[Rd] Typo in tools:::format.check_Rd_metadata()

Duncan Murdoch murdoch.duncan at gmail.com
Tue Mar 29 19:53:01 CEST 2011


On 29/03/2011 1:31 PM, Hervé Pagès wrote:
> Hi,
>
> With R version 2.13.0 alpha (2011-03-24 r55004), 'R CMD check' will
> produce the following output on some packages:
>
>     * checking Rd metadata ... WARNING
>     Error in is.function(FUN) : 'FUN' is missing
>     Calls: print ... format.check_Rd_metadata ->  unlist ->  lapply ->
> match.fun
>     Execution halted
>
> This is happening on packages that contain duplicated \name or \alias
> tags in their Rd files. 'R CMD check' is able to detect this situation
> but is not able to display the warning message correctly because of
> a typo in the tools:::format.check_Rd_metadata() function:
>
>     format.check_Rd_metadata<-
>     function(x, ...)
>     {
>       c(character(),
>         if(length(bad<- x$files_with_duplicated_name)) {
>             unlist(lapply(names(bad)),
>                    function(nm) {
>                        c(gettextf("Rd files with duplicated name '%s':",
>                                   nm),
>                          .pretty_format(bad[[nm]]))
>                    })
>         },
>         if(length(bad<- x$files_with_duplicated_aliases)) {
>             unlist(lapply(names(bad)),
>                    function(nm) {
>                        c(gettextf("Rd files with duplicated alias '%s':",
>                                   nm),
>                          .pretty_format(bad[[nm]]))
>                    })
>         })
>     }
>
> Note the closing parentesis for the lapply() calls? Looks like it
> was intended to be something like:
>
>             unlist(lapply(names(bad),
>                           function(nm) {
>                               c(gettextf("Rd files with duplicated name
> '%s':",
>                                          nm),
>                                 .pretty_format(bad[[nm]]))
>                           }))
>
> Cheers,
> H.
>
>

Thanks, will fix.

Duncan Murdoch



More information about the R-devel mailing list