{Vincent Goulet}
{You (S)wove? Well, (S)tangle now!}
{The concept of literate programming, pionnered by Donald Knuth in
1984, should be familiar to many in the \AllTeX\ community. Very
briefly stated, it consists of a programming paradigm where a program
and its documentation are interspersed in a single file. Source
code and documentation are extracted respectively by \tbcode{tangle} and
\tbcode{weave} procedures. Literate programming plays a central role nowadays
in scientific computing for reproducible research purposes: instead of
being hard coded into a report, results and graphics are woven in
place using the computer code. In the R ecosystem, Sweave and knitr
are widely used to build documents this way from R code. The aim of
this presentation is to shed some light on the perhaps lesser-known
component of literate programming, at least in scientific computing:
the \tbcode{tangle} step. I will describe a use case where a clever
combination of weaving and tangling allows for efficiently maintaining a
set of exercises and solutions.}