[R] "found 4 marked UTF-8 strings" during check of package... but where !

Marc Girondot marc_grt at yahoo.fr
Fri Mar 10 19:44:18 CET 2017

Thanks Duncan and Michael,

Indeed I have data file with utf-8 characters inside. In the 
DESCRIPTION, I have the line Encoding: UTF-8
but it seems to not be sufficient.
In each R page for these data, I have also :
#' @docType data
#' @encoding UTF-8

But I still have the notes during check when I try to submit the package 
in CRAN (not in local --as-cran check).

How I could "say" that these data have utf-8 characters inside?


Le 10/03/2017 à 15:24, Duncan Murdoch a écrit :
> On 10/03/2017 2:52 AM, Marc Girondot via R-help wrote:
>> Dear members,
>> I want submit to CRAN a new version of a package that I maintain. When I
>> check locally "as-cran" no note or error are reported but the link after
>> submission reports several notes and one warning:
>> For example:
>> using R Under development (unstable) (2017-03-05 r72309)
>> using platform: x86_64-apple-darwin16.4.0 (64-bit)
>> using session charset: UTF-8
>> ...
>> checking extension type ... Package
>> this is package ‘embryogrowth’ version ‘6.4’
>> package encoding: UTF-8
>> ...
>> checking data for non-ASCII characters ... NOTE
>>    Note: found 4 marked UTF-8 strings
>> I have the same with
>> using R version 3.3.0 (2016-05-03)
>> using platform: x86_64-apple-darwin13.4.0 (64-bit)
>> but not with some others such as r-devel-linux-x86_64-debian-gcc
>> Based on the message, "Note: found 4 marked UTF-8 strings", it seems
>> that "4 marked UTF-8 strings" are present in the package and it is a
>> problem...
>> Is there any solution to know in which file?
> It's one containing an object coming from your data directory.
> R won't give more detail than that, but if you still can't guess, you 
> could get some idea by debugging the check code:
> debug(tools:::.check_package_datasets)
> tools:::.check_package_datasets(pkg)
> where pkg contains the path to the package source code.  That function 
> does the checking one variable at a time.
> Duncan Murdoch

More information about the R-help mailing list