rPithon is a package that allows you to use Python from within R. To make this possible, it starts a Python executable and communicates with it over a pipe.
More info: http://rpithon.r-forge.r-project.org
This package is my first experiment in using Rcpp, and provides another way to read a CSV file. In principle, you need to specify the number and type (integer, real number, string) of each column, but this may also be guessed from the first data line. On GNU/Linux, parsing the CSV text data can be done using multiple threads (using OpenMP), possibly offering a speedup.
The function provided by this package is somewhat slower than the
fread function from the data.table package when using a single thread, but may very well be faster when using all available processor cores. It may require more memory though.
More info: http://readcsvcolumns.r-forge.r-project.org