My goal was to learn how to produce multiple images for inclusion in a single WW problem. It is possible, of course, to write pg code which will call TikZ from WW and produce excellent images for a problem on-the-fly, post-randomization. Using TikZ from withing WW is under active development. The approach here pre-produces the images, putting less load on the server when the questions are active. Of course, this requires more storage on the server, but disk capacity isn't an issue for us now.
- LaTeX including TikZ produces pdf
- ImageMagik: pdf -> pngs
- dvisvgm: pdf -> svgs
- The LaTeX files are based on a Jinja template
- The pg file is created from a Jinja template
- Configuration of each problem rests in a json file.
- Python files control the Jinja templates
- a make file controls the whole shebang
- python
- pathlib requires >= 3.6
- jinja2
- make (I use choco to install make under windows)
- imagemagick
- which relies on Ghostscript
- a modern LaTeX
- including TikZ and CircuiTikZ
- including latexmk
- including dvisvgm
- Starting LaTeX from C:\Users\Malcolm\Documents\Computing\LaTeX\Sandbox\Circuitkz\Circuitkz_1.tex as logic_circuit_1.tex.jinja
- Starting makefile from C:\Users\Malcolm\Documents\Computing\LaTeX\Examples\digraph
- python control file from C:\Users\Malcolm\Documents\Computing\Jinja\Jinja Experimentation 2020-12-14 and
- C:\Users\Malcolm\Documents\Computing\Jinja\TTL255 Tutorial
Almost no traces of these original sources remain in the current version of the project ;-).