[Rd] Including C Code for Tcl and Tk in an R package

Prof Brian Ripley ripley at stats.ox.ac.uk
Mon Aug 30 09:01:24 CEST 2010


On Mon, 30 Aug 2010, Adrian Waddell wrote:

> Dear R Community,
>
> I adapted the C-Code found at (http://wiki.tcl.tk/25685) implementing
> fast image resizing for Tk images for my R package.
>
> The C Code differs for tcl/tk version 8.4 and 8.5. I have the working
> code for both versions.
>
> How do I integrate this C-Code into my package so that R CMD INSTALL
> takes care of compiling the function correct for each OS and tcl
> version? (such that I can submit the package to CRAN).
>
> Once the two files would be compiled I could load the correct function

You can't do that: very few systems will have both Tcl 8.4 and 8.5 
libraries installed.  So you need to use the package's configure 
script to detect the Tcl/Tk version and also what libraries to compile 
against.

I'm not sure I would try to support Tcl 8.4: 8.5 has been out for a 
long time now and you could simply make it a package 
SystemRequirement.  We are only support 8.4 in R because we have 
existing code to do so that has not yet been removed.

> by checking with
>> library(tcltk)
>> .Tcl('set tcl_version')
> which tcl version R connects to?
>
> For example the gcc command I call under Ubuntu is:
>
> gcc -shared -o imgscale.so -DUSE_TCL_STUBS -I/usr/include/tcl8.5/
> imagescale_85.c -L/usr/lib/ -ltclstub8.5 -fPIC
>
> and
> gcc -shared -o imgscale.so -DUSE_TCL_STUBS -I/usr/include/tcl8.4/
> imagescale_84.c -L/usr/lib/ -ltclstub8.4 -fPIC
>
>
> Thanks,
>
> Adrian Waddell
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595



More information about the R-devel mailing list