Ever needed to create a PDF that includes code snippets together with their
outputs? Spent too much time running the scripts and copying their outputs into
your tex document? No more of that! doc-builder automates the process by
running your scripts, adding their outputs to your tex file and compiling the
result into a nice PDF.
Requirements: Node.js, pdflatex
To build the PDF, run node build.js.
[TODO:] decide how to handle solutions
[TODO:] receive main tex file as input
public-sources/ contains scripts that, together with their output, may appear
in the PDF. They are run by the build script build.js every time before the
PDF compilation and their output is saved temporarily. We provide the following
LaTeX macros to import specific files:
\getenv[\sourcepath]{<script-name>}assigns the actual file path of<script-name>to the LaTeX macro\sourcepath.\codelisting{\sourcepath}puts the source code verbatim in the PDF.\outputlisting{\sourcepath}puts the output of the source code in the PDF.\solution{<script-name>}automates the formatting of source code and its output in a "Solution" box, useful if you're creating versions with and without solutions (e.g. for an exercise). To enable this macro, runnode build.js --with-solutions.
private-sources/ contains all non-LaTeX supporting code that does not appear
in the PDF, e.g. a script that creates some random data to be consumed by a
script in public-sources/.
[TODO:] implement toposort and provide nice API, e.g. YAML file of deps
[TODO:] allow arbitrary directories, as specified in the dependency file, or
scripts/ dir for those who don't bother with the dep file
As far as LaTeX is concerned, all PDF contents are in main.tex. Command
definitions and package declarations are in preamble.sty.
[TODO:] allow arbitrary tex file structures