Ah, sorry, this is my bad - I was trying on an older version of R and it seemed
to work, but it was apparently my fault for not setting the environment
variable correctly in that test, so I take the regression comment back (I had
the recent encoding changes in mind so). I still think it may be w
Thanks for the quick replies.
For me, the main issue is that the value of an environment variable
can causes errors in R. Say that there's an environment variable that
is used as a raw data blob (as Simon mentions) and where that data is
flavor of the day. Then, this would affect R in a random,
n
Tomas,
I think you're not addressing the actual issue which is a clear regression in
Sys.getenv() [because it used to work and still works for single env var, but
not a list] and the cryptic error due to that regression (caused by changes in
R-devel). So in either case, Sys.getenv needs fixing
On 1/30/23 23:01, Henrik Bengtsson wrote:
/Hello.
SUMMARY:
$ BOOM=$'\xFF' LC_ALL=en_US.UTF-8 Rscript --vanilla -e "Sys.getenv()"
Error in substring(x, m + 1L) : invalid multibyte string at ''
$ BOOM=$'\xFF' LC_ALL=en_US.UTF-8 Rscript --vanilla -e "Sys.getenv('BOOM')"
[1] "\xff"
BACKGROUND:
I
/Hello.
SUMMARY:
$ BOOM=$'\xFF' LC_ALL=en_US.UTF-8 Rscript --vanilla -e "Sys.getenv()"
Error in substring(x, m + 1L) : invalid multibyte string at ''
$ BOOM=$'\xFF' LC_ALL=en_US.UTF-8 Rscript --vanilla -e "Sys.getenv('BOOM')"
[1] "\xff"
BACKGROUND:
I launch R through an Son of Grid Engine (SGE)
