R devel - Dec 2007 - (PR#10534 capture.output(), truncated last output without

This only happens if 'file' is a text connection, and is the expected 
behaviour in that case: you cannot capture an incomplete line to a text 
connection.

There seems no reason to break the documented behaviour in other cases to 
change something that you consider to a bug when file=NULL and the user 
does not produce complete output.  It would be possible to make use of 
isIncomplete() to add a final newline only where needed to complete a 
line.  E.g. use

     if(inherits(file, "textConnection") &&
isIncomplete(file)) cat("\n")


On Sat, 29 Dec 2007, phgrosjean at sciviews.org wrote:
> Full_Name: Philippe Grosjean
> Version: 2.6.1
> OS: MacOS X; Windows XP
> Submission from: (NULL) (81.243.237.235)
>
>
> Last output from capture.output() is truncated if it does not end with a
> carriage return:
>
>> capture.output(cat("text\n"))  # Fine
> [1] "text"
>> capture.output(cat("text"))    # Missing output!
> character(0)
>> capture.output({cat("text");1+1})  # Only last output is
affected
> [1] "text[1] 2"
>
> Proposed patch: add a carriage return before exiting capture.output():
>
> capture.output <-
> function (..., file = NULL, append = FALSE) {
>
>
> [...]
>
>        for (item in tmp) if (item$visible)
>            print(item$value)
>    }
>    cat("\n")  ### ADD THIS!
>    rval
> }
>
> This changes the behavior of capture.output() a little bit, since it adds
"" at
> the end of regular outputs, but it solves the problem and it allows to
detect if
> last output line was ended by \n, or not.
>
> Regards,
>
> Philippe Grosjean
>               _
> platform       i386-apple-darwin8.10.1
> arch           i386
> os             darwin8.10.1
> system         i386, darwin8.10.1
> status
> major          2
> minor          6.1
> year           2007
> month          11
> day            26
> svn rev        43537
> language       R
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

>>>>> "BDR" == Brian Ripley <ripley at
stats.ox.ac.uk>
>>>>>     on Sat, 29 Dec 2007 13:20:14 +0100 (CET) writes:
    BDR> This only happens if 'file' is a text connection, and
    BDR> is the expected behaviour in that case: you cannot
    BDR> capture an incomplete line to a text connection.

    BDR> There seems no reason to break the documented behaviour
    BDR> in other cases to change something that you consider to
    BDR> a bug when file=NULL and the user does not produce
    BDR> complete output.  It would be possible to make use of
    BDR> isIncomplete() to add a final newline only where needed
    BDR> to complete a line.  E.g. use

    BDR>   if(inherits(file, "textConnection") &&
isIncomplete(file)) cat("\n")

I think that would be quite a nice improvement.
Martin

    BDR> On Sat, 29 Dec 2007, phgrosjean at sciviews.org wrote:

    >> Full_Name: Philippe Grosjean Version: 2.6.1 OS: MacOS X;
    >> Windows XP Submission from: (NULL) (81.243.237.235)
    >> 
    >> 
    >> Last output from capture.output() is truncated if it does
    >> not end with a carriage return:
    >> 
    >>> capture.output(cat("text\n")) # Fine
    >> [1] "text"
    >>> capture.output(cat("text")) # Missing output!
    >> character(0)
    >>> capture.output({cat("text");1+1}) # Only last output
is
    >>> affected
    >> [1] "text[1] 2"
    >> 
    >> Proposed patch: add a carriage return before exiting
    >> capture.output():
    >> 
    >> capture.output <- function (..., file = NULL, append    
>> FALSE) {
    >> 
    >> 
    >> [...]
    >> 
    >> for (item in tmp) if (item$visible) print(item$value) }
    >> cat("\n") ### ADD THIS!  rval }
    >> 
    >> This changes the behavior of capture.output() a little
    >> bit, since it adds "" at the end of regular outputs, but
    >> it solves the problem and it allows to detect if last
    >> output line was ended by \n, or not.
    >> 
    >> Regards,
    >> 
    >> Philippe Grosjean _ platform i386-apple-darwin8.10.1 arch
    >> i386 os darwin8.10.1 system i386, darwin8.10.1 status
    >> major 2 minor 6.1 year 2007 month 11 day 26 svn rev 43537
    >> language R
    >> 
    >> ______________________________________________
    >> R-devel at r-project.org mailing list
    >> https://stat.ethz.ch/mailman/listinfo/r-devel
    >> 

