[Rd] Building Packages on Windows using .Rbuildignore

[Gabor Grothendieck]

Duncan Murdoch <murdoch <at> stats.uwo.ca> writes:

> 
> On Wed, 17 Nov 2004 20:44:48 +0000 (UTC), Gabor Grothendieck
> <ggrothendieck <at> myway.com> wrote :
> 
> >If one does not use .Rbuildignore then with a pure R package
> >one can run off the original source files, in this case
> >in /wherever/mypkg, like this:
> >...without .Rbuildignore ...
> >R CMD install /wherever/mypkg
> >R CMD check /wherever/mypkg
> >R CMD build /wherever/mypkg
> >R CMD build /wherever/mypkg --binary
> >
> >but if one DOES use .Rbuildignore then one must first build the
> >package to remove the .Rbuildignore cited files and then run off
> >the built files which were created by unpacking the tar.gz file.
> 
> ... or what happens?  I still don't know exactly what you are claiming
> goes wrong.  

QUESTION

I am not claiming anything is wrong.  I was simply describing what
processing is needed to build a package that only uses R, noting that
such processing is different if one uses .Rbuildignore.  My question
was to ask:

- are there already scripts, makefiles or some other software or hooks
  such as configure.win to streamline this

EXAMPLE

For example, in
  http://tolstoy.newcastle.edu.au/R/help/04/11/6729.html
John Fox provided a batch file that automated package creation in the
case that there is only R code and one does not use .Rbuildignore.
I do use .Rbuildignore and need to do the analogous.   Most of my
past two posts described the manual commands that must be done in
each of these two cases to make clear what the process that I want
to streamline is and how it differs in these two cases.

DIFFERENCE BETWEEN USING .RBUILDIGNORE AND NOT

The reason that the processing is different according to whether one
uses .Rbuildignore or not is that 
  R CMD build 
takes the .Rbuildignore file into account but 
  R CMD install
  R CMD check
  R CMD build --binary 
do not take .Rbuildignore into account.  

> 
> It might be that the behaviour should be changed, or maybe there's an
> existing workaround, but I'd like to see a clear statement of the
> problem first.

Certainly its desirable to streamline this whole situation with a 
makefile, script or hook or perhaps one could modify the individual
commands: R CMD install, R CMD check and R CMD build --binary to 
take .Rbuildignore into account, possibly via a switch/flag.  However,
all I was asking about is what is already there.

Thanks.