I believe this is because R uses sQuote() to print package versions
which introduces curly quotes (non-ASCII characters).

> packageVersion('knitr')
[1] ‘0.5’

Unfortunately sQuote() was hard-coded in the printing function, so
there is no direct way to get rid of the two curly quotes, and you
have to use as.character(), or packageDescription('knitr',
fields='Version')

> getAnywhere(print.numeric_version)
A single object matching ‘print.numeric_version’ was found
It was found in the following places
  package:base
  registered S3 method for print from namespace base
  namespace:base
with value

function (x, ...)
{
    y <- as.character(x)
    if (!length(y))
        writeLines(gettext("<0 elements>"))
    else print(noquote(ifelse(is.na(y), NA_character_, sQuote(y))),
        ...)
    invisible(x)
}
<bytecode: 0x1a26bf8>
<environment: namespace:base>


Regards,
Yihui
--
Yihui Xie <xieyi...@gmail.com>
Phone: 515-294-2465 Web: http://yihui.name
Department of Statistics, Iowa State University
2215 Snedecor Hall, Ames, IA


On Sat, Jun 2, 2012 at 8:32 PM, Scott Kostyshak <skost...@princeton.edu> wrote:
> From: Yihui Xie [xieyi...@gmail.com]
> Sent: Saturday, June 02, 2012 8:04 PM
>
>>I cannot reproduce the problem. It compiles successfully with TeXLive
>>2011 under Ubuntu. Can anyone else reproduce it?
>
> I am using Ubuntu (64-bit) 12.04 with TeXLive 2011 and pdflatex version:
> scott@wws-1j1dnk1:~$ pdflatex --version
> pdfTeX 3.1415926-2.3-1.40.12 (TeX Live 2011)
> kpathsea version 6.0.1
>
>>Does this happen to packageVersion() only?
>
> Also with packageDescription(). But I've compiled many knitr documents that 
> have used many packages and functions and this is the first time that I have 
> this problem. If I wrap packageVersion as follows then it works fine:
> as.character(packageVersion('knitr'))
>
> Can you send me the .tex file that is produced when you export the .lyx file 
> I sent to latex (pdflatex)? If they are the same and it works for you then 
> the problem must be my pdflatex. If they are different then the problem is 
> something else.
>
> Thanks,
>
> Scott

Reply via email to