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
    BDR> Professor of Applied Statistics,
    BDR> http://www.stats.ox.ac.uk/~ripley/ University of
    BDR> Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road,
    BDR> +44 1865 272866 (PA) Oxford OX1 3TG, UK Fax: +44 1865
    BDR> 272595

    BDR> ______________________________________________
    BDR> R-devel at r-project.org mailing list
    BDR> https://stat.ethz.ch/mailman/listinfo/r-devel

On Sat, 29 Dec 2007, Martin Maechler wrote:
>>>>>> "BDR" == Brian Ripley <ripley at
stats.ox.ac.uk>
>>>>>>     on Sat, 29 Dec 2007 13:20:14 +0100 (CET) writes:
>
>    BDR> This only happens if 'file' is a text connection, and
>    BDR> is the expected behaviour in that case: you cannot
>    BDR> capture an incomplete line to a text connection.
>
>    BDR> There seems no reason to break the documented behaviour
>    BDR> in other cases to change something that you consider to
>    BDR> a bug when file=NULL and the user does not produce
>    BDR> complete output.  It would be possible to make use of
>    BDR> isIncomplete() to add a final newline only where needed
>    BDR> to complete a line.  E.g. use
>
>    BDR>   if(inherits(file, "textConnection") &&
isIncomplete(file)) cat("\n")
>
> I think that would be quite a nice improvement.
Yes, I'd like to achieve the effect, but I think there is a better way to 
do it.

The nub of the problem is that close() is done in the on.exit action 
after the return value is set, so the internal logic needs changing.

Brian
> Martin
>
>    BDR> On Sat, 29 Dec 2007, phgrosjean at sciviews.org wrote:
>
>    >> Full_Name: Philippe Grosjean Version: 2.6.1 OS: MacOS X;
>    >> Windows XP Submission from: (NULL) (81.243.237.235)
>    >>
>    >>
>    >> Last output from capture.output() is truncated if it does
>    >> not end with a carriage return:
>    >>
>    >>> capture.output(cat("text\n")) # Fine
>    >> [1] "text"
>    >>> capture.output(cat("text")) # Missing output!
>    >> character(0)
>    >>> capture.output({cat("text");1+1}) # Only last
output is
>    >>> affected
>    >> [1] "text[1] 2"
>    >>
>    >> Proposed patch: add a carriage return before exiting
>    >> capture.output():
>    >>
>    >> capture.output <- function (..., file = NULL, append >   
>> FALSE) {
>    >>
>    >>
>    >> [...]
>    >>
>    >> for (item in tmp) if (item$visible) print(item$value) }
>    >> cat("\n") ### ADD THIS!  rval }
>    >>
>    >> This changes the behavior of capture.output() a little
>    >> bit, since it adds "" at the end of regular outputs,
but
>    >> it solves the problem and it allows to detect if last
>    >> output line was ended by \n, or not.
>    >>
>    >> Regards,
>    >>
>    >> Philippe Grosjean _ platform i386-apple-darwin8.10.1 arch
>    >> i386 os darwin8.10.1 system i386, darwin8.10.1 status
>    >> major 2 minor 6.1 year 2007 month 11 day 26 svn rev 43537
>    >> language R
>    >>
>    >> ______________________________________________
>    >> R-devel at r-project.org mailing list
>    >> https://stat.ethz.ch/mailman/listinfo/r-devel
>    >>
>
>
-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595