to invite others in need of a similar framework to share and further improve on the founding ideas.

Examples

First the HTML version:

Next the PDF version:

Notice how the HTML version has two click buttons on the right-hand side just below the header. One for opening the PDF version (better for print) and one for opening the HTML print preview in your browser. Be aware that CSS media print still is very broken in most browsers. Any help improving templates/addiva.html CSS to be able to handle or trick browsers better, is most welcome!

Usage

1. Fork or clone the repository, then jump to section Structure

2. Set up docgen.git as a GIT submodule. E.g., if your project has a subdirectory doc/:

   cd doc/
   git submodule add [email protected]:addiva-elektronik/docgen.git
   ln -s docgen/Makefile .
   

That's it. Provided you have read the Requirements section, you can type make and then find your generated documents in the output/ sub-directory.

Note: the docgen/ sub-directory will be skipped by docgen, only actual substructures in your projects doc/ directory (with .md files) will be used.

Structure

Documents go in their own subdirectory, nesting subdirectories is not supported. Each subdirectory from the top level can have multiple Markdown files. For each Markdown file (.md suffix only) the Makefile generates a HTML and PDF version in the output/ subdirectory using pandoc.

docgen/
 |-- foo/
 |    |-- concepts.md
 |    `-- memo-2022-09-12.md
 |-- bar/
 |    |-- concepts.md
 |    `-- memo-2022-08-01.md
 |-- output/
 |    |-- foo/
 |    |    |-- concepts.html
 |    |    |-- concepts.pdf
 |    |    |-- memo-2022-09-12.html
 |    |    `-- memo-2022-09-12.pdf
 |    `-- bar/
 |          |-- concepts.html
 |          |-- concepts.pdf
 |          |-- memo-2022-08-01.html
 |          `-- memo-2022-08-01.pdf
 |-- templates/
 |    |-- addiva.html
 |    `-- addiva.pdf
 |-- Makefile
 `-- server.sh

Currently two templates, for HTML and PDF generation, are available, but more could be added. For example .docx and slideshows in reveal.js.

Frontmatter

Standard Pandoc variables are supported in the (optional) Yaml frontmatter. However, to enable the Addiva specific document header variants a couple of extra variables have been addded. These are described in this section.

Here is an example of a YAML frontmatter, which goes at the very top of the document. Most relevant options are under document:, the rest can be omitted completely.

Notice the starting and closing --- of the frontmatter:

---
document:
  name: Sample Technical Document
  prepared: Joachim Wiberg
  organisation: Addiva Elektronik AB
  approved:
copyright:
  holder: Addiva Elektronik AB
numbersections: true
classoption: twoside
fontsize: 11pt
mainfont: SourceSerifPro-Regular.otf
mainfontoptions: Numbers=OldStyle
sansfont: SourceSansPro-Regular.otf
sansfontoptions: Numbers=OldStyle
monofont: SourceCodePro-Regular.otf
monofontoptions: Scale=0.75
---

• document: enables the document header
  • name: sets the Document name field
  • prepared: sets the Prepared by field
  • approved: sets the Approved by field, if unset PDF documents will have a DRAFT watermark applied
  • organisation: enables the optional Organisation field right of the logo
  • number: sets Document No field, defaults to an automatically generated string composed from the GIT repository name, the source file's object hash and last change GIT revision of the file: name:object-revision. Note: the -revision part is always added to the document number.
• copyright: enables a copyright footer (type custom text here), or fall back to the default by setting either of the following options:
  • year: sets the optional copyright years field, e.g., "2021-2022", default: automatically sets current year
  • holder: defaults to the organisation (above), or "Addiva Elektronik AB" if organisation is unset
• numbersections: optional, if set to true the PDF will used numbered sections (this is what most readers expect)
• classoption: optional, used for PDF generation (LaTeX class option)
• fontsize: optional
• *font: optional, the xelatex pandoc options are used with fallback to the Adobe Source{Sans,Serif,Code}-Regular.otf family
• *fontoptions: optional, needs the corresponding *font: option to be set to take effect. Default is Numbers=OldStyle for Sans and Serif fonts, and Scale=0.0 for the monofont

Note: the classoption, includes, and font variables are for the PDF generation (via LaTeX). Also, only PDF document support the DRAFT stamp.

Requirements

Debian/Ubuntu Based Systems

sudo apt install build-essential git
sudo apt install pandoc texlive texlive-latex-extra texlive-fonts-extra texlive-font-utils