[Rd] bzip2'ed data under data/

Sebastian P. Luque spluque at gmail.com
Mon Mar 26 20:23:42 CEST 2012


On Mon, 26 Mar 2012 16:44:58 +0100,
Prof Brian Ripley <ripley at stats.ox.ac.uk> wrote:

[...]

>>> Well, the extension is allowed 'optionally' to be .csv.bz2, but that
>>> does not make it good practice and I would suggest not using it.

>> Does this mean we can still compress the files, but leave the file
>> name with suffix *.csv (i.e. not adding the compression-specific
>> suffix)?  The 2 files I'm including in the package are a little over
>> 1 Mb, and bzip2 gets them down to< 150 Kb.

> Yes, that is what the help file says.

If I do that, however, R CMD build builds the package adding the .gz
suffix to these files (i.e. I end up with data/*.csv.gz), presumably
because --resave-data is the default, which uses gzip, provided no
BuildResaveData field is present in DESCRIPTION.  Adding
"BuildResaveData: no" to DESCRIPTION solves this.  For package
maintenance it might be easier to leave the uncompressed data/*.csv and
add "BuildResaveData: bzip2" to DESCRIPTION, but then R CMD build
generates data/*.csv.xz with R-devel (2012-03-22 r58801), so something
seems wrong.  Should we simply stick to doing the compression manually
and adding "BuildResaveData: no" to DESCRIPTION?

Thanks,

-- 
Seb



More information about the R-devel mailing list