Re: One failing test in latest version of PKI

[Simon Urbanek <simon.urbanek@R-project.org>]

Andreas,

thanks for alerting me about the OpenSSL issue - this is now fixed upstream:
https://github.com/s-u/PKI/commit/af3cbcafa13322d170e8ef993b919f57decf45e9

As for other error you see - that comes from
rawToChar(magic[1:5]) == "\xfd7zXZ"
I don't have any details on your setup (I cannot replicate) so I don't know why it happens (maybe that notorious C.UTF-8 locale again?), but should be easily fixed by avoiding strings.
https://github.com/s-u/PKI/issues/27
Please let me know if the latest version works for you.
https://rforge.net/PKI/snapshot/PKI_0.1-11.tar.gz

Cheers,
Simon



> On Mar 13, 2022, at 8:51 PM, Andreas Tille <andreas@an3as.eu> wrote:
> 
> Hi Simon,
> 
> when trying to build PKI for Debian I stumbled upon the following
> error of the test suite:
> 
> LC_ALL=C.UTF-8 R --no-save < tests/test.R
> 
>> ## on some systems using abs paths can break 100 byte limit
>> ## so we must do this in the tempdir
>> wd <- getwd()
>> td <- tempdir()
>> setwd(td)
>> tar(fn, basename(tmpfn), "none")
>> PKI.sign.tar(fn, key)
>    Error: (converted from warning) input string '�7zXZ' cannot be translated to UTF-8, is it valid in 'ANSI_X3.4-1968'?
>    Execution halted
> 
> Do you have any idea how to fix this?
> 
> Kind regards
> 
>      Andreas.
> 
> -- 
> http://fam-tille.de
>