Re: [Rd] package NULL not found

From: Prof Brian Ripley <ripley_at_stats.ox.ac.uk>
Date: Fri, 20 Jul 2007 12:17:19 +0100 (BST)

Note that the message is not about '' and NULL but about '<nl>' and NULL<nl>.

I suspect the line endings are screwed up in the environment in use, which seems not to be our toolset from the comments about Cygwin. It is most likely the shell used by make that is doing this, but it could be Perl.

I can reproduce this on Linux by

gannet% setenv R_DEFAULT_PACKAGES 'NULL\ '
gannet% Rdev --slave
During startup - Warning messages:
1: In library(package, lib.loc = lib.loc, character.only = TRUE, logical.return = TRUE, :

   there is no package called 'NULL
'
2: package NULL
  in options("defaultPackages") was not found

On Thu, 19 Jul 2007, Duncan Murdoch wrote:

> On 19/07/2007 7:16 PM, Gabor Grothendieck wrote:
>> Thanks. I tried performing the check from an empty directory but
>> it still gave the same response. Where can I find the source code
>> for the check process?
>
> On Windows the main wrapper is a Perl script in RHOME/bin/check, but
> most of the work is done by routines in the tools package. The one that
> seems to be dying for you is tools:::.check_package_depends.

It is what is getting passed to Rterm that seems to be the problem.

>
> Duncan Murdoch
>
>>
>> * using log directory 'C:/Rpkgs/check/sqldf.Rcheck'
>> * using ARGUMENT '
>> ' __ignored__ R version 2.5.1 (2007-06-27)
>> * checking for file 'sqldf/DESCRIPTION' ... OK
>> * this is package 'sqldf' version '0-1.0'
>> * checking package dependencies ... ERROR
>> During startup - Warning messages:
>> 1: there is no package called 'NULL
>> ' in: library(package, lib.loc = lib.loc, character.only = TRUE,
>> logical = TRUE,
>> 2: package NULL
>> in options("defaultPackages") was not found
>> See the information on DESCRIPTION files in the chapter 'Creating R
>> packages' of the 'Writing R Extensions' manual.
>>
>>
>> On 7/19/07, Duncan Murdoch <murdoch_at_stats.uwo.ca> wrote:
>>> On 19/07/2007 5:55 PM, Gabor Grothendieck wrote:
>>>> I noticed I am getting the same messages when trying to check other
>>>> packages too such as gsubfn which previously checked ok.
>>>>
>>>> I had recently reinstalled cygwin so its likely something to do with
>>>> that but have not tracked it down.
>>> Another possibility may be that you have a .RData file in the check
>>> directory, and it is masking some important function. This was a
>>> problem for me a few days ago with INSTALL; I don't know if it affects
>>> check.
>>>
>>> Duncan Murdoch
>>>
>>>> On 7/19/07, Gabor Grothendieck <ggrothendieck_at_gmail.com> wrote:
>>>>> In performing Rcmd check I am getting this output regarding using
>>>>> Argument '' and a NULL package not found and it stops with an error:
>>>>>
>>>>> * using log directory 'C:/Rpkgs/sqldf.Rcheck'
>>>>> * using ARGUMENT '
>>>>> ' __ignored__ R version 2.5.1 (2007-06-27)
>>>>> * checking for file 'sqldf/DESCRIPTION' ... OK
>>>>> * this is package 'sqldf' version '0.1-0'
>>>>> * checking package dependencies ... ERROR
>>>>> During startup - Warning messages:
>>>>> 1: there is no package called 'NULL
>>>>> ' in: library(package, lib.loc = lib.loc, character.only = TRUE,
>>>>> logical = TRUE,
>>>>> 2: package NULL
>>>>> in options("defaultPackages") was not found
>>>>> See the information on DESCRIPTION files in the chapter 'Creating R
>>>>> packages' of the 'Writing R Extensions' manual.
>>>>>
>>>>> My DESCRIPTION file looks like this:
>>>>>
>>>>> Package: sqldf
>>>>> Version: 0.1-0
>>>>> Date: 2007-07-19
>>>>> Title: SQL on data frames.
>>>>> Author: G. Grothendieck
>>>>> Maintainer: G. Grothendieck <ggrothendieck_at_gmail.com>
>>>>> Description: Thin layer over RSQLite and RMySQL facilitating use of
>>>>> SQL selects on data frames.
>>>>> Depends: R (>= 2.5.1), RSQLite (>= 0.5-5), gsubfn
>>>>> License: GPL (Version 3 or later)
>>>>> URL: http://code.google.com/p/sqldf/
>>>>>
>>>>> and the entire package is at (click on Source tab):
>>>>>
>>>>> http://code.google.com/p/sqldf/
>>>>>
>>>>> I am using
>>>>>> R.version.string # Windows XP
>>>>> [1] "R version 2.5.1 (2007-06-27)"
>>>>>
>>>>> Does anyone know what this means and how to fix it?
>>>>>
>>>>> The package itself seems to work. Its just getting it past Rcmd check
>>>>> that is the problem.
>>>>>
>>>>> Thanks.
>>>>>
>>>> ______________________________________________
>>>> R-devel_at_r-project.org mailing list
>>>> https://stat.ethz.ch/mailman/listinfo/r-devel
>>>
>
> ______________________________________________
> 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

______________________________________________
R-devel_at_r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Received on Fri 20 Jul 2007 - 12:25:20 GMT

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Sat 21 Jul 2007 - 06:36:46 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-devel. Please read the posting guide before posting to the list.