[R-pkg-devel] [External] RcmdrPlugin.HH_1.1-48.tar.gz

[Richard M. Heiberger]

Almost.

I used 
prompt(".__global__")
to create file
man/.__global__.Rd

This file does not appear in the tar.gz file, but without a note of rejection.
When I checked my disk directory directly
R CMD check RcmdrPlugin.HH
the file was rejected with

Found the following hidden files and directories:
  .DS_Store
  R/.DS_Store
  man/.__global__.Rd
These were most likely included in error. See section ‘Package
structure’ in the ‘Writing R Extensions’ manual.

I looked there
Section 1.1 says that the acceptable characters are
    A-Za-z0-9._!#$%&+,;=@^(){}'[]
and "." and _ are explicitly included.

What should I try next?


> On Mar 5, 2024, at 18:21, Duncan Murdoch <murdoch.duncan using gmail.com> wrote:
> 
> On 05/03/2024 5:41 p.m., Richard M. Heiberger wrote:
>> My package is being rejected by auto-check
>> Flavor: r-devel-linux-x86_64-debian-gcc, r-devel-windows-x86_64
>> Check: for missing documentation entries, Result: WARNING
>>  Undocumented code objects:
>>    '.__global__'
>>  All user-level objects in a package should have documentation entries.
>>  See chapter 'Writing R documentation files' in the 'Writing R
>>  Extensions' manual.
>> The problem is that the string    '.__global__'  is not in the package.
>> I can't find it and John Fox, the maintainer of Rcmdr, can'f find it.
>> Can someone help me understand why a non-existent string is being detected?
> 
> That's the variable modified by the `globalVariables()` function.  So it may well exist in your package.  I'd guess the problem is that your package exports functions by giving a pattern for the names instead of listing each one separately, and it matches that variable.
> 
> Duncan Murdoch
> 
> 
>