A rake task that indexes CRAN packages with a basic web interface and cron scheduler for running the task periodically.

CRAN is a network of ftp and web servers around the world that store identical, up-to-date, versions of code and documentation for R. The use these CRAN Servers to store the R packages.

A CRAN server is just a simple Apache Dir with a bunch of tar.gz files.

Every CRAN server contains a plain file listing all packages in that server.

A package can be accessed via

http://cran.rproject.org/src/contrib/[PACKAGE_NAME]_[PACKAGE_VERSION].tar.gz

A package contains a description file, that contains info about the package.

• Extract useful info from the PACKAGES file of the CRAN server.
• Extract certain info from the DESCRIPTION file of each package.
• On every run of the task, for the first 50 packages, store info about each package, author(s) and maintainers(s).
• Prevents duplicate indexes of a version of same package.
• I used Whenever for scheduling the task. If you use Heroku, Heroku scheduler should do just fine.

rails packages:index

bundle exec rspec

• Pagination for listing packages.
• I currently use DCF for converting the content of the PACKAGES & DESCRIPTION file to hashes. Maybe I could use a faster library but this does the job as well :)
• More, more and more