Today I finished a project I have been working on over the last couple months. I've compiled a lot of my knowledge with I've gained over the last year into a GitHub repository, and created a document using Python 2.7 and pandoc, a free and open source tool for converting documents between common types.

Markdown is a markup language that translates directly into HTML. I wrote units using the language for each module from the assistant technical director apprenticeship standard.

build_pdf.py compiles a pdf using these documents.

The solution I came up with works in 4 stages:

1. The .md (Markdown) units are compiled into a single monolithic Markdown file (./md/out.md).
2. Images from ./image_orig/ are shrunk to a hardcoded width using ffmpeg and are moved to ./images/. (This makes the PDF lighter.)
3. This file is then converted to a LaTeX file (./tex/out.tex), another markup language, which gives me more control of the formatting.
4. Python fixes are run on this file which I have written, which adjust the formatting to my tastes. (i.e. Titles are added, page breaks are added before new sections and at specific points I've specified.)
5. The pdf is generated by pandoc (./pdf/out.pdf).

When you run ./build_pdf.py you have the option of which task to run.

[hbal@terminal6 atd-apprentice-module-summaries]$ python2.7 build_pdf.py
1 md -->md + Shrink Images
2 md_compiled-->tex  + Run Fixes
3 tex_compiled-->pdf
0 Run All
>>> 

I've also created a booklet version (./pdf/booklet.pdf) which makes for an easier way to read this. This can be created by printing 2 pages each on each side of landscape double-sided paper (flip on the short-edge) in this order:

60, 1, 2, 59, 58, 3, 4, 57, 56, 5, 6, 55, 54, 7, 8, 53, 52, 9, 10, 51, 50, 11, 12, 49, 48, 13, 14, 47, 46, 15, 16, 45, 44, 17, 18, 43, 42, 19, 20, 41, 40, 21, 22, 39, 38, 23, 24, 37, 36, 25, 26, 35, 34, 27, 28, 33, 32, 29, 30, 31

(This sequence was generated with ./booklet_pages.py)

Thanks for reading.