R devel - Apr 2020 - How to find detritis rejected by "R CMD check" on Debian?

Hello:


 ????? How can someone help me find and fix the following, contained in 
00check.log on Debian for "https://github.com/JamesRamsay5/fda":


NOTE
Found the following files/directories:
 ? ?fdaMatlabPath.m?
* checking for detritus in the temp directory ... OK


 ????? See:


https://win-builder.r-project.org/incoming_pretest/fda_5.1.3_20200416_225207/Debian/00check.log


 ????? Thanks,
 ????? Spencer Graves

Spencer,

On 17 April 2020 at 09:14, Spencer Graves wrote:
|  ????? How can someone help me find and fix the following, contained in 
| 00check.log on Debian for "https://github.com/JamesRamsay5/fda":
| 
| NOTE
| Found the following files/directories:
|  ? ?fdaMatlabPath.m?
| * checking for detritus in the temp directory ... OK

Obviously from your package, so you could use a tool like 'grep'
locally.

We can just turn to Gabor's handy CRAN mirror at GitHub, open the
repo at github.com/cran/fda and enter the filename in the search.
Second hit is

  writeLines(d2a, 'fdaMatlabPath.m')

which is obviously a file _you write_ and not not delete just as
the package checking diagnostic suggests. As frequently suggested here or on
r-package-devel (where this belonged), prefix paths by `tempdir()` or change
there first.

Dirk



-- 
http://dirk.eddelbuettel.com | @eddelbuettel | edd at debian.org

This should probably have been addressed to R-pkg-devel, not Rd.

On Fri, 17 Apr 2020 09:14:44 -0500
Spencer Graves <spencer.graves at prodsyse.com> wrote:
> Found the following files/directories:
>  ? ?fdaMatlabPath.m?
This is not the "detritus in the temp directory"; the message is
related
to the previous line in the log:
>> * checking for non-standard things in the check directory ... NOTE
I searched the repo for the file name in question and got a few hits
[1]. R CMD check runs the \examples{}, including those in
man/fdaMatlabPath.Rd. The function from R/fdaMatlabPath.R, called from
the examples, writes a file in the current directory [2], which
happens to be the check directory during R CMD check. In order to
conform to the CRAN policy, the function should receive the target file
name as a parameter instead, and the example should pass a file path
somewhere in tempdir().

-- 
Best regards,
Ivan

[1] https://github.com/JamesRamsay5/fda/search?q=fdaMatlabPath.m

[2]
https://github.com/JamesRamsay5/fda/blob/ca077f87f69efcfc434eeea39f4fff6136fcfaeb/R/fdaMatlabPath.R#L33

On 17/04/2020 10:14 a.m., Spencer Graves wrote:
> Hello:
> 
> 
>   ????? How can someone help me find and fix the following, contained in
> 00check.log on Debian for "https://github.com/JamesRamsay5/fda":
> 
> 
> NOTE
> Found the following files/directories:
>   ? ?fdaMatlabPath.m?
> * checking for detritus in the temp directory ... OK
> 
> 
>   ????? See:

That would have resulted from running the example in ?fdaMatlabPath.

You could stop that from happening in a test by using

if (interactive())
   fdaMatlabPath()

but it still potentially wipes out a user's file.  Can't you ask the 
user for the filename to write to, and have the example write to 
tempdir() instead of the current directory?

Duncan Murdoch