[gentoo-soc] Automatically generated overlay of R packages - progress report #10

[André Erdmann]

Hi everyone,

== Brief summary of this project ==

The aim of this project is to create scripts that automate the process
of overlay creation/maintenance for R packages from repositories such
as CRAN and Bioconductor.

Longer:
For the ebuild creation of a single package one needs to extract the
package, copy-paste data from its description file to the ebuild and
look up dependencies, which is time-consuming.
Although trivial for a few number of packages, this is practically
impossible to do by hand for repositories like CRAN (> 3500 packages),
especially 'cause it also requires tracking changes (new / updated /
removed packages).
The solution is to automate that process and this is what this project is about.

== Last week's progress ==

* Implemented incremental overlay writing: creation of an ebuild can
now be skipped if it already exists and added the possibility to keep
only the n-th latest ebuilds per package

* the overlay module can now be used for dependency resolution: all
added/existing packages are part of a DynamicSelfdepPool that resolves
package-deps (e.g. 'seewave' -> 'sci-R/seewave') without any
dependency rule file

* some minor changes based on test results (typos, corner cases like a
few Z compressed packages, ...)

* useful dependency rules: the last test run showed 30094/30921
successful ebuild creations (BIOC, CRAN, CRAN-Archive) with ~400
unresolvable dependencies. Most of them are from the 'Suggests' field
of a package for which ebuild creation ignores unresolvable deps
("collect as many suggestions as possible"). Some are packages from
other repos (omegahat, R-Forge and packages without a repo) and some
are not in the portage tree (e.g. Cytoscape).


== Plans for this week ==

* implement another way to get packages (via http, with the option to
read/parse a remote's PACKAGES file to generate a package list)
* testing

--
Regards,
André E.